daily update
[deliverable/binutils-gdb.git] / gdb / doc / gdb.texinfo
... / ...
CommitLineData
1\input texinfo @c -*-texinfo-*-
2@c Copyright (C) 1988-2013 Free Software Foundation, Inc.
3@c
4@c %**start of header
5@c makeinfo ignores cmds prev to setfilename, so its arg cannot make use
6@c of @set vars. However, you can override filename with makeinfo -o.
7@setfilename gdb.info
8@c
9@c man begin INCLUDE
10@include gdb-cfg.texi
11@c man end
12@c
13@settitle Debugging with @value{GDBN}
14@setchapternewpage odd
15@c %**end of header
16
17@iftex
18@c @smallbook
19@c @cropmarks
20@end iftex
21
22@finalout
23@c To avoid file-name clashes between index.html and Index.html, when
24@c the manual is produced on a Posix host and then moved to a
25@c case-insensitive filesystem (e.g., MS-Windows), we separate the
26@c indices into two: Concept Index and all the rest.
27@syncodeindex ky fn
28@syncodeindex tp fn
29
30@c readline appendices use @vindex, @findex and @ftable,
31@c annotate.texi and gdbmi use @findex.
32@syncodeindex vr fn
33
34@c !!set GDB manual's edition---not the same as GDB version!
35@c This is updated by GNU Press.
36@set EDITION Tenth
37
38@c !!set GDB edit command default editor
39@set EDITOR /bin/ex
40
41@c THIS MANUAL REQUIRES TEXINFO 4.0 OR LATER.
42
43@c This is a dir.info fragment to support semi-automated addition of
44@c manuals to an info tree.
45@dircategory Software development
46@direntry
47* Gdb: (gdb). The GNU debugger.
48@end direntry
49
50@copying
51@c man begin COPYRIGHT
52Copyright @copyright{} 1988-2013 Free Software Foundation, Inc.
53
54Permission is granted to copy, distribute and/or modify this document
55under the terms of the GNU Free Documentation License, Version 1.3 or
56any later version published by the Free Software Foundation; with the
57Invariant Sections being ``Free Software'' and ``Free Software Needs
58Free Documentation'', with the Front-Cover Texts being ``A GNU Manual,''
59and with the Back-Cover Texts as in (a) below.
60
61(a) The FSF's Back-Cover Text is: ``You are free to copy and modify
62this GNU Manual. Buying copies from GNU Press supports the FSF in
63developing GNU and promoting software freedom.''
64@c man end
65@end copying
66
67@ifnottex
68This file documents the @sc{gnu} debugger @value{GDBN}.
69
70This is the @value{EDITION} Edition, of @cite{Debugging with
71@value{GDBN}: the @sc{gnu} Source-Level Debugger} for @value{GDBN}
72@ifset VERSION_PACKAGE
73@value{VERSION_PACKAGE}
74@end ifset
75Version @value{GDBVN}.
76
77@insertcopying
78@end ifnottex
79
80@titlepage
81@title Debugging with @value{GDBN}
82@subtitle The @sc{gnu} Source-Level Debugger
83@sp 1
84@subtitle @value{EDITION} Edition, for @value{GDBN} version @value{GDBVN}
85@ifset VERSION_PACKAGE
86@sp 1
87@subtitle @value{VERSION_PACKAGE}
88@end ifset
89@author Richard Stallman, Roland Pesch, Stan Shebs, et al.
90@page
91@tex
92{\parskip=0pt
93\hfill (Send bugs and comments on @value{GDBN} to @value{BUGURL}.)\par
94\hfill {\it Debugging with @value{GDBN}}\par
95\hfill \TeX{}info \texinfoversion\par
96}
97@end tex
98
99@vskip 0pt plus 1filll
100Published by the Free Software Foundation @*
10151 Franklin Street, Fifth Floor,
102Boston, MA 02110-1301, USA@*
103ISBN 978-0-9831592-3-0 @*
104
105@insertcopying
106@end titlepage
107@page
108
109@ifnottex
110@node Top, Summary, (dir), (dir)
111
112@top Debugging with @value{GDBN}
113
114This file describes @value{GDBN}, the @sc{gnu} symbolic debugger.
115
116This is the @value{EDITION} Edition, for @value{GDBN}
117@ifset VERSION_PACKAGE
118@value{VERSION_PACKAGE}
119@end ifset
120Version @value{GDBVN}.
121
122Copyright (C) 1988-2013 Free Software Foundation, Inc.
123
124This edition of the GDB manual is dedicated to the memory of Fred
125Fish. Fred was a long-standing contributor to GDB and to Free
126software in general. We will miss him.
127
128@menu
129* Summary:: Summary of @value{GDBN}
130* Sample Session:: A sample @value{GDBN} session
131
132* Invocation:: Getting in and out of @value{GDBN}
133* Commands:: @value{GDBN} commands
134* Running:: Running programs under @value{GDBN}
135* Stopping:: Stopping and continuing
136* Reverse Execution:: Running programs backward
137* Process Record and Replay:: Recording inferior's execution and replaying it
138* Stack:: Examining the stack
139* Source:: Examining source files
140* Data:: Examining data
141* Optimized Code:: Debugging optimized code
142* Macros:: Preprocessor Macros
143* Tracepoints:: Debugging remote targets non-intrusively
144* Overlays:: Debugging programs that use overlays
145
146* Languages:: Using @value{GDBN} with different languages
147
148* Symbols:: Examining the symbol table
149* Altering:: Altering execution
150* GDB Files:: @value{GDBN} files
151* Targets:: Specifying a debugging target
152* Remote Debugging:: Debugging remote programs
153* Configurations:: Configuration-specific information
154* Controlling GDB:: Controlling @value{GDBN}
155* Extending GDB:: Extending @value{GDBN}
156* Interpreters:: Command Interpreters
157* TUI:: @value{GDBN} Text User Interface
158* Emacs:: Using @value{GDBN} under @sc{gnu} Emacs
159* GDB/MI:: @value{GDBN}'s Machine Interface.
160* Annotations:: @value{GDBN}'s annotation interface.
161* JIT Interface:: Using the JIT debugging interface.
162* In-Process Agent:: In-Process Agent
163
164* GDB Bugs:: Reporting bugs in @value{GDBN}
165
166@ifset SYSTEM_READLINE
167* Command Line Editing: (rluserman). Command Line Editing
168* Using History Interactively: (history). Using History Interactively
169@end ifset
170@ifclear SYSTEM_READLINE
171* Command Line Editing:: Command Line Editing
172* Using History Interactively:: Using History Interactively
173@end ifclear
174* In Memoriam:: In Memoriam
175* Formatting Documentation:: How to format and print @value{GDBN} documentation
176* Installing GDB:: Installing GDB
177* Maintenance Commands:: Maintenance Commands
178* Remote Protocol:: GDB Remote Serial Protocol
179* Agent Expressions:: The GDB Agent Expression Mechanism
180* Target Descriptions:: How targets can describe themselves to
181 @value{GDBN}
182* Operating System Information:: Getting additional information from
183 the operating system
184* Trace File Format:: GDB trace file format
185* Index Section Format:: .gdb_index section format
186* Man Pages:: Manual pages
187* Copying:: GNU General Public License says
188 how you can copy and share GDB
189* GNU Free Documentation License:: The license for this documentation
190* Concept Index:: Index of @value{GDBN} concepts
191* Command and Variable Index:: Index of @value{GDBN} commands, variables,
192 functions, and Python data types
193@end menu
194
195@end ifnottex
196
197@contents
198
199@node Summary
200@unnumbered Summary of @value{GDBN}
201
202The purpose of a debugger such as @value{GDBN} is to allow you to see what is
203going on ``inside'' another program while it executes---or what another
204program was doing at the moment it crashed.
205
206@value{GDBN} can do four main kinds of things (plus other things in support of
207these) to help you catch bugs in the act:
208
209@itemize @bullet
210@item
211Start your program, specifying anything that might affect its behavior.
212
213@item
214Make your program stop on specified conditions.
215
216@item
217Examine what has happened, when your program has stopped.
218
219@item
220Change things in your program, so you can experiment with correcting the
221effects of one bug and go on to learn about another.
222@end itemize
223
224You can use @value{GDBN} to debug programs written in C and C@t{++}.
225For more information, see @ref{Supported Languages,,Supported Languages}.
226For more information, see @ref{C,,C and C++}.
227
228Support for D is partial. For information on D, see
229@ref{D,,D}.
230
231@cindex Modula-2
232Support for Modula-2 is partial. For information on Modula-2, see
233@ref{Modula-2,,Modula-2}.
234
235Support for OpenCL C is partial. For information on OpenCL C, see
236@ref{OpenCL C,,OpenCL C}.
237
238@cindex Pascal
239Debugging Pascal programs which use sets, subranges, file variables, or
240nested functions does not currently work. @value{GDBN} does not support
241entering expressions, printing values, or similar features using Pascal
242syntax.
243
244@cindex Fortran
245@value{GDBN} can be used to debug programs written in Fortran, although
246it may be necessary to refer to some variables with a trailing
247underscore.
248
249@value{GDBN} can be used to debug programs written in Objective-C,
250using either the Apple/NeXT or the GNU Objective-C runtime.
251
252@menu
253* Free Software:: Freely redistributable software
254* Free Documentation:: Free Software Needs Free Documentation
255* Contributors:: Contributors to GDB
256@end menu
257
258@node Free Software
259@unnumberedsec Free Software
260
261@value{GDBN} is @dfn{free software}, protected by the @sc{gnu}
262General Public License
263(GPL). The GPL gives you the freedom to copy or adapt a licensed
264program---but every person getting a copy also gets with it the
265freedom to modify that copy (which means that they must get access to
266the source code), and the freedom to distribute further copies.
267Typical software companies use copyrights to limit your freedoms; the
268Free Software Foundation uses the GPL to preserve these freedoms.
269
270Fundamentally, the General Public License is a license which says that
271you have these freedoms and that you cannot take these freedoms away
272from anyone else.
273
274@node Free Documentation
275@unnumberedsec Free Software Needs Free Documentation
276
277The biggest deficiency in the free software community today is not in
278the software---it is the lack of good free documentation that we can
279include with the free software. Many of our most important
280programs do not come with free reference manuals and free introductory
281texts. Documentation is an essential part of any software package;
282when an important free software package does not come with a free
283manual and a free tutorial, that is a major gap. We have many such
284gaps today.
285
286Consider Perl, for instance. The tutorial manuals that people
287normally use are non-free. How did this come about? Because the
288authors of those manuals published them with restrictive terms---no
289copying, no modification, source files not available---which exclude
290them from the free software world.
291
292That wasn't the first time this sort of thing happened, and it was far
293from the last. Many times we have heard a GNU user eagerly describe a
294manual that he is writing, his intended contribution to the community,
295only to learn that he had ruined everything by signing a publication
296contract to make it non-free.
297
298Free documentation, like free software, is a matter of freedom, not
299price. The problem with the non-free manual is not that publishers
300charge a price for printed copies---that in itself is fine. (The Free
301Software Foundation sells printed copies of manuals, too.) The
302problem is the restrictions on the use of the manual. Free manuals
303are available in source code form, and give you permission to copy and
304modify. Non-free manuals do not allow this.
305
306The criteria of freedom for a free manual are roughly the same as for
307free software. Redistribution (including the normal kinds of
308commercial redistribution) must be permitted, so that the manual can
309accompany every copy of the program, both on-line and on paper.
310
311Permission for modification of the technical content is crucial too.
312When people modify the software, adding or changing features, if they
313are conscientious they will change the manual too---so they can
314provide accurate and clear documentation for the modified program. A
315manual that leaves you no choice but to write a new manual to document
316a changed version of the program is not really available to our
317community.
318
319Some kinds of limits on the way modification is handled are
320acceptable. For example, requirements to preserve the original
321author's copyright notice, the distribution terms, or the list of
322authors, are ok. It is also no problem to require modified versions
323to include notice that they were modified. Even entire sections that
324may not be deleted or changed are acceptable, as long as they deal
325with nontechnical topics (like this one). These kinds of restrictions
326are acceptable because they don't obstruct the community's normal use
327of the manual.
328
329However, it must be possible to modify all the @emph{technical}
330content of the manual, and then distribute the result in all the usual
331media, through all the usual channels. Otherwise, the restrictions
332obstruct the use of the manual, it is not free, and we need another
333manual to replace it.
334
335Please spread the word about this issue. Our community continues to
336lose manuals to proprietary publishing. If we spread the word that
337free software needs free reference manuals and free tutorials, perhaps
338the next person who wants to contribute by writing documentation will
339realize, before it is too late, that only free manuals contribute to
340the free software community.
341
342If you are writing documentation, please insist on publishing it under
343the GNU Free Documentation License or another free documentation
344license. Remember that this decision requires your approval---you
345don't have to let the publisher decide. Some commercial publishers
346will use a free license if you insist, but they will not propose the
347option; it is up to you to raise the issue and say firmly that this is
348what you want. If the publisher you are dealing with refuses, please
349try other publishers. If you're not sure whether a proposed license
350is free, write to @email{licensing@@gnu.org}.
351
352You can encourage commercial publishers to sell more free, copylefted
353manuals and tutorials by buying them, and particularly by buying
354copies from the publishers that paid for their writing or for major
355improvements. Meanwhile, try to avoid buying non-free documentation
356at all. Check the distribution terms of a manual before you buy it,
357and insist that whoever seeks your business must respect your freedom.
358Check the history of the book, and try to reward the publishers that
359have paid or pay the authors to work on it.
360
361The Free Software Foundation maintains a list of free documentation
362published by other publishers, at
363@url{http://www.fsf.org/doc/other-free-books.html}.
364
365@node Contributors
366@unnumberedsec Contributors to @value{GDBN}
367
368Richard Stallman was the original author of @value{GDBN}, and of many
369other @sc{gnu} programs. Many others have contributed to its
370development. This section attempts to credit major contributors. One
371of the virtues of free software is that everyone is free to contribute
372to it; with regret, we cannot actually acknowledge everyone here. The
373file @file{ChangeLog} in the @value{GDBN} distribution approximates a
374blow-by-blow account.
375
376Changes much prior to version 2.0 are lost in the mists of time.
377
378@quotation
379@emph{Plea:} Additions to this section are particularly welcome. If you
380or your friends (or enemies, to be evenhanded) have been unfairly
381omitted from this list, we would like to add your names!
382@end quotation
383
384So that they may not regard their many labors as thankless, we
385particularly thank those who shepherded @value{GDBN} through major
386releases:
387Andrew Cagney (releases 6.3, 6.2, 6.1, 6.0, 5.3, 5.2, 5.1 and 5.0);
388Jim Blandy (release 4.18);
389Jason Molenda (release 4.17);
390Stan Shebs (release 4.14);
391Fred Fish (releases 4.16, 4.15, 4.13, 4.12, 4.11, 4.10, and 4.9);
392Stu Grossman and John Gilmore (releases 4.8, 4.7, 4.6, 4.5, and 4.4);
393John Gilmore (releases 4.3, 4.2, 4.1, 4.0, and 3.9);
394Jim Kingdon (releases 3.5, 3.4, and 3.3);
395and Randy Smith (releases 3.2, 3.1, and 3.0).
396
397Richard Stallman, assisted at various times by Peter TerMaat, Chris
398Hanson, and Richard Mlynarik, handled releases through 2.8.
399
400Michael Tiemann is the author of most of the @sc{gnu} C@t{++} support
401in @value{GDBN}, with significant additional contributions from Per
402Bothner and Daniel Berlin. James Clark wrote the @sc{gnu} C@t{++}
403demangler. Early work on C@t{++} was by Peter TerMaat (who also did
404much general update work leading to release 3.0).
405
406@value{GDBN} uses the BFD subroutine library to examine multiple
407object-file formats; BFD was a joint project of David V.
408Henkel-Wallace, Rich Pixley, Steve Chamberlain, and John Gilmore.
409
410David Johnson wrote the original COFF support; Pace Willison did
411the original support for encapsulated COFF.
412
413Brent Benson of Harris Computer Systems contributed DWARF 2 support.
414
415Adam de Boor and Bradley Davis contributed the ISI Optimum V support.
416Per Bothner, Noboyuki Hikichi, and Alessandro Forin contributed MIPS
417support.
418Jean-Daniel Fekete contributed Sun 386i support.
419Chris Hanson improved the HP9000 support.
420Noboyuki Hikichi and Tomoyuki Hasei contributed Sony/News OS 3 support.
421David Johnson contributed Encore Umax support.
422Jyrki Kuoppala contributed Altos 3068 support.
423Jeff Law contributed HP PA and SOM support.
424Keith Packard contributed NS32K support.
425Doug Rabson contributed Acorn Risc Machine support.
426Bob Rusk contributed Harris Nighthawk CX-UX support.
427Chris Smith contributed Convex support (and Fortran debugging).
428Jonathan Stone contributed Pyramid support.
429Michael Tiemann contributed SPARC support.
430Tim Tucker contributed support for the Gould NP1 and Gould Powernode.
431Pace Willison contributed Intel 386 support.
432Jay Vosburgh contributed Symmetry support.
433Marko Mlinar contributed OpenRISC 1000 support.
434
435Andreas Schwab contributed M68K @sc{gnu}/Linux support.
436
437Rich Schaefer and Peter Schauer helped with support of SunOS shared
438libraries.
439
440Jay Fenlason and Roland McGrath ensured that @value{GDBN} and GAS agree
441about several machine instruction sets.
442
443Patrick Duval, Ted Goldstein, Vikram Koka and Glenn Engel helped develop
444remote debugging. Intel Corporation, Wind River Systems, AMD, and ARM
445contributed remote debugging modules for the i960, VxWorks, A29K UDI,
446and RDI targets, respectively.
447
448Brian Fox is the author of the readline libraries providing
449command-line editing and command history.
450
451Andrew Beers of SUNY Buffalo wrote the language-switching code, the
452Modula-2 support, and contributed the Languages chapter of this manual.
453
454Fred Fish wrote most of the support for Unix System Vr4.
455He also enhanced the command-completion support to cover C@t{++} overloaded
456symbols.
457
458Hitachi America (now Renesas America), Ltd. sponsored the support for
459H8/300, H8/500, and Super-H processors.
460
461NEC sponsored the support for the v850, Vr4xxx, and Vr5xxx processors.
462
463Mitsubishi (now Renesas) sponsored the support for D10V, D30V, and M32R/D
464processors.
465
466Toshiba sponsored the support for the TX39 Mips processor.
467
468Matsushita sponsored the support for the MN10200 and MN10300 processors.
469
470Fujitsu sponsored the support for SPARClite and FR30 processors.
471
472Kung Hsu, Jeff Law, and Rick Sladkey added support for hardware
473watchpoints.
474
475Michael Snyder added support for tracepoints.
476
477Stu Grossman wrote gdbserver.
478
479Jim Kingdon, Peter Schauer, Ian Taylor, and Stu Grossman made
480nearly innumerable bug fixes and cleanups throughout @value{GDBN}.
481
482The following people at the Hewlett-Packard Company contributed
483support for the PA-RISC 2.0 architecture, HP-UX 10.20, 10.30, and 11.0
484(narrow mode), HP's implementation of kernel threads, HP's aC@t{++}
485compiler, and the Text User Interface (nee Terminal User Interface):
486Ben Krepp, Richard Title, John Bishop, Susan Macchia, Kathy Mann,
487Satish Pai, India Paul, Steve Rehrauer, and Elena Zannoni. Kim Haase
488provided HP-specific information in this manual.
489
490DJ Delorie ported @value{GDBN} to MS-DOS, for the DJGPP project.
491Robert Hoehne made significant contributions to the DJGPP port.
492
493Cygnus Solutions has sponsored @value{GDBN} maintenance and much of its
494development since 1991. Cygnus engineers who have worked on @value{GDBN}
495fulltime include Mark Alexander, Jim Blandy, Per Bothner, Kevin
496Buettner, Edith Epstein, Chris Faylor, Fred Fish, Martin Hunt, Jim
497Ingham, John Gilmore, Stu Grossman, Kung Hsu, Jim Kingdon, John Metzler,
498Fernando Nasser, Geoffrey Noer, Dawn Perchik, Rich Pixley, Zdenek
499Radouch, Keith Seitz, Stan Shebs, David Taylor, and Elena Zannoni. In
500addition, Dave Brolley, Ian Carmichael, Steve Chamberlain, Nick Clifton,
501JT Conklin, Stan Cox, DJ Delorie, Ulrich Drepper, Frank Eigler, Doug
502Evans, Sean Fagan, David Henkel-Wallace, Richard Henderson, Jeff
503Holcomb, Jeff Law, Jim Lemke, Tom Lord, Bob Manson, Michael Meissner,
504Jason Merrill, Catherine Moore, Drew Moseley, Ken Raeburn, Gavin
505Romig-Koch, Rob Savoye, Jamie Smith, Mike Stump, Ian Taylor, Angela
506Thomas, Michael Tiemann, Tom Tromey, Ron Unrau, Jim Wilson, and David
507Zuhn have made contributions both large and small.
508
509Andrew Cagney, Fernando Nasser, and Elena Zannoni, while working for
510Cygnus Solutions, implemented the original @sc{gdb/mi} interface.
511
512Jim Blandy added support for preprocessor macros, while working for Red
513Hat.
514
515Andrew Cagney designed @value{GDBN}'s architecture vector. Many
516people including Andrew Cagney, Stephane Carrez, Randolph Chung, Nick
517Duffek, Richard Henderson, Mark Kettenis, Grace Sainsbury, Kei
518Sakamoto, Yoshinori Sato, Michael Snyder, Andreas Schwab, Jason
519Thorpe, Corinna Vinschen, Ulrich Weigand, and Elena Zannoni, helped
520with the migration of old architectures to this new framework.
521
522Andrew Cagney completely re-designed and re-implemented @value{GDBN}'s
523unwinder framework, this consisting of a fresh new design featuring
524frame IDs, independent frame sniffers, and the sentinel frame. Mark
525Kettenis implemented the @sc{dwarf 2} unwinder, Jeff Johnston the
526libunwind unwinder, and Andrew Cagney the dummy, sentinel, tramp, and
527trad unwinders. The architecture-specific changes, each involving a
528complete rewrite of the architecture's frame code, were carried out by
529Jim Blandy, Joel Brobecker, Kevin Buettner, Andrew Cagney, Stephane
530Carrez, Randolph Chung, Orjan Friberg, Richard Henderson, Daniel
531Jacobowitz, Jeff Johnston, Mark Kettenis, Theodore A. Roth, Kei
532Sakamoto, Yoshinori Sato, Michael Snyder, Corinna Vinschen, and Ulrich
533Weigand.
534
535Christian Zankel, Ross Morley, Bob Wilson, and Maxim Grigoriev from
536Tensilica, Inc.@: contributed support for Xtensa processors. Others
537who have worked on the Xtensa port of @value{GDBN} in the past include
538Steve Tjiang, John Newlin, and Scott Foehner.
539
540Michael Eager and staff of Xilinx, Inc., contributed support for the
541Xilinx MicroBlaze architecture.
542
543@node Sample Session
544@chapter A Sample @value{GDBN} Session
545
546You can use this manual at your leisure to read all about @value{GDBN}.
547However, a handful of commands are enough to get started using the
548debugger. This chapter illustrates those commands.
549
550@iftex
551In this sample session, we emphasize user input like this: @b{input},
552to make it easier to pick out from the surrounding output.
553@end iftex
554
555@c FIXME: this example may not be appropriate for some configs, where
556@c FIXME...primary interest is in remote use.
557
558One of the preliminary versions of @sc{gnu} @code{m4} (a generic macro
559processor) exhibits the following bug: sometimes, when we change its
560quote strings from the default, the commands used to capture one macro
561definition within another stop working. In the following short @code{m4}
562session, we define a macro @code{foo} which expands to @code{0000}; we
563then use the @code{m4} built-in @code{defn} to define @code{bar} as the
564same thing. However, when we change the open quote string to
565@code{<QUOTE>} and the close quote string to @code{<UNQUOTE>}, the same
566procedure fails to define a new synonym @code{baz}:
567
568@smallexample
569$ @b{cd gnu/m4}
570$ @b{./m4}
571@b{define(foo,0000)}
572
573@b{foo}
5740000
575@b{define(bar,defn(`foo'))}
576
577@b{bar}
5780000
579@b{changequote(<QUOTE>,<UNQUOTE>)}
580
581@b{define(baz,defn(<QUOTE>foo<UNQUOTE>))}
582@b{baz}
583@b{Ctrl-d}
584m4: End of input: 0: fatal error: EOF in string
585@end smallexample
586
587@noindent
588Let us use @value{GDBN} to try to see what is going on.
589
590@smallexample
591$ @b{@value{GDBP} m4}
592@c FIXME: this falsifies the exact text played out, to permit smallbook
593@c FIXME... format to come out better.
594@value{GDBN} is free software and you are welcome to distribute copies
595 of it under certain conditions; type "show copying" to see
596 the conditions.
597There is absolutely no warranty for @value{GDBN}; type "show warranty"
598 for details.
599
600@value{GDBN} @value{GDBVN}, Copyright 1999 Free Software Foundation, Inc...
601(@value{GDBP})
602@end smallexample
603
604@noindent
605@value{GDBN} reads only enough symbol data to know where to find the
606rest when needed; as a result, the first prompt comes up very quickly.
607We now tell @value{GDBN} to use a narrower display width than usual, so
608that examples fit in this manual.
609
610@smallexample
611(@value{GDBP}) @b{set width 70}
612@end smallexample
613
614@noindent
615We need to see how the @code{m4} built-in @code{changequote} works.
616Having looked at the source, we know the relevant subroutine is
617@code{m4_changequote}, so we set a breakpoint there with the @value{GDBN}
618@code{break} command.
619
620@smallexample
621(@value{GDBP}) @b{break m4_changequote}
622Breakpoint 1 at 0x62f4: file builtin.c, line 879.
623@end smallexample
624
625@noindent
626Using the @code{run} command, we start @code{m4} running under @value{GDBN}
627control; as long as control does not reach the @code{m4_changequote}
628subroutine, the program runs as usual:
629
630@smallexample
631(@value{GDBP}) @b{run}
632Starting program: /work/Editorial/gdb/gnu/m4/m4
633@b{define(foo,0000)}
634
635@b{foo}
6360000
637@end smallexample
638
639@noindent
640To trigger the breakpoint, we call @code{changequote}. @value{GDBN}
641suspends execution of @code{m4}, displaying information about the
642context where it stops.
643
644@smallexample
645@b{changequote(<QUOTE>,<UNQUOTE>)}
646
647Breakpoint 1, m4_changequote (argc=3, argv=0x33c70)
648 at builtin.c:879
649879 if (bad_argc(TOKEN_DATA_TEXT(argv[0]),argc,1,3))
650@end smallexample
651
652@noindent
653Now we use the command @code{n} (@code{next}) to advance execution to
654the next line of the current function.
655
656@smallexample
657(@value{GDBP}) @b{n}
658882 set_quotes((argc >= 2) ? TOKEN_DATA_TEXT(argv[1])\
659 : nil,
660@end smallexample
661
662@noindent
663@code{set_quotes} looks like a promising subroutine. We can go into it
664by using the command @code{s} (@code{step}) instead of @code{next}.
665@code{step} goes to the next line to be executed in @emph{any}
666subroutine, so it steps into @code{set_quotes}.
667
668@smallexample
669(@value{GDBP}) @b{s}
670set_quotes (lq=0x34c78 "<QUOTE>", rq=0x34c88 "<UNQUOTE>")
671 at input.c:530
672530 if (lquote != def_lquote)
673@end smallexample
674
675@noindent
676The display that shows the subroutine where @code{m4} is now
677suspended (and its arguments) is called a stack frame display. It
678shows a summary of the stack. We can use the @code{backtrace}
679command (which can also be spelled @code{bt}), to see where we are
680in the stack as a whole: the @code{backtrace} command displays a
681stack frame for each active subroutine.
682
683@smallexample
684(@value{GDBP}) @b{bt}
685#0 set_quotes (lq=0x34c78 "<QUOTE>", rq=0x34c88 "<UNQUOTE>")
686 at input.c:530
687#1 0x6344 in m4_changequote (argc=3, argv=0x33c70)
688 at builtin.c:882
689#2 0x8174 in expand_macro (sym=0x33320) at macro.c:242
690#3 0x7a88 in expand_token (obs=0x0, t=209696, td=0xf7fffa30)
691 at macro.c:71
692#4 0x79dc in expand_input () at macro.c:40
693#5 0x2930 in main (argc=0, argv=0xf7fffb20) at m4.c:195
694@end smallexample
695
696@noindent
697We step through a few more lines to see what happens. The first two
698times, we can use @samp{s}; the next two times we use @code{n} to avoid
699falling into the @code{xstrdup} subroutine.
700
701@smallexample
702(@value{GDBP}) @b{s}
7030x3b5c 532 if (rquote != def_rquote)
704(@value{GDBP}) @b{s}
7050x3b80 535 lquote = (lq == nil || *lq == '\0') ? \
706def_lquote : xstrdup(lq);
707(@value{GDBP}) @b{n}
708536 rquote = (rq == nil || *rq == '\0') ? def_rquote\
709 : xstrdup(rq);
710(@value{GDBP}) @b{n}
711538 len_lquote = strlen(rquote);
712@end smallexample
713
714@noindent
715The last line displayed looks a little odd; we can examine the variables
716@code{lquote} and @code{rquote} to see if they are in fact the new left
717and right quotes we specified. We use the command @code{p}
718(@code{print}) to see their values.
719
720@smallexample
721(@value{GDBP}) @b{p lquote}
722$1 = 0x35d40 "<QUOTE>"
723(@value{GDBP}) @b{p rquote}
724$2 = 0x35d50 "<UNQUOTE>"
725@end smallexample
726
727@noindent
728@code{lquote} and @code{rquote} are indeed the new left and right quotes.
729To look at some context, we can display ten lines of source
730surrounding the current line with the @code{l} (@code{list}) command.
731
732@smallexample
733(@value{GDBP}) @b{l}
734533 xfree(rquote);
735534
736535 lquote = (lq == nil || *lq == '\0') ? def_lquote\
737 : xstrdup (lq);
738536 rquote = (rq == nil || *rq == '\0') ? def_rquote\
739 : xstrdup (rq);
740537
741538 len_lquote = strlen(rquote);
742539 len_rquote = strlen(lquote);
743540 @}
744541
745542 void
746@end smallexample
747
748@noindent
749Let us step past the two lines that set @code{len_lquote} and
750@code{len_rquote}, and then examine the values of those variables.
751
752@smallexample
753(@value{GDBP}) @b{n}
754539 len_rquote = strlen(lquote);
755(@value{GDBP}) @b{n}
756540 @}
757(@value{GDBP}) @b{p len_lquote}
758$3 = 9
759(@value{GDBP}) @b{p len_rquote}
760$4 = 7
761@end smallexample
762
763@noindent
764That certainly looks wrong, assuming @code{len_lquote} and
765@code{len_rquote} are meant to be the lengths of @code{lquote} and
766@code{rquote} respectively. We can set them to better values using
767the @code{p} command, since it can print the value of
768any expression---and that expression can include subroutine calls and
769assignments.
770
771@smallexample
772(@value{GDBP}) @b{p len_lquote=strlen(lquote)}
773$5 = 7
774(@value{GDBP}) @b{p len_rquote=strlen(rquote)}
775$6 = 9
776@end smallexample
777
778@noindent
779Is that enough to fix the problem of using the new quotes with the
780@code{m4} built-in @code{defn}? We can allow @code{m4} to continue
781executing with the @code{c} (@code{continue}) command, and then try the
782example that caused trouble initially:
783
784@smallexample
785(@value{GDBP}) @b{c}
786Continuing.
787
788@b{define(baz,defn(<QUOTE>foo<UNQUOTE>))}
789
790baz
7910000
792@end smallexample
793
794@noindent
795Success! The new quotes now work just as well as the default ones. The
796problem seems to have been just the two typos defining the wrong
797lengths. We allow @code{m4} exit by giving it an EOF as input:
798
799@smallexample
800@b{Ctrl-d}
801Program exited normally.
802@end smallexample
803
804@noindent
805The message @samp{Program exited normally.} is from @value{GDBN}; it
806indicates @code{m4} has finished executing. We can end our @value{GDBN}
807session with the @value{GDBN} @code{quit} command.
808
809@smallexample
810(@value{GDBP}) @b{quit}
811@end smallexample
812
813@node Invocation
814@chapter Getting In and Out of @value{GDBN}
815
816This chapter discusses how to start @value{GDBN}, and how to get out of it.
817The essentials are:
818@itemize @bullet
819@item
820type @samp{@value{GDBP}} to start @value{GDBN}.
821@item
822type @kbd{quit} or @kbd{Ctrl-d} to exit.
823@end itemize
824
825@menu
826* Invoking GDB:: How to start @value{GDBN}
827* Quitting GDB:: How to quit @value{GDBN}
828* Shell Commands:: How to use shell commands inside @value{GDBN}
829* Logging Output:: How to log @value{GDBN}'s output to a file
830@end menu
831
832@node Invoking GDB
833@section Invoking @value{GDBN}
834
835Invoke @value{GDBN} by running the program @code{@value{GDBP}}. Once started,
836@value{GDBN} reads commands from the terminal until you tell it to exit.
837
838You can also run @code{@value{GDBP}} with a variety of arguments and options,
839to specify more of your debugging environment at the outset.
840
841The command-line options described here are designed
842to cover a variety of situations; in some environments, some of these
843options may effectively be unavailable.
844
845The most usual way to start @value{GDBN} is with one argument,
846specifying an executable program:
847
848@smallexample
849@value{GDBP} @var{program}
850@end smallexample
851
852@noindent
853You can also start with both an executable program and a core file
854specified:
855
856@smallexample
857@value{GDBP} @var{program} @var{core}
858@end smallexample
859
860You can, instead, specify a process ID as a second argument, if you want
861to debug a running process:
862
863@smallexample
864@value{GDBP} @var{program} 1234
865@end smallexample
866
867@noindent
868would attach @value{GDBN} to process @code{1234} (unless you also have a file
869named @file{1234}; @value{GDBN} does check for a core file first).
870
871Taking advantage of the second command-line argument requires a fairly
872complete operating system; when you use @value{GDBN} as a remote
873debugger attached to a bare board, there may not be any notion of
874``process'', and there is often no way to get a core dump. @value{GDBN}
875will warn you if it is unable to attach or to read core dumps.
876
877You can optionally have @code{@value{GDBP}} pass any arguments after the
878executable file to the inferior using @code{--args}. This option stops
879option processing.
880@smallexample
881@value{GDBP} --args gcc -O2 -c foo.c
882@end smallexample
883This will cause @code{@value{GDBP}} to debug @code{gcc}, and to set
884@code{gcc}'s command-line arguments (@pxref{Arguments}) to @samp{-O2 -c foo.c}.
885
886You can run @code{@value{GDBP}} without printing the front material, which describes
887@value{GDBN}'s non-warranty, by specifying @code{-silent}:
888
889@smallexample
890@value{GDBP} -silent
891@end smallexample
892
893@noindent
894You can further control how @value{GDBN} starts up by using command-line
895options. @value{GDBN} itself can remind you of the options available.
896
897@noindent
898Type
899
900@smallexample
901@value{GDBP} -help
902@end smallexample
903
904@noindent
905to display all available options and briefly describe their use
906(@samp{@value{GDBP} -h} is a shorter equivalent).
907
908All options and command line arguments you give are processed
909in sequential order. The order makes a difference when the
910@samp{-x} option is used.
911
912
913@menu
914* File Options:: Choosing files
915* Mode Options:: Choosing modes
916* Startup:: What @value{GDBN} does during startup
917@end menu
918
919@node File Options
920@subsection Choosing Files
921
922When @value{GDBN} starts, it reads any arguments other than options as
923specifying an executable file and core file (or process ID). This is
924the same as if the arguments were specified by the @samp{-se} and
925@samp{-c} (or @samp{-p}) options respectively. (@value{GDBN} reads the
926first argument that does not have an associated option flag as
927equivalent to the @samp{-se} option followed by that argument; and the
928second argument that does not have an associated option flag, if any, as
929equivalent to the @samp{-c}/@samp{-p} option followed by that argument.)
930If the second argument begins with a decimal digit, @value{GDBN} will
931first attempt to attach to it as a process, and if that fails, attempt
932to open it as a corefile. If you have a corefile whose name begins with
933a digit, you can prevent @value{GDBN} from treating it as a pid by
934prefixing it with @file{./}, e.g.@: @file{./12345}.
935
936If @value{GDBN} has not been configured to included core file support,
937such as for most embedded targets, then it will complain about a second
938argument and ignore it.
939
940Many options have both long and short forms; both are shown in the
941following list. @value{GDBN} also recognizes the long forms if you truncate
942them, so long as enough of the option is present to be unambiguous.
943(If you prefer, you can flag option arguments with @samp{--} rather
944than @samp{-}, though we illustrate the more usual convention.)
945
946@c NOTE: the @cindex entries here use double dashes ON PURPOSE. This
947@c way, both those who look for -foo and --foo in the index, will find
948@c it.
949
950@table @code
951@item -symbols @var{file}
952@itemx -s @var{file}
953@cindex @code{--symbols}
954@cindex @code{-s}
955Read symbol table from file @var{file}.
956
957@item -exec @var{file}
958@itemx -e @var{file}
959@cindex @code{--exec}
960@cindex @code{-e}
961Use file @var{file} as the executable file to execute when appropriate,
962and for examining pure data in conjunction with a core dump.
963
964@item -se @var{file}
965@cindex @code{--se}
966Read symbol table from file @var{file} and use it as the executable
967file.
968
969@item -core @var{file}
970@itemx -c @var{file}
971@cindex @code{--core}
972@cindex @code{-c}
973Use file @var{file} as a core dump to examine.
974
975@item -pid @var{number}
976@itemx -p @var{number}
977@cindex @code{--pid}
978@cindex @code{-p}
979Connect to process ID @var{number}, as with the @code{attach} command.
980
981@item -command @var{file}
982@itemx -x @var{file}
983@cindex @code{--command}
984@cindex @code{-x}
985Execute commands from file @var{file}. The contents of this file is
986evaluated exactly as the @code{source} command would.
987@xref{Command Files,, Command files}.
988
989@item -eval-command @var{command}
990@itemx -ex @var{command}
991@cindex @code{--eval-command}
992@cindex @code{-ex}
993Execute a single @value{GDBN} command.
994
995This option may be used multiple times to call multiple commands. It may
996also be interleaved with @samp{-command} as required.
997
998@smallexample
999@value{GDBP} -ex 'target sim' -ex 'load' \
1000 -x setbreakpoints -ex 'run' a.out
1001@end smallexample
1002
1003@item -init-command @var{file}
1004@itemx -ix @var{file}
1005@cindex @code{--init-command}
1006@cindex @code{-ix}
1007Execute commands from file @var{file} before loading the inferior (but
1008after loading gdbinit files).
1009@xref{Startup}.
1010
1011@item -init-eval-command @var{command}
1012@itemx -iex @var{command}
1013@cindex @code{--init-eval-command}
1014@cindex @code{-iex}
1015Execute a single @value{GDBN} command before loading the inferior (but
1016after loading gdbinit files).
1017@xref{Startup}.
1018
1019@item -directory @var{directory}
1020@itemx -d @var{directory}
1021@cindex @code{--directory}
1022@cindex @code{-d}
1023Add @var{directory} to the path to search for source and script files.
1024
1025@item -r
1026@itemx -readnow
1027@cindex @code{--readnow}
1028@cindex @code{-r}
1029Read each symbol file's entire symbol table immediately, rather than
1030the default, which is to read it incrementally as it is needed.
1031This makes startup slower, but makes future operations faster.
1032
1033@end table
1034
1035@node Mode Options
1036@subsection Choosing Modes
1037
1038You can run @value{GDBN} in various alternative modes---for example, in
1039batch mode or quiet mode.
1040
1041@table @code
1042@anchor{-nx}
1043@item -nx
1044@itemx -n
1045@cindex @code{--nx}
1046@cindex @code{-n}
1047Do not execute commands found in any initialization file.
1048There are three init files, loaded in the following order:
1049
1050@table @code
1051@item @file{system.gdbinit}
1052This is the system-wide init file.
1053Its location is specified with the @code{--with-system-gdbinit}
1054configure option (@pxref{System-wide configuration}).
1055It is loaded first when @value{GDBN} starts, before command line options
1056have been processed.
1057@item @file{~/.gdbinit}
1058This is the init file in your home directory.
1059It is loaded next, after @file{system.gdbinit}, and before
1060command options have been processed.
1061@item @file{./.gdbinit}
1062This is the init file in the current directory.
1063It is loaded last, after command line options other than @code{-x} and
1064@code{-ex} have been processed. Command line options @code{-x} and
1065@code{-ex} are processed last, after @file{./.gdbinit} has been loaded.
1066@end table
1067
1068For further documentation on startup processing, @xref{Startup}.
1069For documentation on how to write command files,
1070@xref{Command Files,,Command Files}.
1071
1072@anchor{-nh}
1073@item -nh
1074@cindex @code{--nh}
1075Do not execute commands found in @file{~/.gdbinit}, the init file
1076in your home directory.
1077@xref{Startup}.
1078
1079@item -quiet
1080@itemx -silent
1081@itemx -q
1082@cindex @code{--quiet}
1083@cindex @code{--silent}
1084@cindex @code{-q}
1085``Quiet''. Do not print the introductory and copyright messages. These
1086messages are also suppressed in batch mode.
1087
1088@item -batch
1089@cindex @code{--batch}
1090Run in batch mode. Exit with status @code{0} after processing all the
1091command files specified with @samp{-x} (and all commands from
1092initialization files, if not inhibited with @samp{-n}). Exit with
1093nonzero status if an error occurs in executing the @value{GDBN} commands
1094in the command files. Batch mode also disables pagination, sets unlimited
1095terminal width and height @pxref{Screen Size}, and acts as if @kbd{set confirm
1096off} were in effect (@pxref{Messages/Warnings}).
1097
1098Batch mode may be useful for running @value{GDBN} as a filter, for
1099example to download and run a program on another computer; in order to
1100make this more useful, the message
1101
1102@smallexample
1103Program exited normally.
1104@end smallexample
1105
1106@noindent
1107(which is ordinarily issued whenever a program running under
1108@value{GDBN} control terminates) is not issued when running in batch
1109mode.
1110
1111@item -batch-silent
1112@cindex @code{--batch-silent}
1113Run in batch mode exactly like @samp{-batch}, but totally silently. All
1114@value{GDBN} output to @code{stdout} is prevented (@code{stderr} is
1115unaffected). This is much quieter than @samp{-silent} and would be useless
1116for an interactive session.
1117
1118This is particularly useful when using targets that give @samp{Loading section}
1119messages, for example.
1120
1121Note that targets that give their output via @value{GDBN}, as opposed to
1122writing directly to @code{stdout}, will also be made silent.
1123
1124@item -return-child-result
1125@cindex @code{--return-child-result}
1126The return code from @value{GDBN} will be the return code from the child
1127process (the process being debugged), with the following exceptions:
1128
1129@itemize @bullet
1130@item
1131@value{GDBN} exits abnormally. E.g., due to an incorrect argument or an
1132internal error. In this case the exit code is the same as it would have been
1133without @samp{-return-child-result}.
1134@item
1135The user quits with an explicit value. E.g., @samp{quit 1}.
1136@item
1137The child process never runs, or is not allowed to terminate, in which case
1138the exit code will be -1.
1139@end itemize
1140
1141This option is useful in conjunction with @samp{-batch} or @samp{-batch-silent},
1142when @value{GDBN} is being used as a remote program loader or simulator
1143interface.
1144
1145@item -nowindows
1146@itemx -nw
1147@cindex @code{--nowindows}
1148@cindex @code{-nw}
1149``No windows''. If @value{GDBN} comes with a graphical user interface
1150(GUI) built in, then this option tells @value{GDBN} to only use the command-line
1151interface. If no GUI is available, this option has no effect.
1152
1153@item -windows
1154@itemx -w
1155@cindex @code{--windows}
1156@cindex @code{-w}
1157If @value{GDBN} includes a GUI, then this option requires it to be
1158used if possible.
1159
1160@item -cd @var{directory}
1161@cindex @code{--cd}
1162Run @value{GDBN} using @var{directory} as its working directory,
1163instead of the current directory.
1164
1165@item -data-directory @var{directory}
1166@cindex @code{--data-directory}
1167Run @value{GDBN} using @var{directory} as its data directory.
1168The data directory is where @value{GDBN} searches for its
1169auxiliary files. @xref{Data Files}.
1170
1171@item -fullname
1172@itemx -f
1173@cindex @code{--fullname}
1174@cindex @code{-f}
1175@sc{gnu} Emacs sets this option when it runs @value{GDBN} as a
1176subprocess. It tells @value{GDBN} to output the full file name and line
1177number in a standard, recognizable fashion each time a stack frame is
1178displayed (which includes each time your program stops). This
1179recognizable format looks like two @samp{\032} characters, followed by
1180the file name, line number and character position separated by colons,
1181and a newline. The Emacs-to-@value{GDBN} interface program uses the two
1182@samp{\032} characters as a signal to display the source code for the
1183frame.
1184
1185@item -annotate @var{level}
1186@cindex @code{--annotate}
1187This option sets the @dfn{annotation level} inside @value{GDBN}. Its
1188effect is identical to using @samp{set annotate @var{level}}
1189(@pxref{Annotations}). The annotation @var{level} controls how much
1190information @value{GDBN} prints together with its prompt, values of
1191expressions, source lines, and other types of output. Level 0 is the
1192normal, level 1 is for use when @value{GDBN} is run as a subprocess of
1193@sc{gnu} Emacs, level 3 is the maximum annotation suitable for programs
1194that control @value{GDBN}, and level 2 has been deprecated.
1195
1196The annotation mechanism has largely been superseded by @sc{gdb/mi}
1197(@pxref{GDB/MI}).
1198
1199@item --args
1200@cindex @code{--args}
1201Change interpretation of command line so that arguments following the
1202executable file are passed as command line arguments to the inferior.
1203This option stops option processing.
1204
1205@item -baud @var{bps}
1206@itemx -b @var{bps}
1207@cindex @code{--baud}
1208@cindex @code{-b}
1209Set the line speed (baud rate or bits per second) of any serial
1210interface used by @value{GDBN} for remote debugging.
1211
1212@item -l @var{timeout}
1213@cindex @code{-l}
1214Set the timeout (in seconds) of any communication used by @value{GDBN}
1215for remote debugging.
1216
1217@item -tty @var{device}
1218@itemx -t @var{device}
1219@cindex @code{--tty}
1220@cindex @code{-t}
1221Run using @var{device} for your program's standard input and output.
1222@c FIXME: kingdon thinks there is more to -tty. Investigate.
1223
1224@c resolve the situation of these eventually
1225@item -tui
1226@cindex @code{--tui}
1227Activate the @dfn{Text User Interface} when starting. The Text User
1228Interface manages several text windows on the terminal, showing
1229source, assembly, registers and @value{GDBN} command outputs
1230(@pxref{TUI, ,@value{GDBN} Text User Interface}). Do not use this
1231option if you run @value{GDBN} from Emacs (@pxref{Emacs, ,
1232Using @value{GDBN} under @sc{gnu} Emacs}).
1233
1234@c @item -xdb
1235@c @cindex @code{--xdb}
1236@c Run in XDB compatibility mode, allowing the use of certain XDB commands.
1237@c For information, see the file @file{xdb_trans.html}, which is usually
1238@c installed in the directory @code{/opt/langtools/wdb/doc} on HP-UX
1239@c systems.
1240
1241@item -interpreter @var{interp}
1242@cindex @code{--interpreter}
1243Use the interpreter @var{interp} for interface with the controlling
1244program or device. This option is meant to be set by programs which
1245communicate with @value{GDBN} using it as a back end.
1246@xref{Interpreters, , Command Interpreters}.
1247
1248@samp{--interpreter=mi} (or @samp{--interpreter=mi2}) causes
1249@value{GDBN} to use the @dfn{@sc{gdb/mi} interface} (@pxref{GDB/MI, ,
1250The @sc{gdb/mi} Interface}) included since @value{GDBN} version 6.0. The
1251previous @sc{gdb/mi} interface, included in @value{GDBN} version 5.3 and
1252selected with @samp{--interpreter=mi1}, is deprecated. Earlier
1253@sc{gdb/mi} interfaces are no longer supported.
1254
1255@item -write
1256@cindex @code{--write}
1257Open the executable and core files for both reading and writing. This
1258is equivalent to the @samp{set write on} command inside @value{GDBN}
1259(@pxref{Patching}).
1260
1261@item -statistics
1262@cindex @code{--statistics}
1263This option causes @value{GDBN} to print statistics about time and
1264memory usage after it completes each command and returns to the prompt.
1265
1266@item -version
1267@cindex @code{--version}
1268This option causes @value{GDBN} to print its version number and
1269no-warranty blurb, and exit.
1270
1271@item -configuration
1272@cindex @code{--configuration}
1273This option causes @value{GDBN} to print details about its build-time
1274configuration parameters, and then exit. These details can be
1275important when reporting @value{GDBN} bugs (@pxref{GDB Bugs}).
1276
1277@end table
1278
1279@node Startup
1280@subsection What @value{GDBN} Does During Startup
1281@cindex @value{GDBN} startup
1282
1283Here's the description of what @value{GDBN} does during session startup:
1284
1285@enumerate
1286@item
1287Sets up the command interpreter as specified by the command line
1288(@pxref{Mode Options, interpreter}).
1289
1290@item
1291@cindex init file
1292Reads the system-wide @dfn{init file} (if @option{--with-system-gdbinit} was
1293used when building @value{GDBN}; @pxref{System-wide configuration,
1294 ,System-wide configuration and settings}) and executes all the commands in
1295that file.
1296
1297@anchor{Home Directory Init File}
1298@item
1299Reads the init file (if any) in your home directory@footnote{On
1300DOS/Windows systems, the home directory is the one pointed to by the
1301@code{HOME} environment variable.} and executes all the commands in
1302that file.
1303
1304@anchor{Option -init-eval-command}
1305@item
1306Executes commands and command files specified by the @samp{-iex} and
1307@samp{-ix} options in their specified order. Usually you should use the
1308@samp{-ex} and @samp{-x} options instead, but this way you can apply
1309settings before @value{GDBN} init files get executed and before inferior
1310gets loaded.
1311
1312@item
1313Processes command line options and operands.
1314
1315@anchor{Init File in the Current Directory during Startup}
1316@item
1317Reads and executes the commands from init file (if any) in the current
1318working directory as long as @samp{set auto-load local-gdbinit} is set to
1319@samp{on} (@pxref{Init File in the Current Directory}).
1320This is only done if the current directory is
1321different from your home directory. Thus, you can have more than one
1322init file, one generic in your home directory, and another, specific
1323to the program you are debugging, in the directory where you invoke
1324@value{GDBN}.
1325
1326@item
1327If the command line specified a program to debug, or a process to
1328attach to, or a core file, @value{GDBN} loads any auto-loaded
1329scripts provided for the program or for its loaded shared libraries.
1330@xref{Auto-loading}.
1331
1332If you wish to disable the auto-loading during startup,
1333you must do something like the following:
1334
1335@smallexample
1336$ gdb -iex "set auto-load python-scripts off" myprogram
1337@end smallexample
1338
1339Option @samp{-ex} does not work because the auto-loading is then turned
1340off too late.
1341
1342@item
1343Executes commands and command files specified by the @samp{-ex} and
1344@samp{-x} options in their specified order. @xref{Command Files}, for
1345more details about @value{GDBN} command files.
1346
1347@item
1348Reads the command history recorded in the @dfn{history file}.
1349@xref{Command History}, for more details about the command history and the
1350files where @value{GDBN} records it.
1351@end enumerate
1352
1353Init files use the same syntax as @dfn{command files} (@pxref{Command
1354Files}) and are processed by @value{GDBN} in the same way. The init
1355file in your home directory can set options (such as @samp{set
1356complaints}) that affect subsequent processing of command line options
1357and operands. Init files are not executed if you use the @samp{-nx}
1358option (@pxref{Mode Options, ,Choosing Modes}).
1359
1360To display the list of init files loaded by gdb at startup, you
1361can use @kbd{gdb --help}.
1362
1363@cindex init file name
1364@cindex @file{.gdbinit}
1365@cindex @file{gdb.ini}
1366The @value{GDBN} init files are normally called @file{.gdbinit}.
1367The DJGPP port of @value{GDBN} uses the name @file{gdb.ini}, due to
1368the limitations of file names imposed by DOS filesystems. The Windows
1369port of @value{GDBN} uses the standard name, but if it finds a
1370@file{gdb.ini} file in your home directory, it warns you about that
1371and suggests to rename the file to the standard name.
1372
1373
1374@node Quitting GDB
1375@section Quitting @value{GDBN}
1376@cindex exiting @value{GDBN}
1377@cindex leaving @value{GDBN}
1378
1379@table @code
1380@kindex quit @r{[}@var{expression}@r{]}
1381@kindex q @r{(@code{quit})}
1382@item quit @r{[}@var{expression}@r{]}
1383@itemx q
1384To exit @value{GDBN}, use the @code{quit} command (abbreviated
1385@code{q}), or type an end-of-file character (usually @kbd{Ctrl-d}). If you
1386do not supply @var{expression}, @value{GDBN} will terminate normally;
1387otherwise it will terminate using the result of @var{expression} as the
1388error code.
1389@end table
1390
1391@cindex interrupt
1392An interrupt (often @kbd{Ctrl-c}) does not exit from @value{GDBN}, but rather
1393terminates the action of any @value{GDBN} command that is in progress and
1394returns to @value{GDBN} command level. It is safe to type the interrupt
1395character at any time because @value{GDBN} does not allow it to take effect
1396until a time when it is safe.
1397
1398If you have been using @value{GDBN} to control an attached process or
1399device, you can release it with the @code{detach} command
1400(@pxref{Attach, ,Debugging an Already-running Process}).
1401
1402@node Shell Commands
1403@section Shell Commands
1404
1405If you need to execute occasional shell commands during your
1406debugging session, there is no need to leave or suspend @value{GDBN}; you can
1407just use the @code{shell} command.
1408
1409@table @code
1410@kindex shell
1411@kindex !
1412@cindex shell escape
1413@item shell @var{command-string}
1414@itemx !@var{command-string}
1415Invoke a standard shell to execute @var{command-string}.
1416Note that no space is needed between @code{!} and @var{command-string}.
1417If it exists, the environment variable @code{SHELL} determines which
1418shell to run. Otherwise @value{GDBN} uses the default shell
1419(@file{/bin/sh} on Unix systems, @file{COMMAND.COM} on MS-DOS, etc.).
1420@end table
1421
1422The utility @code{make} is often needed in development environments.
1423You do not have to use the @code{shell} command for this purpose in
1424@value{GDBN}:
1425
1426@table @code
1427@kindex make
1428@cindex calling make
1429@item make @var{make-args}
1430Execute the @code{make} program with the specified
1431arguments. This is equivalent to @samp{shell make @var{make-args}}.
1432@end table
1433
1434@node Logging Output
1435@section Logging Output
1436@cindex logging @value{GDBN} output
1437@cindex save @value{GDBN} output to a file
1438
1439You may want to save the output of @value{GDBN} commands to a file.
1440There are several commands to control @value{GDBN}'s logging.
1441
1442@table @code
1443@kindex set logging
1444@item set logging on
1445Enable logging.
1446@item set logging off
1447Disable logging.
1448@cindex logging file name
1449@item set logging file @var{file}
1450Change the name of the current logfile. The default logfile is @file{gdb.txt}.
1451@item set logging overwrite [on|off]
1452By default, @value{GDBN} will append to the logfile. Set @code{overwrite} if
1453you want @code{set logging on} to overwrite the logfile instead.
1454@item set logging redirect [on|off]
1455By default, @value{GDBN} output will go to both the terminal and the logfile.
1456Set @code{redirect} if you want output to go only to the log file.
1457@kindex show logging
1458@item show logging
1459Show the current values of the logging settings.
1460@end table
1461
1462@node Commands
1463@chapter @value{GDBN} Commands
1464
1465You can abbreviate a @value{GDBN} command to the first few letters of the command
1466name, if that abbreviation is unambiguous; and you can repeat certain
1467@value{GDBN} commands by typing just @key{RET}. You can also use the @key{TAB}
1468key to get @value{GDBN} to fill out the rest of a word in a command (or to
1469show you the alternatives available, if there is more than one possibility).
1470
1471@menu
1472* Command Syntax:: How to give commands to @value{GDBN}
1473* Completion:: Command completion
1474* Help:: How to ask @value{GDBN} for help
1475@end menu
1476
1477@node Command Syntax
1478@section Command Syntax
1479
1480A @value{GDBN} command is a single line of input. There is no limit on
1481how long it can be. It starts with a command name, which is followed by
1482arguments whose meaning depends on the command name. For example, the
1483command @code{step} accepts an argument which is the number of times to
1484step, as in @samp{step 5}. You can also use the @code{step} command
1485with no arguments. Some commands do not allow any arguments.
1486
1487@cindex abbreviation
1488@value{GDBN} command names may always be truncated if that abbreviation is
1489unambiguous. Other possible command abbreviations are listed in the
1490documentation for individual commands. In some cases, even ambiguous
1491abbreviations are allowed; for example, @code{s} is specially defined as
1492equivalent to @code{step} even though there are other commands whose
1493names start with @code{s}. You can test abbreviations by using them as
1494arguments to the @code{help} command.
1495
1496@cindex repeating commands
1497@kindex RET @r{(repeat last command)}
1498A blank line as input to @value{GDBN} (typing just @key{RET}) means to
1499repeat the previous command. Certain commands (for example, @code{run})
1500will not repeat this way; these are commands whose unintentional
1501repetition might cause trouble and which you are unlikely to want to
1502repeat. User-defined commands can disable this feature; see
1503@ref{Define, dont-repeat}.
1504
1505The @code{list} and @code{x} commands, when you repeat them with
1506@key{RET}, construct new arguments rather than repeating
1507exactly as typed. This permits easy scanning of source or memory.
1508
1509@value{GDBN} can also use @key{RET} in another way: to partition lengthy
1510output, in a way similar to the common utility @code{more}
1511(@pxref{Screen Size,,Screen Size}). Since it is easy to press one
1512@key{RET} too many in this situation, @value{GDBN} disables command
1513repetition after any command that generates this sort of display.
1514
1515@kindex # @r{(a comment)}
1516@cindex comment
1517Any text from a @kbd{#} to the end of the line is a comment; it does
1518nothing. This is useful mainly in command files (@pxref{Command
1519Files,,Command Files}).
1520
1521@cindex repeating command sequences
1522@kindex Ctrl-o @r{(operate-and-get-next)}
1523The @kbd{Ctrl-o} binding is useful for repeating a complex sequence of
1524commands. This command accepts the current line, like @key{RET}, and
1525then fetches the next line relative to the current line from the history
1526for editing.
1527
1528@node Completion
1529@section Command Completion
1530
1531@cindex completion
1532@cindex word completion
1533@value{GDBN} can fill in the rest of a word in a command for you, if there is
1534only one possibility; it can also show you what the valid possibilities
1535are for the next word in a command, at any time. This works for @value{GDBN}
1536commands, @value{GDBN} subcommands, and the names of symbols in your program.
1537
1538Press the @key{TAB} key whenever you want @value{GDBN} to fill out the rest
1539of a word. If there is only one possibility, @value{GDBN} fills in the
1540word, and waits for you to finish the command (or press @key{RET} to
1541enter it). For example, if you type
1542
1543@c FIXME "@key" does not distinguish its argument sufficiently to permit
1544@c complete accuracy in these examples; space introduced for clarity.
1545@c If texinfo enhancements make it unnecessary, it would be nice to
1546@c replace " @key" by "@key" in the following...
1547@smallexample
1548(@value{GDBP}) info bre @key{TAB}
1549@end smallexample
1550
1551@noindent
1552@value{GDBN} fills in the rest of the word @samp{breakpoints}, since that is
1553the only @code{info} subcommand beginning with @samp{bre}:
1554
1555@smallexample
1556(@value{GDBP}) info breakpoints
1557@end smallexample
1558
1559@noindent
1560You can either press @key{RET} at this point, to run the @code{info
1561breakpoints} command, or backspace and enter something else, if
1562@samp{breakpoints} does not look like the command you expected. (If you
1563were sure you wanted @code{info breakpoints} in the first place, you
1564might as well just type @key{RET} immediately after @samp{info bre},
1565to exploit command abbreviations rather than command completion).
1566
1567If there is more than one possibility for the next word when you press
1568@key{TAB}, @value{GDBN} sounds a bell. You can either supply more
1569characters and try again, or just press @key{TAB} a second time;
1570@value{GDBN} displays all the possible completions for that word. For
1571example, you might want to set a breakpoint on a subroutine whose name
1572begins with @samp{make_}, but when you type @kbd{b make_@key{TAB}} @value{GDBN}
1573just sounds the bell. Typing @key{TAB} again displays all the
1574function names in your program that begin with those characters, for
1575example:
1576
1577@smallexample
1578(@value{GDBP}) b make_ @key{TAB}
1579@exdent @value{GDBN} sounds bell; press @key{TAB} again, to see:
1580make_a_section_from_file make_environ
1581make_abs_section make_function_type
1582make_blockvector make_pointer_type
1583make_cleanup make_reference_type
1584make_command make_symbol_completion_list
1585(@value{GDBP}) b make_
1586@end smallexample
1587
1588@noindent
1589After displaying the available possibilities, @value{GDBN} copies your
1590partial input (@samp{b make_} in the example) so you can finish the
1591command.
1592
1593If you just want to see the list of alternatives in the first place, you
1594can press @kbd{M-?} rather than pressing @key{TAB} twice. @kbd{M-?}
1595means @kbd{@key{META} ?}. You can type this either by holding down a
1596key designated as the @key{META} shift on your keyboard (if there is
1597one) while typing @kbd{?}, or as @key{ESC} followed by @kbd{?}.
1598
1599@cindex quotes in commands
1600@cindex completion of quoted strings
1601Sometimes the string you need, while logically a ``word'', may contain
1602parentheses or other characters that @value{GDBN} normally excludes from
1603its notion of a word. To permit word completion to work in this
1604situation, you may enclose words in @code{'} (single quote marks) in
1605@value{GDBN} commands.
1606
1607The most likely situation where you might need this is in typing the
1608name of a C@t{++} function. This is because C@t{++} allows function
1609overloading (multiple definitions of the same function, distinguished
1610by argument type). For example, when you want to set a breakpoint you
1611may need to distinguish whether you mean the version of @code{name}
1612that takes an @code{int} parameter, @code{name(int)}, or the version
1613that takes a @code{float} parameter, @code{name(float)}. To use the
1614word-completion facilities in this situation, type a single quote
1615@code{'} at the beginning of the function name. This alerts
1616@value{GDBN} that it may need to consider more information than usual
1617when you press @key{TAB} or @kbd{M-?} to request word completion:
1618
1619@smallexample
1620(@value{GDBP}) b 'bubble( @kbd{M-?}
1621bubble(double,double) bubble(int,int)
1622(@value{GDBP}) b 'bubble(
1623@end smallexample
1624
1625In some cases, @value{GDBN} can tell that completing a name requires using
1626quotes. When this happens, @value{GDBN} inserts the quote for you (while
1627completing as much as it can) if you do not type the quote in the first
1628place:
1629
1630@smallexample
1631(@value{GDBP}) b bub @key{TAB}
1632@exdent @value{GDBN} alters your input line to the following, and rings a bell:
1633(@value{GDBP}) b 'bubble(
1634@end smallexample
1635
1636@noindent
1637In general, @value{GDBN} can tell that a quote is needed (and inserts it) if
1638you have not yet started typing the argument list when you ask for
1639completion on an overloaded symbol.
1640
1641For more information about overloaded functions, see @ref{C Plus Plus
1642Expressions, ,C@t{++} Expressions}. You can use the command @code{set
1643overload-resolution off} to disable overload resolution;
1644see @ref{Debugging C Plus Plus, ,@value{GDBN} Features for C@t{++}}.
1645
1646@cindex completion of structure field names
1647@cindex structure field name completion
1648@cindex completion of union field names
1649@cindex union field name completion
1650When completing in an expression which looks up a field in a
1651structure, @value{GDBN} also tries@footnote{The completer can be
1652confused by certain kinds of invalid expressions. Also, it only
1653examines the static type of the expression, not the dynamic type.} to
1654limit completions to the field names available in the type of the
1655left-hand-side:
1656
1657@smallexample
1658(@value{GDBP}) p gdb_stdout.@kbd{M-?}
1659magic to_fputs to_rewind
1660to_data to_isatty to_write
1661to_delete to_put to_write_async_safe
1662to_flush to_read
1663@end smallexample
1664
1665@noindent
1666This is because the @code{gdb_stdout} is a variable of the type
1667@code{struct ui_file} that is defined in @value{GDBN} sources as
1668follows:
1669
1670@smallexample
1671struct ui_file
1672@{
1673 int *magic;
1674 ui_file_flush_ftype *to_flush;
1675 ui_file_write_ftype *to_write;
1676 ui_file_write_async_safe_ftype *to_write_async_safe;
1677 ui_file_fputs_ftype *to_fputs;
1678 ui_file_read_ftype *to_read;
1679 ui_file_delete_ftype *to_delete;
1680 ui_file_isatty_ftype *to_isatty;
1681 ui_file_rewind_ftype *to_rewind;
1682 ui_file_put_ftype *to_put;
1683 void *to_data;
1684@}
1685@end smallexample
1686
1687
1688@node Help
1689@section Getting Help
1690@cindex online documentation
1691@kindex help
1692
1693You can always ask @value{GDBN} itself for information on its commands,
1694using the command @code{help}.
1695
1696@table @code
1697@kindex h @r{(@code{help})}
1698@item help
1699@itemx h
1700You can use @code{help} (abbreviated @code{h}) with no arguments to
1701display a short list of named classes of commands:
1702
1703@smallexample
1704(@value{GDBP}) help
1705List of classes of commands:
1706
1707aliases -- Aliases of other commands
1708breakpoints -- Making program stop at certain points
1709data -- Examining data
1710files -- Specifying and examining files
1711internals -- Maintenance commands
1712obscure -- Obscure features
1713running -- Running the program
1714stack -- Examining the stack
1715status -- Status inquiries
1716support -- Support facilities
1717tracepoints -- Tracing of program execution without
1718 stopping the program
1719user-defined -- User-defined commands
1720
1721Type "help" followed by a class name for a list of
1722commands in that class.
1723Type "help" followed by command name for full
1724documentation.
1725Command name abbreviations are allowed if unambiguous.
1726(@value{GDBP})
1727@end smallexample
1728@c the above line break eliminates huge line overfull...
1729
1730@item help @var{class}
1731Using one of the general help classes as an argument, you can get a
1732list of the individual commands in that class. For example, here is the
1733help display for the class @code{status}:
1734
1735@smallexample
1736(@value{GDBP}) help status
1737Status inquiries.
1738
1739List of commands:
1740
1741@c Line break in "show" line falsifies real output, but needed
1742@c to fit in smallbook page size.
1743info -- Generic command for showing things
1744 about the program being debugged
1745show -- Generic command for showing things
1746 about the debugger
1747
1748Type "help" followed by command name for full
1749documentation.
1750Command name abbreviations are allowed if unambiguous.
1751(@value{GDBP})
1752@end smallexample
1753
1754@item help @var{command}
1755With a command name as @code{help} argument, @value{GDBN} displays a
1756short paragraph on how to use that command.
1757
1758@kindex apropos
1759@item apropos @var{args}
1760The @code{apropos} command searches through all of the @value{GDBN}
1761commands, and their documentation, for the regular expression specified in
1762@var{args}. It prints out all matches found. For example:
1763
1764@smallexample
1765apropos alias
1766@end smallexample
1767
1768@noindent
1769results in:
1770
1771@smallexample
1772@c @group
1773alias -- Define a new command that is an alias of an existing command
1774aliases -- Aliases of other commands
1775d -- Delete some breakpoints or auto-display expressions
1776del -- Delete some breakpoints or auto-display expressions
1777delete -- Delete some breakpoints or auto-display expressions
1778@c @end group
1779@end smallexample
1780
1781@kindex complete
1782@item complete @var{args}
1783The @code{complete @var{args}} command lists all the possible completions
1784for the beginning of a command. Use @var{args} to specify the beginning of the
1785command you want completed. For example:
1786
1787@smallexample
1788complete i
1789@end smallexample
1790
1791@noindent results in:
1792
1793@smallexample
1794@group
1795if
1796ignore
1797info
1798inspect
1799@end group
1800@end smallexample
1801
1802@noindent This is intended for use by @sc{gnu} Emacs.
1803@end table
1804
1805In addition to @code{help}, you can use the @value{GDBN} commands @code{info}
1806and @code{show} to inquire about the state of your program, or the state
1807of @value{GDBN} itself. Each command supports many topics of inquiry; this
1808manual introduces each of them in the appropriate context. The listings
1809under @code{info} and under @code{show} in the Command, Variable, and
1810Function Index point to all the sub-commands. @xref{Command and Variable
1811Index}.
1812
1813@c @group
1814@table @code
1815@kindex info
1816@kindex i @r{(@code{info})}
1817@item info
1818This command (abbreviated @code{i}) is for describing the state of your
1819program. For example, you can show the arguments passed to a function
1820with @code{info args}, list the registers currently in use with @code{info
1821registers}, or list the breakpoints you have set with @code{info breakpoints}.
1822You can get a complete list of the @code{info} sub-commands with
1823@w{@code{help info}}.
1824
1825@kindex set
1826@item set
1827You can assign the result of an expression to an environment variable with
1828@code{set}. For example, you can set the @value{GDBN} prompt to a $-sign with
1829@code{set prompt $}.
1830
1831@kindex show
1832@item show
1833In contrast to @code{info}, @code{show} is for describing the state of
1834@value{GDBN} itself.
1835You can change most of the things you can @code{show}, by using the
1836related command @code{set}; for example, you can control what number
1837system is used for displays with @code{set radix}, or simply inquire
1838which is currently in use with @code{show radix}.
1839
1840@kindex info set
1841To display all the settable parameters and their current
1842values, you can use @code{show} with no arguments; you may also use
1843@code{info set}. Both commands produce the same display.
1844@c FIXME: "info set" violates the rule that "info" is for state of
1845@c FIXME...program. Ck w/ GNU: "info set" to be called something else,
1846@c FIXME...or change desc of rule---eg "state of prog and debugging session"?
1847@end table
1848@c @end group
1849
1850Here are several miscellaneous @code{show} subcommands, all of which are
1851exceptional in lacking corresponding @code{set} commands:
1852
1853@table @code
1854@kindex show version
1855@cindex @value{GDBN} version number
1856@item show version
1857Show what version of @value{GDBN} is running. You should include this
1858information in @value{GDBN} bug-reports. If multiple versions of
1859@value{GDBN} are in use at your site, you may need to determine which
1860version of @value{GDBN} you are running; as @value{GDBN} evolves, new
1861commands are introduced, and old ones may wither away. Also, many
1862system vendors ship variant versions of @value{GDBN}, and there are
1863variant versions of @value{GDBN} in @sc{gnu}/Linux distributions as well.
1864The version number is the same as the one announced when you start
1865@value{GDBN}.
1866
1867@kindex show copying
1868@kindex info copying
1869@cindex display @value{GDBN} copyright
1870@item show copying
1871@itemx info copying
1872Display information about permission for copying @value{GDBN}.
1873
1874@kindex show warranty
1875@kindex info warranty
1876@item show warranty
1877@itemx info warranty
1878Display the @sc{gnu} ``NO WARRANTY'' statement, or a warranty,
1879if your version of @value{GDBN} comes with one.
1880
1881@kindex show configuration
1882@item show configuration
1883Display detailed information about the way @value{GDBN} was configured
1884when it was built. This displays the optional arguments passed to the
1885@file{configure} script and also configuration parameters detected
1886automatically by @command{configure}. When reporting a @value{GDBN}
1887bug (@pxref{GDB Bugs}), it is important to include this information in
1888your report.
1889
1890@end table
1891
1892@node Running
1893@chapter Running Programs Under @value{GDBN}
1894
1895When you run a program under @value{GDBN}, you must first generate
1896debugging information when you compile it.
1897
1898You may start @value{GDBN} with its arguments, if any, in an environment
1899of your choice. If you are doing native debugging, you may redirect
1900your program's input and output, debug an already running process, or
1901kill a child process.
1902
1903@menu
1904* Compilation:: Compiling for debugging
1905* Starting:: Starting your program
1906* Arguments:: Your program's arguments
1907* Environment:: Your program's environment
1908
1909* Working Directory:: Your program's working directory
1910* Input/Output:: Your program's input and output
1911* Attach:: Debugging an already-running process
1912* Kill Process:: Killing the child process
1913
1914* Inferiors and Programs:: Debugging multiple inferiors and programs
1915* Threads:: Debugging programs with multiple threads
1916* Forks:: Debugging forks
1917* Checkpoint/Restart:: Setting a @emph{bookmark} to return to later
1918@end menu
1919
1920@node Compilation
1921@section Compiling for Debugging
1922
1923In order to debug a program effectively, you need to generate
1924debugging information when you compile it. This debugging information
1925is stored in the object file; it describes the data type of each
1926variable or function and the correspondence between source line numbers
1927and addresses in the executable code.
1928
1929To request debugging information, specify the @samp{-g} option when you run
1930the compiler.
1931
1932Programs that are to be shipped to your customers are compiled with
1933optimizations, using the @samp{-O} compiler option. However, some
1934compilers are unable to handle the @samp{-g} and @samp{-O} options
1935together. Using those compilers, you cannot generate optimized
1936executables containing debugging information.
1937
1938@value{NGCC}, the @sc{gnu} C/C@t{++} compiler, supports @samp{-g} with or
1939without @samp{-O}, making it possible to debug optimized code. We
1940recommend that you @emph{always} use @samp{-g} whenever you compile a
1941program. You may think your program is correct, but there is no sense
1942in pushing your luck. For more information, see @ref{Optimized Code}.
1943
1944Older versions of the @sc{gnu} C compiler permitted a variant option
1945@w{@samp{-gg}} for debugging information. @value{GDBN} no longer supports this
1946format; if your @sc{gnu} C compiler has this option, do not use it.
1947
1948@value{GDBN} knows about preprocessor macros and can show you their
1949expansion (@pxref{Macros}). Most compilers do not include information
1950about preprocessor macros in the debugging information if you specify
1951the @option{-g} flag alone. Version 3.1 and later of @value{NGCC},
1952the @sc{gnu} C compiler, provides macro information if you are using
1953the DWARF debugging format, and specify the option @option{-g3}.
1954
1955@xref{Debugging Options,,Options for Debugging Your Program or GCC,
1956gcc.info, Using the @sc{gnu} Compiler Collection (GCC)}, for more
1957information on @value{NGCC} options affecting debug information.
1958
1959You will have the best debugging experience if you use the latest
1960version of the DWARF debugging format that your compiler supports.
1961DWARF is currently the most expressive and best supported debugging
1962format in @value{GDBN}.
1963
1964@need 2000
1965@node Starting
1966@section Starting your Program
1967@cindex starting
1968@cindex running
1969
1970@table @code
1971@kindex run
1972@kindex r @r{(@code{run})}
1973@item run
1974@itemx r
1975Use the @code{run} command to start your program under @value{GDBN}.
1976You must first specify the program name (except on VxWorks) with an
1977argument to @value{GDBN} (@pxref{Invocation, ,Getting In and Out of
1978@value{GDBN}}), or by using the @code{file} or @code{exec-file} command
1979(@pxref{Files, ,Commands to Specify Files}).
1980
1981@end table
1982
1983If you are running your program in an execution environment that
1984supports processes, @code{run} creates an inferior process and makes
1985that process run your program. In some environments without processes,
1986@code{run} jumps to the start of your program. Other targets,
1987like @samp{remote}, are always running. If you get an error
1988message like this one:
1989
1990@smallexample
1991The "remote" target does not support "run".
1992Try "help target" or "continue".
1993@end smallexample
1994
1995@noindent
1996then use @code{continue} to run your program. You may need @code{load}
1997first (@pxref{load}).
1998
1999The execution of a program is affected by certain information it
2000receives from its superior. @value{GDBN} provides ways to specify this
2001information, which you must do @emph{before} starting your program. (You
2002can change it after starting your program, but such changes only affect
2003your program the next time you start it.) This information may be
2004divided into four categories:
2005
2006@table @asis
2007@item The @emph{arguments.}
2008Specify the arguments to give your program as the arguments of the
2009@code{run} command. If a shell is available on your target, the shell
2010is used to pass the arguments, so that you may use normal conventions
2011(such as wildcard expansion or variable substitution) in describing
2012the arguments.
2013In Unix systems, you can control which shell is used with the
2014@code{SHELL} environment variable. If you do not define @code{SHELL},
2015@value{GDBN} uses the default shell (@file{/bin/sh}). You can disable
2016use of any shell with the @code{set startup-with-shell} command (see
2017below for details).
2018
2019@item The @emph{environment.}
2020Your program normally inherits its environment from @value{GDBN}, but you can
2021use the @value{GDBN} commands @code{set environment} and @code{unset
2022environment} to change parts of the environment that affect
2023your program. @xref{Environment, ,Your Program's Environment}.
2024
2025@item The @emph{working directory.}
2026Your program inherits its working directory from @value{GDBN}. You can set
2027the @value{GDBN} working directory with the @code{cd} command in @value{GDBN}.
2028@xref{Working Directory, ,Your Program's Working Directory}.
2029
2030@item The @emph{standard input and output.}
2031Your program normally uses the same device for standard input and
2032standard output as @value{GDBN} is using. You can redirect input and output
2033in the @code{run} command line, or you can use the @code{tty} command to
2034set a different device for your program.
2035@xref{Input/Output, ,Your Program's Input and Output}.
2036
2037@cindex pipes
2038@emph{Warning:} While input and output redirection work, you cannot use
2039pipes to pass the output of the program you are debugging to another
2040program; if you attempt this, @value{GDBN} is likely to wind up debugging the
2041wrong program.
2042@end table
2043
2044When you issue the @code{run} command, your program begins to execute
2045immediately. @xref{Stopping, ,Stopping and Continuing}, for discussion
2046of how to arrange for your program to stop. Once your program has
2047stopped, you may call functions in your program, using the @code{print}
2048or @code{call} commands. @xref{Data, ,Examining Data}.
2049
2050If the modification time of your symbol file has changed since the last
2051time @value{GDBN} read its symbols, @value{GDBN} discards its symbol
2052table, and reads it again. When it does this, @value{GDBN} tries to retain
2053your current breakpoints.
2054
2055@table @code
2056@kindex start
2057@item start
2058@cindex run to main procedure
2059The name of the main procedure can vary from language to language.
2060With C or C@t{++}, the main procedure name is always @code{main}, but
2061other languages such as Ada do not require a specific name for their
2062main procedure. The debugger provides a convenient way to start the
2063execution of the program and to stop at the beginning of the main
2064procedure, depending on the language used.
2065
2066The @samp{start} command does the equivalent of setting a temporary
2067breakpoint at the beginning of the main procedure and then invoking
2068the @samp{run} command.
2069
2070@cindex elaboration phase
2071Some programs contain an @dfn{elaboration} phase where some startup code is
2072executed before the main procedure is called. This depends on the
2073languages used to write your program. In C@t{++}, for instance,
2074constructors for static and global objects are executed before
2075@code{main} is called. It is therefore possible that the debugger stops
2076before reaching the main procedure. However, the temporary breakpoint
2077will remain to halt execution.
2078
2079Specify the arguments to give to your program as arguments to the
2080@samp{start} command. These arguments will be given verbatim to the
2081underlying @samp{run} command. Note that the same arguments will be
2082reused if no argument is provided during subsequent calls to
2083@samp{start} or @samp{run}.
2084
2085It is sometimes necessary to debug the program during elaboration. In
2086these cases, using the @code{start} command would stop the execution of
2087your program too late, as the program would have already completed the
2088elaboration phase. Under these circumstances, insert breakpoints in your
2089elaboration code before running your program.
2090
2091@kindex set exec-wrapper
2092@item set exec-wrapper @var{wrapper}
2093@itemx show exec-wrapper
2094@itemx unset exec-wrapper
2095When @samp{exec-wrapper} is set, the specified wrapper is used to
2096launch programs for debugging. @value{GDBN} starts your program
2097with a shell command of the form @kbd{exec @var{wrapper}
2098@var{program}}. Quoting is added to @var{program} and its
2099arguments, but not to @var{wrapper}, so you should add quotes if
2100appropriate for your shell. The wrapper runs until it executes
2101your program, and then @value{GDBN} takes control.
2102
2103You can use any program that eventually calls @code{execve} with
2104its arguments as a wrapper. Several standard Unix utilities do
2105this, e.g.@: @code{env} and @code{nohup}. Any Unix shell script ending
2106with @code{exec "$@@"} will also work.
2107
2108For example, you can use @code{env} to pass an environment variable to
2109the debugged program, without setting the variable in your shell's
2110environment:
2111
2112@smallexample
2113(@value{GDBP}) set exec-wrapper env 'LD_PRELOAD=libtest.so'
2114(@value{GDBP}) run
2115@end smallexample
2116
2117This command is available when debugging locally on most targets, excluding
2118@sc{djgpp}, Cygwin, MS Windows, and QNX Neutrino.
2119
2120@kindex set startup-with-shell
2121@item set startup-with-shell
2122@itemx set startup-with-shell on
2123@itemx set startup-with-shell off
2124@itemx show set startup-with-shell
2125On Unix systems, by default, if a shell is available on your target,
2126@value{GDBN}) uses it to start your program. Arguments of the
2127@code{run} command are passed to the shell, which does variable
2128substitution, expands wildcard characters and performs redirection of
2129I/O. In some circumstances, it may be useful to disable such use of a
2130shell, for example, when debugging the shell itself or diagnosing
2131startup failures such as:
2132
2133@smallexample
2134(@value{GDBP}) run
2135Starting program: ./a.out
2136During startup program terminated with signal SIGSEGV, Segmentation fault.
2137@end smallexample
2138
2139@noindent
2140which indicates the shell or the wrapper specified with
2141@samp{exec-wrapper} crashed, not your program. Most often, this is
2142caused by something odd in your shell's non-interactive mode
2143initialization file---such as @file{.cshrc} for C-shell,
2144$@file{.zshenv} for the Z shell, or the file specified in the
2145@samp{BASH_ENV} environment variable for BASH.
2146
2147@kindex set disable-randomization
2148@item set disable-randomization
2149@itemx set disable-randomization on
2150This option (enabled by default in @value{GDBN}) will turn off the native
2151randomization of the virtual address space of the started program. This option
2152is useful for multiple debugging sessions to make the execution better
2153reproducible and memory addresses reusable across debugging sessions.
2154
2155This feature is implemented only on certain targets, including @sc{gnu}/Linux.
2156On @sc{gnu}/Linux you can get the same behavior using
2157
2158@smallexample
2159(@value{GDBP}) set exec-wrapper setarch `uname -m` -R
2160@end smallexample
2161
2162@item set disable-randomization off
2163Leave the behavior of the started executable unchanged. Some bugs rear their
2164ugly heads only when the program is loaded at certain addresses. If your bug
2165disappears when you run the program under @value{GDBN}, that might be because
2166@value{GDBN} by default disables the address randomization on platforms, such
2167as @sc{gnu}/Linux, which do that for stand-alone programs. Use @kbd{set
2168disable-randomization off} to try to reproduce such elusive bugs.
2169
2170On targets where it is available, virtual address space randomization
2171protects the programs against certain kinds of security attacks. In these
2172cases the attacker needs to know the exact location of a concrete executable
2173code. Randomizing its location makes it impossible to inject jumps misusing
2174a code at its expected addresses.
2175
2176Prelinking shared libraries provides a startup performance advantage but it
2177makes addresses in these libraries predictable for privileged processes by
2178having just unprivileged access at the target system. Reading the shared
2179library binary gives enough information for assembling the malicious code
2180misusing it. Still even a prelinked shared library can get loaded at a new
2181random address just requiring the regular relocation process during the
2182startup. Shared libraries not already prelinked are always loaded at
2183a randomly chosen address.
2184
2185Position independent executables (PIE) contain position independent code
2186similar to the shared libraries and therefore such executables get loaded at
2187a randomly chosen address upon startup. PIE executables always load even
2188already prelinked shared libraries at a random address. You can build such
2189executable using @command{gcc -fPIE -pie}.
2190
2191Heap (malloc storage), stack and custom mmap areas are always placed randomly
2192(as long as the randomization is enabled).
2193
2194@item show disable-randomization
2195Show the current setting of the explicit disable of the native randomization of
2196the virtual address space of the started program.
2197
2198@end table
2199
2200@node Arguments
2201@section Your Program's Arguments
2202
2203@cindex arguments (to your program)
2204The arguments to your program can be specified by the arguments of the
2205@code{run} command.
2206They are passed to a shell, which expands wildcard characters and
2207performs redirection of I/O, and thence to your program. Your
2208@code{SHELL} environment variable (if it exists) specifies what shell
2209@value{GDBN} uses. If you do not define @code{SHELL}, @value{GDBN} uses
2210the default shell (@file{/bin/sh} on Unix).
2211
2212On non-Unix systems, the program is usually invoked directly by
2213@value{GDBN}, which emulates I/O redirection via the appropriate system
2214calls, and the wildcard characters are expanded by the startup code of
2215the program, not by the shell.
2216
2217@code{run} with no arguments uses the same arguments used by the previous
2218@code{run}, or those set by the @code{set args} command.
2219
2220@table @code
2221@kindex set args
2222@item set args
2223Specify the arguments to be used the next time your program is run. If
2224@code{set args} has no arguments, @code{run} executes your program
2225with no arguments. Once you have run your program with arguments,
2226using @code{set args} before the next @code{run} is the only way to run
2227it again without arguments.
2228
2229@kindex show args
2230@item show args
2231Show the arguments to give your program when it is started.
2232@end table
2233
2234@node Environment
2235@section Your Program's Environment
2236
2237@cindex environment (of your program)
2238The @dfn{environment} consists of a set of environment variables and
2239their values. Environment variables conventionally record such things as
2240your user name, your home directory, your terminal type, and your search
2241path for programs to run. Usually you set up environment variables with
2242the shell and they are inherited by all the other programs you run. When
2243debugging, it can be useful to try running your program with a modified
2244environment without having to start @value{GDBN} over again.
2245
2246@table @code
2247@kindex path
2248@item path @var{directory}
2249Add @var{directory} to the front of the @code{PATH} environment variable
2250(the search path for executables) that will be passed to your program.
2251The value of @code{PATH} used by @value{GDBN} does not change.
2252You may specify several directory names, separated by whitespace or by a
2253system-dependent separator character (@samp{:} on Unix, @samp{;} on
2254MS-DOS and MS-Windows). If @var{directory} is already in the path, it
2255is moved to the front, so it is searched sooner.
2256
2257You can use the string @samp{$cwd} to refer to whatever is the current
2258working directory at the time @value{GDBN} searches the path. If you
2259use @samp{.} instead, it refers to the directory where you executed the
2260@code{path} command. @value{GDBN} replaces @samp{.} in the
2261@var{directory} argument (with the current path) before adding
2262@var{directory} to the search path.
2263@c 'path' is explicitly nonrepeatable, but RMS points out it is silly to
2264@c document that, since repeating it would be a no-op.
2265
2266@kindex show paths
2267@item show paths
2268Display the list of search paths for executables (the @code{PATH}
2269environment variable).
2270
2271@kindex show environment
2272@item show environment @r{[}@var{varname}@r{]}
2273Print the value of environment variable @var{varname} to be given to
2274your program when it starts. If you do not supply @var{varname},
2275print the names and values of all environment variables to be given to
2276your program. You can abbreviate @code{environment} as @code{env}.
2277
2278@kindex set environment
2279@item set environment @var{varname} @r{[}=@var{value}@r{]}
2280Set environment variable @var{varname} to @var{value}. The value
2281changes for your program only, not for @value{GDBN} itself. @var{value} may
2282be any string; the values of environment variables are just strings, and
2283any interpretation is supplied by your program itself. The @var{value}
2284parameter is optional; if it is eliminated, the variable is set to a
2285null value.
2286@c "any string" here does not include leading, trailing
2287@c blanks. Gnu asks: does anyone care?
2288
2289For example, this command:
2290
2291@smallexample
2292set env USER = foo
2293@end smallexample
2294
2295@noindent
2296tells the debugged program, when subsequently run, that its user is named
2297@samp{foo}. (The spaces around @samp{=} are used for clarity here; they
2298are not actually required.)
2299
2300@kindex unset environment
2301@item unset environment @var{varname}
2302Remove variable @var{varname} from the environment to be passed to your
2303program. This is different from @samp{set env @var{varname} =};
2304@code{unset environment} removes the variable from the environment,
2305rather than assigning it an empty value.
2306@end table
2307
2308@emph{Warning:} On Unix systems, @value{GDBN} runs your program using
2309the shell indicated by your @code{SHELL} environment variable if it
2310exists (or @code{/bin/sh} if not). If your @code{SHELL} variable
2311names a shell that runs an initialization file when started
2312non-interactively---such as @file{.cshrc} for C-shell, $@file{.zshenv}
2313for the Z shell, or the file specified in the @samp{BASH_ENV}
2314environment variable for BASH---any variables you set in that file
2315affect your program. You may wish to move setting of environment
2316variables to files that are only run when you sign on, such as
2317@file{.login} or @file{.profile}.
2318
2319@node Working Directory
2320@section Your Program's Working Directory
2321
2322@cindex working directory (of your program)
2323Each time you start your program with @code{run}, it inherits its
2324working directory from the current working directory of @value{GDBN}.
2325The @value{GDBN} working directory is initially whatever it inherited
2326from its parent process (typically the shell), but you can specify a new
2327working directory in @value{GDBN} with the @code{cd} command.
2328
2329The @value{GDBN} working directory also serves as a default for the commands
2330that specify files for @value{GDBN} to operate on. @xref{Files, ,Commands to
2331Specify Files}.
2332
2333@table @code
2334@kindex cd
2335@cindex change working directory
2336@item cd @r{[}@var{directory}@r{]}
2337Set the @value{GDBN} working directory to @var{directory}. If not
2338given, @var{directory} uses @file{'~'}.
2339
2340@kindex pwd
2341@item pwd
2342Print the @value{GDBN} working directory.
2343@end table
2344
2345It is generally impossible to find the current working directory of
2346the process being debugged (since a program can change its directory
2347during its run). If you work on a system where @value{GDBN} is
2348configured with the @file{/proc} support, you can use the @code{info
2349proc} command (@pxref{SVR4 Process Information}) to find out the
2350current working directory of the debuggee.
2351
2352@node Input/Output
2353@section Your Program's Input and Output
2354
2355@cindex redirection
2356@cindex i/o
2357@cindex terminal
2358By default, the program you run under @value{GDBN} does input and output to
2359the same terminal that @value{GDBN} uses. @value{GDBN} switches the terminal
2360to its own terminal modes to interact with you, but it records the terminal
2361modes your program was using and switches back to them when you continue
2362running your program.
2363
2364@table @code
2365@kindex info terminal
2366@item info terminal
2367Displays information recorded by @value{GDBN} about the terminal modes your
2368program is using.
2369@end table
2370
2371You can redirect your program's input and/or output using shell
2372redirection with the @code{run} command. For example,
2373
2374@smallexample
2375run > outfile
2376@end smallexample
2377
2378@noindent
2379starts your program, diverting its output to the file @file{outfile}.
2380
2381@kindex tty
2382@cindex controlling terminal
2383Another way to specify where your program should do input and output is
2384with the @code{tty} command. This command accepts a file name as
2385argument, and causes this file to be the default for future @code{run}
2386commands. It also resets the controlling terminal for the child
2387process, for future @code{run} commands. For example,
2388
2389@smallexample
2390tty /dev/ttyb
2391@end smallexample
2392
2393@noindent
2394directs that processes started with subsequent @code{run} commands
2395default to do input and output on the terminal @file{/dev/ttyb} and have
2396that as their controlling terminal.
2397
2398An explicit redirection in @code{run} overrides the @code{tty} command's
2399effect on the input/output device, but not its effect on the controlling
2400terminal.
2401
2402When you use the @code{tty} command or redirect input in the @code{run}
2403command, only the input @emph{for your program} is affected. The input
2404for @value{GDBN} still comes from your terminal. @code{tty} is an alias
2405for @code{set inferior-tty}.
2406
2407@cindex inferior tty
2408@cindex set inferior controlling terminal
2409You can use the @code{show inferior-tty} command to tell @value{GDBN} to
2410display the name of the terminal that will be used for future runs of your
2411program.
2412
2413@table @code
2414@item set inferior-tty /dev/ttyb
2415@kindex set inferior-tty
2416Set the tty for the program being debugged to /dev/ttyb.
2417
2418@item show inferior-tty
2419@kindex show inferior-tty
2420Show the current tty for the program being debugged.
2421@end table
2422
2423@node Attach
2424@section Debugging an Already-running Process
2425@kindex attach
2426@cindex attach
2427
2428@table @code
2429@item attach @var{process-id}
2430This command attaches to a running process---one that was started
2431outside @value{GDBN}. (@code{info files} shows your active
2432targets.) The command takes as argument a process ID. The usual way to
2433find out the @var{process-id} of a Unix process is with the @code{ps} utility,
2434or with the @samp{jobs -l} shell command.
2435
2436@code{attach} does not repeat if you press @key{RET} a second time after
2437executing the command.
2438@end table
2439
2440To use @code{attach}, your program must be running in an environment
2441which supports processes; for example, @code{attach} does not work for
2442programs on bare-board targets that lack an operating system. You must
2443also have permission to send the process a signal.
2444
2445When you use @code{attach}, the debugger finds the program running in
2446the process first by looking in the current working directory, then (if
2447the program is not found) by using the source file search path
2448(@pxref{Source Path, ,Specifying Source Directories}). You can also use
2449the @code{file} command to load the program. @xref{Files, ,Commands to
2450Specify Files}.
2451
2452The first thing @value{GDBN} does after arranging to debug the specified
2453process is to stop it. You can examine and modify an attached process
2454with all the @value{GDBN} commands that are ordinarily available when
2455you start processes with @code{run}. You can insert breakpoints; you
2456can step and continue; you can modify storage. If you would rather the
2457process continue running, you may use the @code{continue} command after
2458attaching @value{GDBN} to the process.
2459
2460@table @code
2461@kindex detach
2462@item detach
2463When you have finished debugging the attached process, you can use the
2464@code{detach} command to release it from @value{GDBN} control. Detaching
2465the process continues its execution. After the @code{detach} command,
2466that process and @value{GDBN} become completely independent once more, and you
2467are ready to @code{attach} another process or start one with @code{run}.
2468@code{detach} does not repeat if you press @key{RET} again after
2469executing the command.
2470@end table
2471
2472If you exit @value{GDBN} while you have an attached process, you detach
2473that process. If you use the @code{run} command, you kill that process.
2474By default, @value{GDBN} asks for confirmation if you try to do either of these
2475things; you can control whether or not you need to confirm by using the
2476@code{set confirm} command (@pxref{Messages/Warnings, ,Optional Warnings and
2477Messages}).
2478
2479@node Kill Process
2480@section Killing the Child Process
2481
2482@table @code
2483@kindex kill
2484@item kill
2485Kill the child process in which your program is running under @value{GDBN}.
2486@end table
2487
2488This command is useful if you wish to debug a core dump instead of a
2489running process. @value{GDBN} ignores any core dump file while your program
2490is running.
2491
2492On some operating systems, a program cannot be executed outside @value{GDBN}
2493while you have breakpoints set on it inside @value{GDBN}. You can use the
2494@code{kill} command in this situation to permit running your program
2495outside the debugger.
2496
2497The @code{kill} command is also useful if you wish to recompile and
2498relink your program, since on many systems it is impossible to modify an
2499executable file while it is running in a process. In this case, when you
2500next type @code{run}, @value{GDBN} notices that the file has changed, and
2501reads the symbol table again (while trying to preserve your current
2502breakpoint settings).
2503
2504@node Inferiors and Programs
2505@section Debugging Multiple Inferiors and Programs
2506
2507@value{GDBN} lets you run and debug multiple programs in a single
2508session. In addition, @value{GDBN} on some systems may let you run
2509several programs simultaneously (otherwise you have to exit from one
2510before starting another). In the most general case, you can have
2511multiple threads of execution in each of multiple processes, launched
2512from multiple executables.
2513
2514@cindex inferior
2515@value{GDBN} represents the state of each program execution with an
2516object called an @dfn{inferior}. An inferior typically corresponds to
2517a process, but is more general and applies also to targets that do not
2518have processes. Inferiors may be created before a process runs, and
2519may be retained after a process exits. Inferiors have unique
2520identifiers that are different from process ids. Usually each
2521inferior will also have its own distinct address space, although some
2522embedded targets may have several inferiors running in different parts
2523of a single address space. Each inferior may in turn have multiple
2524threads running in it.
2525
2526To find out what inferiors exist at any moment, use @w{@code{info
2527inferiors}}:
2528
2529@table @code
2530@kindex info inferiors
2531@item info inferiors
2532Print a list of all inferiors currently being managed by @value{GDBN}.
2533
2534@value{GDBN} displays for each inferior (in this order):
2535
2536@enumerate
2537@item
2538the inferior number assigned by @value{GDBN}
2539
2540@item
2541the target system's inferior identifier
2542
2543@item
2544the name of the executable the inferior is running.
2545
2546@end enumerate
2547
2548@noindent
2549An asterisk @samp{*} preceding the @value{GDBN} inferior number
2550indicates the current inferior.
2551
2552For example,
2553@end table
2554@c end table here to get a little more width for example
2555
2556@smallexample
2557(@value{GDBP}) info inferiors
2558 Num Description Executable
2559 2 process 2307 hello
2560* 1 process 3401 goodbye
2561@end smallexample
2562
2563To switch focus between inferiors, use the @code{inferior} command:
2564
2565@table @code
2566@kindex inferior @var{infno}
2567@item inferior @var{infno}
2568Make inferior number @var{infno} the current inferior. The argument
2569@var{infno} is the inferior number assigned by @value{GDBN}, as shown
2570in the first field of the @samp{info inferiors} display.
2571@end table
2572
2573
2574You can get multiple executables into a debugging session via the
2575@code{add-inferior} and @w{@code{clone-inferior}} commands. On some
2576systems @value{GDBN} can add inferiors to the debug session
2577automatically by following calls to @code{fork} and @code{exec}. To
2578remove inferiors from the debugging session use the
2579@w{@code{remove-inferiors}} command.
2580
2581@table @code
2582@kindex add-inferior
2583@item add-inferior [ -copies @var{n} ] [ -exec @var{executable} ]
2584Adds @var{n} inferiors to be run using @var{executable} as the
2585executable. @var{n} defaults to 1. If no executable is specified,
2586the inferiors begins empty, with no program. You can still assign or
2587change the program assigned to the inferior at any time by using the
2588@code{file} command with the executable name as its argument.
2589
2590@kindex clone-inferior
2591@item clone-inferior [ -copies @var{n} ] [ @var{infno} ]
2592Adds @var{n} inferiors ready to execute the same program as inferior
2593@var{infno}. @var{n} defaults to 1. @var{infno} defaults to the
2594number of the current inferior. This is a convenient command when you
2595want to run another instance of the inferior you are debugging.
2596
2597@smallexample
2598(@value{GDBP}) info inferiors
2599 Num Description Executable
2600* 1 process 29964 helloworld
2601(@value{GDBP}) clone-inferior
2602Added inferior 2.
26031 inferiors added.
2604(@value{GDBP}) info inferiors
2605 Num Description Executable
2606 2 <null> helloworld
2607* 1 process 29964 helloworld
2608@end smallexample
2609
2610You can now simply switch focus to inferior 2 and run it.
2611
2612@kindex remove-inferiors
2613@item remove-inferiors @var{infno}@dots{}
2614Removes the inferior or inferiors @var{infno}@dots{}. It is not
2615possible to remove an inferior that is running with this command. For
2616those, use the @code{kill} or @code{detach} command first.
2617
2618@end table
2619
2620To quit debugging one of the running inferiors that is not the current
2621inferior, you can either detach from it by using the @w{@code{detach
2622inferior}} command (allowing it to run independently), or kill it
2623using the @w{@code{kill inferiors}} command:
2624
2625@table @code
2626@kindex detach inferiors @var{infno}@dots{}
2627@item detach inferior @var{infno}@dots{}
2628Detach from the inferior or inferiors identified by @value{GDBN}
2629inferior number(s) @var{infno}@dots{}. Note that the inferior's entry
2630still stays on the list of inferiors shown by @code{info inferiors},
2631but its Description will show @samp{<null>}.
2632
2633@kindex kill inferiors @var{infno}@dots{}
2634@item kill inferiors @var{infno}@dots{}
2635Kill the inferior or inferiors identified by @value{GDBN} inferior
2636number(s) @var{infno}@dots{}. Note that the inferior's entry still
2637stays on the list of inferiors shown by @code{info inferiors}, but its
2638Description will show @samp{<null>}.
2639@end table
2640
2641After the successful completion of a command such as @code{detach},
2642@code{detach inferiors}, @code{kill} or @code{kill inferiors}, or after
2643a normal process exit, the inferior is still valid and listed with
2644@code{info inferiors}, ready to be restarted.
2645
2646
2647To be notified when inferiors are started or exit under @value{GDBN}'s
2648control use @w{@code{set print inferior-events}}:
2649
2650@table @code
2651@kindex set print inferior-events
2652@cindex print messages on inferior start and exit
2653@item set print inferior-events
2654@itemx set print inferior-events on
2655@itemx set print inferior-events off
2656The @code{set print inferior-events} command allows you to enable or
2657disable printing of messages when @value{GDBN} notices that new
2658inferiors have started or that inferiors have exited or have been
2659detached. By default, these messages will not be printed.
2660
2661@kindex show print inferior-events
2662@item show print inferior-events
2663Show whether messages will be printed when @value{GDBN} detects that
2664inferiors have started, exited or have been detached.
2665@end table
2666
2667Many commands will work the same with multiple programs as with a
2668single program: e.g., @code{print myglobal} will simply display the
2669value of @code{myglobal} in the current inferior.
2670
2671
2672Occasionaly, when debugging @value{GDBN} itself, it may be useful to
2673get more info about the relationship of inferiors, programs, address
2674spaces in a debug session. You can do that with the @w{@code{maint
2675info program-spaces}} command.
2676
2677@table @code
2678@kindex maint info program-spaces
2679@item maint info program-spaces
2680Print a list of all program spaces currently being managed by
2681@value{GDBN}.
2682
2683@value{GDBN} displays for each program space (in this order):
2684
2685@enumerate
2686@item
2687the program space number assigned by @value{GDBN}
2688
2689@item
2690the name of the executable loaded into the program space, with e.g.,
2691the @code{file} command.
2692
2693@end enumerate
2694
2695@noindent
2696An asterisk @samp{*} preceding the @value{GDBN} program space number
2697indicates the current program space.
2698
2699In addition, below each program space line, @value{GDBN} prints extra
2700information that isn't suitable to display in tabular form. For
2701example, the list of inferiors bound to the program space.
2702
2703@smallexample
2704(@value{GDBP}) maint info program-spaces
2705 Id Executable
2706 2 goodbye
2707 Bound inferiors: ID 1 (process 21561)
2708* 1 hello
2709@end smallexample
2710
2711Here we can see that no inferior is running the program @code{hello},
2712while @code{process 21561} is running the program @code{goodbye}. On
2713some targets, it is possible that multiple inferiors are bound to the
2714same program space. The most common example is that of debugging both
2715the parent and child processes of a @code{vfork} call. For example,
2716
2717@smallexample
2718(@value{GDBP}) maint info program-spaces
2719 Id Executable
2720* 1 vfork-test
2721 Bound inferiors: ID 2 (process 18050), ID 1 (process 18045)
2722@end smallexample
2723
2724Here, both inferior 2 and inferior 1 are running in the same program
2725space as a result of inferior 1 having executed a @code{vfork} call.
2726@end table
2727
2728@node Threads
2729@section Debugging Programs with Multiple Threads
2730
2731@cindex threads of execution
2732@cindex multiple threads
2733@cindex switching threads
2734In some operating systems, such as HP-UX and Solaris, a single program
2735may have more than one @dfn{thread} of execution. The precise semantics
2736of threads differ from one operating system to another, but in general
2737the threads of a single program are akin to multiple processes---except
2738that they share one address space (that is, they can all examine and
2739modify the same variables). On the other hand, each thread has its own
2740registers and execution stack, and perhaps private memory.
2741
2742@value{GDBN} provides these facilities for debugging multi-thread
2743programs:
2744
2745@itemize @bullet
2746@item automatic notification of new threads
2747@item @samp{thread @var{threadno}}, a command to switch among threads
2748@item @samp{info threads}, a command to inquire about existing threads
2749@item @samp{thread apply [@var{threadno}] [@var{all}] @var{args}},
2750a command to apply a command to a list of threads
2751@item thread-specific breakpoints
2752@item @samp{set print thread-events}, which controls printing of
2753messages on thread start and exit.
2754@item @samp{set libthread-db-search-path @var{path}}, which lets
2755the user specify which @code{libthread_db} to use if the default choice
2756isn't compatible with the program.
2757@end itemize
2758
2759@quotation
2760@emph{Warning:} These facilities are not yet available on every
2761@value{GDBN} configuration where the operating system supports threads.
2762If your @value{GDBN} does not support threads, these commands have no
2763effect. For example, a system without thread support shows no output
2764from @samp{info threads}, and always rejects the @code{thread} command,
2765like this:
2766
2767@smallexample
2768(@value{GDBP}) info threads
2769(@value{GDBP}) thread 1
2770Thread ID 1 not known. Use the "info threads" command to
2771see the IDs of currently known threads.
2772@end smallexample
2773@c FIXME to implementors: how hard would it be to say "sorry, this GDB
2774@c doesn't support threads"?
2775@end quotation
2776
2777@cindex focus of debugging
2778@cindex current thread
2779The @value{GDBN} thread debugging facility allows you to observe all
2780threads while your program runs---but whenever @value{GDBN} takes
2781control, one thread in particular is always the focus of debugging.
2782This thread is called the @dfn{current thread}. Debugging commands show
2783program information from the perspective of the current thread.
2784
2785@cindex @code{New} @var{systag} message
2786@cindex thread identifier (system)
2787@c FIXME-implementors!! It would be more helpful if the [New...] message
2788@c included GDB's numeric thread handle, so you could just go to that
2789@c thread without first checking `info threads'.
2790Whenever @value{GDBN} detects a new thread in your program, it displays
2791the target system's identification for the thread with a message in the
2792form @samp{[New @var{systag}]}. @var{systag} is a thread identifier
2793whose form varies depending on the particular system. For example, on
2794@sc{gnu}/Linux, you might see
2795
2796@smallexample
2797[New Thread 0x41e02940 (LWP 25582)]
2798@end smallexample
2799
2800@noindent
2801when @value{GDBN} notices a new thread. In contrast, on an SGI system,
2802the @var{systag} is simply something like @samp{process 368}, with no
2803further qualifier.
2804
2805@c FIXME!! (1) Does the [New...] message appear even for the very first
2806@c thread of a program, or does it only appear for the
2807@c second---i.e.@: when it becomes obvious we have a multithread
2808@c program?
2809@c (2) *Is* there necessarily a first thread always? Or do some
2810@c multithread systems permit starting a program with multiple
2811@c threads ab initio?
2812
2813@cindex thread number
2814@cindex thread identifier (GDB)
2815For debugging purposes, @value{GDBN} associates its own thread
2816number---always a single integer---with each thread in your program.
2817
2818@table @code
2819@kindex info threads
2820@item info threads @r{[}@var{id}@dots{}@r{]}
2821Display a summary of all threads currently in your program. Optional
2822argument @var{id}@dots{} is one or more thread ids separated by spaces, and
2823means to print information only about the specified thread or threads.
2824@value{GDBN} displays for each thread (in this order):
2825
2826@enumerate
2827@item
2828the thread number assigned by @value{GDBN}
2829
2830@item
2831the target system's thread identifier (@var{systag})
2832
2833@item
2834the thread's name, if one is known. A thread can either be named by
2835the user (see @code{thread name}, below), or, in some cases, by the
2836program itself.
2837
2838@item
2839the current stack frame summary for that thread
2840@end enumerate
2841
2842@noindent
2843An asterisk @samp{*} to the left of the @value{GDBN} thread number
2844indicates the current thread.
2845
2846For example,
2847@end table
2848@c end table here to get a little more width for example
2849
2850@smallexample
2851(@value{GDBP}) info threads
2852 Id Target Id Frame
2853 3 process 35 thread 27 0x34e5 in sigpause ()
2854 2 process 35 thread 23 0x34e5 in sigpause ()
2855* 1 process 35 thread 13 main (argc=1, argv=0x7ffffff8)
2856 at threadtest.c:68
2857@end smallexample
2858
2859On Solaris, you can display more information about user threads with a
2860Solaris-specific command:
2861
2862@table @code
2863@item maint info sol-threads
2864@kindex maint info sol-threads
2865@cindex thread info (Solaris)
2866Display info on Solaris user threads.
2867@end table
2868
2869@table @code
2870@kindex thread @var{threadno}
2871@item thread @var{threadno}
2872Make thread number @var{threadno} the current thread. The command
2873argument @var{threadno} is the internal @value{GDBN} thread number, as
2874shown in the first field of the @samp{info threads} display.
2875@value{GDBN} responds by displaying the system identifier of the thread
2876you selected, and its current stack frame summary:
2877
2878@smallexample
2879(@value{GDBP}) thread 2
2880[Switching to thread 2 (Thread 0xb7fdab70 (LWP 12747))]
2881#0 some_function (ignore=0x0) at example.c:8
28828 printf ("hello\n");
2883@end smallexample
2884
2885@noindent
2886As with the @samp{[New @dots{}]} message, the form of the text after
2887@samp{Switching to} depends on your system's conventions for identifying
2888threads.
2889
2890@vindex $_thread@r{, convenience variable}
2891The debugger convenience variable @samp{$_thread} contains the number
2892of the current thread. You may find this useful in writing breakpoint
2893conditional expressions, command scripts, and so forth. See
2894@xref{Convenience Vars,, Convenience Variables}, for general
2895information on convenience variables.
2896
2897@kindex thread apply
2898@cindex apply command to several threads
2899@item thread apply [@var{threadno} | all] @var{command}
2900The @code{thread apply} command allows you to apply the named
2901@var{command} to one or more threads. Specify the numbers of the
2902threads that you want affected with the command argument
2903@var{threadno}. It can be a single thread number, one of the numbers
2904shown in the first field of the @samp{info threads} display; or it
2905could be a range of thread numbers, as in @code{2-4}. To apply a
2906command to all threads, type @kbd{thread apply all @var{command}}.
2907
2908@kindex thread name
2909@cindex name a thread
2910@item thread name [@var{name}]
2911This command assigns a name to the current thread. If no argument is
2912given, any existing user-specified name is removed. The thread name
2913appears in the @samp{info threads} display.
2914
2915On some systems, such as @sc{gnu}/Linux, @value{GDBN} is able to
2916determine the name of the thread as given by the OS. On these
2917systems, a name specified with @samp{thread name} will override the
2918system-give name, and removing the user-specified name will cause
2919@value{GDBN} to once again display the system-specified name.
2920
2921@kindex thread find
2922@cindex search for a thread
2923@item thread find [@var{regexp}]
2924Search for and display thread ids whose name or @var{systag}
2925matches the supplied regular expression.
2926
2927As well as being the complement to the @samp{thread name} command,
2928this command also allows you to identify a thread by its target
2929@var{systag}. For instance, on @sc{gnu}/Linux, the target @var{systag}
2930is the LWP id.
2931
2932@smallexample
2933(@value{GDBN}) thread find 26688
2934Thread 4 has target id 'Thread 0x41e02940 (LWP 26688)'
2935(@value{GDBN}) info thread 4
2936 Id Target Id Frame
2937 4 Thread 0x41e02940 (LWP 26688) 0x00000031ca6cd372 in select ()
2938@end smallexample
2939
2940@kindex set print thread-events
2941@cindex print messages on thread start and exit
2942@item set print thread-events
2943@itemx set print thread-events on
2944@itemx set print thread-events off
2945The @code{set print thread-events} command allows you to enable or
2946disable printing of messages when @value{GDBN} notices that new threads have
2947started or that threads have exited. By default, these messages will
2948be printed if detection of these events is supported by the target.
2949Note that these messages cannot be disabled on all targets.
2950
2951@kindex show print thread-events
2952@item show print thread-events
2953Show whether messages will be printed when @value{GDBN} detects that threads
2954have started and exited.
2955@end table
2956
2957@xref{Thread Stops,,Stopping and Starting Multi-thread Programs}, for
2958more information about how @value{GDBN} behaves when you stop and start
2959programs with multiple threads.
2960
2961@xref{Set Watchpoints,,Setting Watchpoints}, for information about
2962watchpoints in programs with multiple threads.
2963
2964@anchor{set libthread-db-search-path}
2965@table @code
2966@kindex set libthread-db-search-path
2967@cindex search path for @code{libthread_db}
2968@item set libthread-db-search-path @r{[}@var{path}@r{]}
2969If this variable is set, @var{path} is a colon-separated list of
2970directories @value{GDBN} will use to search for @code{libthread_db}.
2971If you omit @var{path}, @samp{libthread-db-search-path} will be reset to
2972its default value (@code{$sdir:$pdir} on @sc{gnu}/Linux and Solaris systems).
2973Internally, the default value comes from the @code{LIBTHREAD_DB_SEARCH_PATH}
2974macro.
2975
2976On @sc{gnu}/Linux and Solaris systems, @value{GDBN} uses a ``helper''
2977@code{libthread_db} library to obtain information about threads in the
2978inferior process. @value{GDBN} will use @samp{libthread-db-search-path}
2979to find @code{libthread_db}. @value{GDBN} also consults first if inferior
2980specific thread debugging library loading is enabled
2981by @samp{set auto-load libthread-db} (@pxref{libthread_db.so.1 file}).
2982
2983A special entry @samp{$sdir} for @samp{libthread-db-search-path}
2984refers to the default system directories that are
2985normally searched for loading shared libraries. The @samp{$sdir} entry
2986is the only kind not needing to be enabled by @samp{set auto-load libthread-db}
2987(@pxref{libthread_db.so.1 file}).
2988
2989A special entry @samp{$pdir} for @samp{libthread-db-search-path}
2990refers to the directory from which @code{libpthread}
2991was loaded in the inferior process.
2992
2993For any @code{libthread_db} library @value{GDBN} finds in above directories,
2994@value{GDBN} attempts to initialize it with the current inferior process.
2995If this initialization fails (which could happen because of a version
2996mismatch between @code{libthread_db} and @code{libpthread}), @value{GDBN}
2997will unload @code{libthread_db}, and continue with the next directory.
2998If none of @code{libthread_db} libraries initialize successfully,
2999@value{GDBN} will issue a warning and thread debugging will be disabled.
3000
3001Setting @code{libthread-db-search-path} is currently implemented
3002only on some platforms.
3003
3004@kindex show libthread-db-search-path
3005@item show libthread-db-search-path
3006Display current libthread_db search path.
3007
3008@kindex set debug libthread-db
3009@kindex show debug libthread-db
3010@cindex debugging @code{libthread_db}
3011@item set debug libthread-db
3012@itemx show debug libthread-db
3013Turns on or off display of @code{libthread_db}-related events.
3014Use @code{1} to enable, @code{0} to disable.
3015@end table
3016
3017@node Forks
3018@section Debugging Forks
3019
3020@cindex fork, debugging programs which call
3021@cindex multiple processes
3022@cindex processes, multiple
3023On most systems, @value{GDBN} has no special support for debugging
3024programs which create additional processes using the @code{fork}
3025function. When a program forks, @value{GDBN} will continue to debug the
3026parent process and the child process will run unimpeded. If you have
3027set a breakpoint in any code which the child then executes, the child
3028will get a @code{SIGTRAP} signal which (unless it catches the signal)
3029will cause it to terminate.
3030
3031However, if you want to debug the child process there is a workaround
3032which isn't too painful. Put a call to @code{sleep} in the code which
3033the child process executes after the fork. It may be useful to sleep
3034only if a certain environment variable is set, or a certain file exists,
3035so that the delay need not occur when you don't want to run @value{GDBN}
3036on the child. While the child is sleeping, use the @code{ps} program to
3037get its process ID. Then tell @value{GDBN} (a new invocation of
3038@value{GDBN} if you are also debugging the parent process) to attach to
3039the child process (@pxref{Attach}). From that point on you can debug
3040the child process just like any other process which you attached to.
3041
3042On some systems, @value{GDBN} provides support for debugging programs that
3043create additional processes using the @code{fork} or @code{vfork} functions.
3044Currently, the only platforms with this feature are HP-UX (11.x and later
3045only?) and @sc{gnu}/Linux (kernel version 2.5.60 and later).
3046
3047By default, when a program forks, @value{GDBN} will continue to debug
3048the parent process and the child process will run unimpeded.
3049
3050If you want to follow the child process instead of the parent process,
3051use the command @w{@code{set follow-fork-mode}}.
3052
3053@table @code
3054@kindex set follow-fork-mode
3055@item set follow-fork-mode @var{mode}
3056Set the debugger response to a program call of @code{fork} or
3057@code{vfork}. A call to @code{fork} or @code{vfork} creates a new
3058process. The @var{mode} argument can be:
3059
3060@table @code
3061@item parent
3062The original process is debugged after a fork. The child process runs
3063unimpeded. This is the default.
3064
3065@item child
3066The new process is debugged after a fork. The parent process runs
3067unimpeded.
3068
3069@end table
3070
3071@kindex show follow-fork-mode
3072@item show follow-fork-mode
3073Display the current debugger response to a @code{fork} or @code{vfork} call.
3074@end table
3075
3076@cindex debugging multiple processes
3077On Linux, if you want to debug both the parent and child processes, use the
3078command @w{@code{set detach-on-fork}}.
3079
3080@table @code
3081@kindex set detach-on-fork
3082@item set detach-on-fork @var{mode}
3083Tells gdb whether to detach one of the processes after a fork, or
3084retain debugger control over them both.
3085
3086@table @code
3087@item on
3088The child process (or parent process, depending on the value of
3089@code{follow-fork-mode}) will be detached and allowed to run
3090independently. This is the default.
3091
3092@item off
3093Both processes will be held under the control of @value{GDBN}.
3094One process (child or parent, depending on the value of
3095@code{follow-fork-mode}) is debugged as usual, while the other
3096is held suspended.
3097
3098@end table
3099
3100@kindex show detach-on-fork
3101@item show detach-on-fork
3102Show whether detach-on-fork mode is on/off.
3103@end table
3104
3105If you choose to set @samp{detach-on-fork} mode off, then @value{GDBN}
3106will retain control of all forked processes (including nested forks).
3107You can list the forked processes under the control of @value{GDBN} by
3108using the @w{@code{info inferiors}} command, and switch from one fork
3109to another by using the @code{inferior} command (@pxref{Inferiors and
3110Programs, ,Debugging Multiple Inferiors and Programs}).
3111
3112To quit debugging one of the forked processes, you can either detach
3113from it by using the @w{@code{detach inferiors}} command (allowing it
3114to run independently), or kill it using the @w{@code{kill inferiors}}
3115command. @xref{Inferiors and Programs, ,Debugging Multiple Inferiors
3116and Programs}.
3117
3118If you ask to debug a child process and a @code{vfork} is followed by an
3119@code{exec}, @value{GDBN} executes the new target up to the first
3120breakpoint in the new target. If you have a breakpoint set on
3121@code{main} in your original program, the breakpoint will also be set on
3122the child process's @code{main}.
3123
3124On some systems, when a child process is spawned by @code{vfork}, you
3125cannot debug the child or parent until an @code{exec} call completes.
3126
3127If you issue a @code{run} command to @value{GDBN} after an @code{exec}
3128call executes, the new target restarts. To restart the parent
3129process, use the @code{file} command with the parent executable name
3130as its argument. By default, after an @code{exec} call executes,
3131@value{GDBN} discards the symbols of the previous executable image.
3132You can change this behaviour with the @w{@code{set follow-exec-mode}}
3133command.
3134
3135@table @code
3136@kindex set follow-exec-mode
3137@item set follow-exec-mode @var{mode}
3138
3139Set debugger response to a program call of @code{exec}. An
3140@code{exec} call replaces the program image of a process.
3141
3142@code{follow-exec-mode} can be:
3143
3144@table @code
3145@item new
3146@value{GDBN} creates a new inferior and rebinds the process to this
3147new inferior. The program the process was running before the
3148@code{exec} call can be restarted afterwards by restarting the
3149original inferior.
3150
3151For example:
3152
3153@smallexample
3154(@value{GDBP}) info inferiors
3155(gdb) info inferior
3156 Id Description Executable
3157* 1 <null> prog1
3158(@value{GDBP}) run
3159process 12020 is executing new program: prog2
3160Program exited normally.
3161(@value{GDBP}) info inferiors
3162 Id Description Executable
3163* 2 <null> prog2
3164 1 <null> prog1
3165@end smallexample
3166
3167@item same
3168@value{GDBN} keeps the process bound to the same inferior. The new
3169executable image replaces the previous executable loaded in the
3170inferior. Restarting the inferior after the @code{exec} call, with
3171e.g., the @code{run} command, restarts the executable the process was
3172running after the @code{exec} call. This is the default mode.
3173
3174For example:
3175
3176@smallexample
3177(@value{GDBP}) info inferiors
3178 Id Description Executable
3179* 1 <null> prog1
3180(@value{GDBP}) run
3181process 12020 is executing new program: prog2
3182Program exited normally.
3183(@value{GDBP}) info inferiors
3184 Id Description Executable
3185* 1 <null> prog2
3186@end smallexample
3187
3188@end table
3189@end table
3190
3191You can use the @code{catch} command to make @value{GDBN} stop whenever
3192a @code{fork}, @code{vfork}, or @code{exec} call is made. @xref{Set
3193Catchpoints, ,Setting Catchpoints}.
3194
3195@node Checkpoint/Restart
3196@section Setting a @emph{Bookmark} to Return to Later
3197
3198@cindex checkpoint
3199@cindex restart
3200@cindex bookmark
3201@cindex snapshot of a process
3202@cindex rewind program state
3203
3204On certain operating systems@footnote{Currently, only
3205@sc{gnu}/Linux.}, @value{GDBN} is able to save a @dfn{snapshot} of a
3206program's state, called a @dfn{checkpoint}, and come back to it
3207later.
3208
3209Returning to a checkpoint effectively undoes everything that has
3210happened in the program since the @code{checkpoint} was saved. This
3211includes changes in memory, registers, and even (within some limits)
3212system state. Effectively, it is like going back in time to the
3213moment when the checkpoint was saved.
3214
3215Thus, if you're stepping thru a program and you think you're
3216getting close to the point where things go wrong, you can save
3217a checkpoint. Then, if you accidentally go too far and miss
3218the critical statement, instead of having to restart your program
3219from the beginning, you can just go back to the checkpoint and
3220start again from there.
3221
3222This can be especially useful if it takes a lot of time or
3223steps to reach the point where you think the bug occurs.
3224
3225To use the @code{checkpoint}/@code{restart} method of debugging:
3226
3227@table @code
3228@kindex checkpoint
3229@item checkpoint
3230Save a snapshot of the debugged program's current execution state.
3231The @code{checkpoint} command takes no arguments, but each checkpoint
3232is assigned a small integer id, similar to a breakpoint id.
3233
3234@kindex info checkpoints
3235@item info checkpoints
3236List the checkpoints that have been saved in the current debugging
3237session. For each checkpoint, the following information will be
3238listed:
3239
3240@table @code
3241@item Checkpoint ID
3242@item Process ID
3243@item Code Address
3244@item Source line, or label
3245@end table
3246
3247@kindex restart @var{checkpoint-id}
3248@item restart @var{checkpoint-id}
3249Restore the program state that was saved as checkpoint number
3250@var{checkpoint-id}. All program variables, registers, stack frames
3251etc.@: will be returned to the values that they had when the checkpoint
3252was saved. In essence, gdb will ``wind back the clock'' to the point
3253in time when the checkpoint was saved.
3254
3255Note that breakpoints, @value{GDBN} variables, command history etc.
3256are not affected by restoring a checkpoint. In general, a checkpoint
3257only restores things that reside in the program being debugged, not in
3258the debugger.
3259
3260@kindex delete checkpoint @var{checkpoint-id}
3261@item delete checkpoint @var{checkpoint-id}
3262Delete the previously-saved checkpoint identified by @var{checkpoint-id}.
3263
3264@end table
3265
3266Returning to a previously saved checkpoint will restore the user state
3267of the program being debugged, plus a significant subset of the system
3268(OS) state, including file pointers. It won't ``un-write'' data from
3269a file, but it will rewind the file pointer to the previous location,
3270so that the previously written data can be overwritten. For files
3271opened in read mode, the pointer will also be restored so that the
3272previously read data can be read again.
3273
3274Of course, characters that have been sent to a printer (or other
3275external device) cannot be ``snatched back'', and characters received
3276from eg.@: a serial device can be removed from internal program buffers,
3277but they cannot be ``pushed back'' into the serial pipeline, ready to
3278be received again. Similarly, the actual contents of files that have
3279been changed cannot be restored (at this time).
3280
3281However, within those constraints, you actually can ``rewind'' your
3282program to a previously saved point in time, and begin debugging it
3283again --- and you can change the course of events so as to debug a
3284different execution path this time.
3285
3286@cindex checkpoints and process id
3287Finally, there is one bit of internal program state that will be
3288different when you return to a checkpoint --- the program's process
3289id. Each checkpoint will have a unique process id (or @var{pid}),
3290and each will be different from the program's original @var{pid}.
3291If your program has saved a local copy of its process id, this could
3292potentially pose a problem.
3293
3294@subsection A Non-obvious Benefit of Using Checkpoints
3295
3296On some systems such as @sc{gnu}/Linux, address space randomization
3297is performed on new processes for security reasons. This makes it
3298difficult or impossible to set a breakpoint, or watchpoint, on an
3299absolute address if you have to restart the program, since the
3300absolute location of a symbol will change from one execution to the
3301next.
3302
3303A checkpoint, however, is an @emph{identical} copy of a process.
3304Therefore if you create a checkpoint at (eg.@:) the start of main,
3305and simply return to that checkpoint instead of restarting the
3306process, you can avoid the effects of address randomization and
3307your symbols will all stay in the same place.
3308
3309@node Stopping
3310@chapter Stopping and Continuing
3311
3312The principal purposes of using a debugger are so that you can stop your
3313program before it terminates; or so that, if your program runs into
3314trouble, you can investigate and find out why.
3315
3316Inside @value{GDBN}, your program may stop for any of several reasons,
3317such as a signal, a breakpoint, or reaching a new line after a
3318@value{GDBN} command such as @code{step}. You may then examine and
3319change variables, set new breakpoints or remove old ones, and then
3320continue execution. Usually, the messages shown by @value{GDBN} provide
3321ample explanation of the status of your program---but you can also
3322explicitly request this information at any time.
3323
3324@table @code
3325@kindex info program
3326@item info program
3327Display information about the status of your program: whether it is
3328running or not, what process it is, and why it stopped.
3329@end table
3330
3331@menu
3332* Breakpoints:: Breakpoints, watchpoints, and catchpoints
3333* Continuing and Stepping:: Resuming execution
3334* Skipping Over Functions and Files::
3335 Skipping over functions and files
3336* Signals:: Signals
3337* Thread Stops:: Stopping and starting multi-thread programs
3338@end menu
3339
3340@node Breakpoints
3341@section Breakpoints, Watchpoints, and Catchpoints
3342
3343@cindex breakpoints
3344A @dfn{breakpoint} makes your program stop whenever a certain point in
3345the program is reached. For each breakpoint, you can add conditions to
3346control in finer detail whether your program stops. You can set
3347breakpoints with the @code{break} command and its variants (@pxref{Set
3348Breaks, ,Setting Breakpoints}), to specify the place where your program
3349should stop by line number, function name or exact address in the
3350program.
3351
3352On some systems, you can set breakpoints in shared libraries before
3353the executable is run. There is a minor limitation on HP-UX systems:
3354you must wait until the executable is run in order to set breakpoints
3355in shared library routines that are not called directly by the program
3356(for example, routines that are arguments in a @code{pthread_create}
3357call).
3358
3359@cindex watchpoints
3360@cindex data breakpoints
3361@cindex memory tracing
3362@cindex breakpoint on memory address
3363@cindex breakpoint on variable modification
3364A @dfn{watchpoint} is a special breakpoint that stops your program
3365when the value of an expression changes. The expression may be a value
3366of a variable, or it could involve values of one or more variables
3367combined by operators, such as @samp{a + b}. This is sometimes called
3368@dfn{data breakpoints}. You must use a different command to set
3369watchpoints (@pxref{Set Watchpoints, ,Setting Watchpoints}), but aside
3370from that, you can manage a watchpoint like any other breakpoint: you
3371enable, disable, and delete both breakpoints and watchpoints using the
3372same commands.
3373
3374You can arrange to have values from your program displayed automatically
3375whenever @value{GDBN} stops at a breakpoint. @xref{Auto Display,,
3376Automatic Display}.
3377
3378@cindex catchpoints
3379@cindex breakpoint on events
3380A @dfn{catchpoint} is another special breakpoint that stops your program
3381when a certain kind of event occurs, such as the throwing of a C@t{++}
3382exception or the loading of a library. As with watchpoints, you use a
3383different command to set a catchpoint (@pxref{Set Catchpoints, ,Setting
3384Catchpoints}), but aside from that, you can manage a catchpoint like any
3385other breakpoint. (To stop when your program receives a signal, use the
3386@code{handle} command; see @ref{Signals, ,Signals}.)
3387
3388@cindex breakpoint numbers
3389@cindex numbers for breakpoints
3390@value{GDBN} assigns a number to each breakpoint, watchpoint, or
3391catchpoint when you create it; these numbers are successive integers
3392starting with one. In many of the commands for controlling various
3393features of breakpoints you use the breakpoint number to say which
3394breakpoint you want to change. Each breakpoint may be @dfn{enabled} or
3395@dfn{disabled}; if disabled, it has no effect on your program until you
3396enable it again.
3397
3398@cindex breakpoint ranges
3399@cindex ranges of breakpoints
3400Some @value{GDBN} commands accept a range of breakpoints on which to
3401operate. A breakpoint range is either a single breakpoint number, like
3402@samp{5}, or two such numbers, in increasing order, separated by a
3403hyphen, like @samp{5-7}. When a breakpoint range is given to a command,
3404all breakpoints in that range are operated on.
3405
3406@menu
3407* Set Breaks:: Setting breakpoints
3408* Set Watchpoints:: Setting watchpoints
3409* Set Catchpoints:: Setting catchpoints
3410* Delete Breaks:: Deleting breakpoints
3411* Disabling:: Disabling breakpoints
3412* Conditions:: Break conditions
3413* Break Commands:: Breakpoint command lists
3414* Dynamic Printf:: Dynamic printf
3415* Save Breakpoints:: How to save breakpoints in a file
3416* Static Probe Points:: Listing static probe points
3417* Error in Breakpoints:: ``Cannot insert breakpoints''
3418* Breakpoint-related Warnings:: ``Breakpoint address adjusted...''
3419@end menu
3420
3421@node Set Breaks
3422@subsection Setting Breakpoints
3423
3424@c FIXME LMB what does GDB do if no code on line of breakpt?
3425@c consider in particular declaration with/without initialization.
3426@c
3427@c FIXME 2 is there stuff on this already? break at fun start, already init?
3428
3429@kindex break
3430@kindex b @r{(@code{break})}
3431@vindex $bpnum@r{, convenience variable}
3432@cindex latest breakpoint
3433Breakpoints are set with the @code{break} command (abbreviated
3434@code{b}). The debugger convenience variable @samp{$bpnum} records the
3435number of the breakpoint you've set most recently; see @ref{Convenience
3436Vars,, Convenience Variables}, for a discussion of what you can do with
3437convenience variables.
3438
3439@table @code
3440@item break @var{location}
3441Set a breakpoint at the given @var{location}, which can specify a
3442function name, a line number, or an address of an instruction.
3443(@xref{Specify Location}, for a list of all the possible ways to
3444specify a @var{location}.) The breakpoint will stop your program just
3445before it executes any of the code in the specified @var{location}.
3446
3447When using source languages that permit overloading of symbols, such as
3448C@t{++}, a function name may refer to more than one possible place to break.
3449@xref{Ambiguous Expressions,,Ambiguous Expressions}, for a discussion of
3450that situation.
3451
3452It is also possible to insert a breakpoint that will stop the program
3453only if a specific thread (@pxref{Thread-Specific Breakpoints})
3454or a specific task (@pxref{Ada Tasks}) hits that breakpoint.
3455
3456@item break
3457When called without any arguments, @code{break} sets a breakpoint at
3458the next instruction to be executed in the selected stack frame
3459(@pxref{Stack, ,Examining the Stack}). In any selected frame but the
3460innermost, this makes your program stop as soon as control
3461returns to that frame. This is similar to the effect of a
3462@code{finish} command in the frame inside the selected frame---except
3463that @code{finish} does not leave an active breakpoint. If you use
3464@code{break} without an argument in the innermost frame, @value{GDBN} stops
3465the next time it reaches the current location; this may be useful
3466inside loops.
3467
3468@value{GDBN} normally ignores breakpoints when it resumes execution, until at
3469least one instruction has been executed. If it did not do this, you
3470would be unable to proceed past a breakpoint without first disabling the
3471breakpoint. This rule applies whether or not the breakpoint already
3472existed when your program stopped.
3473
3474@item break @dots{} if @var{cond}
3475Set a breakpoint with condition @var{cond}; evaluate the expression
3476@var{cond} each time the breakpoint is reached, and stop only if the
3477value is nonzero---that is, if @var{cond} evaluates as true.
3478@samp{@dots{}} stands for one of the possible arguments described
3479above (or no argument) specifying where to break. @xref{Conditions,
3480,Break Conditions}, for more information on breakpoint conditions.
3481
3482@kindex tbreak
3483@item tbreak @var{args}
3484Set a breakpoint enabled only for one stop. @var{args} are the
3485same as for the @code{break} command, and the breakpoint is set in the same
3486way, but the breakpoint is automatically deleted after the first time your
3487program stops there. @xref{Disabling, ,Disabling Breakpoints}.
3488
3489@kindex hbreak
3490@cindex hardware breakpoints
3491@item hbreak @var{args}
3492Set a hardware-assisted breakpoint. @var{args} are the same as for the
3493@code{break} command and the breakpoint is set in the same way, but the
3494breakpoint requires hardware support and some target hardware may not
3495have this support. The main purpose of this is EPROM/ROM code
3496debugging, so you can set a breakpoint at an instruction without
3497changing the instruction. This can be used with the new trap-generation
3498provided by SPARClite DSU and most x86-based targets. These targets
3499will generate traps when a program accesses some data or instruction
3500address that is assigned to the debug registers. However the hardware
3501breakpoint registers can take a limited number of breakpoints. For
3502example, on the DSU, only two data breakpoints can be set at a time, and
3503@value{GDBN} will reject this command if more than two are used. Delete
3504or disable unused hardware breakpoints before setting new ones
3505(@pxref{Disabling, ,Disabling Breakpoints}).
3506@xref{Conditions, ,Break Conditions}.
3507For remote targets, you can restrict the number of hardware
3508breakpoints @value{GDBN} will use, see @ref{set remote
3509hardware-breakpoint-limit}.
3510
3511@kindex thbreak
3512@item thbreak @var{args}
3513Set a hardware-assisted breakpoint enabled only for one stop. @var{args}
3514are the same as for the @code{hbreak} command and the breakpoint is set in
3515the same way. However, like the @code{tbreak} command,
3516the breakpoint is automatically deleted after the
3517first time your program stops there. Also, like the @code{hbreak}
3518command, the breakpoint requires hardware support and some target hardware
3519may not have this support. @xref{Disabling, ,Disabling Breakpoints}.
3520See also @ref{Conditions, ,Break Conditions}.
3521
3522@kindex rbreak
3523@cindex regular expression
3524@cindex breakpoints at functions matching a regexp
3525@cindex set breakpoints in many functions
3526@item rbreak @var{regex}
3527Set breakpoints on all functions matching the regular expression
3528@var{regex}. This command sets an unconditional breakpoint on all
3529matches, printing a list of all breakpoints it set. Once these
3530breakpoints are set, they are treated just like the breakpoints set with
3531the @code{break} command. You can delete them, disable them, or make
3532them conditional the same way as any other breakpoint.
3533
3534The syntax of the regular expression is the standard one used with tools
3535like @file{grep}. Note that this is different from the syntax used by
3536shells, so for instance @code{foo*} matches all functions that include
3537an @code{fo} followed by zero or more @code{o}s. There is an implicit
3538@code{.*} leading and trailing the regular expression you supply, so to
3539match only functions that begin with @code{foo}, use @code{^foo}.
3540
3541@cindex non-member C@t{++} functions, set breakpoint in
3542When debugging C@t{++} programs, @code{rbreak} is useful for setting
3543breakpoints on overloaded functions that are not members of any special
3544classes.
3545
3546@cindex set breakpoints on all functions
3547The @code{rbreak} command can be used to set breakpoints in
3548@strong{all} the functions in a program, like this:
3549
3550@smallexample
3551(@value{GDBP}) rbreak .
3552@end smallexample
3553
3554@item rbreak @var{file}:@var{regex}
3555If @code{rbreak} is called with a filename qualification, it limits
3556the search for functions matching the given regular expression to the
3557specified @var{file}. This can be used, for example, to set breakpoints on
3558every function in a given file:
3559
3560@smallexample
3561(@value{GDBP}) rbreak file.c:.
3562@end smallexample
3563
3564The colon separating the filename qualifier from the regex may
3565optionally be surrounded by spaces.
3566
3567@kindex info breakpoints
3568@cindex @code{$_} and @code{info breakpoints}
3569@item info breakpoints @r{[}@var{n}@dots{}@r{]}
3570@itemx info break @r{[}@var{n}@dots{}@r{]}
3571Print a table of all breakpoints, watchpoints, and catchpoints set and
3572not deleted. Optional argument @var{n} means print information only
3573about the specified breakpoint(s) (or watchpoint(s) or catchpoint(s)).
3574For each breakpoint, following columns are printed:
3575
3576@table @emph
3577@item Breakpoint Numbers
3578@item Type
3579Breakpoint, watchpoint, or catchpoint.
3580@item Disposition
3581Whether the breakpoint is marked to be disabled or deleted when hit.
3582@item Enabled or Disabled
3583Enabled breakpoints are marked with @samp{y}. @samp{n} marks breakpoints
3584that are not enabled.
3585@item Address
3586Where the breakpoint is in your program, as a memory address. For a
3587pending breakpoint whose address is not yet known, this field will
3588contain @samp{<PENDING>}. Such breakpoint won't fire until a shared
3589library that has the symbol or line referred by breakpoint is loaded.
3590See below for details. A breakpoint with several locations will
3591have @samp{<MULTIPLE>} in this field---see below for details.
3592@item What
3593Where the breakpoint is in the source for your program, as a file and
3594line number. For a pending breakpoint, the original string passed to
3595the breakpoint command will be listed as it cannot be resolved until
3596the appropriate shared library is loaded in the future.
3597@end table
3598
3599@noindent
3600If a breakpoint is conditional, there are two evaluation modes: ``host'' and
3601``target''. If mode is ``host'', breakpoint condition evaluation is done by
3602@value{GDBN} on the host's side. If it is ``target'', then the condition
3603is evaluated by the target. The @code{info break} command shows
3604the condition on the line following the affected breakpoint, together with
3605its condition evaluation mode in between parentheses.
3606
3607Breakpoint commands, if any, are listed after that. A pending breakpoint is
3608allowed to have a condition specified for it. The condition is not parsed for
3609validity until a shared library is loaded that allows the pending
3610breakpoint to resolve to a valid location.
3611
3612@noindent
3613@code{info break} with a breakpoint
3614number @var{n} as argument lists only that breakpoint. The
3615convenience variable @code{$_} and the default examining-address for
3616the @code{x} command are set to the address of the last breakpoint
3617listed (@pxref{Memory, ,Examining Memory}).
3618
3619@noindent
3620@code{info break} displays a count of the number of times the breakpoint
3621has been hit. This is especially useful in conjunction with the
3622@code{ignore} command. You can ignore a large number of breakpoint
3623hits, look at the breakpoint info to see how many times the breakpoint
3624was hit, and then run again, ignoring one less than that number. This
3625will get you quickly to the last hit of that breakpoint.
3626
3627@noindent
3628For a breakpoints with an enable count (xref) greater than 1,
3629@code{info break} also displays that count.
3630
3631@end table
3632
3633@value{GDBN} allows you to set any number of breakpoints at the same place in
3634your program. There is nothing silly or meaningless about this. When
3635the breakpoints are conditional, this is even useful
3636(@pxref{Conditions, ,Break Conditions}).
3637
3638@cindex multiple locations, breakpoints
3639@cindex breakpoints, multiple locations
3640It is possible that a breakpoint corresponds to several locations
3641in your program. Examples of this situation are:
3642
3643@itemize @bullet
3644@item
3645Multiple functions in the program may have the same name.
3646
3647@item
3648For a C@t{++} constructor, the @value{NGCC} compiler generates several
3649instances of the function body, used in different cases.
3650
3651@item
3652For a C@t{++} template function, a given line in the function can
3653correspond to any number of instantiations.
3654
3655@item
3656For an inlined function, a given source line can correspond to
3657several places where that function is inlined.
3658@end itemize
3659
3660In all those cases, @value{GDBN} will insert a breakpoint at all
3661the relevant locations.
3662
3663A breakpoint with multiple locations is displayed in the breakpoint
3664table using several rows---one header row, followed by one row for
3665each breakpoint location. The header row has @samp{<MULTIPLE>} in the
3666address column. The rows for individual locations contain the actual
3667addresses for locations, and show the functions to which those
3668locations belong. The number column for a location is of the form
3669@var{breakpoint-number}.@var{location-number}.
3670
3671For example:
3672
3673@smallexample
3674Num Type Disp Enb Address What
36751 breakpoint keep y <MULTIPLE>
3676 stop only if i==1
3677 breakpoint already hit 1 time
36781.1 y 0x080486a2 in void foo<int>() at t.cc:8
36791.2 y 0x080486ca in void foo<double>() at t.cc:8
3680@end smallexample
3681
3682Each location can be individually enabled or disabled by passing
3683@var{breakpoint-number}.@var{location-number} as argument to the
3684@code{enable} and @code{disable} commands. Note that you cannot
3685delete the individual locations from the list, you can only delete the
3686entire list of locations that belong to their parent breakpoint (with
3687the @kbd{delete @var{num}} command, where @var{num} is the number of
3688the parent breakpoint, 1 in the above example). Disabling or enabling
3689the parent breakpoint (@pxref{Disabling}) affects all of the locations
3690that belong to that breakpoint.
3691
3692@cindex pending breakpoints
3693It's quite common to have a breakpoint inside a shared library.
3694Shared libraries can be loaded and unloaded explicitly,
3695and possibly repeatedly, as the program is executed. To support
3696this use case, @value{GDBN} updates breakpoint locations whenever
3697any shared library is loaded or unloaded. Typically, you would
3698set a breakpoint in a shared library at the beginning of your
3699debugging session, when the library is not loaded, and when the
3700symbols from the library are not available. When you try to set
3701breakpoint, @value{GDBN} will ask you if you want to set
3702a so called @dfn{pending breakpoint}---breakpoint whose address
3703is not yet resolved.
3704
3705After the program is run, whenever a new shared library is loaded,
3706@value{GDBN} reevaluates all the breakpoints. When a newly loaded
3707shared library contains the symbol or line referred to by some
3708pending breakpoint, that breakpoint is resolved and becomes an
3709ordinary breakpoint. When a library is unloaded, all breakpoints
3710that refer to its symbols or source lines become pending again.
3711
3712This logic works for breakpoints with multiple locations, too. For
3713example, if you have a breakpoint in a C@t{++} template function, and
3714a newly loaded shared library has an instantiation of that template,
3715a new location is added to the list of locations for the breakpoint.
3716
3717Except for having unresolved address, pending breakpoints do not
3718differ from regular breakpoints. You can set conditions or commands,
3719enable and disable them and perform other breakpoint operations.
3720
3721@value{GDBN} provides some additional commands for controlling what
3722happens when the @samp{break} command cannot resolve breakpoint
3723address specification to an address:
3724
3725@kindex set breakpoint pending
3726@kindex show breakpoint pending
3727@table @code
3728@item set breakpoint pending auto
3729This is the default behavior. When @value{GDBN} cannot find the breakpoint
3730location, it queries you whether a pending breakpoint should be created.
3731
3732@item set breakpoint pending on
3733This indicates that an unrecognized breakpoint location should automatically
3734result in a pending breakpoint being created.
3735
3736@item set breakpoint pending off
3737This indicates that pending breakpoints are not to be created. Any
3738unrecognized breakpoint location results in an error. This setting does
3739not affect any pending breakpoints previously created.
3740
3741@item show breakpoint pending
3742Show the current behavior setting for creating pending breakpoints.
3743@end table
3744
3745The settings above only affect the @code{break} command and its
3746variants. Once breakpoint is set, it will be automatically updated
3747as shared libraries are loaded and unloaded.
3748
3749@cindex automatic hardware breakpoints
3750For some targets, @value{GDBN} can automatically decide if hardware or
3751software breakpoints should be used, depending on whether the
3752breakpoint address is read-only or read-write. This applies to
3753breakpoints set with the @code{break} command as well as to internal
3754breakpoints set by commands like @code{next} and @code{finish}. For
3755breakpoints set with @code{hbreak}, @value{GDBN} will always use hardware
3756breakpoints.
3757
3758You can control this automatic behaviour with the following commands::
3759
3760@kindex set breakpoint auto-hw
3761@kindex show breakpoint auto-hw
3762@table @code
3763@item set breakpoint auto-hw on
3764This is the default behavior. When @value{GDBN} sets a breakpoint, it
3765will try to use the target memory map to decide if software or hardware
3766breakpoint must be used.
3767
3768@item set breakpoint auto-hw off
3769This indicates @value{GDBN} should not automatically select breakpoint
3770type. If the target provides a memory map, @value{GDBN} will warn when
3771trying to set software breakpoint at a read-only address.
3772@end table
3773
3774@value{GDBN} normally implements breakpoints by replacing the program code
3775at the breakpoint address with a special instruction, which, when
3776executed, given control to the debugger. By default, the program
3777code is so modified only when the program is resumed. As soon as
3778the program stops, @value{GDBN} restores the original instructions. This
3779behaviour guards against leaving breakpoints inserted in the
3780target should gdb abrubptly disconnect. However, with slow remote
3781targets, inserting and removing breakpoint can reduce the performance.
3782This behavior can be controlled with the following commands::
3783
3784@kindex set breakpoint always-inserted
3785@kindex show breakpoint always-inserted
3786@table @code
3787@item set breakpoint always-inserted off
3788All breakpoints, including newly added by the user, are inserted in
3789the target only when the target is resumed. All breakpoints are
3790removed from the target when it stops.
3791
3792@item set breakpoint always-inserted on
3793Causes all breakpoints to be inserted in the target at all times. If
3794the user adds a new breakpoint, or changes an existing breakpoint, the
3795breakpoints in the target are updated immediately. A breakpoint is
3796removed from the target only when breakpoint itself is removed.
3797
3798@cindex non-stop mode, and @code{breakpoint always-inserted}
3799@item set breakpoint always-inserted auto
3800This is the default mode. If @value{GDBN} is controlling the inferior
3801in non-stop mode (@pxref{Non-Stop Mode}), gdb behaves as if
3802@code{breakpoint always-inserted} mode is on. If @value{GDBN} is
3803controlling the inferior in all-stop mode, @value{GDBN} behaves as if
3804@code{breakpoint always-inserted} mode is off.
3805@end table
3806
3807@value{GDBN} handles conditional breakpoints by evaluating these conditions
3808when a breakpoint breaks. If the condition is true, then the process being
3809debugged stops, otherwise the process is resumed.
3810
3811If the target supports evaluating conditions on its end, @value{GDBN} may
3812download the breakpoint, together with its conditions, to it.
3813
3814This feature can be controlled via the following commands:
3815
3816@kindex set breakpoint condition-evaluation
3817@kindex show breakpoint condition-evaluation
3818@table @code
3819@item set breakpoint condition-evaluation host
3820This option commands @value{GDBN} to evaluate the breakpoint
3821conditions on the host's side. Unconditional breakpoints are sent to
3822the target which in turn receives the triggers and reports them back to GDB
3823for condition evaluation. This is the standard evaluation mode.
3824
3825@item set breakpoint condition-evaluation target
3826This option commands @value{GDBN} to download breakpoint conditions
3827to the target at the moment of their insertion. The target
3828is responsible for evaluating the conditional expression and reporting
3829breakpoint stop events back to @value{GDBN} whenever the condition
3830is true. Due to limitations of target-side evaluation, some conditions
3831cannot be evaluated there, e.g., conditions that depend on local data
3832that is only known to the host. Examples include
3833conditional expressions involving convenience variables, complex types
3834that cannot be handled by the agent expression parser and expressions
3835that are too long to be sent over to the target, specially when the
3836target is a remote system. In these cases, the conditions will be
3837evaluated by @value{GDBN}.
3838
3839@item set breakpoint condition-evaluation auto
3840This is the default mode. If the target supports evaluating breakpoint
3841conditions on its end, @value{GDBN} will download breakpoint conditions to
3842the target (limitations mentioned previously apply). If the target does
3843not support breakpoint condition evaluation, then @value{GDBN} will fallback
3844to evaluating all these conditions on the host's side.
3845@end table
3846
3847
3848@cindex negative breakpoint numbers
3849@cindex internal @value{GDBN} breakpoints
3850@value{GDBN} itself sometimes sets breakpoints in your program for
3851special purposes, such as proper handling of @code{longjmp} (in C
3852programs). These internal breakpoints are assigned negative numbers,
3853starting with @code{-1}; @samp{info breakpoints} does not display them.
3854You can see these breakpoints with the @value{GDBN} maintenance command
3855@samp{maint info breakpoints} (@pxref{maint info breakpoints}).
3856
3857
3858@node Set Watchpoints
3859@subsection Setting Watchpoints
3860
3861@cindex setting watchpoints
3862You can use a watchpoint to stop execution whenever the value of an
3863expression changes, without having to predict a particular place where
3864this may happen. (This is sometimes called a @dfn{data breakpoint}.)
3865The expression may be as simple as the value of a single variable, or
3866as complex as many variables combined by operators. Examples include:
3867
3868@itemize @bullet
3869@item
3870A reference to the value of a single variable.
3871
3872@item
3873An address cast to an appropriate data type. For example,
3874@samp{*(int *)0x12345678} will watch a 4-byte region at the specified
3875address (assuming an @code{int} occupies 4 bytes).
3876
3877@item
3878An arbitrarily complex expression, such as @samp{a*b + c/d}. The
3879expression can use any operators valid in the program's native
3880language (@pxref{Languages}).
3881@end itemize
3882
3883You can set a watchpoint on an expression even if the expression can
3884not be evaluated yet. For instance, you can set a watchpoint on
3885@samp{*global_ptr} before @samp{global_ptr} is initialized.
3886@value{GDBN} will stop when your program sets @samp{global_ptr} and
3887the expression produces a valid value. If the expression becomes
3888valid in some other way than changing a variable (e.g.@: if the memory
3889pointed to by @samp{*global_ptr} becomes readable as the result of a
3890@code{malloc} call), @value{GDBN} may not stop until the next time
3891the expression changes.
3892
3893@cindex software watchpoints
3894@cindex hardware watchpoints
3895Depending on your system, watchpoints may be implemented in software or
3896hardware. @value{GDBN} does software watchpointing by single-stepping your
3897program and testing the variable's value each time, which is hundreds of
3898times slower than normal execution. (But this may still be worth it, to
3899catch errors where you have no clue what part of your program is the
3900culprit.)
3901
3902On some systems, such as HP-UX, PowerPC, @sc{gnu}/Linux and most other
3903x86-based targets, @value{GDBN} includes support for hardware
3904watchpoints, which do not slow down the running of your program.
3905
3906@table @code
3907@kindex watch
3908@item watch @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{threadnum}@r{]} @r{[}mask @var{maskvalue}@r{]}
3909Set a watchpoint for an expression. @value{GDBN} will break when the
3910expression @var{expr} is written into by the program and its value
3911changes. The simplest (and the most popular) use of this command is
3912to watch the value of a single variable:
3913
3914@smallexample
3915(@value{GDBP}) watch foo
3916@end smallexample
3917
3918If the command includes a @code{@r{[}thread @var{threadnum}@r{]}}
3919argument, @value{GDBN} breaks only when the thread identified by
3920@var{threadnum} changes the value of @var{expr}. If any other threads
3921change the value of @var{expr}, @value{GDBN} will not break. Note
3922that watchpoints restricted to a single thread in this way only work
3923with Hardware Watchpoints.
3924
3925Ordinarily a watchpoint respects the scope of variables in @var{expr}
3926(see below). The @code{-location} argument tells @value{GDBN} to
3927instead watch the memory referred to by @var{expr}. In this case,
3928@value{GDBN} will evaluate @var{expr}, take the address of the result,
3929and watch the memory at that address. The type of the result is used
3930to determine the size of the watched memory. If the expression's
3931result does not have an address, then @value{GDBN} will print an
3932error.
3933
3934The @code{@r{[}mask @var{maskvalue}@r{]}} argument allows creation
3935of masked watchpoints, if the current architecture supports this
3936feature (e.g., PowerPC Embedded architecture, see @ref{PowerPC
3937Embedded}.) A @dfn{masked watchpoint} specifies a mask in addition
3938to an address to watch. The mask specifies that some bits of an address
3939(the bits which are reset in the mask) should be ignored when matching
3940the address accessed by the inferior against the watchpoint address.
3941Thus, a masked watchpoint watches many addresses simultaneously---those
3942addresses whose unmasked bits are identical to the unmasked bits in the
3943watchpoint address. The @code{mask} argument implies @code{-location}.
3944Examples:
3945
3946@smallexample
3947(@value{GDBP}) watch foo mask 0xffff00ff
3948(@value{GDBP}) watch *0xdeadbeef mask 0xffffff00
3949@end smallexample
3950
3951@kindex rwatch
3952@item rwatch @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{threadnum}@r{]} @r{[}mask @var{maskvalue}@r{]}
3953Set a watchpoint that will break when the value of @var{expr} is read
3954by the program.
3955
3956@kindex awatch
3957@item awatch @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{threadnum}@r{]} @r{[}mask @var{maskvalue}@r{]}
3958Set a watchpoint that will break when @var{expr} is either read from
3959or written into by the program.
3960
3961@kindex info watchpoints @r{[}@var{n}@dots{}@r{]}
3962@item info watchpoints @r{[}@var{n}@dots{}@r{]}
3963This command prints a list of watchpoints, using the same format as
3964@code{info break} (@pxref{Set Breaks}).
3965@end table
3966
3967If you watch for a change in a numerically entered address you need to
3968dereference it, as the address itself is just a constant number which will
3969never change. @value{GDBN} refuses to create a watchpoint that watches
3970a never-changing value:
3971
3972@smallexample
3973(@value{GDBP}) watch 0x600850
3974Cannot watch constant value 0x600850.
3975(@value{GDBP}) watch *(int *) 0x600850
3976Watchpoint 1: *(int *) 6293584
3977@end smallexample
3978
3979@value{GDBN} sets a @dfn{hardware watchpoint} if possible. Hardware
3980watchpoints execute very quickly, and the debugger reports a change in
3981value at the exact instruction where the change occurs. If @value{GDBN}
3982cannot set a hardware watchpoint, it sets a software watchpoint, which
3983executes more slowly and reports the change in value at the next
3984@emph{statement}, not the instruction, after the change occurs.
3985
3986@cindex use only software watchpoints
3987You can force @value{GDBN} to use only software watchpoints with the
3988@kbd{set can-use-hw-watchpoints 0} command. With this variable set to
3989zero, @value{GDBN} will never try to use hardware watchpoints, even if
3990the underlying system supports them. (Note that hardware-assisted
3991watchpoints that were set @emph{before} setting
3992@code{can-use-hw-watchpoints} to zero will still use the hardware
3993mechanism of watching expression values.)
3994
3995@table @code
3996@item set can-use-hw-watchpoints
3997@kindex set can-use-hw-watchpoints
3998Set whether or not to use hardware watchpoints.
3999
4000@item show can-use-hw-watchpoints
4001@kindex show can-use-hw-watchpoints
4002Show the current mode of using hardware watchpoints.
4003@end table
4004
4005For remote targets, you can restrict the number of hardware
4006watchpoints @value{GDBN} will use, see @ref{set remote
4007hardware-breakpoint-limit}.
4008
4009When you issue the @code{watch} command, @value{GDBN} reports
4010
4011@smallexample
4012Hardware watchpoint @var{num}: @var{expr}
4013@end smallexample
4014
4015@noindent
4016if it was able to set a hardware watchpoint.
4017
4018Currently, the @code{awatch} and @code{rwatch} commands can only set
4019hardware watchpoints, because accesses to data that don't change the
4020value of the watched expression cannot be detected without examining
4021every instruction as it is being executed, and @value{GDBN} does not do
4022that currently. If @value{GDBN} finds that it is unable to set a
4023hardware breakpoint with the @code{awatch} or @code{rwatch} command, it
4024will print a message like this:
4025
4026@smallexample
4027Expression cannot be implemented with read/access watchpoint.
4028@end smallexample
4029
4030Sometimes, @value{GDBN} cannot set a hardware watchpoint because the
4031data type of the watched expression is wider than what a hardware
4032watchpoint on the target machine can handle. For example, some systems
4033can only watch regions that are up to 4 bytes wide; on such systems you
4034cannot set hardware watchpoints for an expression that yields a
4035double-precision floating-point number (which is typically 8 bytes
4036wide). As a work-around, it might be possible to break the large region
4037into a series of smaller ones and watch them with separate watchpoints.
4038
4039If you set too many hardware watchpoints, @value{GDBN} might be unable
4040to insert all of them when you resume the execution of your program.
4041Since the precise number of active watchpoints is unknown until such
4042time as the program is about to be resumed, @value{GDBN} might not be
4043able to warn you about this when you set the watchpoints, and the
4044warning will be printed only when the program is resumed:
4045
4046@smallexample
4047Hardware watchpoint @var{num}: Could not insert watchpoint
4048@end smallexample
4049
4050@noindent
4051If this happens, delete or disable some of the watchpoints.
4052
4053Watching complex expressions that reference many variables can also
4054exhaust the resources available for hardware-assisted watchpoints.
4055That's because @value{GDBN} needs to watch every variable in the
4056expression with separately allocated resources.
4057
4058If you call a function interactively using @code{print} or @code{call},
4059any watchpoints you have set will be inactive until @value{GDBN} reaches another
4060kind of breakpoint or the call completes.
4061
4062@value{GDBN} automatically deletes watchpoints that watch local
4063(automatic) variables, or expressions that involve such variables, when
4064they go out of scope, that is, when the execution leaves the block in
4065which these variables were defined. In particular, when the program
4066being debugged terminates, @emph{all} local variables go out of scope,
4067and so only watchpoints that watch global variables remain set. If you
4068rerun the program, you will need to set all such watchpoints again. One
4069way of doing that would be to set a code breakpoint at the entry to the
4070@code{main} function and when it breaks, set all the watchpoints.
4071
4072@cindex watchpoints and threads
4073@cindex threads and watchpoints
4074In multi-threaded programs, watchpoints will detect changes to the
4075watched expression from every thread.
4076
4077@quotation
4078@emph{Warning:} In multi-threaded programs, software watchpoints
4079have only limited usefulness. If @value{GDBN} creates a software
4080watchpoint, it can only watch the value of an expression @emph{in a
4081single thread}. If you are confident that the expression can only
4082change due to the current thread's activity (and if you are also
4083confident that no other thread can become current), then you can use
4084software watchpoints as usual. However, @value{GDBN} may not notice
4085when a non-current thread's activity changes the expression. (Hardware
4086watchpoints, in contrast, watch an expression in all threads.)
4087@end quotation
4088
4089@xref{set remote hardware-watchpoint-limit}.
4090
4091@node Set Catchpoints
4092@subsection Setting Catchpoints
4093@cindex catchpoints, setting
4094@cindex exception handlers
4095@cindex event handling
4096
4097You can use @dfn{catchpoints} to cause the debugger to stop for certain
4098kinds of program events, such as C@t{++} exceptions or the loading of a
4099shared library. Use the @code{catch} command to set a catchpoint.
4100
4101@table @code
4102@kindex catch
4103@item catch @var{event}
4104Stop when @var{event} occurs. @var{event} can be any of the following:
4105
4106@table @code
4107@item throw @r{[}@var{regexp}@r{]}
4108@itemx rethrow @r{[}@var{regexp}@r{]}
4109@itemx catch @r{[}@var{regexp}@r{]}
4110@cindex stop on C@t{++} exceptions
4111The throwing, re-throwing, or catching of a C@t{++} exception.
4112
4113If @var{regexp} is given, then only exceptions whose type matches the
4114regular expression will be caught.
4115
4116@vindex $_exception@r{, convenience variable}
4117The convenience variable @code{$_exception} is available at an
4118exception-related catchpoint, on some systems. This holds the
4119exception being thrown.
4120
4121There are currently some limitations to C@t{++} exception handling in
4122@value{GDBN}:
4123
4124@itemize @bullet
4125@item
4126The support for these commands is system-dependent. Currently, only
4127systems using the @samp{gnu-v3} C@t{++} ABI (@pxref{ABI}) are
4128supported.
4129
4130@item
4131The regular expression feature and the @code{$_exception} convenience
4132variable rely on the presence of some SDT probes in @code{libstdc++}.
4133If these probes are not present, then these features cannot be used.
4134These probes were first available in the GCC 4.8 release, but whether
4135or not they are available in your GCC also depends on how it was
4136built.
4137
4138@item
4139The @code{$_exception} convenience variable is only valid at the
4140instruction at which an exception-related catchpoint is set.
4141
4142@item
4143When an exception-related catchpoint is hit, @value{GDBN} stops at a
4144location in the system library which implements runtime exception
4145support for C@t{++}, usually @code{libstdc++}. You can use @code{up}
4146(@pxref{Selection}) to get to your code.
4147
4148@item
4149If you call a function interactively, @value{GDBN} normally returns
4150control to you when the function has finished executing. If the call
4151raises an exception, however, the call may bypass the mechanism that
4152returns control to you and cause your program either to abort or to
4153simply continue running until it hits a breakpoint, catches a signal
4154that @value{GDBN} is listening for, or exits. This is the case even if
4155you set a catchpoint for the exception; catchpoints on exceptions are
4156disabled within interactive calls. @xref{Calling}, for information on
4157controlling this with @code{set unwind-on-terminating-exception}.
4158
4159@item
4160You cannot raise an exception interactively.
4161
4162@item
4163You cannot install an exception handler interactively.
4164@end itemize
4165
4166@item exception
4167@cindex Ada exception catching
4168@cindex catch Ada exceptions
4169An Ada exception being raised. If an exception name is specified
4170at the end of the command (eg @code{catch exception Program_Error}),
4171the debugger will stop only when this specific exception is raised.
4172Otherwise, the debugger stops execution when any Ada exception is raised.
4173
4174When inserting an exception catchpoint on a user-defined exception whose
4175name is identical to one of the exceptions defined by the language, the
4176fully qualified name must be used as the exception name. Otherwise,
4177@value{GDBN} will assume that it should stop on the pre-defined exception
4178rather than the user-defined one. For instance, assuming an exception
4179called @code{Constraint_Error} is defined in package @code{Pck}, then
4180the command to use to catch such exceptions is @kbd{catch exception
4181Pck.Constraint_Error}.
4182
4183@item exception unhandled
4184An exception that was raised but is not handled by the program.
4185
4186@item assert
4187A failed Ada assertion.
4188
4189@item exec
4190@cindex break on fork/exec
4191A call to @code{exec}. This is currently only available for HP-UX
4192and @sc{gnu}/Linux.
4193
4194@item syscall
4195@itemx syscall @r{[}@var{name} @r{|} @var{number}@r{]} @dots{}
4196@cindex break on a system call.
4197A call to or return from a system call, a.k.a.@: @dfn{syscall}. A
4198syscall is a mechanism for application programs to request a service
4199from the operating system (OS) or one of the OS system services.
4200@value{GDBN} can catch some or all of the syscalls issued by the
4201debuggee, and show the related information for each syscall. If no
4202argument is specified, calls to and returns from all system calls
4203will be caught.
4204
4205@var{name} can be any system call name that is valid for the
4206underlying OS. Just what syscalls are valid depends on the OS. On
4207GNU and Unix systems, you can find the full list of valid syscall
4208names on @file{/usr/include/asm/unistd.h}.
4209
4210@c For MS-Windows, the syscall names and the corresponding numbers
4211@c can be found, e.g., on this URL:
4212@c http://www.metasploit.com/users/opcode/syscalls.html
4213@c but we don't support Windows syscalls yet.
4214
4215Normally, @value{GDBN} knows in advance which syscalls are valid for
4216each OS, so you can use the @value{GDBN} command-line completion
4217facilities (@pxref{Completion,, command completion}) to list the
4218available choices.
4219
4220You may also specify the system call numerically. A syscall's
4221number is the value passed to the OS's syscall dispatcher to
4222identify the requested service. When you specify the syscall by its
4223name, @value{GDBN} uses its database of syscalls to convert the name
4224into the corresponding numeric code, but using the number directly
4225may be useful if @value{GDBN}'s database does not have the complete
4226list of syscalls on your system (e.g., because @value{GDBN} lags
4227behind the OS upgrades).
4228
4229The example below illustrates how this command works if you don't provide
4230arguments to it:
4231
4232@smallexample
4233(@value{GDBP}) catch syscall
4234Catchpoint 1 (syscall)
4235(@value{GDBP}) r
4236Starting program: /tmp/catch-syscall
4237
4238Catchpoint 1 (call to syscall 'close'), \
4239 0xffffe424 in __kernel_vsyscall ()
4240(@value{GDBP}) c
4241Continuing.
4242
4243Catchpoint 1 (returned from syscall 'close'), \
4244 0xffffe424 in __kernel_vsyscall ()
4245(@value{GDBP})
4246@end smallexample
4247
4248Here is an example of catching a system call by name:
4249
4250@smallexample
4251(@value{GDBP}) catch syscall chroot
4252Catchpoint 1 (syscall 'chroot' [61])
4253(@value{GDBP}) r
4254Starting program: /tmp/catch-syscall
4255
4256Catchpoint 1 (call to syscall 'chroot'), \
4257 0xffffe424 in __kernel_vsyscall ()
4258(@value{GDBP}) c
4259Continuing.
4260
4261Catchpoint 1 (returned from syscall 'chroot'), \
4262 0xffffe424 in __kernel_vsyscall ()
4263(@value{GDBP})
4264@end smallexample
4265
4266An example of specifying a system call numerically. In the case
4267below, the syscall number has a corresponding entry in the XML
4268file, so @value{GDBN} finds its name and prints it:
4269
4270@smallexample
4271(@value{GDBP}) catch syscall 252
4272Catchpoint 1 (syscall(s) 'exit_group')
4273(@value{GDBP}) r
4274Starting program: /tmp/catch-syscall
4275
4276Catchpoint 1 (call to syscall 'exit_group'), \
4277 0xffffe424 in __kernel_vsyscall ()
4278(@value{GDBP}) c
4279Continuing.
4280
4281Program exited normally.
4282(@value{GDBP})
4283@end smallexample
4284
4285However, there can be situations when there is no corresponding name
4286in XML file for that syscall number. In this case, @value{GDBN} prints
4287a warning message saying that it was not able to find the syscall name,
4288but the catchpoint will be set anyway. See the example below:
4289
4290@smallexample
4291(@value{GDBP}) catch syscall 764
4292warning: The number '764' does not represent a known syscall.
4293Catchpoint 2 (syscall 764)
4294(@value{GDBP})
4295@end smallexample
4296
4297If you configure @value{GDBN} using the @samp{--without-expat} option,
4298it will not be able to display syscall names. Also, if your
4299architecture does not have an XML file describing its system calls,
4300you will not be able to see the syscall names. It is important to
4301notice that these two features are used for accessing the syscall
4302name database. In either case, you will see a warning like this:
4303
4304@smallexample
4305(@value{GDBP}) catch syscall
4306warning: Could not open "syscalls/i386-linux.xml"
4307warning: Could not load the syscall XML file 'syscalls/i386-linux.xml'.
4308GDB will not be able to display syscall names.
4309Catchpoint 1 (syscall)
4310(@value{GDBP})
4311@end smallexample
4312
4313Of course, the file name will change depending on your architecture and system.
4314
4315Still using the example above, you can also try to catch a syscall by its
4316number. In this case, you would see something like:
4317
4318@smallexample
4319(@value{GDBP}) catch syscall 252
4320Catchpoint 1 (syscall(s) 252)
4321@end smallexample
4322
4323Again, in this case @value{GDBN} would not be able to display syscall's names.
4324
4325@item fork
4326A call to @code{fork}. This is currently only available for HP-UX
4327and @sc{gnu}/Linux.
4328
4329@item vfork
4330A call to @code{vfork}. This is currently only available for HP-UX
4331and @sc{gnu}/Linux.
4332
4333@item load @r{[}regexp@r{]}
4334@itemx unload @r{[}regexp@r{]}
4335The loading or unloading of a shared library. If @var{regexp} is
4336given, then the catchpoint will stop only if the regular expression
4337matches one of the affected libraries.
4338
4339@item signal @r{[}@var{signal}@dots{} @r{|} @samp{all}@r{]}
4340The delivery of a signal.
4341
4342With no arguments, this catchpoint will catch any signal that is not
4343used internally by @value{GDBN}, specifically, all signals except
4344@samp{SIGTRAP} and @samp{SIGINT}.
4345
4346With the argument @samp{all}, all signals, including those used by
4347@value{GDBN}, will be caught. This argument cannot be used with other
4348signal names.
4349
4350Otherwise, the arguments are a list of signal names as given to
4351@code{handle} (@pxref{Signals}). Only signals specified in this list
4352will be caught.
4353
4354One reason that @code{catch signal} can be more useful than
4355@code{handle} is that you can attach commands and conditions to the
4356catchpoint.
4357
4358When a signal is caught by a catchpoint, the signal's @code{stop} and
4359@code{print} settings, as specified by @code{handle}, are ignored.
4360However, whether the signal is still delivered to the inferior depends
4361on the @code{pass} setting; this can be changed in the catchpoint's
4362commands.
4363
4364@end table
4365
4366@item tcatch @var{event}
4367Set a catchpoint that is enabled only for one stop. The catchpoint is
4368automatically deleted after the first time the event is caught.
4369
4370@end table
4371
4372Use the @code{info break} command to list the current catchpoints.
4373
4374
4375@node Delete Breaks
4376@subsection Deleting Breakpoints
4377
4378@cindex clearing breakpoints, watchpoints, catchpoints
4379@cindex deleting breakpoints, watchpoints, catchpoints
4380It is often necessary to eliminate a breakpoint, watchpoint, or
4381catchpoint once it has done its job and you no longer want your program
4382to stop there. This is called @dfn{deleting} the breakpoint. A
4383breakpoint that has been deleted no longer exists; it is forgotten.
4384
4385With the @code{clear} command you can delete breakpoints according to
4386where they are in your program. With the @code{delete} command you can
4387delete individual breakpoints, watchpoints, or catchpoints by specifying
4388their breakpoint numbers.
4389
4390It is not necessary to delete a breakpoint to proceed past it. @value{GDBN}
4391automatically ignores breakpoints on the first instruction to be executed
4392when you continue execution without changing the execution address.
4393
4394@table @code
4395@kindex clear
4396@item clear
4397Delete any breakpoints at the next instruction to be executed in the
4398selected stack frame (@pxref{Selection, ,Selecting a Frame}). When
4399the innermost frame is selected, this is a good way to delete a
4400breakpoint where your program just stopped.
4401
4402@item clear @var{location}
4403Delete any breakpoints set at the specified @var{location}.
4404@xref{Specify Location}, for the various forms of @var{location}; the
4405most useful ones are listed below:
4406
4407@table @code
4408@item clear @var{function}
4409@itemx clear @var{filename}:@var{function}
4410Delete any breakpoints set at entry to the named @var{function}.
4411
4412@item clear @var{linenum}
4413@itemx clear @var{filename}:@var{linenum}
4414Delete any breakpoints set at or within the code of the specified
4415@var{linenum} of the specified @var{filename}.
4416@end table
4417
4418@cindex delete breakpoints
4419@kindex delete
4420@kindex d @r{(@code{delete})}
4421@item delete @r{[}breakpoints@r{]} @r{[}@var{range}@dots{}@r{]}
4422Delete the breakpoints, watchpoints, or catchpoints of the breakpoint
4423ranges specified as arguments. If no argument is specified, delete all
4424breakpoints (@value{GDBN} asks confirmation, unless you have @code{set
4425confirm off}). You can abbreviate this command as @code{d}.
4426@end table
4427
4428@node Disabling
4429@subsection Disabling Breakpoints
4430
4431@cindex enable/disable a breakpoint
4432Rather than deleting a breakpoint, watchpoint, or catchpoint, you might
4433prefer to @dfn{disable} it. This makes the breakpoint inoperative as if
4434it had been deleted, but remembers the information on the breakpoint so
4435that you can @dfn{enable} it again later.
4436
4437You disable and enable breakpoints, watchpoints, and catchpoints with
4438the @code{enable} and @code{disable} commands, optionally specifying
4439one or more breakpoint numbers as arguments. Use @code{info break} to
4440print a list of all breakpoints, watchpoints, and catchpoints if you
4441do not know which numbers to use.
4442
4443Disabling and enabling a breakpoint that has multiple locations
4444affects all of its locations.
4445
4446A breakpoint, watchpoint, or catchpoint can have any of several
4447different states of enablement:
4448
4449@itemize @bullet
4450@item
4451Enabled. The breakpoint stops your program. A breakpoint set
4452with the @code{break} command starts out in this state.
4453@item
4454Disabled. The breakpoint has no effect on your program.
4455@item
4456Enabled once. The breakpoint stops your program, but then becomes
4457disabled.
4458@item
4459Enabled for a count. The breakpoint stops your program for the next
4460N times, then becomes disabled.
4461@item
4462Enabled for deletion. The breakpoint stops your program, but
4463immediately after it does so it is deleted permanently. A breakpoint
4464set with the @code{tbreak} command starts out in this state.
4465@end itemize
4466
4467You can use the following commands to enable or disable breakpoints,
4468watchpoints, and catchpoints:
4469
4470@table @code
4471@kindex disable
4472@kindex dis @r{(@code{disable})}
4473@item disable @r{[}breakpoints@r{]} @r{[}@var{range}@dots{}@r{]}
4474Disable the specified breakpoints---or all breakpoints, if none are
4475listed. A disabled breakpoint has no effect but is not forgotten. All
4476options such as ignore-counts, conditions and commands are remembered in
4477case the breakpoint is enabled again later. You may abbreviate
4478@code{disable} as @code{dis}.
4479
4480@kindex enable
4481@item enable @r{[}breakpoints@r{]} @r{[}@var{range}@dots{}@r{]}
4482Enable the specified breakpoints (or all defined breakpoints). They
4483become effective once again in stopping your program.
4484
4485@item enable @r{[}breakpoints@r{]} once @var{range}@dots{}
4486Enable the specified breakpoints temporarily. @value{GDBN} disables any
4487of these breakpoints immediately after stopping your program.
4488
4489@item enable @r{[}breakpoints@r{]} count @var{count} @var{range}@dots{}
4490Enable the specified breakpoints temporarily. @value{GDBN} records
4491@var{count} with each of the specified breakpoints, and decrements a
4492breakpoint's count when it is hit. When any count reaches 0,
4493@value{GDBN} disables that breakpoint. If a breakpoint has an ignore
4494count (@pxref{Conditions, ,Break Conditions}), that will be
4495decremented to 0 before @var{count} is affected.
4496
4497@item enable @r{[}breakpoints@r{]} delete @var{range}@dots{}
4498Enable the specified breakpoints to work once, then die. @value{GDBN}
4499deletes any of these breakpoints as soon as your program stops there.
4500Breakpoints set by the @code{tbreak} command start out in this state.
4501@end table
4502
4503@c FIXME: I think the following ``Except for [...] @code{tbreak}'' is
4504@c confusing: tbreak is also initially enabled.
4505Except for a breakpoint set with @code{tbreak} (@pxref{Set Breaks,
4506,Setting Breakpoints}), breakpoints that you set are initially enabled;
4507subsequently, they become disabled or enabled only when you use one of
4508the commands above. (The command @code{until} can set and delete a
4509breakpoint of its own, but it does not change the state of your other
4510breakpoints; see @ref{Continuing and Stepping, ,Continuing and
4511Stepping}.)
4512
4513@node Conditions
4514@subsection Break Conditions
4515@cindex conditional breakpoints
4516@cindex breakpoint conditions
4517
4518@c FIXME what is scope of break condition expr? Context where wanted?
4519@c in particular for a watchpoint?
4520The simplest sort of breakpoint breaks every time your program reaches a
4521specified place. You can also specify a @dfn{condition} for a
4522breakpoint. A condition is just a Boolean expression in your
4523programming language (@pxref{Expressions, ,Expressions}). A breakpoint with
4524a condition evaluates the expression each time your program reaches it,
4525and your program stops only if the condition is @emph{true}.
4526
4527This is the converse of using assertions for program validation; in that
4528situation, you want to stop when the assertion is violated---that is,
4529when the condition is false. In C, if you want to test an assertion expressed
4530by the condition @var{assert}, you should set the condition
4531@samp{! @var{assert}} on the appropriate breakpoint.
4532
4533Conditions are also accepted for watchpoints; you may not need them,
4534since a watchpoint is inspecting the value of an expression anyhow---but
4535it might be simpler, say, to just set a watchpoint on a variable name,
4536and specify a condition that tests whether the new value is an interesting
4537one.
4538
4539Break conditions can have side effects, and may even call functions in
4540your program. This can be useful, for example, to activate functions
4541that log program progress, or to use your own print functions to
4542format special data structures. The effects are completely predictable
4543unless there is another enabled breakpoint at the same address. (In
4544that case, @value{GDBN} might see the other breakpoint first and stop your
4545program without checking the condition of this one.) Note that
4546breakpoint commands are usually more convenient and flexible than break
4547conditions for the
4548purpose of performing side effects when a breakpoint is reached
4549(@pxref{Break Commands, ,Breakpoint Command Lists}).
4550
4551Breakpoint conditions can also be evaluated on the target's side if
4552the target supports it. Instead of evaluating the conditions locally,
4553@value{GDBN} encodes the expression into an agent expression
4554(@pxref{Agent Expressions}) suitable for execution on the target,
4555independently of @value{GDBN}. Global variables become raw memory
4556locations, locals become stack accesses, and so forth.
4557
4558In this case, @value{GDBN} will only be notified of a breakpoint trigger
4559when its condition evaluates to true. This mechanism may provide faster
4560response times depending on the performance characteristics of the target
4561since it does not need to keep @value{GDBN} informed about
4562every breakpoint trigger, even those with false conditions.
4563
4564Break conditions can be specified when a breakpoint is set, by using
4565@samp{if} in the arguments to the @code{break} command. @xref{Set
4566Breaks, ,Setting Breakpoints}. They can also be changed at any time
4567with the @code{condition} command.
4568
4569You can also use the @code{if} keyword with the @code{watch} command.
4570The @code{catch} command does not recognize the @code{if} keyword;
4571@code{condition} is the only way to impose a further condition on a
4572catchpoint.
4573
4574@table @code
4575@kindex condition
4576@item condition @var{bnum} @var{expression}
4577Specify @var{expression} as the break condition for breakpoint,
4578watchpoint, or catchpoint number @var{bnum}. After you set a condition,
4579breakpoint @var{bnum} stops your program only if the value of
4580@var{expression} is true (nonzero, in C). When you use
4581@code{condition}, @value{GDBN} checks @var{expression} immediately for
4582syntactic correctness, and to determine whether symbols in it have
4583referents in the context of your breakpoint. If @var{expression} uses
4584symbols not referenced in the context of the breakpoint, @value{GDBN}
4585prints an error message:
4586
4587@smallexample
4588No symbol "foo" in current context.
4589@end smallexample
4590
4591@noindent
4592@value{GDBN} does
4593not actually evaluate @var{expression} at the time the @code{condition}
4594command (or a command that sets a breakpoint with a condition, like
4595@code{break if @dots{}}) is given, however. @xref{Expressions, ,Expressions}.
4596
4597@item condition @var{bnum}
4598Remove the condition from breakpoint number @var{bnum}. It becomes
4599an ordinary unconditional breakpoint.
4600@end table
4601
4602@cindex ignore count (of breakpoint)
4603A special case of a breakpoint condition is to stop only when the
4604breakpoint has been reached a certain number of times. This is so
4605useful that there is a special way to do it, using the @dfn{ignore
4606count} of the breakpoint. Every breakpoint has an ignore count, which
4607is an integer. Most of the time, the ignore count is zero, and
4608therefore has no effect. But if your program reaches a breakpoint whose
4609ignore count is positive, then instead of stopping, it just decrements
4610the ignore count by one and continues. As a result, if the ignore count
4611value is @var{n}, the breakpoint does not stop the next @var{n} times
4612your program reaches it.
4613
4614@table @code
4615@kindex ignore
4616@item ignore @var{bnum} @var{count}
4617Set the ignore count of breakpoint number @var{bnum} to @var{count}.
4618The next @var{count} times the breakpoint is reached, your program's
4619execution does not stop; other than to decrement the ignore count, @value{GDBN}
4620takes no action.
4621
4622To make the breakpoint stop the next time it is reached, specify
4623a count of zero.
4624
4625When you use @code{continue} to resume execution of your program from a
4626breakpoint, you can specify an ignore count directly as an argument to
4627@code{continue}, rather than using @code{ignore}. @xref{Continuing and
4628Stepping,,Continuing and Stepping}.
4629
4630If a breakpoint has a positive ignore count and a condition, the
4631condition is not checked. Once the ignore count reaches zero,
4632@value{GDBN} resumes checking the condition.
4633
4634You could achieve the effect of the ignore count with a condition such
4635as @w{@samp{$foo-- <= 0}} using a debugger convenience variable that
4636is decremented each time. @xref{Convenience Vars, ,Convenience
4637Variables}.
4638@end table
4639
4640Ignore counts apply to breakpoints, watchpoints, and catchpoints.
4641
4642
4643@node Break Commands
4644@subsection Breakpoint Command Lists
4645
4646@cindex breakpoint commands
4647You can give any breakpoint (or watchpoint or catchpoint) a series of
4648commands to execute when your program stops due to that breakpoint. For
4649example, you might want to print the values of certain expressions, or
4650enable other breakpoints.
4651
4652@table @code
4653@kindex commands
4654@kindex end@r{ (breakpoint commands)}
4655@item commands @r{[}@var{range}@dots{}@r{]}
4656@itemx @dots{} @var{command-list} @dots{}
4657@itemx end
4658Specify a list of commands for the given breakpoints. The commands
4659themselves appear on the following lines. Type a line containing just
4660@code{end} to terminate the commands.
4661
4662To remove all commands from a breakpoint, type @code{commands} and
4663follow it immediately with @code{end}; that is, give no commands.
4664
4665With no argument, @code{commands} refers to the last breakpoint,
4666watchpoint, or catchpoint set (not to the breakpoint most recently
4667encountered). If the most recent breakpoints were set with a single
4668command, then the @code{commands} will apply to all the breakpoints
4669set by that command. This applies to breakpoints set by
4670@code{rbreak}, and also applies when a single @code{break} command
4671creates multiple breakpoints (@pxref{Ambiguous Expressions,,Ambiguous
4672Expressions}).
4673@end table
4674
4675Pressing @key{RET} as a means of repeating the last @value{GDBN} command is
4676disabled within a @var{command-list}.
4677
4678You can use breakpoint commands to start your program up again. Simply
4679use the @code{continue} command, or @code{step}, or any other command
4680that resumes execution.
4681
4682Any other commands in the command list, after a command that resumes
4683execution, are ignored. This is because any time you resume execution
4684(even with a simple @code{next} or @code{step}), you may encounter
4685another breakpoint---which could have its own command list, leading to
4686ambiguities about which list to execute.
4687
4688@kindex silent
4689If the first command you specify in a command list is @code{silent}, the
4690usual message about stopping at a breakpoint is not printed. This may
4691be desirable for breakpoints that are to print a specific message and
4692then continue. If none of the remaining commands print anything, you
4693see no sign that the breakpoint was reached. @code{silent} is
4694meaningful only at the beginning of a breakpoint command list.
4695
4696The commands @code{echo}, @code{output}, and @code{printf} allow you to
4697print precisely controlled output, and are often useful in silent
4698breakpoints. @xref{Output, ,Commands for Controlled Output}.
4699
4700For example, here is how you could use breakpoint commands to print the
4701value of @code{x} at entry to @code{foo} whenever @code{x} is positive.
4702
4703@smallexample
4704break foo if x>0
4705commands
4706silent
4707printf "x is %d\n",x
4708cont
4709end
4710@end smallexample
4711
4712One application for breakpoint commands is to compensate for one bug so
4713you can test for another. Put a breakpoint just after the erroneous line
4714of code, give it a condition to detect the case in which something
4715erroneous has been done, and give it commands to assign correct values
4716to any variables that need them. End with the @code{continue} command
4717so that your program does not stop, and start with the @code{silent}
4718command so that no output is produced. Here is an example:
4719
4720@smallexample
4721break 403
4722commands
4723silent
4724set x = y + 4
4725cont
4726end
4727@end smallexample
4728
4729@node Dynamic Printf
4730@subsection Dynamic Printf
4731
4732@cindex dynamic printf
4733@cindex dprintf
4734The dynamic printf command @code{dprintf} combines a breakpoint with
4735formatted printing of your program's data to give you the effect of
4736inserting @code{printf} calls into your program on-the-fly, without
4737having to recompile it.
4738
4739In its most basic form, the output goes to the GDB console. However,
4740you can set the variable @code{dprintf-style} for alternate handling.
4741For instance, you can ask to format the output by calling your
4742program's @code{printf} function. This has the advantage that the
4743characters go to the program's output device, so they can recorded in
4744redirects to files and so forth.
4745
4746If you are doing remote debugging with a stub or agent, you can also
4747ask to have the printf handled by the remote agent. In addition to
4748ensuring that the output goes to the remote program's device along
4749with any other output the program might produce, you can also ask that
4750the dprintf remain active even after disconnecting from the remote
4751target. Using the stub/agent is also more efficient, as it can do
4752everything without needing to communicate with @value{GDBN}.
4753
4754@table @code
4755@kindex dprintf
4756@item dprintf @var{location},@var{template},@var{expression}[,@var{expression}@dots{}]
4757Whenever execution reaches @var{location}, print the values of one or
4758more @var{expressions} under the control of the string @var{template}.
4759To print several values, separate them with commas.
4760
4761@item set dprintf-style @var{style}
4762Set the dprintf output to be handled in one of several different
4763styles enumerated below. A change of style affects all existing
4764dynamic printfs immediately. (If you need individual control over the
4765print commands, simply define normal breakpoints with
4766explicitly-supplied command lists.)
4767
4768@item gdb
4769@kindex dprintf-style gdb
4770Handle the output using the @value{GDBN} @code{printf} command.
4771
4772@item call
4773@kindex dprintf-style call
4774Handle the output by calling a function in your program (normally
4775@code{printf}).
4776
4777@item agent
4778@kindex dprintf-style agent
4779Have the remote debugging agent (such as @code{gdbserver}) handle
4780the output itself. This style is only available for agents that
4781support running commands on the target.
4782
4783@item set dprintf-function @var{function}
4784Set the function to call if the dprintf style is @code{call}. By
4785default its value is @code{printf}. You may set it to any expression.
4786that @value{GDBN} can evaluate to a function, as per the @code{call}
4787command.
4788
4789@item set dprintf-channel @var{channel}
4790Set a ``channel'' for dprintf. If set to a non-empty value,
4791@value{GDBN} will evaluate it as an expression and pass the result as
4792a first argument to the @code{dprintf-function}, in the manner of
4793@code{fprintf} and similar functions. Otherwise, the dprintf format
4794string will be the first argument, in the manner of @code{printf}.
4795
4796As an example, if you wanted @code{dprintf} output to go to a logfile
4797that is a standard I/O stream assigned to the variable @code{mylog},
4798you could do the following:
4799
4800@example
4801(gdb) set dprintf-style call
4802(gdb) set dprintf-function fprintf
4803(gdb) set dprintf-channel mylog
4804(gdb) dprintf 25,"at line 25, glob=%d\n",glob
4805Dprintf 1 at 0x123456: file main.c, line 25.
4806(gdb) info break
48071 dprintf keep y 0x00123456 in main at main.c:25
4808 call (void) fprintf (mylog,"at line 25, glob=%d\n",glob)
4809 continue
4810(gdb)
4811@end example
4812
4813Note that the @code{info break} displays the dynamic printf commands
4814as normal breakpoint commands; you can thus easily see the effect of
4815the variable settings.
4816
4817@item set disconnected-dprintf on
4818@itemx set disconnected-dprintf off
4819@kindex set disconnected-dprintf
4820Choose whether @code{dprintf} commands should continue to run if
4821@value{GDBN} has disconnected from the target. This only applies
4822if the @code{dprintf-style} is @code{agent}.
4823
4824@item show disconnected-dprintf off
4825@kindex show disconnected-dprintf
4826Show the current choice for disconnected @code{dprintf}.
4827
4828@end table
4829
4830@value{GDBN} does not check the validity of function and channel,
4831relying on you to supply values that are meaningful for the contexts
4832in which they are being used. For instance, the function and channel
4833may be the values of local variables, but if that is the case, then
4834all enabled dynamic prints must be at locations within the scope of
4835those locals. If evaluation fails, @value{GDBN} will report an error.
4836
4837@node Save Breakpoints
4838@subsection How to save breakpoints to a file
4839
4840To save breakpoint definitions to a file use the @w{@code{save
4841breakpoints}} command.
4842
4843@table @code
4844@kindex save breakpoints
4845@cindex save breakpoints to a file for future sessions
4846@item save breakpoints [@var{filename}]
4847This command saves all current breakpoint definitions together with
4848their commands and ignore counts, into a file @file{@var{filename}}
4849suitable for use in a later debugging session. This includes all
4850types of breakpoints (breakpoints, watchpoints, catchpoints,
4851tracepoints). To read the saved breakpoint definitions, use the
4852@code{source} command (@pxref{Command Files}). Note that watchpoints
4853with expressions involving local variables may fail to be recreated
4854because it may not be possible to access the context where the
4855watchpoint is valid anymore. Because the saved breakpoint definitions
4856are simply a sequence of @value{GDBN} commands that recreate the
4857breakpoints, you can edit the file in your favorite editing program,
4858and remove the breakpoint definitions you're not interested in, or
4859that can no longer be recreated.
4860@end table
4861
4862@node Static Probe Points
4863@subsection Static Probe Points
4864
4865@cindex static probe point, SystemTap
4866@value{GDBN} supports @dfn{SDT} probes in the code. @acronym{SDT} stands
4867for Statically Defined Tracing, and the probes are designed to have a tiny
4868runtime code and data footprint, and no dynamic relocations. They are
4869usable from assembly, C and C@t{++} languages. See
4870@uref{http://sourceware.org/systemtap/wiki/UserSpaceProbeImplementation}
4871for a good reference on how the @acronym{SDT} probes are implemented.
4872
4873Currently, @code{SystemTap} (@uref{http://sourceware.org/systemtap/})
4874@acronym{SDT} probes are supported on ELF-compatible systems. See
4875@uref{http://sourceware.org/systemtap/wiki/AddingUserSpaceProbingToApps}
4876for more information on how to add @code{SystemTap} @acronym{SDT} probes
4877in your applications.
4878
4879@cindex semaphores on static probe points
4880Some probes have an associated semaphore variable; for instance, this
4881happens automatically if you defined your probe using a DTrace-style
4882@file{.d} file. If your probe has a semaphore, @value{GDBN} will
4883automatically enable it when you specify a breakpoint using the
4884@samp{-probe-stap} notation. But, if you put a breakpoint at a probe's
4885location by some other method (e.g., @code{break file:line}), then
4886@value{GDBN} will not automatically set the semaphore.
4887
4888You can examine the available static static probes using @code{info
4889probes}, with optional arguments:
4890
4891@table @code
4892@kindex info probes
4893@item info probes stap @r{[}@var{provider} @r{[}@var{name} @r{[}@var{objfile}@r{]}@r{]}@r{]}
4894If given, @var{provider} is a regular expression used to match against provider
4895names when selecting which probes to list. If omitted, probes by all
4896probes from all providers are listed.
4897
4898If given, @var{name} is a regular expression to match against probe names
4899when selecting which probes to list. If omitted, probe names are not
4900considered when deciding whether to display them.
4901
4902If given, @var{objfile} is a regular expression used to select which
4903object files (executable or shared libraries) to examine. If not
4904given, all object files are considered.
4905
4906@item info probes all
4907List the available static probes, from all types.
4908@end table
4909
4910@vindex $_probe_arg@r{, convenience variable}
4911A probe may specify up to twelve arguments. These are available at the
4912point at which the probe is defined---that is, when the current PC is
4913at the probe's location. The arguments are available using the
4914convenience variables (@pxref{Convenience Vars})
4915@code{$_probe_arg0}@dots{}@code{$_probe_arg11}. Each probe argument is
4916an integer of the appropriate size; types are not preserved. The
4917convenience variable @code{$_probe_argc} holds the number of arguments
4918at the current probe point.
4919
4920These variables are always available, but attempts to access them at
4921any location other than a probe point will cause @value{GDBN} to give
4922an error message.
4923
4924
4925@c @ifclear BARETARGET
4926@node Error in Breakpoints
4927@subsection ``Cannot insert breakpoints''
4928
4929If you request too many active hardware-assisted breakpoints and
4930watchpoints, you will see this error message:
4931
4932@c FIXME: the precise wording of this message may change; the relevant
4933@c source change is not committed yet (Sep 3, 1999).
4934@smallexample
4935Stopped; cannot insert breakpoints.
4936You may have requested too many hardware breakpoints and watchpoints.
4937@end smallexample
4938
4939@noindent
4940This message is printed when you attempt to resume the program, since
4941only then @value{GDBN} knows exactly how many hardware breakpoints and
4942watchpoints it needs to insert.
4943
4944When this message is printed, you need to disable or remove some of the
4945hardware-assisted breakpoints and watchpoints, and then continue.
4946
4947@node Breakpoint-related Warnings
4948@subsection ``Breakpoint address adjusted...''
4949@cindex breakpoint address adjusted
4950
4951Some processor architectures place constraints on the addresses at
4952which breakpoints may be placed. For architectures thus constrained,
4953@value{GDBN} will attempt to adjust the breakpoint's address to comply
4954with the constraints dictated by the architecture.
4955
4956One example of such an architecture is the Fujitsu FR-V. The FR-V is
4957a VLIW architecture in which a number of RISC-like instructions may be
4958bundled together for parallel execution. The FR-V architecture
4959constrains the location of a breakpoint instruction within such a
4960bundle to the instruction with the lowest address. @value{GDBN}
4961honors this constraint by adjusting a breakpoint's address to the
4962first in the bundle.
4963
4964It is not uncommon for optimized code to have bundles which contain
4965instructions from different source statements, thus it may happen that
4966a breakpoint's address will be adjusted from one source statement to
4967another. Since this adjustment may significantly alter @value{GDBN}'s
4968breakpoint related behavior from what the user expects, a warning is
4969printed when the breakpoint is first set and also when the breakpoint
4970is hit.
4971
4972A warning like the one below is printed when setting a breakpoint
4973that's been subject to address adjustment:
4974
4975@smallexample
4976warning: Breakpoint address adjusted from 0x00010414 to 0x00010410.
4977@end smallexample
4978
4979Such warnings are printed both for user settable and @value{GDBN}'s
4980internal breakpoints. If you see one of these warnings, you should
4981verify that a breakpoint set at the adjusted address will have the
4982desired affect. If not, the breakpoint in question may be removed and
4983other breakpoints may be set which will have the desired behavior.
4984E.g., it may be sufficient to place the breakpoint at a later
4985instruction. A conditional breakpoint may also be useful in some
4986cases to prevent the breakpoint from triggering too often.
4987
4988@value{GDBN} will also issue a warning when stopping at one of these
4989adjusted breakpoints:
4990
4991@smallexample
4992warning: Breakpoint 1 address previously adjusted from 0x00010414
4993to 0x00010410.
4994@end smallexample
4995
4996When this warning is encountered, it may be too late to take remedial
4997action except in cases where the breakpoint is hit earlier or more
4998frequently than expected.
4999
5000@node Continuing and Stepping
5001@section Continuing and Stepping
5002
5003@cindex stepping
5004@cindex continuing
5005@cindex resuming execution
5006@dfn{Continuing} means resuming program execution until your program
5007completes normally. In contrast, @dfn{stepping} means executing just
5008one more ``step'' of your program, where ``step'' may mean either one
5009line of source code, or one machine instruction (depending on what
5010particular command you use). Either when continuing or when stepping,
5011your program may stop even sooner, due to a breakpoint or a signal. (If
5012it stops due to a signal, you may want to use @code{handle}, or use
5013@samp{signal 0} to resume execution. @xref{Signals, ,Signals}.)
5014
5015@table @code
5016@kindex continue
5017@kindex c @r{(@code{continue})}
5018@kindex fg @r{(resume foreground execution)}
5019@item continue @r{[}@var{ignore-count}@r{]}
5020@itemx c @r{[}@var{ignore-count}@r{]}
5021@itemx fg @r{[}@var{ignore-count}@r{]}
5022Resume program execution, at the address where your program last stopped;
5023any breakpoints set at that address are bypassed. The optional argument
5024@var{ignore-count} allows you to specify a further number of times to
5025ignore a breakpoint at this location; its effect is like that of
5026@code{ignore} (@pxref{Conditions, ,Break Conditions}).
5027
5028The argument @var{ignore-count} is meaningful only when your program
5029stopped due to a breakpoint. At other times, the argument to
5030@code{continue} is ignored.
5031
5032The synonyms @code{c} and @code{fg} (for @dfn{foreground}, as the
5033debugged program is deemed to be the foreground program) are provided
5034purely for convenience, and have exactly the same behavior as
5035@code{continue}.
5036@end table
5037
5038To resume execution at a different place, you can use @code{return}
5039(@pxref{Returning, ,Returning from a Function}) to go back to the
5040calling function; or @code{jump} (@pxref{Jumping, ,Continuing at a
5041Different Address}) to go to an arbitrary location in your program.
5042
5043A typical technique for using stepping is to set a breakpoint
5044(@pxref{Breakpoints, ,Breakpoints; Watchpoints; and Catchpoints}) at the
5045beginning of the function or the section of your program where a problem
5046is believed to lie, run your program until it stops at that breakpoint,
5047and then step through the suspect area, examining the variables that are
5048interesting, until you see the problem happen.
5049
5050@table @code
5051@kindex step
5052@kindex s @r{(@code{step})}
5053@item step
5054Continue running your program until control reaches a different source
5055line, then stop it and return control to @value{GDBN}. This command is
5056abbreviated @code{s}.
5057
5058@quotation
5059@c "without debugging information" is imprecise; actually "without line
5060@c numbers in the debugging information". (gcc -g1 has debugging info but
5061@c not line numbers). But it seems complex to try to make that
5062@c distinction here.
5063@emph{Warning:} If you use the @code{step} command while control is
5064within a function that was compiled without debugging information,
5065execution proceeds until control reaches a function that does have
5066debugging information. Likewise, it will not step into a function which
5067is compiled without debugging information. To step through functions
5068without debugging information, use the @code{stepi} command, described
5069below.
5070@end quotation
5071
5072The @code{step} command only stops at the first instruction of a source
5073line. This prevents the multiple stops that could otherwise occur in
5074@code{switch} statements, @code{for} loops, etc. @code{step} continues
5075to stop if a function that has debugging information is called within
5076the line. In other words, @code{step} @emph{steps inside} any functions
5077called within the line.
5078
5079Also, the @code{step} command only enters a function if there is line
5080number information for the function. Otherwise it acts like the
5081@code{next} command. This avoids problems when using @code{cc -gl}
5082on @acronym{MIPS} machines. Previously, @code{step} entered subroutines if there
5083was any debugging information about the routine.
5084
5085@item step @var{count}
5086Continue running as in @code{step}, but do so @var{count} times. If a
5087breakpoint is reached, or a signal not related to stepping occurs before
5088@var{count} steps, stepping stops right away.
5089
5090@kindex next
5091@kindex n @r{(@code{next})}
5092@item next @r{[}@var{count}@r{]}
5093Continue to the next source line in the current (innermost) stack frame.
5094This is similar to @code{step}, but function calls that appear within
5095the line of code are executed without stopping. Execution stops when
5096control reaches a different line of code at the original stack level
5097that was executing when you gave the @code{next} command. This command
5098is abbreviated @code{n}.
5099
5100An argument @var{count} is a repeat count, as for @code{step}.
5101
5102
5103@c FIX ME!! Do we delete this, or is there a way it fits in with
5104@c the following paragraph? --- Vctoria
5105@c
5106@c @code{next} within a function that lacks debugging information acts like
5107@c @code{step}, but any function calls appearing within the code of the
5108@c function are executed without stopping.
5109
5110The @code{next} command only stops at the first instruction of a
5111source line. This prevents multiple stops that could otherwise occur in
5112@code{switch} statements, @code{for} loops, etc.
5113
5114@kindex set step-mode
5115@item set step-mode
5116@cindex functions without line info, and stepping
5117@cindex stepping into functions with no line info
5118@itemx set step-mode on
5119The @code{set step-mode on} command causes the @code{step} command to
5120stop at the first instruction of a function which contains no debug line
5121information rather than stepping over it.
5122
5123This is useful in cases where you may be interested in inspecting the
5124machine instructions of a function which has no symbolic info and do not
5125want @value{GDBN} to automatically skip over this function.
5126
5127@item set step-mode off
5128Causes the @code{step} command to step over any functions which contains no
5129debug information. This is the default.
5130
5131@item show step-mode
5132Show whether @value{GDBN} will stop in or step over functions without
5133source line debug information.
5134
5135@kindex finish
5136@kindex fin @r{(@code{finish})}
5137@item finish
5138Continue running until just after function in the selected stack frame
5139returns. Print the returned value (if any). This command can be
5140abbreviated as @code{fin}.
5141
5142Contrast this with the @code{return} command (@pxref{Returning,
5143,Returning from a Function}).
5144
5145@kindex until
5146@kindex u @r{(@code{until})}
5147@cindex run until specified location
5148@item until
5149@itemx u
5150Continue running until a source line past the current line, in the
5151current stack frame, is reached. This command is used to avoid single
5152stepping through a loop more than once. It is like the @code{next}
5153command, except that when @code{until} encounters a jump, it
5154automatically continues execution until the program counter is greater
5155than the address of the jump.
5156
5157This means that when you reach the end of a loop after single stepping
5158though it, @code{until} makes your program continue execution until it
5159exits the loop. In contrast, a @code{next} command at the end of a loop
5160simply steps back to the beginning of the loop, which forces you to step
5161through the next iteration.
5162
5163@code{until} always stops your program if it attempts to exit the current
5164stack frame.
5165
5166@code{until} may produce somewhat counterintuitive results if the order
5167of machine code does not match the order of the source lines. For
5168example, in the following excerpt from a debugging session, the @code{f}
5169(@code{frame}) command shows that execution is stopped at line
5170@code{206}; yet when we use @code{until}, we get to line @code{195}:
5171
5172@smallexample
5173(@value{GDBP}) f
5174#0 main (argc=4, argv=0xf7fffae8) at m4.c:206
5175206 expand_input();
5176(@value{GDBP}) until
5177195 for ( ; argc > 0; NEXTARG) @{
5178@end smallexample
5179
5180This happened because, for execution efficiency, the compiler had
5181generated code for the loop closure test at the end, rather than the
5182start, of the loop---even though the test in a C @code{for}-loop is
5183written before the body of the loop. The @code{until} command appeared
5184to step back to the beginning of the loop when it advanced to this
5185expression; however, it has not really gone to an earlier
5186statement---not in terms of the actual machine code.
5187
5188@code{until} with no argument works by means of single
5189instruction stepping, and hence is slower than @code{until} with an
5190argument.
5191
5192@item until @var{location}
5193@itemx u @var{location}
5194Continue running your program until either the specified location is
5195reached, or the current stack frame returns. @var{location} is any of
5196the forms described in @ref{Specify Location}.
5197This form of the command uses temporary breakpoints, and
5198hence is quicker than @code{until} without an argument. The specified
5199location is actually reached only if it is in the current frame. This
5200implies that @code{until} can be used to skip over recursive function
5201invocations. For instance in the code below, if the current location is
5202line @code{96}, issuing @code{until 99} will execute the program up to
5203line @code{99} in the same invocation of factorial, i.e., after the inner
5204invocations have returned.
5205
5206@smallexample
520794 int factorial (int value)
520895 @{
520996 if (value > 1) @{
521097 value *= factorial (value - 1);
521198 @}
521299 return (value);
5213100 @}
5214@end smallexample
5215
5216
5217@kindex advance @var{location}
5218@item advance @var{location}
5219Continue running the program up to the given @var{location}. An argument is
5220required, which should be of one of the forms described in
5221@ref{Specify Location}.
5222Execution will also stop upon exit from the current stack
5223frame. This command is similar to @code{until}, but @code{advance} will
5224not skip over recursive function calls, and the target location doesn't
5225have to be in the same frame as the current one.
5226
5227
5228@kindex stepi
5229@kindex si @r{(@code{stepi})}
5230@item stepi
5231@itemx stepi @var{arg}
5232@itemx si
5233Execute one machine instruction, then stop and return to the debugger.
5234
5235It is often useful to do @samp{display/i $pc} when stepping by machine
5236instructions. This makes @value{GDBN} automatically display the next
5237instruction to be executed, each time your program stops. @xref{Auto
5238Display,, Automatic Display}.
5239
5240An argument is a repeat count, as in @code{step}.
5241
5242@need 750
5243@kindex nexti
5244@kindex ni @r{(@code{nexti})}
5245@item nexti
5246@itemx nexti @var{arg}
5247@itemx ni
5248Execute one machine instruction, but if it is a function call,
5249proceed until the function returns.
5250
5251An argument is a repeat count, as in @code{next}.
5252
5253@end table
5254
5255@anchor{range stepping}
5256@cindex range stepping
5257@cindex target-assisted range stepping
5258By default, and if available, @value{GDBN} makes use of
5259target-assisted @dfn{range stepping}. In other words, whenever you
5260use a stepping command (e.g., @code{step}, @code{next}), @value{GDBN}
5261tells the target to step the corresponding range of instruction
5262addresses instead of issuing multiple single-steps. This speeds up
5263line stepping, particularly for remote targets. Ideally, there should
5264be no reason you would want to turn range stepping off. However, it's
5265possible that a bug in the debug info, a bug in the remote stub (for
5266remote targets), or even a bug in @value{GDBN} could make line
5267stepping behave incorrectly when target-assisted range stepping is
5268enabled. You can use the following command to turn off range stepping
5269if necessary:
5270
5271@table @code
5272@kindex set range-stepping
5273@kindex show range-stepping
5274@item set range-stepping
5275@itemx show range-stepping
5276Control whether range stepping is enabled.
5277
5278If @code{on}, and the target supports it, @value{GDBN} tells the
5279target to step a range of addresses itself, instead of issuing
5280multiple single-steps. If @code{off}, @value{GDBN} always issues
5281single-steps, even if range stepping is supported by the target. The
5282default is @code{on}.
5283
5284@end table
5285
5286@node Skipping Over Functions and Files
5287@section Skipping Over Functions and Files
5288@cindex skipping over functions and files
5289
5290The program you are debugging may contain some functions which are
5291uninteresting to debug. The @code{skip} comand lets you tell @value{GDBN} to
5292skip a function or all functions in a file when stepping.
5293
5294For example, consider the following C function:
5295
5296@smallexample
5297101 int func()
5298102 @{
5299103 foo(boring());
5300104 bar(boring());
5301105 @}
5302@end smallexample
5303
5304@noindent
5305Suppose you wish to step into the functions @code{foo} and @code{bar}, but you
5306are not interested in stepping through @code{boring}. If you run @code{step}
5307at line 103, you'll enter @code{boring()}, but if you run @code{next}, you'll
5308step over both @code{foo} and @code{boring}!
5309
5310One solution is to @code{step} into @code{boring} and use the @code{finish}
5311command to immediately exit it. But this can become tedious if @code{boring}
5312is called from many places.
5313
5314A more flexible solution is to execute @kbd{skip boring}. This instructs
5315@value{GDBN} never to step into @code{boring}. Now when you execute
5316@code{step} at line 103, you'll step over @code{boring} and directly into
5317@code{foo}.
5318
5319You can also instruct @value{GDBN} to skip all functions in a file, with, for
5320example, @code{skip file boring.c}.
5321
5322@table @code
5323@kindex skip function
5324@item skip @r{[}@var{linespec}@r{]}
5325@itemx skip function @r{[}@var{linespec}@r{]}
5326After running this command, the function named by @var{linespec} or the
5327function containing the line named by @var{linespec} will be skipped over when
5328stepping. @xref{Specify Location}.
5329
5330If you do not specify @var{linespec}, the function you're currently debugging
5331will be skipped.
5332
5333(If you have a function called @code{file} that you want to skip, use
5334@kbd{skip function file}.)
5335
5336@kindex skip file
5337@item skip file @r{[}@var{filename}@r{]}
5338After running this command, any function whose source lives in @var{filename}
5339will be skipped over when stepping.
5340
5341If you do not specify @var{filename}, functions whose source lives in the file
5342you're currently debugging will be skipped.
5343@end table
5344
5345Skips can be listed, deleted, disabled, and enabled, much like breakpoints.
5346These are the commands for managing your list of skips:
5347
5348@table @code
5349@kindex info skip
5350@item info skip @r{[}@var{range}@r{]}
5351Print details about the specified skip(s). If @var{range} is not specified,
5352print a table with details about all functions and files marked for skipping.
5353@code{info skip} prints the following information about each skip:
5354
5355@table @emph
5356@item Identifier
5357A number identifying this skip.
5358@item Type
5359The type of this skip, either @samp{function} or @samp{file}.
5360@item Enabled or Disabled
5361Enabled skips are marked with @samp{y}. Disabled skips are marked with @samp{n}.
5362@item Address
5363For function skips, this column indicates the address in memory of the function
5364being skipped. If you've set a function skip on a function which has not yet
5365been loaded, this field will contain @samp{<PENDING>}. Once a shared library
5366which has the function is loaded, @code{info skip} will show the function's
5367address here.
5368@item What
5369For file skips, this field contains the filename being skipped. For functions
5370skips, this field contains the function name and its line number in the file
5371where it is defined.
5372@end table
5373
5374@kindex skip delete
5375@item skip delete @r{[}@var{range}@r{]}
5376Delete the specified skip(s). If @var{range} is not specified, delete all
5377skips.
5378
5379@kindex skip enable
5380@item skip enable @r{[}@var{range}@r{]}
5381Enable the specified skip(s). If @var{range} is not specified, enable all
5382skips.
5383
5384@kindex skip disable
5385@item skip disable @r{[}@var{range}@r{]}
5386Disable the specified skip(s). If @var{range} is not specified, disable all
5387skips.
5388
5389@end table
5390
5391@node Signals
5392@section Signals
5393@cindex signals
5394
5395A signal is an asynchronous event that can happen in a program. The
5396operating system defines the possible kinds of signals, and gives each
5397kind a name and a number. For example, in Unix @code{SIGINT} is the
5398signal a program gets when you type an interrupt character (often @kbd{Ctrl-c});
5399@code{SIGSEGV} is the signal a program gets from referencing a place in
5400memory far away from all the areas in use; @code{SIGALRM} occurs when
5401the alarm clock timer goes off (which happens only if your program has
5402requested an alarm).
5403
5404@cindex fatal signals
5405Some signals, including @code{SIGALRM}, are a normal part of the
5406functioning of your program. Others, such as @code{SIGSEGV}, indicate
5407errors; these signals are @dfn{fatal} (they kill your program immediately) if the
5408program has not specified in advance some other way to handle the signal.
5409@code{SIGINT} does not indicate an error in your program, but it is normally
5410fatal so it can carry out the purpose of the interrupt: to kill the program.
5411
5412@value{GDBN} has the ability to detect any occurrence of a signal in your
5413program. You can tell @value{GDBN} in advance what to do for each kind of
5414signal.
5415
5416@cindex handling signals
5417Normally, @value{GDBN} is set up to let the non-erroneous signals like
5418@code{SIGALRM} be silently passed to your program
5419(so as not to interfere with their role in the program's functioning)
5420but to stop your program immediately whenever an error signal happens.
5421You can change these settings with the @code{handle} command.
5422
5423@table @code
5424@kindex info signals
5425@kindex info handle
5426@item info signals
5427@itemx info handle
5428Print a table of all the kinds of signals and how @value{GDBN} has been told to
5429handle each one. You can use this to see the signal numbers of all
5430the defined types of signals.
5431
5432@item info signals @var{sig}
5433Similar, but print information only about the specified signal number.
5434
5435@code{info handle} is an alias for @code{info signals}.
5436
5437@item catch signal @r{[}@var{signal}@dots{} @r{|} @samp{all}@r{]}
5438Set a catchpoint for the indicated signals. @xref{Set Catchpoints},
5439for details about this command.
5440
5441@kindex handle
5442@item handle @var{signal} @r{[}@var{keywords}@dots{}@r{]}
5443Change the way @value{GDBN} handles signal @var{signal}. @var{signal}
5444can be the number of a signal or its name (with or without the
5445@samp{SIG} at the beginning); a list of signal numbers of the form
5446@samp{@var{low}-@var{high}}; or the word @samp{all}, meaning all the
5447known signals. Optional arguments @var{keywords}, described below,
5448say what change to make.
5449@end table
5450
5451@c @group
5452The keywords allowed by the @code{handle} command can be abbreviated.
5453Their full names are:
5454
5455@table @code
5456@item nostop
5457@value{GDBN} should not stop your program when this signal happens. It may
5458still print a message telling you that the signal has come in.
5459
5460@item stop
5461@value{GDBN} should stop your program when this signal happens. This implies
5462the @code{print} keyword as well.
5463
5464@item print
5465@value{GDBN} should print a message when this signal happens.
5466
5467@item noprint
5468@value{GDBN} should not mention the occurrence of the signal at all. This
5469implies the @code{nostop} keyword as well.
5470
5471@item pass
5472@itemx noignore
5473@value{GDBN} should allow your program to see this signal; your program
5474can handle the signal, or else it may terminate if the signal is fatal
5475and not handled. @code{pass} and @code{noignore} are synonyms.
5476
5477@item nopass
5478@itemx ignore
5479@value{GDBN} should not allow your program to see this signal.
5480@code{nopass} and @code{ignore} are synonyms.
5481@end table
5482@c @end group
5483
5484When a signal stops your program, the signal is not visible to the
5485program until you
5486continue. Your program sees the signal then, if @code{pass} is in
5487effect for the signal in question @emph{at that time}. In other words,
5488after @value{GDBN} reports a signal, you can use the @code{handle}
5489command with @code{pass} or @code{nopass} to control whether your
5490program sees that signal when you continue.
5491
5492The default is set to @code{nostop}, @code{noprint}, @code{pass} for
5493non-erroneous signals such as @code{SIGALRM}, @code{SIGWINCH} and
5494@code{SIGCHLD}, and to @code{stop}, @code{print}, @code{pass} for the
5495erroneous signals.
5496
5497You can also use the @code{signal} command to prevent your program from
5498seeing a signal, or cause it to see a signal it normally would not see,
5499or to give it any signal at any time. For example, if your program stopped
5500due to some sort of memory reference error, you might store correct
5501values into the erroneous variables and continue, hoping to see more
5502execution; but your program would probably terminate immediately as
5503a result of the fatal signal once it saw the signal. To prevent this,
5504you can continue with @samp{signal 0}. @xref{Signaling, ,Giving your
5505Program a Signal}.
5506
5507@cindex extra signal information
5508@anchor{extra signal information}
5509
5510On some targets, @value{GDBN} can inspect extra signal information
5511associated with the intercepted signal, before it is actually
5512delivered to the program being debugged. This information is exported
5513by the convenience variable @code{$_siginfo}, and consists of data
5514that is passed by the kernel to the signal handler at the time of the
5515receipt of a signal. The data type of the information itself is
5516target dependent. You can see the data type using the @code{ptype
5517$_siginfo} command. On Unix systems, it typically corresponds to the
5518standard @code{siginfo_t} type, as defined in the @file{signal.h}
5519system header.
5520
5521Here's an example, on a @sc{gnu}/Linux system, printing the stray
5522referenced address that raised a segmentation fault.
5523
5524@smallexample
5525@group
5526(@value{GDBP}) continue
5527Program received signal SIGSEGV, Segmentation fault.
55280x0000000000400766 in main ()
552969 *(int *)p = 0;
5530(@value{GDBP}) ptype $_siginfo
5531type = struct @{
5532 int si_signo;
5533 int si_errno;
5534 int si_code;
5535 union @{
5536 int _pad[28];
5537 struct @{...@} _kill;
5538 struct @{...@} _timer;
5539 struct @{...@} _rt;
5540 struct @{...@} _sigchld;
5541 struct @{...@} _sigfault;
5542 struct @{...@} _sigpoll;
5543 @} _sifields;
5544@}
5545(@value{GDBP}) ptype $_siginfo._sifields._sigfault
5546type = struct @{
5547 void *si_addr;
5548@}
5549(@value{GDBP}) p $_siginfo._sifields._sigfault.si_addr
5550$1 = (void *) 0x7ffff7ff7000
5551@end group
5552@end smallexample
5553
5554Depending on target support, @code{$_siginfo} may also be writable.
5555
5556@node Thread Stops
5557@section Stopping and Starting Multi-thread Programs
5558
5559@cindex stopped threads
5560@cindex threads, stopped
5561
5562@cindex continuing threads
5563@cindex threads, continuing
5564
5565@value{GDBN} supports debugging programs with multiple threads
5566(@pxref{Threads,, Debugging Programs with Multiple Threads}). There
5567are two modes of controlling execution of your program within the
5568debugger. In the default mode, referred to as @dfn{all-stop mode},
5569when any thread in your program stops (for example, at a breakpoint
5570or while being stepped), all other threads in the program are also stopped by
5571@value{GDBN}. On some targets, @value{GDBN} also supports
5572@dfn{non-stop mode}, in which other threads can continue to run freely while
5573you examine the stopped thread in the debugger.
5574
5575@menu
5576* All-Stop Mode:: All threads stop when GDB takes control
5577* Non-Stop Mode:: Other threads continue to execute
5578* Background Execution:: Running your program asynchronously
5579* Thread-Specific Breakpoints:: Controlling breakpoints
5580* Interrupted System Calls:: GDB may interfere with system calls
5581* Observer Mode:: GDB does not alter program behavior
5582@end menu
5583
5584@node All-Stop Mode
5585@subsection All-Stop Mode
5586
5587@cindex all-stop mode
5588
5589In all-stop mode, whenever your program stops under @value{GDBN} for any reason,
5590@emph{all} threads of execution stop, not just the current thread. This
5591allows you to examine the overall state of the program, including
5592switching between threads, without worrying that things may change
5593underfoot.
5594
5595Conversely, whenever you restart the program, @emph{all} threads start
5596executing. @emph{This is true even when single-stepping} with commands
5597like @code{step} or @code{next}.
5598
5599In particular, @value{GDBN} cannot single-step all threads in lockstep.
5600Since thread scheduling is up to your debugging target's operating
5601system (not controlled by @value{GDBN}), other threads may
5602execute more than one statement while the current thread completes a
5603single step. Moreover, in general other threads stop in the middle of a
5604statement, rather than at a clean statement boundary, when the program
5605stops.
5606
5607You might even find your program stopped in another thread after
5608continuing or even single-stepping. This happens whenever some other
5609thread runs into a breakpoint, a signal, or an exception before the
5610first thread completes whatever you requested.
5611
5612@cindex automatic thread selection
5613@cindex switching threads automatically
5614@cindex threads, automatic switching
5615Whenever @value{GDBN} stops your program, due to a breakpoint or a
5616signal, it automatically selects the thread where that breakpoint or
5617signal happened. @value{GDBN} alerts you to the context switch with a
5618message such as @samp{[Switching to Thread @var{n}]} to identify the
5619thread.
5620
5621On some OSes, you can modify @value{GDBN}'s default behavior by
5622locking the OS scheduler to allow only a single thread to run.
5623
5624@table @code
5625@item set scheduler-locking @var{mode}
5626@cindex scheduler locking mode
5627@cindex lock scheduler
5628Set the scheduler locking mode. If it is @code{off}, then there is no
5629locking and any thread may run at any time. If @code{on}, then only the
5630current thread may run when the inferior is resumed. The @code{step}
5631mode optimizes for single-stepping; it prevents other threads
5632from preempting the current thread while you are stepping, so that
5633the focus of debugging does not change unexpectedly.
5634Other threads only rarely (or never) get a chance to run
5635when you step. They are more likely to run when you @samp{next} over a
5636function call, and they are completely free to run when you use commands
5637like @samp{continue}, @samp{until}, or @samp{finish}. However, unless another
5638thread hits a breakpoint during its timeslice, @value{GDBN} does not change
5639the current thread away from the thread that you are debugging.
5640
5641@item show scheduler-locking
5642Display the current scheduler locking mode.
5643@end table
5644
5645@cindex resume threads of multiple processes simultaneously
5646By default, when you issue one of the execution commands such as
5647@code{continue}, @code{next} or @code{step}, @value{GDBN} allows only
5648threads of the current inferior to run. For example, if @value{GDBN}
5649is attached to two inferiors, each with two threads, the
5650@code{continue} command resumes only the two threads of the current
5651inferior. This is useful, for example, when you debug a program that
5652forks and you want to hold the parent stopped (so that, for instance,
5653it doesn't run to exit), while you debug the child. In other
5654situations, you may not be interested in inspecting the current state
5655of any of the processes @value{GDBN} is attached to, and you may want
5656to resume them all until some breakpoint is hit. In the latter case,
5657you can instruct @value{GDBN} to allow all threads of all the
5658inferiors to run with the @w{@code{set schedule-multiple}} command.
5659
5660@table @code
5661@kindex set schedule-multiple
5662@item set schedule-multiple
5663Set the mode for allowing threads of multiple processes to be resumed
5664when an execution command is issued. When @code{on}, all threads of
5665all processes are allowed to run. When @code{off}, only the threads
5666of the current process are resumed. The default is @code{off}. The
5667@code{scheduler-locking} mode takes precedence when set to @code{on},
5668or while you are stepping and set to @code{step}.
5669
5670@item show schedule-multiple
5671Display the current mode for resuming the execution of threads of
5672multiple processes.
5673@end table
5674
5675@node Non-Stop Mode
5676@subsection Non-Stop Mode
5677
5678@cindex non-stop mode
5679
5680@c This section is really only a place-holder, and needs to be expanded
5681@c with more details.
5682
5683For some multi-threaded targets, @value{GDBN} supports an optional
5684mode of operation in which you can examine stopped program threads in
5685the debugger while other threads continue to execute freely. This
5686minimizes intrusion when debugging live systems, such as programs
5687where some threads have real-time constraints or must continue to
5688respond to external events. This is referred to as @dfn{non-stop} mode.
5689
5690In non-stop mode, when a thread stops to report a debugging event,
5691@emph{only} that thread is stopped; @value{GDBN} does not stop other
5692threads as well, in contrast to the all-stop mode behavior. Additionally,
5693execution commands such as @code{continue} and @code{step} apply by default
5694only to the current thread in non-stop mode, rather than all threads as
5695in all-stop mode. This allows you to control threads explicitly in
5696ways that are not possible in all-stop mode --- for example, stepping
5697one thread while allowing others to run freely, stepping
5698one thread while holding all others stopped, or stepping several threads
5699independently and simultaneously.
5700
5701To enter non-stop mode, use this sequence of commands before you run
5702or attach to your program:
5703
5704@smallexample
5705# Enable the async interface.
5706set target-async 1
5707
5708# If using the CLI, pagination breaks non-stop.
5709set pagination off
5710
5711# Finally, turn it on!
5712set non-stop on
5713@end smallexample
5714
5715You can use these commands to manipulate the non-stop mode setting:
5716
5717@table @code
5718@kindex set non-stop
5719@item set non-stop on
5720Enable selection of non-stop mode.
5721@item set non-stop off
5722Disable selection of non-stop mode.
5723@kindex show non-stop
5724@item show non-stop
5725Show the current non-stop enablement setting.
5726@end table
5727
5728Note these commands only reflect whether non-stop mode is enabled,
5729not whether the currently-executing program is being run in non-stop mode.
5730In particular, the @code{set non-stop} preference is only consulted when
5731@value{GDBN} starts or connects to the target program, and it is generally
5732not possible to switch modes once debugging has started. Furthermore,
5733since not all targets support non-stop mode, even when you have enabled
5734non-stop mode, @value{GDBN} may still fall back to all-stop operation by
5735default.
5736
5737In non-stop mode, all execution commands apply only to the current thread
5738by default. That is, @code{continue} only continues one thread.
5739To continue all threads, issue @code{continue -a} or @code{c -a}.
5740
5741You can use @value{GDBN}'s background execution commands
5742(@pxref{Background Execution}) to run some threads in the background
5743while you continue to examine or step others from @value{GDBN}.
5744The MI execution commands (@pxref{GDB/MI Program Execution}) are
5745always executed asynchronously in non-stop mode.
5746
5747Suspending execution is done with the @code{interrupt} command when
5748running in the background, or @kbd{Ctrl-c} during foreground execution.
5749In all-stop mode, this stops the whole process;
5750but in non-stop mode the interrupt applies only to the current thread.
5751To stop the whole program, use @code{interrupt -a}.
5752
5753Other execution commands do not currently support the @code{-a} option.
5754
5755In non-stop mode, when a thread stops, @value{GDBN} doesn't automatically make
5756that thread current, as it does in all-stop mode. This is because the
5757thread stop notifications are asynchronous with respect to @value{GDBN}'s
5758command interpreter, and it would be confusing if @value{GDBN} unexpectedly
5759changed to a different thread just as you entered a command to operate on the
5760previously current thread.
5761
5762@node Background Execution
5763@subsection Background Execution
5764
5765@cindex foreground execution
5766@cindex background execution
5767@cindex asynchronous execution
5768@cindex execution, foreground, background and asynchronous
5769
5770@value{GDBN}'s execution commands have two variants: the normal
5771foreground (synchronous) behavior, and a background
5772(asynchronous) behavior. In foreground execution, @value{GDBN} waits for
5773the program to report that some thread has stopped before prompting for
5774another command. In background execution, @value{GDBN} immediately gives
5775a command prompt so that you can issue other commands while your program runs.
5776
5777You need to explicitly enable asynchronous mode before you can use
5778background execution commands. You can use these commands to
5779manipulate the asynchronous mode setting:
5780
5781@table @code
5782@kindex set target-async
5783@item set target-async on
5784Enable asynchronous mode.
5785@item set target-async off
5786Disable asynchronous mode.
5787@kindex show target-async
5788@item show target-async
5789Show the current target-async setting.
5790@end table
5791
5792If the target doesn't support async mode, @value{GDBN} issues an error
5793message if you attempt to use the background execution commands.
5794
5795To specify background execution, add a @code{&} to the command. For example,
5796the background form of the @code{continue} command is @code{continue&}, or
5797just @code{c&}. The execution commands that accept background execution
5798are:
5799
5800@table @code
5801@kindex run&
5802@item run
5803@xref{Starting, , Starting your Program}.
5804
5805@item attach
5806@kindex attach&
5807@xref{Attach, , Debugging an Already-running Process}.
5808
5809@item step
5810@kindex step&
5811@xref{Continuing and Stepping, step}.
5812
5813@item stepi
5814@kindex stepi&
5815@xref{Continuing and Stepping, stepi}.
5816
5817@item next
5818@kindex next&
5819@xref{Continuing and Stepping, next}.
5820
5821@item nexti
5822@kindex nexti&
5823@xref{Continuing and Stepping, nexti}.
5824
5825@item continue
5826@kindex continue&
5827@xref{Continuing and Stepping, continue}.
5828
5829@item finish
5830@kindex finish&
5831@xref{Continuing and Stepping, finish}.
5832
5833@item until
5834@kindex until&
5835@xref{Continuing and Stepping, until}.
5836
5837@end table
5838
5839Background execution is especially useful in conjunction with non-stop
5840mode for debugging programs with multiple threads; see @ref{Non-Stop Mode}.
5841However, you can also use these commands in the normal all-stop mode with
5842the restriction that you cannot issue another execution command until the
5843previous one finishes. Examples of commands that are valid in all-stop
5844mode while the program is running include @code{help} and @code{info break}.
5845
5846You can interrupt your program while it is running in the background by
5847using the @code{interrupt} command.
5848
5849@table @code
5850@kindex interrupt
5851@item interrupt
5852@itemx interrupt -a
5853
5854Suspend execution of the running program. In all-stop mode,
5855@code{interrupt} stops the whole process, but in non-stop mode, it stops
5856only the current thread. To stop the whole program in non-stop mode,
5857use @code{interrupt -a}.
5858@end table
5859
5860@node Thread-Specific Breakpoints
5861@subsection Thread-Specific Breakpoints
5862
5863When your program has multiple threads (@pxref{Threads,, Debugging
5864Programs with Multiple Threads}), you can choose whether to set
5865breakpoints on all threads, or on a particular thread.
5866
5867@table @code
5868@cindex breakpoints and threads
5869@cindex thread breakpoints
5870@kindex break @dots{} thread @var{threadno}
5871@item break @var{linespec} thread @var{threadno}
5872@itemx break @var{linespec} thread @var{threadno} if @dots{}
5873@var{linespec} specifies source lines; there are several ways of
5874writing them (@pxref{Specify Location}), but the effect is always to
5875specify some source line.
5876
5877Use the qualifier @samp{thread @var{threadno}} with a breakpoint command
5878to specify that you only want @value{GDBN} to stop the program when a
5879particular thread reaches this breakpoint. @var{threadno} is one of the
5880numeric thread identifiers assigned by @value{GDBN}, shown in the first
5881column of the @samp{info threads} display.
5882
5883If you do not specify @samp{thread @var{threadno}} when you set a
5884breakpoint, the breakpoint applies to @emph{all} threads of your
5885program.
5886
5887You can use the @code{thread} qualifier on conditional breakpoints as
5888well; in this case, place @samp{thread @var{threadno}} before or
5889after the breakpoint condition, like this:
5890
5891@smallexample
5892(@value{GDBP}) break frik.c:13 thread 28 if bartab > lim
5893@end smallexample
5894
5895@end table
5896
5897Thread-specific breakpoints are automatically deleted when
5898@value{GDBN} detects the corresponding thread is no longer in the
5899thread list. For example:
5900
5901@smallexample
5902(@value{GDBP}) c
5903Thread-specific breakpoint 3 deleted - thread 28 no longer in the thread list.
5904@end smallexample
5905
5906There are several ways for a thread to disappear, such as a regular
5907thread exit, but also when you detach from the process with the
5908@code{detach} command (@pxref{Attach, ,Debugging an Already-running
5909Process}), or if @value{GDBN} loses the remote connection
5910(@pxref{Remote Debugging}), etc. Note that with some targets,
5911@value{GDBN} is only able to detect a thread has exited when the user
5912explictly asks for the thread list with the @code{info threads}
5913command.
5914
5915@node Interrupted System Calls
5916@subsection Interrupted System Calls
5917
5918@cindex thread breakpoints and system calls
5919@cindex system calls and thread breakpoints
5920@cindex premature return from system calls
5921There is an unfortunate side effect when using @value{GDBN} to debug
5922multi-threaded programs. If one thread stops for a
5923breakpoint, or for some other reason, and another thread is blocked in a
5924system call, then the system call may return prematurely. This is a
5925consequence of the interaction between multiple threads and the signals
5926that @value{GDBN} uses to implement breakpoints and other events that
5927stop execution.
5928
5929To handle this problem, your program should check the return value of
5930each system call and react appropriately. This is good programming
5931style anyways.
5932
5933For example, do not write code like this:
5934
5935@smallexample
5936 sleep (10);
5937@end smallexample
5938
5939The call to @code{sleep} will return early if a different thread stops
5940at a breakpoint or for some other reason.
5941
5942Instead, write this:
5943
5944@smallexample
5945 int unslept = 10;
5946 while (unslept > 0)
5947 unslept = sleep (unslept);
5948@end smallexample
5949
5950A system call is allowed to return early, so the system is still
5951conforming to its specification. But @value{GDBN} does cause your
5952multi-threaded program to behave differently than it would without
5953@value{GDBN}.
5954
5955Also, @value{GDBN} uses internal breakpoints in the thread library to
5956monitor certain events such as thread creation and thread destruction.
5957When such an event happens, a system call in another thread may return
5958prematurely, even though your program does not appear to stop.
5959
5960@node Observer Mode
5961@subsection Observer Mode
5962
5963If you want to build on non-stop mode and observe program behavior
5964without any chance of disruption by @value{GDBN}, you can set
5965variables to disable all of the debugger's attempts to modify state,
5966whether by writing memory, inserting breakpoints, etc. These operate
5967at a low level, intercepting operations from all commands.
5968
5969When all of these are set to @code{off}, then @value{GDBN} is said to
5970be @dfn{observer mode}. As a convenience, the variable
5971@code{observer} can be set to disable these, plus enable non-stop
5972mode.
5973
5974Note that @value{GDBN} will not prevent you from making nonsensical
5975combinations of these settings. For instance, if you have enabled
5976@code{may-insert-breakpoints} but disabled @code{may-write-memory},
5977then breakpoints that work by writing trap instructions into the code
5978stream will still not be able to be placed.
5979
5980@table @code
5981
5982@kindex observer
5983@item set observer on
5984@itemx set observer off
5985When set to @code{on}, this disables all the permission variables
5986below (except for @code{insert-fast-tracepoints}), plus enables
5987non-stop debugging. Setting this to @code{off} switches back to
5988normal debugging, though remaining in non-stop mode.
5989
5990@item show observer
5991Show whether observer mode is on or off.
5992
5993@kindex may-write-registers
5994@item set may-write-registers on
5995@itemx set may-write-registers off
5996This controls whether @value{GDBN} will attempt to alter the values of
5997registers, such as with assignment expressions in @code{print}, or the
5998@code{jump} command. It defaults to @code{on}.
5999
6000@item show may-write-registers
6001Show the current permission to write registers.
6002
6003@kindex may-write-memory
6004@item set may-write-memory on
6005@itemx set may-write-memory off
6006This controls whether @value{GDBN} will attempt to alter the contents
6007of memory, such as with assignment expressions in @code{print}. It
6008defaults to @code{on}.
6009
6010@item show may-write-memory
6011Show the current permission to write memory.
6012
6013@kindex may-insert-breakpoints
6014@item set may-insert-breakpoints on
6015@itemx set may-insert-breakpoints off
6016This controls whether @value{GDBN} will attempt to insert breakpoints.
6017This affects all breakpoints, including internal breakpoints defined
6018by @value{GDBN}. It defaults to @code{on}.
6019
6020@item show may-insert-breakpoints
6021Show the current permission to insert breakpoints.
6022
6023@kindex may-insert-tracepoints
6024@item set may-insert-tracepoints on
6025@itemx set may-insert-tracepoints off
6026This controls whether @value{GDBN} will attempt to insert (regular)
6027tracepoints at the beginning of a tracing experiment. It affects only
6028non-fast tracepoints, fast tracepoints being under the control of
6029@code{may-insert-fast-tracepoints}. It defaults to @code{on}.
6030
6031@item show may-insert-tracepoints
6032Show the current permission to insert tracepoints.
6033
6034@kindex may-insert-fast-tracepoints
6035@item set may-insert-fast-tracepoints on
6036@itemx set may-insert-fast-tracepoints off
6037This controls whether @value{GDBN} will attempt to insert fast
6038tracepoints at the beginning of a tracing experiment. It affects only
6039fast tracepoints, regular (non-fast) tracepoints being under the
6040control of @code{may-insert-tracepoints}. It defaults to @code{on}.
6041
6042@item show may-insert-fast-tracepoints
6043Show the current permission to insert fast tracepoints.
6044
6045@kindex may-interrupt
6046@item set may-interrupt on
6047@itemx set may-interrupt off
6048This controls whether @value{GDBN} will attempt to interrupt or stop
6049program execution. When this variable is @code{off}, the
6050@code{interrupt} command will have no effect, nor will
6051@kbd{Ctrl-c}. It defaults to @code{on}.
6052
6053@item show may-interrupt
6054Show the current permission to interrupt or stop the program.
6055
6056@end table
6057
6058@node Reverse Execution
6059@chapter Running programs backward
6060@cindex reverse execution
6061@cindex running programs backward
6062
6063When you are debugging a program, it is not unusual to realize that
6064you have gone too far, and some event of interest has already happened.
6065If the target environment supports it, @value{GDBN} can allow you to
6066``rewind'' the program by running it backward.
6067
6068A target environment that supports reverse execution should be able
6069to ``undo'' the changes in machine state that have taken place as the
6070program was executing normally. Variables, registers etc.@: should
6071revert to their previous values. Obviously this requires a great
6072deal of sophistication on the part of the target environment; not
6073all target environments can support reverse execution.
6074
6075When a program is executed in reverse, the instructions that
6076have most recently been executed are ``un-executed'', in reverse
6077order. The program counter runs backward, following the previous
6078thread of execution in reverse. As each instruction is ``un-executed'',
6079the values of memory and/or registers that were changed by that
6080instruction are reverted to their previous states. After executing
6081a piece of source code in reverse, all side effects of that code
6082should be ``undone'', and all variables should be returned to their
6083prior values@footnote{
6084Note that some side effects are easier to undo than others. For instance,
6085memory and registers are relatively easy, but device I/O is hard. Some
6086targets may be able undo things like device I/O, and some may not.
6087
6088The contract between @value{GDBN} and the reverse executing target
6089requires only that the target do something reasonable when
6090@value{GDBN} tells it to execute backwards, and then report the
6091results back to @value{GDBN}. Whatever the target reports back to
6092@value{GDBN}, @value{GDBN} will report back to the user. @value{GDBN}
6093assumes that the memory and registers that the target reports are in a
6094consistant state, but @value{GDBN} accepts whatever it is given.
6095}.
6096
6097If you are debugging in a target environment that supports
6098reverse execution, @value{GDBN} provides the following commands.
6099
6100@table @code
6101@kindex reverse-continue
6102@kindex rc @r{(@code{reverse-continue})}
6103@item reverse-continue @r{[}@var{ignore-count}@r{]}
6104@itemx rc @r{[}@var{ignore-count}@r{]}
6105Beginning at the point where your program last stopped, start executing
6106in reverse. Reverse execution will stop for breakpoints and synchronous
6107exceptions (signals), just like normal execution. Behavior of
6108asynchronous signals depends on the target environment.
6109
6110@kindex reverse-step
6111@kindex rs @r{(@code{step})}
6112@item reverse-step @r{[}@var{count}@r{]}
6113Run the program backward until control reaches the start of a
6114different source line; then stop it, and return control to @value{GDBN}.
6115
6116Like the @code{step} command, @code{reverse-step} will only stop
6117at the beginning of a source line. It ``un-executes'' the previously
6118executed source line. If the previous source line included calls to
6119debuggable functions, @code{reverse-step} will step (backward) into
6120the called function, stopping at the beginning of the @emph{last}
6121statement in the called function (typically a return statement).
6122
6123Also, as with the @code{step} command, if non-debuggable functions are
6124called, @code{reverse-step} will run thru them backward without stopping.
6125
6126@kindex reverse-stepi
6127@kindex rsi @r{(@code{reverse-stepi})}
6128@item reverse-stepi @r{[}@var{count}@r{]}
6129Reverse-execute one machine instruction. Note that the instruction
6130to be reverse-executed is @emph{not} the one pointed to by the program
6131counter, but the instruction executed prior to that one. For instance,
6132if the last instruction was a jump, @code{reverse-stepi} will take you
6133back from the destination of the jump to the jump instruction itself.
6134
6135@kindex reverse-next
6136@kindex rn @r{(@code{reverse-next})}
6137@item reverse-next @r{[}@var{count}@r{]}
6138Run backward to the beginning of the previous line executed in
6139the current (innermost) stack frame. If the line contains function
6140calls, they will be ``un-executed'' without stopping. Starting from
6141the first line of a function, @code{reverse-next} will take you back
6142to the caller of that function, @emph{before} the function was called,
6143just as the normal @code{next} command would take you from the last
6144line of a function back to its return to its caller
6145@footnote{Unless the code is too heavily optimized.}.
6146
6147@kindex reverse-nexti
6148@kindex rni @r{(@code{reverse-nexti})}
6149@item reverse-nexti @r{[}@var{count}@r{]}
6150Like @code{nexti}, @code{reverse-nexti} executes a single instruction
6151in reverse, except that called functions are ``un-executed'' atomically.
6152That is, if the previously executed instruction was a return from
6153another function, @code{reverse-nexti} will continue to execute
6154in reverse until the call to that function (from the current stack
6155frame) is reached.
6156
6157@kindex reverse-finish
6158@item reverse-finish
6159Just as the @code{finish} command takes you to the point where the
6160current function returns, @code{reverse-finish} takes you to the point
6161where it was called. Instead of ending up at the end of the current
6162function invocation, you end up at the beginning.
6163
6164@kindex set exec-direction
6165@item set exec-direction
6166Set the direction of target execution.
6167@item set exec-direction reverse
6168@cindex execute forward or backward in time
6169@value{GDBN} will perform all execution commands in reverse, until the
6170exec-direction mode is changed to ``forward''. Affected commands include
6171@code{step, stepi, next, nexti, continue, and finish}. The @code{return}
6172command cannot be used in reverse mode.
6173@item set exec-direction forward
6174@value{GDBN} will perform all execution commands in the normal fashion.
6175This is the default.
6176@end table
6177
6178
6179@node Process Record and Replay
6180@chapter Recording Inferior's Execution and Replaying It
6181@cindex process record and replay
6182@cindex recording inferior's execution and replaying it
6183
6184On some platforms, @value{GDBN} provides a special @dfn{process record
6185and replay} target that can record a log of the process execution, and
6186replay it later with both forward and reverse execution commands.
6187
6188@cindex replay mode
6189When this target is in use, if the execution log includes the record
6190for the next instruction, @value{GDBN} will debug in @dfn{replay
6191mode}. In the replay mode, the inferior does not really execute code
6192instructions. Instead, all the events that normally happen during
6193code execution are taken from the execution log. While code is not
6194really executed in replay mode, the values of registers (including the
6195program counter register) and the memory of the inferior are still
6196changed as they normally would. Their contents are taken from the
6197execution log.
6198
6199@cindex record mode
6200If the record for the next instruction is not in the execution log,
6201@value{GDBN} will debug in @dfn{record mode}. In this mode, the
6202inferior executes normally, and @value{GDBN} records the execution log
6203for future replay.
6204
6205The process record and replay target supports reverse execution
6206(@pxref{Reverse Execution}), even if the platform on which the
6207inferior runs does not. However, the reverse execution is limited in
6208this case by the range of the instructions recorded in the execution
6209log. In other words, reverse execution on platforms that don't
6210support it directly can only be done in the replay mode.
6211
6212When debugging in the reverse direction, @value{GDBN} will work in
6213replay mode as long as the execution log includes the record for the
6214previous instruction; otherwise, it will work in record mode, if the
6215platform supports reverse execution, or stop if not.
6216
6217For architecture environments that support process record and replay,
6218@value{GDBN} provides the following commands:
6219
6220@table @code
6221@kindex target record
6222@kindex target record-full
6223@kindex target record-btrace
6224@kindex record
6225@kindex record full
6226@kindex record btrace
6227@kindex rec
6228@kindex rec full
6229@kindex rec btrace
6230@item record @var{method}
6231This command starts the process record and replay target. The
6232recording method can be specified as parameter. Without a parameter
6233the command uses the @code{full} recording method. The following
6234recording methods are available:
6235
6236@table @code
6237@item full
6238Full record/replay recording using @value{GDBN}'s software record and
6239replay implementation. This method allows replaying and reverse
6240execution.
6241
6242@item btrace
6243Hardware-supported instruction recording. This method does not allow
6244replaying and reverse execution.
6245
6246This recording method may not be available on all processors.
6247@end table
6248
6249The process record and replay target can only debug a process that is
6250already running. Therefore, you need first to start the process with
6251the @kbd{run} or @kbd{start} commands, and then start the recording
6252with the @kbd{record @var{method}} command.
6253
6254Both @code{record @var{method}} and @code{rec @var{method}} are
6255aliases of @code{target record-@var{method}}.
6256
6257@cindex displaced stepping, and process record and replay
6258Displaced stepping (@pxref{Maintenance Commands,, displaced stepping})
6259will be automatically disabled when process record and replay target
6260is started. That's because the process record and replay target
6261doesn't support displaced stepping.
6262
6263@cindex non-stop mode, and process record and replay
6264@cindex asynchronous execution, and process record and replay
6265If the inferior is in the non-stop mode (@pxref{Non-Stop Mode}) or in
6266the asynchronous execution mode (@pxref{Background Execution}), not
6267all recording methods are available. The @code{full} recording method
6268does not support these two modes.
6269
6270@kindex record stop
6271@kindex rec s
6272@item record stop
6273Stop the process record and replay target. When process record and
6274replay target stops, the entire execution log will be deleted and the
6275inferior will either be terminated, or will remain in its final state.
6276
6277When you stop the process record and replay target in record mode (at
6278the end of the execution log), the inferior will be stopped at the
6279next instruction that would have been recorded. In other words, if
6280you record for a while and then stop recording, the inferior process
6281will be left in the same state as if the recording never happened.
6282
6283On the other hand, if the process record and replay target is stopped
6284while in replay mode (that is, not at the end of the execution log,
6285but at some earlier point), the inferior process will become ``live''
6286at that earlier state, and it will then be possible to continue the
6287usual ``live'' debugging of the process from that state.
6288
6289When the inferior process exits, or @value{GDBN} detaches from it,
6290process record and replay target will automatically stop itself.
6291
6292@kindex record goto
6293@item record goto
6294Go to a specific location in the execution log. There are several
6295ways to specify the location to go to:
6296
6297@table @code
6298@item record goto begin
6299@itemx record goto start
6300Go to the beginning of the execution log.
6301
6302@item record goto end
6303Go to the end of the execution log.
6304
6305@item record goto @var{n}
6306Go to instruction number @var{n} in the execution log.
6307@end table
6308
6309@kindex record save
6310@item record save @var{filename}
6311Save the execution log to a file @file{@var{filename}}.
6312Default filename is @file{gdb_record.@var{process_id}}, where
6313@var{process_id} is the process ID of the inferior.
6314
6315This command may not be available for all recording methods.
6316
6317@kindex record restore
6318@item record restore @var{filename}
6319Restore the execution log from a file @file{@var{filename}}.
6320File must have been created with @code{record save}.
6321
6322@kindex set record full
6323@item set record full insn-number-max @var{limit}
6324@itemx set record full insn-number-max unlimited
6325Set the limit of instructions to be recorded for the @code{full}
6326recording method. Default value is 200000.
6327
6328If @var{limit} is a positive number, then @value{GDBN} will start
6329deleting instructions from the log once the number of the record
6330instructions becomes greater than @var{limit}. For every new recorded
6331instruction, @value{GDBN} will delete the earliest recorded
6332instruction to keep the number of recorded instructions at the limit.
6333(Since deleting recorded instructions loses information, @value{GDBN}
6334lets you control what happens when the limit is reached, by means of
6335the @code{stop-at-limit} option, described below.)
6336
6337If @var{limit} is @code{unlimited} or zero, @value{GDBN} will never
6338delete recorded instructions from the execution log. The number of
6339recorded instructions is limited only by the available memory.
6340
6341@kindex show record full
6342@item show record full insn-number-max
6343Show the limit of instructions to be recorded with the @code{full}
6344recording method.
6345
6346@item set record full stop-at-limit
6347Control the behavior of the @code{full} recording method when the
6348number of recorded instructions reaches the limit. If ON (the
6349default), @value{GDBN} will stop when the limit is reached for the
6350first time and ask you whether you want to stop the inferior or
6351continue running it and recording the execution log. If you decide
6352to continue recording, each new recorded instruction will cause the
6353oldest one to be deleted.
6354
6355If this option is OFF, @value{GDBN} will automatically delete the
6356oldest record to make room for each new one, without asking.
6357
6358@item show record full stop-at-limit
6359Show the current setting of @code{stop-at-limit}.
6360
6361@item set record full memory-query
6362Control the behavior when @value{GDBN} is unable to record memory
6363changes caused by an instruction for the @code{full} recording method.
6364If ON, @value{GDBN} will query whether to stop the inferior in that
6365case.
6366
6367If this option is OFF (the default), @value{GDBN} will automatically
6368ignore the effect of such instructions on memory. Later, when
6369@value{GDBN} replays this execution log, it will mark the log of this
6370instruction as not accessible, and it will not affect the replay
6371results.
6372
6373@item show record full memory-query
6374Show the current setting of @code{memory-query}.
6375
6376@kindex info record
6377@item info record
6378Show various statistics about the recording depending on the recording
6379method:
6380
6381@table @code
6382@item full
6383For the @code{full} recording method, it shows the state of process
6384record and its in-memory execution log buffer, including:
6385
6386@itemize @bullet
6387@item
6388Whether in record mode or replay mode.
6389@item
6390Lowest recorded instruction number (counting from when the current execution log started recording instructions).
6391@item
6392Highest recorded instruction number.
6393@item
6394Current instruction about to be replayed (if in replay mode).
6395@item
6396Number of instructions contained in the execution log.
6397@item
6398Maximum number of instructions that may be contained in the execution log.
6399@end itemize
6400
6401@item btrace
6402For the @code{btrace} recording method, it shows the number of
6403instructions that have been recorded and the number of blocks of
6404sequential control-flow that is formed by the recorded instructions.
6405@end table
6406
6407@kindex record delete
6408@kindex rec del
6409@item record delete
6410When record target runs in replay mode (``in the past''), delete the
6411subsequent execution log and begin to record a new execution log starting
6412from the current address. This means you will abandon the previously
6413recorded ``future'' and begin recording a new ``future''.
6414
6415@kindex record instruction-history
6416@kindex rec instruction-history
6417@item record instruction-history
6418Disassembles instructions from the recorded execution log. By
6419default, ten instructions are disassembled. This can be changed using
6420the @code{set record instruction-history-size} command. Instructions
6421are printed in execution order. There are several ways to specify
6422what part of the execution log to disassemble:
6423
6424@table @code
6425@item record instruction-history @var{insn}
6426Disassembles ten instructions starting from instruction number
6427@var{insn}.
6428
6429@item record instruction-history @var{insn}, +/-@var{n}
6430Disassembles @var{n} instructions around instruction number
6431@var{insn}. If @var{n} is preceded with @code{+}, disassembles
6432@var{n} instructions after instruction number @var{insn}. If
6433@var{n} is preceded with @code{-}, disassembles @var{n}
6434instructions before instruction number @var{insn}.
6435
6436@item record instruction-history
6437Disassembles ten more instructions after the last disassembly.
6438
6439@item record instruction-history -
6440Disassembles ten more instructions before the last disassembly.
6441
6442@item record instruction-history @var{begin} @var{end}
6443Disassembles instructions beginning with instruction number
6444@var{begin} until instruction number @var{end}. The instruction
6445number @var{end} is not included.
6446@end table
6447
6448This command may not be available for all recording methods.
6449
6450@kindex set record
6451@item set record instruction-history-size @var{size}
6452@itemx set record instruction-history-size unlimited
6453Define how many instructions to disassemble in the @code{record
6454instruction-history} command. The default value is 10.
6455A @var{size} of @code{unlimited} means unlimited instructions.
6456
6457@kindex show record
6458@item show record instruction-history-size
6459Show how many instructions to disassemble in the @code{record
6460instruction-history} command.
6461
6462@kindex record function-call-history
6463@kindex rec function-call-history
6464@item record function-call-history
6465Prints the execution history at function granularity. It prints one
6466line for each sequence of instructions that belong to the same
6467function giving the name of that function, the source lines
6468for this instruction sequence (if the @code{/l} modifier is
6469specified), and the instructions numbers that form the sequence (if
6470the @code{/i} modifier is specified).
6471
6472@smallexample
6473(@value{GDBP}) @b{list 1, 10}
64741 void foo (void)
64752 @{
64763 @}
64774
64785 void bar (void)
64796 @{
64807 ...
64818 foo ();
64829 ...
648310 @}
6484(@value{GDBP}) @b{record function-call-history /l}
64851 foo.c:6-8 bar
64862 foo.c:2-3 foo
64873 foo.c:9-10 bar
6488@end smallexample
6489
6490By default, ten lines are printed. This can be changed using the
6491@code{set record function-call-history-size} command. Functions are
6492printed in execution order. There are several ways to specify what
6493to print:
6494
6495@table @code
6496@item record function-call-history @var{func}
6497Prints ten functions starting from function number @var{func}.
6498
6499@item record function-call-history @var{func}, +/-@var{n}
6500Prints @var{n} functions around function number @var{func}. If
6501@var{n} is preceded with @code{+}, prints @var{n} functions after
6502function number @var{func}. If @var{n} is preceded with @code{-},
6503prints @var{n} functions before function number @var{func}.
6504
6505@item record function-call-history
6506Prints ten more functions after the last ten-line print.
6507
6508@item record function-call-history -
6509Prints ten more functions before the last ten-line print.
6510
6511@item record function-call-history @var{begin} @var{end}
6512Prints functions beginning with function number @var{begin} until
6513function number @var{end}. The function number @var{end} is not
6514included.
6515@end table
6516
6517This command may not be available for all recording methods.
6518
6519@item set record function-call-history-size @var{size}
6520@itemx set record function-call-history-size unlimited
6521Define how many lines to print in the
6522@code{record function-call-history} command. The default value is 10.
6523A size of @code{unlimited} means unlimited lines.
6524
6525@item show record function-call-history-size
6526Show how many lines to print in the
6527@code{record function-call-history} command.
6528@end table
6529
6530
6531@node Stack
6532@chapter Examining the Stack
6533
6534When your program has stopped, the first thing you need to know is where it
6535stopped and how it got there.
6536
6537@cindex call stack
6538Each time your program performs a function call, information about the call
6539is generated.
6540That information includes the location of the call in your program,
6541the arguments of the call,
6542and the local variables of the function being called.
6543The information is saved in a block of data called a @dfn{stack frame}.
6544The stack frames are allocated in a region of memory called the @dfn{call
6545stack}.
6546
6547When your program stops, the @value{GDBN} commands for examining the
6548stack allow you to see all of this information.
6549
6550@cindex selected frame
6551One of the stack frames is @dfn{selected} by @value{GDBN} and many
6552@value{GDBN} commands refer implicitly to the selected frame. In
6553particular, whenever you ask @value{GDBN} for the value of a variable in
6554your program, the value is found in the selected frame. There are
6555special @value{GDBN} commands to select whichever frame you are
6556interested in. @xref{Selection, ,Selecting a Frame}.
6557
6558When your program stops, @value{GDBN} automatically selects the
6559currently executing frame and describes it briefly, similar to the
6560@code{frame} command (@pxref{Frame Info, ,Information about a Frame}).
6561
6562@menu
6563* Frames:: Stack frames
6564* Backtrace:: Backtraces
6565* Frame Filter Management:: Managing frame filters
6566* Selection:: Selecting a frame
6567* Frame Info:: Information on a frame
6568
6569@end menu
6570
6571@node Frames
6572@section Stack Frames
6573
6574@cindex frame, definition
6575@cindex stack frame
6576The call stack is divided up into contiguous pieces called @dfn{stack
6577frames}, or @dfn{frames} for short; each frame is the data associated
6578with one call to one function. The frame contains the arguments given
6579to the function, the function's local variables, and the address at
6580which the function is executing.
6581
6582@cindex initial frame
6583@cindex outermost frame
6584@cindex innermost frame
6585When your program is started, the stack has only one frame, that of the
6586function @code{main}. This is called the @dfn{initial} frame or the
6587@dfn{outermost} frame. Each time a function is called, a new frame is
6588made. Each time a function returns, the frame for that function invocation
6589is eliminated. If a function is recursive, there can be many frames for
6590the same function. The frame for the function in which execution is
6591actually occurring is called the @dfn{innermost} frame. This is the most
6592recently created of all the stack frames that still exist.
6593
6594@cindex frame pointer
6595Inside your program, stack frames are identified by their addresses. A
6596stack frame consists of many bytes, each of which has its own address; each
6597kind of computer has a convention for choosing one byte whose
6598address serves as the address of the frame. Usually this address is kept
6599in a register called the @dfn{frame pointer register}
6600(@pxref{Registers, $fp}) while execution is going on in that frame.
6601
6602@cindex frame number
6603@value{GDBN} assigns numbers to all existing stack frames, starting with
6604zero for the innermost frame, one for the frame that called it,
6605and so on upward. These numbers do not really exist in your program;
6606they are assigned by @value{GDBN} to give you a way of designating stack
6607frames in @value{GDBN} commands.
6608
6609@c The -fomit-frame-pointer below perennially causes hbox overflow
6610@c underflow problems.
6611@cindex frameless execution
6612Some compilers provide a way to compile functions so that they operate
6613without stack frames. (For example, the @value{NGCC} option
6614@smallexample
6615@samp{-fomit-frame-pointer}
6616@end smallexample
6617generates functions without a frame.)
6618This is occasionally done with heavily used library functions to save
6619the frame setup time. @value{GDBN} has limited facilities for dealing
6620with these function invocations. If the innermost function invocation
6621has no stack frame, @value{GDBN} nevertheless regards it as though
6622it had a separate frame, which is numbered zero as usual, allowing
6623correct tracing of the function call chain. However, @value{GDBN} has
6624no provision for frameless functions elsewhere in the stack.
6625
6626@table @code
6627@kindex frame@r{, command}
6628@cindex current stack frame
6629@item frame @var{args}
6630The @code{frame} command allows you to move from one stack frame to another,
6631and to print the stack frame you select. @var{args} may be either the
6632address of the frame or the stack frame number. Without an argument,
6633@code{frame} prints the current stack frame.
6634
6635@kindex select-frame
6636@cindex selecting frame silently
6637@item select-frame
6638The @code{select-frame} command allows you to move from one stack frame
6639to another without printing the frame. This is the silent version of
6640@code{frame}.
6641@end table
6642
6643@node Backtrace
6644@section Backtraces
6645
6646@cindex traceback
6647@cindex call stack traces
6648A backtrace is a summary of how your program got where it is. It shows one
6649line per frame, for many frames, starting with the currently executing
6650frame (frame zero), followed by its caller (frame one), and on up the
6651stack.
6652
6653@anchor{backtrace-command}
6654@table @code
6655@kindex backtrace
6656@kindex bt @r{(@code{backtrace})}
6657@item backtrace
6658@itemx bt
6659Print a backtrace of the entire stack: one line per frame for all
6660frames in the stack.
6661
6662You can stop the backtrace at any time by typing the system interrupt
6663character, normally @kbd{Ctrl-c}.
6664
6665@item backtrace @var{n}
6666@itemx bt @var{n}
6667Similar, but print only the innermost @var{n} frames.
6668
6669@item backtrace -@var{n}
6670@itemx bt -@var{n}
6671Similar, but print only the outermost @var{n} frames.
6672
6673@item backtrace full
6674@itemx bt full
6675@itemx bt full @var{n}
6676@itemx bt full -@var{n}
6677Print the values of the local variables also. @var{n} specifies the
6678number of frames to print, as described above.
6679
6680@item backtrace no-filters
6681@itemx bt no-filters
6682@itemx bt no-filters @var{n}
6683@itemx bt no-filters -@var{n}
6684@itemx bt no-filters full
6685@itemx bt no-filters full @var{n}
6686@itemx bt no-filters full -@var{n}
6687Do not run Python frame filters on this backtrace. @xref{Frame
6688Filter API}, for more information. Additionally use @ref{disable
6689frame-filter all} to turn off all frame filters. This is only
6690relevant when @value{GDBN} has been configured with @code{Python}
6691support.
6692@end table
6693
6694@kindex where
6695@kindex info stack
6696The names @code{where} and @code{info stack} (abbreviated @code{info s})
6697are additional aliases for @code{backtrace}.
6698
6699@cindex multiple threads, backtrace
6700In a multi-threaded program, @value{GDBN} by default shows the
6701backtrace only for the current thread. To display the backtrace for
6702several or all of the threads, use the command @code{thread apply}
6703(@pxref{Threads, thread apply}). For example, if you type @kbd{thread
6704apply all backtrace}, @value{GDBN} will display the backtrace for all
6705the threads; this is handy when you debug a core dump of a
6706multi-threaded program.
6707
6708Each line in the backtrace shows the frame number and the function name.
6709The program counter value is also shown---unless you use @code{set
6710print address off}. The backtrace also shows the source file name and
6711line number, as well as the arguments to the function. The program
6712counter value is omitted if it is at the beginning of the code for that
6713line number.
6714
6715Here is an example of a backtrace. It was made with the command
6716@samp{bt 3}, so it shows the innermost three frames.
6717
6718@smallexample
6719@group
6720#0 m4_traceon (obs=0x24eb0, argc=1, argv=0x2b8c8)
6721 at builtin.c:993
6722#1 0x6e38 in expand_macro (sym=0x2b600, data=...) at macro.c:242
6723#2 0x6840 in expand_token (obs=0x0, t=177664, td=0xf7fffb08)
6724 at macro.c:71
6725(More stack frames follow...)
6726@end group
6727@end smallexample
6728
6729@noindent
6730The display for frame zero does not begin with a program counter
6731value, indicating that your program has stopped at the beginning of the
6732code for line @code{993} of @code{builtin.c}.
6733
6734@noindent
6735The value of parameter @code{data} in frame 1 has been replaced by
6736@code{@dots{}}. By default, @value{GDBN} prints the value of a parameter
6737only if it is a scalar (integer, pointer, enumeration, etc). See command
6738@kbd{set print frame-arguments} in @ref{Print Settings} for more details
6739on how to configure the way function parameter values are printed.
6740
6741@cindex optimized out, in backtrace
6742@cindex function call arguments, optimized out
6743If your program was compiled with optimizations, some compilers will
6744optimize away arguments passed to functions if those arguments are
6745never used after the call. Such optimizations generate code that
6746passes arguments through registers, but doesn't store those arguments
6747in the stack frame. @value{GDBN} has no way of displaying such
6748arguments in stack frames other than the innermost one. Here's what
6749such a backtrace might look like:
6750
6751@smallexample
6752@group
6753#0 m4_traceon (obs=0x24eb0, argc=1, argv=0x2b8c8)
6754 at builtin.c:993
6755#1 0x6e38 in expand_macro (sym=<optimized out>) at macro.c:242
6756#2 0x6840 in expand_token (obs=0x0, t=<optimized out>, td=0xf7fffb08)
6757 at macro.c:71
6758(More stack frames follow...)
6759@end group
6760@end smallexample
6761
6762@noindent
6763The values of arguments that were not saved in their stack frames are
6764shown as @samp{<optimized out>}.
6765
6766If you need to display the values of such optimized-out arguments,
6767either deduce that from other variables whose values depend on the one
6768you are interested in, or recompile without optimizations.
6769
6770@cindex backtrace beyond @code{main} function
6771@cindex program entry point
6772@cindex startup code, and backtrace
6773Most programs have a standard user entry point---a place where system
6774libraries and startup code transition into user code. For C this is
6775@code{main}@footnote{
6776Note that embedded programs (the so-called ``free-standing''
6777environment) are not required to have a @code{main} function as the
6778entry point. They could even have multiple entry points.}.
6779When @value{GDBN} finds the entry function in a backtrace
6780it will terminate the backtrace, to avoid tracing into highly
6781system-specific (and generally uninteresting) code.
6782
6783If you need to examine the startup code, or limit the number of levels
6784in a backtrace, you can change this behavior:
6785
6786@table @code
6787@item set backtrace past-main
6788@itemx set backtrace past-main on
6789@kindex set backtrace
6790Backtraces will continue past the user entry point.
6791
6792@item set backtrace past-main off
6793Backtraces will stop when they encounter the user entry point. This is the
6794default.
6795
6796@item show backtrace past-main
6797@kindex show backtrace
6798Display the current user entry point backtrace policy.
6799
6800@item set backtrace past-entry
6801@itemx set backtrace past-entry on
6802Backtraces will continue past the internal entry point of an application.
6803This entry point is encoded by the linker when the application is built,
6804and is likely before the user entry point @code{main} (or equivalent) is called.
6805
6806@item set backtrace past-entry off
6807Backtraces will stop when they encounter the internal entry point of an
6808application. This is the default.
6809
6810@item show backtrace past-entry
6811Display the current internal entry point backtrace policy.
6812
6813@item set backtrace limit @var{n}
6814@itemx set backtrace limit 0
6815@itemx set backtrace limit unlimited
6816@cindex backtrace limit
6817Limit the backtrace to @var{n} levels. A value of @code{unlimited}
6818or zero means unlimited levels.
6819
6820@item show backtrace limit
6821Display the current limit on backtrace levels.
6822@end table
6823
6824You can control how file names are displayed.
6825
6826@table @code
6827@item set filename-display
6828@itemx set filename-display relative
6829@cindex filename-display
6830Display file names relative to the compilation directory. This is the default.
6831
6832@item set filename-display basename
6833Display only basename of a filename.
6834
6835@item set filename-display absolute
6836Display an absolute filename.
6837
6838@item show filename-display
6839Show the current way to display filenames.
6840@end table
6841
6842@node Frame Filter Management
6843@section Management of Frame Filters.
6844@cindex managing frame filters
6845
6846Frame filters are Python based utilities to manage and decorate the
6847output of frames. @xref{Frame Filter API}, for further information.
6848
6849Managing frame filters is performed by several commands available
6850within @value{GDBN}, detailed here.
6851
6852@table @code
6853@kindex info frame-filter
6854@item info frame-filter
6855Print a list of installed frame filters from all dictionaries, showing
6856their name, priority and enabled status.
6857
6858@kindex disable frame-filter
6859@anchor{disable frame-filter all}
6860@item disable frame-filter @var{filter-dictionary} @var{filter-name}
6861Disable a frame filter in the dictionary matching
6862@var{filter-dictionary}, or @code{all}, and @var{filter-name}.
6863@var{filter-dictionary} may be @code{all}, @code{global},
6864@code{progspace} or the name of the object file where the frame filter
6865dictionary resides. When @code{all} is specified, all frame filters
6866across all dictionaries are disabled. @var{filter-name} is the name
6867of the frame filter and is used when @code{all} is not the option for
6868@var{filter-dictionary}. A disabled frame-filter is not deleted, it
6869may be enabled again later.
6870
6871@kindex enable frame-filter
6872@item enable frame-filter @var{filter-dictionary} @var{filter-name}
6873Enable a frame filter in the dictionary matching
6874@var{filter-dictionary}, or @code{all}, and @var{filter-name}.
6875@var{filter-dictionary} may be @code{all}, @code{global},
6876@code{progspace} or the name of the object file where the frame filter
6877dictionary resides. When @code{all} is specified, all frame filters across
6878all dictionaries are enabled. @var{filter-name} is the name of the frame
6879filter and is used when @code{all} is not the option for
6880@var{filter-dictionary}.
6881
6882Example:
6883
6884@smallexample
6885(gdb) info frame-filter
6886
6887global frame-filters:
6888 Priority Enabled Name
6889 1000 No PrimaryFunctionFilter
6890 100 Yes Reverse
6891
6892progspace /build/test frame-filters:
6893 Priority Enabled Name
6894 100 Yes ProgspaceFilter
6895
6896objfile /build/test frame-filters:
6897 Priority Enabled Name
6898 999 Yes BuildProgra Filter
6899
6900(gdb) disable frame-filter /build/test BuildProgramFilter
6901(gdb) info frame-filter
6902
6903global frame-filters:
6904 Priority Enabled Name
6905 1000 No PrimaryFunctionFilter
6906 100 Yes Reverse
6907
6908progspace /build/test frame-filters:
6909 Priority Enabled Name
6910 100 Yes ProgspaceFilter
6911
6912objfile /build/test frame-filters:
6913 Priority Enabled Name
6914 999 No BuildProgramFilter
6915
6916(gdb) enable frame-filter global PrimaryFunctionFilter
6917(gdb) info frame-filter
6918
6919global frame-filters:
6920 Priority Enabled Name
6921 1000 Yes PrimaryFunctionFilter
6922 100 Yes Reverse
6923
6924progspace /build/test frame-filters:
6925 Priority Enabled Name
6926 100 Yes ProgspaceFilter
6927
6928objfile /build/test frame-filters:
6929 Priority Enabled Name
6930 999 No BuildProgramFilter
6931@end smallexample
6932
6933@kindex set frame-filter priority
6934@item set frame-filter priority @var{filter-dictionary} @var{filter-name} @var{priority}
6935Set the @var{priority} of a frame filter in the dictionary matching
6936@var{filter-dictionary}, and the frame filter name matching
6937@var{filter-name}. @var{filter-dictionary} may be @code{global},
6938@code{progspace} or the name of the object file where the frame filter
6939dictionary resides. @var{priority} is an integer.
6940
6941@kindex show frame-filter priority
6942@item show frame-filter priority @var{filter-dictionary} @var{filter-name}
6943Show the @var{priority} of a frame filter in the dictionary matching
6944@var{filter-dictionary}, and the frame filter name matching
6945@var{filter-name}. @var{filter-dictionary} may be @code{global},
6946@code{progspace} or the name of the object file where the frame filter
6947dictionary resides.
6948
6949Example:
6950
6951@smallexample
6952(gdb) info frame-filter
6953
6954global frame-filters:
6955 Priority Enabled Name
6956 1000 Yes PrimaryFunctionFilter
6957 100 Yes Reverse
6958
6959progspace /build/test frame-filters:
6960 Priority Enabled Name
6961 100 Yes ProgspaceFilter
6962
6963objfile /build/test frame-filters:
6964 Priority Enabled Name
6965 999 No BuildProgramFilter
6966
6967(gdb) set frame-filter priority global Reverse 50
6968(gdb) info frame-filter
6969
6970global frame-filters:
6971 Priority Enabled Name
6972 1000 Yes PrimaryFunctionFilter
6973 50 Yes Reverse
6974
6975progspace /build/test frame-filters:
6976 Priority Enabled Name
6977 100 Yes ProgspaceFilter
6978
6979objfile /build/test frame-filters:
6980 Priority Enabled Name
6981 999 No BuildProgramFilter
6982@end smallexample
6983@end table
6984
6985@node Selection
6986@section Selecting a Frame
6987
6988Most commands for examining the stack and other data in your program work on
6989whichever stack frame is selected at the moment. Here are the commands for
6990selecting a stack frame; all of them finish by printing a brief description
6991of the stack frame just selected.
6992
6993@table @code
6994@kindex frame@r{, selecting}
6995@kindex f @r{(@code{frame})}
6996@item frame @var{n}
6997@itemx f @var{n}
6998Select frame number @var{n}. Recall that frame zero is the innermost
6999(currently executing) frame, frame one is the frame that called the
7000innermost one, and so on. The highest-numbered frame is the one for
7001@code{main}.
7002
7003@item frame @var{addr}
7004@itemx f @var{addr}
7005Select the frame at address @var{addr}. This is useful mainly if the
7006chaining of stack frames has been damaged by a bug, making it
7007impossible for @value{GDBN} to assign numbers properly to all frames. In
7008addition, this can be useful when your program has multiple stacks and
7009switches between them.
7010
7011On the SPARC architecture, @code{frame} needs two addresses to
7012select an arbitrary frame: a frame pointer and a stack pointer.
7013
7014On the @acronym{MIPS} and Alpha architecture, it needs two addresses: a stack
7015pointer and a program counter.
7016
7017On the 29k architecture, it needs three addresses: a register stack
7018pointer, a program counter, and a memory stack pointer.
7019
7020@kindex up
7021@item up @var{n}
7022Move @var{n} frames up the stack. For positive numbers @var{n}, this
7023advances toward the outermost frame, to higher frame numbers, to frames
7024that have existed longer. @var{n} defaults to one.
7025
7026@kindex down
7027@kindex do @r{(@code{down})}
7028@item down @var{n}
7029Move @var{n} frames down the stack. For positive numbers @var{n}, this
7030advances toward the innermost frame, to lower frame numbers, to frames
7031that were created more recently. @var{n} defaults to one. You may
7032abbreviate @code{down} as @code{do}.
7033@end table
7034
7035All of these commands end by printing two lines of output describing the
7036frame. The first line shows the frame number, the function name, the
7037arguments, and the source file and line number of execution in that
7038frame. The second line shows the text of that source line.
7039
7040@need 1000
7041For example:
7042
7043@smallexample
7044@group
7045(@value{GDBP}) up
7046#1 0x22f0 in main (argc=1, argv=0xf7fffbf4, env=0xf7fffbfc)
7047 at env.c:10
704810 read_input_file (argv[i]);
7049@end group
7050@end smallexample
7051
7052After such a printout, the @code{list} command with no arguments
7053prints ten lines centered on the point of execution in the frame.
7054You can also edit the program at the point of execution with your favorite
7055editing program by typing @code{edit}.
7056@xref{List, ,Printing Source Lines},
7057for details.
7058
7059@table @code
7060@kindex down-silently
7061@kindex up-silently
7062@item up-silently @var{n}
7063@itemx down-silently @var{n}
7064These two commands are variants of @code{up} and @code{down},
7065respectively; they differ in that they do their work silently, without
7066causing display of the new frame. They are intended primarily for use
7067in @value{GDBN} command scripts, where the output might be unnecessary and
7068distracting.
7069@end table
7070
7071@node Frame Info
7072@section Information About a Frame
7073
7074There are several other commands to print information about the selected
7075stack frame.
7076
7077@table @code
7078@item frame
7079@itemx f
7080When used without any argument, this command does not change which
7081frame is selected, but prints a brief description of the currently
7082selected stack frame. It can be abbreviated @code{f}. With an
7083argument, this command is used to select a stack frame.
7084@xref{Selection, ,Selecting a Frame}.
7085
7086@kindex info frame
7087@kindex info f @r{(@code{info frame})}
7088@item info frame
7089@itemx info f
7090This command prints a verbose description of the selected stack frame,
7091including:
7092
7093@itemize @bullet
7094@item
7095the address of the frame
7096@item
7097the address of the next frame down (called by this frame)
7098@item
7099the address of the next frame up (caller of this frame)
7100@item
7101the language in which the source code corresponding to this frame is written
7102@item
7103the address of the frame's arguments
7104@item
7105the address of the frame's local variables
7106@item
7107the program counter saved in it (the address of execution in the caller frame)
7108@item
7109which registers were saved in the frame
7110@end itemize
7111
7112@noindent The verbose description is useful when
7113something has gone wrong that has made the stack format fail to fit
7114the usual conventions.
7115
7116@item info frame @var{addr}
7117@itemx info f @var{addr}
7118Print a verbose description of the frame at address @var{addr}, without
7119selecting that frame. The selected frame remains unchanged by this
7120command. This requires the same kind of address (more than one for some
7121architectures) that you specify in the @code{frame} command.
7122@xref{Selection, ,Selecting a Frame}.
7123
7124@kindex info args
7125@item info args
7126Print the arguments of the selected frame, each on a separate line.
7127
7128@item info locals
7129@kindex info locals
7130Print the local variables of the selected frame, each on a separate
7131line. These are all variables (declared either static or automatic)
7132accessible at the point of execution of the selected frame.
7133
7134@end table
7135
7136
7137@node Source
7138@chapter Examining Source Files
7139
7140@value{GDBN} can print parts of your program's source, since the debugging
7141information recorded in the program tells @value{GDBN} what source files were
7142used to build it. When your program stops, @value{GDBN} spontaneously prints
7143the line where it stopped. Likewise, when you select a stack frame
7144(@pxref{Selection, ,Selecting a Frame}), @value{GDBN} prints the line where
7145execution in that frame has stopped. You can print other portions of
7146source files by explicit command.
7147
7148If you use @value{GDBN} through its @sc{gnu} Emacs interface, you may
7149prefer to use Emacs facilities to view source; see @ref{Emacs, ,Using
7150@value{GDBN} under @sc{gnu} Emacs}.
7151
7152@menu
7153* List:: Printing source lines
7154* Specify Location:: How to specify code locations
7155* Edit:: Editing source files
7156* Search:: Searching source files
7157* Source Path:: Specifying source directories
7158* Machine Code:: Source and machine code
7159@end menu
7160
7161@node List
7162@section Printing Source Lines
7163
7164@kindex list
7165@kindex l @r{(@code{list})}
7166To print lines from a source file, use the @code{list} command
7167(abbreviated @code{l}). By default, ten lines are printed.
7168There are several ways to specify what part of the file you want to
7169print; see @ref{Specify Location}, for the full list.
7170
7171Here are the forms of the @code{list} command most commonly used:
7172
7173@table @code
7174@item list @var{linenum}
7175Print lines centered around line number @var{linenum} in the
7176current source file.
7177
7178@item list @var{function}
7179Print lines centered around the beginning of function
7180@var{function}.
7181
7182@item list
7183Print more lines. If the last lines printed were printed with a
7184@code{list} command, this prints lines following the last lines
7185printed; however, if the last line printed was a solitary line printed
7186as part of displaying a stack frame (@pxref{Stack, ,Examining the
7187Stack}), this prints lines centered around that line.
7188
7189@item list -
7190Print lines just before the lines last printed.
7191@end table
7192
7193@cindex @code{list}, how many lines to display
7194By default, @value{GDBN} prints ten source lines with any of these forms of
7195the @code{list} command. You can change this using @code{set listsize}:
7196
7197@table @code
7198@kindex set listsize
7199@item set listsize @var{count}
7200@itemx set listsize unlimited
7201Make the @code{list} command display @var{count} source lines (unless
7202the @code{list} argument explicitly specifies some other number).
7203Setting @var{count} to @code{unlimited} or 0 means there's no limit.
7204
7205@kindex show listsize
7206@item show listsize
7207Display the number of lines that @code{list} prints.
7208@end table
7209
7210Repeating a @code{list} command with @key{RET} discards the argument,
7211so it is equivalent to typing just @code{list}. This is more useful
7212than listing the same lines again. An exception is made for an
7213argument of @samp{-}; that argument is preserved in repetition so that
7214each repetition moves up in the source file.
7215
7216In general, the @code{list} command expects you to supply zero, one or two
7217@dfn{linespecs}. Linespecs specify source lines; there are several ways
7218of writing them (@pxref{Specify Location}), but the effect is always
7219to specify some source line.
7220
7221Here is a complete description of the possible arguments for @code{list}:
7222
7223@table @code
7224@item list @var{linespec}
7225Print lines centered around the line specified by @var{linespec}.
7226
7227@item list @var{first},@var{last}
7228Print lines from @var{first} to @var{last}. Both arguments are
7229linespecs. When a @code{list} command has two linespecs, and the
7230source file of the second linespec is omitted, this refers to
7231the same source file as the first linespec.
7232
7233@item list ,@var{last}
7234Print lines ending with @var{last}.
7235
7236@item list @var{first},
7237Print lines starting with @var{first}.
7238
7239@item list +
7240Print lines just after the lines last printed.
7241
7242@item list -
7243Print lines just before the lines last printed.
7244
7245@item list
7246As described in the preceding table.
7247@end table
7248
7249@node Specify Location
7250@section Specifying a Location
7251@cindex specifying location
7252@cindex linespec
7253
7254Several @value{GDBN} commands accept arguments that specify a location
7255of your program's code. Since @value{GDBN} is a source-level
7256debugger, a location usually specifies some line in the source code;
7257for that reason, locations are also known as @dfn{linespecs}.
7258
7259Here are all the different ways of specifying a code location that
7260@value{GDBN} understands:
7261
7262@table @code
7263@item @var{linenum}
7264Specifies the line number @var{linenum} of the current source file.
7265
7266@item -@var{offset}
7267@itemx +@var{offset}
7268Specifies the line @var{offset} lines before or after the @dfn{current
7269line}. For the @code{list} command, the current line is the last one
7270printed; for the breakpoint commands, this is the line at which
7271execution stopped in the currently selected @dfn{stack frame}
7272(@pxref{Frames, ,Frames}, for a description of stack frames.) When
7273used as the second of the two linespecs in a @code{list} command,
7274this specifies the line @var{offset} lines up or down from the first
7275linespec.
7276
7277@item @var{filename}:@var{linenum}
7278Specifies the line @var{linenum} in the source file @var{filename}.
7279If @var{filename} is a relative file name, then it will match any
7280source file name with the same trailing components. For example, if
7281@var{filename} is @samp{gcc/expr.c}, then it will match source file
7282name of @file{/build/trunk/gcc/expr.c}, but not
7283@file{/build/trunk/libcpp/expr.c} or @file{/build/trunk/gcc/x-expr.c}.
7284
7285@item @var{function}
7286Specifies the line that begins the body of the function @var{function}.
7287For example, in C, this is the line with the open brace.
7288
7289@item @var{function}:@var{label}
7290Specifies the line where @var{label} appears in @var{function}.
7291
7292@item @var{filename}:@var{function}
7293Specifies the line that begins the body of the function @var{function}
7294in the file @var{filename}. You only need the file name with a
7295function name to avoid ambiguity when there are identically named
7296functions in different source files.
7297
7298@item @var{label}
7299Specifies the line at which the label named @var{label} appears.
7300@value{GDBN} searches for the label in the function corresponding to
7301the currently selected stack frame. If there is no current selected
7302stack frame (for instance, if the inferior is not running), then
7303@value{GDBN} will not search for a label.
7304
7305@item *@var{address}
7306Specifies the program address @var{address}. For line-oriented
7307commands, such as @code{list} and @code{edit}, this specifies a source
7308line that contains @var{address}. For @code{break} and other
7309breakpoint oriented commands, this can be used to set breakpoints in
7310parts of your program which do not have debugging information or
7311source files.
7312
7313Here @var{address} may be any expression valid in the current working
7314language (@pxref{Languages, working language}) that specifies a code
7315address. In addition, as a convenience, @value{GDBN} extends the
7316semantics of expressions used in locations to cover the situations
7317that frequently happen during debugging. Here are the various forms
7318of @var{address}:
7319
7320@table @code
7321@item @var{expression}
7322Any expression valid in the current working language.
7323
7324@item @var{funcaddr}
7325An address of a function or procedure derived from its name. In C,
7326C@t{++}, Java, Objective-C, Fortran, minimal, and assembly, this is
7327simply the function's name @var{function} (and actually a special case
7328of a valid expression). In Pascal and Modula-2, this is
7329@code{&@var{function}}. In Ada, this is @code{@var{function}'Address}
7330(although the Pascal form also works).
7331
7332This form specifies the address of the function's first instruction,
7333before the stack frame and arguments have been set up.
7334
7335@item '@var{filename}'::@var{funcaddr}
7336Like @var{funcaddr} above, but also specifies the name of the source
7337file explicitly. This is useful if the name of the function does not
7338specify the function unambiguously, e.g., if there are several
7339functions with identical names in different source files.
7340@end table
7341
7342@cindex breakpoint at static probe point
7343@item -pstap|-probe-stap @r{[}@var{objfile}:@r{[}@var{provider}:@r{]}@r{]}@var{name}
7344The @sc{gnu}/Linux tool @code{SystemTap} provides a way for
7345applications to embed static probes. @xref{Static Probe Points}, for more
7346information on finding and using static probes. This form of linespec
7347specifies the location of such a static probe.
7348
7349If @var{objfile} is given, only probes coming from that shared library
7350or executable matching @var{objfile} as a regular expression are considered.
7351If @var{provider} is given, then only probes from that provider are considered.
7352If several probes match the spec, @value{GDBN} will insert a breakpoint at
7353each one of those probes.
7354
7355@end table
7356
7357
7358@node Edit
7359@section Editing Source Files
7360@cindex editing source files
7361
7362@kindex edit
7363@kindex e @r{(@code{edit})}
7364To edit the lines in a source file, use the @code{edit} command.
7365The editing program of your choice
7366is invoked with the current line set to
7367the active line in the program.
7368Alternatively, there are several ways to specify what part of the file you
7369want to print if you want to see other parts of the program:
7370
7371@table @code
7372@item edit @var{location}
7373Edit the source file specified by @code{location}. Editing starts at
7374that @var{location}, e.g., at the specified source line of the
7375specified file. @xref{Specify Location}, for all the possible forms
7376of the @var{location} argument; here are the forms of the @code{edit}
7377command most commonly used:
7378
7379@table @code
7380@item edit @var{number}
7381Edit the current source file with @var{number} as the active line number.
7382
7383@item edit @var{function}
7384Edit the file containing @var{function} at the beginning of its definition.
7385@end table
7386
7387@end table
7388
7389@subsection Choosing your Editor
7390You can customize @value{GDBN} to use any editor you want
7391@footnote{
7392The only restriction is that your editor (say @code{ex}), recognizes the
7393following command-line syntax:
7394@smallexample
7395ex +@var{number} file
7396@end smallexample
7397The optional numeric value +@var{number} specifies the number of the line in
7398the file where to start editing.}.
7399By default, it is @file{@value{EDITOR}}, but you can change this
7400by setting the environment variable @code{EDITOR} before using
7401@value{GDBN}. For example, to configure @value{GDBN} to use the
7402@code{vi} editor, you could use these commands with the @code{sh} shell:
7403@smallexample
7404EDITOR=/usr/bin/vi
7405export EDITOR
7406gdb @dots{}
7407@end smallexample
7408or in the @code{csh} shell,
7409@smallexample
7410setenv EDITOR /usr/bin/vi
7411gdb @dots{}
7412@end smallexample
7413
7414@node Search
7415@section Searching Source Files
7416@cindex searching source files
7417
7418There are two commands for searching through the current source file for a
7419regular expression.
7420
7421@table @code
7422@kindex search
7423@kindex forward-search
7424@kindex fo @r{(@code{forward-search})}
7425@item forward-search @var{regexp}
7426@itemx search @var{regexp}
7427The command @samp{forward-search @var{regexp}} checks each line,
7428starting with the one following the last line listed, for a match for
7429@var{regexp}. It lists the line that is found. You can use the
7430synonym @samp{search @var{regexp}} or abbreviate the command name as
7431@code{fo}.
7432
7433@kindex reverse-search
7434@item reverse-search @var{regexp}
7435The command @samp{reverse-search @var{regexp}} checks each line, starting
7436with the one before the last line listed and going backward, for a match
7437for @var{regexp}. It lists the line that is found. You can abbreviate
7438this command as @code{rev}.
7439@end table
7440
7441@node Source Path
7442@section Specifying Source Directories
7443
7444@cindex source path
7445@cindex directories for source files
7446Executable programs sometimes do not record the directories of the source
7447files from which they were compiled, just the names. Even when they do,
7448the directories could be moved between the compilation and your debugging
7449session. @value{GDBN} has a list of directories to search for source files;
7450this is called the @dfn{source path}. Each time @value{GDBN} wants a source file,
7451it tries all the directories in the list, in the order they are present
7452in the list, until it finds a file with the desired name.
7453
7454For example, suppose an executable references the file
7455@file{/usr/src/foo-1.0/lib/foo.c}, and our source path is
7456@file{/mnt/cross}. The file is first looked up literally; if this
7457fails, @file{/mnt/cross/usr/src/foo-1.0/lib/foo.c} is tried; if this
7458fails, @file{/mnt/cross/foo.c} is opened; if this fails, an error
7459message is printed. @value{GDBN} does not look up the parts of the
7460source file name, such as @file{/mnt/cross/src/foo-1.0/lib/foo.c}.
7461Likewise, the subdirectories of the source path are not searched: if
7462the source path is @file{/mnt/cross}, and the binary refers to
7463@file{foo.c}, @value{GDBN} would not find it under
7464@file{/mnt/cross/usr/src/foo-1.0/lib}.
7465
7466Plain file names, relative file names with leading directories, file
7467names containing dots, etc.@: are all treated as described above; for
7468instance, if the source path is @file{/mnt/cross}, and the source file
7469is recorded as @file{../lib/foo.c}, @value{GDBN} would first try
7470@file{../lib/foo.c}, then @file{/mnt/cross/../lib/foo.c}, and after
7471that---@file{/mnt/cross/foo.c}.
7472
7473Note that the executable search path is @emph{not} used to locate the
7474source files.
7475
7476Whenever you reset or rearrange the source path, @value{GDBN} clears out
7477any information it has cached about where source files are found and where
7478each line is in the file.
7479
7480@kindex directory
7481@kindex dir
7482When you start @value{GDBN}, its source path includes only @samp{cdir}
7483and @samp{cwd}, in that order.
7484To add other directories, use the @code{directory} command.
7485
7486The search path is used to find both program source files and @value{GDBN}
7487script files (read using the @samp{-command} option and @samp{source} command).
7488
7489In addition to the source path, @value{GDBN} provides a set of commands
7490that manage a list of source path substitution rules. A @dfn{substitution
7491rule} specifies how to rewrite source directories stored in the program's
7492debug information in case the sources were moved to a different
7493directory between compilation and debugging. A rule is made of
7494two strings, the first specifying what needs to be rewritten in
7495the path, and the second specifying how it should be rewritten.
7496In @ref{set substitute-path}, we name these two parts @var{from} and
7497@var{to} respectively. @value{GDBN} does a simple string replacement
7498of @var{from} with @var{to} at the start of the directory part of the
7499source file name, and uses that result instead of the original file
7500name to look up the sources.
7501
7502Using the previous example, suppose the @file{foo-1.0} tree has been
7503moved from @file{/usr/src} to @file{/mnt/cross}, then you can tell
7504@value{GDBN} to replace @file{/usr/src} in all source path names with
7505@file{/mnt/cross}. The first lookup will then be
7506@file{/mnt/cross/foo-1.0/lib/foo.c} in place of the original location
7507of @file{/usr/src/foo-1.0/lib/foo.c}. To define a source path
7508substitution rule, use the @code{set substitute-path} command
7509(@pxref{set substitute-path}).
7510
7511To avoid unexpected substitution results, a rule is applied only if the
7512@var{from} part of the directory name ends at a directory separator.
7513For instance, a rule substituting @file{/usr/source} into
7514@file{/mnt/cross} will be applied to @file{/usr/source/foo-1.0} but
7515not to @file{/usr/sourceware/foo-2.0}. And because the substitution
7516is applied only at the beginning of the directory name, this rule will
7517not be applied to @file{/root/usr/source/baz.c} either.
7518
7519In many cases, you can achieve the same result using the @code{directory}
7520command. However, @code{set substitute-path} can be more efficient in
7521the case where the sources are organized in a complex tree with multiple
7522subdirectories. With the @code{directory} command, you need to add each
7523subdirectory of your project. If you moved the entire tree while
7524preserving its internal organization, then @code{set substitute-path}
7525allows you to direct the debugger to all the sources with one single
7526command.
7527
7528@code{set substitute-path} is also more than just a shortcut command.
7529The source path is only used if the file at the original location no
7530longer exists. On the other hand, @code{set substitute-path} modifies
7531the debugger behavior to look at the rewritten location instead. So, if
7532for any reason a source file that is not relevant to your executable is
7533located at the original location, a substitution rule is the only
7534method available to point @value{GDBN} at the new location.
7535
7536@cindex @samp{--with-relocated-sources}
7537@cindex default source path substitution
7538You can configure a default source path substitution rule by
7539configuring @value{GDBN} with the
7540@samp{--with-relocated-sources=@var{dir}} option. The @var{dir}
7541should be the name of a directory under @value{GDBN}'s configured
7542prefix (set with @samp{--prefix} or @samp{--exec-prefix}), and
7543directory names in debug information under @var{dir} will be adjusted
7544automatically if the installed @value{GDBN} is moved to a new
7545location. This is useful if @value{GDBN}, libraries or executables
7546with debug information and corresponding source code are being moved
7547together.
7548
7549@table @code
7550@item directory @var{dirname} @dots{}
7551@item dir @var{dirname} @dots{}
7552Add directory @var{dirname} to the front of the source path. Several
7553directory names may be given to this command, separated by @samp{:}
7554(@samp{;} on MS-DOS and MS-Windows, where @samp{:} usually appears as
7555part of absolute file names) or
7556whitespace. You may specify a directory that is already in the source
7557path; this moves it forward, so @value{GDBN} searches it sooner.
7558
7559@kindex cdir
7560@kindex cwd
7561@vindex $cdir@r{, convenience variable}
7562@vindex $cwd@r{, convenience variable}
7563@cindex compilation directory
7564@cindex current directory
7565@cindex working directory
7566@cindex directory, current
7567@cindex directory, compilation
7568You can use the string @samp{$cdir} to refer to the compilation
7569directory (if one is recorded), and @samp{$cwd} to refer to the current
7570working directory. @samp{$cwd} is not the same as @samp{.}---the former
7571tracks the current working directory as it changes during your @value{GDBN}
7572session, while the latter is immediately expanded to the current
7573directory at the time you add an entry to the source path.
7574
7575@item directory
7576Reset the source path to its default value (@samp{$cdir:$cwd} on Unix systems). This requires confirmation.
7577
7578@c RET-repeat for @code{directory} is explicitly disabled, but since
7579@c repeating it would be a no-op we do not say that. (thanks to RMS)
7580
7581@item set directories @var{path-list}
7582@kindex set directories
7583Set the source path to @var{path-list}.
7584@samp{$cdir:$cwd} are added if missing.
7585
7586@item show directories
7587@kindex show directories
7588Print the source path: show which directories it contains.
7589
7590@anchor{set substitute-path}
7591@item set substitute-path @var{from} @var{to}
7592@kindex set substitute-path
7593Define a source path substitution rule, and add it at the end of the
7594current list of existing substitution rules. If a rule with the same
7595@var{from} was already defined, then the old rule is also deleted.
7596
7597For example, if the file @file{/foo/bar/baz.c} was moved to
7598@file{/mnt/cross/baz.c}, then the command
7599
7600@smallexample
7601(@value{GDBP}) set substitute-path /usr/src /mnt/cross
7602@end smallexample
7603
7604@noindent
7605will tell @value{GDBN} to replace @samp{/usr/src} with
7606@samp{/mnt/cross}, which will allow @value{GDBN} to find the file
7607@file{baz.c} even though it was moved.
7608
7609In the case when more than one substitution rule have been defined,
7610the rules are evaluated one by one in the order where they have been
7611defined. The first one matching, if any, is selected to perform
7612the substitution.
7613
7614For instance, if we had entered the following commands:
7615
7616@smallexample
7617(@value{GDBP}) set substitute-path /usr/src/include /mnt/include
7618(@value{GDBP}) set substitute-path /usr/src /mnt/src
7619@end smallexample
7620
7621@noindent
7622@value{GDBN} would then rewrite @file{/usr/src/include/defs.h} into
7623@file{/mnt/include/defs.h} by using the first rule. However, it would
7624use the second rule to rewrite @file{/usr/src/lib/foo.c} into
7625@file{/mnt/src/lib/foo.c}.
7626
7627
7628@item unset substitute-path [path]
7629@kindex unset substitute-path
7630If a path is specified, search the current list of substitution rules
7631for a rule that would rewrite that path. Delete that rule if found.
7632A warning is emitted by the debugger if no rule could be found.
7633
7634If no path is specified, then all substitution rules are deleted.
7635
7636@item show substitute-path [path]
7637@kindex show substitute-path
7638If a path is specified, then print the source path substitution rule
7639which would rewrite that path, if any.
7640
7641If no path is specified, then print all existing source path substitution
7642rules.
7643
7644@end table
7645
7646If your source path is cluttered with directories that are no longer of
7647interest, @value{GDBN} may sometimes cause confusion by finding the wrong
7648versions of source. You can correct the situation as follows:
7649
7650@enumerate
7651@item
7652Use @code{directory} with no argument to reset the source path to its default value.
7653
7654@item
7655Use @code{directory} with suitable arguments to reinstall the
7656directories you want in the source path. You can add all the
7657directories in one command.
7658@end enumerate
7659
7660@node Machine Code
7661@section Source and Machine Code
7662@cindex source line and its code address
7663
7664You can use the command @code{info line} to map source lines to program
7665addresses (and vice versa), and the command @code{disassemble} to display
7666a range of addresses as machine instructions. You can use the command
7667@code{set disassemble-next-line} to set whether to disassemble next
7668source line when execution stops. When run under @sc{gnu} Emacs
7669mode, the @code{info line} command causes the arrow to point to the
7670line specified. Also, @code{info line} prints addresses in symbolic form as
7671well as hex.
7672
7673@table @code
7674@kindex info line
7675@item info line @var{linespec}
7676Print the starting and ending addresses of the compiled code for
7677source line @var{linespec}. You can specify source lines in any of
7678the ways documented in @ref{Specify Location}.
7679@end table
7680
7681For example, we can use @code{info line} to discover the location of
7682the object code for the first line of function
7683@code{m4_changequote}:
7684
7685@c FIXME: I think this example should also show the addresses in
7686@c symbolic form, as they usually would be displayed.
7687@smallexample
7688(@value{GDBP}) info line m4_changequote
7689Line 895 of "builtin.c" starts at pc 0x634c and ends at 0x6350.
7690@end smallexample
7691
7692@noindent
7693@cindex code address and its source line
7694We can also inquire (using @code{*@var{addr}} as the form for
7695@var{linespec}) what source line covers a particular address:
7696@smallexample
7697(@value{GDBP}) info line *0x63ff
7698Line 926 of "builtin.c" starts at pc 0x63e4 and ends at 0x6404.
7699@end smallexample
7700
7701@cindex @code{$_} and @code{info line}
7702@cindex @code{x} command, default address
7703@kindex x@r{(examine), and} info line
7704After @code{info line}, the default address for the @code{x} command
7705is changed to the starting address of the line, so that @samp{x/i} is
7706sufficient to begin examining the machine code (@pxref{Memory,
7707,Examining Memory}). Also, this address is saved as the value of the
7708convenience variable @code{$_} (@pxref{Convenience Vars, ,Convenience
7709Variables}).
7710
7711@table @code
7712@kindex disassemble
7713@cindex assembly instructions
7714@cindex instructions, assembly
7715@cindex machine instructions
7716@cindex listing machine instructions
7717@item disassemble
7718@itemx disassemble /m
7719@itemx disassemble /r
7720This specialized command dumps a range of memory as machine
7721instructions. It can also print mixed source+disassembly by specifying
7722the @code{/m} modifier and print the raw instructions in hex as well as
7723in symbolic form by specifying the @code{/r}.
7724The default memory range is the function surrounding the
7725program counter of the selected frame. A single argument to this
7726command is a program counter value; @value{GDBN} dumps the function
7727surrounding this value. When two arguments are given, they should
7728be separated by a comma, possibly surrounded by whitespace. The
7729arguments specify a range of addresses to dump, in one of two forms:
7730
7731@table @code
7732@item @var{start},@var{end}
7733the addresses from @var{start} (inclusive) to @var{end} (exclusive)
7734@item @var{start},+@var{length}
7735the addresses from @var{start} (inclusive) to
7736@code{@var{start}+@var{length}} (exclusive).
7737@end table
7738
7739@noindent
7740When 2 arguments are specified, the name of the function is also
7741printed (since there could be several functions in the given range).
7742
7743The argument(s) can be any expression yielding a numeric value, such as
7744@samp{0x32c4}, @samp{&main+10} or @samp{$pc - 8}.
7745
7746If the range of memory being disassembled contains current program counter,
7747the instruction at that location is shown with a @code{=>} marker.
7748@end table
7749
7750The following example shows the disassembly of a range of addresses of
7751HP PA-RISC 2.0 code:
7752
7753@smallexample
7754(@value{GDBP}) disas 0x32c4, 0x32e4
7755Dump of assembler code from 0x32c4 to 0x32e4:
7756 0x32c4 <main+204>: addil 0,dp
7757 0x32c8 <main+208>: ldw 0x22c(sr0,r1),r26
7758 0x32cc <main+212>: ldil 0x3000,r31
7759 0x32d0 <main+216>: ble 0x3f8(sr4,r31)
7760 0x32d4 <main+220>: ldo 0(r31),rp
7761 0x32d8 <main+224>: addil -0x800,dp
7762 0x32dc <main+228>: ldo 0x588(r1),r26
7763 0x32e0 <main+232>: ldil 0x3000,r31
7764End of assembler dump.
7765@end smallexample
7766
7767Here is an example showing mixed source+assembly for Intel x86, when the
7768program is stopped just after function prologue:
7769
7770@smallexample
7771(@value{GDBP}) disas /m main
7772Dump of assembler code for function main:
77735 @{
7774 0x08048330 <+0>: push %ebp
7775 0x08048331 <+1>: mov %esp,%ebp
7776 0x08048333 <+3>: sub $0x8,%esp
7777 0x08048336 <+6>: and $0xfffffff0,%esp
7778 0x08048339 <+9>: sub $0x10,%esp
7779
77806 printf ("Hello.\n");
7781=> 0x0804833c <+12>: movl $0x8048440,(%esp)
7782 0x08048343 <+19>: call 0x8048284 <puts@@plt>
7783
77847 return 0;
77858 @}
7786 0x08048348 <+24>: mov $0x0,%eax
7787 0x0804834d <+29>: leave
7788 0x0804834e <+30>: ret
7789
7790End of assembler dump.
7791@end smallexample
7792
7793Here is another example showing raw instructions in hex for AMD x86-64,
7794
7795@smallexample
7796(gdb) disas /r 0x400281,+10
7797Dump of assembler code from 0x400281 to 0x40028b:
7798 0x0000000000400281: 38 36 cmp %dh,(%rsi)
7799 0x0000000000400283: 2d 36 34 2e 73 sub $0x732e3436,%eax
7800 0x0000000000400288: 6f outsl %ds:(%rsi),(%dx)
7801 0x0000000000400289: 2e 32 00 xor %cs:(%rax),%al
7802End of assembler dump.
7803@end smallexample
7804
7805Addresses cannot be specified as a linespec (@pxref{Specify Location}).
7806So, for example, if you want to disassemble function @code{bar}
7807in file @file{foo.c}, you must type @samp{disassemble 'foo.c'::bar}
7808and not @samp{disassemble foo.c:bar}.
7809
7810Some architectures have more than one commonly-used set of instruction
7811mnemonics or other syntax.
7812
7813For programs that were dynamically linked and use shared libraries,
7814instructions that call functions or branch to locations in the shared
7815libraries might show a seemingly bogus location---it's actually a
7816location of the relocation table. On some architectures, @value{GDBN}
7817might be able to resolve these to actual function names.
7818
7819@table @code
7820@kindex set disassembly-flavor
7821@cindex Intel disassembly flavor
7822@cindex AT&T disassembly flavor
7823@item set disassembly-flavor @var{instruction-set}
7824Select the instruction set to use when disassembling the
7825program via the @code{disassemble} or @code{x/i} commands.
7826
7827Currently this command is only defined for the Intel x86 family. You
7828can set @var{instruction-set} to either @code{intel} or @code{att}.
7829The default is @code{att}, the AT&T flavor used by default by Unix
7830assemblers for x86-based targets.
7831
7832@kindex show disassembly-flavor
7833@item show disassembly-flavor
7834Show the current setting of the disassembly flavor.
7835@end table
7836
7837@table @code
7838@kindex set disassemble-next-line
7839@kindex show disassemble-next-line
7840@item set disassemble-next-line
7841@itemx show disassemble-next-line
7842Control whether or not @value{GDBN} will disassemble the next source
7843line or instruction when execution stops. If ON, @value{GDBN} will
7844display disassembly of the next source line when execution of the
7845program being debugged stops. This is @emph{in addition} to
7846displaying the source line itself, which @value{GDBN} always does if
7847possible. If the next source line cannot be displayed for some reason
7848(e.g., if @value{GDBN} cannot find the source file, or there's no line
7849info in the debug info), @value{GDBN} will display disassembly of the
7850next @emph{instruction} instead of showing the next source line. If
7851AUTO, @value{GDBN} will display disassembly of next instruction only
7852if the source line cannot be displayed. This setting causes
7853@value{GDBN} to display some feedback when you step through a function
7854with no line info or whose source file is unavailable. The default is
7855OFF, which means never display the disassembly of the next line or
7856instruction.
7857@end table
7858
7859
7860@node Data
7861@chapter Examining Data
7862
7863@cindex printing data
7864@cindex examining data
7865@kindex print
7866@kindex inspect
7867The usual way to examine data in your program is with the @code{print}
7868command (abbreviated @code{p}), or its synonym @code{inspect}. It
7869evaluates and prints the value of an expression of the language your
7870program is written in (@pxref{Languages, ,Using @value{GDBN} with
7871Different Languages}). It may also print the expression using a
7872Python-based pretty-printer (@pxref{Pretty Printing}).
7873
7874@table @code
7875@item print @var{expr}
7876@itemx print /@var{f} @var{expr}
7877@var{expr} is an expression (in the source language). By default the
7878value of @var{expr} is printed in a format appropriate to its data type;
7879you can choose a different format by specifying @samp{/@var{f}}, where
7880@var{f} is a letter specifying the format; see @ref{Output Formats,,Output
7881Formats}.
7882
7883@item print
7884@itemx print /@var{f}
7885@cindex reprint the last value
7886If you omit @var{expr}, @value{GDBN} displays the last value again (from the
7887@dfn{value history}; @pxref{Value History, ,Value History}). This allows you to
7888conveniently inspect the same value in an alternative format.
7889@end table
7890
7891A more low-level way of examining data is with the @code{x} command.
7892It examines data in memory at a specified address and prints it in a
7893specified format. @xref{Memory, ,Examining Memory}.
7894
7895If you are interested in information about types, or about how the
7896fields of a struct or a class are declared, use the @code{ptype @var{exp}}
7897command rather than @code{print}. @xref{Symbols, ,Examining the Symbol
7898Table}.
7899
7900@cindex exploring hierarchical data structures
7901@kindex explore
7902Another way of examining values of expressions and type information is
7903through the Python extension command @code{explore} (available only if
7904the @value{GDBN} build is configured with @code{--with-python}). It
7905offers an interactive way to start at the highest level (or, the most
7906abstract level) of the data type of an expression (or, the data type
7907itself) and explore all the way down to leaf scalar values/fields
7908embedded in the higher level data types.
7909
7910@table @code
7911@item explore @var{arg}
7912@var{arg} is either an expression (in the source language), or a type
7913visible in the current context of the program being debugged.
7914@end table
7915
7916The working of the @code{explore} command can be illustrated with an
7917example. If a data type @code{struct ComplexStruct} is defined in your
7918C program as
7919
7920@smallexample
7921struct SimpleStruct
7922@{
7923 int i;
7924 double d;
7925@};
7926
7927struct ComplexStruct
7928@{
7929 struct SimpleStruct *ss_p;
7930 int arr[10];
7931@};
7932@end smallexample
7933
7934@noindent
7935followed by variable declarations as
7936
7937@smallexample
7938struct SimpleStruct ss = @{ 10, 1.11 @};
7939struct ComplexStruct cs = @{ &ss, @{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 @} @};
7940@end smallexample
7941
7942@noindent
7943then, the value of the variable @code{cs} can be explored using the
7944@code{explore} command as follows.
7945
7946@smallexample
7947(gdb) explore cs
7948The value of `cs' is a struct/class of type `struct ComplexStruct' with
7949the following fields:
7950
7951 ss_p = <Enter 0 to explore this field of type `struct SimpleStruct *'>
7952 arr = <Enter 1 to explore this field of type `int [10]'>
7953
7954Enter the field number of choice:
7955@end smallexample
7956
7957@noindent
7958Since the fields of @code{cs} are not scalar values, you are being
7959prompted to chose the field you want to explore. Let's say you choose
7960the field @code{ss_p} by entering @code{0}. Then, since this field is a
7961pointer, you will be asked if it is pointing to a single value. From
7962the declaration of @code{cs} above, it is indeed pointing to a single
7963value, hence you enter @code{y}. If you enter @code{n}, then you will
7964be asked if it were pointing to an array of values, in which case this
7965field will be explored as if it were an array.
7966
7967@smallexample
7968`cs.ss_p' is a pointer to a value of type `struct SimpleStruct'
7969Continue exploring it as a pointer to a single value [y/n]: y
7970The value of `*(cs.ss_p)' is a struct/class of type `struct
7971SimpleStruct' with the following fields:
7972
7973 i = 10 .. (Value of type `int')
7974 d = 1.1100000000000001 .. (Value of type `double')
7975
7976Press enter to return to parent value:
7977@end smallexample
7978
7979@noindent
7980If the field @code{arr} of @code{cs} was chosen for exploration by
7981entering @code{1} earlier, then since it is as array, you will be
7982prompted to enter the index of the element in the array that you want
7983to explore.
7984
7985@smallexample
7986`cs.arr' is an array of `int'.
7987Enter the index of the element you want to explore in `cs.arr': 5
7988
7989`(cs.arr)[5]' is a scalar value of type `int'.
7990
7991(cs.arr)[5] = 4
7992
7993Press enter to return to parent value:
7994@end smallexample
7995
7996In general, at any stage of exploration, you can go deeper towards the
7997leaf values by responding to the prompts appropriately, or hit the
7998return key to return to the enclosing data structure (the @i{higher}
7999level data structure).
8000
8001Similar to exploring values, you can use the @code{explore} command to
8002explore types. Instead of specifying a value (which is typically a
8003variable name or an expression valid in the current context of the
8004program being debugged), you specify a type name. If you consider the
8005same example as above, your can explore the type
8006@code{struct ComplexStruct} by passing the argument
8007@code{struct ComplexStruct} to the @code{explore} command.
8008
8009@smallexample
8010(gdb) explore struct ComplexStruct
8011@end smallexample
8012
8013@noindent
8014By responding to the prompts appropriately in the subsequent interactive
8015session, you can explore the type @code{struct ComplexStruct} in a
8016manner similar to how the value @code{cs} was explored in the above
8017example.
8018
8019The @code{explore} command also has two sub-commands,
8020@code{explore value} and @code{explore type}. The former sub-command is
8021a way to explicitly specify that value exploration of the argument is
8022being invoked, while the latter is a way to explicitly specify that type
8023exploration of the argument is being invoked.
8024
8025@table @code
8026@item explore value @var{expr}
8027@cindex explore value
8028This sub-command of @code{explore} explores the value of the
8029expression @var{expr} (if @var{expr} is an expression valid in the
8030current context of the program being debugged). The behavior of this
8031command is identical to that of the behavior of the @code{explore}
8032command being passed the argument @var{expr}.
8033
8034@item explore type @var{arg}
8035@cindex explore type
8036This sub-command of @code{explore} explores the type of @var{arg} (if
8037@var{arg} is a type visible in the current context of program being
8038debugged), or the type of the value/expression @var{arg} (if @var{arg}
8039is an expression valid in the current context of the program being
8040debugged). If @var{arg} is a type, then the behavior of this command is
8041identical to that of the @code{explore} command being passed the
8042argument @var{arg}. If @var{arg} is an expression, then the behavior of
8043this command will be identical to that of the @code{explore} command
8044being passed the type of @var{arg} as the argument.
8045@end table
8046
8047@menu
8048* Expressions:: Expressions
8049* Ambiguous Expressions:: Ambiguous Expressions
8050* Variables:: Program variables
8051* Arrays:: Artificial arrays
8052* Output Formats:: Output formats
8053* Memory:: Examining memory
8054* Auto Display:: Automatic display
8055* Print Settings:: Print settings
8056* Pretty Printing:: Python pretty printing
8057* Value History:: Value history
8058* Convenience Vars:: Convenience variables
8059* Convenience Funs:: Convenience functions
8060* Registers:: Registers
8061* Floating Point Hardware:: Floating point hardware
8062* Vector Unit:: Vector Unit
8063* OS Information:: Auxiliary data provided by operating system
8064* Memory Region Attributes:: Memory region attributes
8065* Dump/Restore Files:: Copy between memory and a file
8066* Core File Generation:: Cause a program dump its core
8067* Character Sets:: Debugging programs that use a different
8068 character set than GDB does
8069* Caching Target Data:: Data caching for targets
8070* Searching Memory:: Searching memory for a sequence of bytes
8071@end menu
8072
8073@node Expressions
8074@section Expressions
8075
8076@cindex expressions
8077@code{print} and many other @value{GDBN} commands accept an expression and
8078compute its value. Any kind of constant, variable or operator defined
8079by the programming language you are using is valid in an expression in
8080@value{GDBN}. This includes conditional expressions, function calls,
8081casts, and string constants. It also includes preprocessor macros, if
8082you compiled your program to include this information; see
8083@ref{Compilation}.
8084
8085@cindex arrays in expressions
8086@value{GDBN} supports array constants in expressions input by
8087the user. The syntax is @{@var{element}, @var{element}@dots{}@}. For example,
8088you can use the command @code{print @{1, 2, 3@}} to create an array
8089of three integers. If you pass an array to a function or assign it
8090to a program variable, @value{GDBN} copies the array to memory that
8091is @code{malloc}ed in the target program.
8092
8093Because C is so widespread, most of the expressions shown in examples in
8094this manual are in C. @xref{Languages, , Using @value{GDBN} with Different
8095Languages}, for information on how to use expressions in other
8096languages.
8097
8098In this section, we discuss operators that you can use in @value{GDBN}
8099expressions regardless of your programming language.
8100
8101@cindex casts, in expressions
8102Casts are supported in all languages, not just in C, because it is so
8103useful to cast a number into a pointer in order to examine a structure
8104at that address in memory.
8105@c FIXME: casts supported---Mod2 true?
8106
8107@value{GDBN} supports these operators, in addition to those common
8108to programming languages:
8109
8110@table @code
8111@item @@
8112@samp{@@} is a binary operator for treating parts of memory as arrays.
8113@xref{Arrays, ,Artificial Arrays}, for more information.
8114
8115@item ::
8116@samp{::} allows you to specify a variable in terms of the file or
8117function where it is defined. @xref{Variables, ,Program Variables}.
8118
8119@cindex @{@var{type}@}
8120@cindex type casting memory
8121@cindex memory, viewing as typed object
8122@cindex casts, to view memory
8123@item @{@var{type}@} @var{addr}
8124Refers to an object of type @var{type} stored at address @var{addr} in
8125memory. @var{addr} may be any expression whose value is an integer or
8126pointer (but parentheses are required around binary operators, just as in
8127a cast). This construct is allowed regardless of what kind of data is
8128normally supposed to reside at @var{addr}.
8129@end table
8130
8131@node Ambiguous Expressions
8132@section Ambiguous Expressions
8133@cindex ambiguous expressions
8134
8135Expressions can sometimes contain some ambiguous elements. For instance,
8136some programming languages (notably Ada, C@t{++} and Objective-C) permit
8137a single function name to be defined several times, for application in
8138different contexts. This is called @dfn{overloading}. Another example
8139involving Ada is generics. A @dfn{generic package} is similar to C@t{++}
8140templates and is typically instantiated several times, resulting in
8141the same function name being defined in different contexts.
8142
8143In some cases and depending on the language, it is possible to adjust
8144the expression to remove the ambiguity. For instance in C@t{++}, you
8145can specify the signature of the function you want to break on, as in
8146@kbd{break @var{function}(@var{types})}. In Ada, using the fully
8147qualified name of your function often makes the expression unambiguous
8148as well.
8149
8150When an ambiguity that needs to be resolved is detected, the debugger
8151has the capability to display a menu of numbered choices for each
8152possibility, and then waits for the selection with the prompt @samp{>}.
8153The first option is always @samp{[0] cancel}, and typing @kbd{0 @key{RET}}
8154aborts the current command. If the command in which the expression was
8155used allows more than one choice to be selected, the next option in the
8156menu is @samp{[1] all}, and typing @kbd{1 @key{RET}} selects all possible
8157choices.
8158
8159For example, the following session excerpt shows an attempt to set a
8160breakpoint at the overloaded symbol @code{String::after}.
8161We choose three particular definitions of that function name:
8162
8163@c FIXME! This is likely to change to show arg type lists, at least
8164@smallexample
8165@group
8166(@value{GDBP}) b String::after
8167[0] cancel
8168[1] all
8169[2] file:String.cc; line number:867
8170[3] file:String.cc; line number:860
8171[4] file:String.cc; line number:875
8172[5] file:String.cc; line number:853
8173[6] file:String.cc; line number:846
8174[7] file:String.cc; line number:735
8175> 2 4 6
8176Breakpoint 1 at 0xb26c: file String.cc, line 867.
8177Breakpoint 2 at 0xb344: file String.cc, line 875.
8178Breakpoint 3 at 0xafcc: file String.cc, line 846.
8179Multiple breakpoints were set.
8180Use the "delete" command to delete unwanted
8181 breakpoints.
8182(@value{GDBP})
8183@end group
8184@end smallexample
8185
8186@table @code
8187@kindex set multiple-symbols
8188@item set multiple-symbols @var{mode}
8189@cindex multiple-symbols menu
8190
8191This option allows you to adjust the debugger behavior when an expression
8192is ambiguous.
8193
8194By default, @var{mode} is set to @code{all}. If the command with which
8195the expression is used allows more than one choice, then @value{GDBN}
8196automatically selects all possible choices. For instance, inserting
8197a breakpoint on a function using an ambiguous name results in a breakpoint
8198inserted on each possible match. However, if a unique choice must be made,
8199then @value{GDBN} uses the menu to help you disambiguate the expression.
8200For instance, printing the address of an overloaded function will result
8201in the use of the menu.
8202
8203When @var{mode} is set to @code{ask}, the debugger always uses the menu
8204when an ambiguity is detected.
8205
8206Finally, when @var{mode} is set to @code{cancel}, the debugger reports
8207an error due to the ambiguity and the command is aborted.
8208
8209@kindex show multiple-symbols
8210@item show multiple-symbols
8211Show the current value of the @code{multiple-symbols} setting.
8212@end table
8213
8214@node Variables
8215@section Program Variables
8216
8217The most common kind of expression to use is the name of a variable
8218in your program.
8219
8220Variables in expressions are understood in the selected stack frame
8221(@pxref{Selection, ,Selecting a Frame}); they must be either:
8222
8223@itemize @bullet
8224@item
8225global (or file-static)
8226@end itemize
8227
8228@noindent or
8229
8230@itemize @bullet
8231@item
8232visible according to the scope rules of the
8233programming language from the point of execution in that frame
8234@end itemize
8235
8236@noindent This means that in the function
8237
8238@smallexample
8239foo (a)
8240 int a;
8241@{
8242 bar (a);
8243 @{
8244 int b = test ();
8245 bar (b);
8246 @}
8247@}
8248@end smallexample
8249
8250@noindent
8251you can examine and use the variable @code{a} whenever your program is
8252executing within the function @code{foo}, but you can only use or
8253examine the variable @code{b} while your program is executing inside
8254the block where @code{b} is declared.
8255
8256@cindex variable name conflict
8257There is an exception: you can refer to a variable or function whose
8258scope is a single source file even if the current execution point is not
8259in this file. But it is possible to have more than one such variable or
8260function with the same name (in different source files). If that
8261happens, referring to that name has unpredictable effects. If you wish,
8262you can specify a static variable in a particular function or file by
8263using the colon-colon (@code{::}) notation:
8264
8265@cindex colon-colon, context for variables/functions
8266@ifnotinfo
8267@c info cannot cope with a :: index entry, but why deprive hard copy readers?
8268@cindex @code{::}, context for variables/functions
8269@end ifnotinfo
8270@smallexample
8271@var{file}::@var{variable}
8272@var{function}::@var{variable}
8273@end smallexample
8274
8275@noindent
8276Here @var{file} or @var{function} is the name of the context for the
8277static @var{variable}. In the case of file names, you can use quotes to
8278make sure @value{GDBN} parses the file name as a single word---for example,
8279to print a global value of @code{x} defined in @file{f2.c}:
8280
8281@smallexample
8282(@value{GDBP}) p 'f2.c'::x
8283@end smallexample
8284
8285The @code{::} notation is normally used for referring to
8286static variables, since you typically disambiguate uses of local variables
8287in functions by selecting the appropriate frame and using the
8288simple name of the variable. However, you may also use this notation
8289to refer to local variables in frames enclosing the selected frame:
8290
8291@smallexample
8292void
8293foo (int a)
8294@{
8295 if (a < 10)
8296 bar (a);
8297 else
8298 process (a); /* Stop here */
8299@}
8300
8301int
8302bar (int a)
8303@{
8304 foo (a + 5);
8305@}
8306@end smallexample
8307
8308@noindent
8309For example, if there is a breakpoint at the commented line,
8310here is what you might see
8311when the program stops after executing the call @code{bar(0)}:
8312
8313@smallexample
8314(@value{GDBP}) p a
8315$1 = 10
8316(@value{GDBP}) p bar::a
8317$2 = 5
8318(@value{GDBP}) up 2
8319#2 0x080483d0 in foo (a=5) at foobar.c:12
8320(@value{GDBP}) p a
8321$3 = 5
8322(@value{GDBP}) p bar::a
8323$4 = 0
8324@end smallexample
8325
8326@cindex C@t{++} scope resolution
8327These uses of @samp{::} are very rarely in conflict with the very
8328similar use of the same notation in C@t{++}. When they are in
8329conflict, the C@t{++} meaning takes precedence; however, this can be
8330overridden by quoting the file or function name with single quotes.
8331
8332For example, suppose the program is stopped in a method of a class
8333that has a field named @code{includefile}, and there is also an
8334include file named @file{includefile} that defines a variable,
8335@code{some_global}.
8336
8337@smallexample
8338(@value{GDBP}) p includefile
8339$1 = 23
8340(@value{GDBP}) p includefile::some_global
8341A syntax error in expression, near `'.
8342(@value{GDBP}) p 'includefile'::some_global
8343$2 = 27
8344@end smallexample
8345
8346@cindex wrong values
8347@cindex variable values, wrong
8348@cindex function entry/exit, wrong values of variables
8349@cindex optimized code, wrong values of variables
8350@quotation
8351@emph{Warning:} Occasionally, a local variable may appear to have the
8352wrong value at certain points in a function---just after entry to a new
8353scope, and just before exit.
8354@end quotation
8355You may see this problem when you are stepping by machine instructions.
8356This is because, on most machines, it takes more than one instruction to
8357set up a stack frame (including local variable definitions); if you are
8358stepping by machine instructions, variables may appear to have the wrong
8359values until the stack frame is completely built. On exit, it usually
8360also takes more than one machine instruction to destroy a stack frame;
8361after you begin stepping through that group of instructions, local
8362variable definitions may be gone.
8363
8364This may also happen when the compiler does significant optimizations.
8365To be sure of always seeing accurate values, turn off all optimization
8366when compiling.
8367
8368@cindex ``No symbol "foo" in current context''
8369Another possible effect of compiler optimizations is to optimize
8370unused variables out of existence, or assign variables to registers (as
8371opposed to memory addresses). Depending on the support for such cases
8372offered by the debug info format used by the compiler, @value{GDBN}
8373might not be able to display values for such local variables. If that
8374happens, @value{GDBN} will print a message like this:
8375
8376@smallexample
8377No symbol "foo" in current context.
8378@end smallexample
8379
8380To solve such problems, either recompile without optimizations, or use a
8381different debug info format, if the compiler supports several such
8382formats. @xref{Compilation}, for more information on choosing compiler
8383options. @xref{C, ,C and C@t{++}}, for more information about debug
8384info formats that are best suited to C@t{++} programs.
8385
8386If you ask to print an object whose contents are unknown to
8387@value{GDBN}, e.g., because its data type is not completely specified
8388by the debug information, @value{GDBN} will say @samp{<incomplete
8389type>}. @xref{Symbols, incomplete type}, for more about this.
8390
8391If you append @kbd{@@entry} string to a function parameter name you get its
8392value at the time the function got called. If the value is not available an
8393error message is printed. Entry values are available only with some compilers.
8394Entry values are normally also printed at the function parameter list according
8395to @ref{set print entry-values}.
8396
8397@smallexample
8398Breakpoint 1, d (i=30) at gdb.base/entry-value.c:29
839929 i++;
8400(gdb) next
840130 e (i);
8402(gdb) print i
8403$1 = 31
8404(gdb) print i@@entry
8405$2 = 30
8406@end smallexample
8407
8408Strings are identified as arrays of @code{char} values without specified
8409signedness. Arrays of either @code{signed char} or @code{unsigned char} get
8410printed as arrays of 1 byte sized integers. @code{-fsigned-char} or
8411@code{-funsigned-char} @value{NGCC} options have no effect as @value{GDBN}
8412defines literal string type @code{"char"} as @code{char} without a sign.
8413For program code
8414
8415@smallexample
8416char var0[] = "A";
8417signed char var1[] = "A";
8418@end smallexample
8419
8420You get during debugging
8421@smallexample
8422(gdb) print var0
8423$1 = "A"
8424(gdb) print var1
8425$2 = @{65 'A', 0 '\0'@}
8426@end smallexample
8427
8428@node Arrays
8429@section Artificial Arrays
8430
8431@cindex artificial array
8432@cindex arrays
8433@kindex @@@r{, referencing memory as an array}
8434It is often useful to print out several successive objects of the
8435same type in memory; a section of an array, or an array of
8436dynamically determined size for which only a pointer exists in the
8437program.
8438
8439You can do this by referring to a contiguous span of memory as an
8440@dfn{artificial array}, using the binary operator @samp{@@}. The left
8441operand of @samp{@@} should be the first element of the desired array
8442and be an individual object. The right operand should be the desired length
8443of the array. The result is an array value whose elements are all of
8444the type of the left argument. The first element is actually the left
8445argument; the second element comes from bytes of memory immediately
8446following those that hold the first element, and so on. Here is an
8447example. If a program says
8448
8449@smallexample
8450int *array = (int *) malloc (len * sizeof (int));
8451@end smallexample
8452
8453@noindent
8454you can print the contents of @code{array} with
8455
8456@smallexample
8457p *array@@len
8458@end smallexample
8459
8460The left operand of @samp{@@} must reside in memory. Array values made
8461with @samp{@@} in this way behave just like other arrays in terms of
8462subscripting, and are coerced to pointers when used in expressions.
8463Artificial arrays most often appear in expressions via the value history
8464(@pxref{Value History, ,Value History}), after printing one out.
8465
8466Another way to create an artificial array is to use a cast.
8467This re-interprets a value as if it were an array.
8468The value need not be in memory:
8469@smallexample
8470(@value{GDBP}) p/x (short[2])0x12345678
8471$1 = @{0x1234, 0x5678@}
8472@end smallexample
8473
8474As a convenience, if you leave the array length out (as in
8475@samp{(@var{type}[])@var{value}}) @value{GDBN} calculates the size to fill
8476the value (as @samp{sizeof(@var{value})/sizeof(@var{type})}:
8477@smallexample
8478(@value{GDBP}) p/x (short[])0x12345678
8479$2 = @{0x1234, 0x5678@}
8480@end smallexample
8481
8482Sometimes the artificial array mechanism is not quite enough; in
8483moderately complex data structures, the elements of interest may not
8484actually be adjacent---for example, if you are interested in the values
8485of pointers in an array. One useful work-around in this situation is
8486to use a convenience variable (@pxref{Convenience Vars, ,Convenience
8487Variables}) as a counter in an expression that prints the first
8488interesting value, and then repeat that expression via @key{RET}. For
8489instance, suppose you have an array @code{dtab} of pointers to
8490structures, and you are interested in the values of a field @code{fv}
8491in each structure. Here is an example of what you might type:
8492
8493@smallexample
8494set $i = 0
8495p dtab[$i++]->fv
8496@key{RET}
8497@key{RET}
8498@dots{}
8499@end smallexample
8500
8501@node Output Formats
8502@section Output Formats
8503
8504@cindex formatted output
8505@cindex output formats
8506By default, @value{GDBN} prints a value according to its data type. Sometimes
8507this is not what you want. For example, you might want to print a number
8508in hex, or a pointer in decimal. Or you might want to view data in memory
8509at a certain address as a character string or as an instruction. To do
8510these things, specify an @dfn{output format} when you print a value.
8511
8512The simplest use of output formats is to say how to print a value
8513already computed. This is done by starting the arguments of the
8514@code{print} command with a slash and a format letter. The format
8515letters supported are:
8516
8517@table @code
8518@item x
8519Regard the bits of the value as an integer, and print the integer in
8520hexadecimal.
8521
8522@item d
8523Print as integer in signed decimal.
8524
8525@item u
8526Print as integer in unsigned decimal.
8527
8528@item o
8529Print as integer in octal.
8530
8531@item t
8532Print as integer in binary. The letter @samp{t} stands for ``two''.
8533@footnote{@samp{b} cannot be used because these format letters are also
8534used with the @code{x} command, where @samp{b} stands for ``byte'';
8535see @ref{Memory,,Examining Memory}.}
8536
8537@item a
8538@cindex unknown address, locating
8539@cindex locate address
8540Print as an address, both absolute in hexadecimal and as an offset from
8541the nearest preceding symbol. You can use this format used to discover
8542where (in what function) an unknown address is located:
8543
8544@smallexample
8545(@value{GDBP}) p/a 0x54320
8546$3 = 0x54320 <_initialize_vx+396>
8547@end smallexample
8548
8549@noindent
8550The command @code{info symbol 0x54320} yields similar results.
8551@xref{Symbols, info symbol}.
8552
8553@item c
8554Regard as an integer and print it as a character constant. This
8555prints both the numerical value and its character representation. The
8556character representation is replaced with the octal escape @samp{\nnn}
8557for characters outside the 7-bit @sc{ascii} range.
8558
8559Without this format, @value{GDBN} displays @code{char},
8560@w{@code{unsigned char}}, and @w{@code{signed char}} data as character
8561constants. Single-byte members of vectors are displayed as integer
8562data.
8563
8564@item f
8565Regard the bits of the value as a floating point number and print
8566using typical floating point syntax.
8567
8568@item s
8569@cindex printing strings
8570@cindex printing byte arrays
8571Regard as a string, if possible. With this format, pointers to single-byte
8572data are displayed as null-terminated strings and arrays of single-byte data
8573are displayed as fixed-length strings. Other values are displayed in their
8574natural types.
8575
8576Without this format, @value{GDBN} displays pointers to and arrays of
8577@code{char}, @w{@code{unsigned char}}, and @w{@code{signed char}} as
8578strings. Single-byte members of a vector are displayed as an integer
8579array.
8580
8581@item z
8582Like @samp{x} formatting, the value is treated as an integer and
8583printed as hexadecimal, but leading zeros are printed to pad the value
8584to the size of the integer type.
8585
8586@item r
8587@cindex raw printing
8588Print using the @samp{raw} formatting. By default, @value{GDBN} will
8589use a Python-based pretty-printer, if one is available (@pxref{Pretty
8590Printing}). This typically results in a higher-level display of the
8591value's contents. The @samp{r} format bypasses any Python
8592pretty-printer which might exist.
8593@end table
8594
8595For example, to print the program counter in hex (@pxref{Registers}), type
8596
8597@smallexample
8598p/x $pc
8599@end smallexample
8600
8601@noindent
8602Note that no space is required before the slash; this is because command
8603names in @value{GDBN} cannot contain a slash.
8604
8605To reprint the last value in the value history with a different format,
8606you can use the @code{print} command with just a format and no
8607expression. For example, @samp{p/x} reprints the last value in hex.
8608
8609@node Memory
8610@section Examining Memory
8611
8612You can use the command @code{x} (for ``examine'') to examine memory in
8613any of several formats, independently of your program's data types.
8614
8615@cindex examining memory
8616@table @code
8617@kindex x @r{(examine memory)}
8618@item x/@var{nfu} @var{addr}
8619@itemx x @var{addr}
8620@itemx x
8621Use the @code{x} command to examine memory.
8622@end table
8623
8624@var{n}, @var{f}, and @var{u} are all optional parameters that specify how
8625much memory to display and how to format it; @var{addr} is an
8626expression giving the address where you want to start displaying memory.
8627If you use defaults for @var{nfu}, you need not type the slash @samp{/}.
8628Several commands set convenient defaults for @var{addr}.
8629
8630@table @r
8631@item @var{n}, the repeat count
8632The repeat count is a decimal integer; the default is 1. It specifies
8633how much memory (counting by units @var{u}) to display.
8634@c This really is **decimal**; unaffected by 'set radix' as of GDB
8635@c 4.1.2.
8636
8637@item @var{f}, the display format
8638The display format is one of the formats used by @code{print}
8639(@samp{x}, @samp{d}, @samp{u}, @samp{o}, @samp{t}, @samp{a}, @samp{c},
8640@samp{f}, @samp{s}), and in addition @samp{i} (for machine instructions).
8641The default is @samp{x} (hexadecimal) initially. The default changes
8642each time you use either @code{x} or @code{print}.
8643
8644@item @var{u}, the unit size
8645The unit size is any of
8646
8647@table @code
8648@item b
8649Bytes.
8650@item h
8651Halfwords (two bytes).
8652@item w
8653Words (four bytes). This is the initial default.
8654@item g
8655Giant words (eight bytes).
8656@end table
8657
8658Each time you specify a unit size with @code{x}, that size becomes the
8659default unit the next time you use @code{x}. For the @samp{i} format,
8660the unit size is ignored and is normally not written. For the @samp{s} format,
8661the unit size defaults to @samp{b}, unless it is explicitly given.
8662Use @kbd{x /hs} to display 16-bit char strings and @kbd{x /ws} to display
866332-bit strings. The next use of @kbd{x /s} will again display 8-bit strings.
8664Note that the results depend on the programming language of the
8665current compilation unit. If the language is C, the @samp{s}
8666modifier will use the UTF-16 encoding while @samp{w} will use
8667UTF-32. The encoding is set by the programming language and cannot
8668be altered.
8669
8670@item @var{addr}, starting display address
8671@var{addr} is the address where you want @value{GDBN} to begin displaying
8672memory. The expression need not have a pointer value (though it may);
8673it is always interpreted as an integer address of a byte of memory.
8674@xref{Expressions, ,Expressions}, for more information on expressions. The default for
8675@var{addr} is usually just after the last address examined---but several
8676other commands also set the default address: @code{info breakpoints} (to
8677the address of the last breakpoint listed), @code{info line} (to the
8678starting address of a line), and @code{print} (if you use it to display
8679a value from memory).
8680@end table
8681
8682For example, @samp{x/3uh 0x54320} is a request to display three halfwords
8683(@code{h}) of memory, formatted as unsigned decimal integers (@samp{u}),
8684starting at address @code{0x54320}. @samp{x/4xw $sp} prints the four
8685words (@samp{w}) of memory above the stack pointer (here, @samp{$sp};
8686@pxref{Registers, ,Registers}) in hexadecimal (@samp{x}).
8687
8688Since the letters indicating unit sizes are all distinct from the
8689letters specifying output formats, you do not have to remember whether
8690unit size or format comes first; either order works. The output
8691specifications @samp{4xw} and @samp{4wx} mean exactly the same thing.
8692(However, the count @var{n} must come first; @samp{wx4} does not work.)
8693
8694Even though the unit size @var{u} is ignored for the formats @samp{s}
8695and @samp{i}, you might still want to use a count @var{n}; for example,
8696@samp{3i} specifies that you want to see three machine instructions,
8697including any operands. For convenience, especially when used with
8698the @code{display} command, the @samp{i} format also prints branch delay
8699slot instructions, if any, beyond the count specified, which immediately
8700follow the last instruction that is within the count. The command
8701@code{disassemble} gives an alternative way of inspecting machine
8702instructions; see @ref{Machine Code,,Source and Machine Code}.
8703
8704All the defaults for the arguments to @code{x} are designed to make it
8705easy to continue scanning memory with minimal specifications each time
8706you use @code{x}. For example, after you have inspected three machine
8707instructions with @samp{x/3i @var{addr}}, you can inspect the next seven
8708with just @samp{x/7}. If you use @key{RET} to repeat the @code{x} command,
8709the repeat count @var{n} is used again; the other arguments default as
8710for successive uses of @code{x}.
8711
8712When examining machine instructions, the instruction at current program
8713counter is shown with a @code{=>} marker. For example:
8714
8715@smallexample
8716(@value{GDBP}) x/5i $pc-6
8717 0x804837f <main+11>: mov %esp,%ebp
8718 0x8048381 <main+13>: push %ecx
8719 0x8048382 <main+14>: sub $0x4,%esp
8720=> 0x8048385 <main+17>: movl $0x8048460,(%esp)
8721 0x804838c <main+24>: call 0x80482d4 <puts@@plt>
8722@end smallexample
8723
8724@cindex @code{$_}, @code{$__}, and value history
8725The addresses and contents printed by the @code{x} command are not saved
8726in the value history because there is often too much of them and they
8727would get in the way. Instead, @value{GDBN} makes these values available for
8728subsequent use in expressions as values of the convenience variables
8729@code{$_} and @code{$__}. After an @code{x} command, the last address
8730examined is available for use in expressions in the convenience variable
8731@code{$_}. The contents of that address, as examined, are available in
8732the convenience variable @code{$__}.
8733
8734If the @code{x} command has a repeat count, the address and contents saved
8735are from the last memory unit printed; this is not the same as the last
8736address printed if several units were printed on the last line of output.
8737
8738@cindex remote memory comparison
8739@cindex verify remote memory image
8740When you are debugging a program running on a remote target machine
8741(@pxref{Remote Debugging}), you may wish to verify the program's image in the
8742remote machine's memory against the executable file you downloaded to
8743the target. The @code{compare-sections} command is provided for such
8744situations.
8745
8746@table @code
8747@kindex compare-sections
8748@item compare-sections @r{[}@var{section-name}@r{]}
8749Compare the data of a loadable section @var{section-name} in the
8750executable file of the program being debugged with the same section in
8751the remote machine's memory, and report any mismatches. With no
8752arguments, compares all loadable sections. This command's
8753availability depends on the target's support for the @code{"qCRC"}
8754remote request.
8755@end table
8756
8757@node Auto Display
8758@section Automatic Display
8759@cindex automatic display
8760@cindex display of expressions
8761
8762If you find that you want to print the value of an expression frequently
8763(to see how it changes), you might want to add it to the @dfn{automatic
8764display list} so that @value{GDBN} prints its value each time your program stops.
8765Each expression added to the list is given a number to identify it;
8766to remove an expression from the list, you specify that number.
8767The automatic display looks like this:
8768
8769@smallexample
87702: foo = 38
87713: bar[5] = (struct hack *) 0x3804
8772@end smallexample
8773
8774@noindent
8775This display shows item numbers, expressions and their current values. As with
8776displays you request manually using @code{x} or @code{print}, you can
8777specify the output format you prefer; in fact, @code{display} decides
8778whether to use @code{print} or @code{x} depending your format
8779specification---it uses @code{x} if you specify either the @samp{i}
8780or @samp{s} format, or a unit size; otherwise it uses @code{print}.
8781
8782@table @code
8783@kindex display
8784@item display @var{expr}
8785Add the expression @var{expr} to the list of expressions to display
8786each time your program stops. @xref{Expressions, ,Expressions}.
8787
8788@code{display} does not repeat if you press @key{RET} again after using it.
8789
8790@item display/@var{fmt} @var{expr}
8791For @var{fmt} specifying only a display format and not a size or
8792count, add the expression @var{expr} to the auto-display list but
8793arrange to display it each time in the specified format @var{fmt}.
8794@xref{Output Formats,,Output Formats}.
8795
8796@item display/@var{fmt} @var{addr}
8797For @var{fmt} @samp{i} or @samp{s}, or including a unit-size or a
8798number of units, add the expression @var{addr} as a memory address to
8799be examined each time your program stops. Examining means in effect
8800doing @samp{x/@var{fmt} @var{addr}}. @xref{Memory, ,Examining Memory}.
8801@end table
8802
8803For example, @samp{display/i $pc} can be helpful, to see the machine
8804instruction about to be executed each time execution stops (@samp{$pc}
8805is a common name for the program counter; @pxref{Registers, ,Registers}).
8806
8807@table @code
8808@kindex delete display
8809@kindex undisplay
8810@item undisplay @var{dnums}@dots{}
8811@itemx delete display @var{dnums}@dots{}
8812Remove items from the list of expressions to display. Specify the
8813numbers of the displays that you want affected with the command
8814argument @var{dnums}. It can be a single display number, one of the
8815numbers shown in the first field of the @samp{info display} display;
8816or it could be a range of display numbers, as in @code{2-4}.
8817
8818@code{undisplay} does not repeat if you press @key{RET} after using it.
8819(Otherwise you would just get the error @samp{No display number @dots{}}.)
8820
8821@kindex disable display
8822@item disable display @var{dnums}@dots{}
8823Disable the display of item numbers @var{dnums}. A disabled display
8824item is not printed automatically, but is not forgotten. It may be
8825enabled again later. Specify the numbers of the displays that you
8826want affected with the command argument @var{dnums}. It can be a
8827single display number, one of the numbers shown in the first field of
8828the @samp{info display} display; or it could be a range of display
8829numbers, as in @code{2-4}.
8830
8831@kindex enable display
8832@item enable display @var{dnums}@dots{}
8833Enable display of item numbers @var{dnums}. It becomes effective once
8834again in auto display of its expression, until you specify otherwise.
8835Specify the numbers of the displays that you want affected with the
8836command argument @var{dnums}. It can be a single display number, one
8837of the numbers shown in the first field of the @samp{info display}
8838display; or it could be a range of display numbers, as in @code{2-4}.
8839
8840@item display
8841Display the current values of the expressions on the list, just as is
8842done when your program stops.
8843
8844@kindex info display
8845@item info display
8846Print the list of expressions previously set up to display
8847automatically, each one with its item number, but without showing the
8848values. This includes disabled expressions, which are marked as such.
8849It also includes expressions which would not be displayed right now
8850because they refer to automatic variables not currently available.
8851@end table
8852
8853@cindex display disabled out of scope
8854If a display expression refers to local variables, then it does not make
8855sense outside the lexical context for which it was set up. Such an
8856expression is disabled when execution enters a context where one of its
8857variables is not defined. For example, if you give the command
8858@code{display last_char} while inside a function with an argument
8859@code{last_char}, @value{GDBN} displays this argument while your program
8860continues to stop inside that function. When it stops elsewhere---where
8861there is no variable @code{last_char}---the display is disabled
8862automatically. The next time your program stops where @code{last_char}
8863is meaningful, you can enable the display expression once again.
8864
8865@node Print Settings
8866@section Print Settings
8867
8868@cindex format options
8869@cindex print settings
8870@value{GDBN} provides the following ways to control how arrays, structures,
8871and symbols are printed.
8872
8873@noindent
8874These settings are useful for debugging programs in any language:
8875
8876@table @code
8877@kindex set print
8878@item set print address
8879@itemx set print address on
8880@cindex print/don't print memory addresses
8881@value{GDBN} prints memory addresses showing the location of stack
8882traces, structure values, pointer values, breakpoints, and so forth,
8883even when it also displays the contents of those addresses. The default
8884is @code{on}. For example, this is what a stack frame display looks like with
8885@code{set print address on}:
8886
8887@smallexample
8888@group
8889(@value{GDBP}) f
8890#0 set_quotes (lq=0x34c78 "<<", rq=0x34c88 ">>")
8891 at input.c:530
8892530 if (lquote != def_lquote)
8893@end group
8894@end smallexample
8895
8896@item set print address off
8897Do not print addresses when displaying their contents. For example,
8898this is the same stack frame displayed with @code{set print address off}:
8899
8900@smallexample
8901@group
8902(@value{GDBP}) set print addr off
8903(@value{GDBP}) f
8904#0 set_quotes (lq="<<", rq=">>") at input.c:530
8905530 if (lquote != def_lquote)
8906@end group
8907@end smallexample
8908
8909You can use @samp{set print address off} to eliminate all machine
8910dependent displays from the @value{GDBN} interface. For example, with
8911@code{print address off}, you should get the same text for backtraces on
8912all machines---whether or not they involve pointer arguments.
8913
8914@kindex show print
8915@item show print address
8916Show whether or not addresses are to be printed.
8917@end table
8918
8919When @value{GDBN} prints a symbolic address, it normally prints the
8920closest earlier symbol plus an offset. If that symbol does not uniquely
8921identify the address (for example, it is a name whose scope is a single
8922source file), you may need to clarify. One way to do this is with
8923@code{info line}, for example @samp{info line *0x4537}. Alternately,
8924you can set @value{GDBN} to print the source file and line number when
8925it prints a symbolic address:
8926
8927@table @code
8928@item set print symbol-filename on
8929@cindex source file and line of a symbol
8930@cindex symbol, source file and line
8931Tell @value{GDBN} to print the source file name and line number of a
8932symbol in the symbolic form of an address.
8933
8934@item set print symbol-filename off
8935Do not print source file name and line number of a symbol. This is the
8936default.
8937
8938@item show print symbol-filename
8939Show whether or not @value{GDBN} will print the source file name and
8940line number of a symbol in the symbolic form of an address.
8941@end table
8942
8943Another situation where it is helpful to show symbol filenames and line
8944numbers is when disassembling code; @value{GDBN} shows you the line
8945number and source file that corresponds to each instruction.
8946
8947Also, you may wish to see the symbolic form only if the address being
8948printed is reasonably close to the closest earlier symbol:
8949
8950@table @code
8951@item set print max-symbolic-offset @var{max-offset}
8952@itemx set print max-symbolic-offset unlimited
8953@cindex maximum value for offset of closest symbol
8954Tell @value{GDBN} to only display the symbolic form of an address if the
8955offset between the closest earlier symbol and the address is less than
8956@var{max-offset}. The default is @code{unlimited}, which tells @value{GDBN}
8957to always print the symbolic form of an address if any symbol precedes
8958it. Zero is equivalent to @code{unlimited}.
8959
8960@item show print max-symbolic-offset
8961Ask how large the maximum offset is that @value{GDBN} prints in a
8962symbolic address.
8963@end table
8964
8965@cindex wild pointer, interpreting
8966@cindex pointer, finding referent
8967If you have a pointer and you are not sure where it points, try
8968@samp{set print symbol-filename on}. Then you can determine the name
8969and source file location of the variable where it points, using
8970@samp{p/a @var{pointer}}. This interprets the address in symbolic form.
8971For example, here @value{GDBN} shows that a variable @code{ptt} points
8972at another variable @code{t}, defined in @file{hi2.c}:
8973
8974@smallexample
8975(@value{GDBP}) set print symbol-filename on
8976(@value{GDBP}) p/a ptt
8977$4 = 0xe008 <t in hi2.c>
8978@end smallexample
8979
8980@quotation
8981@emph{Warning:} For pointers that point to a local variable, @samp{p/a}
8982does not show the symbol name and filename of the referent, even with
8983the appropriate @code{set print} options turned on.
8984@end quotation
8985
8986You can also enable @samp{/a}-like formatting all the time using
8987@samp{set print symbol on}:
8988
8989@table @code
8990@item set print symbol on
8991Tell @value{GDBN} to print the symbol corresponding to an address, if
8992one exists.
8993
8994@item set print symbol off
8995Tell @value{GDBN} not to print the symbol corresponding to an
8996address. In this mode, @value{GDBN} will still print the symbol
8997corresponding to pointers to functions. This is the default.
8998
8999@item show print symbol
9000Show whether @value{GDBN} will display the symbol corresponding to an
9001address.
9002@end table
9003
9004Other settings control how different kinds of objects are printed:
9005
9006@table @code
9007@item set print array
9008@itemx set print array on
9009@cindex pretty print arrays
9010Pretty print arrays. This format is more convenient to read,
9011but uses more space. The default is off.
9012
9013@item set print array off
9014Return to compressed format for arrays.
9015
9016@item show print array
9017Show whether compressed or pretty format is selected for displaying
9018arrays.
9019
9020@cindex print array indexes
9021@item set print array-indexes
9022@itemx set print array-indexes on
9023Print the index of each element when displaying arrays. May be more
9024convenient to locate a given element in the array or quickly find the
9025index of a given element in that printed array. The default is off.
9026
9027@item set print array-indexes off
9028Stop printing element indexes when displaying arrays.
9029
9030@item show print array-indexes
9031Show whether the index of each element is printed when displaying
9032arrays.
9033
9034@item set print elements @var{number-of-elements}
9035@itemx set print elements unlimited
9036@cindex number of array elements to print
9037@cindex limit on number of printed array elements
9038Set a limit on how many elements of an array @value{GDBN} will print.
9039If @value{GDBN} is printing a large array, it stops printing after it has
9040printed the number of elements set by the @code{set print elements} command.
9041This limit also applies to the display of strings.
9042When @value{GDBN} starts, this limit is set to 200.
9043Setting @var{number-of-elements} to @code{unlimited} or zero means
9044that the number of elements to print is unlimited.
9045
9046@item show print elements
9047Display the number of elements of a large array that @value{GDBN} will print.
9048If the number is 0, then the printing is unlimited.
9049
9050@item set print frame-arguments @var{value}
9051@kindex set print frame-arguments
9052@cindex printing frame argument values
9053@cindex print all frame argument values
9054@cindex print frame argument values for scalars only
9055@cindex do not print frame argument values
9056This command allows to control how the values of arguments are printed
9057when the debugger prints a frame (@pxref{Frames}). The possible
9058values are:
9059
9060@table @code
9061@item all
9062The values of all arguments are printed.
9063
9064@item scalars
9065Print the value of an argument only if it is a scalar. The value of more
9066complex arguments such as arrays, structures, unions, etc, is replaced
9067by @code{@dots{}}. This is the default. Here is an example where
9068only scalar arguments are shown:
9069
9070@smallexample
9071#1 0x08048361 in call_me (i=3, s=@dots{}, ss=0xbf8d508c, u=@dots{}, e=green)
9072 at frame-args.c:23
9073@end smallexample
9074
9075@item none
9076None of the argument values are printed. Instead, the value of each argument
9077is replaced by @code{@dots{}}. In this case, the example above now becomes:
9078
9079@smallexample
9080#1 0x08048361 in call_me (i=@dots{}, s=@dots{}, ss=@dots{}, u=@dots{}, e=@dots{})
9081 at frame-args.c:23
9082@end smallexample
9083@end table
9084
9085By default, only scalar arguments are printed. This command can be used
9086to configure the debugger to print the value of all arguments, regardless
9087of their type. However, it is often advantageous to not print the value
9088of more complex parameters. For instance, it reduces the amount of
9089information printed in each frame, making the backtrace more readable.
9090Also, it improves performance when displaying Ada frames, because
9091the computation of large arguments can sometimes be CPU-intensive,
9092especially in large applications. Setting @code{print frame-arguments}
9093to @code{scalars} (the default) or @code{none} avoids this computation,
9094thus speeding up the display of each Ada frame.
9095
9096@item show print frame-arguments
9097Show how the value of arguments should be displayed when printing a frame.
9098
9099@item set print raw frame-arguments on
9100Print frame arguments in raw, non pretty-printed, form.
9101
9102@item set print raw frame-arguments off
9103Print frame arguments in pretty-printed form, if there is a pretty-printer
9104for the value (@pxref{Pretty Printing}),
9105otherwise print the value in raw form.
9106This is the default.
9107
9108@item show print raw frame-arguments
9109Show whether to print frame arguments in raw form.
9110
9111@anchor{set print entry-values}
9112@item set print entry-values @var{value}
9113@kindex set print entry-values
9114Set printing of frame argument values at function entry. In some cases
9115@value{GDBN} can determine the value of function argument which was passed by
9116the function caller, even if the value was modified inside the called function
9117and therefore is different. With optimized code, the current value could be
9118unavailable, but the entry value may still be known.
9119
9120The default value is @code{default} (see below for its description). Older
9121@value{GDBN} behaved as with the setting @code{no}. Compilers not supporting
9122this feature will behave in the @code{default} setting the same way as with the
9123@code{no} setting.
9124
9125This functionality is currently supported only by DWARF 2 debugging format and
9126the compiler has to produce @samp{DW_TAG_GNU_call_site} tags. With
9127@value{NGCC}, you need to specify @option{-O -g} during compilation, to get
9128this information.
9129
9130The @var{value} parameter can be one of the following:
9131
9132@table @code
9133@item no
9134Print only actual parameter values, never print values from function entry
9135point.
9136@smallexample
9137#0 equal (val=5)
9138#0 different (val=6)
9139#0 lost (val=<optimized out>)
9140#0 born (val=10)
9141#0 invalid (val=<optimized out>)
9142@end smallexample
9143
9144@item only
9145Print only parameter values from function entry point. The actual parameter
9146values are never printed.
9147@smallexample
9148#0 equal (val@@entry=5)
9149#0 different (val@@entry=5)
9150#0 lost (val@@entry=5)
9151#0 born (val@@entry=<optimized out>)
9152#0 invalid (val@@entry=<optimized out>)
9153@end smallexample
9154
9155@item preferred
9156Print only parameter values from function entry point. If value from function
9157entry point is not known while the actual value is known, print the actual
9158value for such parameter.
9159@smallexample
9160#0 equal (val@@entry=5)
9161#0 different (val@@entry=5)
9162#0 lost (val@@entry=5)
9163#0 born (val=10)
9164#0 invalid (val@@entry=<optimized out>)
9165@end smallexample
9166
9167@item if-needed
9168Print actual parameter values. If actual parameter value is not known while
9169value from function entry point is known, print the entry point value for such
9170parameter.
9171@smallexample
9172#0 equal (val=5)
9173#0 different (val=6)
9174#0 lost (val@@entry=5)
9175#0 born (val=10)
9176#0 invalid (val=<optimized out>)
9177@end smallexample
9178
9179@item both
9180Always print both the actual parameter value and its value from function entry
9181point, even if values of one or both are not available due to compiler
9182optimizations.
9183@smallexample
9184#0 equal (val=5, val@@entry=5)
9185#0 different (val=6, val@@entry=5)
9186#0 lost (val=<optimized out>, val@@entry=5)
9187#0 born (val=10, val@@entry=<optimized out>)
9188#0 invalid (val=<optimized out>, val@@entry=<optimized out>)
9189@end smallexample
9190
9191@item compact
9192Print the actual parameter value if it is known and also its value from
9193function entry point if it is known. If neither is known, print for the actual
9194value @code{<optimized out>}. If not in MI mode (@pxref{GDB/MI}) and if both
9195values are known and identical, print the shortened
9196@code{param=param@@entry=VALUE} notation.
9197@smallexample
9198#0 equal (val=val@@entry=5)
9199#0 different (val=6, val@@entry=5)
9200#0 lost (val@@entry=5)
9201#0 born (val=10)
9202#0 invalid (val=<optimized out>)
9203@end smallexample
9204
9205@item default
9206Always print the actual parameter value. Print also its value from function
9207entry point, but only if it is known. If not in MI mode (@pxref{GDB/MI}) and
9208if both values are known and identical, print the shortened
9209@code{param=param@@entry=VALUE} notation.
9210@smallexample
9211#0 equal (val=val@@entry=5)
9212#0 different (val=6, val@@entry=5)
9213#0 lost (val=<optimized out>, val@@entry=5)
9214#0 born (val=10)
9215#0 invalid (val=<optimized out>)
9216@end smallexample
9217@end table
9218
9219For analysis messages on possible failures of frame argument values at function
9220entry resolution see @ref{set debug entry-values}.
9221
9222@item show print entry-values
9223Show the method being used for printing of frame argument values at function
9224entry.
9225
9226@item set print repeats @var{number-of-repeats}
9227@itemx set print repeats unlimited
9228@cindex repeated array elements
9229Set the threshold for suppressing display of repeated array
9230elements. When the number of consecutive identical elements of an
9231array exceeds the threshold, @value{GDBN} prints the string
9232@code{"<repeats @var{n} times>"}, where @var{n} is the number of
9233identical repetitions, instead of displaying the identical elements
9234themselves. Setting the threshold to @code{unlimited} or zero will
9235cause all elements to be individually printed. The default threshold
9236is 10.
9237
9238@item show print repeats
9239Display the current threshold for printing repeated identical
9240elements.
9241
9242@item set print null-stop
9243@cindex @sc{null} elements in arrays
9244Cause @value{GDBN} to stop printing the characters of an array when the first
9245@sc{null} is encountered. This is useful when large arrays actually
9246contain only short strings.
9247The default is off.
9248
9249@item show print null-stop
9250Show whether @value{GDBN} stops printing an array on the first
9251@sc{null} character.
9252
9253@item set print pretty on
9254@cindex print structures in indented form
9255@cindex indentation in structure display
9256Cause @value{GDBN} to print structures in an indented format with one member
9257per line, like this:
9258
9259@smallexample
9260@group
9261$1 = @{
9262 next = 0x0,
9263 flags = @{
9264 sweet = 1,
9265 sour = 1
9266 @},
9267 meat = 0x54 "Pork"
9268@}
9269@end group
9270@end smallexample
9271
9272@item set print pretty off
9273Cause @value{GDBN} to print structures in a compact format, like this:
9274
9275@smallexample
9276@group
9277$1 = @{next = 0x0, flags = @{sweet = 1, sour = 1@}, \
9278meat = 0x54 "Pork"@}
9279@end group
9280@end smallexample
9281
9282@noindent
9283This is the default format.
9284
9285@item show print pretty
9286Show which format @value{GDBN} is using to print structures.
9287
9288@item set print sevenbit-strings on
9289@cindex eight-bit characters in strings
9290@cindex octal escapes in strings
9291Print using only seven-bit characters; if this option is set,
9292@value{GDBN} displays any eight-bit characters (in strings or
9293character values) using the notation @code{\}@var{nnn}. This setting is
9294best if you are working in English (@sc{ascii}) and you use the
9295high-order bit of characters as a marker or ``meta'' bit.
9296
9297@item set print sevenbit-strings off
9298Print full eight-bit characters. This allows the use of more
9299international character sets, and is the default.
9300
9301@item show print sevenbit-strings
9302Show whether or not @value{GDBN} is printing only seven-bit characters.
9303
9304@item set print union on
9305@cindex unions in structures, printing
9306Tell @value{GDBN} to print unions which are contained in structures
9307and other unions. This is the default setting.
9308
9309@item set print union off
9310Tell @value{GDBN} not to print unions which are contained in
9311structures and other unions. @value{GDBN} will print @code{"@{...@}"}
9312instead.
9313
9314@item show print union
9315Ask @value{GDBN} whether or not it will print unions which are contained in
9316structures and other unions.
9317
9318For example, given the declarations
9319
9320@smallexample
9321typedef enum @{Tree, Bug@} Species;
9322typedef enum @{Big_tree, Acorn, Seedling@} Tree_forms;
9323typedef enum @{Caterpillar, Cocoon, Butterfly@}
9324 Bug_forms;
9325
9326struct thing @{
9327 Species it;
9328 union @{
9329 Tree_forms tree;
9330 Bug_forms bug;
9331 @} form;
9332@};
9333
9334struct thing foo = @{Tree, @{Acorn@}@};
9335@end smallexample
9336
9337@noindent
9338with @code{set print union on} in effect @samp{p foo} would print
9339
9340@smallexample
9341$1 = @{it = Tree, form = @{tree = Acorn, bug = Cocoon@}@}
9342@end smallexample
9343
9344@noindent
9345and with @code{set print union off} in effect it would print
9346
9347@smallexample
9348$1 = @{it = Tree, form = @{...@}@}
9349@end smallexample
9350
9351@noindent
9352@code{set print union} affects programs written in C-like languages
9353and in Pascal.
9354@end table
9355
9356@need 1000
9357@noindent
9358These settings are of interest when debugging C@t{++} programs:
9359
9360@table @code
9361@cindex demangling C@t{++} names
9362@item set print demangle
9363@itemx set print demangle on
9364Print C@t{++} names in their source form rather than in the encoded
9365(``mangled'') form passed to the assembler and linker for type-safe
9366linkage. The default is on.
9367
9368@item show print demangle
9369Show whether C@t{++} names are printed in mangled or demangled form.
9370
9371@item set print asm-demangle
9372@itemx set print asm-demangle on
9373Print C@t{++} names in their source form rather than their mangled form, even
9374in assembler code printouts such as instruction disassemblies.
9375The default is off.
9376
9377@item show print asm-demangle
9378Show whether C@t{++} names in assembly listings are printed in mangled
9379or demangled form.
9380
9381@cindex C@t{++} symbol decoding style
9382@cindex symbol decoding style, C@t{++}
9383@kindex set demangle-style
9384@item set demangle-style @var{style}
9385Choose among several encoding schemes used by different compilers to
9386represent C@t{++} names. The choices for @var{style} are currently:
9387
9388@table @code
9389@item auto
9390Allow @value{GDBN} to choose a decoding style by inspecting your program.
9391This is the default.
9392
9393@item gnu
9394Decode based on the @sc{gnu} C@t{++} compiler (@code{g++}) encoding algorithm.
9395
9396@item hp
9397Decode based on the HP ANSI C@t{++} (@code{aCC}) encoding algorithm.
9398
9399@item lucid
9400Decode based on the Lucid C@t{++} compiler (@code{lcc}) encoding algorithm.
9401
9402@item arm
9403Decode using the algorithm in the @cite{C@t{++} Annotated Reference Manual}.
9404@strong{Warning:} this setting alone is not sufficient to allow
9405debugging @code{cfront}-generated executables. @value{GDBN} would
9406require further enhancement to permit that.
9407
9408@end table
9409If you omit @var{style}, you will see a list of possible formats.
9410
9411@item show demangle-style
9412Display the encoding style currently in use for decoding C@t{++} symbols.
9413
9414@item set print object
9415@itemx set print object on
9416@cindex derived type of an object, printing
9417@cindex display derived types
9418When displaying a pointer to an object, identify the @emph{actual}
9419(derived) type of the object rather than the @emph{declared} type, using
9420the virtual function table. Note that the virtual function table is
9421required---this feature can only work for objects that have run-time
9422type identification; a single virtual method in the object's declared
9423type is sufficient. Note that this setting is also taken into account when
9424working with variable objects via MI (@pxref{GDB/MI}).
9425
9426@item set print object off
9427Display only the declared type of objects, without reference to the
9428virtual function table. This is the default setting.
9429
9430@item show print object
9431Show whether actual, or declared, object types are displayed.
9432
9433@item set print static-members
9434@itemx set print static-members on
9435@cindex static members of C@t{++} objects
9436Print static members when displaying a C@t{++} object. The default is on.
9437
9438@item set print static-members off
9439Do not print static members when displaying a C@t{++} object.
9440
9441@item show print static-members
9442Show whether C@t{++} static members are printed or not.
9443
9444@item set print pascal_static-members
9445@itemx set print pascal_static-members on
9446@cindex static members of Pascal objects
9447@cindex Pascal objects, static members display
9448Print static members when displaying a Pascal object. The default is on.
9449
9450@item set print pascal_static-members off
9451Do not print static members when displaying a Pascal object.
9452
9453@item show print pascal_static-members
9454Show whether Pascal static members are printed or not.
9455
9456@c These don't work with HP ANSI C++ yet.
9457@item set print vtbl
9458@itemx set print vtbl on
9459@cindex pretty print C@t{++} virtual function tables
9460@cindex virtual functions (C@t{++}) display
9461@cindex VTBL display
9462Pretty print C@t{++} virtual function tables. The default is off.
9463(The @code{vtbl} commands do not work on programs compiled with the HP
9464ANSI C@t{++} compiler (@code{aCC}).)
9465
9466@item set print vtbl off
9467Do not pretty print C@t{++} virtual function tables.
9468
9469@item show print vtbl
9470Show whether C@t{++} virtual function tables are pretty printed, or not.
9471@end table
9472
9473@node Pretty Printing
9474@section Pretty Printing
9475
9476@value{GDBN} provides a mechanism to allow pretty-printing of values using
9477Python code. It greatly simplifies the display of complex objects. This
9478mechanism works for both MI and the CLI.
9479
9480@menu
9481* Pretty-Printer Introduction:: Introduction to pretty-printers
9482* Pretty-Printer Example:: An example pretty-printer
9483* Pretty-Printer Commands:: Pretty-printer commands
9484@end menu
9485
9486@node Pretty-Printer Introduction
9487@subsection Pretty-Printer Introduction
9488
9489When @value{GDBN} prints a value, it first sees if there is a pretty-printer
9490registered for the value. If there is then @value{GDBN} invokes the
9491pretty-printer to print the value. Otherwise the value is printed normally.
9492
9493Pretty-printers are normally named. This makes them easy to manage.
9494The @samp{info pretty-printer} command will list all the installed
9495pretty-printers with their names.
9496If a pretty-printer can handle multiple data types, then its
9497@dfn{subprinters} are the printers for the individual data types.
9498Each such subprinter has its own name.
9499The format of the name is @var{printer-name};@var{subprinter-name}.
9500
9501Pretty-printers are installed by @dfn{registering} them with @value{GDBN}.
9502Typically they are automatically loaded and registered when the corresponding
9503debug information is loaded, thus making them available without having to
9504do anything special.
9505
9506There are three places where a pretty-printer can be registered.
9507
9508@itemize @bullet
9509@item
9510Pretty-printers registered globally are available when debugging
9511all inferiors.
9512
9513@item
9514Pretty-printers registered with a program space are available only
9515when debugging that program.
9516@xref{Progspaces In Python}, for more details on program spaces in Python.
9517
9518@item
9519Pretty-printers registered with an objfile are loaded and unloaded
9520with the corresponding objfile (e.g., shared library).
9521@xref{Objfiles In Python}, for more details on objfiles in Python.
9522@end itemize
9523
9524@xref{Selecting Pretty-Printers}, for further information on how
9525pretty-printers are selected,
9526
9527@xref{Writing a Pretty-Printer}, for implementing pretty printers
9528for new types.
9529
9530@node Pretty-Printer Example
9531@subsection Pretty-Printer Example
9532
9533Here is how a C@t{++} @code{std::string} looks without a pretty-printer:
9534
9535@smallexample
9536(@value{GDBP}) print s
9537$1 = @{
9538 static npos = 4294967295,
9539 _M_dataplus = @{
9540 <std::allocator<char>> = @{
9541 <__gnu_cxx::new_allocator<char>> = @{
9542 <No data fields>@}, <No data fields>
9543 @},
9544 members of std::basic_string<char, std::char_traits<char>,
9545 std::allocator<char> >::_Alloc_hider:
9546 _M_p = 0x804a014 "abcd"
9547 @}
9548@}
9549@end smallexample
9550
9551With a pretty-printer for @code{std::string} only the contents are printed:
9552
9553@smallexample
9554(@value{GDBP}) print s
9555$2 = "abcd"
9556@end smallexample
9557
9558@node Pretty-Printer Commands
9559@subsection Pretty-Printer Commands
9560@cindex pretty-printer commands
9561
9562@table @code
9563@kindex info pretty-printer
9564@item info pretty-printer [@var{object-regexp} [@var{name-regexp}]]
9565Print the list of installed pretty-printers.
9566This includes disabled pretty-printers, which are marked as such.
9567
9568@var{object-regexp} is a regular expression matching the objects
9569whose pretty-printers to list.
9570Objects can be @code{global}, the program space's file
9571(@pxref{Progspaces In Python}),
9572and the object files within that program space (@pxref{Objfiles In Python}).
9573@xref{Selecting Pretty-Printers}, for details on how @value{GDBN}
9574looks up a printer from these three objects.
9575
9576@var{name-regexp} is a regular expression matching the name of the printers
9577to list.
9578
9579@kindex disable pretty-printer
9580@item disable pretty-printer [@var{object-regexp} [@var{name-regexp}]]
9581Disable pretty-printers matching @var{object-regexp} and @var{name-regexp}.
9582A disabled pretty-printer is not forgotten, it may be enabled again later.
9583
9584@kindex enable pretty-printer
9585@item enable pretty-printer [@var{object-regexp} [@var{name-regexp}]]
9586Enable pretty-printers matching @var{object-regexp} and @var{name-regexp}.
9587@end table
9588
9589Example:
9590
9591Suppose we have three pretty-printers installed: one from library1.so
9592named @code{foo} that prints objects of type @code{foo}, and
9593another from library2.so named @code{bar} that prints two types of objects,
9594@code{bar1} and @code{bar2}.
9595
9596@smallexample
9597(gdb) info pretty-printer
9598library1.so:
9599 foo
9600library2.so:
9601 bar
9602 bar1
9603 bar2
9604(gdb) info pretty-printer library2
9605library2.so:
9606 bar
9607 bar1
9608 bar2
9609(gdb) disable pretty-printer library1
96101 printer disabled
96112 of 3 printers enabled
9612(gdb) info pretty-printer
9613library1.so:
9614 foo [disabled]
9615library2.so:
9616 bar
9617 bar1
9618 bar2
9619(gdb) disable pretty-printer library2 bar:bar1
96201 printer disabled
96211 of 3 printers enabled
9622(gdb) info pretty-printer library2
9623library1.so:
9624 foo [disabled]
9625library2.so:
9626 bar
9627 bar1 [disabled]
9628 bar2
9629(gdb) disable pretty-printer library2 bar
96301 printer disabled
96310 of 3 printers enabled
9632(gdb) info pretty-printer library2
9633library1.so:
9634 foo [disabled]
9635library2.so:
9636 bar [disabled]
9637 bar1 [disabled]
9638 bar2
9639@end smallexample
9640
9641Note that for @code{bar} the entire printer can be disabled,
9642as can each individual subprinter.
9643
9644@node Value History
9645@section Value History
9646
9647@cindex value history
9648@cindex history of values printed by @value{GDBN}
9649Values printed by the @code{print} command are saved in the @value{GDBN}
9650@dfn{value history}. This allows you to refer to them in other expressions.
9651Values are kept until the symbol table is re-read or discarded
9652(for example with the @code{file} or @code{symbol-file} commands).
9653When the symbol table changes, the value history is discarded,
9654since the values may contain pointers back to the types defined in the
9655symbol table.
9656
9657@cindex @code{$}
9658@cindex @code{$$}
9659@cindex history number
9660The values printed are given @dfn{history numbers} by which you can
9661refer to them. These are successive integers starting with one.
9662@code{print} shows you the history number assigned to a value by
9663printing @samp{$@var{num} = } before the value; here @var{num} is the
9664history number.
9665
9666To refer to any previous value, use @samp{$} followed by the value's
9667history number. The way @code{print} labels its output is designed to
9668remind you of this. Just @code{$} refers to the most recent value in
9669the history, and @code{$$} refers to the value before that.
9670@code{$$@var{n}} refers to the @var{n}th value from the end; @code{$$2}
9671is the value just prior to @code{$$}, @code{$$1} is equivalent to
9672@code{$$}, and @code{$$0} is equivalent to @code{$}.
9673
9674For example, suppose you have just printed a pointer to a structure and
9675want to see the contents of the structure. It suffices to type
9676
9677@smallexample
9678p *$
9679@end smallexample
9680
9681If you have a chain of structures where the component @code{next} points
9682to the next one, you can print the contents of the next one with this:
9683
9684@smallexample
9685p *$.next
9686@end smallexample
9687
9688@noindent
9689You can print successive links in the chain by repeating this
9690command---which you can do by just typing @key{RET}.
9691
9692Note that the history records values, not expressions. If the value of
9693@code{x} is 4 and you type these commands:
9694
9695@smallexample
9696print x
9697set x=5
9698@end smallexample
9699
9700@noindent
9701then the value recorded in the value history by the @code{print} command
9702remains 4 even though the value of @code{x} has changed.
9703
9704@table @code
9705@kindex show values
9706@item show values
9707Print the last ten values in the value history, with their item numbers.
9708This is like @samp{p@ $$9} repeated ten times, except that @code{show
9709values} does not change the history.
9710
9711@item show values @var{n}
9712Print ten history values centered on history item number @var{n}.
9713
9714@item show values +
9715Print ten history values just after the values last printed. If no more
9716values are available, @code{show values +} produces no display.
9717@end table
9718
9719Pressing @key{RET} to repeat @code{show values @var{n}} has exactly the
9720same effect as @samp{show values +}.
9721
9722@node Convenience Vars
9723@section Convenience Variables
9724
9725@cindex convenience variables
9726@cindex user-defined variables
9727@value{GDBN} provides @dfn{convenience variables} that you can use within
9728@value{GDBN} to hold on to a value and refer to it later. These variables
9729exist entirely within @value{GDBN}; they are not part of your program, and
9730setting a convenience variable has no direct effect on further execution
9731of your program. That is why you can use them freely.
9732
9733Convenience variables are prefixed with @samp{$}. Any name preceded by
9734@samp{$} can be used for a convenience variable, unless it is one of
9735the predefined machine-specific register names (@pxref{Registers, ,Registers}).
9736(Value history references, in contrast, are @emph{numbers} preceded
9737by @samp{$}. @xref{Value History, ,Value History}.)
9738
9739You can save a value in a convenience variable with an assignment
9740expression, just as you would set a variable in your program.
9741For example:
9742
9743@smallexample
9744set $foo = *object_ptr
9745@end smallexample
9746
9747@noindent
9748would save in @code{$foo} the value contained in the object pointed to by
9749@code{object_ptr}.
9750
9751Using a convenience variable for the first time creates it, but its
9752value is @code{void} until you assign a new value. You can alter the
9753value with another assignment at any time.
9754
9755Convenience variables have no fixed types. You can assign a convenience
9756variable any type of value, including structures and arrays, even if
9757that variable already has a value of a different type. The convenience
9758variable, when used as an expression, has the type of its current value.
9759
9760@table @code
9761@kindex show convenience
9762@cindex show all user variables and functions
9763@item show convenience
9764Print a list of convenience variables used so far, and their values,
9765as well as a list of the convenience functions.
9766Abbreviated @code{show conv}.
9767
9768@kindex init-if-undefined
9769@cindex convenience variables, initializing
9770@item init-if-undefined $@var{variable} = @var{expression}
9771Set a convenience variable if it has not already been set. This is useful
9772for user-defined commands that keep some state. It is similar, in concept,
9773to using local static variables with initializers in C (except that
9774convenience variables are global). It can also be used to allow users to
9775override default values used in a command script.
9776
9777If the variable is already defined then the expression is not evaluated so
9778any side-effects do not occur.
9779@end table
9780
9781One of the ways to use a convenience variable is as a counter to be
9782incremented or a pointer to be advanced. For example, to print
9783a field from successive elements of an array of structures:
9784
9785@smallexample
9786set $i = 0
9787print bar[$i++]->contents
9788@end smallexample
9789
9790@noindent
9791Repeat that command by typing @key{RET}.
9792
9793Some convenience variables are created automatically by @value{GDBN} and given
9794values likely to be useful.
9795
9796@table @code
9797@vindex $_@r{, convenience variable}
9798@item $_
9799The variable @code{$_} is automatically set by the @code{x} command to
9800the last address examined (@pxref{Memory, ,Examining Memory}). Other
9801commands which provide a default address for @code{x} to examine also
9802set @code{$_} to that address; these commands include @code{info line}
9803and @code{info breakpoint}. The type of @code{$_} is @code{void *}
9804except when set by the @code{x} command, in which case it is a pointer
9805to the type of @code{$__}.
9806
9807@vindex $__@r{, convenience variable}
9808@item $__
9809The variable @code{$__} is automatically set by the @code{x} command
9810to the value found in the last address examined. Its type is chosen
9811to match the format in which the data was printed.
9812
9813@item $_exitcode
9814@vindex $_exitcode@r{, convenience variable}
9815When the program being debugged terminates normally, @value{GDBN}
9816automatically sets this variable to the exit code of the program, and
9817resets @code{$_exitsignal} to @code{void}.
9818
9819@item $_exitsignal
9820@vindex $_exitsignal@r{, convenience variable}
9821When the program being debugged dies due to an uncaught signal,
9822@value{GDBN} automatically sets this variable to that signal's number,
9823and resets @code{$_exitcode} to @code{void}.
9824
9825To distinguish between whether the program being debugged has exited
9826(i.e., @code{$_exitcode} is not @code{void}) or signalled (i.e.,
9827@code{$_exitsignal} is not @code{void}), the convenience function
9828@code{$_isvoid} can be used (@pxref{Convenience Funs,, Convenience
9829Functions}). For example, considering the following source code:
9830
9831@smallexample
9832#include <signal.h>
9833
9834int
9835main (int argc, char *argv[])
9836@{
9837 raise (SIGALRM);
9838 return 0;
9839@}
9840@end smallexample
9841
9842A valid way of telling whether the program being debugged has exited
9843or signalled would be:
9844
9845@smallexample
9846(@value{GDBP}) define has_exited_or_signalled
9847Type commands for definition of ``has_exited_or_signalled''.
9848End with a line saying just ``end''.
9849>if $_isvoid ($_exitsignal)
9850 >echo The program has exited\n
9851 >else
9852 >echo The program has signalled\n
9853 >end
9854>end
9855(@value{GDBP}) run
9856Starting program:
9857
9858Program terminated with signal SIGALRM, Alarm clock.
9859The program no longer exists.
9860(@value{GDBP}) has_exited_or_signalled
9861The program has signalled
9862@end smallexample
9863
9864As can be seen, @value{GDBN} correctly informs that the program being
9865debugged has signalled, since it calls @code{raise} and raises a
9866@code{SIGALRM} signal. If the program being debugged had not called
9867@code{raise}, then @value{GDBN} would report a normal exit:
9868
9869@smallexample
9870(@value{GDBP}) has_exited_or_signalled
9871The program has exited
9872@end smallexample
9873
9874@item $_exception
9875The variable @code{$_exception} is set to the exception object being
9876thrown at an exception-related catchpoint. @xref{Set Catchpoints}.
9877
9878@item $_probe_argc
9879@itemx $_probe_arg0@dots{}$_probe_arg11
9880Arguments to a static probe. @xref{Static Probe Points}.
9881
9882@item $_sdata
9883@vindex $_sdata@r{, inspect, convenience variable}
9884The variable @code{$_sdata} contains extra collected static tracepoint
9885data. @xref{Tracepoint Actions,,Tracepoint Action Lists}. Note that
9886@code{$_sdata} could be empty, if not inspecting a trace buffer, or
9887if extra static tracepoint data has not been collected.
9888
9889@item $_siginfo
9890@vindex $_siginfo@r{, convenience variable}
9891The variable @code{$_siginfo} contains extra signal information
9892(@pxref{extra signal information}). Note that @code{$_siginfo}
9893could be empty, if the application has not yet received any signals.
9894For example, it will be empty before you execute the @code{run} command.
9895
9896@item $_tlb
9897@vindex $_tlb@r{, convenience variable}
9898The variable @code{$_tlb} is automatically set when debugging
9899applications running on MS-Windows in native mode or connected to
9900gdbserver that supports the @code{qGetTIBAddr} request.
9901@xref{General Query Packets}.
9902This variable contains the address of the thread information block.
9903
9904@end table
9905
9906On HP-UX systems, if you refer to a function or variable name that
9907begins with a dollar sign, @value{GDBN} searches for a user or system
9908name first, before it searches for a convenience variable.
9909
9910@node Convenience Funs
9911@section Convenience Functions
9912
9913@cindex convenience functions
9914@value{GDBN} also supplies some @dfn{convenience functions}. These
9915have a syntax similar to convenience variables. A convenience
9916function can be used in an expression just like an ordinary function;
9917however, a convenience function is implemented internally to
9918@value{GDBN}.
9919
9920These functions do not require @value{GDBN} to be configured with
9921@code{Python} support, which means that they are always available.
9922
9923@table @code
9924
9925@item $_isvoid (@var{expr})
9926@findex $_isvoid@r{, convenience function}
9927Return one if the expression @var{expr} is @code{void}. Otherwise it
9928returns zero.
9929
9930A @code{void} expression is an expression where the type of the result
9931is @code{void}. For example, you can examine a convenience variable
9932(see @ref{Convenience Vars,, Convenience Variables}) to check whether
9933it is @code{void}:
9934
9935@smallexample
9936(@value{GDBP}) print $_exitcode
9937$1 = void
9938(@value{GDBP}) print $_isvoid ($_exitcode)
9939$2 = 1
9940(@value{GDBP}) run
9941Starting program: ./a.out
9942[Inferior 1 (process 29572) exited normally]
9943(@value{GDBP}) print $_exitcode
9944$3 = 0
9945(@value{GDBP}) print $_isvoid ($_exitcode)
9946$4 = 0
9947@end smallexample
9948
9949In the example above, we used @code{$_isvoid} to check whether
9950@code{$_exitcode} is @code{void} before and after the execution of the
9951program being debugged. Before the execution there is no exit code to
9952be examined, therefore @code{$_exitcode} is @code{void}. After the
9953execution the program being debugged returned zero, therefore
9954@code{$_exitcode} is zero, which means that it is not @code{void}
9955anymore.
9956
9957The @code{void} expression can also be a call of a function from the
9958program being debugged. For example, given the following function:
9959
9960@smallexample
9961void
9962foo (void)
9963@{
9964@}
9965@end smallexample
9966
9967The result of calling it inside @value{GDBN} is @code{void}:
9968
9969@smallexample
9970(@value{GDBP}) print foo ()
9971$1 = void
9972(@value{GDBP}) print $_isvoid (foo ())
9973$2 = 1
9974(@value{GDBP}) set $v = foo ()
9975(@value{GDBP}) print $v
9976$3 = void
9977(@value{GDBP}) print $_isvoid ($v)
9978$4 = 1
9979@end smallexample
9980
9981@end table
9982
9983These functions require @value{GDBN} to be configured with
9984@code{Python} support.
9985
9986@table @code
9987
9988@item $_memeq(@var{buf1}, @var{buf2}, @var{length})
9989@findex $_memeq@r{, convenience function}
9990Returns one if the @var{length} bytes at the addresses given by
9991@var{buf1} and @var{buf2} are equal.
9992Otherwise it returns zero.
9993
9994@item $_regex(@var{str}, @var{regex})
9995@findex $_regex@r{, convenience function}
9996Returns one if the string @var{str} matches the regular expression
9997@var{regex}. Otherwise it returns zero.
9998The syntax of the regular expression is that specified by @code{Python}'s
9999regular expression support.
10000
10001@item $_streq(@var{str1}, @var{str2})
10002@findex $_streq@r{, convenience function}
10003Returns one if the strings @var{str1} and @var{str2} are equal.
10004Otherwise it returns zero.
10005
10006@item $_strlen(@var{str})
10007@findex $_strlen@r{, convenience function}
10008Returns the length of string @var{str}.
10009
10010@end table
10011
10012@value{GDBN} provides the ability to list and get help on
10013convenience functions.
10014
10015@table @code
10016@item help function
10017@kindex help function
10018@cindex show all convenience functions
10019Print a list of all convenience functions.
10020@end table
10021
10022@node Registers
10023@section Registers
10024
10025@cindex registers
10026You can refer to machine register contents, in expressions, as variables
10027with names starting with @samp{$}. The names of registers are different
10028for each machine; use @code{info registers} to see the names used on
10029your machine.
10030
10031@table @code
10032@kindex info registers
10033@item info registers
10034Print the names and values of all registers except floating-point
10035and vector registers (in the selected stack frame).
10036
10037@kindex info all-registers
10038@cindex floating point registers
10039@item info all-registers
10040Print the names and values of all registers, including floating-point
10041and vector registers (in the selected stack frame).
10042
10043@item info registers @var{regname} @dots{}
10044Print the @dfn{relativized} value of each specified register @var{regname}.
10045As discussed in detail below, register values are normally relative to
10046the selected stack frame. @var{regname} may be any register name valid on
10047the machine you are using, with or without the initial @samp{$}.
10048@end table
10049
10050@cindex stack pointer register
10051@cindex program counter register
10052@cindex process status register
10053@cindex frame pointer register
10054@cindex standard registers
10055@value{GDBN} has four ``standard'' register names that are available (in
10056expressions) on most machines---whenever they do not conflict with an
10057architecture's canonical mnemonics for registers. The register names
10058@code{$pc} and @code{$sp} are used for the program counter register and
10059the stack pointer. @code{$fp} is used for a register that contains a
10060pointer to the current stack frame, and @code{$ps} is used for a
10061register that contains the processor status. For example,
10062you could print the program counter in hex with
10063
10064@smallexample
10065p/x $pc
10066@end smallexample
10067
10068@noindent
10069or print the instruction to be executed next with
10070
10071@smallexample
10072x/i $pc
10073@end smallexample
10074
10075@noindent
10076or add four to the stack pointer@footnote{This is a way of removing
10077one word from the stack, on machines where stacks grow downward in
10078memory (most machines, nowadays). This assumes that the innermost
10079stack frame is selected; setting @code{$sp} is not allowed when other
10080stack frames are selected. To pop entire frames off the stack,
10081regardless of machine architecture, use @code{return};
10082see @ref{Returning, ,Returning from a Function}.} with
10083
10084@smallexample
10085set $sp += 4
10086@end smallexample
10087
10088Whenever possible, these four standard register names are available on
10089your machine even though the machine has different canonical mnemonics,
10090so long as there is no conflict. The @code{info registers} command
10091shows the canonical names. For example, on the SPARC, @code{info
10092registers} displays the processor status register as @code{$psr} but you
10093can also refer to it as @code{$ps}; and on x86-based machines @code{$ps}
10094is an alias for the @sc{eflags} register.
10095
10096@value{GDBN} always considers the contents of an ordinary register as an
10097integer when the register is examined in this way. Some machines have
10098special registers which can hold nothing but floating point; these
10099registers are considered to have floating point values. There is no way
10100to refer to the contents of an ordinary register as floating point value
10101(although you can @emph{print} it as a floating point value with
10102@samp{print/f $@var{regname}}).
10103
10104Some registers have distinct ``raw'' and ``virtual'' data formats. This
10105means that the data format in which the register contents are saved by
10106the operating system is not the same one that your program normally
10107sees. For example, the registers of the 68881 floating point
10108coprocessor are always saved in ``extended'' (raw) format, but all C
10109programs expect to work with ``double'' (virtual) format. In such
10110cases, @value{GDBN} normally works with the virtual format only (the format
10111that makes sense for your program), but the @code{info registers} command
10112prints the data in both formats.
10113
10114@cindex SSE registers (x86)
10115@cindex MMX registers (x86)
10116Some machines have special registers whose contents can be interpreted
10117in several different ways. For example, modern x86-based machines
10118have SSE and MMX registers that can hold several values packed
10119together in several different formats. @value{GDBN} refers to such
10120registers in @code{struct} notation:
10121
10122@smallexample
10123(@value{GDBP}) print $xmm1
10124$1 = @{
10125 v4_float = @{0, 3.43859137e-038, 1.54142831e-044, 1.821688e-044@},
10126 v2_double = @{9.92129282474342e-303, 2.7585945287983262e-313@},
10127 v16_int8 = "\000\000\000\000\3706;\001\v\000\000\000\r\000\000",
10128 v8_int16 = @{0, 0, 14072, 315, 11, 0, 13, 0@},
10129 v4_int32 = @{0, 20657912, 11, 13@},
10130 v2_int64 = @{88725056443645952, 55834574859@},
10131 uint128 = 0x0000000d0000000b013b36f800000000
10132@}
10133@end smallexample
10134
10135@noindent
10136To set values of such registers, you need to tell @value{GDBN} which
10137view of the register you wish to change, as if you were assigning
10138value to a @code{struct} member:
10139
10140@smallexample
10141 (@value{GDBP}) set $xmm1.uint128 = 0x000000000000000000000000FFFFFFFF
10142@end smallexample
10143
10144Normally, register values are relative to the selected stack frame
10145(@pxref{Selection, ,Selecting a Frame}). This means that you get the
10146value that the register would contain if all stack frames farther in
10147were exited and their saved registers restored. In order to see the
10148true contents of hardware registers, you must select the innermost
10149frame (with @samp{frame 0}).
10150
10151@cindex caller-saved registers
10152@cindex call-clobbered registers
10153@cindex volatile registers
10154@cindex <not saved> values
10155Usually ABIs reserve some registers as not needed to be saved by the
10156callee (a.k.a.: ``caller-saved'', ``call-clobbered'' or ``volatile''
10157registers). It may therefore not be possible for @value{GDBN} to know
10158the value a register had before the call (in other words, in the outer
10159frame), if the register value has since been changed by the callee.
10160@value{GDBN} tries to deduce where the inner frame saved
10161(``callee-saved'') registers, from the debug info, unwind info, or the
10162machine code generated by your compiler. If some register is not
10163saved, and @value{GDBN} knows the register is ``caller-saved'' (via
10164its own knowledge of the ABI, or because the debug/unwind info
10165explicitly says the register's value is undefined), @value{GDBN}
10166displays @w{@samp{<not saved>}} as the register's value. With targets
10167that @value{GDBN} has no knowledge of the register saving convention,
10168if a register was not saved by the callee, then its value and location
10169in the outer frame are assumed to be the same of the inner frame.
10170This is usually harmless, because if the register is call-clobbered,
10171the caller either does not care what is in the register after the
10172call, or has code to restore the value that it does care about. Note,
10173however, that if you change such a register in the outer frame, you
10174may also be affecting the inner frame. Also, the more ``outer'' the
10175frame is you're looking at, the more likely a call-clobbered
10176register's value is to be wrong, in the sense that it doesn't actually
10177represent the value the register had just before the call.
10178
10179@node Floating Point Hardware
10180@section Floating Point Hardware
10181@cindex floating point
10182
10183Depending on the configuration, @value{GDBN} may be able to give
10184you more information about the status of the floating point hardware.
10185
10186@table @code
10187@kindex info float
10188@item info float
10189Display hardware-dependent information about the floating
10190point unit. The exact contents and layout vary depending on the
10191floating point chip. Currently, @samp{info float} is supported on
10192the ARM and x86 machines.
10193@end table
10194
10195@node Vector Unit
10196@section Vector Unit
10197@cindex vector unit
10198
10199Depending on the configuration, @value{GDBN} may be able to give you
10200more information about the status of the vector unit.
10201
10202@table @code
10203@kindex info vector
10204@item info vector
10205Display information about the vector unit. The exact contents and
10206layout vary depending on the hardware.
10207@end table
10208
10209@node OS Information
10210@section Operating System Auxiliary Information
10211@cindex OS information
10212
10213@value{GDBN} provides interfaces to useful OS facilities that can help
10214you debug your program.
10215
10216@cindex auxiliary vector
10217@cindex vector, auxiliary
10218Some operating systems supply an @dfn{auxiliary vector} to programs at
10219startup. This is akin to the arguments and environment that you
10220specify for a program, but contains a system-dependent variety of
10221binary values that tell system libraries important details about the
10222hardware, operating system, and process. Each value's purpose is
10223identified by an integer tag; the meanings are well-known but system-specific.
10224Depending on the configuration and operating system facilities,
10225@value{GDBN} may be able to show you this information. For remote
10226targets, this functionality may further depend on the remote stub's
10227support of the @samp{qXfer:auxv:read} packet, see
10228@ref{qXfer auxiliary vector read}.
10229
10230@table @code
10231@kindex info auxv
10232@item info auxv
10233Display the auxiliary vector of the inferior, which can be either a
10234live process or a core dump file. @value{GDBN} prints each tag value
10235numerically, and also shows names and text descriptions for recognized
10236tags. Some values in the vector are numbers, some bit masks, and some
10237pointers to strings or other data. @value{GDBN} displays each value in the
10238most appropriate form for a recognized tag, and in hexadecimal for
10239an unrecognized tag.
10240@end table
10241
10242On some targets, @value{GDBN} can access operating system-specific
10243information and show it to you. The types of information available
10244will differ depending on the type of operating system running on the
10245target. The mechanism used to fetch the data is described in
10246@ref{Operating System Information}. For remote targets, this
10247functionality depends on the remote stub's support of the
10248@samp{qXfer:osdata:read} packet, see @ref{qXfer osdata read}.
10249
10250@table @code
10251@kindex info os
10252@item info os @var{infotype}
10253
10254Display OS information of the requested type.
10255
10256On @sc{gnu}/Linux, the following values of @var{infotype} are valid:
10257
10258@anchor{linux info os infotypes}
10259@table @code
10260@kindex info os processes
10261@item processes
10262Display the list of processes on the target. For each process,
10263@value{GDBN} prints the process identifier, the name of the user, the
10264command corresponding to the process, and the list of processor cores
10265that the process is currently running on. (To understand what these
10266properties mean, for this and the following info types, please consult
10267the general @sc{gnu}/Linux documentation.)
10268
10269@kindex info os procgroups
10270@item procgroups
10271Display the list of process groups on the target. For each process,
10272@value{GDBN} prints the identifier of the process group that it belongs
10273to, the command corresponding to the process group leader, the process
10274identifier, and the command line of the process. The list is sorted
10275first by the process group identifier, then by the process identifier,
10276so that processes belonging to the same process group are grouped together
10277and the process group leader is listed first.
10278
10279@kindex info os threads
10280@item threads
10281Display the list of threads running on the target. For each thread,
10282@value{GDBN} prints the identifier of the process that the thread
10283belongs to, the command of the process, the thread identifier, and the
10284processor core that it is currently running on. The main thread of a
10285process is not listed.
10286
10287@kindex info os files
10288@item files
10289Display the list of open file descriptors on the target. For each
10290file descriptor, @value{GDBN} prints the identifier of the process
10291owning the descriptor, the command of the owning process, the value
10292of the descriptor, and the target of the descriptor.
10293
10294@kindex info os sockets
10295@item sockets
10296Display the list of Internet-domain sockets on the target. For each
10297socket, @value{GDBN} prints the address and port of the local and
10298remote endpoints, the current state of the connection, the creator of
10299the socket, the IP address family of the socket, and the type of the
10300connection.
10301
10302@kindex info os shm
10303@item shm
10304Display the list of all System V shared-memory regions on the target.
10305For each shared-memory region, @value{GDBN} prints the region key,
10306the shared-memory identifier, the access permissions, the size of the
10307region, the process that created the region, the process that last
10308attached to or detached from the region, the current number of live
10309attaches to the region, and the times at which the region was last
10310attached to, detach from, and changed.
10311
10312@kindex info os semaphores
10313@item semaphores
10314Display the list of all System V semaphore sets on the target. For each
10315semaphore set, @value{GDBN} prints the semaphore set key, the semaphore
10316set identifier, the access permissions, the number of semaphores in the
10317set, the user and group of the owner and creator of the semaphore set,
10318and the times at which the semaphore set was operated upon and changed.
10319
10320@kindex info os msg
10321@item msg
10322Display the list of all System V message queues on the target. For each
10323message queue, @value{GDBN} prints the message queue key, the message
10324queue identifier, the access permissions, the current number of bytes
10325on the queue, the current number of messages on the queue, the processes
10326that last sent and received a message on the queue, the user and group
10327of the owner and creator of the message queue, the times at which a
10328message was last sent and received on the queue, and the time at which
10329the message queue was last changed.
10330
10331@kindex info os modules
10332@item modules
10333Display the list of all loaded kernel modules on the target. For each
10334module, @value{GDBN} prints the module name, the size of the module in
10335bytes, the number of times the module is used, the dependencies of the
10336module, the status of the module, and the address of the loaded module
10337in memory.
10338@end table
10339
10340@item info os
10341If @var{infotype} is omitted, then list the possible values for
10342@var{infotype} and the kind of OS information available for each
10343@var{infotype}. If the target does not return a list of possible
10344types, this command will report an error.
10345@end table
10346
10347@node Memory Region Attributes
10348@section Memory Region Attributes
10349@cindex memory region attributes
10350
10351@dfn{Memory region attributes} allow you to describe special handling
10352required by regions of your target's memory. @value{GDBN} uses
10353attributes to determine whether to allow certain types of memory
10354accesses; whether to use specific width accesses; and whether to cache
10355target memory. By default the description of memory regions is
10356fetched from the target (if the current target supports this), but the
10357user can override the fetched regions.
10358
10359Defined memory regions can be individually enabled and disabled. When a
10360memory region is disabled, @value{GDBN} uses the default attributes when
10361accessing memory in that region. Similarly, if no memory regions have
10362been defined, @value{GDBN} uses the default attributes when accessing
10363all memory.
10364
10365When a memory region is defined, it is given a number to identify it;
10366to enable, disable, or remove a memory region, you specify that number.
10367
10368@table @code
10369@kindex mem
10370@item mem @var{lower} @var{upper} @var{attributes}@dots{}
10371Define a memory region bounded by @var{lower} and @var{upper} with
10372attributes @var{attributes}@dots{}, and add it to the list of regions
10373monitored by @value{GDBN}. Note that @var{upper} == 0 is a special
10374case: it is treated as the target's maximum memory address.
10375(0xffff on 16 bit targets, 0xffffffff on 32 bit targets, etc.)
10376
10377@item mem auto
10378Discard any user changes to the memory regions and use target-supplied
10379regions, if available, or no regions if the target does not support.
10380
10381@kindex delete mem
10382@item delete mem @var{nums}@dots{}
10383Remove memory regions @var{nums}@dots{} from the list of regions
10384monitored by @value{GDBN}.
10385
10386@kindex disable mem
10387@item disable mem @var{nums}@dots{}
10388Disable monitoring of memory regions @var{nums}@dots{}.
10389A disabled memory region is not forgotten.
10390It may be enabled again later.
10391
10392@kindex enable mem
10393@item enable mem @var{nums}@dots{}
10394Enable monitoring of memory regions @var{nums}@dots{}.
10395
10396@kindex info mem
10397@item info mem
10398Print a table of all defined memory regions, with the following columns
10399for each region:
10400
10401@table @emph
10402@item Memory Region Number
10403@item Enabled or Disabled.
10404Enabled memory regions are marked with @samp{y}.
10405Disabled memory regions are marked with @samp{n}.
10406
10407@item Lo Address
10408The address defining the inclusive lower bound of the memory region.
10409
10410@item Hi Address
10411The address defining the exclusive upper bound of the memory region.
10412
10413@item Attributes
10414The list of attributes set for this memory region.
10415@end table
10416@end table
10417
10418
10419@subsection Attributes
10420
10421@subsubsection Memory Access Mode
10422The access mode attributes set whether @value{GDBN} may make read or
10423write accesses to a memory region.
10424
10425While these attributes prevent @value{GDBN} from performing invalid
10426memory accesses, they do nothing to prevent the target system, I/O DMA,
10427etc.@: from accessing memory.
10428
10429@table @code
10430@item ro
10431Memory is read only.
10432@item wo
10433Memory is write only.
10434@item rw
10435Memory is read/write. This is the default.
10436@end table
10437
10438@subsubsection Memory Access Size
10439The access size attribute tells @value{GDBN} to use specific sized
10440accesses in the memory region. Often memory mapped device registers
10441require specific sized accesses. If no access size attribute is
10442specified, @value{GDBN} may use accesses of any size.
10443
10444@table @code
10445@item 8
10446Use 8 bit memory accesses.
10447@item 16
10448Use 16 bit memory accesses.
10449@item 32
10450Use 32 bit memory accesses.
10451@item 64
10452Use 64 bit memory accesses.
10453@end table
10454
10455@c @subsubsection Hardware/Software Breakpoints
10456@c The hardware/software breakpoint attributes set whether @value{GDBN}
10457@c will use hardware or software breakpoints for the internal breakpoints
10458@c used by the step, next, finish, until, etc. commands.
10459@c
10460@c @table @code
10461@c @item hwbreak
10462@c Always use hardware breakpoints
10463@c @item swbreak (default)
10464@c @end table
10465
10466@subsubsection Data Cache
10467The data cache attributes set whether @value{GDBN} will cache target
10468memory. While this generally improves performance by reducing debug
10469protocol overhead, it can lead to incorrect results because @value{GDBN}
10470does not know about volatile variables or memory mapped device
10471registers.
10472
10473@table @code
10474@item cache
10475Enable @value{GDBN} to cache target memory.
10476@item nocache
10477Disable @value{GDBN} from caching target memory. This is the default.
10478@end table
10479
10480@subsection Memory Access Checking
10481@value{GDBN} can be instructed to refuse accesses to memory that is
10482not explicitly described. This can be useful if accessing such
10483regions has undesired effects for a specific target, or to provide
10484better error checking. The following commands control this behaviour.
10485
10486@table @code
10487@kindex set mem inaccessible-by-default
10488@item set mem inaccessible-by-default [on|off]
10489If @code{on} is specified, make @value{GDBN} treat memory not
10490explicitly described by the memory ranges as non-existent and refuse accesses
10491to such memory. The checks are only performed if there's at least one
10492memory range defined. If @code{off} is specified, make @value{GDBN}
10493treat the memory not explicitly described by the memory ranges as RAM.
10494The default value is @code{on}.
10495@kindex show mem inaccessible-by-default
10496@item show mem inaccessible-by-default
10497Show the current handling of accesses to unknown memory.
10498@end table
10499
10500
10501@c @subsubsection Memory Write Verification
10502@c The memory write verification attributes set whether @value{GDBN}
10503@c will re-reads data after each write to verify the write was successful.
10504@c
10505@c @table @code
10506@c @item verify
10507@c @item noverify (default)
10508@c @end table
10509
10510@node Dump/Restore Files
10511@section Copy Between Memory and a File
10512@cindex dump/restore files
10513@cindex append data to a file
10514@cindex dump data to a file
10515@cindex restore data from a file
10516
10517You can use the commands @code{dump}, @code{append}, and
10518@code{restore} to copy data between target memory and a file. The
10519@code{dump} and @code{append} commands write data to a file, and the
10520@code{restore} command reads data from a file back into the inferior's
10521memory. Files may be in binary, Motorola S-record, Intel hex, or
10522Tektronix Hex format; however, @value{GDBN} can only append to binary
10523files.
10524
10525@table @code
10526
10527@kindex dump
10528@item dump @r{[}@var{format}@r{]} memory @var{filename} @var{start_addr} @var{end_addr}
10529@itemx dump @r{[}@var{format}@r{]} value @var{filename} @var{expr}
10530Dump the contents of memory from @var{start_addr} to @var{end_addr},
10531or the value of @var{expr}, to @var{filename} in the given format.
10532
10533The @var{format} parameter may be any one of:
10534@table @code
10535@item binary
10536Raw binary form.
10537@item ihex
10538Intel hex format.
10539@item srec
10540Motorola S-record format.
10541@item tekhex
10542Tektronix Hex format.
10543@end table
10544
10545@value{GDBN} uses the same definitions of these formats as the
10546@sc{gnu} binary utilities, like @samp{objdump} and @samp{objcopy}. If
10547@var{format} is omitted, @value{GDBN} dumps the data in raw binary
10548form.
10549
10550@kindex append
10551@item append @r{[}binary@r{]} memory @var{filename} @var{start_addr} @var{end_addr}
10552@itemx append @r{[}binary@r{]} value @var{filename} @var{expr}
10553Append the contents of memory from @var{start_addr} to @var{end_addr},
10554or the value of @var{expr}, to the file @var{filename}, in raw binary form.
10555(@value{GDBN} can only append data to files in raw binary form.)
10556
10557@kindex restore
10558@item restore @var{filename} @r{[}binary@r{]} @var{bias} @var{start} @var{end}
10559Restore the contents of file @var{filename} into memory. The
10560@code{restore} command can automatically recognize any known @sc{bfd}
10561file format, except for raw binary. To restore a raw binary file you
10562must specify the optional keyword @code{binary} after the filename.
10563
10564If @var{bias} is non-zero, its value will be added to the addresses
10565contained in the file. Binary files always start at address zero, so
10566they will be restored at address @var{bias}. Other bfd files have
10567a built-in location; they will be restored at offset @var{bias}
10568from that location.
10569
10570If @var{start} and/or @var{end} are non-zero, then only data between
10571file offset @var{start} and file offset @var{end} will be restored.
10572These offsets are relative to the addresses in the file, before
10573the @var{bias} argument is applied.
10574
10575@end table
10576
10577@node Core File Generation
10578@section How to Produce a Core File from Your Program
10579@cindex dump core from inferior
10580
10581A @dfn{core file} or @dfn{core dump} is a file that records the memory
10582image of a running process and its process status (register values
10583etc.). Its primary use is post-mortem debugging of a program that
10584crashed while it ran outside a debugger. A program that crashes
10585automatically produces a core file, unless this feature is disabled by
10586the user. @xref{Files}, for information on invoking @value{GDBN} in
10587the post-mortem debugging mode.
10588
10589Occasionally, you may wish to produce a core file of the program you
10590are debugging in order to preserve a snapshot of its state.
10591@value{GDBN} has a special command for that.
10592
10593@table @code
10594@kindex gcore
10595@kindex generate-core-file
10596@item generate-core-file [@var{file}]
10597@itemx gcore [@var{file}]
10598Produce a core dump of the inferior process. The optional argument
10599@var{file} specifies the file name where to put the core dump. If not
10600specified, the file name defaults to @file{core.@var{pid}}, where
10601@var{pid} is the inferior process ID.
10602
10603Note that this command is implemented only for some systems (as of
10604this writing, @sc{gnu}/Linux, FreeBSD, Solaris, and S390).
10605@end table
10606
10607@node Character Sets
10608@section Character Sets
10609@cindex character sets
10610@cindex charset
10611@cindex translating between character sets
10612@cindex host character set
10613@cindex target character set
10614
10615If the program you are debugging uses a different character set to
10616represent characters and strings than the one @value{GDBN} uses itself,
10617@value{GDBN} can automatically translate between the character sets for
10618you. The character set @value{GDBN} uses we call the @dfn{host
10619character set}; the one the inferior program uses we call the
10620@dfn{target character set}.
10621
10622For example, if you are running @value{GDBN} on a @sc{gnu}/Linux system, which
10623uses the ISO Latin 1 character set, but you are using @value{GDBN}'s
10624remote protocol (@pxref{Remote Debugging}) to debug a program
10625running on an IBM mainframe, which uses the @sc{ebcdic} character set,
10626then the host character set is Latin-1, and the target character set is
10627@sc{ebcdic}. If you give @value{GDBN} the command @code{set
10628target-charset EBCDIC-US}, then @value{GDBN} translates between
10629@sc{ebcdic} and Latin 1 as you print character or string values, or use
10630character and string literals in expressions.
10631
10632@value{GDBN} has no way to automatically recognize which character set
10633the inferior program uses; you must tell it, using the @code{set
10634target-charset} command, described below.
10635
10636Here are the commands for controlling @value{GDBN}'s character set
10637support:
10638
10639@table @code
10640@item set target-charset @var{charset}
10641@kindex set target-charset
10642Set the current target character set to @var{charset}. To display the
10643list of supported target character sets, type
10644@kbd{@w{set target-charset @key{TAB}@key{TAB}}}.
10645
10646@item set host-charset @var{charset}
10647@kindex set host-charset
10648Set the current host character set to @var{charset}.
10649
10650By default, @value{GDBN} uses a host character set appropriate to the
10651system it is running on; you can override that default using the
10652@code{set host-charset} command. On some systems, @value{GDBN} cannot
10653automatically determine the appropriate host character set. In this
10654case, @value{GDBN} uses @samp{UTF-8}.
10655
10656@value{GDBN} can only use certain character sets as its host character
10657set. If you type @kbd{@w{set host-charset @key{TAB}@key{TAB}}},
10658@value{GDBN} will list the host character sets it supports.
10659
10660@item set charset @var{charset}
10661@kindex set charset
10662Set the current host and target character sets to @var{charset}. As
10663above, if you type @kbd{@w{set charset @key{TAB}@key{TAB}}},
10664@value{GDBN} will list the names of the character sets that can be used
10665for both host and target.
10666
10667@item show charset
10668@kindex show charset
10669Show the names of the current host and target character sets.
10670
10671@item show host-charset
10672@kindex show host-charset
10673Show the name of the current host character set.
10674
10675@item show target-charset
10676@kindex show target-charset
10677Show the name of the current target character set.
10678
10679@item set target-wide-charset @var{charset}
10680@kindex set target-wide-charset
10681Set the current target's wide character set to @var{charset}. This is
10682the character set used by the target's @code{wchar_t} type. To
10683display the list of supported wide character sets, type
10684@kbd{@w{set target-wide-charset @key{TAB}@key{TAB}}}.
10685
10686@item show target-wide-charset
10687@kindex show target-wide-charset
10688Show the name of the current target's wide character set.
10689@end table
10690
10691Here is an example of @value{GDBN}'s character set support in action.
10692Assume that the following source code has been placed in the file
10693@file{charset-test.c}:
10694
10695@smallexample
10696#include <stdio.h>
10697
10698char ascii_hello[]
10699 = @{72, 101, 108, 108, 111, 44, 32, 119,
10700 111, 114, 108, 100, 33, 10, 0@};
10701char ibm1047_hello[]
10702 = @{200, 133, 147, 147, 150, 107, 64, 166,
10703 150, 153, 147, 132, 90, 37, 0@};
10704
10705main ()
10706@{
10707 printf ("Hello, world!\n");
10708@}
10709@end smallexample
10710
10711In this program, @code{ascii_hello} and @code{ibm1047_hello} are arrays
10712containing the string @samp{Hello, world!} followed by a newline,
10713encoded in the @sc{ascii} and @sc{ibm1047} character sets.
10714
10715We compile the program, and invoke the debugger on it:
10716
10717@smallexample
10718$ gcc -g charset-test.c -o charset-test
10719$ gdb -nw charset-test
10720GNU gdb 2001-12-19-cvs
10721Copyright 2001 Free Software Foundation, Inc.
10722@dots{}
10723(@value{GDBP})
10724@end smallexample
10725
10726We can use the @code{show charset} command to see what character sets
10727@value{GDBN} is currently using to interpret and display characters and
10728strings:
10729
10730@smallexample
10731(@value{GDBP}) show charset
10732The current host and target character set is `ISO-8859-1'.
10733(@value{GDBP})
10734@end smallexample
10735
10736For the sake of printing this manual, let's use @sc{ascii} as our
10737initial character set:
10738@smallexample
10739(@value{GDBP}) set charset ASCII
10740(@value{GDBP}) show charset
10741The current host and target character set is `ASCII'.
10742(@value{GDBP})
10743@end smallexample
10744
10745Let's assume that @sc{ascii} is indeed the correct character set for our
10746host system --- in other words, let's assume that if @value{GDBN} prints
10747characters using the @sc{ascii} character set, our terminal will display
10748them properly. Since our current target character set is also
10749@sc{ascii}, the contents of @code{ascii_hello} print legibly:
10750
10751@smallexample
10752(@value{GDBP}) print ascii_hello
10753$1 = 0x401698 "Hello, world!\n"
10754(@value{GDBP}) print ascii_hello[0]
10755$2 = 72 'H'
10756(@value{GDBP})
10757@end smallexample
10758
10759@value{GDBN} uses the target character set for character and string
10760literals you use in expressions:
10761
10762@smallexample
10763(@value{GDBP}) print '+'
10764$3 = 43 '+'
10765(@value{GDBP})
10766@end smallexample
10767
10768The @sc{ascii} character set uses the number 43 to encode the @samp{+}
10769character.
10770
10771@value{GDBN} relies on the user to tell it which character set the
10772target program uses. If we print @code{ibm1047_hello} while our target
10773character set is still @sc{ascii}, we get jibberish:
10774
10775@smallexample
10776(@value{GDBP}) print ibm1047_hello
10777$4 = 0x4016a8 "\310\205\223\223\226k@@\246\226\231\223\204Z%"
10778(@value{GDBP}) print ibm1047_hello[0]
10779$5 = 200 '\310'
10780(@value{GDBP})
10781@end smallexample
10782
10783If we invoke the @code{set target-charset} followed by @key{TAB}@key{TAB},
10784@value{GDBN} tells us the character sets it supports:
10785
10786@smallexample
10787(@value{GDBP}) set target-charset
10788ASCII EBCDIC-US IBM1047 ISO-8859-1
10789(@value{GDBP}) set target-charset
10790@end smallexample
10791
10792We can select @sc{ibm1047} as our target character set, and examine the
10793program's strings again. Now the @sc{ascii} string is wrong, but
10794@value{GDBN} translates the contents of @code{ibm1047_hello} from the
10795target character set, @sc{ibm1047}, to the host character set,
10796@sc{ascii}, and they display correctly:
10797
10798@smallexample
10799(@value{GDBP}) set target-charset IBM1047
10800(@value{GDBP}) show charset
10801The current host character set is `ASCII'.
10802The current target character set is `IBM1047'.
10803(@value{GDBP}) print ascii_hello
10804$6 = 0x401698 "\110\145%%?\054\040\167?\162%\144\041\012"
10805(@value{GDBP}) print ascii_hello[0]
10806$7 = 72 '\110'
10807(@value{GDBP}) print ibm1047_hello
10808$8 = 0x4016a8 "Hello, world!\n"
10809(@value{GDBP}) print ibm1047_hello[0]
10810$9 = 200 'H'
10811(@value{GDBP})
10812@end smallexample
10813
10814As above, @value{GDBN} uses the target character set for character and
10815string literals you use in expressions:
10816
10817@smallexample
10818(@value{GDBP}) print '+'
10819$10 = 78 '+'
10820(@value{GDBP})
10821@end smallexample
10822
10823The @sc{ibm1047} character set uses the number 78 to encode the @samp{+}
10824character.
10825
10826@node Caching Target Data
10827@section Caching Data of Targets
10828@cindex caching data of targets
10829
10830@value{GDBN} caches data exchanged between the debugger and a target.
10831Each cache is associated with the address space of the inferior.
10832@xref{Inferiors and Programs}, about inferior and address space.
10833Such caching generally improves performance in remote debugging
10834(@pxref{Remote Debugging}), because it reduces the overhead of the
10835remote protocol by bundling memory reads and writes into large chunks.
10836Unfortunately, simply caching everything would lead to incorrect results,
10837since @value{GDBN} does not necessarily know anything about volatile
10838values, memory-mapped I/O addresses, etc. Furthermore, in non-stop mode
10839(@pxref{Non-Stop Mode}) memory can be changed @emph{while} a gdb command
10840is executing.
10841Therefore, by default, @value{GDBN} only caches data
10842known to be on the stack@footnote{In non-stop mode, it is moderately
10843rare for a running thread to modify the stack of a stopped thread
10844in a way that would interfere with a backtrace, and caching of
10845stack reads provides a significant speed up of remote backtraces.} or
10846in the code segment.
10847Other regions of memory can be explicitly marked as
10848cacheable; @pxref{Memory Region Attributes}.
10849
10850@table @code
10851@kindex set remotecache
10852@item set remotecache on
10853@itemx set remotecache off
10854This option no longer does anything; it exists for compatibility
10855with old scripts.
10856
10857@kindex show remotecache
10858@item show remotecache
10859Show the current state of the obsolete remotecache flag.
10860
10861@kindex set stack-cache
10862@item set stack-cache on
10863@itemx set stack-cache off
10864Enable or disable caching of stack accesses. When @code{on}, use
10865caching. By default, this option is @code{on}.
10866
10867@kindex show stack-cache
10868@item show stack-cache
10869Show the current state of data caching for memory accesses.
10870
10871@kindex set code-cache
10872@item set code-cache on
10873@itemx set code-cache off
10874Enable or disable caching of code segment accesses. When @code{on},
10875use caching. By default, this option is @code{on}. This improves
10876performance of disassembly in remote debugging.
10877
10878@kindex show code-cache
10879@item show code-cache
10880Show the current state of target memory cache for code segment
10881accesses.
10882
10883@kindex info dcache
10884@item info dcache @r{[}line@r{]}
10885Print the information about the performance of data cache of the
10886current inferior's address space. The information displayed
10887includes the dcache width and depth, and for each cache line, its
10888number, address, and how many times it was referenced. This
10889command is useful for debugging the data cache operation.
10890
10891If a line number is specified, the contents of that line will be
10892printed in hex.
10893
10894@item set dcache size @var{size}
10895@cindex dcache size
10896@kindex set dcache size
10897Set maximum number of entries in dcache (dcache depth above).
10898
10899@item set dcache line-size @var{line-size}
10900@cindex dcache line-size
10901@kindex set dcache line-size
10902Set number of bytes each dcache entry caches (dcache width above).
10903Must be a power of 2.
10904
10905@item show dcache size
10906@kindex show dcache size
10907Show maximum number of dcache entries. @xref{Caching Target Data, info dcache}.
10908
10909@item show dcache line-size
10910@kindex show dcache line-size
10911Show default size of dcache lines.
10912
10913@end table
10914
10915@node Searching Memory
10916@section Search Memory
10917@cindex searching memory
10918
10919Memory can be searched for a particular sequence of bytes with the
10920@code{find} command.
10921
10922@table @code
10923@kindex find
10924@item find @r{[}/@var{sn}@r{]} @var{start_addr}, +@var{len}, @var{val1} @r{[}, @var{val2}, @dots{}@r{]}
10925@itemx find @r{[}/@var{sn}@r{]} @var{start_addr}, @var{end_addr}, @var{val1} @r{[}, @var{val2}, @dots{}@r{]}
10926Search memory for the sequence of bytes specified by @var{val1}, @var{val2},
10927etc. The search begins at address @var{start_addr} and continues for either
10928@var{len} bytes or through to @var{end_addr} inclusive.
10929@end table
10930
10931@var{s} and @var{n} are optional parameters.
10932They may be specified in either order, apart or together.
10933
10934@table @r
10935@item @var{s}, search query size
10936The size of each search query value.
10937
10938@table @code
10939@item b
10940bytes
10941@item h
10942halfwords (two bytes)
10943@item w
10944words (four bytes)
10945@item g
10946giant words (eight bytes)
10947@end table
10948
10949All values are interpreted in the current language.
10950This means, for example, that if the current source language is C/C@t{++}
10951then searching for the string ``hello'' includes the trailing '\0'.
10952
10953If the value size is not specified, it is taken from the
10954value's type in the current language.
10955This is useful when one wants to specify the search
10956pattern as a mixture of types.
10957Note that this means, for example, that in the case of C-like languages
10958a search for an untyped 0x42 will search for @samp{(int) 0x42}
10959which is typically four bytes.
10960
10961@item @var{n}, maximum number of finds
10962The maximum number of matches to print. The default is to print all finds.
10963@end table
10964
10965You can use strings as search values. Quote them with double-quotes
10966 (@code{"}).
10967The string value is copied into the search pattern byte by byte,
10968regardless of the endianness of the target and the size specification.
10969
10970The address of each match found is printed as well as a count of the
10971number of matches found.
10972
10973The address of the last value found is stored in convenience variable
10974@samp{$_}.
10975A count of the number of matches is stored in @samp{$numfound}.
10976
10977For example, if stopped at the @code{printf} in this function:
10978
10979@smallexample
10980void
10981hello ()
10982@{
10983 static char hello[] = "hello-hello";
10984 static struct @{ char c; short s; int i; @}
10985 __attribute__ ((packed)) mixed
10986 = @{ 'c', 0x1234, 0x87654321 @};
10987 printf ("%s\n", hello);
10988@}
10989@end smallexample
10990
10991@noindent
10992you get during debugging:
10993
10994@smallexample
10995(gdb) find &hello[0], +sizeof(hello), "hello"
109960x804956d <hello.1620+6>
109971 pattern found
10998(gdb) find &hello[0], +sizeof(hello), 'h', 'e', 'l', 'l', 'o'
109990x8049567 <hello.1620>
110000x804956d <hello.1620+6>
110012 patterns found
11002(gdb) find /b1 &hello[0], +sizeof(hello), 'h', 0x65, 'l'
110030x8049567 <hello.1620>
110041 pattern found
11005(gdb) find &mixed, +sizeof(mixed), (char) 'c', (short) 0x1234, (int) 0x87654321
110060x8049560 <mixed.1625>
110071 pattern found
11008(gdb) print $numfound
11009$1 = 1
11010(gdb) print $_
11011$2 = (void *) 0x8049560
11012@end smallexample
11013
11014@node Optimized Code
11015@chapter Debugging Optimized Code
11016@cindex optimized code, debugging
11017@cindex debugging optimized code
11018
11019Almost all compilers support optimization. With optimization
11020disabled, the compiler generates assembly code that corresponds
11021directly to your source code, in a simplistic way. As the compiler
11022applies more powerful optimizations, the generated assembly code
11023diverges from your original source code. With help from debugging
11024information generated by the compiler, @value{GDBN} can map from
11025the running program back to constructs from your original source.
11026
11027@value{GDBN} is more accurate with optimization disabled. If you
11028can recompile without optimization, it is easier to follow the
11029progress of your program during debugging. But, there are many cases
11030where you may need to debug an optimized version.
11031
11032When you debug a program compiled with @samp{-g -O}, remember that the
11033optimizer has rearranged your code; the debugger shows you what is
11034really there. Do not be too surprised when the execution path does not
11035exactly match your source file! An extreme example: if you define a
11036variable, but never use it, @value{GDBN} never sees that
11037variable---because the compiler optimizes it out of existence.
11038
11039Some things do not work as well with @samp{-g -O} as with just
11040@samp{-g}, particularly on machines with instruction scheduling. If in
11041doubt, recompile with @samp{-g} alone, and if this fixes the problem,
11042please report it to us as a bug (including a test case!).
11043@xref{Variables}, for more information about debugging optimized code.
11044
11045@menu
11046* Inline Functions:: How @value{GDBN} presents inlining
11047* Tail Call Frames:: @value{GDBN} analysis of jumps to functions
11048@end menu
11049
11050@node Inline Functions
11051@section Inline Functions
11052@cindex inline functions, debugging
11053
11054@dfn{Inlining} is an optimization that inserts a copy of the function
11055body directly at each call site, instead of jumping to a shared
11056routine. @value{GDBN} displays inlined functions just like
11057non-inlined functions. They appear in backtraces. You can view their
11058arguments and local variables, step into them with @code{step}, skip
11059them with @code{next}, and escape from them with @code{finish}.
11060You can check whether a function was inlined by using the
11061@code{info frame} command.
11062
11063For @value{GDBN} to support inlined functions, the compiler must
11064record information about inlining in the debug information ---
11065@value{NGCC} using the @sc{dwarf 2} format does this, and several
11066other compilers do also. @value{GDBN} only supports inlined functions
11067when using @sc{dwarf 2}. Versions of @value{NGCC} before 4.1
11068do not emit two required attributes (@samp{DW_AT_call_file} and
11069@samp{DW_AT_call_line}); @value{GDBN} does not display inlined
11070function calls with earlier versions of @value{NGCC}. It instead
11071displays the arguments and local variables of inlined functions as
11072local variables in the caller.
11073
11074The body of an inlined function is directly included at its call site;
11075unlike a non-inlined function, there are no instructions devoted to
11076the call. @value{GDBN} still pretends that the call site and the
11077start of the inlined function are different instructions. Stepping to
11078the call site shows the call site, and then stepping again shows
11079the first line of the inlined function, even though no additional
11080instructions are executed.
11081
11082This makes source-level debugging much clearer; you can see both the
11083context of the call and then the effect of the call. Only stepping by
11084a single instruction using @code{stepi} or @code{nexti} does not do
11085this; single instruction steps always show the inlined body.
11086
11087There are some ways that @value{GDBN} does not pretend that inlined
11088function calls are the same as normal calls:
11089
11090@itemize @bullet
11091@item
11092Setting breakpoints at the call site of an inlined function may not
11093work, because the call site does not contain any code. @value{GDBN}
11094may incorrectly move the breakpoint to the next line of the enclosing
11095function, after the call. This limitation will be removed in a future
11096version of @value{GDBN}; until then, set a breakpoint on an earlier line
11097or inside the inlined function instead.
11098
11099@item
11100@value{GDBN} cannot locate the return value of inlined calls after
11101using the @code{finish} command. This is a limitation of compiler-generated
11102debugging information; after @code{finish}, you can step to the next line
11103and print a variable where your program stored the return value.
11104
11105@end itemize
11106
11107@node Tail Call Frames
11108@section Tail Call Frames
11109@cindex tail call frames, debugging
11110
11111Function @code{B} can call function @code{C} in its very last statement. In
11112unoptimized compilation the call of @code{C} is immediately followed by return
11113instruction at the end of @code{B} code. Optimizing compiler may replace the
11114call and return in function @code{B} into one jump to function @code{C}
11115instead. Such use of a jump instruction is called @dfn{tail call}.
11116
11117During execution of function @code{C}, there will be no indication in the
11118function call stack frames that it was tail-called from @code{B}. If function
11119@code{A} regularly calls function @code{B} which tail-calls function @code{C},
11120then @value{GDBN} will see @code{A} as the caller of @code{C}. However, in
11121some cases @value{GDBN} can determine that @code{C} was tail-called from
11122@code{B}, and it will then create fictitious call frame for that, with the
11123return address set up as if @code{B} called @code{C} normally.
11124
11125This functionality is currently supported only by DWARF 2 debugging format and
11126the compiler has to produce @samp{DW_TAG_GNU_call_site} tags. With
11127@value{NGCC}, you need to specify @option{-O -g} during compilation, to get
11128this information.
11129
11130@kbd{info frame} command (@pxref{Frame Info}) will indicate the tail call frame
11131kind by text @code{tail call frame} such as in this sample @value{GDBN} output:
11132
11133@smallexample
11134(gdb) x/i $pc - 2
11135 0x40066b <b(int, double)+11>: jmp 0x400640 <c(int, double)>
11136(gdb) info frame
11137Stack level 1, frame at 0x7fffffffda30:
11138 rip = 0x40066d in b (amd64-entry-value.cc:59); saved rip 0x4004c5
11139 tail call frame, caller of frame at 0x7fffffffda30
11140 source language c++.
11141 Arglist at unknown address.
11142 Locals at unknown address, Previous frame's sp is 0x7fffffffda30
11143@end smallexample
11144
11145The detection of all the possible code path executions can find them ambiguous.
11146There is no execution history stored (possible @ref{Reverse Execution} is never
11147used for this purpose) and the last known caller could have reached the known
11148callee by multiple different jump sequences. In such case @value{GDBN} still
11149tries to show at least all the unambiguous top tail callers and all the
11150unambiguous bottom tail calees, if any.
11151
11152@table @code
11153@anchor{set debug entry-values}
11154@item set debug entry-values
11155@kindex set debug entry-values
11156When set to on, enables printing of analysis messages for both frame argument
11157values at function entry and tail calls. It will show all the possible valid
11158tail calls code paths it has considered. It will also print the intersection
11159of them with the final unambiguous (possibly partial or even empty) code path
11160result.
11161
11162@item show debug entry-values
11163@kindex show debug entry-values
11164Show the current state of analysis messages printing for both frame argument
11165values at function entry and tail calls.
11166@end table
11167
11168The analysis messages for tail calls can for example show why the virtual tail
11169call frame for function @code{c} has not been recognized (due to the indirect
11170reference by variable @code{x}):
11171
11172@smallexample
11173static void __attribute__((noinline, noclone)) c (void);
11174void (*x) (void) = c;
11175static void __attribute__((noinline, noclone)) a (void) @{ x++; @}
11176static void __attribute__((noinline, noclone)) c (void) @{ a (); @}
11177int main (void) @{ x (); return 0; @}
11178
11179Breakpoint 1, DW_OP_GNU_entry_value resolving cannot find
11180DW_TAG_GNU_call_site 0x40039a in main
11181a () at t.c:3
111823 static void __attribute__((noinline, noclone)) a (void) @{ x++; @}
11183(gdb) bt
11184#0 a () at t.c:3
11185#1 0x000000000040039a in main () at t.c:5
11186@end smallexample
11187
11188Another possibility is an ambiguous virtual tail call frames resolution:
11189
11190@smallexample
11191int i;
11192static void __attribute__((noinline, noclone)) f (void) @{ i++; @}
11193static void __attribute__((noinline, noclone)) e (void) @{ f (); @}
11194static void __attribute__((noinline, noclone)) d (void) @{ f (); @}
11195static void __attribute__((noinline, noclone)) c (void) @{ d (); @}
11196static void __attribute__((noinline, noclone)) b (void)
11197@{ if (i) c (); else e (); @}
11198static void __attribute__((noinline, noclone)) a (void) @{ b (); @}
11199int main (void) @{ a (); return 0; @}
11200
11201tailcall: initial: 0x4004d2(a) 0x4004ce(b) 0x4004b2(c) 0x4004a2(d)
11202tailcall: compare: 0x4004d2(a) 0x4004cc(b) 0x400492(e)
11203tailcall: reduced: 0x4004d2(a) |
11204(gdb) bt
11205#0 f () at t.c:2
11206#1 0x00000000004004d2 in a () at t.c:8
11207#2 0x0000000000400395 in main () at t.c:9
11208@end smallexample
11209
11210@set CALLSEQ1A @code{main@value{ARROW}a@value{ARROW}b@value{ARROW}c@value{ARROW}d@value{ARROW}f}
11211@set CALLSEQ2A @code{main@value{ARROW}a@value{ARROW}b@value{ARROW}e@value{ARROW}f}
11212
11213@c Convert CALLSEQ#A to CALLSEQ#B depending on HAVE_MAKEINFO_CLICK.
11214@ifset HAVE_MAKEINFO_CLICK
11215@set ARROW @click{}
11216@set CALLSEQ1B @clicksequence{@value{CALLSEQ1A}}
11217@set CALLSEQ2B @clicksequence{@value{CALLSEQ2A}}
11218@end ifset
11219@ifclear HAVE_MAKEINFO_CLICK
11220@set ARROW ->
11221@set CALLSEQ1B @value{CALLSEQ1A}
11222@set CALLSEQ2B @value{CALLSEQ2A}
11223@end ifclear
11224
11225Frames #0 and #2 are real, #1 is a virtual tail call frame.
11226The code can have possible execution paths @value{CALLSEQ1B} or
11227@value{CALLSEQ2B}, @value{GDBN} cannot find which one from the inferior state.
11228
11229@code{initial:} state shows some random possible calling sequence @value{GDBN}
11230has found. It then finds another possible calling sequcen - that one is
11231prefixed by @code{compare:}. The non-ambiguous intersection of these two is
11232printed as the @code{reduced:} calling sequence. That one could have many
11233futher @code{compare:} and @code{reduced:} statements as long as there remain
11234any non-ambiguous sequence entries.
11235
11236For the frame of function @code{b} in both cases there are different possible
11237@code{$pc} values (@code{0x4004cc} or @code{0x4004ce}), therefore this frame is
11238also ambigous. The only non-ambiguous frame is the one for function @code{a},
11239therefore this one is displayed to the user while the ambiguous frames are
11240omitted.
11241
11242There can be also reasons why printing of frame argument values at function
11243entry may fail:
11244
11245@smallexample
11246int v;
11247static void __attribute__((noinline, noclone)) c (int i) @{ v++; @}
11248static void __attribute__((noinline, noclone)) a (int i);
11249static void __attribute__((noinline, noclone)) b (int i) @{ a (i); @}
11250static void __attribute__((noinline, noclone)) a (int i)
11251@{ if (i) b (i - 1); else c (0); @}
11252int main (void) @{ a (5); return 0; @}
11253
11254(gdb) bt
11255#0 c (i=i@@entry=0) at t.c:2
11256#1 0x0000000000400428 in a (DW_OP_GNU_entry_value resolving has found
11257function "a" at 0x400420 can call itself via tail calls
11258i=<optimized out>) at t.c:6
11259#2 0x000000000040036e in main () at t.c:7
11260@end smallexample
11261
11262@value{GDBN} cannot find out from the inferior state if and how many times did
11263function @code{a} call itself (via function @code{b}) as these calls would be
11264tail calls. Such tail calls would modify thue @code{i} variable, therefore
11265@value{GDBN} cannot be sure the value it knows would be right - @value{GDBN}
11266prints @code{<optimized out>} instead.
11267
11268@node Macros
11269@chapter C Preprocessor Macros
11270
11271Some languages, such as C and C@t{++}, provide a way to define and invoke
11272``preprocessor macros'' which expand into strings of tokens.
11273@value{GDBN} can evaluate expressions containing macro invocations, show
11274the result of macro expansion, and show a macro's definition, including
11275where it was defined.
11276
11277You may need to compile your program specially to provide @value{GDBN}
11278with information about preprocessor macros. Most compilers do not
11279include macros in their debugging information, even when you compile
11280with the @option{-g} flag. @xref{Compilation}.
11281
11282A program may define a macro at one point, remove that definition later,
11283and then provide a different definition after that. Thus, at different
11284points in the program, a macro may have different definitions, or have
11285no definition at all. If there is a current stack frame, @value{GDBN}
11286uses the macros in scope at that frame's source code line. Otherwise,
11287@value{GDBN} uses the macros in scope at the current listing location;
11288see @ref{List}.
11289
11290Whenever @value{GDBN} evaluates an expression, it always expands any
11291macro invocations present in the expression. @value{GDBN} also provides
11292the following commands for working with macros explicitly.
11293
11294@table @code
11295
11296@kindex macro expand
11297@cindex macro expansion, showing the results of preprocessor
11298@cindex preprocessor macro expansion, showing the results of
11299@cindex expanding preprocessor macros
11300@item macro expand @var{expression}
11301@itemx macro exp @var{expression}
11302Show the results of expanding all preprocessor macro invocations in
11303@var{expression}. Since @value{GDBN} simply expands macros, but does
11304not parse the result, @var{expression} need not be a valid expression;
11305it can be any string of tokens.
11306
11307@kindex macro exp1
11308@item macro expand-once @var{expression}
11309@itemx macro exp1 @var{expression}
11310@cindex expand macro once
11311@i{(This command is not yet implemented.)} Show the results of
11312expanding those preprocessor macro invocations that appear explicitly in
11313@var{expression}. Macro invocations appearing in that expansion are
11314left unchanged. This command allows you to see the effect of a
11315particular macro more clearly, without being confused by further
11316expansions. Since @value{GDBN} simply expands macros, but does not
11317parse the result, @var{expression} need not be a valid expression; it
11318can be any string of tokens.
11319
11320@kindex info macro
11321@cindex macro definition, showing
11322@cindex definition of a macro, showing
11323@cindex macros, from debug info
11324@item info macro [-a|-all] [--] @var{macro}
11325Show the current definition or all definitions of the named @var{macro},
11326and describe the source location or compiler command-line where that
11327definition was established. The optional double dash is to signify the end of
11328argument processing and the beginning of @var{macro} for non C-like macros where
11329the macro may begin with a hyphen.
11330
11331@kindex info macros
11332@item info macros @var{linespec}
11333Show all macro definitions that are in effect at the location specified
11334by @var{linespec}, and describe the source location or compiler
11335command-line where those definitions were established.
11336
11337@kindex macro define
11338@cindex user-defined macros
11339@cindex defining macros interactively
11340@cindex macros, user-defined
11341@item macro define @var{macro} @var{replacement-list}
11342@itemx macro define @var{macro}(@var{arglist}) @var{replacement-list}
11343Introduce a definition for a preprocessor macro named @var{macro},
11344invocations of which are replaced by the tokens given in
11345@var{replacement-list}. The first form of this command defines an
11346``object-like'' macro, which takes no arguments; the second form
11347defines a ``function-like'' macro, which takes the arguments given in
11348@var{arglist}.
11349
11350A definition introduced by this command is in scope in every
11351expression evaluated in @value{GDBN}, until it is removed with the
11352@code{macro undef} command, described below. The definition overrides
11353all definitions for @var{macro} present in the program being debugged,
11354as well as any previous user-supplied definition.
11355
11356@kindex macro undef
11357@item macro undef @var{macro}
11358Remove any user-supplied definition for the macro named @var{macro}.
11359This command only affects definitions provided with the @code{macro
11360define} command, described above; it cannot remove definitions present
11361in the program being debugged.
11362
11363@kindex macro list
11364@item macro list
11365List all the macros defined using the @code{macro define} command.
11366@end table
11367
11368@cindex macros, example of debugging with
11369Here is a transcript showing the above commands in action. First, we
11370show our source files:
11371
11372@smallexample
11373$ cat sample.c
11374#include <stdio.h>
11375#include "sample.h"
11376
11377#define M 42
11378#define ADD(x) (M + x)
11379
11380main ()
11381@{
11382#define N 28
11383 printf ("Hello, world!\n");
11384#undef N
11385 printf ("We're so creative.\n");
11386#define N 1729
11387 printf ("Goodbye, world!\n");
11388@}
11389$ cat sample.h
11390#define Q <
11391$
11392@end smallexample
11393
11394Now, we compile the program using the @sc{gnu} C compiler,
11395@value{NGCC}. We pass the @option{-gdwarf-2}@footnote{This is the
11396minimum. Recent versions of @value{NGCC} support @option{-gdwarf-3}
11397and @option{-gdwarf-4}; we recommend always choosing the most recent
11398version of DWARF.} @emph{and} @option{-g3} flags to ensure the compiler
11399includes information about preprocessor macros in the debugging
11400information.
11401
11402@smallexample
11403$ gcc -gdwarf-2 -g3 sample.c -o sample
11404$
11405@end smallexample
11406
11407Now, we start @value{GDBN} on our sample program:
11408
11409@smallexample
11410$ gdb -nw sample
11411GNU gdb 2002-05-06-cvs
11412Copyright 2002 Free Software Foundation, Inc.
11413GDB is free software, @dots{}
11414(@value{GDBP})
11415@end smallexample
11416
11417We can expand macros and examine their definitions, even when the
11418program is not running. @value{GDBN} uses the current listing position
11419to decide which macro definitions are in scope:
11420
11421@smallexample
11422(@value{GDBP}) list main
114233
114244 #define M 42
114255 #define ADD(x) (M + x)
114266
114277 main ()
114288 @{
114299 #define N 28
1143010 printf ("Hello, world!\n");
1143111 #undef N
1143212 printf ("We're so creative.\n");
11433(@value{GDBP}) info macro ADD
11434Defined at /home/jimb/gdb/macros/play/sample.c:5
11435#define ADD(x) (M + x)
11436(@value{GDBP}) info macro Q
11437Defined at /home/jimb/gdb/macros/play/sample.h:1
11438 included at /home/jimb/gdb/macros/play/sample.c:2
11439#define Q <
11440(@value{GDBP}) macro expand ADD(1)
11441expands to: (42 + 1)
11442(@value{GDBP}) macro expand-once ADD(1)
11443expands to: once (M + 1)
11444(@value{GDBP})
11445@end smallexample
11446
11447In the example above, note that @code{macro expand-once} expands only
11448the macro invocation explicit in the original text --- the invocation of
11449@code{ADD} --- but does not expand the invocation of the macro @code{M},
11450which was introduced by @code{ADD}.
11451
11452Once the program is running, @value{GDBN} uses the macro definitions in
11453force at the source line of the current stack frame:
11454
11455@smallexample
11456(@value{GDBP}) break main
11457Breakpoint 1 at 0x8048370: file sample.c, line 10.
11458(@value{GDBP}) run
11459Starting program: /home/jimb/gdb/macros/play/sample
11460
11461Breakpoint 1, main () at sample.c:10
1146210 printf ("Hello, world!\n");
11463(@value{GDBP})
11464@end smallexample
11465
11466At line 10, the definition of the macro @code{N} at line 9 is in force:
11467
11468@smallexample
11469(@value{GDBP}) info macro N
11470Defined at /home/jimb/gdb/macros/play/sample.c:9
11471#define N 28
11472(@value{GDBP}) macro expand N Q M
11473expands to: 28 < 42
11474(@value{GDBP}) print N Q M
11475$1 = 1
11476(@value{GDBP})
11477@end smallexample
11478
11479As we step over directives that remove @code{N}'s definition, and then
11480give it a new definition, @value{GDBN} finds the definition (or lack
11481thereof) in force at each point:
11482
11483@smallexample
11484(@value{GDBP}) next
11485Hello, world!
1148612 printf ("We're so creative.\n");
11487(@value{GDBP}) info macro N
11488The symbol `N' has no definition as a C/C++ preprocessor macro
11489at /home/jimb/gdb/macros/play/sample.c:12
11490(@value{GDBP}) next
11491We're so creative.
1149214 printf ("Goodbye, world!\n");
11493(@value{GDBP}) info macro N
11494Defined at /home/jimb/gdb/macros/play/sample.c:13
11495#define N 1729
11496(@value{GDBP}) macro expand N Q M
11497expands to: 1729 < 42
11498(@value{GDBP}) print N Q M
11499$2 = 0
11500(@value{GDBP})
11501@end smallexample
11502
11503In addition to source files, macros can be defined on the compilation command
11504line using the @option{-D@var{name}=@var{value}} syntax. For macros defined in
11505such a way, @value{GDBN} displays the location of their definition as line zero
11506of the source file submitted to the compiler.
11507
11508@smallexample
11509(@value{GDBP}) info macro __STDC__
11510Defined at /home/jimb/gdb/macros/play/sample.c:0
11511-D__STDC__=1
11512(@value{GDBP})
11513@end smallexample
11514
11515
11516@node Tracepoints
11517@chapter Tracepoints
11518@c This chapter is based on the documentation written by Michael
11519@c Snyder, David Taylor, Jim Blandy, and Elena Zannoni.
11520
11521@cindex tracepoints
11522In some applications, it is not feasible for the debugger to interrupt
11523the program's execution long enough for the developer to learn
11524anything helpful about its behavior. If the program's correctness
11525depends on its real-time behavior, delays introduced by a debugger
11526might cause the program to change its behavior drastically, or perhaps
11527fail, even when the code itself is correct. It is useful to be able
11528to observe the program's behavior without interrupting it.
11529
11530Using @value{GDBN}'s @code{trace} and @code{collect} commands, you can
11531specify locations in the program, called @dfn{tracepoints}, and
11532arbitrary expressions to evaluate when those tracepoints are reached.
11533Later, using the @code{tfind} command, you can examine the values
11534those expressions had when the program hit the tracepoints. The
11535expressions may also denote objects in memory---structures or arrays,
11536for example---whose values @value{GDBN} should record; while visiting
11537a particular tracepoint, you may inspect those objects as if they were
11538in memory at that moment. However, because @value{GDBN} records these
11539values without interacting with you, it can do so quickly and
11540unobtrusively, hopefully not disturbing the program's behavior.
11541
11542The tracepoint facility is currently available only for remote
11543targets. @xref{Targets}. In addition, your remote target must know
11544how to collect trace data. This functionality is implemented in the
11545remote stub; however, none of the stubs distributed with @value{GDBN}
11546support tracepoints as of this writing. The format of the remote
11547packets used to implement tracepoints are described in @ref{Tracepoint
11548Packets}.
11549
11550It is also possible to get trace data from a file, in a manner reminiscent
11551of corefiles; you specify the filename, and use @code{tfind} to search
11552through the file. @xref{Trace Files}, for more details.
11553
11554This chapter describes the tracepoint commands and features.
11555
11556@menu
11557* Set Tracepoints::
11558* Analyze Collected Data::
11559* Tracepoint Variables::
11560* Trace Files::
11561@end menu
11562
11563@node Set Tracepoints
11564@section Commands to Set Tracepoints
11565
11566Before running such a @dfn{trace experiment}, an arbitrary number of
11567tracepoints can be set. A tracepoint is actually a special type of
11568breakpoint (@pxref{Set Breaks}), so you can manipulate it using
11569standard breakpoint commands. For instance, as with breakpoints,
11570tracepoint numbers are successive integers starting from one, and many
11571of the commands associated with tracepoints take the tracepoint number
11572as their argument, to identify which tracepoint to work on.
11573
11574For each tracepoint, you can specify, in advance, some arbitrary set
11575of data that you want the target to collect in the trace buffer when
11576it hits that tracepoint. The collected data can include registers,
11577local variables, or global data. Later, you can use @value{GDBN}
11578commands to examine the values these data had at the time the
11579tracepoint was hit.
11580
11581Tracepoints do not support every breakpoint feature. Ignore counts on
11582tracepoints have no effect, and tracepoints cannot run @value{GDBN}
11583commands when they are hit. Tracepoints may not be thread-specific
11584either.
11585
11586@cindex fast tracepoints
11587Some targets may support @dfn{fast tracepoints}, which are inserted in
11588a different way (such as with a jump instead of a trap), that is
11589faster but possibly restricted in where they may be installed.
11590
11591@cindex static tracepoints
11592@cindex markers, static tracepoints
11593@cindex probing markers, static tracepoints
11594Regular and fast tracepoints are dynamic tracing facilities, meaning
11595that they can be used to insert tracepoints at (almost) any location
11596in the target. Some targets may also support controlling @dfn{static
11597tracepoints} from @value{GDBN}. With static tracing, a set of
11598instrumentation points, also known as @dfn{markers}, are embedded in
11599the target program, and can be activated or deactivated by name or
11600address. These are usually placed at locations which facilitate
11601investigating what the target is actually doing. @value{GDBN}'s
11602support for static tracing includes being able to list instrumentation
11603points, and attach them with @value{GDBN} defined high level
11604tracepoints that expose the whole range of convenience of
11605@value{GDBN}'s tracepoints support. Namely, support for collecting
11606registers values and values of global or local (to the instrumentation
11607point) variables; tracepoint conditions and trace state variables.
11608The act of installing a @value{GDBN} static tracepoint on an
11609instrumentation point, or marker, is referred to as @dfn{probing} a
11610static tracepoint marker.
11611
11612@code{gdbserver} supports tracepoints on some target systems.
11613@xref{Server,,Tracepoints support in @code{gdbserver}}.
11614
11615This section describes commands to set tracepoints and associated
11616conditions and actions.
11617
11618@menu
11619* Create and Delete Tracepoints::
11620* Enable and Disable Tracepoints::
11621* Tracepoint Passcounts::
11622* Tracepoint Conditions::
11623* Trace State Variables::
11624* Tracepoint Actions::
11625* Listing Tracepoints::
11626* Listing Static Tracepoint Markers::
11627* Starting and Stopping Trace Experiments::
11628* Tracepoint Restrictions::
11629@end menu
11630
11631@node Create and Delete Tracepoints
11632@subsection Create and Delete Tracepoints
11633
11634@table @code
11635@cindex set tracepoint
11636@kindex trace
11637@item trace @var{location}
11638The @code{trace} command is very similar to the @code{break} command.
11639Its argument @var{location} can be a source line, a function name, or
11640an address in the target program. @xref{Specify Location}. The
11641@code{trace} command defines a tracepoint, which is a point in the
11642target program where the debugger will briefly stop, collect some
11643data, and then allow the program to continue. Setting a tracepoint or
11644changing its actions takes effect immediately if the remote stub
11645supports the @samp{InstallInTrace} feature (@pxref{install tracepoint
11646in tracing}).
11647If remote stub doesn't support the @samp{InstallInTrace} feature, all
11648these changes don't take effect until the next @code{tstart}
11649command, and once a trace experiment is running, further changes will
11650not have any effect until the next trace experiment starts. In addition,
11651@value{GDBN} supports @dfn{pending tracepoints}---tracepoints whose
11652address is not yet resolved. (This is similar to pending breakpoints.)
11653Pending tracepoints are not downloaded to the target and not installed
11654until they are resolved. The resolution of pending tracepoints requires
11655@value{GDBN} support---when debugging with the remote target, and
11656@value{GDBN} disconnects from the remote stub (@pxref{disconnected
11657tracing}), pending tracepoints can not be resolved (and downloaded to
11658the remote stub) while @value{GDBN} is disconnected.
11659
11660Here are some examples of using the @code{trace} command:
11661
11662@smallexample
11663(@value{GDBP}) @b{trace foo.c:121} // a source file and line number
11664
11665(@value{GDBP}) @b{trace +2} // 2 lines forward
11666
11667(@value{GDBP}) @b{trace my_function} // first source line of function
11668
11669(@value{GDBP}) @b{trace *my_function} // EXACT start address of function
11670
11671(@value{GDBP}) @b{trace *0x2117c4} // an address
11672@end smallexample
11673
11674@noindent
11675You can abbreviate @code{trace} as @code{tr}.
11676
11677@item trace @var{location} if @var{cond}
11678Set a tracepoint with condition @var{cond}; evaluate the expression
11679@var{cond} each time the tracepoint is reached, and collect data only
11680if the value is nonzero---that is, if @var{cond} evaluates as true.
11681@xref{Tracepoint Conditions, ,Tracepoint Conditions}, for more
11682information on tracepoint conditions.
11683
11684@item ftrace @var{location} [ if @var{cond} ]
11685@cindex set fast tracepoint
11686@cindex fast tracepoints, setting
11687@kindex ftrace
11688The @code{ftrace} command sets a fast tracepoint. For targets that
11689support them, fast tracepoints will use a more efficient but possibly
11690less general technique to trigger data collection, such as a jump
11691instruction instead of a trap, or some sort of hardware support. It
11692may not be possible to create a fast tracepoint at the desired
11693location, in which case the command will exit with an explanatory
11694message.
11695
11696@value{GDBN} handles arguments to @code{ftrace} exactly as for
11697@code{trace}.
11698
11699On 32-bit x86-architecture systems, fast tracepoints normally need to
11700be placed at an instruction that is 5 bytes or longer, but can be
11701placed at 4-byte instructions if the low 64K of memory of the target
11702program is available to install trampolines. Some Unix-type systems,
11703such as @sc{gnu}/Linux, exclude low addresses from the program's
11704address space; but for instance with the Linux kernel it is possible
11705to let @value{GDBN} use this area by doing a @command{sysctl} command
11706to set the @code{mmap_min_addr} kernel parameter, as in
11707
11708@example
11709sudo sysctl -w vm.mmap_min_addr=32768
11710@end example
11711
11712@noindent
11713which sets the low address to 32K, which leaves plenty of room for
11714trampolines. The minimum address should be set to a page boundary.
11715
11716@item strace @var{location} [ if @var{cond} ]
11717@cindex set static tracepoint
11718@cindex static tracepoints, setting
11719@cindex probe static tracepoint marker
11720@kindex strace
11721The @code{strace} command sets a static tracepoint. For targets that
11722support it, setting a static tracepoint probes a static
11723instrumentation point, or marker, found at @var{location}. It may not
11724be possible to set a static tracepoint at the desired location, in
11725which case the command will exit with an explanatory message.
11726
11727@value{GDBN} handles arguments to @code{strace} exactly as for
11728@code{trace}, with the addition that the user can also specify
11729@code{-m @var{marker}} as @var{location}. This probes the marker
11730identified by the @var{marker} string identifier. This identifier
11731depends on the static tracepoint backend library your program is
11732using. You can find all the marker identifiers in the @samp{ID} field
11733of the @code{info static-tracepoint-markers} command output.
11734@xref{Listing Static Tracepoint Markers,,Listing Static Tracepoint
11735Markers}. For example, in the following small program using the UST
11736tracing engine:
11737
11738@smallexample
11739main ()
11740@{
11741 trace_mark(ust, bar33, "str %s", "FOOBAZ");
11742@}
11743@end smallexample
11744
11745@noindent
11746the marker id is composed of joining the first two arguments to the
11747@code{trace_mark} call with a slash, which translates to:
11748
11749@smallexample
11750(@value{GDBP}) info static-tracepoint-markers
11751Cnt Enb ID Address What
117521 n ust/bar33 0x0000000000400ddc in main at stexample.c:22
11753 Data: "str %s"
11754[etc...]
11755@end smallexample
11756
11757@noindent
11758so you may probe the marker above with:
11759
11760@smallexample
11761(@value{GDBP}) strace -m ust/bar33
11762@end smallexample
11763
11764Static tracepoints accept an extra collect action --- @code{collect
11765$_sdata}. This collects arbitrary user data passed in the probe point
11766call to the tracing library. In the UST example above, you'll see
11767that the third argument to @code{trace_mark} is a printf-like format
11768string. The user data is then the result of running that formating
11769string against the following arguments. Note that @code{info
11770static-tracepoint-markers} command output lists that format string in
11771the @samp{Data:} field.
11772
11773You can inspect this data when analyzing the trace buffer, by printing
11774the $_sdata variable like any other variable available to
11775@value{GDBN}. @xref{Tracepoint Actions,,Tracepoint Action Lists}.
11776
11777@vindex $tpnum
11778@cindex last tracepoint number
11779@cindex recent tracepoint number
11780@cindex tracepoint number
11781The convenience variable @code{$tpnum} records the tracepoint number
11782of the most recently set tracepoint.
11783
11784@kindex delete tracepoint
11785@cindex tracepoint deletion
11786@item delete tracepoint @r{[}@var{num}@r{]}
11787Permanently delete one or more tracepoints. With no argument, the
11788default is to delete all tracepoints. Note that the regular
11789@code{delete} command can remove tracepoints also.
11790
11791Examples:
11792
11793@smallexample
11794(@value{GDBP}) @b{delete trace 1 2 3} // remove three tracepoints
11795
11796(@value{GDBP}) @b{delete trace} // remove all tracepoints
11797@end smallexample
11798
11799@noindent
11800You can abbreviate this command as @code{del tr}.
11801@end table
11802
11803@node Enable and Disable Tracepoints
11804@subsection Enable and Disable Tracepoints
11805
11806These commands are deprecated; they are equivalent to plain @code{disable} and @code{enable}.
11807
11808@table @code
11809@kindex disable tracepoint
11810@item disable tracepoint @r{[}@var{num}@r{]}
11811Disable tracepoint @var{num}, or all tracepoints if no argument
11812@var{num} is given. A disabled tracepoint will have no effect during
11813a trace experiment, but it is not forgotten. You can re-enable
11814a disabled tracepoint using the @code{enable tracepoint} command.
11815If the command is issued during a trace experiment and the debug target
11816has support for disabling tracepoints during a trace experiment, then the
11817change will be effective immediately. Otherwise, it will be applied to the
11818next trace experiment.
11819
11820@kindex enable tracepoint
11821@item enable tracepoint @r{[}@var{num}@r{]}
11822Enable tracepoint @var{num}, or all tracepoints. If this command is
11823issued during a trace experiment and the debug target supports enabling
11824tracepoints during a trace experiment, then the enabled tracepoints will
11825become effective immediately. Otherwise, they will become effective the
11826next time a trace experiment is run.
11827@end table
11828
11829@node Tracepoint Passcounts
11830@subsection Tracepoint Passcounts
11831
11832@table @code
11833@kindex passcount
11834@cindex tracepoint pass count
11835@item passcount @r{[}@var{n} @r{[}@var{num}@r{]]}
11836Set the @dfn{passcount} of a tracepoint. The passcount is a way to
11837automatically stop a trace experiment. If a tracepoint's passcount is
11838@var{n}, then the trace experiment will be automatically stopped on
11839the @var{n}'th time that tracepoint is hit. If the tracepoint number
11840@var{num} is not specified, the @code{passcount} command sets the
11841passcount of the most recently defined tracepoint. If no passcount is
11842given, the trace experiment will run until stopped explicitly by the
11843user.
11844
11845Examples:
11846
11847@smallexample
11848(@value{GDBP}) @b{passcount 5 2} // Stop on the 5th execution of
11849@exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// tracepoint 2}
11850
11851(@value{GDBP}) @b{passcount 12} // Stop on the 12th execution of the
11852@exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// most recently defined tracepoint.}
11853(@value{GDBP}) @b{trace foo}
11854(@value{GDBP}) @b{pass 3}
11855(@value{GDBP}) @b{trace bar}
11856(@value{GDBP}) @b{pass 2}
11857(@value{GDBP}) @b{trace baz}
11858(@value{GDBP}) @b{pass 1} // Stop tracing when foo has been
11859@exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// executed 3 times OR when bar has}
11860@exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// been executed 2 times}
11861@exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// OR when baz has been executed 1 time.}
11862@end smallexample
11863@end table
11864
11865@node Tracepoint Conditions
11866@subsection Tracepoint Conditions
11867@cindex conditional tracepoints
11868@cindex tracepoint conditions
11869
11870The simplest sort of tracepoint collects data every time your program
11871reaches a specified place. You can also specify a @dfn{condition} for
11872a tracepoint. A condition is just a Boolean expression in your
11873programming language (@pxref{Expressions, ,Expressions}). A
11874tracepoint with a condition evaluates the expression each time your
11875program reaches it, and data collection happens only if the condition
11876is true.
11877
11878Tracepoint conditions can be specified when a tracepoint is set, by
11879using @samp{if} in the arguments to the @code{trace} command.
11880@xref{Create and Delete Tracepoints, ,Setting Tracepoints}. They can
11881also be set or changed at any time with the @code{condition} command,
11882just as with breakpoints.
11883
11884Unlike breakpoint conditions, @value{GDBN} does not actually evaluate
11885the conditional expression itself. Instead, @value{GDBN} encodes the
11886expression into an agent expression (@pxref{Agent Expressions})
11887suitable for execution on the target, independently of @value{GDBN}.
11888Global variables become raw memory locations, locals become stack
11889accesses, and so forth.
11890
11891For instance, suppose you have a function that is usually called
11892frequently, but should not be called after an error has occurred. You
11893could use the following tracepoint command to collect data about calls
11894of that function that happen while the error code is propagating
11895through the program; an unconditional tracepoint could end up
11896collecting thousands of useless trace frames that you would have to
11897search through.
11898
11899@smallexample
11900(@value{GDBP}) @kbd{trace normal_operation if errcode > 0}
11901@end smallexample
11902
11903@node Trace State Variables
11904@subsection Trace State Variables
11905@cindex trace state variables
11906
11907A @dfn{trace state variable} is a special type of variable that is
11908created and managed by target-side code. The syntax is the same as
11909that for GDB's convenience variables (a string prefixed with ``$''),
11910but they are stored on the target. They must be created explicitly,
11911using a @code{tvariable} command. They are always 64-bit signed
11912integers.
11913
11914Trace state variables are remembered by @value{GDBN}, and downloaded
11915to the target along with tracepoint information when the trace
11916experiment starts. There are no intrinsic limits on the number of
11917trace state variables, beyond memory limitations of the target.
11918
11919@cindex convenience variables, and trace state variables
11920Although trace state variables are managed by the target, you can use
11921them in print commands and expressions as if they were convenience
11922variables; @value{GDBN} will get the current value from the target
11923while the trace experiment is running. Trace state variables share
11924the same namespace as other ``$'' variables, which means that you
11925cannot have trace state variables with names like @code{$23} or
11926@code{$pc}, nor can you have a trace state variable and a convenience
11927variable with the same name.
11928
11929@table @code
11930
11931@item tvariable $@var{name} [ = @var{expression} ]
11932@kindex tvariable
11933The @code{tvariable} command creates a new trace state variable named
11934@code{$@var{name}}, and optionally gives it an initial value of
11935@var{expression}. @var{expression} is evaluated when this command is
11936entered; the result will be converted to an integer if possible,
11937otherwise @value{GDBN} will report an error. A subsequent
11938@code{tvariable} command specifying the same name does not create a
11939variable, but instead assigns the supplied initial value to the
11940existing variable of that name, overwriting any previous initial
11941value. The default initial value is 0.
11942
11943@item info tvariables
11944@kindex info tvariables
11945List all the trace state variables along with their initial values.
11946Their current values may also be displayed, if the trace experiment is
11947currently running.
11948
11949@item delete tvariable @r{[} $@var{name} @dots{} @r{]}
11950@kindex delete tvariable
11951Delete the given trace state variables, or all of them if no arguments
11952are specified.
11953
11954@end table
11955
11956@node Tracepoint Actions
11957@subsection Tracepoint Action Lists
11958
11959@table @code
11960@kindex actions
11961@cindex tracepoint actions
11962@item actions @r{[}@var{num}@r{]}
11963This command will prompt for a list of actions to be taken when the
11964tracepoint is hit. If the tracepoint number @var{num} is not
11965specified, this command sets the actions for the one that was most
11966recently defined (so that you can define a tracepoint and then say
11967@code{actions} without bothering about its number). You specify the
11968actions themselves on the following lines, one action at a time, and
11969terminate the actions list with a line containing just @code{end}. So
11970far, the only defined actions are @code{collect}, @code{teval}, and
11971@code{while-stepping}.
11972
11973@code{actions} is actually equivalent to @code{commands} (@pxref{Break
11974Commands, ,Breakpoint Command Lists}), except that only the defined
11975actions are allowed; any other @value{GDBN} command is rejected.
11976
11977@cindex remove actions from a tracepoint
11978To remove all actions from a tracepoint, type @samp{actions @var{num}}
11979and follow it immediately with @samp{end}.
11980
11981@smallexample
11982(@value{GDBP}) @b{collect @var{data}} // collect some data
11983
11984(@value{GDBP}) @b{while-stepping 5} // single-step 5 times, collect data
11985
11986(@value{GDBP}) @b{end} // signals the end of actions.
11987@end smallexample
11988
11989In the following example, the action list begins with @code{collect}
11990commands indicating the things to be collected when the tracepoint is
11991hit. Then, in order to single-step and collect additional data
11992following the tracepoint, a @code{while-stepping} command is used,
11993followed by the list of things to be collected after each step in a
11994sequence of single steps. The @code{while-stepping} command is
11995terminated by its own separate @code{end} command. Lastly, the action
11996list is terminated by an @code{end} command.
11997
11998@smallexample
11999(@value{GDBP}) @b{trace foo}
12000(@value{GDBP}) @b{actions}
12001Enter actions for tracepoint 1, one per line:
12002> collect bar,baz
12003> collect $regs
12004> while-stepping 12
12005 > collect $pc, arr[i]
12006 > end
12007end
12008@end smallexample
12009
12010@kindex collect @r{(tracepoints)}
12011@item collect@r{[}/@var{mods}@r{]} @var{expr1}, @var{expr2}, @dots{}
12012Collect values of the given expressions when the tracepoint is hit.
12013This command accepts a comma-separated list of any valid expressions.
12014In addition to global, static, or local variables, the following
12015special arguments are supported:
12016
12017@table @code
12018@item $regs
12019Collect all registers.
12020
12021@item $args
12022Collect all function arguments.
12023
12024@item $locals
12025Collect all local variables.
12026
12027@item $_ret
12028Collect the return address. This is helpful if you want to see more
12029of a backtrace.
12030
12031@item $_probe_argc
12032Collects the number of arguments from the static probe at which the
12033tracepoint is located.
12034@xref{Static Probe Points}.
12035
12036@item $_probe_arg@var{n}
12037@var{n} is an integer between 0 and 11. Collects the @var{n}th argument
12038from the static probe at which the tracepoint is located.
12039@xref{Static Probe Points}.
12040
12041@item $_sdata
12042@vindex $_sdata@r{, collect}
12043Collect static tracepoint marker specific data. Only available for
12044static tracepoints. @xref{Tracepoint Actions,,Tracepoint Action
12045Lists}. On the UST static tracepoints library backend, an
12046instrumentation point resembles a @code{printf} function call. The
12047tracing library is able to collect user specified data formatted to a
12048character string using the format provided by the programmer that
12049instrumented the program. Other backends have similar mechanisms.
12050Here's an example of a UST marker call:
12051
12052@smallexample
12053 const char master_name[] = "$your_name";
12054 trace_mark(channel1, marker1, "hello %s", master_name)
12055@end smallexample
12056
12057In this case, collecting @code{$_sdata} collects the string
12058@samp{hello $yourname}. When analyzing the trace buffer, you can
12059inspect @samp{$_sdata} like any other variable available to
12060@value{GDBN}.
12061@end table
12062
12063You can give several consecutive @code{collect} commands, each one
12064with a single argument, or one @code{collect} command with several
12065arguments separated by commas; the effect is the same.
12066
12067The optional @var{mods} changes the usual handling of the arguments.
12068@code{s} requests that pointers to chars be handled as strings, in
12069particular collecting the contents of the memory being pointed at, up
12070to the first zero. The upper bound is by default the value of the
12071@code{print elements} variable; if @code{s} is followed by a decimal
12072number, that is the upper bound instead. So for instance
12073@samp{collect/s25 mystr} collects as many as 25 characters at
12074@samp{mystr}.
12075
12076The command @code{info scope} (@pxref{Symbols, info scope}) is
12077particularly useful for figuring out what data to collect.
12078
12079@kindex teval @r{(tracepoints)}
12080@item teval @var{expr1}, @var{expr2}, @dots{}
12081Evaluate the given expressions when the tracepoint is hit. This
12082command accepts a comma-separated list of expressions. The results
12083are discarded, so this is mainly useful for assigning values to trace
12084state variables (@pxref{Trace State Variables}) without adding those
12085values to the trace buffer, as would be the case if the @code{collect}
12086action were used.
12087
12088@kindex while-stepping @r{(tracepoints)}
12089@item while-stepping @var{n}
12090Perform @var{n} single-step instruction traces after the tracepoint,
12091collecting new data after each step. The @code{while-stepping}
12092command is followed by the list of what to collect while stepping
12093(followed by its own @code{end} command):
12094
12095@smallexample
12096> while-stepping 12
12097 > collect $regs, myglobal
12098 > end
12099>
12100@end smallexample
12101
12102@noindent
12103Note that @code{$pc} is not automatically collected by
12104@code{while-stepping}; you need to explicitly collect that register if
12105you need it. You may abbreviate @code{while-stepping} as @code{ws} or
12106@code{stepping}.
12107
12108@item set default-collect @var{expr1}, @var{expr2}, @dots{}
12109@kindex set default-collect
12110@cindex default collection action
12111This variable is a list of expressions to collect at each tracepoint
12112hit. It is effectively an additional @code{collect} action prepended
12113to every tracepoint action list. The expressions are parsed
12114individually for each tracepoint, so for instance a variable named
12115@code{xyz} may be interpreted as a global for one tracepoint, and a
12116local for another, as appropriate to the tracepoint's location.
12117
12118@item show default-collect
12119@kindex show default-collect
12120Show the list of expressions that are collected by default at each
12121tracepoint hit.
12122
12123@end table
12124
12125@node Listing Tracepoints
12126@subsection Listing Tracepoints
12127
12128@table @code
12129@kindex info tracepoints @r{[}@var{n}@dots{}@r{]}
12130@kindex info tp @r{[}@var{n}@dots{}@r{]}
12131@cindex information about tracepoints
12132@item info tracepoints @r{[}@var{num}@dots{}@r{]}
12133Display information about the tracepoint @var{num}. If you don't
12134specify a tracepoint number, displays information about all the
12135tracepoints defined so far. The format is similar to that used for
12136@code{info breakpoints}; in fact, @code{info tracepoints} is the same
12137command, simply restricting itself to tracepoints.
12138
12139A tracepoint's listing may include additional information specific to
12140tracing:
12141
12142@itemize @bullet
12143@item
12144its passcount as given by the @code{passcount @var{n}} command
12145
12146@item
12147the state about installed on target of each location
12148@end itemize
12149
12150@smallexample
12151(@value{GDBP}) @b{info trace}
12152Num Type Disp Enb Address What
121531 tracepoint keep y 0x0804ab57 in foo() at main.cxx:7
12154 while-stepping 20
12155 collect globfoo, $regs
12156 end
12157 collect globfoo2
12158 end
12159 pass count 1200
121602 tracepoint keep y <MULTIPLE>
12161 collect $eip
121622.1 y 0x0804859c in func4 at change-loc.h:35
12163 installed on target
121642.2 y 0xb7ffc480 in func4 at change-loc.h:35
12165 installed on target
121662.3 y <PENDING> set_tracepoint
121673 tracepoint keep y 0x080485b1 in foo at change-loc.c:29
12168 not installed on target
12169(@value{GDBP})
12170@end smallexample
12171
12172@noindent
12173This command can be abbreviated @code{info tp}.
12174@end table
12175
12176@node Listing Static Tracepoint Markers
12177@subsection Listing Static Tracepoint Markers
12178
12179@table @code
12180@kindex info static-tracepoint-markers
12181@cindex information about static tracepoint markers
12182@item info static-tracepoint-markers
12183Display information about all static tracepoint markers defined in the
12184program.
12185
12186For each marker, the following columns are printed:
12187
12188@table @emph
12189@item Count
12190An incrementing counter, output to help readability. This is not a
12191stable identifier.
12192@item ID
12193The marker ID, as reported by the target.
12194@item Enabled or Disabled
12195Probed markers are tagged with @samp{y}. @samp{n} identifies marks
12196that are not enabled.
12197@item Address
12198Where the marker is in your program, as a memory address.
12199@item What
12200Where the marker is in the source for your program, as a file and line
12201number. If the debug information included in the program does not
12202allow @value{GDBN} to locate the source of the marker, this column
12203will be left blank.
12204@end table
12205
12206@noindent
12207In addition, the following information may be printed for each marker:
12208
12209@table @emph
12210@item Data
12211User data passed to the tracing library by the marker call. In the
12212UST backend, this is the format string passed as argument to the
12213marker call.
12214@item Static tracepoints probing the marker
12215The list of static tracepoints attached to the marker.
12216@end table
12217
12218@smallexample
12219(@value{GDBP}) info static-tracepoint-markers
12220Cnt ID Enb Address What
122211 ust/bar2 y 0x0000000000400e1a in main at stexample.c:25
12222 Data: number1 %d number2 %d
12223 Probed by static tracepoints: #2
122242 ust/bar33 n 0x0000000000400c87 in main at stexample.c:24
12225 Data: str %s
12226(@value{GDBP})
12227@end smallexample
12228@end table
12229
12230@node Starting and Stopping Trace Experiments
12231@subsection Starting and Stopping Trace Experiments
12232
12233@table @code
12234@kindex tstart [ @var{notes} ]
12235@cindex start a new trace experiment
12236@cindex collected data discarded
12237@item tstart
12238This command starts the trace experiment, and begins collecting data.
12239It has the side effect of discarding all the data collected in the
12240trace buffer during the previous trace experiment. If any arguments
12241are supplied, they are taken as a note and stored with the trace
12242experiment's state. The notes may be arbitrary text, and are
12243especially useful with disconnected tracing in a multi-user context;
12244the notes can explain what the trace is doing, supply user contact
12245information, and so forth.
12246
12247@kindex tstop [ @var{notes} ]
12248@cindex stop a running trace experiment
12249@item tstop
12250This command stops the trace experiment. If any arguments are
12251supplied, they are recorded with the experiment as a note. This is
12252useful if you are stopping a trace started by someone else, for
12253instance if the trace is interfering with the system's behavior and
12254needs to be stopped quickly.
12255
12256@strong{Note}: a trace experiment and data collection may stop
12257automatically if any tracepoint's passcount is reached
12258(@pxref{Tracepoint Passcounts}), or if the trace buffer becomes full.
12259
12260@kindex tstatus
12261@cindex status of trace data collection
12262@cindex trace experiment, status of
12263@item tstatus
12264This command displays the status of the current trace data
12265collection.
12266@end table
12267
12268Here is an example of the commands we described so far:
12269
12270@smallexample
12271(@value{GDBP}) @b{trace gdb_c_test}
12272(@value{GDBP}) @b{actions}
12273Enter actions for tracepoint #1, one per line.
12274> collect $regs,$locals,$args
12275> while-stepping 11
12276 > collect $regs
12277 > end
12278> end
12279(@value{GDBP}) @b{tstart}
12280 [time passes @dots{}]
12281(@value{GDBP}) @b{tstop}
12282@end smallexample
12283
12284@anchor{disconnected tracing}
12285@cindex disconnected tracing
12286You can choose to continue running the trace experiment even if
12287@value{GDBN} disconnects from the target, voluntarily or
12288involuntarily. For commands such as @code{detach}, the debugger will
12289ask what you want to do with the trace. But for unexpected
12290terminations (@value{GDBN} crash, network outage), it would be
12291unfortunate to lose hard-won trace data, so the variable
12292@code{disconnected-tracing} lets you decide whether the trace should
12293continue running without @value{GDBN}.
12294
12295@table @code
12296@item set disconnected-tracing on
12297@itemx set disconnected-tracing off
12298@kindex set disconnected-tracing
12299Choose whether a tracing run should continue to run if @value{GDBN}
12300has disconnected from the target. Note that @code{detach} or
12301@code{quit} will ask you directly what to do about a running trace no
12302matter what this variable's setting, so the variable is mainly useful
12303for handling unexpected situations, such as loss of the network.
12304
12305@item show disconnected-tracing
12306@kindex show disconnected-tracing
12307Show the current choice for disconnected tracing.
12308
12309@end table
12310
12311When you reconnect to the target, the trace experiment may or may not
12312still be running; it might have filled the trace buffer in the
12313meantime, or stopped for one of the other reasons. If it is running,
12314it will continue after reconnection.
12315
12316Upon reconnection, the target will upload information about the
12317tracepoints in effect. @value{GDBN} will then compare that
12318information to the set of tracepoints currently defined, and attempt
12319to match them up, allowing for the possibility that the numbers may
12320have changed due to creation and deletion in the meantime. If one of
12321the target's tracepoints does not match any in @value{GDBN}, the
12322debugger will create a new tracepoint, so that you have a number with
12323which to specify that tracepoint. This matching-up process is
12324necessarily heuristic, and it may result in useless tracepoints being
12325created; you may simply delete them if they are of no use.
12326
12327@cindex circular trace buffer
12328If your target agent supports a @dfn{circular trace buffer}, then you
12329can run a trace experiment indefinitely without filling the trace
12330buffer; when space runs out, the agent deletes already-collected trace
12331frames, oldest first, until there is enough room to continue
12332collecting. This is especially useful if your tracepoints are being
12333hit too often, and your trace gets terminated prematurely because the
12334buffer is full. To ask for a circular trace buffer, simply set
12335@samp{circular-trace-buffer} to on. You can set this at any time,
12336including during tracing; if the agent can do it, it will change
12337buffer handling on the fly, otherwise it will not take effect until
12338the next run.
12339
12340@table @code
12341@item set circular-trace-buffer on
12342@itemx set circular-trace-buffer off
12343@kindex set circular-trace-buffer
12344Choose whether a tracing run should use a linear or circular buffer
12345for trace data. A linear buffer will not lose any trace data, but may
12346fill up prematurely, while a circular buffer will discard old trace
12347data, but it will have always room for the latest tracepoint hits.
12348
12349@item show circular-trace-buffer
12350@kindex show circular-trace-buffer
12351Show the current choice for the trace buffer. Note that this may not
12352match the agent's current buffer handling, nor is it guaranteed to
12353match the setting that might have been in effect during a past run,
12354for instance if you are looking at frames from a trace file.
12355
12356@end table
12357
12358@table @code
12359@item set trace-buffer-size @var{n}
12360@itemx set trace-buffer-size unlimited
12361@kindex set trace-buffer-size
12362Request that the target use a trace buffer of @var{n} bytes. Not all
12363targets will honor the request; they may have a compiled-in size for
12364the trace buffer, or some other limitation. Set to a value of
12365@code{unlimited} or @code{-1} to let the target use whatever size it
12366likes. This is also the default.
12367
12368@item show trace-buffer-size
12369@kindex show trace-buffer-size
12370Show the current requested size for the trace buffer. Note that this
12371will only match the actual size if the target supports size-setting,
12372and was able to handle the requested size. For instance, if the
12373target can only change buffer size between runs, this variable will
12374not reflect the change until the next run starts. Use @code{tstatus}
12375to get a report of the actual buffer size.
12376@end table
12377
12378@table @code
12379@item set trace-user @var{text}
12380@kindex set trace-user
12381
12382@item show trace-user
12383@kindex show trace-user
12384
12385@item set trace-notes @var{text}
12386@kindex set trace-notes
12387Set the trace run's notes.
12388
12389@item show trace-notes
12390@kindex show trace-notes
12391Show the trace run's notes.
12392
12393@item set trace-stop-notes @var{text}
12394@kindex set trace-stop-notes
12395Set the trace run's stop notes. The handling of the note is as for
12396@code{tstop} arguments; the set command is convenient way to fix a
12397stop note that is mistaken or incomplete.
12398
12399@item show trace-stop-notes
12400@kindex show trace-stop-notes
12401Show the trace run's stop notes.
12402
12403@end table
12404
12405@node Tracepoint Restrictions
12406@subsection Tracepoint Restrictions
12407
12408@cindex tracepoint restrictions
12409There are a number of restrictions on the use of tracepoints. As
12410described above, tracepoint data gathering occurs on the target
12411without interaction from @value{GDBN}. Thus the full capabilities of
12412the debugger are not available during data gathering, and then at data
12413examination time, you will be limited by only having what was
12414collected. The following items describe some common problems, but it
12415is not exhaustive, and you may run into additional difficulties not
12416mentioned here.
12417
12418@itemize @bullet
12419
12420@item
12421Tracepoint expressions are intended to gather objects (lvalues). Thus
12422the full flexibility of GDB's expression evaluator is not available.
12423You cannot call functions, cast objects to aggregate types, access
12424convenience variables or modify values (except by assignment to trace
12425state variables). Some language features may implicitly call
12426functions (for instance Objective-C fields with accessors), and therefore
12427cannot be collected either.
12428
12429@item
12430Collection of local variables, either individually or in bulk with
12431@code{$locals} or @code{$args}, during @code{while-stepping} may
12432behave erratically. The stepping action may enter a new scope (for
12433instance by stepping into a function), or the location of the variable
12434may change (for instance it is loaded into a register). The
12435tracepoint data recorded uses the location information for the
12436variables that is correct for the tracepoint location. When the
12437tracepoint is created, it is not possible, in general, to determine
12438where the steps of a @code{while-stepping} sequence will advance the
12439program---particularly if a conditional branch is stepped.
12440
12441@item
12442Collection of an incompletely-initialized or partially-destroyed object
12443may result in something that @value{GDBN} cannot display, or displays
12444in a misleading way.
12445
12446@item
12447When @value{GDBN} displays a pointer to character it automatically
12448dereferences the pointer to also display characters of the string
12449being pointed to. However, collecting the pointer during tracing does
12450not automatically collect the string. You need to explicitly
12451dereference the pointer and provide size information if you want to
12452collect not only the pointer, but the memory pointed to. For example,
12453@code{*ptr@@50} can be used to collect the 50 element array pointed to
12454by @code{ptr}.
12455
12456@item
12457It is not possible to collect a complete stack backtrace at a
12458tracepoint. Instead, you may collect the registers and a few hundred
12459bytes from the stack pointer with something like @code{*(unsigned char *)$esp@@300}
12460(adjust to use the name of the actual stack pointer register on your
12461target architecture, and the amount of stack you wish to capture).
12462Then the @code{backtrace} command will show a partial backtrace when
12463using a trace frame. The number of stack frames that can be examined
12464depends on the sizes of the frames in the collected stack. Note that
12465if you ask for a block so large that it goes past the bottom of the
12466stack, the target agent may report an error trying to read from an
12467invalid address.
12468
12469@item
12470If you do not collect registers at a tracepoint, @value{GDBN} can
12471infer that the value of @code{$pc} must be the same as the address of
12472the tracepoint and use that when you are looking at a trace frame
12473for that tracepoint. However, this cannot work if the tracepoint has
12474multiple locations (for instance if it was set in a function that was
12475inlined), or if it has a @code{while-stepping} loop. In those cases
12476@value{GDBN} will warn you that it can't infer @code{$pc}, and default
12477it to zero.
12478
12479@end itemize
12480
12481@node Analyze Collected Data
12482@section Using the Collected Data
12483
12484After the tracepoint experiment ends, you use @value{GDBN} commands
12485for examining the trace data. The basic idea is that each tracepoint
12486collects a trace @dfn{snapshot} every time it is hit and another
12487snapshot every time it single-steps. All these snapshots are
12488consecutively numbered from zero and go into a buffer, and you can
12489examine them later. The way you examine them is to @dfn{focus} on a
12490specific trace snapshot. When the remote stub is focused on a trace
12491snapshot, it will respond to all @value{GDBN} requests for memory and
12492registers by reading from the buffer which belongs to that snapshot,
12493rather than from @emph{real} memory or registers of the program being
12494debugged. This means that @strong{all} @value{GDBN} commands
12495(@code{print}, @code{info registers}, @code{backtrace}, etc.) will
12496behave as if we were currently debugging the program state as it was
12497when the tracepoint occurred. Any requests for data that are not in
12498the buffer will fail.
12499
12500@menu
12501* tfind:: How to select a trace snapshot
12502* tdump:: How to display all data for a snapshot
12503* save tracepoints:: How to save tracepoints for a future run
12504@end menu
12505
12506@node tfind
12507@subsection @code{tfind @var{n}}
12508
12509@kindex tfind
12510@cindex select trace snapshot
12511@cindex find trace snapshot
12512The basic command for selecting a trace snapshot from the buffer is
12513@code{tfind @var{n}}, which finds trace snapshot number @var{n},
12514counting from zero. If no argument @var{n} is given, the next
12515snapshot is selected.
12516
12517Here are the various forms of using the @code{tfind} command.
12518
12519@table @code
12520@item tfind start
12521Find the first snapshot in the buffer. This is a synonym for
12522@code{tfind 0} (since 0 is the number of the first snapshot).
12523
12524@item tfind none
12525Stop debugging trace snapshots, resume @emph{live} debugging.
12526
12527@item tfind end
12528Same as @samp{tfind none}.
12529
12530@item tfind
12531No argument means find the next trace snapshot.
12532
12533@item tfind -
12534Find the previous trace snapshot before the current one. This permits
12535retracing earlier steps.
12536
12537@item tfind tracepoint @var{num}
12538Find the next snapshot associated with tracepoint @var{num}. Search
12539proceeds forward from the last examined trace snapshot. If no
12540argument @var{num} is given, it means find the next snapshot collected
12541for the same tracepoint as the current snapshot.
12542
12543@item tfind pc @var{addr}
12544Find the next snapshot associated with the value @var{addr} of the
12545program counter. Search proceeds forward from the last examined trace
12546snapshot. If no argument @var{addr} is given, it means find the next
12547snapshot with the same value of PC as the current snapshot.
12548
12549@item tfind outside @var{addr1}, @var{addr2}
12550Find the next snapshot whose PC is outside the given range of
12551addresses (exclusive).
12552
12553@item tfind range @var{addr1}, @var{addr2}
12554Find the next snapshot whose PC is between @var{addr1} and
12555@var{addr2} (inclusive).
12556
12557@item tfind line @r{[}@var{file}:@r{]}@var{n}
12558Find the next snapshot associated with the source line @var{n}. If
12559the optional argument @var{file} is given, refer to line @var{n} in
12560that source file. Search proceeds forward from the last examined
12561trace snapshot. If no argument @var{n} is given, it means find the
12562next line other than the one currently being examined; thus saying
12563@code{tfind line} repeatedly can appear to have the same effect as
12564stepping from line to line in a @emph{live} debugging session.
12565@end table
12566
12567The default arguments for the @code{tfind} commands are specifically
12568designed to make it easy to scan through the trace buffer. For
12569instance, @code{tfind} with no argument selects the next trace
12570snapshot, and @code{tfind -} with no argument selects the previous
12571trace snapshot. So, by giving one @code{tfind} command, and then
12572simply hitting @key{RET} repeatedly you can examine all the trace
12573snapshots in order. Or, by saying @code{tfind -} and then hitting
12574@key{RET} repeatedly you can examine the snapshots in reverse order.
12575The @code{tfind line} command with no argument selects the snapshot
12576for the next source line executed. The @code{tfind pc} command with
12577no argument selects the next snapshot with the same program counter
12578(PC) as the current frame. The @code{tfind tracepoint} command with
12579no argument selects the next trace snapshot collected by the same
12580tracepoint as the current one.
12581
12582In addition to letting you scan through the trace buffer manually,
12583these commands make it easy to construct @value{GDBN} scripts that
12584scan through the trace buffer and print out whatever collected data
12585you are interested in. Thus, if we want to examine the PC, FP, and SP
12586registers from each trace frame in the buffer, we can say this:
12587
12588@smallexample
12589(@value{GDBP}) @b{tfind start}
12590(@value{GDBP}) @b{while ($trace_frame != -1)}
12591> printf "Frame %d, PC = %08X, SP = %08X, FP = %08X\n", \
12592 $trace_frame, $pc, $sp, $fp
12593> tfind
12594> end
12595
12596Frame 0, PC = 0020DC64, SP = 0030BF3C, FP = 0030BF44
12597Frame 1, PC = 0020DC6C, SP = 0030BF38, FP = 0030BF44
12598Frame 2, PC = 0020DC70, SP = 0030BF34, FP = 0030BF44
12599Frame 3, PC = 0020DC74, SP = 0030BF30, FP = 0030BF44
12600Frame 4, PC = 0020DC78, SP = 0030BF2C, FP = 0030BF44
12601Frame 5, PC = 0020DC7C, SP = 0030BF28, FP = 0030BF44
12602Frame 6, PC = 0020DC80, SP = 0030BF24, FP = 0030BF44
12603Frame 7, PC = 0020DC84, SP = 0030BF20, FP = 0030BF44
12604Frame 8, PC = 0020DC88, SP = 0030BF1C, FP = 0030BF44
12605Frame 9, PC = 0020DC8E, SP = 0030BF18, FP = 0030BF44
12606Frame 10, PC = 00203F6C, SP = 0030BE3C, FP = 0030BF14
12607@end smallexample
12608
12609Or, if we want to examine the variable @code{X} at each source line in
12610the buffer:
12611
12612@smallexample
12613(@value{GDBP}) @b{tfind start}
12614(@value{GDBP}) @b{while ($trace_frame != -1)}
12615> printf "Frame %d, X == %d\n", $trace_frame, X
12616> tfind line
12617> end
12618
12619Frame 0, X = 1
12620Frame 7, X = 2
12621Frame 13, X = 255
12622@end smallexample
12623
12624@node tdump
12625@subsection @code{tdump}
12626@kindex tdump
12627@cindex dump all data collected at tracepoint
12628@cindex tracepoint data, display
12629
12630This command takes no arguments. It prints all the data collected at
12631the current trace snapshot.
12632
12633@smallexample
12634(@value{GDBP}) @b{trace 444}
12635(@value{GDBP}) @b{actions}
12636Enter actions for tracepoint #2, one per line:
12637> collect $regs, $locals, $args, gdb_long_test
12638> end
12639
12640(@value{GDBP}) @b{tstart}
12641
12642(@value{GDBP}) @b{tfind line 444}
12643#0 gdb_test (p1=0x11, p2=0x22, p3=0x33, p4=0x44, p5=0x55, p6=0x66)
12644at gdb_test.c:444
12645444 printp( "%s: arguments = 0x%X 0x%X 0x%X 0x%X 0x%X 0x%X\n", )
12646
12647(@value{GDBP}) @b{tdump}
12648Data collected at tracepoint 2, trace frame 1:
12649d0 0xc4aa0085 -995491707
12650d1 0x18 24
12651d2 0x80 128
12652d3 0x33 51
12653d4 0x71aea3d 119204413
12654d5 0x22 34
12655d6 0xe0 224
12656d7 0x380035 3670069
12657a0 0x19e24a 1696330
12658a1 0x3000668 50333288
12659a2 0x100 256
12660a3 0x322000 3284992
12661a4 0x3000698 50333336
12662a5 0x1ad3cc 1758156
12663fp 0x30bf3c 0x30bf3c
12664sp 0x30bf34 0x30bf34
12665ps 0x0 0
12666pc 0x20b2c8 0x20b2c8
12667fpcontrol 0x0 0
12668fpstatus 0x0 0
12669fpiaddr 0x0 0
12670p = 0x20e5b4 "gdb-test"
12671p1 = (void *) 0x11
12672p2 = (void *) 0x22
12673p3 = (void *) 0x33
12674p4 = (void *) 0x44
12675p5 = (void *) 0x55
12676p6 = (void *) 0x66
12677gdb_long_test = 17 '\021'
12678
12679(@value{GDBP})
12680@end smallexample
12681
12682@code{tdump} works by scanning the tracepoint's current collection
12683actions and printing the value of each expression listed. So
12684@code{tdump} can fail, if after a run, you change the tracepoint's
12685actions to mention variables that were not collected during the run.
12686
12687Also, for tracepoints with @code{while-stepping} loops, @code{tdump}
12688uses the collected value of @code{$pc} to distinguish between trace
12689frames that were collected at the tracepoint hit, and frames that were
12690collected while stepping. This allows it to correctly choose whether
12691to display the basic list of collections, or the collections from the
12692body of the while-stepping loop. However, if @code{$pc} was not collected,
12693then @code{tdump} will always attempt to dump using the basic collection
12694list, and may fail if a while-stepping frame does not include all the
12695same data that is collected at the tracepoint hit.
12696@c This is getting pretty arcane, example would be good.
12697
12698@node save tracepoints
12699@subsection @code{save tracepoints @var{filename}}
12700@kindex save tracepoints
12701@kindex save-tracepoints
12702@cindex save tracepoints for future sessions
12703
12704This command saves all current tracepoint definitions together with
12705their actions and passcounts, into a file @file{@var{filename}}
12706suitable for use in a later debugging session. To read the saved
12707tracepoint definitions, use the @code{source} command (@pxref{Command
12708Files}). The @w{@code{save-tracepoints}} command is a deprecated
12709alias for @w{@code{save tracepoints}}
12710
12711@node Tracepoint Variables
12712@section Convenience Variables for Tracepoints
12713@cindex tracepoint variables
12714@cindex convenience variables for tracepoints
12715
12716@table @code
12717@vindex $trace_frame
12718@item (int) $trace_frame
12719The current trace snapshot (a.k.a.@: @dfn{frame}) number, or -1 if no
12720snapshot is selected.
12721
12722@vindex $tracepoint
12723@item (int) $tracepoint
12724The tracepoint for the current trace snapshot.
12725
12726@vindex $trace_line
12727@item (int) $trace_line
12728The line number for the current trace snapshot.
12729
12730@vindex $trace_file
12731@item (char []) $trace_file
12732The source file for the current trace snapshot.
12733
12734@vindex $trace_func
12735@item (char []) $trace_func
12736The name of the function containing @code{$tracepoint}.
12737@end table
12738
12739Note: @code{$trace_file} is not suitable for use in @code{printf},
12740use @code{output} instead.
12741
12742Here's a simple example of using these convenience variables for
12743stepping through all the trace snapshots and printing some of their
12744data. Note that these are not the same as trace state variables,
12745which are managed by the target.
12746
12747@smallexample
12748(@value{GDBP}) @b{tfind start}
12749
12750(@value{GDBP}) @b{while $trace_frame != -1}
12751> output $trace_file
12752> printf ", line %d (tracepoint #%d)\n", $trace_line, $tracepoint
12753> tfind
12754> end
12755@end smallexample
12756
12757@node Trace Files
12758@section Using Trace Files
12759@cindex trace files
12760
12761In some situations, the target running a trace experiment may no
12762longer be available; perhaps it crashed, or the hardware was needed
12763for a different activity. To handle these cases, you can arrange to
12764dump the trace data into a file, and later use that file as a source
12765of trace data, via the @code{target tfile} command.
12766
12767@table @code
12768
12769@kindex tsave
12770@item tsave [ -r ] @var{filename}
12771@itemx tsave [-ctf] @var{dirname}
12772Save the trace data to @var{filename}. By default, this command
12773assumes that @var{filename} refers to the host filesystem, so if
12774necessary @value{GDBN} will copy raw trace data up from the target and
12775then save it. If the target supports it, you can also supply the
12776optional argument @code{-r} (``remote'') to direct the target to save
12777the data directly into @var{filename} in its own filesystem, which may be
12778more efficient if the trace buffer is very large. (Note, however, that
12779@code{target tfile} can only read from files accessible to the host.)
12780By default, this command will save trace frame in tfile format.
12781You can supply the optional argument @code{-ctf} to save date in CTF
12782format. The @dfn{Common Trace Format} (CTF) is proposed as a trace format
12783that can be shared by multiple debugging and tracing tools. Please go to
12784@indicateurl{http://www.efficios.com/ctf} to get more information.
12785
12786@kindex target tfile
12787@kindex tfile
12788@kindex target ctf
12789@kindex ctf
12790@item target tfile @var{filename}
12791@itemx target ctf @var{dirname}
12792Use the file named @var{filename} or directory named @var{dirname} as
12793a source of trace data. Commands that examine data work as they do with
12794a live target, but it is not possible to run any new trace experiments.
12795@code{tstatus} will report the state of the trace run at the moment
12796the data was saved, as well as the current trace frame you are examining.
12797@var{filename} or @var{dirname} must be on a filesystem accessible to
12798the host.
12799
12800@smallexample
12801(@value{GDBP}) target ctf ctf.ctf
12802(@value{GDBP}) tfind
12803Found trace frame 0, tracepoint 2
1280439 ++a; /* set tracepoint 1 here */
12805(@value{GDBP}) tdump
12806Data collected at tracepoint 2, trace frame 0:
12807i = 0
12808a = 0
12809b = 1 '\001'
12810c = @{"123", "456", "789", "123", "456", "789"@}
12811d = @{@{@{a = 1, b = 2@}, @{a = 3, b = 4@}@}, @{@{a = 5, b = 6@}, @{a = 7, b = 8@}@}@}
12812(@value{GDBP}) p b
12813$1 = 1
12814@end smallexample
12815
12816@end table
12817
12818@node Overlays
12819@chapter Debugging Programs That Use Overlays
12820@cindex overlays
12821
12822If your program is too large to fit completely in your target system's
12823memory, you can sometimes use @dfn{overlays} to work around this
12824problem. @value{GDBN} provides some support for debugging programs that
12825use overlays.
12826
12827@menu
12828* How Overlays Work:: A general explanation of overlays.
12829* Overlay Commands:: Managing overlays in @value{GDBN}.
12830* Automatic Overlay Debugging:: @value{GDBN} can find out which overlays are
12831 mapped by asking the inferior.
12832* Overlay Sample Program:: A sample program using overlays.
12833@end menu
12834
12835@node How Overlays Work
12836@section How Overlays Work
12837@cindex mapped overlays
12838@cindex unmapped overlays
12839@cindex load address, overlay's
12840@cindex mapped address
12841@cindex overlay area
12842
12843Suppose you have a computer whose instruction address space is only 64
12844kilobytes long, but which has much more memory which can be accessed by
12845other means: special instructions, segment registers, or memory
12846management hardware, for example. Suppose further that you want to
12847adapt a program which is larger than 64 kilobytes to run on this system.
12848
12849One solution is to identify modules of your program which are relatively
12850independent, and need not call each other directly; call these modules
12851@dfn{overlays}. Separate the overlays from the main program, and place
12852their machine code in the larger memory. Place your main program in
12853instruction memory, but leave at least enough space there to hold the
12854largest overlay as well.
12855
12856Now, to call a function located in an overlay, you must first copy that
12857overlay's machine code from the large memory into the space set aside
12858for it in the instruction memory, and then jump to its entry point
12859there.
12860
12861@c NB: In the below the mapped area's size is greater or equal to the
12862@c size of all overlays. This is intentional to remind the developer
12863@c that overlays don't necessarily need to be the same size.
12864
12865@smallexample
12866@group
12867 Data Instruction Larger
12868Address Space Address Space Address Space
12869+-----------+ +-----------+ +-----------+
12870| | | | | |
12871+-----------+ +-----------+ +-----------+<-- overlay 1
12872| program | | main | .----| overlay 1 | load address
12873| variables | | program | | +-----------+
12874| and heap | | | | | |
12875+-----------+ | | | +-----------+<-- overlay 2
12876| | +-----------+ | | | load address
12877+-----------+ | | | .-| overlay 2 |
12878 | | | | | |
12879 mapped --->+-----------+ | | +-----------+
12880 address | | | | | |
12881 | overlay | <-' | | |
12882 | area | <---' +-----------+<-- overlay 3
12883 | | <---. | | load address
12884 +-----------+ `--| overlay 3 |
12885 | | | |
12886 +-----------+ | |
12887 +-----------+
12888 | |
12889 +-----------+
12890
12891 @anchor{A code overlay}A code overlay
12892@end group
12893@end smallexample
12894
12895The diagram (@pxref{A code overlay}) shows a system with separate data
12896and instruction address spaces. To map an overlay, the program copies
12897its code from the larger address space to the instruction address space.
12898Since the overlays shown here all use the same mapped address, only one
12899may be mapped at a time. For a system with a single address space for
12900data and instructions, the diagram would be similar, except that the
12901program variables and heap would share an address space with the main
12902program and the overlay area.
12903
12904An overlay loaded into instruction memory and ready for use is called a
12905@dfn{mapped} overlay; its @dfn{mapped address} is its address in the
12906instruction memory. An overlay not present (or only partially present)
12907in instruction memory is called @dfn{unmapped}; its @dfn{load address}
12908is its address in the larger memory. The mapped address is also called
12909the @dfn{virtual memory address}, or @dfn{VMA}; the load address is also
12910called the @dfn{load memory address}, or @dfn{LMA}.
12911
12912Unfortunately, overlays are not a completely transparent way to adapt a
12913program to limited instruction memory. They introduce a new set of
12914global constraints you must keep in mind as you design your program:
12915
12916@itemize @bullet
12917
12918@item
12919Before calling or returning to a function in an overlay, your program
12920must make sure that overlay is actually mapped. Otherwise, the call or
12921return will transfer control to the right address, but in the wrong
12922overlay, and your program will probably crash.
12923
12924@item
12925If the process of mapping an overlay is expensive on your system, you
12926will need to choose your overlays carefully to minimize their effect on
12927your program's performance.
12928
12929@item
12930The executable file you load onto your system must contain each
12931overlay's instructions, appearing at the overlay's load address, not its
12932mapped address. However, each overlay's instructions must be relocated
12933and its symbols defined as if the overlay were at its mapped address.
12934You can use GNU linker scripts to specify different load and relocation
12935addresses for pieces of your program; see @ref{Overlay Description,,,
12936ld.info, Using ld: the GNU linker}.
12937
12938@item
12939The procedure for loading executable files onto your system must be able
12940to load their contents into the larger address space as well as the
12941instruction and data spaces.
12942
12943@end itemize
12944
12945The overlay system described above is rather simple, and could be
12946improved in many ways:
12947
12948@itemize @bullet
12949
12950@item
12951If your system has suitable bank switch registers or memory management
12952hardware, you could use those facilities to make an overlay's load area
12953contents simply appear at their mapped address in instruction space.
12954This would probably be faster than copying the overlay to its mapped
12955area in the usual way.
12956
12957@item
12958If your overlays are small enough, you could set aside more than one
12959overlay area, and have more than one overlay mapped at a time.
12960
12961@item
12962You can use overlays to manage data, as well as instructions. In
12963general, data overlays are even less transparent to your design than
12964code overlays: whereas code overlays only require care when you call or
12965return to functions, data overlays require care every time you access
12966the data. Also, if you change the contents of a data overlay, you
12967must copy its contents back out to its load address before you can copy a
12968different data overlay into the same mapped area.
12969
12970@end itemize
12971
12972
12973@node Overlay Commands
12974@section Overlay Commands
12975
12976To use @value{GDBN}'s overlay support, each overlay in your program must
12977correspond to a separate section of the executable file. The section's
12978virtual memory address and load memory address must be the overlay's
12979mapped and load addresses. Identifying overlays with sections allows
12980@value{GDBN} to determine the appropriate address of a function or
12981variable, depending on whether the overlay is mapped or not.
12982
12983@value{GDBN}'s overlay commands all start with the word @code{overlay};
12984you can abbreviate this as @code{ov} or @code{ovly}. The commands are:
12985
12986@table @code
12987@item overlay off
12988@kindex overlay
12989Disable @value{GDBN}'s overlay support. When overlay support is
12990disabled, @value{GDBN} assumes that all functions and variables are
12991always present at their mapped addresses. By default, @value{GDBN}'s
12992overlay support is disabled.
12993
12994@item overlay manual
12995@cindex manual overlay debugging
12996Enable @dfn{manual} overlay debugging. In this mode, @value{GDBN}
12997relies on you to tell it which overlays are mapped, and which are not,
12998using the @code{overlay map-overlay} and @code{overlay unmap-overlay}
12999commands described below.
13000
13001@item overlay map-overlay @var{overlay}
13002@itemx overlay map @var{overlay}
13003@cindex map an overlay
13004Tell @value{GDBN} that @var{overlay} is now mapped; @var{overlay} must
13005be the name of the object file section containing the overlay. When an
13006overlay is mapped, @value{GDBN} assumes it can find the overlay's
13007functions and variables at their mapped addresses. @value{GDBN} assumes
13008that any other overlays whose mapped ranges overlap that of
13009@var{overlay} are now unmapped.
13010
13011@item overlay unmap-overlay @var{overlay}
13012@itemx overlay unmap @var{overlay}
13013@cindex unmap an overlay
13014Tell @value{GDBN} that @var{overlay} is no longer mapped; @var{overlay}
13015must be the name of the object file section containing the overlay.
13016When an overlay is unmapped, @value{GDBN} assumes it can find the
13017overlay's functions and variables at their load addresses.
13018
13019@item overlay auto
13020Enable @dfn{automatic} overlay debugging. In this mode, @value{GDBN}
13021consults a data structure the overlay manager maintains in the inferior
13022to see which overlays are mapped. For details, see @ref{Automatic
13023Overlay Debugging}.
13024
13025@item overlay load-target
13026@itemx overlay load
13027@cindex reloading the overlay table
13028Re-read the overlay table from the inferior. Normally, @value{GDBN}
13029re-reads the table @value{GDBN} automatically each time the inferior
13030stops, so this command should only be necessary if you have changed the
13031overlay mapping yourself using @value{GDBN}. This command is only
13032useful when using automatic overlay debugging.
13033
13034@item overlay list-overlays
13035@itemx overlay list
13036@cindex listing mapped overlays
13037Display a list of the overlays currently mapped, along with their mapped
13038addresses, load addresses, and sizes.
13039
13040@end table
13041
13042Normally, when @value{GDBN} prints a code address, it includes the name
13043of the function the address falls in:
13044
13045@smallexample
13046(@value{GDBP}) print main
13047$3 = @{int ()@} 0x11a0 <main>
13048@end smallexample
13049@noindent
13050When overlay debugging is enabled, @value{GDBN} recognizes code in
13051unmapped overlays, and prints the names of unmapped functions with
13052asterisks around them. For example, if @code{foo} is a function in an
13053unmapped overlay, @value{GDBN} prints it this way:
13054
13055@smallexample
13056(@value{GDBP}) overlay list
13057No sections are mapped.
13058(@value{GDBP}) print foo
13059$5 = @{int (int)@} 0x100000 <*foo*>
13060@end smallexample
13061@noindent
13062When @code{foo}'s overlay is mapped, @value{GDBN} prints the function's
13063name normally:
13064
13065@smallexample
13066(@value{GDBP}) overlay list
13067Section .ov.foo.text, loaded at 0x100000 - 0x100034,
13068 mapped at 0x1016 - 0x104a
13069(@value{GDBP}) print foo
13070$6 = @{int (int)@} 0x1016 <foo>
13071@end smallexample
13072
13073When overlay debugging is enabled, @value{GDBN} can find the correct
13074address for functions and variables in an overlay, whether or not the
13075overlay is mapped. This allows most @value{GDBN} commands, like
13076@code{break} and @code{disassemble}, to work normally, even on unmapped
13077code. However, @value{GDBN}'s breakpoint support has some limitations:
13078
13079@itemize @bullet
13080@item
13081@cindex breakpoints in overlays
13082@cindex overlays, setting breakpoints in
13083You can set breakpoints in functions in unmapped overlays, as long as
13084@value{GDBN} can write to the overlay at its load address.
13085@item
13086@value{GDBN} can not set hardware or simulator-based breakpoints in
13087unmapped overlays. However, if you set a breakpoint at the end of your
13088overlay manager (and tell @value{GDBN} which overlays are now mapped, if
13089you are using manual overlay management), @value{GDBN} will re-set its
13090breakpoints properly.
13091@end itemize
13092
13093
13094@node Automatic Overlay Debugging
13095@section Automatic Overlay Debugging
13096@cindex automatic overlay debugging
13097
13098@value{GDBN} can automatically track which overlays are mapped and which
13099are not, given some simple co-operation from the overlay manager in the
13100inferior. If you enable automatic overlay debugging with the
13101@code{overlay auto} command (@pxref{Overlay Commands}), @value{GDBN}
13102looks in the inferior's memory for certain variables describing the
13103current state of the overlays.
13104
13105Here are the variables your overlay manager must define to support
13106@value{GDBN}'s automatic overlay debugging:
13107
13108@table @asis
13109
13110@item @code{_ovly_table}:
13111This variable must be an array of the following structures:
13112
13113@smallexample
13114struct
13115@{
13116 /* The overlay's mapped address. */
13117 unsigned long vma;
13118
13119 /* The size of the overlay, in bytes. */
13120 unsigned long size;
13121
13122 /* The overlay's load address. */
13123 unsigned long lma;
13124
13125 /* Non-zero if the overlay is currently mapped;
13126 zero otherwise. */
13127 unsigned long mapped;
13128@}
13129@end smallexample
13130
13131@item @code{_novlys}:
13132This variable must be a four-byte signed integer, holding the total
13133number of elements in @code{_ovly_table}.
13134
13135@end table
13136
13137To decide whether a particular overlay is mapped or not, @value{GDBN}
13138looks for an entry in @w{@code{_ovly_table}} whose @code{vma} and
13139@code{lma} members equal the VMA and LMA of the overlay's section in the
13140executable file. When @value{GDBN} finds a matching entry, it consults
13141the entry's @code{mapped} member to determine whether the overlay is
13142currently mapped.
13143
13144In addition, your overlay manager may define a function called
13145@code{_ovly_debug_event}. If this function is defined, @value{GDBN}
13146will silently set a breakpoint there. If the overlay manager then
13147calls this function whenever it has changed the overlay table, this
13148will enable @value{GDBN} to accurately keep track of which overlays
13149are in program memory, and update any breakpoints that may be set
13150in overlays. This will allow breakpoints to work even if the
13151overlays are kept in ROM or other non-writable memory while they
13152are not being executed.
13153
13154@node Overlay Sample Program
13155@section Overlay Sample Program
13156@cindex overlay example program
13157
13158When linking a program which uses overlays, you must place the overlays
13159at their load addresses, while relocating them to run at their mapped
13160addresses. To do this, you must write a linker script (@pxref{Overlay
13161Description,,, ld.info, Using ld: the GNU linker}). Unfortunately,
13162since linker scripts are specific to a particular host system, target
13163architecture, and target memory layout, this manual cannot provide
13164portable sample code demonstrating @value{GDBN}'s overlay support.
13165
13166However, the @value{GDBN} source distribution does contain an overlaid
13167program, with linker scripts for a few systems, as part of its test
13168suite. The program consists of the following files from
13169@file{gdb/testsuite/gdb.base}:
13170
13171@table @file
13172@item overlays.c
13173The main program file.
13174@item ovlymgr.c
13175A simple overlay manager, used by @file{overlays.c}.
13176@item foo.c
13177@itemx bar.c
13178@itemx baz.c
13179@itemx grbx.c
13180Overlay modules, loaded and used by @file{overlays.c}.
13181@item d10v.ld
13182@itemx m32r.ld
13183Linker scripts for linking the test program on the @code{d10v-elf}
13184and @code{m32r-elf} targets.
13185@end table
13186
13187You can build the test program using the @code{d10v-elf} GCC
13188cross-compiler like this:
13189
13190@smallexample
13191$ d10v-elf-gcc -g -c overlays.c
13192$ d10v-elf-gcc -g -c ovlymgr.c
13193$ d10v-elf-gcc -g -c foo.c
13194$ d10v-elf-gcc -g -c bar.c
13195$ d10v-elf-gcc -g -c baz.c
13196$ d10v-elf-gcc -g -c grbx.c
13197$ d10v-elf-gcc -g overlays.o ovlymgr.o foo.o bar.o \
13198 baz.o grbx.o -Wl,-Td10v.ld -o overlays
13199@end smallexample
13200
13201The build process is identical for any other architecture, except that
13202you must substitute the appropriate compiler and linker script for the
13203target system for @code{d10v-elf-gcc} and @code{d10v.ld}.
13204
13205
13206@node Languages
13207@chapter Using @value{GDBN} with Different Languages
13208@cindex languages
13209
13210Although programming languages generally have common aspects, they are
13211rarely expressed in the same manner. For instance, in ANSI C,
13212dereferencing a pointer @code{p} is accomplished by @code{*p}, but in
13213Modula-2, it is accomplished by @code{p^}. Values can also be
13214represented (and displayed) differently. Hex numbers in C appear as
13215@samp{0x1ae}, while in Modula-2 they appear as @samp{1AEH}.
13216
13217@cindex working language
13218Language-specific information is built into @value{GDBN} for some languages,
13219allowing you to express operations like the above in your program's
13220native language, and allowing @value{GDBN} to output values in a manner
13221consistent with the syntax of your program's native language. The
13222language you use to build expressions is called the @dfn{working
13223language}.
13224
13225@menu
13226* Setting:: Switching between source languages
13227* Show:: Displaying the language
13228* Checks:: Type and range checks
13229* Supported Languages:: Supported languages
13230* Unsupported Languages:: Unsupported languages
13231@end menu
13232
13233@node Setting
13234@section Switching Between Source Languages
13235
13236There are two ways to control the working language---either have @value{GDBN}
13237set it automatically, or select it manually yourself. You can use the
13238@code{set language} command for either purpose. On startup, @value{GDBN}
13239defaults to setting the language automatically. The working language is
13240used to determine how expressions you type are interpreted, how values
13241are printed, etc.
13242
13243In addition to the working language, every source file that
13244@value{GDBN} knows about has its own working language. For some object
13245file formats, the compiler might indicate which language a particular
13246source file is in. However, most of the time @value{GDBN} infers the
13247language from the name of the file. The language of a source file
13248controls whether C@t{++} names are demangled---this way @code{backtrace} can
13249show each frame appropriately for its own language. There is no way to
13250set the language of a source file from within @value{GDBN}, but you can
13251set the language associated with a filename extension. @xref{Show, ,
13252Displaying the Language}.
13253
13254This is most commonly a problem when you use a program, such
13255as @code{cfront} or @code{f2c}, that generates C but is written in
13256another language. In that case, make the
13257program use @code{#line} directives in its C output; that way
13258@value{GDBN} will know the correct language of the source code of the original
13259program, and will display that source code, not the generated C code.
13260
13261@menu
13262* Filenames:: Filename extensions and languages.
13263* Manually:: Setting the working language manually
13264* Automatically:: Having @value{GDBN} infer the source language
13265@end menu
13266
13267@node Filenames
13268@subsection List of Filename Extensions and Languages
13269
13270If a source file name ends in one of the following extensions, then
13271@value{GDBN} infers that its language is the one indicated.
13272
13273@table @file
13274@item .ada
13275@itemx .ads
13276@itemx .adb
13277@itemx .a
13278Ada source file.
13279
13280@item .c
13281C source file
13282
13283@item .C
13284@itemx .cc
13285@itemx .cp
13286@itemx .cpp
13287@itemx .cxx
13288@itemx .c++
13289C@t{++} source file
13290
13291@item .d
13292D source file
13293
13294@item .m
13295Objective-C source file
13296
13297@item .f
13298@itemx .F
13299Fortran source file
13300
13301@item .mod
13302Modula-2 source file
13303
13304@item .s
13305@itemx .S
13306Assembler source file. This actually behaves almost like C, but
13307@value{GDBN} does not skip over function prologues when stepping.
13308@end table
13309
13310In addition, you may set the language associated with a filename
13311extension. @xref{Show, , Displaying the Language}.
13312
13313@node Manually
13314@subsection Setting the Working Language
13315
13316If you allow @value{GDBN} to set the language automatically,
13317expressions are interpreted the same way in your debugging session and
13318your program.
13319
13320@kindex set language
13321If you wish, you may set the language manually. To do this, issue the
13322command @samp{set language @var{lang}}, where @var{lang} is the name of
13323a language, such as
13324@code{c} or @code{modula-2}.
13325For a list of the supported languages, type @samp{set language}.
13326
13327Setting the language manually prevents @value{GDBN} from updating the working
13328language automatically. This can lead to confusion if you try
13329to debug a program when the working language is not the same as the
13330source language, when an expression is acceptable to both
13331languages---but means different things. For instance, if the current
13332source file were written in C, and @value{GDBN} was parsing Modula-2, a
13333command such as:
13334
13335@smallexample
13336print a = b + c
13337@end smallexample
13338
13339@noindent
13340might not have the effect you intended. In C, this means to add
13341@code{b} and @code{c} and place the result in @code{a}. The result
13342printed would be the value of @code{a}. In Modula-2, this means to compare
13343@code{a} to the result of @code{b+c}, yielding a @code{BOOLEAN} value.
13344
13345@node Automatically
13346@subsection Having @value{GDBN} Infer the Source Language
13347
13348To have @value{GDBN} set the working language automatically, use
13349@samp{set language local} or @samp{set language auto}. @value{GDBN}
13350then infers the working language. That is, when your program stops in a
13351frame (usually by encountering a breakpoint), @value{GDBN} sets the
13352working language to the language recorded for the function in that
13353frame. If the language for a frame is unknown (that is, if the function
13354or block corresponding to the frame was defined in a source file that
13355does not have a recognized extension), the current working language is
13356not changed, and @value{GDBN} issues a warning.
13357
13358This may not seem necessary for most programs, which are written
13359entirely in one source language. However, program modules and libraries
13360written in one source language can be used by a main program written in
13361a different source language. Using @samp{set language auto} in this
13362case frees you from having to set the working language manually.
13363
13364@node Show
13365@section Displaying the Language
13366
13367The following commands help you find out which language is the
13368working language, and also what language source files were written in.
13369
13370@table @code
13371@item show language
13372@anchor{show language}
13373@kindex show language
13374Display the current working language. This is the
13375language you can use with commands such as @code{print} to
13376build and compute expressions that may involve variables in your program.
13377
13378@item info frame
13379@kindex info frame@r{, show the source language}
13380Display the source language for this frame. This language becomes the
13381working language if you use an identifier from this frame.
13382@xref{Frame Info, ,Information about a Frame}, to identify the other
13383information listed here.
13384
13385@item info source
13386@kindex info source@r{, show the source language}
13387Display the source language of this source file.
13388@xref{Symbols, ,Examining the Symbol Table}, to identify the other
13389information listed here.
13390@end table
13391
13392In unusual circumstances, you may have source files with extensions
13393not in the standard list. You can then set the extension associated
13394with a language explicitly:
13395
13396@table @code
13397@item set extension-language @var{ext} @var{language}
13398@kindex set extension-language
13399Tell @value{GDBN} that source files with extension @var{ext} are to be
13400assumed as written in the source language @var{language}.
13401
13402@item info extensions
13403@kindex info extensions
13404List all the filename extensions and the associated languages.
13405@end table
13406
13407@node Checks
13408@section Type and Range Checking
13409
13410Some languages are designed to guard you against making seemingly common
13411errors through a series of compile- and run-time checks. These include
13412checking the type of arguments to functions and operators and making
13413sure mathematical overflows are caught at run time. Checks such as
13414these help to ensure a program's correctness once it has been compiled
13415by eliminating type mismatches and providing active checks for range
13416errors when your program is running.
13417
13418By default @value{GDBN} checks for these errors according to the
13419rules of the current source language. Although @value{GDBN} does not check
13420the statements in your program, it can check expressions entered directly
13421into @value{GDBN} for evaluation via the @code{print} command, for example.
13422
13423@menu
13424* Type Checking:: An overview of type checking
13425* Range Checking:: An overview of range checking
13426@end menu
13427
13428@cindex type checking
13429@cindex checks, type
13430@node Type Checking
13431@subsection An Overview of Type Checking
13432
13433Some languages, such as C and C@t{++}, are strongly typed, meaning that the
13434arguments to operators and functions have to be of the correct type,
13435otherwise an error occurs. These checks prevent type mismatch
13436errors from ever causing any run-time problems. For example,
13437
13438@smallexample
13439int klass::my_method(char *b) @{ return b ? 1 : 2; @}
13440
13441(@value{GDBP}) print obj.my_method (0)
13442$1 = 2
13443@exdent but
13444(@value{GDBP}) print obj.my_method (0x1234)
13445Cannot resolve method klass::my_method to any overloaded instance
13446@end smallexample
13447
13448The second example fails because in C@t{++} the integer constant
13449@samp{0x1234} is not type-compatible with the pointer parameter type.
13450
13451For the expressions you use in @value{GDBN} commands, you can tell
13452@value{GDBN} to not enforce strict type checking or
13453to treat any mismatches as errors and abandon the expression;
13454When type checking is disabled, @value{GDBN} successfully evaluates
13455expressions like the second example above.
13456
13457Even if type checking is off, there may be other reasons
13458related to type that prevent @value{GDBN} from evaluating an expression.
13459For instance, @value{GDBN} does not know how to add an @code{int} and
13460a @code{struct foo}. These particular type errors have nothing to do
13461with the language in use and usually arise from expressions which make
13462little sense to evaluate anyway.
13463
13464@value{GDBN} provides some additional commands for controlling type checking:
13465
13466@kindex set check type
13467@kindex show check type
13468@table @code
13469@item set check type on
13470@itemx set check type off
13471Set strict type checking on or off. If any type mismatches occur in
13472evaluating an expression while type checking is on, @value{GDBN} prints a
13473message and aborts evaluation of the expression.
13474
13475@item show check type
13476Show the current setting of type checking and whether @value{GDBN}
13477is enforcing strict type checking rules.
13478@end table
13479
13480@cindex range checking
13481@cindex checks, range
13482@node Range Checking
13483@subsection An Overview of Range Checking
13484
13485In some languages (such as Modula-2), it is an error to exceed the
13486bounds of a type; this is enforced with run-time checks. Such range
13487checking is meant to ensure program correctness by making sure
13488computations do not overflow, or indices on an array element access do
13489not exceed the bounds of the array.
13490
13491For expressions you use in @value{GDBN} commands, you can tell
13492@value{GDBN} to treat range errors in one of three ways: ignore them,
13493always treat them as errors and abandon the expression, or issue
13494warnings but evaluate the expression anyway.
13495
13496A range error can result from numerical overflow, from exceeding an
13497array index bound, or when you type a constant that is not a member
13498of any type. Some languages, however, do not treat overflows as an
13499error. In many implementations of C, mathematical overflow causes the
13500result to ``wrap around'' to lower values---for example, if @var{m} is
13501the largest integer value, and @var{s} is the smallest, then
13502
13503@smallexample
13504@var{m} + 1 @result{} @var{s}
13505@end smallexample
13506
13507This, too, is specific to individual languages, and in some cases
13508specific to individual compilers or machines. @xref{Supported Languages, ,
13509Supported Languages}, for further details on specific languages.
13510
13511@value{GDBN} provides some additional commands for controlling the range checker:
13512
13513@kindex set check range
13514@kindex show check range
13515@table @code
13516@item set check range auto
13517Set range checking on or off based on the current working language.
13518@xref{Supported Languages, ,Supported Languages}, for the default settings for
13519each language.
13520
13521@item set check range on
13522@itemx set check range off
13523Set range checking on or off, overriding the default setting for the
13524current working language. A warning is issued if the setting does not
13525match the language default. If a range error occurs and range checking is on,
13526then a message is printed and evaluation of the expression is aborted.
13527
13528@item set check range warn
13529Output messages when the @value{GDBN} range checker detects a range error,
13530but attempt to evaluate the expression anyway. Evaluating the
13531expression may still be impossible for other reasons, such as accessing
13532memory that the process does not own (a typical example from many Unix
13533systems).
13534
13535@item show range
13536Show the current setting of the range checker, and whether or not it is
13537being set automatically by @value{GDBN}.
13538@end table
13539
13540@node Supported Languages
13541@section Supported Languages
13542
13543@value{GDBN} supports C, C@t{++}, D, Go, Objective-C, Fortran, Java,
13544OpenCL C, Pascal, assembly, Modula-2, and Ada.
13545@c This is false ...
13546Some @value{GDBN} features may be used in expressions regardless of the
13547language you use: the @value{GDBN} @code{@@} and @code{::} operators,
13548and the @samp{@{type@}addr} construct (@pxref{Expressions,
13549,Expressions}) can be used with the constructs of any supported
13550language.
13551
13552The following sections detail to what degree each source language is
13553supported by @value{GDBN}. These sections are not meant to be language
13554tutorials or references, but serve only as a reference guide to what the
13555@value{GDBN} expression parser accepts, and what input and output
13556formats should look like for different languages. There are many good
13557books written on each of these languages; please look to these for a
13558language reference or tutorial.
13559
13560@menu
13561* C:: C and C@t{++}
13562* D:: D
13563* Go:: Go
13564* Objective-C:: Objective-C
13565* OpenCL C:: OpenCL C
13566* Fortran:: Fortran
13567* Pascal:: Pascal
13568* Modula-2:: Modula-2
13569* Ada:: Ada
13570@end menu
13571
13572@node C
13573@subsection C and C@t{++}
13574
13575@cindex C and C@t{++}
13576@cindex expressions in C or C@t{++}
13577
13578Since C and C@t{++} are so closely related, many features of @value{GDBN} apply
13579to both languages. Whenever this is the case, we discuss those languages
13580together.
13581
13582@cindex C@t{++}
13583@cindex @code{g++}, @sc{gnu} C@t{++} compiler
13584@cindex @sc{gnu} C@t{++}
13585The C@t{++} debugging facilities are jointly implemented by the C@t{++}
13586compiler and @value{GDBN}. Therefore, to debug your C@t{++} code
13587effectively, you must compile your C@t{++} programs with a supported
13588C@t{++} compiler, such as @sc{gnu} @code{g++}, or the HP ANSI C@t{++}
13589compiler (@code{aCC}).
13590
13591@menu
13592* C Operators:: C and C@t{++} operators
13593* C Constants:: C and C@t{++} constants
13594* C Plus Plus Expressions:: C@t{++} expressions
13595* C Defaults:: Default settings for C and C@t{++}
13596* C Checks:: C and C@t{++} type and range checks
13597* Debugging C:: @value{GDBN} and C
13598* Debugging C Plus Plus:: @value{GDBN} features for C@t{++}
13599* Decimal Floating Point:: Numbers in Decimal Floating Point format
13600@end menu
13601
13602@node C Operators
13603@subsubsection C and C@t{++} Operators
13604
13605@cindex C and C@t{++} operators
13606
13607Operators must be defined on values of specific types. For instance,
13608@code{+} is defined on numbers, but not on structures. Operators are
13609often defined on groups of types.
13610
13611For the purposes of C and C@t{++}, the following definitions hold:
13612
13613@itemize @bullet
13614
13615@item
13616@emph{Integral types} include @code{int} with any of its storage-class
13617specifiers; @code{char}; @code{enum}; and, for C@t{++}, @code{bool}.
13618
13619@item
13620@emph{Floating-point types} include @code{float}, @code{double}, and
13621@code{long double} (if supported by the target platform).
13622
13623@item
13624@emph{Pointer types} include all types defined as @code{(@var{type} *)}.
13625
13626@item
13627@emph{Scalar types} include all of the above.
13628
13629@end itemize
13630
13631@noindent
13632The following operators are supported. They are listed here
13633in order of increasing precedence:
13634
13635@table @code
13636@item ,
13637The comma or sequencing operator. Expressions in a comma-separated list
13638are evaluated from left to right, with the result of the entire
13639expression being the last expression evaluated.
13640
13641@item =
13642Assignment. The value of an assignment expression is the value
13643assigned. Defined on scalar types.
13644
13645@item @var{op}=
13646Used in an expression of the form @w{@code{@var{a} @var{op}= @var{b}}},
13647and translated to @w{@code{@var{a} = @var{a op b}}}.
13648@w{@code{@var{op}=}} and @code{=} have the same precedence.
13649@var{op} is any one of the operators @code{|}, @code{^}, @code{&},
13650@code{<<}, @code{>>}, @code{+}, @code{-}, @code{*}, @code{/}, @code{%}.
13651
13652@item ?:
13653The ternary operator. @code{@var{a} ? @var{b} : @var{c}} can be thought
13654of as: if @var{a} then @var{b} else @var{c}. @var{a} should be of an
13655integral type.
13656
13657@item ||
13658Logical @sc{or}. Defined on integral types.
13659
13660@item &&
13661Logical @sc{and}. Defined on integral types.
13662
13663@item |
13664Bitwise @sc{or}. Defined on integral types.
13665
13666@item ^
13667Bitwise exclusive-@sc{or}. Defined on integral types.
13668
13669@item &
13670Bitwise @sc{and}. Defined on integral types.
13671
13672@item ==@r{, }!=
13673Equality and inequality. Defined on scalar types. The value of these
13674expressions is 0 for false and non-zero for true.
13675
13676@item <@r{, }>@r{, }<=@r{, }>=
13677Less than, greater than, less than or equal, greater than or equal.
13678Defined on scalar types. The value of these expressions is 0 for false
13679and non-zero for true.
13680
13681@item <<@r{, }>>
13682left shift, and right shift. Defined on integral types.
13683
13684@item @@
13685The @value{GDBN} ``artificial array'' operator (@pxref{Expressions, ,Expressions}).
13686
13687@item +@r{, }-
13688Addition and subtraction. Defined on integral types, floating-point types and
13689pointer types.
13690
13691@item *@r{, }/@r{, }%
13692Multiplication, division, and modulus. Multiplication and division are
13693defined on integral and floating-point types. Modulus is defined on
13694integral types.
13695
13696@item ++@r{, }--
13697Increment and decrement. When appearing before a variable, the
13698operation is performed before the variable is used in an expression;
13699when appearing after it, the variable's value is used before the
13700operation takes place.
13701
13702@item *
13703Pointer dereferencing. Defined on pointer types. Same precedence as
13704@code{++}.
13705
13706@item &
13707Address operator. Defined on variables. Same precedence as @code{++}.
13708
13709For debugging C@t{++}, @value{GDBN} implements a use of @samp{&} beyond what is
13710allowed in the C@t{++} language itself: you can use @samp{&(&@var{ref})}
13711to examine the address
13712where a C@t{++} reference variable (declared with @samp{&@var{ref}}) is
13713stored.
13714
13715@item -
13716Negative. Defined on integral and floating-point types. Same
13717precedence as @code{++}.
13718
13719@item !
13720Logical negation. Defined on integral types. Same precedence as
13721@code{++}.
13722
13723@item ~
13724Bitwise complement operator. Defined on integral types. Same precedence as
13725@code{++}.
13726
13727
13728@item .@r{, }->
13729Structure member, and pointer-to-structure member. For convenience,
13730@value{GDBN} regards the two as equivalent, choosing whether to dereference a
13731pointer based on the stored type information.
13732Defined on @code{struct} and @code{union} data.
13733
13734@item .*@r{, }->*
13735Dereferences of pointers to members.
13736
13737@item []
13738Array indexing. @code{@var{a}[@var{i}]} is defined as
13739@code{*(@var{a}+@var{i})}. Same precedence as @code{->}.
13740
13741@item ()
13742Function parameter list. Same precedence as @code{->}.
13743
13744@item ::
13745C@t{++} scope resolution operator. Defined on @code{struct}, @code{union},
13746and @code{class} types.
13747
13748@item ::
13749Doubled colons also represent the @value{GDBN} scope operator
13750(@pxref{Expressions, ,Expressions}). Same precedence as @code{::},
13751above.
13752@end table
13753
13754If an operator is redefined in the user code, @value{GDBN} usually
13755attempts to invoke the redefined version instead of using the operator's
13756predefined meaning.
13757
13758@node C Constants
13759@subsubsection C and C@t{++} Constants
13760
13761@cindex C and C@t{++} constants
13762
13763@value{GDBN} allows you to express the constants of C and C@t{++} in the
13764following ways:
13765
13766@itemize @bullet
13767@item
13768Integer constants are a sequence of digits. Octal constants are
13769specified by a leading @samp{0} (i.e.@: zero), and hexadecimal constants
13770by a leading @samp{0x} or @samp{0X}. Constants may also end with a letter
13771@samp{l}, specifying that the constant should be treated as a
13772@code{long} value.
13773
13774@item
13775Floating point constants are a sequence of digits, followed by a decimal
13776point, followed by a sequence of digits, and optionally followed by an
13777exponent. An exponent is of the form:
13778@samp{@w{e@r{[[}+@r{]|}-@r{]}@var{nnn}}}, where @var{nnn} is another
13779sequence of digits. The @samp{+} is optional for positive exponents.
13780A floating-point constant may also end with a letter @samp{f} or
13781@samp{F}, specifying that the constant should be treated as being of
13782the @code{float} (as opposed to the default @code{double}) type; or with
13783a letter @samp{l} or @samp{L}, which specifies a @code{long double}
13784constant.
13785
13786@item
13787Enumerated constants consist of enumerated identifiers, or their
13788integral equivalents.
13789
13790@item
13791Character constants are a single character surrounded by single quotes
13792(@code{'}), or a number---the ordinal value of the corresponding character
13793(usually its @sc{ascii} value). Within quotes, the single character may
13794be represented by a letter or by @dfn{escape sequences}, which are of
13795the form @samp{\@var{nnn}}, where @var{nnn} is the octal representation
13796of the character's ordinal value; or of the form @samp{\@var{x}}, where
13797@samp{@var{x}} is a predefined special character---for example,
13798@samp{\n} for newline.
13799
13800Wide character constants can be written by prefixing a character
13801constant with @samp{L}, as in C. For example, @samp{L'x'} is the wide
13802form of @samp{x}. The target wide character set is used when
13803computing the value of this constant (@pxref{Character Sets}).
13804
13805@item
13806String constants are a sequence of character constants surrounded by
13807double quotes (@code{"}). Any valid character constant (as described
13808above) may appear. Double quotes within the string must be preceded by
13809a backslash, so for instance @samp{"a\"b'c"} is a string of five
13810characters.
13811
13812Wide string constants can be written by prefixing a string constant
13813with @samp{L}, as in C. The target wide character set is used when
13814computing the value of this constant (@pxref{Character Sets}).
13815
13816@item
13817Pointer constants are an integral value. You can also write pointers
13818to constants using the C operator @samp{&}.
13819
13820@item
13821Array constants are comma-separated lists surrounded by braces @samp{@{}
13822and @samp{@}}; for example, @samp{@{1,2,3@}} is a three-element array of
13823integers, @samp{@{@{1,2@}, @{3,4@}, @{5,6@}@}} is a three-by-two array,
13824and @samp{@{&"hi", &"there", &"fred"@}} is a three-element array of pointers.
13825@end itemize
13826
13827@node C Plus Plus Expressions
13828@subsubsection C@t{++} Expressions
13829
13830@cindex expressions in C@t{++}
13831@value{GDBN} expression handling can interpret most C@t{++} expressions.
13832
13833@cindex debugging C@t{++} programs
13834@cindex C@t{++} compilers
13835@cindex debug formats and C@t{++}
13836@cindex @value{NGCC} and C@t{++}
13837@quotation
13838@emph{Warning:} @value{GDBN} can only debug C@t{++} code if you use
13839the proper compiler and the proper debug format. Currently,
13840@value{GDBN} works best when debugging C@t{++} code that is compiled
13841with the most recent version of @value{NGCC} possible. The DWARF
13842debugging format is preferred; @value{NGCC} defaults to this on most
13843popular platforms. Other compilers and/or debug formats are likely to
13844work badly or not at all when using @value{GDBN} to debug C@t{++}
13845code. @xref{Compilation}.
13846@end quotation
13847
13848@enumerate
13849
13850@cindex member functions
13851@item
13852Member function calls are allowed; you can use expressions like
13853
13854@smallexample
13855count = aml->GetOriginal(x, y)
13856@end smallexample
13857
13858@vindex this@r{, inside C@t{++} member functions}
13859@cindex namespace in C@t{++}
13860@item
13861While a member function is active (in the selected stack frame), your
13862expressions have the same namespace available as the member function;
13863that is, @value{GDBN} allows implicit references to the class instance
13864pointer @code{this} following the same rules as C@t{++}. @code{using}
13865declarations in the current scope are also respected by @value{GDBN}.
13866
13867@cindex call overloaded functions
13868@cindex overloaded functions, calling
13869@cindex type conversions in C@t{++}
13870@item
13871You can call overloaded functions; @value{GDBN} resolves the function
13872call to the right definition, with some restrictions. @value{GDBN} does not
13873perform overload resolution involving user-defined type conversions,
13874calls to constructors, or instantiations of templates that do not exist
13875in the program. It also cannot handle ellipsis argument lists or
13876default arguments.
13877
13878It does perform integral conversions and promotions, floating-point
13879promotions, arithmetic conversions, pointer conversions, conversions of
13880class objects to base classes, and standard conversions such as those of
13881functions or arrays to pointers; it requires an exact match on the
13882number of function arguments.
13883
13884Overload resolution is always performed, unless you have specified
13885@code{set overload-resolution off}. @xref{Debugging C Plus Plus,
13886,@value{GDBN} Features for C@t{++}}.
13887
13888You must specify @code{set overload-resolution off} in order to use an
13889explicit function signature to call an overloaded function, as in
13890@smallexample
13891p 'foo(char,int)'('x', 13)
13892@end smallexample
13893
13894The @value{GDBN} command-completion facility can simplify this;
13895see @ref{Completion, ,Command Completion}.
13896
13897@cindex reference declarations
13898@item
13899@value{GDBN} understands variables declared as C@t{++} references; you can use
13900them in expressions just as you do in C@t{++} source---they are automatically
13901dereferenced.
13902
13903In the parameter list shown when @value{GDBN} displays a frame, the values of
13904reference variables are not displayed (unlike other variables); this
13905avoids clutter, since references are often used for large structures.
13906The @emph{address} of a reference variable is always shown, unless
13907you have specified @samp{set print address off}.
13908
13909@item
13910@value{GDBN} supports the C@t{++} name resolution operator @code{::}---your
13911expressions can use it just as expressions in your program do. Since
13912one scope may be defined in another, you can use @code{::} repeatedly if
13913necessary, for example in an expression like
13914@samp{@var{scope1}::@var{scope2}::@var{name}}. @value{GDBN} also allows
13915resolving name scope by reference to source files, in both C and C@t{++}
13916debugging (@pxref{Variables, ,Program Variables}).
13917
13918@item
13919@value{GDBN} performs argument-dependent lookup, following the C@t{++}
13920specification.
13921@end enumerate
13922
13923@node C Defaults
13924@subsubsection C and C@t{++} Defaults
13925
13926@cindex C and C@t{++} defaults
13927
13928If you allow @value{GDBN} to set range checking automatically, it
13929defaults to @code{off} whenever the working language changes to
13930C or C@t{++}. This happens regardless of whether you or @value{GDBN}
13931selects the working language.
13932
13933If you allow @value{GDBN} to set the language automatically, it
13934recognizes source files whose names end with @file{.c}, @file{.C}, or
13935@file{.cc}, etc, and when @value{GDBN} enters code compiled from one of
13936these files, it sets the working language to C or C@t{++}.
13937@xref{Automatically, ,Having @value{GDBN} Infer the Source Language},
13938for further details.
13939
13940@node C Checks
13941@subsubsection C and C@t{++} Type and Range Checks
13942
13943@cindex C and C@t{++} checks
13944
13945By default, when @value{GDBN} parses C or C@t{++} expressions, strict type
13946checking is used. However, if you turn type checking off, @value{GDBN}
13947will allow certain non-standard conversions, such as promoting integer
13948constants to pointers.
13949
13950Range checking, if turned on, is done on mathematical operations. Array
13951indices are not checked, since they are often used to index a pointer
13952that is not itself an array.
13953
13954@node Debugging C
13955@subsubsection @value{GDBN} and C
13956
13957The @code{set print union} and @code{show print union} commands apply to
13958the @code{union} type. When set to @samp{on}, any @code{union} that is
13959inside a @code{struct} or @code{class} is also printed. Otherwise, it
13960appears as @samp{@{...@}}.
13961
13962The @code{@@} operator aids in the debugging of dynamic arrays, formed
13963with pointers and a memory allocation function. @xref{Expressions,
13964,Expressions}.
13965
13966@node Debugging C Plus Plus
13967@subsubsection @value{GDBN} Features for C@t{++}
13968
13969@cindex commands for C@t{++}
13970
13971Some @value{GDBN} commands are particularly useful with C@t{++}, and some are
13972designed specifically for use with C@t{++}. Here is a summary:
13973
13974@table @code
13975@cindex break in overloaded functions
13976@item @r{breakpoint menus}
13977When you want a breakpoint in a function whose name is overloaded,
13978@value{GDBN} has the capability to display a menu of possible breakpoint
13979locations to help you specify which function definition you want.
13980@xref{Ambiguous Expressions,,Ambiguous Expressions}.
13981
13982@cindex overloading in C@t{++}
13983@item rbreak @var{regex}
13984Setting breakpoints using regular expressions is helpful for setting
13985breakpoints on overloaded functions that are not members of any special
13986classes.
13987@xref{Set Breaks, ,Setting Breakpoints}.
13988
13989@cindex C@t{++} exception handling
13990@item catch throw
13991@itemx catch rethrow
13992@itemx catch catch
13993Debug C@t{++} exception handling using these commands. @xref{Set
13994Catchpoints, , Setting Catchpoints}.
13995
13996@cindex inheritance
13997@item ptype @var{typename}
13998Print inheritance relationships as well as other information for type
13999@var{typename}.
14000@xref{Symbols, ,Examining the Symbol Table}.
14001
14002@item info vtbl @var{expression}.
14003The @code{info vtbl} command can be used to display the virtual
14004method tables of the object computed by @var{expression}. This shows
14005one entry per virtual table; there may be multiple virtual tables when
14006multiple inheritance is in use.
14007
14008@cindex C@t{++} symbol display
14009@item set print demangle
14010@itemx show print demangle
14011@itemx set print asm-demangle
14012@itemx show print asm-demangle
14013Control whether C@t{++} symbols display in their source form, both when
14014displaying code as C@t{++} source and when displaying disassemblies.
14015@xref{Print Settings, ,Print Settings}.
14016
14017@item set print object
14018@itemx show print object
14019Choose whether to print derived (actual) or declared types of objects.
14020@xref{Print Settings, ,Print Settings}.
14021
14022@item set print vtbl
14023@itemx show print vtbl
14024Control the format for printing virtual function tables.
14025@xref{Print Settings, ,Print Settings}.
14026(The @code{vtbl} commands do not work on programs compiled with the HP
14027ANSI C@t{++} compiler (@code{aCC}).)
14028
14029@kindex set overload-resolution
14030@cindex overloaded functions, overload resolution
14031@item set overload-resolution on
14032Enable overload resolution for C@t{++} expression evaluation. The default
14033is on. For overloaded functions, @value{GDBN} evaluates the arguments
14034and searches for a function whose signature matches the argument types,
14035using the standard C@t{++} conversion rules (see @ref{C Plus Plus
14036Expressions, ,C@t{++} Expressions}, for details).
14037If it cannot find a match, it emits a message.
14038
14039@item set overload-resolution off
14040Disable overload resolution for C@t{++} expression evaluation. For
14041overloaded functions that are not class member functions, @value{GDBN}
14042chooses the first function of the specified name that it finds in the
14043symbol table, whether or not its arguments are of the correct type. For
14044overloaded functions that are class member functions, @value{GDBN}
14045searches for a function whose signature @emph{exactly} matches the
14046argument types.
14047
14048@kindex show overload-resolution
14049@item show overload-resolution
14050Show the current setting of overload resolution.
14051
14052@item @r{Overloaded symbol names}
14053You can specify a particular definition of an overloaded symbol, using
14054the same notation that is used to declare such symbols in C@t{++}: type
14055@code{@var{symbol}(@var{types})} rather than just @var{symbol}. You can
14056also use the @value{GDBN} command-line word completion facilities to list the
14057available choices, or to finish the type list for you.
14058@xref{Completion,, Command Completion}, for details on how to do this.
14059@end table
14060
14061@node Decimal Floating Point
14062@subsubsection Decimal Floating Point format
14063@cindex decimal floating point format
14064
14065@value{GDBN} can examine, set and perform computations with numbers in
14066decimal floating point format, which in the C language correspond to the
14067@code{_Decimal32}, @code{_Decimal64} and @code{_Decimal128} types as
14068specified by the extension to support decimal floating-point arithmetic.
14069
14070There are two encodings in use, depending on the architecture: BID (Binary
14071Integer Decimal) for x86 and x86-64, and DPD (Densely Packed Decimal) for
14072PowerPC and S/390. @value{GDBN} will use the appropriate encoding for the
14073configured target.
14074
14075Because of a limitation in @file{libdecnumber}, the library used by @value{GDBN}
14076to manipulate decimal floating point numbers, it is not possible to convert
14077(using a cast, for example) integers wider than 32-bit to decimal float.
14078
14079In addition, in order to imitate @value{GDBN}'s behaviour with binary floating
14080point computations, error checking in decimal float operations ignores
14081underflow, overflow and divide by zero exceptions.
14082
14083In the PowerPC architecture, @value{GDBN} provides a set of pseudo-registers
14084to inspect @code{_Decimal128} values stored in floating point registers.
14085See @ref{PowerPC,,PowerPC} for more details.
14086
14087@node D
14088@subsection D
14089
14090@cindex D
14091@value{GDBN} can be used to debug programs written in D and compiled with
14092GDC, LDC or DMD compilers. Currently @value{GDBN} supports only one D
14093specific feature --- dynamic arrays.
14094
14095@node Go
14096@subsection Go
14097
14098@cindex Go (programming language)
14099@value{GDBN} can be used to debug programs written in Go and compiled with
14100@file{gccgo} or @file{6g} compilers.
14101
14102Here is a summary of the Go-specific features and restrictions:
14103
14104@table @code
14105@cindex current Go package
14106@item The current Go package
14107The name of the current package does not need to be specified when
14108specifying global variables and functions.
14109
14110For example, given the program:
14111
14112@example
14113package main
14114var myglob = "Shall we?"
14115func main () @{
14116 // ...
14117@}
14118@end example
14119
14120When stopped inside @code{main} either of these work:
14121
14122@example
14123(gdb) p myglob
14124(gdb) p main.myglob
14125@end example
14126
14127@cindex builtin Go types
14128@item Builtin Go types
14129The @code{string} type is recognized by @value{GDBN} and is printed
14130as a string.
14131
14132@cindex builtin Go functions
14133@item Builtin Go functions
14134The @value{GDBN} expression parser recognizes the @code{unsafe.Sizeof}
14135function and handles it internally.
14136
14137@cindex restrictions on Go expressions
14138@item Restrictions on Go expressions
14139All Go operators are supported except @code{&^}.
14140The Go @code{_} ``blank identifier'' is not supported.
14141Automatic dereferencing of pointers is not supported.
14142@end table
14143
14144@node Objective-C
14145@subsection Objective-C
14146
14147@cindex Objective-C
14148This section provides information about some commands and command
14149options that are useful for debugging Objective-C code. See also
14150@ref{Symbols, info classes}, and @ref{Symbols, info selectors}, for a
14151few more commands specific to Objective-C support.
14152
14153@menu
14154* Method Names in Commands::
14155* The Print Command with Objective-C::
14156@end menu
14157
14158@node Method Names in Commands
14159@subsubsection Method Names in Commands
14160
14161The following commands have been extended to accept Objective-C method
14162names as line specifications:
14163
14164@kindex clear@r{, and Objective-C}
14165@kindex break@r{, and Objective-C}
14166@kindex info line@r{, and Objective-C}
14167@kindex jump@r{, and Objective-C}
14168@kindex list@r{, and Objective-C}
14169@itemize
14170@item @code{clear}
14171@item @code{break}
14172@item @code{info line}
14173@item @code{jump}
14174@item @code{list}
14175@end itemize
14176
14177A fully qualified Objective-C method name is specified as
14178
14179@smallexample
14180-[@var{Class} @var{methodName}]
14181@end smallexample
14182
14183where the minus sign is used to indicate an instance method and a
14184plus sign (not shown) is used to indicate a class method. The class
14185name @var{Class} and method name @var{methodName} are enclosed in
14186brackets, similar to the way messages are specified in Objective-C
14187source code. For example, to set a breakpoint at the @code{create}
14188instance method of class @code{Fruit} in the program currently being
14189debugged, enter:
14190
14191@smallexample
14192break -[Fruit create]
14193@end smallexample
14194
14195To list ten program lines around the @code{initialize} class method,
14196enter:
14197
14198@smallexample
14199list +[NSText initialize]
14200@end smallexample
14201
14202In the current version of @value{GDBN}, the plus or minus sign is
14203required. In future versions of @value{GDBN}, the plus or minus
14204sign will be optional, but you can use it to narrow the search. It
14205is also possible to specify just a method name:
14206
14207@smallexample
14208break create
14209@end smallexample
14210
14211You must specify the complete method name, including any colons. If
14212your program's source files contain more than one @code{create} method,
14213you'll be presented with a numbered list of classes that implement that
14214method. Indicate your choice by number, or type @samp{0} to exit if
14215none apply.
14216
14217As another example, to clear a breakpoint established at the
14218@code{makeKeyAndOrderFront:} method of the @code{NSWindow} class, enter:
14219
14220@smallexample
14221clear -[NSWindow makeKeyAndOrderFront:]
14222@end smallexample
14223
14224@node The Print Command with Objective-C
14225@subsubsection The Print Command With Objective-C
14226@cindex Objective-C, print objects
14227@kindex print-object
14228@kindex po @r{(@code{print-object})}
14229
14230The print command has also been extended to accept methods. For example:
14231
14232@smallexample
14233print -[@var{object} hash]
14234@end smallexample
14235
14236@cindex print an Objective-C object description
14237@cindex @code{_NSPrintForDebugger}, and printing Objective-C objects
14238@noindent
14239will tell @value{GDBN} to send the @code{hash} message to @var{object}
14240and print the result. Also, an additional command has been added,
14241@code{print-object} or @code{po} for short, which is meant to print
14242the description of an object. However, this command may only work
14243with certain Objective-C libraries that have a particular hook
14244function, @code{_NSPrintForDebugger}, defined.
14245
14246@node OpenCL C
14247@subsection OpenCL C
14248
14249@cindex OpenCL C
14250This section provides information about @value{GDBN}s OpenCL C support.
14251
14252@menu
14253* OpenCL C Datatypes::
14254* OpenCL C Expressions::
14255* OpenCL C Operators::
14256@end menu
14257
14258@node OpenCL C Datatypes
14259@subsubsection OpenCL C Datatypes
14260
14261@cindex OpenCL C Datatypes
14262@value{GDBN} supports the builtin scalar and vector datatypes specified
14263by OpenCL 1.1. In addition the half- and double-precision floating point
14264data types of the @code{cl_khr_fp16} and @code{cl_khr_fp64} OpenCL
14265extensions are also known to @value{GDBN}.
14266
14267@node OpenCL C Expressions
14268@subsubsection OpenCL C Expressions
14269
14270@cindex OpenCL C Expressions
14271@value{GDBN} supports accesses to vector components including the access as
14272lvalue where possible. Since OpenCL C is based on C99 most C expressions
14273supported by @value{GDBN} can be used as well.
14274
14275@node OpenCL C Operators
14276@subsubsection OpenCL C Operators
14277
14278@cindex OpenCL C Operators
14279@value{GDBN} supports the operators specified by OpenCL 1.1 for scalar and
14280vector data types.
14281
14282@node Fortran
14283@subsection Fortran
14284@cindex Fortran-specific support in @value{GDBN}
14285
14286@value{GDBN} can be used to debug programs written in Fortran, but it
14287currently supports only the features of Fortran 77 language.
14288
14289@cindex trailing underscore, in Fortran symbols
14290Some Fortran compilers (@sc{gnu} Fortran 77 and Fortran 95 compilers
14291among them) append an underscore to the names of variables and
14292functions. When you debug programs compiled by those compilers, you
14293will need to refer to variables and functions with a trailing
14294underscore.
14295
14296@menu
14297* Fortran Operators:: Fortran operators and expressions
14298* Fortran Defaults:: Default settings for Fortran
14299* Special Fortran Commands:: Special @value{GDBN} commands for Fortran
14300@end menu
14301
14302@node Fortran Operators
14303@subsubsection Fortran Operators and Expressions
14304
14305@cindex Fortran operators and expressions
14306
14307Operators must be defined on values of specific types. For instance,
14308@code{+} is defined on numbers, but not on characters or other non-
14309arithmetic types. Operators are often defined on groups of types.
14310
14311@table @code
14312@item **
14313The exponentiation operator. It raises the first operand to the power
14314of the second one.
14315
14316@item :
14317The range operator. Normally used in the form of array(low:high) to
14318represent a section of array.
14319
14320@item %
14321The access component operator. Normally used to access elements in derived
14322types. Also suitable for unions. As unions aren't part of regular Fortran,
14323this can only happen when accessing a register that uses a gdbarch-defined
14324union type.
14325@end table
14326
14327@node Fortran Defaults
14328@subsubsection Fortran Defaults
14329
14330@cindex Fortran Defaults
14331
14332Fortran symbols are usually case-insensitive, so @value{GDBN} by
14333default uses case-insensitive matches for Fortran symbols. You can
14334change that with the @samp{set case-insensitive} command, see
14335@ref{Symbols}, for the details.
14336
14337@node Special Fortran Commands
14338@subsubsection Special Fortran Commands
14339
14340@cindex Special Fortran commands
14341
14342@value{GDBN} has some commands to support Fortran-specific features,
14343such as displaying common blocks.
14344
14345@table @code
14346@cindex @code{COMMON} blocks, Fortran
14347@kindex info common
14348@item info common @r{[}@var{common-name}@r{]}
14349This command prints the values contained in the Fortran @code{COMMON}
14350block whose name is @var{common-name}. With no argument, the names of
14351all @code{COMMON} blocks visible at the current program location are
14352printed.
14353@end table
14354
14355@node Pascal
14356@subsection Pascal
14357
14358@cindex Pascal support in @value{GDBN}, limitations
14359Debugging Pascal programs which use sets, subranges, file variables, or
14360nested functions does not currently work. @value{GDBN} does not support
14361entering expressions, printing values, or similar features using Pascal
14362syntax.
14363
14364The Pascal-specific command @code{set print pascal_static-members}
14365controls whether static members of Pascal objects are displayed.
14366@xref{Print Settings, pascal_static-members}.
14367
14368@node Modula-2
14369@subsection Modula-2
14370
14371@cindex Modula-2, @value{GDBN} support
14372
14373The extensions made to @value{GDBN} to support Modula-2 only support
14374output from the @sc{gnu} Modula-2 compiler (which is currently being
14375developed). Other Modula-2 compilers are not currently supported, and
14376attempting to debug executables produced by them is most likely
14377to give an error as @value{GDBN} reads in the executable's symbol
14378table.
14379
14380@cindex expressions in Modula-2
14381@menu
14382* M2 Operators:: Built-in operators
14383* Built-In Func/Proc:: Built-in functions and procedures
14384* M2 Constants:: Modula-2 constants
14385* M2 Types:: Modula-2 types
14386* M2 Defaults:: Default settings for Modula-2
14387* Deviations:: Deviations from standard Modula-2
14388* M2 Checks:: Modula-2 type and range checks
14389* M2 Scope:: The scope operators @code{::} and @code{.}
14390* GDB/M2:: @value{GDBN} and Modula-2
14391@end menu
14392
14393@node M2 Operators
14394@subsubsection Operators
14395@cindex Modula-2 operators
14396
14397Operators must be defined on values of specific types. For instance,
14398@code{+} is defined on numbers, but not on structures. Operators are
14399often defined on groups of types. For the purposes of Modula-2, the
14400following definitions hold:
14401
14402@itemize @bullet
14403
14404@item
14405@emph{Integral types} consist of @code{INTEGER}, @code{CARDINAL}, and
14406their subranges.
14407
14408@item
14409@emph{Character types} consist of @code{CHAR} and its subranges.
14410
14411@item
14412@emph{Floating-point types} consist of @code{REAL}.
14413
14414@item
14415@emph{Pointer types} consist of anything declared as @code{POINTER TO
14416@var{type}}.
14417
14418@item
14419@emph{Scalar types} consist of all of the above.
14420
14421@item
14422@emph{Set types} consist of @code{SET} and @code{BITSET} types.
14423
14424@item
14425@emph{Boolean types} consist of @code{BOOLEAN}.
14426@end itemize
14427
14428@noindent
14429The following operators are supported, and appear in order of
14430increasing precedence:
14431
14432@table @code
14433@item ,
14434Function argument or array index separator.
14435
14436@item :=
14437Assignment. The value of @var{var} @code{:=} @var{value} is
14438@var{value}.
14439
14440@item <@r{, }>
14441Less than, greater than on integral, floating-point, or enumerated
14442types.
14443
14444@item <=@r{, }>=
14445Less than or equal to, greater than or equal to
14446on integral, floating-point and enumerated types, or set inclusion on
14447set types. Same precedence as @code{<}.
14448
14449@item =@r{, }<>@r{, }#
14450Equality and two ways of expressing inequality, valid on scalar types.
14451Same precedence as @code{<}. In @value{GDBN} scripts, only @code{<>} is
14452available for inequality, since @code{#} conflicts with the script
14453comment character.
14454
14455@item IN
14456Set membership. Defined on set types and the types of their members.
14457Same precedence as @code{<}.
14458
14459@item OR
14460Boolean disjunction. Defined on boolean types.
14461
14462@item AND@r{, }&
14463Boolean conjunction. Defined on boolean types.
14464
14465@item @@
14466The @value{GDBN} ``artificial array'' operator (@pxref{Expressions, ,Expressions}).
14467
14468@item +@r{, }-
14469Addition and subtraction on integral and floating-point types, or union
14470and difference on set types.
14471
14472@item *
14473Multiplication on integral and floating-point types, or set intersection
14474on set types.
14475
14476@item /
14477Division on floating-point types, or symmetric set difference on set
14478types. Same precedence as @code{*}.
14479
14480@item DIV@r{, }MOD
14481Integer division and remainder. Defined on integral types. Same
14482precedence as @code{*}.
14483
14484@item -
14485Negative. Defined on @code{INTEGER} and @code{REAL} data.
14486
14487@item ^
14488Pointer dereferencing. Defined on pointer types.
14489
14490@item NOT
14491Boolean negation. Defined on boolean types. Same precedence as
14492@code{^}.
14493
14494@item .
14495@code{RECORD} field selector. Defined on @code{RECORD} data. Same
14496precedence as @code{^}.
14497
14498@item []
14499Array indexing. Defined on @code{ARRAY} data. Same precedence as @code{^}.
14500
14501@item ()
14502Procedure argument list. Defined on @code{PROCEDURE} objects. Same precedence
14503as @code{^}.
14504
14505@item ::@r{, }.
14506@value{GDBN} and Modula-2 scope operators.
14507@end table
14508
14509@quotation
14510@emph{Warning:} Set expressions and their operations are not yet supported, so @value{GDBN}
14511treats the use of the operator @code{IN}, or the use of operators
14512@code{+}, @code{-}, @code{*}, @code{/}, @code{=}, , @code{<>}, @code{#},
14513@code{<=}, and @code{>=} on sets as an error.
14514@end quotation
14515
14516
14517@node Built-In Func/Proc
14518@subsubsection Built-in Functions and Procedures
14519@cindex Modula-2 built-ins
14520
14521Modula-2 also makes available several built-in procedures and functions.
14522In describing these, the following metavariables are used:
14523
14524@table @var
14525
14526@item a
14527represents an @code{ARRAY} variable.
14528
14529@item c
14530represents a @code{CHAR} constant or variable.
14531
14532@item i
14533represents a variable or constant of integral type.
14534
14535@item m
14536represents an identifier that belongs to a set. Generally used in the
14537same function with the metavariable @var{s}. The type of @var{s} should
14538be @code{SET OF @var{mtype}} (where @var{mtype} is the type of @var{m}).
14539
14540@item n
14541represents a variable or constant of integral or floating-point type.
14542
14543@item r
14544represents a variable or constant of floating-point type.
14545
14546@item t
14547represents a type.
14548
14549@item v
14550represents a variable.
14551
14552@item x
14553represents a variable or constant of one of many types. See the
14554explanation of the function for details.
14555@end table
14556
14557All Modula-2 built-in procedures also return a result, described below.
14558
14559@table @code
14560@item ABS(@var{n})
14561Returns the absolute value of @var{n}.
14562
14563@item CAP(@var{c})
14564If @var{c} is a lower case letter, it returns its upper case
14565equivalent, otherwise it returns its argument.
14566
14567@item CHR(@var{i})
14568Returns the character whose ordinal value is @var{i}.
14569
14570@item DEC(@var{v})
14571Decrements the value in the variable @var{v} by one. Returns the new value.
14572
14573@item DEC(@var{v},@var{i})
14574Decrements the value in the variable @var{v} by @var{i}. Returns the
14575new value.
14576
14577@item EXCL(@var{m},@var{s})
14578Removes the element @var{m} from the set @var{s}. Returns the new
14579set.
14580
14581@item FLOAT(@var{i})
14582Returns the floating point equivalent of the integer @var{i}.
14583
14584@item HIGH(@var{a})
14585Returns the index of the last member of @var{a}.
14586
14587@item INC(@var{v})
14588Increments the value in the variable @var{v} by one. Returns the new value.
14589
14590@item INC(@var{v},@var{i})
14591Increments the value in the variable @var{v} by @var{i}. Returns the
14592new value.
14593
14594@item INCL(@var{m},@var{s})
14595Adds the element @var{m} to the set @var{s} if it is not already
14596there. Returns the new set.
14597
14598@item MAX(@var{t})
14599Returns the maximum value of the type @var{t}.
14600
14601@item MIN(@var{t})
14602Returns the minimum value of the type @var{t}.
14603
14604@item ODD(@var{i})
14605Returns boolean TRUE if @var{i} is an odd number.
14606
14607@item ORD(@var{x})
14608Returns the ordinal value of its argument. For example, the ordinal
14609value of a character is its @sc{ascii} value (on machines supporting the
14610@sc{ascii} character set). @var{x} must be of an ordered type, which include
14611integral, character and enumerated types.
14612
14613@item SIZE(@var{x})
14614Returns the size of its argument. @var{x} can be a variable or a type.
14615
14616@item TRUNC(@var{r})
14617Returns the integral part of @var{r}.
14618
14619@item TSIZE(@var{x})
14620Returns the size of its argument. @var{x} can be a variable or a type.
14621
14622@item VAL(@var{t},@var{i})
14623Returns the member of the type @var{t} whose ordinal value is @var{i}.
14624@end table
14625
14626@quotation
14627@emph{Warning:} Sets and their operations are not yet supported, so
14628@value{GDBN} treats the use of procedures @code{INCL} and @code{EXCL} as
14629an error.
14630@end quotation
14631
14632@cindex Modula-2 constants
14633@node M2 Constants
14634@subsubsection Constants
14635
14636@value{GDBN} allows you to express the constants of Modula-2 in the following
14637ways:
14638
14639@itemize @bullet
14640
14641@item
14642Integer constants are simply a sequence of digits. When used in an
14643expression, a constant is interpreted to be type-compatible with the
14644rest of the expression. Hexadecimal integers are specified by a
14645trailing @samp{H}, and octal integers by a trailing @samp{B}.
14646
14647@item
14648Floating point constants appear as a sequence of digits, followed by a
14649decimal point and another sequence of digits. An optional exponent can
14650then be specified, in the form @samp{E@r{[}+@r{|}-@r{]}@var{nnn}}, where
14651@samp{@r{[}+@r{|}-@r{]}@var{nnn}} is the desired exponent. All of the
14652digits of the floating point constant must be valid decimal (base 10)
14653digits.
14654
14655@item
14656Character constants consist of a single character enclosed by a pair of
14657like quotes, either single (@code{'}) or double (@code{"}). They may
14658also be expressed by their ordinal value (their @sc{ascii} value, usually)
14659followed by a @samp{C}.
14660
14661@item
14662String constants consist of a sequence of characters enclosed by a
14663pair of like quotes, either single (@code{'}) or double (@code{"}).
14664Escape sequences in the style of C are also allowed. @xref{C
14665Constants, ,C and C@t{++} Constants}, for a brief explanation of escape
14666sequences.
14667
14668@item
14669Enumerated constants consist of an enumerated identifier.
14670
14671@item
14672Boolean constants consist of the identifiers @code{TRUE} and
14673@code{FALSE}.
14674
14675@item
14676Pointer constants consist of integral values only.
14677
14678@item
14679Set constants are not yet supported.
14680@end itemize
14681
14682@node M2 Types
14683@subsubsection Modula-2 Types
14684@cindex Modula-2 types
14685
14686Currently @value{GDBN} can print the following data types in Modula-2
14687syntax: array types, record types, set types, pointer types, procedure
14688types, enumerated types, subrange types and base types. You can also
14689print the contents of variables declared using these type.
14690This section gives a number of simple source code examples together with
14691sample @value{GDBN} sessions.
14692
14693The first example contains the following section of code:
14694
14695@smallexample
14696VAR
14697 s: SET OF CHAR ;
14698 r: [20..40] ;
14699@end smallexample
14700
14701@noindent
14702and you can request @value{GDBN} to interrogate the type and value of
14703@code{r} and @code{s}.
14704
14705@smallexample
14706(@value{GDBP}) print s
14707@{'A'..'C', 'Z'@}
14708(@value{GDBP}) ptype s
14709SET OF CHAR
14710(@value{GDBP}) print r
1471121
14712(@value{GDBP}) ptype r
14713[20..40]
14714@end smallexample
14715
14716@noindent
14717Likewise if your source code declares @code{s} as:
14718
14719@smallexample
14720VAR
14721 s: SET ['A'..'Z'] ;
14722@end smallexample
14723
14724@noindent
14725then you may query the type of @code{s} by:
14726
14727@smallexample
14728(@value{GDBP}) ptype s
14729type = SET ['A'..'Z']
14730@end smallexample
14731
14732@noindent
14733Note that at present you cannot interactively manipulate set
14734expressions using the debugger.
14735
14736The following example shows how you might declare an array in Modula-2
14737and how you can interact with @value{GDBN} to print its type and contents:
14738
14739@smallexample
14740VAR
14741 s: ARRAY [-10..10] OF CHAR ;
14742@end smallexample
14743
14744@smallexample
14745(@value{GDBP}) ptype s
14746ARRAY [-10..10] OF CHAR
14747@end smallexample
14748
14749Note that the array handling is not yet complete and although the type
14750is printed correctly, expression handling still assumes that all
14751arrays have a lower bound of zero and not @code{-10} as in the example
14752above.
14753
14754Here are some more type related Modula-2 examples:
14755
14756@smallexample
14757TYPE
14758 colour = (blue, red, yellow, green) ;
14759 t = [blue..yellow] ;
14760VAR
14761 s: t ;
14762BEGIN
14763 s := blue ;
14764@end smallexample
14765
14766@noindent
14767The @value{GDBN} interaction shows how you can query the data type
14768and value of a variable.
14769
14770@smallexample
14771(@value{GDBP}) print s
14772$1 = blue
14773(@value{GDBP}) ptype t
14774type = [blue..yellow]
14775@end smallexample
14776
14777@noindent
14778In this example a Modula-2 array is declared and its contents
14779displayed. Observe that the contents are written in the same way as
14780their @code{C} counterparts.
14781
14782@smallexample
14783VAR
14784 s: ARRAY [1..5] OF CARDINAL ;
14785BEGIN
14786 s[1] := 1 ;
14787@end smallexample
14788
14789@smallexample
14790(@value{GDBP}) print s
14791$1 = @{1, 0, 0, 0, 0@}
14792(@value{GDBP}) ptype s
14793type = ARRAY [1..5] OF CARDINAL
14794@end smallexample
14795
14796The Modula-2 language interface to @value{GDBN} also understands
14797pointer types as shown in this example:
14798
14799@smallexample
14800VAR
14801 s: POINTER TO ARRAY [1..5] OF CARDINAL ;
14802BEGIN
14803 NEW(s) ;
14804 s^[1] := 1 ;
14805@end smallexample
14806
14807@noindent
14808and you can request that @value{GDBN} describes the type of @code{s}.
14809
14810@smallexample
14811(@value{GDBP}) ptype s
14812type = POINTER TO ARRAY [1..5] OF CARDINAL
14813@end smallexample
14814
14815@value{GDBN} handles compound types as we can see in this example.
14816Here we combine array types, record types, pointer types and subrange
14817types:
14818
14819@smallexample
14820TYPE
14821 foo = RECORD
14822 f1: CARDINAL ;
14823 f2: CHAR ;
14824 f3: myarray ;
14825 END ;
14826
14827 myarray = ARRAY myrange OF CARDINAL ;
14828 myrange = [-2..2] ;
14829VAR
14830 s: POINTER TO ARRAY myrange OF foo ;
14831@end smallexample
14832
14833@noindent
14834and you can ask @value{GDBN} to describe the type of @code{s} as shown
14835below.
14836
14837@smallexample
14838(@value{GDBP}) ptype s
14839type = POINTER TO ARRAY [-2..2] OF foo = RECORD
14840 f1 : CARDINAL;
14841 f2 : CHAR;
14842 f3 : ARRAY [-2..2] OF CARDINAL;
14843END
14844@end smallexample
14845
14846@node M2 Defaults
14847@subsubsection Modula-2 Defaults
14848@cindex Modula-2 defaults
14849
14850If type and range checking are set automatically by @value{GDBN}, they
14851both default to @code{on} whenever the working language changes to
14852Modula-2. This happens regardless of whether you or @value{GDBN}
14853selected the working language.
14854
14855If you allow @value{GDBN} to set the language automatically, then entering
14856code compiled from a file whose name ends with @file{.mod} sets the
14857working language to Modula-2. @xref{Automatically, ,Having @value{GDBN}
14858Infer the Source Language}, for further details.
14859
14860@node Deviations
14861@subsubsection Deviations from Standard Modula-2
14862@cindex Modula-2, deviations from
14863
14864A few changes have been made to make Modula-2 programs easier to debug.
14865This is done primarily via loosening its type strictness:
14866
14867@itemize @bullet
14868@item
14869Unlike in standard Modula-2, pointer constants can be formed by
14870integers. This allows you to modify pointer variables during
14871debugging. (In standard Modula-2, the actual address contained in a
14872pointer variable is hidden from you; it can only be modified
14873through direct assignment to another pointer variable or expression that
14874returned a pointer.)
14875
14876@item
14877C escape sequences can be used in strings and characters to represent
14878non-printable characters. @value{GDBN} prints out strings with these
14879escape sequences embedded. Single non-printable characters are
14880printed using the @samp{CHR(@var{nnn})} format.
14881
14882@item
14883The assignment operator (@code{:=}) returns the value of its right-hand
14884argument.
14885
14886@item
14887All built-in procedures both modify @emph{and} return their argument.
14888@end itemize
14889
14890@node M2 Checks
14891@subsubsection Modula-2 Type and Range Checks
14892@cindex Modula-2 checks
14893
14894@quotation
14895@emph{Warning:} in this release, @value{GDBN} does not yet perform type or
14896range checking.
14897@end quotation
14898@c FIXME remove warning when type/range checks added
14899
14900@value{GDBN} considers two Modula-2 variables type equivalent if:
14901
14902@itemize @bullet
14903@item
14904They are of types that have been declared equivalent via a @code{TYPE
14905@var{t1} = @var{t2}} statement
14906
14907@item
14908They have been declared on the same line. (Note: This is true of the
14909@sc{gnu} Modula-2 compiler, but it may not be true of other compilers.)
14910@end itemize
14911
14912As long as type checking is enabled, any attempt to combine variables
14913whose types are not equivalent is an error.
14914
14915Range checking is done on all mathematical operations, assignment, array
14916index bounds, and all built-in functions and procedures.
14917
14918@node M2 Scope
14919@subsubsection The Scope Operators @code{::} and @code{.}
14920@cindex scope
14921@cindex @code{.}, Modula-2 scope operator
14922@cindex colon, doubled as scope operator
14923@ifinfo
14924@vindex colon-colon@r{, in Modula-2}
14925@c Info cannot handle :: but TeX can.
14926@end ifinfo
14927@ifnotinfo
14928@vindex ::@r{, in Modula-2}
14929@end ifnotinfo
14930
14931There are a few subtle differences between the Modula-2 scope operator
14932(@code{.}) and the @value{GDBN} scope operator (@code{::}). The two have
14933similar syntax:
14934
14935@smallexample
14936
14937@var{module} . @var{id}
14938@var{scope} :: @var{id}
14939@end smallexample
14940
14941@noindent
14942where @var{scope} is the name of a module or a procedure,
14943@var{module} the name of a module, and @var{id} is any declared
14944identifier within your program, except another module.
14945
14946Using the @code{::} operator makes @value{GDBN} search the scope
14947specified by @var{scope} for the identifier @var{id}. If it is not
14948found in the specified scope, then @value{GDBN} searches all scopes
14949enclosing the one specified by @var{scope}.
14950
14951Using the @code{.} operator makes @value{GDBN} search the current scope for
14952the identifier specified by @var{id} that was imported from the
14953definition module specified by @var{module}. With this operator, it is
14954an error if the identifier @var{id} was not imported from definition
14955module @var{module}, or if @var{id} is not an identifier in
14956@var{module}.
14957
14958@node GDB/M2
14959@subsubsection @value{GDBN} and Modula-2
14960
14961Some @value{GDBN} commands have little use when debugging Modula-2 programs.
14962Five subcommands of @code{set print} and @code{show print} apply
14963specifically to C and C@t{++}: @samp{vtbl}, @samp{demangle},
14964@samp{asm-demangle}, @samp{object}, and @samp{union}. The first four
14965apply to C@t{++}, and the last to the C @code{union} type, which has no direct
14966analogue in Modula-2.
14967
14968The @code{@@} operator (@pxref{Expressions, ,Expressions}), while available
14969with any language, is not useful with Modula-2. Its
14970intent is to aid the debugging of @dfn{dynamic arrays}, which cannot be
14971created in Modula-2 as they can in C or C@t{++}. However, because an
14972address can be specified by an integral constant, the construct
14973@samp{@{@var{type}@}@var{adrexp}} is still useful.
14974
14975@cindex @code{#} in Modula-2
14976In @value{GDBN} scripts, the Modula-2 inequality operator @code{#} is
14977interpreted as the beginning of a comment. Use @code{<>} instead.
14978
14979@node Ada
14980@subsection Ada
14981@cindex Ada
14982
14983The extensions made to @value{GDBN} for Ada only support
14984output from the @sc{gnu} Ada (GNAT) compiler.
14985Other Ada compilers are not currently supported, and
14986attempting to debug executables produced by them is most likely
14987to be difficult.
14988
14989
14990@cindex expressions in Ada
14991@menu
14992* Ada Mode Intro:: General remarks on the Ada syntax
14993 and semantics supported by Ada mode
14994 in @value{GDBN}.
14995* Omissions from Ada:: Restrictions on the Ada expression syntax.
14996* Additions to Ada:: Extensions of the Ada expression syntax.
14997* Stopping Before Main Program:: Debugging the program during elaboration.
14998* Ada Exceptions:: Ada Exceptions
14999* Ada Tasks:: Listing and setting breakpoints in tasks.
15000* Ada Tasks and Core Files:: Tasking Support when Debugging Core Files
15001* Ravenscar Profile:: Tasking Support when using the Ravenscar
15002 Profile
15003* Ada Glitches:: Known peculiarities of Ada mode.
15004@end menu
15005
15006@node Ada Mode Intro
15007@subsubsection Introduction
15008@cindex Ada mode, general
15009
15010The Ada mode of @value{GDBN} supports a fairly large subset of Ada expression
15011syntax, with some extensions.
15012The philosophy behind the design of this subset is
15013
15014@itemize @bullet
15015@item
15016That @value{GDBN} should provide basic literals and access to operations for
15017arithmetic, dereferencing, field selection, indexing, and subprogram calls,
15018leaving more sophisticated computations to subprograms written into the
15019program (which therefore may be called from @value{GDBN}).
15020
15021@item
15022That type safety and strict adherence to Ada language restrictions
15023are not particularly important to the @value{GDBN} user.
15024
15025@item
15026That brevity is important to the @value{GDBN} user.
15027@end itemize
15028
15029Thus, for brevity, the debugger acts as if all names declared in
15030user-written packages are directly visible, even if they are not visible
15031according to Ada rules, thus making it unnecessary to fully qualify most
15032names with their packages, regardless of context. Where this causes
15033ambiguity, @value{GDBN} asks the user's intent.
15034
15035The debugger will start in Ada mode if it detects an Ada main program.
15036As for other languages, it will enter Ada mode when stopped in a program that
15037was translated from an Ada source file.
15038
15039While in Ada mode, you may use `@t{--}' for comments. This is useful
15040mostly for documenting command files. The standard @value{GDBN} comment
15041(@samp{#}) still works at the beginning of a line in Ada mode, but not in the
15042middle (to allow based literals).
15043
15044The debugger supports limited overloading. Given a subprogram call in which
15045the function symbol has multiple definitions, it will use the number of
15046actual parameters and some information about their types to attempt to narrow
15047the set of definitions. It also makes very limited use of context, preferring
15048procedures to functions in the context of the @code{call} command, and
15049functions to procedures elsewhere.
15050
15051@node Omissions from Ada
15052@subsubsection Omissions from Ada
15053@cindex Ada, omissions from
15054
15055Here are the notable omissions from the subset:
15056
15057@itemize @bullet
15058@item
15059Only a subset of the attributes are supported:
15060
15061@itemize @minus
15062@item
15063@t{'First}, @t{'Last}, and @t{'Length}
15064 on array objects (not on types and subtypes).
15065
15066@item
15067@t{'Min} and @t{'Max}.
15068
15069@item
15070@t{'Pos} and @t{'Val}.
15071
15072@item
15073@t{'Tag}.
15074
15075@item
15076@t{'Range} on array objects (not subtypes), but only as the right
15077operand of the membership (@code{in}) operator.
15078
15079@item
15080@t{'Access}, @t{'Unchecked_Access}, and
15081@t{'Unrestricted_Access} (a GNAT extension).
15082
15083@item
15084@t{'Address}.
15085@end itemize
15086
15087@item
15088The names in
15089@code{Characters.Latin_1} are not available and
15090concatenation is not implemented. Thus, escape characters in strings are
15091not currently available.
15092
15093@item
15094Equality tests (@samp{=} and @samp{/=}) on arrays test for bitwise
15095equality of representations. They will generally work correctly
15096for strings and arrays whose elements have integer or enumeration types.
15097They may not work correctly for arrays whose element
15098types have user-defined equality, for arrays of real values
15099(in particular, IEEE-conformant floating point, because of negative
15100zeroes and NaNs), and for arrays whose elements contain unused bits with
15101indeterminate values.
15102
15103@item
15104The other component-by-component array operations (@code{and}, @code{or},
15105@code{xor}, @code{not}, and relational tests other than equality)
15106are not implemented.
15107
15108@item
15109@cindex array aggregates (Ada)
15110@cindex record aggregates (Ada)
15111@cindex aggregates (Ada)
15112There is limited support for array and record aggregates. They are
15113permitted only on the right sides of assignments, as in these examples:
15114
15115@smallexample
15116(@value{GDBP}) set An_Array := (1, 2, 3, 4, 5, 6)
15117(@value{GDBP}) set An_Array := (1, others => 0)
15118(@value{GDBP}) set An_Array := (0|4 => 1, 1..3 => 2, 5 => 6)
15119(@value{GDBP}) set A_2D_Array := ((1, 2, 3), (4, 5, 6), (7, 8, 9))
15120(@value{GDBP}) set A_Record := (1, "Peter", True);
15121(@value{GDBP}) set A_Record := (Name => "Peter", Id => 1, Alive => True)
15122@end smallexample
15123
15124Changing a
15125discriminant's value by assigning an aggregate has an
15126undefined effect if that discriminant is used within the record.
15127However, you can first modify discriminants by directly assigning to
15128them (which normally would not be allowed in Ada), and then performing an
15129aggregate assignment. For example, given a variable @code{A_Rec}
15130declared to have a type such as:
15131
15132@smallexample
15133type Rec (Len : Small_Integer := 0) is record
15134 Id : Integer;
15135 Vals : IntArray (1 .. Len);
15136end record;
15137@end smallexample
15138
15139you can assign a value with a different size of @code{Vals} with two
15140assignments:
15141
15142@smallexample
15143(@value{GDBP}) set A_Rec.Len := 4
15144(@value{GDBP}) set A_Rec := (Id => 42, Vals => (1, 2, 3, 4))
15145@end smallexample
15146
15147As this example also illustrates, @value{GDBN} is very loose about the usual
15148rules concerning aggregates. You may leave out some of the
15149components of an array or record aggregate (such as the @code{Len}
15150component in the assignment to @code{A_Rec} above); they will retain their
15151original values upon assignment. You may freely use dynamic values as
15152indices in component associations. You may even use overlapping or
15153redundant component associations, although which component values are
15154assigned in such cases is not defined.
15155
15156@item
15157Calls to dispatching subprograms are not implemented.
15158
15159@item
15160The overloading algorithm is much more limited (i.e., less selective)
15161than that of real Ada. It makes only limited use of the context in
15162which a subexpression appears to resolve its meaning, and it is much
15163looser in its rules for allowing type matches. As a result, some
15164function calls will be ambiguous, and the user will be asked to choose
15165the proper resolution.
15166
15167@item
15168The @code{new} operator is not implemented.
15169
15170@item
15171Entry calls are not implemented.
15172
15173@item
15174Aside from printing, arithmetic operations on the native VAX floating-point
15175formats are not supported.
15176
15177@item
15178It is not possible to slice a packed array.
15179
15180@item
15181The names @code{True} and @code{False}, when not part of a qualified name,
15182are interpreted as if implicitly prefixed by @code{Standard}, regardless of
15183context.
15184Should your program
15185redefine these names in a package or procedure (at best a dubious practice),
15186you will have to use fully qualified names to access their new definitions.
15187@end itemize
15188
15189@node Additions to Ada
15190@subsubsection Additions to Ada
15191@cindex Ada, deviations from
15192
15193As it does for other languages, @value{GDBN} makes certain generic
15194extensions to Ada (@pxref{Expressions}):
15195
15196@itemize @bullet
15197@item
15198If the expression @var{E} is a variable residing in memory (typically
15199a local variable or array element) and @var{N} is a positive integer,
15200then @code{@var{E}@@@var{N}} displays the values of @var{E} and the
15201@var{N}-1 adjacent variables following it in memory as an array. In
15202Ada, this operator is generally not necessary, since its prime use is
15203in displaying parts of an array, and slicing will usually do this in
15204Ada. However, there are occasional uses when debugging programs in
15205which certain debugging information has been optimized away.
15206
15207@item
15208@code{@var{B}::@var{var}} means ``the variable named @var{var} that
15209appears in function or file @var{B}.'' When @var{B} is a file name,
15210you must typically surround it in single quotes.
15211
15212@item
15213The expression @code{@{@var{type}@} @var{addr}} means ``the variable of type
15214@var{type} that appears at address @var{addr}.''
15215
15216@item
15217A name starting with @samp{$} is a convenience variable
15218(@pxref{Convenience Vars}) or a machine register (@pxref{Registers}).
15219@end itemize
15220
15221In addition, @value{GDBN} provides a few other shortcuts and outright
15222additions specific to Ada:
15223
15224@itemize @bullet
15225@item
15226The assignment statement is allowed as an expression, returning
15227its right-hand operand as its value. Thus, you may enter
15228
15229@smallexample
15230(@value{GDBP}) set x := y + 3
15231(@value{GDBP}) print A(tmp := y + 1)
15232@end smallexample
15233
15234@item
15235The semicolon is allowed as an ``operator,'' returning as its value
15236the value of its right-hand operand.
15237This allows, for example,
15238complex conditional breaks:
15239
15240@smallexample
15241(@value{GDBP}) break f
15242(@value{GDBP}) condition 1 (report(i); k += 1; A(k) > 100)
15243@end smallexample
15244
15245@item
15246Rather than use catenation and symbolic character names to introduce special
15247characters into strings, one may instead use a special bracket notation,
15248which is also used to print strings. A sequence of characters of the form
15249@samp{["@var{XX}"]} within a string or character literal denotes the
15250(single) character whose numeric encoding is @var{XX} in hexadecimal. The
15251sequence of characters @samp{["""]} also denotes a single quotation mark
15252in strings. For example,
15253@smallexample
15254 "One line.["0a"]Next line.["0a"]"
15255@end smallexample
15256@noindent
15257contains an ASCII newline character (@code{Ada.Characters.Latin_1.LF})
15258after each period.
15259
15260@item
15261The subtype used as a prefix for the attributes @t{'Pos}, @t{'Min}, and
15262@t{'Max} is optional (and is ignored in any case). For example, it is valid
15263to write
15264
15265@smallexample
15266(@value{GDBP}) print 'max(x, y)
15267@end smallexample
15268
15269@item
15270When printing arrays, @value{GDBN} uses positional notation when the
15271array has a lower bound of 1, and uses a modified named notation otherwise.
15272For example, a one-dimensional array of three integers with a lower bound
15273of 3 might print as
15274
15275@smallexample
15276(3 => 10, 17, 1)
15277@end smallexample
15278
15279@noindent
15280That is, in contrast to valid Ada, only the first component has a @code{=>}
15281clause.
15282
15283@item
15284You may abbreviate attributes in expressions with any unique,
15285multi-character subsequence of
15286their names (an exact match gets preference).
15287For example, you may use @t{a'len}, @t{a'gth}, or @t{a'lh}
15288in place of @t{a'length}.
15289
15290@item
15291@cindex quoting Ada internal identifiers
15292Since Ada is case-insensitive, the debugger normally maps identifiers you type
15293to lower case. The GNAT compiler uses upper-case characters for
15294some of its internal identifiers, which are normally of no interest to users.
15295For the rare occasions when you actually have to look at them,
15296enclose them in angle brackets to avoid the lower-case mapping.
15297For example,
15298@smallexample
15299(@value{GDBP}) print <JMPBUF_SAVE>[0]
15300@end smallexample
15301
15302@item
15303Printing an object of class-wide type or dereferencing an
15304access-to-class-wide value will display all the components of the object's
15305specific type (as indicated by its run-time tag). Likewise, component
15306selection on such a value will operate on the specific type of the
15307object.
15308
15309@end itemize
15310
15311@node Stopping Before Main Program
15312@subsubsection Stopping at the Very Beginning
15313
15314@cindex breakpointing Ada elaboration code
15315It is sometimes necessary to debug the program during elaboration, and
15316before reaching the main procedure.
15317As defined in the Ada Reference
15318Manual, the elaboration code is invoked from a procedure called
15319@code{adainit}. To run your program up to the beginning of
15320elaboration, simply use the following two commands:
15321@code{tbreak adainit} and @code{run}.
15322
15323@node Ada Exceptions
15324@subsubsection Ada Exceptions
15325
15326A command is provided to list all Ada exceptions:
15327
15328@table @code
15329@kindex info exceptions
15330@item info exceptions
15331@itemx info exceptions @var{regexp}
15332The @code{info exceptions} command allows you to list all Ada exceptions
15333defined within the program being debugged, as well as their addresses.
15334With a regular expression, @var{regexp}, as argument, only those exceptions
15335whose names match @var{regexp} are listed.
15336@end table
15337
15338Below is a small example, showing how the command can be used, first
15339without argument, and next with a regular expression passed as an
15340argument.
15341
15342@smallexample
15343(@value{GDBP}) info exceptions
15344All defined Ada exceptions:
15345constraint_error: 0x613da0
15346program_error: 0x613d20
15347storage_error: 0x613ce0
15348tasking_error: 0x613ca0
15349const.aint_global_e: 0x613b00
15350(@value{GDBP}) info exceptions const.aint
15351All Ada exceptions matching regular expression "const.aint":
15352constraint_error: 0x613da0
15353const.aint_global_e: 0x613b00
15354@end smallexample
15355
15356It is also possible to ask @value{GDBN} to stop your program's execution
15357when an exception is raised. For more details, see @ref{Set Catchpoints}.
15358
15359@node Ada Tasks
15360@subsubsection Extensions for Ada Tasks
15361@cindex Ada, tasking
15362
15363Support for Ada tasks is analogous to that for threads (@pxref{Threads}).
15364@value{GDBN} provides the following task-related commands:
15365
15366@table @code
15367@kindex info tasks
15368@item info tasks
15369This command shows a list of current Ada tasks, as in the following example:
15370
15371
15372@smallexample
15373@iftex
15374@leftskip=0.5cm
15375@end iftex
15376(@value{GDBP}) info tasks
15377 ID TID P-ID Pri State Name
15378 1 8088000 0 15 Child Activation Wait main_task
15379 2 80a4000 1 15 Accept Statement b
15380 3 809a800 1 15 Child Activation Wait a
15381* 4 80ae800 3 15 Runnable c
15382
15383@end smallexample
15384
15385@noindent
15386In this listing, the asterisk before the last task indicates it to be the
15387task currently being inspected.
15388
15389@table @asis
15390@item ID
15391Represents @value{GDBN}'s internal task number.
15392
15393@item TID
15394The Ada task ID.
15395
15396@item P-ID
15397The parent's task ID (@value{GDBN}'s internal task number).
15398
15399@item Pri
15400The base priority of the task.
15401
15402@item State
15403Current state of the task.
15404
15405@table @code
15406@item Unactivated
15407The task has been created but has not been activated. It cannot be
15408executing.
15409
15410@item Runnable
15411The task is not blocked for any reason known to Ada. (It may be waiting
15412for a mutex, though.) It is conceptually "executing" in normal mode.
15413
15414@item Terminated
15415The task is terminated, in the sense of ARM 9.3 (5). Any dependents
15416that were waiting on terminate alternatives have been awakened and have
15417terminated themselves.
15418
15419@item Child Activation Wait
15420The task is waiting for created tasks to complete activation.
15421
15422@item Accept Statement
15423The task is waiting on an accept or selective wait statement.
15424
15425@item Waiting on entry call
15426The task is waiting on an entry call.
15427
15428@item Async Select Wait
15429The task is waiting to start the abortable part of an asynchronous
15430select statement.
15431
15432@item Delay Sleep
15433The task is waiting on a select statement with only a delay
15434alternative open.
15435
15436@item Child Termination Wait
15437The task is sleeping having completed a master within itself, and is
15438waiting for the tasks dependent on that master to become terminated or
15439waiting on a terminate Phase.
15440
15441@item Wait Child in Term Alt
15442The task is sleeping waiting for tasks on terminate alternatives to
15443finish terminating.
15444
15445@item Accepting RV with @var{taskno}
15446The task is accepting a rendez-vous with the task @var{taskno}.
15447@end table
15448
15449@item Name
15450Name of the task in the program.
15451
15452@end table
15453
15454@kindex info task @var{taskno}
15455@item info task @var{taskno}
15456This command shows detailled informations on the specified task, as in
15457the following example:
15458@smallexample
15459@iftex
15460@leftskip=0.5cm
15461@end iftex
15462(@value{GDBP}) info tasks
15463 ID TID P-ID Pri State Name
15464 1 8077880 0 15 Child Activation Wait main_task
15465* 2 807c468 1 15 Runnable task_1
15466(@value{GDBP}) info task 2
15467Ada Task: 0x807c468
15468Name: task_1
15469Thread: 0x807f378
15470Parent: 1 (main_task)
15471Base Priority: 15
15472State: Runnable
15473@end smallexample
15474
15475@item task
15476@kindex task@r{ (Ada)}
15477@cindex current Ada task ID
15478This command prints the ID of the current task.
15479
15480@smallexample
15481@iftex
15482@leftskip=0.5cm
15483@end iftex
15484(@value{GDBP}) info tasks
15485 ID TID P-ID Pri State Name
15486 1 8077870 0 15 Child Activation Wait main_task
15487* 2 807c458 1 15 Runnable t
15488(@value{GDBP}) task
15489[Current task is 2]
15490@end smallexample
15491
15492@item task @var{taskno}
15493@cindex Ada task switching
15494This command is like the @code{thread @var{threadno}}
15495command (@pxref{Threads}). It switches the context of debugging
15496from the current task to the given task.
15497
15498@smallexample
15499@iftex
15500@leftskip=0.5cm
15501@end iftex
15502(@value{GDBP}) info tasks
15503 ID TID P-ID Pri State Name
15504 1 8077870 0 15 Child Activation Wait main_task
15505* 2 807c458 1 15 Runnable t
15506(@value{GDBP}) task 1
15507[Switching to task 1]
15508#0 0x8067726 in pthread_cond_wait ()
15509(@value{GDBP}) bt
15510#0 0x8067726 in pthread_cond_wait ()
15511#1 0x8056714 in system.os_interface.pthread_cond_wait ()
15512#2 0x805cb63 in system.task_primitives.operations.sleep ()
15513#3 0x806153e in system.tasking.stages.activate_tasks ()
15514#4 0x804aacc in un () at un.adb:5
15515@end smallexample
15516
15517@item break @var{linespec} task @var{taskno}
15518@itemx break @var{linespec} task @var{taskno} if @dots{}
15519@cindex breakpoints and tasks, in Ada
15520@cindex task breakpoints, in Ada
15521@kindex break @dots{} task @var{taskno}@r{ (Ada)}
15522These commands are like the @code{break @dots{} thread @dots{}}
15523command (@pxref{Thread Stops}).
15524@var{linespec} specifies source lines, as described
15525in @ref{Specify Location}.
15526
15527Use the qualifier @samp{task @var{taskno}} with a breakpoint command
15528to specify that you only want @value{GDBN} to stop the program when a
15529particular Ada task reaches this breakpoint. @var{taskno} is one of the
15530numeric task identifiers assigned by @value{GDBN}, shown in the first
15531column of the @samp{info tasks} display.
15532
15533If you do not specify @samp{task @var{taskno}} when you set a
15534breakpoint, the breakpoint applies to @emph{all} tasks of your
15535program.
15536
15537You can use the @code{task} qualifier on conditional breakpoints as
15538well; in this case, place @samp{task @var{taskno}} before the
15539breakpoint condition (before the @code{if}).
15540
15541For example,
15542
15543@smallexample
15544@iftex
15545@leftskip=0.5cm
15546@end iftex
15547(@value{GDBP}) info tasks
15548 ID TID P-ID Pri State Name
15549 1 140022020 0 15 Child Activation Wait main_task
15550 2 140045060 1 15 Accept/Select Wait t2
15551 3 140044840 1 15 Runnable t1
15552* 4 140056040 1 15 Runnable t3
15553(@value{GDBP}) b 15 task 2
15554Breakpoint 5 at 0x120044cb0: file test_task_debug.adb, line 15.
15555(@value{GDBP}) cont
15556Continuing.
15557task # 1 running
15558task # 2 running
15559
15560Breakpoint 5, test_task_debug () at test_task_debug.adb:15
1556115 flush;
15562(@value{GDBP}) info tasks
15563 ID TID P-ID Pri State Name
15564 1 140022020 0 15 Child Activation Wait main_task
15565* 2 140045060 1 15 Runnable t2
15566 3 140044840 1 15 Runnable t1
15567 4 140056040 1 15 Delay Sleep t3
15568@end smallexample
15569@end table
15570
15571@node Ada Tasks and Core Files
15572@subsubsection Tasking Support when Debugging Core Files
15573@cindex Ada tasking and core file debugging
15574
15575When inspecting a core file, as opposed to debugging a live program,
15576tasking support may be limited or even unavailable, depending on
15577the platform being used.
15578For instance, on x86-linux, the list of tasks is available, but task
15579switching is not supported. On Tru64, however, task switching will work
15580as usual.
15581
15582On certain platforms, including Tru64, the debugger needs to perform some
15583memory writes in order to provide Ada tasking support. When inspecting
15584a core file, this means that the core file must be opened with read-write
15585privileges, using the command @samp{"set write on"} (@pxref{Patching}).
15586Under these circumstances, you should make a backup copy of the core
15587file before inspecting it with @value{GDBN}.
15588
15589@node Ravenscar Profile
15590@subsubsection Tasking Support when using the Ravenscar Profile
15591@cindex Ravenscar Profile
15592
15593The @dfn{Ravenscar Profile} is a subset of the Ada tasking features,
15594specifically designed for systems with safety-critical real-time
15595requirements.
15596
15597@table @code
15598@kindex set ravenscar task-switching on
15599@cindex task switching with program using Ravenscar Profile
15600@item set ravenscar task-switching on
15601Allows task switching when debugging a program that uses the Ravenscar
15602Profile. This is the default.
15603
15604@kindex set ravenscar task-switching off
15605@item set ravenscar task-switching off
15606Turn off task switching when debugging a program that uses the Ravenscar
15607Profile. This is mostly intended to disable the code that adds support
15608for the Ravenscar Profile, in case a bug in either @value{GDBN} or in
15609the Ravenscar runtime is preventing @value{GDBN} from working properly.
15610To be effective, this command should be run before the program is started.
15611
15612@kindex show ravenscar task-switching
15613@item show ravenscar task-switching
15614Show whether it is possible to switch from task to task in a program
15615using the Ravenscar Profile.
15616
15617@end table
15618
15619@node Ada Glitches
15620@subsubsection Known Peculiarities of Ada Mode
15621@cindex Ada, problems
15622
15623Besides the omissions listed previously (@pxref{Omissions from Ada}),
15624we know of several problems with and limitations of Ada mode in
15625@value{GDBN},
15626some of which will be fixed with planned future releases of the debugger
15627and the GNU Ada compiler.
15628
15629@itemize @bullet
15630@item
15631Static constants that the compiler chooses not to materialize as objects in
15632storage are invisible to the debugger.
15633
15634@item
15635Named parameter associations in function argument lists are ignored (the
15636argument lists are treated as positional).
15637
15638@item
15639Many useful library packages are currently invisible to the debugger.
15640
15641@item
15642Fixed-point arithmetic, conversions, input, and output is carried out using
15643floating-point arithmetic, and may give results that only approximate those on
15644the host machine.
15645
15646@item
15647The GNAT compiler never generates the prefix @code{Standard} for any of
15648the standard symbols defined by the Ada language. @value{GDBN} knows about
15649this: it will strip the prefix from names when you use it, and will never
15650look for a name you have so qualified among local symbols, nor match against
15651symbols in other packages or subprograms. If you have
15652defined entities anywhere in your program other than parameters and
15653local variables whose simple names match names in @code{Standard},
15654GNAT's lack of qualification here can cause confusion. When this happens,
15655you can usually resolve the confusion
15656by qualifying the problematic names with package
15657@code{Standard} explicitly.
15658@end itemize
15659
15660Older versions of the compiler sometimes generate erroneous debugging
15661information, resulting in the debugger incorrectly printing the value
15662of affected entities. In some cases, the debugger is able to work
15663around an issue automatically. In other cases, the debugger is able
15664to work around the issue, but the work-around has to be specifically
15665enabled.
15666
15667@kindex set ada trust-PAD-over-XVS
15668@kindex show ada trust-PAD-over-XVS
15669@table @code
15670
15671@item set ada trust-PAD-over-XVS on
15672Configure GDB to strictly follow the GNAT encoding when computing the
15673value of Ada entities, particularly when @code{PAD} and @code{PAD___XVS}
15674types are involved (see @code{ada/exp_dbug.ads} in the GCC sources for
15675a complete description of the encoding used by the GNAT compiler).
15676This is the default.
15677
15678@item set ada trust-PAD-over-XVS off
15679This is related to the encoding using by the GNAT compiler. If @value{GDBN}
15680sometimes prints the wrong value for certain entities, changing @code{ada
15681trust-PAD-over-XVS} to @code{off} activates a work-around which may fix
15682the issue. It is always safe to set @code{ada trust-PAD-over-XVS} to
15683@code{off}, but this incurs a slight performance penalty, so it is
15684recommended to leave this setting to @code{on} unless necessary.
15685
15686@end table
15687
15688@node Unsupported Languages
15689@section Unsupported Languages
15690
15691@cindex unsupported languages
15692@cindex minimal language
15693In addition to the other fully-supported programming languages,
15694@value{GDBN} also provides a pseudo-language, called @code{minimal}.
15695It does not represent a real programming language, but provides a set
15696of capabilities close to what the C or assembly languages provide.
15697This should allow most simple operations to be performed while debugging
15698an application that uses a language currently not supported by @value{GDBN}.
15699
15700If the language is set to @code{auto}, @value{GDBN} will automatically
15701select this language if the current frame corresponds to an unsupported
15702language.
15703
15704@node Symbols
15705@chapter Examining the Symbol Table
15706
15707The commands described in this chapter allow you to inquire about the
15708symbols (names of variables, functions and types) defined in your
15709program. This information is inherent in the text of your program and
15710does not change as your program executes. @value{GDBN} finds it in your
15711program's symbol table, in the file indicated when you started @value{GDBN}
15712(@pxref{File Options, ,Choosing Files}), or by one of the
15713file-management commands (@pxref{Files, ,Commands to Specify Files}).
15714
15715@cindex symbol names
15716@cindex names of symbols
15717@cindex quoting names
15718Occasionally, you may need to refer to symbols that contain unusual
15719characters, which @value{GDBN} ordinarily treats as word delimiters. The
15720most frequent case is in referring to static variables in other
15721source files (@pxref{Variables,,Program Variables}). File names
15722are recorded in object files as debugging symbols, but @value{GDBN} would
15723ordinarily parse a typical file name, like @file{foo.c}, as the three words
15724@samp{foo} @samp{.} @samp{c}. To allow @value{GDBN} to recognize
15725@samp{foo.c} as a single symbol, enclose it in single quotes; for example,
15726
15727@smallexample
15728p 'foo.c'::x
15729@end smallexample
15730
15731@noindent
15732looks up the value of @code{x} in the scope of the file @file{foo.c}.
15733
15734@table @code
15735@cindex case-insensitive symbol names
15736@cindex case sensitivity in symbol names
15737@kindex set case-sensitive
15738@item set case-sensitive on
15739@itemx set case-sensitive off
15740@itemx set case-sensitive auto
15741Normally, when @value{GDBN} looks up symbols, it matches their names
15742with case sensitivity determined by the current source language.
15743Occasionally, you may wish to control that. The command @code{set
15744case-sensitive} lets you do that by specifying @code{on} for
15745case-sensitive matches or @code{off} for case-insensitive ones. If
15746you specify @code{auto}, case sensitivity is reset to the default
15747suitable for the source language. The default is case-sensitive
15748matches for all languages except for Fortran, for which the default is
15749case-insensitive matches.
15750
15751@kindex show case-sensitive
15752@item show case-sensitive
15753This command shows the current setting of case sensitivity for symbols
15754lookups.
15755
15756@kindex set print type methods
15757@item set print type methods
15758@itemx set print type methods on
15759@itemx set print type methods off
15760Normally, when @value{GDBN} prints a class, it displays any methods
15761declared in that class. You can control this behavior either by
15762passing the appropriate flag to @code{ptype}, or using @command{set
15763print type methods}. Specifying @code{on} will cause @value{GDBN} to
15764display the methods; this is the default. Specifying @code{off} will
15765cause @value{GDBN} to omit the methods.
15766
15767@kindex show print type methods
15768@item show print type methods
15769This command shows the current setting of method display when printing
15770classes.
15771
15772@kindex set print type typedefs
15773@item set print type typedefs
15774@itemx set print type typedefs on
15775@itemx set print type typedefs off
15776
15777Normally, when @value{GDBN} prints a class, it displays any typedefs
15778defined in that class. You can control this behavior either by
15779passing the appropriate flag to @code{ptype}, or using @command{set
15780print type typedefs}. Specifying @code{on} will cause @value{GDBN} to
15781display the typedef definitions; this is the default. Specifying
15782@code{off} will cause @value{GDBN} to omit the typedef definitions.
15783Note that this controls whether the typedef definition itself is
15784printed, not whether typedef names are substituted when printing other
15785types.
15786
15787@kindex show print type typedefs
15788@item show print type typedefs
15789This command shows the current setting of typedef display when
15790printing classes.
15791
15792@kindex info address
15793@cindex address of a symbol
15794@item info address @var{symbol}
15795Describe where the data for @var{symbol} is stored. For a register
15796variable, this says which register it is kept in. For a non-register
15797local variable, this prints the stack-frame offset at which the variable
15798is always stored.
15799
15800Note the contrast with @samp{print &@var{symbol}}, which does not work
15801at all for a register variable, and for a stack local variable prints
15802the exact address of the current instantiation of the variable.
15803
15804@kindex info symbol
15805@cindex symbol from address
15806@cindex closest symbol and offset for an address
15807@item info symbol @var{addr}
15808Print the name of a symbol which is stored at the address @var{addr}.
15809If no symbol is stored exactly at @var{addr}, @value{GDBN} prints the
15810nearest symbol and an offset from it:
15811
15812@smallexample
15813(@value{GDBP}) info symbol 0x54320
15814_initialize_vx + 396 in section .text
15815@end smallexample
15816
15817@noindent
15818This is the opposite of the @code{info address} command. You can use
15819it to find out the name of a variable or a function given its address.
15820
15821For dynamically linked executables, the name of executable or shared
15822library containing the symbol is also printed:
15823
15824@smallexample
15825(@value{GDBP}) info symbol 0x400225
15826_start + 5 in section .text of /tmp/a.out
15827(@value{GDBP}) info symbol 0x2aaaac2811cf
15828__read_nocancel + 6 in section .text of /usr/lib64/libc.so.6
15829@end smallexample
15830
15831@kindex whatis
15832@item whatis[/@var{flags}] [@var{arg}]
15833Print the data type of @var{arg}, which can be either an expression
15834or a name of a data type. With no argument, print the data type of
15835@code{$}, the last value in the value history.
15836
15837If @var{arg} is an expression (@pxref{Expressions, ,Expressions}), it
15838is not actually evaluated, and any side-effecting operations (such as
15839assignments or function calls) inside it do not take place.
15840
15841If @var{arg} is a variable or an expression, @code{whatis} prints its
15842literal type as it is used in the source code. If the type was
15843defined using a @code{typedef}, @code{whatis} will @emph{not} print
15844the data type underlying the @code{typedef}. If the type of the
15845variable or the expression is a compound data type, such as
15846@code{struct} or @code{class}, @code{whatis} never prints their
15847fields or methods. It just prints the @code{struct}/@code{class}
15848name (a.k.a.@: its @dfn{tag}). If you want to see the members of
15849such a compound data type, use @code{ptype}.
15850
15851If @var{arg} is a type name that was defined using @code{typedef},
15852@code{whatis} @dfn{unrolls} only one level of that @code{typedef}.
15853Unrolling means that @code{whatis} will show the underlying type used
15854in the @code{typedef} declaration of @var{arg}. However, if that
15855underlying type is also a @code{typedef}, @code{whatis} will not
15856unroll it.
15857
15858For C code, the type names may also have the form @samp{class
15859@var{class-name}}, @samp{struct @var{struct-tag}}, @samp{union
15860@var{union-tag}} or @samp{enum @var{enum-tag}}.
15861
15862@var{flags} can be used to modify how the type is displayed.
15863Available flags are:
15864
15865@table @code
15866@item r
15867Display in ``raw'' form. Normally, @value{GDBN} substitutes template
15868parameters and typedefs defined in a class when printing the class'
15869members. The @code{/r} flag disables this.
15870
15871@item m
15872Do not print methods defined in the class.
15873
15874@item M
15875Print methods defined in the class. This is the default, but the flag
15876exists in case you change the default with @command{set print type methods}.
15877
15878@item t
15879Do not print typedefs defined in the class. Note that this controls
15880whether the typedef definition itself is printed, not whether typedef
15881names are substituted when printing other types.
15882
15883@item T
15884Print typedefs defined in the class. This is the default, but the flag
15885exists in case you change the default with @command{set print type typedefs}.
15886@end table
15887
15888@kindex ptype
15889@item ptype[/@var{flags}] [@var{arg}]
15890@code{ptype} accepts the same arguments as @code{whatis}, but prints a
15891detailed description of the type, instead of just the name of the type.
15892@xref{Expressions, ,Expressions}.
15893
15894Contrary to @code{whatis}, @code{ptype} always unrolls any
15895@code{typedef}s in its argument declaration, whether the argument is
15896a variable, expression, or a data type. This means that @code{ptype}
15897of a variable or an expression will not print literally its type as
15898present in the source code---use @code{whatis} for that. @code{typedef}s at
15899the pointer or reference targets are also unrolled. Only @code{typedef}s of
15900fields, methods and inner @code{class typedef}s of @code{struct}s,
15901@code{class}es and @code{union}s are not unrolled even with @code{ptype}.
15902
15903For example, for this variable declaration:
15904
15905@smallexample
15906typedef double real_t;
15907struct complex @{ real_t real; double imag; @};
15908typedef struct complex complex_t;
15909complex_t var;
15910real_t *real_pointer_var;
15911@end smallexample
15912
15913@noindent
15914the two commands give this output:
15915
15916@smallexample
15917@group
15918(@value{GDBP}) whatis var
15919type = complex_t
15920(@value{GDBP}) ptype var
15921type = struct complex @{
15922 real_t real;
15923 double imag;
15924@}
15925(@value{GDBP}) whatis complex_t
15926type = struct complex
15927(@value{GDBP}) whatis struct complex
15928type = struct complex
15929(@value{GDBP}) ptype struct complex
15930type = struct complex @{
15931 real_t real;
15932 double imag;
15933@}
15934(@value{GDBP}) whatis real_pointer_var
15935type = real_t *
15936(@value{GDBP}) ptype real_pointer_var
15937type = double *
15938@end group
15939@end smallexample
15940
15941@noindent
15942As with @code{whatis}, using @code{ptype} without an argument refers to
15943the type of @code{$}, the last value in the value history.
15944
15945@cindex incomplete type
15946Sometimes, programs use opaque data types or incomplete specifications
15947of complex data structure. If the debug information included in the
15948program does not allow @value{GDBN} to display a full declaration of
15949the data type, it will say @samp{<incomplete type>}. For example,
15950given these declarations:
15951
15952@smallexample
15953 struct foo;
15954 struct foo *fooptr;
15955@end smallexample
15956
15957@noindent
15958but no definition for @code{struct foo} itself, @value{GDBN} will say:
15959
15960@smallexample
15961 (@value{GDBP}) ptype foo
15962 $1 = <incomplete type>
15963@end smallexample
15964
15965@noindent
15966``Incomplete type'' is C terminology for data types that are not
15967completely specified.
15968
15969@kindex info types
15970@item info types @var{regexp}
15971@itemx info types
15972Print a brief description of all types whose names match the regular
15973expression @var{regexp} (or all types in your program, if you supply
15974no argument). Each complete typename is matched as though it were a
15975complete line; thus, @samp{i type value} gives information on all
15976types in your program whose names include the string @code{value}, but
15977@samp{i type ^value$} gives information only on types whose complete
15978name is @code{value}.
15979
15980This command differs from @code{ptype} in two ways: first, like
15981@code{whatis}, it does not print a detailed description; second, it
15982lists all source files where a type is defined.
15983
15984@kindex info type-printers
15985@item info type-printers
15986Versions of @value{GDBN} that ship with Python scripting enabled may
15987have ``type printers'' available. When using @command{ptype} or
15988@command{whatis}, these printers are consulted when the name of a type
15989is needed. @xref{Type Printing API}, for more information on writing
15990type printers.
15991
15992@code{info type-printers} displays all the available type printers.
15993
15994@kindex enable type-printer
15995@kindex disable type-printer
15996@item enable type-printer @var{name}@dots{}
15997@item disable type-printer @var{name}@dots{}
15998These commands can be used to enable or disable type printers.
15999
16000@kindex info scope
16001@cindex local variables
16002@item info scope @var{location}
16003List all the variables local to a particular scope. This command
16004accepts a @var{location} argument---a function name, a source line, or
16005an address preceded by a @samp{*}, and prints all the variables local
16006to the scope defined by that location. (@xref{Specify Location}, for
16007details about supported forms of @var{location}.) For example:
16008
16009@smallexample
16010(@value{GDBP}) @b{info scope command_line_handler}
16011Scope for command_line_handler:
16012Symbol rl is an argument at stack/frame offset 8, length 4.
16013Symbol linebuffer is in static storage at address 0x150a18, length 4.
16014Symbol linelength is in static storage at address 0x150a1c, length 4.
16015Symbol p is a local variable in register $esi, length 4.
16016Symbol p1 is a local variable in register $ebx, length 4.
16017Symbol nline is a local variable in register $edx, length 4.
16018Symbol repeat is a local variable at frame offset -8, length 4.
16019@end smallexample
16020
16021@noindent
16022This command is especially useful for determining what data to collect
16023during a @dfn{trace experiment}, see @ref{Tracepoint Actions,
16024collect}.
16025
16026@kindex info source
16027@item info source
16028Show information about the current source file---that is, the source file for
16029the function containing the current point of execution:
16030@itemize @bullet
16031@item
16032the name of the source file, and the directory containing it,
16033@item
16034the directory it was compiled in,
16035@item
16036its length, in lines,
16037@item
16038which programming language it is written in,
16039@item
16040whether the executable includes debugging information for that file, and
16041if so, what format the information is in (e.g., STABS, Dwarf 2, etc.), and
16042@item
16043whether the debugging information includes information about
16044preprocessor macros.
16045@end itemize
16046
16047
16048@kindex info sources
16049@item info sources
16050Print the names of all source files in your program for which there is
16051debugging information, organized into two lists: files whose symbols
16052have already been read, and files whose symbols will be read when needed.
16053
16054@kindex info functions
16055@item info functions
16056Print the names and data types of all defined functions.
16057
16058@item info functions @var{regexp}
16059Print the names and data types of all defined functions
16060whose names contain a match for regular expression @var{regexp}.
16061Thus, @samp{info fun step} finds all functions whose names
16062include @code{step}; @samp{info fun ^step} finds those whose names
16063start with @code{step}. If a function name contains characters
16064that conflict with the regular expression language (e.g.@:
16065@samp{operator*()}), they may be quoted with a backslash.
16066
16067@kindex info variables
16068@item info variables
16069Print the names and data types of all variables that are defined
16070outside of functions (i.e.@: excluding local variables).
16071
16072@item info variables @var{regexp}
16073Print the names and data types of all variables (except for local
16074variables) whose names contain a match for regular expression
16075@var{regexp}.
16076
16077@kindex info classes
16078@cindex Objective-C, classes and selectors
16079@item info classes
16080@itemx info classes @var{regexp}
16081Display all Objective-C classes in your program, or
16082(with the @var{regexp} argument) all those matching a particular regular
16083expression.
16084
16085@kindex info selectors
16086@item info selectors
16087@itemx info selectors @var{regexp}
16088Display all Objective-C selectors in your program, or
16089(with the @var{regexp} argument) all those matching a particular regular
16090expression.
16091
16092@ignore
16093This was never implemented.
16094@kindex info methods
16095@item info methods
16096@itemx info methods @var{regexp}
16097The @code{info methods} command permits the user to examine all defined
16098methods within C@t{++} program, or (with the @var{regexp} argument) a
16099specific set of methods found in the various C@t{++} classes. Many
16100C@t{++} classes provide a large number of methods. Thus, the output
16101from the @code{ptype} command can be overwhelming and hard to use. The
16102@code{info-methods} command filters the methods, printing only those
16103which match the regular-expression @var{regexp}.
16104@end ignore
16105
16106@cindex opaque data types
16107@kindex set opaque-type-resolution
16108@item set opaque-type-resolution on
16109Tell @value{GDBN} to resolve opaque types. An opaque type is a type
16110declared as a pointer to a @code{struct}, @code{class}, or
16111@code{union}---for example, @code{struct MyType *}---that is used in one
16112source file although the full declaration of @code{struct MyType} is in
16113another source file. The default is on.
16114
16115A change in the setting of this subcommand will not take effect until
16116the next time symbols for a file are loaded.
16117
16118@item set opaque-type-resolution off
16119Tell @value{GDBN} not to resolve opaque types. In this case, the type
16120is printed as follows:
16121@smallexample
16122@{<no data fields>@}
16123@end smallexample
16124
16125@kindex show opaque-type-resolution
16126@item show opaque-type-resolution
16127Show whether opaque types are resolved or not.
16128
16129@kindex maint print symbols
16130@cindex symbol dump
16131@kindex maint print psymbols
16132@cindex partial symbol dump
16133@kindex maint print msymbols
16134@cindex minimal symbol dump
16135@item maint print symbols @var{filename}
16136@itemx maint print psymbols @var{filename}
16137@itemx maint print msymbols @var{filename}
16138Write a dump of debugging symbol data into the file @var{filename}.
16139These commands are used to debug the @value{GDBN} symbol-reading code. Only
16140symbols with debugging data are included. If you use @samp{maint print
16141symbols}, @value{GDBN} includes all the symbols for which it has already
16142collected full details: that is, @var{filename} reflects symbols for
16143only those files whose symbols @value{GDBN} has read. You can use the
16144command @code{info sources} to find out which files these are. If you
16145use @samp{maint print psymbols} instead, the dump shows information about
16146symbols that @value{GDBN} only knows partially---that is, symbols defined in
16147files that @value{GDBN} has skimmed, but not yet read completely. Finally,
16148@samp{maint print msymbols} dumps just the minimal symbol information
16149required for each object file from which @value{GDBN} has read some symbols.
16150@xref{Files, ,Commands to Specify Files}, for a discussion of how
16151@value{GDBN} reads symbols (in the description of @code{symbol-file}).
16152
16153@kindex maint info symtabs
16154@kindex maint info psymtabs
16155@cindex listing @value{GDBN}'s internal symbol tables
16156@cindex symbol tables, listing @value{GDBN}'s internal
16157@cindex full symbol tables, listing @value{GDBN}'s internal
16158@cindex partial symbol tables, listing @value{GDBN}'s internal
16159@item maint info symtabs @r{[} @var{regexp} @r{]}
16160@itemx maint info psymtabs @r{[} @var{regexp} @r{]}
16161
16162List the @code{struct symtab} or @code{struct partial_symtab}
16163structures whose names match @var{regexp}. If @var{regexp} is not
16164given, list them all. The output includes expressions which you can
16165copy into a @value{GDBN} debugging this one to examine a particular
16166structure in more detail. For example:
16167
16168@smallexample
16169(@value{GDBP}) maint info psymtabs dwarf2read
16170@{ objfile /home/gnu/build/gdb/gdb
16171 ((struct objfile *) 0x82e69d0)
16172 @{ psymtab /home/gnu/src/gdb/dwarf2read.c
16173 ((struct partial_symtab *) 0x8474b10)
16174 readin no
16175 fullname (null)
16176 text addresses 0x814d3c8 -- 0x8158074
16177 globals (* (struct partial_symbol **) 0x8507a08 @@ 9)
16178 statics (* (struct partial_symbol **) 0x40e95b78 @@ 2882)
16179 dependencies (none)
16180 @}
16181@}
16182(@value{GDBP}) maint info symtabs
16183(@value{GDBP})
16184@end smallexample
16185@noindent
16186We see that there is one partial symbol table whose filename contains
16187the string @samp{dwarf2read}, belonging to the @samp{gdb} executable;
16188and we see that @value{GDBN} has not read in any symtabs yet at all.
16189If we set a breakpoint on a function, that will cause @value{GDBN} to
16190read the symtab for the compilation unit containing that function:
16191
16192@smallexample
16193(@value{GDBP}) break dwarf2_psymtab_to_symtab
16194Breakpoint 1 at 0x814e5da: file /home/gnu/src/gdb/dwarf2read.c,
16195line 1574.
16196(@value{GDBP}) maint info symtabs
16197@{ objfile /home/gnu/build/gdb/gdb
16198 ((struct objfile *) 0x82e69d0)
16199 @{ symtab /home/gnu/src/gdb/dwarf2read.c
16200 ((struct symtab *) 0x86c1f38)
16201 dirname (null)
16202 fullname (null)
16203 blockvector ((struct blockvector *) 0x86c1bd0) (primary)
16204 linetable ((struct linetable *) 0x8370fa0)
16205 debugformat DWARF 2
16206 @}
16207@}
16208(@value{GDBP})
16209@end smallexample
16210@end table
16211
16212
16213@node Altering
16214@chapter Altering Execution
16215
16216Once you think you have found an error in your program, you might want to
16217find out for certain whether correcting the apparent error would lead to
16218correct results in the rest of the run. You can find the answer by
16219experiment, using the @value{GDBN} features for altering execution of the
16220program.
16221
16222For example, you can store new values into variables or memory
16223locations, give your program a signal, restart it at a different
16224address, or even return prematurely from a function.
16225
16226@menu
16227* Assignment:: Assignment to variables
16228* Jumping:: Continuing at a different address
16229* Signaling:: Giving your program a signal
16230* Returning:: Returning from a function
16231* Calling:: Calling your program's functions
16232* Patching:: Patching your program
16233@end menu
16234
16235@node Assignment
16236@section Assignment to Variables
16237
16238@cindex assignment
16239@cindex setting variables
16240To alter the value of a variable, evaluate an assignment expression.
16241@xref{Expressions, ,Expressions}. For example,
16242
16243@smallexample
16244print x=4
16245@end smallexample
16246
16247@noindent
16248stores the value 4 into the variable @code{x}, and then prints the
16249value of the assignment expression (which is 4).
16250@xref{Languages, ,Using @value{GDBN} with Different Languages}, for more
16251information on operators in supported languages.
16252
16253@kindex set variable
16254@cindex variables, setting
16255If you are not interested in seeing the value of the assignment, use the
16256@code{set} command instead of the @code{print} command. @code{set} is
16257really the same as @code{print} except that the expression's value is
16258not printed and is not put in the value history (@pxref{Value History,
16259,Value History}). The expression is evaluated only for its effects.
16260
16261If the beginning of the argument string of the @code{set} command
16262appears identical to a @code{set} subcommand, use the @code{set
16263variable} command instead of just @code{set}. This command is identical
16264to @code{set} except for its lack of subcommands. For example, if your
16265program has a variable @code{width}, you get an error if you try to set
16266a new value with just @samp{set width=13}, because @value{GDBN} has the
16267command @code{set width}:
16268
16269@smallexample
16270(@value{GDBP}) whatis width
16271type = double
16272(@value{GDBP}) p width
16273$4 = 13
16274(@value{GDBP}) set width=47
16275Invalid syntax in expression.
16276@end smallexample
16277
16278@noindent
16279The invalid expression, of course, is @samp{=47}. In
16280order to actually set the program's variable @code{width}, use
16281
16282@smallexample
16283(@value{GDBP}) set var width=47
16284@end smallexample
16285
16286Because the @code{set} command has many subcommands that can conflict
16287with the names of program variables, it is a good idea to use the
16288@code{set variable} command instead of just @code{set}. For example, if
16289your program has a variable @code{g}, you run into problems if you try
16290to set a new value with just @samp{set g=4}, because @value{GDBN} has
16291the command @code{set gnutarget}, abbreviated @code{set g}:
16292
16293@smallexample
16294@group
16295(@value{GDBP}) whatis g
16296type = double
16297(@value{GDBP}) p g
16298$1 = 1
16299(@value{GDBP}) set g=4
16300(@value{GDBP}) p g
16301$2 = 1
16302(@value{GDBP}) r
16303The program being debugged has been started already.
16304Start it from the beginning? (y or n) y
16305Starting program: /home/smith/cc_progs/a.out
16306"/home/smith/cc_progs/a.out": can't open to read symbols:
16307 Invalid bfd target.
16308(@value{GDBP}) show g
16309The current BFD target is "=4".
16310@end group
16311@end smallexample
16312
16313@noindent
16314The program variable @code{g} did not change, and you silently set the
16315@code{gnutarget} to an invalid value. In order to set the variable
16316@code{g}, use
16317
16318@smallexample
16319(@value{GDBP}) set var g=4
16320@end smallexample
16321
16322@value{GDBN} allows more implicit conversions in assignments than C; you can
16323freely store an integer value into a pointer variable or vice versa,
16324and you can convert any structure to any other structure that is the
16325same length or shorter.
16326@comment FIXME: how do structs align/pad in these conversions?
16327@comment /doc@cygnus.com 18dec1990
16328
16329To store values into arbitrary places in memory, use the @samp{@{@dots{}@}}
16330construct to generate a value of specified type at a specified address
16331(@pxref{Expressions, ,Expressions}). For example, @code{@{int@}0x83040} refers
16332to memory location @code{0x83040} as an integer (which implies a certain size
16333and representation in memory), and
16334
16335@smallexample
16336set @{int@}0x83040 = 4
16337@end smallexample
16338
16339@noindent
16340stores the value 4 into that memory location.
16341
16342@node Jumping
16343@section Continuing at a Different Address
16344
16345Ordinarily, when you continue your program, you do so at the place where
16346it stopped, with the @code{continue} command. You can instead continue at
16347an address of your own choosing, with the following commands:
16348
16349@table @code
16350@kindex jump
16351@kindex j @r{(@code{jump})}
16352@item jump @var{linespec}
16353@itemx j @var{linespec}
16354@itemx jump @var{location}
16355@itemx j @var{location}
16356Resume execution at line @var{linespec} or at address given by
16357@var{location}. Execution stops again immediately if there is a
16358breakpoint there. @xref{Specify Location}, for a description of the
16359different forms of @var{linespec} and @var{location}. It is common
16360practice to use the @code{tbreak} command in conjunction with
16361@code{jump}. @xref{Set Breaks, ,Setting Breakpoints}.
16362
16363The @code{jump} command does not change the current stack frame, or
16364the stack pointer, or the contents of any memory location or any
16365register other than the program counter. If line @var{linespec} is in
16366a different function from the one currently executing, the results may
16367be bizarre if the two functions expect different patterns of arguments or
16368of local variables. For this reason, the @code{jump} command requests
16369confirmation if the specified line is not in the function currently
16370executing. However, even bizarre results are predictable if you are
16371well acquainted with the machine-language code of your program.
16372@end table
16373
16374@c Doesn't work on HP-UX; have to set $pcoqh and $pcoqt.
16375On many systems, you can get much the same effect as the @code{jump}
16376command by storing a new value into the register @code{$pc}. The
16377difference is that this does not start your program running; it only
16378changes the address of where it @emph{will} run when you continue. For
16379example,
16380
16381@smallexample
16382set $pc = 0x485
16383@end smallexample
16384
16385@noindent
16386makes the next @code{continue} command or stepping command execute at
16387address @code{0x485}, rather than at the address where your program stopped.
16388@xref{Continuing and Stepping, ,Continuing and Stepping}.
16389
16390The most common occasion to use the @code{jump} command is to back
16391up---perhaps with more breakpoints set---over a portion of a program
16392that has already executed, in order to examine its execution in more
16393detail.
16394
16395@c @group
16396@node Signaling
16397@section Giving your Program a Signal
16398@cindex deliver a signal to a program
16399
16400@table @code
16401@kindex signal
16402@item signal @var{signal}
16403Resume execution where your program stopped, but immediately give it the
16404signal @var{signal}. @var{signal} can be the name or the number of a
16405signal. For example, on many systems @code{signal 2} and @code{signal
16406SIGINT} are both ways of sending an interrupt signal.
16407
16408Alternatively, if @var{signal} is zero, continue execution without
16409giving a signal. This is useful when your program stopped on account of
16410a signal and would ordinarily see the signal when resumed with the
16411@code{continue} command; @samp{signal 0} causes it to resume without a
16412signal.
16413
16414@code{signal} does not repeat when you press @key{RET} a second time
16415after executing the command.
16416@end table
16417@c @end group
16418
16419Invoking the @code{signal} command is not the same as invoking the
16420@code{kill} utility from the shell. Sending a signal with @code{kill}
16421causes @value{GDBN} to decide what to do with the signal depending on
16422the signal handling tables (@pxref{Signals}). The @code{signal} command
16423passes the signal directly to your program.
16424
16425
16426@node Returning
16427@section Returning from a Function
16428
16429@table @code
16430@cindex returning from a function
16431@kindex return
16432@item return
16433@itemx return @var{expression}
16434You can cancel execution of a function call with the @code{return}
16435command. If you give an
16436@var{expression} argument, its value is used as the function's return
16437value.
16438@end table
16439
16440When you use @code{return}, @value{GDBN} discards the selected stack frame
16441(and all frames within it). You can think of this as making the
16442discarded frame return prematurely. If you wish to specify a value to
16443be returned, give that value as the argument to @code{return}.
16444
16445This pops the selected stack frame (@pxref{Selection, ,Selecting a
16446Frame}), and any other frames inside of it, leaving its caller as the
16447innermost remaining frame. That frame becomes selected. The
16448specified value is stored in the registers used for returning values
16449of functions.
16450
16451The @code{return} command does not resume execution; it leaves the
16452program stopped in the state that would exist if the function had just
16453returned. In contrast, the @code{finish} command (@pxref{Continuing
16454and Stepping, ,Continuing and Stepping}) resumes execution until the
16455selected stack frame returns naturally.
16456
16457@value{GDBN} needs to know how the @var{expression} argument should be set for
16458the inferior. The concrete registers assignment depends on the OS ABI and the
16459type being returned by the selected stack frame. For example it is common for
16460OS ABI to return floating point values in FPU registers while integer values in
16461CPU registers. Still some ABIs return even floating point values in CPU
16462registers. Larger integer widths (such as @code{long long int}) also have
16463specific placement rules. @value{GDBN} already knows the OS ABI from its
16464current target so it needs to find out also the type being returned to make the
16465assignment into the right register(s).
16466
16467Normally, the selected stack frame has debug info. @value{GDBN} will always
16468use the debug info instead of the implicit type of @var{expression} when the
16469debug info is available. For example, if you type @kbd{return -1}, and the
16470function in the current stack frame is declared to return a @code{long long
16471int}, @value{GDBN} transparently converts the implicit @code{int} value of -1
16472into a @code{long long int}:
16473
16474@smallexample
16475Breakpoint 1, func () at gdb.base/return-nodebug.c:29
1647629 return 31;
16477(@value{GDBP}) return -1
16478Make func return now? (y or n) y
16479#0 0x004004f6 in main () at gdb.base/return-nodebug.c:43
1648043 printf ("result=%lld\n", func ());
16481(@value{GDBP})
16482@end smallexample
16483
16484However, if the selected stack frame does not have a debug info, e.g., if the
16485function was compiled without debug info, @value{GDBN} has to find out the type
16486to return from user. Specifying a different type by mistake may set the value
16487in different inferior registers than the caller code expects. For example,
16488typing @kbd{return -1} with its implicit type @code{int} would set only a part
16489of a @code{long long int} result for a debug info less function (on 32-bit
16490architectures). Therefore the user is required to specify the return type by
16491an appropriate cast explicitly:
16492
16493@smallexample
16494Breakpoint 2, 0x0040050b in func ()
16495(@value{GDBP}) return -1
16496Return value type not available for selected stack frame.
16497Please use an explicit cast of the value to return.
16498(@value{GDBP}) return (long long int) -1
16499Make selected stack frame return now? (y or n) y
16500#0 0x00400526 in main ()
16501(@value{GDBP})
16502@end smallexample
16503
16504@node Calling
16505@section Calling Program Functions
16506
16507@table @code
16508@cindex calling functions
16509@cindex inferior functions, calling
16510@item print @var{expr}
16511Evaluate the expression @var{expr} and display the resulting value.
16512@var{expr} may include calls to functions in the program being
16513debugged.
16514
16515@kindex call
16516@item call @var{expr}
16517Evaluate the expression @var{expr} without displaying @code{void}
16518returned values.
16519
16520You can use this variant of the @code{print} command if you want to
16521execute a function from your program that does not return anything
16522(a.k.a.@: @dfn{a void function}), but without cluttering the output
16523with @code{void} returned values that @value{GDBN} will otherwise
16524print. If the result is not void, it is printed and saved in the
16525value history.
16526@end table
16527
16528It is possible for the function you call via the @code{print} or
16529@code{call} command to generate a signal (e.g., if there's a bug in
16530the function, or if you passed it incorrect arguments). What happens
16531in that case is controlled by the @code{set unwindonsignal} command.
16532
16533Similarly, with a C@t{++} program it is possible for the function you
16534call via the @code{print} or @code{call} command to generate an
16535exception that is not handled due to the constraints of the dummy
16536frame. In this case, any exception that is raised in the frame, but has
16537an out-of-frame exception handler will not be found. GDB builds a
16538dummy-frame for the inferior function call, and the unwinder cannot
16539seek for exception handlers outside of this dummy-frame. What happens
16540in that case is controlled by the
16541@code{set unwind-on-terminating-exception} command.
16542
16543@table @code
16544@item set unwindonsignal
16545@kindex set unwindonsignal
16546@cindex unwind stack in called functions
16547@cindex call dummy stack unwinding
16548Set unwinding of the stack if a signal is received while in a function
16549that @value{GDBN} called in the program being debugged. If set to on,
16550@value{GDBN} unwinds the stack it created for the call and restores
16551the context to what it was before the call. If set to off (the
16552default), @value{GDBN} stops in the frame where the signal was
16553received.
16554
16555@item show unwindonsignal
16556@kindex show unwindonsignal
16557Show the current setting of stack unwinding in the functions called by
16558@value{GDBN}.
16559
16560@item set unwind-on-terminating-exception
16561@kindex set unwind-on-terminating-exception
16562@cindex unwind stack in called functions with unhandled exceptions
16563@cindex call dummy stack unwinding on unhandled exception.
16564Set unwinding of the stack if a C@t{++} exception is raised, but left
16565unhandled while in a function that @value{GDBN} called in the program being
16566debugged. If set to on (the default), @value{GDBN} unwinds the stack
16567it created for the call and restores the context to what it was before
16568the call. If set to off, @value{GDBN} the exception is delivered to
16569the default C@t{++} exception handler and the inferior terminated.
16570
16571@item show unwind-on-terminating-exception
16572@kindex show unwind-on-terminating-exception
16573Show the current setting of stack unwinding in the functions called by
16574@value{GDBN}.
16575
16576@end table
16577
16578@cindex weak alias functions
16579Sometimes, a function you wish to call is actually a @dfn{weak alias}
16580for another function. In such case, @value{GDBN} might not pick up
16581the type information, including the types of the function arguments,
16582which causes @value{GDBN} to call the inferior function incorrectly.
16583As a result, the called function will function erroneously and may
16584even crash. A solution to that is to use the name of the aliased
16585function instead.
16586
16587@node Patching
16588@section Patching Programs
16589
16590@cindex patching binaries
16591@cindex writing into executables
16592@cindex writing into corefiles
16593
16594By default, @value{GDBN} opens the file containing your program's
16595executable code (or the corefile) read-only. This prevents accidental
16596alterations to machine code; but it also prevents you from intentionally
16597patching your program's binary.
16598
16599If you'd like to be able to patch the binary, you can specify that
16600explicitly with the @code{set write} command. For example, you might
16601want to turn on internal debugging flags, or even to make emergency
16602repairs.
16603
16604@table @code
16605@kindex set write
16606@item set write on
16607@itemx set write off
16608If you specify @samp{set write on}, @value{GDBN} opens executable and
16609core files for both reading and writing; if you specify @kbd{set write
16610off} (the default), @value{GDBN} opens them read-only.
16611
16612If you have already loaded a file, you must load it again (using the
16613@code{exec-file} or @code{core-file} command) after changing @code{set
16614write}, for your new setting to take effect.
16615
16616@item show write
16617@kindex show write
16618Display whether executable files and core files are opened for writing
16619as well as reading.
16620@end table
16621
16622@node GDB Files
16623@chapter @value{GDBN} Files
16624
16625@value{GDBN} needs to know the file name of the program to be debugged,
16626both in order to read its symbol table and in order to start your
16627program. To debug a core dump of a previous run, you must also tell
16628@value{GDBN} the name of the core dump file.
16629
16630@menu
16631* Files:: Commands to specify files
16632* Separate Debug Files:: Debugging information in separate files
16633* MiniDebugInfo:: Debugging information in a special section
16634* Index Files:: Index files speed up GDB
16635* Symbol Errors:: Errors reading symbol files
16636* Data Files:: GDB data files
16637@end menu
16638
16639@node Files
16640@section Commands to Specify Files
16641
16642@cindex symbol table
16643@cindex core dump file
16644
16645You may want to specify executable and core dump file names. The usual
16646way to do this is at start-up time, using the arguments to
16647@value{GDBN}'s start-up commands (@pxref{Invocation, , Getting In and
16648Out of @value{GDBN}}).
16649
16650Occasionally it is necessary to change to a different file during a
16651@value{GDBN} session. Or you may run @value{GDBN} and forget to
16652specify a file you want to use. Or you are debugging a remote target
16653via @code{gdbserver} (@pxref{Server, file, Using the @code{gdbserver}
16654Program}). In these situations the @value{GDBN} commands to specify
16655new files are useful.
16656
16657@table @code
16658@cindex executable file
16659@kindex file
16660@item file @var{filename}
16661Use @var{filename} as the program to be debugged. It is read for its
16662symbols and for the contents of pure memory. It is also the program
16663executed when you use the @code{run} command. If you do not specify a
16664directory and the file is not found in the @value{GDBN} working directory,
16665@value{GDBN} uses the environment variable @code{PATH} as a list of
16666directories to search, just as the shell does when looking for a program
16667to run. You can change the value of this variable, for both @value{GDBN}
16668and your program, using the @code{path} command.
16669
16670@cindex unlinked object files
16671@cindex patching object files
16672You can load unlinked object @file{.o} files into @value{GDBN} using
16673the @code{file} command. You will not be able to ``run'' an object
16674file, but you can disassemble functions and inspect variables. Also,
16675if the underlying BFD functionality supports it, you could use
16676@kbd{gdb -write} to patch object files using this technique. Note
16677that @value{GDBN} can neither interpret nor modify relocations in this
16678case, so branches and some initialized variables will appear to go to
16679the wrong place. But this feature is still handy from time to time.
16680
16681@item file
16682@code{file} with no argument makes @value{GDBN} discard any information it
16683has on both executable file and the symbol table.
16684
16685@kindex exec-file
16686@item exec-file @r{[} @var{filename} @r{]}
16687Specify that the program to be run (but not the symbol table) is found
16688in @var{filename}. @value{GDBN} searches the environment variable @code{PATH}
16689if necessary to locate your program. Omitting @var{filename} means to
16690discard information on the executable file.
16691
16692@kindex symbol-file
16693@item symbol-file @r{[} @var{filename} @r{]}
16694Read symbol table information from file @var{filename}. @code{PATH} is
16695searched when necessary. Use the @code{file} command to get both symbol
16696table and program to run from the same file.
16697
16698@code{symbol-file} with no argument clears out @value{GDBN} information on your
16699program's symbol table.
16700
16701The @code{symbol-file} command causes @value{GDBN} to forget the contents of
16702some breakpoints and auto-display expressions. This is because they may
16703contain pointers to the internal data recording symbols and data types,
16704which are part of the old symbol table data being discarded inside
16705@value{GDBN}.
16706
16707@code{symbol-file} does not repeat if you press @key{RET} again after
16708executing it once.
16709
16710When @value{GDBN} is configured for a particular environment, it
16711understands debugging information in whatever format is the standard
16712generated for that environment; you may use either a @sc{gnu} compiler, or
16713other compilers that adhere to the local conventions.
16714Best results are usually obtained from @sc{gnu} compilers; for example,
16715using @code{@value{NGCC}} you can generate debugging information for
16716optimized code.
16717
16718For most kinds of object files, with the exception of old SVR3 systems
16719using COFF, the @code{symbol-file} command does not normally read the
16720symbol table in full right away. Instead, it scans the symbol table
16721quickly to find which source files and which symbols are present. The
16722details are read later, one source file at a time, as they are needed.
16723
16724The purpose of this two-stage reading strategy is to make @value{GDBN}
16725start up faster. For the most part, it is invisible except for
16726occasional pauses while the symbol table details for a particular source
16727file are being read. (The @code{set verbose} command can turn these
16728pauses into messages if desired. @xref{Messages/Warnings, ,Optional
16729Warnings and Messages}.)
16730
16731We have not implemented the two-stage strategy for COFF yet. When the
16732symbol table is stored in COFF format, @code{symbol-file} reads the
16733symbol table data in full right away. Note that ``stabs-in-COFF''
16734still does the two-stage strategy, since the debug info is actually
16735in stabs format.
16736
16737@kindex readnow
16738@cindex reading symbols immediately
16739@cindex symbols, reading immediately
16740@item symbol-file @r{[} -readnow @r{]} @var{filename}
16741@itemx file @r{[} -readnow @r{]} @var{filename}
16742You can override the @value{GDBN} two-stage strategy for reading symbol
16743tables by using the @samp{-readnow} option with any of the commands that
16744load symbol table information, if you want to be sure @value{GDBN} has the
16745entire symbol table available.
16746
16747@c FIXME: for now no mention of directories, since this seems to be in
16748@c flux. 13mar1992 status is that in theory GDB would look either in
16749@c current dir or in same dir as myprog; but issues like competing
16750@c GDB's, or clutter in system dirs, mean that in practice right now
16751@c only current dir is used. FFish says maybe a special GDB hierarchy
16752@c (eg rooted in val of env var GDBSYMS) could exist for mappable symbol
16753@c files.
16754
16755@kindex core-file
16756@item core-file @r{[}@var{filename}@r{]}
16757@itemx core
16758Specify the whereabouts of a core dump file to be used as the ``contents
16759of memory''. Traditionally, core files contain only some parts of the
16760address space of the process that generated them; @value{GDBN} can access the
16761executable file itself for other parts.
16762
16763@code{core-file} with no argument specifies that no core file is
16764to be used.
16765
16766Note that the core file is ignored when your program is actually running
16767under @value{GDBN}. So, if you have been running your program and you
16768wish to debug a core file instead, you must kill the subprocess in which
16769the program is running. To do this, use the @code{kill} command
16770(@pxref{Kill Process, ,Killing the Child Process}).
16771
16772@kindex add-symbol-file
16773@cindex dynamic linking
16774@item add-symbol-file @var{filename} @var{address}
16775@itemx add-symbol-file @var{filename} @var{address} @r{[} -readnow @r{]}
16776@itemx add-symbol-file @var{filename} @var{address} -s @var{section} @var{address} @dots{}
16777The @code{add-symbol-file} command reads additional symbol table
16778information from the file @var{filename}. You would use this command
16779when @var{filename} has been dynamically loaded (by some other means)
16780into the program that is running. @var{address} should be the memory
16781address at which the file has been loaded; @value{GDBN} cannot figure
16782this out for itself. You can additionally specify an arbitrary number
16783of @samp{-s @var{section} @var{address}} pairs, to give an explicit
16784section name and base address for that section. You can specify any
16785@var{address} as an expression.
16786
16787The symbol table of the file @var{filename} is added to the symbol table
16788originally read with the @code{symbol-file} command. You can use the
16789@code{add-symbol-file} command any number of times; the new symbol data
16790thus read is kept in addition to the old.
16791
16792Changes can be reverted using the command @code{remove-symbol-file}.
16793
16794@cindex relocatable object files, reading symbols from
16795@cindex object files, relocatable, reading symbols from
16796@cindex reading symbols from relocatable object files
16797@cindex symbols, reading from relocatable object files
16798@cindex @file{.o} files, reading symbols from
16799Although @var{filename} is typically a shared library file, an
16800executable file, or some other object file which has been fully
16801relocated for loading into a process, you can also load symbolic
16802information from relocatable @file{.o} files, as long as:
16803
16804@itemize @bullet
16805@item
16806the file's symbolic information refers only to linker symbols defined in
16807that file, not to symbols defined by other object files,
16808@item
16809every section the file's symbolic information refers to has actually
16810been loaded into the inferior, as it appears in the file, and
16811@item
16812you can determine the address at which every section was loaded, and
16813provide these to the @code{add-symbol-file} command.
16814@end itemize
16815
16816@noindent
16817Some embedded operating systems, like Sun Chorus and VxWorks, can load
16818relocatable files into an already running program; such systems
16819typically make the requirements above easy to meet. However, it's
16820important to recognize that many native systems use complex link
16821procedures (@code{.linkonce} section factoring and C@t{++} constructor table
16822assembly, for example) that make the requirements difficult to meet. In
16823general, one cannot assume that using @code{add-symbol-file} to read a
16824relocatable object file's symbolic information will have the same effect
16825as linking the relocatable object file into the program in the normal
16826way.
16827
16828@code{add-symbol-file} does not repeat if you press @key{RET} after using it.
16829
16830@kindex remove-symbol-file
16831@item remove-symbol-file @var{filename}
16832@item remove-symbol-file -a @var{address}
16833Remove a symbol file added via the @code{add-symbol-file} command. The
16834file to remove can be identified by its @var{filename} or by an @var{address}
16835that lies within the boundaries of this symbol file in memory. Example:
16836
16837@smallexample
16838(gdb) add-symbol-file /home/user/gdb/mylib.so 0x7ffff7ff9480
16839add symbol table from file "/home/user/gdb/mylib.so" at
16840 .text_addr = 0x7ffff7ff9480
16841(y or n) y
16842Reading symbols from /home/user/gdb/mylib.so...done.
16843(gdb) remove-symbol-file -a 0x7ffff7ff9480
16844Remove symbol table from file "/home/user/gdb/mylib.so"? (y or n) y
16845(gdb)
16846@end smallexample
16847
16848
16849@code{remove-symbol-file} does not repeat if you press @key{RET} after using it.
16850
16851@kindex add-symbol-file-from-memory
16852@cindex @code{syscall DSO}
16853@cindex load symbols from memory
16854@item add-symbol-file-from-memory @var{address}
16855Load symbols from the given @var{address} in a dynamically loaded
16856object file whose image is mapped directly into the inferior's memory.
16857For example, the Linux kernel maps a @code{syscall DSO} into each
16858process's address space; this DSO provides kernel-specific code for
16859some system calls. The argument can be any expression whose
16860evaluation yields the address of the file's shared object file header.
16861For this command to work, you must have used @code{symbol-file} or
16862@code{exec-file} commands in advance.
16863
16864@kindex add-shared-symbol-files
16865@kindex assf
16866@item add-shared-symbol-files @var{library-file}
16867@itemx assf @var{library-file}
16868The @code{add-shared-symbol-files} command can currently be used only
16869in the Cygwin build of @value{GDBN} on MS-Windows OS, where it is an
16870alias for the @code{dll-symbols} command (@pxref{Cygwin Native}).
16871@value{GDBN} automatically looks for shared libraries, however if
16872@value{GDBN} does not find yours, you can invoke
16873@code{add-shared-symbol-files}. It takes one argument: the shared
16874library's file name. @code{assf} is a shorthand alias for
16875@code{add-shared-symbol-files}.
16876
16877@kindex section
16878@item section @var{section} @var{addr}
16879The @code{section} command changes the base address of the named
16880@var{section} of the exec file to @var{addr}. This can be used if the
16881exec file does not contain section addresses, (such as in the
16882@code{a.out} format), or when the addresses specified in the file
16883itself are wrong. Each section must be changed separately. The
16884@code{info files} command, described below, lists all the sections and
16885their addresses.
16886
16887@kindex info files
16888@kindex info target
16889@item info files
16890@itemx info target
16891@code{info files} and @code{info target} are synonymous; both print the
16892current target (@pxref{Targets, ,Specifying a Debugging Target}),
16893including the names of the executable and core dump files currently in
16894use by @value{GDBN}, and the files from which symbols were loaded. The
16895command @code{help target} lists all possible targets rather than
16896current ones.
16897
16898@kindex maint info sections
16899@item maint info sections
16900Another command that can give you extra information about program sections
16901is @code{maint info sections}. In addition to the section information
16902displayed by @code{info files}, this command displays the flags and file
16903offset of each section in the executable and core dump files. In addition,
16904@code{maint info sections} provides the following command options (which
16905may be arbitrarily combined):
16906
16907@table @code
16908@item ALLOBJ
16909Display sections for all loaded object files, including shared libraries.
16910@item @var{sections}
16911Display info only for named @var{sections}.
16912@item @var{section-flags}
16913Display info only for sections for which @var{section-flags} are true.
16914The section flags that @value{GDBN} currently knows about are:
16915@table @code
16916@item ALLOC
16917Section will have space allocated in the process when loaded.
16918Set for all sections except those containing debug information.
16919@item LOAD
16920Section will be loaded from the file into the child process memory.
16921Set for pre-initialized code and data, clear for @code{.bss} sections.
16922@item RELOC
16923Section needs to be relocated before loading.
16924@item READONLY
16925Section cannot be modified by the child process.
16926@item CODE
16927Section contains executable code only.
16928@item DATA
16929Section contains data only (no executable code).
16930@item ROM
16931Section will reside in ROM.
16932@item CONSTRUCTOR
16933Section contains data for constructor/destructor lists.
16934@item HAS_CONTENTS
16935Section is not empty.
16936@item NEVER_LOAD
16937An instruction to the linker to not output the section.
16938@item COFF_SHARED_LIBRARY
16939A notification to the linker that the section contains
16940COFF shared library information.
16941@item IS_COMMON
16942Section contains common symbols.
16943@end table
16944@end table
16945@kindex set trust-readonly-sections
16946@cindex read-only sections
16947@item set trust-readonly-sections on
16948Tell @value{GDBN} that readonly sections in your object file
16949really are read-only (i.e.@: that their contents will not change).
16950In that case, @value{GDBN} can fetch values from these sections
16951out of the object file, rather than from the target program.
16952For some targets (notably embedded ones), this can be a significant
16953enhancement to debugging performance.
16954
16955The default is off.
16956
16957@item set trust-readonly-sections off
16958Tell @value{GDBN} not to trust readonly sections. This means that
16959the contents of the section might change while the program is running,
16960and must therefore be fetched from the target when needed.
16961
16962@item show trust-readonly-sections
16963Show the current setting of trusting readonly sections.
16964@end table
16965
16966All file-specifying commands allow both absolute and relative file names
16967as arguments. @value{GDBN} always converts the file name to an absolute file
16968name and remembers it that way.
16969
16970@cindex shared libraries
16971@anchor{Shared Libraries}
16972@value{GDBN} supports @sc{gnu}/Linux, MS-Windows, HP-UX, SunOS, SVr4, Irix,
16973and IBM RS/6000 AIX shared libraries.
16974
16975On MS-Windows @value{GDBN} must be linked with the Expat library to support
16976shared libraries. @xref{Expat}.
16977
16978@value{GDBN} automatically loads symbol definitions from shared libraries
16979when you use the @code{run} command, or when you examine a core file.
16980(Before you issue the @code{run} command, @value{GDBN} does not understand
16981references to a function in a shared library, however---unless you are
16982debugging a core file).
16983
16984On HP-UX, if the program loads a library explicitly, @value{GDBN}
16985automatically loads the symbols at the time of the @code{shl_load} call.
16986
16987@c FIXME: some @value{GDBN} release may permit some refs to undef
16988@c FIXME...symbols---eg in a break cmd---assuming they are from a shared
16989@c FIXME...lib; check this from time to time when updating manual
16990
16991There are times, however, when you may wish to not automatically load
16992symbol definitions from shared libraries, such as when they are
16993particularly large or there are many of them.
16994
16995To control the automatic loading of shared library symbols, use the
16996commands:
16997
16998@table @code
16999@kindex set auto-solib-add
17000@item set auto-solib-add @var{mode}
17001If @var{mode} is @code{on}, symbols from all shared object libraries
17002will be loaded automatically when the inferior begins execution, you
17003attach to an independently started inferior, or when the dynamic linker
17004informs @value{GDBN} that a new library has been loaded. If @var{mode}
17005is @code{off}, symbols must be loaded manually, using the
17006@code{sharedlibrary} command. The default value is @code{on}.
17007
17008@cindex memory used for symbol tables
17009If your program uses lots of shared libraries with debug info that
17010takes large amounts of memory, you can decrease the @value{GDBN}
17011memory footprint by preventing it from automatically loading the
17012symbols from shared libraries. To that end, type @kbd{set
17013auto-solib-add off} before running the inferior, then load each
17014library whose debug symbols you do need with @kbd{sharedlibrary
17015@var{regexp}}, where @var{regexp} is a regular expression that matches
17016the libraries whose symbols you want to be loaded.
17017
17018@kindex show auto-solib-add
17019@item show auto-solib-add
17020Display the current autoloading mode.
17021@end table
17022
17023@cindex load shared library
17024To explicitly load shared library symbols, use the @code{sharedlibrary}
17025command:
17026
17027@table @code
17028@kindex info sharedlibrary
17029@kindex info share
17030@item info share @var{regex}
17031@itemx info sharedlibrary @var{regex}
17032Print the names of the shared libraries which are currently loaded
17033that match @var{regex}. If @var{regex} is omitted then print
17034all shared libraries that are loaded.
17035
17036@kindex sharedlibrary
17037@kindex share
17038@item sharedlibrary @var{regex}
17039@itemx share @var{regex}
17040Load shared object library symbols for files matching a
17041Unix regular expression.
17042As with files loaded automatically, it only loads shared libraries
17043required by your program for a core file or after typing @code{run}. If
17044@var{regex} is omitted all shared libraries required by your program are
17045loaded.
17046
17047@item nosharedlibrary
17048@kindex nosharedlibrary
17049@cindex unload symbols from shared libraries
17050Unload all shared object library symbols. This discards all symbols
17051that have been loaded from all shared libraries. Symbols from shared
17052libraries that were loaded by explicit user requests are not
17053discarded.
17054@end table
17055
17056Sometimes you may wish that @value{GDBN} stops and gives you control
17057when any of shared library events happen. The best way to do this is
17058to use @code{catch load} and @code{catch unload} (@pxref{Set
17059Catchpoints}).
17060
17061@value{GDBN} also supports the the @code{set stop-on-solib-events}
17062command for this. This command exists for historical reasons. It is
17063less useful than setting a catchpoint, because it does not allow for
17064conditions or commands as a catchpoint does.
17065
17066@table @code
17067@item set stop-on-solib-events
17068@kindex set stop-on-solib-events
17069This command controls whether @value{GDBN} should give you control
17070when the dynamic linker notifies it about some shared library event.
17071The most common event of interest is loading or unloading of a new
17072shared library.
17073
17074@item show stop-on-solib-events
17075@kindex show stop-on-solib-events
17076Show whether @value{GDBN} stops and gives you control when shared
17077library events happen.
17078@end table
17079
17080Shared libraries are also supported in many cross or remote debugging
17081configurations. @value{GDBN} needs to have access to the target's libraries;
17082this can be accomplished either by providing copies of the libraries
17083on the host system, or by asking @value{GDBN} to automatically retrieve the
17084libraries from the target. If copies of the target libraries are
17085provided, they need to be the same as the target libraries, although the
17086copies on the target can be stripped as long as the copies on the host are
17087not.
17088
17089@cindex where to look for shared libraries
17090For remote debugging, you need to tell @value{GDBN} where the target
17091libraries are, so that it can load the correct copies---otherwise, it
17092may try to load the host's libraries. @value{GDBN} has two variables
17093to specify the search directories for target libraries.
17094
17095@table @code
17096@cindex prefix for shared library file names
17097@cindex system root, alternate
17098@kindex set solib-absolute-prefix
17099@kindex set sysroot
17100@item set sysroot @var{path}
17101Use @var{path} as the system root for the program being debugged. Any
17102absolute shared library paths will be prefixed with @var{path}; many
17103runtime loaders store the absolute paths to the shared library in the
17104target program's memory. If you use @code{set sysroot} to find shared
17105libraries, they need to be laid out in the same way that they are on
17106the target, with e.g.@: a @file{/lib} and @file{/usr/lib} hierarchy
17107under @var{path}.
17108
17109If @var{path} starts with the sequence @file{remote:}, @value{GDBN} will
17110retrieve the target libraries from the remote system. This is only
17111supported when using a remote target that supports the @code{remote get}
17112command (@pxref{File Transfer,,Sending files to a remote system}).
17113The part of @var{path} following the initial @file{remote:}
17114(if present) is used as system root prefix on the remote file system.
17115@footnote{If you want to specify a local system root using a directory
17116that happens to be named @file{remote:}, you need to use some equivalent
17117variant of the name like @file{./remote:}.}
17118
17119For targets with an MS-DOS based filesystem, such as MS-Windows and
17120SymbianOS, @value{GDBN} tries prefixing a few variants of the target
17121absolute file name with @var{path}. But first, on Unix hosts,
17122@value{GDBN} converts all backslash directory separators into forward
17123slashes, because the backslash is not a directory separator on Unix:
17124
17125@smallexample
17126 c:\foo\bar.dll @result{} c:/foo/bar.dll
17127@end smallexample
17128
17129Then, @value{GDBN} attempts prefixing the target file name with
17130@var{path}, and looks for the resulting file name in the host file
17131system:
17132
17133@smallexample
17134 c:/foo/bar.dll @result{} /path/to/sysroot/c:/foo/bar.dll
17135@end smallexample
17136
17137If that does not find the shared library, @value{GDBN} tries removing
17138the @samp{:} character from the drive spec, both for convenience, and,
17139for the case of the host file system not supporting file names with
17140colons:
17141
17142@smallexample
17143 c:/foo/bar.dll @result{} /path/to/sysroot/c/foo/bar.dll
17144@end smallexample
17145
17146This makes it possible to have a system root that mirrors a target
17147with more than one drive. E.g., you may want to setup your local
17148copies of the target system shared libraries like so (note @samp{c} vs
17149@samp{z}):
17150
17151@smallexample
17152 @file{/path/to/sysroot/c/sys/bin/foo.dll}
17153 @file{/path/to/sysroot/c/sys/bin/bar.dll}
17154 @file{/path/to/sysroot/z/sys/bin/bar.dll}
17155@end smallexample
17156
17157@noindent
17158and point the system root at @file{/path/to/sysroot}, so that
17159@value{GDBN} can find the correct copies of both
17160@file{c:\sys\bin\foo.dll}, and @file{z:\sys\bin\bar.dll}.
17161
17162If that still does not find the shared library, @value{GDBN} tries
17163removing the whole drive spec from the target file name:
17164
17165@smallexample
17166 c:/foo/bar.dll @result{} /path/to/sysroot/foo/bar.dll
17167@end smallexample
17168
17169This last lookup makes it possible to not care about the drive name,
17170if you don't want or need to.
17171
17172The @code{set solib-absolute-prefix} command is an alias for @code{set
17173sysroot}.
17174
17175@cindex default system root
17176@cindex @samp{--with-sysroot}
17177You can set the default system root by using the configure-time
17178@samp{--with-sysroot} option. If the system root is inside
17179@value{GDBN}'s configured binary prefix (set with @samp{--prefix} or
17180@samp{--exec-prefix}), then the default system root will be updated
17181automatically if the installed @value{GDBN} is moved to a new
17182location.
17183
17184@kindex show sysroot
17185@item show sysroot
17186Display the current shared library prefix.
17187
17188@kindex set solib-search-path
17189@item set solib-search-path @var{path}
17190If this variable is set, @var{path} is a colon-separated list of
17191directories to search for shared libraries. @samp{solib-search-path}
17192is used after @samp{sysroot} fails to locate the library, or if the
17193path to the library is relative instead of absolute. If you want to
17194use @samp{solib-search-path} instead of @samp{sysroot}, be sure to set
17195@samp{sysroot} to a nonexistent directory to prevent @value{GDBN} from
17196finding your host's libraries. @samp{sysroot} is preferred; setting
17197it to a nonexistent directory may interfere with automatic loading
17198of shared library symbols.
17199
17200@kindex show solib-search-path
17201@item show solib-search-path
17202Display the current shared library search path.
17203
17204@cindex DOS file-name semantics of file names.
17205@kindex set target-file-system-kind (unix|dos-based|auto)
17206@kindex show target-file-system-kind
17207@item set target-file-system-kind @var{kind}
17208Set assumed file system kind for target reported file names.
17209
17210Shared library file names as reported by the target system may not
17211make sense as is on the system @value{GDBN} is running on. For
17212example, when remote debugging a target that has MS-DOS based file
17213system semantics, from a Unix host, the target may be reporting to
17214@value{GDBN} a list of loaded shared libraries with file names such as
17215@file{c:\Windows\kernel32.dll}. On Unix hosts, there's no concept of
17216drive letters, so the @samp{c:\} prefix is not normally understood as
17217indicating an absolute file name, and neither is the backslash
17218normally considered a directory separator character. In that case,
17219the native file system would interpret this whole absolute file name
17220as a relative file name with no directory components. This would make
17221it impossible to point @value{GDBN} at a copy of the remote target's
17222shared libraries on the host using @code{set sysroot}, and impractical
17223with @code{set solib-search-path}. Setting
17224@code{target-file-system-kind} to @code{dos-based} tells @value{GDBN}
17225to interpret such file names similarly to how the target would, and to
17226map them to file names valid on @value{GDBN}'s native file system
17227semantics. The value of @var{kind} can be @code{"auto"}, in addition
17228to one of the supported file system kinds. In that case, @value{GDBN}
17229tries to determine the appropriate file system variant based on the
17230current target's operating system (@pxref{ABI, ,Configuring the
17231Current ABI}). The supported file system settings are:
17232
17233@table @code
17234@item unix
17235Instruct @value{GDBN} to assume the target file system is of Unix
17236kind. Only file names starting the forward slash (@samp{/}) character
17237are considered absolute, and the directory separator character is also
17238the forward slash.
17239
17240@item dos-based
17241Instruct @value{GDBN} to assume the target file system is DOS based.
17242File names starting with either a forward slash, or a drive letter
17243followed by a colon (e.g., @samp{c:}), are considered absolute, and
17244both the slash (@samp{/}) and the backslash (@samp{\\}) characters are
17245considered directory separators.
17246
17247@item auto
17248Instruct @value{GDBN} to use the file system kind associated with the
17249target operating system (@pxref{ABI, ,Configuring the Current ABI}).
17250This is the default.
17251@end table
17252@end table
17253
17254@cindex file name canonicalization
17255@cindex base name differences
17256When processing file names provided by the user, @value{GDBN}
17257frequently needs to compare them to the file names recorded in the
17258program's debug info. Normally, @value{GDBN} compares just the
17259@dfn{base names} of the files as strings, which is reasonably fast
17260even for very large programs. (The base name of a file is the last
17261portion of its name, after stripping all the leading directories.)
17262This shortcut in comparison is based upon the assumption that files
17263cannot have more than one base name. This is usually true, but
17264references to files that use symlinks or similar filesystem
17265facilities violate that assumption. If your program records files
17266using such facilities, or if you provide file names to @value{GDBN}
17267using symlinks etc., you can set @code{basenames-may-differ} to
17268@code{true} to instruct @value{GDBN} to completely canonicalize each
17269pair of file names it needs to compare. This will make file-name
17270comparisons accurate, but at a price of a significant slowdown.
17271
17272@table @code
17273@item set basenames-may-differ
17274@kindex set basenames-may-differ
17275Set whether a source file may have multiple base names.
17276
17277@item show basenames-may-differ
17278@kindex show basenames-may-differ
17279Show whether a source file may have multiple base names.
17280@end table
17281
17282@node Separate Debug Files
17283@section Debugging Information in Separate Files
17284@cindex separate debugging information files
17285@cindex debugging information in separate files
17286@cindex @file{.debug} subdirectories
17287@cindex debugging information directory, global
17288@cindex global debugging information directories
17289@cindex build ID, and separate debugging files
17290@cindex @file{.build-id} directory
17291
17292@value{GDBN} allows you to put a program's debugging information in a
17293file separate from the executable itself, in a way that allows
17294@value{GDBN} to find and load the debugging information automatically.
17295Since debugging information can be very large---sometimes larger
17296than the executable code itself---some systems distribute debugging
17297information for their executables in separate files, which users can
17298install only when they need to debug a problem.
17299
17300@value{GDBN} supports two ways of specifying the separate debug info
17301file:
17302
17303@itemize @bullet
17304@item
17305The executable contains a @dfn{debug link} that specifies the name of
17306the separate debug info file. The separate debug file's name is
17307usually @file{@var{executable}.debug}, where @var{executable} is the
17308name of the corresponding executable file without leading directories
17309(e.g., @file{ls.debug} for @file{/usr/bin/ls}). In addition, the
17310debug link specifies a 32-bit @dfn{Cyclic Redundancy Check} (CRC)
17311checksum for the debug file, which @value{GDBN} uses to validate that
17312the executable and the debug file came from the same build.
17313
17314@item
17315The executable contains a @dfn{build ID}, a unique bit string that is
17316also present in the corresponding debug info file. (This is supported
17317only on some operating systems, notably those which use the ELF format
17318for binary files and the @sc{gnu} Binutils.) For more details about
17319this feature, see the description of the @option{--build-id}
17320command-line option in @ref{Options, , Command Line Options, ld.info,
17321The GNU Linker}. The debug info file's name is not specified
17322explicitly by the build ID, but can be computed from the build ID, see
17323below.
17324@end itemize
17325
17326Depending on the way the debug info file is specified, @value{GDBN}
17327uses two different methods of looking for the debug file:
17328
17329@itemize @bullet
17330@item
17331For the ``debug link'' method, @value{GDBN} looks up the named file in
17332the directory of the executable file, then in a subdirectory of that
17333directory named @file{.debug}, and finally under each one of the global debug
17334directories, in a subdirectory whose name is identical to the leading
17335directories of the executable's absolute file name.
17336
17337@item
17338For the ``build ID'' method, @value{GDBN} looks in the
17339@file{.build-id} subdirectory of each one of the global debug directories for
17340a file named @file{@var{nn}/@var{nnnnnnnn}.debug}, where @var{nn} are the
17341first 2 hex characters of the build ID bit string, and @var{nnnnnnnn}
17342are the rest of the bit string. (Real build ID strings are 32 or more
17343hex characters, not 10.)
17344@end itemize
17345
17346So, for example, suppose you ask @value{GDBN} to debug
17347@file{/usr/bin/ls}, which has a debug link that specifies the
17348file @file{ls.debug}, and a build ID whose value in hex is
17349@code{abcdef1234}. If the list of the global debug directories includes
17350@file{/usr/lib/debug}, then @value{GDBN} will look for the following
17351debug information files, in the indicated order:
17352
17353@itemize @minus
17354@item
17355@file{/usr/lib/debug/.build-id/ab/cdef1234.debug}
17356@item
17357@file{/usr/bin/ls.debug}
17358@item
17359@file{/usr/bin/.debug/ls.debug}
17360@item
17361@file{/usr/lib/debug/usr/bin/ls.debug}.
17362@end itemize
17363
17364@anchor{debug-file-directory}
17365Global debugging info directories default to what is set by @value{GDBN}
17366configure option @option{--with-separate-debug-dir}. During @value{GDBN} run
17367you can also set the global debugging info directories, and view the list
17368@value{GDBN} is currently using.
17369
17370@table @code
17371
17372@kindex set debug-file-directory
17373@item set debug-file-directory @var{directories}
17374Set the directories which @value{GDBN} searches for separate debugging
17375information files to @var{directory}. Multiple path components can be set
17376concatenating them by a path separator.
17377
17378@kindex show debug-file-directory
17379@item show debug-file-directory
17380Show the directories @value{GDBN} searches for separate debugging
17381information files.
17382
17383@end table
17384
17385@cindex @code{.gnu_debuglink} sections
17386@cindex debug link sections
17387A debug link is a special section of the executable file named
17388@code{.gnu_debuglink}. The section must contain:
17389
17390@itemize
17391@item
17392A filename, with any leading directory components removed, followed by
17393a zero byte,
17394@item
17395zero to three bytes of padding, as needed to reach the next four-byte
17396boundary within the section, and
17397@item
17398a four-byte CRC checksum, stored in the same endianness used for the
17399executable file itself. The checksum is computed on the debugging
17400information file's full contents by the function given below, passing
17401zero as the @var{crc} argument.
17402@end itemize
17403
17404Any executable file format can carry a debug link, as long as it can
17405contain a section named @code{.gnu_debuglink} with the contents
17406described above.
17407
17408@cindex @code{.note.gnu.build-id} sections
17409@cindex build ID sections
17410The build ID is a special section in the executable file (and in other
17411ELF binary files that @value{GDBN} may consider). This section is
17412often named @code{.note.gnu.build-id}, but that name is not mandatory.
17413It contains unique identification for the built files---the ID remains
17414the same across multiple builds of the same build tree. The default
17415algorithm SHA1 produces 160 bits (40 hexadecimal characters) of the
17416content for the build ID string. The same section with an identical
17417value is present in the original built binary with symbols, in its
17418stripped variant, and in the separate debugging information file.
17419
17420The debugging information file itself should be an ordinary
17421executable, containing a full set of linker symbols, sections, and
17422debugging information. The sections of the debugging information file
17423should have the same names, addresses, and sizes as the original file,
17424but they need not contain any data---much like a @code{.bss} section
17425in an ordinary executable.
17426
17427The @sc{gnu} binary utilities (Binutils) package includes the
17428@samp{objcopy} utility that can produce
17429the separated executable / debugging information file pairs using the
17430following commands:
17431
17432@smallexample
17433@kbd{objcopy --only-keep-debug foo foo.debug}
17434@kbd{strip -g foo}
17435@end smallexample
17436
17437@noindent
17438These commands remove the debugging
17439information from the executable file @file{foo} and place it in the file
17440@file{foo.debug}. You can use the first, second or both methods to link the
17441two files:
17442
17443@itemize @bullet
17444@item
17445The debug link method needs the following additional command to also leave
17446behind a debug link in @file{foo}:
17447
17448@smallexample
17449@kbd{objcopy --add-gnu-debuglink=foo.debug foo}
17450@end smallexample
17451
17452Ulrich Drepper's @file{elfutils} package, starting with version 0.53, contains
17453a version of the @code{strip} command such that the command @kbd{strip foo -f
17454foo.debug} has the same functionality as the two @code{objcopy} commands and
17455the @code{ln -s} command above, together.
17456
17457@item
17458Build ID gets embedded into the main executable using @code{ld --build-id} or
17459the @value{NGCC} counterpart @code{gcc -Wl,--build-id}. Build ID support plus
17460compatibility fixes for debug files separation are present in @sc{gnu} binary
17461utilities (Binutils) package since version 2.18.
17462@end itemize
17463
17464@noindent
17465
17466@cindex CRC algorithm definition
17467The CRC used in @code{.gnu_debuglink} is the CRC-32 defined in
17468IEEE 802.3 using the polynomial:
17469
17470@c TexInfo requires naked braces for multi-digit exponents for Tex
17471@c output, but this causes HTML output to barf. HTML has to be set using
17472@c raw commands. So we end up having to specify this equation in 2
17473@c different ways!
17474@ifhtml
17475@display
17476@html
17477 <em>x</em><sup>32</sup> + <em>x</em><sup>26</sup> + <em>x</em><sup>23</sup> + <em>x</em><sup>22</sup> + <em>x</em><sup>16</sup> + <em>x</em><sup>12</sup> + <em>x</em><sup>11</sup>
17478 + <em>x</em><sup>10</sup> + <em>x</em><sup>8</sup> + <em>x</em><sup>7</sup> + <em>x</em><sup>5</sup> + <em>x</em><sup>4</sup> + <em>x</em><sup>2</sup> + <em>x</em> + 1
17479@end html
17480@end display
17481@end ifhtml
17482@ifnothtml
17483@display
17484 @math{x^{32} + x^{26} + x^{23} + x^{22} + x^{16} + x^{12} + x^{11}}
17485 @math{+ x^{10} + x^8 + x^7 + x^5 + x^4 + x^2 + x + 1}
17486@end display
17487@end ifnothtml
17488
17489The function is computed byte at a time, taking the least
17490significant bit of each byte first. The initial pattern
17491@code{0xffffffff} is used, to ensure leading zeros affect the CRC and
17492the final result is inverted to ensure trailing zeros also affect the
17493CRC.
17494
17495@emph{Note:} This is the same CRC polynomial as used in handling the
17496@dfn{Remote Serial Protocol} @code{qCRC} packet (@pxref{Remote Protocol,
17497, @value{GDBN} Remote Serial Protocol}). However in the
17498case of the Remote Serial Protocol, the CRC is computed @emph{most}
17499significant bit first, and the result is not inverted, so trailing
17500zeros have no effect on the CRC value.
17501
17502To complete the description, we show below the code of the function
17503which produces the CRC used in @code{.gnu_debuglink}. Inverting the
17504initially supplied @code{crc} argument means that an initial call to
17505this function passing in zero will start computing the CRC using
17506@code{0xffffffff}.
17507
17508@kindex gnu_debuglink_crc32
17509@smallexample
17510unsigned long
17511gnu_debuglink_crc32 (unsigned long crc,
17512 unsigned char *buf, size_t len)
17513@{
17514 static const unsigned long crc32_table[256] =
17515 @{
17516 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419,
17517 0x706af48f, 0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4,
17518 0xe0d5e91e, 0x97d2d988, 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07,
17519 0x90bf1d91, 0x1db71064, 0x6ab020f2, 0xf3b97148, 0x84be41de,
17520 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7, 0x136c9856,
17521 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9,
17522 0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4,
17523 0xa2677172, 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b,
17524 0x35b5a8fa, 0x42b2986c, 0xdbbbc9d6, 0xacbcf940, 0x32d86ce3,
17525 0x45df5c75, 0xdcd60dcf, 0xabd13d59, 0x26d930ac, 0x51de003a,
17526 0xc8d75180, 0xbfd06116, 0x21b4f4b5, 0x56b3c423, 0xcfba9599,
17527 0xb8bda50f, 0x2802b89e, 0x5f058808, 0xc60cd9b2, 0xb10be924,
17528 0x2f6f7c87, 0x58684c11, 0xc1611dab, 0xb6662d3d, 0x76dc4190,
17529 0x01db7106, 0x98d220bc, 0xefd5102a, 0x71b18589, 0x06b6b51f,
17530 0x9fbfe4a5, 0xe8b8d433, 0x7807c9a2, 0x0f00f934, 0x9609a88e,
17531 0xe10e9818, 0x7f6a0dbb, 0x086d3d2d, 0x91646c97, 0xe6635c01,
17532 0x6b6b51f4, 0x1c6c6162, 0x856530d8, 0xf262004e, 0x6c0695ed,
17533 0x1b01a57b, 0x8208f4c1, 0xf50fc457, 0x65b0d9c6, 0x12b7e950,
17534 0x8bbeb8ea, 0xfcb9887c, 0x62dd1ddf, 0x15da2d49, 0x8cd37cf3,
17535 0xfbd44c65, 0x4db26158, 0x3ab551ce, 0xa3bc0074, 0xd4bb30e2,
17536 0x4adfa541, 0x3dd895d7, 0xa4d1c46d, 0xd3d6f4fb, 0x4369e96a,
17537 0x346ed9fc, 0xad678846, 0xda60b8d0, 0x44042d73, 0x33031de5,
17538 0xaa0a4c5f, 0xdd0d7cc9, 0x5005713c, 0x270241aa, 0xbe0b1010,
17539 0xc90c2086, 0x5768b525, 0x206f85b3, 0xb966d409, 0xce61e49f,
17540 0x5edef90e, 0x29d9c998, 0xb0d09822, 0xc7d7a8b4, 0x59b33d17,
17541 0x2eb40d81, 0xb7bd5c3b, 0xc0ba6cad, 0xedb88320, 0x9abfb3b6,
17542 0x03b6e20c, 0x74b1d29a, 0xead54739, 0x9dd277af, 0x04db2615,
17543 0x73dc1683, 0xe3630b12, 0x94643b84, 0x0d6d6a3e, 0x7a6a5aa8,
17544 0xe40ecf0b, 0x9309ff9d, 0x0a00ae27, 0x7d079eb1, 0xf00f9344,
17545 0x8708a3d2, 0x1e01f268, 0x6906c2fe, 0xf762575d, 0x806567cb,
17546 0x196c3671, 0x6e6b06e7, 0xfed41b76, 0x89d32be0, 0x10da7a5a,
17547 0x67dd4acc, 0xf9b9df6f, 0x8ebeeff9, 0x17b7be43, 0x60b08ed5,
17548 0xd6d6a3e8, 0xa1d1937e, 0x38d8c2c4, 0x4fdff252, 0xd1bb67f1,
17549 0xa6bc5767, 0x3fb506dd, 0x48b2364b, 0xd80d2bda, 0xaf0a1b4c,
17550 0x36034af6, 0x41047a60, 0xdf60efc3, 0xa867df55, 0x316e8eef,
17551 0x4669be79, 0xcb61b38c, 0xbc66831a, 0x256fd2a0, 0x5268e236,
17552 0xcc0c7795, 0xbb0b4703, 0x220216b9, 0x5505262f, 0xc5ba3bbe,
17553 0xb2bd0b28, 0x2bb45a92, 0x5cb36a04, 0xc2d7ffa7, 0xb5d0cf31,
17554 0x2cd99e8b, 0x5bdeae1d, 0x9b64c2b0, 0xec63f226, 0x756aa39c,
17555 0x026d930a, 0x9c0906a9, 0xeb0e363f, 0x72076785, 0x05005713,
17556 0x95bf4a82, 0xe2b87a14, 0x7bb12bae, 0x0cb61b38, 0x92d28e9b,
17557 0xe5d5be0d, 0x7cdcefb7, 0x0bdbdf21, 0x86d3d2d4, 0xf1d4e242,
17558 0x68ddb3f8, 0x1fda836e, 0x81be16cd, 0xf6b9265b, 0x6fb077e1,
17559 0x18b74777, 0x88085ae6, 0xff0f6a70, 0x66063bca, 0x11010b5c,
17560 0x8f659eff, 0xf862ae69, 0x616bffd3, 0x166ccf45, 0xa00ae278,
17561 0xd70dd2ee, 0x4e048354, 0x3903b3c2, 0xa7672661, 0xd06016f7,
17562 0x4969474d, 0x3e6e77db, 0xaed16a4a, 0xd9d65adc, 0x40df0b66,
17563 0x37d83bf0, 0xa9bcae53, 0xdebb9ec5, 0x47b2cf7f, 0x30b5ffe9,
17564 0xbdbdf21c, 0xcabac28a, 0x53b39330, 0x24b4a3a6, 0xbad03605,
17565 0xcdd70693, 0x54de5729, 0x23d967bf, 0xb3667a2e, 0xc4614ab8,
17566 0x5d681b02, 0x2a6f2b94, 0xb40bbe37, 0xc30c8ea1, 0x5a05df1b,
17567 0x2d02ef8d
17568 @};
17569 unsigned char *end;
17570
17571 crc = ~crc & 0xffffffff;
17572 for (end = buf + len; buf < end; ++buf)
17573 crc = crc32_table[(crc ^ *buf) & 0xff] ^ (crc >> 8);
17574 return ~crc & 0xffffffff;
17575@}
17576@end smallexample
17577
17578@noindent
17579This computation does not apply to the ``build ID'' method.
17580
17581@node MiniDebugInfo
17582@section Debugging information in a special section
17583@cindex separate debug sections
17584@cindex @samp{.gnu_debugdata} section
17585
17586Some systems ship pre-built executables and libraries that have a
17587special @samp{.gnu_debugdata} section. This feature is called
17588@dfn{MiniDebugInfo}. This section holds an LZMA-compressed object and
17589is used to supply extra symbols for backtraces.
17590
17591The intent of this section is to provide extra minimal debugging
17592information for use in simple backtraces. It is not intended to be a
17593replacement for full separate debugging information (@pxref{Separate
17594Debug Files}). The example below shows the intended use; however,
17595@value{GDBN} does not currently put restrictions on what sort of
17596debugging information might be included in the section.
17597
17598@value{GDBN} has support for this extension. If the section exists,
17599then it is used provided that no other source of debugging information
17600can be found, and that @value{GDBN} was configured with LZMA support.
17601
17602This section can be easily created using @command{objcopy} and other
17603standard utilities:
17604
17605@smallexample
17606# Extract the dynamic symbols from the main binary, there is no need
17607# to also have these in the normal symbol table.
17608nm -D @var{binary} --format=posix --defined-only \
17609 | awk '@{ print $1 @}' | sort > dynsyms
17610
17611# Extract all the text (i.e. function) symbols from the debuginfo.
17612# (Note that we actually also accept "D" symbols, for the benefit
17613# of platforms like PowerPC64 that use function descriptors.)
17614nm @var{binary} --format=posix --defined-only \
17615 | awk '@{ if ($2 == "T" || $2 == "t" || $2 == "D") print $1 @}' \
17616 | sort > funcsyms
17617
17618# Keep all the function symbols not already in the dynamic symbol
17619# table.
17620comm -13 dynsyms funcsyms > keep_symbols
17621
17622# Separate full debug info into debug binary.
17623objcopy --only-keep-debug @var{binary} debug
17624
17625# Copy the full debuginfo, keeping only a minimal set of symbols and
17626# removing some unnecessary sections.
17627objcopy -S --remove-section .gdb_index --remove-section .comment \
17628 --keep-symbols=keep_symbols debug mini_debuginfo
17629
17630# Drop the full debug info from the original binary.
17631strip --strip-all -R .comment @var{binary}
17632
17633# Inject the compressed data into the .gnu_debugdata section of the
17634# original binary.
17635xz mini_debuginfo
17636objcopy --add-section .gnu_debugdata=mini_debuginfo.xz @var{binary}
17637@end smallexample
17638
17639@node Index Files
17640@section Index Files Speed Up @value{GDBN}
17641@cindex index files
17642@cindex @samp{.gdb_index} section
17643
17644When @value{GDBN} finds a symbol file, it scans the symbols in the
17645file in order to construct an internal symbol table. This lets most
17646@value{GDBN} operations work quickly---at the cost of a delay early
17647on. For large programs, this delay can be quite lengthy, so
17648@value{GDBN} provides a way to build an index, which speeds up
17649startup.
17650
17651The index is stored as a section in the symbol file. @value{GDBN} can
17652write the index to a file, then you can put it into the symbol file
17653using @command{objcopy}.
17654
17655To create an index file, use the @code{save gdb-index} command:
17656
17657@table @code
17658@item save gdb-index @var{directory}
17659@kindex save gdb-index
17660Create an index file for each symbol file currently known by
17661@value{GDBN}. Each file is named after its corresponding symbol file,
17662with @samp{.gdb-index} appended, and is written into the given
17663@var{directory}.
17664@end table
17665
17666Once you have created an index file you can merge it into your symbol
17667file, here named @file{symfile}, using @command{objcopy}:
17668
17669@smallexample
17670$ objcopy --add-section .gdb_index=symfile.gdb-index \
17671 --set-section-flags .gdb_index=readonly symfile symfile
17672@end smallexample
17673
17674@value{GDBN} will normally ignore older versions of @file{.gdb_index}
17675sections that have been deprecated. Usually they are deprecated because
17676they are missing a new feature or have performance issues.
17677To tell @value{GDBN} to use a deprecated index section anyway
17678specify @code{set use-deprecated-index-sections on}.
17679The default is @code{off}.
17680This can speed up startup, but may result in some functionality being lost.
17681@xref{Index Section Format}.
17682
17683@emph{Warning:} Setting @code{use-deprecated-index-sections} to @code{on}
17684must be done before gdb reads the file. The following will not work:
17685
17686@smallexample
17687$ gdb -ex "set use-deprecated-index-sections on" <program>
17688@end smallexample
17689
17690Instead you must do, for example,
17691
17692@smallexample
17693$ gdb -iex "set use-deprecated-index-sections on" <program>
17694@end smallexample
17695
17696There are currently some limitation on indices. They only work when
17697for DWARF debugging information, not stabs. And, they do not
17698currently work for programs using Ada.
17699
17700@node Symbol Errors
17701@section Errors Reading Symbol Files
17702
17703While reading a symbol file, @value{GDBN} occasionally encounters problems,
17704such as symbol types it does not recognize, or known bugs in compiler
17705output. By default, @value{GDBN} does not notify you of such problems, since
17706they are relatively common and primarily of interest to people
17707debugging compilers. If you are interested in seeing information
17708about ill-constructed symbol tables, you can either ask @value{GDBN} to print
17709only one message about each such type of problem, no matter how many
17710times the problem occurs; or you can ask @value{GDBN} to print more messages,
17711to see how many times the problems occur, with the @code{set
17712complaints} command (@pxref{Messages/Warnings, ,Optional Warnings and
17713Messages}).
17714
17715The messages currently printed, and their meanings, include:
17716
17717@table @code
17718@item inner block not inside outer block in @var{symbol}
17719
17720The symbol information shows where symbol scopes begin and end
17721(such as at the start of a function or a block of statements). This
17722error indicates that an inner scope block is not fully contained
17723in its outer scope blocks.
17724
17725@value{GDBN} circumvents the problem by treating the inner block as if it had
17726the same scope as the outer block. In the error message, @var{symbol}
17727may be shown as ``@code{(don't know)}'' if the outer block is not a
17728function.
17729
17730@item block at @var{address} out of order
17731
17732The symbol information for symbol scope blocks should occur in
17733order of increasing addresses. This error indicates that it does not
17734do so.
17735
17736@value{GDBN} does not circumvent this problem, and has trouble
17737locating symbols in the source file whose symbols it is reading. (You
17738can often determine what source file is affected by specifying
17739@code{set verbose on}. @xref{Messages/Warnings, ,Optional Warnings and
17740Messages}.)
17741
17742@item bad block start address patched
17743
17744The symbol information for a symbol scope block has a start address
17745smaller than the address of the preceding source line. This is known
17746to occur in the SunOS 4.1.1 (and earlier) C compiler.
17747
17748@value{GDBN} circumvents the problem by treating the symbol scope block as
17749starting on the previous source line.
17750
17751@item bad string table offset in symbol @var{n}
17752
17753@cindex foo
17754Symbol number @var{n} contains a pointer into the string table which is
17755larger than the size of the string table.
17756
17757@value{GDBN} circumvents the problem by considering the symbol to have the
17758name @code{foo}, which may cause other problems if many symbols end up
17759with this name.
17760
17761@item unknown symbol type @code{0x@var{nn}}
17762
17763The symbol information contains new data types that @value{GDBN} does
17764not yet know how to read. @code{0x@var{nn}} is the symbol type of the
17765uncomprehended information, in hexadecimal.
17766
17767@value{GDBN} circumvents the error by ignoring this symbol information.
17768This usually allows you to debug your program, though certain symbols
17769are not accessible. If you encounter such a problem and feel like
17770debugging it, you can debug @code{@value{GDBP}} with itself, breakpoint
17771on @code{complain}, then go up to the function @code{read_dbx_symtab}
17772and examine @code{*bufp} to see the symbol.
17773
17774@item stub type has NULL name
17775
17776@value{GDBN} could not find the full definition for a struct or class.
17777
17778@item const/volatile indicator missing (ok if using g++ v1.x), got@dots{}
17779The symbol information for a C@t{++} member function is missing some
17780information that recent versions of the compiler should have output for
17781it.
17782
17783@item info mismatch between compiler and debugger
17784
17785@value{GDBN} could not parse a type specification output by the compiler.
17786
17787@end table
17788
17789@node Data Files
17790@section GDB Data Files
17791
17792@cindex prefix for data files
17793@value{GDBN} will sometimes read an auxiliary data file. These files
17794are kept in a directory known as the @dfn{data directory}.
17795
17796You can set the data directory's name, and view the name @value{GDBN}
17797is currently using.
17798
17799@table @code
17800@kindex set data-directory
17801@item set data-directory @var{directory}
17802Set the directory which @value{GDBN} searches for auxiliary data files
17803to @var{directory}.
17804
17805@kindex show data-directory
17806@item show data-directory
17807Show the directory @value{GDBN} searches for auxiliary data files.
17808@end table
17809
17810@cindex default data directory
17811@cindex @samp{--with-gdb-datadir}
17812You can set the default data directory by using the configure-time
17813@samp{--with-gdb-datadir} option. If the data directory is inside
17814@value{GDBN}'s configured binary prefix (set with @samp{--prefix} or
17815@samp{--exec-prefix}), then the default data directory will be updated
17816automatically if the installed @value{GDBN} is moved to a new
17817location.
17818
17819The data directory may also be specified with the
17820@code{--data-directory} command line option.
17821@xref{Mode Options}.
17822
17823@node Targets
17824@chapter Specifying a Debugging Target
17825
17826@cindex debugging target
17827A @dfn{target} is the execution environment occupied by your program.
17828
17829Often, @value{GDBN} runs in the same host environment as your program;
17830in that case, the debugging target is specified as a side effect when
17831you use the @code{file} or @code{core} commands. When you need more
17832flexibility---for example, running @value{GDBN} on a physically separate
17833host, or controlling a standalone system over a serial port or a
17834realtime system over a TCP/IP connection---you can use the @code{target}
17835command to specify one of the target types configured for @value{GDBN}
17836(@pxref{Target Commands, ,Commands for Managing Targets}).
17837
17838@cindex target architecture
17839It is possible to build @value{GDBN} for several different @dfn{target
17840architectures}. When @value{GDBN} is built like that, you can choose
17841one of the available architectures with the @kbd{set architecture}
17842command.
17843
17844@table @code
17845@kindex set architecture
17846@kindex show architecture
17847@item set architecture @var{arch}
17848This command sets the current target architecture to @var{arch}. The
17849value of @var{arch} can be @code{"auto"}, in addition to one of the
17850supported architectures.
17851
17852@item show architecture
17853Show the current target architecture.
17854
17855@item set processor
17856@itemx processor
17857@kindex set processor
17858@kindex show processor
17859These are alias commands for, respectively, @code{set architecture}
17860and @code{show architecture}.
17861@end table
17862
17863@menu
17864* Active Targets:: Active targets
17865* Target Commands:: Commands for managing targets
17866* Byte Order:: Choosing target byte order
17867@end menu
17868
17869@node Active Targets
17870@section Active Targets
17871
17872@cindex stacking targets
17873@cindex active targets
17874@cindex multiple targets
17875
17876There are multiple classes of targets such as: processes, executable files or
17877recording sessions. Core files belong to the process class, making core file
17878and process mutually exclusive. Otherwise, @value{GDBN} can work concurrently
17879on multiple active targets, one in each class. This allows you to (for
17880example) start a process and inspect its activity, while still having access to
17881the executable file after the process finishes. Or if you start process
17882recording (@pxref{Reverse Execution}) and @code{reverse-step} there, you are
17883presented a virtual layer of the recording target, while the process target
17884remains stopped at the chronologically last point of the process execution.
17885
17886Use the @code{core-file} and @code{exec-file} commands to select a new core
17887file or executable target (@pxref{Files, ,Commands to Specify Files}). To
17888specify as a target a process that is already running, use the @code{attach}
17889command (@pxref{Attach, ,Debugging an Already-running Process}).
17890
17891@node Target Commands
17892@section Commands for Managing Targets
17893
17894@table @code
17895@item target @var{type} @var{parameters}
17896Connects the @value{GDBN} host environment to a target machine or
17897process. A target is typically a protocol for talking to debugging
17898facilities. You use the argument @var{type} to specify the type or
17899protocol of the target machine.
17900
17901Further @var{parameters} are interpreted by the target protocol, but
17902typically include things like device names or host names to connect
17903with, process numbers, and baud rates.
17904
17905The @code{target} command does not repeat if you press @key{RET} again
17906after executing the command.
17907
17908@kindex help target
17909@item help target
17910Displays the names of all targets available. To display targets
17911currently selected, use either @code{info target} or @code{info files}
17912(@pxref{Files, ,Commands to Specify Files}).
17913
17914@item help target @var{name}
17915Describe a particular target, including any parameters necessary to
17916select it.
17917
17918@kindex set gnutarget
17919@item set gnutarget @var{args}
17920@value{GDBN} uses its own library BFD to read your files. @value{GDBN}
17921knows whether it is reading an @dfn{executable},
17922a @dfn{core}, or a @dfn{.o} file; however, you can specify the file format
17923with the @code{set gnutarget} command. Unlike most @code{target} commands,
17924with @code{gnutarget} the @code{target} refers to a program, not a machine.
17925
17926@quotation
17927@emph{Warning:} To specify a file format with @code{set gnutarget},
17928you must know the actual BFD name.
17929@end quotation
17930
17931@noindent
17932@xref{Files, , Commands to Specify Files}.
17933
17934@kindex show gnutarget
17935@item show gnutarget
17936Use the @code{show gnutarget} command to display what file format
17937@code{gnutarget} is set to read. If you have not set @code{gnutarget},
17938@value{GDBN} will determine the file format for each file automatically,
17939and @code{show gnutarget} displays @samp{The current BFD target is "auto"}.
17940@end table
17941
17942@cindex common targets
17943Here are some common targets (available, or not, depending on the GDB
17944configuration):
17945
17946@table @code
17947@kindex target
17948@item target exec @var{program}
17949@cindex executable file target
17950An executable file. @samp{target exec @var{program}} is the same as
17951@samp{exec-file @var{program}}.
17952
17953@item target core @var{filename}
17954@cindex core dump file target
17955A core dump file. @samp{target core @var{filename}} is the same as
17956@samp{core-file @var{filename}}.
17957
17958@item target remote @var{medium}
17959@cindex remote target
17960A remote system connected to @value{GDBN} via a serial line or network
17961connection. This command tells @value{GDBN} to use its own remote
17962protocol over @var{medium} for debugging. @xref{Remote Debugging}.
17963
17964For example, if you have a board connected to @file{/dev/ttya} on the
17965machine running @value{GDBN}, you could say:
17966
17967@smallexample
17968target remote /dev/ttya
17969@end smallexample
17970
17971@code{target remote} supports the @code{load} command. This is only
17972useful if you have some other way of getting the stub to the target
17973system, and you can put it somewhere in memory where it won't get
17974clobbered by the download.
17975
17976@item target sim @r{[}@var{simargs}@r{]} @dots{}
17977@cindex built-in simulator target
17978Builtin CPU simulator. @value{GDBN} includes simulators for most architectures.
17979In general,
17980@smallexample
17981 target sim
17982 load
17983 run
17984@end smallexample
17985@noindent
17986works; however, you cannot assume that a specific memory map, device
17987drivers, or even basic I/O is available, although some simulators do
17988provide these. For info about any processor-specific simulator details,
17989see the appropriate section in @ref{Embedded Processors, ,Embedded
17990Processors}.
17991
17992@end table
17993
17994Different targets are available on different configurations of @value{GDBN};
17995your configuration may have more or fewer targets.
17996
17997Many remote targets require you to download the executable's code once
17998you've successfully established a connection. You may wish to control
17999various aspects of this process.
18000
18001@table @code
18002
18003@item set hash
18004@kindex set hash@r{, for remote monitors}
18005@cindex hash mark while downloading
18006This command controls whether a hash mark @samp{#} is displayed while
18007downloading a file to the remote monitor. If on, a hash mark is
18008displayed after each S-record is successfully downloaded to the
18009monitor.
18010
18011@item show hash
18012@kindex show hash@r{, for remote monitors}
18013Show the current status of displaying the hash mark.
18014
18015@item set debug monitor
18016@kindex set debug monitor
18017@cindex display remote monitor communications
18018Enable or disable display of communications messages between
18019@value{GDBN} and the remote monitor.
18020
18021@item show debug monitor
18022@kindex show debug monitor
18023Show the current status of displaying communications between
18024@value{GDBN} and the remote monitor.
18025@end table
18026
18027@table @code
18028
18029@kindex load @var{filename}
18030@item load @var{filename}
18031@anchor{load}
18032Depending on what remote debugging facilities are configured into
18033@value{GDBN}, the @code{load} command may be available. Where it exists, it
18034is meant to make @var{filename} (an executable) available for debugging
18035on the remote system---by downloading, or dynamic linking, for example.
18036@code{load} also records the @var{filename} symbol table in @value{GDBN}, like
18037the @code{add-symbol-file} command.
18038
18039If your @value{GDBN} does not have a @code{load} command, attempting to
18040execute it gets the error message ``@code{You can't do that when your
18041target is @dots{}}''
18042
18043The file is loaded at whatever address is specified in the executable.
18044For some object file formats, you can specify the load address when you
18045link the program; for other formats, like a.out, the object file format
18046specifies a fixed address.
18047@c FIXME! This would be a good place for an xref to the GNU linker doc.
18048
18049Depending on the remote side capabilities, @value{GDBN} may be able to
18050load programs into flash memory.
18051
18052@code{load} does not repeat if you press @key{RET} again after using it.
18053@end table
18054
18055@node Byte Order
18056@section Choosing Target Byte Order
18057
18058@cindex choosing target byte order
18059@cindex target byte order
18060
18061Some types of processors, such as the @acronym{MIPS}, PowerPC, and Renesas SH,
18062offer the ability to run either big-endian or little-endian byte
18063orders. Usually the executable or symbol will include a bit to
18064designate the endian-ness, and you will not need to worry about
18065which to use. However, you may still find it useful to adjust
18066@value{GDBN}'s idea of processor endian-ness manually.
18067
18068@table @code
18069@kindex set endian
18070@item set endian big
18071Instruct @value{GDBN} to assume the target is big-endian.
18072
18073@item set endian little
18074Instruct @value{GDBN} to assume the target is little-endian.
18075
18076@item set endian auto
18077Instruct @value{GDBN} to use the byte order associated with the
18078executable.
18079
18080@item show endian
18081Display @value{GDBN}'s current idea of the target byte order.
18082
18083@end table
18084
18085Note that these commands merely adjust interpretation of symbolic
18086data on the host, and that they have absolutely no effect on the
18087target system.
18088
18089
18090@node Remote Debugging
18091@chapter Debugging Remote Programs
18092@cindex remote debugging
18093
18094If you are trying to debug a program running on a machine that cannot run
18095@value{GDBN} in the usual way, it is often useful to use remote debugging.
18096For example, you might use remote debugging on an operating system kernel,
18097or on a small system which does not have a general purpose operating system
18098powerful enough to run a full-featured debugger.
18099
18100Some configurations of @value{GDBN} have special serial or TCP/IP interfaces
18101to make this work with particular debugging targets. In addition,
18102@value{GDBN} comes with a generic serial protocol (specific to @value{GDBN},
18103but not specific to any particular target system) which you can use if you
18104write the remote stubs---the code that runs on the remote system to
18105communicate with @value{GDBN}.
18106
18107Other remote targets may be available in your
18108configuration of @value{GDBN}; use @code{help target} to list them.
18109
18110@menu
18111* Connecting:: Connecting to a remote target
18112* File Transfer:: Sending files to a remote system
18113* Server:: Using the gdbserver program
18114* Remote Configuration:: Remote configuration
18115* Remote Stub:: Implementing a remote stub
18116@end menu
18117
18118@node Connecting
18119@section Connecting to a Remote Target
18120
18121On the @value{GDBN} host machine, you will need an unstripped copy of
18122your program, since @value{GDBN} needs symbol and debugging information.
18123Start up @value{GDBN} as usual, using the name of the local copy of your
18124program as the first argument.
18125
18126@cindex @code{target remote}
18127@value{GDBN} can communicate with the target over a serial line, or
18128over an @acronym{IP} network using @acronym{TCP} or @acronym{UDP}. In
18129each case, @value{GDBN} uses the same protocol for debugging your
18130program; only the medium carrying the debugging packets varies. The
18131@code{target remote} command establishes a connection to the target.
18132Its arguments indicate which medium to use:
18133
18134@table @code
18135
18136@item target remote @var{serial-device}
18137@cindex serial line, @code{target remote}
18138Use @var{serial-device} to communicate with the target. For example,
18139to use a serial line connected to the device named @file{/dev/ttyb}:
18140
18141@smallexample
18142target remote /dev/ttyb
18143@end smallexample
18144
18145If you're using a serial line, you may want to give @value{GDBN} the
18146@samp{--baud} option, or use the @code{set serial baud} command
18147(@pxref{Remote Configuration, set serial baud}) before the
18148@code{target} command.
18149
18150@item target remote @code{@var{host}:@var{port}}
18151@itemx target remote @code{tcp:@var{host}:@var{port}}
18152@cindex @acronym{TCP} port, @code{target remote}
18153Debug using a @acronym{TCP} connection to @var{port} on @var{host}.
18154The @var{host} may be either a host name or a numeric @acronym{IP}
18155address; @var{port} must be a decimal number. The @var{host} could be
18156the target machine itself, if it is directly connected to the net, or
18157it might be a terminal server which in turn has a serial line to the
18158target.
18159
18160For example, to connect to port 2828 on a terminal server named
18161@code{manyfarms}:
18162
18163@smallexample
18164target remote manyfarms:2828
18165@end smallexample
18166
18167If your remote target is actually running on the same machine as your
18168debugger session (e.g.@: a simulator for your target running on the
18169same host), you can omit the hostname. For example, to connect to
18170port 1234 on your local machine:
18171
18172@smallexample
18173target remote :1234
18174@end smallexample
18175@noindent
18176
18177Note that the colon is still required here.
18178
18179@item target remote @code{udp:@var{host}:@var{port}}
18180@cindex @acronym{UDP} port, @code{target remote}
18181Debug using @acronym{UDP} packets to @var{port} on @var{host}. For example, to
18182connect to @acronym{UDP} port 2828 on a terminal server named @code{manyfarms}:
18183
18184@smallexample
18185target remote udp:manyfarms:2828
18186@end smallexample
18187
18188When using a @acronym{UDP} connection for remote debugging, you should
18189keep in mind that the `U' stands for ``Unreliable''. @acronym{UDP}
18190can silently drop packets on busy or unreliable networks, which will
18191cause havoc with your debugging session.
18192
18193@item target remote | @var{command}
18194@cindex pipe, @code{target remote} to
18195Run @var{command} in the background and communicate with it using a
18196pipe. The @var{command} is a shell command, to be parsed and expanded
18197by the system's command shell, @code{/bin/sh}; it should expect remote
18198protocol packets on its standard input, and send replies on its
18199standard output. You could use this to run a stand-alone simulator
18200that speaks the remote debugging protocol, to make net connections
18201using programs like @code{ssh}, or for other similar tricks.
18202
18203If @var{command} closes its standard output (perhaps by exiting),
18204@value{GDBN} will try to send it a @code{SIGTERM} signal. (If the
18205program has already exited, this will have no effect.)
18206
18207@end table
18208
18209Once the connection has been established, you can use all the usual
18210commands to examine and change data. The remote program is already
18211running; you can use @kbd{step} and @kbd{continue}, and you do not
18212need to use @kbd{run}.
18213
18214@cindex interrupting remote programs
18215@cindex remote programs, interrupting
18216Whenever @value{GDBN} is waiting for the remote program, if you type the
18217interrupt character (often @kbd{Ctrl-c}), @value{GDBN} attempts to stop the
18218program. This may or may not succeed, depending in part on the hardware
18219and the serial drivers the remote system uses. If you type the
18220interrupt character once again, @value{GDBN} displays this prompt:
18221
18222@smallexample
18223Interrupted while waiting for the program.
18224Give up (and stop debugging it)? (y or n)
18225@end smallexample
18226
18227If you type @kbd{y}, @value{GDBN} abandons the remote debugging session.
18228(If you decide you want to try again later, you can use @samp{target
18229remote} again to connect once more.) If you type @kbd{n}, @value{GDBN}
18230goes back to waiting.
18231
18232@table @code
18233@kindex detach (remote)
18234@item detach
18235When you have finished debugging the remote program, you can use the
18236@code{detach} command to release it from @value{GDBN} control.
18237Detaching from the target normally resumes its execution, but the results
18238will depend on your particular remote stub. After the @code{detach}
18239command, @value{GDBN} is free to connect to another target.
18240
18241@kindex disconnect
18242@item disconnect
18243The @code{disconnect} command behaves like @code{detach}, except that
18244the target is generally not resumed. It will wait for @value{GDBN}
18245(this instance or another one) to connect and continue debugging. After
18246the @code{disconnect} command, @value{GDBN} is again free to connect to
18247another target.
18248
18249@cindex send command to remote monitor
18250@cindex extend @value{GDBN} for remote targets
18251@cindex add new commands for external monitor
18252@kindex monitor
18253@item monitor @var{cmd}
18254This command allows you to send arbitrary commands directly to the
18255remote monitor. Since @value{GDBN} doesn't care about the commands it
18256sends like this, this command is the way to extend @value{GDBN}---you
18257can add new commands that only the external monitor will understand
18258and implement.
18259@end table
18260
18261@node File Transfer
18262@section Sending files to a remote system
18263@cindex remote target, file transfer
18264@cindex file transfer
18265@cindex sending files to remote systems
18266
18267Some remote targets offer the ability to transfer files over the same
18268connection used to communicate with @value{GDBN}. This is convenient
18269for targets accessible through other means, e.g.@: @sc{gnu}/Linux systems
18270running @code{gdbserver} over a network interface. For other targets,
18271e.g.@: embedded devices with only a single serial port, this may be
18272the only way to upload or download files.
18273
18274Not all remote targets support these commands.
18275
18276@table @code
18277@kindex remote put
18278@item remote put @var{hostfile} @var{targetfile}
18279Copy file @var{hostfile} from the host system (the machine running
18280@value{GDBN}) to @var{targetfile} on the target system.
18281
18282@kindex remote get
18283@item remote get @var{targetfile} @var{hostfile}
18284Copy file @var{targetfile} from the target system to @var{hostfile}
18285on the host system.
18286
18287@kindex remote delete
18288@item remote delete @var{targetfile}
18289Delete @var{targetfile} from the target system.
18290
18291@end table
18292
18293@node Server
18294@section Using the @code{gdbserver} Program
18295
18296@kindex gdbserver
18297@cindex remote connection without stubs
18298@code{gdbserver} is a control program for Unix-like systems, which
18299allows you to connect your program with a remote @value{GDBN} via
18300@code{target remote}---but without linking in the usual debugging stub.
18301
18302@code{gdbserver} is not a complete replacement for the debugging stubs,
18303because it requires essentially the same operating-system facilities
18304that @value{GDBN} itself does. In fact, a system that can run
18305@code{gdbserver} to connect to a remote @value{GDBN} could also run
18306@value{GDBN} locally! @code{gdbserver} is sometimes useful nevertheless,
18307because it is a much smaller program than @value{GDBN} itself. It is
18308also easier to port than all of @value{GDBN}, so you may be able to get
18309started more quickly on a new system by using @code{gdbserver}.
18310Finally, if you develop code for real-time systems, you may find that
18311the tradeoffs involved in real-time operation make it more convenient to
18312do as much development work as possible on another system, for example
18313by cross-compiling. You can use @code{gdbserver} to make a similar
18314choice for debugging.
18315
18316@value{GDBN} and @code{gdbserver} communicate via either a serial line
18317or a TCP connection, using the standard @value{GDBN} remote serial
18318protocol.
18319
18320@quotation
18321@emph{Warning:} @code{gdbserver} does not have any built-in security.
18322Do not run @code{gdbserver} connected to any public network; a
18323@value{GDBN} connection to @code{gdbserver} provides access to the
18324target system with the same privileges as the user running
18325@code{gdbserver}.
18326@end quotation
18327
18328@subsection Running @code{gdbserver}
18329@cindex arguments, to @code{gdbserver}
18330@cindex @code{gdbserver}, command-line arguments
18331
18332Run @code{gdbserver} on the target system. You need a copy of the
18333program you want to debug, including any libraries it requires.
18334@code{gdbserver} does not need your program's symbol table, so you can
18335strip the program if necessary to save space. @value{GDBN} on the host
18336system does all the symbol handling.
18337
18338To use the server, you must tell it how to communicate with @value{GDBN};
18339the name of your program; and the arguments for your program. The usual
18340syntax is:
18341
18342@smallexample
18343target> gdbserver @var{comm} @var{program} [ @var{args} @dots{} ]
18344@end smallexample
18345
18346@var{comm} is either a device name (to use a serial line), or a TCP
18347hostname and portnumber, or @code{-} or @code{stdio} to use
18348stdin/stdout of @code{gdbserver}.
18349For example, to debug Emacs with the argument
18350@samp{foo.txt} and communicate with @value{GDBN} over the serial port
18351@file{/dev/com1}:
18352
18353@smallexample
18354target> gdbserver /dev/com1 emacs foo.txt
18355@end smallexample
18356
18357@code{gdbserver} waits passively for the host @value{GDBN} to communicate
18358with it.
18359
18360To use a TCP connection instead of a serial line:
18361
18362@smallexample
18363target> gdbserver host:2345 emacs foo.txt
18364@end smallexample
18365
18366The only difference from the previous example is the first argument,
18367specifying that you are communicating with the host @value{GDBN} via
18368TCP. The @samp{host:2345} argument means that @code{gdbserver} is to
18369expect a TCP connection from machine @samp{host} to local TCP port 2345.
18370(Currently, the @samp{host} part is ignored.) You can choose any number
18371you want for the port number as long as it does not conflict with any
18372TCP ports already in use on the target system (for example, @code{23} is
18373reserved for @code{telnet}).@footnote{If you choose a port number that
18374conflicts with another service, @code{gdbserver} prints an error message
18375and exits.} You must use the same port number with the host @value{GDBN}
18376@code{target remote} command.
18377
18378The @code{stdio} connection is useful when starting @code{gdbserver}
18379with ssh:
18380
18381@smallexample
18382(gdb) target remote | ssh -T hostname gdbserver - hello
18383@end smallexample
18384
18385The @samp{-T} option to ssh is provided because we don't need a remote pty,
18386and we don't want escape-character handling. Ssh does this by default when
18387a command is provided, the flag is provided to make it explicit.
18388You could elide it if you want to.
18389
18390Programs started with stdio-connected gdbserver have @file{/dev/null} for
18391@code{stdin}, and @code{stdout},@code{stderr} are sent back to gdb for
18392display through a pipe connected to gdbserver.
18393Both @code{stdout} and @code{stderr} use the same pipe.
18394
18395@subsubsection Attaching to a Running Program
18396@cindex attach to a program, @code{gdbserver}
18397@cindex @option{--attach}, @code{gdbserver} option
18398
18399On some targets, @code{gdbserver} can also attach to running programs.
18400This is accomplished via the @code{--attach} argument. The syntax is:
18401
18402@smallexample
18403target> gdbserver --attach @var{comm} @var{pid}
18404@end smallexample
18405
18406@var{pid} is the process ID of a currently running process. It isn't necessary
18407to point @code{gdbserver} at a binary for the running process.
18408
18409@pindex pidof
18410You can debug processes by name instead of process ID if your target has the
18411@code{pidof} utility:
18412
18413@smallexample
18414target> gdbserver --attach @var{comm} `pidof @var{program}`
18415@end smallexample
18416
18417In case more than one copy of @var{program} is running, or @var{program}
18418has multiple threads, most versions of @code{pidof} support the
18419@code{-s} option to only return the first process ID.
18420
18421@subsubsection Multi-Process Mode for @code{gdbserver}
18422@cindex @code{gdbserver}, multiple processes
18423@cindex multiple processes with @code{gdbserver}
18424
18425When you connect to @code{gdbserver} using @code{target remote},
18426@code{gdbserver} debugs the specified program only once. When the
18427program exits, or you detach from it, @value{GDBN} closes the connection
18428and @code{gdbserver} exits.
18429
18430If you connect using @kbd{target extended-remote}, @code{gdbserver}
18431enters multi-process mode. When the debugged program exits, or you
18432detach from it, @value{GDBN} stays connected to @code{gdbserver} even
18433though no program is running. The @code{run} and @code{attach}
18434commands instruct @code{gdbserver} to run or attach to a new program.
18435The @code{run} command uses @code{set remote exec-file} (@pxref{set
18436remote exec-file}) to select the program to run. Command line
18437arguments are supported, except for wildcard expansion and I/O
18438redirection (@pxref{Arguments}).
18439
18440@cindex @option{--multi}, @code{gdbserver} option
18441To start @code{gdbserver} without supplying an initial command to run
18442or process ID to attach, use the @option{--multi} command line option.
18443Then you can connect using @kbd{target extended-remote} and start
18444the program you want to debug.
18445
18446In multi-process mode @code{gdbserver} does not automatically exit unless you
18447use the option @option{--once}. You can terminate it by using
18448@code{monitor exit} (@pxref{Monitor Commands for gdbserver}). Note that the
18449conditions under which @code{gdbserver} terminates depend on how @value{GDBN}
18450connects to it (@kbd{target remote} or @kbd{target extended-remote}). The
18451@option{--multi} option to @code{gdbserver} has no influence on that.
18452
18453@subsubsection TCP port allocation lifecycle of @code{gdbserver}
18454
18455This section applies only when @code{gdbserver} is run to listen on a TCP port.
18456
18457@code{gdbserver} normally terminates after all of its debugged processes have
18458terminated in @kbd{target remote} mode. On the other hand, for @kbd{target
18459extended-remote}, @code{gdbserver} stays running even with no processes left.
18460@value{GDBN} normally terminates the spawned debugged process on its exit,
18461which normally also terminates @code{gdbserver} in the @kbd{target remote}
18462mode. Therefore, when the connection drops unexpectedly, and @value{GDBN}
18463cannot ask @code{gdbserver} to kill its debugged processes, @code{gdbserver}
18464stays running even in the @kbd{target remote} mode.
18465
18466When @code{gdbserver} stays running, @value{GDBN} can connect to it again later.
18467Such reconnecting is useful for features like @ref{disconnected tracing}. For
18468completeness, at most one @value{GDBN} can be connected at a time.
18469
18470@cindex @option{--once}, @code{gdbserver} option
18471By default, @code{gdbserver} keeps the listening TCP port open, so that
18472subsequent connections are possible. However, if you start @code{gdbserver}
18473with the @option{--once} option, it will stop listening for any further
18474connection attempts after connecting to the first @value{GDBN} session. This
18475means no further connections to @code{gdbserver} will be possible after the
18476first one. It also means @code{gdbserver} will terminate after the first
18477connection with remote @value{GDBN} has closed, even for unexpectedly closed
18478connections and even in the @kbd{target extended-remote} mode. The
18479@option{--once} option allows reusing the same port number for connecting to
18480multiple instances of @code{gdbserver} running on the same host, since each
18481instance closes its port after the first connection.
18482
18483@subsubsection Other Command-Line Arguments for @code{gdbserver}
18484
18485@cindex @option{--debug}, @code{gdbserver} option
18486The @option{--debug} option tells @code{gdbserver} to display extra
18487status information about the debugging process.
18488@cindex @option{--remote-debug}, @code{gdbserver} option
18489The @option{--remote-debug} option tells @code{gdbserver} to display
18490remote protocol debug output. These options are intended for
18491@code{gdbserver} development and for bug reports to the developers.
18492
18493@cindex @option{--wrapper}, @code{gdbserver} option
18494The @option{--wrapper} option specifies a wrapper to launch programs
18495for debugging. The option should be followed by the name of the
18496wrapper, then any command-line arguments to pass to the wrapper, then
18497@kbd{--} indicating the end of the wrapper arguments.
18498
18499@code{gdbserver} runs the specified wrapper program with a combined
18500command line including the wrapper arguments, then the name of the
18501program to debug, then any arguments to the program. The wrapper
18502runs until it executes your program, and then @value{GDBN} gains control.
18503
18504You can use any program that eventually calls @code{execve} with
18505its arguments as a wrapper. Several standard Unix utilities do
18506this, e.g.@: @code{env} and @code{nohup}. Any Unix shell script ending
18507with @code{exec "$@@"} will also work.
18508
18509For example, you can use @code{env} to pass an environment variable to
18510the debugged program, without setting the variable in @code{gdbserver}'s
18511environment:
18512
18513@smallexample
18514$ gdbserver --wrapper env LD_PRELOAD=libtest.so -- :2222 ./testprog
18515@end smallexample
18516
18517@subsection Connecting to @code{gdbserver}
18518
18519Run @value{GDBN} on the host system.
18520
18521First make sure you have the necessary symbol files. Load symbols for
18522your application using the @code{file} command before you connect. Use
18523@code{set sysroot} to locate target libraries (unless your @value{GDBN}
18524was compiled with the correct sysroot using @code{--with-sysroot}).
18525
18526The symbol file and target libraries must exactly match the executable
18527and libraries on the target, with one exception: the files on the host
18528system should not be stripped, even if the files on the target system
18529are. Mismatched or missing files will lead to confusing results
18530during debugging. On @sc{gnu}/Linux targets, mismatched or missing
18531files may also prevent @code{gdbserver} from debugging multi-threaded
18532programs.
18533
18534Connect to your target (@pxref{Connecting,,Connecting to a Remote Target}).
18535For TCP connections, you must start up @code{gdbserver} prior to using
18536the @code{target remote} command. Otherwise you may get an error whose
18537text depends on the host system, but which usually looks something like
18538@samp{Connection refused}. Don't use the @code{load}
18539command in @value{GDBN} when using @code{gdbserver}, since the program is
18540already on the target.
18541
18542@subsection Monitor Commands for @code{gdbserver}
18543@cindex monitor commands, for @code{gdbserver}
18544@anchor{Monitor Commands for gdbserver}
18545
18546During a @value{GDBN} session using @code{gdbserver}, you can use the
18547@code{monitor} command to send special requests to @code{gdbserver}.
18548Here are the available commands.
18549
18550@table @code
18551@item monitor help
18552List the available monitor commands.
18553
18554@item monitor set debug 0
18555@itemx monitor set debug 1
18556Disable or enable general debugging messages.
18557
18558@item monitor set remote-debug 0
18559@itemx monitor set remote-debug 1
18560Disable or enable specific debugging messages associated with the remote
18561protocol (@pxref{Remote Protocol}).
18562
18563@item monitor set libthread-db-search-path [PATH]
18564@cindex gdbserver, search path for @code{libthread_db}
18565When this command is issued, @var{path} is a colon-separated list of
18566directories to search for @code{libthread_db} (@pxref{Threads,,set
18567libthread-db-search-path}). If you omit @var{path},
18568@samp{libthread-db-search-path} will be reset to its default value.
18569
18570The special entry @samp{$pdir} for @samp{libthread-db-search-path} is
18571not supported in @code{gdbserver}.
18572
18573@item monitor exit
18574Tell gdbserver to exit immediately. This command should be followed by
18575@code{disconnect} to close the debugging session. @code{gdbserver} will
18576detach from any attached processes and kill any processes it created.
18577Use @code{monitor exit} to terminate @code{gdbserver} at the end
18578of a multi-process mode debug session.
18579
18580@end table
18581
18582@subsection Tracepoints support in @code{gdbserver}
18583@cindex tracepoints support in @code{gdbserver}
18584
18585On some targets, @code{gdbserver} supports tracepoints, fast
18586tracepoints and static tracepoints.
18587
18588For fast or static tracepoints to work, a special library called the
18589@dfn{in-process agent} (IPA), must be loaded in the inferior process.
18590This library is built and distributed as an integral part of
18591@code{gdbserver}. In addition, support for static tracepoints
18592requires building the in-process agent library with static tracepoints
18593support. At present, the UST (LTTng Userspace Tracer,
18594@url{http://lttng.org/ust}) tracing engine is supported. This support
18595is automatically available if UST development headers are found in the
18596standard include path when @code{gdbserver} is built, or if
18597@code{gdbserver} was explicitly configured using @option{--with-ust}
18598to point at such headers. You can explicitly disable the support
18599using @option{--with-ust=no}.
18600
18601There are several ways to load the in-process agent in your program:
18602
18603@table @code
18604@item Specifying it as dependency at link time
18605
18606You can link your program dynamically with the in-process agent
18607library. On most systems, this is accomplished by adding
18608@code{-linproctrace} to the link command.
18609
18610@item Using the system's preloading mechanisms
18611
18612You can force loading the in-process agent at startup time by using
18613your system's support for preloading shared libraries. Many Unixes
18614support the concept of preloading user defined libraries. In most
18615cases, you do that by specifying @code{LD_PRELOAD=libinproctrace.so}
18616in the environment. See also the description of @code{gdbserver}'s
18617@option{--wrapper} command line option.
18618
18619@item Using @value{GDBN} to force loading the agent at run time
18620
18621On some systems, you can force the inferior to load a shared library,
18622by calling a dynamic loader function in the inferior that takes care
18623of dynamically looking up and loading a shared library. On most Unix
18624systems, the function is @code{dlopen}. You'll use the @code{call}
18625command for that. For example:
18626
18627@smallexample
18628(@value{GDBP}) call dlopen ("libinproctrace.so", ...)
18629@end smallexample
18630
18631Note that on most Unix systems, for the @code{dlopen} function to be
18632available, the program needs to be linked with @code{-ldl}.
18633@end table
18634
18635On systems that have a userspace dynamic loader, like most Unix
18636systems, when you connect to @code{gdbserver} using @code{target
18637remote}, you'll find that the program is stopped at the dynamic
18638loader's entry point, and no shared library has been loaded in the
18639program's address space yet, including the in-process agent. In that
18640case, before being able to use any of the fast or static tracepoints
18641features, you need to let the loader run and load the shared
18642libraries. The simplest way to do that is to run the program to the
18643main procedure. E.g., if debugging a C or C@t{++} program, start
18644@code{gdbserver} like so:
18645
18646@smallexample
18647$ gdbserver :9999 myprogram
18648@end smallexample
18649
18650Start GDB and connect to @code{gdbserver} like so, and run to main:
18651
18652@smallexample
18653$ gdb myprogram
18654(@value{GDBP}) target remote myhost:9999
186550x00007f215893ba60 in ?? () from /lib64/ld-linux-x86-64.so.2
18656(@value{GDBP}) b main
18657(@value{GDBP}) continue
18658@end smallexample
18659
18660The in-process tracing agent library should now be loaded into the
18661process; you can confirm it with the @code{info sharedlibrary}
18662command, which will list @file{libinproctrace.so} as loaded in the
18663process. You are now ready to install fast tracepoints, list static
18664tracepoint markers, probe static tracepoints markers, and start
18665tracing.
18666
18667@node Remote Configuration
18668@section Remote Configuration
18669
18670@kindex set remote
18671@kindex show remote
18672This section documents the configuration options available when
18673debugging remote programs. For the options related to the File I/O
18674extensions of the remote protocol, see @ref{system,
18675system-call-allowed}.
18676
18677@table @code
18678@item set remoteaddresssize @var{bits}
18679@cindex address size for remote targets
18680@cindex bits in remote address
18681Set the maximum size of address in a memory packet to the specified
18682number of bits. @value{GDBN} will mask off the address bits above
18683that number, when it passes addresses to the remote target. The
18684default value is the number of bits in the target's address.
18685
18686@item show remoteaddresssize
18687Show the current value of remote address size in bits.
18688
18689@item set serial baud @var{n}
18690@cindex baud rate for remote targets
18691Set the baud rate for the remote serial I/O to @var{n} baud. The
18692value is used to set the speed of the serial port used for debugging
18693remote targets.
18694
18695@item show serial baud
18696Show the current speed of the remote connection.
18697
18698@item set remotebreak
18699@cindex interrupt remote programs
18700@cindex BREAK signal instead of Ctrl-C
18701@anchor{set remotebreak}
18702If set to on, @value{GDBN} sends a @code{BREAK} signal to the remote
18703when you type @kbd{Ctrl-c} to interrupt the program running
18704on the remote. If set to off, @value{GDBN} sends the @samp{Ctrl-C}
18705character instead. The default is off, since most remote systems
18706expect to see @samp{Ctrl-C} as the interrupt signal.
18707
18708@item show remotebreak
18709Show whether @value{GDBN} sends @code{BREAK} or @samp{Ctrl-C} to
18710interrupt the remote program.
18711
18712@item set remoteflow on
18713@itemx set remoteflow off
18714@kindex set remoteflow
18715Enable or disable hardware flow control (@code{RTS}/@code{CTS})
18716on the serial port used to communicate to the remote target.
18717
18718@item show remoteflow
18719@kindex show remoteflow
18720Show the current setting of hardware flow control.
18721
18722@item set remotelogbase @var{base}
18723Set the base (a.k.a.@: radix) of logging serial protocol
18724communications to @var{base}. Supported values of @var{base} are:
18725@code{ascii}, @code{octal}, and @code{hex}. The default is
18726@code{ascii}.
18727
18728@item show remotelogbase
18729Show the current setting of the radix for logging remote serial
18730protocol.
18731
18732@item set remotelogfile @var{file}
18733@cindex record serial communications on file
18734Record remote serial communications on the named @var{file}. The
18735default is not to record at all.
18736
18737@item show remotelogfile.
18738Show the current setting of the file name on which to record the
18739serial communications.
18740
18741@item set remotetimeout @var{num}
18742@cindex timeout for serial communications
18743@cindex remote timeout
18744Set the timeout limit to wait for the remote target to respond to
18745@var{num} seconds. The default is 2 seconds.
18746
18747@item show remotetimeout
18748Show the current number of seconds to wait for the remote target
18749responses.
18750
18751@cindex limit hardware breakpoints and watchpoints
18752@cindex remote target, limit break- and watchpoints
18753@anchor{set remote hardware-watchpoint-limit}
18754@anchor{set remote hardware-breakpoint-limit}
18755@item set remote hardware-watchpoint-limit @var{limit}
18756@itemx set remote hardware-breakpoint-limit @var{limit}
18757Restrict @value{GDBN} to using @var{limit} remote hardware breakpoint or
18758watchpoints. A limit of -1, the default, is treated as unlimited.
18759
18760@cindex limit hardware watchpoints length
18761@cindex remote target, limit watchpoints length
18762@anchor{set remote hardware-watchpoint-length-limit}
18763@item set remote hardware-watchpoint-length-limit @var{limit}
18764Restrict @value{GDBN} to using @var{limit} bytes for the maximum length of
18765a remote hardware watchpoint. A limit of -1, the default, is treated
18766as unlimited.
18767
18768@item show remote hardware-watchpoint-length-limit
18769Show the current limit (in bytes) of the maximum length of
18770a remote hardware watchpoint.
18771
18772@item set remote exec-file @var{filename}
18773@itemx show remote exec-file
18774@anchor{set remote exec-file}
18775@cindex executable file, for remote target
18776Select the file used for @code{run} with @code{target
18777extended-remote}. This should be set to a filename valid on the
18778target system. If it is not set, the target will use a default
18779filename (e.g.@: the last program run).
18780
18781@item set remote interrupt-sequence
18782@cindex interrupt remote programs
18783@cindex select Ctrl-C, BREAK or BREAK-g
18784Allow the user to select one of @samp{Ctrl-C}, a @code{BREAK} or
18785@samp{BREAK-g} as the
18786sequence to the remote target in order to interrupt the execution.
18787@samp{Ctrl-C} is a default. Some system prefers @code{BREAK} which
18788is high level of serial line for some certain time.
18789Linux kernel prefers @samp{BREAK-g}, a.k.a Magic SysRq g.
18790It is @code{BREAK} signal followed by character @code{g}.
18791
18792@item show interrupt-sequence
18793Show which of @samp{Ctrl-C}, @code{BREAK} or @code{BREAK-g}
18794is sent by @value{GDBN} to interrupt the remote program.
18795@code{BREAK-g} is BREAK signal followed by @code{g} and
18796also known as Magic SysRq g.
18797
18798@item set remote interrupt-on-connect
18799@cindex send interrupt-sequence on start
18800Specify whether interrupt-sequence is sent to remote target when
18801@value{GDBN} connects to it. This is mostly needed when you debug
18802Linux kernel. Linux kernel expects @code{BREAK} followed by @code{g}
18803which is known as Magic SysRq g in order to connect @value{GDBN}.
18804
18805@item show interrupt-on-connect
18806Show whether interrupt-sequence is sent
18807to remote target when @value{GDBN} connects to it.
18808
18809@kindex set tcp
18810@kindex show tcp
18811@item set tcp auto-retry on
18812@cindex auto-retry, for remote TCP target
18813Enable auto-retry for remote TCP connections. This is useful if the remote
18814debugging agent is launched in parallel with @value{GDBN}; there is a race
18815condition because the agent may not become ready to accept the connection
18816before @value{GDBN} attempts to connect. When auto-retry is
18817enabled, if the initial attempt to connect fails, @value{GDBN} reattempts
18818to establish the connection using the timeout specified by
18819@code{set tcp connect-timeout}.
18820
18821@item set tcp auto-retry off
18822Do not auto-retry failed TCP connections.
18823
18824@item show tcp auto-retry
18825Show the current auto-retry setting.
18826
18827@item set tcp connect-timeout @var{seconds}
18828@itemx set tcp connect-timeout unlimited
18829@cindex connection timeout, for remote TCP target
18830@cindex timeout, for remote target connection
18831Set the timeout for establishing a TCP connection to the remote target to
18832@var{seconds}. The timeout affects both polling to retry failed connections
18833(enabled by @code{set tcp auto-retry on}) and waiting for connections
18834that are merely slow to complete, and represents an approximate cumulative
18835value. If @var{seconds} is @code{unlimited}, there is no timeout and
18836@value{GDBN} will keep attempting to establish a connection forever,
18837unless interrupted with @kbd{Ctrl-c}. The default is 15 seconds.
18838
18839@item show tcp connect-timeout
18840Show the current connection timeout setting.
18841@end table
18842
18843@cindex remote packets, enabling and disabling
18844The @value{GDBN} remote protocol autodetects the packets supported by
18845your debugging stub. If you need to override the autodetection, you
18846can use these commands to enable or disable individual packets. Each
18847packet can be set to @samp{on} (the remote target supports this
18848packet), @samp{off} (the remote target does not support this packet),
18849or @samp{auto} (detect remote target support for this packet). They
18850all default to @samp{auto}. For more information about each packet,
18851see @ref{Remote Protocol}.
18852
18853During normal use, you should not have to use any of these commands.
18854If you do, that may be a bug in your remote debugging stub, or a bug
18855in @value{GDBN}. You may want to report the problem to the
18856@value{GDBN} developers.
18857
18858For each packet @var{name}, the command to enable or disable the
18859packet is @code{set remote @var{name}-packet}. The available settings
18860are:
18861
18862@multitable @columnfractions 0.28 0.32 0.25
18863@item Command Name
18864@tab Remote Packet
18865@tab Related Features
18866
18867@item @code{fetch-register}
18868@tab @code{p}
18869@tab @code{info registers}
18870
18871@item @code{set-register}
18872@tab @code{P}
18873@tab @code{set}
18874
18875@item @code{binary-download}
18876@tab @code{X}
18877@tab @code{load}, @code{set}
18878
18879@item @code{read-aux-vector}
18880@tab @code{qXfer:auxv:read}
18881@tab @code{info auxv}
18882
18883@item @code{symbol-lookup}
18884@tab @code{qSymbol}
18885@tab Detecting multiple threads
18886
18887@item @code{attach}
18888@tab @code{vAttach}
18889@tab @code{attach}
18890
18891@item @code{verbose-resume}
18892@tab @code{vCont}
18893@tab Stepping or resuming multiple threads
18894
18895@item @code{run}
18896@tab @code{vRun}
18897@tab @code{run}
18898
18899@item @code{software-breakpoint}
18900@tab @code{Z0}
18901@tab @code{break}
18902
18903@item @code{hardware-breakpoint}
18904@tab @code{Z1}
18905@tab @code{hbreak}
18906
18907@item @code{write-watchpoint}
18908@tab @code{Z2}
18909@tab @code{watch}
18910
18911@item @code{read-watchpoint}
18912@tab @code{Z3}
18913@tab @code{rwatch}
18914
18915@item @code{access-watchpoint}
18916@tab @code{Z4}
18917@tab @code{awatch}
18918
18919@item @code{target-features}
18920@tab @code{qXfer:features:read}
18921@tab @code{set architecture}
18922
18923@item @code{library-info}
18924@tab @code{qXfer:libraries:read}
18925@tab @code{info sharedlibrary}
18926
18927@item @code{memory-map}
18928@tab @code{qXfer:memory-map:read}
18929@tab @code{info mem}
18930
18931@item @code{read-sdata-object}
18932@tab @code{qXfer:sdata:read}
18933@tab @code{print $_sdata}
18934
18935@item @code{read-spu-object}
18936@tab @code{qXfer:spu:read}
18937@tab @code{info spu}
18938
18939@item @code{write-spu-object}
18940@tab @code{qXfer:spu:write}
18941@tab @code{info spu}
18942
18943@item @code{read-siginfo-object}
18944@tab @code{qXfer:siginfo:read}
18945@tab @code{print $_siginfo}
18946
18947@item @code{write-siginfo-object}
18948@tab @code{qXfer:siginfo:write}
18949@tab @code{set $_siginfo}
18950
18951@item @code{threads}
18952@tab @code{qXfer:threads:read}
18953@tab @code{info threads}
18954
18955@item @code{get-thread-local-@*storage-address}
18956@tab @code{qGetTLSAddr}
18957@tab Displaying @code{__thread} variables
18958
18959@item @code{get-thread-information-block-address}
18960@tab @code{qGetTIBAddr}
18961@tab Display MS-Windows Thread Information Block.
18962
18963@item @code{search-memory}
18964@tab @code{qSearch:memory}
18965@tab @code{find}
18966
18967@item @code{supported-packets}
18968@tab @code{qSupported}
18969@tab Remote communications parameters
18970
18971@item @code{pass-signals}
18972@tab @code{QPassSignals}
18973@tab @code{handle @var{signal}}
18974
18975@item @code{program-signals}
18976@tab @code{QProgramSignals}
18977@tab @code{handle @var{signal}}
18978
18979@item @code{hostio-close-packet}
18980@tab @code{vFile:close}
18981@tab @code{remote get}, @code{remote put}
18982
18983@item @code{hostio-open-packet}
18984@tab @code{vFile:open}
18985@tab @code{remote get}, @code{remote put}
18986
18987@item @code{hostio-pread-packet}
18988@tab @code{vFile:pread}
18989@tab @code{remote get}, @code{remote put}
18990
18991@item @code{hostio-pwrite-packet}
18992@tab @code{vFile:pwrite}
18993@tab @code{remote get}, @code{remote put}
18994
18995@item @code{hostio-unlink-packet}
18996@tab @code{vFile:unlink}
18997@tab @code{remote delete}
18998
18999@item @code{hostio-readlink-packet}
19000@tab @code{vFile:readlink}
19001@tab Host I/O
19002
19003@item @code{noack-packet}
19004@tab @code{QStartNoAckMode}
19005@tab Packet acknowledgment
19006
19007@item @code{osdata}
19008@tab @code{qXfer:osdata:read}
19009@tab @code{info os}
19010
19011@item @code{query-attached}
19012@tab @code{qAttached}
19013@tab Querying remote process attach state.
19014
19015@item @code{trace-buffer-size}
19016@tab @code{QTBuffer:size}
19017@tab @code{set trace-buffer-size}
19018
19019@item @code{trace-status}
19020@tab @code{qTStatus}
19021@tab @code{tstatus}
19022
19023@item @code{traceframe-info}
19024@tab @code{qXfer:traceframe-info:read}
19025@tab Traceframe info
19026
19027@item @code{install-in-trace}
19028@tab @code{InstallInTrace}
19029@tab Install tracepoint in tracing
19030
19031@item @code{disable-randomization}
19032@tab @code{QDisableRandomization}
19033@tab @code{set disable-randomization}
19034
19035@item @code{conditional-breakpoints-packet}
19036@tab @code{Z0 and Z1}
19037@tab @code{Support for target-side breakpoint condition evaluation}
19038@end multitable
19039
19040@node Remote Stub
19041@section Implementing a Remote Stub
19042
19043@cindex debugging stub, example
19044@cindex remote stub, example
19045@cindex stub example, remote debugging
19046The stub files provided with @value{GDBN} implement the target side of the
19047communication protocol, and the @value{GDBN} side is implemented in the
19048@value{GDBN} source file @file{remote.c}. Normally, you can simply allow
19049these subroutines to communicate, and ignore the details. (If you're
19050implementing your own stub file, you can still ignore the details: start
19051with one of the existing stub files. @file{sparc-stub.c} is the best
19052organized, and therefore the easiest to read.)
19053
19054@cindex remote serial debugging, overview
19055To debug a program running on another machine (the debugging
19056@dfn{target} machine), you must first arrange for all the usual
19057prerequisites for the program to run by itself. For example, for a C
19058program, you need:
19059
19060@enumerate
19061@item
19062A startup routine to set up the C runtime environment; these usually
19063have a name like @file{crt0}. The startup routine may be supplied by
19064your hardware supplier, or you may have to write your own.
19065
19066@item
19067A C subroutine library to support your program's
19068subroutine calls, notably managing input and output.
19069
19070@item
19071A way of getting your program to the other machine---for example, a
19072download program. These are often supplied by the hardware
19073manufacturer, but you may have to write your own from hardware
19074documentation.
19075@end enumerate
19076
19077The next step is to arrange for your program to use a serial port to
19078communicate with the machine where @value{GDBN} is running (the @dfn{host}
19079machine). In general terms, the scheme looks like this:
19080
19081@table @emph
19082@item On the host,
19083@value{GDBN} already understands how to use this protocol; when everything
19084else is set up, you can simply use the @samp{target remote} command
19085(@pxref{Targets,,Specifying a Debugging Target}).
19086
19087@item On the target,
19088you must link with your program a few special-purpose subroutines that
19089implement the @value{GDBN} remote serial protocol. The file containing these
19090subroutines is called a @dfn{debugging stub}.
19091
19092On certain remote targets, you can use an auxiliary program
19093@code{gdbserver} instead of linking a stub into your program.
19094@xref{Server,,Using the @code{gdbserver} Program}, for details.
19095@end table
19096
19097The debugging stub is specific to the architecture of the remote
19098machine; for example, use @file{sparc-stub.c} to debug programs on
19099@sc{sparc} boards.
19100
19101@cindex remote serial stub list
19102These working remote stubs are distributed with @value{GDBN}:
19103
19104@table @code
19105
19106@item i386-stub.c
19107@cindex @file{i386-stub.c}
19108@cindex Intel
19109@cindex i386
19110For Intel 386 and compatible architectures.
19111
19112@item m68k-stub.c
19113@cindex @file{m68k-stub.c}
19114@cindex Motorola 680x0
19115@cindex m680x0
19116For Motorola 680x0 architectures.
19117
19118@item sh-stub.c
19119@cindex @file{sh-stub.c}
19120@cindex Renesas
19121@cindex SH
19122For Renesas SH architectures.
19123
19124@item sparc-stub.c
19125@cindex @file{sparc-stub.c}
19126@cindex Sparc
19127For @sc{sparc} architectures.
19128
19129@item sparcl-stub.c
19130@cindex @file{sparcl-stub.c}
19131@cindex Fujitsu
19132@cindex SparcLite
19133For Fujitsu @sc{sparclite} architectures.
19134
19135@end table
19136
19137The @file{README} file in the @value{GDBN} distribution may list other
19138recently added stubs.
19139
19140@menu
19141* Stub Contents:: What the stub can do for you
19142* Bootstrapping:: What you must do for the stub
19143* Debug Session:: Putting it all together
19144@end menu
19145
19146@node Stub Contents
19147@subsection What the Stub Can Do for You
19148
19149@cindex remote serial stub
19150The debugging stub for your architecture supplies these three
19151subroutines:
19152
19153@table @code
19154@item set_debug_traps
19155@findex set_debug_traps
19156@cindex remote serial stub, initialization
19157This routine arranges for @code{handle_exception} to run when your
19158program stops. You must call this subroutine explicitly in your
19159program's startup code.
19160
19161@item handle_exception
19162@findex handle_exception
19163@cindex remote serial stub, main routine
19164This is the central workhorse, but your program never calls it
19165explicitly---the setup code arranges for @code{handle_exception} to
19166run when a trap is triggered.
19167
19168@code{handle_exception} takes control when your program stops during
19169execution (for example, on a breakpoint), and mediates communications
19170with @value{GDBN} on the host machine. This is where the communications
19171protocol is implemented; @code{handle_exception} acts as the @value{GDBN}
19172representative on the target machine. It begins by sending summary
19173information on the state of your program, then continues to execute,
19174retrieving and transmitting any information @value{GDBN} needs, until you
19175execute a @value{GDBN} command that makes your program resume; at that point,
19176@code{handle_exception} returns control to your own code on the target
19177machine.
19178
19179@item breakpoint
19180@cindex @code{breakpoint} subroutine, remote
19181Use this auxiliary subroutine to make your program contain a
19182breakpoint. Depending on the particular situation, this may be the only
19183way for @value{GDBN} to get control. For instance, if your target
19184machine has some sort of interrupt button, you won't need to call this;
19185pressing the interrupt button transfers control to
19186@code{handle_exception}---in effect, to @value{GDBN}. On some machines,
19187simply receiving characters on the serial port may also trigger a trap;
19188again, in that situation, you don't need to call @code{breakpoint} from
19189your own program---simply running @samp{target remote} from the host
19190@value{GDBN} session gets control.
19191
19192Call @code{breakpoint} if none of these is true, or if you simply want
19193to make certain your program stops at a predetermined point for the
19194start of your debugging session.
19195@end table
19196
19197@node Bootstrapping
19198@subsection What You Must Do for the Stub
19199
19200@cindex remote stub, support routines
19201The debugging stubs that come with @value{GDBN} are set up for a particular
19202chip architecture, but they have no information about the rest of your
19203debugging target machine.
19204
19205First of all you need to tell the stub how to communicate with the
19206serial port.
19207
19208@table @code
19209@item int getDebugChar()
19210@findex getDebugChar
19211Write this subroutine to read a single character from the serial port.
19212It may be identical to @code{getchar} for your target system; a
19213different name is used to allow you to distinguish the two if you wish.
19214
19215@item void putDebugChar(int)
19216@findex putDebugChar
19217Write this subroutine to write a single character to the serial port.
19218It may be identical to @code{putchar} for your target system; a
19219different name is used to allow you to distinguish the two if you wish.
19220@end table
19221
19222@cindex control C, and remote debugging
19223@cindex interrupting remote targets
19224If you want @value{GDBN} to be able to stop your program while it is
19225running, you need to use an interrupt-driven serial driver, and arrange
19226for it to stop when it receives a @code{^C} (@samp{\003}, the control-C
19227character). That is the character which @value{GDBN} uses to tell the
19228remote system to stop.
19229
19230Getting the debugging target to return the proper status to @value{GDBN}
19231probably requires changes to the standard stub; one quick and dirty way
19232is to just execute a breakpoint instruction (the ``dirty'' part is that
19233@value{GDBN} reports a @code{SIGTRAP} instead of a @code{SIGINT}).
19234
19235Other routines you need to supply are:
19236
19237@table @code
19238@item void exceptionHandler (int @var{exception_number}, void *@var{exception_address})
19239@findex exceptionHandler
19240Write this function to install @var{exception_address} in the exception
19241handling tables. You need to do this because the stub does not have any
19242way of knowing what the exception handling tables on your target system
19243are like (for example, the processor's table might be in @sc{rom},
19244containing entries which point to a table in @sc{ram}).
19245@var{exception_number} is the exception number which should be changed;
19246its meaning is architecture-dependent (for example, different numbers
19247might represent divide by zero, misaligned access, etc). When this
19248exception occurs, control should be transferred directly to
19249@var{exception_address}, and the processor state (stack, registers,
19250and so on) should be just as it is when a processor exception occurs. So if
19251you want to use a jump instruction to reach @var{exception_address}, it
19252should be a simple jump, not a jump to subroutine.
19253
19254For the 386, @var{exception_address} should be installed as an interrupt
19255gate so that interrupts are masked while the handler runs. The gate
19256should be at privilege level 0 (the most privileged level). The
19257@sc{sparc} and 68k stubs are able to mask interrupts themselves without
19258help from @code{exceptionHandler}.
19259
19260@item void flush_i_cache()
19261@findex flush_i_cache
19262On @sc{sparc} and @sc{sparclite} only, write this subroutine to flush the
19263instruction cache, if any, on your target machine. If there is no
19264instruction cache, this subroutine may be a no-op.
19265
19266On target machines that have instruction caches, @value{GDBN} requires this
19267function to make certain that the state of your program is stable.
19268@end table
19269
19270@noindent
19271You must also make sure this library routine is available:
19272
19273@table @code
19274@item void *memset(void *, int, int)
19275@findex memset
19276This is the standard library function @code{memset} that sets an area of
19277memory to a known value. If you have one of the free versions of
19278@code{libc.a}, @code{memset} can be found there; otherwise, you must
19279either obtain it from your hardware manufacturer, or write your own.
19280@end table
19281
19282If you do not use the GNU C compiler, you may need other standard
19283library subroutines as well; this varies from one stub to another,
19284but in general the stubs are likely to use any of the common library
19285subroutines which @code{@value{NGCC}} generates as inline code.
19286
19287
19288@node Debug Session
19289@subsection Putting it All Together
19290
19291@cindex remote serial debugging summary
19292In summary, when your program is ready to debug, you must follow these
19293steps.
19294
19295@enumerate
19296@item
19297Make sure you have defined the supporting low-level routines
19298(@pxref{Bootstrapping,,What You Must Do for the Stub}):
19299@display
19300@code{getDebugChar}, @code{putDebugChar},
19301@code{flush_i_cache}, @code{memset}, @code{exceptionHandler}.
19302@end display
19303
19304@item
19305Insert these lines in your program's startup code, before the main
19306procedure is called:
19307
19308@smallexample
19309set_debug_traps();
19310breakpoint();
19311@end smallexample
19312
19313On some machines, when a breakpoint trap is raised, the hardware
19314automatically makes the PC point to the instruction after the
19315breakpoint. If your machine doesn't do that, you may need to adjust
19316@code{handle_exception} to arrange for it to return to the instruction
19317after the breakpoint on this first invocation, so that your program
19318doesn't keep hitting the initial breakpoint instead of making
19319progress.
19320
19321@item
19322For the 680x0 stub only, you need to provide a variable called
19323@code{exceptionHook}. Normally you just use:
19324
19325@smallexample
19326void (*exceptionHook)() = 0;
19327@end smallexample
19328
19329@noindent
19330but if before calling @code{set_debug_traps}, you set it to point to a
19331function in your program, that function is called when
19332@code{@value{GDBN}} continues after stopping on a trap (for example, bus
19333error). The function indicated by @code{exceptionHook} is called with
19334one parameter: an @code{int} which is the exception number.
19335
19336@item
19337Compile and link together: your program, the @value{GDBN} debugging stub for
19338your target architecture, and the supporting subroutines.
19339
19340@item
19341Make sure you have a serial connection between your target machine and
19342the @value{GDBN} host, and identify the serial port on the host.
19343
19344@item
19345@c The "remote" target now provides a `load' command, so we should
19346@c document that. FIXME.
19347Download your program to your target machine (or get it there by
19348whatever means the manufacturer provides), and start it.
19349
19350@item
19351Start @value{GDBN} on the host, and connect to the target
19352(@pxref{Connecting,,Connecting to a Remote Target}).
19353
19354@end enumerate
19355
19356@node Configurations
19357@chapter Configuration-Specific Information
19358
19359While nearly all @value{GDBN} commands are available for all native and
19360cross versions of the debugger, there are some exceptions. This chapter
19361describes things that are only available in certain configurations.
19362
19363There are three major categories of configurations: native
19364configurations, where the host and target are the same, embedded
19365operating system configurations, which are usually the same for several
19366different processor architectures, and bare embedded processors, which
19367are quite different from each other.
19368
19369@menu
19370* Native::
19371* Embedded OS::
19372* Embedded Processors::
19373* Architectures::
19374@end menu
19375
19376@node Native
19377@section Native
19378
19379This section describes details specific to particular native
19380configurations.
19381
19382@menu
19383* HP-UX:: HP-UX
19384* BSD libkvm Interface:: Debugging BSD kernel memory images
19385* SVR4 Process Information:: SVR4 process information
19386* DJGPP Native:: Features specific to the DJGPP port
19387* Cygwin Native:: Features specific to the Cygwin port
19388* Hurd Native:: Features specific to @sc{gnu} Hurd
19389* Darwin:: Features specific to Darwin
19390@end menu
19391
19392@node HP-UX
19393@subsection HP-UX
19394
19395On HP-UX systems, if you refer to a function or variable name that
19396begins with a dollar sign, @value{GDBN} searches for a user or system
19397name first, before it searches for a convenience variable.
19398
19399
19400@node BSD libkvm Interface
19401@subsection BSD libkvm Interface
19402
19403@cindex libkvm
19404@cindex kernel memory image
19405@cindex kernel crash dump
19406
19407BSD-derived systems (FreeBSD/NetBSD/OpenBSD) have a kernel memory
19408interface that provides a uniform interface for accessing kernel virtual
19409memory images, including live systems and crash dumps. @value{GDBN}
19410uses this interface to allow you to debug live kernels and kernel crash
19411dumps on many native BSD configurations. This is implemented as a
19412special @code{kvm} debugging target. For debugging a live system, load
19413the currently running kernel into @value{GDBN} and connect to the
19414@code{kvm} target:
19415
19416@smallexample
19417(@value{GDBP}) @b{target kvm}
19418@end smallexample
19419
19420For debugging crash dumps, provide the file name of the crash dump as an
19421argument:
19422
19423@smallexample
19424(@value{GDBP}) @b{target kvm /var/crash/bsd.0}
19425@end smallexample
19426
19427Once connected to the @code{kvm} target, the following commands are
19428available:
19429
19430@table @code
19431@kindex kvm
19432@item kvm pcb
19433Set current context from the @dfn{Process Control Block} (PCB) address.
19434
19435@item kvm proc
19436Set current context from proc address. This command isn't available on
19437modern FreeBSD systems.
19438@end table
19439
19440@node SVR4 Process Information
19441@subsection SVR4 Process Information
19442@cindex /proc
19443@cindex examine process image
19444@cindex process info via @file{/proc}
19445
19446Many versions of SVR4 and compatible systems provide a facility called
19447@samp{/proc} that can be used to examine the image of a running
19448process using file-system subroutines.
19449
19450If @value{GDBN} is configured for an operating system with this
19451facility, the command @code{info proc} is available to report
19452information about the process running your program, or about any
19453process running on your system. This includes, as of this writing,
19454@sc{gnu}/Linux, OSF/1 (Digital Unix), Solaris, and Irix, but
19455not HP-UX, for example.
19456
19457This command may also work on core files that were created on a system
19458that has the @samp{/proc} facility.
19459
19460@table @code
19461@kindex info proc
19462@cindex process ID
19463@item info proc
19464@itemx info proc @var{process-id}
19465Summarize available information about any running process. If a
19466process ID is specified by @var{process-id}, display information about
19467that process; otherwise display information about the program being
19468debugged. The summary includes the debugged process ID, the command
19469line used to invoke it, its current working directory, and its
19470executable file's absolute file name.
19471
19472On some systems, @var{process-id} can be of the form
19473@samp{[@var{pid}]/@var{tid}} which specifies a certain thread ID
19474within a process. If the optional @var{pid} part is missing, it means
19475a thread from the process being debugged (the leading @samp{/} still
19476needs to be present, or else @value{GDBN} will interpret the number as
19477a process ID rather than a thread ID).
19478
19479@item info proc cmdline
19480@cindex info proc cmdline
19481Show the original command line of the process. This command is
19482specific to @sc{gnu}/Linux.
19483
19484@item info proc cwd
19485@cindex info proc cwd
19486Show the current working directory of the process. This command is
19487specific to @sc{gnu}/Linux.
19488
19489@item info proc exe
19490@cindex info proc exe
19491Show the name of executable of the process. This command is specific
19492to @sc{gnu}/Linux.
19493
19494@item info proc mappings
19495@cindex memory address space mappings
19496Report the memory address space ranges accessible in the program, with
19497information on whether the process has read, write, or execute access
19498rights to each range. On @sc{gnu}/Linux systems, each memory range
19499includes the object file which is mapped to that range, instead of the
19500memory access rights to that range.
19501
19502@item info proc stat
19503@itemx info proc status
19504@cindex process detailed status information
19505These subcommands are specific to @sc{gnu}/Linux systems. They show
19506the process-related information, including the user ID and group ID;
19507how many threads are there in the process; its virtual memory usage;
19508the signals that are pending, blocked, and ignored; its TTY; its
19509consumption of system and user time; its stack size; its @samp{nice}
19510value; etc. For more information, see the @samp{proc} man page
19511(type @kbd{man 5 proc} from your shell prompt).
19512
19513@item info proc all
19514Show all the information about the process described under all of the
19515above @code{info proc} subcommands.
19516
19517@ignore
19518@comment These sub-options of 'info proc' were not included when
19519@comment procfs.c was re-written. Keep their descriptions around
19520@comment against the day when someone finds the time to put them back in.
19521@kindex info proc times
19522@item info proc times
19523Starting time, user CPU time, and system CPU time for your program and
19524its children.
19525
19526@kindex info proc id
19527@item info proc id
19528Report on the process IDs related to your program: its own process ID,
19529the ID of its parent, the process group ID, and the session ID.
19530@end ignore
19531
19532@item set procfs-trace
19533@kindex set procfs-trace
19534@cindex @code{procfs} API calls
19535This command enables and disables tracing of @code{procfs} API calls.
19536
19537@item show procfs-trace
19538@kindex show procfs-trace
19539Show the current state of @code{procfs} API call tracing.
19540
19541@item set procfs-file @var{file}
19542@kindex set procfs-file
19543Tell @value{GDBN} to write @code{procfs} API trace to the named
19544@var{file}. @value{GDBN} appends the trace info to the previous
19545contents of the file. The default is to display the trace on the
19546standard output.
19547
19548@item show procfs-file
19549@kindex show procfs-file
19550Show the file to which @code{procfs} API trace is written.
19551
19552@item proc-trace-entry
19553@itemx proc-trace-exit
19554@itemx proc-untrace-entry
19555@itemx proc-untrace-exit
19556@kindex proc-trace-entry
19557@kindex proc-trace-exit
19558@kindex proc-untrace-entry
19559@kindex proc-untrace-exit
19560These commands enable and disable tracing of entries into and exits
19561from the @code{syscall} interface.
19562
19563@item info pidlist
19564@kindex info pidlist
19565@cindex process list, QNX Neutrino
19566For QNX Neutrino only, this command displays the list of all the
19567processes and all the threads within each process.
19568
19569@item info meminfo
19570@kindex info meminfo
19571@cindex mapinfo list, QNX Neutrino
19572For QNX Neutrino only, this command displays the list of all mapinfos.
19573@end table
19574
19575@node DJGPP Native
19576@subsection Features for Debugging @sc{djgpp} Programs
19577@cindex @sc{djgpp} debugging
19578@cindex native @sc{djgpp} debugging
19579@cindex MS-DOS-specific commands
19580
19581@cindex DPMI
19582@sc{djgpp} is a port of the @sc{gnu} development tools to MS-DOS and
19583MS-Windows. @sc{djgpp} programs are 32-bit protected-mode programs
19584that use the @dfn{DPMI} (DOS Protected-Mode Interface) API to run on
19585top of real-mode DOS systems and their emulations.
19586
19587@value{GDBN} supports native debugging of @sc{djgpp} programs, and
19588defines a few commands specific to the @sc{djgpp} port. This
19589subsection describes those commands.
19590
19591@table @code
19592@kindex info dos
19593@item info dos
19594This is a prefix of @sc{djgpp}-specific commands which print
19595information about the target system and important OS structures.
19596
19597@kindex sysinfo
19598@cindex MS-DOS system info
19599@cindex free memory information (MS-DOS)
19600@item info dos sysinfo
19601This command displays assorted information about the underlying
19602platform: the CPU type and features, the OS version and flavor, the
19603DPMI version, and the available conventional and DPMI memory.
19604
19605@cindex GDT
19606@cindex LDT
19607@cindex IDT
19608@cindex segment descriptor tables
19609@cindex descriptor tables display
19610@item info dos gdt
19611@itemx info dos ldt
19612@itemx info dos idt
19613These 3 commands display entries from, respectively, Global, Local,
19614and Interrupt Descriptor Tables (GDT, LDT, and IDT). The descriptor
19615tables are data structures which store a descriptor for each segment
19616that is currently in use. The segment's selector is an index into a
19617descriptor table; the table entry for that index holds the
19618descriptor's base address and limit, and its attributes and access
19619rights.
19620
19621A typical @sc{djgpp} program uses 3 segments: a code segment, a data
19622segment (used for both data and the stack), and a DOS segment (which
19623allows access to DOS/BIOS data structures and absolute addresses in
19624conventional memory). However, the DPMI host will usually define
19625additional segments in order to support the DPMI environment.
19626
19627@cindex garbled pointers
19628These commands allow to display entries from the descriptor tables.
19629Without an argument, all entries from the specified table are
19630displayed. An argument, which should be an integer expression, means
19631display a single entry whose index is given by the argument. For
19632example, here's a convenient way to display information about the
19633debugged program's data segment:
19634
19635@smallexample
19636@exdent @code{(@value{GDBP}) info dos ldt $ds}
19637@exdent @code{0x13f: base=0x11970000 limit=0x0009ffff 32-Bit Data (Read/Write, Exp-up)}
19638@end smallexample
19639
19640@noindent
19641This comes in handy when you want to see whether a pointer is outside
19642the data segment's limit (i.e.@: @dfn{garbled}).
19643
19644@cindex page tables display (MS-DOS)
19645@item info dos pde
19646@itemx info dos pte
19647These two commands display entries from, respectively, the Page
19648Directory and the Page Tables. Page Directories and Page Tables are
19649data structures which control how virtual memory addresses are mapped
19650into physical addresses. A Page Table includes an entry for every
19651page of memory that is mapped into the program's address space; there
19652may be several Page Tables, each one holding up to 4096 entries. A
19653Page Directory has up to 4096 entries, one each for every Page Table
19654that is currently in use.
19655
19656Without an argument, @kbd{info dos pde} displays the entire Page
19657Directory, and @kbd{info dos pte} displays all the entries in all of
19658the Page Tables. An argument, an integer expression, given to the
19659@kbd{info dos pde} command means display only that entry from the Page
19660Directory table. An argument given to the @kbd{info dos pte} command
19661means display entries from a single Page Table, the one pointed to by
19662the specified entry in the Page Directory.
19663
19664@cindex direct memory access (DMA) on MS-DOS
19665These commands are useful when your program uses @dfn{DMA} (Direct
19666Memory Access), which needs physical addresses to program the DMA
19667controller.
19668
19669These commands are supported only with some DPMI servers.
19670
19671@cindex physical address from linear address
19672@item info dos address-pte @var{addr}
19673This command displays the Page Table entry for a specified linear
19674address. The argument @var{addr} is a linear address which should
19675already have the appropriate segment's base address added to it,
19676because this command accepts addresses which may belong to @emph{any}
19677segment. For example, here's how to display the Page Table entry for
19678the page where a variable @code{i} is stored:
19679
19680@smallexample
19681@exdent @code{(@value{GDBP}) info dos address-pte __djgpp_base_address + (char *)&i}
19682@exdent @code{Page Table entry for address 0x11a00d30:}
19683@exdent @code{Base=0x02698000 Dirty Acc. Not-Cached Write-Back Usr Read-Write +0xd30}
19684@end smallexample
19685
19686@noindent
19687This says that @code{i} is stored at offset @code{0xd30} from the page
19688whose physical base address is @code{0x02698000}, and shows all the
19689attributes of that page.
19690
19691Note that you must cast the addresses of variables to a @code{char *},
19692since otherwise the value of @code{__djgpp_base_address}, the base
19693address of all variables and functions in a @sc{djgpp} program, will
19694be added using the rules of C pointer arithmetics: if @code{i} is
19695declared an @code{int}, @value{GDBN} will add 4 times the value of
19696@code{__djgpp_base_address} to the address of @code{i}.
19697
19698Here's another example, it displays the Page Table entry for the
19699transfer buffer:
19700
19701@smallexample
19702@exdent @code{(@value{GDBP}) info dos address-pte *((unsigned *)&_go32_info_block + 3)}
19703@exdent @code{Page Table entry for address 0x29110:}
19704@exdent @code{Base=0x00029000 Dirty Acc. Not-Cached Write-Back Usr Read-Write +0x110}
19705@end smallexample
19706
19707@noindent
19708(The @code{+ 3} offset is because the transfer buffer's address is the
197093rd member of the @code{_go32_info_block} structure.) The output
19710clearly shows that this DPMI server maps the addresses in conventional
19711memory 1:1, i.e.@: the physical (@code{0x00029000} + @code{0x110}) and
19712linear (@code{0x29110}) addresses are identical.
19713
19714This command is supported only with some DPMI servers.
19715@end table
19716
19717@cindex DOS serial data link, remote debugging
19718In addition to native debugging, the DJGPP port supports remote
19719debugging via a serial data link. The following commands are specific
19720to remote serial debugging in the DJGPP port of @value{GDBN}.
19721
19722@table @code
19723@kindex set com1base
19724@kindex set com1irq
19725@kindex set com2base
19726@kindex set com2irq
19727@kindex set com3base
19728@kindex set com3irq
19729@kindex set com4base
19730@kindex set com4irq
19731@item set com1base @var{addr}
19732This command sets the base I/O port address of the @file{COM1} serial
19733port.
19734
19735@item set com1irq @var{irq}
19736This command sets the @dfn{Interrupt Request} (@code{IRQ}) line to use
19737for the @file{COM1} serial port.
19738
19739There are similar commands @samp{set com2base}, @samp{set com3irq},
19740etc.@: for setting the port address and the @code{IRQ} lines for the
19741other 3 COM ports.
19742
19743@kindex show com1base
19744@kindex show com1irq
19745@kindex show com2base
19746@kindex show com2irq
19747@kindex show com3base
19748@kindex show com3irq
19749@kindex show com4base
19750@kindex show com4irq
19751The related commands @samp{show com1base}, @samp{show com1irq} etc.@:
19752display the current settings of the base address and the @code{IRQ}
19753lines used by the COM ports.
19754
19755@item info serial
19756@kindex info serial
19757@cindex DOS serial port status
19758This command prints the status of the 4 DOS serial ports. For each
19759port, it prints whether it's active or not, its I/O base address and
19760IRQ number, whether it uses a 16550-style FIFO, its baudrate, and the
19761counts of various errors encountered so far.
19762@end table
19763
19764
19765@node Cygwin Native
19766@subsection Features for Debugging MS Windows PE Executables
19767@cindex MS Windows debugging
19768@cindex native Cygwin debugging
19769@cindex Cygwin-specific commands
19770
19771@value{GDBN} supports native debugging of MS Windows programs, including
19772DLLs with and without symbolic debugging information.
19773
19774@cindex Ctrl-BREAK, MS-Windows
19775@cindex interrupt debuggee on MS-Windows
19776MS-Windows programs that call @code{SetConsoleMode} to switch off the
19777special meaning of the @samp{Ctrl-C} keystroke cannot be interrupted
19778by typing @kbd{C-c}. For this reason, @value{GDBN} on MS-Windows
19779supports @kbd{C-@key{BREAK}} as an alternative interrupt key
19780sequence, which can be used to interrupt the debuggee even if it
19781ignores @kbd{C-c}.
19782
19783There are various additional Cygwin-specific commands, described in
19784this section. Working with DLLs that have no debugging symbols is
19785described in @ref{Non-debug DLL Symbols}.
19786
19787@table @code
19788@kindex info w32
19789@item info w32
19790This is a prefix of MS Windows-specific commands which print
19791information about the target system and important OS structures.
19792
19793@item info w32 selector
19794This command displays information returned by
19795the Win32 API @code{GetThreadSelectorEntry} function.
19796It takes an optional argument that is evaluated to
19797a long value to give the information about this given selector.
19798Without argument, this command displays information
19799about the six segment registers.
19800
19801@item info w32 thread-information-block
19802This command displays thread specific information stored in the
19803Thread Information Block (readable on the X86 CPU family using @code{$fs}
19804selector for 32-bit programs and @code{$gs} for 64-bit programs).
19805
19806@kindex info dll
19807@item info dll
19808This is a Cygwin-specific alias of @code{info shared}.
19809
19810@kindex dll-symbols
19811@item dll-symbols
19812This command loads symbols from a dll similarly to
19813add-sym command but without the need to specify a base address.
19814
19815@kindex set cygwin-exceptions
19816@cindex debugging the Cygwin DLL
19817@cindex Cygwin DLL, debugging
19818@item set cygwin-exceptions @var{mode}
19819If @var{mode} is @code{on}, @value{GDBN} will break on exceptions that
19820happen inside the Cygwin DLL. If @var{mode} is @code{off},
19821@value{GDBN} will delay recognition of exceptions, and may ignore some
19822exceptions which seem to be caused by internal Cygwin DLL
19823``bookkeeping''. This option is meant primarily for debugging the
19824Cygwin DLL itself; the default value is @code{off} to avoid annoying
19825@value{GDBN} users with false @code{SIGSEGV} signals.
19826
19827@kindex show cygwin-exceptions
19828@item show cygwin-exceptions
19829Displays whether @value{GDBN} will break on exceptions that happen
19830inside the Cygwin DLL itself.
19831
19832@kindex set new-console
19833@item set new-console @var{mode}
19834If @var{mode} is @code{on} the debuggee will
19835be started in a new console on next start.
19836If @var{mode} is @code{off}, the debuggee will
19837be started in the same console as the debugger.
19838
19839@kindex show new-console
19840@item show new-console
19841Displays whether a new console is used
19842when the debuggee is started.
19843
19844@kindex set new-group
19845@item set new-group @var{mode}
19846This boolean value controls whether the debuggee should
19847start a new group or stay in the same group as the debugger.
19848This affects the way the Windows OS handles
19849@samp{Ctrl-C}.
19850
19851@kindex show new-group
19852@item show new-group
19853Displays current value of new-group boolean.
19854
19855@kindex set debugevents
19856@item set debugevents
19857This boolean value adds debug output concerning kernel events related
19858to the debuggee seen by the debugger. This includes events that
19859signal thread and process creation and exit, DLL loading and
19860unloading, console interrupts, and debugging messages produced by the
19861Windows @code{OutputDebugString} API call.
19862
19863@kindex set debugexec
19864@item set debugexec
19865This boolean value adds debug output concerning execute events
19866(such as resume thread) seen by the debugger.
19867
19868@kindex set debugexceptions
19869@item set debugexceptions
19870This boolean value adds debug output concerning exceptions in the
19871debuggee seen by the debugger.
19872
19873@kindex set debugmemory
19874@item set debugmemory
19875This boolean value adds debug output concerning debuggee memory reads
19876and writes by the debugger.
19877
19878@kindex set shell
19879@item set shell
19880This boolean values specifies whether the debuggee is called
19881via a shell or directly (default value is on).
19882
19883@kindex show shell
19884@item show shell
19885Displays if the debuggee will be started with a shell.
19886
19887@end table
19888
19889@menu
19890* Non-debug DLL Symbols:: Support for DLLs without debugging symbols
19891@end menu
19892
19893@node Non-debug DLL Symbols
19894@subsubsection Support for DLLs without Debugging Symbols
19895@cindex DLLs with no debugging symbols
19896@cindex Minimal symbols and DLLs
19897
19898Very often on windows, some of the DLLs that your program relies on do
19899not include symbolic debugging information (for example,
19900@file{kernel32.dll}). When @value{GDBN} doesn't recognize any debugging
19901symbols in a DLL, it relies on the minimal amount of symbolic
19902information contained in the DLL's export table. This section
19903describes working with such symbols, known internally to @value{GDBN} as
19904``minimal symbols''.
19905
19906Note that before the debugged program has started execution, no DLLs
19907will have been loaded. The easiest way around this problem is simply to
19908start the program --- either by setting a breakpoint or letting the
19909program run once to completion. It is also possible to force
19910@value{GDBN} to load a particular DLL before starting the executable ---
19911see the shared library information in @ref{Files}, or the
19912@code{dll-symbols} command in @ref{Cygwin Native}. Currently,
19913explicitly loading symbols from a DLL with no debugging information will
19914cause the symbol names to be duplicated in @value{GDBN}'s lookup table,
19915which may adversely affect symbol lookup performance.
19916
19917@subsubsection DLL Name Prefixes
19918
19919In keeping with the naming conventions used by the Microsoft debugging
19920tools, DLL export symbols are made available with a prefix based on the
19921DLL name, for instance @code{KERNEL32!CreateFileA}. The plain name is
19922also entered into the symbol table, so @code{CreateFileA} is often
19923sufficient. In some cases there will be name clashes within a program
19924(particularly if the executable itself includes full debugging symbols)
19925necessitating the use of the fully qualified name when referring to the
19926contents of the DLL. Use single-quotes around the name to avoid the
19927exclamation mark (``!'') being interpreted as a language operator.
19928
19929Note that the internal name of the DLL may be all upper-case, even
19930though the file name of the DLL is lower-case, or vice-versa. Since
19931symbols within @value{GDBN} are @emph{case-sensitive} this may cause
19932some confusion. If in doubt, try the @code{info functions} and
19933@code{info variables} commands or even @code{maint print msymbols}
19934(@pxref{Symbols}). Here's an example:
19935
19936@smallexample
19937(@value{GDBP}) info function CreateFileA
19938All functions matching regular expression "CreateFileA":
19939
19940Non-debugging symbols:
199410x77e885f4 CreateFileA
199420x77e885f4 KERNEL32!CreateFileA
19943@end smallexample
19944
19945@smallexample
19946(@value{GDBP}) info function !
19947All functions matching regular expression "!":
19948
19949Non-debugging symbols:
199500x6100114c cygwin1!__assert
199510x61004034 cygwin1!_dll_crt0@@0
199520x61004240 cygwin1!dll_crt0(per_process *)
19953[etc...]
19954@end smallexample
19955
19956@subsubsection Working with Minimal Symbols
19957
19958Symbols extracted from a DLL's export table do not contain very much
19959type information. All that @value{GDBN} can do is guess whether a symbol
19960refers to a function or variable depending on the linker section that
19961contains the symbol. Also note that the actual contents of the memory
19962contained in a DLL are not available unless the program is running. This
19963means that you cannot examine the contents of a variable or disassemble
19964a function within a DLL without a running program.
19965
19966Variables are generally treated as pointers and dereferenced
19967automatically. For this reason, it is often necessary to prefix a
19968variable name with the address-of operator (``&'') and provide explicit
19969type information in the command. Here's an example of the type of
19970problem:
19971
19972@smallexample
19973(@value{GDBP}) print 'cygwin1!__argv'
19974$1 = 268572168
19975@end smallexample
19976
19977@smallexample
19978(@value{GDBP}) x 'cygwin1!__argv'
199790x10021610: "\230y\""
19980@end smallexample
19981
19982And two possible solutions:
19983
19984@smallexample
19985(@value{GDBP}) print ((char **)'cygwin1!__argv')[0]
19986$2 = 0x22fd98 "/cygdrive/c/mydirectory/myprogram"
19987@end smallexample
19988
19989@smallexample
19990(@value{GDBP}) x/2x &'cygwin1!__argv'
199910x610c0aa8 <cygwin1!__argv>: 0x10021608 0x00000000
19992(@value{GDBP}) x/x 0x10021608
199930x10021608: 0x0022fd98
19994(@value{GDBP}) x/s 0x0022fd98
199950x22fd98: "/cygdrive/c/mydirectory/myprogram"
19996@end smallexample
19997
19998Setting a break point within a DLL is possible even before the program
19999starts execution. However, under these circumstances, @value{GDBN} can't
20000examine the initial instructions of the function in order to skip the
20001function's frame set-up code. You can work around this by using ``*&''
20002to set the breakpoint at a raw memory address:
20003
20004@smallexample
20005(@value{GDBP}) break *&'python22!PyOS_Readline'
20006Breakpoint 1 at 0x1e04eff0
20007@end smallexample
20008
20009The author of these extensions is not entirely convinced that setting a
20010break point within a shared DLL like @file{kernel32.dll} is completely
20011safe.
20012
20013@node Hurd Native
20014@subsection Commands Specific to @sc{gnu} Hurd Systems
20015@cindex @sc{gnu} Hurd debugging
20016
20017This subsection describes @value{GDBN} commands specific to the
20018@sc{gnu} Hurd native debugging.
20019
20020@table @code
20021@item set signals
20022@itemx set sigs
20023@kindex set signals@r{, Hurd command}
20024@kindex set sigs@r{, Hurd command}
20025This command toggles the state of inferior signal interception by
20026@value{GDBN}. Mach exceptions, such as breakpoint traps, are not
20027affected by this command. @code{sigs} is a shorthand alias for
20028@code{signals}.
20029
20030@item show signals
20031@itemx show sigs
20032@kindex show signals@r{, Hurd command}
20033@kindex show sigs@r{, Hurd command}
20034Show the current state of intercepting inferior's signals.
20035
20036@item set signal-thread
20037@itemx set sigthread
20038@kindex set signal-thread
20039@kindex set sigthread
20040This command tells @value{GDBN} which thread is the @code{libc} signal
20041thread. That thread is run when a signal is delivered to a running
20042process. @code{set sigthread} is the shorthand alias of @code{set
20043signal-thread}.
20044
20045@item show signal-thread
20046@itemx show sigthread
20047@kindex show signal-thread
20048@kindex show sigthread
20049These two commands show which thread will run when the inferior is
20050delivered a signal.
20051
20052@item set stopped
20053@kindex set stopped@r{, Hurd command}
20054This commands tells @value{GDBN} that the inferior process is stopped,
20055as with the @code{SIGSTOP} signal. The stopped process can be
20056continued by delivering a signal to it.
20057
20058@item show stopped
20059@kindex show stopped@r{, Hurd command}
20060This command shows whether @value{GDBN} thinks the debuggee is
20061stopped.
20062
20063@item set exceptions
20064@kindex set exceptions@r{, Hurd command}
20065Use this command to turn off trapping of exceptions in the inferior.
20066When exception trapping is off, neither breakpoints nor
20067single-stepping will work. To restore the default, set exception
20068trapping on.
20069
20070@item show exceptions
20071@kindex show exceptions@r{, Hurd command}
20072Show the current state of trapping exceptions in the inferior.
20073
20074@item set task pause
20075@kindex set task@r{, Hurd commands}
20076@cindex task attributes (@sc{gnu} Hurd)
20077@cindex pause current task (@sc{gnu} Hurd)
20078This command toggles task suspension when @value{GDBN} has control.
20079Setting it to on takes effect immediately, and the task is suspended
20080whenever @value{GDBN} gets control. Setting it to off will take
20081effect the next time the inferior is continued. If this option is set
20082to off, you can use @code{set thread default pause on} or @code{set
20083thread pause on} (see below) to pause individual threads.
20084
20085@item show task pause
20086@kindex show task@r{, Hurd commands}
20087Show the current state of task suspension.
20088
20089@item set task detach-suspend-count
20090@cindex task suspend count
20091@cindex detach from task, @sc{gnu} Hurd
20092This command sets the suspend count the task will be left with when
20093@value{GDBN} detaches from it.
20094
20095@item show task detach-suspend-count
20096Show the suspend count the task will be left with when detaching.
20097
20098@item set task exception-port
20099@itemx set task excp
20100@cindex task exception port, @sc{gnu} Hurd
20101This command sets the task exception port to which @value{GDBN} will
20102forward exceptions. The argument should be the value of the @dfn{send
20103rights} of the task. @code{set task excp} is a shorthand alias.
20104
20105@item set noninvasive
20106@cindex noninvasive task options
20107This command switches @value{GDBN} to a mode that is the least
20108invasive as far as interfering with the inferior is concerned. This
20109is the same as using @code{set task pause}, @code{set exceptions}, and
20110@code{set signals} to values opposite to the defaults.
20111
20112@item info send-rights
20113@itemx info receive-rights
20114@itemx info port-rights
20115@itemx info port-sets
20116@itemx info dead-names
20117@itemx info ports
20118@itemx info psets
20119@cindex send rights, @sc{gnu} Hurd
20120@cindex receive rights, @sc{gnu} Hurd
20121@cindex port rights, @sc{gnu} Hurd
20122@cindex port sets, @sc{gnu} Hurd
20123@cindex dead names, @sc{gnu} Hurd
20124These commands display information about, respectively, send rights,
20125receive rights, port rights, port sets, and dead names of a task.
20126There are also shorthand aliases: @code{info ports} for @code{info
20127port-rights} and @code{info psets} for @code{info port-sets}.
20128
20129@item set thread pause
20130@kindex set thread@r{, Hurd command}
20131@cindex thread properties, @sc{gnu} Hurd
20132@cindex pause current thread (@sc{gnu} Hurd)
20133This command toggles current thread suspension when @value{GDBN} has
20134control. Setting it to on takes effect immediately, and the current
20135thread is suspended whenever @value{GDBN} gets control. Setting it to
20136off will take effect the next time the inferior is continued.
20137Normally, this command has no effect, since when @value{GDBN} has
20138control, the whole task is suspended. However, if you used @code{set
20139task pause off} (see above), this command comes in handy to suspend
20140only the current thread.
20141
20142@item show thread pause
20143@kindex show thread@r{, Hurd command}
20144This command shows the state of current thread suspension.
20145
20146@item set thread run
20147This command sets whether the current thread is allowed to run.
20148
20149@item show thread run
20150Show whether the current thread is allowed to run.
20151
20152@item set thread detach-suspend-count
20153@cindex thread suspend count, @sc{gnu} Hurd
20154@cindex detach from thread, @sc{gnu} Hurd
20155This command sets the suspend count @value{GDBN} will leave on a
20156thread when detaching. This number is relative to the suspend count
20157found by @value{GDBN} when it notices the thread; use @code{set thread
20158takeover-suspend-count} to force it to an absolute value.
20159
20160@item show thread detach-suspend-count
20161Show the suspend count @value{GDBN} will leave on the thread when
20162detaching.
20163
20164@item set thread exception-port
20165@itemx set thread excp
20166Set the thread exception port to which to forward exceptions. This
20167overrides the port set by @code{set task exception-port} (see above).
20168@code{set thread excp} is the shorthand alias.
20169
20170@item set thread takeover-suspend-count
20171Normally, @value{GDBN}'s thread suspend counts are relative to the
20172value @value{GDBN} finds when it notices each thread. This command
20173changes the suspend counts to be absolute instead.
20174
20175@item set thread default
20176@itemx show thread default
20177@cindex thread default settings, @sc{gnu} Hurd
20178Each of the above @code{set thread} commands has a @code{set thread
20179default} counterpart (e.g., @code{set thread default pause}, @code{set
20180thread default exception-port}, etc.). The @code{thread default}
20181variety of commands sets the default thread properties for all
20182threads; you can then change the properties of individual threads with
20183the non-default commands.
20184@end table
20185
20186@node Darwin
20187@subsection Darwin
20188@cindex Darwin
20189
20190@value{GDBN} provides the following commands specific to the Darwin target:
20191
20192@table @code
20193@item set debug darwin @var{num}
20194@kindex set debug darwin
20195When set to a non zero value, enables debugging messages specific to
20196the Darwin support. Higher values produce more verbose output.
20197
20198@item show debug darwin
20199@kindex show debug darwin
20200Show the current state of Darwin messages.
20201
20202@item set debug mach-o @var{num}
20203@kindex set debug mach-o
20204When set to a non zero value, enables debugging messages while
20205@value{GDBN} is reading Darwin object files. (@dfn{Mach-O} is the
20206file format used on Darwin for object and executable files.) Higher
20207values produce more verbose output. This is a command to diagnose
20208problems internal to @value{GDBN} and should not be needed in normal
20209usage.
20210
20211@item show debug mach-o
20212@kindex show debug mach-o
20213Show the current state of Mach-O file messages.
20214
20215@item set mach-exceptions on
20216@itemx set mach-exceptions off
20217@kindex set mach-exceptions
20218On Darwin, faults are first reported as a Mach exception and are then
20219mapped to a Posix signal. Use this command to turn on trapping of
20220Mach exceptions in the inferior. This might be sometimes useful to
20221better understand the cause of a fault. The default is off.
20222
20223@item show mach-exceptions
20224@kindex show mach-exceptions
20225Show the current state of exceptions trapping.
20226@end table
20227
20228
20229@node Embedded OS
20230@section Embedded Operating Systems
20231
20232This section describes configurations involving the debugging of
20233embedded operating systems that are available for several different
20234architectures.
20235
20236@menu
20237* VxWorks:: Using @value{GDBN} with VxWorks
20238@end menu
20239
20240@value{GDBN} includes the ability to debug programs running on
20241various real-time operating systems.
20242
20243@node VxWorks
20244@subsection Using @value{GDBN} with VxWorks
20245
20246@cindex VxWorks
20247
20248@table @code
20249
20250@kindex target vxworks
20251@item target vxworks @var{machinename}
20252A VxWorks system, attached via TCP/IP. The argument @var{machinename}
20253is the target system's machine name or IP address.
20254
20255@end table
20256
20257On VxWorks, @code{load} links @var{filename} dynamically on the
20258current target system as well as adding its symbols in @value{GDBN}.
20259
20260@value{GDBN} enables developers to spawn and debug tasks running on networked
20261VxWorks targets from a Unix host. Already-running tasks spawned from
20262the VxWorks shell can also be debugged. @value{GDBN} uses code that runs on
20263both the Unix host and on the VxWorks target. The program
20264@code{@value{GDBP}} is installed and executed on the Unix host. (It may be
20265installed with the name @code{vxgdb}, to distinguish it from a
20266@value{GDBN} for debugging programs on the host itself.)
20267
20268@table @code
20269@item VxWorks-timeout @var{args}
20270@kindex vxworks-timeout
20271All VxWorks-based targets now support the option @code{vxworks-timeout}.
20272This option is set by the user, and @var{args} represents the number of
20273seconds @value{GDBN} waits for responses to rpc's. You might use this if
20274your VxWorks target is a slow software simulator or is on the far side
20275of a thin network line.
20276@end table
20277
20278The following information on connecting to VxWorks was current when
20279this manual was produced; newer releases of VxWorks may use revised
20280procedures.
20281
20282@findex INCLUDE_RDB
20283To use @value{GDBN} with VxWorks, you must rebuild your VxWorks kernel
20284to include the remote debugging interface routines in the VxWorks
20285library @file{rdb.a}. To do this, define @code{INCLUDE_RDB} in the
20286VxWorks configuration file @file{configAll.h} and rebuild your VxWorks
20287kernel. The resulting kernel contains @file{rdb.a}, and spawns the
20288source debugging task @code{tRdbTask} when VxWorks is booted. For more
20289information on configuring and remaking VxWorks, see the manufacturer's
20290manual.
20291@c VxWorks, see the @cite{VxWorks Programmer's Guide}.
20292
20293Once you have included @file{rdb.a} in your VxWorks system image and set
20294your Unix execution search path to find @value{GDBN}, you are ready to
20295run @value{GDBN}. From your Unix host, run @code{@value{GDBP}} (or
20296@code{vxgdb}, depending on your installation).
20297
20298@value{GDBN} comes up showing the prompt:
20299
20300@smallexample
20301(vxgdb)
20302@end smallexample
20303
20304@menu
20305* VxWorks Connection:: Connecting to VxWorks
20306* VxWorks Download:: VxWorks download
20307* VxWorks Attach:: Running tasks
20308@end menu
20309
20310@node VxWorks Connection
20311@subsubsection Connecting to VxWorks
20312
20313The @value{GDBN} command @code{target} lets you connect to a VxWorks target on the
20314network. To connect to a target whose host name is ``@code{tt}'', type:
20315
20316@smallexample
20317(vxgdb) target vxworks tt
20318@end smallexample
20319
20320@need 750
20321@value{GDBN} displays messages like these:
20322
20323@smallexample
20324Attaching remote machine across net...
20325Connected to tt.
20326@end smallexample
20327
20328@need 1000
20329@value{GDBN} then attempts to read the symbol tables of any object modules
20330loaded into the VxWorks target since it was last booted. @value{GDBN} locates
20331these files by searching the directories listed in the command search
20332path (@pxref{Environment, ,Your Program's Environment}); if it fails
20333to find an object file, it displays a message such as:
20334
20335@smallexample
20336prog.o: No such file or directory.
20337@end smallexample
20338
20339When this happens, add the appropriate directory to the search path with
20340the @value{GDBN} command @code{path}, and execute the @code{target}
20341command again.
20342
20343@node VxWorks Download
20344@subsubsection VxWorks Download
20345
20346@cindex download to VxWorks
20347If you have connected to the VxWorks target and you want to debug an
20348object that has not yet been loaded, you can use the @value{GDBN}
20349@code{load} command to download a file from Unix to VxWorks
20350incrementally. The object file given as an argument to the @code{load}
20351command is actually opened twice: first by the VxWorks target in order
20352to download the code, then by @value{GDBN} in order to read the symbol
20353table. This can lead to problems if the current working directories on
20354the two systems differ. If both systems have NFS mounted the same
20355filesystems, you can avoid these problems by using absolute paths.
20356Otherwise, it is simplest to set the working directory on both systems
20357to the directory in which the object file resides, and then to reference
20358the file by its name, without any path. For instance, a program
20359@file{prog.o} may reside in @file{@var{vxpath}/vw/demo/rdb} in VxWorks
20360and in @file{@var{hostpath}/vw/demo/rdb} on the host. To load this
20361program, type this on VxWorks:
20362
20363@smallexample
20364-> cd "@var{vxpath}/vw/demo/rdb"
20365@end smallexample
20366
20367@noindent
20368Then, in @value{GDBN}, type:
20369
20370@smallexample
20371(vxgdb) cd @var{hostpath}/vw/demo/rdb
20372(vxgdb) load prog.o
20373@end smallexample
20374
20375@value{GDBN} displays a response similar to this:
20376
20377@smallexample
20378Reading symbol data from wherever/vw/demo/rdb/prog.o... done.
20379@end smallexample
20380
20381You can also use the @code{load} command to reload an object module
20382after editing and recompiling the corresponding source file. Note that
20383this makes @value{GDBN} delete all currently-defined breakpoints,
20384auto-displays, and convenience variables, and to clear the value
20385history. (This is necessary in order to preserve the integrity of
20386debugger's data structures that reference the target system's symbol
20387table.)
20388
20389@node VxWorks Attach
20390@subsubsection Running Tasks
20391
20392@cindex running VxWorks tasks
20393You can also attach to an existing task using the @code{attach} command as
20394follows:
20395
20396@smallexample
20397(vxgdb) attach @var{task}
20398@end smallexample
20399
20400@noindent
20401where @var{task} is the VxWorks hexadecimal task ID. The task can be running
20402or suspended when you attach to it. Running tasks are suspended at
20403the time of attachment.
20404
20405@node Embedded Processors
20406@section Embedded Processors
20407
20408This section goes into details specific to particular embedded
20409configurations.
20410
20411@cindex send command to simulator
20412Whenever a specific embedded processor has a simulator, @value{GDBN}
20413allows to send an arbitrary command to the simulator.
20414
20415@table @code
20416@item sim @var{command}
20417@kindex sim@r{, a command}
20418Send an arbitrary @var{command} string to the simulator. Consult the
20419documentation for the specific simulator in use for information about
20420acceptable commands.
20421@end table
20422
20423
20424@menu
20425* ARM:: ARM RDI
20426* M32R/D:: Renesas M32R/D
20427* M68K:: Motorola M68K
20428* MicroBlaze:: Xilinx MicroBlaze
20429* MIPS Embedded:: MIPS Embedded
20430* PowerPC Embedded:: PowerPC Embedded
20431* PA:: HP PA Embedded
20432* Sparclet:: Tsqware Sparclet
20433* Sparclite:: Fujitsu Sparclite
20434* Z8000:: Zilog Z8000
20435* AVR:: Atmel AVR
20436* CRIS:: CRIS
20437* Super-H:: Renesas Super-H
20438@end menu
20439
20440@node ARM
20441@subsection ARM
20442@cindex ARM RDI
20443
20444@table @code
20445@kindex target rdi
20446@item target rdi @var{dev}
20447ARM Angel monitor, via RDI library interface to ADP protocol. You may
20448use this target to communicate with both boards running the Angel
20449monitor, or with the EmbeddedICE JTAG debug device.
20450
20451@kindex target rdp
20452@item target rdp @var{dev}
20453ARM Demon monitor.
20454
20455@end table
20456
20457@value{GDBN} provides the following ARM-specific commands:
20458
20459@table @code
20460@item set arm disassembler
20461@kindex set arm
20462This commands selects from a list of disassembly styles. The
20463@code{"std"} style is the standard style.
20464
20465@item show arm disassembler
20466@kindex show arm
20467Show the current disassembly style.
20468
20469@item set arm apcs32
20470@cindex ARM 32-bit mode
20471This command toggles ARM operation mode between 32-bit and 26-bit.
20472
20473@item show arm apcs32
20474Display the current usage of the ARM 32-bit mode.
20475
20476@item set arm fpu @var{fputype}
20477This command sets the ARM floating-point unit (FPU) type. The
20478argument @var{fputype} can be one of these:
20479
20480@table @code
20481@item auto
20482Determine the FPU type by querying the OS ABI.
20483@item softfpa
20484Software FPU, with mixed-endian doubles on little-endian ARM
20485processors.
20486@item fpa
20487GCC-compiled FPA co-processor.
20488@item softvfp
20489Software FPU with pure-endian doubles.
20490@item vfp
20491VFP co-processor.
20492@end table
20493
20494@item show arm fpu
20495Show the current type of the FPU.
20496
20497@item set arm abi
20498This command forces @value{GDBN} to use the specified ABI.
20499
20500@item show arm abi
20501Show the currently used ABI.
20502
20503@item set arm fallback-mode (arm|thumb|auto)
20504@value{GDBN} uses the symbol table, when available, to determine
20505whether instructions are ARM or Thumb. This command controls
20506@value{GDBN}'s default behavior when the symbol table is not
20507available. The default is @samp{auto}, which causes @value{GDBN} to
20508use the current execution mode (from the @code{T} bit in the @code{CPSR}
20509register).
20510
20511@item show arm fallback-mode
20512Show the current fallback instruction mode.
20513
20514@item set arm force-mode (arm|thumb|auto)
20515This command overrides use of the symbol table to determine whether
20516instructions are ARM or Thumb. The default is @samp{auto}, which
20517causes @value{GDBN} to use the symbol table and then the setting
20518of @samp{set arm fallback-mode}.
20519
20520@item show arm force-mode
20521Show the current forced instruction mode.
20522
20523@item set debug arm
20524Toggle whether to display ARM-specific debugging messages from the ARM
20525target support subsystem.
20526
20527@item show debug arm
20528Show whether ARM-specific debugging messages are enabled.
20529@end table
20530
20531The following commands are available when an ARM target is debugged
20532using the RDI interface:
20533
20534@table @code
20535@item rdilogfile @r{[}@var{file}@r{]}
20536@kindex rdilogfile
20537@cindex ADP (Angel Debugger Protocol) logging
20538Set the filename for the ADP (Angel Debugger Protocol) packet log.
20539With an argument, sets the log file to the specified @var{file}. With
20540no argument, show the current log file name. The default log file is
20541@file{rdi.log}.
20542
20543@item rdilogenable @r{[}@var{arg}@r{]}
20544@kindex rdilogenable
20545Control logging of ADP packets. With an argument of 1 or @code{"yes"}
20546enables logging, with an argument 0 or @code{"no"} disables it. With
20547no arguments displays the current setting. When logging is enabled,
20548ADP packets exchanged between @value{GDBN} and the RDI target device
20549are logged to a file.
20550
20551@item set rdiromatzero
20552@kindex set rdiromatzero
20553@cindex ROM at zero address, RDI
20554Tell @value{GDBN} whether the target has ROM at address 0. If on,
20555vector catching is disabled, so that zero address can be used. If off
20556(the default), vector catching is enabled. For this command to take
20557effect, it needs to be invoked prior to the @code{target rdi} command.
20558
20559@item show rdiromatzero
20560@kindex show rdiromatzero
20561Show the current setting of ROM at zero address.
20562
20563@item set rdiheartbeat
20564@kindex set rdiheartbeat
20565@cindex RDI heartbeat
20566Enable or disable RDI heartbeat packets. It is not recommended to
20567turn on this option, since it confuses ARM and EPI JTAG interface, as
20568well as the Angel monitor.
20569
20570@item show rdiheartbeat
20571@kindex show rdiheartbeat
20572Show the setting of RDI heartbeat packets.
20573@end table
20574
20575@table @code
20576@item target sim @r{[}@var{simargs}@r{]} @dots{}
20577The @value{GDBN} ARM simulator accepts the following optional arguments.
20578
20579@table @code
20580@item --swi-support=@var{type}
20581Tell the simulator which SWI interfaces to support.
20582@var{type} may be a comma separated list of the following values.
20583The default value is @code{all}.
20584
20585@table @code
20586@item none
20587@item demon
20588@item angel
20589@item redboot
20590@item all
20591@end table
20592@end table
20593@end table
20594
20595@node M32R/D
20596@subsection Renesas M32R/D and M32R/SDI
20597
20598@table @code
20599@kindex target m32r
20600@item target m32r @var{dev}
20601Renesas M32R/D ROM monitor.
20602
20603@kindex target m32rsdi
20604@item target m32rsdi @var{dev}
20605Renesas M32R SDI server, connected via parallel port to the board.
20606@end table
20607
20608The following @value{GDBN} commands are specific to the M32R monitor:
20609
20610@table @code
20611@item set download-path @var{path}
20612@kindex set download-path
20613@cindex find downloadable @sc{srec} files (M32R)
20614Set the default path for finding downloadable @sc{srec} files.
20615
20616@item show download-path
20617@kindex show download-path
20618Show the default path for downloadable @sc{srec} files.
20619
20620@item set board-address @var{addr}
20621@kindex set board-address
20622@cindex M32-EVA target board address
20623Set the IP address for the M32R-EVA target board.
20624
20625@item show board-address
20626@kindex show board-address
20627Show the current IP address of the target board.
20628
20629@item set server-address @var{addr}
20630@kindex set server-address
20631@cindex download server address (M32R)
20632Set the IP address for the download server, which is the @value{GDBN}'s
20633host machine.
20634
20635@item show server-address
20636@kindex show server-address
20637Display the IP address of the download server.
20638
20639@item upload @r{[}@var{file}@r{]}
20640@kindex upload@r{, M32R}
20641Upload the specified @sc{srec} @var{file} via the monitor's Ethernet
20642upload capability. If no @var{file} argument is given, the current
20643executable file is uploaded.
20644
20645@item tload @r{[}@var{file}@r{]}
20646@kindex tload@r{, M32R}
20647Test the @code{upload} command.
20648@end table
20649
20650The following commands are available for M32R/SDI:
20651
20652@table @code
20653@item sdireset
20654@kindex sdireset
20655@cindex reset SDI connection, M32R
20656This command resets the SDI connection.
20657
20658@item sdistatus
20659@kindex sdistatus
20660This command shows the SDI connection status.
20661
20662@item debug_chaos
20663@kindex debug_chaos
20664@cindex M32R/Chaos debugging
20665Instructs the remote that M32R/Chaos debugging is to be used.
20666
20667@item use_debug_dma
20668@kindex use_debug_dma
20669Instructs the remote to use the DEBUG_DMA method of accessing memory.
20670
20671@item use_mon_code
20672@kindex use_mon_code
20673Instructs the remote to use the MON_CODE method of accessing memory.
20674
20675@item use_ib_break
20676@kindex use_ib_break
20677Instructs the remote to set breakpoints by IB break.
20678
20679@item use_dbt_break
20680@kindex use_dbt_break
20681Instructs the remote to set breakpoints by DBT.
20682@end table
20683
20684@node M68K
20685@subsection M68k
20686
20687The Motorola m68k configuration includes ColdFire support, and a
20688target command for the following ROM monitor.
20689
20690@table @code
20691
20692@kindex target dbug
20693@item target dbug @var{dev}
20694dBUG ROM monitor for Motorola ColdFire.
20695
20696@end table
20697
20698@node MicroBlaze
20699@subsection MicroBlaze
20700@cindex Xilinx MicroBlaze
20701@cindex XMD, Xilinx Microprocessor Debugger
20702
20703The MicroBlaze is a soft-core processor supported on various Xilinx
20704FPGAs, such as Spartan or Virtex series. Boards with these processors
20705usually have JTAG ports which connect to a host system running the Xilinx
20706Embedded Development Kit (EDK) or Software Development Kit (SDK).
20707This host system is used to download the configuration bitstream to
20708the target FPGA. The Xilinx Microprocessor Debugger (XMD) program
20709communicates with the target board using the JTAG interface and
20710presents a @code{gdbserver} interface to the board. By default
20711@code{xmd} uses port @code{1234}. (While it is possible to change
20712this default port, it requires the use of undocumented @code{xmd}
20713commands. Contact Xilinx support if you need to do this.)
20714
20715Use these GDB commands to connect to the MicroBlaze target processor.
20716
20717@table @code
20718@item target remote :1234
20719Use this command to connect to the target if you are running @value{GDBN}
20720on the same system as @code{xmd}.
20721
20722@item target remote @var{xmd-host}:1234
20723Use this command to connect to the target if it is connected to @code{xmd}
20724running on a different system named @var{xmd-host}.
20725
20726@item load
20727Use this command to download a program to the MicroBlaze target.
20728
20729@item set debug microblaze @var{n}
20730Enable MicroBlaze-specific debugging messages if non-zero.
20731
20732@item show debug microblaze @var{n}
20733Show MicroBlaze-specific debugging level.
20734@end table
20735
20736@node MIPS Embedded
20737@subsection @acronym{MIPS} Embedded
20738
20739@cindex @acronym{MIPS} boards
20740@value{GDBN} can use the @acronym{MIPS} remote debugging protocol to talk to a
20741@acronym{MIPS} board attached to a serial line. This is available when
20742you configure @value{GDBN} with @samp{--target=mips-elf}.
20743
20744@need 1000
20745Use these @value{GDBN} commands to specify the connection to your target board:
20746
20747@table @code
20748@item target mips @var{port}
20749@kindex target mips @var{port}
20750To run a program on the board, start up @code{@value{GDBP}} with the
20751name of your program as the argument. To connect to the board, use the
20752command @samp{target mips @var{port}}, where @var{port} is the name of
20753the serial port connected to the board. If the program has not already
20754been downloaded to the board, you may use the @code{load} command to
20755download it. You can then use all the usual @value{GDBN} commands.
20756
20757For example, this sequence connects to the target board through a serial
20758port, and loads and runs a program called @var{prog} through the
20759debugger:
20760
20761@smallexample
20762host$ @value{GDBP} @var{prog}
20763@value{GDBN} is free software and @dots{}
20764(@value{GDBP}) target mips /dev/ttyb
20765(@value{GDBP}) load @var{prog}
20766(@value{GDBP}) run
20767@end smallexample
20768
20769@item target mips @var{hostname}:@var{portnumber}
20770On some @value{GDBN} host configurations, you can specify a TCP
20771connection (for instance, to a serial line managed by a terminal
20772concentrator) instead of a serial port, using the syntax
20773@samp{@var{hostname}:@var{portnumber}}.
20774
20775@item target pmon @var{port}
20776@kindex target pmon @var{port}
20777PMON ROM monitor.
20778
20779@item target ddb @var{port}
20780@kindex target ddb @var{port}
20781NEC's DDB variant of PMON for Vr4300.
20782
20783@item target lsi @var{port}
20784@kindex target lsi @var{port}
20785LSI variant of PMON.
20786
20787@kindex target r3900
20788@item target r3900 @var{dev}
20789Densan DVE-R3900 ROM monitor for Toshiba R3900 Mips.
20790
20791@kindex target array
20792@item target array @var{dev}
20793Array Tech LSI33K RAID controller board.
20794
20795@end table
20796
20797
20798@noindent
20799@value{GDBN} also supports these special commands for @acronym{MIPS} targets:
20800
20801@table @code
20802@item set mipsfpu double
20803@itemx set mipsfpu single
20804@itemx set mipsfpu none
20805@itemx set mipsfpu auto
20806@itemx show mipsfpu
20807@kindex set mipsfpu
20808@kindex show mipsfpu
20809@cindex @acronym{MIPS} remote floating point
20810@cindex floating point, @acronym{MIPS} remote
20811If your target board does not support the @acronym{MIPS} floating point
20812coprocessor, you should use the command @samp{set mipsfpu none} (if you
20813need this, you may wish to put the command in your @value{GDBN} init
20814file). This tells @value{GDBN} how to find the return value of
20815functions which return floating point values. It also allows
20816@value{GDBN} to avoid saving the floating point registers when calling
20817functions on the board. If you are using a floating point coprocessor
20818with only single precision floating point support, as on the @sc{r4650}
20819processor, use the command @samp{set mipsfpu single}. The default
20820double precision floating point coprocessor may be selected using
20821@samp{set mipsfpu double}.
20822
20823In previous versions the only choices were double precision or no
20824floating point, so @samp{set mipsfpu on} will select double precision
20825and @samp{set mipsfpu off} will select no floating point.
20826
20827As usual, you can inquire about the @code{mipsfpu} variable with
20828@samp{show mipsfpu}.
20829
20830@item set timeout @var{seconds}
20831@itemx set retransmit-timeout @var{seconds}
20832@itemx show timeout
20833@itemx show retransmit-timeout
20834@cindex @code{timeout}, @acronym{MIPS} protocol
20835@cindex @code{retransmit-timeout}, @acronym{MIPS} protocol
20836@kindex set timeout
20837@kindex show timeout
20838@kindex set retransmit-timeout
20839@kindex show retransmit-timeout
20840You can control the timeout used while waiting for a packet, in the @acronym{MIPS}
20841remote protocol, with the @code{set timeout @var{seconds}} command. The
20842default is 5 seconds. Similarly, you can control the timeout used while
20843waiting for an acknowledgment of a packet with the @code{set
20844retransmit-timeout @var{seconds}} command. The default is 3 seconds.
20845You can inspect both values with @code{show timeout} and @code{show
20846retransmit-timeout}. (These commands are @emph{only} available when
20847@value{GDBN} is configured for @samp{--target=mips-elf}.)
20848
20849The timeout set by @code{set timeout} does not apply when @value{GDBN}
20850is waiting for your program to stop. In that case, @value{GDBN} waits
20851forever because it has no way of knowing how long the program is going
20852to run before stopping.
20853
20854@item set syn-garbage-limit @var{num}
20855@kindex set syn-garbage-limit@r{, @acronym{MIPS} remote}
20856@cindex synchronize with remote @acronym{MIPS} target
20857Limit the maximum number of characters @value{GDBN} should ignore when
20858it tries to synchronize with the remote target. The default is 10
20859characters. Setting the limit to -1 means there's no limit.
20860
20861@item show syn-garbage-limit
20862@kindex show syn-garbage-limit@r{, @acronym{MIPS} remote}
20863Show the current limit on the number of characters to ignore when
20864trying to synchronize with the remote system.
20865
20866@item set monitor-prompt @var{prompt}
20867@kindex set monitor-prompt@r{, @acronym{MIPS} remote}
20868@cindex remote monitor prompt
20869Tell @value{GDBN} to expect the specified @var{prompt} string from the
20870remote monitor. The default depends on the target:
20871@table @asis
20872@item pmon target
20873@samp{PMON}
20874@item ddb target
20875@samp{NEC010}
20876@item lsi target
20877@samp{PMON>}
20878@end table
20879
20880@item show monitor-prompt
20881@kindex show monitor-prompt@r{, @acronym{MIPS} remote}
20882Show the current strings @value{GDBN} expects as the prompt from the
20883remote monitor.
20884
20885@item set monitor-warnings
20886@kindex set monitor-warnings@r{, @acronym{MIPS} remote}
20887Enable or disable monitor warnings about hardware breakpoints. This
20888has effect only for the @code{lsi} target. When on, @value{GDBN} will
20889display warning messages whose codes are returned by the @code{lsi}
20890PMON monitor for breakpoint commands.
20891
20892@item show monitor-warnings
20893@kindex show monitor-warnings@r{, @acronym{MIPS} remote}
20894Show the current setting of printing monitor warnings.
20895
20896@item pmon @var{command}
20897@kindex pmon@r{, @acronym{MIPS} remote}
20898@cindex send PMON command
20899This command allows sending an arbitrary @var{command} string to the
20900monitor. The monitor must be in debug mode for this to work.
20901@end table
20902
20903@node PowerPC Embedded
20904@subsection PowerPC Embedded
20905
20906@cindex DVC register
20907@value{GDBN} supports using the DVC (Data Value Compare) register to
20908implement in hardware simple hardware watchpoint conditions of the form:
20909
20910@smallexample
20911(@value{GDBP}) watch @var{ADDRESS|VARIABLE} \
20912 if @var{ADDRESS|VARIABLE} == @var{CONSTANT EXPRESSION}
20913@end smallexample
20914
20915The DVC register will be automatically used when @value{GDBN} detects
20916such pattern in a condition expression, and the created watchpoint uses one
20917debug register (either the @code{exact-watchpoints} option is on and the
20918variable is scalar, or the variable has a length of one byte). This feature
20919is available in native @value{GDBN} running on a Linux kernel version 2.6.34
20920or newer.
20921
20922When running on PowerPC embedded processors, @value{GDBN} automatically uses
20923ranged hardware watchpoints, unless the @code{exact-watchpoints} option is on,
20924in which case watchpoints using only one debug register are created when
20925watching variables of scalar types.
20926
20927You can create an artificial array to watch an arbitrary memory
20928region using one of the following commands (@pxref{Expressions}):
20929
20930@smallexample
20931(@value{GDBP}) watch *((char *) @var{address})@@@var{length}
20932(@value{GDBP}) watch @{char[@var{length}]@} @var{address}
20933@end smallexample
20934
20935PowerPC embedded processors support masked watchpoints. See the discussion
20936about the @code{mask} argument in @ref{Set Watchpoints}.
20937
20938@cindex ranged breakpoint
20939PowerPC embedded processors support hardware accelerated
20940@dfn{ranged breakpoints}. A ranged breakpoint stops execution of
20941the inferior whenever it executes an instruction at any address within
20942the range it specifies. To set a ranged breakpoint in @value{GDBN},
20943use the @code{break-range} command.
20944
20945@value{GDBN} provides the following PowerPC-specific commands:
20946
20947@table @code
20948@kindex break-range
20949@item break-range @var{start-location}, @var{end-location}
20950Set a breakpoint for an address range.
20951@var{start-location} and @var{end-location} can specify a function name,
20952a line number, an offset of lines from the current line or from the start
20953location, or an address of an instruction (see @ref{Specify Location},
20954for a list of all the possible ways to specify a @var{location}.)
20955The breakpoint will stop execution of the inferior whenever it
20956executes an instruction at any address within the specified range,
20957(including @var{start-location} and @var{end-location}.)
20958
20959@kindex set powerpc
20960@item set powerpc soft-float
20961@itemx show powerpc soft-float
20962Force @value{GDBN} to use (or not use) a software floating point calling
20963convention. By default, @value{GDBN} selects the calling convention based
20964on the selected architecture and the provided executable file.
20965
20966@item set powerpc vector-abi
20967@itemx show powerpc vector-abi
20968Force @value{GDBN} to use the specified calling convention for vector
20969arguments and return values. The valid options are @samp{auto};
20970@samp{generic}, to avoid vector registers even if they are present;
20971@samp{altivec}, to use AltiVec registers; and @samp{spe} to use SPE
20972registers. By default, @value{GDBN} selects the calling convention
20973based on the selected architecture and the provided executable file.
20974
20975@item set powerpc exact-watchpoints
20976@itemx show powerpc exact-watchpoints
20977Allow @value{GDBN} to use only one debug register when watching a variable
20978of scalar type, thus assuming that the variable is accessed through the
20979address of its first byte.
20980
20981@kindex target dink32
20982@item target dink32 @var{dev}
20983DINK32 ROM monitor.
20984
20985@kindex target ppcbug
20986@item target ppcbug @var{dev}
20987@kindex target ppcbug1
20988@item target ppcbug1 @var{dev}
20989PPCBUG ROM monitor for PowerPC.
20990
20991@kindex target sds
20992@item target sds @var{dev}
20993SDS monitor, running on a PowerPC board (such as Motorola's ADS).
20994@end table
20995
20996@cindex SDS protocol
20997The following commands specific to the SDS protocol are supported
20998by @value{GDBN}:
20999
21000@table @code
21001@item set sdstimeout @var{nsec}
21002@kindex set sdstimeout
21003Set the timeout for SDS protocol reads to be @var{nsec} seconds. The
21004default is 2 seconds.
21005
21006@item show sdstimeout
21007@kindex show sdstimeout
21008Show the current value of the SDS timeout.
21009
21010@item sds @var{command}
21011@kindex sds@r{, a command}
21012Send the specified @var{command} string to the SDS monitor.
21013@end table
21014
21015
21016@node PA
21017@subsection HP PA Embedded
21018
21019@table @code
21020
21021@kindex target op50n
21022@item target op50n @var{dev}
21023OP50N monitor, running on an OKI HPPA board.
21024
21025@kindex target w89k
21026@item target w89k @var{dev}
21027W89K monitor, running on a Winbond HPPA board.
21028
21029@end table
21030
21031@node Sparclet
21032@subsection Tsqware Sparclet
21033
21034@cindex Sparclet
21035
21036@value{GDBN} enables developers to debug tasks running on
21037Sparclet targets from a Unix host.
21038@value{GDBN} uses code that runs on
21039both the Unix host and on the Sparclet target. The program
21040@code{@value{GDBP}} is installed and executed on the Unix host.
21041
21042@table @code
21043@item remotetimeout @var{args}
21044@kindex remotetimeout
21045@value{GDBN} supports the option @code{remotetimeout}.
21046This option is set by the user, and @var{args} represents the number of
21047seconds @value{GDBN} waits for responses.
21048@end table
21049
21050@cindex compiling, on Sparclet
21051When compiling for debugging, include the options @samp{-g} to get debug
21052information and @samp{-Ttext} to relocate the program to where you wish to
21053load it on the target. You may also want to add the options @samp{-n} or
21054@samp{-N} in order to reduce the size of the sections. Example:
21055
21056@smallexample
21057sparclet-aout-gcc prog.c -Ttext 0x12010000 -g -o prog -N
21058@end smallexample
21059
21060You can use @code{objdump} to verify that the addresses are what you intended:
21061
21062@smallexample
21063sparclet-aout-objdump --headers --syms prog
21064@end smallexample
21065
21066@cindex running, on Sparclet
21067Once you have set
21068your Unix execution search path to find @value{GDBN}, you are ready to
21069run @value{GDBN}. From your Unix host, run @code{@value{GDBP}}
21070(or @code{sparclet-aout-gdb}, depending on your installation).
21071
21072@value{GDBN} comes up showing the prompt:
21073
21074@smallexample
21075(gdbslet)
21076@end smallexample
21077
21078@menu
21079* Sparclet File:: Setting the file to debug
21080* Sparclet Connection:: Connecting to Sparclet
21081* Sparclet Download:: Sparclet download
21082* Sparclet Execution:: Running and debugging
21083@end menu
21084
21085@node Sparclet File
21086@subsubsection Setting File to Debug
21087
21088The @value{GDBN} command @code{file} lets you choose with program to debug.
21089
21090@smallexample
21091(gdbslet) file prog
21092@end smallexample
21093
21094@need 1000
21095@value{GDBN} then attempts to read the symbol table of @file{prog}.
21096@value{GDBN} locates
21097the file by searching the directories listed in the command search
21098path.
21099If the file was compiled with debug information (option @samp{-g}), source
21100files will be searched as well.
21101@value{GDBN} locates
21102the source files by searching the directories listed in the directory search
21103path (@pxref{Environment, ,Your Program's Environment}).
21104If it fails
21105to find a file, it displays a message such as:
21106
21107@smallexample
21108prog: No such file or directory.
21109@end smallexample
21110
21111When this happens, add the appropriate directories to the search paths with
21112the @value{GDBN} commands @code{path} and @code{dir}, and execute the
21113@code{target} command again.
21114
21115@node Sparclet Connection
21116@subsubsection Connecting to Sparclet
21117
21118The @value{GDBN} command @code{target} lets you connect to a Sparclet target.
21119To connect to a target on serial port ``@code{ttya}'', type:
21120
21121@smallexample
21122(gdbslet) target sparclet /dev/ttya
21123Remote target sparclet connected to /dev/ttya
21124main () at ../prog.c:3
21125@end smallexample
21126
21127@need 750
21128@value{GDBN} displays messages like these:
21129
21130@smallexample
21131Connected to ttya.
21132@end smallexample
21133
21134@node Sparclet Download
21135@subsubsection Sparclet Download
21136
21137@cindex download to Sparclet
21138Once connected to the Sparclet target,
21139you can use the @value{GDBN}
21140@code{load} command to download the file from the host to the target.
21141The file name and load offset should be given as arguments to the @code{load}
21142command.
21143Since the file format is aout, the program must be loaded to the starting
21144address. You can use @code{objdump} to find out what this value is. The load
21145offset is an offset which is added to the VMA (virtual memory address)
21146of each of the file's sections.
21147For instance, if the program
21148@file{prog} was linked to text address 0x1201000, with data at 0x12010160
21149and bss at 0x12010170, in @value{GDBN}, type:
21150
21151@smallexample
21152(gdbslet) load prog 0x12010000
21153Loading section .text, size 0xdb0 vma 0x12010000
21154@end smallexample
21155
21156If the code is loaded at a different address then what the program was linked
21157to, you may need to use the @code{section} and @code{add-symbol-file} commands
21158to tell @value{GDBN} where to map the symbol table.
21159
21160@node Sparclet Execution
21161@subsubsection Running and Debugging
21162
21163@cindex running and debugging Sparclet programs
21164You can now begin debugging the task using @value{GDBN}'s execution control
21165commands, @code{b}, @code{step}, @code{run}, etc. See the @value{GDBN}
21166manual for the list of commands.
21167
21168@smallexample
21169(gdbslet) b main
21170Breakpoint 1 at 0x12010000: file prog.c, line 3.
21171(gdbslet) run
21172Starting program: prog
21173Breakpoint 1, main (argc=1, argv=0xeffff21c) at prog.c:3
211743 char *symarg = 0;
21175(gdbslet) step
211764 char *execarg = "hello!";
21177(gdbslet)
21178@end smallexample
21179
21180@node Sparclite
21181@subsection Fujitsu Sparclite
21182
21183@table @code
21184
21185@kindex target sparclite
21186@item target sparclite @var{dev}
21187Fujitsu sparclite boards, used only for the purpose of loading.
21188You must use an additional command to debug the program.
21189For example: target remote @var{dev} using @value{GDBN} standard
21190remote protocol.
21191
21192@end table
21193
21194@node Z8000
21195@subsection Zilog Z8000
21196
21197@cindex Z8000
21198@cindex simulator, Z8000
21199@cindex Zilog Z8000 simulator
21200
21201When configured for debugging Zilog Z8000 targets, @value{GDBN} includes
21202a Z8000 simulator.
21203
21204For the Z8000 family, @samp{target sim} simulates either the Z8002 (the
21205unsegmented variant of the Z8000 architecture) or the Z8001 (the
21206segmented variant). The simulator recognizes which architecture is
21207appropriate by inspecting the object code.
21208
21209@table @code
21210@item target sim @var{args}
21211@kindex sim
21212@kindex target sim@r{, with Z8000}
21213Debug programs on a simulated CPU. If the simulator supports setup
21214options, specify them via @var{args}.
21215@end table
21216
21217@noindent
21218After specifying this target, you can debug programs for the simulated
21219CPU in the same style as programs for your host computer; use the
21220@code{file} command to load a new program image, the @code{run} command
21221to run your program, and so on.
21222
21223As well as making available all the usual machine registers
21224(@pxref{Registers, ,Registers}), the Z8000 simulator provides three
21225additional items of information as specially named registers:
21226
21227@table @code
21228
21229@item cycles
21230Counts clock-ticks in the simulator.
21231
21232@item insts
21233Counts instructions run in the simulator.
21234
21235@item time
21236Execution time in 60ths of a second.
21237
21238@end table
21239
21240You can refer to these values in @value{GDBN} expressions with the usual
21241conventions; for example, @w{@samp{b fputc if $cycles>5000}} sets a
21242conditional breakpoint that suspends only after at least 5000
21243simulated clock ticks.
21244
21245@node AVR
21246@subsection Atmel AVR
21247@cindex AVR
21248
21249When configured for debugging the Atmel AVR, @value{GDBN} supports the
21250following AVR-specific commands:
21251
21252@table @code
21253@item info io_registers
21254@kindex info io_registers@r{, AVR}
21255@cindex I/O registers (Atmel AVR)
21256This command displays information about the AVR I/O registers. For
21257each register, @value{GDBN} prints its number and value.
21258@end table
21259
21260@node CRIS
21261@subsection CRIS
21262@cindex CRIS
21263
21264When configured for debugging CRIS, @value{GDBN} provides the
21265following CRIS-specific commands:
21266
21267@table @code
21268@item set cris-version @var{ver}
21269@cindex CRIS version
21270Set the current CRIS version to @var{ver}, either @samp{10} or @samp{32}.
21271The CRIS version affects register names and sizes. This command is useful in
21272case autodetection of the CRIS version fails.
21273
21274@item show cris-version
21275Show the current CRIS version.
21276
21277@item set cris-dwarf2-cfi
21278@cindex DWARF-2 CFI and CRIS
21279Set the usage of DWARF-2 CFI for CRIS debugging. The default is @samp{on}.
21280Change to @samp{off} when using @code{gcc-cris} whose version is below
21281@code{R59}.
21282
21283@item show cris-dwarf2-cfi
21284Show the current state of using DWARF-2 CFI.
21285
21286@item set cris-mode @var{mode}
21287@cindex CRIS mode
21288Set the current CRIS mode to @var{mode}. It should only be changed when
21289debugging in guru mode, in which case it should be set to
21290@samp{guru} (the default is @samp{normal}).
21291
21292@item show cris-mode
21293Show the current CRIS mode.
21294@end table
21295
21296@node Super-H
21297@subsection Renesas Super-H
21298@cindex Super-H
21299
21300For the Renesas Super-H processor, @value{GDBN} provides these
21301commands:
21302
21303@table @code
21304@item set sh calling-convention @var{convention}
21305@kindex set sh calling-convention
21306Set the calling-convention used when calling functions from @value{GDBN}.
21307Allowed values are @samp{gcc}, which is the default setting, and @samp{renesas}.
21308With the @samp{gcc} setting, functions are called using the @value{NGCC} calling
21309convention. If the DWARF-2 information of the called function specifies
21310that the function follows the Renesas calling convention, the function
21311is called using the Renesas calling convention. If the calling convention
21312is set to @samp{renesas}, the Renesas calling convention is always used,
21313regardless of the DWARF-2 information. This can be used to override the
21314default of @samp{gcc} if debug information is missing, or the compiler
21315does not emit the DWARF-2 calling convention entry for a function.
21316
21317@item show sh calling-convention
21318@kindex show sh calling-convention
21319Show the current calling convention setting.
21320
21321@end table
21322
21323
21324@node Architectures
21325@section Architectures
21326
21327This section describes characteristics of architectures that affect
21328all uses of @value{GDBN} with the architecture, both native and cross.
21329
21330@menu
21331* AArch64::
21332* i386::
21333* Alpha::
21334* MIPS::
21335* HPPA:: HP PA architecture
21336* SPU:: Cell Broadband Engine SPU architecture
21337* PowerPC::
21338* Nios II::
21339@end menu
21340
21341@node AArch64
21342@subsection AArch64
21343@cindex AArch64 support
21344
21345When @value{GDBN} is debugging the AArch64 architecture, it provides the
21346following special commands:
21347
21348@table @code
21349@item set debug aarch64
21350@kindex set debug aarch64
21351This command determines whether AArch64 architecture-specific debugging
21352messages are to be displayed.
21353
21354@item show debug aarch64
21355Show whether AArch64 debugging messages are displayed.
21356
21357@end table
21358
21359@node i386
21360@subsection x86 Architecture-specific Issues
21361
21362@table @code
21363@item set struct-convention @var{mode}
21364@kindex set struct-convention
21365@cindex struct return convention
21366@cindex struct/union returned in registers
21367Set the convention used by the inferior to return @code{struct}s and
21368@code{union}s from functions to @var{mode}. Possible values of
21369@var{mode} are @code{"pcc"}, @code{"reg"}, and @code{"default"} (the
21370default). @code{"default"} or @code{"pcc"} means that @code{struct}s
21371are returned on the stack, while @code{"reg"} means that a
21372@code{struct} or a @code{union} whose size is 1, 2, 4, or 8 bytes will
21373be returned in a register.
21374
21375@item show struct-convention
21376@kindex show struct-convention
21377Show the current setting of the convention to return @code{struct}s
21378from functions.
21379@end table
21380
21381@node Alpha
21382@subsection Alpha
21383
21384See the following section.
21385
21386@node MIPS
21387@subsection @acronym{MIPS}
21388
21389@cindex stack on Alpha
21390@cindex stack on @acronym{MIPS}
21391@cindex Alpha stack
21392@cindex @acronym{MIPS} stack
21393Alpha- and @acronym{MIPS}-based computers use an unusual stack frame, which
21394sometimes requires @value{GDBN} to search backward in the object code to
21395find the beginning of a function.
21396
21397@cindex response time, @acronym{MIPS} debugging
21398To improve response time (especially for embedded applications, where
21399@value{GDBN} may be restricted to a slow serial line for this search)
21400you may want to limit the size of this search, using one of these
21401commands:
21402
21403@table @code
21404@cindex @code{heuristic-fence-post} (Alpha, @acronym{MIPS})
21405@item set heuristic-fence-post @var{limit}
21406Restrict @value{GDBN} to examining at most @var{limit} bytes in its
21407search for the beginning of a function. A value of @var{0} (the
21408default) means there is no limit. However, except for @var{0}, the
21409larger the limit the more bytes @code{heuristic-fence-post} must search
21410and therefore the longer it takes to run. You should only need to use
21411this command when debugging a stripped executable.
21412
21413@item show heuristic-fence-post
21414Display the current limit.
21415@end table
21416
21417@noindent
21418These commands are available @emph{only} when @value{GDBN} is configured
21419for debugging programs on Alpha or @acronym{MIPS} processors.
21420
21421Several @acronym{MIPS}-specific commands are available when debugging @acronym{MIPS}
21422programs:
21423
21424@table @code
21425@item set mips abi @var{arg}
21426@kindex set mips abi
21427@cindex set ABI for @acronym{MIPS}
21428Tell @value{GDBN} which @acronym{MIPS} ABI is used by the inferior. Possible
21429values of @var{arg} are:
21430
21431@table @samp
21432@item auto
21433The default ABI associated with the current binary (this is the
21434default).
21435@item o32
21436@item o64
21437@item n32
21438@item n64
21439@item eabi32
21440@item eabi64
21441@end table
21442
21443@item show mips abi
21444@kindex show mips abi
21445Show the @acronym{MIPS} ABI used by @value{GDBN} to debug the inferior.
21446
21447@item set mips compression @var{arg}
21448@kindex set mips compression
21449@cindex code compression, @acronym{MIPS}
21450Tell @value{GDBN} which @acronym{MIPS} compressed
21451@acronym{ISA, Instruction Set Architecture} encoding is used by the
21452inferior. @value{GDBN} uses this for code disassembly and other
21453internal interpretation purposes. This setting is only referred to
21454when no executable has been associated with the debugging session or
21455the executable does not provide information about the encoding it uses.
21456Otherwise this setting is automatically updated from information
21457provided by the executable.
21458
21459Possible values of @var{arg} are @samp{mips16} and @samp{micromips}.
21460The default compressed @acronym{ISA} encoding is @samp{mips16}, as
21461executables containing @acronym{MIPS16} code frequently are not
21462identified as such.
21463
21464This setting is ``sticky''; that is, it retains its value across
21465debugging sessions until reset either explicitly with this command or
21466implicitly from an executable.
21467
21468The compiler and/or assembler typically add symbol table annotations to
21469identify functions compiled for the @acronym{MIPS16} or
21470@acronym{microMIPS} @acronym{ISA}s. If these function-scope annotations
21471are present, @value{GDBN} uses them in preference to the global
21472compressed @acronym{ISA} encoding setting.
21473
21474@item show mips compression
21475@kindex show mips compression
21476Show the @acronym{MIPS} compressed @acronym{ISA} encoding used by
21477@value{GDBN} to debug the inferior.
21478
21479@item set mipsfpu
21480@itemx show mipsfpu
21481@xref{MIPS Embedded, set mipsfpu}.
21482
21483@item set mips mask-address @var{arg}
21484@kindex set mips mask-address
21485@cindex @acronym{MIPS} addresses, masking
21486This command determines whether the most-significant 32 bits of 64-bit
21487@acronym{MIPS} addresses are masked off. The argument @var{arg} can be
21488@samp{on}, @samp{off}, or @samp{auto}. The latter is the default
21489setting, which lets @value{GDBN} determine the correct value.
21490
21491@item show mips mask-address
21492@kindex show mips mask-address
21493Show whether the upper 32 bits of @acronym{MIPS} addresses are masked off or
21494not.
21495
21496@item set remote-mips64-transfers-32bit-regs
21497@kindex set remote-mips64-transfers-32bit-regs
21498This command controls compatibility with 64-bit @acronym{MIPS} targets that
21499transfer data in 32-bit quantities. If you have an old @acronym{MIPS} 64 target
21500that transfers 32 bits for some registers, like @sc{sr} and @sc{fsr},
21501and 64 bits for other registers, set this option to @samp{on}.
21502
21503@item show remote-mips64-transfers-32bit-regs
21504@kindex show remote-mips64-transfers-32bit-regs
21505Show the current setting of compatibility with older @acronym{MIPS} 64 targets.
21506
21507@item set debug mips
21508@kindex set debug mips
21509This command turns on and off debugging messages for the @acronym{MIPS}-specific
21510target code in @value{GDBN}.
21511
21512@item show debug mips
21513@kindex show debug mips
21514Show the current setting of @acronym{MIPS} debugging messages.
21515@end table
21516
21517
21518@node HPPA
21519@subsection HPPA
21520@cindex HPPA support
21521
21522When @value{GDBN} is debugging the HP PA architecture, it provides the
21523following special commands:
21524
21525@table @code
21526@item set debug hppa
21527@kindex set debug hppa
21528This command determines whether HPPA architecture-specific debugging
21529messages are to be displayed.
21530
21531@item show debug hppa
21532Show whether HPPA debugging messages are displayed.
21533
21534@item maint print unwind @var{address}
21535@kindex maint print unwind@r{, HPPA}
21536This command displays the contents of the unwind table entry at the
21537given @var{address}.
21538
21539@end table
21540
21541
21542@node SPU
21543@subsection Cell Broadband Engine SPU architecture
21544@cindex Cell Broadband Engine
21545@cindex SPU
21546
21547When @value{GDBN} is debugging the Cell Broadband Engine SPU architecture,
21548it provides the following special commands:
21549
21550@table @code
21551@item info spu event
21552@kindex info spu
21553Display SPU event facility status. Shows current event mask
21554and pending event status.
21555
21556@item info spu signal
21557Display SPU signal notification facility status. Shows pending
21558signal-control word and signal notification mode of both signal
21559notification channels.
21560
21561@item info spu mailbox
21562Display SPU mailbox facility status. Shows all pending entries,
21563in order of processing, in each of the SPU Write Outbound,
21564SPU Write Outbound Interrupt, and SPU Read Inbound mailboxes.
21565
21566@item info spu dma
21567Display MFC DMA status. Shows all pending commands in the MFC
21568DMA queue. For each entry, opcode, tag, class IDs, effective
21569and local store addresses and transfer size are shown.
21570
21571@item info spu proxydma
21572Display MFC Proxy-DMA status. Shows all pending commands in the MFC
21573Proxy-DMA queue. For each entry, opcode, tag, class IDs, effective
21574and local store addresses and transfer size are shown.
21575
21576@end table
21577
21578When @value{GDBN} is debugging a combined PowerPC/SPU application
21579on the Cell Broadband Engine, it provides in addition the following
21580special commands:
21581
21582@table @code
21583@item set spu stop-on-load @var{arg}
21584@kindex set spu
21585Set whether to stop for new SPE threads. When set to @code{on}, @value{GDBN}
21586will give control to the user when a new SPE thread enters its @code{main}
21587function. The default is @code{off}.
21588
21589@item show spu stop-on-load
21590@kindex show spu
21591Show whether to stop for new SPE threads.
21592
21593@item set spu auto-flush-cache @var{arg}
21594Set whether to automatically flush the software-managed cache. When set to
21595@code{on}, @value{GDBN} will automatically cause the SPE software-managed
21596cache to be flushed whenever SPE execution stops. This provides a consistent
21597view of PowerPC memory that is accessed via the cache. If an application
21598does not use the software-managed cache, this option has no effect.
21599
21600@item show spu auto-flush-cache
21601Show whether to automatically flush the software-managed cache.
21602
21603@end table
21604
21605@node PowerPC
21606@subsection PowerPC
21607@cindex PowerPC architecture
21608
21609When @value{GDBN} is debugging the PowerPC architecture, it provides a set of
21610pseudo-registers to enable inspection of 128-bit wide Decimal Floating Point
21611numbers stored in the floating point registers. These values must be stored
21612in two consecutive registers, always starting at an even register like
21613@code{f0} or @code{f2}.
21614
21615The pseudo-registers go from @code{$dl0} through @code{$dl15}, and are formed
21616by joining the even/odd register pairs @code{f0} and @code{f1} for @code{$dl0},
21617@code{f2} and @code{f3} for @code{$dl1} and so on.
21618
21619For POWER7 processors, @value{GDBN} provides a set of pseudo-registers, the 64-bit
21620wide Extended Floating Point Registers (@samp{f32} through @samp{f63}).
21621
21622@node Nios II
21623@subsection Nios II
21624@cindex Nios II architecture
21625
21626When @value{GDBN} is debugging the Nios II architecture,
21627it provides the following special commands:
21628
21629@table @code
21630
21631@item set debug nios2
21632@kindex set debug nios2
21633This command turns on and off debugging messages for the Nios II
21634target code in @value{GDBN}.
21635
21636@item show debug nios2
21637@kindex show debug nios2
21638Show the current setting of Nios II debugging messages.
21639@end table
21640
21641@node Controlling GDB
21642@chapter Controlling @value{GDBN}
21643
21644You can alter the way @value{GDBN} interacts with you by using the
21645@code{set} command. For commands controlling how @value{GDBN} displays
21646data, see @ref{Print Settings, ,Print Settings}. Other settings are
21647described here.
21648
21649@menu
21650* Prompt:: Prompt
21651* Editing:: Command editing
21652* Command History:: Command history
21653* Screen Size:: Screen size
21654* Numbers:: Numbers
21655* ABI:: Configuring the current ABI
21656* Auto-loading:: Automatically loading associated files
21657* Messages/Warnings:: Optional warnings and messages
21658* Debugging Output:: Optional messages about internal happenings
21659* Other Misc Settings:: Other Miscellaneous Settings
21660@end menu
21661
21662@node Prompt
21663@section Prompt
21664
21665@cindex prompt
21666
21667@value{GDBN} indicates its readiness to read a command by printing a string
21668called the @dfn{prompt}. This string is normally @samp{(@value{GDBP})}. You
21669can change the prompt string with the @code{set prompt} command. For
21670instance, when debugging @value{GDBN} with @value{GDBN}, it is useful to change
21671the prompt in one of the @value{GDBN} sessions so that you can always tell
21672which one you are talking to.
21673
21674@emph{Note:} @code{set prompt} does not add a space for you after the
21675prompt you set. This allows you to set a prompt which ends in a space
21676or a prompt that does not.
21677
21678@table @code
21679@kindex set prompt
21680@item set prompt @var{newprompt}
21681Directs @value{GDBN} to use @var{newprompt} as its prompt string henceforth.
21682
21683@kindex show prompt
21684@item show prompt
21685Prints a line of the form: @samp{Gdb's prompt is: @var{your-prompt}}
21686@end table
21687
21688Versions of @value{GDBN} that ship with Python scripting enabled have
21689prompt extensions. The commands for interacting with these extensions
21690are:
21691
21692@table @code
21693@kindex set extended-prompt
21694@item set extended-prompt @var{prompt}
21695Set an extended prompt that allows for substitutions.
21696@xref{gdb.prompt}, for a list of escape sequences that can be used for
21697substitution. Any escape sequences specified as part of the prompt
21698string are replaced with the corresponding strings each time the prompt
21699is displayed.
21700
21701For example:
21702
21703@smallexample
21704set extended-prompt Current working directory: \w (gdb)
21705@end smallexample
21706
21707Note that when an extended-prompt is set, it takes control of the
21708@var{prompt_hook} hook. @xref{prompt_hook}, for further information.
21709
21710@kindex show extended-prompt
21711@item show extended-prompt
21712Prints the extended prompt. Any escape sequences specified as part of
21713the prompt string with @code{set extended-prompt}, are replaced with the
21714corresponding strings each time the prompt is displayed.
21715@end table
21716
21717@node Editing
21718@section Command Editing
21719@cindex readline
21720@cindex command line editing
21721
21722@value{GDBN} reads its input commands via the @dfn{Readline} interface. This
21723@sc{gnu} library provides consistent behavior for programs which provide a
21724command line interface to the user. Advantages are @sc{gnu} Emacs-style
21725or @dfn{vi}-style inline editing of commands, @code{csh}-like history
21726substitution, and a storage and recall of command history across
21727debugging sessions.
21728
21729You may control the behavior of command line editing in @value{GDBN} with the
21730command @code{set}.
21731
21732@table @code
21733@kindex set editing
21734@cindex editing
21735@item set editing
21736@itemx set editing on
21737Enable command line editing (enabled by default).
21738
21739@item set editing off
21740Disable command line editing.
21741
21742@kindex show editing
21743@item show editing
21744Show whether command line editing is enabled.
21745@end table
21746
21747@ifset SYSTEM_READLINE
21748@xref{Command Line Editing, , , rluserman, GNU Readline Library},
21749@end ifset
21750@ifclear SYSTEM_READLINE
21751@xref{Command Line Editing},
21752@end ifclear
21753for more details about the Readline
21754interface. Users unfamiliar with @sc{gnu} Emacs or @code{vi} are
21755encouraged to read that chapter.
21756
21757@node Command History
21758@section Command History
21759@cindex command history
21760
21761@value{GDBN} can keep track of the commands you type during your
21762debugging sessions, so that you can be certain of precisely what
21763happened. Use these commands to manage the @value{GDBN} command
21764history facility.
21765
21766@value{GDBN} uses the @sc{gnu} History library, a part of the Readline
21767package, to provide the history facility.
21768@ifset SYSTEM_READLINE
21769@xref{Using History Interactively, , , history, GNU History Library},
21770@end ifset
21771@ifclear SYSTEM_READLINE
21772@xref{Using History Interactively},
21773@end ifclear
21774for the detailed description of the History library.
21775
21776To issue a command to @value{GDBN} without affecting certain aspects of
21777the state which is seen by users, prefix it with @samp{server }
21778(@pxref{Server Prefix}). This
21779means that this command will not affect the command history, nor will it
21780affect @value{GDBN}'s notion of which command to repeat if @key{RET} is
21781pressed on a line by itself.
21782
21783@cindex @code{server}, command prefix
21784The server prefix does not affect the recording of values into the value
21785history; to print a value without recording it into the value history,
21786use the @code{output} command instead of the @code{print} command.
21787
21788Here is the description of @value{GDBN} commands related to command
21789history.
21790
21791@table @code
21792@cindex history substitution
21793@cindex history file
21794@kindex set history filename
21795@cindex @env{GDBHISTFILE}, environment variable
21796@item set history filename @var{fname}
21797Set the name of the @value{GDBN} command history file to @var{fname}.
21798This is the file where @value{GDBN} reads an initial command history
21799list, and where it writes the command history from this session when it
21800exits. You can access this list through history expansion or through
21801the history command editing characters listed below. This file defaults
21802to the value of the environment variable @code{GDBHISTFILE}, or to
21803@file{./.gdb_history} (@file{./_gdb_history} on MS-DOS) if this variable
21804is not set.
21805
21806@cindex save command history
21807@kindex set history save
21808@item set history save
21809@itemx set history save on
21810Record command history in a file, whose name may be specified with the
21811@code{set history filename} command. By default, this option is disabled.
21812
21813@item set history save off
21814Stop recording command history in a file.
21815
21816@cindex history size
21817@kindex set history size
21818@cindex @env{HISTSIZE}, environment variable
21819@item set history size @var{size}
21820@itemx set history size unlimited
21821Set the number of commands which @value{GDBN} keeps in its history list.
21822This defaults to the value of the environment variable
21823@code{HISTSIZE}, or to 256 if this variable is not set. If @var{size}
21824is @code{unlimited}, the number of commands @value{GDBN} keeps in the
21825history list is unlimited.
21826@end table
21827
21828History expansion assigns special meaning to the character @kbd{!}.
21829@ifset SYSTEM_READLINE
21830@xref{Event Designators, , , history, GNU History Library},
21831@end ifset
21832@ifclear SYSTEM_READLINE
21833@xref{Event Designators},
21834@end ifclear
21835for more details.
21836
21837@cindex history expansion, turn on/off
21838Since @kbd{!} is also the logical not operator in C, history expansion
21839is off by default. If you decide to enable history expansion with the
21840@code{set history expansion on} command, you may sometimes need to
21841follow @kbd{!} (when it is used as logical not, in an expression) with
21842a space or a tab to prevent it from being expanded. The readline
21843history facilities do not attempt substitution on the strings
21844@kbd{!=} and @kbd{!(}, even when history expansion is enabled.
21845
21846The commands to control history expansion are:
21847
21848@table @code
21849@item set history expansion on
21850@itemx set history expansion
21851@kindex set history expansion
21852Enable history expansion. History expansion is off by default.
21853
21854@item set history expansion off
21855Disable history expansion.
21856
21857@c @group
21858@kindex show history
21859@item show history
21860@itemx show history filename
21861@itemx show history save
21862@itemx show history size
21863@itemx show history expansion
21864These commands display the state of the @value{GDBN} history parameters.
21865@code{show history} by itself displays all four states.
21866@c @end group
21867@end table
21868
21869@table @code
21870@kindex show commands
21871@cindex show last commands
21872@cindex display command history
21873@item show commands
21874Display the last ten commands in the command history.
21875
21876@item show commands @var{n}
21877Print ten commands centered on command number @var{n}.
21878
21879@item show commands +
21880Print ten commands just after the commands last printed.
21881@end table
21882
21883@node Screen Size
21884@section Screen Size
21885@cindex size of screen
21886@cindex pauses in output
21887
21888Certain commands to @value{GDBN} may produce large amounts of
21889information output to the screen. To help you read all of it,
21890@value{GDBN} pauses and asks you for input at the end of each page of
21891output. Type @key{RET} when you want to continue the output, or @kbd{q}
21892to discard the remaining output. Also, the screen width setting
21893determines when to wrap lines of output. Depending on what is being
21894printed, @value{GDBN} tries to break the line at a readable place,
21895rather than simply letting it overflow onto the following line.
21896
21897Normally @value{GDBN} knows the size of the screen from the terminal
21898driver software. For example, on Unix @value{GDBN} uses the termcap data base
21899together with the value of the @code{TERM} environment variable and the
21900@code{stty rows} and @code{stty cols} settings. If this is not correct,
21901you can override it with the @code{set height} and @code{set
21902width} commands:
21903
21904@table @code
21905@kindex set height
21906@kindex set width
21907@kindex show width
21908@kindex show height
21909@item set height @var{lpp}
21910@itemx set height unlimited
21911@itemx show height
21912@itemx set width @var{cpl}
21913@itemx set width unlimited
21914@itemx show width
21915These @code{set} commands specify a screen height of @var{lpp} lines and
21916a screen width of @var{cpl} characters. The associated @code{show}
21917commands display the current settings.
21918
21919If you specify a height of either @code{unlimited} or zero lines,
21920@value{GDBN} does not pause during output no matter how long the
21921output is. This is useful if output is to a file or to an editor
21922buffer.
21923
21924Likewise, you can specify @samp{set width unlimited} or @samp{set
21925width 0} to prevent @value{GDBN} from wrapping its output.
21926
21927@item set pagination on
21928@itemx set pagination off
21929@kindex set pagination
21930Turn the output pagination on or off; the default is on. Turning
21931pagination off is the alternative to @code{set height unlimited}. Note that
21932running @value{GDBN} with the @option{--batch} option (@pxref{Mode
21933Options, -batch}) also automatically disables pagination.
21934
21935@item show pagination
21936@kindex show pagination
21937Show the current pagination mode.
21938@end table
21939
21940@node Numbers
21941@section Numbers
21942@cindex number representation
21943@cindex entering numbers
21944
21945You can always enter numbers in octal, decimal, or hexadecimal in
21946@value{GDBN} by the usual conventions: octal numbers begin with
21947@samp{0}, decimal numbers end with @samp{.}, and hexadecimal numbers
21948begin with @samp{0x}. Numbers that neither begin with @samp{0} or
21949@samp{0x}, nor end with a @samp{.} are, by default, entered in base
2195010; likewise, the default display for numbers---when no particular
21951format is specified---is base 10. You can change the default base for
21952both input and output with the commands described below.
21953
21954@table @code
21955@kindex set input-radix
21956@item set input-radix @var{base}
21957Set the default base for numeric input. Supported choices
21958for @var{base} are decimal 8, 10, or 16. @var{base} must itself be
21959specified either unambiguously or using the current input radix; for
21960example, any of
21961
21962@smallexample
21963set input-radix 012
21964set input-radix 10.
21965set input-radix 0xa
21966@end smallexample
21967
21968@noindent
21969sets the input base to decimal. On the other hand, @samp{set input-radix 10}
21970leaves the input radix unchanged, no matter what it was, since
21971@samp{10}, being without any leading or trailing signs of its base, is
21972interpreted in the current radix. Thus, if the current radix is 16,
21973@samp{10} is interpreted in hex, i.e.@: as 16 decimal, which doesn't
21974change the radix.
21975
21976@kindex set output-radix
21977@item set output-radix @var{base}
21978Set the default base for numeric display. Supported choices
21979for @var{base} are decimal 8, 10, or 16. @var{base} must itself be
21980specified either unambiguously or using the current input radix.
21981
21982@kindex show input-radix
21983@item show input-radix
21984Display the current default base for numeric input.
21985
21986@kindex show output-radix
21987@item show output-radix
21988Display the current default base for numeric display.
21989
21990@item set radix @r{[}@var{base}@r{]}
21991@itemx show radix
21992@kindex set radix
21993@kindex show radix
21994These commands set and show the default base for both input and output
21995of numbers. @code{set radix} sets the radix of input and output to
21996the same base; without an argument, it resets the radix back to its
21997default value of 10.
21998
21999@end table
22000
22001@node ABI
22002@section Configuring the Current ABI
22003
22004@value{GDBN} can determine the @dfn{ABI} (Application Binary Interface) of your
22005application automatically. However, sometimes you need to override its
22006conclusions. Use these commands to manage @value{GDBN}'s view of the
22007current ABI.
22008
22009@cindex OS ABI
22010@kindex set osabi
22011@kindex show osabi
22012@cindex Newlib OS ABI and its influence on the longjmp handling
22013
22014One @value{GDBN} configuration can debug binaries for multiple operating
22015system targets, either via remote debugging or native emulation.
22016@value{GDBN} will autodetect the @dfn{OS ABI} (Operating System ABI) in use,
22017but you can override its conclusion using the @code{set osabi} command.
22018One example where this is useful is in debugging of binaries which use
22019an alternate C library (e.g.@: @sc{uClibc} for @sc{gnu}/Linux) which does
22020not have the same identifying marks that the standard C library for your
22021platform provides.
22022
22023When @value{GDBN} is debugging the AArch64 architecture, it provides a
22024``Newlib'' OS ABI. This is useful for handling @code{setjmp} and
22025@code{longjmp} when debugging binaries that use the @sc{newlib} C library.
22026The ``Newlib'' OS ABI can be selected by @code{set osabi Newlib}.
22027
22028@table @code
22029@item show osabi
22030Show the OS ABI currently in use.
22031
22032@item set osabi
22033With no argument, show the list of registered available OS ABI's.
22034
22035@item set osabi @var{abi}
22036Set the current OS ABI to @var{abi}.
22037@end table
22038
22039@cindex float promotion
22040
22041Generally, the way that an argument of type @code{float} is passed to a
22042function depends on whether the function is prototyped. For a prototyped
22043(i.e.@: ANSI/ISO style) function, @code{float} arguments are passed unchanged,
22044according to the architecture's convention for @code{float}. For unprototyped
22045(i.e.@: K&R style) functions, @code{float} arguments are first promoted to type
22046@code{double} and then passed.
22047
22048Unfortunately, some forms of debug information do not reliably indicate whether
22049a function is prototyped. If @value{GDBN} calls a function that is not marked
22050as prototyped, it consults @kbd{set coerce-float-to-double}.
22051
22052@table @code
22053@kindex set coerce-float-to-double
22054@item set coerce-float-to-double
22055@itemx set coerce-float-to-double on
22056Arguments of type @code{float} will be promoted to @code{double} when passed
22057to an unprototyped function. This is the default setting.
22058
22059@item set coerce-float-to-double off
22060Arguments of type @code{float} will be passed directly to unprototyped
22061functions.
22062
22063@kindex show coerce-float-to-double
22064@item show coerce-float-to-double
22065Show the current setting of promoting @code{float} to @code{double}.
22066@end table
22067
22068@kindex set cp-abi
22069@kindex show cp-abi
22070@value{GDBN} needs to know the ABI used for your program's C@t{++}
22071objects. The correct C@t{++} ABI depends on which C@t{++} compiler was
22072used to build your application. @value{GDBN} only fully supports
22073programs with a single C@t{++} ABI; if your program contains code using
22074multiple C@t{++} ABI's or if @value{GDBN} can not identify your
22075program's ABI correctly, you can tell @value{GDBN} which ABI to use.
22076Currently supported ABI's include ``gnu-v2'', for @code{g++} versions
22077before 3.0, ``gnu-v3'', for @code{g++} versions 3.0 and later, and
22078``hpaCC'' for the HP ANSI C@t{++} compiler. Other C@t{++} compilers may
22079use the ``gnu-v2'' or ``gnu-v3'' ABI's as well. The default setting is
22080``auto''.
22081
22082@table @code
22083@item show cp-abi
22084Show the C@t{++} ABI currently in use.
22085
22086@item set cp-abi
22087With no argument, show the list of supported C@t{++} ABI's.
22088
22089@item set cp-abi @var{abi}
22090@itemx set cp-abi auto
22091Set the current C@t{++} ABI to @var{abi}, or return to automatic detection.
22092@end table
22093
22094@node Auto-loading
22095@section Automatically loading associated files
22096@cindex auto-loading
22097
22098@value{GDBN} sometimes reads files with commands and settings automatically,
22099without being explicitly told so by the user. We call this feature
22100@dfn{auto-loading}. While auto-loading is useful for automatically adapting
22101@value{GDBN} to the needs of your project, it can sometimes produce unexpected
22102results or introduce security risks (e.g., if the file comes from untrusted
22103sources).
22104
22105Note that loading of these associated files (including the local @file{.gdbinit}
22106file) requires accordingly configured @code{auto-load safe-path}
22107(@pxref{Auto-loading safe path}).
22108
22109For these reasons, @value{GDBN} includes commands and options to let you
22110control when to auto-load files and which files should be auto-loaded.
22111
22112@table @code
22113@anchor{set auto-load off}
22114@kindex set auto-load off
22115@item set auto-load off
22116Globally disable loading of all auto-loaded files.
22117You may want to use this command with the @samp{-iex} option
22118(@pxref{Option -init-eval-command}) such as:
22119@smallexample
22120$ @kbd{gdb -iex "set auto-load off" untrusted-executable corefile}
22121@end smallexample
22122
22123Be aware that system init file (@pxref{System-wide configuration})
22124and init files from your home directory (@pxref{Home Directory Init File})
22125still get read (as they come from generally trusted directories).
22126To prevent @value{GDBN} from auto-loading even those init files, use the
22127@option{-nx} option (@pxref{Mode Options}), in addition to
22128@code{set auto-load no}.
22129
22130@anchor{show auto-load}
22131@kindex show auto-load
22132@item show auto-load
22133Show whether auto-loading of each specific @samp{auto-load} file(s) is enabled
22134or disabled.
22135
22136@smallexample
22137(gdb) show auto-load
22138gdb-scripts: Auto-loading of canned sequences of commands scripts is on.
22139libthread-db: Auto-loading of inferior specific libthread_db is on.
22140local-gdbinit: Auto-loading of .gdbinit script from current directory
22141 is on.
22142python-scripts: Auto-loading of Python scripts is on.
22143safe-path: List of directories from which it is safe to auto-load files
22144 is $debugdir:$datadir/auto-load.
22145scripts-directory: List of directories from which to load auto-loaded scripts
22146 is $debugdir:$datadir/auto-load.
22147@end smallexample
22148
22149@anchor{info auto-load}
22150@kindex info auto-load
22151@item info auto-load
22152Print whether each specific @samp{auto-load} file(s) have been auto-loaded or
22153not.
22154
22155@smallexample
22156(gdb) info auto-load
22157gdb-scripts:
22158Loaded Script
22159Yes /home/user/gdb/gdb-gdb.gdb
22160libthread-db: No auto-loaded libthread-db.
22161local-gdbinit: Local .gdbinit file "/home/user/gdb/.gdbinit" has been
22162 loaded.
22163python-scripts:
22164Loaded Script
22165Yes /home/user/gdb/gdb-gdb.py
22166@end smallexample
22167@end table
22168
22169These are various kinds of files @value{GDBN} can automatically load:
22170
22171@itemize @bullet
22172@item
22173@xref{objfile-gdb.py file}, controlled by @ref{set auto-load python-scripts}.
22174@item
22175@xref{objfile-gdb.gdb file}, controlled by @ref{set auto-load gdb-scripts}.
22176@item
22177@xref{dotdebug_gdb_scripts section},
22178controlled by @ref{set auto-load python-scripts}.
22179@item
22180@xref{Init File in the Current Directory},
22181controlled by @ref{set auto-load local-gdbinit}.
22182@item
22183@xref{libthread_db.so.1 file}, controlled by @ref{set auto-load libthread-db}.
22184@end itemize
22185
22186These are @value{GDBN} control commands for the auto-loading:
22187
22188@multitable @columnfractions .5 .5
22189@item @xref{set auto-load off}.
22190@tab Disable auto-loading globally.
22191@item @xref{show auto-load}.
22192@tab Show setting of all kinds of files.
22193@item @xref{info auto-load}.
22194@tab Show state of all kinds of files.
22195@item @xref{set auto-load gdb-scripts}.
22196@tab Control for @value{GDBN} command scripts.
22197@item @xref{show auto-load gdb-scripts}.
22198@tab Show setting of @value{GDBN} command scripts.
22199@item @xref{info auto-load gdb-scripts}.
22200@tab Show state of @value{GDBN} command scripts.
22201@item @xref{set auto-load python-scripts}.
22202@tab Control for @value{GDBN} Python scripts.
22203@item @xref{show auto-load python-scripts}.
22204@tab Show setting of @value{GDBN} Python scripts.
22205@item @xref{info auto-load python-scripts}.
22206@tab Show state of @value{GDBN} Python scripts.
22207@item @xref{set auto-load scripts-directory}.
22208@tab Control for @value{GDBN} auto-loaded scripts location.
22209@item @xref{show auto-load scripts-directory}.
22210@tab Show @value{GDBN} auto-loaded scripts location.
22211@item @xref{set auto-load local-gdbinit}.
22212@tab Control for init file in the current directory.
22213@item @xref{show auto-load local-gdbinit}.
22214@tab Show setting of init file in the current directory.
22215@item @xref{info auto-load local-gdbinit}.
22216@tab Show state of init file in the current directory.
22217@item @xref{set auto-load libthread-db}.
22218@tab Control for thread debugging library.
22219@item @xref{show auto-load libthread-db}.
22220@tab Show setting of thread debugging library.
22221@item @xref{info auto-load libthread-db}.
22222@tab Show state of thread debugging library.
22223@item @xref{set auto-load safe-path}.
22224@tab Control directories trusted for automatic loading.
22225@item @xref{show auto-load safe-path}.
22226@tab Show directories trusted for automatic loading.
22227@item @xref{add-auto-load-safe-path}.
22228@tab Add directory trusted for automatic loading.
22229@end multitable
22230
22231@menu
22232* Init File in the Current Directory:: @samp{set/show/info auto-load local-gdbinit}
22233* libthread_db.so.1 file:: @samp{set/show/info auto-load libthread-db}
22234* objfile-gdb.gdb file:: @samp{set/show/info auto-load gdb-script}
22235* Auto-loading safe path:: @samp{set/show/info auto-load safe-path}
22236* Auto-loading verbose mode:: @samp{set/show debug auto-load}
22237@xref{Python Auto-loading}.
22238@end menu
22239
22240@node Init File in the Current Directory
22241@subsection Automatically loading init file in the current directory
22242@cindex auto-loading init file in the current directory
22243
22244By default, @value{GDBN} reads and executes the canned sequences of commands
22245from init file (if any) in the current working directory,
22246see @ref{Init File in the Current Directory during Startup}.
22247
22248Note that loading of this local @file{.gdbinit} file also requires accordingly
22249configured @code{auto-load safe-path} (@pxref{Auto-loading safe path}).
22250
22251@table @code
22252@anchor{set auto-load local-gdbinit}
22253@kindex set auto-load local-gdbinit
22254@item set auto-load local-gdbinit [on|off]
22255Enable or disable the auto-loading of canned sequences of commands
22256(@pxref{Sequences}) found in init file in the current directory.
22257
22258@anchor{show auto-load local-gdbinit}
22259@kindex show auto-load local-gdbinit
22260@item show auto-load local-gdbinit
22261Show whether auto-loading of canned sequences of commands from init file in the
22262current directory is enabled or disabled.
22263
22264@anchor{info auto-load local-gdbinit}
22265@kindex info auto-load local-gdbinit
22266@item info auto-load local-gdbinit
22267Print whether canned sequences of commands from init file in the
22268current directory have been auto-loaded.
22269@end table
22270
22271@node libthread_db.so.1 file
22272@subsection Automatically loading thread debugging library
22273@cindex auto-loading libthread_db.so.1
22274
22275This feature is currently present only on @sc{gnu}/Linux native hosts.
22276
22277@value{GDBN} reads in some cases thread debugging library from places specific
22278to the inferior (@pxref{set libthread-db-search-path}).
22279
22280The special @samp{libthread-db-search-path} entry @samp{$sdir} is processed
22281without checking this @samp{set auto-load libthread-db} switch as system
22282libraries have to be trusted in general. In all other cases of
22283@samp{libthread-db-search-path} entries @value{GDBN} checks first if @samp{set
22284auto-load libthread-db} is enabled before trying to open such thread debugging
22285library.
22286
22287Note that loading of this debugging library also requires accordingly configured
22288@code{auto-load safe-path} (@pxref{Auto-loading safe path}).
22289
22290@table @code
22291@anchor{set auto-load libthread-db}
22292@kindex set auto-load libthread-db
22293@item set auto-load libthread-db [on|off]
22294Enable or disable the auto-loading of inferior specific thread debugging library.
22295
22296@anchor{show auto-load libthread-db}
22297@kindex show auto-load libthread-db
22298@item show auto-load libthread-db
22299Show whether auto-loading of inferior specific thread debugging library is
22300enabled or disabled.
22301
22302@anchor{info auto-load libthread-db}
22303@kindex info auto-load libthread-db
22304@item info auto-load libthread-db
22305Print the list of all loaded inferior specific thread debugging libraries and
22306for each such library print list of inferior @var{pid}s using it.
22307@end table
22308
22309@node objfile-gdb.gdb file
22310@subsection The @file{@var{objfile}-gdb.gdb} file
22311@cindex auto-loading @file{@var{objfile}-gdb.gdb}
22312
22313@value{GDBN} tries to load an @file{@var{objfile}-gdb.gdb} file containing
22314canned sequences of commands (@pxref{Sequences}), as long as @samp{set
22315auto-load gdb-scripts} is set to @samp{on}.
22316
22317Note that loading of this script file also requires accordingly configured
22318@code{auto-load safe-path} (@pxref{Auto-loading safe path}).
22319
22320For more background refer to the similar Python scripts auto-loading
22321description (@pxref{objfile-gdb.py file}).
22322
22323@table @code
22324@anchor{set auto-load gdb-scripts}
22325@kindex set auto-load gdb-scripts
22326@item set auto-load gdb-scripts [on|off]
22327Enable or disable the auto-loading of canned sequences of commands scripts.
22328
22329@anchor{show auto-load gdb-scripts}
22330@kindex show auto-load gdb-scripts
22331@item show auto-load gdb-scripts
22332Show whether auto-loading of canned sequences of commands scripts is enabled or
22333disabled.
22334
22335@anchor{info auto-load gdb-scripts}
22336@kindex info auto-load gdb-scripts
22337@cindex print list of auto-loaded canned sequences of commands scripts
22338@item info auto-load gdb-scripts [@var{regexp}]
22339Print the list of all canned sequences of commands scripts that @value{GDBN}
22340auto-loaded.
22341@end table
22342
22343If @var{regexp} is supplied only canned sequences of commands scripts with
22344matching names are printed.
22345
22346@node Auto-loading safe path
22347@subsection Security restriction for auto-loading
22348@cindex auto-loading safe-path
22349
22350As the files of inferior can come from untrusted source (such as submitted by
22351an application user) @value{GDBN} does not always load any files automatically.
22352@value{GDBN} provides the @samp{set auto-load safe-path} setting to list
22353directories trusted for loading files not explicitly requested by user.
22354Each directory can also be a shell wildcard pattern.
22355
22356If the path is not set properly you will see a warning and the file will not
22357get loaded:
22358
22359@smallexample
22360$ ./gdb -q ./gdb
22361Reading symbols from /home/user/gdb/gdb...done.
22362warning: File "/home/user/gdb/gdb-gdb.gdb" auto-loading has been
22363 declined by your `auto-load safe-path' set
22364 to "$debugdir:$datadir/auto-load".
22365warning: File "/home/user/gdb/gdb-gdb.py" auto-loading has been
22366 declined by your `auto-load safe-path' set
22367 to "$debugdir:$datadir/auto-load".
22368@end smallexample
22369
22370@noindent
22371To instruct @value{GDBN} to go ahead and use the init files anyway,
22372invoke @value{GDBN} like this:
22373
22374@smallexample
22375$ gdb -q -iex "set auto-load safe-path /home/user/gdb" ./gdb
22376@end smallexample
22377
22378The list of trusted directories is controlled by the following commands:
22379
22380@table @code
22381@anchor{set auto-load safe-path}
22382@kindex set auto-load safe-path
22383@item set auto-load safe-path @r{[}@var{directories}@r{]}
22384Set the list of directories (and their subdirectories) trusted for automatic
22385loading and execution of scripts. You can also enter a specific trusted file.
22386Each directory can also be a shell wildcard pattern; wildcards do not match
22387directory separator - see @code{FNM_PATHNAME} for system function @code{fnmatch}
22388(@pxref{Wildcard Matching, fnmatch, , libc, GNU C Library Reference Manual}).
22389If you omit @var{directories}, @samp{auto-load safe-path} will be reset to
22390its default value as specified during @value{GDBN} compilation.
22391
22392The list of directories uses path separator (@samp{:} on GNU and Unix
22393systems, @samp{;} on MS-Windows and MS-DOS) to separate directories, similarly
22394to the @env{PATH} environment variable.
22395
22396@anchor{show auto-load safe-path}
22397@kindex show auto-load safe-path
22398@item show auto-load safe-path
22399Show the list of directories trusted for automatic loading and execution of
22400scripts.
22401
22402@anchor{add-auto-load-safe-path}
22403@kindex add-auto-load-safe-path
22404@item add-auto-load-safe-path
22405Add an entry (or list of entries) the list of directories trusted for automatic
22406loading and execution of scripts. Multiple entries may be delimited by the
22407host platform path separator in use.
22408@end table
22409
22410This variable defaults to what @code{--with-auto-load-dir} has been configured
22411to (@pxref{with-auto-load-dir}). @file{$debugdir} and @file{$datadir}
22412substitution applies the same as for @ref{set auto-load scripts-directory}.
22413The default @code{set auto-load safe-path} value can be also overriden by
22414@value{GDBN} configuration option @option{--with-auto-load-safe-path}.
22415
22416Setting this variable to @file{/} disables this security protection,
22417corresponding @value{GDBN} configuration option is
22418@option{--without-auto-load-safe-path}.
22419This variable is supposed to be set to the system directories writable by the
22420system superuser only. Users can add their source directories in init files in
22421their home directories (@pxref{Home Directory Init File}). See also deprecated
22422init file in the current directory
22423(@pxref{Init File in the Current Directory during Startup}).
22424
22425To force @value{GDBN} to load the files it declined to load in the previous
22426example, you could use one of the following ways:
22427
22428@table @asis
22429@item @file{~/.gdbinit}: @samp{add-auto-load-safe-path ~/src/gdb}
22430Specify this trusted directory (or a file) as additional component of the list.
22431You have to specify also any existing directories displayed by
22432by @samp{show auto-load safe-path} (such as @samp{/usr:/bin} in this example).
22433
22434@item @kbd{gdb -iex "set auto-load safe-path /usr:/bin:~/src/gdb" @dots{}}
22435Specify this directory as in the previous case but just for a single
22436@value{GDBN} session.
22437
22438@item @kbd{gdb -iex "set auto-load safe-path /" @dots{}}
22439Disable auto-loading safety for a single @value{GDBN} session.
22440This assumes all the files you debug during this @value{GDBN} session will come
22441from trusted sources.
22442
22443@item @kbd{./configure --without-auto-load-safe-path}
22444During compilation of @value{GDBN} you may disable any auto-loading safety.
22445This assumes all the files you will ever debug with this @value{GDBN} come from
22446trusted sources.
22447@end table
22448
22449On the other hand you can also explicitly forbid automatic files loading which
22450also suppresses any such warning messages:
22451
22452@table @asis
22453@item @kbd{gdb -iex "set auto-load no" @dots{}}
22454You can use @value{GDBN} command-line option for a single @value{GDBN} session.
22455
22456@item @file{~/.gdbinit}: @samp{set auto-load no}
22457Disable auto-loading globally for the user
22458(@pxref{Home Directory Init File}). While it is improbable, you could also
22459use system init file instead (@pxref{System-wide configuration}).
22460@end table
22461
22462This setting applies to the file names as entered by user. If no entry matches
22463@value{GDBN} tries as a last resort to also resolve all the file names into
22464their canonical form (typically resolving symbolic links) and compare the
22465entries again. @value{GDBN} already canonicalizes most of the filenames on its
22466own before starting the comparison so a canonical form of directories is
22467recommended to be entered.
22468
22469@node Auto-loading verbose mode
22470@subsection Displaying files tried for auto-load
22471@cindex auto-loading verbose mode
22472
22473For better visibility of all the file locations where you can place scripts to
22474be auto-loaded with inferior --- or to protect yourself against accidental
22475execution of untrusted scripts --- @value{GDBN} provides a feature for printing
22476all the files attempted to be loaded. Both existing and non-existing files may
22477be printed.
22478
22479For example the list of directories from which it is safe to auto-load files
22480(@pxref{Auto-loading safe path}) applies also to canonicalized filenames which
22481may not be too obvious while setting it up.
22482
22483@smallexample
22484(gdb) set debug auto-load on
22485(gdb) file ~/src/t/true
22486auto-load: Loading canned sequences of commands script "/tmp/true-gdb.gdb"
22487 for objfile "/tmp/true".
22488auto-load: Updating directories of "/usr:/opt".
22489auto-load: Using directory "/usr".
22490auto-load: Using directory "/opt".
22491warning: File "/tmp/true-gdb.gdb" auto-loading has been declined
22492 by your `auto-load safe-path' set to "/usr:/opt".
22493@end smallexample
22494
22495@table @code
22496@anchor{set debug auto-load}
22497@kindex set debug auto-load
22498@item set debug auto-load [on|off]
22499Set whether to print the filenames attempted to be auto-loaded.
22500
22501@anchor{show debug auto-load}
22502@kindex show debug auto-load
22503@item show debug auto-load
22504Show whether printing of the filenames attempted to be auto-loaded is turned
22505on or off.
22506@end table
22507
22508@node Messages/Warnings
22509@section Optional Warnings and Messages
22510
22511@cindex verbose operation
22512@cindex optional warnings
22513By default, @value{GDBN} is silent about its inner workings. If you are
22514running on a slow machine, you may want to use the @code{set verbose}
22515command. This makes @value{GDBN} tell you when it does a lengthy
22516internal operation, so you will not think it has crashed.
22517
22518Currently, the messages controlled by @code{set verbose} are those
22519which announce that the symbol table for a source file is being read;
22520see @code{symbol-file} in @ref{Files, ,Commands to Specify Files}.
22521
22522@table @code
22523@kindex set verbose
22524@item set verbose on
22525Enables @value{GDBN} output of certain informational messages.
22526
22527@item set verbose off
22528Disables @value{GDBN} output of certain informational messages.
22529
22530@kindex show verbose
22531@item show verbose
22532Displays whether @code{set verbose} is on or off.
22533@end table
22534
22535By default, if @value{GDBN} encounters bugs in the symbol table of an
22536object file, it is silent; but if you are debugging a compiler, you may
22537find this information useful (@pxref{Symbol Errors, ,Errors Reading
22538Symbol Files}).
22539
22540@table @code
22541
22542@kindex set complaints
22543@item set complaints @var{limit}
22544Permits @value{GDBN} to output @var{limit} complaints about each type of
22545unusual symbols before becoming silent about the problem. Set
22546@var{limit} to zero to suppress all complaints; set it to a large number
22547to prevent complaints from being suppressed.
22548
22549@kindex show complaints
22550@item show complaints
22551Displays how many symbol complaints @value{GDBN} is permitted to produce.
22552
22553@end table
22554
22555@anchor{confirmation requests}
22556By default, @value{GDBN} is cautious, and asks what sometimes seems to be a
22557lot of stupid questions to confirm certain commands. For example, if
22558you try to run a program which is already running:
22559
22560@smallexample
22561(@value{GDBP}) run
22562The program being debugged has been started already.
22563Start it from the beginning? (y or n)
22564@end smallexample
22565
22566If you are willing to unflinchingly face the consequences of your own
22567commands, you can disable this ``feature'':
22568
22569@table @code
22570
22571@kindex set confirm
22572@cindex flinching
22573@cindex confirmation
22574@cindex stupid questions
22575@item set confirm off
22576Disables confirmation requests. Note that running @value{GDBN} with
22577the @option{--batch} option (@pxref{Mode Options, -batch}) also
22578automatically disables confirmation requests.
22579
22580@item set confirm on
22581Enables confirmation requests (the default).
22582
22583@kindex show confirm
22584@item show confirm
22585Displays state of confirmation requests.
22586
22587@end table
22588
22589@cindex command tracing
22590If you need to debug user-defined commands or sourced files you may find it
22591useful to enable @dfn{command tracing}. In this mode each command will be
22592printed as it is executed, prefixed with one or more @samp{+} symbols, the
22593quantity denoting the call depth of each command.
22594
22595@table @code
22596@kindex set trace-commands
22597@cindex command scripts, debugging
22598@item set trace-commands on
22599Enable command tracing.
22600@item set trace-commands off
22601Disable command tracing.
22602@item show trace-commands
22603Display the current state of command tracing.
22604@end table
22605
22606@node Debugging Output
22607@section Optional Messages about Internal Happenings
22608@cindex optional debugging messages
22609
22610@value{GDBN} has commands that enable optional debugging messages from
22611various @value{GDBN} subsystems; normally these commands are of
22612interest to @value{GDBN} maintainers, or when reporting a bug. This
22613section documents those commands.
22614
22615@table @code
22616@kindex set exec-done-display
22617@item set exec-done-display
22618Turns on or off the notification of asynchronous commands'
22619completion. When on, @value{GDBN} will print a message when an
22620asynchronous command finishes its execution. The default is off.
22621@kindex show exec-done-display
22622@item show exec-done-display
22623Displays the current setting of asynchronous command completion
22624notification.
22625@kindex set debug
22626@cindex ARM AArch64
22627@item set debug aarch64
22628Turns on or off display of debugging messages related to ARM AArch64.
22629The default is off.
22630@kindex show debug
22631@item show debug aarch64
22632Displays the current state of displaying debugging messages related to
22633ARM AArch64.
22634@cindex gdbarch debugging info
22635@cindex architecture debugging info
22636@item set debug arch
22637Turns on or off display of gdbarch debugging info. The default is off
22638@item show debug arch
22639Displays the current state of displaying gdbarch debugging info.
22640@item set debug aix-solib
22641@cindex AIX shared library debugging
22642Control display of debugging messages from the AIX shared library
22643support module. The default is off.
22644@item show debug aix-thread
22645Show the current state of displaying AIX shared library debugging messages.
22646@item set debug aix-thread
22647@cindex AIX threads
22648Display debugging messages about inner workings of the AIX thread
22649module.
22650@item show debug aix-thread
22651Show the current state of AIX thread debugging info display.
22652@item set debug check-physname
22653@cindex physname
22654Check the results of the ``physname'' computation. When reading DWARF
22655debugging information for C@t{++}, @value{GDBN} attempts to compute
22656each entity's name. @value{GDBN} can do this computation in two
22657different ways, depending on exactly what information is present.
22658When enabled, this setting causes @value{GDBN} to compute the names
22659both ways and display any discrepancies.
22660@item show debug check-physname
22661Show the current state of ``physname'' checking.
22662@item set debug coff-pe-read
22663@cindex COFF/PE exported symbols
22664Control display of debugging messages related to reading of COFF/PE
22665exported symbols. The default is off.
22666@item show debug coff-pe-read
22667Displays the current state of displaying debugging messages related to
22668reading of COFF/PE exported symbols.
22669@item set debug dwarf2-die
22670@cindex DWARF2 DIEs
22671Dump DWARF2 DIEs after they are read in.
22672The value is the number of nesting levels to print.
22673A value of zero turns off the display.
22674@item show debug dwarf2-die
22675Show the current state of DWARF2 DIE debugging.
22676@item set debug dwarf2-read
22677@cindex DWARF2 Reading
22678Turns on or off display of debugging messages related to reading
22679DWARF debug info. The default is 0 (off).
22680A value of 1 provides basic information.
22681A value greater than 1 provides more verbose information.
22682@item show debug dwarf2-read
22683Show the current state of DWARF2 reader debugging.
22684@item set debug displaced
22685@cindex displaced stepping debugging info
22686Turns on or off display of @value{GDBN} debugging info for the
22687displaced stepping support. The default is off.
22688@item show debug displaced
22689Displays the current state of displaying @value{GDBN} debugging info
22690related to displaced stepping.
22691@item set debug event
22692@cindex event debugging info
22693Turns on or off display of @value{GDBN} event debugging info. The
22694default is off.
22695@item show debug event
22696Displays the current state of displaying @value{GDBN} event debugging
22697info.
22698@item set debug expression
22699@cindex expression debugging info
22700Turns on or off display of debugging info about @value{GDBN}
22701expression parsing. The default is off.
22702@item show debug expression
22703Displays the current state of displaying debugging info about
22704@value{GDBN} expression parsing.
22705@item set debug frame
22706@cindex frame debugging info
22707Turns on or off display of @value{GDBN} frame debugging info. The
22708default is off.
22709@item show debug frame
22710Displays the current state of displaying @value{GDBN} frame debugging
22711info.
22712@item set debug gnu-nat
22713@cindex @sc{gnu}/Hurd debug messages
22714Turns on or off debugging messages from the @sc{gnu}/Hurd debug support.
22715@item show debug gnu-nat
22716Show the current state of @sc{gnu}/Hurd debugging messages.
22717@item set debug infrun
22718@cindex inferior debugging info
22719Turns on or off display of @value{GDBN} debugging info for running the inferior.
22720The default is off. @file{infrun.c} contains GDB's runtime state machine used
22721for implementing operations such as single-stepping the inferior.
22722@item show debug infrun
22723Displays the current state of @value{GDBN} inferior debugging.
22724@item set debug jit
22725@cindex just-in-time compilation, debugging messages
22726Turns on or off debugging messages from JIT debug support.
22727@item show debug jit
22728Displays the current state of @value{GDBN} JIT debugging.
22729@item set debug lin-lwp
22730@cindex @sc{gnu}/Linux LWP debug messages
22731@cindex Linux lightweight processes
22732Turns on or off debugging messages from the Linux LWP debug support.
22733@item show debug lin-lwp
22734Show the current state of Linux LWP debugging messages.
22735@item set debug mach-o
22736@cindex Mach-O symbols processing
22737Control display of debugging messages related to Mach-O symbols
22738processing. The default is off.
22739@item show debug mach-o
22740Displays the current state of displaying debugging messages related to
22741reading of COFF/PE exported symbols.
22742@item set debug notification
22743@cindex remote async notification debugging info
22744Turns on or off debugging messages about remote async notification.
22745The default is off.
22746@item show debug notification
22747Displays the current state of remote async notification debugging messages.
22748@item set debug observer
22749@cindex observer debugging info
22750Turns on or off display of @value{GDBN} observer debugging. This
22751includes info such as the notification of observable events.
22752@item show debug observer
22753Displays the current state of observer debugging.
22754@item set debug overload
22755@cindex C@t{++} overload debugging info
22756Turns on or off display of @value{GDBN} C@t{++} overload debugging
22757info. This includes info such as ranking of functions, etc. The default
22758is off.
22759@item show debug overload
22760Displays the current state of displaying @value{GDBN} C@t{++} overload
22761debugging info.
22762@cindex expression parser, debugging info
22763@cindex debug expression parser
22764@item set debug parser
22765Turns on or off the display of expression parser debugging output.
22766Internally, this sets the @code{yydebug} variable in the expression
22767parser. @xref{Tracing, , Tracing Your Parser, bison, Bison}, for
22768details. The default is off.
22769@item show debug parser
22770Show the current state of expression parser debugging.
22771@cindex packets, reporting on stdout
22772@cindex serial connections, debugging
22773@cindex debug remote protocol
22774@cindex remote protocol debugging
22775@cindex display remote packets
22776@item set debug remote
22777Turns on or off display of reports on all packets sent back and forth across
22778the serial line to the remote machine. The info is printed on the
22779@value{GDBN} standard output stream. The default is off.
22780@item show debug remote
22781Displays the state of display of remote packets.
22782@item set debug serial
22783Turns on or off display of @value{GDBN} serial debugging info. The
22784default is off.
22785@item show debug serial
22786Displays the current state of displaying @value{GDBN} serial debugging
22787info.
22788@item set debug solib-frv
22789@cindex FR-V shared-library debugging
22790Turns on or off debugging messages for FR-V shared-library code.
22791@item show debug solib-frv
22792Display the current state of FR-V shared-library code debugging
22793messages.
22794@item set debug symfile
22795@cindex symbol file functions
22796Turns on or off display of debugging messages related to symbol file functions.
22797The default is off. @xref{Files}.
22798@item show debug symfile
22799Show the current state of symbol file debugging messages.
22800@item set debug symtab-create
22801@cindex symbol table creation
22802Turns on or off display of debugging messages related to symbol table creation.
22803The default is 0 (off).
22804A value of 1 provides basic information.
22805A value greater than 1 provides more verbose information.
22806@item show debug symtab-create
22807Show the current state of symbol table creation debugging.
22808@item set debug target
22809@cindex target debugging info
22810Turns on or off display of @value{GDBN} target debugging info. This info
22811includes what is going on at the target level of GDB, as it happens. The
22812default is 0. Set it to 1 to track events, and to 2 to also track the
22813value of large memory transfers. Changes to this flag do not take effect
22814until the next time you connect to a target or use the @code{run} command.
22815@item show debug target
22816Displays the current state of displaying @value{GDBN} target debugging
22817info.
22818@item set debug timestamp
22819@cindex timestampping debugging info
22820Turns on or off display of timestamps with @value{GDBN} debugging info.
22821When enabled, seconds and microseconds are displayed before each debugging
22822message.
22823@item show debug timestamp
22824Displays the current state of displaying timestamps with @value{GDBN}
22825debugging info.
22826@item set debugvarobj
22827@cindex variable object debugging info
22828Turns on or off display of @value{GDBN} variable object debugging
22829info. The default is off.
22830@item show debugvarobj
22831Displays the current state of displaying @value{GDBN} variable object
22832debugging info.
22833@item set debug xml
22834@cindex XML parser debugging
22835Turns on or off debugging messages for built-in XML parsers.
22836@item show debug xml
22837Displays the current state of XML debugging messages.
22838@end table
22839
22840@node Other Misc Settings
22841@section Other Miscellaneous Settings
22842@cindex miscellaneous settings
22843
22844@table @code
22845@kindex set interactive-mode
22846@item set interactive-mode
22847If @code{on}, forces @value{GDBN} to assume that GDB was started
22848in a terminal. In practice, this means that @value{GDBN} should wait
22849for the user to answer queries generated by commands entered at
22850the command prompt. If @code{off}, forces @value{GDBN} to operate
22851in the opposite mode, and it uses the default answers to all queries.
22852If @code{auto} (the default), @value{GDBN} tries to determine whether
22853its standard input is a terminal, and works in interactive-mode if it
22854is, non-interactively otherwise.
22855
22856In the vast majority of cases, the debugger should be able to guess
22857correctly which mode should be used. But this setting can be useful
22858in certain specific cases, such as running a MinGW @value{GDBN}
22859inside a cygwin window.
22860
22861@kindex show interactive-mode
22862@item show interactive-mode
22863Displays whether the debugger is operating in interactive mode or not.
22864@end table
22865
22866@node Extending GDB
22867@chapter Extending @value{GDBN}
22868@cindex extending GDB
22869
22870@value{GDBN} provides three mechanisms for extension. The first is based
22871on composition of @value{GDBN} commands, the second is based on the
22872Python scripting language, and the third is for defining new aliases of
22873existing commands.
22874
22875To facilitate the use of the first two extensions, @value{GDBN} is capable
22876of evaluating the contents of a file. When doing so, @value{GDBN}
22877can recognize which scripting language is being used by looking at
22878the filename extension. Files with an unrecognized filename extension
22879are always treated as a @value{GDBN} Command Files.
22880@xref{Command Files,, Command files}.
22881
22882You can control how @value{GDBN} evaluates these files with the following
22883setting:
22884
22885@table @code
22886@kindex set script-extension
22887@kindex show script-extension
22888@item set script-extension off
22889All scripts are always evaluated as @value{GDBN} Command Files.
22890
22891@item set script-extension soft
22892The debugger determines the scripting language based on filename
22893extension. If this scripting language is supported, @value{GDBN}
22894evaluates the script using that language. Otherwise, it evaluates
22895the file as a @value{GDBN} Command File.
22896
22897@item set script-extension strict
22898The debugger determines the scripting language based on filename
22899extension, and evaluates the script using that language. If the
22900language is not supported, then the evaluation fails.
22901
22902@item show script-extension
22903Display the current value of the @code{script-extension} option.
22904
22905@end table
22906
22907@menu
22908* Sequences:: Canned Sequences of Commands
22909* Python:: Scripting @value{GDBN} using Python
22910* Aliases:: Creating new spellings of existing commands
22911@end menu
22912
22913@node Sequences
22914@section Canned Sequences of Commands
22915
22916Aside from breakpoint commands (@pxref{Break Commands, ,Breakpoint
22917Command Lists}), @value{GDBN} provides two ways to store sequences of
22918commands for execution as a unit: user-defined commands and command
22919files.
22920
22921@menu
22922* Define:: How to define your own commands
22923* Hooks:: Hooks for user-defined commands
22924* Command Files:: How to write scripts of commands to be stored in a file
22925* Output:: Commands for controlled output
22926@end menu
22927
22928@node Define
22929@subsection User-defined Commands
22930
22931@cindex user-defined command
22932@cindex arguments, to user-defined commands
22933A @dfn{user-defined command} is a sequence of @value{GDBN} commands to
22934which you assign a new name as a command. This is done with the
22935@code{define} command. User commands may accept up to 10 arguments
22936separated by whitespace. Arguments are accessed within the user command
22937via @code{$arg0@dots{}$arg9}. A trivial example:
22938
22939@smallexample
22940define adder
22941 print $arg0 + $arg1 + $arg2
22942end
22943@end smallexample
22944
22945@noindent
22946To execute the command use:
22947
22948@smallexample
22949adder 1 2 3
22950@end smallexample
22951
22952@noindent
22953This defines the command @code{adder}, which prints the sum of
22954its three arguments. Note the arguments are text substitutions, so they may
22955reference variables, use complex expressions, or even perform inferior
22956functions calls.
22957
22958@cindex argument count in user-defined commands
22959@cindex how many arguments (user-defined commands)
22960In addition, @code{$argc} may be used to find out how many arguments have
22961been passed. This expands to a number in the range 0@dots{}10.
22962
22963@smallexample
22964define adder
22965 if $argc == 2
22966 print $arg0 + $arg1
22967 end
22968 if $argc == 3
22969 print $arg0 + $arg1 + $arg2
22970 end
22971end
22972@end smallexample
22973
22974@table @code
22975
22976@kindex define
22977@item define @var{commandname}
22978Define a command named @var{commandname}. If there is already a command
22979by that name, you are asked to confirm that you want to redefine it.
22980@var{commandname} may be a bare command name consisting of letters,
22981numbers, dashes, and underscores. It may also start with any predefined
22982prefix command. For example, @samp{define target my-target} creates
22983a user-defined @samp{target my-target} command.
22984
22985The definition of the command is made up of other @value{GDBN} command lines,
22986which are given following the @code{define} command. The end of these
22987commands is marked by a line containing @code{end}.
22988
22989@kindex document
22990@kindex end@r{ (user-defined commands)}
22991@item document @var{commandname}
22992Document the user-defined command @var{commandname}, so that it can be
22993accessed by @code{help}. The command @var{commandname} must already be
22994defined. This command reads lines of documentation just as @code{define}
22995reads the lines of the command definition, ending with @code{end}.
22996After the @code{document} command is finished, @code{help} on command
22997@var{commandname} displays the documentation you have written.
22998
22999You may use the @code{document} command again to change the
23000documentation of a command. Redefining the command with @code{define}
23001does not change the documentation.
23002
23003@kindex dont-repeat
23004@cindex don't repeat command
23005@item dont-repeat
23006Used inside a user-defined command, this tells @value{GDBN} that this
23007command should not be repeated when the user hits @key{RET}
23008(@pxref{Command Syntax, repeat last command}).
23009
23010@kindex help user-defined
23011@item help user-defined
23012List all user-defined commands and all python commands defined in class
23013COMAND_USER. The first line of the documentation or docstring is
23014included (if any).
23015
23016@kindex show user
23017@item show user
23018@itemx show user @var{commandname}
23019Display the @value{GDBN} commands used to define @var{commandname} (but
23020not its documentation). If no @var{commandname} is given, display the
23021definitions for all user-defined commands.
23022This does not work for user-defined python commands.
23023
23024@cindex infinite recursion in user-defined commands
23025@kindex show max-user-call-depth
23026@kindex set max-user-call-depth
23027@item show max-user-call-depth
23028@itemx set max-user-call-depth
23029The value of @code{max-user-call-depth} controls how many recursion
23030levels are allowed in user-defined commands before @value{GDBN} suspects an
23031infinite recursion and aborts the command.
23032This does not apply to user-defined python commands.
23033@end table
23034
23035In addition to the above commands, user-defined commands frequently
23036use control flow commands, described in @ref{Command Files}.
23037
23038When user-defined commands are executed, the
23039commands of the definition are not printed. An error in any command
23040stops execution of the user-defined command.
23041
23042If used interactively, commands that would ask for confirmation proceed
23043without asking when used inside a user-defined command. Many @value{GDBN}
23044commands that normally print messages to say what they are doing omit the
23045messages when used in a user-defined command.
23046
23047@node Hooks
23048@subsection User-defined Command Hooks
23049@cindex command hooks
23050@cindex hooks, for commands
23051@cindex hooks, pre-command
23052
23053@kindex hook
23054You may define @dfn{hooks}, which are a special kind of user-defined
23055command. Whenever you run the command @samp{foo}, if the user-defined
23056command @samp{hook-foo} exists, it is executed (with no arguments)
23057before that command.
23058
23059@cindex hooks, post-command
23060@kindex hookpost
23061A hook may also be defined which is run after the command you executed.
23062Whenever you run the command @samp{foo}, if the user-defined command
23063@samp{hookpost-foo} exists, it is executed (with no arguments) after
23064that command. Post-execution hooks may exist simultaneously with
23065pre-execution hooks, for the same command.
23066
23067It is valid for a hook to call the command which it hooks. If this
23068occurs, the hook is not re-executed, thereby avoiding infinite recursion.
23069
23070@c It would be nice if hookpost could be passed a parameter indicating
23071@c if the command it hooks executed properly or not. FIXME!
23072
23073@kindex stop@r{, a pseudo-command}
23074In addition, a pseudo-command, @samp{stop} exists. Defining
23075(@samp{hook-stop}) makes the associated commands execute every time
23076execution stops in your program: before breakpoint commands are run,
23077displays are printed, or the stack frame is printed.
23078
23079For example, to ignore @code{SIGALRM} signals while
23080single-stepping, but treat them normally during normal execution,
23081you could define:
23082
23083@smallexample
23084define hook-stop
23085handle SIGALRM nopass
23086end
23087
23088define hook-run
23089handle SIGALRM pass
23090end
23091
23092define hook-continue
23093handle SIGALRM pass
23094end
23095@end smallexample
23096
23097As a further example, to hook at the beginning and end of the @code{echo}
23098command, and to add extra text to the beginning and end of the message,
23099you could define:
23100
23101@smallexample
23102define hook-echo
23103echo <<<---
23104end
23105
23106define hookpost-echo
23107echo --->>>\n
23108end
23109
23110(@value{GDBP}) echo Hello World
23111<<<---Hello World--->>>
23112(@value{GDBP})
23113
23114@end smallexample
23115
23116You can define a hook for any single-word command in @value{GDBN}, but
23117not for command aliases; you should define a hook for the basic command
23118name, e.g.@: @code{backtrace} rather than @code{bt}.
23119@c FIXME! So how does Joe User discover whether a command is an alias
23120@c or not?
23121You can hook a multi-word command by adding @code{hook-} or
23122@code{hookpost-} to the last word of the command, e.g.@:
23123@samp{define target hook-remote} to add a hook to @samp{target remote}.
23124
23125If an error occurs during the execution of your hook, execution of
23126@value{GDBN} commands stops and @value{GDBN} issues a prompt
23127(before the command that you actually typed had a chance to run).
23128
23129If you try to define a hook which does not match any known command, you
23130get a warning from the @code{define} command.
23131
23132@node Command Files
23133@subsection Command Files
23134
23135@cindex command files
23136@cindex scripting commands
23137A command file for @value{GDBN} is a text file made of lines that are
23138@value{GDBN} commands. Comments (lines starting with @kbd{#}) may
23139also be included. An empty line in a command file does nothing; it
23140does not mean to repeat the last command, as it would from the
23141terminal.
23142
23143You can request the execution of a command file with the @code{source}
23144command. Note that the @code{source} command is also used to evaluate
23145scripts that are not Command Files. The exact behavior can be configured
23146using the @code{script-extension} setting.
23147@xref{Extending GDB,, Extending GDB}.
23148
23149@table @code
23150@kindex source
23151@cindex execute commands from a file
23152@item source [-s] [-v] @var{filename}
23153Execute the command file @var{filename}.
23154@end table
23155
23156The lines in a command file are generally executed sequentially,
23157unless the order of execution is changed by one of the
23158@emph{flow-control commands} described below. The commands are not
23159printed as they are executed. An error in any command terminates
23160execution of the command file and control is returned to the console.
23161
23162@value{GDBN} first searches for @var{filename} in the current directory.
23163If the file is not found there, and @var{filename} does not specify a
23164directory, then @value{GDBN} also looks for the file on the source search path
23165(specified with the @samp{directory} command);
23166except that @file{$cdir} is not searched because the compilation directory
23167is not relevant to scripts.
23168
23169If @code{-s} is specified, then @value{GDBN} searches for @var{filename}
23170on the search path even if @var{filename} specifies a directory.
23171The search is done by appending @var{filename} to each element of the
23172search path. So, for example, if @var{filename} is @file{mylib/myscript}
23173and the search path contains @file{/home/user} then @value{GDBN} will
23174look for the script @file{/home/user/mylib/myscript}.
23175The search is also done if @var{filename} is an absolute path.
23176For example, if @var{filename} is @file{/tmp/myscript} and
23177the search path contains @file{/home/user} then @value{GDBN} will
23178look for the script @file{/home/user/tmp/myscript}.
23179For DOS-like systems, if @var{filename} contains a drive specification,
23180it is stripped before concatenation. For example, if @var{filename} is
23181@file{d:myscript} and the search path contains @file{c:/tmp} then @value{GDBN}
23182will look for the script @file{c:/tmp/myscript}.
23183
23184If @code{-v}, for verbose mode, is given then @value{GDBN} displays
23185each command as it is executed. The option must be given before
23186@var{filename}, and is interpreted as part of the filename anywhere else.
23187
23188Commands that would ask for confirmation if used interactively proceed
23189without asking when used in a command file. Many @value{GDBN} commands that
23190normally print messages to say what they are doing omit the messages
23191when called from command files.
23192
23193@value{GDBN} also accepts command input from standard input. In this
23194mode, normal output goes to standard output and error output goes to
23195standard error. Errors in a command file supplied on standard input do
23196not terminate execution of the command file---execution continues with
23197the next command.
23198
23199@smallexample
23200gdb < cmds > log 2>&1
23201@end smallexample
23202
23203(The syntax above will vary depending on the shell used.) This example
23204will execute commands from the file @file{cmds}. All output and errors
23205would be directed to @file{log}.
23206
23207Since commands stored on command files tend to be more general than
23208commands typed interactively, they frequently need to deal with
23209complicated situations, such as different or unexpected values of
23210variables and symbols, changes in how the program being debugged is
23211built, etc. @value{GDBN} provides a set of flow-control commands to
23212deal with these complexities. Using these commands, you can write
23213complex scripts that loop over data structures, execute commands
23214conditionally, etc.
23215
23216@table @code
23217@kindex if
23218@kindex else
23219@item if
23220@itemx else
23221This command allows to include in your script conditionally executed
23222commands. The @code{if} command takes a single argument, which is an
23223expression to evaluate. It is followed by a series of commands that
23224are executed only if the expression is true (its value is nonzero).
23225There can then optionally be an @code{else} line, followed by a series
23226of commands that are only executed if the expression was false. The
23227end of the list is marked by a line containing @code{end}.
23228
23229@kindex while
23230@item while
23231This command allows to write loops. Its syntax is similar to
23232@code{if}: the command takes a single argument, which is an expression
23233to evaluate, and must be followed by the commands to execute, one per
23234line, terminated by an @code{end}. These commands are called the
23235@dfn{body} of the loop. The commands in the body of @code{while} are
23236executed repeatedly as long as the expression evaluates to true.
23237
23238@kindex loop_break
23239@item loop_break
23240This command exits the @code{while} loop in whose body it is included.
23241Execution of the script continues after that @code{while}s @code{end}
23242line.
23243
23244@kindex loop_continue
23245@item loop_continue
23246This command skips the execution of the rest of the body of commands
23247in the @code{while} loop in whose body it is included. Execution
23248branches to the beginning of the @code{while} loop, where it evaluates
23249the controlling expression.
23250
23251@kindex end@r{ (if/else/while commands)}
23252@item end
23253Terminate the block of commands that are the body of @code{if},
23254@code{else}, or @code{while} flow-control commands.
23255@end table
23256
23257
23258@node Output
23259@subsection Commands for Controlled Output
23260
23261During the execution of a command file or a user-defined command, normal
23262@value{GDBN} output is suppressed; the only output that appears is what is
23263explicitly printed by the commands in the definition. This section
23264describes three commands useful for generating exactly the output you
23265want.
23266
23267@table @code
23268@kindex echo
23269@item echo @var{text}
23270@c I do not consider backslash-space a standard C escape sequence
23271@c because it is not in ANSI.
23272Print @var{text}. Nonprinting characters can be included in
23273@var{text} using C escape sequences, such as @samp{\n} to print a
23274newline. @strong{No newline is printed unless you specify one.}
23275In addition to the standard C escape sequences, a backslash followed
23276by a space stands for a space. This is useful for displaying a
23277string with spaces at the beginning or the end, since leading and
23278trailing spaces are otherwise trimmed from all arguments.
23279To print @samp{@w{ }and foo =@w{ }}, use the command
23280@samp{echo \@w{ }and foo = \@w{ }}.
23281
23282A backslash at the end of @var{text} can be used, as in C, to continue
23283the command onto subsequent lines. For example,
23284
23285@smallexample
23286echo This is some text\n\
23287which is continued\n\
23288onto several lines.\n
23289@end smallexample
23290
23291produces the same output as
23292
23293@smallexample
23294echo This is some text\n
23295echo which is continued\n
23296echo onto several lines.\n
23297@end smallexample
23298
23299@kindex output
23300@item output @var{expression}
23301Print the value of @var{expression} and nothing but that value: no
23302newlines, no @samp{$@var{nn} = }. The value is not entered in the
23303value history either. @xref{Expressions, ,Expressions}, for more information
23304on expressions.
23305
23306@item output/@var{fmt} @var{expression}
23307Print the value of @var{expression} in format @var{fmt}. You can use
23308the same formats as for @code{print}. @xref{Output Formats,,Output
23309Formats}, for more information.
23310
23311@kindex printf
23312@item printf @var{template}, @var{expressions}@dots{}
23313Print the values of one or more @var{expressions} under the control of
23314the string @var{template}. To print several values, make
23315@var{expressions} be a comma-separated list of individual expressions,
23316which may be either numbers or pointers. Their values are printed as
23317specified by @var{template}, exactly as a C program would do by
23318executing the code below:
23319
23320@smallexample
23321printf (@var{template}, @var{expressions}@dots{});
23322@end smallexample
23323
23324As in @code{C} @code{printf}, ordinary characters in @var{template}
23325are printed verbatim, while @dfn{conversion specification} introduced
23326by the @samp{%} character cause subsequent @var{expressions} to be
23327evaluated, their values converted and formatted according to type and
23328style information encoded in the conversion specifications, and then
23329printed.
23330
23331For example, you can print two values in hex like this:
23332
23333@smallexample
23334printf "foo, bar-foo = 0x%x, 0x%x\n", foo, bar-foo
23335@end smallexample
23336
23337@code{printf} supports all the standard @code{C} conversion
23338specifications, including the flags and modifiers between the @samp{%}
23339character and the conversion letter, with the following exceptions:
23340
23341@itemize @bullet
23342@item
23343The argument-ordering modifiers, such as @samp{2$}, are not supported.
23344
23345@item
23346The modifier @samp{*} is not supported for specifying precision or
23347width.
23348
23349@item
23350The @samp{'} flag (for separation of digits into groups according to
23351@code{LC_NUMERIC'}) is not supported.
23352
23353@item
23354The type modifiers @samp{hh}, @samp{j}, @samp{t}, and @samp{z} are not
23355supported.
23356
23357@item
23358The conversion letter @samp{n} (as in @samp{%n}) is not supported.
23359
23360@item
23361The conversion letters @samp{a} and @samp{A} are not supported.
23362@end itemize
23363
23364@noindent
23365Note that the @samp{ll} type modifier is supported only if the
23366underlying @code{C} implementation used to build @value{GDBN} supports
23367the @code{long long int} type, and the @samp{L} type modifier is
23368supported only if @code{long double} type is available.
23369
23370As in @code{C}, @code{printf} supports simple backslash-escape
23371sequences, such as @code{\n}, @samp{\t}, @samp{\\}, @samp{\"},
23372@samp{\a}, and @samp{\f}, that consist of backslash followed by a
23373single character. Octal and hexadecimal escape sequences are not
23374supported.
23375
23376Additionally, @code{printf} supports conversion specifications for DFP
23377(@dfn{Decimal Floating Point}) types using the following length modifiers
23378together with a floating point specifier.
23379letters:
23380
23381@itemize @bullet
23382@item
23383@samp{H} for printing @code{Decimal32} types.
23384
23385@item
23386@samp{D} for printing @code{Decimal64} types.
23387
23388@item
23389@samp{DD} for printing @code{Decimal128} types.
23390@end itemize
23391
23392If the underlying @code{C} implementation used to build @value{GDBN} has
23393support for the three length modifiers for DFP types, other modifiers
23394such as width and precision will also be available for @value{GDBN} to use.
23395
23396In case there is no such @code{C} support, no additional modifiers will be
23397available and the value will be printed in the standard way.
23398
23399Here's an example of printing DFP types using the above conversion letters:
23400@smallexample
23401printf "D32: %Hf - D64: %Df - D128: %DDf\n",1.2345df,1.2E10dd,1.2E1dl
23402@end smallexample
23403
23404@kindex eval
23405@item eval @var{template}, @var{expressions}@dots{}
23406Convert the values of one or more @var{expressions} under the control of
23407the string @var{template} to a command line, and call it.
23408
23409@end table
23410
23411@node Python
23412@section Scripting @value{GDBN} using Python
23413@cindex python scripting
23414@cindex scripting with python
23415
23416You can script @value{GDBN} using the @uref{http://www.python.org/,
23417Python programming language}. This feature is available only if
23418@value{GDBN} was configured using @option{--with-python}.
23419
23420@cindex python directory
23421Python scripts used by @value{GDBN} should be installed in
23422@file{@var{data-directory}/python}, where @var{data-directory} is
23423the data directory as determined at @value{GDBN} startup (@pxref{Data Files}).
23424This directory, known as the @dfn{python directory},
23425is automatically added to the Python Search Path in order to allow
23426the Python interpreter to locate all scripts installed at this location.
23427
23428Additionally, @value{GDBN} commands and convenience functions which
23429are written in Python and are located in the
23430@file{@var{data-directory}/python/gdb/command} or
23431@file{@var{data-directory}/python/gdb/function} directories are
23432automatically imported when @value{GDBN} starts.
23433
23434@menu
23435* Python Commands:: Accessing Python from @value{GDBN}.
23436* Python API:: Accessing @value{GDBN} from Python.
23437* Python Auto-loading:: Automatically loading Python code.
23438* Python modules:: Python modules provided by @value{GDBN}.
23439@end menu
23440
23441@node Python Commands
23442@subsection Python Commands
23443@cindex python commands
23444@cindex commands to access python
23445
23446@value{GDBN} provides two commands for accessing the Python interpreter,
23447and one related setting:
23448
23449@table @code
23450@kindex python-interactive
23451@kindex pi
23452@item python-interactive @r{[}@var{command}@r{]}
23453@itemx pi @r{[}@var{command}@r{]}
23454Without an argument, the @code{python-interactive} command can be used
23455to start an interactive Python prompt. To return to @value{GDBN},
23456type the @code{EOF} character (e.g., @kbd{Ctrl-D} on an empty prompt).
23457
23458Alternatively, a single-line Python command can be given as an
23459argument and evaluated. If the command is an expression, the result
23460will be printed; otherwise, nothing will be printed. For example:
23461
23462@smallexample
23463(@value{GDBP}) python-interactive 2 + 3
234645
23465@end smallexample
23466
23467@kindex python
23468@kindex py
23469@item python @r{[}@var{command}@r{]}
23470@itemx py @r{[}@var{command}@r{]}
23471The @code{python} command can be used to evaluate Python code.
23472
23473If given an argument, the @code{python} command will evaluate the
23474argument as a Python command. For example:
23475
23476@smallexample
23477(@value{GDBP}) python print 23
2347823
23479@end smallexample
23480
23481If you do not provide an argument to @code{python}, it will act as a
23482multi-line command, like @code{define}. In this case, the Python
23483script is made up of subsequent command lines, given after the
23484@code{python} command. This command list is terminated using a line
23485containing @code{end}. For example:
23486
23487@smallexample
23488(@value{GDBP}) python
23489Type python script
23490End with a line saying just "end".
23491>print 23
23492>end
2349323
23494@end smallexample
23495
23496@kindex set python print-stack
23497@item set python print-stack
23498By default, @value{GDBN} will print only the message component of a
23499Python exception when an error occurs in a Python script. This can be
23500controlled using @code{set python print-stack}: if @code{full}, then
23501full Python stack printing is enabled; if @code{none}, then Python stack
23502and message printing is disabled; if @code{message}, the default, only
23503the message component of the error is printed.
23504@end table
23505
23506It is also possible to execute a Python script from the @value{GDBN}
23507interpreter:
23508
23509@table @code
23510@item source @file{script-name}
23511The script name must end with @samp{.py} and @value{GDBN} must be configured
23512to recognize the script language based on filename extension using
23513the @code{script-extension} setting. @xref{Extending GDB, ,Extending GDB}.
23514
23515@item python execfile ("script-name")
23516This method is based on the @code{execfile} Python built-in function,
23517and thus is always available.
23518@end table
23519
23520@node Python API
23521@subsection Python API
23522@cindex python api
23523@cindex programming in python
23524
23525You can get quick online help for @value{GDBN}'s Python API by issuing
23526the command @w{@kbd{python help (gdb)}}.
23527
23528Functions and methods which have two or more optional arguments allow
23529them to be specified using keyword syntax. This allows passing some
23530optional arguments while skipping others. Example:
23531@w{@code{gdb.some_function ('foo', bar = 1, baz = 2)}}.
23532
23533@menu
23534* Basic Python:: Basic Python Functions.
23535* Exception Handling:: How Python exceptions are translated.
23536* Values From Inferior:: Python representation of values.
23537* Types In Python:: Python representation of types.
23538* Pretty Printing API:: Pretty-printing values.
23539* Selecting Pretty-Printers:: How GDB chooses a pretty-printer.
23540* Writing a Pretty-Printer:: Writing a Pretty-Printer.
23541* Type Printing API:: Pretty-printing types.
23542* Frame Filter API:: Filtering Frames.
23543* Frame Decorator API:: Decorating Frames.
23544* Writing a Frame Filter:: Writing a Frame Filter.
23545* Inferiors In Python:: Python representation of inferiors (processes)
23546* Events In Python:: Listening for events from @value{GDBN}.
23547* Threads In Python:: Accessing inferior threads from Python.
23548* Commands In Python:: Implementing new commands in Python.
23549* Parameters In Python:: Adding new @value{GDBN} parameters.
23550* Functions In Python:: Writing new convenience functions.
23551* Progspaces In Python:: Program spaces.
23552* Objfiles In Python:: Object files.
23553* Frames In Python:: Accessing inferior stack frames from Python.
23554* Blocks In Python:: Accessing blocks from Python.
23555* Symbols In Python:: Python representation of symbols.
23556* Symbol Tables In Python:: Python representation of symbol tables.
23557* Line Tables In Python:: Python representation of line tables.
23558* Breakpoints In Python:: Manipulating breakpoints using Python.
23559* Finish Breakpoints in Python:: Setting Breakpoints on function return
23560 using Python.
23561* Lazy Strings In Python:: Python representation of lazy strings.
23562* Architectures In Python:: Python representation of architectures.
23563@end menu
23564
23565@node Basic Python
23566@subsubsection Basic Python
23567
23568@cindex python stdout
23569@cindex python pagination
23570At startup, @value{GDBN} overrides Python's @code{sys.stdout} and
23571@code{sys.stderr} to print using @value{GDBN}'s output-paging streams.
23572A Python program which outputs to one of these streams may have its
23573output interrupted by the user (@pxref{Screen Size}). In this
23574situation, a Python @code{KeyboardInterrupt} exception is thrown.
23575
23576Some care must be taken when writing Python code to run in
23577@value{GDBN}. Two things worth noting in particular:
23578
23579@itemize @bullet
23580@item
23581@value{GDBN} install handlers for @code{SIGCHLD} and @code{SIGINT}.
23582Python code must not override these, or even change the options using
23583@code{sigaction}. If your program changes the handling of these
23584signals, @value{GDBN} will most likely stop working correctly. Note
23585that it is unfortunately common for GUI toolkits to install a
23586@code{SIGCHLD} handler.
23587
23588@item
23589@value{GDBN} takes care to mark its internal file descriptors as
23590close-on-exec. However, this cannot be done in a thread-safe way on
23591all platforms. Your Python programs should be aware of this and
23592should both create new file descriptors with the close-on-exec flag
23593set and arrange to close unneeded file descriptors before starting a
23594child process.
23595@end itemize
23596
23597@cindex python functions
23598@cindex python module
23599@cindex gdb module
23600@value{GDBN} introduces a new Python module, named @code{gdb}. All
23601methods and classes added by @value{GDBN} are placed in this module.
23602@value{GDBN} automatically @code{import}s the @code{gdb} module for
23603use in all scripts evaluated by the @code{python} command.
23604
23605@findex gdb.PYTHONDIR
23606@defvar gdb.PYTHONDIR
23607A string containing the python directory (@pxref{Python}).
23608@end defvar
23609
23610@findex gdb.execute
23611@defun gdb.execute (command @r{[}, from_tty @r{[}, to_string@r{]]})
23612Evaluate @var{command}, a string, as a @value{GDBN} CLI command.
23613If a GDB exception happens while @var{command} runs, it is
23614translated as described in @ref{Exception Handling,,Exception Handling}.
23615
23616@var{from_tty} specifies whether @value{GDBN} ought to consider this
23617command as having originated from the user invoking it interactively.
23618It must be a boolean value. If omitted, it defaults to @code{False}.
23619
23620By default, any output produced by @var{command} is sent to
23621@value{GDBN}'s standard output. If the @var{to_string} parameter is
23622@code{True}, then output will be collected by @code{gdb.execute} and
23623returned as a string. The default is @code{False}, in which case the
23624return value is @code{None}. If @var{to_string} is @code{True}, the
23625@value{GDBN} virtual terminal will be temporarily set to unlimited width
23626and height, and its pagination will be disabled; @pxref{Screen Size}.
23627@end defun
23628
23629@findex gdb.breakpoints
23630@defun gdb.breakpoints ()
23631Return a sequence holding all of @value{GDBN}'s breakpoints.
23632@xref{Breakpoints In Python}, for more information.
23633@end defun
23634
23635@findex gdb.parameter
23636@defun gdb.parameter (parameter)
23637Return the value of a @value{GDBN} parameter. @var{parameter} is a
23638string naming the parameter to look up; @var{parameter} may contain
23639spaces if the parameter has a multi-part name. For example,
23640@samp{print object} is a valid parameter name.
23641
23642If the named parameter does not exist, this function throws a
23643@code{gdb.error} (@pxref{Exception Handling}). Otherwise, the
23644parameter's value is converted to a Python value of the appropriate
23645type, and returned.
23646@end defun
23647
23648@findex gdb.history
23649@defun gdb.history (number)
23650Return a value from @value{GDBN}'s value history (@pxref{Value
23651History}). @var{number} indicates which history element to return.
23652If @var{number} is negative, then @value{GDBN} will take its absolute value
23653and count backward from the last element (i.e., the most recent element) to
23654find the value to return. If @var{number} is zero, then @value{GDBN} will
23655return the most recent element. If the element specified by @var{number}
23656doesn't exist in the value history, a @code{gdb.error} exception will be
23657raised.
23658
23659If no exception is raised, the return value is always an instance of
23660@code{gdb.Value} (@pxref{Values From Inferior}).
23661@end defun
23662
23663@findex gdb.parse_and_eval
23664@defun gdb.parse_and_eval (expression)
23665Parse @var{expression} as an expression in the current language,
23666evaluate it, and return the result as a @code{gdb.Value}.
23667@var{expression} must be a string.
23668
23669This function can be useful when implementing a new command
23670(@pxref{Commands In Python}), as it provides a way to parse the
23671command's argument as an expression. It is also useful simply to
23672compute values, for example, it is the only way to get the value of a
23673convenience variable (@pxref{Convenience Vars}) as a @code{gdb.Value}.
23674@end defun
23675
23676@findex gdb.find_pc_line
23677@defun gdb.find_pc_line (pc)
23678Return the @code{gdb.Symtab_and_line} object corresponding to the
23679@var{pc} value. @xref{Symbol Tables In Python}. If an invalid
23680value of @var{pc} is passed as an argument, then the @code{symtab} and
23681@code{line} attributes of the returned @code{gdb.Symtab_and_line} object
23682will be @code{None} and 0 respectively.
23683@end defun
23684
23685@findex gdb.post_event
23686@defun gdb.post_event (event)
23687Put @var{event}, a callable object taking no arguments, into
23688@value{GDBN}'s internal event queue. This callable will be invoked at
23689some later point, during @value{GDBN}'s event processing. Events
23690posted using @code{post_event} will be run in the order in which they
23691were posted; however, there is no way to know when they will be
23692processed relative to other events inside @value{GDBN}.
23693
23694@value{GDBN} is not thread-safe. If your Python program uses multiple
23695threads, you must be careful to only call @value{GDBN}-specific
23696functions in the main @value{GDBN} thread. @code{post_event} ensures
23697this. For example:
23698
23699@smallexample
23700(@value{GDBP}) python
23701>import threading
23702>
23703>class Writer():
23704> def __init__(self, message):
23705> self.message = message;
23706> def __call__(self):
23707> gdb.write(self.message)
23708>
23709>class MyThread1 (threading.Thread):
23710> def run (self):
23711> gdb.post_event(Writer("Hello "))
23712>
23713>class MyThread2 (threading.Thread):
23714> def run (self):
23715> gdb.post_event(Writer("World\n"))
23716>
23717>MyThread1().start()
23718>MyThread2().start()
23719>end
23720(@value{GDBP}) Hello World
23721@end smallexample
23722@end defun
23723
23724@findex gdb.write
23725@defun gdb.write (string @r{[}, stream{]})
23726Print a string to @value{GDBN}'s paginated output stream. The
23727optional @var{stream} determines the stream to print to. The default
23728stream is @value{GDBN}'s standard output stream. Possible stream
23729values are:
23730
23731@table @code
23732@findex STDOUT
23733@findex gdb.STDOUT
23734@item gdb.STDOUT
23735@value{GDBN}'s standard output stream.
23736
23737@findex STDERR
23738@findex gdb.STDERR
23739@item gdb.STDERR
23740@value{GDBN}'s standard error stream.
23741
23742@findex STDLOG
23743@findex gdb.STDLOG
23744@item gdb.STDLOG
23745@value{GDBN}'s log stream (@pxref{Logging Output}).
23746@end table
23747
23748Writing to @code{sys.stdout} or @code{sys.stderr} will automatically
23749call this function and will automatically direct the output to the
23750relevant stream.
23751@end defun
23752
23753@findex gdb.flush
23754@defun gdb.flush ()
23755Flush the buffer of a @value{GDBN} paginated stream so that the
23756contents are displayed immediately. @value{GDBN} will flush the
23757contents of a stream automatically when it encounters a newline in the
23758buffer. The optional @var{stream} determines the stream to flush. The
23759default stream is @value{GDBN}'s standard output stream. Possible
23760stream values are:
23761
23762@table @code
23763@findex STDOUT
23764@findex gdb.STDOUT
23765@item gdb.STDOUT
23766@value{GDBN}'s standard output stream.
23767
23768@findex STDERR
23769@findex gdb.STDERR
23770@item gdb.STDERR
23771@value{GDBN}'s standard error stream.
23772
23773@findex STDLOG
23774@findex gdb.STDLOG
23775@item gdb.STDLOG
23776@value{GDBN}'s log stream (@pxref{Logging Output}).
23777
23778@end table
23779
23780Flushing @code{sys.stdout} or @code{sys.stderr} will automatically
23781call this function for the relevant stream.
23782@end defun
23783
23784@findex gdb.target_charset
23785@defun gdb.target_charset ()
23786Return the name of the current target character set (@pxref{Character
23787Sets}). This differs from @code{gdb.parameter('target-charset')} in
23788that @samp{auto} is never returned.
23789@end defun
23790
23791@findex gdb.target_wide_charset
23792@defun gdb.target_wide_charset ()
23793Return the name of the current target wide character set
23794(@pxref{Character Sets}). This differs from
23795@code{gdb.parameter('target-wide-charset')} in that @samp{auto} is
23796never returned.
23797@end defun
23798
23799@findex gdb.solib_name
23800@defun gdb.solib_name (address)
23801Return the name of the shared library holding the given @var{address}
23802as a string, or @code{None}.
23803@end defun
23804
23805@findex gdb.decode_line
23806@defun gdb.decode_line @r{[}expression@r{]}
23807Return locations of the line specified by @var{expression}, or of the
23808current line if no argument was given. This function returns a Python
23809tuple containing two elements. The first element contains a string
23810holding any unparsed section of @var{expression} (or @code{None} if
23811the expression has been fully parsed). The second element contains
23812either @code{None} or another tuple that contains all the locations
23813that match the expression represented as @code{gdb.Symtab_and_line}
23814objects (@pxref{Symbol Tables In Python}). If @var{expression} is
23815provided, it is decoded the way that @value{GDBN}'s inbuilt
23816@code{break} or @code{edit} commands do (@pxref{Specify Location}).
23817@end defun
23818
23819@defun gdb.prompt_hook (current_prompt)
23820@anchor{prompt_hook}
23821
23822If @var{prompt_hook} is callable, @value{GDBN} will call the method
23823assigned to this operation before a prompt is displayed by
23824@value{GDBN}.
23825
23826The parameter @code{current_prompt} contains the current @value{GDBN}
23827prompt. This method must return a Python string, or @code{None}. If
23828a string is returned, the @value{GDBN} prompt will be set to that
23829string. If @code{None} is returned, @value{GDBN} will continue to use
23830the current prompt.
23831
23832Some prompts cannot be substituted in @value{GDBN}. Secondary prompts
23833such as those used by readline for command input, and annotation
23834related prompts are prohibited from being changed.
23835@end defun
23836
23837@node Exception Handling
23838@subsubsection Exception Handling
23839@cindex python exceptions
23840@cindex exceptions, python
23841
23842When executing the @code{python} command, Python exceptions
23843uncaught within the Python code are translated to calls to
23844@value{GDBN} error-reporting mechanism. If the command that called
23845@code{python} does not handle the error, @value{GDBN} will
23846terminate it and print an error message containing the Python
23847exception name, the associated value, and the Python call stack
23848backtrace at the point where the exception was raised. Example:
23849
23850@smallexample
23851(@value{GDBP}) python print foo
23852Traceback (most recent call last):
23853 File "<string>", line 1, in <module>
23854NameError: name 'foo' is not defined
23855@end smallexample
23856
23857@value{GDBN} errors that happen in @value{GDBN} commands invoked by
23858Python code are converted to Python exceptions. The type of the
23859Python exception depends on the error.
23860
23861@ftable @code
23862@item gdb.error
23863This is the base class for most exceptions generated by @value{GDBN}.
23864It is derived from @code{RuntimeError}, for compatibility with earlier
23865versions of @value{GDBN}.
23866
23867If an error occurring in @value{GDBN} does not fit into some more
23868specific category, then the generated exception will have this type.
23869
23870@item gdb.MemoryError
23871This is a subclass of @code{gdb.error} which is thrown when an
23872operation tried to access invalid memory in the inferior.
23873
23874@item KeyboardInterrupt
23875User interrupt (via @kbd{C-c} or by typing @kbd{q} at a pagination
23876prompt) is translated to a Python @code{KeyboardInterrupt} exception.
23877@end ftable
23878
23879In all cases, your exception handler will see the @value{GDBN} error
23880message as its value and the Python call stack backtrace at the Python
23881statement closest to where the @value{GDBN} error occured as the
23882traceback.
23883
23884@findex gdb.GdbError
23885When implementing @value{GDBN} commands in Python via @code{gdb.Command},
23886it is useful to be able to throw an exception that doesn't cause a
23887traceback to be printed. For example, the user may have invoked the
23888command incorrectly. Use the @code{gdb.GdbError} exception
23889to handle this case. Example:
23890
23891@smallexample
23892(gdb) python
23893>class HelloWorld (gdb.Command):
23894> """Greet the whole world."""
23895> def __init__ (self):
23896> super (HelloWorld, self).__init__ ("hello-world", gdb.COMMAND_USER)
23897> def invoke (self, args, from_tty):
23898> argv = gdb.string_to_argv (args)
23899> if len (argv) != 0:
23900> raise gdb.GdbError ("hello-world takes no arguments")
23901> print "Hello, World!"
23902>HelloWorld ()
23903>end
23904(gdb) hello-world 42
23905hello-world takes no arguments
23906@end smallexample
23907
23908@node Values From Inferior
23909@subsubsection Values From Inferior
23910@cindex values from inferior, with Python
23911@cindex python, working with values from inferior
23912
23913@cindex @code{gdb.Value}
23914@value{GDBN} provides values it obtains from the inferior program in
23915an object of type @code{gdb.Value}. @value{GDBN} uses this object
23916for its internal bookkeeping of the inferior's values, and for
23917fetching values when necessary.
23918
23919Inferior values that are simple scalars can be used directly in
23920Python expressions that are valid for the value's data type. Here's
23921an example for an integer or floating-point value @code{some_val}:
23922
23923@smallexample
23924bar = some_val + 2
23925@end smallexample
23926
23927@noindent
23928As result of this, @code{bar} will also be a @code{gdb.Value} object
23929whose values are of the same type as those of @code{some_val}.
23930
23931Inferior values that are structures or instances of some class can
23932be accessed using the Python @dfn{dictionary syntax}. For example, if
23933@code{some_val} is a @code{gdb.Value} instance holding a structure, you
23934can access its @code{foo} element with:
23935
23936@smallexample
23937bar = some_val['foo']
23938@end smallexample
23939
23940Again, @code{bar} will also be a @code{gdb.Value} object.
23941
23942A @code{gdb.Value} that represents a function can be executed via
23943inferior function call. Any arguments provided to the call must match
23944the function's prototype, and must be provided in the order specified
23945by that prototype.
23946
23947For example, @code{some_val} is a @code{gdb.Value} instance
23948representing a function that takes two integers as arguments. To
23949execute this function, call it like so:
23950
23951@smallexample
23952result = some_val (10,20)
23953@end smallexample
23954
23955Any values returned from a function call will be stored as a
23956@code{gdb.Value}.
23957
23958The following attributes are provided:
23959
23960@defvar Value.address
23961If this object is addressable, this read-only attribute holds a
23962@code{gdb.Value} object representing the address. Otherwise,
23963this attribute holds @code{None}.
23964@end defvar
23965
23966@cindex optimized out value in Python
23967@defvar Value.is_optimized_out
23968This read-only boolean attribute is true if the compiler optimized out
23969this value, thus it is not available for fetching from the inferior.
23970@end defvar
23971
23972@defvar Value.type
23973The type of this @code{gdb.Value}. The value of this attribute is a
23974@code{gdb.Type} object (@pxref{Types In Python}).
23975@end defvar
23976
23977@defvar Value.dynamic_type
23978The dynamic type of this @code{gdb.Value}. This uses C@t{++} run-time
23979type information (@acronym{RTTI}) to determine the dynamic type of the
23980value. If this value is of class type, it will return the class in
23981which the value is embedded, if any. If this value is of pointer or
23982reference to a class type, it will compute the dynamic type of the
23983referenced object, and return a pointer or reference to that type,
23984respectively. In all other cases, it will return the value's static
23985type.
23986
23987Note that this feature will only work when debugging a C@t{++} program
23988that includes @acronym{RTTI} for the object in question. Otherwise,
23989it will just return the static type of the value as in @kbd{ptype foo}
23990(@pxref{Symbols, ptype}).
23991@end defvar
23992
23993@defvar Value.is_lazy
23994The value of this read-only boolean attribute is @code{True} if this
23995@code{gdb.Value} has not yet been fetched from the inferior.
23996@value{GDBN} does not fetch values until necessary, for efficiency.
23997For example:
23998
23999@smallexample
24000myval = gdb.parse_and_eval ('somevar')
24001@end smallexample
24002
24003The value of @code{somevar} is not fetched at this time. It will be
24004fetched when the value is needed, or when the @code{fetch_lazy}
24005method is invoked.
24006@end defvar
24007
24008The following methods are provided:
24009
24010@defun Value.__init__ (@var{val})
24011Many Python values can be converted directly to a @code{gdb.Value} via
24012this object initializer. Specifically:
24013
24014@table @asis
24015@item Python boolean
24016A Python boolean is converted to the boolean type from the current
24017language.
24018
24019@item Python integer
24020A Python integer is converted to the C @code{long} type for the
24021current architecture.
24022
24023@item Python long
24024A Python long is converted to the C @code{long long} type for the
24025current architecture.
24026
24027@item Python float
24028A Python float is converted to the C @code{double} type for the
24029current architecture.
24030
24031@item Python string
24032A Python string is converted to a target string, using the current
24033target encoding.
24034
24035@item @code{gdb.Value}
24036If @code{val} is a @code{gdb.Value}, then a copy of the value is made.
24037
24038@item @code{gdb.LazyString}
24039If @code{val} is a @code{gdb.LazyString} (@pxref{Lazy Strings In
24040Python}), then the lazy string's @code{value} method is called, and
24041its result is used.
24042@end table
24043@end defun
24044
24045@defun Value.cast (type)
24046Return a new instance of @code{gdb.Value} that is the result of
24047casting this instance to the type described by @var{type}, which must
24048be a @code{gdb.Type} object. If the cast cannot be performed for some
24049reason, this method throws an exception.
24050@end defun
24051
24052@defun Value.dereference ()
24053For pointer data types, this method returns a new @code{gdb.Value} object
24054whose contents is the object pointed to by the pointer. For example, if
24055@code{foo} is a C pointer to an @code{int}, declared in your C program as
24056
24057@smallexample
24058int *foo;
24059@end smallexample
24060
24061@noindent
24062then you can use the corresponding @code{gdb.Value} to access what
24063@code{foo} points to like this:
24064
24065@smallexample
24066bar = foo.dereference ()
24067@end smallexample
24068
24069The result @code{bar} will be a @code{gdb.Value} object holding the
24070value pointed to by @code{foo}.
24071
24072A similar function @code{Value.referenced_value} exists which also
24073returns @code{gdb.Value} objects corresonding to the values pointed to
24074by pointer values (and additionally, values referenced by reference
24075values). However, the behavior of @code{Value.dereference}
24076differs from @code{Value.referenced_value} by the fact that the
24077behavior of @code{Value.dereference} is identical to applying the C
24078unary operator @code{*} on a given value. For example, consider a
24079reference to a pointer @code{ptrref}, declared in your C@t{++} program
24080as
24081
24082@smallexample
24083typedef int *intptr;
24084...
24085int val = 10;
24086intptr ptr = &val;
24087intptr &ptrref = ptr;
24088@end smallexample
24089
24090Though @code{ptrref} is a reference value, one can apply the method
24091@code{Value.dereference} to the @code{gdb.Value} object corresponding
24092to it and obtain a @code{gdb.Value} which is identical to that
24093corresponding to @code{val}. However, if you apply the method
24094@code{Value.referenced_value}, the result would be a @code{gdb.Value}
24095object identical to that corresponding to @code{ptr}.
24096
24097@smallexample
24098py_ptrref = gdb.parse_and_eval ("ptrref")
24099py_val = py_ptrref.dereference ()
24100py_ptr = py_ptrref.referenced_value ()
24101@end smallexample
24102
24103The @code{gdb.Value} object @code{py_val} is identical to that
24104corresponding to @code{val}, and @code{py_ptr} is identical to that
24105corresponding to @code{ptr}. In general, @code{Value.dereference} can
24106be applied whenever the C unary operator @code{*} can be applied
24107to the corresponding C value. For those cases where applying both
24108@code{Value.dereference} and @code{Value.referenced_value} is allowed,
24109the results obtained need not be identical (as we have seen in the above
24110example). The results are however identical when applied on
24111@code{gdb.Value} objects corresponding to pointers (@code{gdb.Value}
24112objects with type code @code{TYPE_CODE_PTR}) in a C/C@t{++} program.
24113@end defun
24114
24115@defun Value.referenced_value ()
24116For pointer or reference data types, this method returns a new
24117@code{gdb.Value} object corresponding to the value referenced by the
24118pointer/reference value. For pointer data types,
24119@code{Value.dereference} and @code{Value.referenced_value} produce
24120identical results. The difference between these methods is that
24121@code{Value.dereference} cannot get the values referenced by reference
24122values. For example, consider a reference to an @code{int}, declared
24123in your C@t{++} program as
24124
24125@smallexample
24126int val = 10;
24127int &ref = val;
24128@end smallexample
24129
24130@noindent
24131then applying @code{Value.dereference} to the @code{gdb.Value} object
24132corresponding to @code{ref} will result in an error, while applying
24133@code{Value.referenced_value} will result in a @code{gdb.Value} object
24134identical to that corresponding to @code{val}.
24135
24136@smallexample
24137py_ref = gdb.parse_and_eval ("ref")
24138er_ref = py_ref.dereference () # Results in error
24139py_val = py_ref.referenced_value () # Returns the referenced value
24140@end smallexample
24141
24142The @code{gdb.Value} object @code{py_val} is identical to that
24143corresponding to @code{val}.
24144@end defun
24145
24146@defun Value.dynamic_cast (type)
24147Like @code{Value.cast}, but works as if the C@t{++} @code{dynamic_cast}
24148operator were used. Consult a C@t{++} reference for details.
24149@end defun
24150
24151@defun Value.reinterpret_cast (type)
24152Like @code{Value.cast}, but works as if the C@t{++} @code{reinterpret_cast}
24153operator were used. Consult a C@t{++} reference for details.
24154@end defun
24155
24156@defun Value.string (@r{[}encoding@r{[}, errors@r{[}, length@r{]]]})
24157If this @code{gdb.Value} represents a string, then this method
24158converts the contents to a Python string. Otherwise, this method will
24159throw an exception.
24160
24161Strings are recognized in a language-specific way; whether a given
24162@code{gdb.Value} represents a string is determined by the current
24163language.
24164
24165For C-like languages, a value is a string if it is a pointer to or an
24166array of characters or ints. The string is assumed to be terminated
24167by a zero of the appropriate width. However if the optional length
24168argument is given, the string will be converted to that given length,
24169ignoring any embedded zeros that the string may contain.
24170
24171If the optional @var{encoding} argument is given, it must be a string
24172naming the encoding of the string in the @code{gdb.Value}, such as
24173@code{"ascii"}, @code{"iso-8859-6"} or @code{"utf-8"}. It accepts
24174the same encodings as the corresponding argument to Python's
24175@code{string.decode} method, and the Python codec machinery will be used
24176to convert the string. If @var{encoding} is not given, or if
24177@var{encoding} is the empty string, then either the @code{target-charset}
24178(@pxref{Character Sets}) will be used, or a language-specific encoding
24179will be used, if the current language is able to supply one.
24180
24181The optional @var{errors} argument is the same as the corresponding
24182argument to Python's @code{string.decode} method.
24183
24184If the optional @var{length} argument is given, the string will be
24185fetched and converted to the given length.
24186@end defun
24187
24188@defun Value.lazy_string (@r{[}encoding @r{[}, length@r{]]})
24189If this @code{gdb.Value} represents a string, then this method
24190converts the contents to a @code{gdb.LazyString} (@pxref{Lazy Strings
24191In Python}). Otherwise, this method will throw an exception.
24192
24193If the optional @var{encoding} argument is given, it must be a string
24194naming the encoding of the @code{gdb.LazyString}. Some examples are:
24195@samp{ascii}, @samp{iso-8859-6} or @samp{utf-8}. If the
24196@var{encoding} argument is an encoding that @value{GDBN} does
24197recognize, @value{GDBN} will raise an error.
24198
24199When a lazy string is printed, the @value{GDBN} encoding machinery is
24200used to convert the string during printing. If the optional
24201@var{encoding} argument is not provided, or is an empty string,
24202@value{GDBN} will automatically select the encoding most suitable for
24203the string type. For further information on encoding in @value{GDBN}
24204please see @ref{Character Sets}.
24205
24206If the optional @var{length} argument is given, the string will be
24207fetched and encoded to the length of characters specified. If
24208the @var{length} argument is not provided, the string will be fetched
24209and encoded until a null of appropriate width is found.
24210@end defun
24211
24212@defun Value.fetch_lazy ()
24213If the @code{gdb.Value} object is currently a lazy value
24214(@code{gdb.Value.is_lazy} is @code{True}), then the value is
24215fetched from the inferior. Any errors that occur in the process
24216will produce a Python exception.
24217
24218If the @code{gdb.Value} object is not a lazy value, this method
24219has no effect.
24220
24221This method does not return a value.
24222@end defun
24223
24224
24225@node Types In Python
24226@subsubsection Types In Python
24227@cindex types in Python
24228@cindex Python, working with types
24229
24230@tindex gdb.Type
24231@value{GDBN} represents types from the inferior using the class
24232@code{gdb.Type}.
24233
24234The following type-related functions are available in the @code{gdb}
24235module:
24236
24237@findex gdb.lookup_type
24238@defun gdb.lookup_type (name @r{[}, block@r{]})
24239This function looks up a type by name. @var{name} is the name of the
24240type to look up. It must be a string.
24241
24242If @var{block} is given, then @var{name} is looked up in that scope.
24243Otherwise, it is searched for globally.
24244
24245Ordinarily, this function will return an instance of @code{gdb.Type}.
24246If the named type cannot be found, it will throw an exception.
24247@end defun
24248
24249If the type is a structure or class type, or an enum type, the fields
24250of that type can be accessed using the Python @dfn{dictionary syntax}.
24251For example, if @code{some_type} is a @code{gdb.Type} instance holding
24252a structure type, you can access its @code{foo} field with:
24253
24254@smallexample
24255bar = some_type['foo']
24256@end smallexample
24257
24258@code{bar} will be a @code{gdb.Field} object; see below under the
24259description of the @code{Type.fields} method for a description of the
24260@code{gdb.Field} class.
24261
24262An instance of @code{Type} has the following attributes:
24263
24264@defvar Type.code
24265The type code for this type. The type code will be one of the
24266@code{TYPE_CODE_} constants defined below.
24267@end defvar
24268
24269@defvar Type.sizeof
24270The size of this type, in target @code{char} units. Usually, a
24271target's @code{char} type will be an 8-bit byte. However, on some
24272unusual platforms, this type may have a different size.
24273@end defvar
24274
24275@defvar Type.tag
24276The tag name for this type. The tag name is the name after
24277@code{struct}, @code{union}, or @code{enum} in C and C@t{++}; not all
24278languages have this concept. If this type has no tag name, then
24279@code{None} is returned.
24280@end defvar
24281
24282The following methods are provided:
24283
24284@defun Type.fields ()
24285For structure and union types, this method returns the fields. Range
24286types have two fields, the minimum and maximum values. Enum types
24287have one field per enum constant. Function and method types have one
24288field per parameter. The base types of C@t{++} classes are also
24289represented as fields. If the type has no fields, or does not fit
24290into one of these categories, an empty sequence will be returned.
24291
24292Each field is a @code{gdb.Field} object, with some pre-defined attributes:
24293@table @code
24294@item bitpos
24295This attribute is not available for @code{static} fields (as in
24296C@t{++} or Java). For non-@code{static} fields, the value is the bit
24297position of the field. For @code{enum} fields, the value is the
24298enumeration member's integer representation.
24299
24300@item name
24301The name of the field, or @code{None} for anonymous fields.
24302
24303@item artificial
24304This is @code{True} if the field is artificial, usually meaning that
24305it was provided by the compiler and not the user. This attribute is
24306always provided, and is @code{False} if the field is not artificial.
24307
24308@item is_base_class
24309This is @code{True} if the field represents a base class of a C@t{++}
24310structure. This attribute is always provided, and is @code{False}
24311if the field is not a base class of the type that is the argument of
24312@code{fields}, or if that type was not a C@t{++} class.
24313
24314@item bitsize
24315If the field is packed, or is a bitfield, then this will have a
24316non-zero value, which is the size of the field in bits. Otherwise,
24317this will be zero; in this case the field's size is given by its type.
24318
24319@item type
24320The type of the field. This is usually an instance of @code{Type},
24321but it can be @code{None} in some situations.
24322@end table
24323@end defun
24324
24325@defun Type.array (@var{n1} @r{[}, @var{n2}@r{]})
24326Return a new @code{gdb.Type} object which represents an array of this
24327type. If one argument is given, it is the inclusive upper bound of
24328the array; in this case the lower bound is zero. If two arguments are
24329given, the first argument is the lower bound of the array, and the
24330second argument is the upper bound of the array. An array's length
24331must not be negative, but the bounds can be.
24332@end defun
24333
24334@defun Type.vector (@var{n1} @r{[}, @var{n2}@r{]})
24335Return a new @code{gdb.Type} object which represents a vector of this
24336type. If one argument is given, it is the inclusive upper bound of
24337the vector; in this case the lower bound is zero. If two arguments are
24338given, the first argument is the lower bound of the vector, and the
24339second argument is the upper bound of the vector. A vector's length
24340must not be negative, but the bounds can be.
24341
24342The difference between an @code{array} and a @code{vector} is that
24343arrays behave like in C: when used in expressions they decay to a pointer
24344to the first element whereas vectors are treated as first class values.
24345@end defun
24346
24347@defun Type.const ()
24348Return a new @code{gdb.Type} object which represents a
24349@code{const}-qualified variant of this type.
24350@end defun
24351
24352@defun Type.volatile ()
24353Return a new @code{gdb.Type} object which represents a
24354@code{volatile}-qualified variant of this type.
24355@end defun
24356
24357@defun Type.unqualified ()
24358Return a new @code{gdb.Type} object which represents an unqualified
24359variant of this type. That is, the result is neither @code{const} nor
24360@code{volatile}.
24361@end defun
24362
24363@defun Type.range ()
24364Return a Python @code{Tuple} object that contains two elements: the
24365low bound of the argument type and the high bound of that type. If
24366the type does not have a range, @value{GDBN} will raise a
24367@code{gdb.error} exception (@pxref{Exception Handling}).
24368@end defun
24369
24370@defun Type.reference ()
24371Return a new @code{gdb.Type} object which represents a reference to this
24372type.
24373@end defun
24374
24375@defun Type.pointer ()
24376Return a new @code{gdb.Type} object which represents a pointer to this
24377type.
24378@end defun
24379
24380@defun Type.strip_typedefs ()
24381Return a new @code{gdb.Type} that represents the real type,
24382after removing all layers of typedefs.
24383@end defun
24384
24385@defun Type.target ()
24386Return a new @code{gdb.Type} object which represents the target type
24387of this type.
24388
24389For a pointer type, the target type is the type of the pointed-to
24390object. For an array type (meaning C-like arrays), the target type is
24391the type of the elements of the array. For a function or method type,
24392the target type is the type of the return value. For a complex type,
24393the target type is the type of the elements. For a typedef, the
24394target type is the aliased type.
24395
24396If the type does not have a target, this method will throw an
24397exception.
24398@end defun
24399
24400@defun Type.template_argument (n @r{[}, block@r{]})
24401If this @code{gdb.Type} is an instantiation of a template, this will
24402return a new @code{gdb.Type} which represents the type of the
24403@var{n}th template argument.
24404
24405If this @code{gdb.Type} is not a template type, this will throw an
24406exception. Ordinarily, only C@t{++} code will have template types.
24407
24408If @var{block} is given, then @var{name} is looked up in that scope.
24409Otherwise, it is searched for globally.
24410@end defun
24411
24412
24413Each type has a code, which indicates what category this type falls
24414into. The available type categories are represented by constants
24415defined in the @code{gdb} module:
24416
24417@table @code
24418@findex TYPE_CODE_PTR
24419@findex gdb.TYPE_CODE_PTR
24420@item gdb.TYPE_CODE_PTR
24421The type is a pointer.
24422
24423@findex TYPE_CODE_ARRAY
24424@findex gdb.TYPE_CODE_ARRAY
24425@item gdb.TYPE_CODE_ARRAY
24426The type is an array.
24427
24428@findex TYPE_CODE_STRUCT
24429@findex gdb.TYPE_CODE_STRUCT
24430@item gdb.TYPE_CODE_STRUCT
24431The type is a structure.
24432
24433@findex TYPE_CODE_UNION
24434@findex gdb.TYPE_CODE_UNION
24435@item gdb.TYPE_CODE_UNION
24436The type is a union.
24437
24438@findex TYPE_CODE_ENUM
24439@findex gdb.TYPE_CODE_ENUM
24440@item gdb.TYPE_CODE_ENUM
24441The type is an enum.
24442
24443@findex TYPE_CODE_FLAGS
24444@findex gdb.TYPE_CODE_FLAGS
24445@item gdb.TYPE_CODE_FLAGS
24446A bit flags type, used for things such as status registers.
24447
24448@findex TYPE_CODE_FUNC
24449@findex gdb.TYPE_CODE_FUNC
24450@item gdb.TYPE_CODE_FUNC
24451The type is a function.
24452
24453@findex TYPE_CODE_INT
24454@findex gdb.TYPE_CODE_INT
24455@item gdb.TYPE_CODE_INT
24456The type is an integer type.
24457
24458@findex TYPE_CODE_FLT
24459@findex gdb.TYPE_CODE_FLT
24460@item gdb.TYPE_CODE_FLT
24461A floating point type.
24462
24463@findex TYPE_CODE_VOID
24464@findex gdb.TYPE_CODE_VOID
24465@item gdb.TYPE_CODE_VOID
24466The special type @code{void}.
24467
24468@findex TYPE_CODE_SET
24469@findex gdb.TYPE_CODE_SET
24470@item gdb.TYPE_CODE_SET
24471A Pascal set type.
24472
24473@findex TYPE_CODE_RANGE
24474@findex gdb.TYPE_CODE_RANGE
24475@item gdb.TYPE_CODE_RANGE
24476A range type, that is, an integer type with bounds.
24477
24478@findex TYPE_CODE_STRING
24479@findex gdb.TYPE_CODE_STRING
24480@item gdb.TYPE_CODE_STRING
24481A string type. Note that this is only used for certain languages with
24482language-defined string types; C strings are not represented this way.
24483
24484@findex TYPE_CODE_BITSTRING
24485@findex gdb.TYPE_CODE_BITSTRING
24486@item gdb.TYPE_CODE_BITSTRING
24487A string of bits. It is deprecated.
24488
24489@findex TYPE_CODE_ERROR
24490@findex gdb.TYPE_CODE_ERROR
24491@item gdb.TYPE_CODE_ERROR
24492An unknown or erroneous type.
24493
24494@findex TYPE_CODE_METHOD
24495@findex gdb.TYPE_CODE_METHOD
24496@item gdb.TYPE_CODE_METHOD
24497A method type, as found in C@t{++} or Java.
24498
24499@findex TYPE_CODE_METHODPTR
24500@findex gdb.TYPE_CODE_METHODPTR
24501@item gdb.TYPE_CODE_METHODPTR
24502A pointer-to-member-function.
24503
24504@findex TYPE_CODE_MEMBERPTR
24505@findex gdb.TYPE_CODE_MEMBERPTR
24506@item gdb.TYPE_CODE_MEMBERPTR
24507A pointer-to-member.
24508
24509@findex TYPE_CODE_REF
24510@findex gdb.TYPE_CODE_REF
24511@item gdb.TYPE_CODE_REF
24512A reference type.
24513
24514@findex TYPE_CODE_CHAR
24515@findex gdb.TYPE_CODE_CHAR
24516@item gdb.TYPE_CODE_CHAR
24517A character type.
24518
24519@findex TYPE_CODE_BOOL
24520@findex gdb.TYPE_CODE_BOOL
24521@item gdb.TYPE_CODE_BOOL
24522A boolean type.
24523
24524@findex TYPE_CODE_COMPLEX
24525@findex gdb.TYPE_CODE_COMPLEX
24526@item gdb.TYPE_CODE_COMPLEX
24527A complex float type.
24528
24529@findex TYPE_CODE_TYPEDEF
24530@findex gdb.TYPE_CODE_TYPEDEF
24531@item gdb.TYPE_CODE_TYPEDEF
24532A typedef to some other type.
24533
24534@findex TYPE_CODE_NAMESPACE
24535@findex gdb.TYPE_CODE_NAMESPACE
24536@item gdb.TYPE_CODE_NAMESPACE
24537A C@t{++} namespace.
24538
24539@findex TYPE_CODE_DECFLOAT
24540@findex gdb.TYPE_CODE_DECFLOAT
24541@item gdb.TYPE_CODE_DECFLOAT
24542A decimal floating point type.
24543
24544@findex TYPE_CODE_INTERNAL_FUNCTION
24545@findex gdb.TYPE_CODE_INTERNAL_FUNCTION
24546@item gdb.TYPE_CODE_INTERNAL_FUNCTION
24547A function internal to @value{GDBN}. This is the type used to represent
24548convenience functions.
24549@end table
24550
24551Further support for types is provided in the @code{gdb.types}
24552Python module (@pxref{gdb.types}).
24553
24554@node Pretty Printing API
24555@subsubsection Pretty Printing API
24556
24557An example output is provided (@pxref{Pretty Printing}).
24558
24559A pretty-printer is just an object that holds a value and implements a
24560specific interface, defined here.
24561
24562@defun pretty_printer.children (self)
24563@value{GDBN} will call this method on a pretty-printer to compute the
24564children of the pretty-printer's value.
24565
24566This method must return an object conforming to the Python iterator
24567protocol. Each item returned by the iterator must be a tuple holding
24568two elements. The first element is the ``name'' of the child; the
24569second element is the child's value. The value can be any Python
24570object which is convertible to a @value{GDBN} value.
24571
24572This method is optional. If it does not exist, @value{GDBN} will act
24573as though the value has no children.
24574@end defun
24575
24576@defun pretty_printer.display_hint (self)
24577The CLI may call this method and use its result to change the
24578formatting of a value. The result will also be supplied to an MI
24579consumer as a @samp{displayhint} attribute of the variable being
24580printed.
24581
24582This method is optional. If it does exist, this method must return a
24583string.
24584
24585Some display hints are predefined by @value{GDBN}:
24586
24587@table @samp
24588@item array
24589Indicate that the object being printed is ``array-like''. The CLI
24590uses this to respect parameters such as @code{set print elements} and
24591@code{set print array}.
24592
24593@item map
24594Indicate that the object being printed is ``map-like'', and that the
24595children of this value can be assumed to alternate between keys and
24596values.
24597
24598@item string
24599Indicate that the object being printed is ``string-like''. If the
24600printer's @code{to_string} method returns a Python string of some
24601kind, then @value{GDBN} will call its internal language-specific
24602string-printing function to format the string. For the CLI this means
24603adding quotation marks, possibly escaping some characters, respecting
24604@code{set print elements}, and the like.
24605@end table
24606@end defun
24607
24608@defun pretty_printer.to_string (self)
24609@value{GDBN} will call this method to display the string
24610representation of the value passed to the object's constructor.
24611
24612When printing from the CLI, if the @code{to_string} method exists,
24613then @value{GDBN} will prepend its result to the values returned by
24614@code{children}. Exactly how this formatting is done is dependent on
24615the display hint, and may change as more hints are added. Also,
24616depending on the print settings (@pxref{Print Settings}), the CLI may
24617print just the result of @code{to_string} in a stack trace, omitting
24618the result of @code{children}.
24619
24620If this method returns a string, it is printed verbatim.
24621
24622Otherwise, if this method returns an instance of @code{gdb.Value},
24623then @value{GDBN} prints this value. This may result in a call to
24624another pretty-printer.
24625
24626If instead the method returns a Python value which is convertible to a
24627@code{gdb.Value}, then @value{GDBN} performs the conversion and prints
24628the resulting value. Again, this may result in a call to another
24629pretty-printer. Python scalars (integers, floats, and booleans) and
24630strings are convertible to @code{gdb.Value}; other types are not.
24631
24632Finally, if this method returns @code{None} then no further operations
24633are peformed in this method and nothing is printed.
24634
24635If the result is not one of these types, an exception is raised.
24636@end defun
24637
24638@value{GDBN} provides a function which can be used to look up the
24639default pretty-printer for a @code{gdb.Value}:
24640
24641@findex gdb.default_visualizer
24642@defun gdb.default_visualizer (value)
24643This function takes a @code{gdb.Value} object as an argument. If a
24644pretty-printer for this value exists, then it is returned. If no such
24645printer exists, then this returns @code{None}.
24646@end defun
24647
24648@node Selecting Pretty-Printers
24649@subsubsection Selecting Pretty-Printers
24650
24651The Python list @code{gdb.pretty_printers} contains an array of
24652functions or callable objects that have been registered via addition
24653as a pretty-printer. Printers in this list are called @code{global}
24654printers, they're available when debugging all inferiors.
24655Each @code{gdb.Progspace} contains a @code{pretty_printers} attribute.
24656Each @code{gdb.Objfile} also contains a @code{pretty_printers}
24657attribute.
24658
24659Each function on these lists is passed a single @code{gdb.Value}
24660argument and should return a pretty-printer object conforming to the
24661interface definition above (@pxref{Pretty Printing API}). If a function
24662cannot create a pretty-printer for the value, it should return
24663@code{None}.
24664
24665@value{GDBN} first checks the @code{pretty_printers} attribute of each
24666@code{gdb.Objfile} in the current program space and iteratively calls
24667each enabled lookup routine in the list for that @code{gdb.Objfile}
24668until it receives a pretty-printer object.
24669If no pretty-printer is found in the objfile lists, @value{GDBN} then
24670searches the pretty-printer list of the current program space,
24671calling each enabled function until an object is returned.
24672After these lists have been exhausted, it tries the global
24673@code{gdb.pretty_printers} list, again calling each enabled function until an
24674object is returned.
24675
24676The order in which the objfiles are searched is not specified. For a
24677given list, functions are always invoked from the head of the list,
24678and iterated over sequentially until the end of the list, or a printer
24679object is returned.
24680
24681For various reasons a pretty-printer may not work.
24682For example, the underlying data structure may have changed and
24683the pretty-printer is out of date.
24684
24685The consequences of a broken pretty-printer are severe enough that
24686@value{GDBN} provides support for enabling and disabling individual
24687printers. For example, if @code{print frame-arguments} is on,
24688a backtrace can become highly illegible if any argument is printed
24689with a broken printer.
24690
24691Pretty-printers are enabled and disabled by attaching an @code{enabled}
24692attribute to the registered function or callable object. If this attribute
24693is present and its value is @code{False}, the printer is disabled, otherwise
24694the printer is enabled.
24695
24696@node Writing a Pretty-Printer
24697@subsubsection Writing a Pretty-Printer
24698@cindex writing a pretty-printer
24699
24700A pretty-printer consists of two parts: a lookup function to detect
24701if the type is supported, and the printer itself.
24702
24703Here is an example showing how a @code{std::string} printer might be
24704written. @xref{Pretty Printing API}, for details on the API this class
24705must provide.
24706
24707@smallexample
24708class StdStringPrinter(object):
24709 "Print a std::string"
24710
24711 def __init__(self, val):
24712 self.val = val
24713
24714 def to_string(self):
24715 return self.val['_M_dataplus']['_M_p']
24716
24717 def display_hint(self):
24718 return 'string'
24719@end smallexample
24720
24721And here is an example showing how a lookup function for the printer
24722example above might be written.
24723
24724@smallexample
24725def str_lookup_function(val):
24726 lookup_tag = val.type.tag
24727 if lookup_tag == None:
24728 return None
24729 regex = re.compile("^std::basic_string<char,.*>$")
24730 if regex.match(lookup_tag):
24731 return StdStringPrinter(val)
24732 return None
24733@end smallexample
24734
24735The example lookup function extracts the value's type, and attempts to
24736match it to a type that it can pretty-print. If it is a type the
24737printer can pretty-print, it will return a printer object. If not, it
24738returns @code{None}.
24739
24740We recommend that you put your core pretty-printers into a Python
24741package. If your pretty-printers are for use with a library, we
24742further recommend embedding a version number into the package name.
24743This practice will enable @value{GDBN} to load multiple versions of
24744your pretty-printers at the same time, because they will have
24745different names.
24746
24747You should write auto-loaded code (@pxref{Python Auto-loading}) such that it
24748can be evaluated multiple times without changing its meaning. An
24749ideal auto-load file will consist solely of @code{import}s of your
24750printer modules, followed by a call to a register pretty-printers with
24751the current objfile.
24752
24753Taken as a whole, this approach will scale nicely to multiple
24754inferiors, each potentially using a different library version.
24755Embedding a version number in the Python package name will ensure that
24756@value{GDBN} is able to load both sets of printers simultaneously.
24757Then, because the search for pretty-printers is done by objfile, and
24758because your auto-loaded code took care to register your library's
24759printers with a specific objfile, @value{GDBN} will find the correct
24760printers for the specific version of the library used by each
24761inferior.
24762
24763To continue the @code{std::string} example (@pxref{Pretty Printing API}),
24764this code might appear in @code{gdb.libstdcxx.v6}:
24765
24766@smallexample
24767def register_printers(objfile):
24768 objfile.pretty_printers.append(str_lookup_function)
24769@end smallexample
24770
24771@noindent
24772And then the corresponding contents of the auto-load file would be:
24773
24774@smallexample
24775import gdb.libstdcxx.v6
24776gdb.libstdcxx.v6.register_printers(gdb.current_objfile())
24777@end smallexample
24778
24779The previous example illustrates a basic pretty-printer.
24780There are a few things that can be improved on.
24781The printer doesn't have a name, making it hard to identify in a
24782list of installed printers. The lookup function has a name, but
24783lookup functions can have arbitrary, even identical, names.
24784
24785Second, the printer only handles one type, whereas a library typically has
24786several types. One could install a lookup function for each desired type
24787in the library, but one could also have a single lookup function recognize
24788several types. The latter is the conventional way this is handled.
24789If a pretty-printer can handle multiple data types, then its
24790@dfn{subprinters} are the printers for the individual data types.
24791
24792The @code{gdb.printing} module provides a formal way of solving these
24793problems (@pxref{gdb.printing}).
24794Here is another example that handles multiple types.
24795
24796These are the types we are going to pretty-print:
24797
24798@smallexample
24799struct foo @{ int a, b; @};
24800struct bar @{ struct foo x, y; @};
24801@end smallexample
24802
24803Here are the printers:
24804
24805@smallexample
24806class fooPrinter:
24807 """Print a foo object."""
24808
24809 def __init__(self, val):
24810 self.val = val
24811
24812 def to_string(self):
24813 return ("a=<" + str(self.val["a"]) +
24814 "> b=<" + str(self.val["b"]) + ">")
24815
24816class barPrinter:
24817 """Print a bar object."""
24818
24819 def __init__(self, val):
24820 self.val = val
24821
24822 def to_string(self):
24823 return ("x=<" + str(self.val["x"]) +
24824 "> y=<" + str(self.val["y"]) + ">")
24825@end smallexample
24826
24827This example doesn't need a lookup function, that is handled by the
24828@code{gdb.printing} module. Instead a function is provided to build up
24829the object that handles the lookup.
24830
24831@smallexample
24832import gdb.printing
24833
24834def build_pretty_printer():
24835 pp = gdb.printing.RegexpCollectionPrettyPrinter(
24836 "my_library")
24837 pp.add_printer('foo', '^foo$', fooPrinter)
24838 pp.add_printer('bar', '^bar$', barPrinter)
24839 return pp
24840@end smallexample
24841
24842And here is the autoload support:
24843
24844@smallexample
24845import gdb.printing
24846import my_library
24847gdb.printing.register_pretty_printer(
24848 gdb.current_objfile(),
24849 my_library.build_pretty_printer())
24850@end smallexample
24851
24852Finally, when this printer is loaded into @value{GDBN}, here is the
24853corresponding output of @samp{info pretty-printer}:
24854
24855@smallexample
24856(gdb) info pretty-printer
24857my_library.so:
24858 my_library
24859 foo
24860 bar
24861@end smallexample
24862
24863@node Type Printing API
24864@subsubsection Type Printing API
24865@cindex type printing API for Python
24866
24867@value{GDBN} provides a way for Python code to customize type display.
24868This is mainly useful for substituting canonical typedef names for
24869types.
24870
24871@cindex type printer
24872A @dfn{type printer} is just a Python object conforming to a certain
24873protocol. A simple base class implementing the protocol is provided;
24874see @ref{gdb.types}. A type printer must supply at least:
24875
24876@defivar type_printer enabled
24877A boolean which is True if the printer is enabled, and False
24878otherwise. This is manipulated by the @code{enable type-printer}
24879and @code{disable type-printer} commands.
24880@end defivar
24881
24882@defivar type_printer name
24883The name of the type printer. This must be a string. This is used by
24884the @code{enable type-printer} and @code{disable type-printer}
24885commands.
24886@end defivar
24887
24888@defmethod type_printer instantiate (self)
24889This is called by @value{GDBN} at the start of type-printing. It is
24890only called if the type printer is enabled. This method must return a
24891new object that supplies a @code{recognize} method, as described below.
24892@end defmethod
24893
24894
24895When displaying a type, say via the @code{ptype} command, @value{GDBN}
24896will compute a list of type recognizers. This is done by iterating
24897first over the per-objfile type printers (@pxref{Objfiles In Python}),
24898followed by the per-progspace type printers (@pxref{Progspaces In
24899Python}), and finally the global type printers.
24900
24901@value{GDBN} will call the @code{instantiate} method of each enabled
24902type printer. If this method returns @code{None}, then the result is
24903ignored; otherwise, it is appended to the list of recognizers.
24904
24905Then, when @value{GDBN} is going to display a type name, it iterates
24906over the list of recognizers. For each one, it calls the recognition
24907function, stopping if the function returns a non-@code{None} value.
24908The recognition function is defined as:
24909
24910@defmethod type_recognizer recognize (self, type)
24911If @var{type} is not recognized, return @code{None}. Otherwise,
24912return a string which is to be printed as the name of @var{type}.
24913@var{type} will be an instance of @code{gdb.Type} (@pxref{Types In
24914Python}).
24915@end defmethod
24916
24917@value{GDBN} uses this two-pass approach so that type printers can
24918efficiently cache information without holding on to it too long. For
24919example, it can be convenient to look up type information in a type
24920printer and hold it for a recognizer's lifetime; if a single pass were
24921done then type printers would have to make use of the event system in
24922order to avoid holding information that could become stale as the
24923inferior changed.
24924
24925@node Frame Filter API
24926@subsubsection Filtering Frames.
24927@cindex frame filters api
24928
24929Frame filters are Python objects that manipulate the visibility of a
24930frame or frames when a backtrace (@pxref{Backtrace}) is printed by
24931@value{GDBN}.
24932
24933Only commands that print a backtrace, or, in the case of @sc{gdb/mi}
24934commands (@pxref{GDB/MI}), those that return a collection of frames
24935are affected. The commands that work with frame filters are:
24936
24937@code{backtrace} (@pxref{backtrace-command,, The backtrace command}),
24938@code{-stack-list-frames}
24939(@pxref{-stack-list-frames,, The -stack-list-frames command}),
24940@code{-stack-list-variables} (@pxref{-stack-list-variables,, The
24941-stack-list-variables command}), @code{-stack-list-arguments}
24942@pxref{-stack-list-arguments,, The -stack-list-arguments command}) and
24943@code{-stack-list-locals} (@pxref{-stack-list-locals,, The
24944-stack-list-locals command}).
24945
24946A frame filter works by taking an iterator as an argument, applying
24947actions to the contents of that iterator, and returning another
24948iterator (or, possibly, the same iterator it was provided in the case
24949where the filter does not perform any operations). Typically, frame
24950filters utilize tools such as the Python's @code{itertools} module to
24951work with and create new iterators from the source iterator.
24952Regardless of how a filter chooses to apply actions, it must not alter
24953the underlying @value{GDBN} frame or frames, or attempt to alter the
24954call-stack within @value{GDBN}. This preserves data integrity within
24955@value{GDBN}. Frame filters are executed on a priority basis and care
24956should be taken that some frame filters may have been executed before,
24957and that some frame filters will be executed after.
24958
24959An important consideration when designing frame filters, and well
24960worth reflecting upon, is that frame filters should avoid unwinding
24961the call stack if possible. Some stacks can run very deep, into the
24962tens of thousands in some cases. To search every frame when a frame
24963filter executes may be too expensive at that step. The frame filter
24964cannot know how many frames it has to iterate over, and it may have to
24965iterate through them all. This ends up duplicating effort as
24966@value{GDBN} performs this iteration when it prints the frames. If
24967the filter can defer unwinding frames until frame decorators are
24968executed, after the last filter has executed, it should. @xref{Frame
24969Decorator API}, for more information on decorators. Also, there are
24970examples for both frame decorators and filters in later chapters.
24971@xref{Writing a Frame Filter}, for more information.
24972
24973The Python dictionary @code{gdb.frame_filters} contains key/object
24974pairings that comprise a frame filter. Frame filters in this
24975dictionary are called @code{global} frame filters, and they are
24976available when debugging all inferiors. These frame filters must
24977register with the dictionary directly. In addition to the
24978@code{global} dictionary, there are other dictionaries that are loaded
24979with different inferiors via auto-loading (@pxref{Python
24980Auto-loading}). The two other areas where frame filter dictionaries
24981can be found are: @code{gdb.Progspace} which contains a
24982@code{frame_filters} dictionary attribute, and each @code{gdb.Objfile}
24983object which also contains a @code{frame_filters} dictionary
24984attribute.
24985
24986When a command is executed from @value{GDBN} that is compatible with
24987frame filters, @value{GDBN} combines the @code{global},
24988@code{gdb.Progspace} and all @code{gdb.Objfile} dictionaries currently
24989loaded. All of the @code{gdb.Objfile} dictionaries are combined, as
24990several frames, and thus several object files, might be in use.
24991@value{GDBN} then prunes any frame filter whose @code{enabled}
24992attribute is @code{False}. This pruned list is then sorted according
24993to the @code{priority} attribute in each filter.
24994
24995Once the dictionaries are combined, pruned and sorted, @value{GDBN}
24996creates an iterator which wraps each frame in the call stack in a
24997@code{FrameDecorator} object, and calls each filter in order. The
24998output from the previous filter will always be the input to the next
24999filter, and so on.
25000
25001Frame filters have a mandatory interface which each frame filter must
25002implement, defined here:
25003
25004@defun FrameFilter.filter (iterator)
25005@value{GDBN} will call this method on a frame filter when it has
25006reached the order in the priority list for that filter.
25007
25008For example, if there are four frame filters:
25009
25010@smallexample
25011Name Priority
25012
25013Filter1 5
25014Filter2 10
25015Filter3 100
25016Filter4 1
25017@end smallexample
25018
25019The order that the frame filters will be called is:
25020
25021@smallexample
25022Filter3 -> Filter2 -> Filter1 -> Filter4
25023@end smallexample
25024
25025Note that the output from @code{Filter3} is passed to the input of
25026@code{Filter2}, and so on.
25027
25028This @code{filter} method is passed a Python iterator. This iterator
25029contains a sequence of frame decorators that wrap each
25030@code{gdb.Frame}, or a frame decorator that wraps another frame
25031decorator. The first filter that is executed in the sequence of frame
25032filters will receive an iterator entirely comprised of default
25033@code{FrameDecorator} objects. However, after each frame filter is
25034executed, the previous frame filter may have wrapped some or all of
25035the frame decorators with their own frame decorator. As frame
25036decorators must also conform to a mandatory interface, these
25037decorators can be assumed to act in a uniform manner (@pxref{Frame
25038Decorator API}).
25039
25040This method must return an object conforming to the Python iterator
25041protocol. Each item in the iterator must be an object conforming to
25042the frame decorator interface. If a frame filter does not wish to
25043perform any operations on this iterator, it should return that
25044iterator untouched.
25045
25046This method is not optional. If it does not exist, @value{GDBN} will
25047raise and print an error.
25048@end defun
25049
25050@defvar FrameFilter.name
25051The @code{name} attribute must be Python string which contains the
25052name of the filter displayed by @value{GDBN} (@pxref{Frame Filter
25053Management}). This attribute may contain any combination of letters
25054or numbers. Care should be taken to ensure that it is unique. This
25055attribute is mandatory.
25056@end defvar
25057
25058@defvar FrameFilter.enabled
25059The @code{enabled} attribute must be Python boolean. This attribute
25060indicates to @value{GDBN} whether the frame filter is enabled, and
25061should be considered when frame filters are executed. If
25062@code{enabled} is @code{True}, then the frame filter will be executed
25063when any of the backtrace commands detailed earlier in this chapter
25064are executed. If @code{enabled} is @code{False}, then the frame
25065filter will not be executed. This attribute is mandatory.
25066@end defvar
25067
25068@defvar FrameFilter.priority
25069The @code{priority} attribute must be Python integer. This attribute
25070controls the order of execution in relation to other frame filters.
25071There are no imposed limits on the range of @code{priority} other than
25072it must be a valid integer. The higher the @code{priority} attribute,
25073the sooner the frame filter will be executed in relation to other
25074frame filters. Although @code{priority} can be negative, it is
25075recommended practice to assume zero is the lowest priority that a
25076frame filter can be assigned. Frame filters that have the same
25077priority are executed in unsorted order in that priority slot. This
25078attribute is mandatory.
25079@end defvar
25080
25081@node Frame Decorator API
25082@subsubsection Decorating Frames.
25083@cindex frame decorator api
25084
25085Frame decorators are sister objects to frame filters (@pxref{Frame
25086Filter API}). Frame decorators are applied by a frame filter and can
25087only be used in conjunction with frame filters.
25088
25089The purpose of a frame decorator is to customize the printed content
25090of each @code{gdb.Frame} in commands where frame filters are executed.
25091This concept is called decorating a frame. Frame decorators decorate
25092a @code{gdb.Frame} with Python code contained within each API call.
25093This separates the actual data contained in a @code{gdb.Frame} from
25094the decorated data produced by a frame decorator. This abstraction is
25095necessary to maintain integrity of the data contained in each
25096@code{gdb.Frame}.
25097
25098Frame decorators have a mandatory interface, defined below.
25099
25100@value{GDBN} already contains a frame decorator called
25101@code{FrameDecorator}. This contains substantial amounts of
25102boilerplate code to decorate the content of a @code{gdb.Frame}. It is
25103recommended that other frame decorators inherit and extend this
25104object, and only to override the methods needed.
25105
25106@defun FrameDecorator.elided (self)
25107
25108The @code{elided} method groups frames together in a hierarchical
25109system. An example would be an interpreter, where multiple low-level
25110frames make up a single call in the interpreted language. In this
25111example, the frame filter would elide the low-level frames and present
25112a single high-level frame, representing the call in the interpreted
25113language, to the user.
25114
25115The @code{elided} function must return an iterable and this iterable
25116must contain the frames that are being elided wrapped in a suitable
25117frame decorator. If no frames are being elided this function may
25118return an empty iterable, or @code{None}. Elided frames are indented
25119from normal frames in a @code{CLI} backtrace, or in the case of
25120@code{GDB/MI}, are placed in the @code{children} field of the eliding
25121frame.
25122
25123It is the frame filter's task to also filter out the elided frames from
25124the source iterator. This will avoid printing the frame twice.
25125@end defun
25126
25127@defun FrameDecorator.function (self)
25128
25129This method returns the name of the function in the frame that is to
25130be printed.
25131
25132This method must return a Python string describing the function, or
25133@code{None}.
25134
25135If this function returns @code{None}, @value{GDBN} will not print any
25136data for this field.
25137@end defun
25138
25139@defun FrameDecorator.address (self)
25140
25141This method returns the address of the frame that is to be printed.
25142
25143This method must return a Python numeric integer type of sufficient
25144size to describe the address of the frame, or @code{None}.
25145
25146If this function returns a @code{None}, @value{GDBN} will not print
25147any data for this field.
25148@end defun
25149
25150@defun FrameDecorator.filename (self)
25151
25152This method returns the filename and path associated with this frame.
25153
25154This method must return a Python string containing the filename and
25155the path to the object file backing the frame, or @code{None}.
25156
25157If this function returns a @code{None}, @value{GDBN} will not print
25158any data for this field.
25159@end defun
25160
25161@defun FrameDecorator.line (self):
25162
25163This method returns the line number associated with the current
25164position within the function addressed by this frame.
25165
25166This method must return a Python integer type, or @code{None}.
25167
25168If this function returns a @code{None}, @value{GDBN} will not print
25169any data for this field.
25170@end defun
25171
25172@defun FrameDecorator.frame_args (self)
25173@anchor{frame_args}
25174
25175This method must return an iterable, or @code{None}. Returning an
25176empty iterable, or @code{None} means frame arguments will not be
25177printed for this frame. This iterable must contain objects that
25178implement two methods, described here.
25179
25180This object must implement a @code{argument} method which takes a
25181single @code{self} parameter and must return a @code{gdb.Symbol}
25182(@pxref{Symbols In Python}), or a Python string. The object must also
25183implement a @code{value} method which takes a single @code{self}
25184parameter and must return a @code{gdb.Value} (@pxref{Values From
25185Inferior}), a Python value, or @code{None}. If the @code{value}
25186method returns @code{None}, and the @code{argument} method returns a
25187@code{gdb.Symbol}, @value{GDBN} will look-up and print the value of
25188the @code{gdb.Symbol} automatically.
25189
25190A brief example:
25191
25192@smallexample
25193class SymValueWrapper():
25194
25195 def __init__(self, symbol, value):
25196 self.sym = symbol
25197 self.val = value
25198
25199 def value(self):
25200 return self.val
25201
25202 def symbol(self):
25203 return self.sym
25204
25205class SomeFrameDecorator()
25206...
25207...
25208 def frame_args(self):
25209 args = []
25210 try:
25211 block = self.inferior_frame.block()
25212 except:
25213 return None
25214
25215 # Iterate over all symbols in a block. Only add
25216 # symbols that are arguments.
25217 for sym in block:
25218 if not sym.is_argument:
25219 continue
25220 args.append(SymValueWrapper(sym,None))
25221
25222 # Add example synthetic argument.
25223 args.append(SymValueWrapper(``foo'', 42))
25224
25225 return args
25226@end smallexample
25227@end defun
25228
25229@defun FrameDecorator.frame_locals (self)
25230
25231This method must return an iterable or @code{None}. Returning an
25232empty iterable, or @code{None} means frame local arguments will not be
25233printed for this frame.
25234
25235The object interface, the description of the various strategies for
25236reading frame locals, and the example are largely similar to those
25237described in the @code{frame_args} function, (@pxref{frame_args,,The
25238frame filter frame_args function}). Below is a modified example:
25239
25240@smallexample
25241class SomeFrameDecorator()
25242...
25243...
25244 def frame_locals(self):
25245 vars = []
25246 try:
25247 block = self.inferior_frame.block()
25248 except:
25249 return None
25250
25251 # Iterate over all symbols in a block. Add all
25252 # symbols, except arguments.
25253 for sym in block:
25254 if sym.is_argument:
25255 continue
25256 vars.append(SymValueWrapper(sym,None))
25257
25258 # Add an example of a synthetic local variable.
25259 vars.append(SymValueWrapper(``bar'', 99))
25260
25261 return vars
25262@end smallexample
25263@end defun
25264
25265@defun FrameDecorator.inferior_frame (self):
25266
25267This method must return the underlying @code{gdb.Frame} that this
25268frame decorator is decorating. @value{GDBN} requires the underlying
25269frame for internal frame information to determine how to print certain
25270values when printing a frame.
25271@end defun
25272
25273@node Writing a Frame Filter
25274@subsubsection Writing a Frame Filter
25275@cindex writing a frame filter
25276
25277There are three basic elements that a frame filter must implement: it
25278must correctly implement the documented interface (@pxref{Frame Filter
25279API}), it must register itself with @value{GDBN}, and finally, it must
25280decide if it is to work on the data provided by @value{GDBN}. In all
25281cases, whether it works on the iterator or not, each frame filter must
25282return an iterator. A bare-bones frame filter follows the pattern in
25283the following example.
25284
25285@smallexample
25286import gdb
25287
25288class FrameFilter():
25289
25290 def __init__(self):
25291 # Frame filter attribute creation.
25292 #
25293 # 'name' is the name of the filter that GDB will display.
25294 #
25295 # 'priority' is the priority of the filter relative to other
25296 # filters.
25297 #
25298 # 'enabled' is a boolean that indicates whether this filter is
25299 # enabled and should be executed.
25300
25301 self.name = "Foo"
25302 self.priority = 100
25303 self.enabled = True
25304
25305 # Register this frame filter with the global frame_filters
25306 # dictionary.
25307 gdb.frame_filters[self.name] = self
25308
25309 def filter(self, frame_iter):
25310 # Just return the iterator.
25311 return frame_iter
25312@end smallexample
25313
25314The frame filter in the example above implements the three
25315requirements for all frame filters. It implements the API, self
25316registers, and makes a decision on the iterator (in this case, it just
25317returns the iterator untouched).
25318
25319The first step is attribute creation and assignment, and as shown in
25320the comments the filter assigns the following attributes: @code{name},
25321@code{priority} and whether the filter should be enabled with the
25322@code{enabled} attribute.
25323
25324The second step is registering the frame filter with the dictionary or
25325dictionaries that the frame filter has interest in. As shown in the
25326comments, this filter just registers itself with the global dictionary
25327@code{gdb.frame_filters}. As noted earlier, @code{gdb.frame_filters}
25328is a dictionary that is initialized in the @code{gdb} module when
25329@value{GDBN} starts. What dictionary a filter registers with is an
25330important consideration. Generally, if a filter is specific to a set
25331of code, it should be registered either in the @code{objfile} or
25332@code{progspace} dictionaries as they are specific to the program
25333currently loaded in @value{GDBN}. The global dictionary is always
25334present in @value{GDBN} and is never unloaded. Any filters registered
25335with the global dictionary will exist until @value{GDBN} exits. To
25336avoid filters that may conflict, it is generally better to register
25337frame filters against the dictionaries that more closely align with
25338the usage of the filter currently in question. @xref{Python
25339Auto-loading}, for further information on auto-loading Python scripts.
25340
25341@value{GDBN} takes a hands-off approach to frame filter registration,
25342therefore it is the frame filter's responsibility to ensure
25343registration has occurred, and that any exceptions are handled
25344appropriately. In particular, you may wish to handle exceptions
25345relating to Python dictionary key uniqueness. It is mandatory that
25346the dictionary key is the same as frame filter's @code{name}
25347attribute. When a user manages frame filters (@pxref{Frame Filter
25348Management}), the names @value{GDBN} will display are those contained
25349in the @code{name} attribute.
25350
25351The final step of this example is the implementation of the
25352@code{filter} method. As shown in the example comments, we define the
25353@code{filter} method and note that the method must take an iterator,
25354and also must return an iterator. In this bare-bones example, the
25355frame filter is not very useful as it just returns the iterator
25356untouched. However this is a valid operation for frame filters that
25357have the @code{enabled} attribute set, but decide not to operate on
25358any frames.
25359
25360In the next example, the frame filter operates on all frames and
25361utilizes a frame decorator to perform some work on the frames.
25362@xref{Frame Decorator API}, for further information on the frame
25363decorator interface.
25364
25365This example works on inlined frames. It highlights frames which are
25366inlined by tagging them with an ``[inlined]'' tag. By applying a
25367frame decorator to all frames with the Python @code{itertools imap}
25368method, the example defers actions to the frame decorator. Frame
25369decorators are only processed when @value{GDBN} prints the backtrace.
25370
25371This introduces a new decision making topic: whether to perform
25372decision making operations at the filtering step, or at the printing
25373step. In this example's approach, it does not perform any filtering
25374decisions at the filtering step beyond mapping a frame decorator to
25375each frame. This allows the actual decision making to be performed
25376when each frame is printed. This is an important consideration, and
25377well worth reflecting upon when designing a frame filter. An issue
25378that frame filters should avoid is unwinding the stack if possible.
25379Some stacks can run very deep, into the tens of thousands in some
25380cases. To search every frame to determine if it is inlined ahead of
25381time may be too expensive at the filtering step. The frame filter
25382cannot know how many frames it has to iterate over, and it would have
25383to iterate through them all. This ends up duplicating effort as
25384@value{GDBN} performs this iteration when it prints the frames.
25385
25386In this example decision making can be deferred to the printing step.
25387As each frame is printed, the frame decorator can examine each frame
25388in turn when @value{GDBN} iterates. From a performance viewpoint,
25389this is the most appropriate decision to make as it avoids duplicating
25390the effort that the printing step would undertake anyway. Also, if
25391there are many frame filters unwinding the stack during filtering, it
25392can substantially delay the printing of the backtrace which will
25393result in large memory usage, and a poor user experience.
25394
25395@smallexample
25396class InlineFilter():
25397
25398 def __init__(self):
25399 self.name = "InlinedFrameFilter"
25400 self.priority = 100
25401 self.enabled = True
25402 gdb.frame_filters[self.name] = self
25403
25404 def filter(self, frame_iter):
25405 frame_iter = itertools.imap(InlinedFrameDecorator,
25406 frame_iter)
25407 return frame_iter
25408@end smallexample
25409
25410This frame filter is somewhat similar to the earlier example, except
25411that the @code{filter} method applies a frame decorator object called
25412@code{InlinedFrameDecorator} to each element in the iterator. The
25413@code{imap} Python method is light-weight. It does not proactively
25414iterate over the iterator, but rather creates a new iterator which
25415wraps the existing one.
25416
25417Below is the frame decorator for this example.
25418
25419@smallexample
25420class InlinedFrameDecorator(FrameDecorator):
25421
25422 def __init__(self, fobj):
25423 super(InlinedFrameDecorator, self).__init__(fobj)
25424
25425 def function(self):
25426 frame = fobj.inferior_frame()
25427 name = str(frame.name())
25428
25429 if frame.type() == gdb.INLINE_FRAME:
25430 name = name + " [inlined]"
25431
25432 return name
25433@end smallexample
25434
25435This frame decorator only defines and overrides the @code{function}
25436method. It lets the supplied @code{FrameDecorator}, which is shipped
25437with @value{GDBN}, perform the other work associated with printing
25438this frame.
25439
25440The combination of these two objects create this output from a
25441backtrace:
25442
25443@smallexample
25444#0 0x004004e0 in bar () at inline.c:11
25445#1 0x00400566 in max [inlined] (b=6, a=12) at inline.c:21
25446#2 0x00400566 in main () at inline.c:31
25447@end smallexample
25448
25449So in the case of this example, a frame decorator is applied to all
25450frames, regardless of whether they may be inlined or not. As
25451@value{GDBN} iterates over the iterator produced by the frame filters,
25452@value{GDBN} executes each frame decorator which then makes a decision
25453on what to print in the @code{function} callback. Using a strategy
25454like this is a way to defer decisions on the frame content to printing
25455time.
25456
25457@subheading Eliding Frames
25458
25459It might be that the above example is not desirable for representing
25460inlined frames, and a hierarchical approach may be preferred. If we
25461want to hierarchically represent frames, the @code{elided} frame
25462decorator interface might be preferable.
25463
25464This example approaches the issue with the @code{elided} method. This
25465example is quite long, but very simplistic. It is out-of-scope for
25466this section to write a complete example that comprehensively covers
25467all approaches of finding and printing inlined frames. However, this
25468example illustrates the approach an author might use.
25469
25470This example comprises of three sections.
25471
25472@smallexample
25473class InlineFrameFilter():
25474
25475 def __init__(self):
25476 self.name = "InlinedFrameFilter"
25477 self.priority = 100
25478 self.enabled = True
25479 gdb.frame_filters[self.name] = self
25480
25481 def filter(self, frame_iter):
25482 return ElidingInlineIterator(frame_iter)
25483@end smallexample
25484
25485This frame filter is very similar to the other examples. The only
25486difference is this frame filter is wrapping the iterator provided to
25487it (@code{frame_iter}) with a custom iterator called
25488@code{ElidingInlineIterator}. This again defers actions to when
25489@value{GDBN} prints the backtrace, as the iterator is not traversed
25490until printing.
25491
25492The iterator for this example is as follows. It is in this section of
25493the example where decisions are made on the content of the backtrace.
25494
25495@smallexample
25496class ElidingInlineIterator:
25497 def __init__(self, ii):
25498 self.input_iterator = ii
25499
25500 def __iter__(self):
25501 return self
25502
25503 def next(self):
25504 frame = next(self.input_iterator)
25505
25506 if frame.inferior_frame().type() != gdb.INLINE_FRAME:
25507 return frame
25508
25509 try:
25510 eliding_frame = next(self.input_iterator)
25511 except StopIteration:
25512 return frame
25513 return ElidingFrameDecorator(eliding_frame, [frame])
25514@end smallexample
25515
25516This iterator implements the Python iterator protocol. When the
25517@code{next} function is called (when @value{GDBN} prints each frame),
25518the iterator checks if this frame decorator, @code{frame}, is wrapping
25519an inlined frame. If it is not, it returns the existing frame decorator
25520untouched. If it is wrapping an inlined frame, it assumes that the
25521inlined frame was contained within the next oldest frame,
25522@code{eliding_frame}, which it fetches. It then creates and returns a
25523frame decorator, @code{ElidingFrameDecorator}, which contains both the
25524elided frame, and the eliding frame.
25525
25526@smallexample
25527class ElidingInlineDecorator(FrameDecorator):
25528
25529 def __init__(self, frame, elided_frames):
25530 super(ElidingInlineDecorator, self).__init__(frame)
25531 self.frame = frame
25532 self.elided_frames = elided_frames
25533
25534 def elided(self):
25535 return iter(self.elided_frames)
25536@end smallexample
25537
25538This frame decorator overrides one function and returns the inlined
25539frame in the @code{elided} method. As before it lets
25540@code{FrameDecorator} do the rest of the work involved in printing
25541this frame. This produces the following output.
25542
25543@smallexample
25544#0 0x004004e0 in bar () at inline.c:11
25545#2 0x00400529 in main () at inline.c:25
25546 #1 0x00400529 in max (b=6, a=12) at inline.c:15
25547@end smallexample
25548
25549In that output, @code{max} which has been inlined into @code{main} is
25550printed hierarchically. Another approach would be to combine the
25551@code{function} method, and the @code{elided} method to both print a
25552marker in the inlined frame, and also show the hierarchical
25553relationship.
25554
25555@node Inferiors In Python
25556@subsubsection Inferiors In Python
25557@cindex inferiors in Python
25558
25559@findex gdb.Inferior
25560Programs which are being run under @value{GDBN} are called inferiors
25561(@pxref{Inferiors and Programs}). Python scripts can access
25562information about and manipulate inferiors controlled by @value{GDBN}
25563via objects of the @code{gdb.Inferior} class.
25564
25565The following inferior-related functions are available in the @code{gdb}
25566module:
25567
25568@defun gdb.inferiors ()
25569Return a tuple containing all inferior objects.
25570@end defun
25571
25572@defun gdb.selected_inferior ()
25573Return an object representing the current inferior.
25574@end defun
25575
25576A @code{gdb.Inferior} object has the following attributes:
25577
25578@defvar Inferior.num
25579ID of inferior, as assigned by GDB.
25580@end defvar
25581
25582@defvar Inferior.pid
25583Process ID of the inferior, as assigned by the underlying operating
25584system.
25585@end defvar
25586
25587@defvar Inferior.was_attached
25588Boolean signaling whether the inferior was created using `attach', or
25589started by @value{GDBN} itself.
25590@end defvar
25591
25592A @code{gdb.Inferior} object has the following methods:
25593
25594@defun Inferior.is_valid ()
25595Returns @code{True} if the @code{gdb.Inferior} object is valid,
25596@code{False} if not. A @code{gdb.Inferior} object will become invalid
25597if the inferior no longer exists within @value{GDBN}. All other
25598@code{gdb.Inferior} methods will throw an exception if it is invalid
25599at the time the method is called.
25600@end defun
25601
25602@defun Inferior.threads ()
25603This method returns a tuple holding all the threads which are valid
25604when it is called. If there are no valid threads, the method will
25605return an empty tuple.
25606@end defun
25607
25608@findex Inferior.read_memory
25609@defun Inferior.read_memory (address, length)
25610Read @var{length} bytes of memory from the inferior, starting at
25611@var{address}. Returns a buffer object, which behaves much like an array
25612or a string. It can be modified and given to the
25613@code{Inferior.write_memory} function. In @code{Python} 3, the return
25614value is a @code{memoryview} object.
25615@end defun
25616
25617@findex Inferior.write_memory
25618@defun Inferior.write_memory (address, buffer @r{[}, length@r{]})
25619Write the contents of @var{buffer} to the inferior, starting at
25620@var{address}. The @var{buffer} parameter must be a Python object
25621which supports the buffer protocol, i.e., a string, an array or the
25622object returned from @code{Inferior.read_memory}. If given, @var{length}
25623determines the number of bytes from @var{buffer} to be written.
25624@end defun
25625
25626@findex gdb.search_memory
25627@defun Inferior.search_memory (address, length, pattern)
25628Search a region of the inferior memory starting at @var{address} with
25629the given @var{length} using the search pattern supplied in
25630@var{pattern}. The @var{pattern} parameter must be a Python object
25631which supports the buffer protocol, i.e., a string, an array or the
25632object returned from @code{gdb.read_memory}. Returns a Python @code{Long}
25633containing the address where the pattern was found, or @code{None} if
25634the pattern could not be found.
25635@end defun
25636
25637@node Events In Python
25638@subsubsection Events In Python
25639@cindex inferior events in Python
25640
25641@value{GDBN} provides a general event facility so that Python code can be
25642notified of various state changes, particularly changes that occur in
25643the inferior.
25644
25645An @dfn{event} is just an object that describes some state change. The
25646type of the object and its attributes will vary depending on the details
25647of the change. All the existing events are described below.
25648
25649In order to be notified of an event, you must register an event handler
25650with an @dfn{event registry}. An event registry is an object in the
25651@code{gdb.events} module which dispatches particular events. A registry
25652provides methods to register and unregister event handlers:
25653
25654@defun EventRegistry.connect (object)
25655Add the given callable @var{object} to the registry. This object will be
25656called when an event corresponding to this registry occurs.
25657@end defun
25658
25659@defun EventRegistry.disconnect (object)
25660Remove the given @var{object} from the registry. Once removed, the object
25661will no longer receive notifications of events.
25662@end defun
25663
25664Here is an example:
25665
25666@smallexample
25667def exit_handler (event):
25668 print "event type: exit"
25669 print "exit code: %d" % (event.exit_code)
25670
25671gdb.events.exited.connect (exit_handler)
25672@end smallexample
25673
25674In the above example we connect our handler @code{exit_handler} to the
25675registry @code{events.exited}. Once connected, @code{exit_handler} gets
25676called when the inferior exits. The argument @dfn{event} in this example is
25677of type @code{gdb.ExitedEvent}. As you can see in the example the
25678@code{ExitedEvent} object has an attribute which indicates the exit code of
25679the inferior.
25680
25681The following is a listing of the event registries that are available and
25682details of the events they emit:
25683
25684@table @code
25685
25686@item events.cont
25687Emits @code{gdb.ThreadEvent}.
25688
25689Some events can be thread specific when @value{GDBN} is running in non-stop
25690mode. When represented in Python, these events all extend
25691@code{gdb.ThreadEvent}. Note, this event is not emitted directly; instead,
25692events which are emitted by this or other modules might extend this event.
25693Examples of these events are @code{gdb.BreakpointEvent} and
25694@code{gdb.ContinueEvent}.
25695
25696@defvar ThreadEvent.inferior_thread
25697In non-stop mode this attribute will be set to the specific thread which was
25698involved in the emitted event. Otherwise, it will be set to @code{None}.
25699@end defvar
25700
25701Emits @code{gdb.ContinueEvent} which extends @code{gdb.ThreadEvent}.
25702
25703This event indicates that the inferior has been continued after a stop. For
25704inherited attribute refer to @code{gdb.ThreadEvent} above.
25705
25706@item events.exited
25707Emits @code{events.ExitedEvent} which indicates that the inferior has exited.
25708@code{events.ExitedEvent} has two attributes:
25709@defvar ExitedEvent.exit_code
25710An integer representing the exit code, if available, which the inferior
25711has returned. (The exit code could be unavailable if, for example,
25712@value{GDBN} detaches from the inferior.) If the exit code is unavailable,
25713the attribute does not exist.
25714@end defvar
25715@defvar ExitedEvent inferior
25716A reference to the inferior which triggered the @code{exited} event.
25717@end defvar
25718
25719@item events.stop
25720Emits @code{gdb.StopEvent} which extends @code{gdb.ThreadEvent}.
25721
25722Indicates that the inferior has stopped. All events emitted by this registry
25723extend StopEvent. As a child of @code{gdb.ThreadEvent}, @code{gdb.StopEvent}
25724will indicate the stopped thread when @value{GDBN} is running in non-stop
25725mode. Refer to @code{gdb.ThreadEvent} above for more details.
25726
25727Emits @code{gdb.SignalEvent} which extends @code{gdb.StopEvent}.
25728
25729This event indicates that the inferior or one of its threads has received as
25730signal. @code{gdb.SignalEvent} has the following attributes:
25731
25732@defvar SignalEvent.stop_signal
25733A string representing the signal received by the inferior. A list of possible
25734signal values can be obtained by running the command @code{info signals} in
25735the @value{GDBN} command prompt.
25736@end defvar
25737
25738Also emits @code{gdb.BreakpointEvent} which extends @code{gdb.StopEvent}.
25739
25740@code{gdb.BreakpointEvent} event indicates that one or more breakpoints have
25741been hit, and has the following attributes:
25742
25743@defvar BreakpointEvent.breakpoints
25744A sequence containing references to all the breakpoints (type
25745@code{gdb.Breakpoint}) that were hit.
25746@xref{Breakpoints In Python}, for details of the @code{gdb.Breakpoint} object.
25747@end defvar
25748@defvar BreakpointEvent.breakpoint
25749A reference to the first breakpoint that was hit.
25750This function is maintained for backward compatibility and is now deprecated
25751in favor of the @code{gdb.BreakpointEvent.breakpoints} attribute.
25752@end defvar
25753
25754@item events.new_objfile
25755Emits @code{gdb.NewObjFileEvent} which indicates that a new object file has
25756been loaded by @value{GDBN}. @code{gdb.NewObjFileEvent} has one attribute:
25757
25758@defvar NewObjFileEvent.new_objfile
25759A reference to the object file (@code{gdb.Objfile}) which has been loaded.
25760@xref{Objfiles In Python}, for details of the @code{gdb.Objfile} object.
25761@end defvar
25762
25763@end table
25764
25765@node Threads In Python
25766@subsubsection Threads In Python
25767@cindex threads in python
25768
25769@findex gdb.InferiorThread
25770Python scripts can access information about, and manipulate inferior threads
25771controlled by @value{GDBN}, via objects of the @code{gdb.InferiorThread} class.
25772
25773The following thread-related functions are available in the @code{gdb}
25774module:
25775
25776@findex gdb.selected_thread
25777@defun gdb.selected_thread ()
25778This function returns the thread object for the selected thread. If there
25779is no selected thread, this will return @code{None}.
25780@end defun
25781
25782A @code{gdb.InferiorThread} object has the following attributes:
25783
25784@defvar InferiorThread.name
25785The name of the thread. If the user specified a name using
25786@code{thread name}, then this returns that name. Otherwise, if an
25787OS-supplied name is available, then it is returned. Otherwise, this
25788returns @code{None}.
25789
25790This attribute can be assigned to. The new value must be a string
25791object, which sets the new name, or @code{None}, which removes any
25792user-specified thread name.
25793@end defvar
25794
25795@defvar InferiorThread.num
25796ID of the thread, as assigned by GDB.
25797@end defvar
25798
25799@defvar InferiorThread.ptid
25800ID of the thread, as assigned by the operating system. This attribute is a
25801tuple containing three integers. The first is the Process ID (PID); the second
25802is the Lightweight Process ID (LWPID), and the third is the Thread ID (TID).
25803Either the LWPID or TID may be 0, which indicates that the operating system
25804does not use that identifier.
25805@end defvar
25806
25807A @code{gdb.InferiorThread} object has the following methods:
25808
25809@defun InferiorThread.is_valid ()
25810Returns @code{True} if the @code{gdb.InferiorThread} object is valid,
25811@code{False} if not. A @code{gdb.InferiorThread} object will become
25812invalid if the thread exits, or the inferior that the thread belongs
25813is deleted. All other @code{gdb.InferiorThread} methods will throw an
25814exception if it is invalid at the time the method is called.
25815@end defun
25816
25817@defun InferiorThread.switch ()
25818This changes @value{GDBN}'s currently selected thread to the one represented
25819by this object.
25820@end defun
25821
25822@defun InferiorThread.is_stopped ()
25823Return a Boolean indicating whether the thread is stopped.
25824@end defun
25825
25826@defun InferiorThread.is_running ()
25827Return a Boolean indicating whether the thread is running.
25828@end defun
25829
25830@defun InferiorThread.is_exited ()
25831Return a Boolean indicating whether the thread is exited.
25832@end defun
25833
25834@node Commands In Python
25835@subsubsection Commands In Python
25836
25837@cindex commands in python
25838@cindex python commands
25839You can implement new @value{GDBN} CLI commands in Python. A CLI
25840command is implemented using an instance of the @code{gdb.Command}
25841class, most commonly using a subclass.
25842
25843@defun Command.__init__ (name, @var{command_class} @r{[}, @var{completer_class} @r{[}, @var{prefix}@r{]]})
25844The object initializer for @code{Command} registers the new command
25845with @value{GDBN}. This initializer is normally invoked from the
25846subclass' own @code{__init__} method.
25847
25848@var{name} is the name of the command. If @var{name} consists of
25849multiple words, then the initial words are looked for as prefix
25850commands. In this case, if one of the prefix commands does not exist,
25851an exception is raised.
25852
25853There is no support for multi-line commands.
25854
25855@var{command_class} should be one of the @samp{COMMAND_} constants
25856defined below. This argument tells @value{GDBN} how to categorize the
25857new command in the help system.
25858
25859@var{completer_class} is an optional argument. If given, it should be
25860one of the @samp{COMPLETE_} constants defined below. This argument
25861tells @value{GDBN} how to perform completion for this command. If not
25862given, @value{GDBN} will attempt to complete using the object's
25863@code{complete} method (see below); if no such method is found, an
25864error will occur when completion is attempted.
25865
25866@var{prefix} is an optional argument. If @code{True}, then the new
25867command is a prefix command; sub-commands of this command may be
25868registered.
25869
25870The help text for the new command is taken from the Python
25871documentation string for the command's class, if there is one. If no
25872documentation string is provided, the default value ``This command is
25873not documented.'' is used.
25874@end defun
25875
25876@cindex don't repeat Python command
25877@defun Command.dont_repeat ()
25878By default, a @value{GDBN} command is repeated when the user enters a
25879blank line at the command prompt. A command can suppress this
25880behavior by invoking the @code{dont_repeat} method. This is similar
25881to the user command @code{dont-repeat}, see @ref{Define, dont-repeat}.
25882@end defun
25883
25884@defun Command.invoke (argument, from_tty)
25885This method is called by @value{GDBN} when this command is invoked.
25886
25887@var{argument} is a string. It is the argument to the command, after
25888leading and trailing whitespace has been stripped.
25889
25890@var{from_tty} is a boolean argument. When true, this means that the
25891command was entered by the user at the terminal; when false it means
25892that the command came from elsewhere.
25893
25894If this method throws an exception, it is turned into a @value{GDBN}
25895@code{error} call. Otherwise, the return value is ignored.
25896
25897@findex gdb.string_to_argv
25898To break @var{argument} up into an argv-like string use
25899@code{gdb.string_to_argv}. This function behaves identically to
25900@value{GDBN}'s internal argument lexer @code{buildargv}.
25901It is recommended to use this for consistency.
25902Arguments are separated by spaces and may be quoted.
25903Example:
25904
25905@smallexample
25906print gdb.string_to_argv ("1 2\ \\\"3 '4 \"5' \"6 '7\"")
25907['1', '2 "3', '4 "5', "6 '7"]
25908@end smallexample
25909
25910@end defun
25911
25912@cindex completion of Python commands
25913@defun Command.complete (text, word)
25914This method is called by @value{GDBN} when the user attempts
25915completion on this command. All forms of completion are handled by
25916this method, that is, the @key{TAB} and @key{M-?} key bindings
25917(@pxref{Completion}), and the @code{complete} command (@pxref{Help,
25918complete}).
25919
25920The arguments @var{text} and @var{word} are both strings. @var{text}
25921holds the complete command line up to the cursor's location.
25922@var{word} holds the last word of the command line; this is computed
25923using a word-breaking heuristic.
25924
25925The @code{complete} method can return several values:
25926@itemize @bullet
25927@item
25928If the return value is a sequence, the contents of the sequence are
25929used as the completions. It is up to @code{complete} to ensure that the
25930contents actually do complete the word. A zero-length sequence is
25931allowed, it means that there were no completions available. Only
25932string elements of the sequence are used; other elements in the
25933sequence are ignored.
25934
25935@item
25936If the return value is one of the @samp{COMPLETE_} constants defined
25937below, then the corresponding @value{GDBN}-internal completion
25938function is invoked, and its result is used.
25939
25940@item
25941All other results are treated as though there were no available
25942completions.
25943@end itemize
25944@end defun
25945
25946When a new command is registered, it must be declared as a member of
25947some general class of commands. This is used to classify top-level
25948commands in the on-line help system; note that prefix commands are not
25949listed under their own category but rather that of their top-level
25950command. The available classifications are represented by constants
25951defined in the @code{gdb} module:
25952
25953@table @code
25954@findex COMMAND_NONE
25955@findex gdb.COMMAND_NONE
25956@item gdb.COMMAND_NONE
25957The command does not belong to any particular class. A command in
25958this category will not be displayed in any of the help categories.
25959
25960@findex COMMAND_RUNNING
25961@findex gdb.COMMAND_RUNNING
25962@item gdb.COMMAND_RUNNING
25963The command is related to running the inferior. For example,
25964@code{start}, @code{step}, and @code{continue} are in this category.
25965Type @kbd{help running} at the @value{GDBN} prompt to see a list of
25966commands in this category.
25967
25968@findex COMMAND_DATA
25969@findex gdb.COMMAND_DATA
25970@item gdb.COMMAND_DATA
25971The command is related to data or variables. For example,
25972@code{call}, @code{find}, and @code{print} are in this category. Type
25973@kbd{help data} at the @value{GDBN} prompt to see a list of commands
25974in this category.
25975
25976@findex COMMAND_STACK
25977@findex gdb.COMMAND_STACK
25978@item gdb.COMMAND_STACK
25979The command has to do with manipulation of the stack. For example,
25980@code{backtrace}, @code{frame}, and @code{return} are in this
25981category. Type @kbd{help stack} at the @value{GDBN} prompt to see a
25982list of commands in this category.
25983
25984@findex COMMAND_FILES
25985@findex gdb.COMMAND_FILES
25986@item gdb.COMMAND_FILES
25987This class is used for file-related commands. For example,
25988@code{file}, @code{list} and @code{section} are in this category.
25989Type @kbd{help files} at the @value{GDBN} prompt to see a list of
25990commands in this category.
25991
25992@findex COMMAND_SUPPORT
25993@findex gdb.COMMAND_SUPPORT
25994@item gdb.COMMAND_SUPPORT
25995This should be used for ``support facilities'', generally meaning
25996things that are useful to the user when interacting with @value{GDBN},
25997but not related to the state of the inferior. For example,
25998@code{help}, @code{make}, and @code{shell} are in this category. Type
25999@kbd{help support} at the @value{GDBN} prompt to see a list of
26000commands in this category.
26001
26002@findex COMMAND_STATUS
26003@findex gdb.COMMAND_STATUS
26004@item gdb.COMMAND_STATUS
26005The command is an @samp{info}-related command, that is, related to the
26006state of @value{GDBN} itself. For example, @code{info}, @code{macro},
26007and @code{show} are in this category. Type @kbd{help status} at the
26008@value{GDBN} prompt to see a list of commands in this category.
26009
26010@findex COMMAND_BREAKPOINTS
26011@findex gdb.COMMAND_BREAKPOINTS
26012@item gdb.COMMAND_BREAKPOINTS
26013The command has to do with breakpoints. For example, @code{break},
26014@code{clear}, and @code{delete} are in this category. Type @kbd{help
26015breakpoints} at the @value{GDBN} prompt to see a list of commands in
26016this category.
26017
26018@findex COMMAND_TRACEPOINTS
26019@findex gdb.COMMAND_TRACEPOINTS
26020@item gdb.COMMAND_TRACEPOINTS
26021The command has to do with tracepoints. For example, @code{trace},
26022@code{actions}, and @code{tfind} are in this category. Type
26023@kbd{help tracepoints} at the @value{GDBN} prompt to see a list of
26024commands in this category.
26025
26026@findex COMMAND_USER
26027@findex gdb.COMMAND_USER
26028@item gdb.COMMAND_USER
26029The command is a general purpose command for the user, and typically
26030does not fit in one of the other categories.
26031Type @kbd{help user-defined} at the @value{GDBN} prompt to see
26032a list of commands in this category, as well as the list of gdb macros
26033(@pxref{Sequences}).
26034
26035@findex COMMAND_OBSCURE
26036@findex gdb.COMMAND_OBSCURE
26037@item gdb.COMMAND_OBSCURE
26038The command is only used in unusual circumstances, or is not of
26039general interest to users. For example, @code{checkpoint},
26040@code{fork}, and @code{stop} are in this category. Type @kbd{help
26041obscure} at the @value{GDBN} prompt to see a list of commands in this
26042category.
26043
26044@findex COMMAND_MAINTENANCE
26045@findex gdb.COMMAND_MAINTENANCE
26046@item gdb.COMMAND_MAINTENANCE
26047The command is only useful to @value{GDBN} maintainers. The
26048@code{maintenance} and @code{flushregs} commands are in this category.
26049Type @kbd{help internals} at the @value{GDBN} prompt to see a list of
26050commands in this category.
26051@end table
26052
26053A new command can use a predefined completion function, either by
26054specifying it via an argument at initialization, or by returning it
26055from the @code{complete} method. These predefined completion
26056constants are all defined in the @code{gdb} module:
26057
26058@table @code
26059@findex COMPLETE_NONE
26060@findex gdb.COMPLETE_NONE
26061@item gdb.COMPLETE_NONE
26062This constant means that no completion should be done.
26063
26064@findex COMPLETE_FILENAME
26065@findex gdb.COMPLETE_FILENAME
26066@item gdb.COMPLETE_FILENAME
26067This constant means that filename completion should be performed.
26068
26069@findex COMPLETE_LOCATION
26070@findex gdb.COMPLETE_LOCATION
26071@item gdb.COMPLETE_LOCATION
26072This constant means that location completion should be done.
26073@xref{Specify Location}.
26074
26075@findex COMPLETE_COMMAND
26076@findex gdb.COMPLETE_COMMAND
26077@item gdb.COMPLETE_COMMAND
26078This constant means that completion should examine @value{GDBN}
26079command names.
26080
26081@findex COMPLETE_SYMBOL
26082@findex gdb.COMPLETE_SYMBOL
26083@item gdb.COMPLETE_SYMBOL
26084This constant means that completion should be done using symbol names
26085as the source.
26086
26087@findex COMPLETE_EXPRESSION
26088@findex gdb.COMPLETE_EXPRESSION
26089@item gdb.COMPLETE_EXPRESSION
26090This constant means that completion should be done on expressions.
26091Often this means completing on symbol names, but some language
26092parsers also have support for completing on field names.
26093@end table
26094
26095The following code snippet shows how a trivial CLI command can be
26096implemented in Python:
26097
26098@smallexample
26099class HelloWorld (gdb.Command):
26100 """Greet the whole world."""
26101
26102 def __init__ (self):
26103 super (HelloWorld, self).__init__ ("hello-world", gdb.COMMAND_USER)
26104
26105 def invoke (self, arg, from_tty):
26106 print "Hello, World!"
26107
26108HelloWorld ()
26109@end smallexample
26110
26111The last line instantiates the class, and is necessary to trigger the
26112registration of the command with @value{GDBN}. Depending on how the
26113Python code is read into @value{GDBN}, you may need to import the
26114@code{gdb} module explicitly.
26115
26116@node Parameters In Python
26117@subsubsection Parameters In Python
26118
26119@cindex parameters in python
26120@cindex python parameters
26121@tindex gdb.Parameter
26122@tindex Parameter
26123You can implement new @value{GDBN} parameters using Python. A new
26124parameter is implemented as an instance of the @code{gdb.Parameter}
26125class.
26126
26127Parameters are exposed to the user via the @code{set} and
26128@code{show} commands. @xref{Help}.
26129
26130There are many parameters that already exist and can be set in
26131@value{GDBN}. Two examples are: @code{set follow fork} and
26132@code{set charset}. Setting these parameters influences certain
26133behavior in @value{GDBN}. Similarly, you can define parameters that
26134can be used to influence behavior in custom Python scripts and commands.
26135
26136@defun Parameter.__init__ (name, @var{command-class}, @var{parameter-class} @r{[}, @var{enum-sequence}@r{]})
26137The object initializer for @code{Parameter} registers the new
26138parameter with @value{GDBN}. This initializer is normally invoked
26139from the subclass' own @code{__init__} method.
26140
26141@var{name} is the name of the new parameter. If @var{name} consists
26142of multiple words, then the initial words are looked for as prefix
26143parameters. An example of this can be illustrated with the
26144@code{set print} set of parameters. If @var{name} is
26145@code{print foo}, then @code{print} will be searched as the prefix
26146parameter. In this case the parameter can subsequently be accessed in
26147@value{GDBN} as @code{set print foo}.
26148
26149If @var{name} consists of multiple words, and no prefix parameter group
26150can be found, an exception is raised.
26151
26152@var{command-class} should be one of the @samp{COMMAND_} constants
26153(@pxref{Commands In Python}). This argument tells @value{GDBN} how to
26154categorize the new parameter in the help system.
26155
26156@var{parameter-class} should be one of the @samp{PARAM_} constants
26157defined below. This argument tells @value{GDBN} the type of the new
26158parameter; this information is used for input validation and
26159completion.
26160
26161If @var{parameter-class} is @code{PARAM_ENUM}, then
26162@var{enum-sequence} must be a sequence of strings. These strings
26163represent the possible values for the parameter.
26164
26165If @var{parameter-class} is not @code{PARAM_ENUM}, then the presence
26166of a fourth argument will cause an exception to be thrown.
26167
26168The help text for the new parameter is taken from the Python
26169documentation string for the parameter's class, if there is one. If
26170there is no documentation string, a default value is used.
26171@end defun
26172
26173@defvar Parameter.set_doc
26174If this attribute exists, and is a string, then its value is used as
26175the help text for this parameter's @code{set} command. The value is
26176examined when @code{Parameter.__init__} is invoked; subsequent changes
26177have no effect.
26178@end defvar
26179
26180@defvar Parameter.show_doc
26181If this attribute exists, and is a string, then its value is used as
26182the help text for this parameter's @code{show} command. The value is
26183examined when @code{Parameter.__init__} is invoked; subsequent changes
26184have no effect.
26185@end defvar
26186
26187@defvar Parameter.value
26188The @code{value} attribute holds the underlying value of the
26189parameter. It can be read and assigned to just as any other
26190attribute. @value{GDBN} does validation when assignments are made.
26191@end defvar
26192
26193There are two methods that should be implemented in any
26194@code{Parameter} class. These are:
26195
26196@defun Parameter.get_set_string (self)
26197@value{GDBN} will call this method when a @var{parameter}'s value has
26198been changed via the @code{set} API (for example, @kbd{set foo off}).
26199The @code{value} attribute has already been populated with the new
26200value and may be used in output. This method must return a string.
26201@end defun
26202
26203@defun Parameter.get_show_string (self, svalue)
26204@value{GDBN} will call this method when a @var{parameter}'s
26205@code{show} API has been invoked (for example, @kbd{show foo}). The
26206argument @code{svalue} receives the string representation of the
26207current value. This method must return a string.
26208@end defun
26209
26210When a new parameter is defined, its type must be specified. The
26211available types are represented by constants defined in the @code{gdb}
26212module:
26213
26214@table @code
26215@findex PARAM_BOOLEAN
26216@findex gdb.PARAM_BOOLEAN
26217@item gdb.PARAM_BOOLEAN
26218The value is a plain boolean. The Python boolean values, @code{True}
26219and @code{False} are the only valid values.
26220
26221@findex PARAM_AUTO_BOOLEAN
26222@findex gdb.PARAM_AUTO_BOOLEAN
26223@item gdb.PARAM_AUTO_BOOLEAN
26224The value has three possible states: true, false, and @samp{auto}. In
26225Python, true and false are represented using boolean constants, and
26226@samp{auto} is represented using @code{None}.
26227
26228@findex PARAM_UINTEGER
26229@findex gdb.PARAM_UINTEGER
26230@item gdb.PARAM_UINTEGER
26231The value is an unsigned integer. The value of 0 should be
26232interpreted to mean ``unlimited''.
26233
26234@findex PARAM_INTEGER
26235@findex gdb.PARAM_INTEGER
26236@item gdb.PARAM_INTEGER
26237The value is a signed integer. The value of 0 should be interpreted
26238to mean ``unlimited''.
26239
26240@findex PARAM_STRING
26241@findex gdb.PARAM_STRING
26242@item gdb.PARAM_STRING
26243The value is a string. When the user modifies the string, any escape
26244sequences, such as @samp{\t}, @samp{\f}, and octal escapes, are
26245translated into corresponding characters and encoded into the current
26246host charset.
26247
26248@findex PARAM_STRING_NOESCAPE
26249@findex gdb.PARAM_STRING_NOESCAPE
26250@item gdb.PARAM_STRING_NOESCAPE
26251The value is a string. When the user modifies the string, escapes are
26252passed through untranslated.
26253
26254@findex PARAM_OPTIONAL_FILENAME
26255@findex gdb.PARAM_OPTIONAL_FILENAME
26256@item gdb.PARAM_OPTIONAL_FILENAME
26257The value is a either a filename (a string), or @code{None}.
26258
26259@findex PARAM_FILENAME
26260@findex gdb.PARAM_FILENAME
26261@item gdb.PARAM_FILENAME
26262The value is a filename. This is just like
26263@code{PARAM_STRING_NOESCAPE}, but uses file names for completion.
26264
26265@findex PARAM_ZINTEGER
26266@findex gdb.PARAM_ZINTEGER
26267@item gdb.PARAM_ZINTEGER
26268The value is an integer. This is like @code{PARAM_INTEGER}, except 0
26269is interpreted as itself.
26270
26271@findex PARAM_ENUM
26272@findex gdb.PARAM_ENUM
26273@item gdb.PARAM_ENUM
26274The value is a string, which must be one of a collection string
26275constants provided when the parameter is created.
26276@end table
26277
26278@node Functions In Python
26279@subsubsection Writing new convenience functions
26280
26281@cindex writing convenience functions
26282@cindex convenience functions in python
26283@cindex python convenience functions
26284@tindex gdb.Function
26285@tindex Function
26286You can implement new convenience functions (@pxref{Convenience Vars})
26287in Python. A convenience function is an instance of a subclass of the
26288class @code{gdb.Function}.
26289
26290@defun Function.__init__ (name)
26291The initializer for @code{Function} registers the new function with
26292@value{GDBN}. The argument @var{name} is the name of the function,
26293a string. The function will be visible to the user as a convenience
26294variable of type @code{internal function}, whose name is the same as
26295the given @var{name}.
26296
26297The documentation for the new function is taken from the documentation
26298string for the new class.
26299@end defun
26300
26301@defun Function.invoke (@var{*args})
26302When a convenience function is evaluated, its arguments are converted
26303to instances of @code{gdb.Value}, and then the function's
26304@code{invoke} method is called. Note that @value{GDBN} does not
26305predetermine the arity of convenience functions. Instead, all
26306available arguments are passed to @code{invoke}, following the
26307standard Python calling convention. In particular, a convenience
26308function can have default values for parameters without ill effect.
26309
26310The return value of this method is used as its value in the enclosing
26311expression. If an ordinary Python value is returned, it is converted
26312to a @code{gdb.Value} following the usual rules.
26313@end defun
26314
26315The following code snippet shows how a trivial convenience function can
26316be implemented in Python:
26317
26318@smallexample
26319class Greet (gdb.Function):
26320 """Return string to greet someone.
26321Takes a name as argument."""
26322
26323 def __init__ (self):
26324 super (Greet, self).__init__ ("greet")
26325
26326 def invoke (self, name):
26327 return "Hello, %s!" % name.string ()
26328
26329Greet ()
26330@end smallexample
26331
26332The last line instantiates the class, and is necessary to trigger the
26333registration of the function with @value{GDBN}. Depending on how the
26334Python code is read into @value{GDBN}, you may need to import the
26335@code{gdb} module explicitly.
26336
26337Now you can use the function in an expression:
26338
26339@smallexample
26340(gdb) print $greet("Bob")
26341$1 = "Hello, Bob!"
26342@end smallexample
26343
26344@node Progspaces In Python
26345@subsubsection Program Spaces In Python
26346
26347@cindex progspaces in python
26348@tindex gdb.Progspace
26349@tindex Progspace
26350A program space, or @dfn{progspace}, represents a symbolic view
26351of an address space.
26352It consists of all of the objfiles of the program.
26353@xref{Objfiles In Python}.
26354@xref{Inferiors and Programs, program spaces}, for more details
26355about program spaces.
26356
26357The following progspace-related functions are available in the
26358@code{gdb} module:
26359
26360@findex gdb.current_progspace
26361@defun gdb.current_progspace ()
26362This function returns the program space of the currently selected inferior.
26363@xref{Inferiors and Programs}.
26364@end defun
26365
26366@findex gdb.progspaces
26367@defun gdb.progspaces ()
26368Return a sequence of all the progspaces currently known to @value{GDBN}.
26369@end defun
26370
26371Each progspace is represented by an instance of the @code{gdb.Progspace}
26372class.
26373
26374@defvar Progspace.filename
26375The file name of the progspace as a string.
26376@end defvar
26377
26378@defvar Progspace.pretty_printers
26379The @code{pretty_printers} attribute is a list of functions. It is
26380used to look up pretty-printers. A @code{Value} is passed to each
26381function in order; if the function returns @code{None}, then the
26382search continues. Otherwise, the return value should be an object
26383which is used to format the value. @xref{Pretty Printing API}, for more
26384information.
26385@end defvar
26386
26387@defvar Progspace.type_printers
26388The @code{type_printers} attribute is a list of type printer objects.
26389@xref{Type Printing API}, for more information.
26390@end defvar
26391
26392@defvar Progspace.frame_filters
26393The @code{frame_filters} attribute is a dictionary of frame filter
26394objects. @xref{Frame Filter API}, for more information.
26395@end defvar
26396
26397@node Objfiles In Python
26398@subsubsection Objfiles In Python
26399
26400@cindex objfiles in python
26401@tindex gdb.Objfile
26402@tindex Objfile
26403@value{GDBN} loads symbols for an inferior from various
26404symbol-containing files (@pxref{Files}). These include the primary
26405executable file, any shared libraries used by the inferior, and any
26406separate debug info files (@pxref{Separate Debug Files}).
26407@value{GDBN} calls these symbol-containing files @dfn{objfiles}.
26408
26409The following objfile-related functions are available in the
26410@code{gdb} module:
26411
26412@findex gdb.current_objfile
26413@defun gdb.current_objfile ()
26414When auto-loading a Python script (@pxref{Python Auto-loading}), @value{GDBN}
26415sets the ``current objfile'' to the corresponding objfile. This
26416function returns the current objfile. If there is no current objfile,
26417this function returns @code{None}.
26418@end defun
26419
26420@findex gdb.objfiles
26421@defun gdb.objfiles ()
26422Return a sequence of all the objfiles current known to @value{GDBN}.
26423@xref{Objfiles In Python}.
26424@end defun
26425
26426Each objfile is represented by an instance of the @code{gdb.Objfile}
26427class.
26428
26429@defvar Objfile.filename
26430The file name of the objfile as a string.
26431@end defvar
26432
26433@defvar Objfile.pretty_printers
26434The @code{pretty_printers} attribute is a list of functions. It is
26435used to look up pretty-printers. A @code{Value} is passed to each
26436function in order; if the function returns @code{None}, then the
26437search continues. Otherwise, the return value should be an object
26438which is used to format the value. @xref{Pretty Printing API}, for more
26439information.
26440@end defvar
26441
26442@defvar Objfile.type_printers
26443The @code{type_printers} attribute is a list of type printer objects.
26444@xref{Type Printing API}, for more information.
26445@end defvar
26446
26447@defvar Objfile.frame_filters
26448The @code{frame_filters} attribute is a dictionary of frame filter
26449objects. @xref{Frame Filter API}, for more information.
26450@end defvar
26451
26452A @code{gdb.Objfile} object has the following methods:
26453
26454@defun Objfile.is_valid ()
26455Returns @code{True} if the @code{gdb.Objfile} object is valid,
26456@code{False} if not. A @code{gdb.Objfile} object can become invalid
26457if the object file it refers to is not loaded in @value{GDBN} any
26458longer. All other @code{gdb.Objfile} methods will throw an exception
26459if it is invalid at the time the method is called.
26460@end defun
26461
26462@node Frames In Python
26463@subsubsection Accessing inferior stack frames from Python.
26464
26465@cindex frames in python
26466When the debugged program stops, @value{GDBN} is able to analyze its call
26467stack (@pxref{Frames,,Stack frames}). The @code{gdb.Frame} class
26468represents a frame in the stack. A @code{gdb.Frame} object is only valid
26469while its corresponding frame exists in the inferior's stack. If you try
26470to use an invalid frame object, @value{GDBN} will throw a @code{gdb.error}
26471exception (@pxref{Exception Handling}).
26472
26473Two @code{gdb.Frame} objects can be compared for equality with the @code{==}
26474operator, like:
26475
26476@smallexample
26477(@value{GDBP}) python print gdb.newest_frame() == gdb.selected_frame ()
26478True
26479@end smallexample
26480
26481The following frame-related functions are available in the @code{gdb} module:
26482
26483@findex gdb.selected_frame
26484@defun gdb.selected_frame ()
26485Return the selected frame object. (@pxref{Selection,,Selecting a Frame}).
26486@end defun
26487
26488@findex gdb.newest_frame
26489@defun gdb.newest_frame ()
26490Return the newest frame object for the selected thread.
26491@end defun
26492
26493@defun gdb.frame_stop_reason_string (reason)
26494Return a string explaining the reason why @value{GDBN} stopped unwinding
26495frames, as expressed by the given @var{reason} code (an integer, see the
26496@code{unwind_stop_reason} method further down in this section).
26497@end defun
26498
26499A @code{gdb.Frame} object has the following methods:
26500
26501@defun Frame.is_valid ()
26502Returns true if the @code{gdb.Frame} object is valid, false if not.
26503A frame object can become invalid if the frame it refers to doesn't
26504exist anymore in the inferior. All @code{gdb.Frame} methods will throw
26505an exception if it is invalid at the time the method is called.
26506@end defun
26507
26508@defun Frame.name ()
26509Returns the function name of the frame, or @code{None} if it can't be
26510obtained.
26511@end defun
26512
26513@defun Frame.architecture ()
26514Returns the @code{gdb.Architecture} object corresponding to the frame's
26515architecture. @xref{Architectures In Python}.
26516@end defun
26517
26518@defun Frame.type ()
26519Returns the type of the frame. The value can be one of:
26520@table @code
26521@item gdb.NORMAL_FRAME
26522An ordinary stack frame.
26523
26524@item gdb.DUMMY_FRAME
26525A fake stack frame that was created by @value{GDBN} when performing an
26526inferior function call.
26527
26528@item gdb.INLINE_FRAME
26529A frame representing an inlined function. The function was inlined
26530into a @code{gdb.NORMAL_FRAME} that is older than this one.
26531
26532@item gdb.TAILCALL_FRAME
26533A frame representing a tail call. @xref{Tail Call Frames}.
26534
26535@item gdb.SIGTRAMP_FRAME
26536A signal trampoline frame. This is the frame created by the OS when
26537it calls into a signal handler.
26538
26539@item gdb.ARCH_FRAME
26540A fake stack frame representing a cross-architecture call.
26541
26542@item gdb.SENTINEL_FRAME
26543This is like @code{gdb.NORMAL_FRAME}, but it is only used for the
26544newest frame.
26545@end table
26546@end defun
26547
26548@defun Frame.unwind_stop_reason ()
26549Return an integer representing the reason why it's not possible to find
26550more frames toward the outermost frame. Use
26551@code{gdb.frame_stop_reason_string} to convert the value returned by this
26552function to a string. The value can be one of:
26553
26554@table @code
26555@item gdb.FRAME_UNWIND_NO_REASON
26556No particular reason (older frames should be available).
26557
26558@item gdb.FRAME_UNWIND_NULL_ID
26559The previous frame's analyzer returns an invalid result. This is no
26560longer used by @value{GDBN}, and is kept only for backward
26561compatibility.
26562
26563@item gdb.FRAME_UNWIND_OUTERMOST
26564This frame is the outermost.
26565
26566@item gdb.FRAME_UNWIND_UNAVAILABLE
26567Cannot unwind further, because that would require knowing the
26568values of registers or memory that have not been collected.
26569
26570@item gdb.FRAME_UNWIND_INNER_ID
26571This frame ID looks like it ought to belong to a NEXT frame,
26572but we got it for a PREV frame. Normally, this is a sign of
26573unwinder failure. It could also indicate stack corruption.
26574
26575@item gdb.FRAME_UNWIND_SAME_ID
26576This frame has the same ID as the previous one. That means
26577that unwinding further would almost certainly give us another
26578frame with exactly the same ID, so break the chain. Normally,
26579this is a sign of unwinder failure. It could also indicate
26580stack corruption.
26581
26582@item gdb.FRAME_UNWIND_NO_SAVED_PC
26583The frame unwinder did not find any saved PC, but we needed
26584one to unwind further.
26585
26586@item gdb.FRAME_UNWIND_FIRST_ERROR
26587Any stop reason greater or equal to this value indicates some kind
26588of error. This special value facilitates writing code that tests
26589for errors in unwinding in a way that will work correctly even if
26590the list of the other values is modified in future @value{GDBN}
26591versions. Using it, you could write:
26592@smallexample
26593reason = gdb.selected_frame().unwind_stop_reason ()
26594reason_str = gdb.frame_stop_reason_string (reason)
26595if reason >= gdb.FRAME_UNWIND_FIRST_ERROR:
26596 print "An error occured: %s" % reason_str
26597@end smallexample
26598@end table
26599
26600@end defun
26601
26602@defun Frame.pc ()
26603Returns the frame's resume address.
26604@end defun
26605
26606@defun Frame.block ()
26607Return the frame's code block. @xref{Blocks In Python}.
26608@end defun
26609
26610@defun Frame.function ()
26611Return the symbol for the function corresponding to this frame.
26612@xref{Symbols In Python}.
26613@end defun
26614
26615@defun Frame.older ()
26616Return the frame that called this frame.
26617@end defun
26618
26619@defun Frame.newer ()
26620Return the frame called by this frame.
26621@end defun
26622
26623@defun Frame.find_sal ()
26624Return the frame's symtab and line object.
26625@xref{Symbol Tables In Python}.
26626@end defun
26627
26628@defun Frame.read_var (variable @r{[}, block@r{]})
26629Return the value of @var{variable} in this frame. If the optional
26630argument @var{block} is provided, search for the variable from that
26631block; otherwise start at the frame's current block (which is
26632determined by the frame's current program counter). @var{variable}
26633must be a string or a @code{gdb.Symbol} object. @var{block} must be a
26634@code{gdb.Block} object.
26635@end defun
26636
26637@defun Frame.select ()
26638Set this frame to be the selected frame. @xref{Stack, ,Examining the
26639Stack}.
26640@end defun
26641
26642@node Blocks In Python
26643@subsubsection Accessing blocks from Python.
26644
26645@cindex blocks in python
26646@tindex gdb.Block
26647
26648In @value{GDBN}, symbols are stored in blocks. A block corresponds
26649roughly to a scope in the source code. Blocks are organized
26650hierarchically, and are represented individually in Python as a
26651@code{gdb.Block}. Blocks rely on debugging information being
26652available.
26653
26654A frame has a block. Please see @ref{Frames In Python}, for a more
26655in-depth discussion of frames.
26656
26657The outermost block is known as the @dfn{global block}. The global
26658block typically holds public global variables and functions.
26659
26660The block nested just inside the global block is the @dfn{static
26661block}. The static block typically holds file-scoped variables and
26662functions.
26663
26664@value{GDBN} provides a method to get a block's superblock, but there
26665is currently no way to examine the sub-blocks of a block, or to
26666iterate over all the blocks in a symbol table (@pxref{Symbol Tables In
26667Python}).
26668
26669Here is a short example that should help explain blocks:
26670
26671@smallexample
26672/* This is in the global block. */
26673int global;
26674
26675/* This is in the static block. */
26676static int file_scope;
26677
26678/* 'function' is in the global block, and 'argument' is
26679 in a block nested inside of 'function'. */
26680int function (int argument)
26681@{
26682 /* 'local' is in a block inside 'function'. It may or may
26683 not be in the same block as 'argument'. */
26684 int local;
26685
26686 @{
26687 /* 'inner' is in a block whose superblock is the one holding
26688 'local'. */
26689 int inner;
26690
26691 /* If this call is expanded by the compiler, you may see
26692 a nested block here whose function is 'inline_function'
26693 and whose superblock is the one holding 'inner'. */
26694 inline_function ();
26695 @}
26696@}
26697@end smallexample
26698
26699A @code{gdb.Block} is iterable. The iterator returns the symbols
26700(@pxref{Symbols In Python}) local to the block. Python programs
26701should not assume that a specific block object will always contain a
26702given symbol, since changes in @value{GDBN} features and
26703infrastructure may cause symbols move across blocks in a symbol
26704table.
26705
26706The following block-related functions are available in the @code{gdb}
26707module:
26708
26709@findex gdb.block_for_pc
26710@defun gdb.block_for_pc (pc)
26711Return the innermost @code{gdb.Block} containing the given @var{pc}
26712value. If the block cannot be found for the @var{pc} value specified,
26713the function will return @code{None}.
26714@end defun
26715
26716A @code{gdb.Block} object has the following methods:
26717
26718@defun Block.is_valid ()
26719Returns @code{True} if the @code{gdb.Block} object is valid,
26720@code{False} if not. A block object can become invalid if the block it
26721refers to doesn't exist anymore in the inferior. All other
26722@code{gdb.Block} methods will throw an exception if it is invalid at
26723the time the method is called. The block's validity is also checked
26724during iteration over symbols of the block.
26725@end defun
26726
26727A @code{gdb.Block} object has the following attributes:
26728
26729@defvar Block.start
26730The start address of the block. This attribute is not writable.
26731@end defvar
26732
26733@defvar Block.end
26734The end address of the block. This attribute is not writable.
26735@end defvar
26736
26737@defvar Block.function
26738The name of the block represented as a @code{gdb.Symbol}. If the
26739block is not named, then this attribute holds @code{None}. This
26740attribute is not writable.
26741
26742For ordinary function blocks, the superblock is the static block.
26743However, you should note that it is possible for a function block to
26744have a superblock that is not the static block -- for instance this
26745happens for an inlined function.
26746@end defvar
26747
26748@defvar Block.superblock
26749The block containing this block. If this parent block does not exist,
26750this attribute holds @code{None}. This attribute is not writable.
26751@end defvar
26752
26753@defvar Block.global_block
26754The global block associated with this block. This attribute is not
26755writable.
26756@end defvar
26757
26758@defvar Block.static_block
26759The static block associated with this block. This attribute is not
26760writable.
26761@end defvar
26762
26763@defvar Block.is_global
26764@code{True} if the @code{gdb.Block} object is a global block,
26765@code{False} if not. This attribute is not
26766writable.
26767@end defvar
26768
26769@defvar Block.is_static
26770@code{True} if the @code{gdb.Block} object is a static block,
26771@code{False} if not. This attribute is not writable.
26772@end defvar
26773
26774@node Symbols In Python
26775@subsubsection Python representation of Symbols.
26776
26777@cindex symbols in python
26778@tindex gdb.Symbol
26779
26780@value{GDBN} represents every variable, function and type as an
26781entry in a symbol table. @xref{Symbols, ,Examining the Symbol Table}.
26782Similarly, Python represents these symbols in @value{GDBN} with the
26783@code{gdb.Symbol} object.
26784
26785The following symbol-related functions are available in the @code{gdb}
26786module:
26787
26788@findex gdb.lookup_symbol
26789@defun gdb.lookup_symbol (name @r{[}, block @r{[}, domain@r{]]})
26790This function searches for a symbol by name. The search scope can be
26791restricted to the parameters defined in the optional domain and block
26792arguments.
26793
26794@var{name} is the name of the symbol. It must be a string. The
26795optional @var{block} argument restricts the search to symbols visible
26796in that @var{block}. The @var{block} argument must be a
26797@code{gdb.Block} object. If omitted, the block for the current frame
26798is used. The optional @var{domain} argument restricts
26799the search to the domain type. The @var{domain} argument must be a
26800domain constant defined in the @code{gdb} module and described later
26801in this chapter.
26802
26803The result is a tuple of two elements.
26804The first element is a @code{gdb.Symbol} object or @code{None} if the symbol
26805is not found.
26806If the symbol is found, the second element is @code{True} if the symbol
26807is a field of a method's object (e.g., @code{this} in C@t{++}),
26808otherwise it is @code{False}.
26809If the symbol is not found, the second element is @code{False}.
26810@end defun
26811
26812@findex gdb.lookup_global_symbol
26813@defun gdb.lookup_global_symbol (name @r{[}, domain@r{]})
26814This function searches for a global symbol by name.
26815The search scope can be restricted to by the domain argument.
26816
26817@var{name} is the name of the symbol. It must be a string.
26818The optional @var{domain} argument restricts the search to the domain type.
26819The @var{domain} argument must be a domain constant defined in the @code{gdb}
26820module and described later in this chapter.
26821
26822The result is a @code{gdb.Symbol} object or @code{None} if the symbol
26823is not found.
26824@end defun
26825
26826A @code{gdb.Symbol} object has the following attributes:
26827
26828@defvar Symbol.type
26829The type of the symbol or @code{None} if no type is recorded.
26830This attribute is represented as a @code{gdb.Type} object.
26831@xref{Types In Python}. This attribute is not writable.
26832@end defvar
26833
26834@defvar Symbol.symtab
26835The symbol table in which the symbol appears. This attribute is
26836represented as a @code{gdb.Symtab} object. @xref{Symbol Tables In
26837Python}. This attribute is not writable.
26838@end defvar
26839
26840@defvar Symbol.line
26841The line number in the source code at which the symbol was defined.
26842This is an integer.
26843@end defvar
26844
26845@defvar Symbol.name
26846The name of the symbol as a string. This attribute is not writable.
26847@end defvar
26848
26849@defvar Symbol.linkage_name
26850The name of the symbol, as used by the linker (i.e., may be mangled).
26851This attribute is not writable.
26852@end defvar
26853
26854@defvar Symbol.print_name
26855The name of the symbol in a form suitable for output. This is either
26856@code{name} or @code{linkage_name}, depending on whether the user
26857asked @value{GDBN} to display demangled or mangled names.
26858@end defvar
26859
26860@defvar Symbol.addr_class
26861The address class of the symbol. This classifies how to find the value
26862of a symbol. Each address class is a constant defined in the
26863@code{gdb} module and described later in this chapter.
26864@end defvar
26865
26866@defvar Symbol.needs_frame
26867This is @code{True} if evaluating this symbol's value requires a frame
26868(@pxref{Frames In Python}) and @code{False} otherwise. Typically,
26869local variables will require a frame, but other symbols will not.
26870@end defvar
26871
26872@defvar Symbol.is_argument
26873@code{True} if the symbol is an argument of a function.
26874@end defvar
26875
26876@defvar Symbol.is_constant
26877@code{True} if the symbol is a constant.
26878@end defvar
26879
26880@defvar Symbol.is_function
26881@code{True} if the symbol is a function or a method.
26882@end defvar
26883
26884@defvar Symbol.is_variable
26885@code{True} if the symbol is a variable.
26886@end defvar
26887
26888A @code{gdb.Symbol} object has the following methods:
26889
26890@defun Symbol.is_valid ()
26891Returns @code{True} if the @code{gdb.Symbol} object is valid,
26892@code{False} if not. A @code{gdb.Symbol} object can become invalid if
26893the symbol it refers to does not exist in @value{GDBN} any longer.
26894All other @code{gdb.Symbol} methods will throw an exception if it is
26895invalid at the time the method is called.
26896@end defun
26897
26898@defun Symbol.value (@r{[}frame@r{]})
26899Compute the value of the symbol, as a @code{gdb.Value}. For
26900functions, this computes the address of the function, cast to the
26901appropriate type. If the symbol requires a frame in order to compute
26902its value, then @var{frame} must be given. If @var{frame} is not
26903given, or if @var{frame} is invalid, then this method will throw an
26904exception.
26905@end defun
26906
26907The available domain categories in @code{gdb.Symbol} are represented
26908as constants in the @code{gdb} module:
26909
26910@table @code
26911@findex SYMBOL_UNDEF_DOMAIN
26912@findex gdb.SYMBOL_UNDEF_DOMAIN
26913@item gdb.SYMBOL_UNDEF_DOMAIN
26914This is used when a domain has not been discovered or none of the
26915following domains apply. This usually indicates an error either
26916in the symbol information or in @value{GDBN}'s handling of symbols.
26917@findex SYMBOL_VAR_DOMAIN
26918@findex gdb.SYMBOL_VAR_DOMAIN
26919@item gdb.SYMBOL_VAR_DOMAIN
26920This domain contains variables, function names, typedef names and enum
26921type values.
26922@findex SYMBOL_STRUCT_DOMAIN
26923@findex gdb.SYMBOL_STRUCT_DOMAIN
26924@item gdb.SYMBOL_STRUCT_DOMAIN
26925This domain holds struct, union and enum type names.
26926@findex SYMBOL_LABEL_DOMAIN
26927@findex gdb.SYMBOL_LABEL_DOMAIN
26928@item gdb.SYMBOL_LABEL_DOMAIN
26929This domain contains names of labels (for gotos).
26930@findex SYMBOL_VARIABLES_DOMAIN
26931@findex gdb.SYMBOL_VARIABLES_DOMAIN
26932@item gdb.SYMBOL_VARIABLES_DOMAIN
26933This domain holds a subset of the @code{SYMBOLS_VAR_DOMAIN}; it
26934contains everything minus functions and types.
26935@findex SYMBOL_FUNCTIONS_DOMAIN
26936@findex gdb.SYMBOL_FUNCTIONS_DOMAIN
26937@item gdb.SYMBOL_FUNCTION_DOMAIN
26938This domain contains all functions.
26939@findex SYMBOL_TYPES_DOMAIN
26940@findex gdb.SYMBOL_TYPES_DOMAIN
26941@item gdb.SYMBOL_TYPES_DOMAIN
26942This domain contains all types.
26943@end table
26944
26945The available address class categories in @code{gdb.Symbol} are represented
26946as constants in the @code{gdb} module:
26947
26948@table @code
26949@findex SYMBOL_LOC_UNDEF
26950@findex gdb.SYMBOL_LOC_UNDEF
26951@item gdb.SYMBOL_LOC_UNDEF
26952If this is returned by address class, it indicates an error either in
26953the symbol information or in @value{GDBN}'s handling of symbols.
26954@findex SYMBOL_LOC_CONST
26955@findex gdb.SYMBOL_LOC_CONST
26956@item gdb.SYMBOL_LOC_CONST
26957Value is constant int.
26958@findex SYMBOL_LOC_STATIC
26959@findex gdb.SYMBOL_LOC_STATIC
26960@item gdb.SYMBOL_LOC_STATIC
26961Value is at a fixed address.
26962@findex SYMBOL_LOC_REGISTER
26963@findex gdb.SYMBOL_LOC_REGISTER
26964@item gdb.SYMBOL_LOC_REGISTER
26965Value is in a register.
26966@findex SYMBOL_LOC_ARG
26967@findex gdb.SYMBOL_LOC_ARG
26968@item gdb.SYMBOL_LOC_ARG
26969Value is an argument. This value is at the offset stored within the
26970symbol inside the frame's argument list.
26971@findex SYMBOL_LOC_REF_ARG
26972@findex gdb.SYMBOL_LOC_REF_ARG
26973@item gdb.SYMBOL_LOC_REF_ARG
26974Value address is stored in the frame's argument list. Just like
26975@code{LOC_ARG} except that the value's address is stored at the
26976offset, not the value itself.
26977@findex SYMBOL_LOC_REGPARM_ADDR
26978@findex gdb.SYMBOL_LOC_REGPARM_ADDR
26979@item gdb.SYMBOL_LOC_REGPARM_ADDR
26980Value is a specified register. Just like @code{LOC_REGISTER} except
26981the register holds the address of the argument instead of the argument
26982itself.
26983@findex SYMBOL_LOC_LOCAL
26984@findex gdb.SYMBOL_LOC_LOCAL
26985@item gdb.SYMBOL_LOC_LOCAL
26986Value is a local variable.
26987@findex SYMBOL_LOC_TYPEDEF
26988@findex gdb.SYMBOL_LOC_TYPEDEF
26989@item gdb.SYMBOL_LOC_TYPEDEF
26990Value not used. Symbols in the domain @code{SYMBOL_STRUCT_DOMAIN} all
26991have this class.
26992@findex SYMBOL_LOC_BLOCK
26993@findex gdb.SYMBOL_LOC_BLOCK
26994@item gdb.SYMBOL_LOC_BLOCK
26995Value is a block.
26996@findex SYMBOL_LOC_CONST_BYTES
26997@findex gdb.SYMBOL_LOC_CONST_BYTES
26998@item gdb.SYMBOL_LOC_CONST_BYTES
26999Value is a byte-sequence.
27000@findex SYMBOL_LOC_UNRESOLVED
27001@findex gdb.SYMBOL_LOC_UNRESOLVED
27002@item gdb.SYMBOL_LOC_UNRESOLVED
27003Value is at a fixed address, but the address of the variable has to be
27004determined from the minimal symbol table whenever the variable is
27005referenced.
27006@findex SYMBOL_LOC_OPTIMIZED_OUT
27007@findex gdb.SYMBOL_LOC_OPTIMIZED_OUT
27008@item gdb.SYMBOL_LOC_OPTIMIZED_OUT
27009The value does not actually exist in the program.
27010@findex SYMBOL_LOC_COMPUTED
27011@findex gdb.SYMBOL_LOC_COMPUTED
27012@item gdb.SYMBOL_LOC_COMPUTED
27013The value's address is a computed location.
27014@end table
27015
27016@node Symbol Tables In Python
27017@subsubsection Symbol table representation in Python.
27018
27019@cindex symbol tables in python
27020@tindex gdb.Symtab
27021@tindex gdb.Symtab_and_line
27022
27023Access to symbol table data maintained by @value{GDBN} on the inferior
27024is exposed to Python via two objects: @code{gdb.Symtab_and_line} and
27025@code{gdb.Symtab}. Symbol table and line data for a frame is returned
27026from the @code{find_sal} method in @code{gdb.Frame} object.
27027@xref{Frames In Python}.
27028
27029For more information on @value{GDBN}'s symbol table management, see
27030@ref{Symbols, ,Examining the Symbol Table}, for more information.
27031
27032A @code{gdb.Symtab_and_line} object has the following attributes:
27033
27034@defvar Symtab_and_line.symtab
27035The symbol table object (@code{gdb.Symtab}) for this frame.
27036This attribute is not writable.
27037@end defvar
27038
27039@defvar Symtab_and_line.pc
27040Indicates the start of the address range occupied by code for the
27041current source line. This attribute is not writable.
27042@end defvar
27043
27044@defvar Symtab_and_line.last
27045Indicates the end of the address range occupied by code for the current
27046source line. This attribute is not writable.
27047@end defvar
27048
27049@defvar Symtab_and_line.line
27050Indicates the current line number for this object. This
27051attribute is not writable.
27052@end defvar
27053
27054A @code{gdb.Symtab_and_line} object has the following methods:
27055
27056@defun Symtab_and_line.is_valid ()
27057Returns @code{True} if the @code{gdb.Symtab_and_line} object is valid,
27058@code{False} if not. A @code{gdb.Symtab_and_line} object can become
27059invalid if the Symbol table and line object it refers to does not
27060exist in @value{GDBN} any longer. All other
27061@code{gdb.Symtab_and_line} methods will throw an exception if it is
27062invalid at the time the method is called.
27063@end defun
27064
27065A @code{gdb.Symtab} object has the following attributes:
27066
27067@defvar Symtab.filename
27068The symbol table's source filename. This attribute is not writable.
27069@end defvar
27070
27071@defvar Symtab.objfile
27072The symbol table's backing object file. @xref{Objfiles In Python}.
27073This attribute is not writable.
27074@end defvar
27075
27076A @code{gdb.Symtab} object has the following methods:
27077
27078@defun Symtab.is_valid ()
27079Returns @code{True} if the @code{gdb.Symtab} object is valid,
27080@code{False} if not. A @code{gdb.Symtab} object can become invalid if
27081the symbol table it refers to does not exist in @value{GDBN} any
27082longer. All other @code{gdb.Symtab} methods will throw an exception
27083if it is invalid at the time the method is called.
27084@end defun
27085
27086@defun Symtab.fullname ()
27087Return the symbol table's source absolute file name.
27088@end defun
27089
27090@defun Symtab.global_block ()
27091Return the global block of the underlying symbol table.
27092@xref{Blocks In Python}.
27093@end defun
27094
27095@defun Symtab.static_block ()
27096Return the static block of the underlying symbol table.
27097@xref{Blocks In Python}.
27098@end defun
27099
27100@defun Symtab.linetable ()
27101Return the line table associated with the symbol table.
27102@xref{Line Tables In Python}.
27103@end defun
27104
27105@node Line Tables In Python
27106@subsubsection Manipulating line tables using Python
27107
27108@cindex line tables in python
27109@tindex gdb.LineTable
27110
27111Python code can request and inspect line table information from a
27112symbol table that is loaded in @value{GDBN}. A line table is a
27113mapping of source lines to their executable locations in memory. To
27114acquire the line table information for a particular symbol table, use
27115the @code{linetable} function (@pxref{Symbol Tables In Python}).
27116
27117A @code{gdb.LineTable} is iterable. The iterator returns
27118@code{LineTableEntry} objects that correspond to the source line and
27119address for each line table entry. @code{LineTableEntry} objects have
27120the following attributes:
27121
27122@defvar LineTableEntry.line
27123The source line number for this line table entry. This number
27124corresponds to the actual line of source. This attribute is not
27125writable.
27126@end defvar
27127
27128@defvar LineTableEntry.pc
27129The address that is associated with the line table entry where the
27130executable code for that source line resides in memory. This
27131attribute is not writable.
27132@end defvar
27133
27134As there can be multiple addresses for a single source line, you may
27135receive multiple @code{LineTableEntry} objects with matching
27136@code{line} attributes, but with different @code{pc} attributes. The
27137iterator is sorted in ascending @code{pc} order. Here is a small
27138example illustrating iterating over a line table.
27139
27140@smallexample
27141symtab = gdb.selected_frame().find_sal().symtab
27142linetable = symtab.linetable()
27143for line in linetable:
27144 print "Line: "+str(line.line)+" Address: "+hex(line.pc)
27145@end smallexample
27146
27147This will have the following output:
27148
27149@smallexample
27150Line: 33 Address: 0x4005c8L
27151Line: 37 Address: 0x4005caL
27152Line: 39 Address: 0x4005d2L
27153Line: 40 Address: 0x4005f8L
27154Line: 42 Address: 0x4005ffL
27155Line: 44 Address: 0x400608L
27156Line: 42 Address: 0x40060cL
27157Line: 45 Address: 0x400615L
27158@end smallexample
27159
27160In addition to being able to iterate over a @code{LineTable}, it also
27161has the following direct access methods:
27162
27163@defun LineTable.line (line)
27164Return a Python @code{Tuple} of @code{LineTableEntry} objects for any
27165entries in the line table for the given @var{line}. @var{line} refers
27166to the source code line. If there are no entries for that source code
27167@var{line}, the Python @code{None} is returned.
27168@end defun
27169
27170@defun LineTable.has_line (line)
27171Return a Python @code{Boolean} indicating whether there is an entry in
27172the line table for this source line. Return @code{True} if an entry
27173is found, or @code{False} if not.
27174@end defun
27175
27176@defun LineTable.source_lines ()
27177Return a Python @code{List} of the source line numbers in the symbol
27178table. Only lines with executable code locations are returned. The
27179contents of the @code{List} will just be the source line entries
27180represented as Python @code{Long} values.
27181@end defun
27182
27183@node Breakpoints In Python
27184@subsubsection Manipulating breakpoints using Python
27185
27186@cindex breakpoints in python
27187@tindex gdb.Breakpoint
27188
27189Python code can manipulate breakpoints via the @code{gdb.Breakpoint}
27190class.
27191
27192@defun Breakpoint.__init__ (spec @r{[}, type @r{[}, wp_class @r{[},internal @r{[},temporary@r{]]]]})
27193Create a new breakpoint. @var{spec} is a string naming the location
27194of the breakpoint, or an expression that defines a watchpoint. The
27195contents can be any location recognized by the @code{break} command,
27196or in the case of a watchpoint, by the @code{watch} command. The
27197optional @var{type} denotes the breakpoint to create from the types
27198defined later in this chapter. This argument can be either:
27199@code{gdb.BP_BREAKPOINT} or @code{gdb.BP_WATCHPOINT}. @var{type}
27200defaults to @code{gdb.BP_BREAKPOINT}. The optional @var{internal}
27201argument allows the breakpoint to become invisible to the user. The
27202breakpoint will neither be reported when created, nor will it be
27203listed in the output from @code{info breakpoints} (but will be listed
27204with the @code{maint info breakpoints} command). The optional
27205@var{temporary} argument makes the breakpoint a temporary breakpoint.
27206Temporary breakpoints are deleted after they have been hit. Any
27207further access to the Python breakpoint after it has been hit will
27208result in a runtime error (as that breakpoint has now been
27209automatically deleted). The optional @var{wp_class} argument defines
27210the class of watchpoint to create, if @var{type} is
27211@code{gdb.BP_WATCHPOINT}. If a watchpoint class is not provided, it
27212is assumed to be a @code{gdb.WP_WRITE} class.
27213@end defun
27214
27215@defun Breakpoint.stop (self)
27216The @code{gdb.Breakpoint} class can be sub-classed and, in
27217particular, you may choose to implement the @code{stop} method.
27218If this method is defined in a sub-class of @code{gdb.Breakpoint},
27219it will be called when the inferior reaches any location of a
27220breakpoint which instantiates that sub-class. If the method returns
27221@code{True}, the inferior will be stopped at the location of the
27222breakpoint, otherwise the inferior will continue.
27223
27224If there are multiple breakpoints at the same location with a
27225@code{stop} method, each one will be called regardless of the
27226return status of the previous. This ensures that all @code{stop}
27227methods have a chance to execute at that location. In this scenario
27228if one of the methods returns @code{True} but the others return
27229@code{False}, the inferior will still be stopped.
27230
27231You should not alter the execution state of the inferior (i.e.@:, step,
27232next, etc.), alter the current frame context (i.e.@:, change the current
27233active frame), or alter, add or delete any breakpoint. As a general
27234rule, you should not alter any data within @value{GDBN} or the inferior
27235at this time.
27236
27237Example @code{stop} implementation:
27238
27239@smallexample
27240class MyBreakpoint (gdb.Breakpoint):
27241 def stop (self):
27242 inf_val = gdb.parse_and_eval("foo")
27243 if inf_val == 3:
27244 return True
27245 return False
27246@end smallexample
27247@end defun
27248
27249The available watchpoint types represented by constants are defined in the
27250@code{gdb} module:
27251
27252@table @code
27253@findex WP_READ
27254@findex gdb.WP_READ
27255@item gdb.WP_READ
27256Read only watchpoint.
27257
27258@findex WP_WRITE
27259@findex gdb.WP_WRITE
27260@item gdb.WP_WRITE
27261Write only watchpoint.
27262
27263@findex WP_ACCESS
27264@findex gdb.WP_ACCESS
27265@item gdb.WP_ACCESS
27266Read/Write watchpoint.
27267@end table
27268
27269@defun Breakpoint.is_valid ()
27270Return @code{True} if this @code{Breakpoint} object is valid,
27271@code{False} otherwise. A @code{Breakpoint} object can become invalid
27272if the user deletes the breakpoint. In this case, the object still
27273exists, but the underlying breakpoint does not. In the cases of
27274watchpoint scope, the watchpoint remains valid even if execution of the
27275inferior leaves the scope of that watchpoint.
27276@end defun
27277
27278@defun Breakpoint.delete
27279Permanently deletes the @value{GDBN} breakpoint. This also
27280invalidates the Python @code{Breakpoint} object. Any further access
27281to this object's attributes or methods will raise an error.
27282@end defun
27283
27284@defvar Breakpoint.enabled
27285This attribute is @code{True} if the breakpoint is enabled, and
27286@code{False} otherwise. This attribute is writable.
27287@end defvar
27288
27289@defvar Breakpoint.silent
27290This attribute is @code{True} if the breakpoint is silent, and
27291@code{False} otherwise. This attribute is writable.
27292
27293Note that a breakpoint can also be silent if it has commands and the
27294first command is @code{silent}. This is not reported by the
27295@code{silent} attribute.
27296@end defvar
27297
27298@defvar Breakpoint.thread
27299If the breakpoint is thread-specific, this attribute holds the thread
27300id. If the breakpoint is not thread-specific, this attribute is
27301@code{None}. This attribute is writable.
27302@end defvar
27303
27304@defvar Breakpoint.task
27305If the breakpoint is Ada task-specific, this attribute holds the Ada task
27306id. If the breakpoint is not task-specific (or the underlying
27307language is not Ada), this attribute is @code{None}. This attribute
27308is writable.
27309@end defvar
27310
27311@defvar Breakpoint.ignore_count
27312This attribute holds the ignore count for the breakpoint, an integer.
27313This attribute is writable.
27314@end defvar
27315
27316@defvar Breakpoint.number
27317This attribute holds the breakpoint's number --- the identifier used by
27318the user to manipulate the breakpoint. This attribute is not writable.
27319@end defvar
27320
27321@defvar Breakpoint.type
27322This attribute holds the breakpoint's type --- the identifier used to
27323determine the actual breakpoint type or use-case. This attribute is not
27324writable.
27325@end defvar
27326
27327@defvar Breakpoint.visible
27328This attribute tells whether the breakpoint is visible to the user
27329when set, or when the @samp{info breakpoints} command is run. This
27330attribute is not writable.
27331@end defvar
27332
27333@defvar Breakpoint.temporary
27334This attribute indicates whether the breakpoint was created as a
27335temporary breakpoint. Temporary breakpoints are automatically deleted
27336after that breakpoint has been hit. Access to this attribute, and all
27337other attributes and functions other than the @code{is_valid}
27338function, will result in an error after the breakpoint has been hit
27339(as it has been automatically deleted). This attribute is not
27340writable.
27341@end defvar
27342
27343The available types are represented by constants defined in the @code{gdb}
27344module:
27345
27346@table @code
27347@findex BP_BREAKPOINT
27348@findex gdb.BP_BREAKPOINT
27349@item gdb.BP_BREAKPOINT
27350Normal code breakpoint.
27351
27352@findex BP_WATCHPOINT
27353@findex gdb.BP_WATCHPOINT
27354@item gdb.BP_WATCHPOINT
27355Watchpoint breakpoint.
27356
27357@findex BP_HARDWARE_WATCHPOINT
27358@findex gdb.BP_HARDWARE_WATCHPOINT
27359@item gdb.BP_HARDWARE_WATCHPOINT
27360Hardware assisted watchpoint.
27361
27362@findex BP_READ_WATCHPOINT
27363@findex gdb.BP_READ_WATCHPOINT
27364@item gdb.BP_READ_WATCHPOINT
27365Hardware assisted read watchpoint.
27366
27367@findex BP_ACCESS_WATCHPOINT
27368@findex gdb.BP_ACCESS_WATCHPOINT
27369@item gdb.BP_ACCESS_WATCHPOINT
27370Hardware assisted access watchpoint.
27371@end table
27372
27373@defvar Breakpoint.hit_count
27374This attribute holds the hit count for the breakpoint, an integer.
27375This attribute is writable, but currently it can only be set to zero.
27376@end defvar
27377
27378@defvar Breakpoint.location
27379This attribute holds the location of the breakpoint, as specified by
27380the user. It is a string. If the breakpoint does not have a location
27381(that is, it is a watchpoint) the attribute's value is @code{None}. This
27382attribute is not writable.
27383@end defvar
27384
27385@defvar Breakpoint.expression
27386This attribute holds a breakpoint expression, as specified by
27387the user. It is a string. If the breakpoint does not have an
27388expression (the breakpoint is not a watchpoint) the attribute's value
27389is @code{None}. This attribute is not writable.
27390@end defvar
27391
27392@defvar Breakpoint.condition
27393This attribute holds the condition of the breakpoint, as specified by
27394the user. It is a string. If there is no condition, this attribute's
27395value is @code{None}. This attribute is writable.
27396@end defvar
27397
27398@defvar Breakpoint.commands
27399This attribute holds the commands attached to the breakpoint. If
27400there are commands, this attribute's value is a string holding all the
27401commands, separated by newlines. If there are no commands, this
27402attribute is @code{None}. This attribute is not writable.
27403@end defvar
27404
27405@node Finish Breakpoints in Python
27406@subsubsection Finish Breakpoints
27407
27408@cindex python finish breakpoints
27409@tindex gdb.FinishBreakpoint
27410
27411A finish breakpoint is a temporary breakpoint set at the return address of
27412a frame, based on the @code{finish} command. @code{gdb.FinishBreakpoint}
27413extends @code{gdb.Breakpoint}. The underlying breakpoint will be disabled
27414and deleted when the execution will run out of the breakpoint scope (i.e.@:
27415@code{Breakpoint.stop} or @code{FinishBreakpoint.out_of_scope} triggered).
27416Finish breakpoints are thread specific and must be create with the right
27417thread selected.
27418
27419@defun FinishBreakpoint.__init__ (@r{[}frame@r{]} @r{[}, internal@r{]})
27420Create a finish breakpoint at the return address of the @code{gdb.Frame}
27421object @var{frame}. If @var{frame} is not provided, this defaults to the
27422newest frame. The optional @var{internal} argument allows the breakpoint to
27423become invisible to the user. @xref{Breakpoints In Python}, for further
27424details about this argument.
27425@end defun
27426
27427@defun FinishBreakpoint.out_of_scope (self)
27428In some circumstances (e.g.@: @code{longjmp}, C@t{++} exceptions, @value{GDBN}
27429@code{return} command, @dots{}), a function may not properly terminate, and
27430thus never hit the finish breakpoint. When @value{GDBN} notices such a
27431situation, the @code{out_of_scope} callback will be triggered.
27432
27433You may want to sub-class @code{gdb.FinishBreakpoint} and override this
27434method:
27435
27436@smallexample
27437class MyFinishBreakpoint (gdb.FinishBreakpoint)
27438 def stop (self):
27439 print "normal finish"
27440 return True
27441
27442 def out_of_scope ():
27443 print "abnormal finish"
27444@end smallexample
27445@end defun
27446
27447@defvar FinishBreakpoint.return_value
27448When @value{GDBN} is stopped at a finish breakpoint and the frame
27449used to build the @code{gdb.FinishBreakpoint} object had debug symbols, this
27450attribute will contain a @code{gdb.Value} object corresponding to the return
27451value of the function. The value will be @code{None} if the function return
27452type is @code{void} or if the return value was not computable. This attribute
27453is not writable.
27454@end defvar
27455
27456@node Lazy Strings In Python
27457@subsubsection Python representation of lazy strings.
27458
27459@cindex lazy strings in python
27460@tindex gdb.LazyString
27461
27462A @dfn{lazy string} is a string whose contents is not retrieved or
27463encoded until it is needed.
27464
27465A @code{gdb.LazyString} is represented in @value{GDBN} as an
27466@code{address} that points to a region of memory, an @code{encoding}
27467that will be used to encode that region of memory, and a @code{length}
27468to delimit the region of memory that represents the string. The
27469difference between a @code{gdb.LazyString} and a string wrapped within
27470a @code{gdb.Value} is that a @code{gdb.LazyString} will be treated
27471differently by @value{GDBN} when printing. A @code{gdb.LazyString} is
27472retrieved and encoded during printing, while a @code{gdb.Value}
27473wrapping a string is immediately retrieved and encoded on creation.
27474
27475A @code{gdb.LazyString} object has the following functions:
27476
27477@defun LazyString.value ()
27478Convert the @code{gdb.LazyString} to a @code{gdb.Value}. This value
27479will point to the string in memory, but will lose all the delayed
27480retrieval, encoding and handling that @value{GDBN} applies to a
27481@code{gdb.LazyString}.
27482@end defun
27483
27484@defvar LazyString.address
27485This attribute holds the address of the string. This attribute is not
27486writable.
27487@end defvar
27488
27489@defvar LazyString.length
27490This attribute holds the length of the string in characters. If the
27491length is -1, then the string will be fetched and encoded up to the
27492first null of appropriate width. This attribute is not writable.
27493@end defvar
27494
27495@defvar LazyString.encoding
27496This attribute holds the encoding that will be applied to the string
27497when the string is printed by @value{GDBN}. If the encoding is not
27498set, or contains an empty string, then @value{GDBN} will select the
27499most appropriate encoding when the string is printed. This attribute
27500is not writable.
27501@end defvar
27502
27503@defvar LazyString.type
27504This attribute holds the type that is represented by the lazy string's
27505type. For a lazy string this will always be a pointer type. To
27506resolve this to the lazy string's character type, use the type's
27507@code{target} method. @xref{Types In Python}. This attribute is not
27508writable.
27509@end defvar
27510
27511@node Architectures In Python
27512@subsubsection Python representation of architectures
27513@cindex Python architectures
27514
27515@value{GDBN} uses architecture specific parameters and artifacts in a
27516number of its various computations. An architecture is represented
27517by an instance of the @code{gdb.Architecture} class.
27518
27519A @code{gdb.Architecture} class has the following methods:
27520
27521@defun Architecture.name ()
27522Return the name (string value) of the architecture.
27523@end defun
27524
27525@defun Architecture.disassemble (@var{start_pc} @r{[}, @var{end_pc} @r{[}, @var{count}@r{]]})
27526Return a list of disassembled instructions starting from the memory
27527address @var{start_pc}. The optional arguments @var{end_pc} and
27528@var{count} determine the number of instructions in the returned list.
27529If both the optional arguments @var{end_pc} and @var{count} are
27530specified, then a list of at most @var{count} disassembled instructions
27531whose start address falls in the closed memory address interval from
27532@var{start_pc} to @var{end_pc} are returned. If @var{end_pc} is not
27533specified, but @var{count} is specified, then @var{count} number of
27534instructions starting from the address @var{start_pc} are returned. If
27535@var{count} is not specified but @var{end_pc} is specified, then all
27536instructions whose start address falls in the closed memory address
27537interval from @var{start_pc} to @var{end_pc} are returned. If neither
27538@var{end_pc} nor @var{count} are specified, then a single instruction at
27539@var{start_pc} is returned. For all of these cases, each element of the
27540returned list is a Python @code{dict} with the following string keys:
27541
27542@table @code
27543
27544@item addr
27545The value corresponding to this key is a Python long integer capturing
27546the memory address of the instruction.
27547
27548@item asm
27549The value corresponding to this key is a string value which represents
27550the instruction with assembly language mnemonics. The assembly
27551language flavor used is the same as that specified by the current CLI
27552variable @code{disassembly-flavor}. @xref{Machine Code}.
27553
27554@item length
27555The value corresponding to this key is the length (integer value) of the
27556instruction in bytes.
27557
27558@end table
27559@end defun
27560
27561@node Python Auto-loading
27562@subsection Python Auto-loading
27563@cindex Python auto-loading
27564
27565When a new object file is read (for example, due to the @code{file}
27566command, or because the inferior has loaded a shared library),
27567@value{GDBN} will look for Python support scripts in several ways:
27568@file{@var{objfile}-gdb.py} (@pxref{objfile-gdb.py file})
27569and @code{.debug_gdb_scripts} section
27570(@pxref{dotdebug_gdb_scripts section}).
27571
27572The auto-loading feature is useful for supplying application-specific
27573debugging commands and scripts.
27574
27575Auto-loading can be enabled or disabled,
27576and the list of auto-loaded scripts can be printed.
27577
27578@table @code
27579@anchor{set auto-load python-scripts}
27580@kindex set auto-load python-scripts
27581@item set auto-load python-scripts [on|off]
27582Enable or disable the auto-loading of Python scripts.
27583
27584@anchor{show auto-load python-scripts}
27585@kindex show auto-load python-scripts
27586@item show auto-load python-scripts
27587Show whether auto-loading of Python scripts is enabled or disabled.
27588
27589@anchor{info auto-load python-scripts}
27590@kindex info auto-load python-scripts
27591@cindex print list of auto-loaded Python scripts
27592@item info auto-load python-scripts [@var{regexp}]
27593Print the list of all Python scripts that @value{GDBN} auto-loaded.
27594
27595Also printed is the list of Python scripts that were mentioned in
27596the @code{.debug_gdb_scripts} section and were not found
27597(@pxref{dotdebug_gdb_scripts section}).
27598This is useful because their names are not printed when @value{GDBN}
27599tries to load them and fails. There may be many of them, and printing
27600an error message for each one is problematic.
27601
27602If @var{regexp} is supplied only Python scripts with matching names are printed.
27603
27604Example:
27605
27606@smallexample
27607(gdb) info auto-load python-scripts
27608Loaded Script
27609Yes py-section-script.py
27610 full name: /tmp/py-section-script.py
27611No my-foo-pretty-printers.py
27612@end smallexample
27613@end table
27614
27615When reading an auto-loaded file, @value{GDBN} sets the
27616@dfn{current objfile}. This is available via the @code{gdb.current_objfile}
27617function (@pxref{Objfiles In Python}). This can be useful for
27618registering objfile-specific pretty-printers and frame-filters.
27619
27620@menu
27621* objfile-gdb.py file:: The @file{@var{objfile}-gdb.py} file
27622* dotdebug_gdb_scripts section:: The @code{.debug_gdb_scripts} section
27623* Which flavor to choose?::
27624@end menu
27625
27626@node objfile-gdb.py file
27627@subsubsection The @file{@var{objfile}-gdb.py} file
27628@cindex @file{@var{objfile}-gdb.py}
27629
27630When a new object file is read, @value{GDBN} looks for
27631a file named @file{@var{objfile}-gdb.py} (we call it @var{script-name} below),
27632where @var{objfile} is the object file's real name, formed by ensuring
27633that the file name is absolute, following all symlinks, and resolving
27634@code{.} and @code{..} components. If this file exists and is
27635readable, @value{GDBN} will evaluate it as a Python script.
27636
27637If this file does not exist, then @value{GDBN} will look for
27638@var{script-name} file in all of the directories as specified below.
27639
27640Note that loading of this script file also requires accordingly configured
27641@code{auto-load safe-path} (@pxref{Auto-loading safe path}).
27642
27643For object files using @file{.exe} suffix @value{GDBN} tries to load first the
27644scripts normally according to its @file{.exe} filename. But if no scripts are
27645found @value{GDBN} also tries script filenames matching the object file without
27646its @file{.exe} suffix. This @file{.exe} stripping is case insensitive and it
27647is attempted on any platform. This makes the script filenames compatible
27648between Unix and MS-Windows hosts.
27649
27650@table @code
27651@anchor{set auto-load scripts-directory}
27652@kindex set auto-load scripts-directory
27653@item set auto-load scripts-directory @r{[}@var{directories}@r{]}
27654Control @value{GDBN} auto-loaded scripts location. Multiple directory entries
27655may be delimited by the host platform path separator in use
27656(@samp{:} on Unix, @samp{;} on MS-Windows and MS-DOS).
27657
27658Each entry here needs to be covered also by the security setting
27659@code{set auto-load safe-path} (@pxref{set auto-load safe-path}).
27660
27661@anchor{with-auto-load-dir}
27662This variable defaults to @file{$debugdir:$datadir/auto-load}. The default
27663@code{set auto-load safe-path} value can be also overriden by @value{GDBN}
27664configuration option @option{--with-auto-load-dir}.
27665
27666Any reference to @file{$debugdir} will get replaced by
27667@var{debug-file-directory} value (@pxref{Separate Debug Files}) and any
27668reference to @file{$datadir} will get replaced by @var{data-directory} which is
27669determined at @value{GDBN} startup (@pxref{Data Files}). @file{$debugdir} and
27670@file{$datadir} must be placed as a directory component --- either alone or
27671delimited by @file{/} or @file{\} directory separators, depending on the host
27672platform.
27673
27674The list of directories uses path separator (@samp{:} on GNU and Unix
27675systems, @samp{;} on MS-Windows and MS-DOS) to separate directories, similarly
27676to the @env{PATH} environment variable.
27677
27678@anchor{show auto-load scripts-directory}
27679@kindex show auto-load scripts-directory
27680@item show auto-load scripts-directory
27681Show @value{GDBN} auto-loaded scripts location.
27682@end table
27683
27684@value{GDBN} does not track which files it has already auto-loaded this way.
27685@value{GDBN} will load the associated script every time the corresponding
27686@var{objfile} is opened.
27687So your @file{-gdb.py} file should be careful to avoid errors if it
27688is evaluated more than once.
27689
27690@node dotdebug_gdb_scripts section
27691@subsubsection The @code{.debug_gdb_scripts} section
27692@cindex @code{.debug_gdb_scripts} section
27693
27694For systems using file formats like ELF and COFF,
27695when @value{GDBN} loads a new object file
27696it will look for a special section named @samp{.debug_gdb_scripts}.
27697If this section exists, its contents is a list of names of scripts to load.
27698
27699@value{GDBN} will look for each specified script file first in the
27700current directory and then along the source search path
27701(@pxref{Source Path, ,Specifying Source Directories}),
27702except that @file{$cdir} is not searched, since the compilation
27703directory is not relevant to scripts.
27704
27705Entries can be placed in section @code{.debug_gdb_scripts} with,
27706for example, this GCC macro:
27707
27708@example
27709/* Note: The "MS" section flags are to remove duplicates. */
27710#define DEFINE_GDB_SCRIPT(script_name) \
27711 asm("\
27712.pushsection \".debug_gdb_scripts\", \"MS\",@@progbits,1\n\
27713.byte 1\n\
27714.asciz \"" script_name "\"\n\
27715.popsection \n\
27716");
27717@end example
27718
27719@noindent
27720Then one can reference the macro in a header or source file like this:
27721
27722@example
27723DEFINE_GDB_SCRIPT ("my-app-scripts.py")
27724@end example
27725
27726The script name may include directories if desired.
27727
27728Note that loading of this script file also requires accordingly configured
27729@code{auto-load safe-path} (@pxref{Auto-loading safe path}).
27730
27731If the macro is put in a header, any application or library
27732using this header will get a reference to the specified script.
27733
27734@node Which flavor to choose?
27735@subsubsection Which flavor to choose?
27736
27737Given the multiple ways of auto-loading Python scripts, it might not always
27738be clear which one to choose. This section provides some guidance.
27739
27740Benefits of the @file{-gdb.py} way:
27741
27742@itemize @bullet
27743@item
27744Can be used with file formats that don't support multiple sections.
27745
27746@item
27747Ease of finding scripts for public libraries.
27748
27749Scripts specified in the @code{.debug_gdb_scripts} section are searched for
27750in the source search path.
27751For publicly installed libraries, e.g., @file{libstdc++}, there typically
27752isn't a source directory in which to find the script.
27753
27754@item
27755Doesn't require source code additions.
27756@end itemize
27757
27758Benefits of the @code{.debug_gdb_scripts} way:
27759
27760@itemize @bullet
27761@item
27762Works with static linking.
27763
27764Scripts for libraries done the @file{-gdb.py} way require an objfile to
27765trigger their loading. When an application is statically linked the only
27766objfile available is the executable, and it is cumbersome to attach all the
27767scripts from all the input libraries to the executable's @file{-gdb.py} script.
27768
27769@item
27770Works with classes that are entirely inlined.
27771
27772Some classes can be entirely inlined, and thus there may not be an associated
27773shared library to attach a @file{-gdb.py} script to.
27774
27775@item
27776Scripts needn't be copied out of the source tree.
27777
27778In some circumstances, apps can be built out of large collections of internal
27779libraries, and the build infrastructure necessary to install the
27780@file{-gdb.py} scripts in a place where @value{GDBN} can find them is
27781cumbersome. It may be easier to specify the scripts in the
27782@code{.debug_gdb_scripts} section as relative paths, and add a path to the
27783top of the source tree to the source search path.
27784@end itemize
27785
27786@node Python modules
27787@subsection Python modules
27788@cindex python modules
27789
27790@value{GDBN} comes with several modules to assist writing Python code.
27791
27792@menu
27793* gdb.printing:: Building and registering pretty-printers.
27794* gdb.types:: Utilities for working with types.
27795* gdb.prompt:: Utilities for prompt value substitution.
27796@end menu
27797
27798@node gdb.printing
27799@subsubsection gdb.printing
27800@cindex gdb.printing
27801
27802This module provides a collection of utilities for working with
27803pretty-printers.
27804
27805@table @code
27806@item PrettyPrinter (@var{name}, @var{subprinters}=None)
27807This class specifies the API that makes @samp{info pretty-printer},
27808@samp{enable pretty-printer} and @samp{disable pretty-printer} work.
27809Pretty-printers should generally inherit from this class.
27810
27811@item SubPrettyPrinter (@var{name})
27812For printers that handle multiple types, this class specifies the
27813corresponding API for the subprinters.
27814
27815@item RegexpCollectionPrettyPrinter (@var{name})
27816Utility class for handling multiple printers, all recognized via
27817regular expressions.
27818@xref{Writing a Pretty-Printer}, for an example.
27819
27820@item FlagEnumerationPrinter (@var{name})
27821A pretty-printer which handles printing of @code{enum} values. Unlike
27822@value{GDBN}'s built-in @code{enum} printing, this printer attempts to
27823work properly when there is some overlap between the enumeration
27824constants. @var{name} is the name of the printer and also the name of
27825the @code{enum} type to look up.
27826
27827@item register_pretty_printer (@var{obj}, @var{printer}, @var{replace}=False)
27828Register @var{printer} with the pretty-printer list of @var{obj}.
27829If @var{replace} is @code{True} then any existing copy of the printer
27830is replaced. Otherwise a @code{RuntimeError} exception is raised
27831if a printer with the same name already exists.
27832@end table
27833
27834@node gdb.types
27835@subsubsection gdb.types
27836@cindex gdb.types
27837
27838This module provides a collection of utilities for working with
27839@code{gdb.Type} objects.
27840
27841@table @code
27842@item get_basic_type (@var{type})
27843Return @var{type} with const and volatile qualifiers stripped,
27844and with typedefs and C@t{++} references converted to the underlying type.
27845
27846C@t{++} example:
27847
27848@smallexample
27849typedef const int const_int;
27850const_int foo (3);
27851const_int& foo_ref (foo);
27852int main () @{ return 0; @}
27853@end smallexample
27854
27855Then in gdb:
27856
27857@smallexample
27858(gdb) start
27859(gdb) python import gdb.types
27860(gdb) python foo_ref = gdb.parse_and_eval("foo_ref")
27861(gdb) python print gdb.types.get_basic_type(foo_ref.type)
27862int
27863@end smallexample
27864
27865@item has_field (@var{type}, @var{field})
27866Return @code{True} if @var{type}, assumed to be a type with fields
27867(e.g., a structure or union), has field @var{field}.
27868
27869@item make_enum_dict (@var{enum_type})
27870Return a Python @code{dictionary} type produced from @var{enum_type}.
27871
27872@item deep_items (@var{type})
27873Returns a Python iterator similar to the standard
27874@code{gdb.Type.iteritems} method, except that the iterator returned
27875by @code{deep_items} will recursively traverse anonymous struct or
27876union fields. For example:
27877
27878@smallexample
27879struct A
27880@{
27881 int a;
27882 union @{
27883 int b0;
27884 int b1;
27885 @};
27886@};
27887@end smallexample
27888
27889@noindent
27890Then in @value{GDBN}:
27891@smallexample
27892(@value{GDBP}) python import gdb.types
27893(@value{GDBP}) python struct_a = gdb.lookup_type("struct A")
27894(@value{GDBP}) python print struct_a.keys ()
27895@{['a', '']@}
27896(@value{GDBP}) python print [k for k,v in gdb.types.deep_items(struct_a)]
27897@{['a', 'b0', 'b1']@}
27898@end smallexample
27899
27900@item get_type_recognizers ()
27901Return a list of the enabled type recognizers for the current context.
27902This is called by @value{GDBN} during the type-printing process
27903(@pxref{Type Printing API}).
27904
27905@item apply_type_recognizers (recognizers, type_obj)
27906Apply the type recognizers, @var{recognizers}, to the type object
27907@var{type_obj}. If any recognizer returns a string, return that
27908string. Otherwise, return @code{None}. This is called by
27909@value{GDBN} during the type-printing process (@pxref{Type Printing
27910API}).
27911
27912@item register_type_printer (locus, printer)
27913This is a convenience function to register a type printer.
27914@var{printer} is the type printer to register. It must implement the
27915type printer protocol. @var{locus} is either a @code{gdb.Objfile}, in
27916which case the printer is registered with that objfile; a
27917@code{gdb.Progspace}, in which case the printer is registered with
27918that progspace; or @code{None}, in which case the printer is
27919registered globally.
27920
27921@item TypePrinter
27922This is a base class that implements the type printer protocol. Type
27923printers are encouraged, but not required, to derive from this class.
27924It defines a constructor:
27925
27926@defmethod TypePrinter __init__ (self, name)
27927Initialize the type printer with the given name. The new printer
27928starts in the enabled state.
27929@end defmethod
27930
27931@end table
27932
27933@node gdb.prompt
27934@subsubsection gdb.prompt
27935@cindex gdb.prompt
27936
27937This module provides a method for prompt value-substitution.
27938
27939@table @code
27940@item substitute_prompt (@var{string})
27941Return @var{string} with escape sequences substituted by values. Some
27942escape sequences take arguments. You can specify arguments inside
27943``@{@}'' immediately following the escape sequence.
27944
27945The escape sequences you can pass to this function are:
27946
27947@table @code
27948@item \\
27949Substitute a backslash.
27950@item \e
27951Substitute an ESC character.
27952@item \f
27953Substitute the selected frame; an argument names a frame parameter.
27954@item \n
27955Substitute a newline.
27956@item \p
27957Substitute a parameter's value; the argument names the parameter.
27958@item \r
27959Substitute a carriage return.
27960@item \t
27961Substitute the selected thread; an argument names a thread parameter.
27962@item \v
27963Substitute the version of GDB.
27964@item \w
27965Substitute the current working directory.
27966@item \[
27967Begin a sequence of non-printing characters. These sequences are
27968typically used with the ESC character, and are not counted in the string
27969length. Example: ``\[\e[0;34m\](gdb)\[\e[0m\]'' will return a
27970blue-colored ``(gdb)'' prompt where the length is five.
27971@item \]
27972End a sequence of non-printing characters.
27973@end table
27974
27975For example:
27976
27977@smallexample
27978substitute_prompt (``frame: \f,
27979 print arguments: \p@{print frame-arguments@}'')
27980@end smallexample
27981
27982@exdent will return the string:
27983
27984@smallexample
27985"frame: main, print arguments: scalars"
27986@end smallexample
27987@end table
27988
27989@node Aliases
27990@section Creating new spellings of existing commands
27991@cindex aliases for commands
27992
27993It is often useful to define alternate spellings of existing commands.
27994For example, if a new @value{GDBN} command defined in Python has
27995a long name to type, it is handy to have an abbreviated version of it
27996that involves less typing.
27997
27998@value{GDBN} itself uses aliases. For example @samp{s} is an alias
27999of the @samp{step} command even though it is otherwise an ambiguous
28000abbreviation of other commands like @samp{set} and @samp{show}.
28001
28002Aliases are also used to provide shortened or more common versions
28003of multi-word commands. For example, @value{GDBN} provides the
28004@samp{tty} alias of the @samp{set inferior-tty} command.
28005
28006You can define a new alias with the @samp{alias} command.
28007
28008@table @code
28009
28010@kindex alias
28011@item alias [-a] [--] @var{ALIAS} = @var{COMMAND}
28012
28013@end table
28014
28015@var{ALIAS} specifies the name of the new alias.
28016Each word of @var{ALIAS} must consist of letters, numbers, dashes and
28017underscores.
28018
28019@var{COMMAND} specifies the name of an existing command
28020that is being aliased.
28021
28022The @samp{-a} option specifies that the new alias is an abbreviation
28023of the command. Abbreviations are not shown in command
28024lists displayed by the @samp{help} command.
28025
28026The @samp{--} option specifies the end of options,
28027and is useful when @var{ALIAS} begins with a dash.
28028
28029Here is a simple example showing how to make an abbreviation
28030of a command so that there is less to type.
28031Suppose you were tired of typing @samp{disas}, the current
28032shortest unambiguous abbreviation of the @samp{disassemble} command
28033and you wanted an even shorter version named @samp{di}.
28034The following will accomplish this.
28035
28036@smallexample
28037(gdb) alias -a di = disas
28038@end smallexample
28039
28040Note that aliases are different from user-defined commands.
28041With a user-defined command, you also need to write documentation
28042for it with the @samp{document} command.
28043An alias automatically picks up the documentation of the existing command.
28044
28045Here is an example where we make @samp{elms} an abbreviation of
28046@samp{elements} in the @samp{set print elements} command.
28047This is to show that you can make an abbreviation of any part
28048of a command.
28049
28050@smallexample
28051(gdb) alias -a set print elms = set print elements
28052(gdb) alias -a show print elms = show print elements
28053(gdb) set p elms 20
28054(gdb) show p elms
28055Limit on string chars or array elements to print is 200.
28056@end smallexample
28057
28058Note that if you are defining an alias of a @samp{set} command,
28059and you want to have an alias for the corresponding @samp{show}
28060command, then you need to define the latter separately.
28061
28062Unambiguously abbreviated commands are allowed in @var{COMMAND} and
28063@var{ALIAS}, just as they are normally.
28064
28065@smallexample
28066(gdb) alias -a set pr elms = set p ele
28067@end smallexample
28068
28069Finally, here is an example showing the creation of a one word
28070alias for a more complex command.
28071This creates alias @samp{spe} of the command @samp{set print elements}.
28072
28073@smallexample
28074(gdb) alias spe = set print elements
28075(gdb) spe 20
28076@end smallexample
28077
28078@node Interpreters
28079@chapter Command Interpreters
28080@cindex command interpreters
28081
28082@value{GDBN} supports multiple command interpreters, and some command
28083infrastructure to allow users or user interface writers to switch
28084between interpreters or run commands in other interpreters.
28085
28086@value{GDBN} currently supports two command interpreters, the console
28087interpreter (sometimes called the command-line interpreter or @sc{cli})
28088and the machine interface interpreter (or @sc{gdb/mi}). This manual
28089describes both of these interfaces in great detail.
28090
28091By default, @value{GDBN} will start with the console interpreter.
28092However, the user may choose to start @value{GDBN} with another
28093interpreter by specifying the @option{-i} or @option{--interpreter}
28094startup options. Defined interpreters include:
28095
28096@table @code
28097@item console
28098@cindex console interpreter
28099The traditional console or command-line interpreter. This is the most often
28100used interpreter with @value{GDBN}. With no interpreter specified at runtime,
28101@value{GDBN} will use this interpreter.
28102
28103@item mi
28104@cindex mi interpreter
28105The newest @sc{gdb/mi} interface (currently @code{mi2}). Used primarily
28106by programs wishing to use @value{GDBN} as a backend for a debugger GUI
28107or an IDE. For more information, see @ref{GDB/MI, ,The @sc{gdb/mi}
28108Interface}.
28109
28110@item mi2
28111@cindex mi2 interpreter
28112The current @sc{gdb/mi} interface.
28113
28114@item mi1
28115@cindex mi1 interpreter
28116The @sc{gdb/mi} interface included in @value{GDBN} 5.1, 5.2, and 5.3.
28117
28118@end table
28119
28120@cindex invoke another interpreter
28121The interpreter being used by @value{GDBN} may not be dynamically
28122switched at runtime. Although possible, this could lead to a very
28123precarious situation. Consider an IDE using @sc{gdb/mi}. If a user
28124enters the command "interpreter-set console" in a console view,
28125@value{GDBN} would switch to using the console interpreter, rendering
28126the IDE inoperable!
28127
28128@kindex interpreter-exec
28129Although you may only choose a single interpreter at startup, you may execute
28130commands in any interpreter from the current interpreter using the appropriate
28131command. If you are running the console interpreter, simply use the
28132@code{interpreter-exec} command:
28133
28134@smallexample
28135interpreter-exec mi "-data-list-register-names"
28136@end smallexample
28137
28138@sc{gdb/mi} has a similar command, although it is only available in versions of
28139@value{GDBN} which support @sc{gdb/mi} version 2 (or greater).
28140
28141@node TUI
28142@chapter @value{GDBN} Text User Interface
28143@cindex TUI
28144@cindex Text User Interface
28145
28146@menu
28147* TUI Overview:: TUI overview
28148* TUI Keys:: TUI key bindings
28149* TUI Single Key Mode:: TUI single key mode
28150* TUI Commands:: TUI-specific commands
28151* TUI Configuration:: TUI configuration variables
28152@end menu
28153
28154The @value{GDBN} Text User Interface (TUI) is a terminal
28155interface which uses the @code{curses} library to show the source
28156file, the assembly output, the program registers and @value{GDBN}
28157commands in separate text windows. The TUI mode is supported only
28158on platforms where a suitable version of the @code{curses} library
28159is available.
28160
28161The TUI mode is enabled by default when you invoke @value{GDBN} as
28162@samp{@value{GDBP} -tui}.
28163You can also switch in and out of TUI mode while @value{GDBN} runs by
28164using various TUI commands and key bindings, such as @kbd{C-x C-a}.
28165@xref{TUI Keys, ,TUI Key Bindings}.
28166
28167@node TUI Overview
28168@section TUI Overview
28169
28170In TUI mode, @value{GDBN} can display several text windows:
28171
28172@table @emph
28173@item command
28174This window is the @value{GDBN} command window with the @value{GDBN}
28175prompt and the @value{GDBN} output. The @value{GDBN} input is still
28176managed using readline.
28177
28178@item source
28179The source window shows the source file of the program. The current
28180line and active breakpoints are displayed in this window.
28181
28182@item assembly
28183The assembly window shows the disassembly output of the program.
28184
28185@item register
28186This window shows the processor registers. Registers are highlighted
28187when their values change.
28188@end table
28189
28190The source and assembly windows show the current program position
28191by highlighting the current line and marking it with a @samp{>} marker.
28192Breakpoints are indicated with two markers. The first marker
28193indicates the breakpoint type:
28194
28195@table @code
28196@item B
28197Breakpoint which was hit at least once.
28198
28199@item b
28200Breakpoint which was never hit.
28201
28202@item H
28203Hardware breakpoint which was hit at least once.
28204
28205@item h
28206Hardware breakpoint which was never hit.
28207@end table
28208
28209The second marker indicates whether the breakpoint is enabled or not:
28210
28211@table @code
28212@item +
28213Breakpoint is enabled.
28214
28215@item -
28216Breakpoint is disabled.
28217@end table
28218
28219The source, assembly and register windows are updated when the current
28220thread changes, when the frame changes, or when the program counter
28221changes.
28222
28223These windows are not all visible at the same time. The command
28224window is always visible. The others can be arranged in several
28225layouts:
28226
28227@itemize @bullet
28228@item
28229source only,
28230
28231@item
28232assembly only,
28233
28234@item
28235source and assembly,
28236
28237@item
28238source and registers, or
28239
28240@item
28241assembly and registers.
28242@end itemize
28243
28244A status line above the command window shows the following information:
28245
28246@table @emph
28247@item target
28248Indicates the current @value{GDBN} target.
28249(@pxref{Targets, ,Specifying a Debugging Target}).
28250
28251@item process
28252Gives the current process or thread number.
28253When no process is being debugged, this field is set to @code{No process}.
28254
28255@item function
28256Gives the current function name for the selected frame.
28257The name is demangled if demangling is turned on (@pxref{Print Settings}).
28258When there is no symbol corresponding to the current program counter,
28259the string @code{??} is displayed.
28260
28261@item line
28262Indicates the current line number for the selected frame.
28263When the current line number is not known, the string @code{??} is displayed.
28264
28265@item pc
28266Indicates the current program counter address.
28267@end table
28268
28269@node TUI Keys
28270@section TUI Key Bindings
28271@cindex TUI key bindings
28272
28273The TUI installs several key bindings in the readline keymaps
28274@ifset SYSTEM_READLINE
28275(@pxref{Command Line Editing, , , rluserman, GNU Readline Library}).
28276@end ifset
28277@ifclear SYSTEM_READLINE
28278(@pxref{Command Line Editing}).
28279@end ifclear
28280The following key bindings are installed for both TUI mode and the
28281@value{GDBN} standard mode.
28282
28283@table @kbd
28284@kindex C-x C-a
28285@item C-x C-a
28286@kindex C-x a
28287@itemx C-x a
28288@kindex C-x A
28289@itemx C-x A
28290Enter or leave the TUI mode. When leaving the TUI mode,
28291the curses window management stops and @value{GDBN} operates using
28292its standard mode, writing on the terminal directly. When reentering
28293the TUI mode, control is given back to the curses windows.
28294The screen is then refreshed.
28295
28296@kindex C-x 1
28297@item C-x 1
28298Use a TUI layout with only one window. The layout will
28299either be @samp{source} or @samp{assembly}. When the TUI mode
28300is not active, it will switch to the TUI mode.
28301
28302Think of this key binding as the Emacs @kbd{C-x 1} binding.
28303
28304@kindex C-x 2
28305@item C-x 2
28306Use a TUI layout with at least two windows. When the current
28307layout already has two windows, the next layout with two windows is used.
28308When a new layout is chosen, one window will always be common to the
28309previous layout and the new one.
28310
28311Think of it as the Emacs @kbd{C-x 2} binding.
28312
28313@kindex C-x o
28314@item C-x o
28315Change the active window. The TUI associates several key bindings
28316(like scrolling and arrow keys) with the active window. This command
28317gives the focus to the next TUI window.
28318
28319Think of it as the Emacs @kbd{C-x o} binding.
28320
28321@kindex C-x s
28322@item C-x s
28323Switch in and out of the TUI SingleKey mode that binds single
28324keys to @value{GDBN} commands (@pxref{TUI Single Key Mode}).
28325@end table
28326
28327The following key bindings only work in the TUI mode:
28328
28329@table @asis
28330@kindex PgUp
28331@item @key{PgUp}
28332Scroll the active window one page up.
28333
28334@kindex PgDn
28335@item @key{PgDn}
28336Scroll the active window one page down.
28337
28338@kindex Up
28339@item @key{Up}
28340Scroll the active window one line up.
28341
28342@kindex Down
28343@item @key{Down}
28344Scroll the active window one line down.
28345
28346@kindex Left
28347@item @key{Left}
28348Scroll the active window one column left.
28349
28350@kindex Right
28351@item @key{Right}
28352Scroll the active window one column right.
28353
28354@kindex C-L
28355@item @kbd{C-L}
28356Refresh the screen.
28357@end table
28358
28359Because the arrow keys scroll the active window in the TUI mode, they
28360are not available for their normal use by readline unless the command
28361window has the focus. When another window is active, you must use
28362other readline key bindings such as @kbd{C-p}, @kbd{C-n}, @kbd{C-b}
28363and @kbd{C-f} to control the command window.
28364
28365@node TUI Single Key Mode
28366@section TUI Single Key Mode
28367@cindex TUI single key mode
28368
28369The TUI also provides a @dfn{SingleKey} mode, which binds several
28370frequently used @value{GDBN} commands to single keys. Type @kbd{C-x s} to
28371switch into this mode, where the following key bindings are used:
28372
28373@table @kbd
28374@kindex c @r{(SingleKey TUI key)}
28375@item c
28376continue
28377
28378@kindex d @r{(SingleKey TUI key)}
28379@item d
28380down
28381
28382@kindex f @r{(SingleKey TUI key)}
28383@item f
28384finish
28385
28386@kindex n @r{(SingleKey TUI key)}
28387@item n
28388next
28389
28390@kindex q @r{(SingleKey TUI key)}
28391@item q
28392exit the SingleKey mode.
28393
28394@kindex r @r{(SingleKey TUI key)}
28395@item r
28396run
28397
28398@kindex s @r{(SingleKey TUI key)}
28399@item s
28400step
28401
28402@kindex u @r{(SingleKey TUI key)}
28403@item u
28404up
28405
28406@kindex v @r{(SingleKey TUI key)}
28407@item v
28408info locals
28409
28410@kindex w @r{(SingleKey TUI key)}
28411@item w
28412where
28413@end table
28414
28415Other keys temporarily switch to the @value{GDBN} command prompt.
28416The key that was pressed is inserted in the editing buffer so that
28417it is possible to type most @value{GDBN} commands without interaction
28418with the TUI SingleKey mode. Once the command is entered the TUI
28419SingleKey mode is restored. The only way to permanently leave
28420this mode is by typing @kbd{q} or @kbd{C-x s}.
28421
28422
28423@node TUI Commands
28424@section TUI-specific Commands
28425@cindex TUI commands
28426
28427The TUI has specific commands to control the text windows.
28428These commands are always available, even when @value{GDBN} is not in
28429the TUI mode. When @value{GDBN} is in the standard mode, most
28430of these commands will automatically switch to the TUI mode.
28431
28432Note that if @value{GDBN}'s @code{stdout} is not connected to a
28433terminal, or @value{GDBN} has been started with the machine interface
28434interpreter (@pxref{GDB/MI, ,The @sc{gdb/mi} Interface}), most of
28435these commands will fail with an error, because it would not be
28436possible or desirable to enable curses window management.
28437
28438@table @code
28439@item info win
28440@kindex info win
28441List and give the size of all displayed windows.
28442
28443@item layout next
28444@kindex layout
28445Display the next layout.
28446
28447@item layout prev
28448Display the previous layout.
28449
28450@item layout src
28451Display the source window only.
28452
28453@item layout asm
28454Display the assembly window only.
28455
28456@item layout split
28457Display the source and assembly window.
28458
28459@item layout regs
28460Display the register window together with the source or assembly window.
28461
28462@item focus next
28463@kindex focus
28464Make the next window active for scrolling.
28465
28466@item focus prev
28467Make the previous window active for scrolling.
28468
28469@item focus src
28470Make the source window active for scrolling.
28471
28472@item focus asm
28473Make the assembly window active for scrolling.
28474
28475@item focus regs
28476Make the register window active for scrolling.
28477
28478@item focus cmd
28479Make the command window active for scrolling.
28480
28481@item refresh
28482@kindex refresh
28483Refresh the screen. This is similar to typing @kbd{C-L}.
28484
28485@item tui reg float
28486@kindex tui reg
28487Show the floating point registers in the register window.
28488
28489@item tui reg general
28490Show the general registers in the register window.
28491
28492@item tui reg next
28493Show the next register group. The list of register groups as well as
28494their order is target specific. The predefined register groups are the
28495following: @code{general}, @code{float}, @code{system}, @code{vector},
28496@code{all}, @code{save}, @code{restore}.
28497
28498@item tui reg system
28499Show the system registers in the register window.
28500
28501@item update
28502@kindex update
28503Update the source window and the current execution point.
28504
28505@item winheight @var{name} +@var{count}
28506@itemx winheight @var{name} -@var{count}
28507@kindex winheight
28508Change the height of the window @var{name} by @var{count}
28509lines. Positive counts increase the height, while negative counts
28510decrease it.
28511
28512@item tabset @var{nchars}
28513@kindex tabset
28514Set the width of tab stops to be @var{nchars} characters.
28515@end table
28516
28517@node TUI Configuration
28518@section TUI Configuration Variables
28519@cindex TUI configuration variables
28520
28521Several configuration variables control the appearance of TUI windows.
28522
28523@table @code
28524@item set tui border-kind @var{kind}
28525@kindex set tui border-kind
28526Select the border appearance for the source, assembly and register windows.
28527The possible values are the following:
28528@table @code
28529@item space
28530Use a space character to draw the border.
28531
28532@item ascii
28533Use @sc{ascii} characters @samp{+}, @samp{-} and @samp{|} to draw the border.
28534
28535@item acs
28536Use the Alternate Character Set to draw the border. The border is
28537drawn using character line graphics if the terminal supports them.
28538@end table
28539
28540@item set tui border-mode @var{mode}
28541@kindex set tui border-mode
28542@itemx set tui active-border-mode @var{mode}
28543@kindex set tui active-border-mode
28544Select the display attributes for the borders of the inactive windows
28545or the active window. The @var{mode} can be one of the following:
28546@table @code
28547@item normal
28548Use normal attributes to display the border.
28549
28550@item standout
28551Use standout mode.
28552
28553@item reverse
28554Use reverse video mode.
28555
28556@item half
28557Use half bright mode.
28558
28559@item half-standout
28560Use half bright and standout mode.
28561
28562@item bold
28563Use extra bright or bold mode.
28564
28565@item bold-standout
28566Use extra bright or bold and standout mode.
28567@end table
28568@end table
28569
28570@node Emacs
28571@chapter Using @value{GDBN} under @sc{gnu} Emacs
28572
28573@cindex Emacs
28574@cindex @sc{gnu} Emacs
28575A special interface allows you to use @sc{gnu} Emacs to view (and
28576edit) the source files for the program you are debugging with
28577@value{GDBN}.
28578
28579To use this interface, use the command @kbd{M-x gdb} in Emacs. Give the
28580executable file you want to debug as an argument. This command starts
28581@value{GDBN} as a subprocess of Emacs, with input and output through a newly
28582created Emacs buffer.
28583@c (Do not use the @code{-tui} option to run @value{GDBN} from Emacs.)
28584
28585Running @value{GDBN} under Emacs can be just like running @value{GDBN} normally except for two
28586things:
28587
28588@itemize @bullet
28589@item
28590All ``terminal'' input and output goes through an Emacs buffer, called
28591the GUD buffer.
28592
28593This applies both to @value{GDBN} commands and their output, and to the input
28594and output done by the program you are debugging.
28595
28596This is useful because it means that you can copy the text of previous
28597commands and input them again; you can even use parts of the output
28598in this way.
28599
28600All the facilities of Emacs' Shell mode are available for interacting
28601with your program. In particular, you can send signals the usual
28602way---for example, @kbd{C-c C-c} for an interrupt, @kbd{C-c C-z} for a
28603stop.
28604
28605@item
28606@value{GDBN} displays source code through Emacs.
28607
28608Each time @value{GDBN} displays a stack frame, Emacs automatically finds the
28609source file for that frame and puts an arrow (@samp{=>}) at the
28610left margin of the current line. Emacs uses a separate buffer for
28611source display, and splits the screen to show both your @value{GDBN} session
28612and the source.
28613
28614Explicit @value{GDBN} @code{list} or search commands still produce output as
28615usual, but you probably have no reason to use them from Emacs.
28616@end itemize
28617
28618We call this @dfn{text command mode}. Emacs 22.1, and later, also uses
28619a graphical mode, enabled by default, which provides further buffers
28620that can control the execution and describe the state of your program.
28621@xref{GDB Graphical Interface,,, Emacs, The @sc{gnu} Emacs Manual}.
28622
28623If you specify an absolute file name when prompted for the @kbd{M-x
28624gdb} argument, then Emacs sets your current working directory to where
28625your program resides. If you only specify the file name, then Emacs
28626sets your current working directory to the directory associated
28627with the previous buffer. In this case, @value{GDBN} may find your
28628program by searching your environment's @code{PATH} variable, but on
28629some operating systems it might not find the source. So, although the
28630@value{GDBN} input and output session proceeds normally, the auxiliary
28631buffer does not display the current source and line of execution.
28632
28633The initial working directory of @value{GDBN} is printed on the top
28634line of the GUD buffer and this serves as a default for the commands
28635that specify files for @value{GDBN} to operate on. @xref{Files,
28636,Commands to Specify Files}.
28637
28638By default, @kbd{M-x gdb} calls the program called @file{gdb}. If you
28639need to call @value{GDBN} by a different name (for example, if you
28640keep several configurations around, with different names) you can
28641customize the Emacs variable @code{gud-gdb-command-name} to run the
28642one you want.
28643
28644In the GUD buffer, you can use these special Emacs commands in
28645addition to the standard Shell mode commands:
28646
28647@table @kbd
28648@item C-h m
28649Describe the features of Emacs' GUD Mode.
28650
28651@item C-c C-s
28652Execute to another source line, like the @value{GDBN} @code{step} command; also
28653update the display window to show the current file and location.
28654
28655@item C-c C-n
28656Execute to next source line in this function, skipping all function
28657calls, like the @value{GDBN} @code{next} command. Then update the display window
28658to show the current file and location.
28659
28660@item C-c C-i
28661Execute one instruction, like the @value{GDBN} @code{stepi} command; update
28662display window accordingly.
28663
28664@item C-c C-f
28665Execute until exit from the selected stack frame, like the @value{GDBN}
28666@code{finish} command.
28667
28668@item C-c C-r
28669Continue execution of your program, like the @value{GDBN} @code{continue}
28670command.
28671
28672@item C-c <
28673Go up the number of frames indicated by the numeric argument
28674(@pxref{Arguments, , Numeric Arguments, Emacs, The @sc{gnu} Emacs Manual}),
28675like the @value{GDBN} @code{up} command.
28676
28677@item C-c >
28678Go down the number of frames indicated by the numeric argument, like the
28679@value{GDBN} @code{down} command.
28680@end table
28681
28682In any source file, the Emacs command @kbd{C-x @key{SPC}} (@code{gud-break})
28683tells @value{GDBN} to set a breakpoint on the source line point is on.
28684
28685In text command mode, if you type @kbd{M-x speedbar}, Emacs displays a
28686separate frame which shows a backtrace when the GUD buffer is current.
28687Move point to any frame in the stack and type @key{RET} to make it
28688become the current frame and display the associated source in the
28689source buffer. Alternatively, click @kbd{Mouse-2} to make the
28690selected frame become the current one. In graphical mode, the
28691speedbar displays watch expressions.
28692
28693If you accidentally delete the source-display buffer, an easy way to get
28694it back is to type the command @code{f} in the @value{GDBN} buffer, to
28695request a frame display; when you run under Emacs, this recreates
28696the source buffer if necessary to show you the context of the current
28697frame.
28698
28699The source files displayed in Emacs are in ordinary Emacs buffers
28700which are visiting the source files in the usual way. You can edit
28701the files with these buffers if you wish; but keep in mind that @value{GDBN}
28702communicates with Emacs in terms of line numbers. If you add or
28703delete lines from the text, the line numbers that @value{GDBN} knows cease
28704to correspond properly with the code.
28705
28706A more detailed description of Emacs' interaction with @value{GDBN} is
28707given in the Emacs manual (@pxref{Debuggers,,, Emacs, The @sc{gnu}
28708Emacs Manual}).
28709
28710@node GDB/MI
28711@chapter The @sc{gdb/mi} Interface
28712
28713@unnumberedsec Function and Purpose
28714
28715@cindex @sc{gdb/mi}, its purpose
28716@sc{gdb/mi} is a line based machine oriented text interface to
28717@value{GDBN} and is activated by specifying using the
28718@option{--interpreter} command line option (@pxref{Mode Options}). It
28719is specifically intended to support the development of systems which
28720use the debugger as just one small component of a larger system.
28721
28722This chapter is a specification of the @sc{gdb/mi} interface. It is written
28723in the form of a reference manual.
28724
28725Note that @sc{gdb/mi} is still under construction, so some of the
28726features described below are incomplete and subject to change
28727(@pxref{GDB/MI Development and Front Ends, , @sc{gdb/mi} Development and Front Ends}).
28728
28729@unnumberedsec Notation and Terminology
28730
28731@cindex notational conventions, for @sc{gdb/mi}
28732This chapter uses the following notation:
28733
28734@itemize @bullet
28735@item
28736@code{|} separates two alternatives.
28737
28738@item
28739@code{[ @var{something} ]} indicates that @var{something} is optional:
28740it may or may not be given.
28741
28742@item
28743@code{( @var{group} )*} means that @var{group} inside the parentheses
28744may repeat zero or more times.
28745
28746@item
28747@code{( @var{group} )+} means that @var{group} inside the parentheses
28748may repeat one or more times.
28749
28750@item
28751@code{"@var{string}"} means a literal @var{string}.
28752@end itemize
28753
28754@ignore
28755@heading Dependencies
28756@end ignore
28757
28758@menu
28759* GDB/MI General Design::
28760* GDB/MI Command Syntax::
28761* GDB/MI Compatibility with CLI::
28762* GDB/MI Development and Front Ends::
28763* GDB/MI Output Records::
28764* GDB/MI Simple Examples::
28765* GDB/MI Command Description Format::
28766* GDB/MI Breakpoint Commands::
28767* GDB/MI Catchpoint Commands::
28768* GDB/MI Program Context::
28769* GDB/MI Thread Commands::
28770* GDB/MI Ada Tasking Commands::
28771* GDB/MI Program Execution::
28772* GDB/MI Stack Manipulation::
28773* GDB/MI Variable Objects::
28774* GDB/MI Data Manipulation::
28775* GDB/MI Tracepoint Commands::
28776* GDB/MI Symbol Query::
28777* GDB/MI File Commands::
28778@ignore
28779* GDB/MI Kod Commands::
28780* GDB/MI Memory Overlay Commands::
28781* GDB/MI Signal Handling Commands::
28782@end ignore
28783* GDB/MI Target Manipulation::
28784* GDB/MI File Transfer Commands::
28785* GDB/MI Ada Exceptions Commands::
28786* GDB/MI Miscellaneous Commands::
28787@end menu
28788
28789@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
28790@node GDB/MI General Design
28791@section @sc{gdb/mi} General Design
28792@cindex GDB/MI General Design
28793
28794Interaction of a @sc{GDB/MI} frontend with @value{GDBN} involves three
28795parts---commands sent to @value{GDBN}, responses to those commands
28796and notifications. Each command results in exactly one response,
28797indicating either successful completion of the command, or an error.
28798For the commands that do not resume the target, the response contains the
28799requested information. For the commands that resume the target, the
28800response only indicates whether the target was successfully resumed.
28801Notifications is the mechanism for reporting changes in the state of the
28802target, or in @value{GDBN} state, that cannot conveniently be associated with
28803a command and reported as part of that command response.
28804
28805The important examples of notifications are:
28806@itemize @bullet
28807
28808@item
28809Exec notifications. These are used to report changes in
28810target state---when a target is resumed, or stopped. It would not
28811be feasible to include this information in response of resuming
28812commands, because one resume commands can result in multiple events in
28813different threads. Also, quite some time may pass before any event
28814happens in the target, while a frontend needs to know whether the resuming
28815command itself was successfully executed.
28816
28817@item
28818Console output, and status notifications. Console output
28819notifications are used to report output of CLI commands, as well as
28820diagnostics for other commands. Status notifications are used to
28821report the progress of a long-running operation. Naturally, including
28822this information in command response would mean no output is produced
28823until the command is finished, which is undesirable.
28824
28825@item
28826General notifications. Commands may have various side effects on
28827the @value{GDBN} or target state beyond their official purpose. For example,
28828a command may change the selected thread. Although such changes can
28829be included in command response, using notification allows for more
28830orthogonal frontend design.
28831
28832@end itemize
28833
28834There's no guarantee that whenever an MI command reports an error,
28835@value{GDBN} or the target are in any specific state, and especially,
28836the state is not reverted to the state before the MI command was
28837processed. Therefore, whenever an MI command results in an error,
28838we recommend that the frontend refreshes all the information shown in
28839the user interface.
28840
28841
28842@menu
28843* Context management::
28844* Asynchronous and non-stop modes::
28845* Thread groups::
28846@end menu
28847
28848@node Context management
28849@subsection Context management
28850
28851@subsubsection Threads and Frames
28852
28853In most cases when @value{GDBN} accesses the target, this access is
28854done in context of a specific thread and frame (@pxref{Frames}).
28855Often, even when accessing global data, the target requires that a thread
28856be specified. The CLI interface maintains the selected thread and frame,
28857and supplies them to target on each command. This is convenient,
28858because a command line user would not want to specify that information
28859explicitly on each command, and because user interacts with
28860@value{GDBN} via a single terminal, so no confusion is possible as
28861to what thread and frame are the current ones.
28862
28863In the case of MI, the concept of selected thread and frame is less
28864useful. First, a frontend can easily remember this information
28865itself. Second, a graphical frontend can have more than one window,
28866each one used for debugging a different thread, and the frontend might
28867want to access additional threads for internal purposes. This
28868increases the risk that by relying on implicitly selected thread, the
28869frontend may be operating on a wrong one. Therefore, each MI command
28870should explicitly specify which thread and frame to operate on. To
28871make it possible, each MI command accepts the @samp{--thread} and
28872@samp{--frame} options, the value to each is @value{GDBN} identifier
28873for thread and frame to operate on.
28874
28875Usually, each top-level window in a frontend allows the user to select
28876a thread and a frame, and remembers the user selection for further
28877operations. However, in some cases @value{GDBN} may suggest that the
28878current thread be changed. For example, when stopping on a breakpoint
28879it is reasonable to switch to the thread where breakpoint is hit. For
28880another example, if the user issues the CLI @samp{thread} command via
28881the frontend, it is desirable to change the frontend's selected thread to the
28882one specified by user. @value{GDBN} communicates the suggestion to
28883change current thread using the @samp{=thread-selected} notification.
28884No such notification is available for the selected frame at the moment.
28885
28886Note that historically, MI shares the selected thread with CLI, so
28887frontends used the @code{-thread-select} to execute commands in the
28888right context. However, getting this to work right is cumbersome. The
28889simplest way is for frontend to emit @code{-thread-select} command
28890before every command. This doubles the number of commands that need
28891to be sent. The alternative approach is to suppress @code{-thread-select}
28892if the selected thread in @value{GDBN} is supposed to be identical to the
28893thread the frontend wants to operate on. However, getting this
28894optimization right can be tricky. In particular, if the frontend
28895sends several commands to @value{GDBN}, and one of the commands changes the
28896selected thread, then the behaviour of subsequent commands will
28897change. So, a frontend should either wait for response from such
28898problematic commands, or explicitly add @code{-thread-select} for
28899all subsequent commands. No frontend is known to do this exactly
28900right, so it is suggested to just always pass the @samp{--thread} and
28901@samp{--frame} options.
28902
28903@subsubsection Language
28904
28905The execution of several commands depends on which language is selected.
28906By default, the current language (@pxref{show language}) is used.
28907But for commands known to be language-sensitive, it is recommended
28908to use the @samp{--language} option. This option takes one argument,
28909which is the name of the language to use while executing the command.
28910For instance:
28911
28912@smallexample
28913-data-evaluate-expression --language c "sizeof (void*)"
28914^done,value="4"
28915(gdb)
28916@end smallexample
28917
28918The valid language names are the same names accepted by the
28919@samp{set language} command (@pxref{Manually}), excluding @samp{auto},
28920@samp{local} or @samp{unknown}.
28921
28922@node Asynchronous and non-stop modes
28923@subsection Asynchronous command execution and non-stop mode
28924
28925On some targets, @value{GDBN} is capable of processing MI commands
28926even while the target is running. This is called @dfn{asynchronous
28927command execution} (@pxref{Background Execution}). The frontend may
28928specify a preferrence for asynchronous execution using the
28929@code{-gdb-set target-async 1} command, which should be emitted before
28930either running the executable or attaching to the target. After the
28931frontend has started the executable or attached to the target, it can
28932find if asynchronous execution is enabled using the
28933@code{-list-target-features} command.
28934
28935Even if @value{GDBN} can accept a command while target is running,
28936many commands that access the target do not work when the target is
28937running. Therefore, asynchronous command execution is most useful
28938when combined with non-stop mode (@pxref{Non-Stop Mode}). Then,
28939it is possible to examine the state of one thread, while other threads
28940are running.
28941
28942When a given thread is running, MI commands that try to access the
28943target in the context of that thread may not work, or may work only on
28944some targets. In particular, commands that try to operate on thread's
28945stack will not work, on any target. Commands that read memory, or
28946modify breakpoints, may work or not work, depending on the target. Note
28947that even commands that operate on global state, such as @code{print},
28948@code{set}, and breakpoint commands, still access the target in the
28949context of a specific thread, so frontend should try to find a
28950stopped thread and perform the operation on that thread (using the
28951@samp{--thread} option).
28952
28953Which commands will work in the context of a running thread is
28954highly target dependent. However, the two commands
28955@code{-exec-interrupt}, to stop a thread, and @code{-thread-info},
28956to find the state of a thread, will always work.
28957
28958@node Thread groups
28959@subsection Thread groups
28960@value{GDBN} may be used to debug several processes at the same time.
28961On some platfroms, @value{GDBN} may support debugging of several
28962hardware systems, each one having several cores with several different
28963processes running on each core. This section describes the MI
28964mechanism to support such debugging scenarios.
28965
28966The key observation is that regardless of the structure of the
28967target, MI can have a global list of threads, because most commands that
28968accept the @samp{--thread} option do not need to know what process that
28969thread belongs to. Therefore, it is not necessary to introduce
28970neither additional @samp{--process} option, nor an notion of the
28971current process in the MI interface. The only strictly new feature
28972that is required is the ability to find how the threads are grouped
28973into processes.
28974
28975To allow the user to discover such grouping, and to support arbitrary
28976hierarchy of machines/cores/processes, MI introduces the concept of a
28977@dfn{thread group}. Thread group is a collection of threads and other
28978thread groups. A thread group always has a string identifier, a type,
28979and may have additional attributes specific to the type. A new
28980command, @code{-list-thread-groups}, returns the list of top-level
28981thread groups, which correspond to processes that @value{GDBN} is
28982debugging at the moment. By passing an identifier of a thread group
28983to the @code{-list-thread-groups} command, it is possible to obtain
28984the members of specific thread group.
28985
28986To allow the user to easily discover processes, and other objects, he
28987wishes to debug, a concept of @dfn{available thread group} is
28988introduced. Available thread group is an thread group that
28989@value{GDBN} is not debugging, but that can be attached to, using the
28990@code{-target-attach} command. The list of available top-level thread
28991groups can be obtained using @samp{-list-thread-groups --available}.
28992In general, the content of a thread group may be only retrieved only
28993after attaching to that thread group.
28994
28995Thread groups are related to inferiors (@pxref{Inferiors and
28996Programs}). Each inferior corresponds to a thread group of a special
28997type @samp{process}, and some additional operations are permitted on
28998such thread groups.
28999
29000@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
29001@node GDB/MI Command Syntax
29002@section @sc{gdb/mi} Command Syntax
29003
29004@menu
29005* GDB/MI Input Syntax::
29006* GDB/MI Output Syntax::
29007@end menu
29008
29009@node GDB/MI Input Syntax
29010@subsection @sc{gdb/mi} Input Syntax
29011
29012@cindex input syntax for @sc{gdb/mi}
29013@cindex @sc{gdb/mi}, input syntax
29014@table @code
29015@item @var{command} @expansion{}
29016@code{@var{cli-command} | @var{mi-command}}
29017
29018@item @var{cli-command} @expansion{}
29019@code{[ @var{token} ] @var{cli-command} @var{nl}}, where
29020@var{cli-command} is any existing @value{GDBN} CLI command.
29021
29022@item @var{mi-command} @expansion{}
29023@code{[ @var{token} ] "-" @var{operation} ( " " @var{option} )*
29024@code{[} " --" @code{]} ( " " @var{parameter} )* @var{nl}}
29025
29026@item @var{token} @expansion{}
29027"any sequence of digits"
29028
29029@item @var{option} @expansion{}
29030@code{"-" @var{parameter} [ " " @var{parameter} ]}
29031
29032@item @var{parameter} @expansion{}
29033@code{@var{non-blank-sequence} | @var{c-string}}
29034
29035@item @var{operation} @expansion{}
29036@emph{any of the operations described in this chapter}
29037
29038@item @var{non-blank-sequence} @expansion{}
29039@emph{anything, provided it doesn't contain special characters such as
29040"-", @var{nl}, """ and of course " "}
29041
29042@item @var{c-string} @expansion{}
29043@code{""" @var{seven-bit-iso-c-string-content} """}
29044
29045@item @var{nl} @expansion{}
29046@code{CR | CR-LF}
29047@end table
29048
29049@noindent
29050Notes:
29051
29052@itemize @bullet
29053@item
29054The CLI commands are still handled by the @sc{mi} interpreter; their
29055output is described below.
29056
29057@item
29058The @code{@var{token}}, when present, is passed back when the command
29059finishes.
29060
29061@item
29062Some @sc{mi} commands accept optional arguments as part of the parameter
29063list. Each option is identified by a leading @samp{-} (dash) and may be
29064followed by an optional argument parameter. Options occur first in the
29065parameter list and can be delimited from normal parameters using
29066@samp{--} (this is useful when some parameters begin with a dash).
29067@end itemize
29068
29069Pragmatics:
29070
29071@itemize @bullet
29072@item
29073We want easy access to the existing CLI syntax (for debugging).
29074
29075@item
29076We want it to be easy to spot a @sc{mi} operation.
29077@end itemize
29078
29079@node GDB/MI Output Syntax
29080@subsection @sc{gdb/mi} Output Syntax
29081
29082@cindex output syntax of @sc{gdb/mi}
29083@cindex @sc{gdb/mi}, output syntax
29084The output from @sc{gdb/mi} consists of zero or more out-of-band records
29085followed, optionally, by a single result record. This result record
29086is for the most recent command. The sequence of output records is
29087terminated by @samp{(gdb)}.
29088
29089If an input command was prefixed with a @code{@var{token}} then the
29090corresponding output for that command will also be prefixed by that same
29091@var{token}.
29092
29093@table @code
29094@item @var{output} @expansion{}
29095@code{( @var{out-of-band-record} )* [ @var{result-record} ] "(gdb)" @var{nl}}
29096
29097@item @var{result-record} @expansion{}
29098@code{ [ @var{token} ] "^" @var{result-class} ( "," @var{result} )* @var{nl}}
29099
29100@item @var{out-of-band-record} @expansion{}
29101@code{@var{async-record} | @var{stream-record}}
29102
29103@item @var{async-record} @expansion{}
29104@code{@var{exec-async-output} | @var{status-async-output} | @var{notify-async-output}}
29105
29106@item @var{exec-async-output} @expansion{}
29107@code{[ @var{token} ] "*" @var{async-output}}
29108
29109@item @var{status-async-output} @expansion{}
29110@code{[ @var{token} ] "+" @var{async-output}}
29111
29112@item @var{notify-async-output} @expansion{}
29113@code{[ @var{token} ] "=" @var{async-output}}
29114
29115@item @var{async-output} @expansion{}
29116@code{@var{async-class} ( "," @var{result} )* @var{nl}}
29117
29118@item @var{result-class} @expansion{}
29119@code{"done" | "running" | "connected" | "error" | "exit"}
29120
29121@item @var{async-class} @expansion{}
29122@code{"stopped" | @var{others}} (where @var{others} will be added
29123depending on the needs---this is still in development).
29124
29125@item @var{result} @expansion{}
29126@code{ @var{variable} "=" @var{value}}
29127
29128@item @var{variable} @expansion{}
29129@code{ @var{string} }
29130
29131@item @var{value} @expansion{}
29132@code{ @var{const} | @var{tuple} | @var{list} }
29133
29134@item @var{const} @expansion{}
29135@code{@var{c-string}}
29136
29137@item @var{tuple} @expansion{}
29138@code{ "@{@}" | "@{" @var{result} ( "," @var{result} )* "@}" }
29139
29140@item @var{list} @expansion{}
29141@code{ "[]" | "[" @var{value} ( "," @var{value} )* "]" | "["
29142@var{result} ( "," @var{result} )* "]" }
29143
29144@item @var{stream-record} @expansion{}
29145@code{@var{console-stream-output} | @var{target-stream-output} | @var{log-stream-output}}
29146
29147@item @var{console-stream-output} @expansion{}
29148@code{"~" @var{c-string}}
29149
29150@item @var{target-stream-output} @expansion{}
29151@code{"@@" @var{c-string}}
29152
29153@item @var{log-stream-output} @expansion{}
29154@code{"&" @var{c-string}}
29155
29156@item @var{nl} @expansion{}
29157@code{CR | CR-LF}
29158
29159@item @var{token} @expansion{}
29160@emph{any sequence of digits}.
29161@end table
29162
29163@noindent
29164Notes:
29165
29166@itemize @bullet
29167@item
29168All output sequences end in a single line containing a period.
29169
29170@item
29171The @code{@var{token}} is from the corresponding request. Note that
29172for all async output, while the token is allowed by the grammar and
29173may be output by future versions of @value{GDBN} for select async
29174output messages, it is generally omitted. Frontends should treat
29175all async output as reporting general changes in the state of the
29176target and there should be no need to associate async output to any
29177prior command.
29178
29179@item
29180@cindex status output in @sc{gdb/mi}
29181@var{status-async-output} contains on-going status information about the
29182progress of a slow operation. It can be discarded. All status output is
29183prefixed by @samp{+}.
29184
29185@item
29186@cindex async output in @sc{gdb/mi}
29187@var{exec-async-output} contains asynchronous state change on the target
29188(stopped, started, disappeared). All async output is prefixed by
29189@samp{*}.
29190
29191@item
29192@cindex notify output in @sc{gdb/mi}
29193@var{notify-async-output} contains supplementary information that the
29194client should handle (e.g., a new breakpoint information). All notify
29195output is prefixed by @samp{=}.
29196
29197@item
29198@cindex console output in @sc{gdb/mi}
29199@var{console-stream-output} is output that should be displayed as is in the
29200console. It is the textual response to a CLI command. All the console
29201output is prefixed by @samp{~}.
29202
29203@item
29204@cindex target output in @sc{gdb/mi}
29205@var{target-stream-output} is the output produced by the target program.
29206All the target output is prefixed by @samp{@@}.
29207
29208@item
29209@cindex log output in @sc{gdb/mi}
29210@var{log-stream-output} is output text coming from @value{GDBN}'s internals, for
29211instance messages that should be displayed as part of an error log. All
29212the log output is prefixed by @samp{&}.
29213
29214@item
29215@cindex list output in @sc{gdb/mi}
29216New @sc{gdb/mi} commands should only output @var{lists} containing
29217@var{values}.
29218
29219
29220@end itemize
29221
29222@xref{GDB/MI Stream Records, , @sc{gdb/mi} Stream Records}, for more
29223details about the various output records.
29224
29225@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
29226@node GDB/MI Compatibility with CLI
29227@section @sc{gdb/mi} Compatibility with CLI
29228
29229@cindex compatibility, @sc{gdb/mi} and CLI
29230@cindex @sc{gdb/mi}, compatibility with CLI
29231
29232For the developers convenience CLI commands can be entered directly,
29233but there may be some unexpected behaviour. For example, commands
29234that query the user will behave as if the user replied yes, breakpoint
29235command lists are not executed and some CLI commands, such as
29236@code{if}, @code{when} and @code{define}, prompt for further input with
29237@samp{>}, which is not valid MI output.
29238
29239This feature may be removed at some stage in the future and it is
29240recommended that front ends use the @code{-interpreter-exec} command
29241(@pxref{-interpreter-exec}).
29242
29243@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
29244@node GDB/MI Development and Front Ends
29245@section @sc{gdb/mi} Development and Front Ends
29246@cindex @sc{gdb/mi} development
29247
29248The application which takes the MI output and presents the state of the
29249program being debugged to the user is called a @dfn{front end}.
29250
29251Although @sc{gdb/mi} is still incomplete, it is currently being used
29252by a variety of front ends to @value{GDBN}. This makes it difficult
29253to introduce new functionality without breaking existing usage. This
29254section tries to minimize the problems by describing how the protocol
29255might change.
29256
29257Some changes in MI need not break a carefully designed front end, and
29258for these the MI version will remain unchanged. The following is a
29259list of changes that may occur within one level, so front ends should
29260parse MI output in a way that can handle them:
29261
29262@itemize @bullet
29263@item
29264New MI commands may be added.
29265
29266@item
29267New fields may be added to the output of any MI command.
29268
29269@item
29270The range of values for fields with specified values, e.g.,
29271@code{in_scope} (@pxref{-var-update}) may be extended.
29272
29273@c The format of field's content e.g type prefix, may change so parse it
29274@c at your own risk. Yes, in general?
29275
29276@c The order of fields may change? Shouldn't really matter but it might
29277@c resolve inconsistencies.
29278@end itemize
29279
29280If the changes are likely to break front ends, the MI version level
29281will be increased by one. This will allow the front end to parse the
29282output according to the MI version. Apart from mi0, new versions of
29283@value{GDBN} will not support old versions of MI and it will be the
29284responsibility of the front end to work with the new one.
29285
29286@c Starting with mi3, add a new command -mi-version that prints the MI
29287@c version?
29288
29289The best way to avoid unexpected changes in MI that might break your front
29290end is to make your project known to @value{GDBN} developers and
29291follow development on @email{gdb@@sourceware.org} and
29292@email{gdb-patches@@sourceware.org}.
29293@cindex mailing lists
29294
29295@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
29296@node GDB/MI Output Records
29297@section @sc{gdb/mi} Output Records
29298
29299@menu
29300* GDB/MI Result Records::
29301* GDB/MI Stream Records::
29302* GDB/MI Async Records::
29303* GDB/MI Breakpoint Information::
29304* GDB/MI Frame Information::
29305* GDB/MI Thread Information::
29306* GDB/MI Ada Exception Information::
29307@end menu
29308
29309@node GDB/MI Result Records
29310@subsection @sc{gdb/mi} Result Records
29311
29312@cindex result records in @sc{gdb/mi}
29313@cindex @sc{gdb/mi}, result records
29314In addition to a number of out-of-band notifications, the response to a
29315@sc{gdb/mi} command includes one of the following result indications:
29316
29317@table @code
29318@findex ^done
29319@item "^done" [ "," @var{results} ]
29320The synchronous operation was successful, @code{@var{results}} are the return
29321values.
29322
29323@item "^running"
29324@findex ^running
29325This result record is equivalent to @samp{^done}. Historically, it
29326was output instead of @samp{^done} if the command has resumed the
29327target. This behaviour is maintained for backward compatibility, but
29328all frontends should treat @samp{^done} and @samp{^running}
29329identically and rely on the @samp{*running} output record to determine
29330which threads are resumed.
29331
29332@item "^connected"
29333@findex ^connected
29334@value{GDBN} has connected to a remote target.
29335
29336@item "^error" "," @var{c-string}
29337@findex ^error
29338The operation failed. The @code{@var{c-string}} contains the corresponding
29339error message.
29340
29341@item "^exit"
29342@findex ^exit
29343@value{GDBN} has terminated.
29344
29345@end table
29346
29347@node GDB/MI Stream Records
29348@subsection @sc{gdb/mi} Stream Records
29349
29350@cindex @sc{gdb/mi}, stream records
29351@cindex stream records in @sc{gdb/mi}
29352@value{GDBN} internally maintains a number of output streams: the console, the
29353target, and the log. The output intended for each of these streams is
29354funneled through the @sc{gdb/mi} interface using @dfn{stream records}.
29355
29356Each stream record begins with a unique @dfn{prefix character} which
29357identifies its stream (@pxref{GDB/MI Output Syntax, , @sc{gdb/mi} Output
29358Syntax}). In addition to the prefix, each stream record contains a
29359@code{@var{string-output}}. This is either raw text (with an implicit new
29360line) or a quoted C string (which does not contain an implicit newline).
29361
29362@table @code
29363@item "~" @var{string-output}
29364The console output stream contains text that should be displayed in the
29365CLI console window. It contains the textual responses to CLI commands.
29366
29367@item "@@" @var{string-output}
29368The target output stream contains any textual output from the running
29369target. This is only present when GDB's event loop is truly
29370asynchronous, which is currently only the case for remote targets.
29371
29372@item "&" @var{string-output}
29373The log stream contains debugging messages being produced by @value{GDBN}'s
29374internals.
29375@end table
29376
29377@node GDB/MI Async Records
29378@subsection @sc{gdb/mi} Async Records
29379
29380@cindex async records in @sc{gdb/mi}
29381@cindex @sc{gdb/mi}, async records
29382@dfn{Async} records are used to notify the @sc{gdb/mi} client of
29383additional changes that have occurred. Those changes can either be a
29384consequence of @sc{gdb/mi} commands (e.g., a breakpoint modified) or a result of
29385target activity (e.g., target stopped).
29386
29387The following is the list of possible async records:
29388
29389@table @code
29390
29391@item *running,thread-id="@var{thread}"
29392The target is now running. The @var{thread} field tells which
29393specific thread is now running, and can be @samp{all} if all threads
29394are running. The frontend should assume that no interaction with a
29395running thread is possible after this notification is produced.
29396The frontend should not assume that this notification is output
29397only once for any command. @value{GDBN} may emit this notification
29398several times, either for different threads, because it cannot resume
29399all threads together, or even for a single thread, if the thread must
29400be stepped though some code before letting it run freely.
29401
29402@item *stopped,reason="@var{reason}",thread-id="@var{id}",stopped-threads="@var{stopped}",core="@var{core}"
29403The target has stopped. The @var{reason} field can have one of the
29404following values:
29405
29406@table @code
29407@item breakpoint-hit
29408A breakpoint was reached.
29409@item watchpoint-trigger
29410A watchpoint was triggered.
29411@item read-watchpoint-trigger
29412A read watchpoint was triggered.
29413@item access-watchpoint-trigger
29414An access watchpoint was triggered.
29415@item function-finished
29416An -exec-finish or similar CLI command was accomplished.
29417@item location-reached
29418An -exec-until or similar CLI command was accomplished.
29419@item watchpoint-scope
29420A watchpoint has gone out of scope.
29421@item end-stepping-range
29422An -exec-next, -exec-next-instruction, -exec-step, -exec-step-instruction or
29423similar CLI command was accomplished.
29424@item exited-signalled
29425The inferior exited because of a signal.
29426@item exited
29427The inferior exited.
29428@item exited-normally
29429The inferior exited normally.
29430@item signal-received
29431A signal was received by the inferior.
29432@item solib-event
29433The inferior has stopped due to a library being loaded or unloaded.
29434This can happen when @code{stop-on-solib-events} (@pxref{Files}) is
29435set or when a @code{catch load} or @code{catch unload} catchpoint is
29436in use (@pxref{Set Catchpoints}).
29437@item fork
29438The inferior has forked. This is reported when @code{catch fork}
29439(@pxref{Set Catchpoints}) has been used.
29440@item vfork
29441The inferior has vforked. This is reported in when @code{catch vfork}
29442(@pxref{Set Catchpoints}) has been used.
29443@item syscall-entry
29444The inferior entered a system call. This is reported when @code{catch
29445syscall} (@pxref{Set Catchpoints}) has been used.
29446@item syscall-entry
29447The inferior returned from a system call. This is reported when
29448@code{catch syscall} (@pxref{Set Catchpoints}) has been used.
29449@item exec
29450The inferior called @code{exec}. This is reported when @code{catch exec}
29451(@pxref{Set Catchpoints}) has been used.
29452@end table
29453
29454The @var{id} field identifies the thread that directly caused the stop
29455-- for example by hitting a breakpoint. Depending on whether all-stop
29456mode is in effect (@pxref{All-Stop Mode}), @value{GDBN} may either
29457stop all threads, or only the thread that directly triggered the stop.
29458If all threads are stopped, the @var{stopped} field will have the
29459value of @code{"all"}. Otherwise, the value of the @var{stopped}
29460field will be a list of thread identifiers. Presently, this list will
29461always include a single thread, but frontend should be prepared to see
29462several threads in the list. The @var{core} field reports the
29463processor core on which the stop event has happened. This field may be absent
29464if such information is not available.
29465
29466@item =thread-group-added,id="@var{id}"
29467@itemx =thread-group-removed,id="@var{id}"
29468A thread group was either added or removed. The @var{id} field
29469contains the @value{GDBN} identifier of the thread group. When a thread
29470group is added, it generally might not be associated with a running
29471process. When a thread group is removed, its id becomes invalid and
29472cannot be used in any way.
29473
29474@item =thread-group-started,id="@var{id}",pid="@var{pid}"
29475A thread group became associated with a running program,
29476either because the program was just started or the thread group
29477was attached to a program. The @var{id} field contains the
29478@value{GDBN} identifier of the thread group. The @var{pid} field
29479contains process identifier, specific to the operating system.
29480
29481@item =thread-group-exited,id="@var{id}"[,exit-code="@var{code}"]
29482A thread group is no longer associated with a running program,
29483either because the program has exited, or because it was detached
29484from. The @var{id} field contains the @value{GDBN} identifier of the
29485thread group. @var{code} is the exit code of the inferior; it exists
29486only when the inferior exited with some code.
29487
29488@item =thread-created,id="@var{id}",group-id="@var{gid}"
29489@itemx =thread-exited,id="@var{id}",group-id="@var{gid}"
29490A thread either was created, or has exited. The @var{id} field
29491contains the @value{GDBN} identifier of the thread. The @var{gid}
29492field identifies the thread group this thread belongs to.
29493
29494@item =thread-selected,id="@var{id}"
29495Informs that the selected thread was changed as result of the last
29496command. This notification is not emitted as result of @code{-thread-select}
29497command but is emitted whenever an MI command that is not documented
29498to change the selected thread actually changes it. In particular,
29499invoking, directly or indirectly (via user-defined command), the CLI
29500@code{thread} command, will generate this notification.
29501
29502We suggest that in response to this notification, front ends
29503highlight the selected thread and cause subsequent commands to apply to
29504that thread.
29505
29506@item =library-loaded,...
29507Reports that a new library file was loaded by the program. This
29508notification has 4 fields---@var{id}, @var{target-name},
29509@var{host-name}, and @var{symbols-loaded}. The @var{id} field is an
29510opaque identifier of the library. For remote debugging case,
29511@var{target-name} and @var{host-name} fields give the name of the
29512library file on the target, and on the host respectively. For native
29513debugging, both those fields have the same value. The
29514@var{symbols-loaded} field is emitted only for backward compatibility
29515and should not be relied on to convey any useful information. The
29516@var{thread-group} field, if present, specifies the id of the thread
29517group in whose context the library was loaded. If the field is
29518absent, it means the library was loaded in the context of all present
29519thread groups.
29520
29521@item =library-unloaded,...
29522Reports that a library was unloaded by the program. This notification
29523has 3 fields---@var{id}, @var{target-name} and @var{host-name} with
29524the same meaning as for the @code{=library-loaded} notification.
29525The @var{thread-group} field, if present, specifies the id of the
29526thread group in whose context the library was unloaded. If the field is
29527absent, it means the library was unloaded in the context of all present
29528thread groups.
29529
29530@item =traceframe-changed,num=@var{tfnum},tracepoint=@var{tpnum}
29531@itemx =traceframe-changed,end
29532Reports that the trace frame was changed and its new number is
29533@var{tfnum}. The number of the tracepoint associated with this trace
29534frame is @var{tpnum}.
29535
29536@item =tsv-created,name=@var{name},initial=@var{initial}
29537Reports that the new trace state variable @var{name} is created with
29538initial value @var{initial}.
29539
29540@item =tsv-deleted,name=@var{name}
29541@itemx =tsv-deleted
29542Reports that the trace state variable @var{name} is deleted or all
29543trace state variables are deleted.
29544
29545@item =tsv-modified,name=@var{name},initial=@var{initial}[,current=@var{current}]
29546Reports that the trace state variable @var{name} is modified with
29547the initial value @var{initial}. The current value @var{current} of
29548trace state variable is optional and is reported if the current
29549value of trace state variable is known.
29550
29551@item =breakpoint-created,bkpt=@{...@}
29552@itemx =breakpoint-modified,bkpt=@{...@}
29553@itemx =breakpoint-deleted,id=@var{number}
29554Reports that a breakpoint was created, modified, or deleted,
29555respectively. Only user-visible breakpoints are reported to the MI
29556user.
29557
29558The @var{bkpt} argument is of the same form as returned by the various
29559breakpoint commands; @xref{GDB/MI Breakpoint Commands}. The
29560@var{number} is the ordinal number of the breakpoint.
29561
29562Note that if a breakpoint is emitted in the result record of a
29563command, then it will not also be emitted in an async record.
29564
29565@item =record-started,thread-group="@var{id}"
29566@itemx =record-stopped,thread-group="@var{id}"
29567Execution log recording was either started or stopped on an
29568inferior. The @var{id} is the @value{GDBN} identifier of the thread
29569group corresponding to the affected inferior.
29570
29571@item =cmd-param-changed,param=@var{param},value=@var{value}
29572Reports that a parameter of the command @code{set @var{param}} is
29573changed to @var{value}. In the multi-word @code{set} command,
29574the @var{param} is the whole parameter list to @code{set} command.
29575For example, In command @code{set check type on}, @var{param}
29576is @code{check type} and @var{value} is @code{on}.
29577
29578@item =memory-changed,thread-group=@var{id},addr=@var{addr},len=@var{len}[,type="code"]
29579Reports that bytes from @var{addr} to @var{data} + @var{len} were
29580written in an inferior. The @var{id} is the identifier of the
29581thread group corresponding to the affected inferior. The optional
29582@code{type="code"} part is reported if the memory written to holds
29583executable code.
29584@end table
29585
29586@node GDB/MI Breakpoint Information
29587@subsection @sc{gdb/mi} Breakpoint Information
29588
29589When @value{GDBN} reports information about a breakpoint, a
29590tracepoint, a watchpoint, or a catchpoint, it uses a tuple with the
29591following fields:
29592
29593@table @code
29594@item number
29595The breakpoint number. For a breakpoint that represents one location
29596of a multi-location breakpoint, this will be a dotted pair, like
29597@samp{1.2}.
29598
29599@item type
29600The type of the breakpoint. For ordinary breakpoints this will be
29601@samp{breakpoint}, but many values are possible.
29602
29603@item catch-type
29604If the type of the breakpoint is @samp{catchpoint}, then this
29605indicates the exact type of catchpoint.
29606
29607@item disp
29608This is the breakpoint disposition---either @samp{del}, meaning that
29609the breakpoint will be deleted at the next stop, or @samp{keep},
29610meaning that the breakpoint will not be deleted.
29611
29612@item enabled
29613This indicates whether the breakpoint is enabled, in which case the
29614value is @samp{y}, or disabled, in which case the value is @samp{n}.
29615Note that this is not the same as the field @code{enable}.
29616
29617@item addr
29618The address of the breakpoint. This may be a hexidecimal number,
29619giving the address; or the string @samp{<PENDING>}, for a pending
29620breakpoint; or the string @samp{<MULTIPLE>}, for a breakpoint with
29621multiple locations. This field will not be present if no address can
29622be determined. For example, a watchpoint does not have an address.
29623
29624@item func
29625If known, the function in which the breakpoint appears.
29626If not known, this field is not present.
29627
29628@item filename
29629The name of the source file which contains this function, if known.
29630If not known, this field is not present.
29631
29632@item fullname
29633The full file name of the source file which contains this function, if
29634known. If not known, this field is not present.
29635
29636@item line
29637The line number at which this breakpoint appears, if known.
29638If not known, this field is not present.
29639
29640@item at
29641If the source file is not known, this field may be provided. If
29642provided, this holds the address of the breakpoint, possibly followed
29643by a symbol name.
29644
29645@item pending
29646If this breakpoint is pending, this field is present and holds the
29647text used to set the breakpoint, as entered by the user.
29648
29649@item evaluated-by
29650Where this breakpoint's condition is evaluated, either @samp{host} or
29651@samp{target}.
29652
29653@item thread
29654If this is a thread-specific breakpoint, then this identifies the
29655thread in which the breakpoint can trigger.
29656
29657@item task
29658If this breakpoint is restricted to a particular Ada task, then this
29659field will hold the task identifier.
29660
29661@item cond
29662If the breakpoint is conditional, this is the condition expression.
29663
29664@item ignore
29665The ignore count of the breakpoint.
29666
29667@item enable
29668The enable count of the breakpoint.
29669
29670@item traceframe-usage
29671FIXME.
29672
29673@item static-tracepoint-marker-string-id
29674For a static tracepoint, the name of the static tracepoint marker.
29675
29676@item mask
29677For a masked watchpoint, this is the mask.
29678
29679@item pass
29680A tracepoint's pass count.
29681
29682@item original-location
29683The location of the breakpoint as originally specified by the user.
29684This field is optional.
29685
29686@item times
29687The number of times the breakpoint has been hit.
29688
29689@item installed
29690This field is only given for tracepoints. This is either @samp{y},
29691meaning that the tracepoint is installed, or @samp{n}, meaning that it
29692is not.
29693
29694@item what
29695Some extra data, the exact contents of which are type-dependent.
29696
29697@end table
29698
29699For example, here is what the output of @code{-break-insert}
29700(@pxref{GDB/MI Breakpoint Commands}) might be:
29701
29702@smallexample
29703-> -break-insert main
29704<- ^done,bkpt=@{number="1",type="breakpoint",disp="keep",
29705 enabled="y",addr="0x08048564",func="main",file="myprog.c",
29706 fullname="/home/nickrob/myprog.c",line="68",thread-groups=["i1"],
29707 times="0"@}
29708<- (gdb)
29709@end smallexample
29710
29711@node GDB/MI Frame Information
29712@subsection @sc{gdb/mi} Frame Information
29713
29714Response from many MI commands includes an information about stack
29715frame. This information is a tuple that may have the following
29716fields:
29717
29718@table @code
29719@item level
29720The level of the stack frame. The innermost frame has the level of
29721zero. This field is always present.
29722
29723@item func
29724The name of the function corresponding to the frame. This field may
29725be absent if @value{GDBN} is unable to determine the function name.
29726
29727@item addr
29728The code address for the frame. This field is always present.
29729
29730@item file
29731The name of the source files that correspond to the frame's code
29732address. This field may be absent.
29733
29734@item line
29735The source line corresponding to the frames' code address. This field
29736may be absent.
29737
29738@item from
29739The name of the binary file (either executable or shared library) the
29740corresponds to the frame's code address. This field may be absent.
29741
29742@end table
29743
29744@node GDB/MI Thread Information
29745@subsection @sc{gdb/mi} Thread Information
29746
29747Whenever @value{GDBN} has to report an information about a thread, it
29748uses a tuple with the following fields:
29749
29750@table @code
29751@item id
29752The numeric id assigned to the thread by @value{GDBN}. This field is
29753always present.
29754
29755@item target-id
29756Target-specific string identifying the thread. This field is always present.
29757
29758@item details
29759Additional information about the thread provided by the target.
29760It is supposed to be human-readable and not interpreted by the
29761frontend. This field is optional.
29762
29763@item state
29764Either @samp{stopped} or @samp{running}, depending on whether the
29765thread is presently running. This field is always present.
29766
29767@item core
29768The value of this field is an integer number of the processor core the
29769thread was last seen on. This field is optional.
29770@end table
29771
29772@node GDB/MI Ada Exception Information
29773@subsection @sc{gdb/mi} Ada Exception Information
29774
29775Whenever a @code{*stopped} record is emitted because the program
29776stopped after hitting an exception catchpoint (@pxref{Set Catchpoints}),
29777@value{GDBN} provides the name of the exception that was raised via
29778the @code{exception-name} field.
29779
29780@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
29781@node GDB/MI Simple Examples
29782@section Simple Examples of @sc{gdb/mi} Interaction
29783@cindex @sc{gdb/mi}, simple examples
29784
29785This subsection presents several simple examples of interaction using
29786the @sc{gdb/mi} interface. In these examples, @samp{->} means that the
29787following line is passed to @sc{gdb/mi} as input, while @samp{<-} means
29788the output received from @sc{gdb/mi}.
29789
29790Note the line breaks shown in the examples are here only for
29791readability, they don't appear in the real output.
29792
29793@subheading Setting a Breakpoint
29794
29795Setting a breakpoint generates synchronous output which contains detailed
29796information of the breakpoint.
29797
29798@smallexample
29799-> -break-insert main
29800<- ^done,bkpt=@{number="1",type="breakpoint",disp="keep",
29801 enabled="y",addr="0x08048564",func="main",file="myprog.c",
29802 fullname="/home/nickrob/myprog.c",line="68",thread-groups=["i1"],
29803 times="0"@}
29804<- (gdb)
29805@end smallexample
29806
29807@subheading Program Execution
29808
29809Program execution generates asynchronous records and MI gives the
29810reason that execution stopped.
29811
29812@smallexample
29813-> -exec-run
29814<- ^running
29815<- (gdb)
29816<- *stopped,reason="breakpoint-hit",disp="keep",bkptno="1",thread-id="0",
29817 frame=@{addr="0x08048564",func="main",
29818 args=[@{name="argc",value="1"@},@{name="argv",value="0xbfc4d4d4"@}],
29819 file="myprog.c",fullname="/home/nickrob/myprog.c",line="68"@}
29820<- (gdb)
29821-> -exec-continue
29822<- ^running
29823<- (gdb)
29824<- *stopped,reason="exited-normally"
29825<- (gdb)
29826@end smallexample
29827
29828@subheading Quitting @value{GDBN}
29829
29830Quitting @value{GDBN} just prints the result class @samp{^exit}.
29831
29832@smallexample
29833-> (gdb)
29834<- -gdb-exit
29835<- ^exit
29836@end smallexample
29837
29838Please note that @samp{^exit} is printed immediately, but it might
29839take some time for @value{GDBN} to actually exit. During that time, @value{GDBN}
29840performs necessary cleanups, including killing programs being debugged
29841or disconnecting from debug hardware, so the frontend should wait till
29842@value{GDBN} exits and should only forcibly kill @value{GDBN} if it
29843fails to exit in reasonable time.
29844
29845@subheading A Bad Command
29846
29847Here's what happens if you pass a non-existent command:
29848
29849@smallexample
29850-> -rubbish
29851<- ^error,msg="Undefined MI command: rubbish"
29852<- (gdb)
29853@end smallexample
29854
29855
29856@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
29857@node GDB/MI Command Description Format
29858@section @sc{gdb/mi} Command Description Format
29859
29860The remaining sections describe blocks of commands. Each block of
29861commands is laid out in a fashion similar to this section.
29862
29863@subheading Motivation
29864
29865The motivation for this collection of commands.
29866
29867@subheading Introduction
29868
29869A brief introduction to this collection of commands as a whole.
29870
29871@subheading Commands
29872
29873For each command in the block, the following is described:
29874
29875@subsubheading Synopsis
29876
29877@smallexample
29878 -command @var{args}@dots{}
29879@end smallexample
29880
29881@subsubheading Result
29882
29883@subsubheading @value{GDBN} Command
29884
29885The corresponding @value{GDBN} CLI command(s), if any.
29886
29887@subsubheading Example
29888
29889Example(s) formatted for readability. Some of the described commands have
29890not been implemented yet and these are labeled N.A.@: (not available).
29891
29892
29893@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
29894@node GDB/MI Breakpoint Commands
29895@section @sc{gdb/mi} Breakpoint Commands
29896
29897@cindex breakpoint commands for @sc{gdb/mi}
29898@cindex @sc{gdb/mi}, breakpoint commands
29899This section documents @sc{gdb/mi} commands for manipulating
29900breakpoints.
29901
29902@subheading The @code{-break-after} Command
29903@findex -break-after
29904
29905@subsubheading Synopsis
29906
29907@smallexample
29908 -break-after @var{number} @var{count}
29909@end smallexample
29910
29911The breakpoint number @var{number} is not in effect until it has been
29912hit @var{count} times. To see how this is reflected in the output of
29913the @samp{-break-list} command, see the description of the
29914@samp{-break-list} command below.
29915
29916@subsubheading @value{GDBN} Command
29917
29918The corresponding @value{GDBN} command is @samp{ignore}.
29919
29920@subsubheading Example
29921
29922@smallexample
29923(gdb)
29924-break-insert main
29925^done,bkpt=@{number="1",type="breakpoint",disp="keep",
29926enabled="y",addr="0x000100d0",func="main",file="hello.c",
29927fullname="/home/foo/hello.c",line="5",thread-groups=["i1"],
29928times="0"@}
29929(gdb)
29930-break-after 1 3
29931~
29932^done
29933(gdb)
29934-break-list
29935^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
29936hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
29937@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
29938@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
29939@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
29940@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
29941@{width="40",alignment="2",col_name="what",colhdr="What"@}],
29942body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
29943addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
29944line="5",thread-groups=["i1"],times="0",ignore="3"@}]@}
29945(gdb)
29946@end smallexample
29947
29948@ignore
29949@subheading The @code{-break-catch} Command
29950@findex -break-catch
29951@end ignore
29952
29953@subheading The @code{-break-commands} Command
29954@findex -break-commands
29955
29956@subsubheading Synopsis
29957
29958@smallexample
29959 -break-commands @var{number} [ @var{command1} ... @var{commandN} ]
29960@end smallexample
29961
29962Specifies the CLI commands that should be executed when breakpoint
29963@var{number} is hit. The parameters @var{command1} to @var{commandN}
29964are the commands. If no command is specified, any previously-set
29965commands are cleared. @xref{Break Commands}. Typical use of this
29966functionality is tracing a program, that is, printing of values of
29967some variables whenever breakpoint is hit and then continuing.
29968
29969@subsubheading @value{GDBN} Command
29970
29971The corresponding @value{GDBN} command is @samp{commands}.
29972
29973@subsubheading Example
29974
29975@smallexample
29976(gdb)
29977-break-insert main
29978^done,bkpt=@{number="1",type="breakpoint",disp="keep",
29979enabled="y",addr="0x000100d0",func="main",file="hello.c",
29980fullname="/home/foo/hello.c",line="5",thread-groups=["i1"],
29981times="0"@}
29982(gdb)
29983-break-commands 1 "print v" "continue"
29984^done
29985(gdb)
29986@end smallexample
29987
29988@subheading The @code{-break-condition} Command
29989@findex -break-condition
29990
29991@subsubheading Synopsis
29992
29993@smallexample
29994 -break-condition @var{number} @var{expr}
29995@end smallexample
29996
29997Breakpoint @var{number} will stop the program only if the condition in
29998@var{expr} is true. The condition becomes part of the
29999@samp{-break-list} output (see the description of the @samp{-break-list}
30000command below).
30001
30002@subsubheading @value{GDBN} Command
30003
30004The corresponding @value{GDBN} command is @samp{condition}.
30005
30006@subsubheading Example
30007
30008@smallexample
30009(gdb)
30010-break-condition 1 1
30011^done
30012(gdb)
30013-break-list
30014^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
30015hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30016@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30017@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30018@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30019@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30020@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30021body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
30022addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
30023line="5",cond="1",thread-groups=["i1"],times="0",ignore="3"@}]@}
30024(gdb)
30025@end smallexample
30026
30027@subheading The @code{-break-delete} Command
30028@findex -break-delete
30029
30030@subsubheading Synopsis
30031
30032@smallexample
30033 -break-delete ( @var{breakpoint} )+
30034@end smallexample
30035
30036Delete the breakpoint(s) whose number(s) are specified in the argument
30037list. This is obviously reflected in the breakpoint list.
30038
30039@subsubheading @value{GDBN} Command
30040
30041The corresponding @value{GDBN} command is @samp{delete}.
30042
30043@subsubheading Example
30044
30045@smallexample
30046(gdb)
30047-break-delete 1
30048^done
30049(gdb)
30050-break-list
30051^done,BreakpointTable=@{nr_rows="0",nr_cols="6",
30052hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30053@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30054@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30055@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30056@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30057@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30058body=[]@}
30059(gdb)
30060@end smallexample
30061
30062@subheading The @code{-break-disable} Command
30063@findex -break-disable
30064
30065@subsubheading Synopsis
30066
30067@smallexample
30068 -break-disable ( @var{breakpoint} )+
30069@end smallexample
30070
30071Disable the named @var{breakpoint}(s). The field @samp{enabled} in the
30072break list is now set to @samp{n} for the named @var{breakpoint}(s).
30073
30074@subsubheading @value{GDBN} Command
30075
30076The corresponding @value{GDBN} command is @samp{disable}.
30077
30078@subsubheading Example
30079
30080@smallexample
30081(gdb)
30082-break-disable 2
30083^done
30084(gdb)
30085-break-list
30086^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
30087hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30088@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30089@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30090@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30091@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30092@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30093body=[bkpt=@{number="2",type="breakpoint",disp="keep",enabled="n",
30094addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
30095line="5",thread-groups=["i1"],times="0"@}]@}
30096(gdb)
30097@end smallexample
30098
30099@subheading The @code{-break-enable} Command
30100@findex -break-enable
30101
30102@subsubheading Synopsis
30103
30104@smallexample
30105 -break-enable ( @var{breakpoint} )+
30106@end smallexample
30107
30108Enable (previously disabled) @var{breakpoint}(s).
30109
30110@subsubheading @value{GDBN} Command
30111
30112The corresponding @value{GDBN} command is @samp{enable}.
30113
30114@subsubheading Example
30115
30116@smallexample
30117(gdb)
30118-break-enable 2
30119^done
30120(gdb)
30121-break-list
30122^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
30123hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30124@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30125@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30126@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30127@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30128@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30129body=[bkpt=@{number="2",type="breakpoint",disp="keep",enabled="y",
30130addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
30131line="5",thread-groups=["i1"],times="0"@}]@}
30132(gdb)
30133@end smallexample
30134
30135@subheading The @code{-break-info} Command
30136@findex -break-info
30137
30138@subsubheading Synopsis
30139
30140@smallexample
30141 -break-info @var{breakpoint}
30142@end smallexample
30143
30144@c REDUNDANT???
30145Get information about a single breakpoint.
30146
30147The result is a table of breakpoints. @xref{GDB/MI Breakpoint
30148Information}, for details on the format of each breakpoint in the
30149table.
30150
30151@subsubheading @value{GDBN} Command
30152
30153The corresponding @value{GDBN} command is @samp{info break @var{breakpoint}}.
30154
30155@subsubheading Example
30156N.A.
30157
30158@subheading The @code{-break-insert} Command
30159@findex -break-insert
30160
30161@subsubheading Synopsis
30162
30163@smallexample
30164 -break-insert [ -t ] [ -h ] [ -f ] [ -d ] [ -a ]
30165 [ -c @var{condition} ] [ -i @var{ignore-count} ]
30166 [ -p @var{thread-id} ] [ @var{location} ]
30167@end smallexample
30168
30169@noindent
30170If specified, @var{location}, can be one of:
30171
30172@itemize @bullet
30173@item function
30174@c @item +offset
30175@c @item -offset
30176@c @item linenum
30177@item filename:linenum
30178@item filename:function
30179@item *address
30180@end itemize
30181
30182The possible optional parameters of this command are:
30183
30184@table @samp
30185@item -t
30186Insert a temporary breakpoint.
30187@item -h
30188Insert a hardware breakpoint.
30189@item -f
30190If @var{location} cannot be parsed (for example if it
30191refers to unknown files or functions), create a pending
30192breakpoint. Without this flag, @value{GDBN} will report
30193an error, and won't create a breakpoint, if @var{location}
30194cannot be parsed.
30195@item -d
30196Create a disabled breakpoint.
30197@item -a
30198Create a tracepoint. @xref{Tracepoints}. When this parameter
30199is used together with @samp{-h}, a fast tracepoint is created.
30200@item -c @var{condition}
30201Make the breakpoint conditional on @var{condition}.
30202@item -i @var{ignore-count}
30203Initialize the @var{ignore-count}.
30204@item -p @var{thread-id}
30205Restrict the breakpoint to the specified @var{thread-id}.
30206@end table
30207
30208@subsubheading Result
30209
30210@xref{GDB/MI Breakpoint Information}, for details on the format of the
30211resulting breakpoint.
30212
30213Note: this format is open to change.
30214@c An out-of-band breakpoint instead of part of the result?
30215
30216@subsubheading @value{GDBN} Command
30217
30218The corresponding @value{GDBN} commands are @samp{break}, @samp{tbreak},
30219@samp{hbreak}, and @samp{thbreak}. @c and @samp{rbreak}.
30220
30221@subsubheading Example
30222
30223@smallexample
30224(gdb)
30225-break-insert main
30226^done,bkpt=@{number="1",addr="0x0001072c",file="recursive2.c",
30227fullname="/home/foo/recursive2.c,line="4",thread-groups=["i1"],
30228times="0"@}
30229(gdb)
30230-break-insert -t foo
30231^done,bkpt=@{number="2",addr="0x00010774",file="recursive2.c",
30232fullname="/home/foo/recursive2.c,line="11",thread-groups=["i1"],
30233times="0"@}
30234(gdb)
30235-break-list
30236^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
30237hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30238@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30239@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30240@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30241@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30242@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30243body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
30244addr="0x0001072c", func="main",file="recursive2.c",
30245fullname="/home/foo/recursive2.c,"line="4",thread-groups=["i1"],
30246times="0"@},
30247bkpt=@{number="2",type="breakpoint",disp="del",enabled="y",
30248addr="0x00010774",func="foo",file="recursive2.c",
30249fullname="/home/foo/recursive2.c",line="11",thread-groups=["i1"],
30250times="0"@}]@}
30251(gdb)
30252@c -break-insert -r foo.*
30253@c ~int foo(int, int);
30254@c ^done,bkpt=@{number="3",addr="0x00010774",file="recursive2.c,
30255@c "fullname="/home/foo/recursive2.c",line="11",thread-groups=["i1"],
30256@c times="0"@}
30257@c (gdb)
30258@end smallexample
30259
30260@subheading The @code{-dprintf-insert} Command
30261@findex -dprintf-insert
30262
30263@subsubheading Synopsis
30264
30265@smallexample
30266 -dprintf-insert [ -t ] [ -f ] [ -d ]
30267 [ -c @var{condition} ] [ -i @var{ignore-count} ]
30268 [ -p @var{thread-id} ] [ @var{location} ] [ @var{format} ]
30269 [ @var{argument} ]
30270@end smallexample
30271
30272@noindent
30273If specified, @var{location}, can be one of:
30274
30275@itemize @bullet
30276@item @var{function}
30277@c @item +offset
30278@c @item -offset
30279@c @item @var{linenum}
30280@item @var{filename}:@var{linenum}
30281@item @var{filename}:function
30282@item *@var{address}
30283@end itemize
30284
30285The possible optional parameters of this command are:
30286
30287@table @samp
30288@item -t
30289Insert a temporary breakpoint.
30290@item -f
30291If @var{location} cannot be parsed (for example, if it
30292refers to unknown files or functions), create a pending
30293breakpoint. Without this flag, @value{GDBN} will report
30294an error, and won't create a breakpoint, if @var{location}
30295cannot be parsed.
30296@item -d
30297Create a disabled breakpoint.
30298@item -c @var{condition}
30299Make the breakpoint conditional on @var{condition}.
30300@item -i @var{ignore-count}
30301Set the ignore count of the breakpoint (@pxref{Conditions, ignore count})
30302to @var{ignore-count}.
30303@item -p @var{thread-id}
30304Restrict the breakpoint to the specified @var{thread-id}.
30305@end table
30306
30307@subsubheading Result
30308
30309@xref{GDB/MI Breakpoint Information}, for details on the format of the
30310resulting breakpoint.
30311
30312@c An out-of-band breakpoint instead of part of the result?
30313
30314@subsubheading @value{GDBN} Command
30315
30316The corresponding @value{GDBN} command is @samp{dprintf}.
30317
30318@subsubheading Example
30319
30320@smallexample
30321(gdb)
303224-dprintf-insert foo "At foo entry\n"
303234^done,bkpt=@{number="1",type="dprintf",disp="keep",enabled="y",
30324addr="0x000000000040061b",func="foo",file="mi-dprintf.c",
30325fullname="mi-dprintf.c",line="25",thread-groups=["i1"],
30326times="0",script=@{"printf \"At foo entry\\n\"","continue"@},
30327original-location="foo"@}
30328(gdb)
303295-dprintf-insert 26 "arg=%d, g=%d\n" arg g
303305^done,bkpt=@{number="2",type="dprintf",disp="keep",enabled="y",
30331addr="0x000000000040062a",func="foo",file="mi-dprintf.c",
30332fullname="mi-dprintf.c",line="26",thread-groups=["i1"],
30333times="0",script=@{"printf \"arg=%d, g=%d\\n\", arg, g","continue"@},
30334original-location="mi-dprintf.c:26"@}
30335(gdb)
30336@end smallexample
30337
30338@subheading The @code{-break-list} Command
30339@findex -break-list
30340
30341@subsubheading Synopsis
30342
30343@smallexample
30344 -break-list
30345@end smallexample
30346
30347Displays the list of inserted breakpoints, showing the following fields:
30348
30349@table @samp
30350@item Number
30351number of the breakpoint
30352@item Type
30353type of the breakpoint: @samp{breakpoint} or @samp{watchpoint}
30354@item Disposition
30355should the breakpoint be deleted or disabled when it is hit: @samp{keep}
30356or @samp{nokeep}
30357@item Enabled
30358is the breakpoint enabled or no: @samp{y} or @samp{n}
30359@item Address
30360memory location at which the breakpoint is set
30361@item What
30362logical location of the breakpoint, expressed by function name, file
30363name, line number
30364@item Thread-groups
30365list of thread groups to which this breakpoint applies
30366@item Times
30367number of times the breakpoint has been hit
30368@end table
30369
30370If there are no breakpoints or watchpoints, the @code{BreakpointTable}
30371@code{body} field is an empty list.
30372
30373@subsubheading @value{GDBN} Command
30374
30375The corresponding @value{GDBN} command is @samp{info break}.
30376
30377@subsubheading Example
30378
30379@smallexample
30380(gdb)
30381-break-list
30382^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
30383hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30384@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30385@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30386@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30387@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30388@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30389body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
30390addr="0x000100d0",func="main",file="hello.c",line="5",thread-groups=["i1"],
30391times="0"@},
30392bkpt=@{number="2",type="breakpoint",disp="keep",enabled="y",
30393addr="0x00010114",func="foo",file="hello.c",fullname="/home/foo/hello.c",
30394line="13",thread-groups=["i1"],times="0"@}]@}
30395(gdb)
30396@end smallexample
30397
30398Here's an example of the result when there are no breakpoints:
30399
30400@smallexample
30401(gdb)
30402-break-list
30403^done,BreakpointTable=@{nr_rows="0",nr_cols="6",
30404hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30405@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30406@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30407@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30408@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30409@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30410body=[]@}
30411(gdb)
30412@end smallexample
30413
30414@subheading The @code{-break-passcount} Command
30415@findex -break-passcount
30416
30417@subsubheading Synopsis
30418
30419@smallexample
30420 -break-passcount @var{tracepoint-number} @var{passcount}
30421@end smallexample
30422
30423Set the passcount for tracepoint @var{tracepoint-number} to
30424@var{passcount}. If the breakpoint referred to by @var{tracepoint-number}
30425is not a tracepoint, error is emitted. This corresponds to CLI
30426command @samp{passcount}.
30427
30428@subheading The @code{-break-watch} Command
30429@findex -break-watch
30430
30431@subsubheading Synopsis
30432
30433@smallexample
30434 -break-watch [ -a | -r ]
30435@end smallexample
30436
30437Create a watchpoint. With the @samp{-a} option it will create an
30438@dfn{access} watchpoint, i.e., a watchpoint that triggers either on a
30439read from or on a write to the memory location. With the @samp{-r}
30440option, the watchpoint created is a @dfn{read} watchpoint, i.e., it will
30441trigger only when the memory location is accessed for reading. Without
30442either of the options, the watchpoint created is a regular watchpoint,
30443i.e., it will trigger when the memory location is accessed for writing.
30444@xref{Set Watchpoints, , Setting Watchpoints}.
30445
30446Note that @samp{-break-list} will report a single list of watchpoints and
30447breakpoints inserted.
30448
30449@subsubheading @value{GDBN} Command
30450
30451The corresponding @value{GDBN} commands are @samp{watch}, @samp{awatch}, and
30452@samp{rwatch}.
30453
30454@subsubheading Example
30455
30456Setting a watchpoint on a variable in the @code{main} function:
30457
30458@smallexample
30459(gdb)
30460-break-watch x
30461^done,wpt=@{number="2",exp="x"@}
30462(gdb)
30463-exec-continue
30464^running
30465(gdb)
30466*stopped,reason="watchpoint-trigger",wpt=@{number="2",exp="x"@},
30467value=@{old="-268439212",new="55"@},
30468frame=@{func="main",args=[],file="recursive2.c",
30469fullname="/home/foo/bar/recursive2.c",line="5"@}
30470(gdb)
30471@end smallexample
30472
30473Setting a watchpoint on a variable local to a function. @value{GDBN} will stop
30474the program execution twice: first for the variable changing value, then
30475for the watchpoint going out of scope.
30476
30477@smallexample
30478(gdb)
30479-break-watch C
30480^done,wpt=@{number="5",exp="C"@}
30481(gdb)
30482-exec-continue
30483^running
30484(gdb)
30485*stopped,reason="watchpoint-trigger",
30486wpt=@{number="5",exp="C"@},value=@{old="-276895068",new="3"@},
30487frame=@{func="callee4",args=[],
30488file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
30489fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="13"@}
30490(gdb)
30491-exec-continue
30492^running
30493(gdb)
30494*stopped,reason="watchpoint-scope",wpnum="5",
30495frame=@{func="callee3",args=[@{name="strarg",
30496value="0x11940 \"A string argument.\""@}],
30497file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
30498fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="18"@}
30499(gdb)
30500@end smallexample
30501
30502Listing breakpoints and watchpoints, at different points in the program
30503execution. Note that once the watchpoint goes out of scope, it is
30504deleted.
30505
30506@smallexample
30507(gdb)
30508-break-watch C
30509^done,wpt=@{number="2",exp="C"@}
30510(gdb)
30511-break-list
30512^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
30513hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30514@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30515@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30516@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30517@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30518@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30519body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
30520addr="0x00010734",func="callee4",
30521file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
30522fullname="/home/foo/devo/gdb/testsuite/gdb.mi/basics.c"line="8",thread-groups=["i1"],
30523times="1"@},
30524bkpt=@{number="2",type="watchpoint",disp="keep",
30525enabled="y",addr="",what="C",thread-groups=["i1"],times="0"@}]@}
30526(gdb)
30527-exec-continue
30528^running
30529(gdb)
30530*stopped,reason="watchpoint-trigger",wpt=@{number="2",exp="C"@},
30531value=@{old="-276895068",new="3"@},
30532frame=@{func="callee4",args=[],
30533file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
30534fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="13"@}
30535(gdb)
30536-break-list
30537^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
30538hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30539@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30540@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30541@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30542@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30543@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30544body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
30545addr="0x00010734",func="callee4",
30546file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
30547fullname="/home/foo/devo/gdb/testsuite/gdb.mi/basics.c",line="8",thread-groups=["i1"],
30548times="1"@},
30549bkpt=@{number="2",type="watchpoint",disp="keep",
30550enabled="y",addr="",what="C",thread-groups=["i1"],times="-5"@}]@}
30551(gdb)
30552-exec-continue
30553^running
30554^done,reason="watchpoint-scope",wpnum="2",
30555frame=@{func="callee3",args=[@{name="strarg",
30556value="0x11940 \"A string argument.\""@}],
30557file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
30558fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="18"@}
30559(gdb)
30560-break-list
30561^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
30562hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
30563@{width="14",alignment="-1",col_name="type",colhdr="Type"@},
30564@{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
30565@{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
30566@{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
30567@{width="40",alignment="2",col_name="what",colhdr="What"@}],
30568body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
30569addr="0x00010734",func="callee4",
30570file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
30571fullname="/home/foo/devo/gdb/testsuite/gdb.mi/basics.c",line="8",
30572thread-groups=["i1"],times="1"@}]@}
30573(gdb)
30574@end smallexample
30575
30576
30577@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30578@node GDB/MI Catchpoint Commands
30579@section @sc{gdb/mi} Catchpoint Commands
30580
30581This section documents @sc{gdb/mi} commands for manipulating
30582catchpoints.
30583
30584@menu
30585* Shared Library GDB/MI Catchpoint Commands::
30586* Ada Exception GDB/MI Catchpoint Commands::
30587@end menu
30588
30589@node Shared Library GDB/MI Catchpoint Commands
30590@subsection Shared Library @sc{gdb/mi} Catchpoints
30591
30592@subheading The @code{-catch-load} Command
30593@findex -catch-load
30594
30595@subsubheading Synopsis
30596
30597@smallexample
30598 -catch-load [ -t ] [ -d ] @var{regexp}
30599@end smallexample
30600
30601Add a catchpoint for library load events. If the @samp{-t} option is used,
30602the catchpoint is a temporary one (@pxref{Set Breaks, ,Setting
30603Breakpoints}). If the @samp{-d} option is used, the catchpoint is created
30604in a disabled state. The @samp{regexp} argument is a regular
30605expression used to match the name of the loaded library.
30606
30607
30608@subsubheading @value{GDBN} Command
30609
30610The corresponding @value{GDBN} command is @samp{catch load}.
30611
30612@subsubheading Example
30613
30614@smallexample
30615-catch-load -t foo.so
30616^done,bkpt=@{number="1",type="catchpoint",disp="del",enabled="y",
30617what="load of library matching foo.so",catch-type="load",times="0"@}
30618(gdb)
30619@end smallexample
30620
30621
30622@subheading The @code{-catch-unload} Command
30623@findex -catch-unload
30624
30625@subsubheading Synopsis
30626
30627@smallexample
30628 -catch-unload [ -t ] [ -d ] @var{regexp}
30629@end smallexample
30630
30631Add a catchpoint for library unload events. If the @samp{-t} option is
30632used, the catchpoint is a temporary one (@pxref{Set Breaks, ,Setting
30633Breakpoints}). If the @samp{-d} option is used, the catchpoint is
30634created in a disabled state. The @samp{regexp} argument is a regular
30635expression used to match the name of the unloaded library.
30636
30637@subsubheading @value{GDBN} Command
30638
30639The corresponding @value{GDBN} command is @samp{catch unload}.
30640
30641@subsubheading Example
30642
30643@smallexample
30644-catch-unload -d bar.so
30645^done,bkpt=@{number="2",type="catchpoint",disp="keep",enabled="n",
30646what="load of library matching bar.so",catch-type="unload",times="0"@}
30647(gdb)
30648@end smallexample
30649
30650@node Ada Exception GDB/MI Catchpoint Commands
30651@subsection Ada Exception @sc{gdb/mi} Catchpoints
30652
30653The following @sc{gdb/mi} commands can be used to create catchpoints
30654that stop the execution when Ada exceptions are being raised.
30655
30656@subheading The @code{-catch-assert} Command
30657@findex -catch-assert
30658
30659@subsubheading Synopsis
30660
30661@smallexample
30662 -catch-assert [ -c @var{condition}] [ -d ] [ -t ]
30663@end smallexample
30664
30665Add a catchpoint for failed Ada assertions.
30666
30667The possible optional parameters for this command are:
30668
30669@table @samp
30670@item -c @var{condition}
30671Make the catchpoint conditional on @var{condition}.
30672@item -d
30673Create a disabled catchpoint.
30674@item -t
30675Create a temporary catchpoint.
30676@end table
30677
30678@subsubheading @value{GDBN} Command
30679
30680The corresponding @value{GDBN} command is @samp{catch assert}.
30681
30682@subsubheading Example
30683
30684@smallexample
30685-catch-assert
30686^done,bkptno="5",bkpt=@{number="5",type="breakpoint",disp="keep",
30687enabled="y",addr="0x0000000000404888",what="failed Ada assertions",
30688thread-groups=["i1"],times="0",
30689original-location="__gnat_debug_raise_assert_failure"@}
30690(gdb)
30691@end smallexample
30692
30693@subheading The @code{-catch-exception} Command
30694@findex -catch-exception
30695
30696@subsubheading Synopsis
30697
30698@smallexample
30699 -catch-exception [ -c @var{condition}] [ -d ] [ -e @var{exception-name} ]
30700 [ -t ] [ -u ]
30701@end smallexample
30702
30703Add a catchpoint stopping when Ada exceptions are raised.
30704By default, the command stops the program when any Ada exception
30705gets raised. But it is also possible, by using some of the
30706optional parameters described below, to create more selective
30707catchpoints.
30708
30709The possible optional parameters for this command are:
30710
30711@table @samp
30712@item -c @var{condition}
30713Make the catchpoint conditional on @var{condition}.
30714@item -d
30715Create a disabled catchpoint.
30716@item -e @var{exception-name}
30717Only stop when @var{exception-name} is raised. This option cannot
30718be used combined with @samp{-u}.
30719@item -t
30720Create a temporary catchpoint.
30721@item -u
30722Stop only when an unhandled exception gets raised. This option
30723cannot be used combined with @samp{-e}.
30724@end table
30725
30726@subsubheading @value{GDBN} Command
30727
30728The corresponding @value{GDBN} commands are @samp{catch exception}
30729and @samp{catch exception unhandled}.
30730
30731@subsubheading Example
30732
30733@smallexample
30734-catch-exception -e Program_Error
30735^done,bkptno="4",bkpt=@{number="4",type="breakpoint",disp="keep",
30736enabled="y",addr="0x0000000000404874",
30737what="`Program_Error' Ada exception", thread-groups=["i1"],
30738times="0",original-location="__gnat_debug_raise_exception"@}
30739(gdb)
30740@end smallexample
30741
30742@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30743@node GDB/MI Program Context
30744@section @sc{gdb/mi} Program Context
30745
30746@subheading The @code{-exec-arguments} Command
30747@findex -exec-arguments
30748
30749
30750@subsubheading Synopsis
30751
30752@smallexample
30753 -exec-arguments @var{args}
30754@end smallexample
30755
30756Set the inferior program arguments, to be used in the next
30757@samp{-exec-run}.
30758
30759@subsubheading @value{GDBN} Command
30760
30761The corresponding @value{GDBN} command is @samp{set args}.
30762
30763@subsubheading Example
30764
30765@smallexample
30766(gdb)
30767-exec-arguments -v word
30768^done
30769(gdb)
30770@end smallexample
30771
30772
30773@ignore
30774@subheading The @code{-exec-show-arguments} Command
30775@findex -exec-show-arguments
30776
30777@subsubheading Synopsis
30778
30779@smallexample
30780 -exec-show-arguments
30781@end smallexample
30782
30783Print the arguments of the program.
30784
30785@subsubheading @value{GDBN} Command
30786
30787The corresponding @value{GDBN} command is @samp{show args}.
30788
30789@subsubheading Example
30790N.A.
30791@end ignore
30792
30793
30794@subheading The @code{-environment-cd} Command
30795@findex -environment-cd
30796
30797@subsubheading Synopsis
30798
30799@smallexample
30800 -environment-cd @var{pathdir}
30801@end smallexample
30802
30803Set @value{GDBN}'s working directory.
30804
30805@subsubheading @value{GDBN} Command
30806
30807The corresponding @value{GDBN} command is @samp{cd}.
30808
30809@subsubheading Example
30810
30811@smallexample
30812(gdb)
30813-environment-cd /kwikemart/marge/ezannoni/flathead-dev/devo/gdb
30814^done
30815(gdb)
30816@end smallexample
30817
30818
30819@subheading The @code{-environment-directory} Command
30820@findex -environment-directory
30821
30822@subsubheading Synopsis
30823
30824@smallexample
30825 -environment-directory [ -r ] [ @var{pathdir} ]+
30826@end smallexample
30827
30828Add directories @var{pathdir} to beginning of search path for source files.
30829If the @samp{-r} option is used, the search path is reset to the default
30830search path. If directories @var{pathdir} are supplied in addition to the
30831@samp{-r} option, the search path is first reset and then addition
30832occurs as normal.
30833Multiple directories may be specified, separated by blanks. Specifying
30834multiple directories in a single command
30835results in the directories added to the beginning of the
30836search path in the same order they were presented in the command.
30837If blanks are needed as
30838part of a directory name, double-quotes should be used around
30839the name. In the command output, the path will show up separated
30840by the system directory-separator character. The directory-separator
30841character must not be used
30842in any directory name.
30843If no directories are specified, the current search path is displayed.
30844
30845@subsubheading @value{GDBN} Command
30846
30847The corresponding @value{GDBN} command is @samp{dir}.
30848
30849@subsubheading Example
30850
30851@smallexample
30852(gdb)
30853-environment-directory /kwikemart/marge/ezannoni/flathead-dev/devo/gdb
30854^done,source-path="/kwikemart/marge/ezannoni/flathead-dev/devo/gdb:$cdir:$cwd"
30855(gdb)
30856-environment-directory ""
30857^done,source-path="/kwikemart/marge/ezannoni/flathead-dev/devo/gdb:$cdir:$cwd"
30858(gdb)
30859-environment-directory -r /home/jjohnstn/src/gdb /usr/src
30860^done,source-path="/home/jjohnstn/src/gdb:/usr/src:$cdir:$cwd"
30861(gdb)
30862-environment-directory -r
30863^done,source-path="$cdir:$cwd"
30864(gdb)
30865@end smallexample
30866
30867
30868@subheading The @code{-environment-path} Command
30869@findex -environment-path
30870
30871@subsubheading Synopsis
30872
30873@smallexample
30874 -environment-path [ -r ] [ @var{pathdir} ]+
30875@end smallexample
30876
30877Add directories @var{pathdir} to beginning of search path for object files.
30878If the @samp{-r} option is used, the search path is reset to the original
30879search path that existed at gdb start-up. If directories @var{pathdir} are
30880supplied in addition to the
30881@samp{-r} option, the search path is first reset and then addition
30882occurs as normal.
30883Multiple directories may be specified, separated by blanks. Specifying
30884multiple directories in a single command
30885results in the directories added to the beginning of the
30886search path in the same order they were presented in the command.
30887If blanks are needed as
30888part of a directory name, double-quotes should be used around
30889the name. In the command output, the path will show up separated
30890by the system directory-separator character. The directory-separator
30891character must not be used
30892in any directory name.
30893If no directories are specified, the current path is displayed.
30894
30895
30896@subsubheading @value{GDBN} Command
30897
30898The corresponding @value{GDBN} command is @samp{path}.
30899
30900@subsubheading Example
30901
30902@smallexample
30903(gdb)
30904-environment-path
30905^done,path="/usr/bin"
30906(gdb)
30907-environment-path /kwikemart/marge/ezannoni/flathead-dev/ppc-eabi/gdb /bin
30908^done,path="/kwikemart/marge/ezannoni/flathead-dev/ppc-eabi/gdb:/bin:/usr/bin"
30909(gdb)
30910-environment-path -r /usr/local/bin
30911^done,path="/usr/local/bin:/usr/bin"
30912(gdb)
30913@end smallexample
30914
30915
30916@subheading The @code{-environment-pwd} Command
30917@findex -environment-pwd
30918
30919@subsubheading Synopsis
30920
30921@smallexample
30922 -environment-pwd
30923@end smallexample
30924
30925Show the current working directory.
30926
30927@subsubheading @value{GDBN} Command
30928
30929The corresponding @value{GDBN} command is @samp{pwd}.
30930
30931@subsubheading Example
30932
30933@smallexample
30934(gdb)
30935-environment-pwd
30936^done,cwd="/kwikemart/marge/ezannoni/flathead-dev/devo/gdb"
30937(gdb)
30938@end smallexample
30939
30940@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30941@node GDB/MI Thread Commands
30942@section @sc{gdb/mi} Thread Commands
30943
30944
30945@subheading The @code{-thread-info} Command
30946@findex -thread-info
30947
30948@subsubheading Synopsis
30949
30950@smallexample
30951 -thread-info [ @var{thread-id} ]
30952@end smallexample
30953
30954Reports information about either a specific thread, if
30955the @var{thread-id} parameter is present, or about all
30956threads. When printing information about all threads,
30957also reports the current thread.
30958
30959@subsubheading @value{GDBN} Command
30960
30961The @samp{info thread} command prints the same information
30962about all threads.
30963
30964@subsubheading Result
30965
30966The result is a list of threads. The following attributes are
30967defined for a given thread:
30968
30969@table @samp
30970@item current
30971This field exists only for the current thread. It has the value @samp{*}.
30972
30973@item id
30974The identifier that @value{GDBN} uses to refer to the thread.
30975
30976@item target-id
30977The identifier that the target uses to refer to the thread.
30978
30979@item details
30980Extra information about the thread, in a target-specific format. This
30981field is optional.
30982
30983@item name
30984The name of the thread. If the user specified a name using the
30985@code{thread name} command, then this name is given. Otherwise, if
30986@value{GDBN} can extract the thread name from the target, then that
30987name is given. If @value{GDBN} cannot find the thread name, then this
30988field is omitted.
30989
30990@item frame
30991The stack frame currently executing in the thread.
30992
30993@item state
30994The thread's state. The @samp{state} field may have the following
30995values:
30996
30997@table @code
30998@item stopped
30999The thread is stopped. Frame information is available for stopped
31000threads.
31001
31002@item running
31003The thread is running. There's no frame information for running
31004threads.
31005
31006@end table
31007
31008@item core
31009If @value{GDBN} can find the CPU core on which this thread is running,
31010then this field is the core identifier. This field is optional.
31011
31012@end table
31013
31014@subsubheading Example
31015
31016@smallexample
31017-thread-info
31018^done,threads=[
31019@{id="2",target-id="Thread 0xb7e14b90 (LWP 21257)",
31020 frame=@{level="0",addr="0xffffe410",func="__kernel_vsyscall",
31021 args=[]@},state="running"@},
31022@{id="1",target-id="Thread 0xb7e156b0 (LWP 21254)",
31023 frame=@{level="0",addr="0x0804891f",func="foo",
31024 args=[@{name="i",value="10"@}],
31025 file="/tmp/a.c",fullname="/tmp/a.c",line="158"@},
31026 state="running"@}],
31027current-thread-id="1"
31028(gdb)
31029@end smallexample
31030
31031@subheading The @code{-thread-list-ids} Command
31032@findex -thread-list-ids
31033
31034@subsubheading Synopsis
31035
31036@smallexample
31037 -thread-list-ids
31038@end smallexample
31039
31040Produces a list of the currently known @value{GDBN} thread ids. At the
31041end of the list it also prints the total number of such threads.
31042
31043This command is retained for historical reasons, the
31044@code{-thread-info} command should be used instead.
31045
31046@subsubheading @value{GDBN} Command
31047
31048Part of @samp{info threads} supplies the same information.
31049
31050@subsubheading Example
31051
31052@smallexample
31053(gdb)
31054-thread-list-ids
31055^done,thread-ids=@{thread-id="3",thread-id="2",thread-id="1"@},
31056current-thread-id="1",number-of-threads="3"
31057(gdb)
31058@end smallexample
31059
31060
31061@subheading The @code{-thread-select} Command
31062@findex -thread-select
31063
31064@subsubheading Synopsis
31065
31066@smallexample
31067 -thread-select @var{threadnum}
31068@end smallexample
31069
31070Make @var{threadnum} the current thread. It prints the number of the new
31071current thread, and the topmost frame for that thread.
31072
31073This command is deprecated in favor of explicitly using the
31074@samp{--thread} option to each command.
31075
31076@subsubheading @value{GDBN} Command
31077
31078The corresponding @value{GDBN} command is @samp{thread}.
31079
31080@subsubheading Example
31081
31082@smallexample
31083(gdb)
31084-exec-next
31085^running
31086(gdb)
31087*stopped,reason="end-stepping-range",thread-id="2",line="187",
31088file="../../../devo/gdb/testsuite/gdb.threads/linux-dp.c"
31089(gdb)
31090-thread-list-ids
31091^done,
31092thread-ids=@{thread-id="3",thread-id="2",thread-id="1"@},
31093number-of-threads="3"
31094(gdb)
31095-thread-select 3
31096^done,new-thread-id="3",
31097frame=@{level="0",func="vprintf",
31098args=[@{name="format",value="0x8048e9c \"%*s%c %d %c\\n\""@},
31099@{name="arg",value="0x2"@}],file="vprintf.c",line="31"@}
31100(gdb)
31101@end smallexample
31102
31103@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
31104@node GDB/MI Ada Tasking Commands
31105@section @sc{gdb/mi} Ada Tasking Commands
31106
31107@subheading The @code{-ada-task-info} Command
31108@findex -ada-task-info
31109
31110@subsubheading Synopsis
31111
31112@smallexample
31113 -ada-task-info [ @var{task-id} ]
31114@end smallexample
31115
31116Reports information about either a specific Ada task, if the
31117@var{task-id} parameter is present, or about all Ada tasks.
31118
31119@subsubheading @value{GDBN} Command
31120
31121The @samp{info tasks} command prints the same information
31122about all Ada tasks (@pxref{Ada Tasks}).
31123
31124@subsubheading Result
31125
31126The result is a table of Ada tasks. The following columns are
31127defined for each Ada task:
31128
31129@table @samp
31130@item current
31131This field exists only for the current thread. It has the value @samp{*}.
31132
31133@item id
31134The identifier that @value{GDBN} uses to refer to the Ada task.
31135
31136@item task-id
31137The identifier that the target uses to refer to the Ada task.
31138
31139@item thread-id
31140The identifier of the thread corresponding to the Ada task.
31141
31142This field should always exist, as Ada tasks are always implemented
31143on top of a thread. But if @value{GDBN} cannot find this corresponding
31144thread for any reason, the field is omitted.
31145
31146@item parent-id
31147This field exists only when the task was created by another task.
31148In this case, it provides the ID of the parent task.
31149
31150@item priority
31151The base priority of the task.
31152
31153@item state
31154The current state of the task. For a detailed description of the
31155possible states, see @ref{Ada Tasks}.
31156
31157@item name
31158The name of the task.
31159
31160@end table
31161
31162@subsubheading Example
31163
31164@smallexample
31165-ada-task-info
31166^done,tasks=@{nr_rows="3",nr_cols="8",
31167hdr=[@{width="1",alignment="-1",col_name="current",colhdr=""@},
31168@{width="3",alignment="1",col_name="id",colhdr="ID"@},
31169@{width="9",alignment="1",col_name="task-id",colhdr="TID"@},
31170@{width="4",alignment="1",col_name="thread-id",colhdr=""@},
31171@{width="4",alignment="1",col_name="parent-id",colhdr="P-ID"@},
31172@{width="3",alignment="1",col_name="priority",colhdr="Pri"@},
31173@{width="22",alignment="-1",col_name="state",colhdr="State"@},
31174@{width="1",alignment="2",col_name="name",colhdr="Name"@}],
31175body=[@{current="*",id="1",task-id=" 644010",thread-id="1",priority="48",
31176state="Child Termination Wait",name="main_task"@}]@}
31177(gdb)
31178@end smallexample
31179
31180@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
31181@node GDB/MI Program Execution
31182@section @sc{gdb/mi} Program Execution
31183
31184These are the asynchronous commands which generate the out-of-band
31185record @samp{*stopped}. Currently @value{GDBN} only really executes
31186asynchronously with remote targets and this interaction is mimicked in
31187other cases.
31188
31189@subheading The @code{-exec-continue} Command
31190@findex -exec-continue
31191
31192@subsubheading Synopsis
31193
31194@smallexample
31195 -exec-continue [--reverse] [--all|--thread-group N]
31196@end smallexample
31197
31198Resumes the execution of the inferior program, which will continue
31199to execute until it reaches a debugger stop event. If the
31200@samp{--reverse} option is specified, execution resumes in reverse until
31201it reaches a stop event. Stop events may include
31202@itemize @bullet
31203@item
31204breakpoints or watchpoints
31205@item
31206signals or exceptions
31207@item
31208the end of the process (or its beginning under @samp{--reverse})
31209@item
31210the end or beginning of a replay log if one is being used.
31211@end itemize
31212In all-stop mode (@pxref{All-Stop
31213Mode}), may resume only one thread, or all threads, depending on the
31214value of the @samp{scheduler-locking} variable. If @samp{--all} is
31215specified, all threads (in all inferiors) will be resumed. The @samp{--all} option is
31216ignored in all-stop mode. If the @samp{--thread-group} options is
31217specified, then all threads in that thread group are resumed.
31218
31219@subsubheading @value{GDBN} Command
31220
31221The corresponding @value{GDBN} corresponding is @samp{continue}.
31222
31223@subsubheading Example
31224
31225@smallexample
31226-exec-continue
31227^running
31228(gdb)
31229@@Hello world
31230*stopped,reason="breakpoint-hit",disp="keep",bkptno="2",frame=@{
31231func="foo",args=[],file="hello.c",fullname="/home/foo/bar/hello.c",
31232line="13"@}
31233(gdb)
31234@end smallexample
31235
31236
31237@subheading The @code{-exec-finish} Command
31238@findex -exec-finish
31239
31240@subsubheading Synopsis
31241
31242@smallexample
31243 -exec-finish [--reverse]
31244@end smallexample
31245
31246Resumes the execution of the inferior program until the current
31247function is exited. Displays the results returned by the function.
31248If the @samp{--reverse} option is specified, resumes the reverse
31249execution of the inferior program until the point where current
31250function was called.
31251
31252@subsubheading @value{GDBN} Command
31253
31254The corresponding @value{GDBN} command is @samp{finish}.
31255
31256@subsubheading Example
31257
31258Function returning @code{void}.
31259
31260@smallexample
31261-exec-finish
31262^running
31263(gdb)
31264@@hello from foo
31265*stopped,reason="function-finished",frame=@{func="main",args=[],
31266file="hello.c",fullname="/home/foo/bar/hello.c",line="7"@}
31267(gdb)
31268@end smallexample
31269
31270Function returning other than @code{void}. The name of the internal
31271@value{GDBN} variable storing the result is printed, together with the
31272value itself.
31273
31274@smallexample
31275-exec-finish
31276^running
31277(gdb)
31278*stopped,reason="function-finished",frame=@{addr="0x000107b0",func="foo",
31279args=[@{name="a",value="1"],@{name="b",value="9"@}@},
31280file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31281gdb-result-var="$1",return-value="0"
31282(gdb)
31283@end smallexample
31284
31285
31286@subheading The @code{-exec-interrupt} Command
31287@findex -exec-interrupt
31288
31289@subsubheading Synopsis
31290
31291@smallexample
31292 -exec-interrupt [--all|--thread-group N]
31293@end smallexample
31294
31295Interrupts the background execution of the target. Note how the token
31296associated with the stop message is the one for the execution command
31297that has been interrupted. The token for the interrupt itself only
31298appears in the @samp{^done} output. If the user is trying to
31299interrupt a non-running program, an error message will be printed.
31300
31301Note that when asynchronous execution is enabled, this command is
31302asynchronous just like other execution commands. That is, first the
31303@samp{^done} response will be printed, and the target stop will be
31304reported after that using the @samp{*stopped} notification.
31305
31306In non-stop mode, only the context thread is interrupted by default.
31307All threads (in all inferiors) will be interrupted if the
31308@samp{--all} option is specified. If the @samp{--thread-group}
31309option is specified, all threads in that group will be interrupted.
31310
31311@subsubheading @value{GDBN} Command
31312
31313The corresponding @value{GDBN} command is @samp{interrupt}.
31314
31315@subsubheading Example
31316
31317@smallexample
31318(gdb)
31319111-exec-continue
31320111^running
31321
31322(gdb)
31323222-exec-interrupt
31324222^done
31325(gdb)
31326111*stopped,signal-name="SIGINT",signal-meaning="Interrupt",
31327frame=@{addr="0x00010140",func="foo",args=[],file="try.c",
31328fullname="/home/foo/bar/try.c",line="13"@}
31329(gdb)
31330
31331(gdb)
31332-exec-interrupt
31333^error,msg="mi_cmd_exec_interrupt: Inferior not executing."
31334(gdb)
31335@end smallexample
31336
31337@subheading The @code{-exec-jump} Command
31338@findex -exec-jump
31339
31340@subsubheading Synopsis
31341
31342@smallexample
31343 -exec-jump @var{location}
31344@end smallexample
31345
31346Resumes execution of the inferior program at the location specified by
31347parameter. @xref{Specify Location}, for a description of the
31348different forms of @var{location}.
31349
31350@subsubheading @value{GDBN} Command
31351
31352The corresponding @value{GDBN} command is @samp{jump}.
31353
31354@subsubheading Example
31355
31356@smallexample
31357-exec-jump foo.c:10
31358*running,thread-id="all"
31359^running
31360@end smallexample
31361
31362
31363@subheading The @code{-exec-next} Command
31364@findex -exec-next
31365
31366@subsubheading Synopsis
31367
31368@smallexample
31369 -exec-next [--reverse]
31370@end smallexample
31371
31372Resumes execution of the inferior program, stopping when the beginning
31373of the next source line is reached.
31374
31375If the @samp{--reverse} option is specified, resumes reverse execution
31376of the inferior program, stopping at the beginning of the previous
31377source line. If you issue this command on the first line of a
31378function, it will take you back to the caller of that function, to the
31379source line where the function was called.
31380
31381
31382@subsubheading @value{GDBN} Command
31383
31384The corresponding @value{GDBN} command is @samp{next}.
31385
31386@subsubheading Example
31387
31388@smallexample
31389-exec-next
31390^running
31391(gdb)
31392*stopped,reason="end-stepping-range",line="8",file="hello.c"
31393(gdb)
31394@end smallexample
31395
31396
31397@subheading The @code{-exec-next-instruction} Command
31398@findex -exec-next-instruction
31399
31400@subsubheading Synopsis
31401
31402@smallexample
31403 -exec-next-instruction [--reverse]
31404@end smallexample
31405
31406Executes one machine instruction. If the instruction is a function
31407call, continues until the function returns. If the program stops at an
31408instruction in the middle of a source line, the address will be
31409printed as well.
31410
31411If the @samp{--reverse} option is specified, resumes reverse execution
31412of the inferior program, stopping at the previous instruction. If the
31413previously executed instruction was a return from another function,
31414it will continue to execute in reverse until the call to that function
31415(from the current stack frame) is reached.
31416
31417@subsubheading @value{GDBN} Command
31418
31419The corresponding @value{GDBN} command is @samp{nexti}.
31420
31421@subsubheading Example
31422
31423@smallexample
31424(gdb)
31425-exec-next-instruction
31426^running
31427
31428(gdb)
31429*stopped,reason="end-stepping-range",
31430addr="0x000100d4",line="5",file="hello.c"
31431(gdb)
31432@end smallexample
31433
31434
31435@subheading The @code{-exec-return} Command
31436@findex -exec-return
31437
31438@subsubheading Synopsis
31439
31440@smallexample
31441 -exec-return
31442@end smallexample
31443
31444Makes current function return immediately. Doesn't execute the inferior.
31445Displays the new current frame.
31446
31447@subsubheading @value{GDBN} Command
31448
31449The corresponding @value{GDBN} command is @samp{return}.
31450
31451@subsubheading Example
31452
31453@smallexample
31454(gdb)
31455200-break-insert callee4
31456200^done,bkpt=@{number="1",addr="0x00010734",
31457file="../../../devo/gdb/testsuite/gdb.mi/basics.c",line="8"@}
31458(gdb)
31459000-exec-run
31460000^running
31461(gdb)
31462000*stopped,reason="breakpoint-hit",disp="keep",bkptno="1",
31463frame=@{func="callee4",args=[],
31464file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31465fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="8"@}
31466(gdb)
31467205-break-delete
31468205^done
31469(gdb)
31470111-exec-return
31471111^done,frame=@{level="0",func="callee3",
31472args=[@{name="strarg",
31473value="0x11940 \"A string argument.\""@}],
31474file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31475fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="18"@}
31476(gdb)
31477@end smallexample
31478
31479
31480@subheading The @code{-exec-run} Command
31481@findex -exec-run
31482
31483@subsubheading Synopsis
31484
31485@smallexample
31486 -exec-run [ --all | --thread-group N ] [ --start ]
31487@end smallexample
31488
31489Starts execution of the inferior from the beginning. The inferior
31490executes until either a breakpoint is encountered or the program
31491exits. In the latter case the output will include an exit code, if
31492the program has exited exceptionally.
31493
31494When neither the @samp{--all} nor the @samp{--thread-group} option
31495is specified, the current inferior is started. If the
31496@samp{--thread-group} option is specified, it should refer to a thread
31497group of type @samp{process}, and that thread group will be started.
31498If the @samp{--all} option is specified, then all inferiors will be started.
31499
31500Using the @samp{--start} option instructs the debugger to stop
31501the execution at the start of the inferior's main subprogram,
31502following the same behavior as the @code{start} command
31503(@pxref{Starting}).
31504
31505@subsubheading @value{GDBN} Command
31506
31507The corresponding @value{GDBN} command is @samp{run}.
31508
31509@subsubheading Examples
31510
31511@smallexample
31512(gdb)
31513-break-insert main
31514^done,bkpt=@{number="1",addr="0x0001072c",file="recursive2.c",line="4"@}
31515(gdb)
31516-exec-run
31517^running
31518(gdb)
31519*stopped,reason="breakpoint-hit",disp="keep",bkptno="1",
31520frame=@{func="main",args=[],file="recursive2.c",
31521fullname="/home/foo/bar/recursive2.c",line="4"@}
31522(gdb)
31523@end smallexample
31524
31525@noindent
31526Program exited normally:
31527
31528@smallexample
31529(gdb)
31530-exec-run
31531^running
31532(gdb)
31533x = 55
31534*stopped,reason="exited-normally"
31535(gdb)
31536@end smallexample
31537
31538@noindent
31539Program exited exceptionally:
31540
31541@smallexample
31542(gdb)
31543-exec-run
31544^running
31545(gdb)
31546x = 55
31547*stopped,reason="exited",exit-code="01"
31548(gdb)
31549@end smallexample
31550
31551Another way the program can terminate is if it receives a signal such as
31552@code{SIGINT}. In this case, @sc{gdb/mi} displays this:
31553
31554@smallexample
31555(gdb)
31556*stopped,reason="exited-signalled",signal-name="SIGINT",
31557signal-meaning="Interrupt"
31558@end smallexample
31559
31560
31561@c @subheading -exec-signal
31562
31563
31564@subheading The @code{-exec-step} Command
31565@findex -exec-step
31566
31567@subsubheading Synopsis
31568
31569@smallexample
31570 -exec-step [--reverse]
31571@end smallexample
31572
31573Resumes execution of the inferior program, stopping when the beginning
31574of the next source line is reached, if the next source line is not a
31575function call. If it is, stop at the first instruction of the called
31576function. If the @samp{--reverse} option is specified, resumes reverse
31577execution of the inferior program, stopping at the beginning of the
31578previously executed source line.
31579
31580@subsubheading @value{GDBN} Command
31581
31582The corresponding @value{GDBN} command is @samp{step}.
31583
31584@subsubheading Example
31585
31586Stepping into a function:
31587
31588@smallexample
31589-exec-step
31590^running
31591(gdb)
31592*stopped,reason="end-stepping-range",
31593frame=@{func="foo",args=[@{name="a",value="10"@},
31594@{name="b",value="0"@}],file="recursive2.c",
31595fullname="/home/foo/bar/recursive2.c",line="11"@}
31596(gdb)
31597@end smallexample
31598
31599Regular stepping:
31600
31601@smallexample
31602-exec-step
31603^running
31604(gdb)
31605*stopped,reason="end-stepping-range",line="14",file="recursive2.c"
31606(gdb)
31607@end smallexample
31608
31609
31610@subheading The @code{-exec-step-instruction} Command
31611@findex -exec-step-instruction
31612
31613@subsubheading Synopsis
31614
31615@smallexample
31616 -exec-step-instruction [--reverse]
31617@end smallexample
31618
31619Resumes the inferior which executes one machine instruction. If the
31620@samp{--reverse} option is specified, resumes reverse execution of the
31621inferior program, stopping at the previously executed instruction.
31622The output, once @value{GDBN} has stopped, will vary depending on
31623whether we have stopped in the middle of a source line or not. In the
31624former case, the address at which the program stopped will be printed
31625as well.
31626
31627@subsubheading @value{GDBN} Command
31628
31629The corresponding @value{GDBN} command is @samp{stepi}.
31630
31631@subsubheading Example
31632
31633@smallexample
31634(gdb)
31635-exec-step-instruction
31636^running
31637
31638(gdb)
31639*stopped,reason="end-stepping-range",
31640frame=@{func="foo",args=[],file="try.c",
31641fullname="/home/foo/bar/try.c",line="10"@}
31642(gdb)
31643-exec-step-instruction
31644^running
31645
31646(gdb)
31647*stopped,reason="end-stepping-range",
31648frame=@{addr="0x000100f4",func="foo",args=[],file="try.c",
31649fullname="/home/foo/bar/try.c",line="10"@}
31650(gdb)
31651@end smallexample
31652
31653
31654@subheading The @code{-exec-until} Command
31655@findex -exec-until
31656
31657@subsubheading Synopsis
31658
31659@smallexample
31660 -exec-until [ @var{location} ]
31661@end smallexample
31662
31663Executes the inferior until the @var{location} specified in the
31664argument is reached. If there is no argument, the inferior executes
31665until a source line greater than the current one is reached. The
31666reason for stopping in this case will be @samp{location-reached}.
31667
31668@subsubheading @value{GDBN} Command
31669
31670The corresponding @value{GDBN} command is @samp{until}.
31671
31672@subsubheading Example
31673
31674@smallexample
31675(gdb)
31676-exec-until recursive2.c:6
31677^running
31678(gdb)
31679x = 55
31680*stopped,reason="location-reached",frame=@{func="main",args=[],
31681file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="6"@}
31682(gdb)
31683@end smallexample
31684
31685@ignore
31686@subheading -file-clear
31687Is this going away????
31688@end ignore
31689
31690@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
31691@node GDB/MI Stack Manipulation
31692@section @sc{gdb/mi} Stack Manipulation Commands
31693
31694@subheading The @code{-enable-frame-filters} Command
31695@findex -enable-frame-filters
31696
31697@smallexample
31698-enable-frame-filters
31699@end smallexample
31700
31701@value{GDBN} allows Python-based frame filters to affect the output of
31702the MI commands relating to stack traces. As there is no way to
31703implement this in a fully backward-compatible way, a front end must
31704request that this functionality be enabled.
31705
31706Once enabled, this feature cannot be disabled.
31707
31708Note that if Python support has not been compiled into @value{GDBN},
31709this command will still succeed (and do nothing).
31710
31711@subheading The @code{-stack-info-frame} Command
31712@findex -stack-info-frame
31713
31714@subsubheading Synopsis
31715
31716@smallexample
31717 -stack-info-frame
31718@end smallexample
31719
31720Get info on the selected frame.
31721
31722@subsubheading @value{GDBN} Command
31723
31724The corresponding @value{GDBN} command is @samp{info frame} or @samp{frame}
31725(without arguments).
31726
31727@subsubheading Example
31728
31729@smallexample
31730(gdb)
31731-stack-info-frame
31732^done,frame=@{level="1",addr="0x0001076c",func="callee3",
31733file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31734fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="17"@}
31735(gdb)
31736@end smallexample
31737
31738@subheading The @code{-stack-info-depth} Command
31739@findex -stack-info-depth
31740
31741@subsubheading Synopsis
31742
31743@smallexample
31744 -stack-info-depth [ @var{max-depth} ]
31745@end smallexample
31746
31747Return the depth of the stack. If the integer argument @var{max-depth}
31748is specified, do not count beyond @var{max-depth} frames.
31749
31750@subsubheading @value{GDBN} Command
31751
31752There's no equivalent @value{GDBN} command.
31753
31754@subsubheading Example
31755
31756For a stack with frame levels 0 through 11:
31757
31758@smallexample
31759(gdb)
31760-stack-info-depth
31761^done,depth="12"
31762(gdb)
31763-stack-info-depth 4
31764^done,depth="4"
31765(gdb)
31766-stack-info-depth 12
31767^done,depth="12"
31768(gdb)
31769-stack-info-depth 11
31770^done,depth="11"
31771(gdb)
31772-stack-info-depth 13
31773^done,depth="12"
31774(gdb)
31775@end smallexample
31776
31777@anchor{-stack-list-arguments}
31778@subheading The @code{-stack-list-arguments} Command
31779@findex -stack-list-arguments
31780
31781@subsubheading Synopsis
31782
31783@smallexample
31784 -stack-list-arguments [ --no-frame-filters ] [ --skip-unavailable ] @var{print-values}
31785 [ @var{low-frame} @var{high-frame} ]
31786@end smallexample
31787
31788Display a list of the arguments for the frames between @var{low-frame}
31789and @var{high-frame} (inclusive). If @var{low-frame} and
31790@var{high-frame} are not provided, list the arguments for the whole
31791call stack. If the two arguments are equal, show the single frame
31792at the corresponding level. It is an error if @var{low-frame} is
31793larger than the actual number of frames. On the other hand,
31794@var{high-frame} may be larger than the actual number of frames, in
31795which case only existing frames will be returned.
31796
31797If @var{print-values} is 0 or @code{--no-values}, print only the names of
31798the variables; if it is 1 or @code{--all-values}, print also their
31799values; and if it is 2 or @code{--simple-values}, print the name,
31800type and value for simple data types, and the name and type for arrays,
31801structures and unions. If the option @code{--no-frame-filters} is
31802supplied, then Python frame filters will not be executed.
31803
31804If the @code{--skip-unavailable} option is specified, arguments that
31805are not available are not listed. Partially available arguments
31806are still displayed, however.
31807
31808Use of this command to obtain arguments in a single frame is
31809deprecated in favor of the @samp{-stack-list-variables} command.
31810
31811@subsubheading @value{GDBN} Command
31812
31813@value{GDBN} does not have an equivalent command. @code{gdbtk} has a
31814@samp{gdb_get_args} command which partially overlaps with the
31815functionality of @samp{-stack-list-arguments}.
31816
31817@subsubheading Example
31818
31819@smallexample
31820(gdb)
31821-stack-list-frames
31822^done,
31823stack=[
31824frame=@{level="0",addr="0x00010734",func="callee4",
31825file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31826fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="8"@},
31827frame=@{level="1",addr="0x0001076c",func="callee3",
31828file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31829fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="17"@},
31830frame=@{level="2",addr="0x0001078c",func="callee2",
31831file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31832fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="22"@},
31833frame=@{level="3",addr="0x000107b4",func="callee1",
31834file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31835fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="27"@},
31836frame=@{level="4",addr="0x000107e0",func="main",
31837file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31838fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="32"@}]
31839(gdb)
31840-stack-list-arguments 0
31841^done,
31842stack-args=[
31843frame=@{level="0",args=[]@},
31844frame=@{level="1",args=[name="strarg"]@},
31845frame=@{level="2",args=[name="intarg",name="strarg"]@},
31846frame=@{level="3",args=[name="intarg",name="strarg",name="fltarg"]@},
31847frame=@{level="4",args=[]@}]
31848(gdb)
31849-stack-list-arguments 1
31850^done,
31851stack-args=[
31852frame=@{level="0",args=[]@},
31853frame=@{level="1",
31854 args=[@{name="strarg",value="0x11940 \"A string argument.\""@}]@},
31855frame=@{level="2",args=[
31856@{name="intarg",value="2"@},
31857@{name="strarg",value="0x11940 \"A string argument.\""@}]@},
31858@{frame=@{level="3",args=[
31859@{name="intarg",value="2"@},
31860@{name="strarg",value="0x11940 \"A string argument.\""@},
31861@{name="fltarg",value="3.5"@}]@},
31862frame=@{level="4",args=[]@}]
31863(gdb)
31864-stack-list-arguments 0 2 2
31865^done,stack-args=[frame=@{level="2",args=[name="intarg",name="strarg"]@}]
31866(gdb)
31867-stack-list-arguments 1 2 2
31868^done,stack-args=[frame=@{level="2",
31869args=[@{name="intarg",value="2"@},
31870@{name="strarg",value="0x11940 \"A string argument.\""@}]@}]
31871(gdb)
31872@end smallexample
31873
31874@c @subheading -stack-list-exception-handlers
31875
31876
31877@anchor{-stack-list-frames}
31878@subheading The @code{-stack-list-frames} Command
31879@findex -stack-list-frames
31880
31881@subsubheading Synopsis
31882
31883@smallexample
31884 -stack-list-frames [ --no-frame-filters @var{low-frame} @var{high-frame} ]
31885@end smallexample
31886
31887List the frames currently on the stack. For each frame it displays the
31888following info:
31889
31890@table @samp
31891@item @var{level}
31892The frame number, 0 being the topmost frame, i.e., the innermost function.
31893@item @var{addr}
31894The @code{$pc} value for that frame.
31895@item @var{func}
31896Function name.
31897@item @var{file}
31898File name of the source file where the function lives.
31899@item @var{fullname}
31900The full file name of the source file where the function lives.
31901@item @var{line}
31902Line number corresponding to the @code{$pc}.
31903@item @var{from}
31904The shared library where this function is defined. This is only given
31905if the frame's function is not known.
31906@end table
31907
31908If invoked without arguments, this command prints a backtrace for the
31909whole stack. If given two integer arguments, it shows the frames whose
31910levels are between the two arguments (inclusive). If the two arguments
31911are equal, it shows the single frame at the corresponding level. It is
31912an error if @var{low-frame} is larger than the actual number of
31913frames. On the other hand, @var{high-frame} may be larger than the
31914actual number of frames, in which case only existing frames will be
31915returned. If the option @code{--no-frame-filters} is supplied, then
31916Python frame filters will not be executed.
31917
31918@subsubheading @value{GDBN} Command
31919
31920The corresponding @value{GDBN} commands are @samp{backtrace} and @samp{where}.
31921
31922@subsubheading Example
31923
31924Full stack backtrace:
31925
31926@smallexample
31927(gdb)
31928-stack-list-frames
31929^done,stack=
31930[frame=@{level="0",addr="0x0001076c",func="foo",
31931 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="11"@},
31932frame=@{level="1",addr="0x000107a4",func="foo",
31933 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31934frame=@{level="2",addr="0x000107a4",func="foo",
31935 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31936frame=@{level="3",addr="0x000107a4",func="foo",
31937 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31938frame=@{level="4",addr="0x000107a4",func="foo",
31939 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31940frame=@{level="5",addr="0x000107a4",func="foo",
31941 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31942frame=@{level="6",addr="0x000107a4",func="foo",
31943 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31944frame=@{level="7",addr="0x000107a4",func="foo",
31945 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31946frame=@{level="8",addr="0x000107a4",func="foo",
31947 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31948frame=@{level="9",addr="0x000107a4",func="foo",
31949 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31950frame=@{level="10",addr="0x000107a4",func="foo",
31951 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31952frame=@{level="11",addr="0x00010738",func="main",
31953 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="4"@}]
31954(gdb)
31955@end smallexample
31956
31957Show frames between @var{low_frame} and @var{high_frame}:
31958
31959@smallexample
31960(gdb)
31961-stack-list-frames 3 5
31962^done,stack=
31963[frame=@{level="3",addr="0x000107a4",func="foo",
31964 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31965frame=@{level="4",addr="0x000107a4",func="foo",
31966 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@},
31967frame=@{level="5",addr="0x000107a4",func="foo",
31968 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@}]
31969(gdb)
31970@end smallexample
31971
31972Show a single frame:
31973
31974@smallexample
31975(gdb)
31976-stack-list-frames 3 3
31977^done,stack=
31978[frame=@{level="3",addr="0x000107a4",func="foo",
31979 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14"@}]
31980(gdb)
31981@end smallexample
31982
31983
31984@subheading The @code{-stack-list-locals} Command
31985@findex -stack-list-locals
31986@anchor{-stack-list-locals}
31987
31988@subsubheading Synopsis
31989
31990@smallexample
31991 -stack-list-locals [ --no-frame-filters ] [ --skip-unavailable ] @var{print-values}
31992@end smallexample
31993
31994Display the local variable names for the selected frame. If
31995@var{print-values} is 0 or @code{--no-values}, print only the names of
31996the variables; if it is 1 or @code{--all-values}, print also their
31997values; and if it is 2 or @code{--simple-values}, print the name,
31998type and value for simple data types, and the name and type for arrays,
31999structures and unions. In this last case, a frontend can immediately
32000display the value of simple data types and create variable objects for
32001other data types when the user wishes to explore their values in
32002more detail. If the option @code{--no-frame-filters} is supplied, then
32003Python frame filters will not be executed.
32004
32005If the @code{--skip-unavailable} option is specified, local variables
32006that are not available are not listed. Partially available local
32007variables are still displayed, however.
32008
32009This command is deprecated in favor of the
32010@samp{-stack-list-variables} command.
32011
32012@subsubheading @value{GDBN} Command
32013
32014@samp{info locals} in @value{GDBN}, @samp{gdb_get_locals} in @code{gdbtk}.
32015
32016@subsubheading Example
32017
32018@smallexample
32019(gdb)
32020-stack-list-locals 0
32021^done,locals=[name="A",name="B",name="C"]
32022(gdb)
32023-stack-list-locals --all-values
32024^done,locals=[@{name="A",value="1"@},@{name="B",value="2"@},
32025 @{name="C",value="@{1, 2, 3@}"@}]
32026-stack-list-locals --simple-values
32027^done,locals=[@{name="A",type="int",value="1"@},
32028 @{name="B",type="int",value="2"@},@{name="C",type="int [3]"@}]
32029(gdb)
32030@end smallexample
32031
32032@anchor{-stack-list-variables}
32033@subheading The @code{-stack-list-variables} Command
32034@findex -stack-list-variables
32035
32036@subsubheading Synopsis
32037
32038@smallexample
32039 -stack-list-variables [ --no-frame-filters ] [ --skip-unavailable ] @var{print-values}
32040@end smallexample
32041
32042Display the names of local variables and function arguments for the selected frame. If
32043@var{print-values} is 0 or @code{--no-values}, print only the names of
32044the variables; if it is 1 or @code{--all-values}, print also their
32045values; and if it is 2 or @code{--simple-values}, print the name,
32046type and value for simple data types, and the name and type for arrays,
32047structures and unions. If the option @code{--no-frame-filters} is
32048supplied, then Python frame filters will not be executed.
32049
32050If the @code{--skip-unavailable} option is specified, local variables
32051and arguments that are not available are not listed. Partially
32052available arguments and local variables are still displayed, however.
32053
32054@subsubheading Example
32055
32056@smallexample
32057(gdb)
32058-stack-list-variables --thread 1 --frame 0 --all-values
32059^done,variables=[@{name="x",value="11"@},@{name="s",value="@{a = 1, b = 2@}"@}]
32060(gdb)
32061@end smallexample
32062
32063
32064@subheading The @code{-stack-select-frame} Command
32065@findex -stack-select-frame
32066
32067@subsubheading Synopsis
32068
32069@smallexample
32070 -stack-select-frame @var{framenum}
32071@end smallexample
32072
32073Change the selected frame. Select a different frame @var{framenum} on
32074the stack.
32075
32076This command in deprecated in favor of passing the @samp{--frame}
32077option to every command.
32078
32079@subsubheading @value{GDBN} Command
32080
32081The corresponding @value{GDBN} commands are @samp{frame}, @samp{up},
32082@samp{down}, @samp{select-frame}, @samp{up-silent}, and @samp{down-silent}.
32083
32084@subsubheading Example
32085
32086@smallexample
32087(gdb)
32088-stack-select-frame 2
32089^done
32090(gdb)
32091@end smallexample
32092
32093@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32094@node GDB/MI Variable Objects
32095@section @sc{gdb/mi} Variable Objects
32096
32097@ignore
32098
32099@subheading Motivation for Variable Objects in @sc{gdb/mi}
32100
32101For the implementation of a variable debugger window (locals, watched
32102expressions, etc.), we are proposing the adaptation of the existing code
32103used by @code{Insight}.
32104
32105The two main reasons for that are:
32106
32107@enumerate 1
32108@item
32109It has been proven in practice (it is already on its second generation).
32110
32111@item
32112It will shorten development time (needless to say how important it is
32113now).
32114@end enumerate
32115
32116The original interface was designed to be used by Tcl code, so it was
32117slightly changed so it could be used through @sc{gdb/mi}. This section
32118describes the @sc{gdb/mi} operations that will be available and gives some
32119hints about their use.
32120
32121@emph{Note}: In addition to the set of operations described here, we
32122expect the @sc{gui} implementation of a variable window to require, at
32123least, the following operations:
32124
32125@itemize @bullet
32126@item @code{-gdb-show} @code{output-radix}
32127@item @code{-stack-list-arguments}
32128@item @code{-stack-list-locals}
32129@item @code{-stack-select-frame}
32130@end itemize
32131
32132@end ignore
32133
32134@subheading Introduction to Variable Objects
32135
32136@cindex variable objects in @sc{gdb/mi}
32137
32138Variable objects are "object-oriented" MI interface for examining and
32139changing values of expressions. Unlike some other MI interfaces that
32140work with expressions, variable objects are specifically designed for
32141simple and efficient presentation in the frontend. A variable object
32142is identified by string name. When a variable object is created, the
32143frontend specifies the expression for that variable object. The
32144expression can be a simple variable, or it can be an arbitrary complex
32145expression, and can even involve CPU registers. After creating a
32146variable object, the frontend can invoke other variable object
32147operations---for example to obtain or change the value of a variable
32148object, or to change display format.
32149
32150Variable objects have hierarchical tree structure. Any variable object
32151that corresponds to a composite type, such as structure in C, has
32152a number of child variable objects, for example corresponding to each
32153element of a structure. A child variable object can itself have
32154children, recursively. Recursion ends when we reach
32155leaf variable objects, which always have built-in types. Child variable
32156objects are created only by explicit request, so if a frontend
32157is not interested in the children of a particular variable object, no
32158child will be created.
32159
32160For a leaf variable object it is possible to obtain its value as a
32161string, or set the value from a string. String value can be also
32162obtained for a non-leaf variable object, but it's generally a string
32163that only indicates the type of the object, and does not list its
32164contents. Assignment to a non-leaf variable object is not allowed.
32165
32166A frontend does not need to read the values of all variable objects each time
32167the program stops. Instead, MI provides an update command that lists all
32168variable objects whose values has changed since the last update
32169operation. This considerably reduces the amount of data that must
32170be transferred to the frontend. As noted above, children variable
32171objects are created on demand, and only leaf variable objects have a
32172real value. As result, gdb will read target memory only for leaf
32173variables that frontend has created.
32174
32175The automatic update is not always desirable. For example, a frontend
32176might want to keep a value of some expression for future reference,
32177and never update it. For another example, fetching memory is
32178relatively slow for embedded targets, so a frontend might want
32179to disable automatic update for the variables that are either not
32180visible on the screen, or ``closed''. This is possible using so
32181called ``frozen variable objects''. Such variable objects are never
32182implicitly updated.
32183
32184Variable objects can be either @dfn{fixed} or @dfn{floating}. For the
32185fixed variable object, the expression is parsed when the variable
32186object is created, including associating identifiers to specific
32187variables. The meaning of expression never changes. For a floating
32188variable object the values of variables whose names appear in the
32189expressions are re-evaluated every time in the context of the current
32190frame. Consider this example:
32191
32192@smallexample
32193void do_work(...)
32194@{
32195 struct work_state state;
32196
32197 if (...)
32198 do_work(...);
32199@}
32200@end smallexample
32201
32202If a fixed variable object for the @code{state} variable is created in
32203this function, and we enter the recursive call, the variable
32204object will report the value of @code{state} in the top-level
32205@code{do_work} invocation. On the other hand, a floating variable
32206object will report the value of @code{state} in the current frame.
32207
32208If an expression specified when creating a fixed variable object
32209refers to a local variable, the variable object becomes bound to the
32210thread and frame in which the variable object is created. When such
32211variable object is updated, @value{GDBN} makes sure that the
32212thread/frame combination the variable object is bound to still exists,
32213and re-evaluates the variable object in context of that thread/frame.
32214
32215The following is the complete set of @sc{gdb/mi} operations defined to
32216access this functionality:
32217
32218@multitable @columnfractions .4 .6
32219@item @strong{Operation}
32220@tab @strong{Description}
32221
32222@item @code{-enable-pretty-printing}
32223@tab enable Python-based pretty-printing
32224@item @code{-var-create}
32225@tab create a variable object
32226@item @code{-var-delete}
32227@tab delete the variable object and/or its children
32228@item @code{-var-set-format}
32229@tab set the display format of this variable
32230@item @code{-var-show-format}
32231@tab show the display format of this variable
32232@item @code{-var-info-num-children}
32233@tab tells how many children this object has
32234@item @code{-var-list-children}
32235@tab return a list of the object's children
32236@item @code{-var-info-type}
32237@tab show the type of this variable object
32238@item @code{-var-info-expression}
32239@tab print parent-relative expression that this variable object represents
32240@item @code{-var-info-path-expression}
32241@tab print full expression that this variable object represents
32242@item @code{-var-show-attributes}
32243@tab is this variable editable? does it exist here?
32244@item @code{-var-evaluate-expression}
32245@tab get the value of this variable
32246@item @code{-var-assign}
32247@tab set the value of this variable
32248@item @code{-var-update}
32249@tab update the variable and its children
32250@item @code{-var-set-frozen}
32251@tab set frozeness attribute
32252@item @code{-var-set-update-range}
32253@tab set range of children to display on update
32254@end multitable
32255
32256In the next subsection we describe each operation in detail and suggest
32257how it can be used.
32258
32259@subheading Description And Use of Operations on Variable Objects
32260
32261@subheading The @code{-enable-pretty-printing} Command
32262@findex -enable-pretty-printing
32263
32264@smallexample
32265-enable-pretty-printing
32266@end smallexample
32267
32268@value{GDBN} allows Python-based visualizers to affect the output of the
32269MI variable object commands. However, because there was no way to
32270implement this in a fully backward-compatible way, a front end must
32271request that this functionality be enabled.
32272
32273Once enabled, this feature cannot be disabled.
32274
32275Note that if Python support has not been compiled into @value{GDBN},
32276this command will still succeed (and do nothing).
32277
32278This feature is currently (as of @value{GDBN} 7.0) experimental, and
32279may work differently in future versions of @value{GDBN}.
32280
32281@subheading The @code{-var-create} Command
32282@findex -var-create
32283
32284@subsubheading Synopsis
32285
32286@smallexample
32287 -var-create @{@var{name} | "-"@}
32288 @{@var{frame-addr} | "*" | "@@"@} @var{expression}
32289@end smallexample
32290
32291This operation creates a variable object, which allows the monitoring of
32292a variable, the result of an expression, a memory cell or a CPU
32293register.
32294
32295The @var{name} parameter is the string by which the object can be
32296referenced. It must be unique. If @samp{-} is specified, the varobj
32297system will generate a string ``varNNNNNN'' automatically. It will be
32298unique provided that one does not specify @var{name} of that format.
32299The command fails if a duplicate name is found.
32300
32301The frame under which the expression should be evaluated can be
32302specified by @var{frame-addr}. A @samp{*} indicates that the current
32303frame should be used. A @samp{@@} indicates that a floating variable
32304object must be created.
32305
32306@var{expression} is any expression valid on the current language set (must not
32307begin with a @samp{*}), or one of the following:
32308
32309@itemize @bullet
32310@item
32311@samp{*@var{addr}}, where @var{addr} is the address of a memory cell
32312
32313@item
32314@samp{*@var{addr}-@var{addr}} --- a memory address range (TBD)
32315
32316@item
32317@samp{$@var{regname}} --- a CPU register name
32318@end itemize
32319
32320@cindex dynamic varobj
32321A varobj's contents may be provided by a Python-based pretty-printer. In this
32322case the varobj is known as a @dfn{dynamic varobj}. Dynamic varobjs
32323have slightly different semantics in some cases. If the
32324@code{-enable-pretty-printing} command is not sent, then @value{GDBN}
32325will never create a dynamic varobj. This ensures backward
32326compatibility for existing clients.
32327
32328@subsubheading Result
32329
32330This operation returns attributes of the newly-created varobj. These
32331are:
32332
32333@table @samp
32334@item name
32335The name of the varobj.
32336
32337@item numchild
32338The number of children of the varobj. This number is not necessarily
32339reliable for a dynamic varobj. Instead, you must examine the
32340@samp{has_more} attribute.
32341
32342@item value
32343The varobj's scalar value. For a varobj whose type is some sort of
32344aggregate (e.g., a @code{struct}), or for a dynamic varobj, this value
32345will not be interesting.
32346
32347@item type
32348The varobj's type. This is a string representation of the type, as
32349would be printed by the @value{GDBN} CLI. If @samp{print object}
32350(@pxref{Print Settings, set print object}) is set to @code{on}, the
32351@emph{actual} (derived) type of the object is shown rather than the
32352@emph{declared} one.
32353
32354@item thread-id
32355If a variable object is bound to a specific thread, then this is the
32356thread's identifier.
32357
32358@item has_more
32359For a dynamic varobj, this indicates whether there appear to be any
32360children available. For a non-dynamic varobj, this will be 0.
32361
32362@item dynamic
32363This attribute will be present and have the value @samp{1} if the
32364varobj is a dynamic varobj. If the varobj is not a dynamic varobj,
32365then this attribute will not be present.
32366
32367@item displayhint
32368A dynamic varobj can supply a display hint to the front end. The
32369value comes directly from the Python pretty-printer object's
32370@code{display_hint} method. @xref{Pretty Printing API}.
32371@end table
32372
32373Typical output will look like this:
32374
32375@smallexample
32376 name="@var{name}",numchild="@var{N}",type="@var{type}",thread-id="@var{M}",
32377 has_more="@var{has_more}"
32378@end smallexample
32379
32380
32381@subheading The @code{-var-delete} Command
32382@findex -var-delete
32383
32384@subsubheading Synopsis
32385
32386@smallexample
32387 -var-delete [ -c ] @var{name}
32388@end smallexample
32389
32390Deletes a previously created variable object and all of its children.
32391With the @samp{-c} option, just deletes the children.
32392
32393Returns an error if the object @var{name} is not found.
32394
32395
32396@subheading The @code{-var-set-format} Command
32397@findex -var-set-format
32398
32399@subsubheading Synopsis
32400
32401@smallexample
32402 -var-set-format @var{name} @var{format-spec}
32403@end smallexample
32404
32405Sets the output format for the value of the object @var{name} to be
32406@var{format-spec}.
32407
32408@anchor{-var-set-format}
32409The syntax for the @var{format-spec} is as follows:
32410
32411@smallexample
32412 @var{format-spec} @expansion{}
32413 @{binary | decimal | hexadecimal | octal | natural@}
32414@end smallexample
32415
32416The natural format is the default format choosen automatically
32417based on the variable type (like decimal for an @code{int}, hex
32418for pointers, etc.).
32419
32420For a variable with children, the format is set only on the
32421variable itself, and the children are not affected.
32422
32423@subheading The @code{-var-show-format} Command
32424@findex -var-show-format
32425
32426@subsubheading Synopsis
32427
32428@smallexample
32429 -var-show-format @var{name}
32430@end smallexample
32431
32432Returns the format used to display the value of the object @var{name}.
32433
32434@smallexample
32435 @var{format} @expansion{}
32436 @var{format-spec}
32437@end smallexample
32438
32439
32440@subheading The @code{-var-info-num-children} Command
32441@findex -var-info-num-children
32442
32443@subsubheading Synopsis
32444
32445@smallexample
32446 -var-info-num-children @var{name}
32447@end smallexample
32448
32449Returns the number of children of a variable object @var{name}:
32450
32451@smallexample
32452 numchild=@var{n}
32453@end smallexample
32454
32455Note that this number is not completely reliable for a dynamic varobj.
32456It will return the current number of children, but more children may
32457be available.
32458
32459
32460@subheading The @code{-var-list-children} Command
32461@findex -var-list-children
32462
32463@subsubheading Synopsis
32464
32465@smallexample
32466 -var-list-children [@var{print-values}] @var{name} [@var{from} @var{to}]
32467@end smallexample
32468@anchor{-var-list-children}
32469
32470Return a list of the children of the specified variable object and
32471create variable objects for them, if they do not already exist. With
32472a single argument or if @var{print-values} has a value of 0 or
32473@code{--no-values}, print only the names of the variables; if
32474@var{print-values} is 1 or @code{--all-values}, also print their
32475values; and if it is 2 or @code{--simple-values} print the name and
32476value for simple data types and just the name for arrays, structures
32477and unions.
32478
32479@var{from} and @var{to}, if specified, indicate the range of children
32480to report. If @var{from} or @var{to} is less than zero, the range is
32481reset and all children will be reported. Otherwise, children starting
32482at @var{from} (zero-based) and up to and excluding @var{to} will be
32483reported.
32484
32485If a child range is requested, it will only affect the current call to
32486@code{-var-list-children}, but not future calls to @code{-var-update}.
32487For this, you must instead use @code{-var-set-update-range}. The
32488intent of this approach is to enable a front end to implement any
32489update approach it likes; for example, scrolling a view may cause the
32490front end to request more children with @code{-var-list-children}, and
32491then the front end could call @code{-var-set-update-range} with a
32492different range to ensure that future updates are restricted to just
32493the visible items.
32494
32495For each child the following results are returned:
32496
32497@table @var
32498
32499@item name
32500Name of the variable object created for this child.
32501
32502@item exp
32503The expression to be shown to the user by the front end to designate this child.
32504For example this may be the name of a structure member.
32505
32506For a dynamic varobj, this value cannot be used to form an
32507expression. There is no way to do this at all with a dynamic varobj.
32508
32509For C/C@t{++} structures there are several pseudo children returned to
32510designate access qualifiers. For these pseudo children @var{exp} is
32511@samp{public}, @samp{private}, or @samp{protected}. In this case the
32512type and value are not present.
32513
32514A dynamic varobj will not report the access qualifying
32515pseudo-children, regardless of the language. This information is not
32516available at all with a dynamic varobj.
32517
32518@item numchild
32519Number of children this child has. For a dynamic varobj, this will be
325200.
32521
32522@item type
32523The type of the child. If @samp{print object}
32524(@pxref{Print Settings, set print object}) is set to @code{on}, the
32525@emph{actual} (derived) type of the object is shown rather than the
32526@emph{declared} one.
32527
32528@item value
32529If values were requested, this is the value.
32530
32531@item thread-id
32532If this variable object is associated with a thread, this is the thread id.
32533Otherwise this result is not present.
32534
32535@item frozen
32536If the variable object is frozen, this variable will be present with a value of 1.
32537
32538@item displayhint
32539A dynamic varobj can supply a display hint to the front end. The
32540value comes directly from the Python pretty-printer object's
32541@code{display_hint} method. @xref{Pretty Printing API}.
32542
32543@item dynamic
32544This attribute will be present and have the value @samp{1} if the
32545varobj is a dynamic varobj. If the varobj is not a dynamic varobj,
32546then this attribute will not be present.
32547
32548@end table
32549
32550The result may have its own attributes:
32551
32552@table @samp
32553@item displayhint
32554A dynamic varobj can supply a display hint to the front end. The
32555value comes directly from the Python pretty-printer object's
32556@code{display_hint} method. @xref{Pretty Printing API}.
32557
32558@item has_more
32559This is an integer attribute which is nonzero if there are children
32560remaining after the end of the selected range.
32561@end table
32562
32563@subsubheading Example
32564
32565@smallexample
32566(gdb)
32567 -var-list-children n
32568 ^done,numchild=@var{n},children=[child=@{name=@var{name},exp=@var{exp},
32569 numchild=@var{n},type=@var{type}@},@r{(repeats N times)}]
32570(gdb)
32571 -var-list-children --all-values n
32572 ^done,numchild=@var{n},children=[child=@{name=@var{name},exp=@var{exp},
32573 numchild=@var{n},value=@var{value},type=@var{type}@},@r{(repeats N times)}]
32574@end smallexample
32575
32576
32577@subheading The @code{-var-info-type} Command
32578@findex -var-info-type
32579
32580@subsubheading Synopsis
32581
32582@smallexample
32583 -var-info-type @var{name}
32584@end smallexample
32585
32586Returns the type of the specified variable @var{name}. The type is
32587returned as a string in the same format as it is output by the
32588@value{GDBN} CLI:
32589
32590@smallexample
32591 type=@var{typename}
32592@end smallexample
32593
32594
32595@subheading The @code{-var-info-expression} Command
32596@findex -var-info-expression
32597
32598@subsubheading Synopsis
32599
32600@smallexample
32601 -var-info-expression @var{name}
32602@end smallexample
32603
32604Returns a string that is suitable for presenting this
32605variable object in user interface. The string is generally
32606not valid expression in the current language, and cannot be evaluated.
32607
32608For example, if @code{a} is an array, and variable object
32609@code{A} was created for @code{a}, then we'll get this output:
32610
32611@smallexample
32612(gdb) -var-info-expression A.1
32613^done,lang="C",exp="1"
32614@end smallexample
32615
32616@noindent
32617Here, the value of @code{lang} is the language name, which can be
32618found in @ref{Supported Languages}.
32619
32620Note that the output of the @code{-var-list-children} command also
32621includes those expressions, so the @code{-var-info-expression} command
32622is of limited use.
32623
32624@subheading The @code{-var-info-path-expression} Command
32625@findex -var-info-path-expression
32626
32627@subsubheading Synopsis
32628
32629@smallexample
32630 -var-info-path-expression @var{name}
32631@end smallexample
32632
32633Returns an expression that can be evaluated in the current
32634context and will yield the same value that a variable object has.
32635Compare this with the @code{-var-info-expression} command, which
32636result can be used only for UI presentation. Typical use of
32637the @code{-var-info-path-expression} command is creating a
32638watchpoint from a variable object.
32639
32640This command is currently not valid for children of a dynamic varobj,
32641and will give an error when invoked on one.
32642
32643For example, suppose @code{C} is a C@t{++} class, derived from class
32644@code{Base}, and that the @code{Base} class has a member called
32645@code{m_size}. Assume a variable @code{c} is has the type of
32646@code{C} and a variable object @code{C} was created for variable
32647@code{c}. Then, we'll get this output:
32648@smallexample
32649(gdb) -var-info-path-expression C.Base.public.m_size
32650^done,path_expr=((Base)c).m_size)
32651@end smallexample
32652
32653@subheading The @code{-var-show-attributes} Command
32654@findex -var-show-attributes
32655
32656@subsubheading Synopsis
32657
32658@smallexample
32659 -var-show-attributes @var{name}
32660@end smallexample
32661
32662List attributes of the specified variable object @var{name}:
32663
32664@smallexample
32665 status=@var{attr} [ ( ,@var{attr} )* ]
32666@end smallexample
32667
32668@noindent
32669where @var{attr} is @code{@{ @{ editable | noneditable @} | TBD @}}.
32670
32671@subheading The @code{-var-evaluate-expression} Command
32672@findex -var-evaluate-expression
32673
32674@subsubheading Synopsis
32675
32676@smallexample
32677 -var-evaluate-expression [-f @var{format-spec}] @var{name}
32678@end smallexample
32679
32680Evaluates the expression that is represented by the specified variable
32681object and returns its value as a string. The format of the string
32682can be specified with the @samp{-f} option. The possible values of
32683this option are the same as for @code{-var-set-format}
32684(@pxref{-var-set-format}). If the @samp{-f} option is not specified,
32685the current display format will be used. The current display format
32686can be changed using the @code{-var-set-format} command.
32687
32688@smallexample
32689 value=@var{value}
32690@end smallexample
32691
32692Note that one must invoke @code{-var-list-children} for a variable
32693before the value of a child variable can be evaluated.
32694
32695@subheading The @code{-var-assign} Command
32696@findex -var-assign
32697
32698@subsubheading Synopsis
32699
32700@smallexample
32701 -var-assign @var{name} @var{expression}
32702@end smallexample
32703
32704Assigns the value of @var{expression} to the variable object specified
32705by @var{name}. The object must be @samp{editable}. If the variable's
32706value is altered by the assign, the variable will show up in any
32707subsequent @code{-var-update} list.
32708
32709@subsubheading Example
32710
32711@smallexample
32712(gdb)
32713-var-assign var1 3
32714^done,value="3"
32715(gdb)
32716-var-update *
32717^done,changelist=[@{name="var1",in_scope="true",type_changed="false"@}]
32718(gdb)
32719@end smallexample
32720
32721@subheading The @code{-var-update} Command
32722@findex -var-update
32723
32724@subsubheading Synopsis
32725
32726@smallexample
32727 -var-update [@var{print-values}] @{@var{name} | "*"@}
32728@end smallexample
32729
32730Reevaluate the expressions corresponding to the variable object
32731@var{name} and all its direct and indirect children, and return the
32732list of variable objects whose values have changed; @var{name} must
32733be a root variable object. Here, ``changed'' means that the result of
32734@code{-var-evaluate-expression} before and after the
32735@code{-var-update} is different. If @samp{*} is used as the variable
32736object names, all existing variable objects are updated, except
32737for frozen ones (@pxref{-var-set-frozen}). The option
32738@var{print-values} determines whether both names and values, or just
32739names are printed. The possible values of this option are the same
32740as for @code{-var-list-children} (@pxref{-var-list-children}). It is
32741recommended to use the @samp{--all-values} option, to reduce the
32742number of MI commands needed on each program stop.
32743
32744With the @samp{*} parameter, if a variable object is bound to a
32745currently running thread, it will not be updated, without any
32746diagnostic.
32747
32748If @code{-var-set-update-range} was previously used on a varobj, then
32749only the selected range of children will be reported.
32750
32751@code{-var-update} reports all the changed varobjs in a tuple named
32752@samp{changelist}.
32753
32754Each item in the change list is itself a tuple holding:
32755
32756@table @samp
32757@item name
32758The name of the varobj.
32759
32760@item value
32761If values were requested for this update, then this field will be
32762present and will hold the value of the varobj.
32763
32764@item in_scope
32765@anchor{-var-update}
32766This field is a string which may take one of three values:
32767
32768@table @code
32769@item "true"
32770The variable object's current value is valid.
32771
32772@item "false"
32773The variable object does not currently hold a valid value but it may
32774hold one in the future if its associated expression comes back into
32775scope.
32776
32777@item "invalid"
32778The variable object no longer holds a valid value.
32779This can occur when the executable file being debugged has changed,
32780either through recompilation or by using the @value{GDBN} @code{file}
32781command. The front end should normally choose to delete these variable
32782objects.
32783@end table
32784
32785In the future new values may be added to this list so the front should
32786be prepared for this possibility. @xref{GDB/MI Development and Front Ends, ,@sc{GDB/MI} Development and Front Ends}.
32787
32788@item type_changed
32789This is only present if the varobj is still valid. If the type
32790changed, then this will be the string @samp{true}; otherwise it will
32791be @samp{false}.
32792
32793When a varobj's type changes, its children are also likely to have
32794become incorrect. Therefore, the varobj's children are automatically
32795deleted when this attribute is @samp{true}. Also, the varobj's update
32796range, when set using the @code{-var-set-update-range} command, is
32797unset.
32798
32799@item new_type
32800If the varobj's type changed, then this field will be present and will
32801hold the new type.
32802
32803@item new_num_children
32804For a dynamic varobj, if the number of children changed, or if the
32805type changed, this will be the new number of children.
32806
32807The @samp{numchild} field in other varobj responses is generally not
32808valid for a dynamic varobj -- it will show the number of children that
32809@value{GDBN} knows about, but because dynamic varobjs lazily
32810instantiate their children, this will not reflect the number of
32811children which may be available.
32812
32813The @samp{new_num_children} attribute only reports changes to the
32814number of children known by @value{GDBN}. This is the only way to
32815detect whether an update has removed children (which necessarily can
32816only happen at the end of the update range).
32817
32818@item displayhint
32819The display hint, if any.
32820
32821@item has_more
32822This is an integer value, which will be 1 if there are more children
32823available outside the varobj's update range.
32824
32825@item dynamic
32826This attribute will be present and have the value @samp{1} if the
32827varobj is a dynamic varobj. If the varobj is not a dynamic varobj,
32828then this attribute will not be present.
32829
32830@item new_children
32831If new children were added to a dynamic varobj within the selected
32832update range (as set by @code{-var-set-update-range}), then they will
32833be listed in this attribute.
32834@end table
32835
32836@subsubheading Example
32837
32838@smallexample
32839(gdb)
32840-var-assign var1 3
32841^done,value="3"
32842(gdb)
32843-var-update --all-values var1
32844^done,changelist=[@{name="var1",value="3",in_scope="true",
32845type_changed="false"@}]
32846(gdb)
32847@end smallexample
32848
32849@subheading The @code{-var-set-frozen} Command
32850@findex -var-set-frozen
32851@anchor{-var-set-frozen}
32852
32853@subsubheading Synopsis
32854
32855@smallexample
32856 -var-set-frozen @var{name} @var{flag}
32857@end smallexample
32858
32859Set the frozenness flag on the variable object @var{name}. The
32860@var{flag} parameter should be either @samp{1} to make the variable
32861frozen or @samp{0} to make it unfrozen. If a variable object is
32862frozen, then neither itself, nor any of its children, are
32863implicitly updated by @code{-var-update} of
32864a parent variable or by @code{-var-update *}. Only
32865@code{-var-update} of the variable itself will update its value and
32866values of its children. After a variable object is unfrozen, it is
32867implicitly updated by all subsequent @code{-var-update} operations.
32868Unfreezing a variable does not update it, only subsequent
32869@code{-var-update} does.
32870
32871@subsubheading Example
32872
32873@smallexample
32874(gdb)
32875-var-set-frozen V 1
32876^done
32877(gdb)
32878@end smallexample
32879
32880@subheading The @code{-var-set-update-range} command
32881@findex -var-set-update-range
32882@anchor{-var-set-update-range}
32883
32884@subsubheading Synopsis
32885
32886@smallexample
32887 -var-set-update-range @var{name} @var{from} @var{to}
32888@end smallexample
32889
32890Set the range of children to be returned by future invocations of
32891@code{-var-update}.
32892
32893@var{from} and @var{to} indicate the range of children to report. If
32894@var{from} or @var{to} is less than zero, the range is reset and all
32895children will be reported. Otherwise, children starting at @var{from}
32896(zero-based) and up to and excluding @var{to} will be reported.
32897
32898@subsubheading Example
32899
32900@smallexample
32901(gdb)
32902-var-set-update-range V 1 2
32903^done
32904@end smallexample
32905
32906@subheading The @code{-var-set-visualizer} command
32907@findex -var-set-visualizer
32908@anchor{-var-set-visualizer}
32909
32910@subsubheading Synopsis
32911
32912@smallexample
32913 -var-set-visualizer @var{name} @var{visualizer}
32914@end smallexample
32915
32916Set a visualizer for the variable object @var{name}.
32917
32918@var{visualizer} is the visualizer to use. The special value
32919@samp{None} means to disable any visualizer in use.
32920
32921If not @samp{None}, @var{visualizer} must be a Python expression.
32922This expression must evaluate to a callable object which accepts a
32923single argument. @value{GDBN} will call this object with the value of
32924the varobj @var{name} as an argument (this is done so that the same
32925Python pretty-printing code can be used for both the CLI and MI).
32926When called, this object must return an object which conforms to the
32927pretty-printing interface (@pxref{Pretty Printing API}).
32928
32929The pre-defined function @code{gdb.default_visualizer} may be used to
32930select a visualizer by following the built-in process
32931(@pxref{Selecting Pretty-Printers}). This is done automatically when
32932a varobj is created, and so ordinarily is not needed.
32933
32934This feature is only available if Python support is enabled. The MI
32935command @code{-list-features} (@pxref{GDB/MI Miscellaneous Commands})
32936can be used to check this.
32937
32938@subsubheading Example
32939
32940Resetting the visualizer:
32941
32942@smallexample
32943(gdb)
32944-var-set-visualizer V None
32945^done
32946@end smallexample
32947
32948Reselecting the default (type-based) visualizer:
32949
32950@smallexample
32951(gdb)
32952-var-set-visualizer V gdb.default_visualizer
32953^done
32954@end smallexample
32955
32956Suppose @code{SomeClass} is a visualizer class. A lambda expression
32957can be used to instantiate this class for a varobj:
32958
32959@smallexample
32960(gdb)
32961-var-set-visualizer V "lambda val: SomeClass()"
32962^done
32963@end smallexample
32964
32965@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32966@node GDB/MI Data Manipulation
32967@section @sc{gdb/mi} Data Manipulation
32968
32969@cindex data manipulation, in @sc{gdb/mi}
32970@cindex @sc{gdb/mi}, data manipulation
32971This section describes the @sc{gdb/mi} commands that manipulate data:
32972examine memory and registers, evaluate expressions, etc.
32973
32974@c REMOVED FROM THE INTERFACE.
32975@c @subheading -data-assign
32976@c Change the value of a program variable. Plenty of side effects.
32977@c @subsubheading GDB Command
32978@c set variable
32979@c @subsubheading Example
32980@c N.A.
32981
32982@subheading The @code{-data-disassemble} Command
32983@findex -data-disassemble
32984
32985@subsubheading Synopsis
32986
32987@smallexample
32988 -data-disassemble
32989 [ -s @var{start-addr} -e @var{end-addr} ]
32990 | [ -f @var{filename} -l @var{linenum} [ -n @var{lines} ] ]
32991 -- @var{mode}
32992@end smallexample
32993
32994@noindent
32995Where:
32996
32997@table @samp
32998@item @var{start-addr}
32999is the beginning address (or @code{$pc})
33000@item @var{end-addr}
33001is the end address
33002@item @var{filename}
33003is the name of the file to disassemble
33004@item @var{linenum}
33005is the line number to disassemble around
33006@item @var{lines}
33007is the number of disassembly lines to be produced. If it is -1,
33008the whole function will be disassembled, in case no @var{end-addr} is
33009specified. If @var{end-addr} is specified as a non-zero value, and
33010@var{lines} is lower than the number of disassembly lines between
33011@var{start-addr} and @var{end-addr}, only @var{lines} lines are
33012displayed; if @var{lines} is higher than the number of lines between
33013@var{start-addr} and @var{end-addr}, only the lines up to @var{end-addr}
33014are displayed.
33015@item @var{mode}
33016is either 0 (meaning only disassembly), 1 (meaning mixed source and
33017disassembly), 2 (meaning disassembly with raw opcodes), or 3 (meaning
33018mixed source and disassembly with raw opcodes).
33019@end table
33020
33021@subsubheading Result
33022
33023The result of the @code{-data-disassemble} command will be a list named
33024@samp{asm_insns}, the contents of this list depend on the @var{mode}
33025used with the @code{-data-disassemble} command.
33026
33027For modes 0 and 2 the @samp{asm_insns} list contains tuples with the
33028following fields:
33029
33030@table @code
33031@item address
33032The address at which this instruction was disassembled.
33033
33034@item func-name
33035The name of the function this instruction is within.
33036
33037@item offset
33038The decimal offset in bytes from the start of @samp{func-name}.
33039
33040@item inst
33041The text disassembly for this @samp{address}.
33042
33043@item opcodes
33044This field is only present for mode 2. This contains the raw opcode
33045bytes for the @samp{inst} field.
33046
33047@end table
33048
33049For modes 1 and 3 the @samp{asm_insns} list contains tuples named
33050@samp{src_and_asm_line}, each of which has the following fields:
33051
33052@table @code
33053@item line
33054The line number within @samp{file}.
33055
33056@item file
33057The file name from the compilation unit. This might be an absolute
33058file name or a relative file name depending on the compile command
33059used.
33060
33061@item fullname
33062Absolute file name of @samp{file}. It is converted to a canonical form
33063using the source file search path
33064(@pxref{Source Path, ,Specifying Source Directories})
33065and after resolving all the symbolic links.
33066
33067If the source file is not found this field will contain the path as
33068present in the debug information.
33069
33070@item line_asm_insn
33071This is a list of tuples containing the disassembly for @samp{line} in
33072@samp{file}. The fields of each tuple are the same as for
33073@code{-data-disassemble} in @var{mode} 0 and 2, so @samp{address},
33074@samp{func-name}, @samp{offset}, @samp{inst}, and optionally
33075@samp{opcodes}.
33076
33077@end table
33078
33079Note that whatever included in the @samp{inst} field, is not
33080manipulated directly by @sc{gdb/mi}, i.e., it is not possible to
33081adjust its format.
33082
33083@subsubheading @value{GDBN} Command
33084
33085The corresponding @value{GDBN} command is @samp{disassemble}.
33086
33087@subsubheading Example
33088
33089Disassemble from the current value of @code{$pc} to @code{$pc + 20}:
33090
33091@smallexample
33092(gdb)
33093-data-disassemble -s $pc -e "$pc + 20" -- 0
33094^done,
33095asm_insns=[
33096@{address="0x000107c0",func-name="main",offset="4",
33097inst="mov 2, %o0"@},
33098@{address="0x000107c4",func-name="main",offset="8",
33099inst="sethi %hi(0x11800), %o2"@},
33100@{address="0x000107c8",func-name="main",offset="12",
33101inst="or %o2, 0x140, %o1\t! 0x11940 <_lib_version+8>"@},
33102@{address="0x000107cc",func-name="main",offset="16",
33103inst="sethi %hi(0x11800), %o2"@},
33104@{address="0x000107d0",func-name="main",offset="20",
33105inst="or %o2, 0x168, %o4\t! 0x11968 <_lib_version+48>"@}]
33106(gdb)
33107@end smallexample
33108
33109Disassemble the whole @code{main} function. Line 32 is part of
33110@code{main}.
33111
33112@smallexample
33113-data-disassemble -f basics.c -l 32 -- 0
33114^done,asm_insns=[
33115@{address="0x000107bc",func-name="main",offset="0",
33116inst="save %sp, -112, %sp"@},
33117@{address="0x000107c0",func-name="main",offset="4",
33118inst="mov 2, %o0"@},
33119@{address="0x000107c4",func-name="main",offset="8",
33120inst="sethi %hi(0x11800), %o2"@},
33121[@dots{}]
33122@{address="0x0001081c",func-name="main",offset="96",inst="ret "@},
33123@{address="0x00010820",func-name="main",offset="100",inst="restore "@}]
33124(gdb)
33125@end smallexample
33126
33127Disassemble 3 instructions from the start of @code{main}:
33128
33129@smallexample
33130(gdb)
33131-data-disassemble -f basics.c -l 32 -n 3 -- 0
33132^done,asm_insns=[
33133@{address="0x000107bc",func-name="main",offset="0",
33134inst="save %sp, -112, %sp"@},
33135@{address="0x000107c0",func-name="main",offset="4",
33136inst="mov 2, %o0"@},
33137@{address="0x000107c4",func-name="main",offset="8",
33138inst="sethi %hi(0x11800), %o2"@}]
33139(gdb)
33140@end smallexample
33141
33142Disassemble 3 instructions from the start of @code{main} in mixed mode:
33143
33144@smallexample
33145(gdb)
33146-data-disassemble -f basics.c -l 32 -n 3 -- 1
33147^done,asm_insns=[
33148src_and_asm_line=@{line="31",
33149file="../../../src/gdb/testsuite/gdb.mi/basics.c",
33150fullname="/absolute/path/to/src/gdb/testsuite/gdb.mi/basics.c",
33151line_asm_insn=[@{address="0x000107bc",
33152func-name="main",offset="0",inst="save %sp, -112, %sp"@}]@},
33153src_and_asm_line=@{line="32",
33154file="../../../src/gdb/testsuite/gdb.mi/basics.c",
33155fullname="/absolute/path/to/src/gdb/testsuite/gdb.mi/basics.c",
33156line_asm_insn=[@{address="0x000107c0",
33157func-name="main",offset="4",inst="mov 2, %o0"@},
33158@{address="0x000107c4",func-name="main",offset="8",
33159inst="sethi %hi(0x11800), %o2"@}]@}]
33160(gdb)
33161@end smallexample
33162
33163
33164@subheading The @code{-data-evaluate-expression} Command
33165@findex -data-evaluate-expression
33166
33167@subsubheading Synopsis
33168
33169@smallexample
33170 -data-evaluate-expression @var{expr}
33171@end smallexample
33172
33173Evaluate @var{expr} as an expression. The expression could contain an
33174inferior function call. The function call will execute synchronously.
33175If the expression contains spaces, it must be enclosed in double quotes.
33176
33177@subsubheading @value{GDBN} Command
33178
33179The corresponding @value{GDBN} commands are @samp{print}, @samp{output}, and
33180@samp{call}. In @code{gdbtk} only, there's a corresponding
33181@samp{gdb_eval} command.
33182
33183@subsubheading Example
33184
33185In the following example, the numbers that precede the commands are the
33186@dfn{tokens} described in @ref{GDB/MI Command Syntax, ,@sc{gdb/mi}
33187Command Syntax}. Notice how @sc{gdb/mi} returns the same tokens in its
33188output.
33189
33190@smallexample
33191211-data-evaluate-expression A
33192211^done,value="1"
33193(gdb)
33194311-data-evaluate-expression &A
33195311^done,value="0xefffeb7c"
33196(gdb)
33197411-data-evaluate-expression A+3
33198411^done,value="4"
33199(gdb)
33200511-data-evaluate-expression "A + 3"
33201511^done,value="4"
33202(gdb)
33203@end smallexample
33204
33205
33206@subheading The @code{-data-list-changed-registers} Command
33207@findex -data-list-changed-registers
33208
33209@subsubheading Synopsis
33210
33211@smallexample
33212 -data-list-changed-registers
33213@end smallexample
33214
33215Display a list of the registers that have changed.
33216
33217@subsubheading @value{GDBN} Command
33218
33219@value{GDBN} doesn't have a direct analog for this command; @code{gdbtk}
33220has the corresponding command @samp{gdb_changed_register_list}.
33221
33222@subsubheading Example
33223
33224On a PPC MBX board:
33225
33226@smallexample
33227(gdb)
33228-exec-continue
33229^running
33230
33231(gdb)
33232*stopped,reason="breakpoint-hit",disp="keep",bkptno="1",frame=@{
33233func="main",args=[],file="try.c",fullname="/home/foo/bar/try.c",
33234line="5"@}
33235(gdb)
33236-data-list-changed-registers
33237^done,changed-registers=["0","1","2","4","5","6","7","8","9",
33238"10","11","13","14","15","16","17","18","19","20","21","22","23",
33239"24","25","26","27","28","30","31","64","65","66","67","69"]
33240(gdb)
33241@end smallexample
33242
33243
33244@subheading The @code{-data-list-register-names} Command
33245@findex -data-list-register-names
33246
33247@subsubheading Synopsis
33248
33249@smallexample
33250 -data-list-register-names [ ( @var{regno} )+ ]
33251@end smallexample
33252
33253Show a list of register names for the current target. If no arguments
33254are given, it shows a list of the names of all the registers. If
33255integer numbers are given as arguments, it will print a list of the
33256names of the registers corresponding to the arguments. To ensure
33257consistency between a register name and its number, the output list may
33258include empty register names.
33259
33260@subsubheading @value{GDBN} Command
33261
33262@value{GDBN} does not have a command which corresponds to
33263@samp{-data-list-register-names}. In @code{gdbtk} there is a
33264corresponding command @samp{gdb_regnames}.
33265
33266@subsubheading Example
33267
33268For the PPC MBX board:
33269@smallexample
33270(gdb)
33271-data-list-register-names
33272^done,register-names=["r0","r1","r2","r3","r4","r5","r6","r7",
33273"r8","r9","r10","r11","r12","r13","r14","r15","r16","r17","r18",
33274"r19","r20","r21","r22","r23","r24","r25","r26","r27","r28","r29",
33275"r30","r31","f0","f1","f2","f3","f4","f5","f6","f7","f8","f9",
33276"f10","f11","f12","f13","f14","f15","f16","f17","f18","f19","f20",
33277"f21","f22","f23","f24","f25","f26","f27","f28","f29","f30","f31",
33278"", "pc","ps","cr","lr","ctr","xer"]
33279(gdb)
33280-data-list-register-names 1 2 3
33281^done,register-names=["r1","r2","r3"]
33282(gdb)
33283@end smallexample
33284
33285@subheading The @code{-data-list-register-values} Command
33286@findex -data-list-register-values
33287
33288@subsubheading Synopsis
33289
33290@smallexample
33291 -data-list-register-values
33292 [ @code{--skip-unavailable} ] @var{fmt} [ ( @var{regno} )*]
33293@end smallexample
33294
33295Display the registers' contents. @var{fmt} is the format according to
33296which the registers' contents are to be returned, followed by an optional
33297list of numbers specifying the registers to display. A missing list of
33298numbers indicates that the contents of all the registers must be
33299returned. The @code{--skip-unavailable} option indicates that only
33300the available registers are to be returned.
33301
33302Allowed formats for @var{fmt} are:
33303
33304@table @code
33305@item x
33306Hexadecimal
33307@item o
33308Octal
33309@item t
33310Binary
33311@item d
33312Decimal
33313@item r
33314Raw
33315@item N
33316Natural
33317@end table
33318
33319@subsubheading @value{GDBN} Command
33320
33321The corresponding @value{GDBN} commands are @samp{info reg}, @samp{info
33322all-reg}, and (in @code{gdbtk}) @samp{gdb_fetch_registers}.
33323
33324@subsubheading Example
33325
33326For a PPC MBX board (note: line breaks are for readability only, they
33327don't appear in the actual output):
33328
33329@smallexample
33330(gdb)
33331-data-list-register-values r 64 65
33332^done,register-values=[@{number="64",value="0xfe00a300"@},
33333@{number="65",value="0x00029002"@}]
33334(gdb)
33335-data-list-register-values x
33336^done,register-values=[@{number="0",value="0xfe0043c8"@},
33337@{number="1",value="0x3fff88"@},@{number="2",value="0xfffffffe"@},
33338@{number="3",value="0x0"@},@{number="4",value="0xa"@},
33339@{number="5",value="0x3fff68"@},@{number="6",value="0x3fff58"@},
33340@{number="7",value="0xfe011e98"@},@{number="8",value="0x2"@},
33341@{number="9",value="0xfa202820"@},@{number="10",value="0xfa202808"@},
33342@{number="11",value="0x1"@},@{number="12",value="0x0"@},
33343@{number="13",value="0x4544"@},@{number="14",value="0xffdfffff"@},
33344@{number="15",value="0xffffffff"@},@{number="16",value="0xfffffeff"@},
33345@{number="17",value="0xefffffed"@},@{number="18",value="0xfffffffe"@},
33346@{number="19",value="0xffffffff"@},@{number="20",value="0xffffffff"@},
33347@{number="21",value="0xffffffff"@},@{number="22",value="0xfffffff7"@},
33348@{number="23",value="0xffffffff"@},@{number="24",value="0xffffffff"@},
33349@{number="25",value="0xffffffff"@},@{number="26",value="0xfffffffb"@},
33350@{number="27",value="0xffffffff"@},@{number="28",value="0xf7bfffff"@},
33351@{number="29",value="0x0"@},@{number="30",value="0xfe010000"@},
33352@{number="31",value="0x0"@},@{number="32",value="0x0"@},
33353@{number="33",value="0x0"@},@{number="34",value="0x0"@},
33354@{number="35",value="0x0"@},@{number="36",value="0x0"@},
33355@{number="37",value="0x0"@},@{number="38",value="0x0"@},
33356@{number="39",value="0x0"@},@{number="40",value="0x0"@},
33357@{number="41",value="0x0"@},@{number="42",value="0x0"@},
33358@{number="43",value="0x0"@},@{number="44",value="0x0"@},
33359@{number="45",value="0x0"@},@{number="46",value="0x0"@},
33360@{number="47",value="0x0"@},@{number="48",value="0x0"@},
33361@{number="49",value="0x0"@},@{number="50",value="0x0"@},
33362@{number="51",value="0x0"@},@{number="52",value="0x0"@},
33363@{number="53",value="0x0"@},@{number="54",value="0x0"@},
33364@{number="55",value="0x0"@},@{number="56",value="0x0"@},
33365@{number="57",value="0x0"@},@{number="58",value="0x0"@},
33366@{number="59",value="0x0"@},@{number="60",value="0x0"@},
33367@{number="61",value="0x0"@},@{number="62",value="0x0"@},
33368@{number="63",value="0x0"@},@{number="64",value="0xfe00a300"@},
33369@{number="65",value="0x29002"@},@{number="66",value="0x202f04b5"@},
33370@{number="67",value="0xfe0043b0"@},@{number="68",value="0xfe00b3e4"@},
33371@{number="69",value="0x20002b03"@}]
33372(gdb)
33373@end smallexample
33374
33375
33376@subheading The @code{-data-read-memory} Command
33377@findex -data-read-memory
33378
33379This command is deprecated, use @code{-data-read-memory-bytes} instead.
33380
33381@subsubheading Synopsis
33382
33383@smallexample
33384 -data-read-memory [ -o @var{byte-offset} ]
33385 @var{address} @var{word-format} @var{word-size}
33386 @var{nr-rows} @var{nr-cols} [ @var{aschar} ]
33387@end smallexample
33388
33389@noindent
33390where:
33391
33392@table @samp
33393@item @var{address}
33394An expression specifying the address of the first memory word to be
33395read. Complex expressions containing embedded white space should be
33396quoted using the C convention.
33397
33398@item @var{word-format}
33399The format to be used to print the memory words. The notation is the
33400same as for @value{GDBN}'s @code{print} command (@pxref{Output Formats,
33401,Output Formats}).
33402
33403@item @var{word-size}
33404The size of each memory word in bytes.
33405
33406@item @var{nr-rows}
33407The number of rows in the output table.
33408
33409@item @var{nr-cols}
33410The number of columns in the output table.
33411
33412@item @var{aschar}
33413If present, indicates that each row should include an @sc{ascii} dump. The
33414value of @var{aschar} is used as a padding character when a byte is not a
33415member of the printable @sc{ascii} character set (printable @sc{ascii}
33416characters are those whose code is between 32 and 126, inclusively).
33417
33418@item @var{byte-offset}
33419An offset to add to the @var{address} before fetching memory.
33420@end table
33421
33422This command displays memory contents as a table of @var{nr-rows} by
33423@var{nr-cols} words, each word being @var{word-size} bytes. In total,
33424@code{@var{nr-rows} * @var{nr-cols} * @var{word-size}} bytes are read
33425(returned as @samp{total-bytes}). Should less than the requested number
33426of bytes be returned by the target, the missing words are identified
33427using @samp{N/A}. The number of bytes read from the target is returned
33428in @samp{nr-bytes} and the starting address used to read memory in
33429@samp{addr}.
33430
33431The address of the next/previous row or page is available in
33432@samp{next-row} and @samp{prev-row}, @samp{next-page} and
33433@samp{prev-page}.
33434
33435@subsubheading @value{GDBN} Command
33436
33437The corresponding @value{GDBN} command is @samp{x}. @code{gdbtk} has
33438@samp{gdb_get_mem} memory read command.
33439
33440@subsubheading Example
33441
33442Read six bytes of memory starting at @code{bytes+6} but then offset by
33443@code{-6} bytes. Format as three rows of two columns. One byte per
33444word. Display each word in hex.
33445
33446@smallexample
33447(gdb)
334489-data-read-memory -o -6 -- bytes+6 x 1 3 2
334499^done,addr="0x00001390",nr-bytes="6",total-bytes="6",
33450next-row="0x00001396",prev-row="0x0000138e",next-page="0x00001396",
33451prev-page="0x0000138a",memory=[
33452@{addr="0x00001390",data=["0x00","0x01"]@},
33453@{addr="0x00001392",data=["0x02","0x03"]@},
33454@{addr="0x00001394",data=["0x04","0x05"]@}]
33455(gdb)
33456@end smallexample
33457
33458Read two bytes of memory starting at address @code{shorts + 64} and
33459display as a single word formatted in decimal.
33460
33461@smallexample
33462(gdb)
334635-data-read-memory shorts+64 d 2 1 1
334645^done,addr="0x00001510",nr-bytes="2",total-bytes="2",
33465next-row="0x00001512",prev-row="0x0000150e",
33466next-page="0x00001512",prev-page="0x0000150e",memory=[
33467@{addr="0x00001510",data=["128"]@}]
33468(gdb)
33469@end smallexample
33470
33471Read thirty two bytes of memory starting at @code{bytes+16} and format
33472as eight rows of four columns. Include a string encoding with @samp{x}
33473used as the non-printable character.
33474
33475@smallexample
33476(gdb)
334774-data-read-memory bytes+16 x 1 8 4 x
334784^done,addr="0x000013a0",nr-bytes="32",total-bytes="32",
33479next-row="0x000013c0",prev-row="0x0000139c",
33480next-page="0x000013c0",prev-page="0x00001380",memory=[
33481@{addr="0x000013a0",data=["0x10","0x11","0x12","0x13"],ascii="xxxx"@},
33482@{addr="0x000013a4",data=["0x14","0x15","0x16","0x17"],ascii="xxxx"@},
33483@{addr="0x000013a8",data=["0x18","0x19","0x1a","0x1b"],ascii="xxxx"@},
33484@{addr="0x000013ac",data=["0x1c","0x1d","0x1e","0x1f"],ascii="xxxx"@},
33485@{addr="0x000013b0",data=["0x20","0x21","0x22","0x23"],ascii=" !\"#"@},
33486@{addr="0x000013b4",data=["0x24","0x25","0x26","0x27"],ascii="$%&'"@},
33487@{addr="0x000013b8",data=["0x28","0x29","0x2a","0x2b"],ascii="()*+"@},
33488@{addr="0x000013bc",data=["0x2c","0x2d","0x2e","0x2f"],ascii=",-./"@}]
33489(gdb)
33490@end smallexample
33491
33492@subheading The @code{-data-read-memory-bytes} Command
33493@findex -data-read-memory-bytes
33494
33495@subsubheading Synopsis
33496
33497@smallexample
33498 -data-read-memory-bytes [ -o @var{byte-offset} ]
33499 @var{address} @var{count}
33500@end smallexample
33501
33502@noindent
33503where:
33504
33505@table @samp
33506@item @var{address}
33507An expression specifying the address of the first memory word to be
33508read. Complex expressions containing embedded white space should be
33509quoted using the C convention.
33510
33511@item @var{count}
33512The number of bytes to read. This should be an integer literal.
33513
33514@item @var{byte-offset}
33515The offsets in bytes relative to @var{address} at which to start
33516reading. This should be an integer literal. This option is provided
33517so that a frontend is not required to first evaluate address and then
33518perform address arithmetics itself.
33519
33520@end table
33521
33522This command attempts to read all accessible memory regions in the
33523specified range. First, all regions marked as unreadable in the memory
33524map (if one is defined) will be skipped. @xref{Memory Region
33525Attributes}. Second, @value{GDBN} will attempt to read the remaining
33526regions. For each one, if reading full region results in an errors,
33527@value{GDBN} will try to read a subset of the region.
33528
33529In general, every single byte in the region may be readable or not,
33530and the only way to read every readable byte is to try a read at
33531every address, which is not practical. Therefore, @value{GDBN} will
33532attempt to read all accessible bytes at either beginning or the end
33533of the region, using a binary division scheme. This heuristic works
33534well for reading accross a memory map boundary. Note that if a region
33535has a readable range that is neither at the beginning or the end,
33536@value{GDBN} will not read it.
33537
33538The result record (@pxref{GDB/MI Result Records}) that is output of
33539the command includes a field named @samp{memory} whose content is a
33540list of tuples. Each tuple represent a successfully read memory block
33541and has the following fields:
33542
33543@table @code
33544@item begin
33545The start address of the memory block, as hexadecimal literal.
33546
33547@item end
33548The end address of the memory block, as hexadecimal literal.
33549
33550@item offset
33551The offset of the memory block, as hexadecimal literal, relative to
33552the start address passed to @code{-data-read-memory-bytes}.
33553
33554@item contents
33555The contents of the memory block, in hex.
33556
33557@end table
33558
33559
33560
33561@subsubheading @value{GDBN} Command
33562
33563The corresponding @value{GDBN} command is @samp{x}.
33564
33565@subsubheading Example
33566
33567@smallexample
33568(gdb)
33569-data-read-memory-bytes &a 10
33570^done,memory=[@{begin="0xbffff154",offset="0x00000000",
33571 end="0xbffff15e",
33572 contents="01000000020000000300"@}]
33573(gdb)
33574@end smallexample
33575
33576
33577@subheading The @code{-data-write-memory-bytes} Command
33578@findex -data-write-memory-bytes
33579
33580@subsubheading Synopsis
33581
33582@smallexample
33583 -data-write-memory-bytes @var{address} @var{contents}
33584 -data-write-memory-bytes @var{address} @var{contents} @r{[}@var{count}@r{]}
33585@end smallexample
33586
33587@noindent
33588where:
33589
33590@table @samp
33591@item @var{address}
33592An expression specifying the address of the first memory word to be
33593read. Complex expressions containing embedded white space should be
33594quoted using the C convention.
33595
33596@item @var{contents}
33597The hex-encoded bytes to write.
33598
33599@item @var{count}
33600Optional argument indicating the number of bytes to be written. If @var{count}
33601is greater than @var{contents}' length, @value{GDBN} will repeatedly
33602write @var{contents} until it fills @var{count} bytes.
33603
33604@end table
33605
33606@subsubheading @value{GDBN} Command
33607
33608There's no corresponding @value{GDBN} command.
33609
33610@subsubheading Example
33611
33612@smallexample
33613(gdb)
33614-data-write-memory-bytes &a "aabbccdd"
33615^done
33616(gdb)
33617@end smallexample
33618
33619@smallexample
33620(gdb)
33621-data-write-memory-bytes &a "aabbccdd" 16e
33622^done
33623(gdb)
33624@end smallexample
33625
33626@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
33627@node GDB/MI Tracepoint Commands
33628@section @sc{gdb/mi} Tracepoint Commands
33629
33630The commands defined in this section implement MI support for
33631tracepoints. For detailed introduction, see @ref{Tracepoints}.
33632
33633@subheading The @code{-trace-find} Command
33634@findex -trace-find
33635
33636@subsubheading Synopsis
33637
33638@smallexample
33639 -trace-find @var{mode} [@var{parameters}@dots{}]
33640@end smallexample
33641
33642Find a trace frame using criteria defined by @var{mode} and
33643@var{parameters}. The following table lists permissible
33644modes and their parameters. For details of operation, see @ref{tfind}.
33645
33646@table @samp
33647
33648@item none
33649No parameters are required. Stops examining trace frames.
33650
33651@item frame-number
33652An integer is required as parameter. Selects tracepoint frame with
33653that index.
33654
33655@item tracepoint-number
33656An integer is required as parameter. Finds next
33657trace frame that corresponds to tracepoint with the specified number.
33658
33659@item pc
33660An address is required as parameter. Finds
33661next trace frame that corresponds to any tracepoint at the specified
33662address.
33663
33664@item pc-inside-range
33665Two addresses are required as parameters. Finds next trace
33666frame that corresponds to a tracepoint at an address inside the
33667specified range. Both bounds are considered to be inside the range.
33668
33669@item pc-outside-range
33670Two addresses are required as parameters. Finds
33671next trace frame that corresponds to a tracepoint at an address outside
33672the specified range. Both bounds are considered to be inside the range.
33673
33674@item line
33675Line specification is required as parameter. @xref{Specify Location}.
33676Finds next trace frame that corresponds to a tracepoint at
33677the specified location.
33678
33679@end table
33680
33681If @samp{none} was passed as @var{mode}, the response does not
33682have fields. Otherwise, the response may have the following fields:
33683
33684@table @samp
33685@item found
33686This field has either @samp{0} or @samp{1} as the value, depending
33687on whether a matching tracepoint was found.
33688
33689@item traceframe
33690The index of the found traceframe. This field is present iff
33691the @samp{found} field has value of @samp{1}.
33692
33693@item tracepoint
33694The index of the found tracepoint. This field is present iff
33695the @samp{found} field has value of @samp{1}.
33696
33697@item frame
33698The information about the frame corresponding to the found trace
33699frame. This field is present only if a trace frame was found.
33700@xref{GDB/MI Frame Information}, for description of this field.
33701
33702@end table
33703
33704@subsubheading @value{GDBN} Command
33705
33706The corresponding @value{GDBN} command is @samp{tfind}.
33707
33708@subheading -trace-define-variable
33709@findex -trace-define-variable
33710
33711@subsubheading Synopsis
33712
33713@smallexample
33714 -trace-define-variable @var{name} [ @var{value} ]
33715@end smallexample
33716
33717Create trace variable @var{name} if it does not exist. If
33718@var{value} is specified, sets the initial value of the specified
33719trace variable to that value. Note that the @var{name} should start
33720with the @samp{$} character.
33721
33722@subsubheading @value{GDBN} Command
33723
33724The corresponding @value{GDBN} command is @samp{tvariable}.
33725
33726@subheading The @code{-trace-frame-collected} Command
33727@findex -trace-frame-collected
33728
33729@subsubheading Synopsis
33730
33731@smallexample
33732 -trace-frame-collected
33733 [--var-print-values @var{var_pval}]
33734 [--comp-print-values @var{comp_pval}]
33735 [--registers-format @var{regformat}]
33736 [--memory-contents]
33737@end smallexample
33738
33739This command returns the set of collected objects, register names,
33740trace state variable names, memory ranges and computed expressions
33741that have been collected at a particular trace frame. The optional
33742parameters to the command affect the output format in different ways.
33743See the output description table below for more details.
33744
33745The reported names can be used in the normal manner to create
33746varobjs and inspect the objects themselves. The items returned by
33747this command are categorized so that it is clear which is a variable,
33748which is a register, which is a trace state variable, which is a
33749memory range and which is a computed expression.
33750
33751For instance, if the actions were
33752@smallexample
33753collect myVar, myArray[myIndex], myObj.field, myPtr->field, myCount + 2
33754collect *(int*)0xaf02bef0@@40
33755@end smallexample
33756
33757@noindent
33758the object collected in its entirety would be @code{myVar}. The
33759object @code{myArray} would be partially collected, because only the
33760element at index @code{myIndex} would be collected. The remaining
33761objects would be computed expressions.
33762
33763An example output would be:
33764
33765@smallexample
33766(gdb)
33767-trace-frame-collected
33768^done,
33769 explicit-variables=[@{name="myVar",value="1"@}],
33770 computed-expressions=[@{name="myArray[myIndex]",value="0"@},
33771 @{name="myObj.field",value="0"@},
33772 @{name="myPtr->field",value="1"@},
33773 @{name="myCount + 2",value="3"@},
33774 @{name="$tvar1 + 1",value="43970027"@}],
33775 registers=[@{number="0",value="0x7fe2c6e79ec8"@},
33776 @{number="1",value="0x0"@},
33777 @{number="2",value="0x4"@},
33778 ...
33779 @{number="125",value="0x0"@}],
33780 tvars=[@{name="$tvar1",current="43970026"@}],
33781 memory=[@{address="0x0000000000602264",length="4"@},
33782 @{address="0x0000000000615bc0",length="4"@}]
33783(gdb)
33784@end smallexample
33785
33786Where:
33787
33788@table @code
33789@item explicit-variables
33790The set of objects that have been collected in their entirety (as
33791opposed to collecting just a few elements of an array or a few struct
33792members). For each object, its name and value are printed.
33793The @code{--var-print-values} option affects how or whether the value
33794field is output. If @var{var_pval} is 0, then print only the names;
33795if it is 1, print also their values; and if it is 2, print the name,
33796type and value for simple data types, and the name and type for
33797arrays, structures and unions.
33798
33799@item computed-expressions
33800The set of computed expressions that have been collected at the
33801current trace frame. The @code{--comp-print-values} option affects
33802this set like the @code{--var-print-values} option affects the
33803@code{explicit-variables} set. See above.
33804
33805@item registers
33806The registers that have been collected at the current trace frame.
33807For each register collected, the name and current value are returned.
33808The value is formatted according to the @code{--registers-format}
33809option. See the @command{-data-list-register-values} command for a
33810list of the allowed formats. The default is @samp{x}.
33811
33812@item tvars
33813The trace state variables that have been collected at the current
33814trace frame. For each trace state variable collected, the name and
33815current value are returned.
33816
33817@item memory
33818The set of memory ranges that have been collected at the current trace
33819frame. Its content is a list of tuples. Each tuple represents a
33820collected memory range and has the following fields:
33821
33822@table @code
33823@item address
33824The start address of the memory range, as hexadecimal literal.
33825
33826@item length
33827The length of the memory range, as decimal literal.
33828
33829@item contents
33830The contents of the memory block, in hex. This field is only present
33831if the @code{--memory-contents} option is specified.
33832
33833@end table
33834
33835@end table
33836
33837@subsubheading @value{GDBN} Command
33838
33839There is no corresponding @value{GDBN} command.
33840
33841@subsubheading Example
33842
33843@subheading -trace-list-variables
33844@findex -trace-list-variables
33845
33846@subsubheading Synopsis
33847
33848@smallexample
33849 -trace-list-variables
33850@end smallexample
33851
33852Return a table of all defined trace variables. Each element of the
33853table has the following fields:
33854
33855@table @samp
33856@item name
33857The name of the trace variable. This field is always present.
33858
33859@item initial
33860The initial value. This is a 64-bit signed integer. This
33861field is always present.
33862
33863@item current
33864The value the trace variable has at the moment. This is a 64-bit
33865signed integer. This field is absent iff current value is
33866not defined, for example if the trace was never run, or is
33867presently running.
33868
33869@end table
33870
33871@subsubheading @value{GDBN} Command
33872
33873The corresponding @value{GDBN} command is @samp{tvariables}.
33874
33875@subsubheading Example
33876
33877@smallexample
33878(gdb)
33879-trace-list-variables
33880^done,trace-variables=@{nr_rows="1",nr_cols="3",
33881hdr=[@{width="15",alignment="-1",col_name="name",colhdr="Name"@},
33882 @{width="11",alignment="-1",col_name="initial",colhdr="Initial"@},
33883 @{width="11",alignment="-1",col_name="current",colhdr="Current"@}],
33884body=[variable=@{name="$trace_timestamp",initial="0"@}
33885 variable=@{name="$foo",initial="10",current="15"@}]@}
33886(gdb)
33887@end smallexample
33888
33889@subheading -trace-save
33890@findex -trace-save
33891
33892@subsubheading Synopsis
33893
33894@smallexample
33895 -trace-save [-r ] @var{filename}
33896@end smallexample
33897
33898Saves the collected trace data to @var{filename}. Without the
33899@samp{-r} option, the data is downloaded from the target and saved
33900in a local file. With the @samp{-r} option the target is asked
33901to perform the save.
33902
33903@subsubheading @value{GDBN} Command
33904
33905The corresponding @value{GDBN} command is @samp{tsave}.
33906
33907
33908@subheading -trace-start
33909@findex -trace-start
33910
33911@subsubheading Synopsis
33912
33913@smallexample
33914 -trace-start
33915@end smallexample
33916
33917Starts a tracing experiments. The result of this command does not
33918have any fields.
33919
33920@subsubheading @value{GDBN} Command
33921
33922The corresponding @value{GDBN} command is @samp{tstart}.
33923
33924@subheading -trace-status
33925@findex -trace-status
33926
33927@subsubheading Synopsis
33928
33929@smallexample
33930 -trace-status
33931@end smallexample
33932
33933Obtains the status of a tracing experiment. The result may include
33934the following fields:
33935
33936@table @samp
33937
33938@item supported
33939May have a value of either @samp{0}, when no tracing operations are
33940supported, @samp{1}, when all tracing operations are supported, or
33941@samp{file} when examining trace file. In the latter case, examining
33942of trace frame is possible but new tracing experiement cannot be
33943started. This field is always present.
33944
33945@item running
33946May have a value of either @samp{0} or @samp{1} depending on whether
33947tracing experiement is in progress on target. This field is present
33948if @samp{supported} field is not @samp{0}.
33949
33950@item stop-reason
33951Report the reason why the tracing was stopped last time. This field
33952may be absent iff tracing was never stopped on target yet. The
33953value of @samp{request} means the tracing was stopped as result of
33954the @code{-trace-stop} command. The value of @samp{overflow} means
33955the tracing buffer is full. The value of @samp{disconnection} means
33956tracing was automatically stopped when @value{GDBN} has disconnected.
33957The value of @samp{passcount} means tracing was stopped when a
33958tracepoint was passed a maximal number of times for that tracepoint.
33959This field is present if @samp{supported} field is not @samp{0}.
33960
33961@item stopping-tracepoint
33962The number of tracepoint whose passcount as exceeded. This field is
33963present iff the @samp{stop-reason} field has the value of
33964@samp{passcount}.
33965
33966@item frames
33967@itemx frames-created
33968The @samp{frames} field is a count of the total number of trace frames
33969in the trace buffer, while @samp{frames-created} is the total created
33970during the run, including ones that were discarded, such as when a
33971circular trace buffer filled up. Both fields are optional.
33972
33973@item buffer-size
33974@itemx buffer-free
33975These fields tell the current size of the tracing buffer and the
33976remaining space. These fields are optional.
33977
33978@item circular
33979The value of the circular trace buffer flag. @code{1} means that the
33980trace buffer is circular and old trace frames will be discarded if
33981necessary to make room, @code{0} means that the trace buffer is linear
33982and may fill up.
33983
33984@item disconnected
33985The value of the disconnected tracing flag. @code{1} means that
33986tracing will continue after @value{GDBN} disconnects, @code{0} means
33987that the trace run will stop.
33988
33989@item trace-file
33990The filename of the trace file being examined. This field is
33991optional, and only present when examining a trace file.
33992
33993@end table
33994
33995@subsubheading @value{GDBN} Command
33996
33997The corresponding @value{GDBN} command is @samp{tstatus}.
33998
33999@subheading -trace-stop
34000@findex -trace-stop
34001
34002@subsubheading Synopsis
34003
34004@smallexample
34005 -trace-stop
34006@end smallexample
34007
34008Stops a tracing experiment. The result of this command has the same
34009fields as @code{-trace-status}, except that the @samp{supported} and
34010@samp{running} fields are not output.
34011
34012@subsubheading @value{GDBN} Command
34013
34014The corresponding @value{GDBN} command is @samp{tstop}.
34015
34016
34017@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34018@node GDB/MI Symbol Query
34019@section @sc{gdb/mi} Symbol Query Commands
34020
34021
34022@ignore
34023@subheading The @code{-symbol-info-address} Command
34024@findex -symbol-info-address
34025
34026@subsubheading Synopsis
34027
34028@smallexample
34029 -symbol-info-address @var{symbol}
34030@end smallexample
34031
34032Describe where @var{symbol} is stored.
34033
34034@subsubheading @value{GDBN} Command
34035
34036The corresponding @value{GDBN} command is @samp{info address}.
34037
34038@subsubheading Example
34039N.A.
34040
34041
34042@subheading The @code{-symbol-info-file} Command
34043@findex -symbol-info-file
34044
34045@subsubheading Synopsis
34046
34047@smallexample
34048 -symbol-info-file
34049@end smallexample
34050
34051Show the file for the symbol.
34052
34053@subsubheading @value{GDBN} Command
34054
34055There's no equivalent @value{GDBN} command. @code{gdbtk} has
34056@samp{gdb_find_file}.
34057
34058@subsubheading Example
34059N.A.
34060
34061
34062@subheading The @code{-symbol-info-function} Command
34063@findex -symbol-info-function
34064
34065@subsubheading Synopsis
34066
34067@smallexample
34068 -symbol-info-function
34069@end smallexample
34070
34071Show which function the symbol lives in.
34072
34073@subsubheading @value{GDBN} Command
34074
34075@samp{gdb_get_function} in @code{gdbtk}.
34076
34077@subsubheading Example
34078N.A.
34079
34080
34081@subheading The @code{-symbol-info-line} Command
34082@findex -symbol-info-line
34083
34084@subsubheading Synopsis
34085
34086@smallexample
34087 -symbol-info-line
34088@end smallexample
34089
34090Show the core addresses of the code for a source line.
34091
34092@subsubheading @value{GDBN} Command
34093
34094The corresponding @value{GDBN} command is @samp{info line}.
34095@code{gdbtk} has the @samp{gdb_get_line} and @samp{gdb_get_file} commands.
34096
34097@subsubheading Example
34098N.A.
34099
34100
34101@subheading The @code{-symbol-info-symbol} Command
34102@findex -symbol-info-symbol
34103
34104@subsubheading Synopsis
34105
34106@smallexample
34107 -symbol-info-symbol @var{addr}
34108@end smallexample
34109
34110Describe what symbol is at location @var{addr}.
34111
34112@subsubheading @value{GDBN} Command
34113
34114The corresponding @value{GDBN} command is @samp{info symbol}.
34115
34116@subsubheading Example
34117N.A.
34118
34119
34120@subheading The @code{-symbol-list-functions} Command
34121@findex -symbol-list-functions
34122
34123@subsubheading Synopsis
34124
34125@smallexample
34126 -symbol-list-functions
34127@end smallexample
34128
34129List the functions in the executable.
34130
34131@subsubheading @value{GDBN} Command
34132
34133@samp{info functions} in @value{GDBN}, @samp{gdb_listfunc} and
34134@samp{gdb_search} in @code{gdbtk}.
34135
34136@subsubheading Example
34137N.A.
34138@end ignore
34139
34140
34141@subheading The @code{-symbol-list-lines} Command
34142@findex -symbol-list-lines
34143
34144@subsubheading Synopsis
34145
34146@smallexample
34147 -symbol-list-lines @var{filename}
34148@end smallexample
34149
34150Print the list of lines that contain code and their associated program
34151addresses for the given source filename. The entries are sorted in
34152ascending PC order.
34153
34154@subsubheading @value{GDBN} Command
34155
34156There is no corresponding @value{GDBN} command.
34157
34158@subsubheading Example
34159@smallexample
34160(gdb)
34161-symbol-list-lines basics.c
34162^done,lines=[@{pc="0x08048554",line="7"@},@{pc="0x0804855a",line="8"@}]
34163(gdb)
34164@end smallexample
34165
34166
34167@ignore
34168@subheading The @code{-symbol-list-types} Command
34169@findex -symbol-list-types
34170
34171@subsubheading Synopsis
34172
34173@smallexample
34174 -symbol-list-types
34175@end smallexample
34176
34177List all the type names.
34178
34179@subsubheading @value{GDBN} Command
34180
34181The corresponding commands are @samp{info types} in @value{GDBN},
34182@samp{gdb_search} in @code{gdbtk}.
34183
34184@subsubheading Example
34185N.A.
34186
34187
34188@subheading The @code{-symbol-list-variables} Command
34189@findex -symbol-list-variables
34190
34191@subsubheading Synopsis
34192
34193@smallexample
34194 -symbol-list-variables
34195@end smallexample
34196
34197List all the global and static variable names.
34198
34199@subsubheading @value{GDBN} Command
34200
34201@samp{info variables} in @value{GDBN}, @samp{gdb_search} in @code{gdbtk}.
34202
34203@subsubheading Example
34204N.A.
34205
34206
34207@subheading The @code{-symbol-locate} Command
34208@findex -symbol-locate
34209
34210@subsubheading Synopsis
34211
34212@smallexample
34213 -symbol-locate
34214@end smallexample
34215
34216@subsubheading @value{GDBN} Command
34217
34218@samp{gdb_loc} in @code{gdbtk}.
34219
34220@subsubheading Example
34221N.A.
34222
34223
34224@subheading The @code{-symbol-type} Command
34225@findex -symbol-type
34226
34227@subsubheading Synopsis
34228
34229@smallexample
34230 -symbol-type @var{variable}
34231@end smallexample
34232
34233Show type of @var{variable}.
34234
34235@subsubheading @value{GDBN} Command
34236
34237The corresponding @value{GDBN} command is @samp{ptype}, @code{gdbtk} has
34238@samp{gdb_obj_variable}.
34239
34240@subsubheading Example
34241N.A.
34242@end ignore
34243
34244
34245@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34246@node GDB/MI File Commands
34247@section @sc{gdb/mi} File Commands
34248
34249This section describes the GDB/MI commands to specify executable file names
34250and to read in and obtain symbol table information.
34251
34252@subheading The @code{-file-exec-and-symbols} Command
34253@findex -file-exec-and-symbols
34254
34255@subsubheading Synopsis
34256
34257@smallexample
34258 -file-exec-and-symbols @var{file}
34259@end smallexample
34260
34261Specify the executable file to be debugged. This file is the one from
34262which the symbol table is also read. If no file is specified, the
34263command clears the executable and symbol information. If breakpoints
34264are set when using this command with no arguments, @value{GDBN} will produce
34265error messages. Otherwise, no output is produced, except a completion
34266notification.
34267
34268@subsubheading @value{GDBN} Command
34269
34270The corresponding @value{GDBN} command is @samp{file}.
34271
34272@subsubheading Example
34273
34274@smallexample
34275(gdb)
34276-file-exec-and-symbols /kwikemart/marge/ezannoni/TRUNK/mbx/hello.mbx
34277^done
34278(gdb)
34279@end smallexample
34280
34281
34282@subheading The @code{-file-exec-file} Command
34283@findex -file-exec-file
34284
34285@subsubheading Synopsis
34286
34287@smallexample
34288 -file-exec-file @var{file}
34289@end smallexample
34290
34291Specify the executable file to be debugged. Unlike
34292@samp{-file-exec-and-symbols}, the symbol table is @emph{not} read
34293from this file. If used without argument, @value{GDBN} clears the information
34294about the executable file. No output is produced, except a completion
34295notification.
34296
34297@subsubheading @value{GDBN} Command
34298
34299The corresponding @value{GDBN} command is @samp{exec-file}.
34300
34301@subsubheading Example
34302
34303@smallexample
34304(gdb)
34305-file-exec-file /kwikemart/marge/ezannoni/TRUNK/mbx/hello.mbx
34306^done
34307(gdb)
34308@end smallexample
34309
34310
34311@ignore
34312@subheading The @code{-file-list-exec-sections} Command
34313@findex -file-list-exec-sections
34314
34315@subsubheading Synopsis
34316
34317@smallexample
34318 -file-list-exec-sections
34319@end smallexample
34320
34321List the sections of the current executable file.
34322
34323@subsubheading @value{GDBN} Command
34324
34325The @value{GDBN} command @samp{info file} shows, among the rest, the same
34326information as this command. @code{gdbtk} has a corresponding command
34327@samp{gdb_load_info}.
34328
34329@subsubheading Example
34330N.A.
34331@end ignore
34332
34333
34334@subheading The @code{-file-list-exec-source-file} Command
34335@findex -file-list-exec-source-file
34336
34337@subsubheading Synopsis
34338
34339@smallexample
34340 -file-list-exec-source-file
34341@end smallexample
34342
34343List the line number, the current source file, and the absolute path
34344to the current source file for the current executable. The macro
34345information field has a value of @samp{1} or @samp{0} depending on
34346whether or not the file includes preprocessor macro information.
34347
34348@subsubheading @value{GDBN} Command
34349
34350The @value{GDBN} equivalent is @samp{info source}
34351
34352@subsubheading Example
34353
34354@smallexample
34355(gdb)
34356123-file-list-exec-source-file
34357123^done,line="1",file="foo.c",fullname="/home/bar/foo.c,macro-info="1"
34358(gdb)
34359@end smallexample
34360
34361
34362@subheading The @code{-file-list-exec-source-files} Command
34363@findex -file-list-exec-source-files
34364
34365@subsubheading Synopsis
34366
34367@smallexample
34368 -file-list-exec-source-files
34369@end smallexample
34370
34371List the source files for the current executable.
34372
34373It will always output both the filename and fullname (absolute file
34374name) of a source file.
34375
34376@subsubheading @value{GDBN} Command
34377
34378The @value{GDBN} equivalent is @samp{info sources}.
34379@code{gdbtk} has an analogous command @samp{gdb_listfiles}.
34380
34381@subsubheading Example
34382@smallexample
34383(gdb)
34384-file-list-exec-source-files
34385^done,files=[
34386@{file=foo.c,fullname=/home/foo.c@},
34387@{file=/home/bar.c,fullname=/home/bar.c@},
34388@{file=gdb_could_not_find_fullpath.c@}]
34389(gdb)
34390@end smallexample
34391
34392@ignore
34393@subheading The @code{-file-list-shared-libraries} Command
34394@findex -file-list-shared-libraries
34395
34396@subsubheading Synopsis
34397
34398@smallexample
34399 -file-list-shared-libraries
34400@end smallexample
34401
34402List the shared libraries in the program.
34403
34404@subsubheading @value{GDBN} Command
34405
34406The corresponding @value{GDBN} command is @samp{info shared}.
34407
34408@subsubheading Example
34409N.A.
34410
34411
34412@subheading The @code{-file-list-symbol-files} Command
34413@findex -file-list-symbol-files
34414
34415@subsubheading Synopsis
34416
34417@smallexample
34418 -file-list-symbol-files
34419@end smallexample
34420
34421List symbol files.
34422
34423@subsubheading @value{GDBN} Command
34424
34425The corresponding @value{GDBN} command is @samp{info file} (part of it).
34426
34427@subsubheading Example
34428N.A.
34429@end ignore
34430
34431
34432@subheading The @code{-file-symbol-file} Command
34433@findex -file-symbol-file
34434
34435@subsubheading Synopsis
34436
34437@smallexample
34438 -file-symbol-file @var{file}
34439@end smallexample
34440
34441Read symbol table info from the specified @var{file} argument. When
34442used without arguments, clears @value{GDBN}'s symbol table info. No output is
34443produced, except for a completion notification.
34444
34445@subsubheading @value{GDBN} Command
34446
34447The corresponding @value{GDBN} command is @samp{symbol-file}.
34448
34449@subsubheading Example
34450
34451@smallexample
34452(gdb)
34453-file-symbol-file /kwikemart/marge/ezannoni/TRUNK/mbx/hello.mbx
34454^done
34455(gdb)
34456@end smallexample
34457
34458@ignore
34459@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34460@node GDB/MI Memory Overlay Commands
34461@section @sc{gdb/mi} Memory Overlay Commands
34462
34463The memory overlay commands are not implemented.
34464
34465@c @subheading -overlay-auto
34466
34467@c @subheading -overlay-list-mapping-state
34468
34469@c @subheading -overlay-list-overlays
34470
34471@c @subheading -overlay-map
34472
34473@c @subheading -overlay-off
34474
34475@c @subheading -overlay-on
34476
34477@c @subheading -overlay-unmap
34478
34479@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34480@node GDB/MI Signal Handling Commands
34481@section @sc{gdb/mi} Signal Handling Commands
34482
34483Signal handling commands are not implemented.
34484
34485@c @subheading -signal-handle
34486
34487@c @subheading -signal-list-handle-actions
34488
34489@c @subheading -signal-list-signal-types
34490@end ignore
34491
34492
34493@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34494@node GDB/MI Target Manipulation
34495@section @sc{gdb/mi} Target Manipulation Commands
34496
34497
34498@subheading The @code{-target-attach} Command
34499@findex -target-attach
34500
34501@subsubheading Synopsis
34502
34503@smallexample
34504 -target-attach @var{pid} | @var{gid} | @var{file}
34505@end smallexample
34506
34507Attach to a process @var{pid} or a file @var{file} outside of
34508@value{GDBN}, or a thread group @var{gid}. If attaching to a thread
34509group, the id previously returned by
34510@samp{-list-thread-groups --available} must be used.
34511
34512@subsubheading @value{GDBN} Command
34513
34514The corresponding @value{GDBN} command is @samp{attach}.
34515
34516@subsubheading Example
34517@smallexample
34518(gdb)
34519-target-attach 34
34520=thread-created,id="1"
34521*stopped,thread-id="1",frame=@{addr="0xb7f7e410",func="bar",args=[]@}
34522^done
34523(gdb)
34524@end smallexample
34525
34526@ignore
34527@subheading The @code{-target-compare-sections} Command
34528@findex -target-compare-sections
34529
34530@subsubheading Synopsis
34531
34532@smallexample
34533 -target-compare-sections [ @var{section} ]
34534@end smallexample
34535
34536Compare data of section @var{section} on target to the exec file.
34537Without the argument, all sections are compared.
34538
34539@subsubheading @value{GDBN} Command
34540
34541The @value{GDBN} equivalent is @samp{compare-sections}.
34542
34543@subsubheading Example
34544N.A.
34545@end ignore
34546
34547
34548@subheading The @code{-target-detach} Command
34549@findex -target-detach
34550
34551@subsubheading Synopsis
34552
34553@smallexample
34554 -target-detach [ @var{pid} | @var{gid} ]
34555@end smallexample
34556
34557Detach from the remote target which normally resumes its execution.
34558If either @var{pid} or @var{gid} is specified, detaches from either
34559the specified process, or specified thread group. There's no output.
34560
34561@subsubheading @value{GDBN} Command
34562
34563The corresponding @value{GDBN} command is @samp{detach}.
34564
34565@subsubheading Example
34566
34567@smallexample
34568(gdb)
34569-target-detach
34570^done
34571(gdb)
34572@end smallexample
34573
34574
34575@subheading The @code{-target-disconnect} Command
34576@findex -target-disconnect
34577
34578@subsubheading Synopsis
34579
34580@smallexample
34581 -target-disconnect
34582@end smallexample
34583
34584Disconnect from the remote target. There's no output and the target is
34585generally not resumed.
34586
34587@subsubheading @value{GDBN} Command
34588
34589The corresponding @value{GDBN} command is @samp{disconnect}.
34590
34591@subsubheading Example
34592
34593@smallexample
34594(gdb)
34595-target-disconnect
34596^done
34597(gdb)
34598@end smallexample
34599
34600
34601@subheading The @code{-target-download} Command
34602@findex -target-download
34603
34604@subsubheading Synopsis
34605
34606@smallexample
34607 -target-download
34608@end smallexample
34609
34610Loads the executable onto the remote target.
34611It prints out an update message every half second, which includes the fields:
34612
34613@table @samp
34614@item section
34615The name of the section.
34616@item section-sent
34617The size of what has been sent so far for that section.
34618@item section-size
34619The size of the section.
34620@item total-sent
34621The total size of what was sent so far (the current and the previous sections).
34622@item total-size
34623The size of the overall executable to download.
34624@end table
34625
34626@noindent
34627Each message is sent as status record (@pxref{GDB/MI Output Syntax, ,
34628@sc{gdb/mi} Output Syntax}).
34629
34630In addition, it prints the name and size of the sections, as they are
34631downloaded. These messages include the following fields:
34632
34633@table @samp
34634@item section
34635The name of the section.
34636@item section-size
34637The size of the section.
34638@item total-size
34639The size of the overall executable to download.
34640@end table
34641
34642@noindent
34643At the end, a summary is printed.
34644
34645@subsubheading @value{GDBN} Command
34646
34647The corresponding @value{GDBN} command is @samp{load}.
34648
34649@subsubheading Example
34650
34651Note: each status message appears on a single line. Here the messages
34652have been broken down so that they can fit onto a page.
34653
34654@smallexample
34655(gdb)
34656-target-download
34657+download,@{section=".text",section-size="6668",total-size="9880"@}
34658+download,@{section=".text",section-sent="512",section-size="6668",
34659total-sent="512",total-size="9880"@}
34660+download,@{section=".text",section-sent="1024",section-size="6668",
34661total-sent="1024",total-size="9880"@}
34662+download,@{section=".text",section-sent="1536",section-size="6668",
34663total-sent="1536",total-size="9880"@}
34664+download,@{section=".text",section-sent="2048",section-size="6668",
34665total-sent="2048",total-size="9880"@}
34666+download,@{section=".text",section-sent="2560",section-size="6668",
34667total-sent="2560",total-size="9880"@}
34668+download,@{section=".text",section-sent="3072",section-size="6668",
34669total-sent="3072",total-size="9880"@}
34670+download,@{section=".text",section-sent="3584",section-size="6668",
34671total-sent="3584",total-size="9880"@}
34672+download,@{section=".text",section-sent="4096",section-size="6668",
34673total-sent="4096",total-size="9880"@}
34674+download,@{section=".text",section-sent="4608",section-size="6668",
34675total-sent="4608",total-size="9880"@}
34676+download,@{section=".text",section-sent="5120",section-size="6668",
34677total-sent="5120",total-size="9880"@}
34678+download,@{section=".text",section-sent="5632",section-size="6668",
34679total-sent="5632",total-size="9880"@}
34680+download,@{section=".text",section-sent="6144",section-size="6668",
34681total-sent="6144",total-size="9880"@}
34682+download,@{section=".text",section-sent="6656",section-size="6668",
34683total-sent="6656",total-size="9880"@}
34684+download,@{section=".init",section-size="28",total-size="9880"@}
34685+download,@{section=".fini",section-size="28",total-size="9880"@}
34686+download,@{section=".data",section-size="3156",total-size="9880"@}
34687+download,@{section=".data",section-sent="512",section-size="3156",
34688total-sent="7236",total-size="9880"@}
34689+download,@{section=".data",section-sent="1024",section-size="3156",
34690total-sent="7748",total-size="9880"@}
34691+download,@{section=".data",section-sent="1536",section-size="3156",
34692total-sent="8260",total-size="9880"@}
34693+download,@{section=".data",section-sent="2048",section-size="3156",
34694total-sent="8772",total-size="9880"@}
34695+download,@{section=".data",section-sent="2560",section-size="3156",
34696total-sent="9284",total-size="9880"@}
34697+download,@{section=".data",section-sent="3072",section-size="3156",
34698total-sent="9796",total-size="9880"@}
34699^done,address="0x10004",load-size="9880",transfer-rate="6586",
34700write-rate="429"
34701(gdb)
34702@end smallexample
34703
34704
34705@ignore
34706@subheading The @code{-target-exec-status} Command
34707@findex -target-exec-status
34708
34709@subsubheading Synopsis
34710
34711@smallexample
34712 -target-exec-status
34713@end smallexample
34714
34715Provide information on the state of the target (whether it is running or
34716not, for instance).
34717
34718@subsubheading @value{GDBN} Command
34719
34720There's no equivalent @value{GDBN} command.
34721
34722@subsubheading Example
34723N.A.
34724
34725
34726@subheading The @code{-target-list-available-targets} Command
34727@findex -target-list-available-targets
34728
34729@subsubheading Synopsis
34730
34731@smallexample
34732 -target-list-available-targets
34733@end smallexample
34734
34735List the possible targets to connect to.
34736
34737@subsubheading @value{GDBN} Command
34738
34739The corresponding @value{GDBN} command is @samp{help target}.
34740
34741@subsubheading Example
34742N.A.
34743
34744
34745@subheading The @code{-target-list-current-targets} Command
34746@findex -target-list-current-targets
34747
34748@subsubheading Synopsis
34749
34750@smallexample
34751 -target-list-current-targets
34752@end smallexample
34753
34754Describe the current target.
34755
34756@subsubheading @value{GDBN} Command
34757
34758The corresponding information is printed by @samp{info file} (among
34759other things).
34760
34761@subsubheading Example
34762N.A.
34763
34764
34765@subheading The @code{-target-list-parameters} Command
34766@findex -target-list-parameters
34767
34768@subsubheading Synopsis
34769
34770@smallexample
34771 -target-list-parameters
34772@end smallexample
34773
34774@c ????
34775@end ignore
34776
34777@subsubheading @value{GDBN} Command
34778
34779No equivalent.
34780
34781@subsubheading Example
34782N.A.
34783
34784
34785@subheading The @code{-target-select} Command
34786@findex -target-select
34787
34788@subsubheading Synopsis
34789
34790@smallexample
34791 -target-select @var{type} @var{parameters @dots{}}
34792@end smallexample
34793
34794Connect @value{GDBN} to the remote target. This command takes two args:
34795
34796@table @samp
34797@item @var{type}
34798The type of target, for instance @samp{remote}, etc.
34799@item @var{parameters}
34800Device names, host names and the like. @xref{Target Commands, ,
34801Commands for Managing Targets}, for more details.
34802@end table
34803
34804The output is a connection notification, followed by the address at
34805which the target program is, in the following form:
34806
34807@smallexample
34808^connected,addr="@var{address}",func="@var{function name}",
34809 args=[@var{arg list}]
34810@end smallexample
34811
34812@subsubheading @value{GDBN} Command
34813
34814The corresponding @value{GDBN} command is @samp{target}.
34815
34816@subsubheading Example
34817
34818@smallexample
34819(gdb)
34820-target-select remote /dev/ttya
34821^connected,addr="0xfe00a300",func="??",args=[]
34822(gdb)
34823@end smallexample
34824
34825@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34826@node GDB/MI File Transfer Commands
34827@section @sc{gdb/mi} File Transfer Commands
34828
34829
34830@subheading The @code{-target-file-put} Command
34831@findex -target-file-put
34832
34833@subsubheading Synopsis
34834
34835@smallexample
34836 -target-file-put @var{hostfile} @var{targetfile}
34837@end smallexample
34838
34839Copy file @var{hostfile} from the host system (the machine running
34840@value{GDBN}) to @var{targetfile} on the target system.
34841
34842@subsubheading @value{GDBN} Command
34843
34844The corresponding @value{GDBN} command is @samp{remote put}.
34845
34846@subsubheading Example
34847
34848@smallexample
34849(gdb)
34850-target-file-put localfile remotefile
34851^done
34852(gdb)
34853@end smallexample
34854
34855
34856@subheading The @code{-target-file-get} Command
34857@findex -target-file-get
34858
34859@subsubheading Synopsis
34860
34861@smallexample
34862 -target-file-get @var{targetfile} @var{hostfile}
34863@end smallexample
34864
34865Copy file @var{targetfile} from the target system to @var{hostfile}
34866on the host system.
34867
34868@subsubheading @value{GDBN} Command
34869
34870The corresponding @value{GDBN} command is @samp{remote get}.
34871
34872@subsubheading Example
34873
34874@smallexample
34875(gdb)
34876-target-file-get remotefile localfile
34877^done
34878(gdb)
34879@end smallexample
34880
34881
34882@subheading The @code{-target-file-delete} Command
34883@findex -target-file-delete
34884
34885@subsubheading Synopsis
34886
34887@smallexample
34888 -target-file-delete @var{targetfile}
34889@end smallexample
34890
34891Delete @var{targetfile} from the target system.
34892
34893@subsubheading @value{GDBN} Command
34894
34895The corresponding @value{GDBN} command is @samp{remote delete}.
34896
34897@subsubheading Example
34898
34899@smallexample
34900(gdb)
34901-target-file-delete remotefile
34902^done
34903(gdb)
34904@end smallexample
34905
34906
34907@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34908@node GDB/MI Ada Exceptions Commands
34909@section Ada Exceptions @sc{gdb/mi} Commands
34910
34911@subheading The @code{-info-ada-exceptions} Command
34912@findex -info-ada-exceptions
34913
34914@subsubheading Synopsis
34915
34916@smallexample
34917 -info-ada-exceptions [ @var{regexp}]
34918@end smallexample
34919
34920List all Ada exceptions defined within the program being debugged.
34921With a regular expression @var{regexp}, only those exceptions whose
34922names match @var{regexp} are listed.
34923
34924@subsubheading @value{GDBN} Command
34925
34926The corresponding @value{GDBN} command is @samp{info exceptions}.
34927
34928@subsubheading Result
34929
34930The result is a table of Ada exceptions. The following columns are
34931defined for each exception:
34932
34933@table @samp
34934@item name
34935The name of the exception.
34936
34937@item address
34938The address of the exception.
34939
34940@end table
34941
34942@subsubheading Example
34943
34944@smallexample
34945-info-ada-exceptions aint
34946^done,ada-exceptions=@{nr_rows="2",nr_cols="2",
34947hdr=[@{width="1",alignment="-1",col_name="name",colhdr="Name"@},
34948@{width="1",alignment="-1",col_name="address",colhdr="Address"@}],
34949body=[@{name="constraint_error",address="0x0000000000613da0"@},
34950@{name="const.aint_global_e",address="0x0000000000613b00"@}]@}
34951@end smallexample
34952
34953@subheading Catching Ada Exceptions
34954
34955The commands describing how to ask @value{GDBN} to stop when a program
34956raises an exception are described at @ref{Ada Exception GDB/MI
34957Catchpoint Commands}.
34958
34959
34960@c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34961@node GDB/MI Miscellaneous Commands
34962@section Miscellaneous @sc{gdb/mi} Commands
34963
34964@c @subheading -gdb-complete
34965
34966@subheading The @code{-gdb-exit} Command
34967@findex -gdb-exit
34968
34969@subsubheading Synopsis
34970
34971@smallexample
34972 -gdb-exit
34973@end smallexample
34974
34975Exit @value{GDBN} immediately.
34976
34977@subsubheading @value{GDBN} Command
34978
34979Approximately corresponds to @samp{quit}.
34980
34981@subsubheading Example
34982
34983@smallexample
34984(gdb)
34985-gdb-exit
34986^exit
34987@end smallexample
34988
34989
34990@ignore
34991@subheading The @code{-exec-abort} Command
34992@findex -exec-abort
34993
34994@subsubheading Synopsis
34995
34996@smallexample
34997 -exec-abort
34998@end smallexample
34999
35000Kill the inferior running program.
35001
35002@subsubheading @value{GDBN} Command
35003
35004The corresponding @value{GDBN} command is @samp{kill}.
35005
35006@subsubheading Example
35007N.A.
35008@end ignore
35009
35010
35011@subheading The @code{-gdb-set} Command
35012@findex -gdb-set
35013
35014@subsubheading Synopsis
35015
35016@smallexample
35017 -gdb-set
35018@end smallexample
35019
35020Set an internal @value{GDBN} variable.
35021@c IS THIS A DOLLAR VARIABLE? OR SOMETHING LIKE ANNOTATE ?????
35022
35023@subsubheading @value{GDBN} Command
35024
35025The corresponding @value{GDBN} command is @samp{set}.
35026
35027@subsubheading Example
35028
35029@smallexample
35030(gdb)
35031-gdb-set $foo=3
35032^done
35033(gdb)
35034@end smallexample
35035
35036
35037@subheading The @code{-gdb-show} Command
35038@findex -gdb-show
35039
35040@subsubheading Synopsis
35041
35042@smallexample
35043 -gdb-show
35044@end smallexample
35045
35046Show the current value of a @value{GDBN} variable.
35047
35048@subsubheading @value{GDBN} Command
35049
35050The corresponding @value{GDBN} command is @samp{show}.
35051
35052@subsubheading Example
35053
35054@smallexample
35055(gdb)
35056-gdb-show annotate
35057^done,value="0"
35058(gdb)
35059@end smallexample
35060
35061@c @subheading -gdb-source
35062
35063
35064@subheading The @code{-gdb-version} Command
35065@findex -gdb-version
35066
35067@subsubheading Synopsis
35068
35069@smallexample
35070 -gdb-version
35071@end smallexample
35072
35073Show version information for @value{GDBN}. Used mostly in testing.
35074
35075@subsubheading @value{GDBN} Command
35076
35077The @value{GDBN} equivalent is @samp{show version}. @value{GDBN} by
35078default shows this information when you start an interactive session.
35079
35080@subsubheading Example
35081
35082@c This example modifies the actual output from GDB to avoid overfull
35083@c box in TeX.
35084@smallexample
35085(gdb)
35086-gdb-version
35087~GNU gdb 5.2.1
35088~Copyright 2000 Free Software Foundation, Inc.
35089~GDB is free software, covered by the GNU General Public License, and
35090~you are welcome to change it and/or distribute copies of it under
35091~ certain conditions.
35092~Type "show copying" to see the conditions.
35093~There is absolutely no warranty for GDB. Type "show warranty" for
35094~ details.
35095~This GDB was configured as
35096 "--host=sparc-sun-solaris2.5.1 --target=ppc-eabi".
35097^done
35098(gdb)
35099@end smallexample
35100
35101@subheading The @code{-list-features} Command
35102@findex -list-features
35103
35104Returns a list of particular features of the MI protocol that
35105this version of gdb implements. A feature can be a command,
35106or a new field in an output of some command, or even an
35107important bugfix. While a frontend can sometimes detect presence
35108of a feature at runtime, it is easier to perform detection at debugger
35109startup.
35110
35111The command returns a list of strings, with each string naming an
35112available feature. Each returned string is just a name, it does not
35113have any internal structure. The list of possible feature names
35114is given below.
35115
35116Example output:
35117
35118@smallexample
35119(gdb) -list-features
35120^done,result=["feature1","feature2"]
35121@end smallexample
35122
35123The current list of features is:
35124
35125@table @samp
35126@item frozen-varobjs
35127Indicates support for the @code{-var-set-frozen} command, as well
35128as possible presense of the @code{frozen} field in the output
35129of @code{-varobj-create}.
35130@item pending-breakpoints
35131Indicates support for the @option{-f} option to the @code{-break-insert}
35132command.
35133@item python
35134Indicates Python scripting support, Python-based
35135pretty-printing commands, and possible presence of the
35136@samp{display_hint} field in the output of @code{-var-list-children}
35137@item thread-info
35138Indicates support for the @code{-thread-info} command.
35139@item data-read-memory-bytes
35140Indicates support for the @code{-data-read-memory-bytes} and the
35141@code{-data-write-memory-bytes} commands.
35142@item breakpoint-notifications
35143Indicates that changes to breakpoints and breakpoints created via the
35144CLI will be announced via async records.
35145@item ada-task-info
35146Indicates support for the @code{-ada-task-info} command.
35147@item ada-exceptions
35148Indicates support for the following commands, all of them related to Ada
35149exceptions: @code{-info-ada-exceptions}, @code{-catch-assert} and
35150@code{-catch-exception}.
35151@item language-option
35152Indicates that all @sc{gdb/mi} commands accept the @option{--language}
35153option (@pxref{Context management}).
35154@end table
35155
35156@subheading The @code{-list-target-features} Command
35157@findex -list-target-features
35158
35159Returns a list of particular features that are supported by the
35160target. Those features affect the permitted MI commands, but
35161unlike the features reported by the @code{-list-features} command, the
35162features depend on which target GDB is using at the moment. Whenever
35163a target can change, due to commands such as @code{-target-select},
35164@code{-target-attach} or @code{-exec-run}, the list of target features
35165may change, and the frontend should obtain it again.
35166Example output:
35167
35168@smallexample
35169(gdb) -list-target-features
35170^done,result=["async"]
35171@end smallexample
35172
35173The current list of features is:
35174
35175@table @samp
35176@item async
35177Indicates that the target is capable of asynchronous command
35178execution, which means that @value{GDBN} will accept further commands
35179while the target is running.
35180
35181@item reverse
35182Indicates that the target is capable of reverse execution.
35183@xref{Reverse Execution}, for more information.
35184
35185@end table
35186
35187@subheading The @code{-list-thread-groups} Command
35188@findex -list-thread-groups
35189
35190@subheading Synopsis
35191
35192@smallexample
35193-list-thread-groups [ --available ] [ --recurse 1 ] [ @var{group} ... ]
35194@end smallexample
35195
35196Lists thread groups (@pxref{Thread groups}). When a single thread
35197group is passed as the argument, lists the children of that group.
35198When several thread group are passed, lists information about those
35199thread groups. Without any parameters, lists information about all
35200top-level thread groups.
35201
35202Normally, thread groups that are being debugged are reported.
35203With the @samp{--available} option, @value{GDBN} reports thread groups
35204available on the target.
35205
35206The output of this command may have either a @samp{threads} result or
35207a @samp{groups} result. The @samp{thread} result has a list of tuples
35208as value, with each tuple describing a thread (@pxref{GDB/MI Thread
35209Information}). The @samp{groups} result has a list of tuples as value,
35210each tuple describing a thread group. If top-level groups are
35211requested (that is, no parameter is passed), or when several groups
35212are passed, the output always has a @samp{groups} result. The format
35213of the @samp{group} result is described below.
35214
35215To reduce the number of roundtrips it's possible to list thread groups
35216together with their children, by passing the @samp{--recurse} option
35217and the recursion depth. Presently, only recursion depth of 1 is
35218permitted. If this option is present, then every reported thread group
35219will also include its children, either as @samp{group} or
35220@samp{threads} field.
35221
35222In general, any combination of option and parameters is permitted, with
35223the following caveats:
35224
35225@itemize @bullet
35226@item
35227When a single thread group is passed, the output will typically
35228be the @samp{threads} result. Because threads may not contain
35229anything, the @samp{recurse} option will be ignored.
35230
35231@item
35232When the @samp{--available} option is passed, limited information may
35233be available. In particular, the list of threads of a process might
35234be inaccessible. Further, specifying specific thread groups might
35235not give any performance advantage over listing all thread groups.
35236The frontend should assume that @samp{-list-thread-groups --available}
35237is always an expensive operation and cache the results.
35238
35239@end itemize
35240
35241The @samp{groups} result is a list of tuples, where each tuple may
35242have the following fields:
35243
35244@table @code
35245@item id
35246Identifier of the thread group. This field is always present.
35247The identifier is an opaque string; frontends should not try to
35248convert it to an integer, even though it might look like one.
35249
35250@item type
35251The type of the thread group. At present, only @samp{process} is a
35252valid type.
35253
35254@item pid
35255The target-specific process identifier. This field is only present
35256for thread groups of type @samp{process} and only if the process exists.
35257
35258@item num_children
35259The number of children this thread group has. This field may be
35260absent for an available thread group.
35261
35262@item threads
35263This field has a list of tuples as value, each tuple describing a
35264thread. It may be present if the @samp{--recurse} option is
35265specified, and it's actually possible to obtain the threads.
35266
35267@item cores
35268This field is a list of integers, each identifying a core that one
35269thread of the group is running on. This field may be absent if
35270such information is not available.
35271
35272@item executable
35273The name of the executable file that corresponds to this thread group.
35274The field is only present for thread groups of type @samp{process},
35275and only if there is a corresponding executable file.
35276
35277@end table
35278
35279@subheading Example
35280
35281@smallexample
35282@value{GDBP}
35283-list-thread-groups
35284^done,groups=[@{id="17",type="process",pid="yyy",num_children="2"@}]
35285-list-thread-groups 17
35286^done,threads=[@{id="2",target-id="Thread 0xb7e14b90 (LWP 21257)",
35287 frame=@{level="0",addr="0xffffe410",func="__kernel_vsyscall",args=[]@},state="running"@},
35288@{id="1",target-id="Thread 0xb7e156b0 (LWP 21254)",
35289 frame=@{level="0",addr="0x0804891f",func="foo",args=[@{name="i",value="10"@}],
35290 file="/tmp/a.c",fullname="/tmp/a.c",line="158"@},state="running"@}]]
35291-list-thread-groups --available
35292^done,groups=[@{id="17",type="process",pid="yyy",num_children="2",cores=[1,2]@}]
35293-list-thread-groups --available --recurse 1
35294 ^done,groups=[@{id="17", types="process",pid="yyy",num_children="2",cores=[1,2],
35295 threads=[@{id="1",target-id="Thread 0xb7e14b90",cores=[1]@},
35296 @{id="2",target-id="Thread 0xb7e14b90",cores=[2]@}]@},..]
35297-list-thread-groups --available --recurse 1 17 18
35298^done,groups=[@{id="17", types="process",pid="yyy",num_children="2",cores=[1,2],
35299 threads=[@{id="1",target-id="Thread 0xb7e14b90",cores=[1]@},
35300 @{id="2",target-id="Thread 0xb7e14b90",cores=[2]@}]@},...]
35301@end smallexample
35302
35303@subheading The @code{-info-os} Command
35304@findex -info-os
35305
35306@subsubheading Synopsis
35307
35308@smallexample
35309-info-os [ @var{type} ]
35310@end smallexample
35311
35312If no argument is supplied, the command returns a table of available
35313operating-system-specific information types. If one of these types is
35314supplied as an argument @var{type}, then the command returns a table
35315of data of that type.
35316
35317The types of information available depend on the target operating
35318system.
35319
35320@subsubheading @value{GDBN} Command
35321
35322The corresponding @value{GDBN} command is @samp{info os}.
35323
35324@subsubheading Example
35325
35326When run on a @sc{gnu}/Linux system, the output will look something
35327like this:
35328
35329@smallexample
35330@value{GDBP}
35331-info-os
35332^done,OSDataTable=@{nr_rows="9",nr_cols="3",
35333hdr=[@{width="10",alignment="-1",col_name="col0",colhdr="Type"@},
35334 @{width="10",alignment="-1",col_name="col1",colhdr="Description"@},
35335 @{width="10",alignment="-1",col_name="col2",colhdr="Title"@}],
35336body=[item=@{col0="processes",col1="Listing of all processes",
35337 col2="Processes"@},
35338 item=@{col0="procgroups",col1="Listing of all process groups",
35339 col2="Process groups"@},
35340 item=@{col0="threads",col1="Listing of all threads",
35341 col2="Threads"@},
35342 item=@{col0="files",col1="Listing of all file descriptors",
35343 col2="File descriptors"@},
35344 item=@{col0="sockets",col1="Listing of all internet-domain sockets",
35345 col2="Sockets"@},
35346 item=@{col0="shm",col1="Listing of all shared-memory regions",
35347 col2="Shared-memory regions"@},
35348 item=@{col0="semaphores",col1="Listing of all semaphores",
35349 col2="Semaphores"@},
35350 item=@{col0="msg",col1="Listing of all message queues",
35351 col2="Message queues"@},
35352 item=@{col0="modules",col1="Listing of all loaded kernel modules",
35353 col2="Kernel modules"@}]@}
35354@value{GDBP}
35355-info-os processes
35356^done,OSDataTable=@{nr_rows="190",nr_cols="4",
35357hdr=[@{width="10",alignment="-1",col_name="col0",colhdr="pid"@},
35358 @{width="10",alignment="-1",col_name="col1",colhdr="user"@},
35359 @{width="10",alignment="-1",col_name="col2",colhdr="command"@},
35360 @{width="10",alignment="-1",col_name="col3",colhdr="cores"@}],
35361body=[item=@{col0="1",col1="root",col2="/sbin/init",col3="0"@},
35362 item=@{col0="2",col1="root",col2="[kthreadd]",col3="1"@},
35363 item=@{col0="3",col1="root",col2="[ksoftirqd/0]",col3="0"@},
35364 ...
35365 item=@{col0="26446",col1="stan",col2="bash",col3="0"@},
35366 item=@{col0="28152",col1="stan",col2="bash",col3="1"@}]@}
35367(gdb)
35368@end smallexample
35369
35370(Note that the MI output here includes a @code{"Title"} column that
35371does not appear in command-line @code{info os}; this column is useful
35372for MI clients that want to enumerate the types of data, such as in a
35373popup menu, but is needless clutter on the command line, and
35374@code{info os} omits it.)
35375
35376@subheading The @code{-add-inferior} Command
35377@findex -add-inferior
35378
35379@subheading Synopsis
35380
35381@smallexample
35382-add-inferior
35383@end smallexample
35384
35385Creates a new inferior (@pxref{Inferiors and Programs}). The created
35386inferior is not associated with any executable. Such association may
35387be established with the @samp{-file-exec-and-symbols} command
35388(@pxref{GDB/MI File Commands}). The command response has a single
35389field, @samp{inferior}, whose value is the identifier of the
35390thread group corresponding to the new inferior.
35391
35392@subheading Example
35393
35394@smallexample
35395@value{GDBP}
35396-add-inferior
35397^done,inferior="i3"
35398@end smallexample
35399
35400@subheading The @code{-interpreter-exec} Command
35401@findex -interpreter-exec
35402
35403@subheading Synopsis
35404
35405@smallexample
35406-interpreter-exec @var{interpreter} @var{command}
35407@end smallexample
35408@anchor{-interpreter-exec}
35409
35410Execute the specified @var{command} in the given @var{interpreter}.
35411
35412@subheading @value{GDBN} Command
35413
35414The corresponding @value{GDBN} command is @samp{interpreter-exec}.
35415
35416@subheading Example
35417
35418@smallexample
35419(gdb)
35420-interpreter-exec console "break main"
35421&"During symbol reading, couldn't parse type; debugger out of date?.\n"
35422&"During symbol reading, bad structure-type format.\n"
35423~"Breakpoint 1 at 0x8074fc6: file ../../src/gdb/main.c, line 743.\n"
35424^done
35425(gdb)
35426@end smallexample
35427
35428@subheading The @code{-inferior-tty-set} Command
35429@findex -inferior-tty-set
35430
35431@subheading Synopsis
35432
35433@smallexample
35434-inferior-tty-set /dev/pts/1
35435@end smallexample
35436
35437Set terminal for future runs of the program being debugged.
35438
35439@subheading @value{GDBN} Command
35440
35441The corresponding @value{GDBN} command is @samp{set inferior-tty} /dev/pts/1.
35442
35443@subheading Example
35444
35445@smallexample
35446(gdb)
35447-inferior-tty-set /dev/pts/1
35448^done
35449(gdb)
35450@end smallexample
35451
35452@subheading The @code{-inferior-tty-show} Command
35453@findex -inferior-tty-show
35454
35455@subheading Synopsis
35456
35457@smallexample
35458-inferior-tty-show
35459@end smallexample
35460
35461Show terminal for future runs of program being debugged.
35462
35463@subheading @value{GDBN} Command
35464
35465The corresponding @value{GDBN} command is @samp{show inferior-tty}.
35466
35467@subheading Example
35468
35469@smallexample
35470(gdb)
35471-inferior-tty-set /dev/pts/1
35472^done
35473(gdb)
35474-inferior-tty-show
35475^done,inferior_tty_terminal="/dev/pts/1"
35476(gdb)
35477@end smallexample
35478
35479@subheading The @code{-enable-timings} Command
35480@findex -enable-timings
35481
35482@subheading Synopsis
35483
35484@smallexample
35485-enable-timings [yes | no]
35486@end smallexample
35487
35488Toggle the printing of the wallclock, user and system times for an MI
35489command as a field in its output. This command is to help frontend
35490developers optimize the performance of their code. No argument is
35491equivalent to @samp{yes}.
35492
35493@subheading @value{GDBN} Command
35494
35495No equivalent.
35496
35497@subheading Example
35498
35499@smallexample
35500(gdb)
35501-enable-timings
35502^done
35503(gdb)
35504-break-insert main
35505^done,bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
35506addr="0x080484ed",func="main",file="myprog.c",
35507fullname="/home/nickrob/myprog.c",line="73",thread-groups=["i1"],
35508times="0"@},
35509time=@{wallclock="0.05185",user="0.00800",system="0.00000"@}
35510(gdb)
35511-enable-timings no
35512^done
35513(gdb)
35514-exec-run
35515^running
35516(gdb)
35517*stopped,reason="breakpoint-hit",disp="keep",bkptno="1",thread-id="0",
35518frame=@{addr="0x080484ed",func="main",args=[@{name="argc",value="1"@},
35519@{name="argv",value="0xbfb60364"@}],file="myprog.c",
35520fullname="/home/nickrob/myprog.c",line="73"@}
35521(gdb)
35522@end smallexample
35523
35524@node Annotations
35525@chapter @value{GDBN} Annotations
35526
35527This chapter describes annotations in @value{GDBN}. Annotations were
35528designed to interface @value{GDBN} to graphical user interfaces or other
35529similar programs which want to interact with @value{GDBN} at a
35530relatively high level.
35531
35532The annotation mechanism has largely been superseded by @sc{gdb/mi}
35533(@pxref{GDB/MI}).
35534
35535@ignore
35536This is Edition @value{EDITION}, @value{DATE}.
35537@end ignore
35538
35539@menu
35540* Annotations Overview:: What annotations are; the general syntax.
35541* Server Prefix:: Issuing a command without affecting user state.
35542* Prompting:: Annotations marking @value{GDBN}'s need for input.
35543* Errors:: Annotations for error messages.
35544* Invalidation:: Some annotations describe things now invalid.
35545* Annotations for Running::
35546 Whether the program is running, how it stopped, etc.
35547* Source Annotations:: Annotations describing source code.
35548@end menu
35549
35550@node Annotations Overview
35551@section What is an Annotation?
35552@cindex annotations
35553
35554Annotations start with a newline character, two @samp{control-z}
35555characters, and the name of the annotation. If there is no additional
35556information associated with this annotation, the name of the annotation
35557is followed immediately by a newline. If there is additional
35558information, the name of the annotation is followed by a space, the
35559additional information, and a newline. The additional information
35560cannot contain newline characters.
35561
35562Any output not beginning with a newline and two @samp{control-z}
35563characters denotes literal output from @value{GDBN}. Currently there is
35564no need for @value{GDBN} to output a newline followed by two
35565@samp{control-z} characters, but if there was such a need, the
35566annotations could be extended with an @samp{escape} annotation which
35567means those three characters as output.
35568
35569The annotation @var{level}, which is specified using the
35570@option{--annotate} command line option (@pxref{Mode Options}), controls
35571how much information @value{GDBN} prints together with its prompt,
35572values of expressions, source lines, and other types of output. Level 0
35573is for no annotations, level 1 is for use when @value{GDBN} is run as a
35574subprocess of @sc{gnu} Emacs, level 3 is the maximum annotation suitable
35575for programs that control @value{GDBN}, and level 2 annotations have
35576been made obsolete (@pxref{Limitations, , Limitations of the Annotation
35577Interface, annotate, GDB's Obsolete Annotations}).
35578
35579@table @code
35580@kindex set annotate
35581@item set annotate @var{level}
35582The @value{GDBN} command @code{set annotate} sets the level of
35583annotations to the specified @var{level}.
35584
35585@item show annotate
35586@kindex show annotate
35587Show the current annotation level.
35588@end table
35589
35590This chapter describes level 3 annotations.
35591
35592A simple example of starting up @value{GDBN} with annotations is:
35593
35594@smallexample
35595$ @kbd{gdb --annotate=3}
35596GNU gdb 6.0
35597Copyright 2003 Free Software Foundation, Inc.
35598GDB is free software, covered by the GNU General Public License,
35599and you are welcome to change it and/or distribute copies of it
35600under certain conditions.
35601Type "show copying" to see the conditions.
35602There is absolutely no warranty for GDB. Type "show warranty"
35603for details.
35604This GDB was configured as "i386-pc-linux-gnu"
35605
35606^Z^Zpre-prompt
35607(@value{GDBP})
35608^Z^Zprompt
35609@kbd{quit}
35610
35611^Z^Zpost-prompt
35612$
35613@end smallexample
35614
35615Here @samp{quit} is input to @value{GDBN}; the rest is output from
35616@value{GDBN}. The three lines beginning @samp{^Z^Z} (where @samp{^Z}
35617denotes a @samp{control-z} character) are annotations; the rest is
35618output from @value{GDBN}.
35619
35620@node Server Prefix
35621@section The Server Prefix
35622@cindex server prefix
35623
35624If you prefix a command with @samp{server } then it will not affect
35625the command history, nor will it affect @value{GDBN}'s notion of which
35626command to repeat if @key{RET} is pressed on a line by itself. This
35627means that commands can be run behind a user's back by a front-end in
35628a transparent manner.
35629
35630The @code{server } prefix does not affect the recording of values into
35631the value history; to print a value without recording it into the
35632value history, use the @code{output} command instead of the
35633@code{print} command.
35634
35635Using this prefix also disables confirmation requests
35636(@pxref{confirmation requests}).
35637
35638@node Prompting
35639@section Annotation for @value{GDBN} Input
35640
35641@cindex annotations for prompts
35642When @value{GDBN} prompts for input, it annotates this fact so it is possible
35643to know when to send output, when the output from a given command is
35644over, etc.
35645
35646Different kinds of input each have a different @dfn{input type}. Each
35647input type has three annotations: a @code{pre-} annotation, which
35648denotes the beginning of any prompt which is being output, a plain
35649annotation, which denotes the end of the prompt, and then a @code{post-}
35650annotation which denotes the end of any echo which may (or may not) be
35651associated with the input. For example, the @code{prompt} input type
35652features the following annotations:
35653
35654@smallexample
35655^Z^Zpre-prompt
35656^Z^Zprompt
35657^Z^Zpost-prompt
35658@end smallexample
35659
35660The input types are
35661
35662@table @code
35663@findex pre-prompt annotation
35664@findex prompt annotation
35665@findex post-prompt annotation
35666@item prompt
35667When @value{GDBN} is prompting for a command (the main @value{GDBN} prompt).
35668
35669@findex pre-commands annotation
35670@findex commands annotation
35671@findex post-commands annotation
35672@item commands
35673When @value{GDBN} prompts for a set of commands, like in the @code{commands}
35674command. The annotations are repeated for each command which is input.
35675
35676@findex pre-overload-choice annotation
35677@findex overload-choice annotation
35678@findex post-overload-choice annotation
35679@item overload-choice
35680When @value{GDBN} wants the user to select between various overloaded functions.
35681
35682@findex pre-query annotation
35683@findex query annotation
35684@findex post-query annotation
35685@item query
35686When @value{GDBN} wants the user to confirm a potentially dangerous operation.
35687
35688@findex pre-prompt-for-continue annotation
35689@findex prompt-for-continue annotation
35690@findex post-prompt-for-continue annotation
35691@item prompt-for-continue
35692When @value{GDBN} is asking the user to press return to continue. Note: Don't
35693expect this to work well; instead use @code{set height 0} to disable
35694prompting. This is because the counting of lines is buggy in the
35695presence of annotations.
35696@end table
35697
35698@node Errors
35699@section Errors
35700@cindex annotations for errors, warnings and interrupts
35701
35702@findex quit annotation
35703@smallexample
35704^Z^Zquit
35705@end smallexample
35706
35707This annotation occurs right before @value{GDBN} responds to an interrupt.
35708
35709@findex error annotation
35710@smallexample
35711^Z^Zerror
35712@end smallexample
35713
35714This annotation occurs right before @value{GDBN} responds to an error.
35715
35716Quit and error annotations indicate that any annotations which @value{GDBN} was
35717in the middle of may end abruptly. For example, if a
35718@code{value-history-begin} annotation is followed by a @code{error}, one
35719cannot expect to receive the matching @code{value-history-end}. One
35720cannot expect not to receive it either, however; an error annotation
35721does not necessarily mean that @value{GDBN} is immediately returning all the way
35722to the top level.
35723
35724@findex error-begin annotation
35725A quit or error annotation may be preceded by
35726
35727@smallexample
35728^Z^Zerror-begin
35729@end smallexample
35730
35731Any output between that and the quit or error annotation is the error
35732message.
35733
35734Warning messages are not yet annotated.
35735@c If we want to change that, need to fix warning(), type_error(),
35736@c range_error(), and possibly other places.
35737
35738@node Invalidation
35739@section Invalidation Notices
35740
35741@cindex annotations for invalidation messages
35742The following annotations say that certain pieces of state may have
35743changed.
35744
35745@table @code
35746@findex frames-invalid annotation
35747@item ^Z^Zframes-invalid
35748
35749The frames (for example, output from the @code{backtrace} command) may
35750have changed.
35751
35752@findex breakpoints-invalid annotation
35753@item ^Z^Zbreakpoints-invalid
35754
35755The breakpoints may have changed. For example, the user just added or
35756deleted a breakpoint.
35757@end table
35758
35759@node Annotations for Running
35760@section Running the Program
35761@cindex annotations for running programs
35762
35763@findex starting annotation
35764@findex stopping annotation
35765When the program starts executing due to a @value{GDBN} command such as
35766@code{step} or @code{continue},
35767
35768@smallexample
35769^Z^Zstarting
35770@end smallexample
35771
35772is output. When the program stops,
35773
35774@smallexample
35775^Z^Zstopped
35776@end smallexample
35777
35778is output. Before the @code{stopped} annotation, a variety of
35779annotations describe how the program stopped.
35780
35781@table @code
35782@findex exited annotation
35783@item ^Z^Zexited @var{exit-status}
35784The program exited, and @var{exit-status} is the exit status (zero for
35785successful exit, otherwise nonzero).
35786
35787@findex signalled annotation
35788@findex signal-name annotation
35789@findex signal-name-end annotation
35790@findex signal-string annotation
35791@findex signal-string-end annotation
35792@item ^Z^Zsignalled
35793The program exited with a signal. After the @code{^Z^Zsignalled}, the
35794annotation continues:
35795
35796@smallexample
35797@var{intro-text}
35798^Z^Zsignal-name
35799@var{name}
35800^Z^Zsignal-name-end
35801@var{middle-text}
35802^Z^Zsignal-string
35803@var{string}
35804^Z^Zsignal-string-end
35805@var{end-text}
35806@end smallexample
35807
35808@noindent
35809where @var{name} is the name of the signal, such as @code{SIGILL} or
35810@code{SIGSEGV}, and @var{string} is the explanation of the signal, such
35811as @code{Illegal Instruction} or @code{Segmentation fault}.
35812@var{intro-text}, @var{middle-text}, and @var{end-text} are for the
35813user's benefit and have no particular format.
35814
35815@findex signal annotation
35816@item ^Z^Zsignal
35817The syntax of this annotation is just like @code{signalled}, but @value{GDBN} is
35818just saying that the program received the signal, not that it was
35819terminated with it.
35820
35821@findex breakpoint annotation
35822@item ^Z^Zbreakpoint @var{number}
35823The program hit breakpoint number @var{number}.
35824
35825@findex watchpoint annotation
35826@item ^Z^Zwatchpoint @var{number}
35827The program hit watchpoint number @var{number}.
35828@end table
35829
35830@node Source Annotations
35831@section Displaying Source
35832@cindex annotations for source display
35833
35834@findex source annotation
35835The following annotation is used instead of displaying source code:
35836
35837@smallexample
35838^Z^Zsource @var{filename}:@var{line}:@var{character}:@var{middle}:@var{addr}
35839@end smallexample
35840
35841where @var{filename} is an absolute file name indicating which source
35842file, @var{line} is the line number within that file (where 1 is the
35843first line in the file), @var{character} is the character position
35844within the file (where 0 is the first character in the file) (for most
35845debug formats this will necessarily point to the beginning of a line),
35846@var{middle} is @samp{middle} if @var{addr} is in the middle of the
35847line, or @samp{beg} if @var{addr} is at the beginning of the line, and
35848@var{addr} is the address in the target program associated with the
35849source which is being displayed. @var{addr} is in the form @samp{0x}
35850followed by one or more lowercase hex digits (note that this does not
35851depend on the language).
35852
35853@node JIT Interface
35854@chapter JIT Compilation Interface
35855@cindex just-in-time compilation
35856@cindex JIT compilation interface
35857
35858This chapter documents @value{GDBN}'s @dfn{just-in-time} (JIT) compilation
35859interface. A JIT compiler is a program or library that generates native
35860executable code at runtime and executes it, usually in order to achieve good
35861performance while maintaining platform independence.
35862
35863Programs that use JIT compilation are normally difficult to debug because
35864portions of their code are generated at runtime, instead of being loaded from
35865object files, which is where @value{GDBN} normally finds the program's symbols
35866and debug information. In order to debug programs that use JIT compilation,
35867@value{GDBN} has an interface that allows the program to register in-memory
35868symbol files with @value{GDBN} at runtime.
35869
35870If you are using @value{GDBN} to debug a program that uses this interface, then
35871it should work transparently so long as you have not stripped the binary. If
35872you are developing a JIT compiler, then the interface is documented in the rest
35873of this chapter. At this time, the only known client of this interface is the
35874LLVM JIT.
35875
35876Broadly speaking, the JIT interface mirrors the dynamic loader interface. The
35877JIT compiler communicates with @value{GDBN} by writing data into a global
35878variable and calling a fuction at a well-known symbol. When @value{GDBN}
35879attaches, it reads a linked list of symbol files from the global variable to
35880find existing code, and puts a breakpoint in the function so that it can find
35881out about additional code.
35882
35883@menu
35884* Declarations:: Relevant C struct declarations
35885* Registering Code:: Steps to register code
35886* Unregistering Code:: Steps to unregister code
35887* Custom Debug Info:: Emit debug information in a custom format
35888@end menu
35889
35890@node Declarations
35891@section JIT Declarations
35892
35893These are the relevant struct declarations that a C program should include to
35894implement the interface:
35895
35896@smallexample
35897typedef enum
35898@{
35899 JIT_NOACTION = 0,
35900 JIT_REGISTER_FN,
35901 JIT_UNREGISTER_FN
35902@} jit_actions_t;
35903
35904struct jit_code_entry
35905@{
35906 struct jit_code_entry *next_entry;
35907 struct jit_code_entry *prev_entry;
35908 const char *symfile_addr;
35909 uint64_t symfile_size;
35910@};
35911
35912struct jit_descriptor
35913@{
35914 uint32_t version;
35915 /* This type should be jit_actions_t, but we use uint32_t
35916 to be explicit about the bitwidth. */
35917 uint32_t action_flag;
35918 struct jit_code_entry *relevant_entry;
35919 struct jit_code_entry *first_entry;
35920@};
35921
35922/* GDB puts a breakpoint in this function. */
35923void __attribute__((noinline)) __jit_debug_register_code() @{ @};
35924
35925/* Make sure to specify the version statically, because the
35926 debugger may check the version before we can set it. */
35927struct jit_descriptor __jit_debug_descriptor = @{ 1, 0, 0, 0 @};
35928@end smallexample
35929
35930If the JIT is multi-threaded, then it is important that the JIT synchronize any
35931modifications to this global data properly, which can easily be done by putting
35932a global mutex around modifications to these structures.
35933
35934@node Registering Code
35935@section Registering Code
35936
35937To register code with @value{GDBN}, the JIT should follow this protocol:
35938
35939@itemize @bullet
35940@item
35941Generate an object file in memory with symbols and other desired debug
35942information. The file must include the virtual addresses of the sections.
35943
35944@item
35945Create a code entry for the file, which gives the start and size of the symbol
35946file.
35947
35948@item
35949Add it to the linked list in the JIT descriptor.
35950
35951@item
35952Point the relevant_entry field of the descriptor at the entry.
35953
35954@item
35955Set @code{action_flag} to @code{JIT_REGISTER} and call
35956@code{__jit_debug_register_code}.
35957@end itemize
35958
35959When @value{GDBN} is attached and the breakpoint fires, @value{GDBN} uses the
35960@code{relevant_entry} pointer so it doesn't have to walk the list looking for
35961new code. However, the linked list must still be maintained in order to allow
35962@value{GDBN} to attach to a running process and still find the symbol files.
35963
35964@node Unregistering Code
35965@section Unregistering Code
35966
35967If code is freed, then the JIT should use the following protocol:
35968
35969@itemize @bullet
35970@item
35971Remove the code entry corresponding to the code from the linked list.
35972
35973@item
35974Point the @code{relevant_entry} field of the descriptor at the code entry.
35975
35976@item
35977Set @code{action_flag} to @code{JIT_UNREGISTER} and call
35978@code{__jit_debug_register_code}.
35979@end itemize
35980
35981If the JIT frees or recompiles code without unregistering it, then @value{GDBN}
35982and the JIT will leak the memory used for the associated symbol files.
35983
35984@node Custom Debug Info
35985@section Custom Debug Info
35986@cindex custom JIT debug info
35987@cindex JIT debug info reader
35988
35989Generating debug information in platform-native file formats (like ELF
35990or COFF) may be an overkill for JIT compilers; especially if all the
35991debug info is used for is displaying a meaningful backtrace. The
35992issue can be resolved by having the JIT writers decide on a debug info
35993format and also provide a reader that parses the debug info generated
35994by the JIT compiler. This section gives a brief overview on writing
35995such a parser. More specific details can be found in the source file
35996@file{gdb/jit-reader.in}, which is also installed as a header at
35997@file{@var{includedir}/gdb/jit-reader.h} for easy inclusion.
35998
35999The reader is implemented as a shared object (so this functionality is
36000not available on platforms which don't allow loading shared objects at
36001runtime). Two @value{GDBN} commands, @code{jit-reader-load} and
36002@code{jit-reader-unload} are provided, to be used to load and unload
36003the readers from a preconfigured directory. Once loaded, the shared
36004object is used the parse the debug information emitted by the JIT
36005compiler.
36006
36007@menu
36008* Using JIT Debug Info Readers:: How to use supplied readers correctly
36009* Writing JIT Debug Info Readers:: Creating a debug-info reader
36010@end menu
36011
36012@node Using JIT Debug Info Readers
36013@subsection Using JIT Debug Info Readers
36014@kindex jit-reader-load
36015@kindex jit-reader-unload
36016
36017Readers can be loaded and unloaded using the @code{jit-reader-load}
36018and @code{jit-reader-unload} commands.
36019
36020@table @code
36021@item jit-reader-load @var{reader}
36022Load the JIT reader named @var{reader}. @var{reader} is a shared
36023object specified as either an absolute or a relative file name. In
36024the latter case, @value{GDBN} will try to load the reader from a
36025pre-configured directory, usually @file{@var{libdir}/gdb/} on a UNIX
36026system (here @var{libdir} is the system library directory, often
36027@file{/usr/local/lib}).
36028
36029Only one reader can be active at a time; trying to load a second
36030reader when one is already loaded will result in @value{GDBN}
36031reporting an error. A new JIT reader can be loaded by first unloading
36032the current one using @code{jit-reader-unload} and then invoking
36033@code{jit-reader-load}.
36034
36035@item jit-reader-unload
36036Unload the currently loaded JIT reader.
36037
36038@end table
36039
36040@node Writing JIT Debug Info Readers
36041@subsection Writing JIT Debug Info Readers
36042@cindex writing JIT debug info readers
36043
36044As mentioned, a reader is essentially a shared object conforming to a
36045certain ABI. This ABI is described in @file{jit-reader.h}.
36046
36047@file{jit-reader.h} defines the structures, macros and functions
36048required to write a reader. It is installed (along with
36049@value{GDBN}), in @file{@var{includedir}/gdb} where @var{includedir} is
36050the system include directory.
36051
36052Readers need to be released under a GPL compatible license. A reader
36053can be declared as released under such a license by placing the macro
36054@code{GDB_DECLARE_GPL_COMPATIBLE_READER} in a source file.
36055
36056The entry point for readers is the symbol @code{gdb_init_reader},
36057which is expected to be a function with the prototype
36058
36059@findex gdb_init_reader
36060@smallexample
36061extern struct gdb_reader_funcs *gdb_init_reader (void);
36062@end smallexample
36063
36064@cindex @code{struct gdb_reader_funcs}
36065
36066@code{struct gdb_reader_funcs} contains a set of pointers to callback
36067functions. These functions are executed to read the debug info
36068generated by the JIT compiler (@code{read}), to unwind stack frames
36069(@code{unwind}) and to create canonical frame IDs
36070(@code{get_Frame_id}). It also has a callback that is called when the
36071reader is being unloaded (@code{destroy}). The struct looks like this
36072
36073@smallexample
36074struct gdb_reader_funcs
36075@{
36076 /* Must be set to GDB_READER_INTERFACE_VERSION. */
36077 int reader_version;
36078
36079 /* For use by the reader. */
36080 void *priv_data;
36081
36082 gdb_read_debug_info *read;
36083 gdb_unwind_frame *unwind;
36084 gdb_get_frame_id *get_frame_id;
36085 gdb_destroy_reader *destroy;
36086@};
36087@end smallexample
36088
36089@cindex @code{struct gdb_symbol_callbacks}
36090@cindex @code{struct gdb_unwind_callbacks}
36091
36092The callbacks are provided with another set of callbacks by
36093@value{GDBN} to do their job. For @code{read}, these callbacks are
36094passed in a @code{struct gdb_symbol_callbacks} and for @code{unwind}
36095and @code{get_frame_id}, in a @code{struct gdb_unwind_callbacks}.
36096@code{struct gdb_symbol_callbacks} has callbacks to create new object
36097files and new symbol tables inside those object files. @code{struct
36098gdb_unwind_callbacks} has callbacks to read registers off the current
36099frame and to write out the values of the registers in the previous
36100frame. Both have a callback (@code{target_read}) to read bytes off the
36101target's address space.
36102
36103@node In-Process Agent
36104@chapter In-Process Agent
36105@cindex debugging agent
36106The traditional debugging model is conceptually low-speed, but works fine,
36107because most bugs can be reproduced in debugging-mode execution. However,
36108as multi-core or many-core processors are becoming mainstream, and
36109multi-threaded programs become more and more popular, there should be more
36110and more bugs that only manifest themselves at normal-mode execution, for
36111example, thread races, because debugger's interference with the program's
36112timing may conceal the bugs. On the other hand, in some applications,
36113it is not feasible for the debugger to interrupt the program's execution
36114long enough for the developer to learn anything helpful about its behavior.
36115If the program's correctness depends on its real-time behavior, delays
36116introduced by a debugger might cause the program to fail, even when the
36117code itself is correct. It is useful to be able to observe the program's
36118behavior without interrupting it.
36119
36120Therefore, traditional debugging model is too intrusive to reproduce
36121some bugs. In order to reduce the interference with the program, we can
36122reduce the number of operations performed by debugger. The
36123@dfn{In-Process Agent}, a shared library, is running within the same
36124process with inferior, and is able to perform some debugging operations
36125itself. As a result, debugger is only involved when necessary, and
36126performance of debugging can be improved accordingly. Note that
36127interference with program can be reduced but can't be removed completely,
36128because the in-process agent will still stop or slow down the program.
36129
36130The in-process agent can interpret and execute Agent Expressions
36131(@pxref{Agent Expressions}) during performing debugging operations. The
36132agent expressions can be used for different purposes, such as collecting
36133data in tracepoints, and condition evaluation in breakpoints.
36134
36135@anchor{Control Agent}
36136You can control whether the in-process agent is used as an aid for
36137debugging with the following commands:
36138
36139@table @code
36140@kindex set agent on
36141@item set agent on
36142Causes the in-process agent to perform some operations on behalf of the
36143debugger. Just which operations requested by the user will be done
36144by the in-process agent depends on the its capabilities. For example,
36145if you request to evaluate breakpoint conditions in the in-process agent,
36146and the in-process agent has such capability as well, then breakpoint
36147conditions will be evaluated in the in-process agent.
36148
36149@kindex set agent off
36150@item set agent off
36151Disables execution of debugging operations by the in-process agent. All
36152of the operations will be performed by @value{GDBN}.
36153
36154@kindex show agent
36155@item show agent
36156Display the current setting of execution of debugging operations by
36157the in-process agent.
36158@end table
36159
36160@menu
36161* In-Process Agent Protocol::
36162@end menu
36163
36164@node In-Process Agent Protocol
36165@section In-Process Agent Protocol
36166@cindex in-process agent protocol
36167
36168The in-process agent is able to communicate with both @value{GDBN} and
36169GDBserver (@pxref{In-Process Agent}). This section documents the protocol
36170used for communications between @value{GDBN} or GDBserver and the IPA.
36171In general, @value{GDBN} or GDBserver sends commands
36172(@pxref{IPA Protocol Commands}) and data to in-process agent, and then
36173in-process agent replies back with the return result of the command, or
36174some other information. The data sent to in-process agent is composed
36175of primitive data types, such as 4-byte or 8-byte type, and composite
36176types, which are called objects (@pxref{IPA Protocol Objects}).
36177
36178@menu
36179* IPA Protocol Objects::
36180* IPA Protocol Commands::
36181@end menu
36182
36183@node IPA Protocol Objects
36184@subsection IPA Protocol Objects
36185@cindex ipa protocol objects
36186
36187The commands sent to and results received from agent may contain some
36188complex data types called @dfn{objects}.
36189
36190The in-process agent is running on the same machine with @value{GDBN}
36191or GDBserver, so it doesn't have to handle as much differences between
36192two ends as remote protocol (@pxref{Remote Protocol}) tries to handle.
36193However, there are still some differences of two ends in two processes:
36194
36195@enumerate
36196@item
36197word size. On some 64-bit machines, @value{GDBN} or GDBserver can be
36198compiled as a 64-bit executable, while in-process agent is a 32-bit one.
36199@item
36200ABI. Some machines may have multiple types of ABI, @value{GDBN} or
36201GDBserver is compiled with one, and in-process agent is compiled with
36202the other one.
36203@end enumerate
36204
36205Here are the IPA Protocol Objects:
36206
36207@enumerate
36208@item
36209agent expression object. It represents an agent expression
36210(@pxref{Agent Expressions}).
36211@anchor{agent expression object}
36212@item
36213tracepoint action object. It represents a tracepoint action
36214(@pxref{Tracepoint Actions,,Tracepoint Action Lists}) to collect registers,
36215memory, static trace data and to evaluate expression.
36216@anchor{tracepoint action object}
36217@item
36218tracepoint object. It represents a tracepoint (@pxref{Tracepoints}).
36219@anchor{tracepoint object}
36220
36221@end enumerate
36222
36223The following table describes important attributes of each IPA protocol
36224object:
36225
36226@multitable @columnfractions .30 .20 .50
36227@headitem Name @tab Size @tab Description
36228@item @emph{agent expression object} @tab @tab
36229@item length @tab 4 @tab length of bytes code
36230@item byte code @tab @var{length} @tab contents of byte code
36231@item @emph{tracepoint action for collecting memory} @tab @tab
36232@item 'M' @tab 1 @tab type of tracepoint action
36233@item addr @tab 8 @tab if @var{basereg} is @samp{-1}, @var{addr} is the
36234address of the lowest byte to collect, otherwise @var{addr} is the offset
36235of @var{basereg} for memory collecting.
36236@item len @tab 8 @tab length of memory for collecting
36237@item basereg @tab 4 @tab the register number containing the starting
36238memory address for collecting.
36239@item @emph{tracepoint action for collecting registers} @tab @tab
36240@item 'R' @tab 1 @tab type of tracepoint action
36241@item @emph{tracepoint action for collecting static trace data} @tab @tab
36242@item 'L' @tab 1 @tab type of tracepoint action
36243@item @emph{tracepoint action for expression evaluation} @tab @tab
36244@item 'X' @tab 1 @tab type of tracepoint action
36245@item agent expression @tab length of @tab @ref{agent expression object}
36246@item @emph{tracepoint object} @tab @tab
36247@item number @tab 4 @tab number of tracepoint
36248@item address @tab 8 @tab address of tracepoint inserted on
36249@item type @tab 4 @tab type of tracepoint
36250@item enabled @tab 1 @tab enable or disable of tracepoint
36251@item step_count @tab 8 @tab step
36252@item pass_count @tab 8 @tab pass
36253@item numactions @tab 4 @tab number of tracepoint actions
36254@item hit count @tab 8 @tab hit count
36255@item trace frame usage @tab 8 @tab trace frame usage
36256@item compiled_cond @tab 8 @tab compiled condition
36257@item orig_size @tab 8 @tab orig size
36258@item condition @tab 4 if condition is NULL otherwise length of
36259@ref{agent expression object}
36260@tab zero if condition is NULL, otherwise is
36261@ref{agent expression object}
36262@item actions @tab variable
36263@tab numactions number of @ref{tracepoint action object}
36264@end multitable
36265
36266@node IPA Protocol Commands
36267@subsection IPA Protocol Commands
36268@cindex ipa protocol commands
36269
36270The spaces in each command are delimiters to ease reading this commands
36271specification. They don't exist in real commands.
36272
36273@table @samp
36274
36275@item FastTrace:@var{tracepoint_object} @var{gdb_jump_pad_head}
36276Installs a new fast tracepoint described by @var{tracepoint_object}
36277(@pxref{tracepoint object}). @var{gdb_jump_pad_head}, 8-byte long, is the
36278head of @dfn{jumppad}, which is used to jump to data collection routine
36279in IPA finally.
36280
36281Replies:
36282@table @samp
36283@item OK @var{target_address} @var{gdb_jump_pad_head} @var{fjump_size} @var{fjump}
36284@var{target_address} is address of tracepoint in the inferior.
36285@var{gdb_jump_pad_head} is updated head of jumppad. Both of
36286@var{target_address} and @var{gdb_jump_pad_head} are 8-byte long.
36287@var{fjump} contains a sequence of instructions jump to jumppad entry.
36288@var{fjump_size}, 4-byte long, is the size of @var{fjump}.
36289@item E @var{NN}
36290for an error
36291
36292@end table
36293
36294@item close
36295Closes the in-process agent. This command is sent when @value{GDBN} or GDBserver
36296is about to kill inferiors.
36297
36298@item qTfSTM
36299@xref{qTfSTM}.
36300@item qTsSTM
36301@xref{qTsSTM}.
36302@item qTSTMat
36303@xref{qTSTMat}.
36304@item probe_marker_at:@var{address}
36305Asks in-process agent to probe the marker at @var{address}.
36306
36307Replies:
36308@table @samp
36309@item E @var{NN}
36310for an error
36311@end table
36312@item unprobe_marker_at:@var{address}
36313Asks in-process agent to unprobe the marker at @var{address}.
36314@end table
36315
36316@node GDB Bugs
36317@chapter Reporting Bugs in @value{GDBN}
36318@cindex bugs in @value{GDBN}
36319@cindex reporting bugs in @value{GDBN}
36320
36321Your bug reports play an essential role in making @value{GDBN} reliable.
36322
36323Reporting a bug may help you by bringing a solution to your problem, or it
36324may not. But in any case the principal function of a bug report is to help
36325the entire community by making the next version of @value{GDBN} work better. Bug
36326reports are your contribution to the maintenance of @value{GDBN}.
36327
36328In order for a bug report to serve its purpose, you must include the
36329information that enables us to fix the bug.
36330
36331@menu
36332* Bug Criteria:: Have you found a bug?
36333* Bug Reporting:: How to report bugs
36334@end menu
36335
36336@node Bug Criteria
36337@section Have You Found a Bug?
36338@cindex bug criteria
36339
36340If you are not sure whether you have found a bug, here are some guidelines:
36341
36342@itemize @bullet
36343@cindex fatal signal
36344@cindex debugger crash
36345@cindex crash of debugger
36346@item
36347If the debugger gets a fatal signal, for any input whatever, that is a
36348@value{GDBN} bug. Reliable debuggers never crash.
36349
36350@cindex error on valid input
36351@item
36352If @value{GDBN} produces an error message for valid input, that is a
36353bug. (Note that if you're cross debugging, the problem may also be
36354somewhere in the connection to the target.)
36355
36356@cindex invalid input
36357@item
36358If @value{GDBN} does not produce an error message for invalid input,
36359that is a bug. However, you should note that your idea of
36360``invalid input'' might be our idea of ``an extension'' or ``support
36361for traditional practice''.
36362
36363@item
36364If you are an experienced user of debugging tools, your suggestions
36365for improvement of @value{GDBN} are welcome in any case.
36366@end itemize
36367
36368@node Bug Reporting
36369@section How to Report Bugs
36370@cindex bug reports
36371@cindex @value{GDBN} bugs, reporting
36372
36373A number of companies and individuals offer support for @sc{gnu} products.
36374If you obtained @value{GDBN} from a support organization, we recommend you
36375contact that organization first.
36376
36377You can find contact information for many support companies and
36378individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
36379distribution.
36380@c should add a web page ref...
36381
36382@ifset BUGURL
36383@ifset BUGURL_DEFAULT
36384In any event, we also recommend that you submit bug reports for
36385@value{GDBN}. The preferred method is to submit them directly using
36386@uref{http://www.gnu.org/software/gdb/bugs/, @value{GDBN}'s Bugs web
36387page}. Alternatively, the @email{bug-gdb@@gnu.org, e-mail gateway} can
36388be used.
36389
36390@strong{Do not send bug reports to @samp{info-gdb}, or to
36391@samp{help-gdb}, or to any newsgroups.} Most users of @value{GDBN} do
36392not want to receive bug reports. Those that do have arranged to receive
36393@samp{bug-gdb}.
36394
36395The mailing list @samp{bug-gdb} has a newsgroup @samp{gnu.gdb.bug} which
36396serves as a repeater. The mailing list and the newsgroup carry exactly
36397the same messages. Often people think of posting bug reports to the
36398newsgroup instead of mailing them. This appears to work, but it has one
36399problem which can be crucial: a newsgroup posting often lacks a mail
36400path back to the sender. Thus, if we need to ask for more information,
36401we may be unable to reach you. For this reason, it is better to send
36402bug reports to the mailing list.
36403@end ifset
36404@ifclear BUGURL_DEFAULT
36405In any event, we also recommend that you submit bug reports for
36406@value{GDBN} to @value{BUGURL}.
36407@end ifclear
36408@end ifset
36409
36410The fundamental principle of reporting bugs usefully is this:
36411@strong{report all the facts}. If you are not sure whether to state a
36412fact or leave it out, state it!
36413
36414Often people omit facts because they think they know what causes the
36415problem and assume that some details do not matter. Thus, you might
36416assume that the name of the variable you use in an example does not matter.
36417Well, probably it does not, but one cannot be sure. Perhaps the bug is a
36418stray memory reference which happens to fetch from the location where that
36419name is stored in memory; perhaps, if the name were different, the contents
36420of that location would fool the debugger into doing the right thing despite
36421the bug. Play it safe and give a specific, complete example. That is the
36422easiest thing for you to do, and the most helpful.
36423
36424Keep in mind that the purpose of a bug report is to enable us to fix the
36425bug. It may be that the bug has been reported previously, but neither
36426you nor we can know that unless your bug report is complete and
36427self-contained.
36428
36429Sometimes people give a few sketchy facts and ask, ``Does this ring a
36430bell?'' Those bug reports are useless, and we urge everyone to
36431@emph{refuse to respond to them} except to chide the sender to report
36432bugs properly.
36433
36434To enable us to fix the bug, you should include all these things:
36435
36436@itemize @bullet
36437@item
36438The version of @value{GDBN}. @value{GDBN} announces it if you start
36439with no arguments; you can also print it at any time using @code{show
36440version}.
36441
36442Without this, we will not know whether there is any point in looking for
36443the bug in the current version of @value{GDBN}.
36444
36445@item
36446The type of machine you are using, and the operating system name and
36447version number.
36448
36449@item
36450The details of the @value{GDBN} build-time configuration.
36451@value{GDBN} shows these details if you invoke it with the
36452@option{--configuration} command-line option, or if you type
36453@code{show configuration} at @value{GDBN}'s prompt.
36454
36455@item
36456What compiler (and its version) was used to compile @value{GDBN}---e.g.@:
36457``@value{GCC}--2.8.1''.
36458
36459@item
36460What compiler (and its version) was used to compile the program you are
36461debugging---e.g.@: ``@value{GCC}--2.8.1'', or ``HP92453-01 A.10.32.03 HP
36462C Compiler''. For @value{NGCC}, you can say @kbd{@value{GCC} --version}
36463to get this information; for other compilers, see the documentation for
36464those compilers.
36465
36466@item
36467The command arguments you gave the compiler to compile your example and
36468observe the bug. For example, did you use @samp{-O}? To guarantee
36469you will not omit something important, list them all. A copy of the
36470Makefile (or the output from make) is sufficient.
36471
36472If we were to try to guess the arguments, we would probably guess wrong
36473and then we might not encounter the bug.
36474
36475@item
36476A complete input script, and all necessary source files, that will
36477reproduce the bug.
36478
36479@item
36480A description of what behavior you observe that you believe is
36481incorrect. For example, ``It gets a fatal signal.''
36482
36483Of course, if the bug is that @value{GDBN} gets a fatal signal, then we
36484will certainly notice it. But if the bug is incorrect output, we might
36485not notice unless it is glaringly wrong. You might as well not give us
36486a chance to make a mistake.
36487
36488Even if the problem you experience is a fatal signal, you should still
36489say so explicitly. Suppose something strange is going on, such as, your
36490copy of @value{GDBN} is out of synch, or you have encountered a bug in
36491the C library on your system. (This has happened!) Your copy might
36492crash and ours would not. If you told us to expect a crash, then when
36493ours fails to crash, we would know that the bug was not happening for
36494us. If you had not told us to expect a crash, then we would not be able
36495to draw any conclusion from our observations.
36496
36497@pindex script
36498@cindex recording a session script
36499To collect all this information, you can use a session recording program
36500such as @command{script}, which is available on many Unix systems.
36501Just run your @value{GDBN} session inside @command{script} and then
36502include the @file{typescript} file with your bug report.
36503
36504Another way to record a @value{GDBN} session is to run @value{GDBN}
36505inside Emacs and then save the entire buffer to a file.
36506
36507@item
36508If you wish to suggest changes to the @value{GDBN} source, send us context
36509diffs. If you even discuss something in the @value{GDBN} source, refer to
36510it by context, not by line number.
36511
36512The line numbers in our development sources will not match those in your
36513sources. Your line numbers would convey no useful information to us.
36514
36515@end itemize
36516
36517Here are some things that are not necessary:
36518
36519@itemize @bullet
36520@item
36521A description of the envelope of the bug.
36522
36523Often people who encounter a bug spend a lot of time investigating
36524which changes to the input file will make the bug go away and which
36525changes will not affect it.
36526
36527This is often time consuming and not very useful, because the way we
36528will find the bug is by running a single example under the debugger
36529with breakpoints, not by pure deduction from a series of examples.
36530We recommend that you save your time for something else.
36531
36532Of course, if you can find a simpler example to report @emph{instead}
36533of the original one, that is a convenience for us. Errors in the
36534output will be easier to spot, running under the debugger will take
36535less time, and so on.
36536
36537However, simplification is not vital; if you do not want to do this,
36538report the bug anyway and send us the entire test case you used.
36539
36540@item
36541A patch for the bug.
36542
36543A patch for the bug does help us if it is a good one. But do not omit
36544the necessary information, such as the test case, on the assumption that
36545a patch is all we need. We might see problems with your patch and decide
36546to fix the problem another way, or we might not understand it at all.
36547
36548Sometimes with a program as complicated as @value{GDBN} it is very hard to
36549construct an example that will make the program follow a certain path
36550through the code. If you do not send us the example, we will not be able
36551to construct one, so we will not be able to verify that the bug is fixed.
36552
36553And if we cannot understand what bug you are trying to fix, or why your
36554patch should be an improvement, we will not install it. A test case will
36555help us to understand.
36556
36557@item
36558A guess about what the bug is or what it depends on.
36559
36560Such guesses are usually wrong. Even we cannot guess right about such
36561things without first using the debugger to find the facts.
36562@end itemize
36563
36564@c The readline documentation is distributed with the readline code
36565@c and consists of the two following files:
36566@c rluser.texi
36567@c hsuser.texi
36568@c Use -I with makeinfo to point to the appropriate directory,
36569@c environment var TEXINPUTS with TeX.
36570@ifclear SYSTEM_READLINE
36571@include rluser.texi
36572@include hsuser.texi
36573@end ifclear
36574
36575@node In Memoriam
36576@appendix In Memoriam
36577
36578The @value{GDBN} project mourns the loss of the following long-time
36579contributors:
36580
36581@table @code
36582@item Fred Fish
36583Fred was a long-standing contributor to @value{GDBN} (1991-2006), and
36584to Free Software in general. Outside of @value{GDBN}, he was known in
36585the Amiga world for his series of Fish Disks, and the GeekGadget project.
36586
36587@item Michael Snyder
36588Michael was one of the Global Maintainers of the @value{GDBN} project,
36589with contributions recorded as early as 1996, until 2011. In addition
36590to his day to day participation, he was a large driving force behind
36591adding Reverse Debugging to @value{GDBN}.
36592@end table
36593
36594Beyond their technical contributions to the project, they were also
36595enjoyable members of the Free Software Community. We will miss them.
36596
36597@node Formatting Documentation
36598@appendix Formatting Documentation
36599
36600@cindex @value{GDBN} reference card
36601@cindex reference card
36602The @value{GDBN} 4 release includes an already-formatted reference card, ready
36603for printing with PostScript or Ghostscript, in the @file{gdb}
36604subdirectory of the main source directory@footnote{In
36605@file{gdb-@value{GDBVN}/gdb/refcard.ps} of the version @value{GDBVN}
36606release.}. If you can use PostScript or Ghostscript with your printer,
36607you can print the reference card immediately with @file{refcard.ps}.
36608
36609The release also includes the source for the reference card. You
36610can format it, using @TeX{}, by typing:
36611
36612@smallexample
36613make refcard.dvi
36614@end smallexample
36615
36616The @value{GDBN} reference card is designed to print in @dfn{landscape}
36617mode on US ``letter'' size paper;
36618that is, on a sheet 11 inches wide by 8.5 inches
36619high. You will need to specify this form of printing as an option to
36620your @sc{dvi} output program.
36621
36622@cindex documentation
36623
36624All the documentation for @value{GDBN} comes as part of the machine-readable
36625distribution. The documentation is written in Texinfo format, which is
36626a documentation system that uses a single source file to produce both
36627on-line information and a printed manual. You can use one of the Info
36628formatting commands to create the on-line version of the documentation
36629and @TeX{} (or @code{texi2roff}) to typeset the printed version.
36630
36631@value{GDBN} includes an already formatted copy of the on-line Info
36632version of this manual in the @file{gdb} subdirectory. The main Info
36633file is @file{gdb-@value{GDBVN}/gdb/gdb.info}, and it refers to
36634subordinate files matching @samp{gdb.info*} in the same directory. If
36635necessary, you can print out these files, or read them with any editor;
36636but they are easier to read using the @code{info} subsystem in @sc{gnu}
36637Emacs or the standalone @code{info} program, available as part of the
36638@sc{gnu} Texinfo distribution.
36639
36640If you want to format these Info files yourself, you need one of the
36641Info formatting programs, such as @code{texinfo-format-buffer} or
36642@code{makeinfo}.
36643
36644If you have @code{makeinfo} installed, and are in the top level
36645@value{GDBN} source directory (@file{gdb-@value{GDBVN}}, in the case of
36646version @value{GDBVN}), you can make the Info file by typing:
36647
36648@smallexample
36649cd gdb
36650make gdb.info
36651@end smallexample
36652
36653If you want to typeset and print copies of this manual, you need @TeX{},
36654a program to print its @sc{dvi} output files, and @file{texinfo.tex}, the
36655Texinfo definitions file.
36656
36657@TeX{} is a typesetting program; it does not print files directly, but
36658produces output files called @sc{dvi} files. To print a typeset
36659document, you need a program to print @sc{dvi} files. If your system
36660has @TeX{} installed, chances are it has such a program. The precise
36661command to use depends on your system; @kbd{lpr -d} is common; another
36662(for PostScript devices) is @kbd{dvips}. The @sc{dvi} print command may
36663require a file name without any extension or a @samp{.dvi} extension.
36664
36665@TeX{} also requires a macro definitions file called
36666@file{texinfo.tex}. This file tells @TeX{} how to typeset a document
36667written in Texinfo format. On its own, @TeX{} cannot either read or
36668typeset a Texinfo file. @file{texinfo.tex} is distributed with GDB
36669and is located in the @file{gdb-@var{version-number}/texinfo}
36670directory.
36671
36672If you have @TeX{} and a @sc{dvi} printer program installed, you can
36673typeset and print this manual. First switch to the @file{gdb}
36674subdirectory of the main source directory (for example, to
36675@file{gdb-@value{GDBVN}/gdb}) and type:
36676
36677@smallexample
36678make gdb.dvi
36679@end smallexample
36680
36681Then give @file{gdb.dvi} to your @sc{dvi} printing program.
36682
36683@node Installing GDB
36684@appendix Installing @value{GDBN}
36685@cindex installation
36686
36687@menu
36688* Requirements:: Requirements for building @value{GDBN}
36689* Running Configure:: Invoking the @value{GDBN} @file{configure} script
36690* Separate Objdir:: Compiling @value{GDBN} in another directory
36691* Config Names:: Specifying names for hosts and targets
36692* Configure Options:: Summary of options for configure
36693* System-wide configuration:: Having a system-wide init file
36694@end menu
36695
36696@node Requirements
36697@section Requirements for Building @value{GDBN}
36698@cindex building @value{GDBN}, requirements for
36699
36700Building @value{GDBN} requires various tools and packages to be available.
36701Other packages will be used only if they are found.
36702
36703@heading Tools/Packages Necessary for Building @value{GDBN}
36704@table @asis
36705@item ISO C90 compiler
36706@value{GDBN} is written in ISO C90. It should be buildable with any
36707working C90 compiler, e.g.@: GCC.
36708
36709@end table
36710
36711@heading Tools/Packages Optional for Building @value{GDBN}
36712@table @asis
36713@item Expat
36714@anchor{Expat}
36715@value{GDBN} can use the Expat XML parsing library. This library may be
36716included with your operating system distribution; if it is not, you
36717can get the latest version from @url{http://expat.sourceforge.net}.
36718The @file{configure} script will search for this library in several
36719standard locations; if it is installed in an unusual path, you can
36720use the @option{--with-libexpat-prefix} option to specify its location.
36721
36722Expat is used for:
36723
36724@itemize @bullet
36725@item
36726Remote protocol memory maps (@pxref{Memory Map Format})
36727@item
36728Target descriptions (@pxref{Target Descriptions})
36729@item
36730Remote shared library lists (@xref{Library List Format},
36731or alternatively @pxref{Library List Format for SVR4 Targets})
36732@item
36733MS-Windows shared libraries (@pxref{Shared Libraries})
36734@item
36735Traceframe info (@pxref{Traceframe Info Format})
36736@item
36737Branch trace (@pxref{Branch Trace Format})
36738@end itemize
36739
36740@item zlib
36741@cindex compressed debug sections
36742@value{GDBN} will use the @samp{zlib} library, if available, to read
36743compressed debug sections. Some linkers, such as GNU gold, are capable
36744of producing binaries with compressed debug sections. If @value{GDBN}
36745is compiled with @samp{zlib}, it will be able to read the debug
36746information in such binaries.
36747
36748The @samp{zlib} library is likely included with your operating system
36749distribution; if it is not, you can get the latest version from
36750@url{http://zlib.net}.
36751
36752@item iconv
36753@value{GDBN}'s features related to character sets (@pxref{Character
36754Sets}) require a functioning @code{iconv} implementation. If you are
36755on a GNU system, then this is provided by the GNU C Library. Some
36756other systems also provide a working @code{iconv}.
36757
36758If @value{GDBN} is using the @code{iconv} program which is installed
36759in a non-standard place, you will need to tell @value{GDBN} where to find it.
36760This is done with @option{--with-iconv-bin} which specifies the
36761directory that contains the @code{iconv} program.
36762
36763On systems without @code{iconv}, you can install GNU Libiconv. If you
36764have previously installed Libiconv, you can use the
36765@option{--with-libiconv-prefix} option to configure.
36766
36767@value{GDBN}'s top-level @file{configure} and @file{Makefile} will
36768arrange to build Libiconv if a directory named @file{libiconv} appears
36769in the top-most source directory. If Libiconv is built this way, and
36770if the operating system does not provide a suitable @code{iconv}
36771implementation, then the just-built library will automatically be used
36772by @value{GDBN}. One easy way to set this up is to download GNU
36773Libiconv, unpack it, and then rename the directory holding the
36774Libiconv source code to @samp{libiconv}.
36775@end table
36776
36777@node Running Configure
36778@section Invoking the @value{GDBN} @file{configure} Script
36779@cindex configuring @value{GDBN}
36780@value{GDBN} comes with a @file{configure} script that automates the process
36781of preparing @value{GDBN} for installation; you can then use @code{make} to
36782build the @code{gdb} program.
36783@iftex
36784@c irrelevant in info file; it's as current as the code it lives with.
36785@footnote{If you have a more recent version of @value{GDBN} than @value{GDBVN},
36786look at the @file{README} file in the sources; we may have improved the
36787installation procedures since publishing this manual.}
36788@end iftex
36789
36790The @value{GDBN} distribution includes all the source code you need for
36791@value{GDBN} in a single directory, whose name is usually composed by
36792appending the version number to @samp{gdb}.
36793
36794For example, the @value{GDBN} version @value{GDBVN} distribution is in the
36795@file{gdb-@value{GDBVN}} directory. That directory contains:
36796
36797@table @code
36798@item gdb-@value{GDBVN}/configure @r{(and supporting files)}
36799script for configuring @value{GDBN} and all its supporting libraries
36800
36801@item gdb-@value{GDBVN}/gdb
36802the source specific to @value{GDBN} itself
36803
36804@item gdb-@value{GDBVN}/bfd
36805source for the Binary File Descriptor library
36806
36807@item gdb-@value{GDBVN}/include
36808@sc{gnu} include files
36809
36810@item gdb-@value{GDBVN}/libiberty
36811source for the @samp{-liberty} free software library
36812
36813@item gdb-@value{GDBVN}/opcodes
36814source for the library of opcode tables and disassemblers
36815
36816@item gdb-@value{GDBVN}/readline
36817source for the @sc{gnu} command-line interface
36818
36819@item gdb-@value{GDBVN}/glob
36820source for the @sc{gnu} filename pattern-matching subroutine
36821
36822@item gdb-@value{GDBVN}/mmalloc
36823source for the @sc{gnu} memory-mapped malloc package
36824@end table
36825
36826The simplest way to configure and build @value{GDBN} is to run @file{configure}
36827from the @file{gdb-@var{version-number}} source directory, which in
36828this example is the @file{gdb-@value{GDBVN}} directory.
36829
36830First switch to the @file{gdb-@var{version-number}} source directory
36831if you are not already in it; then run @file{configure}. Pass the
36832identifier for the platform on which @value{GDBN} will run as an
36833argument.
36834
36835For example:
36836
36837@smallexample
36838cd gdb-@value{GDBVN}
36839./configure @var{host}
36840make
36841@end smallexample
36842
36843@noindent
36844where @var{host} is an identifier such as @samp{sun4} or
36845@samp{decstation}, that identifies the platform where @value{GDBN} will run.
36846(You can often leave off @var{host}; @file{configure} tries to guess the
36847correct value by examining your system.)
36848
36849Running @samp{configure @var{host}} and then running @code{make} builds the
36850@file{bfd}, @file{readline}, @file{mmalloc}, and @file{libiberty}
36851libraries, then @code{gdb} itself. The configured source files, and the
36852binaries, are left in the corresponding source directories.
36853
36854@need 750
36855@file{configure} is a Bourne-shell (@code{/bin/sh}) script; if your
36856system does not recognize this automatically when you run a different
36857shell, you may need to run @code{sh} on it explicitly:
36858
36859@smallexample
36860sh configure @var{host}
36861@end smallexample
36862
36863If you run @file{configure} from a directory that contains source
36864directories for multiple libraries or programs, such as the
36865@file{gdb-@value{GDBVN}} source directory for version @value{GDBVN},
36866@file{configure}
36867creates configuration files for every directory level underneath (unless
36868you tell it not to, with the @samp{--norecursion} option).
36869
36870You should run the @file{configure} script from the top directory in the
36871source tree, the @file{gdb-@var{version-number}} directory. If you run
36872@file{configure} from one of the subdirectories, you will configure only
36873that subdirectory. That is usually not what you want. In particular,
36874if you run the first @file{configure} from the @file{gdb} subdirectory
36875of the @file{gdb-@var{version-number}} directory, you will omit the
36876configuration of @file{bfd}, @file{readline}, and other sibling
36877directories of the @file{gdb} subdirectory. This leads to build errors
36878about missing include files such as @file{bfd/bfd.h}.
36879
36880You can install @code{@value{GDBP}} anywhere; it has no hardwired paths.
36881However, you should make sure that the shell on your path (named by
36882the @samp{SHELL} environment variable) is publicly readable. Remember
36883that @value{GDBN} uses the shell to start your program---some systems refuse to
36884let @value{GDBN} debug child processes whose programs are not readable.
36885
36886@node Separate Objdir
36887@section Compiling @value{GDBN} in Another Directory
36888
36889If you want to run @value{GDBN} versions for several host or target machines,
36890you need a different @code{gdb} compiled for each combination of
36891host and target. @file{configure} is designed to make this easy by
36892allowing you to generate each configuration in a separate subdirectory,
36893rather than in the source directory. If your @code{make} program
36894handles the @samp{VPATH} feature (@sc{gnu} @code{make} does), running
36895@code{make} in each of these directories builds the @code{gdb}
36896program specified there.
36897
36898To build @code{gdb} in a separate directory, run @file{configure}
36899with the @samp{--srcdir} option to specify where to find the source.
36900(You also need to specify a path to find @file{configure}
36901itself from your working directory. If the path to @file{configure}
36902would be the same as the argument to @samp{--srcdir}, you can leave out
36903the @samp{--srcdir} option; it is assumed.)
36904
36905For example, with version @value{GDBVN}, you can build @value{GDBN} in a
36906separate directory for a Sun 4 like this:
36907
36908@smallexample
36909@group
36910cd gdb-@value{GDBVN}
36911mkdir ../gdb-sun4
36912cd ../gdb-sun4
36913../gdb-@value{GDBVN}/configure sun4
36914make
36915@end group
36916@end smallexample
36917
36918When @file{configure} builds a configuration using a remote source
36919directory, it creates a tree for the binaries with the same structure
36920(and using the same names) as the tree under the source directory. In
36921the example, you'd find the Sun 4 library @file{libiberty.a} in the
36922directory @file{gdb-sun4/libiberty}, and @value{GDBN} itself in
36923@file{gdb-sun4/gdb}.
36924
36925Make sure that your path to the @file{configure} script has just one
36926instance of @file{gdb} in it. If your path to @file{configure} looks
36927like @file{../gdb-@value{GDBVN}/gdb/configure}, you are configuring only
36928one subdirectory of @value{GDBN}, not the whole package. This leads to
36929build errors about missing include files such as @file{bfd/bfd.h}.
36930
36931One popular reason to build several @value{GDBN} configurations in separate
36932directories is to configure @value{GDBN} for cross-compiling (where
36933@value{GDBN} runs on one machine---the @dfn{host}---while debugging
36934programs that run on another machine---the @dfn{target}).
36935You specify a cross-debugging target by
36936giving the @samp{--target=@var{target}} option to @file{configure}.
36937
36938When you run @code{make} to build a program or library, you must run
36939it in a configured directory---whatever directory you were in when you
36940called @file{configure} (or one of its subdirectories).
36941
36942The @code{Makefile} that @file{configure} generates in each source
36943directory also runs recursively. If you type @code{make} in a source
36944directory such as @file{gdb-@value{GDBVN}} (or in a separate configured
36945directory configured with @samp{--srcdir=@var{dirname}/gdb-@value{GDBVN}}), you
36946will build all the required libraries, and then build GDB.
36947
36948When you have multiple hosts or targets configured in separate
36949directories, you can run @code{make} on them in parallel (for example,
36950if they are NFS-mounted on each of the hosts); they will not interfere
36951with each other.
36952
36953@node Config Names
36954@section Specifying Names for Hosts and Targets
36955
36956The specifications used for hosts and targets in the @file{configure}
36957script are based on a three-part naming scheme, but some short predefined
36958aliases are also supported. The full naming scheme encodes three pieces
36959of information in the following pattern:
36960
36961@smallexample
36962@var{architecture}-@var{vendor}-@var{os}
36963@end smallexample
36964
36965For example, you can use the alias @code{sun4} as a @var{host} argument,
36966or as the value for @var{target} in a @code{--target=@var{target}}
36967option. The equivalent full name is @samp{sparc-sun-sunos4}.
36968
36969The @file{configure} script accompanying @value{GDBN} does not provide
36970any query facility to list all supported host and target names or
36971aliases. @file{configure} calls the Bourne shell script
36972@code{config.sub} to map abbreviations to full names; you can read the
36973script, if you wish, or you can use it to test your guesses on
36974abbreviations---for example:
36975
36976@smallexample
36977% sh config.sub i386-linux
36978i386-pc-linux-gnu
36979% sh config.sub alpha-linux
36980alpha-unknown-linux-gnu
36981% sh config.sub hp9k700
36982hppa1.1-hp-hpux
36983% sh config.sub sun4
36984sparc-sun-sunos4.1.1
36985% sh config.sub sun3
36986m68k-sun-sunos4.1.1
36987% sh config.sub i986v
36988Invalid configuration `i986v': machine `i986v' not recognized
36989@end smallexample
36990
36991@noindent
36992@code{config.sub} is also distributed in the @value{GDBN} source
36993directory (@file{gdb-@value{GDBVN}}, for version @value{GDBVN}).
36994
36995@node Configure Options
36996@section @file{configure} Options
36997
36998Here is a summary of the @file{configure} options and arguments that
36999are most often useful for building @value{GDBN}. @file{configure} also has
37000several other options not listed here. @inforef{What Configure
37001Does,,configure.info}, for a full explanation of @file{configure}.
37002
37003@smallexample
37004configure @r{[}--help@r{]}
37005 @r{[}--prefix=@var{dir}@r{]}
37006 @r{[}--exec-prefix=@var{dir}@r{]}
37007 @r{[}--srcdir=@var{dirname}@r{]}
37008 @r{[}--norecursion@r{]} @r{[}--rm@r{]}
37009 @r{[}--target=@var{target}@r{]}
37010 @var{host}
37011@end smallexample
37012
37013@noindent
37014You may introduce options with a single @samp{-} rather than
37015@samp{--} if you prefer; but you may abbreviate option names if you use
37016@samp{--}.
37017
37018@table @code
37019@item --help
37020Display a quick summary of how to invoke @file{configure}.
37021
37022@item --prefix=@var{dir}
37023Configure the source to install programs and files under directory
37024@file{@var{dir}}.
37025
37026@item --exec-prefix=@var{dir}
37027Configure the source to install programs under directory
37028@file{@var{dir}}.
37029
37030@c avoid splitting the warning from the explanation:
37031@need 2000
37032@item --srcdir=@var{dirname}
37033@strong{Warning: using this option requires @sc{gnu} @code{make}, or another
37034@code{make} that implements the @code{VPATH} feature.}@*
37035Use this option to make configurations in directories separate from the
37036@value{GDBN} source directories. Among other things, you can use this to
37037build (or maintain) several configurations simultaneously, in separate
37038directories. @file{configure} writes configuration-specific files in
37039the current directory, but arranges for them to use the source in the
37040directory @var{dirname}. @file{configure} creates directories under
37041the working directory in parallel to the source directories below
37042@var{dirname}.
37043
37044@item --norecursion
37045Configure only the directory level where @file{configure} is executed; do not
37046propagate configuration to subdirectories.
37047
37048@item --target=@var{target}
37049Configure @value{GDBN} for cross-debugging programs running on the specified
37050@var{target}. Without this option, @value{GDBN} is configured to debug
37051programs that run on the same machine (@var{host}) as @value{GDBN} itself.
37052
37053There is no convenient way to generate a list of all available targets.
37054
37055@item @var{host} @dots{}
37056Configure @value{GDBN} to run on the specified @var{host}.
37057
37058There is no convenient way to generate a list of all available hosts.
37059@end table
37060
37061There are many other options available as well, but they are generally
37062needed for special purposes only.
37063
37064@node System-wide configuration
37065@section System-wide configuration and settings
37066@cindex system-wide init file
37067
37068@value{GDBN} can be configured to have a system-wide init file;
37069this file will be read and executed at startup (@pxref{Startup, , What
37070@value{GDBN} does during startup}).
37071
37072Here is the corresponding configure option:
37073
37074@table @code
37075@item --with-system-gdbinit=@var{file}
37076Specify that the default location of the system-wide init file is
37077@var{file}.
37078@end table
37079
37080If @value{GDBN} has been configured with the option @option{--prefix=$prefix},
37081it may be subject to relocation. Two possible cases:
37082
37083@itemize @bullet
37084@item
37085If the default location of this init file contains @file{$prefix},
37086it will be subject to relocation. Suppose that the configure options
37087are @option{--prefix=$prefix --with-system-gdbinit=$prefix/etc/gdbinit};
37088if @value{GDBN} is moved from @file{$prefix} to @file{$install}, the system
37089init file is looked for as @file{$install/etc/gdbinit} instead of
37090@file{$prefix/etc/gdbinit}.
37091
37092@item
37093By contrast, if the default location does not contain the prefix,
37094it will not be relocated. E.g.@: if @value{GDBN} has been configured with
37095@option{--prefix=/usr/local --with-system-gdbinit=/usr/share/gdb/gdbinit},
37096then @value{GDBN} will always look for @file{/usr/share/gdb/gdbinit},
37097wherever @value{GDBN} is installed.
37098@end itemize
37099
37100If the configured location of the system-wide init file (as given by the
37101@option{--with-system-gdbinit} option at configure time) is in the
37102data-directory (as specified by @option{--with-gdb-datadir} at configure
37103time) or in one of its subdirectories, then @value{GDBN} will look for the
37104system-wide init file in the directory specified by the
37105@option{--data-directory} command-line option.
37106Note that the system-wide init file is only read once, during @value{GDBN}
37107initialization. If the data-directory is changed after @value{GDBN} has
37108started with the @code{set data-directory} command, the file will not be
37109reread.
37110
37111@menu
37112* System-wide Configuration Scripts:: Installed System-wide Configuration Scripts
37113@end menu
37114
37115@node System-wide Configuration Scripts
37116@subsection Installed System-wide Configuration Scripts
37117@cindex system-wide configuration scripts
37118
37119The @file{system-gdbinit} directory, located inside the data-directory
37120(as specified by @option{--with-gdb-datadir} at configure time) contains
37121a number of scripts which can be used as system-wide init files. To
37122automatically source those scripts at startup, @value{GDBN} should be
37123configured with @option{--with-system-gdbinit}. Otherwise, any user
37124should be able to source them by hand as needed.
37125
37126The following scripts are currently available:
37127@itemize @bullet
37128
37129@item @file{elinos.py}
37130@pindex elinos.py
37131@cindex ELinOS system-wide configuration script
37132This script is useful when debugging a program on an ELinOS target.
37133It takes advantage of the environment variables defined in a standard
37134ELinOS environment in order to determine the location of the system
37135shared libraries, and then sets the @samp{solib-absolute-prefix}
37136and @samp{solib-search-path} variables appropriately.
37137
37138@item @file{wrs-linux.py}
37139@pindex wrs-linux.py
37140@cindex Wind River Linux system-wide configuration script
37141This script is useful when debugging a program on a target running
37142Wind River Linux. It expects the @env{ENV_PREFIX} to be set to
37143the host-side sysroot used by the target system.
37144
37145@end itemize
37146
37147@node Maintenance Commands
37148@appendix Maintenance Commands
37149@cindex maintenance commands
37150@cindex internal commands
37151
37152In addition to commands intended for @value{GDBN} users, @value{GDBN}
37153includes a number of commands intended for @value{GDBN} developers,
37154that are not documented elsewhere in this manual. These commands are
37155provided here for reference. (For commands that turn on debugging
37156messages, see @ref{Debugging Output}.)
37157
37158@table @code
37159@kindex maint agent
37160@kindex maint agent-eval
37161@item maint agent @r{[}-at @var{location}@r{,}@r{]} @var{expression}
37162@itemx maint agent-eval @r{[}-at @var{location}@r{,}@r{]} @var{expression}
37163Translate the given @var{expression} into remote agent bytecodes.
37164This command is useful for debugging the Agent Expression mechanism
37165(@pxref{Agent Expressions}). The @samp{agent} version produces an
37166expression useful for data collection, such as by tracepoints, while
37167@samp{maint agent-eval} produces an expression that evaluates directly
37168to a result. For instance, a collection expression for @code{globa +
37169globb} will include bytecodes to record four bytes of memory at each
37170of the addresses of @code{globa} and @code{globb}, while discarding
37171the result of the addition, while an evaluation expression will do the
37172addition and return the sum.
37173If @code{-at} is given, generate remote agent bytecode for @var{location}.
37174If not, generate remote agent bytecode for current frame PC address.
37175
37176@kindex maint agent-printf
37177@item maint agent-printf @var{format},@var{expr},...
37178Translate the given format string and list of argument expressions
37179into remote agent bytecodes and display them as a disassembled list.
37180This command is useful for debugging the agent version of dynamic
37181printf (@pxref{Dynamic Printf}).
37182
37183@kindex maint info breakpoints
37184@item @anchor{maint info breakpoints}maint info breakpoints
37185Using the same format as @samp{info breakpoints}, display both the
37186breakpoints you've set explicitly, and those @value{GDBN} is using for
37187internal purposes. Internal breakpoints are shown with negative
37188breakpoint numbers. The type column identifies what kind of breakpoint
37189is shown:
37190
37191@table @code
37192@item breakpoint
37193Normal, explicitly set breakpoint.
37194
37195@item watchpoint
37196Normal, explicitly set watchpoint.
37197
37198@item longjmp
37199Internal breakpoint, used to handle correctly stepping through
37200@code{longjmp} calls.
37201
37202@item longjmp resume
37203Internal breakpoint at the target of a @code{longjmp}.
37204
37205@item until
37206Temporary internal breakpoint used by the @value{GDBN} @code{until} command.
37207
37208@item finish
37209Temporary internal breakpoint used by the @value{GDBN} @code{finish} command.
37210
37211@item shlib events
37212Shared library events.
37213
37214@end table
37215
37216@kindex maint info bfds
37217@item maint info bfds
37218This prints information about each @code{bfd} object that is known to
37219@value{GDBN}. @xref{Top, , BFD, bfd, The Binary File Descriptor Library}.
37220
37221@kindex set displaced-stepping
37222@kindex show displaced-stepping
37223@cindex displaced stepping support
37224@cindex out-of-line single-stepping
37225@item set displaced-stepping
37226@itemx show displaced-stepping
37227Control whether or not @value{GDBN} will do @dfn{displaced stepping}
37228if the target supports it. Displaced stepping is a way to single-step
37229over breakpoints without removing them from the inferior, by executing
37230an out-of-line copy of the instruction that was originally at the
37231breakpoint location. It is also known as out-of-line single-stepping.
37232
37233@table @code
37234@item set displaced-stepping on
37235If the target architecture supports it, @value{GDBN} will use
37236displaced stepping to step over breakpoints.
37237
37238@item set displaced-stepping off
37239@value{GDBN} will not use displaced stepping to step over breakpoints,
37240even if such is supported by the target architecture.
37241
37242@cindex non-stop mode, and @samp{set displaced-stepping}
37243@item set displaced-stepping auto
37244This is the default mode. @value{GDBN} will use displaced stepping
37245only if non-stop mode is active (@pxref{Non-Stop Mode}) and the target
37246architecture supports displaced stepping.
37247@end table
37248
37249@kindex maint check-psymtabs
37250@item maint check-psymtabs
37251Check the consistency of currently expanded psymtabs versus symtabs.
37252Use this to check, for example, whether a symbol is in one but not the other.
37253
37254@kindex maint check-symtabs
37255@item maint check-symtabs
37256Check the consistency of currently expanded symtabs.
37257
37258@kindex maint expand-symtabs
37259@item maint expand-symtabs [@var{regexp}]
37260Expand symbol tables.
37261If @var{regexp} is specified, only expand symbol tables for file
37262names matching @var{regexp}.
37263
37264@kindex maint cplus first_component
37265@item maint cplus first_component @var{name}
37266Print the first C@t{++} class/namespace component of @var{name}.
37267
37268@kindex maint cplus namespace
37269@item maint cplus namespace
37270Print the list of possible C@t{++} namespaces.
37271
37272@kindex maint demangle
37273@item maint demangle @var{name}
37274Demangle a C@t{++} or Objective-C mangled @var{name}.
37275
37276@kindex maint deprecate
37277@kindex maint undeprecate
37278@cindex deprecated commands
37279@item maint deprecate @var{command} @r{[}@var{replacement}@r{]}
37280@itemx maint undeprecate @var{command}
37281Deprecate or undeprecate the named @var{command}. Deprecated commands
37282cause @value{GDBN} to issue a warning when you use them. The optional
37283argument @var{replacement} says which newer command should be used in
37284favor of the deprecated one; if it is given, @value{GDBN} will mention
37285the replacement as part of the warning.
37286
37287@kindex maint dump-me
37288@item maint dump-me
37289@cindex @code{SIGQUIT} signal, dump core of @value{GDBN}
37290Cause a fatal signal in the debugger and force it to dump its core.
37291This is supported only on systems which support aborting a program
37292with the @code{SIGQUIT} signal.
37293
37294@kindex maint internal-error
37295@kindex maint internal-warning
37296@item maint internal-error @r{[}@var{message-text}@r{]}
37297@itemx maint internal-warning @r{[}@var{message-text}@r{]}
37298Cause @value{GDBN} to call the internal function @code{internal_error}
37299or @code{internal_warning} and hence behave as though an internal error
37300or internal warning has been detected. In addition to reporting the
37301internal problem, these functions give the user the opportunity to
37302either quit @value{GDBN} or create a core file of the current
37303@value{GDBN} session.
37304
37305These commands take an optional parameter @var{message-text} that is
37306used as the text of the error or warning message.
37307
37308Here's an example of using @code{internal-error}:
37309
37310@smallexample
37311(@value{GDBP}) @kbd{maint internal-error testing, 1, 2}
37312@dots{}/maint.c:121: internal-error: testing, 1, 2
37313A problem internal to GDB has been detected. Further
37314debugging may prove unreliable.
37315Quit this debugging session? (y or n) @kbd{n}
37316Create a core file? (y or n) @kbd{n}
37317(@value{GDBP})
37318@end smallexample
37319
37320@cindex @value{GDBN} internal error
37321@cindex internal errors, control of @value{GDBN} behavior
37322
37323@kindex maint set internal-error
37324@kindex maint show internal-error
37325@kindex maint set internal-warning
37326@kindex maint show internal-warning
37327@item maint set internal-error @var{action} [ask|yes|no]
37328@itemx maint show internal-error @var{action}
37329@itemx maint set internal-warning @var{action} [ask|yes|no]
37330@itemx maint show internal-warning @var{action}
37331When @value{GDBN} reports an internal problem (error or warning) it
37332gives the user the opportunity to both quit @value{GDBN} and create a
37333core file of the current @value{GDBN} session. These commands let you
37334override the default behaviour for each particular @var{action},
37335described in the table below.
37336
37337@table @samp
37338@item quit
37339You can specify that @value{GDBN} should always (yes) or never (no)
37340quit. The default is to ask the user what to do.
37341
37342@item corefile
37343You can specify that @value{GDBN} should always (yes) or never (no)
37344create a core file. The default is to ask the user what to do.
37345@end table
37346
37347@kindex maint packet
37348@item maint packet @var{text}
37349If @value{GDBN} is talking to an inferior via the serial protocol,
37350then this command sends the string @var{text} to the inferior, and
37351displays the response packet. @value{GDBN} supplies the initial
37352@samp{$} character, the terminating @samp{#} character, and the
37353checksum.
37354
37355@kindex maint print architecture
37356@item maint print architecture @r{[}@var{file}@r{]}
37357Print the entire architecture configuration. The optional argument
37358@var{file} names the file where the output goes.
37359
37360@kindex maint print c-tdesc
37361@item maint print c-tdesc
37362Print the current target description (@pxref{Target Descriptions}) as
37363a C source file. The created source file can be used in @value{GDBN}
37364when an XML parser is not available to parse the description.
37365
37366@kindex maint print dummy-frames
37367@item maint print dummy-frames
37368Prints the contents of @value{GDBN}'s internal dummy-frame stack.
37369
37370@smallexample
37371(@value{GDBP}) @kbd{b add}
37372@dots{}
37373(@value{GDBP}) @kbd{print add(2,3)}
37374Breakpoint 2, add (a=2, b=3) at @dots{}
3737558 return (a + b);
37376The program being debugged stopped while in a function called from GDB.
37377@dots{}
37378(@value{GDBP}) @kbd{maint print dummy-frames}
373790x1a57c80: pc=0x01014068 fp=0x0200bddc sp=0x0200bdd6
37380 top=0x0200bdd4 id=@{stack=0x200bddc,code=0x101405c@}
37381 call_lo=0x01014000 call_hi=0x01014001
37382(@value{GDBP})
37383@end smallexample
37384
37385Takes an optional file parameter.
37386
37387@kindex maint print registers
37388@kindex maint print raw-registers
37389@kindex maint print cooked-registers
37390@kindex maint print register-groups
37391@kindex maint print remote-registers
37392@item maint print registers @r{[}@var{file}@r{]}
37393@itemx maint print raw-registers @r{[}@var{file}@r{]}
37394@itemx maint print cooked-registers @r{[}@var{file}@r{]}
37395@itemx maint print register-groups @r{[}@var{file}@r{]}
37396@itemx maint print remote-registers @r{[}@var{file}@r{]}
37397Print @value{GDBN}'s internal register data structures.
37398
37399The command @code{maint print raw-registers} includes the contents of
37400the raw register cache; the command @code{maint print
37401cooked-registers} includes the (cooked) value of all registers,
37402including registers which aren't available on the target nor visible
37403to user; the command @code{maint print register-groups} includes the
37404groups that each register is a member of; and the command @code{maint
37405print remote-registers} includes the remote target's register numbers
37406and offsets in the `G' packets.
37407
37408These commands take an optional parameter, a file name to which to
37409write the information.
37410
37411@kindex maint print reggroups
37412@item maint print reggroups @r{[}@var{file}@r{]}
37413Print @value{GDBN}'s internal register group data structures. The
37414optional argument @var{file} tells to what file to write the
37415information.
37416
37417The register groups info looks like this:
37418
37419@smallexample
37420(@value{GDBP}) @kbd{maint print reggroups}
37421 Group Type
37422 general user
37423 float user
37424 all user
37425 vector user
37426 system user
37427 save internal
37428 restore internal
37429@end smallexample
37430
37431@kindex flushregs
37432@item flushregs
37433This command forces @value{GDBN} to flush its internal register cache.
37434
37435@kindex maint print objfiles
37436@cindex info for known object files
37437@item maint print objfiles @r{[}@var{regexp}@r{]}
37438Print a dump of all known object files.
37439If @var{regexp} is specified, only print object files whose names
37440match @var{regexp}. For each object file, this command prints its name,
37441address in memory, and all of its psymtabs and symtabs.
37442
37443@kindex maint print section-scripts
37444@cindex info for known .debug_gdb_scripts-loaded scripts
37445@item maint print section-scripts [@var{regexp}]
37446Print a dump of scripts specified in the @code{.debug_gdb_section} section.
37447If @var{regexp} is specified, only print scripts loaded by object files
37448matching @var{regexp}.
37449For each script, this command prints its name as specified in the objfile,
37450and the full path if known.
37451@xref{dotdebug_gdb_scripts section}.
37452
37453@kindex maint print statistics
37454@cindex bcache statistics
37455@item maint print statistics
37456This command prints, for each object file in the program, various data
37457about that object file followed by the byte cache (@dfn{bcache})
37458statistics for the object file. The objfile data includes the number
37459of minimal, partial, full, and stabs symbols, the number of types
37460defined by the objfile, the number of as yet unexpanded psym tables,
37461the number of line tables and string tables, and the amount of memory
37462used by the various tables. The bcache statistics include the counts,
37463sizes, and counts of duplicates of all and unique objects, max,
37464average, and median entry size, total memory used and its overhead and
37465savings, and various measures of the hash table size and chain
37466lengths.
37467
37468@kindex maint print target-stack
37469@cindex target stack description
37470@item maint print target-stack
37471A @dfn{target} is an interface between the debugger and a particular
37472kind of file or process. Targets can be stacked in @dfn{strata},
37473so that more than one target can potentially respond to a request.
37474In particular, memory accesses will walk down the stack of targets
37475until they find a target that is interested in handling that particular
37476address.
37477
37478This command prints a short description of each layer that was pushed on
37479the @dfn{target stack}, starting from the top layer down to the bottom one.
37480
37481@kindex maint print type
37482@cindex type chain of a data type
37483@item maint print type @var{expr}
37484Print the type chain for a type specified by @var{expr}. The argument
37485can be either a type name or a symbol. If it is a symbol, the type of
37486that symbol is described. The type chain produced by this command is
37487a recursive definition of the data type as stored in @value{GDBN}'s
37488data structures, including its flags and contained types.
37489
37490@kindex maint set dwarf2 always-disassemble
37491@kindex maint show dwarf2 always-disassemble
37492@item maint set dwarf2 always-disassemble
37493@item maint show dwarf2 always-disassemble
37494Control the behavior of @code{info address} when using DWARF debugging
37495information.
37496
37497The default is @code{off}, which means that @value{GDBN} should try to
37498describe a variable's location in an easily readable format. When
37499@code{on}, @value{GDBN} will instead display the DWARF location
37500expression in an assembly-like format. Note that some locations are
37501too complex for @value{GDBN} to describe simply; in this case you will
37502always see the disassembly form.
37503
37504Here is an example of the resulting disassembly:
37505
37506@smallexample
37507(gdb) info addr argc
37508Symbol "argc" is a complex DWARF expression:
37509 1: DW_OP_fbreg 0
37510@end smallexample
37511
37512For more information on these expressions, see
37513@uref{http://www.dwarfstd.org/, the DWARF standard}.
37514
37515@kindex maint set dwarf2 max-cache-age
37516@kindex maint show dwarf2 max-cache-age
37517@item maint set dwarf2 max-cache-age
37518@itemx maint show dwarf2 max-cache-age
37519Control the DWARF 2 compilation unit cache.
37520
37521@cindex DWARF 2 compilation units cache
37522In object files with inter-compilation-unit references, such as those
37523produced by the GCC option @samp{-feliminate-dwarf2-dups}, the DWARF 2
37524reader needs to frequently refer to previously read compilation units.
37525This setting controls how long a compilation unit will remain in the
37526cache if it is not referenced. A higher limit means that cached
37527compilation units will be stored in memory longer, and more total
37528memory will be used. Setting it to zero disables caching, which will
37529slow down @value{GDBN} startup, but reduce memory consumption.
37530
37531@kindex maint set profile
37532@kindex maint show profile
37533@cindex profiling GDB
37534@item maint set profile
37535@itemx maint show profile
37536Control profiling of @value{GDBN}.
37537
37538Profiling will be disabled until you use the @samp{maint set profile}
37539command to enable it. When you enable profiling, the system will begin
37540collecting timing and execution count data; when you disable profiling or
37541exit @value{GDBN}, the results will be written to a log file. Remember that
37542if you use profiling, @value{GDBN} will overwrite the profiling log file
37543(often called @file{gmon.out}). If you have a record of important profiling
37544data in a @file{gmon.out} file, be sure to move it to a safe location.
37545
37546Configuring with @samp{--enable-profiling} arranges for @value{GDBN} to be
37547compiled with the @samp{-pg} compiler option.
37548
37549@kindex maint set show-debug-regs
37550@kindex maint show show-debug-regs
37551@cindex hardware debug registers
37552@item maint set show-debug-regs
37553@itemx maint show show-debug-regs
37554Control whether to show variables that mirror the hardware debug
37555registers. Use @code{on} to enable, @code{off} to disable. If
37556enabled, the debug registers values are shown when @value{GDBN} inserts or
37557removes a hardware breakpoint or watchpoint, and when the inferior
37558triggers a hardware-assisted breakpoint or watchpoint.
37559
37560@kindex maint set show-all-tib
37561@kindex maint show show-all-tib
37562@item maint set show-all-tib
37563@itemx maint show show-all-tib
37564Control whether to show all non zero areas within a 1k block starting
37565at thread local base, when using the @samp{info w32 thread-information-block}
37566command.
37567
37568@kindex maint set per-command
37569@kindex maint show per-command
37570@item maint set per-command
37571@itemx maint show per-command
37572@cindex resources used by commands
37573
37574@value{GDBN} can display the resources used by each command.
37575This is useful in debugging performance problems.
37576
37577@table @code
37578@item maint set per-command space [on|off]
37579@itemx maint show per-command space
37580Enable or disable the printing of the memory used by GDB for each command.
37581If enabled, @value{GDBN} will display how much memory each command
37582took, following the command's own output.
37583This can also be requested by invoking @value{GDBN} with the
37584@option{--statistics} command-line switch (@pxref{Mode Options}).
37585
37586@item maint set per-command time [on|off]
37587@itemx maint show per-command time
37588Enable or disable the printing of the execution time of @value{GDBN}
37589for each command.
37590If enabled, @value{GDBN} will display how much time it
37591took to execute each command, following the command's own output.
37592Both CPU time and wallclock time are printed.
37593Printing both is useful when trying to determine whether the cost is
37594CPU or, e.g., disk/network latency.
37595Note that the CPU time printed is for @value{GDBN} only, it does not include
37596the execution time of the inferior because there's no mechanism currently
37597to compute how much time was spent by @value{GDBN} and how much time was
37598spent by the program been debugged.
37599This can also be requested by invoking @value{GDBN} with the
37600@option{--statistics} command-line switch (@pxref{Mode Options}).
37601
37602@item maint set per-command symtab [on|off]
37603@itemx maint show per-command symtab
37604Enable or disable the printing of basic symbol table statistics
37605for each command.
37606If enabled, @value{GDBN} will display the following information:
37607
37608@enumerate a
37609@item
37610number of symbol tables
37611@item
37612number of primary symbol tables
37613@item
37614number of blocks in the blockvector
37615@end enumerate
37616@end table
37617
37618@kindex maint space
37619@cindex memory used by commands
37620@item maint space @var{value}
37621An alias for @code{maint set per-command space}.
37622A non-zero value enables it, zero disables it.
37623
37624@kindex maint time
37625@cindex time of command execution
37626@item maint time @var{value}
37627An alias for @code{maint set per-command time}.
37628A non-zero value enables it, zero disables it.
37629
37630@kindex maint translate-address
37631@item maint translate-address @r{[}@var{section}@r{]} @var{addr}
37632Find the symbol stored at the location specified by the address
37633@var{addr} and an optional section name @var{section}. If found,
37634@value{GDBN} prints the name of the closest symbol and an offset from
37635the symbol's location to the specified address. This is similar to
37636the @code{info address} command (@pxref{Symbols}), except that this
37637command also allows to find symbols in other sections.
37638
37639If section was not specified, the section in which the symbol was found
37640is also printed. For dynamically linked executables, the name of
37641executable or shared library containing the symbol is printed as well.
37642
37643@end table
37644
37645The following command is useful for non-interactive invocations of
37646@value{GDBN}, such as in the test suite.
37647
37648@table @code
37649@item set watchdog @var{nsec}
37650@kindex set watchdog
37651@cindex watchdog timer
37652@cindex timeout for commands
37653Set the maximum number of seconds @value{GDBN} will wait for the
37654target operation to finish. If this time expires, @value{GDBN}
37655reports and error and the command is aborted.
37656
37657@item show watchdog
37658Show the current setting of the target wait timeout.
37659@end table
37660
37661@node Remote Protocol
37662@appendix @value{GDBN} Remote Serial Protocol
37663
37664@menu
37665* Overview::
37666* Packets::
37667* Stop Reply Packets::
37668* General Query Packets::
37669* Architecture-Specific Protocol Details::
37670* Tracepoint Packets::
37671* Host I/O Packets::
37672* Interrupts::
37673* Notification Packets::
37674* Remote Non-Stop::
37675* Packet Acknowledgment::
37676* Examples::
37677* File-I/O Remote Protocol Extension::
37678* Library List Format::
37679* Library List Format for SVR4 Targets::
37680* Memory Map Format::
37681* Thread List Format::
37682* Traceframe Info Format::
37683* Branch Trace Format::
37684@end menu
37685
37686@node Overview
37687@section Overview
37688
37689There may be occasions when you need to know something about the
37690protocol---for example, if there is only one serial port to your target
37691machine, you might want your program to do something special if it
37692recognizes a packet meant for @value{GDBN}.
37693
37694In the examples below, @samp{->} and @samp{<-} are used to indicate
37695transmitted and received data, respectively.
37696
37697@cindex protocol, @value{GDBN} remote serial
37698@cindex serial protocol, @value{GDBN} remote
37699@cindex remote serial protocol
37700All @value{GDBN} commands and responses (other than acknowledgments
37701and notifications, see @ref{Notification Packets}) are sent as a
37702@var{packet}. A @var{packet} is introduced with the character
37703@samp{$}, the actual @var{packet-data}, and the terminating character
37704@samp{#} followed by a two-digit @var{checksum}:
37705
37706@smallexample
37707@code{$}@var{packet-data}@code{#}@var{checksum}
37708@end smallexample
37709@noindent
37710
37711@cindex checksum, for @value{GDBN} remote
37712@noindent
37713The two-digit @var{checksum} is computed as the modulo 256 sum of all
37714characters between the leading @samp{$} and the trailing @samp{#} (an
37715eight bit unsigned checksum).
37716
37717Implementors should note that prior to @value{GDBN} 5.0 the protocol
37718specification also included an optional two-digit @var{sequence-id}:
37719
37720@smallexample
37721@code{$}@var{sequence-id}@code{:}@var{packet-data}@code{#}@var{checksum}
37722@end smallexample
37723
37724@cindex sequence-id, for @value{GDBN} remote
37725@noindent
37726That @var{sequence-id} was appended to the acknowledgment. @value{GDBN}
37727has never output @var{sequence-id}s. Stubs that handle packets added
37728since @value{GDBN} 5.0 must not accept @var{sequence-id}.
37729
37730When either the host or the target machine receives a packet, the first
37731response expected is an acknowledgment: either @samp{+} (to indicate
37732the package was received correctly) or @samp{-} (to request
37733retransmission):
37734
37735@smallexample
37736-> @code{$}@var{packet-data}@code{#}@var{checksum}
37737<- @code{+}
37738@end smallexample
37739@noindent
37740
37741The @samp{+}/@samp{-} acknowledgments can be disabled
37742once a connection is established.
37743@xref{Packet Acknowledgment}, for details.
37744
37745The host (@value{GDBN}) sends @var{command}s, and the target (the
37746debugging stub incorporated in your program) sends a @var{response}. In
37747the case of step and continue @var{command}s, the response is only sent
37748when the operation has completed, and the target has again stopped all
37749threads in all attached processes. This is the default all-stop mode
37750behavior, but the remote protocol also supports @value{GDBN}'s non-stop
37751execution mode; see @ref{Remote Non-Stop}, for details.
37752
37753@var{packet-data} consists of a sequence of characters with the
37754exception of @samp{#} and @samp{$} (see @samp{X} packet for additional
37755exceptions).
37756
37757@cindex remote protocol, field separator
37758Fields within the packet should be separated using @samp{,} @samp{;} or
37759@samp{:}. Except where otherwise noted all numbers are represented in
37760@sc{hex} with leading zeros suppressed.
37761
37762Implementors should note that prior to @value{GDBN} 5.0, the character
37763@samp{:} could not appear as the third character in a packet (as it
37764would potentially conflict with the @var{sequence-id}).
37765
37766@cindex remote protocol, binary data
37767@anchor{Binary Data}
37768Binary data in most packets is encoded either as two hexadecimal
37769digits per byte of binary data. This allowed the traditional remote
37770protocol to work over connections which were only seven-bit clean.
37771Some packets designed more recently assume an eight-bit clean
37772connection, and use a more efficient encoding to send and receive
37773binary data.
37774
37775The binary data representation uses @code{7d} (@sc{ascii} @samp{@}})
37776as an escape character. Any escaped byte is transmitted as the escape
37777character followed by the original character XORed with @code{0x20}.
37778For example, the byte @code{0x7d} would be transmitted as the two
37779bytes @code{0x7d 0x5d}. The bytes @code{0x23} (@sc{ascii} @samp{#}),
37780@code{0x24} (@sc{ascii} @samp{$}), and @code{0x7d} (@sc{ascii}
37781@samp{@}}) must always be escaped. Responses sent by the stub
37782must also escape @code{0x2a} (@sc{ascii} @samp{*}), so that it
37783is not interpreted as the start of a run-length encoded sequence
37784(described next).
37785
37786Response @var{data} can be run-length encoded to save space.
37787Run-length encoding replaces runs of identical characters with one
37788instance of the repeated character, followed by a @samp{*} and a
37789repeat count. The repeat count is itself sent encoded, to avoid
37790binary characters in @var{data}: a value of @var{n} is sent as
37791@code{@var{n}+29}. For a repeat count greater or equal to 3, this
37792produces a printable @sc{ascii} character, e.g.@: a space (@sc{ascii}
37793code 32) for a repeat count of 3. (This is because run-length
37794encoding starts to win for counts 3 or more.) Thus, for example,
37795@samp{0* } is a run-length encoding of ``0000'': the space character
37796after @samp{*} means repeat the leading @code{0} @w{@code{32 - 29 =
377973}} more times.
37798
37799The printable characters @samp{#} and @samp{$} or with a numeric value
37800greater than 126 must not be used. Runs of six repeats (@samp{#}) or
37801seven repeats (@samp{$}) can be expanded using a repeat count of only
37802five (@samp{"}). For example, @samp{00000000} can be encoded as
37803@samp{0*"00}.
37804
37805The error response returned for some packets includes a two character
37806error number. That number is not well defined.
37807
37808@cindex empty response, for unsupported packets
37809For any @var{command} not supported by the stub, an empty response
37810(@samp{$#00}) should be returned. That way it is possible to extend the
37811protocol. A newer @value{GDBN} can tell if a packet is supported based
37812on that response.
37813
37814At a minimum, a stub is required to support the @samp{g} and @samp{G}
37815commands for register access, and the @samp{m} and @samp{M} commands
37816for memory access. Stubs that only control single-threaded targets
37817can implement run control with the @samp{c} (continue), and @samp{s}
37818(step) commands. Stubs that support multi-threading targets should
37819support the @samp{vCont} command. All other commands are optional.
37820
37821@node Packets
37822@section Packets
37823
37824The following table provides a complete list of all currently defined
37825@var{command}s and their corresponding response @var{data}.
37826@xref{File-I/O Remote Protocol Extension}, for details about the File
37827I/O extension of the remote protocol.
37828
37829Each packet's description has a template showing the packet's overall
37830syntax, followed by an explanation of the packet's meaning. We
37831include spaces in some of the templates for clarity; these are not
37832part of the packet's syntax. No @value{GDBN} packet uses spaces to
37833separate its components. For example, a template like @samp{foo
37834@var{bar} @var{baz}} describes a packet beginning with the three ASCII
37835bytes @samp{foo}, followed by a @var{bar}, followed directly by a
37836@var{baz}. @value{GDBN} does not transmit a space character between the
37837@samp{foo} and the @var{bar}, or between the @var{bar} and the
37838@var{baz}.
37839
37840@cindex @var{thread-id}, in remote protocol
37841@anchor{thread-id syntax}
37842Several packets and replies include a @var{thread-id} field to identify
37843a thread. Normally these are positive numbers with a target-specific
37844interpretation, formatted as big-endian hex strings. A @var{thread-id}
37845can also be a literal @samp{-1} to indicate all threads, or @samp{0} to
37846pick any thread.
37847
37848In addition, the remote protocol supports a multiprocess feature in
37849which the @var{thread-id} syntax is extended to optionally include both
37850process and thread ID fields, as @samp{p@var{pid}.@var{tid}}.
37851The @var{pid} (process) and @var{tid} (thread) components each have the
37852format described above: a positive number with target-specific
37853interpretation formatted as a big-endian hex string, literal @samp{-1}
37854to indicate all processes or threads (respectively), or @samp{0} to
37855indicate an arbitrary process or thread. Specifying just a process, as
37856@samp{p@var{pid}}, is equivalent to @samp{p@var{pid}.-1}. It is an
37857error to specify all processes but a specific thread, such as
37858@samp{p-1.@var{tid}}. Note that the @samp{p} prefix is @emph{not} used
37859for those packets and replies explicitly documented to include a process
37860ID, rather than a @var{thread-id}.
37861
37862The multiprocess @var{thread-id} syntax extensions are only used if both
37863@value{GDBN} and the stub report support for the @samp{multiprocess}
37864feature using @samp{qSupported}. @xref{multiprocess extensions}, for
37865more information.
37866
37867Note that all packet forms beginning with an upper- or lower-case
37868letter, other than those described here, are reserved for future use.
37869
37870Here are the packet descriptions.
37871
37872@table @samp
37873
37874@item !
37875@cindex @samp{!} packet
37876@anchor{extended mode}
37877Enable extended mode. In extended mode, the remote server is made
37878persistent. The @samp{R} packet is used to restart the program being
37879debugged.
37880
37881Reply:
37882@table @samp
37883@item OK
37884The remote target both supports and has enabled extended mode.
37885@end table
37886
37887@item ?
37888@cindex @samp{?} packet
37889Indicate the reason the target halted. The reply is the same as for
37890step and continue. This packet has a special interpretation when the
37891target is in non-stop mode; see @ref{Remote Non-Stop}.
37892
37893Reply:
37894@xref{Stop Reply Packets}, for the reply specifications.
37895
37896@item A @var{arglen},@var{argnum},@var{arg},@dots{}
37897@cindex @samp{A} packet
37898Initialized @code{argv[]} array passed into program. @var{arglen}
37899specifies the number of bytes in the hex encoded byte stream
37900@var{arg}. See @code{gdbserver} for more details.
37901
37902Reply:
37903@table @samp
37904@item OK
37905The arguments were set.
37906@item E @var{NN}
37907An error occurred.
37908@end table
37909
37910@item b @var{baud}
37911@cindex @samp{b} packet
37912(Don't use this packet; its behavior is not well-defined.)
37913Change the serial line speed to @var{baud}.
37914
37915JTC: @emph{When does the transport layer state change? When it's
37916received, or after the ACK is transmitted. In either case, there are
37917problems if the command or the acknowledgment packet is dropped.}
37918
37919Stan: @emph{If people really wanted to add something like this, and get
37920it working for the first time, they ought to modify ser-unix.c to send
37921some kind of out-of-band message to a specially-setup stub and have the
37922switch happen "in between" packets, so that from remote protocol's point
37923of view, nothing actually happened.}
37924
37925@item B @var{addr},@var{mode}
37926@cindex @samp{B} packet
37927Set (@var{mode} is @samp{S}) or clear (@var{mode} is @samp{C}) a
37928breakpoint at @var{addr}.
37929
37930Don't use this packet. Use the @samp{Z} and @samp{z} packets instead
37931(@pxref{insert breakpoint or watchpoint packet}).
37932
37933@cindex @samp{bc} packet
37934@anchor{bc}
37935@item bc
37936Backward continue. Execute the target system in reverse. No parameter.
37937@xref{Reverse Execution}, for more information.
37938
37939Reply:
37940@xref{Stop Reply Packets}, for the reply specifications.
37941
37942@cindex @samp{bs} packet
37943@anchor{bs}
37944@item bs
37945Backward single step. Execute one instruction in reverse. No parameter.
37946@xref{Reverse Execution}, for more information.
37947
37948Reply:
37949@xref{Stop Reply Packets}, for the reply specifications.
37950
37951@item c @r{[}@var{addr}@r{]}
37952@cindex @samp{c} packet
37953Continue. @var{addr} is address to resume. If @var{addr} is omitted,
37954resume at current address.
37955
37956This packet is deprecated for multi-threading support. @xref{vCont
37957packet}.
37958
37959Reply:
37960@xref{Stop Reply Packets}, for the reply specifications.
37961
37962@item C @var{sig}@r{[};@var{addr}@r{]}
37963@cindex @samp{C} packet
37964Continue with signal @var{sig} (hex signal number). If
37965@samp{;@var{addr}} is omitted, resume at same address.
37966
37967This packet is deprecated for multi-threading support. @xref{vCont
37968packet}.
37969
37970Reply:
37971@xref{Stop Reply Packets}, for the reply specifications.
37972
37973@item d
37974@cindex @samp{d} packet
37975Toggle debug flag.
37976
37977Don't use this packet; instead, define a general set packet
37978(@pxref{General Query Packets}).
37979
37980@item D
37981@itemx D;@var{pid}
37982@cindex @samp{D} packet
37983The first form of the packet is used to detach @value{GDBN} from the
37984remote system. It is sent to the remote target
37985before @value{GDBN} disconnects via the @code{detach} command.
37986
37987The second form, including a process ID, is used when multiprocess
37988protocol extensions are enabled (@pxref{multiprocess extensions}), to
37989detach only a specific process. The @var{pid} is specified as a
37990big-endian hex string.
37991
37992Reply:
37993@table @samp
37994@item OK
37995for success
37996@item E @var{NN}
37997for an error
37998@end table
37999
38000@item F @var{RC},@var{EE},@var{CF};@var{XX}
38001@cindex @samp{F} packet
38002A reply from @value{GDBN} to an @samp{F} packet sent by the target.
38003This is part of the File-I/O protocol extension. @xref{File-I/O
38004Remote Protocol Extension}, for the specification.
38005
38006@item g
38007@anchor{read registers packet}
38008@cindex @samp{g} packet
38009Read general registers.
38010
38011Reply:
38012@table @samp
38013@item @var{XX@dots{}}
38014Each byte of register data is described by two hex digits. The bytes
38015with the register are transmitted in target byte order. The size of
38016each register and their position within the @samp{g} packet are
38017determined by the @value{GDBN} internal gdbarch functions
38018@code{DEPRECATED_REGISTER_RAW_SIZE} and @code{gdbarch_register_name}. The
38019specification of several standard @samp{g} packets is specified below.
38020
38021When reading registers from a trace frame (@pxref{Analyze Collected
38022Data,,Using the Collected Data}), the stub may also return a string of
38023literal @samp{x}'s in place of the register data digits, to indicate
38024that the corresponding register has not been collected, thus its value
38025is unavailable. For example, for an architecture with 4 registers of
380264 bytes each, the following reply indicates to @value{GDBN} that
38027registers 0 and 2 have not been collected, while registers 1 and 3
38028have been collected, and both have zero value:
38029
38030@smallexample
38031-> @code{g}
38032<- @code{xxxxxxxx00000000xxxxxxxx00000000}
38033@end smallexample
38034
38035@item E @var{NN}
38036for an error.
38037@end table
38038
38039@item G @var{XX@dots{}}
38040@cindex @samp{G} packet
38041Write general registers. @xref{read registers packet}, for a
38042description of the @var{XX@dots{}} data.
38043
38044Reply:
38045@table @samp
38046@item OK
38047for success
38048@item E @var{NN}
38049for an error
38050@end table
38051
38052@item H @var{op} @var{thread-id}
38053@cindex @samp{H} packet
38054Set thread for subsequent operations (@samp{m}, @samp{M}, @samp{g},
38055@samp{G}, et.al.). @var{op} depends on the operation to be performed:
38056it should be @samp{c} for step and continue operations (note that this
38057is deprecated, supporting the @samp{vCont} command is a better
38058option), @samp{g} for other operations. The thread designator
38059@var{thread-id} has the format and interpretation described in
38060@ref{thread-id syntax}.
38061
38062Reply:
38063@table @samp
38064@item OK
38065for success
38066@item E @var{NN}
38067for an error
38068@end table
38069
38070@c FIXME: JTC:
38071@c 'H': How restrictive (or permissive) is the thread model. If a
38072@c thread is selected and stopped, are other threads allowed
38073@c to continue to execute? As I mentioned above, I think the
38074@c semantics of each command when a thread is selected must be
38075@c described. For example:
38076@c
38077@c 'g': If the stub supports threads and a specific thread is
38078@c selected, returns the register block from that thread;
38079@c otherwise returns current registers.
38080@c
38081@c 'G' If the stub supports threads and a specific thread is
38082@c selected, sets the registers of the register block of
38083@c that thread; otherwise sets current registers.
38084
38085@item i @r{[}@var{addr}@r{[},@var{nnn}@r{]]}
38086@anchor{cycle step packet}
38087@cindex @samp{i} packet
38088Step the remote target by a single clock cycle. If @samp{,@var{nnn}} is
38089present, cycle step @var{nnn} cycles. If @var{addr} is present, cycle
38090step starting at that address.
38091
38092@item I
38093@cindex @samp{I} packet
38094Signal, then cycle step. @xref{step with signal packet}. @xref{cycle
38095step packet}.
38096
38097@item k
38098@cindex @samp{k} packet
38099Kill request.
38100
38101FIXME: @emph{There is no description of how to operate when a specific
38102thread context has been selected (i.e.@: does 'k' kill only that
38103thread?)}.
38104
38105@item m @var{addr},@var{length}
38106@cindex @samp{m} packet
38107Read @var{length} bytes of memory starting at address @var{addr}.
38108Note that @var{addr} may not be aligned to any particular boundary.
38109
38110The stub need not use any particular size or alignment when gathering
38111data from memory for the response; even if @var{addr} is word-aligned
38112and @var{length} is a multiple of the word size, the stub is free to
38113use byte accesses, or not. For this reason, this packet may not be
38114suitable for accessing memory-mapped I/O devices.
38115@cindex alignment of remote memory accesses
38116@cindex size of remote memory accesses
38117@cindex memory, alignment and size of remote accesses
38118
38119Reply:
38120@table @samp
38121@item @var{XX@dots{}}
38122Memory contents; each byte is transmitted as a two-digit hexadecimal
38123number. The reply may contain fewer bytes than requested if the
38124server was able to read only part of the region of memory.
38125@item E @var{NN}
38126@var{NN} is errno
38127@end table
38128
38129@item M @var{addr},@var{length}:@var{XX@dots{}}
38130@cindex @samp{M} packet
38131Write @var{length} bytes of memory starting at address @var{addr}.
38132@var{XX@dots{}} is the data; each byte is transmitted as a two-digit
38133hexadecimal number.
38134
38135Reply:
38136@table @samp
38137@item OK
38138for success
38139@item E @var{NN}
38140for an error (this includes the case where only part of the data was
38141written).
38142@end table
38143
38144@item p @var{n}
38145@cindex @samp{p} packet
38146Read the value of register @var{n}; @var{n} is in hex.
38147@xref{read registers packet}, for a description of how the returned
38148register value is encoded.
38149
38150Reply:
38151@table @samp
38152@item @var{XX@dots{}}
38153the register's value
38154@item E @var{NN}
38155for an error
38156@item @w{}
38157Indicating an unrecognized @var{query}.
38158@end table
38159
38160@item P @var{n@dots{}}=@var{r@dots{}}
38161@anchor{write register packet}
38162@cindex @samp{P} packet
38163Write register @var{n@dots{}} with value @var{r@dots{}}. The register
38164number @var{n} is in hexadecimal, and @var{r@dots{}} contains two hex
38165digits for each byte in the register (target byte order).
38166
38167Reply:
38168@table @samp
38169@item OK
38170for success
38171@item E @var{NN}
38172for an error
38173@end table
38174
38175@item q @var{name} @var{params}@dots{}
38176@itemx Q @var{name} @var{params}@dots{}
38177@cindex @samp{q} packet
38178@cindex @samp{Q} packet
38179General query (@samp{q}) and set (@samp{Q}). These packets are
38180described fully in @ref{General Query Packets}.
38181
38182@item r
38183@cindex @samp{r} packet
38184Reset the entire system.
38185
38186Don't use this packet; use the @samp{R} packet instead.
38187
38188@item R @var{XX}
38189@cindex @samp{R} packet
38190Restart the program being debugged. @var{XX}, while needed, is ignored.
38191This packet is only available in extended mode (@pxref{extended mode}).
38192
38193The @samp{R} packet has no reply.
38194
38195@item s @r{[}@var{addr}@r{]}
38196@cindex @samp{s} packet
38197Single step. @var{addr} is the address at which to resume. If
38198@var{addr} is omitted, resume at same address.
38199
38200This packet is deprecated for multi-threading support. @xref{vCont
38201packet}.
38202
38203Reply:
38204@xref{Stop Reply Packets}, for the reply specifications.
38205
38206@item S @var{sig}@r{[};@var{addr}@r{]}
38207@anchor{step with signal packet}
38208@cindex @samp{S} packet
38209Step with signal. This is analogous to the @samp{C} packet, but
38210requests a single-step, rather than a normal resumption of execution.
38211
38212This packet is deprecated for multi-threading support. @xref{vCont
38213packet}.
38214
38215Reply:
38216@xref{Stop Reply Packets}, for the reply specifications.
38217
38218@item t @var{addr}:@var{PP},@var{MM}
38219@cindex @samp{t} packet
38220Search backwards starting at address @var{addr} for a match with pattern
38221@var{PP} and mask @var{MM}. @var{PP} and @var{MM} are 4 bytes.
38222@var{addr} must be at least 3 digits.
38223
38224@item T @var{thread-id}
38225@cindex @samp{T} packet
38226Find out if the thread @var{thread-id} is alive. @xref{thread-id syntax}.
38227
38228Reply:
38229@table @samp
38230@item OK
38231thread is still alive
38232@item E @var{NN}
38233thread is dead
38234@end table
38235
38236@item v
38237Packets starting with @samp{v} are identified by a multi-letter name,
38238up to the first @samp{;} or @samp{?} (or the end of the packet).
38239
38240@item vAttach;@var{pid}
38241@cindex @samp{vAttach} packet
38242Attach to a new process with the specified process ID @var{pid}.
38243The process ID is a
38244hexadecimal integer identifying the process. In all-stop mode, all
38245threads in the attached process are stopped; in non-stop mode, it may be
38246attached without being stopped if that is supported by the target.
38247
38248@c In non-stop mode, on a successful vAttach, the stub should set the
38249@c current thread to a thread of the newly-attached process. After
38250@c attaching, GDB queries for the attached process's thread ID with qC.
38251@c Also note that, from a user perspective, whether or not the
38252@c target is stopped on attach in non-stop mode depends on whether you
38253@c use the foreground or background version of the attach command, not
38254@c on what vAttach does; GDB does the right thing with respect to either
38255@c stopping or restarting threads.
38256
38257This packet is only available in extended mode (@pxref{extended mode}).
38258
38259Reply:
38260@table @samp
38261@item E @var{nn}
38262for an error
38263@item @r{Any stop packet}
38264for success in all-stop mode (@pxref{Stop Reply Packets})
38265@item OK
38266for success in non-stop mode (@pxref{Remote Non-Stop})
38267@end table
38268
38269@item vCont@r{[};@var{action}@r{[}:@var{thread-id}@r{]]}@dots{}
38270@cindex @samp{vCont} packet
38271@anchor{vCont packet}
38272Resume the inferior, specifying different actions for each thread.
38273If an action is specified with no @var{thread-id}, then it is applied to any
38274threads that don't have a specific action specified; if no default action is
38275specified then other threads should remain stopped in all-stop mode and
38276in their current state in non-stop mode.
38277Specifying multiple
38278default actions is an error; specifying no actions is also an error.
38279Thread IDs are specified using the syntax described in @ref{thread-id syntax}.
38280
38281Currently supported actions are:
38282
38283@table @samp
38284@item c
38285Continue.
38286@item C @var{sig}
38287Continue with signal @var{sig}. The signal @var{sig} should be two hex digits.
38288@item s
38289Step.
38290@item S @var{sig}
38291Step with signal @var{sig}. The signal @var{sig} should be two hex digits.
38292@item t
38293Stop.
38294@item r @var{start},@var{end}
38295Step once, and then keep stepping as long as the thread stops at
38296addresses between @var{start} (inclusive) and @var{end} (exclusive).
38297The remote stub reports a stop reply when either the thread goes out
38298of the range or is stopped due to an unrelated reason, such as hitting
38299a breakpoint. @xref{range stepping}.
38300
38301If the range is empty (@var{start} == @var{end}), then the action
38302becomes equivalent to the @samp{s} action. In other words,
38303single-step once, and report the stop (even if the stepped instruction
38304jumps to @var{start}).
38305
38306(A stop reply may be sent at any point even if the PC is still within
38307the stepping range; for example, it is valid to implement this packet
38308in a degenerate way as a single instruction step operation.)
38309
38310@end table
38311
38312The optional argument @var{addr} normally associated with the
38313@samp{c}, @samp{C}, @samp{s}, and @samp{S} packets is
38314not supported in @samp{vCont}.
38315
38316The @samp{t} action is only relevant in non-stop mode
38317(@pxref{Remote Non-Stop}) and may be ignored by the stub otherwise.
38318A stop reply should be generated for any affected thread not already stopped.
38319When a thread is stopped by means of a @samp{t} action,
38320the corresponding stop reply should indicate that the thread has stopped with
38321signal @samp{0}, regardless of whether the target uses some other signal
38322as an implementation detail.
38323
38324The stub must support @samp{vCont} if it reports support for
38325multiprocess extensions (@pxref{multiprocess extensions}). Note that in
38326this case @samp{vCont} actions can be specified to apply to all threads
38327in a process by using the @samp{p@var{pid}.-1} form of the
38328@var{thread-id}.
38329
38330Reply:
38331@xref{Stop Reply Packets}, for the reply specifications.
38332
38333@item vCont?
38334@cindex @samp{vCont?} packet
38335Request a list of actions supported by the @samp{vCont} packet.
38336
38337Reply:
38338@table @samp
38339@item vCont@r{[};@var{action}@dots{}@r{]}
38340The @samp{vCont} packet is supported. Each @var{action} is a supported
38341command in the @samp{vCont} packet.
38342@item @w{}
38343The @samp{vCont} packet is not supported.
38344@end table
38345
38346@item vFile:@var{operation}:@var{parameter}@dots{}
38347@cindex @samp{vFile} packet
38348Perform a file operation on the target system. For details,
38349see @ref{Host I/O Packets}.
38350
38351@item vFlashErase:@var{addr},@var{length}
38352@cindex @samp{vFlashErase} packet
38353Direct the stub to erase @var{length} bytes of flash starting at
38354@var{addr}. The region may enclose any number of flash blocks, but
38355its start and end must fall on block boundaries, as indicated by the
38356flash block size appearing in the memory map (@pxref{Memory Map
38357Format}). @value{GDBN} groups flash memory programming operations
38358together, and sends a @samp{vFlashDone} request after each group; the
38359stub is allowed to delay erase operation until the @samp{vFlashDone}
38360packet is received.
38361
38362Reply:
38363@table @samp
38364@item OK
38365for success
38366@item E @var{NN}
38367for an error
38368@end table
38369
38370@item vFlashWrite:@var{addr}:@var{XX@dots{}}
38371@cindex @samp{vFlashWrite} packet
38372Direct the stub to write data to flash address @var{addr}. The data
38373is passed in binary form using the same encoding as for the @samp{X}
38374packet (@pxref{Binary Data}). The memory ranges specified by
38375@samp{vFlashWrite} packets preceding a @samp{vFlashDone} packet must
38376not overlap, and must appear in order of increasing addresses
38377(although @samp{vFlashErase} packets for higher addresses may already
38378have been received; the ordering is guaranteed only between
38379@samp{vFlashWrite} packets). If a packet writes to an address that was
38380neither erased by a preceding @samp{vFlashErase} packet nor by some other
38381target-specific method, the results are unpredictable.
38382
38383
38384Reply:
38385@table @samp
38386@item OK
38387for success
38388@item E.memtype
38389for vFlashWrite addressing non-flash memory
38390@item E @var{NN}
38391for an error
38392@end table
38393
38394@item vFlashDone
38395@cindex @samp{vFlashDone} packet
38396Indicate to the stub that flash programming operation is finished.
38397The stub is permitted to delay or batch the effects of a group of
38398@samp{vFlashErase} and @samp{vFlashWrite} packets until a
38399@samp{vFlashDone} packet is received. The contents of the affected
38400regions of flash memory are unpredictable until the @samp{vFlashDone}
38401request is completed.
38402
38403@item vKill;@var{pid}
38404@cindex @samp{vKill} packet
38405Kill the process with the specified process ID. @var{pid} is a
38406hexadecimal integer identifying the process. This packet is used in
38407preference to @samp{k} when multiprocess protocol extensions are
38408supported; see @ref{multiprocess extensions}.
38409
38410Reply:
38411@table @samp
38412@item E @var{nn}
38413for an error
38414@item OK
38415for success
38416@end table
38417
38418@item vRun;@var{filename}@r{[};@var{argument}@r{]}@dots{}
38419@cindex @samp{vRun} packet
38420Run the program @var{filename}, passing it each @var{argument} on its
38421command line. The file and arguments are hex-encoded strings. If
38422@var{filename} is an empty string, the stub may use a default program
38423(e.g.@: the last program run). The program is created in the stopped
38424state.
38425
38426@c FIXME: What about non-stop mode?
38427
38428This packet is only available in extended mode (@pxref{extended mode}).
38429
38430Reply:
38431@table @samp
38432@item E @var{nn}
38433for an error
38434@item @r{Any stop packet}
38435for success (@pxref{Stop Reply Packets})
38436@end table
38437
38438@item vStopped
38439@cindex @samp{vStopped} packet
38440@xref{Notification Packets}.
38441
38442@item X @var{addr},@var{length}:@var{XX@dots{}}
38443@anchor{X packet}
38444@cindex @samp{X} packet
38445Write data to memory, where the data is transmitted in binary.
38446@var{addr} is address, @var{length} is number of bytes,
38447@samp{@var{XX}@dots{}} is binary data (@pxref{Binary Data}).
38448
38449Reply:
38450@table @samp
38451@item OK
38452for success
38453@item E @var{NN}
38454for an error
38455@end table
38456
38457@item z @var{type},@var{addr},@var{kind}
38458@itemx Z @var{type},@var{addr},@var{kind}
38459@anchor{insert breakpoint or watchpoint packet}
38460@cindex @samp{z} packet
38461@cindex @samp{Z} packets
38462Insert (@samp{Z}) or remove (@samp{z}) a @var{type} breakpoint or
38463watchpoint starting at address @var{address} of kind @var{kind}.
38464
38465Each breakpoint and watchpoint packet @var{type} is documented
38466separately.
38467
38468@emph{Implementation notes: A remote target shall return an empty string
38469for an unrecognized breakpoint or watchpoint packet @var{type}. A
38470remote target shall support either both or neither of a given
38471@samp{Z@var{type}@dots{}} and @samp{z@var{type}@dots{}} packet pair. To
38472avoid potential problems with duplicate packets, the operations should
38473be implemented in an idempotent way.}
38474
38475@item z0,@var{addr},@var{kind}
38476@itemx Z0,@var{addr},@var{kind}@r{[};@var{cond_list}@dots{}@r{]}@r{[};cmds:@var{persist},@var{cmd_list}@dots{}@r{]}
38477@cindex @samp{z0} packet
38478@cindex @samp{Z0} packet
38479Insert (@samp{Z0}) or remove (@samp{z0}) a memory breakpoint at address
38480@var{addr} of type @var{kind}.
38481
38482A memory breakpoint is implemented by replacing the instruction at
38483@var{addr} with a software breakpoint or trap instruction. The
38484@var{kind} is target-specific and typically indicates the size of
38485the breakpoint in bytes that should be inserted. E.g., the @sc{arm}
38486and @sc{mips} can insert either a 2 or 4 byte breakpoint. Some
38487architectures have additional meanings for @var{kind};
38488@var{cond_list} is an optional list of conditional expressions in bytecode
38489form that should be evaluated on the target's side. These are the
38490conditions that should be taken into consideration when deciding if
38491the breakpoint trigger should be reported back to @var{GDBN}.
38492
38493The @var{cond_list} parameter is comprised of a series of expressions,
38494concatenated without separators. Each expression has the following form:
38495
38496@table @samp
38497
38498@item X @var{len},@var{expr}
38499@var{len} is the length of the bytecode expression and @var{expr} is the
38500actual conditional expression in bytecode form.
38501
38502@end table
38503
38504The optional @var{cmd_list} parameter introduces commands that may be
38505run on the target, rather than being reported back to @value{GDBN}.
38506The parameter starts with a numeric flag @var{persist}; if the flag is
38507nonzero, then the breakpoint may remain active and the commands
38508continue to be run even when @value{GDBN} disconnects from the target.
38509Following this flag is a series of expressions concatenated with no
38510separators. Each expression has the following form:
38511
38512@table @samp
38513
38514@item X @var{len},@var{expr}
38515@var{len} is the length of the bytecode expression and @var{expr} is the
38516actual conditional expression in bytecode form.
38517
38518@end table
38519
38520see @ref{Architecture-Specific Protocol Details}.
38521
38522@emph{Implementation note: It is possible for a target to copy or move
38523code that contains memory breakpoints (e.g., when implementing
38524overlays). The behavior of this packet, in the presence of such a
38525target, is not defined.}
38526
38527Reply:
38528@table @samp
38529@item OK
38530success
38531@item @w{}
38532not supported
38533@item E @var{NN}
38534for an error
38535@end table
38536
38537@item z1,@var{addr},@var{kind}
38538@itemx Z1,@var{addr},@var{kind}@r{[};@var{cond_list}@dots{}@r{]}
38539@cindex @samp{z1} packet
38540@cindex @samp{Z1} packet
38541Insert (@samp{Z1}) or remove (@samp{z1}) a hardware breakpoint at
38542address @var{addr}.
38543
38544A hardware breakpoint is implemented using a mechanism that is not
38545dependant on being able to modify the target's memory. @var{kind}
38546and @var{cond_list} have the same meaning as in @samp{Z0} packets.
38547
38548@emph{Implementation note: A hardware breakpoint is not affected by code
38549movement.}
38550
38551Reply:
38552@table @samp
38553@item OK
38554success
38555@item @w{}
38556not supported
38557@item E @var{NN}
38558for an error
38559@end table
38560
38561@item z2,@var{addr},@var{kind}
38562@itemx Z2,@var{addr},@var{kind}
38563@cindex @samp{z2} packet
38564@cindex @samp{Z2} packet
38565Insert (@samp{Z2}) or remove (@samp{z2}) a write watchpoint at @var{addr}.
38566@var{kind} is interpreted as the number of bytes to watch.
38567
38568Reply:
38569@table @samp
38570@item OK
38571success
38572@item @w{}
38573not supported
38574@item E @var{NN}
38575for an error
38576@end table
38577
38578@item z3,@var{addr},@var{kind}
38579@itemx Z3,@var{addr},@var{kind}
38580@cindex @samp{z3} packet
38581@cindex @samp{Z3} packet
38582Insert (@samp{Z3}) or remove (@samp{z3}) a read watchpoint at @var{addr}.
38583@var{kind} is interpreted as the number of bytes to watch.
38584
38585Reply:
38586@table @samp
38587@item OK
38588success
38589@item @w{}
38590not supported
38591@item E @var{NN}
38592for an error
38593@end table
38594
38595@item z4,@var{addr},@var{kind}
38596@itemx Z4,@var{addr},@var{kind}
38597@cindex @samp{z4} packet
38598@cindex @samp{Z4} packet
38599Insert (@samp{Z4}) or remove (@samp{z4}) an access watchpoint at @var{addr}.
38600@var{kind} is interpreted as the number of bytes to watch.
38601
38602Reply:
38603@table @samp
38604@item OK
38605success
38606@item @w{}
38607not supported
38608@item E @var{NN}
38609for an error
38610@end table
38611
38612@end table
38613
38614@node Stop Reply Packets
38615@section Stop Reply Packets
38616@cindex stop reply packets
38617
38618The @samp{C}, @samp{c}, @samp{S}, @samp{s}, @samp{vCont},
38619@samp{vAttach}, @samp{vRun}, @samp{vStopped}, and @samp{?} packets can
38620receive any of the below as a reply. Except for @samp{?}
38621and @samp{vStopped}, that reply is only returned
38622when the target halts. In the below the exact meaning of @dfn{signal
38623number} is defined by the header @file{include/gdb/signals.h} in the
38624@value{GDBN} source code.
38625
38626As in the description of request packets, we include spaces in the
38627reply templates for clarity; these are not part of the reply packet's
38628syntax. No @value{GDBN} stop reply packet uses spaces to separate its
38629components.
38630
38631@table @samp
38632
38633@item S @var{AA}
38634The program received signal number @var{AA} (a two-digit hexadecimal
38635number). This is equivalent to a @samp{T} response with no
38636@var{n}:@var{r} pairs.
38637
38638@item T @var{AA} @var{n1}:@var{r1};@var{n2}:@var{r2};@dots{}
38639@cindex @samp{T} packet reply
38640The program received signal number @var{AA} (a two-digit hexadecimal
38641number). This is equivalent to an @samp{S} response, except that the
38642@samp{@var{n}:@var{r}} pairs can carry values of important registers
38643and other information directly in the stop reply packet, reducing
38644round-trip latency. Single-step and breakpoint traps are reported
38645this way. Each @samp{@var{n}:@var{r}} pair is interpreted as follows:
38646
38647@itemize @bullet
38648@item
38649If @var{n} is a hexadecimal number, it is a register number, and the
38650corresponding @var{r} gives that register's value. @var{r} is a
38651series of bytes in target byte order, with each byte given by a
38652two-digit hex number.
38653
38654@item
38655If @var{n} is @samp{thread}, then @var{r} is the @var{thread-id} of
38656the stopped thread, as specified in @ref{thread-id syntax}.
38657
38658@item
38659If @var{n} is @samp{core}, then @var{r} is the hexadecimal number of
38660the core on which the stop event was detected.
38661
38662@item
38663If @var{n} is a recognized @dfn{stop reason}, it describes a more
38664specific event that stopped the target. The currently defined stop
38665reasons are listed below. @var{aa} should be @samp{05}, the trap
38666signal. At most one stop reason should be present.
38667
38668@item
38669Otherwise, @value{GDBN} should ignore this @samp{@var{n}:@var{r}} pair
38670and go on to the next; this allows us to extend the protocol in the
38671future.
38672@end itemize
38673
38674The currently defined stop reasons are:
38675
38676@table @samp
38677@item watch
38678@itemx rwatch
38679@itemx awatch
38680The packet indicates a watchpoint hit, and @var{r} is the data address, in
38681hex.
38682
38683@cindex shared library events, remote reply
38684@item library
38685The packet indicates that the loaded libraries have changed.
38686@value{GDBN} should use @samp{qXfer:libraries:read} to fetch a new
38687list of loaded libraries. @var{r} is ignored.
38688
38689@cindex replay log events, remote reply
38690@item replaylog
38691The packet indicates that the target cannot continue replaying
38692logged execution events, because it has reached the end (or the
38693beginning when executing backward) of the log. The value of @var{r}
38694will be either @samp{begin} or @samp{end}. @xref{Reverse Execution},
38695for more information.
38696@end table
38697
38698@item W @var{AA}
38699@itemx W @var{AA} ; process:@var{pid}
38700The process exited, and @var{AA} is the exit status. This is only
38701applicable to certain targets.
38702
38703The second form of the response, including the process ID of the exited
38704process, can be used only when @value{GDBN} has reported support for
38705multiprocess protocol extensions; see @ref{multiprocess extensions}.
38706The @var{pid} is formatted as a big-endian hex string.
38707
38708@item X @var{AA}
38709@itemx X @var{AA} ; process:@var{pid}
38710The process terminated with signal @var{AA}.
38711
38712The second form of the response, including the process ID of the
38713terminated process, can be used only when @value{GDBN} has reported
38714support for multiprocess protocol extensions; see @ref{multiprocess
38715extensions}. The @var{pid} is formatted as a big-endian hex string.
38716
38717@item O @var{XX}@dots{}
38718@samp{@var{XX}@dots{}} is hex encoding of @sc{ascii} data, to be
38719written as the program's console output. This can happen at any time
38720while the program is running and the debugger should continue to wait
38721for @samp{W}, @samp{T}, etc. This reply is not permitted in non-stop mode.
38722
38723@item F @var{call-id},@var{parameter}@dots{}
38724@var{call-id} is the identifier which says which host system call should
38725be called. This is just the name of the function. Translation into the
38726correct system call is only applicable as it's defined in @value{GDBN}.
38727@xref{File-I/O Remote Protocol Extension}, for a list of implemented
38728system calls.
38729
38730@samp{@var{parameter}@dots{}} is a list of parameters as defined for
38731this very system call.
38732
38733The target replies with this packet when it expects @value{GDBN} to
38734call a host system call on behalf of the target. @value{GDBN} replies
38735with an appropriate @samp{F} packet and keeps up waiting for the next
38736reply packet from the target. The latest @samp{C}, @samp{c}, @samp{S}
38737or @samp{s} action is expected to be continued. @xref{File-I/O Remote
38738Protocol Extension}, for more details.
38739
38740@end table
38741
38742@node General Query Packets
38743@section General Query Packets
38744@cindex remote query requests
38745
38746Packets starting with @samp{q} are @dfn{general query packets};
38747packets starting with @samp{Q} are @dfn{general set packets}. General
38748query and set packets are a semi-unified form for retrieving and
38749sending information to and from the stub.
38750
38751The initial letter of a query or set packet is followed by a name
38752indicating what sort of thing the packet applies to. For example,
38753@value{GDBN} may use a @samp{qSymbol} packet to exchange symbol
38754definitions with the stub. These packet names follow some
38755conventions:
38756
38757@itemize @bullet
38758@item
38759The name must not contain commas, colons or semicolons.
38760@item
38761Most @value{GDBN} query and set packets have a leading upper case
38762letter.
38763@item
38764The names of custom vendor packets should use a company prefix, in
38765lower case, followed by a period. For example, packets designed at
38766the Acme Corporation might begin with @samp{qacme.foo} (for querying
38767foos) or @samp{Qacme.bar} (for setting bars).
38768@end itemize
38769
38770The name of a query or set packet should be separated from any
38771parameters by a @samp{:}; the parameters themselves should be
38772separated by @samp{,} or @samp{;}. Stubs must be careful to match the
38773full packet name, and check for a separator or the end of the packet,
38774in case two packet names share a common prefix. New packets should not begin
38775with @samp{qC}, @samp{qP}, or @samp{qL}@footnote{The @samp{qP} and @samp{qL}
38776packets predate these conventions, and have arguments without any terminator
38777for the packet name; we suspect they are in widespread use in places that
38778are difficult to upgrade. The @samp{qC} packet has no arguments, but some
38779existing stubs (e.g.@: RedBoot) are known to not check for the end of the
38780packet.}.
38781
38782Like the descriptions of the other packets, each description here
38783has a template showing the packet's overall syntax, followed by an
38784explanation of the packet's meaning. We include spaces in some of the
38785templates for clarity; these are not part of the packet's syntax. No
38786@value{GDBN} packet uses spaces to separate its components.
38787
38788Here are the currently defined query and set packets:
38789
38790@table @samp
38791
38792@item QAgent:1
38793@itemx QAgent:0
38794Turn on or off the agent as a helper to perform some debugging operations
38795delegated from @value{GDBN} (@pxref{Control Agent}).
38796
38797@item QAllow:@var{op}:@var{val}@dots{}
38798@cindex @samp{QAllow} packet
38799Specify which operations @value{GDBN} expects to request of the
38800target, as a semicolon-separated list of operation name and value
38801pairs. Possible values for @var{op} include @samp{WriteReg},
38802@samp{WriteMem}, @samp{InsertBreak}, @samp{InsertTrace},
38803@samp{InsertFastTrace}, and @samp{Stop}. @var{val} is either 0,
38804indicating that @value{GDBN} will not request the operation, or 1,
38805indicating that it may. (The target can then use this to set up its
38806own internals optimally, for instance if the debugger never expects to
38807insert breakpoints, it may not need to install its own trap handler.)
38808
38809@item qC
38810@cindex current thread, remote request
38811@cindex @samp{qC} packet
38812Return the current thread ID.
38813
38814Reply:
38815@table @samp
38816@item QC @var{thread-id}
38817Where @var{thread-id} is a thread ID as documented in
38818@ref{thread-id syntax}.
38819@item @r{(anything else)}
38820Any other reply implies the old thread ID.
38821@end table
38822
38823@item qCRC:@var{addr},@var{length}
38824@cindex CRC of memory block, remote request
38825@cindex @samp{qCRC} packet
38826Compute the CRC checksum of a block of memory using CRC-32 defined in
38827IEEE 802.3. The CRC is computed byte at a time, taking the most
38828significant bit of each byte first. The initial pattern code
38829@code{0xffffffff} is used to ensure leading zeros affect the CRC.
38830
38831@emph{Note:} This is the same CRC used in validating separate debug
38832files (@pxref{Separate Debug Files, , Debugging Information in Separate
38833Files}). However the algorithm is slightly different. When validating
38834separate debug files, the CRC is computed taking the @emph{least}
38835significant bit of each byte first, and the final result is inverted to
38836detect trailing zeros.
38837
38838Reply:
38839@table @samp
38840@item E @var{NN}
38841An error (such as memory fault)
38842@item C @var{crc32}
38843The specified memory region's checksum is @var{crc32}.
38844@end table
38845
38846@item QDisableRandomization:@var{value}
38847@cindex disable address space randomization, remote request
38848@cindex @samp{QDisableRandomization} packet
38849Some target operating systems will randomize the virtual address space
38850of the inferior process as a security feature, but provide a feature
38851to disable such randomization, e.g.@: to allow for a more deterministic
38852debugging experience. On such systems, this packet with a @var{value}
38853of 1 directs the target to disable address space randomization for
38854processes subsequently started via @samp{vRun} packets, while a packet
38855with a @var{value} of 0 tells the target to enable address space
38856randomization.
38857
38858This packet is only available in extended mode (@pxref{extended mode}).
38859
38860Reply:
38861@table @samp
38862@item OK
38863The request succeeded.
38864
38865@item E @var{nn}
38866An error occurred. @var{nn} are hex digits.
38867
38868@item @w{}
38869An empty reply indicates that @samp{QDisableRandomization} is not supported
38870by the stub.
38871@end table
38872
38873This packet is not probed by default; the remote stub must request it,
38874by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
38875This should only be done on targets that actually support disabling
38876address space randomization.
38877
38878@item qfThreadInfo
38879@itemx qsThreadInfo
38880@cindex list active threads, remote request
38881@cindex @samp{qfThreadInfo} packet
38882@cindex @samp{qsThreadInfo} packet
38883Obtain a list of all active thread IDs from the target (OS). Since there
38884may be too many active threads to fit into one reply packet, this query
38885works iteratively: it may require more than one query/reply sequence to
38886obtain the entire list of threads. The first query of the sequence will
38887be the @samp{qfThreadInfo} query; subsequent queries in the
38888sequence will be the @samp{qsThreadInfo} query.
38889
38890NOTE: This packet replaces the @samp{qL} query (see below).
38891
38892Reply:
38893@table @samp
38894@item m @var{thread-id}
38895A single thread ID
38896@item m @var{thread-id},@var{thread-id}@dots{}
38897a comma-separated list of thread IDs
38898@item l
38899(lower case letter @samp{L}) denotes end of list.
38900@end table
38901
38902In response to each query, the target will reply with a list of one or
38903more thread IDs, separated by commas.
38904@value{GDBN} will respond to each reply with a request for more thread
38905ids (using the @samp{qs} form of the query), until the target responds
38906with @samp{l} (lower-case ell, for @dfn{last}).
38907Refer to @ref{thread-id syntax}, for the format of the @var{thread-id}
38908fields.
38909
38910@item qGetTLSAddr:@var{thread-id},@var{offset},@var{lm}
38911@cindex get thread-local storage address, remote request
38912@cindex @samp{qGetTLSAddr} packet
38913Fetch the address associated with thread local storage specified
38914by @var{thread-id}, @var{offset}, and @var{lm}.
38915
38916@var{thread-id} is the thread ID associated with the
38917thread for which to fetch the TLS address. @xref{thread-id syntax}.
38918
38919@var{offset} is the (big endian, hex encoded) offset associated with the
38920thread local variable. (This offset is obtained from the debug
38921information associated with the variable.)
38922
38923@var{lm} is the (big endian, hex encoded) OS/ABI-specific encoding of the
38924load module associated with the thread local storage. For example,
38925a @sc{gnu}/Linux system will pass the link map address of the shared
38926object associated with the thread local storage under consideration.
38927Other operating environments may choose to represent the load module
38928differently, so the precise meaning of this parameter will vary.
38929
38930Reply:
38931@table @samp
38932@item @var{XX}@dots{}
38933Hex encoded (big endian) bytes representing the address of the thread
38934local storage requested.
38935
38936@item E @var{nn}
38937An error occurred. @var{nn} are hex digits.
38938
38939@item @w{}
38940An empty reply indicates that @samp{qGetTLSAddr} is not supported by the stub.
38941@end table
38942
38943@item qGetTIBAddr:@var{thread-id}
38944@cindex get thread information block address
38945@cindex @samp{qGetTIBAddr} packet
38946Fetch address of the Windows OS specific Thread Information Block.
38947
38948@var{thread-id} is the thread ID associated with the thread.
38949
38950Reply:
38951@table @samp
38952@item @var{XX}@dots{}
38953Hex encoded (big endian) bytes representing the linear address of the
38954thread information block.
38955
38956@item E @var{nn}
38957An error occured. This means that either the thread was not found, or the
38958address could not be retrieved.
38959
38960@item @w{}
38961An empty reply indicates that @samp{qGetTIBAddr} is not supported by the stub.
38962@end table
38963
38964@item qL @var{startflag} @var{threadcount} @var{nextthread}
38965Obtain thread information from RTOS. Where: @var{startflag} (one hex
38966digit) is one to indicate the first query and zero to indicate a
38967subsequent query; @var{threadcount} (two hex digits) is the maximum
38968number of threads the response packet can contain; and @var{nextthread}
38969(eight hex digits), for subsequent queries (@var{startflag} is zero), is
38970returned in the response as @var{argthread}.
38971
38972Don't use this packet; use the @samp{qfThreadInfo} query instead (see above).
38973
38974Reply:
38975@table @samp
38976@item qM @var{count} @var{done} @var{argthread} @var{thread}@dots{}
38977Where: @var{count} (two hex digits) is the number of threads being
38978returned; @var{done} (one hex digit) is zero to indicate more threads
38979and one indicates no further threads; @var{argthreadid} (eight hex
38980digits) is @var{nextthread} from the request packet; @var{thread}@dots{}
38981is a sequence of thread IDs from the target. @var{threadid} (eight hex
38982digits). See @code{remote.c:parse_threadlist_response()}.
38983@end table
38984
38985@item qOffsets
38986@cindex section offsets, remote request
38987@cindex @samp{qOffsets} packet
38988Get section offsets that the target used when relocating the downloaded
38989image.
38990
38991Reply:
38992@table @samp
38993@item Text=@var{xxx};Data=@var{yyy}@r{[};Bss=@var{zzz}@r{]}
38994Relocate the @code{Text} section by @var{xxx} from its original address.
38995Relocate the @code{Data} section by @var{yyy} from its original address.
38996If the object file format provides segment information (e.g.@: @sc{elf}
38997@samp{PT_LOAD} program headers), @value{GDBN} will relocate entire
38998segments by the supplied offsets.
38999
39000@emph{Note: while a @code{Bss} offset may be included in the response,
39001@value{GDBN} ignores this and instead applies the @code{Data} offset
39002to the @code{Bss} section.}
39003
39004@item TextSeg=@var{xxx}@r{[};DataSeg=@var{yyy}@r{]}
39005Relocate the first segment of the object file, which conventionally
39006contains program code, to a starting address of @var{xxx}. If
39007@samp{DataSeg} is specified, relocate the second segment, which
39008conventionally contains modifiable data, to a starting address of
39009@var{yyy}. @value{GDBN} will report an error if the object file
39010does not contain segment information, or does not contain at least
39011as many segments as mentioned in the reply. Extra segments are
39012kept at fixed offsets relative to the last relocated segment.
39013@end table
39014
39015@item qP @var{mode} @var{thread-id}
39016@cindex thread information, remote request
39017@cindex @samp{qP} packet
39018Returns information on @var{thread-id}. Where: @var{mode} is a hex
39019encoded 32 bit mode; @var{thread-id} is a thread ID
39020(@pxref{thread-id syntax}).
39021
39022Don't use this packet; use the @samp{qThreadExtraInfo} query instead
39023(see below).
39024
39025Reply: see @code{remote.c:remote_unpack_thread_info_response()}.
39026
39027@item QNonStop:1
39028@itemx QNonStop:0
39029@cindex non-stop mode, remote request
39030@cindex @samp{QNonStop} packet
39031@anchor{QNonStop}
39032Enter non-stop (@samp{QNonStop:1}) or all-stop (@samp{QNonStop:0}) mode.
39033@xref{Remote Non-Stop}, for more information.
39034
39035Reply:
39036@table @samp
39037@item OK
39038The request succeeded.
39039
39040@item E @var{nn}
39041An error occurred. @var{nn} are hex digits.
39042
39043@item @w{}
39044An empty reply indicates that @samp{QNonStop} is not supported by
39045the stub.
39046@end table
39047
39048This packet is not probed by default; the remote stub must request it,
39049by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39050Use of this packet is controlled by the @code{set non-stop} command;
39051@pxref{Non-Stop Mode}.
39052
39053@item QPassSignals: @var{signal} @r{[};@var{signal}@r{]}@dots{}
39054@cindex pass signals to inferior, remote request
39055@cindex @samp{QPassSignals} packet
39056@anchor{QPassSignals}
39057Each listed @var{signal} should be passed directly to the inferior process.
39058Signals are numbered identically to continue packets and stop replies
39059(@pxref{Stop Reply Packets}). Each @var{signal} list item should be
39060strictly greater than the previous item. These signals do not need to stop
39061the inferior, or be reported to @value{GDBN}. All other signals should be
39062reported to @value{GDBN}. Multiple @samp{QPassSignals} packets do not
39063combine; any earlier @samp{QPassSignals} list is completely replaced by the
39064new list. This packet improves performance when using @samp{handle
39065@var{signal} nostop noprint pass}.
39066
39067Reply:
39068@table @samp
39069@item OK
39070The request succeeded.
39071
39072@item E @var{nn}
39073An error occurred. @var{nn} are hex digits.
39074
39075@item @w{}
39076An empty reply indicates that @samp{QPassSignals} is not supported by
39077the stub.
39078@end table
39079
39080Use of this packet is controlled by the @code{set remote pass-signals}
39081command (@pxref{Remote Configuration, set remote pass-signals}).
39082This packet is not probed by default; the remote stub must request it,
39083by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39084
39085@item QProgramSignals: @var{signal} @r{[};@var{signal}@r{]}@dots{}
39086@cindex signals the inferior may see, remote request
39087@cindex @samp{QProgramSignals} packet
39088@anchor{QProgramSignals}
39089Each listed @var{signal} may be delivered to the inferior process.
39090Others should be silently discarded.
39091
39092In some cases, the remote stub may need to decide whether to deliver a
39093signal to the program or not without @value{GDBN} involvement. One
39094example of that is while detaching --- the program's threads may have
39095stopped for signals that haven't yet had a chance of being reported to
39096@value{GDBN}, and so the remote stub can use the signal list specified
39097by this packet to know whether to deliver or ignore those pending
39098signals.
39099
39100This does not influence whether to deliver a signal as requested by a
39101resumption packet (@pxref{vCont packet}).
39102
39103Signals are numbered identically to continue packets and stop replies
39104(@pxref{Stop Reply Packets}). Each @var{signal} list item should be
39105strictly greater than the previous item. Multiple
39106@samp{QProgramSignals} packets do not combine; any earlier
39107@samp{QProgramSignals} list is completely replaced by the new list.
39108
39109Reply:
39110@table @samp
39111@item OK
39112The request succeeded.
39113
39114@item E @var{nn}
39115An error occurred. @var{nn} are hex digits.
39116
39117@item @w{}
39118An empty reply indicates that @samp{QProgramSignals} is not supported
39119by the stub.
39120@end table
39121
39122Use of this packet is controlled by the @code{set remote program-signals}
39123command (@pxref{Remote Configuration, set remote program-signals}).
39124This packet is not probed by default; the remote stub must request it,
39125by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39126
39127@item qRcmd,@var{command}
39128@cindex execute remote command, remote request
39129@cindex @samp{qRcmd} packet
39130@var{command} (hex encoded) is passed to the local interpreter for
39131execution. Invalid commands should be reported using the output
39132string. Before the final result packet, the target may also respond
39133with a number of intermediate @samp{O@var{output}} console output
39134packets. @emph{Implementors should note that providing access to a
39135stubs's interpreter may have security implications}.
39136
39137Reply:
39138@table @samp
39139@item OK
39140A command response with no output.
39141@item @var{OUTPUT}
39142A command response with the hex encoded output string @var{OUTPUT}.
39143@item E @var{NN}
39144Indicate a badly formed request.
39145@item @w{}
39146An empty reply indicates that @samp{qRcmd} is not recognized.
39147@end table
39148
39149(Note that the @code{qRcmd} packet's name is separated from the
39150command by a @samp{,}, not a @samp{:}, contrary to the naming
39151conventions above. Please don't use this packet as a model for new
39152packets.)
39153
39154@item qSearch:memory:@var{address};@var{length};@var{search-pattern}
39155@cindex searching memory, in remote debugging
39156@ifnotinfo
39157@cindex @samp{qSearch:memory} packet
39158@end ifnotinfo
39159@cindex @samp{qSearch memory} packet
39160@anchor{qSearch memory}
39161Search @var{length} bytes at @var{address} for @var{search-pattern}.
39162@var{address} and @var{length} are encoded in hex.
39163@var{search-pattern} is a sequence of bytes, hex encoded.
39164
39165Reply:
39166@table @samp
39167@item 0
39168The pattern was not found.
39169@item 1,address
39170The pattern was found at @var{address}.
39171@item E @var{NN}
39172A badly formed request or an error was encountered while searching memory.
39173@item @w{}
39174An empty reply indicates that @samp{qSearch:memory} is not recognized.
39175@end table
39176
39177@item QStartNoAckMode
39178@cindex @samp{QStartNoAckMode} packet
39179@anchor{QStartNoAckMode}
39180Request that the remote stub disable the normal @samp{+}/@samp{-}
39181protocol acknowledgments (@pxref{Packet Acknowledgment}).
39182
39183Reply:
39184@table @samp
39185@item OK
39186The stub has switched to no-acknowledgment mode.
39187@value{GDBN} acknowledges this reponse,
39188but neither the stub nor @value{GDBN} shall send or expect further
39189@samp{+}/@samp{-} acknowledgments in the current connection.
39190@item @w{}
39191An empty reply indicates that the stub does not support no-acknowledgment mode.
39192@end table
39193
39194@item qSupported @r{[}:@var{gdbfeature} @r{[};@var{gdbfeature}@r{]}@dots{} @r{]}
39195@cindex supported packets, remote query
39196@cindex features of the remote protocol
39197@cindex @samp{qSupported} packet
39198@anchor{qSupported}
39199Tell the remote stub about features supported by @value{GDBN}, and
39200query the stub for features it supports. This packet allows
39201@value{GDBN} and the remote stub to take advantage of each others'
39202features. @samp{qSupported} also consolidates multiple feature probes
39203at startup, to improve @value{GDBN} performance---a single larger
39204packet performs better than multiple smaller probe packets on
39205high-latency links. Some features may enable behavior which must not
39206be on by default, e.g.@: because it would confuse older clients or
39207stubs. Other features may describe packets which could be
39208automatically probed for, but are not. These features must be
39209reported before @value{GDBN} will use them. This ``default
39210unsupported'' behavior is not appropriate for all packets, but it
39211helps to keep the initial connection time under control with new
39212versions of @value{GDBN} which support increasing numbers of packets.
39213
39214Reply:
39215@table @samp
39216@item @var{stubfeature} @r{[};@var{stubfeature}@r{]}@dots{}
39217The stub supports or does not support each returned @var{stubfeature},
39218depending on the form of each @var{stubfeature} (see below for the
39219possible forms).
39220@item @w{}
39221An empty reply indicates that @samp{qSupported} is not recognized,
39222or that no features needed to be reported to @value{GDBN}.
39223@end table
39224
39225The allowed forms for each feature (either a @var{gdbfeature} in the
39226@samp{qSupported} packet, or a @var{stubfeature} in the response)
39227are:
39228
39229@table @samp
39230@item @var{name}=@var{value}
39231The remote protocol feature @var{name} is supported, and associated
39232with the specified @var{value}. The format of @var{value} depends
39233on the feature, but it must not include a semicolon.
39234@item @var{name}+
39235The remote protocol feature @var{name} is supported, and does not
39236need an associated value.
39237@item @var{name}-
39238The remote protocol feature @var{name} is not supported.
39239@item @var{name}?
39240The remote protocol feature @var{name} may be supported, and
39241@value{GDBN} should auto-detect support in some other way when it is
39242needed. This form will not be used for @var{gdbfeature} notifications,
39243but may be used for @var{stubfeature} responses.
39244@end table
39245
39246Whenever the stub receives a @samp{qSupported} request, the
39247supplied set of @value{GDBN} features should override any previous
39248request. This allows @value{GDBN} to put the stub in a known
39249state, even if the stub had previously been communicating with
39250a different version of @value{GDBN}.
39251
39252The following values of @var{gdbfeature} (for the packet sent by @value{GDBN})
39253are defined:
39254
39255@table @samp
39256@item multiprocess
39257This feature indicates whether @value{GDBN} supports multiprocess
39258extensions to the remote protocol. @value{GDBN} does not use such
39259extensions unless the stub also reports that it supports them by
39260including @samp{multiprocess+} in its @samp{qSupported} reply.
39261@xref{multiprocess extensions}, for details.
39262
39263@item xmlRegisters
39264This feature indicates that @value{GDBN} supports the XML target
39265description. If the stub sees @samp{xmlRegisters=} with target
39266specific strings separated by a comma, it will report register
39267description.
39268
39269@item qRelocInsn
39270This feature indicates whether @value{GDBN} supports the
39271@samp{qRelocInsn} packet (@pxref{Tracepoint Packets,,Relocate
39272instruction reply packet}).
39273@end table
39274
39275Stubs should ignore any unknown values for
39276@var{gdbfeature}. Any @value{GDBN} which sends a @samp{qSupported}
39277packet supports receiving packets of unlimited length (earlier
39278versions of @value{GDBN} may reject overly long responses). Additional values
39279for @var{gdbfeature} may be defined in the future to let the stub take
39280advantage of new features in @value{GDBN}, e.g.@: incompatible
39281improvements in the remote protocol---the @samp{multiprocess} feature is
39282an example of such a feature. The stub's reply should be independent
39283of the @var{gdbfeature} entries sent by @value{GDBN}; first @value{GDBN}
39284describes all the features it supports, and then the stub replies with
39285all the features it supports.
39286
39287Similarly, @value{GDBN} will silently ignore unrecognized stub feature
39288responses, as long as each response uses one of the standard forms.
39289
39290Some features are flags. A stub which supports a flag feature
39291should respond with a @samp{+} form response. Other features
39292require values, and the stub should respond with an @samp{=}
39293form response.
39294
39295Each feature has a default value, which @value{GDBN} will use if
39296@samp{qSupported} is not available or if the feature is not mentioned
39297in the @samp{qSupported} response. The default values are fixed; a
39298stub is free to omit any feature responses that match the defaults.
39299
39300Not all features can be probed, but for those which can, the probing
39301mechanism is useful: in some cases, a stub's internal
39302architecture may not allow the protocol layer to know some information
39303about the underlying target in advance. This is especially common in
39304stubs which may be configured for multiple targets.
39305
39306These are the currently defined stub features and their properties:
39307
39308@multitable @columnfractions 0.35 0.2 0.12 0.2
39309@c NOTE: The first row should be @headitem, but we do not yet require
39310@c a new enough version of Texinfo (4.7) to use @headitem.
39311@item Feature Name
39312@tab Value Required
39313@tab Default
39314@tab Probe Allowed
39315
39316@item @samp{PacketSize}
39317@tab Yes
39318@tab @samp{-}
39319@tab No
39320
39321@item @samp{qXfer:auxv:read}
39322@tab No
39323@tab @samp{-}
39324@tab Yes
39325
39326@item @samp{qXfer:btrace:read}
39327@tab No
39328@tab @samp{-}
39329@tab Yes
39330
39331@item @samp{qXfer:features:read}
39332@tab No
39333@tab @samp{-}
39334@tab Yes
39335
39336@item @samp{qXfer:libraries:read}
39337@tab No
39338@tab @samp{-}
39339@tab Yes
39340
39341@item @samp{qXfer:libraries-svr4:read}
39342@tab No
39343@tab @samp{-}
39344@tab Yes
39345
39346@item @samp{augmented-libraries-svr4-read}
39347@tab No
39348@tab @samp{-}
39349@tab No
39350
39351@item @samp{qXfer:memory-map:read}
39352@tab No
39353@tab @samp{-}
39354@tab Yes
39355
39356@item @samp{qXfer:sdata:read}
39357@tab No
39358@tab @samp{-}
39359@tab Yes
39360
39361@item @samp{qXfer:spu:read}
39362@tab No
39363@tab @samp{-}
39364@tab Yes
39365
39366@item @samp{qXfer:spu:write}
39367@tab No
39368@tab @samp{-}
39369@tab Yes
39370
39371@item @samp{qXfer:siginfo:read}
39372@tab No
39373@tab @samp{-}
39374@tab Yes
39375
39376@item @samp{qXfer:siginfo:write}
39377@tab No
39378@tab @samp{-}
39379@tab Yes
39380
39381@item @samp{qXfer:threads:read}
39382@tab No
39383@tab @samp{-}
39384@tab Yes
39385
39386@item @samp{qXfer:traceframe-info:read}
39387@tab No
39388@tab @samp{-}
39389@tab Yes
39390
39391@item @samp{qXfer:uib:read}
39392@tab No
39393@tab @samp{-}
39394@tab Yes
39395
39396@item @samp{qXfer:fdpic:read}
39397@tab No
39398@tab @samp{-}
39399@tab Yes
39400
39401@item @samp{Qbtrace:off}
39402@tab Yes
39403@tab @samp{-}
39404@tab Yes
39405
39406@item @samp{Qbtrace:bts}
39407@tab Yes
39408@tab @samp{-}
39409@tab Yes
39410
39411@item @samp{QNonStop}
39412@tab No
39413@tab @samp{-}
39414@tab Yes
39415
39416@item @samp{QPassSignals}
39417@tab No
39418@tab @samp{-}
39419@tab Yes
39420
39421@item @samp{QStartNoAckMode}
39422@tab No
39423@tab @samp{-}
39424@tab Yes
39425
39426@item @samp{multiprocess}
39427@tab No
39428@tab @samp{-}
39429@tab No
39430
39431@item @samp{ConditionalBreakpoints}
39432@tab No
39433@tab @samp{-}
39434@tab No
39435
39436@item @samp{ConditionalTracepoints}
39437@tab No
39438@tab @samp{-}
39439@tab No
39440
39441@item @samp{ReverseContinue}
39442@tab No
39443@tab @samp{-}
39444@tab No
39445
39446@item @samp{ReverseStep}
39447@tab No
39448@tab @samp{-}
39449@tab No
39450
39451@item @samp{TracepointSource}
39452@tab No
39453@tab @samp{-}
39454@tab No
39455
39456@item @samp{QAgent}
39457@tab No
39458@tab @samp{-}
39459@tab No
39460
39461@item @samp{QAllow}
39462@tab No
39463@tab @samp{-}
39464@tab No
39465
39466@item @samp{QDisableRandomization}
39467@tab No
39468@tab @samp{-}
39469@tab No
39470
39471@item @samp{EnableDisableTracepoints}
39472@tab No
39473@tab @samp{-}
39474@tab No
39475
39476@item @samp{QTBuffer:size}
39477@tab No
39478@tab @samp{-}
39479@tab No
39480
39481@item @samp{tracenz}
39482@tab No
39483@tab @samp{-}
39484@tab No
39485
39486@item @samp{BreakpointCommands}
39487@tab No
39488@tab @samp{-}
39489@tab No
39490
39491@end multitable
39492
39493These are the currently defined stub features, in more detail:
39494
39495@table @samp
39496@cindex packet size, remote protocol
39497@item PacketSize=@var{bytes}
39498The remote stub can accept packets up to at least @var{bytes} in
39499length. @value{GDBN} will send packets up to this size for bulk
39500transfers, and will never send larger packets. This is a limit on the
39501data characters in the packet, including the frame and checksum.
39502There is no trailing NUL byte in a remote protocol packet; if the stub
39503stores packets in a NUL-terminated format, it should allow an extra
39504byte in its buffer for the NUL. If this stub feature is not supported,
39505@value{GDBN} guesses based on the size of the @samp{g} packet response.
39506
39507@item qXfer:auxv:read
39508The remote stub understands the @samp{qXfer:auxv:read} packet
39509(@pxref{qXfer auxiliary vector read}).
39510
39511@item qXfer:btrace:read
39512The remote stub understands the @samp{qXfer:btrace:read}
39513packet (@pxref{qXfer btrace read}).
39514
39515@item qXfer:features:read
39516The remote stub understands the @samp{qXfer:features:read} packet
39517(@pxref{qXfer target description read}).
39518
39519@item qXfer:libraries:read
39520The remote stub understands the @samp{qXfer:libraries:read} packet
39521(@pxref{qXfer library list read}).
39522
39523@item qXfer:libraries-svr4:read
39524The remote stub understands the @samp{qXfer:libraries-svr4:read} packet
39525(@pxref{qXfer svr4 library list read}).
39526
39527@item augmented-libraries-svr4-read
39528The remote stub understands the augmented form of the
39529@samp{qXfer:libraries-svr4:read} packet
39530(@pxref{qXfer svr4 library list read}).
39531
39532@item qXfer:memory-map:read
39533The remote stub understands the @samp{qXfer:memory-map:read} packet
39534(@pxref{qXfer memory map read}).
39535
39536@item qXfer:sdata:read
39537The remote stub understands the @samp{qXfer:sdata:read} packet
39538(@pxref{qXfer sdata read}).
39539
39540@item qXfer:spu:read
39541The remote stub understands the @samp{qXfer:spu:read} packet
39542(@pxref{qXfer spu read}).
39543
39544@item qXfer:spu:write
39545The remote stub understands the @samp{qXfer:spu:write} packet
39546(@pxref{qXfer spu write}).
39547
39548@item qXfer:siginfo:read
39549The remote stub understands the @samp{qXfer:siginfo:read} packet
39550(@pxref{qXfer siginfo read}).
39551
39552@item qXfer:siginfo:write
39553The remote stub understands the @samp{qXfer:siginfo:write} packet
39554(@pxref{qXfer siginfo write}).
39555
39556@item qXfer:threads:read
39557The remote stub understands the @samp{qXfer:threads:read} packet
39558(@pxref{qXfer threads read}).
39559
39560@item qXfer:traceframe-info:read
39561The remote stub understands the @samp{qXfer:traceframe-info:read}
39562packet (@pxref{qXfer traceframe info read}).
39563
39564@item qXfer:uib:read
39565The remote stub understands the @samp{qXfer:uib:read}
39566packet (@pxref{qXfer unwind info block}).
39567
39568@item qXfer:fdpic:read
39569The remote stub understands the @samp{qXfer:fdpic:read}
39570packet (@pxref{qXfer fdpic loadmap read}).
39571
39572@item QNonStop
39573The remote stub understands the @samp{QNonStop} packet
39574(@pxref{QNonStop}).
39575
39576@item QPassSignals
39577The remote stub understands the @samp{QPassSignals} packet
39578(@pxref{QPassSignals}).
39579
39580@item QStartNoAckMode
39581The remote stub understands the @samp{QStartNoAckMode} packet and
39582prefers to operate in no-acknowledgment mode. @xref{Packet Acknowledgment}.
39583
39584@item multiprocess
39585@anchor{multiprocess extensions}
39586@cindex multiprocess extensions, in remote protocol
39587The remote stub understands the multiprocess extensions to the remote
39588protocol syntax. The multiprocess extensions affect the syntax of
39589thread IDs in both packets and replies (@pxref{thread-id syntax}), and
39590add process IDs to the @samp{D} packet and @samp{W} and @samp{X}
39591replies. Note that reporting this feature indicates support for the
39592syntactic extensions only, not that the stub necessarily supports
39593debugging of more than one process at a time. The stub must not use
39594multiprocess extensions in packet replies unless @value{GDBN} has also
39595indicated it supports them in its @samp{qSupported} request.
39596
39597@item qXfer:osdata:read
39598The remote stub understands the @samp{qXfer:osdata:read} packet
39599((@pxref{qXfer osdata read}).
39600
39601@item ConditionalBreakpoints
39602The target accepts and implements evaluation of conditional expressions
39603defined for breakpoints. The target will only report breakpoint triggers
39604when such conditions are true (@pxref{Conditions, ,Break Conditions}).
39605
39606@item ConditionalTracepoints
39607The remote stub accepts and implements conditional expressions defined
39608for tracepoints (@pxref{Tracepoint Conditions}).
39609
39610@item ReverseContinue
39611The remote stub accepts and implements the reverse continue packet
39612(@pxref{bc}).
39613
39614@item ReverseStep
39615The remote stub accepts and implements the reverse step packet
39616(@pxref{bs}).
39617
39618@item TracepointSource
39619The remote stub understands the @samp{QTDPsrc} packet that supplies
39620the source form of tracepoint definitions.
39621
39622@item QAgent
39623The remote stub understands the @samp{QAgent} packet.
39624
39625@item QAllow
39626The remote stub understands the @samp{QAllow} packet.
39627
39628@item QDisableRandomization
39629The remote stub understands the @samp{QDisableRandomization} packet.
39630
39631@item StaticTracepoint
39632@cindex static tracepoints, in remote protocol
39633The remote stub supports static tracepoints.
39634
39635@item InstallInTrace
39636@anchor{install tracepoint in tracing}
39637The remote stub supports installing tracepoint in tracing.
39638
39639@item EnableDisableTracepoints
39640The remote stub supports the @samp{QTEnable} (@pxref{QTEnable}) and
39641@samp{QTDisable} (@pxref{QTDisable}) packets that allow tracepoints
39642to be enabled and disabled while a trace experiment is running.
39643
39644@item QTBuffer:size
39645The remote stub supports the @samp{QTBuffer:size} (@pxref{QTBuffer-size})
39646packet that allows to change the size of the trace buffer.
39647
39648@item tracenz
39649@cindex string tracing, in remote protocol
39650The remote stub supports the @samp{tracenz} bytecode for collecting strings.
39651See @ref{Bytecode Descriptions} for details about the bytecode.
39652
39653@item BreakpointCommands
39654@cindex breakpoint commands, in remote protocol
39655The remote stub supports running a breakpoint's command list itself,
39656rather than reporting the hit to @value{GDBN}.
39657
39658@item Qbtrace:off
39659The remote stub understands the @samp{Qbtrace:off} packet.
39660
39661@item Qbtrace:bts
39662The remote stub understands the @samp{Qbtrace:bts} packet.
39663
39664@end table
39665
39666@item qSymbol::
39667@cindex symbol lookup, remote request
39668@cindex @samp{qSymbol} packet
39669Notify the target that @value{GDBN} is prepared to serve symbol lookup
39670requests. Accept requests from the target for the values of symbols.
39671
39672Reply:
39673@table @samp
39674@item OK
39675The target does not need to look up any (more) symbols.
39676@item qSymbol:@var{sym_name}
39677The target requests the value of symbol @var{sym_name} (hex encoded).
39678@value{GDBN} may provide the value by using the
39679@samp{qSymbol:@var{sym_value}:@var{sym_name}} message, described
39680below.
39681@end table
39682
39683@item qSymbol:@var{sym_value}:@var{sym_name}
39684Set the value of @var{sym_name} to @var{sym_value}.
39685
39686@var{sym_name} (hex encoded) is the name of a symbol whose value the
39687target has previously requested.
39688
39689@var{sym_value} (hex) is the value for symbol @var{sym_name}. If
39690@value{GDBN} cannot supply a value for @var{sym_name}, then this field
39691will be empty.
39692
39693Reply:
39694@table @samp
39695@item OK
39696The target does not need to look up any (more) symbols.
39697@item qSymbol:@var{sym_name}
39698The target requests the value of a new symbol @var{sym_name} (hex
39699encoded). @value{GDBN} will continue to supply the values of symbols
39700(if available), until the target ceases to request them.
39701@end table
39702
39703@item qTBuffer
39704@itemx QTBuffer
39705@itemx QTDisconnected
39706@itemx QTDP
39707@itemx QTDPsrc
39708@itemx QTDV
39709@itemx qTfP
39710@itemx qTfV
39711@itemx QTFrame
39712@itemx qTMinFTPILen
39713
39714@xref{Tracepoint Packets}.
39715
39716@item qThreadExtraInfo,@var{thread-id}
39717@cindex thread attributes info, remote request
39718@cindex @samp{qThreadExtraInfo} packet
39719Obtain a printable string description of a thread's attributes from
39720the target OS. @var{thread-id} is a thread ID;
39721see @ref{thread-id syntax}. This
39722string may contain anything that the target OS thinks is interesting
39723for @value{GDBN} to tell the user about the thread. The string is
39724displayed in @value{GDBN}'s @code{info threads} display. Some
39725examples of possible thread extra info strings are @samp{Runnable}, or
39726@samp{Blocked on Mutex}.
39727
39728Reply:
39729@table @samp
39730@item @var{XX}@dots{}
39731Where @samp{@var{XX}@dots{}} is a hex encoding of @sc{ascii} data,
39732comprising the printable string containing the extra information about
39733the thread's attributes.
39734@end table
39735
39736(Note that the @code{qThreadExtraInfo} packet's name is separated from
39737the command by a @samp{,}, not a @samp{:}, contrary to the naming
39738conventions above. Please don't use this packet as a model for new
39739packets.)
39740
39741@item QTNotes
39742@itemx qTP
39743@itemx QTSave
39744@itemx qTsP
39745@itemx qTsV
39746@itemx QTStart
39747@itemx QTStop
39748@itemx QTEnable
39749@itemx QTDisable
39750@itemx QTinit
39751@itemx QTro
39752@itemx qTStatus
39753@itemx qTV
39754@itemx qTfSTM
39755@itemx qTsSTM
39756@itemx qTSTMat
39757@xref{Tracepoint Packets}.
39758
39759@item qXfer:@var{object}:read:@var{annex}:@var{offset},@var{length}
39760@cindex read special object, remote request
39761@cindex @samp{qXfer} packet
39762@anchor{qXfer read}
39763Read uninterpreted bytes from the target's special data area
39764identified by the keyword @var{object}. Request @var{length} bytes
39765starting at @var{offset} bytes into the data. The content and
39766encoding of @var{annex} is specific to @var{object}; it can supply
39767additional details about what data to access.
39768
39769Here are the specific requests of this form defined so far. All
39770@samp{qXfer:@var{object}:read:@dots{}} requests use the same reply
39771formats, listed below.
39772
39773@table @samp
39774@item qXfer:auxv:read::@var{offset},@var{length}
39775@anchor{qXfer auxiliary vector read}
39776Access the target's @dfn{auxiliary vector}. @xref{OS Information,
39777auxiliary vector}. Note @var{annex} must be empty.
39778
39779This packet is not probed by default; the remote stub must request it,
39780by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39781
39782@item qXfer:btrace:read:@var{annex}:@var{offset},@var{length}
39783@anchor{qXfer btrace read}
39784
39785Return a description of the current branch trace.
39786@xref{Branch Trace Format}. The annex part of the generic @samp{qXfer}
39787packet may have one of the following values:
39788
39789@table @code
39790@item all
39791Returns all available branch trace.
39792
39793@item new
39794Returns all available branch trace if the branch trace changed since
39795the last read request.
39796@end table
39797
39798This packet is not probed by default; the remote stub must request it
39799by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39800
39801@item qXfer:features:read:@var{annex}:@var{offset},@var{length}
39802@anchor{qXfer target description read}
39803Access the @dfn{target description}. @xref{Target Descriptions}. The
39804annex specifies which XML document to access. The main description is
39805always loaded from the @samp{target.xml} annex.
39806
39807This packet is not probed by default; the remote stub must request it,
39808by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39809
39810@item qXfer:libraries:read:@var{annex}:@var{offset},@var{length}
39811@anchor{qXfer library list read}
39812Access the target's list of loaded libraries. @xref{Library List Format}.
39813The annex part of the generic @samp{qXfer} packet must be empty
39814(@pxref{qXfer read}).
39815
39816Targets which maintain a list of libraries in the program's memory do
39817not need to implement this packet; it is designed for platforms where
39818the operating system manages the list of loaded libraries.
39819
39820This packet is not probed by default; the remote stub must request it,
39821by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39822
39823@item qXfer:libraries-svr4:read:@var{annex}:@var{offset},@var{length}
39824@anchor{qXfer svr4 library list read}
39825Access the target's list of loaded libraries when the target is an SVR4
39826platform. @xref{Library List Format for SVR4 Targets}. The annex part
39827of the generic @samp{qXfer} packet must be empty unless the remote
39828stub indicated it supports the augmented form of this packet
39829by supplying an appropriate @samp{qSupported} response
39830(@pxref{qXfer read}, @ref{qSupported}).
39831
39832This packet is optional for better performance on SVR4 targets.
39833@value{GDBN} uses memory read packets to read the SVR4 library list otherwise.
39834
39835This packet is not probed by default; the remote stub must request it,
39836by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39837
39838If the remote stub indicates it supports the augmented form of this
39839packet then the annex part of the generic @samp{qXfer} packet may
39840contain a semicolon-separated list of @samp{@var{name}=@var{value}}
39841arguments. The currently supported arguments are:
39842
39843@table @code
39844@item start=@var{address}
39845A hexadecimal number specifying the address of the @samp{struct
39846link_map} to start reading the library list from. If unset or zero
39847then the first @samp{struct link_map} in the library list will be
39848chosen as the starting point.
39849
39850@item prev=@var{address}
39851A hexadecimal number specifying the address of the @samp{struct
39852link_map} immediately preceding the @samp{struct link_map}
39853specified by the @samp{start} argument. If unset or zero then
39854the remote stub will expect that no @samp{struct link_map}
39855exists prior to the starting point.
39856
39857@end table
39858
39859Arguments that are not understood by the remote stub will be silently
39860ignored.
39861
39862@item qXfer:memory-map:read::@var{offset},@var{length}
39863@anchor{qXfer memory map read}
39864Access the target's @dfn{memory-map}. @xref{Memory Map Format}. The
39865annex part of the generic @samp{qXfer} packet must be empty
39866(@pxref{qXfer read}).
39867
39868This packet is not probed by default; the remote stub must request it,
39869by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39870
39871@item qXfer:sdata:read::@var{offset},@var{length}
39872@anchor{qXfer sdata read}
39873
39874Read contents of the extra collected static tracepoint marker
39875information. The annex part of the generic @samp{qXfer} packet must
39876be empty (@pxref{qXfer read}). @xref{Tracepoint Actions,,Tracepoint
39877Action Lists}.
39878
39879This packet is not probed by default; the remote stub must request it,
39880by supplying an appropriate @samp{qSupported} response
39881(@pxref{qSupported}).
39882
39883@item qXfer:siginfo:read::@var{offset},@var{length}
39884@anchor{qXfer siginfo read}
39885Read contents of the extra signal information on the target
39886system. The annex part of the generic @samp{qXfer} packet must be
39887empty (@pxref{qXfer read}).
39888
39889This packet is not probed by default; the remote stub must request it,
39890by supplying an appropriate @samp{qSupported} response
39891(@pxref{qSupported}).
39892
39893@item qXfer:spu:read:@var{annex}:@var{offset},@var{length}
39894@anchor{qXfer spu read}
39895Read contents of an @code{spufs} file on the target system. The
39896annex specifies which file to read; it must be of the form
39897@file{@var{id}/@var{name}}, where @var{id} specifies an SPU context ID
39898in the target process, and @var{name} identifes the @code{spufs} file
39899in that context to be accessed.
39900
39901This packet is not probed by default; the remote stub must request it,
39902by supplying an appropriate @samp{qSupported} response
39903(@pxref{qSupported}).
39904
39905@item qXfer:threads:read::@var{offset},@var{length}
39906@anchor{qXfer threads read}
39907Access the list of threads on target. @xref{Thread List Format}. The
39908annex part of the generic @samp{qXfer} packet must be empty
39909(@pxref{qXfer read}).
39910
39911This packet is not probed by default; the remote stub must request it,
39912by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39913
39914@item qXfer:traceframe-info:read::@var{offset},@var{length}
39915@anchor{qXfer traceframe info read}
39916
39917Return a description of the current traceframe's contents.
39918@xref{Traceframe Info Format}. The annex part of the generic
39919@samp{qXfer} packet must be empty (@pxref{qXfer read}).
39920
39921This packet is not probed by default; the remote stub must request it,
39922by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39923
39924@item qXfer:uib:read:@var{pc}:@var{offset},@var{length}
39925@anchor{qXfer unwind info block}
39926
39927Return the unwind information block for @var{pc}. This packet is used
39928on OpenVMS/ia64 to ask the kernel unwind information.
39929
39930This packet is not probed by default.
39931
39932@item qXfer:fdpic:read:@var{annex}:@var{offset},@var{length}
39933@anchor{qXfer fdpic loadmap read}
39934Read contents of @code{loadmap}s on the target system. The
39935annex, either @samp{exec} or @samp{interp}, specifies which @code{loadmap},
39936executable @code{loadmap} or interpreter @code{loadmap} to read.
39937
39938This packet is not probed by default; the remote stub must request it,
39939by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
39940
39941@item qXfer:osdata:read::@var{offset},@var{length}
39942@anchor{qXfer osdata read}
39943Access the target's @dfn{operating system information}.
39944@xref{Operating System Information}.
39945
39946@end table
39947
39948Reply:
39949@table @samp
39950@item m @var{data}
39951Data @var{data} (@pxref{Binary Data}) has been read from the
39952target. There may be more data at a higher address (although
39953it is permitted to return @samp{m} even for the last valid
39954block of data, as long as at least one byte of data was read).
39955@var{data} may have fewer bytes than the @var{length} in the
39956request.
39957
39958@item l @var{data}
39959Data @var{data} (@pxref{Binary Data}) has been read from the target.
39960There is no more data to be read. @var{data} may have fewer bytes
39961than the @var{length} in the request.
39962
39963@item l
39964The @var{offset} in the request is at the end of the data.
39965There is no more data to be read.
39966
39967@item E00
39968The request was malformed, or @var{annex} was invalid.
39969
39970@item E @var{nn}
39971The offset was invalid, or there was an error encountered reading the data.
39972@var{nn} is a hex-encoded @code{errno} value.
39973
39974@item @w{}
39975An empty reply indicates the @var{object} string was not recognized by
39976the stub, or that the object does not support reading.
39977@end table
39978
39979@item qXfer:@var{object}:write:@var{annex}:@var{offset}:@var{data}@dots{}
39980@cindex write data into object, remote request
39981@anchor{qXfer write}
39982Write uninterpreted bytes into the target's special data area
39983identified by the keyword @var{object}, starting at @var{offset} bytes
39984into the data. @var{data}@dots{} is the binary-encoded data
39985(@pxref{Binary Data}) to be written. The content and encoding of @var{annex}
39986is specific to @var{object}; it can supply additional details about what data
39987to access.
39988
39989Here are the specific requests of this form defined so far. All
39990@samp{qXfer:@var{object}:write:@dots{}} requests use the same reply
39991formats, listed below.
39992
39993@table @samp
39994@item qXfer:siginfo:write::@var{offset}:@var{data}@dots{}
39995@anchor{qXfer siginfo write}
39996Write @var{data} to the extra signal information on the target system.
39997The annex part of the generic @samp{qXfer} packet must be
39998empty (@pxref{qXfer write}).
39999
40000This packet is not probed by default; the remote stub must request it,
40001by supplying an appropriate @samp{qSupported} response
40002(@pxref{qSupported}).
40003
40004@item qXfer:spu:write:@var{annex}:@var{offset}:@var{data}@dots{}
40005@anchor{qXfer spu write}
40006Write @var{data} to an @code{spufs} file on the target system. The
40007annex specifies which file to write; it must be of the form
40008@file{@var{id}/@var{name}}, where @var{id} specifies an SPU context ID
40009in the target process, and @var{name} identifes the @code{spufs} file
40010in that context to be accessed.
40011
40012This packet is not probed by default; the remote stub must request it,
40013by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
40014@end table
40015
40016Reply:
40017@table @samp
40018@item @var{nn}
40019@var{nn} (hex encoded) is the number of bytes written.
40020This may be fewer bytes than supplied in the request.
40021
40022@item E00
40023The request was malformed, or @var{annex} was invalid.
40024
40025@item E @var{nn}
40026The offset was invalid, or there was an error encountered writing the data.
40027@var{nn} is a hex-encoded @code{errno} value.
40028
40029@item @w{}
40030An empty reply indicates the @var{object} string was not
40031recognized by the stub, or that the object does not support writing.
40032@end table
40033
40034@item qXfer:@var{object}:@var{operation}:@dots{}
40035Requests of this form may be added in the future. When a stub does
40036not recognize the @var{object} keyword, or its support for
40037@var{object} does not recognize the @var{operation} keyword, the stub
40038must respond with an empty packet.
40039
40040@item qAttached:@var{pid}
40041@cindex query attached, remote request
40042@cindex @samp{qAttached} packet
40043Return an indication of whether the remote server attached to an
40044existing process or created a new process. When the multiprocess
40045protocol extensions are supported (@pxref{multiprocess extensions}),
40046@var{pid} is an integer in hexadecimal format identifying the target
40047process. Otherwise, @value{GDBN} will omit the @var{pid} field and
40048the query packet will be simplified as @samp{qAttached}.
40049
40050This query is used, for example, to know whether the remote process
40051should be detached or killed when a @value{GDBN} session is ended with
40052the @code{quit} command.
40053
40054Reply:
40055@table @samp
40056@item 1
40057The remote server attached to an existing process.
40058@item 0
40059The remote server created a new process.
40060@item E @var{NN}
40061A badly formed request or an error was encountered.
40062@end table
40063
40064@item Qbtrace:bts
40065Enable branch tracing for the current thread using bts tracing.
40066
40067Reply:
40068@table @samp
40069@item OK
40070Branch tracing has been enabled.
40071@item E.errtext
40072A badly formed request or an error was encountered.
40073@end table
40074
40075@item Qbtrace:off
40076Disable branch tracing for the current thread.
40077
40078Reply:
40079@table @samp
40080@item OK
40081Branch tracing has been disabled.
40082@item E.errtext
40083A badly formed request or an error was encountered.
40084@end table
40085
40086@end table
40087
40088@node Architecture-Specific Protocol Details
40089@section Architecture-Specific Protocol Details
40090
40091This section describes how the remote protocol is applied to specific
40092target architectures. Also see @ref{Standard Target Features}, for
40093details of XML target descriptions for each architecture.
40094
40095@menu
40096* ARM-Specific Protocol Details::
40097* MIPS-Specific Protocol Details::
40098@end menu
40099
40100@node ARM-Specific Protocol Details
40101@subsection @acronym{ARM}-specific Protocol Details
40102
40103@menu
40104* ARM Breakpoint Kinds::
40105@end menu
40106
40107@node ARM Breakpoint Kinds
40108@subsubsection @acronym{ARM} Breakpoint Kinds
40109@cindex breakpoint kinds, @acronym{ARM}
40110
40111These breakpoint kinds are defined for the @samp{Z0} and @samp{Z1} packets.
40112
40113@table @r
40114
40115@item 2
4011616-bit Thumb mode breakpoint.
40117
40118@item 3
4011932-bit Thumb mode (Thumb-2) breakpoint.
40120
40121@item 4
4012232-bit @acronym{ARM} mode breakpoint.
40123
40124@end table
40125
40126@node MIPS-Specific Protocol Details
40127@subsection @acronym{MIPS}-specific Protocol Details
40128
40129@menu
40130* MIPS Register packet Format::
40131* MIPS Breakpoint Kinds::
40132@end menu
40133
40134@node MIPS Register packet Format
40135@subsubsection @acronym{MIPS} Register Packet Format
40136@cindex register packet format, @acronym{MIPS}
40137
40138The following @code{g}/@code{G} packets have previously been defined.
40139In the below, some thirty-two bit registers are transferred as
40140sixty-four bits. Those registers should be zero/sign extended (which?)
40141to fill the space allocated. Register bytes are transferred in target
40142byte order. The two nibbles within a register byte are transferred
40143most-significant -- least-significant.
40144
40145@table @r
40146
40147@item MIPS32
40148All registers are transferred as thirty-two bit quantities in the order:
4014932 general-purpose; sr; lo; hi; bad; cause; pc; 32 floating-point
40150registers; fsr; fir; fp.
40151
40152@item MIPS64
40153All registers are transferred as sixty-four bit quantities (including
40154thirty-two bit registers such as @code{sr}). The ordering is the same
40155as @code{MIPS32}.
40156
40157@end table
40158
40159@node MIPS Breakpoint Kinds
40160@subsubsection @acronym{MIPS} Breakpoint Kinds
40161@cindex breakpoint kinds, @acronym{MIPS}
40162
40163These breakpoint kinds are defined for the @samp{Z0} and @samp{Z1} packets.
40164
40165@table @r
40166
40167@item 2
4016816-bit @acronym{MIPS16} mode breakpoint.
40169
40170@item 3
4017116-bit @acronym{microMIPS} mode breakpoint.
40172
40173@item 4
4017432-bit standard @acronym{MIPS} mode breakpoint.
40175
40176@item 5
4017732-bit @acronym{microMIPS} mode breakpoint.
40178
40179@end table
40180
40181@node Tracepoint Packets
40182@section Tracepoint Packets
40183@cindex tracepoint packets
40184@cindex packets, tracepoint
40185
40186Here we describe the packets @value{GDBN} uses to implement
40187tracepoints (@pxref{Tracepoints}).
40188
40189@table @samp
40190
40191@item QTDP:@var{n}:@var{addr}:@var{ena}:@var{step}:@var{pass}[:F@var{flen}][:X@var{len},@var{bytes}]@r{[}-@r{]}
40192@cindex @samp{QTDP} packet
40193Create a new tracepoint, number @var{n}, at @var{addr}. If @var{ena}
40194is @samp{E}, then the tracepoint is enabled; if it is @samp{D}, then
40195the tracepoint is disabled. @var{step} is the tracepoint's step
40196count, and @var{pass} is its pass count. If an @samp{F} is present,
40197then the tracepoint is to be a fast tracepoint, and the @var{flen} is
40198the number of bytes that the target should copy elsewhere to make room
40199for the tracepoint. If an @samp{X} is present, it introduces a
40200tracepoint condition, which consists of a hexadecimal length, followed
40201by a comma and hex-encoded bytes, in a manner similar to action
40202encodings as described below. If the trailing @samp{-} is present,
40203further @samp{QTDP} packets will follow to specify this tracepoint's
40204actions.
40205
40206Replies:
40207@table @samp
40208@item OK
40209The packet was understood and carried out.
40210@item qRelocInsn
40211@xref{Tracepoint Packets,,Relocate instruction reply packet}.
40212@item @w{}
40213The packet was not recognized.
40214@end table
40215
40216@item QTDP:-@var{n}:@var{addr}:@r{[}S@r{]}@var{action}@dots{}@r{[}-@r{]}
40217Define actions to be taken when a tracepoint is hit. @var{n} and
40218@var{addr} must be the same as in the initial @samp{QTDP} packet for
40219this tracepoint. This packet may only be sent immediately after
40220another @samp{QTDP} packet that ended with a @samp{-}. If the
40221trailing @samp{-} is present, further @samp{QTDP} packets will follow,
40222specifying more actions for this tracepoint.
40223
40224In the series of action packets for a given tracepoint, at most one
40225can have an @samp{S} before its first @var{action}. If such a packet
40226is sent, it and the following packets define ``while-stepping''
40227actions. Any prior packets define ordinary actions --- that is, those
40228taken when the tracepoint is first hit. If no action packet has an
40229@samp{S}, then all the packets in the series specify ordinary
40230tracepoint actions.
40231
40232The @samp{@var{action}@dots{}} portion of the packet is a series of
40233actions, concatenated without separators. Each action has one of the
40234following forms:
40235
40236@table @samp
40237
40238@item R @var{mask}
40239Collect the registers whose bits are set in @var{mask}. @var{mask} is
40240a hexadecimal number whose @var{i}'th bit is set if register number
40241@var{i} should be collected. (The least significant bit is numbered
40242zero.) Note that @var{mask} may be any number of digits long; it may
40243not fit in a 32-bit word.
40244
40245@item M @var{basereg},@var{offset},@var{len}
40246Collect @var{len} bytes of memory starting at the address in register
40247number @var{basereg}, plus @var{offset}. If @var{basereg} is
40248@samp{-1}, then the range has a fixed address: @var{offset} is the
40249address of the lowest byte to collect. The @var{basereg},
40250@var{offset}, and @var{len} parameters are all unsigned hexadecimal
40251values (the @samp{-1} value for @var{basereg} is a special case).
40252
40253@item X @var{len},@var{expr}
40254Evaluate @var{expr}, whose length is @var{len}, and collect memory as
40255it directs. @var{expr} is an agent expression, as described in
40256@ref{Agent Expressions}. Each byte of the expression is encoded as a
40257two-digit hex number in the packet; @var{len} is the number of bytes
40258in the expression (and thus one-half the number of hex digits in the
40259packet).
40260
40261@end table
40262
40263Any number of actions may be packed together in a single @samp{QTDP}
40264packet, as long as the packet does not exceed the maximum packet
40265length (400 bytes, for many stubs). There may be only one @samp{R}
40266action per tracepoint, and it must precede any @samp{M} or @samp{X}
40267actions. Any registers referred to by @samp{M} and @samp{X} actions
40268must be collected by a preceding @samp{R} action. (The
40269``while-stepping'' actions are treated as if they were attached to a
40270separate tracepoint, as far as these restrictions are concerned.)
40271
40272Replies:
40273@table @samp
40274@item OK
40275The packet was understood and carried out.
40276@item qRelocInsn
40277@xref{Tracepoint Packets,,Relocate instruction reply packet}.
40278@item @w{}
40279The packet was not recognized.
40280@end table
40281
40282@item QTDPsrc:@var{n}:@var{addr}:@var{type}:@var{start}:@var{slen}:@var{bytes}
40283@cindex @samp{QTDPsrc} packet
40284Specify a source string of tracepoint @var{n} at address @var{addr}.
40285This is useful to get accurate reproduction of the tracepoints
40286originally downloaded at the beginning of the trace run. @var{type}
40287is the name of the tracepoint part, such as @samp{cond} for the
40288tracepoint's conditional expression (see below for a list of types), while
40289@var{bytes} is the string, encoded in hexadecimal.
40290
40291@var{start} is the offset of the @var{bytes} within the overall source
40292string, while @var{slen} is the total length of the source string.
40293This is intended for handling source strings that are longer than will
40294fit in a single packet.
40295@c Add detailed example when this info is moved into a dedicated
40296@c tracepoint descriptions section.
40297
40298The available string types are @samp{at} for the location,
40299@samp{cond} for the conditional, and @samp{cmd} for an action command.
40300@value{GDBN} sends a separate packet for each command in the action
40301list, in the same order in which the commands are stored in the list.
40302
40303The target does not need to do anything with source strings except
40304report them back as part of the replies to the @samp{qTfP}/@samp{qTsP}
40305query packets.
40306
40307Although this packet is optional, and @value{GDBN} will only send it
40308if the target replies with @samp{TracepointSource} @xref{General
40309Query Packets}, it makes both disconnected tracing and trace files
40310much easier to use. Otherwise the user must be careful that the
40311tracepoints in effect while looking at trace frames are identical to
40312the ones in effect during the trace run; even a small discrepancy
40313could cause @samp{tdump} not to work, or a particular trace frame not
40314be found.
40315
40316@item QTDV:@var{n}:@var{value}
40317@cindex define trace state variable, remote request
40318@cindex @samp{QTDV} packet
40319Create a new trace state variable, number @var{n}, with an initial
40320value of @var{value}, which is a 64-bit signed integer. Both @var{n}
40321and @var{value} are encoded as hexadecimal values. @value{GDBN} has
40322the option of not using this packet for initial values of zero; the
40323target should simply create the trace state variables as they are
40324mentioned in expressions.
40325
40326@item QTFrame:@var{n}
40327@cindex @samp{QTFrame} packet
40328Select the @var{n}'th tracepoint frame from the buffer, and use the
40329register and memory contents recorded there to answer subsequent
40330request packets from @value{GDBN}.
40331
40332A successful reply from the stub indicates that the stub has found the
40333requested frame. The response is a series of parts, concatenated
40334without separators, describing the frame we selected. Each part has
40335one of the following forms:
40336
40337@table @samp
40338@item F @var{f}
40339The selected frame is number @var{n} in the trace frame buffer;
40340@var{f} is a hexadecimal number. If @var{f} is @samp{-1}, then there
40341was no frame matching the criteria in the request packet.
40342
40343@item T @var{t}
40344The selected trace frame records a hit of tracepoint number @var{t};
40345@var{t} is a hexadecimal number.
40346
40347@end table
40348
40349@item QTFrame:pc:@var{addr}
40350Like @samp{QTFrame:@var{n}}, but select the first tracepoint frame after the
40351currently selected frame whose PC is @var{addr};
40352@var{addr} is a hexadecimal number.
40353
40354@item QTFrame:tdp:@var{t}
40355Like @samp{QTFrame:@var{n}}, but select the first tracepoint frame after the
40356currently selected frame that is a hit of tracepoint @var{t}; @var{t}
40357is a hexadecimal number.
40358
40359@item QTFrame:range:@var{start}:@var{end}
40360Like @samp{QTFrame:@var{n}}, but select the first tracepoint frame after the
40361currently selected frame whose PC is between @var{start} (inclusive)
40362and @var{end} (inclusive); @var{start} and @var{end} are hexadecimal
40363numbers.
40364
40365@item QTFrame:outside:@var{start}:@var{end}
40366Like @samp{QTFrame:range:@var{start}:@var{end}}, but select the first
40367frame @emph{outside} the given range of addresses (exclusive).
40368
40369@item qTMinFTPILen
40370@cindex @samp{qTMinFTPILen} packet
40371This packet requests the minimum length of instruction at which a fast
40372tracepoint (@pxref{Set Tracepoints}) may be placed. For instance, on
40373the 32-bit x86 architecture, it is possible to use a 4-byte jump, but
40374it depends on the target system being able to create trampolines in
40375the first 64K of memory, which might or might not be possible for that
40376system. So the reply to this packet will be 4 if it is able to
40377arrange for that.
40378
40379Replies:
40380
40381@table @samp
40382@item 0
40383The minimum instruction length is currently unknown.
40384@item @var{length}
40385The minimum instruction length is @var{length}, where @var{length} is greater
40386or equal to 1. @var{length} is a hexadecimal number. A reply of 1 means
40387that a fast tracepoint may be placed on any instruction regardless of size.
40388@item E
40389An error has occurred.
40390@item @w{}
40391An empty reply indicates that the request is not supported by the stub.
40392@end table
40393
40394@item QTStart
40395@cindex @samp{QTStart} packet
40396Begin the tracepoint experiment. Begin collecting data from
40397tracepoint hits in the trace frame buffer. This packet supports the
40398@samp{qRelocInsn} reply (@pxref{Tracepoint Packets,,Relocate
40399instruction reply packet}).
40400
40401@item QTStop
40402@cindex @samp{QTStop} packet
40403End the tracepoint experiment. Stop collecting trace frames.
40404
40405@item QTEnable:@var{n}:@var{addr}
40406@anchor{QTEnable}
40407@cindex @samp{QTEnable} packet
40408Enable tracepoint @var{n} at address @var{addr} in a started tracepoint
40409experiment. If the tracepoint was previously disabled, then collection
40410of data from it will resume.
40411
40412@item QTDisable:@var{n}:@var{addr}
40413@anchor{QTDisable}
40414@cindex @samp{QTDisable} packet
40415Disable tracepoint @var{n} at address @var{addr} in a started tracepoint
40416experiment. No more data will be collected from the tracepoint unless
40417@samp{QTEnable:@var{n}:@var{addr}} is subsequently issued.
40418
40419@item QTinit
40420@cindex @samp{QTinit} packet
40421Clear the table of tracepoints, and empty the trace frame buffer.
40422
40423@item QTro:@var{start1},@var{end1}:@var{start2},@var{end2}:@dots{}
40424@cindex @samp{QTro} packet
40425Establish the given ranges of memory as ``transparent''. The stub
40426will answer requests for these ranges from memory's current contents,
40427if they were not collected as part of the tracepoint hit.
40428
40429@value{GDBN} uses this to mark read-only regions of memory, like those
40430containing program code. Since these areas never change, they should
40431still have the same contents they did when the tracepoint was hit, so
40432there's no reason for the stub to refuse to provide their contents.
40433
40434@item QTDisconnected:@var{value}
40435@cindex @samp{QTDisconnected} packet
40436Set the choice to what to do with the tracing run when @value{GDBN}
40437disconnects from the target. A @var{value} of 1 directs the target to
40438continue the tracing run, while 0 tells the target to stop tracing if
40439@value{GDBN} is no longer in the picture.
40440
40441@item qTStatus
40442@cindex @samp{qTStatus} packet
40443Ask the stub if there is a trace experiment running right now.
40444
40445The reply has the form:
40446
40447@table @samp
40448
40449@item T@var{running}@r{[};@var{field}@r{]}@dots{}
40450@var{running} is a single digit @code{1} if the trace is presently
40451running, or @code{0} if not. It is followed by semicolon-separated
40452optional fields that an agent may use to report additional status.
40453
40454@end table
40455
40456If the trace is not running, the agent may report any of several
40457explanations as one of the optional fields:
40458
40459@table @samp
40460
40461@item tnotrun:0
40462No trace has been run yet.
40463
40464@item tstop[:@var{text}]:0
40465The trace was stopped by a user-originated stop command. The optional
40466@var{text} field is a user-supplied string supplied as part of the
40467stop command (for instance, an explanation of why the trace was
40468stopped manually). It is hex-encoded.
40469
40470@item tfull:0
40471The trace stopped because the trace buffer filled up.
40472
40473@item tdisconnected:0
40474The trace stopped because @value{GDBN} disconnected from the target.
40475
40476@item tpasscount:@var{tpnum}
40477The trace stopped because tracepoint @var{tpnum} exceeded its pass count.
40478
40479@item terror:@var{text}:@var{tpnum}
40480The trace stopped because tracepoint @var{tpnum} had an error. The
40481string @var{text} is available to describe the nature of the error
40482(for instance, a divide by zero in the condition expression).
40483@var{text} is hex encoded.
40484
40485@item tunknown:0
40486The trace stopped for some other reason.
40487
40488@end table
40489
40490Additional optional fields supply statistical and other information.
40491Although not required, they are extremely useful for users monitoring
40492the progress of a trace run. If a trace has stopped, and these
40493numbers are reported, they must reflect the state of the just-stopped
40494trace.
40495
40496@table @samp
40497
40498@item tframes:@var{n}
40499The number of trace frames in the buffer.
40500
40501@item tcreated:@var{n}
40502The total number of trace frames created during the run. This may
40503be larger than the trace frame count, if the buffer is circular.
40504
40505@item tsize:@var{n}
40506The total size of the trace buffer, in bytes.
40507
40508@item tfree:@var{n}
40509The number of bytes still unused in the buffer.
40510
40511@item circular:@var{n}
40512The value of the circular trace buffer flag. @code{1} means that the
40513trace buffer is circular and old trace frames will be discarded if
40514necessary to make room, @code{0} means that the trace buffer is linear
40515and may fill up.
40516
40517@item disconn:@var{n}
40518The value of the disconnected tracing flag. @code{1} means that
40519tracing will continue after @value{GDBN} disconnects, @code{0} means
40520that the trace run will stop.
40521
40522@end table
40523
40524@item qTP:@var{tp}:@var{addr}
40525@cindex tracepoint status, remote request
40526@cindex @samp{qTP} packet
40527Ask the stub for the current state of tracepoint number @var{tp} at
40528address @var{addr}.
40529
40530Replies:
40531@table @samp
40532@item V@var{hits}:@var{usage}
40533The tracepoint has been hit @var{hits} times so far during the trace
40534run, and accounts for @var{usage} in the trace buffer. Note that
40535@code{while-stepping} steps are not counted as separate hits, but the
40536steps' space consumption is added into the usage number.
40537
40538@end table
40539
40540@item qTV:@var{var}
40541@cindex trace state variable value, remote request
40542@cindex @samp{qTV} packet
40543Ask the stub for the value of the trace state variable number @var{var}.
40544
40545Replies:
40546@table @samp
40547@item V@var{value}
40548The value of the variable is @var{value}. This will be the current
40549value of the variable if the user is examining a running target, or a
40550saved value if the variable was collected in the trace frame that the
40551user is looking at. Note that multiple requests may result in
40552different reply values, such as when requesting values while the
40553program is running.
40554
40555@item U
40556The value of the variable is unknown. This would occur, for example,
40557if the user is examining a trace frame in which the requested variable
40558was not collected.
40559@end table
40560
40561@item qTfP
40562@cindex @samp{qTfP} packet
40563@itemx qTsP
40564@cindex @samp{qTsP} packet
40565These packets request data about tracepoints that are being used by
40566the target. @value{GDBN} sends @code{qTfP} to get the first piece
40567of data, and multiple @code{qTsP} to get additional pieces. Replies
40568to these packets generally take the form of the @code{QTDP} packets
40569that define tracepoints. (FIXME add detailed syntax)
40570
40571@item qTfV
40572@cindex @samp{qTfV} packet
40573@itemx qTsV
40574@cindex @samp{qTsV} packet
40575These packets request data about trace state variables that are on the
40576target. @value{GDBN} sends @code{qTfV} to get the first vari of data,
40577and multiple @code{qTsV} to get additional variables. Replies to
40578these packets follow the syntax of the @code{QTDV} packets that define
40579trace state variables.
40580
40581@item qTfSTM
40582@itemx qTsSTM
40583@anchor{qTfSTM}
40584@anchor{qTsSTM}
40585@cindex @samp{qTfSTM} packet
40586@cindex @samp{qTsSTM} packet
40587These packets request data about static tracepoint markers that exist
40588in the target program. @value{GDBN} sends @code{qTfSTM} to get the
40589first piece of data, and multiple @code{qTsSTM} to get additional
40590pieces. Replies to these packets take the following form:
40591
40592Reply:
40593@table @samp
40594@item m @var{address}:@var{id}:@var{extra}
40595A single marker
40596@item m @var{address}:@var{id}:@var{extra},@var{address}:@var{id}:@var{extra}@dots{}
40597a comma-separated list of markers
40598@item l
40599(lower case letter @samp{L}) denotes end of list.
40600@item E @var{nn}
40601An error occurred. @var{nn} are hex digits.
40602@item @w{}
40603An empty reply indicates that the request is not supported by the
40604stub.
40605@end table
40606
40607@var{address} is encoded in hex.
40608@var{id} and @var{extra} are strings encoded in hex.
40609
40610In response to each query, the target will reply with a list of one or
40611more markers, separated by commas. @value{GDBN} will respond to each
40612reply with a request for more markers (using the @samp{qs} form of the
40613query), until the target responds with @samp{l} (lower-case ell, for
40614@dfn{last}).
40615
40616@item qTSTMat:@var{address}
40617@anchor{qTSTMat}
40618@cindex @samp{qTSTMat} packet
40619This packets requests data about static tracepoint markers in the
40620target program at @var{address}. Replies to this packet follow the
40621syntax of the @samp{qTfSTM} and @code{qTsSTM} packets that list static
40622tracepoint markers.
40623
40624@item QTSave:@var{filename}
40625@cindex @samp{QTSave} packet
40626This packet directs the target to save trace data to the file name
40627@var{filename} in the target's filesystem. @var{filename} is encoded
40628as a hex string; the interpretation of the file name (relative vs
40629absolute, wild cards, etc) is up to the target.
40630
40631@item qTBuffer:@var{offset},@var{len}
40632@cindex @samp{qTBuffer} packet
40633Return up to @var{len} bytes of the current contents of trace buffer,
40634starting at @var{offset}. The trace buffer is treated as if it were
40635a contiguous collection of traceframes, as per the trace file format.
40636The reply consists as many hex-encoded bytes as the target can deliver
40637in a packet; it is not an error to return fewer than were asked for.
40638A reply consisting of just @code{l} indicates that no bytes are
40639available.
40640
40641@item QTBuffer:circular:@var{value}
40642This packet directs the target to use a circular trace buffer if
40643@var{value} is 1, or a linear buffer if the value is 0.
40644
40645@item QTBuffer:size:@var{size}
40646@anchor{QTBuffer-size}
40647@cindex @samp{QTBuffer size} packet
40648This packet directs the target to make the trace buffer be of size
40649@var{size} if possible. A value of @code{-1} tells the target to
40650use whatever size it prefers.
40651
40652@item QTNotes:@r{[}@var{type}:@var{text}@r{]}@r{[};@var{type}:@var{text}@r{]}@dots{}
40653@cindex @samp{QTNotes} packet
40654This packet adds optional textual notes to the trace run. Allowable
40655types include @code{user}, @code{notes}, and @code{tstop}, the
40656@var{text} fields are arbitrary strings, hex-encoded.
40657
40658@end table
40659
40660@subsection Relocate instruction reply packet
40661When installing fast tracepoints in memory, the target may need to
40662relocate the instruction currently at the tracepoint address to a
40663different address in memory. For most instructions, a simple copy is
40664enough, but, for example, call instructions that implicitly push the
40665return address on the stack, and relative branches or other
40666PC-relative instructions require offset adjustment, so that the effect
40667of executing the instruction at a different address is the same as if
40668it had executed in the original location.
40669
40670In response to several of the tracepoint packets, the target may also
40671respond with a number of intermediate @samp{qRelocInsn} request
40672packets before the final result packet, to have @value{GDBN} handle
40673this relocation operation. If a packet supports this mechanism, its
40674documentation will explicitly say so. See for example the above
40675descriptions for the @samp{QTStart} and @samp{QTDP} packets. The
40676format of the request is:
40677
40678@table @samp
40679@item qRelocInsn:@var{from};@var{to}
40680
40681This requests @value{GDBN} to copy instruction at address @var{from}
40682to address @var{to}, possibly adjusted so that executing the
40683instruction at @var{to} has the same effect as executing it at
40684@var{from}. @value{GDBN} writes the adjusted instruction to target
40685memory starting at @var{to}.
40686@end table
40687
40688Replies:
40689@table @samp
40690@item qRelocInsn:@var{adjusted_size}
40691Informs the stub the relocation is complete. @var{adjusted_size} is
40692the length in bytes of resulting relocated instruction sequence.
40693@item E @var{NN}
40694A badly formed request was detected, or an error was encountered while
40695relocating the instruction.
40696@end table
40697
40698@node Host I/O Packets
40699@section Host I/O Packets
40700@cindex Host I/O, remote protocol
40701@cindex file transfer, remote protocol
40702
40703The @dfn{Host I/O} packets allow @value{GDBN} to perform I/O
40704operations on the far side of a remote link. For example, Host I/O is
40705used to upload and download files to a remote target with its own
40706filesystem. Host I/O uses the same constant values and data structure
40707layout as the target-initiated File-I/O protocol. However, the
40708Host I/O packets are structured differently. The target-initiated
40709protocol relies on target memory to store parameters and buffers.
40710Host I/O requests are initiated by @value{GDBN}, and the
40711target's memory is not involved. @xref{File-I/O Remote Protocol
40712Extension}, for more details on the target-initiated protocol.
40713
40714The Host I/O request packets all encode a single operation along with
40715its arguments. They have this format:
40716
40717@table @samp
40718
40719@item vFile:@var{operation}: @var{parameter}@dots{}
40720@var{operation} is the name of the particular request; the target
40721should compare the entire packet name up to the second colon when checking
40722for a supported operation. The format of @var{parameter} depends on
40723the operation. Numbers are always passed in hexadecimal. Negative
40724numbers have an explicit minus sign (i.e.@: two's complement is not
40725used). Strings (e.g.@: filenames) are encoded as a series of
40726hexadecimal bytes. The last argument to a system call may be a
40727buffer of escaped binary data (@pxref{Binary Data}).
40728
40729@end table
40730
40731The valid responses to Host I/O packets are:
40732
40733@table @samp
40734
40735@item F @var{result} [, @var{errno}] [; @var{attachment}]
40736@var{result} is the integer value returned by this operation, usually
40737non-negative for success and -1 for errors. If an error has occured,
40738@var{errno} will be included in the result. @var{errno} will have a
40739value defined by the File-I/O protocol (@pxref{Errno Values}). For
40740operations which return data, @var{attachment} supplies the data as a
40741binary buffer. Binary buffers in response packets are escaped in the
40742normal way (@pxref{Binary Data}). See the individual packet
40743documentation for the interpretation of @var{result} and
40744@var{attachment}.
40745
40746@item @w{}
40747An empty response indicates that this operation is not recognized.
40748
40749@end table
40750
40751These are the supported Host I/O operations:
40752
40753@table @samp
40754@item vFile:open: @var{pathname}, @var{flags}, @var{mode}
40755Open a file at @var{pathname} and return a file descriptor for it, or
40756return -1 if an error occurs. @var{pathname} is a string,
40757@var{flags} is an integer indicating a mask of open flags
40758(@pxref{Open Flags}), and @var{mode} is an integer indicating a mask
40759of mode bits to use if the file is created (@pxref{mode_t Values}).
40760@xref{open}, for details of the open flags and mode values.
40761
40762@item vFile:close: @var{fd}
40763Close the open file corresponding to @var{fd} and return 0, or
40764-1 if an error occurs.
40765
40766@item vFile:pread: @var{fd}, @var{count}, @var{offset}
40767Read data from the open file corresponding to @var{fd}. Up to
40768@var{count} bytes will be read from the file, starting at @var{offset}
40769relative to the start of the file. The target may read fewer bytes;
40770common reasons include packet size limits and an end-of-file
40771condition. The number of bytes read is returned. Zero should only be
40772returned for a successful read at the end of the file, or if
40773@var{count} was zero.
40774
40775The data read should be returned as a binary attachment on success.
40776If zero bytes were read, the response should include an empty binary
40777attachment (i.e.@: a trailing semicolon). The return value is the
40778number of target bytes read; the binary attachment may be longer if
40779some characters were escaped.
40780
40781@item vFile:pwrite: @var{fd}, @var{offset}, @var{data}
40782Write @var{data} (a binary buffer) to the open file corresponding
40783to @var{fd}. Start the write at @var{offset} from the start of the
40784file. Unlike many @code{write} system calls, there is no
40785separate @var{count} argument; the length of @var{data} in the
40786packet is used. @samp{vFile:write} returns the number of bytes written,
40787which may be shorter than the length of @var{data}, or -1 if an
40788error occurred.
40789
40790@item vFile:unlink: @var{pathname}
40791Delete the file at @var{pathname} on the target. Return 0,
40792or -1 if an error occurs. @var{pathname} is a string.
40793
40794@item vFile:readlink: @var{filename}
40795Read value of symbolic link @var{filename} on the target. Return
40796the number of bytes read, or -1 if an error occurs.
40797
40798The data read should be returned as a binary attachment on success.
40799If zero bytes were read, the response should include an empty binary
40800attachment (i.e.@: a trailing semicolon). The return value is the
40801number of target bytes read; the binary attachment may be longer if
40802some characters were escaped.
40803
40804@end table
40805
40806@node Interrupts
40807@section Interrupts
40808@cindex interrupts (remote protocol)
40809
40810When a program on the remote target is running, @value{GDBN} may
40811attempt to interrupt it by sending a @samp{Ctrl-C}, @code{BREAK} or
40812a @code{BREAK} followed by @code{g},
40813control of which is specified via @value{GDBN}'s @samp{interrupt-sequence}.
40814
40815The precise meaning of @code{BREAK} is defined by the transport
40816mechanism and may, in fact, be undefined. @value{GDBN} does not
40817currently define a @code{BREAK} mechanism for any of the network
40818interfaces except for TCP, in which case @value{GDBN} sends the
40819@code{telnet} BREAK sequence.
40820
40821@samp{Ctrl-C}, on the other hand, is defined and implemented for all
40822transport mechanisms. It is represented by sending the single byte
40823@code{0x03} without any of the usual packet overhead described in
40824the Overview section (@pxref{Overview}). When a @code{0x03} byte is
40825transmitted as part of a packet, it is considered to be packet data
40826and does @emph{not} represent an interrupt. E.g., an @samp{X} packet
40827(@pxref{X packet}), used for binary downloads, may include an unescaped
40828@code{0x03} as part of its packet.
40829
40830@code{BREAK} followed by @code{g} is also known as Magic SysRq g.
40831When Linux kernel receives this sequence from serial port,
40832it stops execution and connects to gdb.
40833
40834Stubs are not required to recognize these interrupt mechanisms and the
40835precise meaning associated with receipt of the interrupt is
40836implementation defined. If the target supports debugging of multiple
40837threads and/or processes, it should attempt to interrupt all
40838currently-executing threads and processes.
40839If the stub is successful at interrupting the
40840running program, it should send one of the stop
40841reply packets (@pxref{Stop Reply Packets}) to @value{GDBN} as a result
40842of successfully stopping the program in all-stop mode, and a stop reply
40843for each stopped thread in non-stop mode.
40844Interrupts received while the
40845program is stopped are discarded.
40846
40847@node Notification Packets
40848@section Notification Packets
40849@cindex notification packets
40850@cindex packets, notification
40851
40852The @value{GDBN} remote serial protocol includes @dfn{notifications},
40853packets that require no acknowledgment. Both the GDB and the stub
40854may send notifications (although the only notifications defined at
40855present are sent by the stub). Notifications carry information
40856without incurring the round-trip latency of an acknowledgment, and so
40857are useful for low-impact communications where occasional packet loss
40858is not a problem.
40859
40860A notification packet has the form @samp{% @var{data} #
40861@var{checksum}}, where @var{data} is the content of the notification,
40862and @var{checksum} is a checksum of @var{data}, computed and formatted
40863as for ordinary @value{GDBN} packets. A notification's @var{data}
40864never contains @samp{$}, @samp{%} or @samp{#} characters. Upon
40865receiving a notification, the recipient sends no @samp{+} or @samp{-}
40866to acknowledge the notification's receipt or to report its corruption.
40867
40868Every notification's @var{data} begins with a name, which contains no
40869colon characters, followed by a colon character.
40870
40871Recipients should silently ignore corrupted notifications and
40872notifications they do not understand. Recipients should restart
40873timeout periods on receipt of a well-formed notification, whether or
40874not they understand it.
40875
40876Senders should only send the notifications described here when this
40877protocol description specifies that they are permitted. In the
40878future, we may extend the protocol to permit existing notifications in
40879new contexts; this rule helps older senders avoid confusing newer
40880recipients.
40881
40882(Older versions of @value{GDBN} ignore bytes received until they see
40883the @samp{$} byte that begins an ordinary packet, so new stubs may
40884transmit notifications without fear of confusing older clients. There
40885are no notifications defined for @value{GDBN} to send at the moment, but we
40886assume that most older stubs would ignore them, as well.)
40887
40888Each notification is comprised of three parts:
40889@table @samp
40890@item @var{name}:@var{event}
40891The notification packet is sent by the side that initiates the
40892exchange (currently, only the stub does that), with @var{event}
40893carrying the specific information about the notification.
40894@var{name} is the name of the notification.
40895@item @var{ack}
40896The acknowledge sent by the other side, usually @value{GDBN}, to
40897acknowledge the exchange and request the event.
40898@end table
40899
40900The purpose of an asynchronous notification mechanism is to report to
40901@value{GDBN} that something interesting happened in the remote stub.
40902
40903The remote stub may send notification @var{name}:@var{event}
40904at any time, but @value{GDBN} acknowledges the notification when
40905appropriate. The notification event is pending before @value{GDBN}
40906acknowledges. Only one notification at a time may be pending; if
40907additional events occur before @value{GDBN} has acknowledged the
40908previous notification, they must be queued by the stub for later
40909synchronous transmission in response to @var{ack} packets from
40910@value{GDBN}. Because the notification mechanism is unreliable,
40911the stub is permitted to resend a notification if it believes
40912@value{GDBN} may not have received it.
40913
40914Specifically, notifications may appear when @value{GDBN} is not
40915otherwise reading input from the stub, or when @value{GDBN} is
40916expecting to read a normal synchronous response or a
40917@samp{+}/@samp{-} acknowledgment to a packet it has sent.
40918Notification packets are distinct from any other communication from
40919the stub so there is no ambiguity.
40920
40921After receiving a notification, @value{GDBN} shall acknowledge it by
40922sending a @var{ack} packet as a regular, synchronous request to the
40923stub. Such acknowledgment is not required to happen immediately, as
40924@value{GDBN} is permitted to send other, unrelated packets to the
40925stub first, which the stub should process normally.
40926
40927Upon receiving a @var{ack} packet, if the stub has other queued
40928events to report to @value{GDBN}, it shall respond by sending a
40929normal @var{event}. @value{GDBN} shall then send another @var{ack}
40930packet to solicit further responses; again, it is permitted to send
40931other, unrelated packets as well which the stub should process
40932normally.
40933
40934If the stub receives a @var{ack} packet and there are no additional
40935@var{event} to report, the stub shall return an @samp{OK} response.
40936At this point, @value{GDBN} has finished processing a notification
40937and the stub has completed sending any queued events. @value{GDBN}
40938won't accept any new notifications until the final @samp{OK} is
40939received . If further notification events occur, the stub shall send
40940a new notification, @value{GDBN} shall accept the notification, and
40941the process shall be repeated.
40942
40943The process of asynchronous notification can be illustrated by the
40944following example:
40945@smallexample
40946<- @code{%%Stop:T0505:98e7ffbf;04:4ce6ffbf;08:b1b6e54c;thread:p7526.7526;core:0;}
40947@code{...}
40948-> @code{vStopped}
40949<- @code{T0505:68f37db7;04:40f37db7;08:63850408;thread:p7526.7528;core:0;}
40950-> @code{vStopped}
40951<- @code{T0505:68e3fdb6;04:40e3fdb6;08:63850408;thread:p7526.7529;core:0;}
40952-> @code{vStopped}
40953<- @code{OK}
40954@end smallexample
40955
40956The following notifications are defined:
40957@multitable @columnfractions 0.12 0.12 0.38 0.38
40958
40959@item Notification
40960@tab Ack
40961@tab Event
40962@tab Description
40963
40964@item Stop
40965@tab vStopped
40966@tab @var{reply}. The @var{reply} has the form of a stop reply, as
40967described in @ref{Stop Reply Packets}. Refer to @ref{Remote Non-Stop},
40968for information on how these notifications are acknowledged by
40969@value{GDBN}.
40970@tab Report an asynchronous stop event in non-stop mode.
40971
40972@end multitable
40973
40974@node Remote Non-Stop
40975@section Remote Protocol Support for Non-Stop Mode
40976
40977@value{GDBN}'s remote protocol supports non-stop debugging of
40978multi-threaded programs, as described in @ref{Non-Stop Mode}. If the stub
40979supports non-stop mode, it should report that to @value{GDBN} by including
40980@samp{QNonStop+} in its @samp{qSupported} response (@pxref{qSupported}).
40981
40982@value{GDBN} typically sends a @samp{QNonStop} packet only when
40983establishing a new connection with the stub. Entering non-stop mode
40984does not alter the state of any currently-running threads, but targets
40985must stop all threads in any already-attached processes when entering
40986all-stop mode. @value{GDBN} uses the @samp{?} packet as necessary to
40987probe the target state after a mode change.
40988
40989In non-stop mode, when an attached process encounters an event that
40990would otherwise be reported with a stop reply, it uses the
40991asynchronous notification mechanism (@pxref{Notification Packets}) to
40992inform @value{GDBN}. In contrast to all-stop mode, where all threads
40993in all processes are stopped when a stop reply is sent, in non-stop
40994mode only the thread reporting the stop event is stopped. That is,
40995when reporting a @samp{S} or @samp{T} response to indicate completion
40996of a step operation, hitting a breakpoint, or a fault, only the
40997affected thread is stopped; any other still-running threads continue
40998to run. When reporting a @samp{W} or @samp{X} response, all running
40999threads belonging to other attached processes continue to run.
41000
41001In non-stop mode, the target shall respond to the @samp{?} packet as
41002follows. First, any incomplete stop reply notification/@samp{vStopped}
41003sequence in progress is abandoned. The target must begin a new
41004sequence reporting stop events for all stopped threads, whether or not
41005it has previously reported those events to @value{GDBN}. The first
41006stop reply is sent as a synchronous reply to the @samp{?} packet, and
41007subsequent stop replies are sent as responses to @samp{vStopped} packets
41008using the mechanism described above. The target must not send
41009asynchronous stop reply notifications until the sequence is complete.
41010If all threads are running when the target receives the @samp{?} packet,
41011or if the target is not attached to any process, it shall respond
41012@samp{OK}.
41013
41014@node Packet Acknowledgment
41015@section Packet Acknowledgment
41016
41017@cindex acknowledgment, for @value{GDBN} remote
41018@cindex packet acknowledgment, for @value{GDBN} remote
41019By default, when either the host or the target machine receives a packet,
41020the first response expected is an acknowledgment: either @samp{+} (to indicate
41021the package was received correctly) or @samp{-} (to request retransmission).
41022This mechanism allows the @value{GDBN} remote protocol to operate over
41023unreliable transport mechanisms, such as a serial line.
41024
41025In cases where the transport mechanism is itself reliable (such as a pipe or
41026TCP connection), the @samp{+}/@samp{-} acknowledgments are redundant.
41027It may be desirable to disable them in that case to reduce communication
41028overhead, or for other reasons. This can be accomplished by means of the
41029@samp{QStartNoAckMode} packet; @pxref{QStartNoAckMode}.
41030
41031When in no-acknowledgment mode, neither the stub nor @value{GDBN} shall send or
41032expect @samp{+}/@samp{-} protocol acknowledgments. The packet
41033and response format still includes the normal checksum, as described in
41034@ref{Overview}, but the checksum may be ignored by the receiver.
41035
41036If the stub supports @samp{QStartNoAckMode} and prefers to operate in
41037no-acknowledgment mode, it should report that to @value{GDBN}
41038by including @samp{QStartNoAckMode+} in its response to @samp{qSupported};
41039@pxref{qSupported}.
41040If @value{GDBN} also supports @samp{QStartNoAckMode} and it has not been
41041disabled via the @code{set remote noack-packet off} command
41042(@pxref{Remote Configuration}),
41043@value{GDBN} may then send a @samp{QStartNoAckMode} packet to the stub.
41044Only then may the stub actually turn off packet acknowledgments.
41045@value{GDBN} sends a final @samp{+} acknowledgment of the stub's @samp{OK}
41046response, which can be safely ignored by the stub.
41047
41048Note that @code{set remote noack-packet} command only affects negotiation
41049between @value{GDBN} and the stub when subsequent connections are made;
41050it does not affect the protocol acknowledgment state for any current
41051connection.
41052Since @samp{+}/@samp{-} acknowledgments are enabled by default when a
41053new connection is established,
41054there is also no protocol request to re-enable the acknowledgments
41055for the current connection, once disabled.
41056
41057@node Examples
41058@section Examples
41059
41060Example sequence of a target being re-started. Notice how the restart
41061does not get any direct output:
41062
41063@smallexample
41064-> @code{R00}
41065<- @code{+}
41066@emph{target restarts}
41067-> @code{?}
41068<- @code{+}
41069<- @code{T001:1234123412341234}
41070-> @code{+}
41071@end smallexample
41072
41073Example sequence of a target being stepped by a single instruction:
41074
41075@smallexample
41076-> @code{G1445@dots{}}
41077<- @code{+}
41078-> @code{s}
41079<- @code{+}
41080@emph{time passes}
41081<- @code{T001:1234123412341234}
41082-> @code{+}
41083-> @code{g}
41084<- @code{+}
41085<- @code{1455@dots{}}
41086-> @code{+}
41087@end smallexample
41088
41089@node File-I/O Remote Protocol Extension
41090@section File-I/O Remote Protocol Extension
41091@cindex File-I/O remote protocol extension
41092
41093@menu
41094* File-I/O Overview::
41095* Protocol Basics::
41096* The F Request Packet::
41097* The F Reply Packet::
41098* The Ctrl-C Message::
41099* Console I/O::
41100* List of Supported Calls::
41101* Protocol-specific Representation of Datatypes::
41102* Constants::
41103* File-I/O Examples::
41104@end menu
41105
41106@node File-I/O Overview
41107@subsection File-I/O Overview
41108@cindex file-i/o overview
41109
41110The @dfn{File I/O remote protocol extension} (short: File-I/O) allows the
41111target to use the host's file system and console I/O to perform various
41112system calls. System calls on the target system are translated into a
41113remote protocol packet to the host system, which then performs the needed
41114actions and returns a response packet to the target system.
41115This simulates file system operations even on targets that lack file systems.
41116
41117The protocol is defined to be independent of both the host and target systems.
41118It uses its own internal representation of datatypes and values. Both
41119@value{GDBN} and the target's @value{GDBN} stub are responsible for
41120translating the system-dependent value representations into the internal
41121protocol representations when data is transmitted.
41122
41123The communication is synchronous. A system call is possible only when
41124@value{GDBN} is waiting for a response from the @samp{C}, @samp{c}, @samp{S}
41125or @samp{s} packets. While @value{GDBN} handles the request for a system call,
41126the target is stopped to allow deterministic access to the target's
41127memory. Therefore File-I/O is not interruptible by target signals. On
41128the other hand, it is possible to interrupt File-I/O by a user interrupt
41129(@samp{Ctrl-C}) within @value{GDBN}.
41130
41131The target's request to perform a host system call does not finish
41132the latest @samp{C}, @samp{c}, @samp{S} or @samp{s} action. That means,
41133after finishing the system call, the target returns to continuing the
41134previous activity (continue, step). No additional continue or step
41135request from @value{GDBN} is required.
41136
41137@smallexample
41138(@value{GDBP}) continue
41139 <- target requests 'system call X'
41140 target is stopped, @value{GDBN} executes system call
41141 -> @value{GDBN} returns result
41142 ... target continues, @value{GDBN} returns to wait for the target
41143 <- target hits breakpoint and sends a Txx packet
41144@end smallexample
41145
41146The protocol only supports I/O on the console and to regular files on
41147the host file system. Character or block special devices, pipes,
41148named pipes, sockets or any other communication method on the host
41149system are not supported by this protocol.
41150
41151File I/O is not supported in non-stop mode.
41152
41153@node Protocol Basics
41154@subsection Protocol Basics
41155@cindex protocol basics, file-i/o
41156
41157The File-I/O protocol uses the @code{F} packet as the request as well
41158as reply packet. Since a File-I/O system call can only occur when
41159@value{GDBN} is waiting for a response from the continuing or stepping target,
41160the File-I/O request is a reply that @value{GDBN} has to expect as a result
41161of a previous @samp{C}, @samp{c}, @samp{S} or @samp{s} packet.
41162This @code{F} packet contains all information needed to allow @value{GDBN}
41163to call the appropriate host system call:
41164
41165@itemize @bullet
41166@item
41167A unique identifier for the requested system call.
41168
41169@item
41170All parameters to the system call. Pointers are given as addresses
41171in the target memory address space. Pointers to strings are given as
41172pointer/length pair. Numerical values are given as they are.
41173Numerical control flags are given in a protocol-specific representation.
41174
41175@end itemize
41176
41177At this point, @value{GDBN} has to perform the following actions.
41178
41179@itemize @bullet
41180@item
41181If the parameters include pointer values to data needed as input to a
41182system call, @value{GDBN} requests this data from the target with a
41183standard @code{m} packet request. This additional communication has to be
41184expected by the target implementation and is handled as any other @code{m}
41185packet.
41186
41187@item
41188@value{GDBN} translates all value from protocol representation to host
41189representation as needed. Datatypes are coerced into the host types.
41190
41191@item
41192@value{GDBN} calls the system call.
41193
41194@item
41195It then coerces datatypes back to protocol representation.
41196
41197@item
41198If the system call is expected to return data in buffer space specified
41199by pointer parameters to the call, the data is transmitted to the
41200target using a @code{M} or @code{X} packet. This packet has to be expected
41201by the target implementation and is handled as any other @code{M} or @code{X}
41202packet.
41203
41204@end itemize
41205
41206Eventually @value{GDBN} replies with another @code{F} packet which contains all
41207necessary information for the target to continue. This at least contains
41208
41209@itemize @bullet
41210@item
41211Return value.
41212
41213@item
41214@code{errno}, if has been changed by the system call.
41215
41216@item
41217``Ctrl-C'' flag.
41218
41219@end itemize
41220
41221After having done the needed type and value coercion, the target continues
41222the latest continue or step action.
41223
41224@node The F Request Packet
41225@subsection The @code{F} Request Packet
41226@cindex file-i/o request packet
41227@cindex @code{F} request packet
41228
41229The @code{F} request packet has the following format:
41230
41231@table @samp
41232@item F@var{call-id},@var{parameter@dots{}}
41233
41234@var{call-id} is the identifier to indicate the host system call to be called.
41235This is just the name of the function.
41236
41237@var{parameter@dots{}} are the parameters to the system call.
41238Parameters are hexadecimal integer values, either the actual values in case
41239of scalar datatypes, pointers to target buffer space in case of compound
41240datatypes and unspecified memory areas, or pointer/length pairs in case
41241of string parameters. These are appended to the @var{call-id} as a
41242comma-delimited list. All values are transmitted in ASCII
41243string representation, pointer/length pairs separated by a slash.
41244
41245@end table
41246
41247
41248
41249@node The F Reply Packet
41250@subsection The @code{F} Reply Packet
41251@cindex file-i/o reply packet
41252@cindex @code{F} reply packet
41253
41254The @code{F} reply packet has the following format:
41255
41256@table @samp
41257
41258@item F@var{retcode},@var{errno},@var{Ctrl-C flag};@var{call-specific attachment}
41259
41260@var{retcode} is the return code of the system call as hexadecimal value.
41261
41262@var{errno} is the @code{errno} set by the call, in protocol-specific
41263representation.
41264This parameter can be omitted if the call was successful.
41265
41266@var{Ctrl-C flag} is only sent if the user requested a break. In this
41267case, @var{errno} must be sent as well, even if the call was successful.
41268The @var{Ctrl-C flag} itself consists of the character @samp{C}:
41269
41270@smallexample
41271F0,0,C
41272@end smallexample
41273
41274@noindent
41275or, if the call was interrupted before the host call has been performed:
41276
41277@smallexample
41278F-1,4,C
41279@end smallexample
41280
41281@noindent
41282assuming 4 is the protocol-specific representation of @code{EINTR}.
41283
41284@end table
41285
41286
41287@node The Ctrl-C Message
41288@subsection The @samp{Ctrl-C} Message
41289@cindex ctrl-c message, in file-i/o protocol
41290
41291If the @samp{Ctrl-C} flag is set in the @value{GDBN}
41292reply packet (@pxref{The F Reply Packet}),
41293the target should behave as if it had
41294gotten a break message. The meaning for the target is ``system call
41295interrupted by @code{SIGINT}''. Consequentially, the target should actually stop
41296(as with a break message) and return to @value{GDBN} with a @code{T02}
41297packet.
41298
41299It's important for the target to know in which
41300state the system call was interrupted. There are two possible cases:
41301
41302@itemize @bullet
41303@item
41304The system call hasn't been performed on the host yet.
41305
41306@item
41307The system call on the host has been finished.
41308
41309@end itemize
41310
41311These two states can be distinguished by the target by the value of the
41312returned @code{errno}. If it's the protocol representation of @code{EINTR}, the system
41313call hasn't been performed. This is equivalent to the @code{EINTR} handling
41314on POSIX systems. In any other case, the target may presume that the
41315system call has been finished --- successfully or not --- and should behave
41316as if the break message arrived right after the system call.
41317
41318@value{GDBN} must behave reliably. If the system call has not been called
41319yet, @value{GDBN} may send the @code{F} reply immediately, setting @code{EINTR} as
41320@code{errno} in the packet. If the system call on the host has been finished
41321before the user requests a break, the full action must be finished by
41322@value{GDBN}. This requires sending @code{M} or @code{X} packets as necessary.
41323The @code{F} packet may only be sent when either nothing has happened
41324or the full action has been completed.
41325
41326@node Console I/O
41327@subsection Console I/O
41328@cindex console i/o as part of file-i/o
41329
41330By default and if not explicitly closed by the target system, the file
41331descriptors 0, 1 and 2 are connected to the @value{GDBN} console. Output
41332on the @value{GDBN} console is handled as any other file output operation
41333(@code{write(1, @dots{})} or @code{write(2, @dots{})}). Console input is handled
41334by @value{GDBN} so that after the target read request from file descriptor
413350 all following typing is buffered until either one of the following
41336conditions is met:
41337
41338@itemize @bullet
41339@item
41340The user types @kbd{Ctrl-c}. The behaviour is as explained above, and the
41341@code{read}
41342system call is treated as finished.
41343
41344@item
41345The user presses @key{RET}. This is treated as end of input with a trailing
41346newline.
41347
41348@item
41349The user types @kbd{Ctrl-d}. This is treated as end of input. No trailing
41350character (neither newline nor @samp{Ctrl-D}) is appended to the input.
41351
41352@end itemize
41353
41354If the user has typed more characters than fit in the buffer given to
41355the @code{read} call, the trailing characters are buffered in @value{GDBN} until
41356either another @code{read(0, @dots{})} is requested by the target, or debugging
41357is stopped at the user's request.
41358
41359
41360@node List of Supported Calls
41361@subsection List of Supported Calls
41362@cindex list of supported file-i/o calls
41363
41364@menu
41365* open::
41366* close::
41367* read::
41368* write::
41369* lseek::
41370* rename::
41371* unlink::
41372* stat/fstat::
41373* gettimeofday::
41374* isatty::
41375* system::
41376@end menu
41377
41378@node open
41379@unnumberedsubsubsec open
41380@cindex open, file-i/o system call
41381
41382@table @asis
41383@item Synopsis:
41384@smallexample
41385int open(const char *pathname, int flags);
41386int open(const char *pathname, int flags, mode_t mode);
41387@end smallexample
41388
41389@item Request:
41390@samp{Fopen,@var{pathptr}/@var{len},@var{flags},@var{mode}}
41391
41392@noindent
41393@var{flags} is the bitwise @code{OR} of the following values:
41394
41395@table @code
41396@item O_CREAT
41397If the file does not exist it will be created. The host
41398rules apply as far as file ownership and time stamps
41399are concerned.
41400
41401@item O_EXCL
41402When used with @code{O_CREAT}, if the file already exists it is
41403an error and open() fails.
41404
41405@item O_TRUNC
41406If the file already exists and the open mode allows
41407writing (@code{O_RDWR} or @code{O_WRONLY} is given) it will be
41408truncated to zero length.
41409
41410@item O_APPEND
41411The file is opened in append mode.
41412
41413@item O_RDONLY
41414The file is opened for reading only.
41415
41416@item O_WRONLY
41417The file is opened for writing only.
41418
41419@item O_RDWR
41420The file is opened for reading and writing.
41421@end table
41422
41423@noindent
41424Other bits are silently ignored.
41425
41426
41427@noindent
41428@var{mode} is the bitwise @code{OR} of the following values:
41429
41430@table @code
41431@item S_IRUSR
41432User has read permission.
41433
41434@item S_IWUSR
41435User has write permission.
41436
41437@item S_IRGRP
41438Group has read permission.
41439
41440@item S_IWGRP
41441Group has write permission.
41442
41443@item S_IROTH
41444Others have read permission.
41445
41446@item S_IWOTH
41447Others have write permission.
41448@end table
41449
41450@noindent
41451Other bits are silently ignored.
41452
41453
41454@item Return value:
41455@code{open} returns the new file descriptor or -1 if an error
41456occurred.
41457
41458@item Errors:
41459
41460@table @code
41461@item EEXIST
41462@var{pathname} already exists and @code{O_CREAT} and @code{O_EXCL} were used.
41463
41464@item EISDIR
41465@var{pathname} refers to a directory.
41466
41467@item EACCES
41468The requested access is not allowed.
41469
41470@item ENAMETOOLONG
41471@var{pathname} was too long.
41472
41473@item ENOENT
41474A directory component in @var{pathname} does not exist.
41475
41476@item ENODEV
41477@var{pathname} refers to a device, pipe, named pipe or socket.
41478
41479@item EROFS
41480@var{pathname} refers to a file on a read-only filesystem and
41481write access was requested.
41482
41483@item EFAULT
41484@var{pathname} is an invalid pointer value.
41485
41486@item ENOSPC
41487No space on device to create the file.
41488
41489@item EMFILE
41490The process already has the maximum number of files open.
41491
41492@item ENFILE
41493The limit on the total number of files open on the system
41494has been reached.
41495
41496@item EINTR
41497The call was interrupted by the user.
41498@end table
41499
41500@end table
41501
41502@node close
41503@unnumberedsubsubsec close
41504@cindex close, file-i/o system call
41505
41506@table @asis
41507@item Synopsis:
41508@smallexample
41509int close(int fd);
41510@end smallexample
41511
41512@item Request:
41513@samp{Fclose,@var{fd}}
41514
41515@item Return value:
41516@code{close} returns zero on success, or -1 if an error occurred.
41517
41518@item Errors:
41519
41520@table @code
41521@item EBADF
41522@var{fd} isn't a valid open file descriptor.
41523
41524@item EINTR
41525The call was interrupted by the user.
41526@end table
41527
41528@end table
41529
41530@node read
41531@unnumberedsubsubsec read
41532@cindex read, file-i/o system call
41533
41534@table @asis
41535@item Synopsis:
41536@smallexample
41537int read(int fd, void *buf, unsigned int count);
41538@end smallexample
41539
41540@item Request:
41541@samp{Fread,@var{fd},@var{bufptr},@var{count}}
41542
41543@item Return value:
41544On success, the number of bytes read is returned.
41545Zero indicates end of file. If count is zero, read
41546returns zero as well. On error, -1 is returned.
41547
41548@item Errors:
41549
41550@table @code
41551@item EBADF
41552@var{fd} is not a valid file descriptor or is not open for
41553reading.
41554
41555@item EFAULT
41556@var{bufptr} is an invalid pointer value.
41557
41558@item EINTR
41559The call was interrupted by the user.
41560@end table
41561
41562@end table
41563
41564@node write
41565@unnumberedsubsubsec write
41566@cindex write, file-i/o system call
41567
41568@table @asis
41569@item Synopsis:
41570@smallexample
41571int write(int fd, const void *buf, unsigned int count);
41572@end smallexample
41573
41574@item Request:
41575@samp{Fwrite,@var{fd},@var{bufptr},@var{count}}
41576
41577@item Return value:
41578On success, the number of bytes written are returned.
41579Zero indicates nothing was written. On error, -1
41580is returned.
41581
41582@item Errors:
41583
41584@table @code
41585@item EBADF
41586@var{fd} is not a valid file descriptor or is not open for
41587writing.
41588
41589@item EFAULT
41590@var{bufptr} is an invalid pointer value.
41591
41592@item EFBIG
41593An attempt was made to write a file that exceeds the
41594host-specific maximum file size allowed.
41595
41596@item ENOSPC
41597No space on device to write the data.
41598
41599@item EINTR
41600The call was interrupted by the user.
41601@end table
41602
41603@end table
41604
41605@node lseek
41606@unnumberedsubsubsec lseek
41607@cindex lseek, file-i/o system call
41608
41609@table @asis
41610@item Synopsis:
41611@smallexample
41612long lseek (int fd, long offset, int flag);
41613@end smallexample
41614
41615@item Request:
41616@samp{Flseek,@var{fd},@var{offset},@var{flag}}
41617
41618@var{flag} is one of:
41619
41620@table @code
41621@item SEEK_SET
41622The offset is set to @var{offset} bytes.
41623
41624@item SEEK_CUR
41625The offset is set to its current location plus @var{offset}
41626bytes.
41627
41628@item SEEK_END
41629The offset is set to the size of the file plus @var{offset}
41630bytes.
41631@end table
41632
41633@item Return value:
41634On success, the resulting unsigned offset in bytes from
41635the beginning of the file is returned. Otherwise, a
41636value of -1 is returned.
41637
41638@item Errors:
41639
41640@table @code
41641@item EBADF
41642@var{fd} is not a valid open file descriptor.
41643
41644@item ESPIPE
41645@var{fd} is associated with the @value{GDBN} console.
41646
41647@item EINVAL
41648@var{flag} is not a proper value.
41649
41650@item EINTR
41651The call was interrupted by the user.
41652@end table
41653
41654@end table
41655
41656@node rename
41657@unnumberedsubsubsec rename
41658@cindex rename, file-i/o system call
41659
41660@table @asis
41661@item Synopsis:
41662@smallexample
41663int rename(const char *oldpath, const char *newpath);
41664@end smallexample
41665
41666@item Request:
41667@samp{Frename,@var{oldpathptr}/@var{len},@var{newpathptr}/@var{len}}
41668
41669@item Return value:
41670On success, zero is returned. On error, -1 is returned.
41671
41672@item Errors:
41673
41674@table @code
41675@item EISDIR
41676@var{newpath} is an existing directory, but @var{oldpath} is not a
41677directory.
41678
41679@item EEXIST
41680@var{newpath} is a non-empty directory.
41681
41682@item EBUSY
41683@var{oldpath} or @var{newpath} is a directory that is in use by some
41684process.
41685
41686@item EINVAL
41687An attempt was made to make a directory a subdirectory
41688of itself.
41689
41690@item ENOTDIR
41691A component used as a directory in @var{oldpath} or new
41692path is not a directory. Or @var{oldpath} is a directory
41693and @var{newpath} exists but is not a directory.
41694
41695@item EFAULT
41696@var{oldpathptr} or @var{newpathptr} are invalid pointer values.
41697
41698@item EACCES
41699No access to the file or the path of the file.
41700
41701@item ENAMETOOLONG
41702
41703@var{oldpath} or @var{newpath} was too long.
41704
41705@item ENOENT
41706A directory component in @var{oldpath} or @var{newpath} does not exist.
41707
41708@item EROFS
41709The file is on a read-only filesystem.
41710
41711@item ENOSPC
41712The device containing the file has no room for the new
41713directory entry.
41714
41715@item EINTR
41716The call was interrupted by the user.
41717@end table
41718
41719@end table
41720
41721@node unlink
41722@unnumberedsubsubsec unlink
41723@cindex unlink, file-i/o system call
41724
41725@table @asis
41726@item Synopsis:
41727@smallexample
41728int unlink(const char *pathname);
41729@end smallexample
41730
41731@item Request:
41732@samp{Funlink,@var{pathnameptr}/@var{len}}
41733
41734@item Return value:
41735On success, zero is returned. On error, -1 is returned.
41736
41737@item Errors:
41738
41739@table @code
41740@item EACCES
41741No access to the file or the path of the file.
41742
41743@item EPERM
41744The system does not allow unlinking of directories.
41745
41746@item EBUSY
41747The file @var{pathname} cannot be unlinked because it's
41748being used by another process.
41749
41750@item EFAULT
41751@var{pathnameptr} is an invalid pointer value.
41752
41753@item ENAMETOOLONG
41754@var{pathname} was too long.
41755
41756@item ENOENT
41757A directory component in @var{pathname} does not exist.
41758
41759@item ENOTDIR
41760A component of the path is not a directory.
41761
41762@item EROFS
41763The file is on a read-only filesystem.
41764
41765@item EINTR
41766The call was interrupted by the user.
41767@end table
41768
41769@end table
41770
41771@node stat/fstat
41772@unnumberedsubsubsec stat/fstat
41773@cindex fstat, file-i/o system call
41774@cindex stat, file-i/o system call
41775
41776@table @asis
41777@item Synopsis:
41778@smallexample
41779int stat(const char *pathname, struct stat *buf);
41780int fstat(int fd, struct stat *buf);
41781@end smallexample
41782
41783@item Request:
41784@samp{Fstat,@var{pathnameptr}/@var{len},@var{bufptr}}@*
41785@samp{Ffstat,@var{fd},@var{bufptr}}
41786
41787@item Return value:
41788On success, zero is returned. On error, -1 is returned.
41789
41790@item Errors:
41791
41792@table @code
41793@item EBADF
41794@var{fd} is not a valid open file.
41795
41796@item ENOENT
41797A directory component in @var{pathname} does not exist or the
41798path is an empty string.
41799
41800@item ENOTDIR
41801A component of the path is not a directory.
41802
41803@item EFAULT
41804@var{pathnameptr} is an invalid pointer value.
41805
41806@item EACCES
41807No access to the file or the path of the file.
41808
41809@item ENAMETOOLONG
41810@var{pathname} was too long.
41811
41812@item EINTR
41813The call was interrupted by the user.
41814@end table
41815
41816@end table
41817
41818@node gettimeofday
41819@unnumberedsubsubsec gettimeofday
41820@cindex gettimeofday, file-i/o system call
41821
41822@table @asis
41823@item Synopsis:
41824@smallexample
41825int gettimeofday(struct timeval *tv, void *tz);
41826@end smallexample
41827
41828@item Request:
41829@samp{Fgettimeofday,@var{tvptr},@var{tzptr}}
41830
41831@item Return value:
41832On success, 0 is returned, -1 otherwise.
41833
41834@item Errors:
41835
41836@table @code
41837@item EINVAL
41838@var{tz} is a non-NULL pointer.
41839
41840@item EFAULT
41841@var{tvptr} and/or @var{tzptr} is an invalid pointer value.
41842@end table
41843
41844@end table
41845
41846@node isatty
41847@unnumberedsubsubsec isatty
41848@cindex isatty, file-i/o system call
41849
41850@table @asis
41851@item Synopsis:
41852@smallexample
41853int isatty(int fd);
41854@end smallexample
41855
41856@item Request:
41857@samp{Fisatty,@var{fd}}
41858
41859@item Return value:
41860Returns 1 if @var{fd} refers to the @value{GDBN} console, 0 otherwise.
41861
41862@item Errors:
41863
41864@table @code
41865@item EINTR
41866The call was interrupted by the user.
41867@end table
41868
41869@end table
41870
41871Note that the @code{isatty} call is treated as a special case: it returns
418721 to the target if the file descriptor is attached
41873to the @value{GDBN} console, 0 otherwise. Implementing through system calls
41874would require implementing @code{ioctl} and would be more complex than
41875needed.
41876
41877
41878@node system
41879@unnumberedsubsubsec system
41880@cindex system, file-i/o system call
41881
41882@table @asis
41883@item Synopsis:
41884@smallexample
41885int system(const char *command);
41886@end smallexample
41887
41888@item Request:
41889@samp{Fsystem,@var{commandptr}/@var{len}}
41890
41891@item Return value:
41892If @var{len} is zero, the return value indicates whether a shell is
41893available. A zero return value indicates a shell is not available.
41894For non-zero @var{len}, the value returned is -1 on error and the
41895return status of the command otherwise. Only the exit status of the
41896command is returned, which is extracted from the host's @code{system}
41897return value by calling @code{WEXITSTATUS(retval)}. In case
41898@file{/bin/sh} could not be executed, 127 is returned.
41899
41900@item Errors:
41901
41902@table @code
41903@item EINTR
41904The call was interrupted by the user.
41905@end table
41906
41907@end table
41908
41909@value{GDBN} takes over the full task of calling the necessary host calls
41910to perform the @code{system} call. The return value of @code{system} on
41911the host is simplified before it's returned
41912to the target. Any termination signal information from the child process
41913is discarded, and the return value consists
41914entirely of the exit status of the called command.
41915
41916Due to security concerns, the @code{system} call is by default refused
41917by @value{GDBN}. The user has to allow this call explicitly with the
41918@code{set remote system-call-allowed 1} command.
41919
41920@table @code
41921@item set remote system-call-allowed
41922@kindex set remote system-call-allowed
41923Control whether to allow the @code{system} calls in the File I/O
41924protocol for the remote target. The default is zero (disabled).
41925
41926@item show remote system-call-allowed
41927@kindex show remote system-call-allowed
41928Show whether the @code{system} calls are allowed in the File I/O
41929protocol.
41930@end table
41931
41932@node Protocol-specific Representation of Datatypes
41933@subsection Protocol-specific Representation of Datatypes
41934@cindex protocol-specific representation of datatypes, in file-i/o protocol
41935
41936@menu
41937* Integral Datatypes::
41938* Pointer Values::
41939* Memory Transfer::
41940* struct stat::
41941* struct timeval::
41942@end menu
41943
41944@node Integral Datatypes
41945@unnumberedsubsubsec Integral Datatypes
41946@cindex integral datatypes, in file-i/o protocol
41947
41948The integral datatypes used in the system calls are @code{int},
41949@code{unsigned int}, @code{long}, @code{unsigned long},
41950@code{mode_t}, and @code{time_t}.
41951
41952@code{int}, @code{unsigned int}, @code{mode_t} and @code{time_t} are
41953implemented as 32 bit values in this protocol.
41954
41955@code{long} and @code{unsigned long} are implemented as 64 bit types.
41956
41957@xref{Limits}, for corresponding MIN and MAX values (similar to those
41958in @file{limits.h}) to allow range checking on host and target.
41959
41960@code{time_t} datatypes are defined as seconds since the Epoch.
41961
41962All integral datatypes transferred as part of a memory read or write of a
41963structured datatype e.g.@: a @code{struct stat} have to be given in big endian
41964byte order.
41965
41966@node Pointer Values
41967@unnumberedsubsubsec Pointer Values
41968@cindex pointer values, in file-i/o protocol
41969
41970Pointers to target data are transmitted as they are. An exception
41971is made for pointers to buffers for which the length isn't
41972transmitted as part of the function call, namely strings. Strings
41973are transmitted as a pointer/length pair, both as hex values, e.g.@:
41974
41975@smallexample
41976@code{1aaf/12}
41977@end smallexample
41978
41979@noindent
41980which is a pointer to data of length 18 bytes at position 0x1aaf.
41981The length is defined as the full string length in bytes, including
41982the trailing null byte. For example, the string @code{"hello world"}
41983at address 0x123456 is transmitted as
41984
41985@smallexample
41986@code{123456/d}
41987@end smallexample
41988
41989@node Memory Transfer
41990@unnumberedsubsubsec Memory Transfer
41991@cindex memory transfer, in file-i/o protocol
41992
41993Structured data which is transferred using a memory read or write (for
41994example, a @code{struct stat}) is expected to be in a protocol-specific format
41995with all scalar multibyte datatypes being big endian. Translation to
41996this representation needs to be done both by the target before the @code{F}
41997packet is sent, and by @value{GDBN} before
41998it transfers memory to the target. Transferred pointers to structured
41999data should point to the already-coerced data at any time.
42000
42001
42002@node struct stat
42003@unnumberedsubsubsec struct stat
42004@cindex struct stat, in file-i/o protocol
42005
42006The buffer of type @code{struct stat} used by the target and @value{GDBN}
42007is defined as follows:
42008
42009@smallexample
42010struct stat @{
42011 unsigned int st_dev; /* device */
42012 unsigned int st_ino; /* inode */
42013 mode_t st_mode; /* protection */
42014 unsigned int st_nlink; /* number of hard links */
42015 unsigned int st_uid; /* user ID of owner */
42016 unsigned int st_gid; /* group ID of owner */
42017 unsigned int st_rdev; /* device type (if inode device) */
42018 unsigned long st_size; /* total size, in bytes */
42019 unsigned long st_blksize; /* blocksize for filesystem I/O */
42020 unsigned long st_blocks; /* number of blocks allocated */
42021 time_t st_atime; /* time of last access */
42022 time_t st_mtime; /* time of last modification */
42023 time_t st_ctime; /* time of last change */
42024@};
42025@end smallexample
42026
42027The integral datatypes conform to the definitions given in the
42028appropriate section (see @ref{Integral Datatypes}, for details) so this
42029structure is of size 64 bytes.
42030
42031The values of several fields have a restricted meaning and/or
42032range of values.
42033
42034@table @code
42035
42036@item st_dev
42037A value of 0 represents a file, 1 the console.
42038
42039@item st_ino
42040No valid meaning for the target. Transmitted unchanged.
42041
42042@item st_mode
42043Valid mode bits are described in @ref{Constants}. Any other
42044bits have currently no meaning for the target.
42045
42046@item st_uid
42047@itemx st_gid
42048@itemx st_rdev
42049No valid meaning for the target. Transmitted unchanged.
42050
42051@item st_atime
42052@itemx st_mtime
42053@itemx st_ctime
42054These values have a host and file system dependent
42055accuracy. Especially on Windows hosts, the file system may not
42056support exact timing values.
42057@end table
42058
42059The target gets a @code{struct stat} of the above representation and is
42060responsible for coercing it to the target representation before
42061continuing.
42062
42063Note that due to size differences between the host, target, and protocol
42064representations of @code{struct stat} members, these members could eventually
42065get truncated on the target.
42066
42067@node struct timeval
42068@unnumberedsubsubsec struct timeval
42069@cindex struct timeval, in file-i/o protocol
42070
42071The buffer of type @code{struct timeval} used by the File-I/O protocol
42072is defined as follows:
42073
42074@smallexample
42075struct timeval @{
42076 time_t tv_sec; /* second */
42077 long tv_usec; /* microsecond */
42078@};
42079@end smallexample
42080
42081The integral datatypes conform to the definitions given in the
42082appropriate section (see @ref{Integral Datatypes}, for details) so this
42083structure is of size 8 bytes.
42084
42085@node Constants
42086@subsection Constants
42087@cindex constants, in file-i/o protocol
42088
42089The following values are used for the constants inside of the
42090protocol. @value{GDBN} and target are responsible for translating these
42091values before and after the call as needed.
42092
42093@menu
42094* Open Flags::
42095* mode_t Values::
42096* Errno Values::
42097* Lseek Flags::
42098* Limits::
42099@end menu
42100
42101@node Open Flags
42102@unnumberedsubsubsec Open Flags
42103@cindex open flags, in file-i/o protocol
42104
42105All values are given in hexadecimal representation.
42106
42107@smallexample
42108 O_RDONLY 0x0
42109 O_WRONLY 0x1
42110 O_RDWR 0x2
42111 O_APPEND 0x8
42112 O_CREAT 0x200
42113 O_TRUNC 0x400
42114 O_EXCL 0x800
42115@end smallexample
42116
42117@node mode_t Values
42118@unnumberedsubsubsec mode_t Values
42119@cindex mode_t values, in file-i/o protocol
42120
42121All values are given in octal representation.
42122
42123@smallexample
42124 S_IFREG 0100000
42125 S_IFDIR 040000
42126 S_IRUSR 0400
42127 S_IWUSR 0200
42128 S_IXUSR 0100
42129 S_IRGRP 040
42130 S_IWGRP 020
42131 S_IXGRP 010
42132 S_IROTH 04
42133 S_IWOTH 02
42134 S_IXOTH 01
42135@end smallexample
42136
42137@node Errno Values
42138@unnumberedsubsubsec Errno Values
42139@cindex errno values, in file-i/o protocol
42140
42141All values are given in decimal representation.
42142
42143@smallexample
42144 EPERM 1
42145 ENOENT 2
42146 EINTR 4
42147 EBADF 9
42148 EACCES 13
42149 EFAULT 14
42150 EBUSY 16
42151 EEXIST 17
42152 ENODEV 19
42153 ENOTDIR 20
42154 EISDIR 21
42155 EINVAL 22
42156 ENFILE 23
42157 EMFILE 24
42158 EFBIG 27
42159 ENOSPC 28
42160 ESPIPE 29
42161 EROFS 30
42162 ENAMETOOLONG 91
42163 EUNKNOWN 9999
42164@end smallexample
42165
42166 @code{EUNKNOWN} is used as a fallback error value if a host system returns
42167 any error value not in the list of supported error numbers.
42168
42169@node Lseek Flags
42170@unnumberedsubsubsec Lseek Flags
42171@cindex lseek flags, in file-i/o protocol
42172
42173@smallexample
42174 SEEK_SET 0
42175 SEEK_CUR 1
42176 SEEK_END 2
42177@end smallexample
42178
42179@node Limits
42180@unnumberedsubsubsec Limits
42181@cindex limits, in file-i/o protocol
42182
42183All values are given in decimal representation.
42184
42185@smallexample
42186 INT_MIN -2147483648
42187 INT_MAX 2147483647
42188 UINT_MAX 4294967295
42189 LONG_MIN -9223372036854775808
42190 LONG_MAX 9223372036854775807
42191 ULONG_MAX 18446744073709551615
42192@end smallexample
42193
42194@node File-I/O Examples
42195@subsection File-I/O Examples
42196@cindex file-i/o examples
42197
42198Example sequence of a write call, file descriptor 3, buffer is at target
42199address 0x1234, 6 bytes should be written:
42200
42201@smallexample
42202<- @code{Fwrite,3,1234,6}
42203@emph{request memory read from target}
42204-> @code{m1234,6}
42205<- XXXXXX
42206@emph{return "6 bytes written"}
42207-> @code{F6}
42208@end smallexample
42209
42210Example sequence of a read call, file descriptor 3, buffer is at target
42211address 0x1234, 6 bytes should be read:
42212
42213@smallexample
42214<- @code{Fread,3,1234,6}
42215@emph{request memory write to target}
42216-> @code{X1234,6:XXXXXX}
42217@emph{return "6 bytes read"}
42218-> @code{F6}
42219@end smallexample
42220
42221Example sequence of a read call, call fails on the host due to invalid
42222file descriptor (@code{EBADF}):
42223
42224@smallexample
42225<- @code{Fread,3,1234,6}
42226-> @code{F-1,9}
42227@end smallexample
42228
42229Example sequence of a read call, user presses @kbd{Ctrl-c} before syscall on
42230host is called:
42231
42232@smallexample
42233<- @code{Fread,3,1234,6}
42234-> @code{F-1,4,C}
42235<- @code{T02}
42236@end smallexample
42237
42238Example sequence of a read call, user presses @kbd{Ctrl-c} after syscall on
42239host is called:
42240
42241@smallexample
42242<- @code{Fread,3,1234,6}
42243-> @code{X1234,6:XXXXXX}
42244<- @code{T02}
42245@end smallexample
42246
42247@node Library List Format
42248@section Library List Format
42249@cindex library list format, remote protocol
42250
42251On some platforms, a dynamic loader (e.g.@: @file{ld.so}) runs in the
42252same process as your application to manage libraries. In this case,
42253@value{GDBN} can use the loader's symbol table and normal memory
42254operations to maintain a list of shared libraries. On other
42255platforms, the operating system manages loaded libraries.
42256@value{GDBN} can not retrieve the list of currently loaded libraries
42257through memory operations, so it uses the @samp{qXfer:libraries:read}
42258packet (@pxref{qXfer library list read}) instead. The remote stub
42259queries the target's operating system and reports which libraries
42260are loaded.
42261
42262The @samp{qXfer:libraries:read} packet returns an XML document which
42263lists loaded libraries and their offsets. Each library has an
42264associated name and one or more segment or section base addresses,
42265which report where the library was loaded in memory.
42266
42267For the common case of libraries that are fully linked binaries, the
42268library should have a list of segments. If the target supports
42269dynamic linking of a relocatable object file, its library XML element
42270should instead include a list of allocated sections. The segment or
42271section bases are start addresses, not relocation offsets; they do not
42272depend on the library's link-time base addresses.
42273
42274@value{GDBN} must be linked with the Expat library to support XML
42275library lists. @xref{Expat}.
42276
42277A simple memory map, with one loaded library relocated by a single
42278offset, looks like this:
42279
42280@smallexample
42281<library-list>
42282 <library name="/lib/libc.so.6">
42283 <segment address="0x10000000"/>
42284 </library>
42285</library-list>
42286@end smallexample
42287
42288Another simple memory map, with one loaded library with three
42289allocated sections (.text, .data, .bss), looks like this:
42290
42291@smallexample
42292<library-list>
42293 <library name="sharedlib.o">
42294 <section address="0x10000000"/>
42295 <section address="0x20000000"/>
42296 <section address="0x30000000"/>
42297 </library>
42298</library-list>
42299@end smallexample
42300
42301The format of a library list is described by this DTD:
42302
42303@smallexample
42304<!-- library-list: Root element with versioning -->
42305<!ELEMENT library-list (library)*>
42306<!ATTLIST library-list version CDATA #FIXED "1.0">
42307<!ELEMENT library (segment*, section*)>
42308<!ATTLIST library name CDATA #REQUIRED>
42309<!ELEMENT segment EMPTY>
42310<!ATTLIST segment address CDATA #REQUIRED>
42311<!ELEMENT section EMPTY>
42312<!ATTLIST section address CDATA #REQUIRED>
42313@end smallexample
42314
42315In addition, segments and section descriptors cannot be mixed within a
42316single library element, and you must supply at least one segment or
42317section for each library.
42318
42319@node Library List Format for SVR4 Targets
42320@section Library List Format for SVR4 Targets
42321@cindex library list format, remote protocol
42322
42323On SVR4 platforms @value{GDBN} can use the symbol table of a dynamic loader
42324(e.g.@: @file{ld.so}) and normal memory operations to maintain a list of
42325shared libraries. Still a special library list provided by this packet is
42326more efficient for the @value{GDBN} remote protocol.
42327
42328The @samp{qXfer:libraries-svr4:read} packet returns an XML document which lists
42329loaded libraries and their SVR4 linker parameters. For each library on SVR4
42330target, the following parameters are reported:
42331
42332@itemize @minus
42333@item
42334@code{name}, the absolute file name from the @code{l_name} field of
42335@code{struct link_map}.
42336@item
42337@code{lm} with address of @code{struct link_map} used for TLS
42338(Thread Local Storage) access.
42339@item
42340@code{l_addr}, the displacement as read from the field @code{l_addr} of
42341@code{struct link_map}. For prelinked libraries this is not an absolute
42342memory address. It is a displacement of absolute memory address against
42343address the file was prelinked to during the library load.
42344@item
42345@code{l_ld}, which is memory address of the @code{PT_DYNAMIC} segment
42346@end itemize
42347
42348Additionally the single @code{main-lm} attribute specifies address of
42349@code{struct link_map} used for the main executable. This parameter is used
42350for TLS access and its presence is optional.
42351
42352@value{GDBN} must be linked with the Expat library to support XML
42353SVR4 library lists. @xref{Expat}.
42354
42355A simple memory map, with two loaded libraries (which do not use prelink),
42356looks like this:
42357
42358@smallexample
42359<library-list-svr4 version="1.0" main-lm="0xe4f8f8">
42360 <library name="/lib/ld-linux.so.2" lm="0xe4f51c" l_addr="0xe2d000"
42361 l_ld="0xe4eefc"/>
42362 <library name="/lib/libc.so.6" lm="0xe4fbe8" l_addr="0x154000"
42363 l_ld="0x152350"/>
42364</library-list-svr>
42365@end smallexample
42366
42367The format of an SVR4 library list is described by this DTD:
42368
42369@smallexample
42370<!-- library-list-svr4: Root element with versioning -->
42371<!ELEMENT library-list-svr4 (library)*>
42372<!ATTLIST library-list-svr4 version CDATA #FIXED "1.0">
42373<!ATTLIST library-list-svr4 main-lm CDATA #IMPLIED>
42374<!ELEMENT library EMPTY>
42375<!ATTLIST library name CDATA #REQUIRED>
42376<!ATTLIST library lm CDATA #REQUIRED>
42377<!ATTLIST library l_addr CDATA #REQUIRED>
42378<!ATTLIST library l_ld CDATA #REQUIRED>
42379@end smallexample
42380
42381@node Memory Map Format
42382@section Memory Map Format
42383@cindex memory map format
42384
42385To be able to write into flash memory, @value{GDBN} needs to obtain a
42386memory map from the target. This section describes the format of the
42387memory map.
42388
42389The memory map is obtained using the @samp{qXfer:memory-map:read}
42390(@pxref{qXfer memory map read}) packet and is an XML document that
42391lists memory regions.
42392
42393@value{GDBN} must be linked with the Expat library to support XML
42394memory maps. @xref{Expat}.
42395
42396The top-level structure of the document is shown below:
42397
42398@smallexample
42399<?xml version="1.0"?>
42400<!DOCTYPE memory-map
42401 PUBLIC "+//IDN gnu.org//DTD GDB Memory Map V1.0//EN"
42402 "http://sourceware.org/gdb/gdb-memory-map.dtd">
42403<memory-map>
42404 region...
42405</memory-map>
42406@end smallexample
42407
42408Each region can be either:
42409
42410@itemize
42411
42412@item
42413A region of RAM starting at @var{addr} and extending for @var{length}
42414bytes from there:
42415
42416@smallexample
42417<memory type="ram" start="@var{addr}" length="@var{length}"/>
42418@end smallexample
42419
42420
42421@item
42422A region of read-only memory:
42423
42424@smallexample
42425<memory type="rom" start="@var{addr}" length="@var{length}"/>
42426@end smallexample
42427
42428
42429@item
42430A region of flash memory, with erasure blocks @var{blocksize}
42431bytes in length:
42432
42433@smallexample
42434<memory type="flash" start="@var{addr}" length="@var{length}">
42435 <property name="blocksize">@var{blocksize}</property>
42436</memory>
42437@end smallexample
42438
42439@end itemize
42440
42441Regions must not overlap. @value{GDBN} assumes that areas of memory not covered
42442by the memory map are RAM, and uses the ordinary @samp{M} and @samp{X}
42443packets to write to addresses in such ranges.
42444
42445The formal DTD for memory map format is given below:
42446
42447@smallexample
42448<!-- ................................................... -->
42449<!-- Memory Map XML DTD ................................ -->
42450<!-- File: memory-map.dtd .............................. -->
42451<!-- .................................... .............. -->
42452<!-- memory-map.dtd -->
42453<!-- memory-map: Root element with versioning -->
42454<!ELEMENT memory-map (memory | property)>
42455<!ATTLIST memory-map version CDATA #FIXED "1.0.0">
42456<!ELEMENT memory (property)>
42457<!-- memory: Specifies a memory region,
42458 and its type, or device. -->
42459<!ATTLIST memory type CDATA #REQUIRED
42460 start CDATA #REQUIRED
42461 length CDATA #REQUIRED
42462 device CDATA #IMPLIED>
42463<!-- property: Generic attribute tag -->
42464<!ELEMENT property (#PCDATA | property)*>
42465<!ATTLIST property name CDATA #REQUIRED>
42466@end smallexample
42467
42468@node Thread List Format
42469@section Thread List Format
42470@cindex thread list format
42471
42472To efficiently update the list of threads and their attributes,
42473@value{GDBN} issues the @samp{qXfer:threads:read} packet
42474(@pxref{qXfer threads read}) and obtains the XML document with
42475the following structure:
42476
42477@smallexample
42478<?xml version="1.0"?>
42479<threads>
42480 <thread id="id" core="0">
42481 ... description ...
42482 </thread>
42483</threads>
42484@end smallexample
42485
42486Each @samp{thread} element must have the @samp{id} attribute that
42487identifies the thread (@pxref{thread-id syntax}). The
42488@samp{core} attribute, if present, specifies which processor core
42489the thread was last executing on. The content of the of @samp{thread}
42490element is interpreted as human-readable auxilliary information.
42491
42492@node Traceframe Info Format
42493@section Traceframe Info Format
42494@cindex traceframe info format
42495
42496To be able to know which objects in the inferior can be examined when
42497inspecting a tracepoint hit, @value{GDBN} needs to obtain the list of
42498memory ranges, registers and trace state variables that have been
42499collected in a traceframe.
42500
42501This list is obtained using the @samp{qXfer:traceframe-info:read}
42502(@pxref{qXfer traceframe info read}) packet and is an XML document.
42503
42504@value{GDBN} must be linked with the Expat library to support XML
42505traceframe info discovery. @xref{Expat}.
42506
42507The top-level structure of the document is shown below:
42508
42509@smallexample
42510<?xml version="1.0"?>
42511<!DOCTYPE traceframe-info
42512 PUBLIC "+//IDN gnu.org//DTD GDB Memory Map V1.0//EN"
42513 "http://sourceware.org/gdb/gdb-traceframe-info.dtd">
42514<traceframe-info>
42515 block...
42516</traceframe-info>
42517@end smallexample
42518
42519Each traceframe block can be either:
42520
42521@itemize
42522
42523@item
42524A region of collected memory starting at @var{addr} and extending for
42525@var{length} bytes from there:
42526
42527@smallexample
42528<memory start="@var{addr}" length="@var{length}"/>
42529@end smallexample
42530
42531@item
42532A block indicating trace state variable numbered @var{number} has been
42533collected:
42534
42535@smallexample
42536<tvar id="@var{number}"/>
42537@end smallexample
42538
42539@end itemize
42540
42541The formal DTD for the traceframe info format is given below:
42542
42543@smallexample
42544<!ELEMENT traceframe-info (memory | tvar)* >
42545<!ATTLIST traceframe-info version CDATA #FIXED "1.0">
42546
42547<!ELEMENT memory EMPTY>
42548<!ATTLIST memory start CDATA #REQUIRED
42549 length CDATA #REQUIRED>
42550<!ELEMENT tvar>
42551<!ATTLIST tvar id CDATA #REQUIRED>
42552@end smallexample
42553
42554@node Branch Trace Format
42555@section Branch Trace Format
42556@cindex branch trace format
42557
42558In order to display the branch trace of an inferior thread,
42559@value{GDBN} needs to obtain the list of branches. This list is
42560represented as list of sequential code blocks that are connected via
42561branches. The code in each block has been executed sequentially.
42562
42563This list is obtained using the @samp{qXfer:btrace:read}
42564(@pxref{qXfer btrace read}) packet and is an XML document.
42565
42566@value{GDBN} must be linked with the Expat library to support XML
42567traceframe info discovery. @xref{Expat}.
42568
42569The top-level structure of the document is shown below:
42570
42571@smallexample
42572<?xml version="1.0"?>
42573<!DOCTYPE btrace
42574 PUBLIC "+//IDN gnu.org//DTD GDB Branch Trace V1.0//EN"
42575 "http://sourceware.org/gdb/gdb-btrace.dtd">
42576<btrace>
42577 block...
42578</btrace>
42579@end smallexample
42580
42581@itemize
42582
42583@item
42584A block of sequentially executed instructions starting at @var{begin}
42585and ending at @var{end}:
42586
42587@smallexample
42588<block begin="@var{begin}" end="@var{end}"/>
42589@end smallexample
42590
42591@end itemize
42592
42593The formal DTD for the branch trace format is given below:
42594
42595@smallexample
42596<!ELEMENT btrace (block)* >
42597<!ATTLIST btrace version CDATA #FIXED "1.0">
42598
42599<!ELEMENT block EMPTY>
42600<!ATTLIST block begin CDATA #REQUIRED
42601 end CDATA #REQUIRED>
42602@end smallexample
42603
42604@include agentexpr.texi
42605
42606@node Target Descriptions
42607@appendix Target Descriptions
42608@cindex target descriptions
42609
42610One of the challenges of using @value{GDBN} to debug embedded systems
42611is that there are so many minor variants of each processor
42612architecture in use. It is common practice for vendors to start with
42613a standard processor core --- ARM, PowerPC, or @acronym{MIPS}, for example ---
42614and then make changes to adapt it to a particular market niche. Some
42615architectures have hundreds of variants, available from dozens of
42616vendors. This leads to a number of problems:
42617
42618@itemize @bullet
42619@item
42620With so many different customized processors, it is difficult for
42621the @value{GDBN} maintainers to keep up with the changes.
42622@item
42623Since individual variants may have short lifetimes or limited
42624audiences, it may not be worthwhile to carry information about every
42625variant in the @value{GDBN} source tree.
42626@item
42627When @value{GDBN} does support the architecture of the embedded system
42628at hand, the task of finding the correct architecture name to give the
42629@command{set architecture} command can be error-prone.
42630@end itemize
42631
42632To address these problems, the @value{GDBN} remote protocol allows a
42633target system to not only identify itself to @value{GDBN}, but to
42634actually describe its own features. This lets @value{GDBN} support
42635processor variants it has never seen before --- to the extent that the
42636descriptions are accurate, and that @value{GDBN} understands them.
42637
42638@value{GDBN} must be linked with the Expat library to support XML
42639target descriptions. @xref{Expat}.
42640
42641@menu
42642* Retrieving Descriptions:: How descriptions are fetched from a target.
42643* Target Description Format:: The contents of a target description.
42644* Predefined Target Types:: Standard types available for target
42645 descriptions.
42646* Standard Target Features:: Features @value{GDBN} knows about.
42647@end menu
42648
42649@node Retrieving Descriptions
42650@section Retrieving Descriptions
42651
42652Target descriptions can be read from the target automatically, or
42653specified by the user manually. The default behavior is to read the
42654description from the target. @value{GDBN} retrieves it via the remote
42655protocol using @samp{qXfer} requests (@pxref{General Query Packets,
42656qXfer}). The @var{annex} in the @samp{qXfer} packet will be
42657@samp{target.xml}. The contents of the @samp{target.xml} annex are an
42658XML document, of the form described in @ref{Target Description
42659Format}.
42660
42661Alternatively, you can specify a file to read for the target description.
42662If a file is set, the target will not be queried. The commands to
42663specify a file are:
42664
42665@table @code
42666@cindex set tdesc filename
42667@item set tdesc filename @var{path}
42668Read the target description from @var{path}.
42669
42670@cindex unset tdesc filename
42671@item unset tdesc filename
42672Do not read the XML target description from a file. @value{GDBN}
42673will use the description supplied by the current target.
42674
42675@cindex show tdesc filename
42676@item show tdesc filename
42677Show the filename to read for a target description, if any.
42678@end table
42679
42680
42681@node Target Description Format
42682@section Target Description Format
42683@cindex target descriptions, XML format
42684
42685A target description annex is an @uref{http://www.w3.org/XML/, XML}
42686document which complies with the Document Type Definition provided in
42687the @value{GDBN} sources in @file{gdb/features/gdb-target.dtd}. This
42688means you can use generally available tools like @command{xmllint} to
42689check that your feature descriptions are well-formed and valid.
42690However, to help people unfamiliar with XML write descriptions for
42691their targets, we also describe the grammar here.
42692
42693Target descriptions can identify the architecture of the remote target
42694and (for some architectures) provide information about custom register
42695sets. They can also identify the OS ABI of the remote target.
42696@value{GDBN} can use this information to autoconfigure for your
42697target, or to warn you if you connect to an unsupported target.
42698
42699Here is a simple target description:
42700
42701@smallexample
42702<target version="1.0">
42703 <architecture>i386:x86-64</architecture>
42704</target>
42705@end smallexample
42706
42707@noindent
42708This minimal description only says that the target uses
42709the x86-64 architecture.
42710
42711A target description has the following overall form, with [ ] marking
42712optional elements and @dots{} marking repeatable elements. The elements
42713are explained further below.
42714
42715@smallexample
42716<?xml version="1.0"?>
42717<!DOCTYPE target SYSTEM "gdb-target.dtd">
42718<target version="1.0">
42719 @r{[}@var{architecture}@r{]}
42720 @r{[}@var{osabi}@r{]}
42721 @r{[}@var{compatible}@r{]}
42722 @r{[}@var{feature}@dots{}@r{]}
42723</target>
42724@end smallexample
42725
42726@noindent
42727The description is generally insensitive to whitespace and line
42728breaks, under the usual common-sense rules. The XML version
42729declaration and document type declaration can generally be omitted
42730(@value{GDBN} does not require them), but specifying them may be
42731useful for XML validation tools. The @samp{version} attribute for
42732@samp{<target>} may also be omitted, but we recommend
42733including it; if future versions of @value{GDBN} use an incompatible
42734revision of @file{gdb-target.dtd}, they will detect and report
42735the version mismatch.
42736
42737@subsection Inclusion
42738@cindex target descriptions, inclusion
42739@cindex XInclude
42740@ifnotinfo
42741@cindex <xi:include>
42742@end ifnotinfo
42743
42744It can sometimes be valuable to split a target description up into
42745several different annexes, either for organizational purposes, or to
42746share files between different possible target descriptions. You can
42747divide a description into multiple files by replacing any element of
42748the target description with an inclusion directive of the form:
42749
42750@smallexample
42751<xi:include href="@var{document}"/>
42752@end smallexample
42753
42754@noindent
42755When @value{GDBN} encounters an element of this form, it will retrieve
42756the named XML @var{document}, and replace the inclusion directive with
42757the contents of that document. If the current description was read
42758using @samp{qXfer}, then so will be the included document;
42759@var{document} will be interpreted as the name of an annex. If the
42760current description was read from a file, @value{GDBN} will look for
42761@var{document} as a file in the same directory where it found the
42762original description.
42763
42764@subsection Architecture
42765@cindex <architecture>
42766
42767An @samp{<architecture>} element has this form:
42768
42769@smallexample
42770 <architecture>@var{arch}</architecture>
42771@end smallexample
42772
42773@var{arch} is one of the architectures from the set accepted by
42774@code{set architecture} (@pxref{Targets, ,Specifying a Debugging Target}).
42775
42776@subsection OS ABI
42777@cindex @code{<osabi>}
42778
42779This optional field was introduced in @value{GDBN} version 7.0.
42780Previous versions of @value{GDBN} ignore it.
42781
42782An @samp{<osabi>} element has this form:
42783
42784@smallexample
42785 <osabi>@var{abi-name}</osabi>
42786@end smallexample
42787
42788@var{abi-name} is an OS ABI name from the same selection accepted by
42789@w{@code{set osabi}} (@pxref{ABI, ,Configuring the Current ABI}).
42790
42791@subsection Compatible Architecture
42792@cindex @code{<compatible>}
42793
42794This optional field was introduced in @value{GDBN} version 7.0.
42795Previous versions of @value{GDBN} ignore it.
42796
42797A @samp{<compatible>} element has this form:
42798
42799@smallexample
42800 <compatible>@var{arch}</compatible>
42801@end smallexample
42802
42803@var{arch} is one of the architectures from the set accepted by
42804@code{set architecture} (@pxref{Targets, ,Specifying a Debugging Target}).
42805
42806A @samp{<compatible>} element is used to specify that the target
42807is able to run binaries in some other than the main target architecture
42808given by the @samp{<architecture>} element. For example, on the
42809Cell Broadband Engine, the main architecture is @code{powerpc:common}
42810or @code{powerpc:common64}, but the system is able to run binaries
42811in the @code{spu} architecture as well. The way to describe this
42812capability with @samp{<compatible>} is as follows:
42813
42814@smallexample
42815 <architecture>powerpc:common</architecture>
42816 <compatible>spu</compatible>
42817@end smallexample
42818
42819@subsection Features
42820@cindex <feature>
42821
42822Each @samp{<feature>} describes some logical portion of the target
42823system. Features are currently used to describe available CPU
42824registers and the types of their contents. A @samp{<feature>} element
42825has this form:
42826
42827@smallexample
42828<feature name="@var{name}">
42829 @r{[}@var{type}@dots{}@r{]}
42830 @var{reg}@dots{}
42831</feature>
42832@end smallexample
42833
42834@noindent
42835Each feature's name should be unique within the description. The name
42836of a feature does not matter unless @value{GDBN} has some special
42837knowledge of the contents of that feature; if it does, the feature
42838should have its standard name. @xref{Standard Target Features}.
42839
42840@subsection Types
42841
42842Any register's value is a collection of bits which @value{GDBN} must
42843interpret. The default interpretation is a two's complement integer,
42844but other types can be requested by name in the register description.
42845Some predefined types are provided by @value{GDBN} (@pxref{Predefined
42846Target Types}), and the description can define additional composite types.
42847
42848Each type element must have an @samp{id} attribute, which gives
42849a unique (within the containing @samp{<feature>}) name to the type.
42850Types must be defined before they are used.
42851
42852@cindex <vector>
42853Some targets offer vector registers, which can be treated as arrays
42854of scalar elements. These types are written as @samp{<vector>} elements,
42855specifying the array element type, @var{type}, and the number of elements,
42856@var{count}:
42857
42858@smallexample
42859<vector id="@var{id}" type="@var{type}" count="@var{count}"/>
42860@end smallexample
42861
42862@cindex <union>
42863If a register's value is usefully viewed in multiple ways, define it
42864with a union type containing the useful representations. The
42865@samp{<union>} element contains one or more @samp{<field>} elements,
42866each of which has a @var{name} and a @var{type}:
42867
42868@smallexample
42869<union id="@var{id}">
42870 <field name="@var{name}" type="@var{type}"/>
42871 @dots{}
42872</union>
42873@end smallexample
42874
42875@cindex <struct>
42876If a register's value is composed from several separate values, define
42877it with a structure type. There are two forms of the @samp{<struct>}
42878element; a @samp{<struct>} element must either contain only bitfields
42879or contain no bitfields. If the structure contains only bitfields,
42880its total size in bytes must be specified, each bitfield must have an
42881explicit start and end, and bitfields are automatically assigned an
42882integer type. The field's @var{start} should be less than or
42883equal to its @var{end}, and zero represents the least significant bit.
42884
42885@smallexample
42886<struct id="@var{id}" size="@var{size}">
42887 <field name="@var{name}" start="@var{start}" end="@var{end}"/>
42888 @dots{}
42889</struct>
42890@end smallexample
42891
42892If the structure contains no bitfields, then each field has an
42893explicit type, and no implicit padding is added.
42894
42895@smallexample
42896<struct id="@var{id}">
42897 <field name="@var{name}" type="@var{type}"/>
42898 @dots{}
42899</struct>
42900@end smallexample
42901
42902@cindex <flags>
42903If a register's value is a series of single-bit flags, define it with
42904a flags type. The @samp{<flags>} element has an explicit @var{size}
42905and contains one or more @samp{<field>} elements. Each field has a
42906@var{name}, a @var{start}, and an @var{end}. Only single-bit flags
42907are supported.
42908
42909@smallexample
42910<flags id="@var{id}" size="@var{size}">
42911 <field name="@var{name}" start="@var{start}" end="@var{end}"/>
42912 @dots{}
42913</flags>
42914@end smallexample
42915
42916@subsection Registers
42917@cindex <reg>
42918
42919Each register is represented as an element with this form:
42920
42921@smallexample
42922<reg name="@var{name}"
42923 bitsize="@var{size}"
42924 @r{[}regnum="@var{num}"@r{]}
42925 @r{[}save-restore="@var{save-restore}"@r{]}
42926 @r{[}type="@var{type}"@r{]}
42927 @r{[}group="@var{group}"@r{]}/>
42928@end smallexample
42929
42930@noindent
42931The components are as follows:
42932
42933@table @var
42934
42935@item name
42936The register's name; it must be unique within the target description.
42937
42938@item bitsize
42939The register's size, in bits.
42940
42941@item regnum
42942The register's number. If omitted, a register's number is one greater
42943than that of the previous register (either in the current feature or in
42944a preceding feature); the first register in the target description
42945defaults to zero. This register number is used to read or write
42946the register; e.g.@: it is used in the remote @code{p} and @code{P}
42947packets, and registers appear in the @code{g} and @code{G} packets
42948in order of increasing register number.
42949
42950@item save-restore
42951Whether the register should be preserved across inferior function
42952calls; this must be either @code{yes} or @code{no}. The default is
42953@code{yes}, which is appropriate for most registers except for
42954some system control registers; this is not related to the target's
42955ABI.
42956
42957@item type
42958The type of the register. @var{type} may be a predefined type, a type
42959defined in the current feature, or one of the special types @code{int}
42960and @code{float}. @code{int} is an integer type of the correct size
42961for @var{bitsize}, and @code{float} is a floating point type (in the
42962architecture's normal floating point format) of the correct size for
42963@var{bitsize}. The default is @code{int}.
42964
42965@item group
42966The register group to which this register belongs. @var{group} must
42967be either @code{general}, @code{float}, or @code{vector}. If no
42968@var{group} is specified, @value{GDBN} will not display the register
42969in @code{info registers}.
42970
42971@end table
42972
42973@node Predefined Target Types
42974@section Predefined Target Types
42975@cindex target descriptions, predefined types
42976
42977Type definitions in the self-description can build up composite types
42978from basic building blocks, but can not define fundamental types. Instead,
42979standard identifiers are provided by @value{GDBN} for the fundamental
42980types. The currently supported types are:
42981
42982@table @code
42983
42984@item int8
42985@itemx int16
42986@itemx int32
42987@itemx int64
42988@itemx int128
42989Signed integer types holding the specified number of bits.
42990
42991@item uint8
42992@itemx uint16
42993@itemx uint32
42994@itemx uint64
42995@itemx uint128
42996Unsigned integer types holding the specified number of bits.
42997
42998@item code_ptr
42999@itemx data_ptr
43000Pointers to unspecified code and data. The program counter and
43001any dedicated return address register may be marked as code
43002pointers; printing a code pointer converts it into a symbolic
43003address. The stack pointer and any dedicated address registers
43004may be marked as data pointers.
43005
43006@item ieee_single
43007Single precision IEEE floating point.
43008
43009@item ieee_double
43010Double precision IEEE floating point.
43011
43012@item arm_fpa_ext
43013The 12-byte extended precision format used by ARM FPA registers.
43014
43015@item i387_ext
43016The 10-byte extended precision format used by x87 registers.
43017
43018@item i386_eflags
4301932bit @sc{eflags} register used by x86.
43020
43021@item i386_mxcsr
4302232bit @sc{mxcsr} register used by x86.
43023
43024@end table
43025
43026@node Standard Target Features
43027@section Standard Target Features
43028@cindex target descriptions, standard features
43029
43030A target description must contain either no registers or all the
43031target's registers. If the description contains no registers, then
43032@value{GDBN} will assume a default register layout, selected based on
43033the architecture. If the description contains any registers, the
43034default layout will not be used; the standard registers must be
43035described in the target description, in such a way that @value{GDBN}
43036can recognize them.
43037
43038This is accomplished by giving specific names to feature elements
43039which contain standard registers. @value{GDBN} will look for features
43040with those names and verify that they contain the expected registers;
43041if any known feature is missing required registers, or if any required
43042feature is missing, @value{GDBN} will reject the target
43043description. You can add additional registers to any of the
43044standard features --- @value{GDBN} will display them just as if
43045they were added to an unrecognized feature.
43046
43047This section lists the known features and their expected contents.
43048Sample XML documents for these features are included in the
43049@value{GDBN} source tree, in the directory @file{gdb/features}.
43050
43051Names recognized by @value{GDBN} should include the name of the
43052company or organization which selected the name, and the overall
43053architecture to which the feature applies; so e.g.@: the feature
43054containing ARM core registers is named @samp{org.gnu.gdb.arm.core}.
43055
43056The names of registers are not case sensitive for the purpose
43057of recognizing standard features, but @value{GDBN} will only display
43058registers using the capitalization used in the description.
43059
43060@menu
43061* AArch64 Features::
43062* ARM Features::
43063* i386 Features::
43064* MIPS Features::
43065* M68K Features::
43066* Nios II Features::
43067* PowerPC Features::
43068* S/390 and System z Features::
43069* TIC6x Features::
43070@end menu
43071
43072
43073@node AArch64 Features
43074@subsection AArch64 Features
43075@cindex target descriptions, AArch64 features
43076
43077The @samp{org.gnu.gdb.aarch64.core} feature is required for AArch64
43078targets. It should contain registers @samp{x0} through @samp{x30},
43079@samp{sp}, @samp{pc}, and @samp{cpsr}.
43080
43081The @samp{org.gnu.gdb.aarch64.fpu} feature is optional. If present,
43082it should contain registers @samp{v0} through @samp{v31}, @samp{fpsr},
43083and @samp{fpcr}.
43084
43085@node ARM Features
43086@subsection ARM Features
43087@cindex target descriptions, ARM features
43088
43089The @samp{org.gnu.gdb.arm.core} feature is required for non-M-profile
43090ARM targets.
43091It should contain registers @samp{r0} through @samp{r13}, @samp{sp},
43092@samp{lr}, @samp{pc}, and @samp{cpsr}.
43093
43094For M-profile targets (e.g. Cortex-M3), the @samp{org.gnu.gdb.arm.core}
43095feature is replaced by @samp{org.gnu.gdb.arm.m-profile}. It should contain
43096registers @samp{r0} through @samp{r13}, @samp{sp}, @samp{lr}, @samp{pc},
43097and @samp{xpsr}.
43098
43099The @samp{org.gnu.gdb.arm.fpa} feature is optional. If present, it
43100should contain registers @samp{f0} through @samp{f7} and @samp{fps}.
43101
43102The @samp{org.gnu.gdb.xscale.iwmmxt} feature is optional. If present,
43103it should contain at least registers @samp{wR0} through @samp{wR15} and
43104@samp{wCGR0} through @samp{wCGR3}. The @samp{wCID}, @samp{wCon},
43105@samp{wCSSF}, and @samp{wCASF} registers are optional.
43106
43107The @samp{org.gnu.gdb.arm.vfp} feature is optional. If present, it
43108should contain at least registers @samp{d0} through @samp{d15}. If
43109they are present, @samp{d16} through @samp{d31} should also be included.
43110@value{GDBN} will synthesize the single-precision registers from
43111halves of the double-precision registers.
43112
43113The @samp{org.gnu.gdb.arm.neon} feature is optional. It does not
43114need to contain registers; it instructs @value{GDBN} to display the
43115VFP double-precision registers as vectors and to synthesize the
43116quad-precision registers from pairs of double-precision registers.
43117If this feature is present, @samp{org.gnu.gdb.arm.vfp} must also
43118be present and include 32 double-precision registers.
43119
43120@node i386 Features
43121@subsection i386 Features
43122@cindex target descriptions, i386 features
43123
43124The @samp{org.gnu.gdb.i386.core} feature is required for i386/amd64
43125targets. It should describe the following registers:
43126
43127@itemize @minus
43128@item
43129@samp{eax} through @samp{edi} plus @samp{eip} for i386
43130@item
43131@samp{rax} through @samp{r15} plus @samp{rip} for amd64
43132@item
43133@samp{eflags}, @samp{cs}, @samp{ss}, @samp{ds}, @samp{es},
43134@samp{fs}, @samp{gs}
43135@item
43136@samp{st0} through @samp{st7}
43137@item
43138@samp{fctrl}, @samp{fstat}, @samp{ftag}, @samp{fiseg}, @samp{fioff},
43139@samp{foseg}, @samp{fooff} and @samp{fop}
43140@end itemize
43141
43142The register sets may be different, depending on the target.
43143
43144The @samp{org.gnu.gdb.i386.sse} feature is optional. It should
43145describe registers:
43146
43147@itemize @minus
43148@item
43149@samp{xmm0} through @samp{xmm7} for i386
43150@item
43151@samp{xmm0} through @samp{xmm15} for amd64
43152@item
43153@samp{mxcsr}
43154@end itemize
43155
43156The @samp{org.gnu.gdb.i386.avx} feature is optional and requires the
43157@samp{org.gnu.gdb.i386.sse} feature. It should
43158describe the upper 128 bits of @sc{ymm} registers:
43159
43160@itemize @minus
43161@item
43162@samp{ymm0h} through @samp{ymm7h} for i386
43163@item
43164@samp{ymm0h} through @samp{ymm15h} for amd64
43165@end itemize
43166
43167The @samp{org.gnu.gdb.i386.linux} feature is optional. It should
43168describe a single register, @samp{orig_eax}.
43169
43170@node MIPS Features
43171@subsection @acronym{MIPS} Features
43172@cindex target descriptions, @acronym{MIPS} features
43173
43174The @samp{org.gnu.gdb.mips.cpu} feature is required for @acronym{MIPS} targets.
43175It should contain registers @samp{r0} through @samp{r31}, @samp{lo},
43176@samp{hi}, and @samp{pc}. They may be 32-bit or 64-bit depending
43177on the target.
43178
43179The @samp{org.gnu.gdb.mips.cp0} feature is also required. It should
43180contain at least the @samp{status}, @samp{badvaddr}, and @samp{cause}
43181registers. They may be 32-bit or 64-bit depending on the target.
43182
43183The @samp{org.gnu.gdb.mips.fpu} feature is currently required, though
43184it may be optional in a future version of @value{GDBN}. It should
43185contain registers @samp{f0} through @samp{f31}, @samp{fcsr}, and
43186@samp{fir}. They may be 32-bit or 64-bit depending on the target.
43187
43188The @samp{org.gnu.gdb.mips.dsp} feature is optional. It should
43189contain registers @samp{hi1} through @samp{hi3}, @samp{lo1} through
43190@samp{lo3}, and @samp{dspctl}. The @samp{dspctl} register should
43191be 32-bit and the rest may be 32-bit or 64-bit depending on the target.
43192
43193The @samp{org.gnu.gdb.mips.linux} feature is optional. It should
43194contain a single register, @samp{restart}, which is used by the
43195Linux kernel to control restartable syscalls.
43196
43197@node M68K Features
43198@subsection M68K Features
43199@cindex target descriptions, M68K features
43200
43201@table @code
43202@item @samp{org.gnu.gdb.m68k.core}
43203@itemx @samp{org.gnu.gdb.coldfire.core}
43204@itemx @samp{org.gnu.gdb.fido.core}
43205One of those features must be always present.
43206The feature that is present determines which flavor of m68k is
43207used. The feature that is present should contain registers
43208@samp{d0} through @samp{d7}, @samp{a0} through @samp{a5}, @samp{fp},
43209@samp{sp}, @samp{ps} and @samp{pc}.
43210
43211@item @samp{org.gnu.gdb.coldfire.fp}
43212This feature is optional. If present, it should contain registers
43213@samp{fp0} through @samp{fp7}, @samp{fpcontrol}, @samp{fpstatus} and
43214@samp{fpiaddr}.
43215@end table
43216
43217@node Nios II Features
43218@subsection Nios II Features
43219@cindex target descriptions, Nios II features
43220
43221The @samp{org.gnu.gdb.nios2.cpu} feature is required for Nios II
43222targets. It should contain the 32 core registers (@samp{zero},
43223@samp{at}, @samp{r2} through @samp{r23}, @samp{et} through @samp{ra}),
43224@samp{pc}, and the 16 control registers (@samp{status} through
43225@samp{mpuacc}).
43226
43227@node PowerPC Features
43228@subsection PowerPC Features
43229@cindex target descriptions, PowerPC features
43230
43231The @samp{org.gnu.gdb.power.core} feature is required for PowerPC
43232targets. It should contain registers @samp{r0} through @samp{r31},
43233@samp{pc}, @samp{msr}, @samp{cr}, @samp{lr}, @samp{ctr}, and
43234@samp{xer}. They may be 32-bit or 64-bit depending on the target.
43235
43236The @samp{org.gnu.gdb.power.fpu} feature is optional. It should
43237contain registers @samp{f0} through @samp{f31} and @samp{fpscr}.
43238
43239The @samp{org.gnu.gdb.power.altivec} feature is optional. It should
43240contain registers @samp{vr0} through @samp{vr31}, @samp{vscr},
43241and @samp{vrsave}.
43242
43243The @samp{org.gnu.gdb.power.vsx} feature is optional. It should
43244contain registers @samp{vs0h} through @samp{vs31h}. @value{GDBN}
43245will combine these registers with the floating point registers
43246(@samp{f0} through @samp{f31}) and the altivec registers (@samp{vr0}
43247through @samp{vr31}) to present the 128-bit wide registers @samp{vs0}
43248through @samp{vs63}, the set of vector registers for POWER7.
43249
43250The @samp{org.gnu.gdb.power.spe} feature is optional. It should
43251contain registers @samp{ev0h} through @samp{ev31h}, @samp{acc}, and
43252@samp{spefscr}. SPE targets should provide 32-bit registers in
43253@samp{org.gnu.gdb.power.core} and provide the upper halves in
43254@samp{ev0h} through @samp{ev31h}. @value{GDBN} will combine
43255these to present registers @samp{ev0} through @samp{ev31} to the
43256user.
43257
43258@node S/390 and System z Features
43259@subsection S/390 and System z Features
43260@cindex target descriptions, S/390 features
43261@cindex target descriptions, System z features
43262
43263The @samp{org.gnu.gdb.s390.core} feature is required for S/390 and
43264System z targets. It should contain the PSW and the 16 general
43265registers. In particular, System z targets should provide the 64-bit
43266registers @samp{pswm}, @samp{pswa}, and @samp{r0} through @samp{r15}.
43267S/390 targets should provide the 32-bit versions of these registers.
43268A System z target that runs in 31-bit addressing mode should provide
4326932-bit versions of @samp{pswm} and @samp{pswa}, as well as the general
43270register's upper halves @samp{r0h} through @samp{r15h}, and their
43271lower halves @samp{r0l} through @samp{r15l}.
43272
43273The @samp{org.gnu.gdb.s390.fpr} feature is required. It should
43274contain the 64-bit registers @samp{f0} through @samp{f15}, and
43275@samp{fpc}.
43276
43277The @samp{org.gnu.gdb.s390.acr} feature is required. It should
43278contain the 32-bit registers @samp{acr0} through @samp{acr15}.
43279
43280The @samp{org.gnu.gdb.s390.linux} feature is optional. It should
43281contain the register @samp{orig_r2}, which is 64-bit wide on System z
43282targets and 32-bit otherwise. In addition, the feature may contain
43283the @samp{last_break} register, whose width depends on the addressing
43284mode, as well as the @samp{system_call} register, which is always
4328532-bit wide.
43286
43287The @samp{org.gnu.gdb.s390.tdb} feature is optional. It should
43288contain the 64-bit registers @samp{tdb0}, @samp{tac}, @samp{tct},
43289@samp{atia}, and @samp{tr0} through @samp{tr15}.
43290
43291@node TIC6x Features
43292@subsection TMS320C6x Features
43293@cindex target descriptions, TIC6x features
43294@cindex target descriptions, TMS320C6x features
43295The @samp{org.gnu.gdb.tic6x.core} feature is required for TMS320C6x
43296targets. It should contain registers @samp{A0} through @samp{A15},
43297registers @samp{B0} through @samp{B15}, @samp{CSR} and @samp{PC}.
43298
43299The @samp{org.gnu.gdb.tic6x.gp} feature is optional. It should
43300contain registers @samp{A16} through @samp{A31} and @samp{B16}
43301through @samp{B31}.
43302
43303The @samp{org.gnu.gdb.tic6x.c6xp} feature is optional. It should
43304contain registers @samp{TSR}, @samp{ILC} and @samp{RILC}.
43305
43306@node Operating System Information
43307@appendix Operating System Information
43308@cindex operating system information
43309
43310@menu
43311* Process list::
43312@end menu
43313
43314Users of @value{GDBN} often wish to obtain information about the state of
43315the operating system running on the target---for example the list of
43316processes, or the list of open files. This section describes the
43317mechanism that makes it possible. This mechanism is similar to the
43318target features mechanism (@pxref{Target Descriptions}), but focuses
43319on a different aspect of target.
43320
43321Operating system information is retrived from the target via the
43322remote protocol, using @samp{qXfer} requests (@pxref{qXfer osdata
43323read}). The object name in the request should be @samp{osdata}, and
43324the @var{annex} identifies the data to be fetched.
43325
43326@node Process list
43327@appendixsection Process list
43328@cindex operating system information, process list
43329
43330When requesting the process list, the @var{annex} field in the
43331@samp{qXfer} request should be @samp{processes}. The returned data is
43332an XML document. The formal syntax of this document is defined in
43333@file{gdb/features/osdata.dtd}.
43334
43335An example document is:
43336
43337@smallexample
43338<?xml version="1.0"?>
43339<!DOCTYPE target SYSTEM "osdata.dtd">
43340<osdata type="processes">
43341 <item>
43342 <column name="pid">1</column>
43343 <column name="user">root</column>
43344 <column name="command">/sbin/init</column>
43345 <column name="cores">1,2,3</column>
43346 </item>
43347</osdata>
43348@end smallexample
43349
43350Each item should include a column whose name is @samp{pid}. The value
43351of that column should identify the process on the target. The
43352@samp{user} and @samp{command} columns are optional, and will be
43353displayed by @value{GDBN}. The @samp{cores} column, if present,
43354should contain a comma-separated list of cores that this process
43355is running on. Target may provide additional columns,
43356which @value{GDBN} currently ignores.
43357
43358@node Trace File Format
43359@appendix Trace File Format
43360@cindex trace file format
43361
43362The trace file comes in three parts: a header, a textual description
43363section, and a trace frame section with binary data.
43364
43365The header has the form @code{\x7fTRACE0\n}. The first byte is
43366@code{0x7f} so as to indicate that the file contains binary data,
43367while the @code{0} is a version number that may have different values
43368in the future.
43369
43370The description section consists of multiple lines of @sc{ascii} text
43371separated by newline characters (@code{0xa}). The lines may include a
43372variety of optional descriptive or context-setting information, such
43373as tracepoint definitions or register set size. @value{GDBN} will
43374ignore any line that it does not recognize. An empty line marks the end
43375of this section.
43376
43377@c FIXME add some specific types of data
43378
43379The trace frame section consists of a number of consecutive frames.
43380Each frame begins with a two-byte tracepoint number, followed by a
43381four-byte size giving the amount of data in the frame. The data in
43382the frame consists of a number of blocks, each introduced by a
43383character indicating its type (at least register, memory, and trace
43384state variable). The data in this section is raw binary, not a
43385hexadecimal or other encoding; its endianness matches the target's
43386endianness.
43387
43388@c FIXME bi-arch may require endianness/arch info in description section
43389
43390@table @code
43391@item R @var{bytes}
43392Register block. The number and ordering of bytes matches that of a
43393@code{g} packet in the remote protocol. Note that these are the
43394actual bytes, in target order and @value{GDBN} register order, not a
43395hexadecimal encoding.
43396
43397@item M @var{address} @var{length} @var{bytes}...
43398Memory block. This is a contiguous block of memory, at the 8-byte
43399address @var{address}, with a 2-byte length @var{length}, followed by
43400@var{length} bytes.
43401
43402@item V @var{number} @var{value}
43403Trace state variable block. This records the 8-byte signed value
43404@var{value} of trace state variable numbered @var{number}.
43405
43406@end table
43407
43408Future enhancements of the trace file format may include additional types
43409of blocks.
43410
43411@node Index Section Format
43412@appendix @code{.gdb_index} section format
43413@cindex .gdb_index section format
43414@cindex index section format
43415
43416This section documents the index section that is created by @code{save
43417gdb-index} (@pxref{Index Files}). The index section is
43418DWARF-specific; some knowledge of DWARF is assumed in this
43419description.
43420
43421The mapped index file format is designed to be directly
43422@code{mmap}able on any architecture. In most cases, a datum is
43423represented using a little-endian 32-bit integer value, called an
43424@code{offset_type}. Big endian machines must byte-swap the values
43425before using them. Exceptions to this rule are noted. The data is
43426laid out such that alignment is always respected.
43427
43428A mapped index consists of several areas, laid out in order.
43429
43430@enumerate
43431@item
43432The file header. This is a sequence of values, of @code{offset_type}
43433unless otherwise noted:
43434
43435@enumerate
43436@item
43437The version number, currently 8. Versions 1, 2 and 3 are obsolete.
43438Version 4 uses a different hashing function from versions 5 and 6.
43439Version 6 includes symbols for inlined functions, whereas versions 4
43440and 5 do not. Version 7 adds attributes to the CU indices in the
43441symbol table. Version 8 specifies that symbols from DWARF type units
43442(@samp{DW_TAG_type_unit}) refer to the type unit's symbol table and not the
43443compilation unit (@samp{DW_TAG_comp_unit}) using the type.
43444
43445@value{GDBN} will only read version 4, 5, or 6 indices
43446by specifying @code{set use-deprecated-index-sections on}.
43447GDB has a workaround for potentially broken version 7 indices so it is
43448currently not flagged as deprecated.
43449
43450@item
43451The offset, from the start of the file, of the CU list.
43452
43453@item
43454The offset, from the start of the file, of the types CU list. Note
43455that this area can be empty, in which case this offset will be equal
43456to the next offset.
43457
43458@item
43459The offset, from the start of the file, of the address area.
43460
43461@item
43462The offset, from the start of the file, of the symbol table.
43463
43464@item
43465The offset, from the start of the file, of the constant pool.
43466@end enumerate
43467
43468@item
43469The CU list. This is a sequence of pairs of 64-bit little-endian
43470values, sorted by the CU offset. The first element in each pair is
43471the offset of a CU in the @code{.debug_info} section. The second
43472element in each pair is the length of that CU. References to a CU
43473elsewhere in the map are done using a CU index, which is just the
434740-based index into this table. Note that if there are type CUs, then
43475conceptually CUs and type CUs form a single list for the purposes of
43476CU indices.
43477
43478@item
43479The types CU list. This is a sequence of triplets of 64-bit
43480little-endian values. In a triplet, the first value is the CU offset,
43481the second value is the type offset in the CU, and the third value is
43482the type signature. The types CU list is not sorted.
43483
43484@item
43485The address area. The address area consists of a sequence of address
43486entries. Each address entry has three elements:
43487
43488@enumerate
43489@item
43490The low address. This is a 64-bit little-endian value.
43491
43492@item
43493The high address. This is a 64-bit little-endian value. Like
43494@code{DW_AT_high_pc}, the value is one byte beyond the end.
43495
43496@item
43497The CU index. This is an @code{offset_type} value.
43498@end enumerate
43499
43500@item
43501The symbol table. This is an open-addressed hash table. The size of
43502the hash table is always a power of 2.
43503
43504Each slot in the hash table consists of a pair of @code{offset_type}
43505values. The first value is the offset of the symbol's name in the
43506constant pool. The second value is the offset of the CU vector in the
43507constant pool.
43508
43509If both values are 0, then this slot in the hash table is empty. This
43510is ok because while 0 is a valid constant pool index, it cannot be a
43511valid index for both a string and a CU vector.
43512
43513The hash value for a table entry is computed by applying an
43514iterative hash function to the symbol's name. Starting with an
43515initial value of @code{r = 0}, each (unsigned) character @samp{c} in
43516the string is incorporated into the hash using the formula depending on the
43517index version:
43518
43519@table @asis
43520@item Version 4
43521The formula is @code{r = r * 67 + c - 113}.
43522
43523@item Versions 5 to 7
43524The formula is @code{r = r * 67 + tolower (c) - 113}.
43525@end table
43526
43527The terminating @samp{\0} is not incorporated into the hash.
43528
43529The step size used in the hash table is computed via
43530@code{((hash * 17) & (size - 1)) | 1}, where @samp{hash} is the hash
43531value, and @samp{size} is the size of the hash table. The step size
43532is used to find the next candidate slot when handling a hash
43533collision.
43534
43535The names of C@t{++} symbols in the hash table are canonicalized. We
43536don't currently have a simple description of the canonicalization
43537algorithm; if you intend to create new index sections, you must read
43538the code.
43539
43540@item
43541The constant pool. This is simply a bunch of bytes. It is organized
43542so that alignment is correct: CU vectors are stored first, followed by
43543strings.
43544
43545A CU vector in the constant pool is a sequence of @code{offset_type}
43546values. The first value is the number of CU indices in the vector.
43547Each subsequent value is the index and symbol attributes of a CU in
43548the CU list. This element in the hash table is used to indicate which
43549CUs define the symbol and how the symbol is used.
43550See below for the format of each CU index+attributes entry.
43551
43552A string in the constant pool is zero-terminated.
43553@end enumerate
43554
43555Attributes were added to CU index values in @code{.gdb_index} version 7.
43556If a symbol has multiple uses within a CU then there is one
43557CU index+attributes value for each use.
43558
43559The format of each CU index+attributes entry is as follows
43560(bit 0 = LSB):
43561
43562@table @asis
43563
43564@item Bits 0-23
43565This is the index of the CU in the CU list.
43566@item Bits 24-27
43567These bits are reserved for future purposes and must be zero.
43568@item Bits 28-30
43569The kind of the symbol in the CU.
43570
43571@table @asis
43572@item 0
43573This value is reserved and should not be used.
43574By reserving zero the full @code{offset_type} value is backwards compatible
43575with previous versions of the index.
43576@item 1
43577The symbol is a type.
43578@item 2
43579The symbol is a variable or an enum value.
43580@item 3
43581The symbol is a function.
43582@item 4
43583Any other kind of symbol.
43584@item 5,6,7
43585These values are reserved.
43586@end table
43587
43588@item Bit 31
43589This bit is zero if the value is global and one if it is static.
43590
43591The determination of whether a symbol is global or static is complicated.
43592The authorative reference is the file @file{dwarf2read.c} in
43593@value{GDBN} sources.
43594
43595@end table
43596
43597This pseudo-code describes the computation of a symbol's kind and
43598global/static attributes in the index.
43599
43600@smallexample
43601is_external = get_attribute (die, DW_AT_external);
43602language = get_attribute (cu_die, DW_AT_language);
43603switch (die->tag)
43604 @{
43605 case DW_TAG_typedef:
43606 case DW_TAG_base_type:
43607 case DW_TAG_subrange_type:
43608 kind = TYPE;
43609 is_static = 1;
43610 break;
43611 case DW_TAG_enumerator:
43612 kind = VARIABLE;
43613 is_static = (language != CPLUS && language != JAVA);
43614 break;
43615 case DW_TAG_subprogram:
43616 kind = FUNCTION;
43617 is_static = ! (is_external || language == ADA);
43618 break;
43619 case DW_TAG_constant:
43620 kind = VARIABLE;
43621 is_static = ! is_external;
43622 break;
43623 case DW_TAG_variable:
43624 kind = VARIABLE;
43625 is_static = ! is_external;
43626 break;
43627 case DW_TAG_namespace:
43628 kind = TYPE;
43629 is_static = 0;
43630 break;
43631 case DW_TAG_class_type:
43632 case DW_TAG_interface_type:
43633 case DW_TAG_structure_type:
43634 case DW_TAG_union_type:
43635 case DW_TAG_enumeration_type:
43636 kind = TYPE;
43637 is_static = (language != CPLUS && language != JAVA);
43638 break;
43639 default:
43640 assert (0);
43641 @}
43642@end smallexample
43643
43644@node Man Pages
43645@appendix Manual pages
43646@cindex Man pages
43647
43648@menu
43649* gdb man:: The GNU Debugger man page
43650* gdbserver man:: Remote Server for the GNU Debugger man page
43651* gcore man:: Generate a core file of a running program
43652* gdbinit man:: gdbinit scripts
43653@end menu
43654
43655@node gdb man
43656@heading gdb man
43657
43658@c man title gdb The GNU Debugger
43659
43660@c man begin SYNOPSIS gdb
43661gdb [@option{-help}] [@option{-nh}] [@option{-nx}] [@option{-q}]
43662[@option{-batch}] [@option{-cd=}@var{dir}] [@option{-f}]
43663[@option{-b}@w{ }@var{bps}]
43664 [@option{-tty=}@var{dev}] [@option{-s} @var{symfile}]
43665[@option{-e}@w{ }@var{prog}] [@option{-se}@w{ }@var{prog}]
43666[@option{-c}@w{ }@var{core}] [@option{-p}@w{ }@var{procID}]
43667 [@option{-x}@w{ }@var{cmds}] [@option{-d}@w{ }@var{dir}]
43668[@var{prog}|@var{prog} @var{procID}|@var{prog} @var{core}]
43669@c man end
43670
43671@c man begin DESCRIPTION gdb
43672The purpose of a debugger such as @value{GDBN} is to allow you to see what is
43673going on ``inside'' another program while it executes -- or what another
43674program was doing at the moment it crashed.
43675
43676@value{GDBN} can do four main kinds of things (plus other things in support of
43677these) to help you catch bugs in the act:
43678
43679@itemize @bullet
43680@item
43681Start your program, specifying anything that might affect its behavior.
43682
43683@item
43684Make your program stop on specified conditions.
43685
43686@item
43687Examine what has happened, when your program has stopped.
43688
43689@item
43690Change things in your program, so you can experiment with correcting the
43691effects of one bug and go on to learn about another.
43692@end itemize
43693
43694You can use @value{GDBN} to debug programs written in C, C@t{++}, Fortran and
43695Modula-2.
43696
43697@value{GDBN} is invoked with the shell command @code{gdb}. Once started, it reads
43698commands from the terminal until you tell it to exit with the @value{GDBN}
43699command @code{quit}. You can get online help from @value{GDBN} itself
43700by using the command @code{help}.
43701
43702You can run @code{gdb} with no arguments or options; but the most
43703usual way to start @value{GDBN} is with one argument or two, specifying an
43704executable program as the argument:
43705
43706@smallexample
43707gdb program
43708@end smallexample
43709
43710You can also start with both an executable program and a core file specified:
43711
43712@smallexample
43713gdb program core
43714@end smallexample
43715
43716You can, instead, specify a process ID as a second argument, if you want
43717to debug a running process:
43718
43719@smallexample
43720gdb program 1234
43721gdb -p 1234
43722@end smallexample
43723
43724@noindent
43725would attach @value{GDBN} to process @code{1234} (unless you also have a file
43726named @file{1234}; @value{GDBN} does check for a core file first).
43727With option @option{-p} you can omit the @var{program} filename.
43728
43729Here are some of the most frequently needed @value{GDBN} commands:
43730
43731@c pod2man highlights the right hand side of the @item lines.
43732@table @env
43733@item break [@var{file}:]@var{functiop}
43734Set a breakpoint at @var{function} (in @var{file}).
43735
43736@item run [@var{arglist}]
43737Start your program (with @var{arglist}, if specified).
43738
43739@item bt
43740Backtrace: display the program stack.
43741
43742@item print @var{expr}
43743Display the value of an expression.
43744
43745@item c
43746Continue running your program (after stopping, e.g. at a breakpoint).
43747
43748@item next
43749Execute next program line (after stopping); step @emph{over} any
43750function calls in the line.
43751
43752@item edit [@var{file}:]@var{function}
43753look at the program line where it is presently stopped.
43754
43755@item list [@var{file}:]@var{function}
43756type the text of the program in the vicinity of where it is presently stopped.
43757
43758@item step
43759Execute next program line (after stopping); step @emph{into} any
43760function calls in the line.
43761
43762@item help [@var{name}]
43763Show information about @value{GDBN} command @var{name}, or general information
43764about using @value{GDBN}.
43765
43766@item quit
43767Exit from @value{GDBN}.
43768@end table
43769
43770@ifset man
43771For full details on @value{GDBN},
43772see @cite{Using GDB: A Guide to the GNU Source-Level Debugger},
43773by Richard M. Stallman and Roland H. Pesch. The same text is available online
43774as the @code{gdb} entry in the @code{info} program.
43775@end ifset
43776@c man end
43777
43778@c man begin OPTIONS gdb
43779Any arguments other than options specify an executable
43780file and core file (or process ID); that is, the first argument
43781encountered with no
43782associated option flag is equivalent to a @option{-se} option, and the second,
43783if any, is equivalent to a @option{-c} option if it's the name of a file.
43784Many options have
43785both long and short forms; both are shown here. The long forms are also
43786recognized if you truncate them, so long as enough of the option is
43787present to be unambiguous. (If you prefer, you can flag option
43788arguments with @option{+} rather than @option{-}, though we illustrate the
43789more usual convention.)
43790
43791All the options and command line arguments you give are processed
43792in sequential order. The order makes a difference when the @option{-x}
43793option is used.
43794
43795@table @env
43796@item -help
43797@itemx -h
43798List all options, with brief explanations.
43799
43800@item -symbols=@var{file}
43801@itemx -s @var{file}
43802Read symbol table from file @var{file}.
43803
43804@item -write
43805Enable writing into executable and core files.
43806
43807@item -exec=@var{file}
43808@itemx -e @var{file}
43809Use file @var{file} as the executable file to execute when
43810appropriate, and for examining pure data in conjunction with a core
43811dump.
43812
43813@item -se=@var{file}
43814Read symbol table from file @var{file} and use it as the executable
43815file.
43816
43817@item -core=@var{file}
43818@itemx -c @var{file}
43819Use file @var{file} as a core dump to examine.
43820
43821@item -command=@var{file}
43822@itemx -x @var{file}
43823Execute @value{GDBN} commands from file @var{file}.
43824
43825@item -ex @var{command}
43826Execute given @value{GDBN} @var{command}.
43827
43828@item -directory=@var{directory}
43829@itemx -d @var{directory}
43830Add @var{directory} to the path to search for source files.
43831
43832@item -nh
43833Do not execute commands from @file{~/.gdbinit}.
43834
43835@item -nx
43836@itemx -n
43837Do not execute commands from any @file{.gdbinit} initialization files.
43838
43839@item -quiet
43840@itemx -q
43841``Quiet''. Do not print the introductory and copyright messages. These
43842messages are also suppressed in batch mode.
43843
43844@item -batch
43845Run in batch mode. Exit with status @code{0} after processing all the command
43846files specified with @option{-x} (and @file{.gdbinit}, if not inhibited).
43847Exit with nonzero status if an error occurs in executing the @value{GDBN}
43848commands in the command files.
43849
43850Batch mode may be useful for running @value{GDBN} as a filter, for example to
43851download and run a program on another computer; in order to make this
43852more useful, the message
43853
43854@smallexample
43855Program exited normally.
43856@end smallexample
43857
43858@noindent
43859(which is ordinarily issued whenever a program running under @value{GDBN} control
43860terminates) is not issued when running in batch mode.
43861
43862@item -cd=@var{directory}
43863Run @value{GDBN} using @var{directory} as its working directory,
43864instead of the current directory.
43865
43866@item -fullname
43867@itemx -f
43868Emacs sets this option when it runs @value{GDBN} as a subprocess. It tells
43869@value{GDBN} to output the full file name and line number in a standard,
43870recognizable fashion each time a stack frame is displayed (which
43871includes each time the program stops). This recognizable format looks
43872like two @samp{\032} characters, followed by the file name, line number
43873and character position separated by colons, and a newline. The
43874Emacs-to-@value{GDBN} interface program uses the two @samp{\032}
43875characters as a signal to display the source code for the frame.
43876
43877@item -b @var{bps}
43878Set the line speed (baud rate or bits per second) of any serial
43879interface used by @value{GDBN} for remote debugging.
43880
43881@item -tty=@var{device}
43882Run using @var{device} for your program's standard input and output.
43883@end table
43884@c man end
43885
43886@c man begin SEEALSO gdb
43887@ifset man
43888The full documentation for @value{GDBN} is maintained as a Texinfo manual.
43889If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
43890documentation are properly installed at your site, the command
43891
43892@smallexample
43893info gdb
43894@end smallexample
43895
43896@noindent
43897should give you access to the complete manual.
43898
43899@cite{Using GDB: A Guide to the GNU Source-Level Debugger},
43900Richard M. Stallman and Roland H. Pesch, July 1991.
43901@end ifset
43902@c man end
43903
43904@node gdbserver man
43905@heading gdbserver man
43906
43907@c man title gdbserver Remote Server for the GNU Debugger
43908@format
43909@c man begin SYNOPSIS gdbserver
43910gdbserver @var{comm} @var{prog} [@var{args}@dots{}]
43911
43912gdbserver --attach @var{comm} @var{pid}
43913
43914gdbserver --multi @var{comm}
43915@c man end
43916@end format
43917
43918@c man begin DESCRIPTION gdbserver
43919@command{gdbserver} is a program that allows you to run @value{GDBN} on a different machine
43920than the one which is running the program being debugged.
43921
43922@ifclear man
43923@subheading Usage (server (target) side)
43924@end ifclear
43925@ifset man
43926Usage (server (target) side):
43927@end ifset
43928
43929First, you need to have a copy of the program you want to debug put onto
43930the target system. The program can be stripped to save space if needed, as
43931@command{gdbserver} doesn't care about symbols. All symbol handling is taken care of by
43932the @value{GDBN} running on the host system.
43933
43934To use the server, you log on to the target system, and run the @command{gdbserver}
43935program. You must tell it (a) how to communicate with @value{GDBN}, (b) the name of
43936your program, and (c) its arguments. The general syntax is:
43937
43938@smallexample
43939target> gdbserver @var{comm} @var{program} [@var{args} ...]
43940@end smallexample
43941
43942For example, using a serial port, you might say:
43943
43944@smallexample
43945@ifset man
43946@c @file would wrap it as F</dev/com1>.
43947target> gdbserver /dev/com1 emacs foo.txt
43948@end ifset
43949@ifclear man
43950target> gdbserver @file{/dev/com1} emacs foo.txt
43951@end ifclear
43952@end smallexample
43953
43954This tells @command{gdbserver} to debug emacs with an argument of foo.txt, and
43955to communicate with @value{GDBN} via @file{/dev/com1}. @command{gdbserver} now
43956waits patiently for the host @value{GDBN} to communicate with it.
43957
43958To use a TCP connection, you could say:
43959
43960@smallexample
43961target> gdbserver host:2345 emacs foo.txt
43962@end smallexample
43963
43964This says pretty much the same thing as the last example, except that we are
43965going to communicate with the @code{host} @value{GDBN} via TCP. The @code{host:2345} argument means
43966that we are expecting to see a TCP connection from @code{host} to local TCP port
439672345. (Currently, the @code{host} part is ignored.) You can choose any number you
43968want for the port number as long as it does not conflict with any existing TCP
43969ports on the target system. This same port number must be used in the host
43970@value{GDBN}s @code{target remote} command, which will be described shortly. Note that if
43971you chose a port number that conflicts with another service, @command{gdbserver} will
43972print an error message and exit.
43973
43974@command{gdbserver} can also attach to running programs.
43975This is accomplished via the @option{--attach} argument. The syntax is:
43976
43977@smallexample
43978target> gdbserver --attach @var{comm} @var{pid}
43979@end smallexample
43980
43981@var{pid} is the process ID of a currently running process. It isn't
43982necessary to point @command{gdbserver} at a binary for the running process.
43983
43984To start @code{gdbserver} without supplying an initial command to run
43985or process ID to attach, use the @option{--multi} command line option.
43986In such case you should connect using @kbd{target extended-remote} to start
43987the program you want to debug.
43988
43989@smallexample
43990target> gdbserver --multi @var{comm}
43991@end smallexample
43992
43993@ifclear man
43994@subheading Usage (host side)
43995@end ifclear
43996@ifset man
43997Usage (host side):
43998@end ifset
43999
44000You need an unstripped copy of the target program on your host system, since
44001@value{GDBN} needs to examine it's symbol tables and such. Start up @value{GDBN} as you normally
44002would, with the target program as the first argument. (You may need to use the
44003@option{--baud} option if the serial line is running at anything except 9600 baud.)
44004That is @code{gdb TARGET-PROG}, or @code{gdb --baud BAUD TARGET-PROG}. After that, the only
44005new command you need to know about is @code{target remote}
44006(or @code{target extended-remote}). Its argument is either
44007a device name (usually a serial device, like @file{/dev/ttyb}), or a @code{HOST:PORT}
44008descriptor. For example:
44009
44010@smallexample
44011@ifset man
44012@c @file would wrap it as F</dev/ttyb>.
44013(gdb) target remote /dev/ttyb
44014@end ifset
44015@ifclear man
44016(gdb) target remote @file{/dev/ttyb}
44017@end ifclear
44018@end smallexample
44019
44020@noindent
44021communicates with the server via serial line @file{/dev/ttyb}, and:
44022
44023@smallexample
44024(gdb) target remote the-target:2345
44025@end smallexample
44026
44027@noindent
44028communicates via a TCP connection to port 2345 on host `the-target', where
44029you previously started up @command{gdbserver} with the same port number. Note that for
44030TCP connections, you must start up @command{gdbserver} prior to using the `target remote'
44031command, otherwise you may get an error that looks something like
44032`Connection refused'.
44033
44034@command{gdbserver} can also debug multiple inferiors at once,
44035described in
44036@ifset man
44037the @value{GDBN} manual in node @code{Inferiors and Programs}
44038-- shell command @code{info -f gdb -n 'Inferiors and Programs'}.
44039@end ifset
44040@ifclear man
44041@ref{Inferiors and Programs}.
44042@end ifclear
44043In such case use the @code{extended-remote} @value{GDBN} command variant:
44044
44045@smallexample
44046(gdb) target extended-remote the-target:2345
44047@end smallexample
44048
44049The @command{gdbserver} option @option{--multi} may or may not be used in such
44050case.
44051@c man end
44052
44053@c man begin OPTIONS gdbserver
44054There are three different modes for invoking @command{gdbserver}:
44055
44056@itemize @bullet
44057
44058@item
44059Debug a specific program specified by its program name:
44060
44061@smallexample
44062gdbserver @var{comm} @var{prog} [@var{args}@dots{}]
44063@end smallexample
44064
44065The @var{comm} parameter specifies how should the server communicate
44066with @value{GDBN}; it is either a device name (to use a serial line),
44067a TCP port number (@code{:1234}), or @code{-} or @code{stdio} to use
44068stdin/stdout of @code{gdbserver}. Specify the name of the program to
44069debug in @var{prog}. Any remaining arguments will be passed to the
44070program verbatim. When the program exits, @value{GDBN} will close the
44071connection, and @code{gdbserver} will exit.
44072
44073@item
44074Debug a specific program by specifying the process ID of a running
44075program:
44076
44077@smallexample
44078gdbserver --attach @var{comm} @var{pid}
44079@end smallexample
44080
44081The @var{comm} parameter is as described above. Supply the process ID
44082of a running program in @var{pid}; @value{GDBN} will do everything
44083else. Like with the previous mode, when the process @var{pid} exits,
44084@value{GDBN} will close the connection, and @code{gdbserver} will exit.
44085
44086@item
44087Multi-process mode -- debug more than one program/process:
44088
44089@smallexample
44090gdbserver --multi @var{comm}
44091@end smallexample
44092
44093In this mode, @value{GDBN} can instruct @command{gdbserver} which
44094command(s) to run. Unlike the other 2 modes, @value{GDBN} will not
44095close the connection when a process being debugged exits, so you can
44096debug several processes in the same session.
44097@end itemize
44098
44099In each of the modes you may specify these options:
44100
44101@table @env
44102
44103@item --help
44104List all options, with brief explanations.
44105
44106@item --version
44107This option causes @command{gdbserver} to print its version number and exit.
44108
44109@item --attach
44110@command{gdbserver} will attach to a running program. The syntax is:
44111
44112@smallexample
44113target> gdbserver --attach @var{comm} @var{pid}
44114@end smallexample
44115
44116@var{pid} is the process ID of a currently running process. It isn't
44117necessary to point @command{gdbserver} at a binary for the running process.
44118
44119@item --multi
44120To start @code{gdbserver} without supplying an initial command to run
44121or process ID to attach, use this command line option.
44122Then you can connect using @kbd{target extended-remote} and start
44123the program you want to debug. The syntax is:
44124
44125@smallexample
44126target> gdbserver --multi @var{comm}
44127@end smallexample
44128
44129@item --debug
44130Instruct @code{gdbserver} to display extra status information about the debugging
44131process.
44132This option is intended for @code{gdbserver} development and for bug reports to
44133the developers.
44134
44135@item --remote-debug
44136Instruct @code{gdbserver} to display remote protocol debug output.
44137This option is intended for @code{gdbserver} development and for bug reports to
44138the developers.
44139
44140@item --wrapper
44141Specify a wrapper to launch programs
44142for debugging. The option should be followed by the name of the
44143wrapper, then any command-line arguments to pass to the wrapper, then
44144@kbd{--} indicating the end of the wrapper arguments.
44145
44146@item --once
44147By default, @command{gdbserver} keeps the listening TCP port open, so that
44148additional connections are possible. However, if you start @code{gdbserver}
44149with the @option{--once} option, it will stop listening for any further
44150connection attempts after connecting to the first @value{GDBN} session.
44151
44152@c --disable-packet is not documented for users.
44153
44154@c --disable-randomization and --no-disable-randomization are superseded by
44155@c QDisableRandomization.
44156
44157@end table
44158@c man end
44159
44160@c man begin SEEALSO gdbserver
44161@ifset man
44162The full documentation for @value{GDBN} is maintained as a Texinfo manual.
44163If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
44164documentation are properly installed at your site, the command
44165
44166@smallexample
44167info gdb
44168@end smallexample
44169
44170should give you access to the complete manual.
44171
44172@cite{Using GDB: A Guide to the GNU Source-Level Debugger},
44173Richard M. Stallman and Roland H. Pesch, July 1991.
44174@end ifset
44175@c man end
44176
44177@node gcore man
44178@heading gcore
44179
44180@c man title gcore Generate a core file of a running program
44181
44182@format
44183@c man begin SYNOPSIS gcore
44184gcore [-o @var{filename}] @var{pid}
44185@c man end
44186@end format
44187
44188@c man begin DESCRIPTION gcore
44189Generate a core dump of a running program with process ID @var{pid}.
44190Produced file is equivalent to a kernel produced core file as if the process
44191crashed (and if @kbd{ulimit -c} were used to set up an appropriate core dump
44192limit). Unlike after a crash, after @command{gcore} the program remains
44193running without any change.
44194@c man end
44195
44196@c man begin OPTIONS gcore
44197@table @env
44198@item -o @var{filename}
44199The optional argument
44200@var{filename} specifies the file name where to put the core dump.
44201If not specified, the file name defaults to @file{core.@var{pid}},
44202where @var{pid} is the running program process ID.
44203@end table
44204@c man end
44205
44206@c man begin SEEALSO gcore
44207@ifset man
44208The full documentation for @value{GDBN} is maintained as a Texinfo manual.
44209If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
44210documentation are properly installed at your site, the command
44211
44212@smallexample
44213info gdb
44214@end smallexample
44215
44216@noindent
44217should give you access to the complete manual.
44218
44219@cite{Using GDB: A Guide to the GNU Source-Level Debugger},
44220Richard M. Stallman and Roland H. Pesch, July 1991.
44221@end ifset
44222@c man end
44223
44224@node gdbinit man
44225@heading gdbinit
44226
44227@c man title gdbinit GDB initialization scripts
44228
44229@format
44230@c man begin SYNOPSIS gdbinit
44231@ifset SYSTEM_GDBINIT
44232@value{SYSTEM_GDBINIT}
44233@end ifset
44234
44235~/.gdbinit
44236
44237./.gdbinit
44238@c man end
44239@end format
44240
44241@c man begin DESCRIPTION gdbinit
44242These files contain @value{GDBN} commands to automatically execute during
44243@value{GDBN} startup. The lines of contents are canned sequences of commands,
44244described in
44245@ifset man
44246the @value{GDBN} manual in node @code{Sequences}
44247-- shell command @code{info -f gdb -n Sequences}.
44248@end ifset
44249@ifclear man
44250@ref{Sequences}.
44251@end ifclear
44252
44253Please read more in
44254@ifset man
44255the @value{GDBN} manual in node @code{Startup}
44256-- shell command @code{info -f gdb -n Startup}.
44257@end ifset
44258@ifclear man
44259@ref{Startup}.
44260@end ifclear
44261
44262@table @env
44263@ifset SYSTEM_GDBINIT
44264@item @value{SYSTEM_GDBINIT}
44265@end ifset
44266@ifclear SYSTEM_GDBINIT
44267@item (not enabled with @code{--with-system-gdbinit} during compilation)
44268@end ifclear
44269System-wide initialization file. It is executed unless user specified
44270@value{GDBN} option @code{-nx} or @code{-n}.
44271See more in
44272@ifset man
44273the @value{GDBN} manual in node @code{System-wide configuration}
44274-- shell command @code{info -f gdb -n 'System-wide configuration'}.
44275@end ifset
44276@ifclear man
44277@ref{System-wide configuration}.
44278@end ifclear
44279
44280@item ~/.gdbinit
44281User initialization file. It is executed unless user specified
44282@value{GDBN} options @code{-nx}, @code{-n} or @code{-nh}.
44283
44284@item ./.gdbinit
44285Initialization file for current directory. It may need to be enabled with
44286@value{GDBN} security command @code{set auto-load local-gdbinit}.
44287See more in
44288@ifset man
44289the @value{GDBN} manual in node @code{Init File in the Current Directory}
44290-- shell command @code{info -f gdb -n 'Init File in the Current Directory'}.
44291@end ifset
44292@ifclear man
44293@ref{Init File in the Current Directory}.
44294@end ifclear
44295@end table
44296@c man end
44297
44298@c man begin SEEALSO gdbinit
44299@ifset man
44300gdb(1), @code{info -f gdb -n Startup}
44301
44302The full documentation for @value{GDBN} is maintained as a Texinfo manual.
44303If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
44304documentation are properly installed at your site, the command
44305
44306@smallexample
44307info gdb
44308@end smallexample
44309
44310should give you access to the complete manual.
44311
44312@cite{Using GDB: A Guide to the GNU Source-Level Debugger},
44313Richard M. Stallman and Roland H. Pesch, July 1991.
44314@end ifset
44315@c man end
44316
44317@include gpl.texi
44318
44319@node GNU Free Documentation License
44320@appendix GNU Free Documentation License
44321@include fdl.texi
44322
44323@node Concept Index
44324@unnumbered Concept Index
44325
44326@printindex cp
44327
44328@node Command and Variable Index
44329@unnumbered Command, Variable, and Function Index
44330
44331@printindex fn
44332
44333@tex
44334% I think something like @@colophon should be in texinfo. In the
44335% meantime:
44336\long\def\colophon{\hbox to0pt{}\vfill
44337\centerline{The body of this manual is set in}
44338\centerline{\fontname\tenrm,}
44339\centerline{with headings in {\bf\fontname\tenbf}}
44340\centerline{and examples in {\tt\fontname\tentt}.}
44341\centerline{{\it\fontname\tenit\/},}
44342\centerline{{\bf\fontname\tenbf}, and}
44343\centerline{{\sl\fontname\tensl\/}}
44344\centerline{are used for emphasis.}\vfill}
44345\page\colophon
44346% Blame: doc@@cygnus.com, 1991.
44347@end tex
44348
44349@bye
This page took 0.209535 seconds and 4 git commands to generate.