1 \input texinfo @c -*-texinfo-*-
2 @c Copyright (C) 1988-2020 Free Software Foundation, Inc.
3 @c Copyright (C) 2020 Advanced Micro Devices, Inc. All rights reserved.
6 @c makeinfo ignores cmds prev to setfilename, so its arg cannot make use
7 @c of @set vars. However, you can override filename with makeinfo -o.
14 @settitle Debugging with @value{GDBN}
15 @setchapternewpage odd
24 @c To avoid file-name clashes between index.html and Index.html, when
25 @c the manual is produced on a Posix host and then moved to a
26 @c case-insensitive filesystem (e.g., MS-Windows), we separate the
27 @c indices into two: Concept Index and all the rest.
31 @c readline appendices use @vindex, @findex and @ftable,
32 @c annotate.texi and gdbmi use @findex.
35 @c !!set GDB manual's edition---not the same as GDB version!
36 @c This is updated by GNU Press.
39 @c !!set GDB edit command default editor
42 @c THIS MANUAL REQUIRES TEXINFO 4.0 OR LATER.
44 @c This is a dir.info fragment to support semi-automated addition of
45 @c manuals to an info tree.
46 @dircategory Software development
48 * ROCgdb: (gdb). The ROCm GNU debugger.
49 @c * gdbserver: (gdb) Server. The GNU debugging server.
53 @c man begin COPYRIGHT
54 Copyright @copyright{} 1988-2020 Free Software Foundation, Inc.
56 Copyright @copyright{} 2020 Advanced Micro Devices, Inc. All rights reserved.
58 Permission is granted to copy, distribute and/or modify this document
59 under the terms of the GNU Free Documentation License, Version 1.3 or
60 any later version published by the Free Software Foundation; with the
61 Invariant Sections being ``Free Software'' and ``Free Software Needs
62 Free Documentation'', with the Front-Cover Texts being ``A GNU Manual,''
63 and with the Back-Cover Texts as in (a) below.
65 (a) The FSF's Back-Cover Text is: ``You are free to copy and modify
66 this GNU Manual. Buying copies from GNU Press supports the FSF in
67 developing GNU and promoting software freedom.''
72 This file documents the @sc{gnu} debugger @value{GDBN}.
74 This is the @value{EDITION} Edition, of @cite{Debugging with
75 @value{GDBN}: the @sc{gnu} Source-Level Debugger} for @value{GDBN}
76 @ifset VERSION_PACKAGE
77 @value{VERSION_PACKAGE}
79 Version @value{GDBVN}.
85 @title Debugging with @value{GDBN}
86 @subtitle The @sc{gnu} Source-Level Debugger
88 @subtitle @value{EDITION} Edition, for @value{GDBN} version @value{GDBVN}
89 @ifset VERSION_PACKAGE
91 @subtitle @value{VERSION_PACKAGE}
93 @author Richard Stallman, Roland Pesch, Stan Shebs, et al.
97 \hfill (Send bugs and comments on @value{GDBN} to @value{BUGURL}.)\par
98 \hfill {\it Debugging with @value{GDBN}}\par
99 \hfill \TeX{}info \texinfoversion\par
103 @c Comment out publisher until upstreamed:
104 @c @vskip 0pt plus 1filll
105 @c Published by the Free Software Foundation @*
106 @c 51 Franklin Street, Fifth Floor,
107 @c Boston, MA 02110-1301, USA@*
108 @c ISBN 978-0-9831592-3-0 @*
115 @node Top, Summary, (dir), (dir)
117 @top Debugging with @value{GDBN}
119 This file describes @value{GDBN}, the @sc{gnu} symbolic debugger.
121 This is the @value{EDITION} Edition, for @value{GDBN}
122 @ifset VERSION_PACKAGE
123 @value{VERSION_PACKAGE}
125 Version @value{GDBVN}.
127 Copyright (C) 1988-2020 Free Software Foundation, Inc.
129 This edition of the GDB manual is dedicated to the memory of Fred
130 Fish. Fred was a long-standing contributor to GDB and to Free
131 software in general. We will miss him.
134 * Summary:: Summary of @value{GDBN}
135 * Sample Session:: A sample @value{GDBN} session
137 * Invocation:: Getting in and out of @value{GDBN}
138 * Commands:: @value{GDBN} commands
139 * Running:: Running programs under @value{GDBN}
140 * Stopping:: Stopping and continuing
141 * Reverse Execution:: Running programs backward
142 * Process Record and Replay:: Recording inferior's execution and replaying it
143 * Stack:: Examining the stack
144 * Source:: Examining source files
145 * Data:: Examining data
146 * Optimized Code:: Debugging optimized code
147 * Macros:: Preprocessor Macros
148 * Tracepoints:: Debugging remote targets non-intrusively
149 * Overlays:: Debugging programs that use overlays
151 * Languages:: Using @value{GDBN} with different languages
153 * Symbols:: Examining the symbol table
154 * Altering:: Altering execution
155 * GDB Files:: @value{GDBN} files
156 * Targets:: Specifying a debugging target
157 * Heterogeneous Debugging:: Debugging Heterogeneous Programs
158 * Remote Debugging:: Debugging remote programs
159 * Configurations:: Configuration-specific information
160 * Controlling GDB:: Controlling @value{GDBN}
161 * Extending GDB:: Extending @value{GDBN}
162 * Interpreters:: Command Interpreters
163 * TUI:: @value{GDBN} Text User Interface
164 * Emacs:: Using @value{GDBN} under @sc{gnu} Emacs
165 * GDB/MI:: @value{GDBN}'s Machine Interface.
166 * Annotations:: @value{GDBN}'s annotation interface.
167 * JIT Interface:: Using the JIT debugging interface.
168 * In-Process Agent:: In-Process Agent
170 * GDB Bugs:: Reporting bugs in @value{GDBN}
172 @ifset SYSTEM_READLINE
173 * Command Line Editing: (rluserman). Command Line Editing
174 * Using History Interactively: (history). Using History Interactively
176 @ifclear SYSTEM_READLINE
177 * Command Line Editing:: Command Line Editing
178 * Using History Interactively:: Using History Interactively
180 * In Memoriam:: In Memoriam
181 * Formatting Documentation:: How to format and print @value{GDBN} documentation
182 * Installing GDB:: Installing @value{GDBN}
183 * Maintenance Commands:: Maintenance Commands
184 * Remote Protocol:: GDB Remote Serial Protocol
185 * Agent Expressions:: The @value{GDBN} Agent Expression Mechanism
186 * Target Descriptions:: How targets can describe themselves to
188 * Operating System Information:: Getting additional information from
190 * Trace File Format:: @value{GDBN} trace file format
191 * Index Section Format:: .gdb_index section format
192 * Man Pages:: Manual pages
193 * Copying:: GNU General Public License says
194 how you can copy and share @value{GDBN}
195 * GNU Free Documentation License:: The license for this documentation
196 * Concept Index:: Index of @value{GDBN} concepts
197 * Command and Variable Index:: Index of @value{GDBN} commands, variables,
198 functions, and Python data types
206 @unnumbered Summary of @value{GDBN}
208 The purpose of a debugger such as @value{GDBN} is to allow you to see what is
209 going on ``inside'' another program while it executes---or what another
210 program was doing at the moment it crashed.
212 @value{GDBN} can do four main kinds of things (plus other things in support of
213 these) to help you catch bugs in the act:
217 Start your program, specifying anything that might affect its behavior.
220 Make your program stop on specified conditions.
223 Examine what has happened, when your program has stopped.
226 Change things in your program, so you can experiment with correcting the
227 effects of one bug and go on to learn about another.
230 You can use @value{GDBN} to debug programs written in C and C@t{++}.
231 For more information, see @ref{Supported Languages,,Supported Languages}.
232 For more information, see @ref{C,,C and C++}.
234 Support for D is partial. For information on D, see
238 Support for Modula-2 is partial. For information on Modula-2, see
239 @ref{Modula-2,,Modula-2}.
241 Support for OpenCL C is partial. For information on OpenCL C, see
242 @ref{OpenCL C,,OpenCL C}.
245 Debugging Pascal programs which use sets, subranges, file variables, or
246 nested functions does not currently work. @value{GDBN} does not support
247 entering expressions, printing values, or similar features using Pascal
251 @value{GDBN} can be used to debug programs written in Fortran, although
252 it may be necessary to refer to some variables with a trailing
255 @value{GDBN} can be used to debug programs written in Objective-C,
256 using either the Apple/NeXT or the GNU Objective-C runtime.
259 * Free Software:: Freely redistributable software
260 * Free Documentation:: Free Software Needs Free Documentation
261 * Contributors:: Contributors to GDB
265 @unnumberedsec Free Software
267 @value{GDBN} is @dfn{free software}, protected by the @sc{gnu}
268 General Public License
269 (GPL). The GPL gives you the freedom to copy or adapt a licensed
270 program---but every person getting a copy also gets with it the
271 freedom to modify that copy (which means that they must get access to
272 the source code), and the freedom to distribute further copies.
273 Typical software companies use copyrights to limit your freedoms; the
274 Free Software Foundation uses the GPL to preserve these freedoms.
276 Fundamentally, the General Public License is a license which says that
277 you have these freedoms and that you cannot take these freedoms away
280 @node Free Documentation
281 @unnumberedsec Free Software Needs Free Documentation
283 The biggest deficiency in the free software community today is not in
284 the software---it is the lack of good free documentation that we can
285 include with the free software. Many of our most important
286 programs do not come with free reference manuals and free introductory
287 texts. Documentation is an essential part of any software package;
288 when an important free software package does not come with a free
289 manual and a free tutorial, that is a major gap. We have many such
292 Consider Perl, for instance. The tutorial manuals that people
293 normally use are non-free. How did this come about? Because the
294 authors of those manuals published them with restrictive terms---no
295 copying, no modification, source files not available---which exclude
296 them from the free software world.
298 That wasn't the first time this sort of thing happened, and it was far
299 from the last. Many times we have heard a GNU user eagerly describe a
300 manual that he is writing, his intended contribution to the community,
301 only to learn that he had ruined everything by signing a publication
302 contract to make it non-free.
304 Free documentation, like free software, is a matter of freedom, not
305 price. The problem with the non-free manual is not that publishers
306 charge a price for printed copies---that in itself is fine. (The Free
307 Software Foundation sells printed copies of manuals, too.) The
308 problem is the restrictions on the use of the manual. Free manuals
309 are available in source code form, and give you permission to copy and
310 modify. Non-free manuals do not allow this.
312 The criteria of freedom for a free manual are roughly the same as for
313 free software. Redistribution (including the normal kinds of
314 commercial redistribution) must be permitted, so that the manual can
315 accompany every copy of the program, both on-line and on paper.
317 Permission for modification of the technical content is crucial too.
318 When people modify the software, adding or changing features, if they
319 are conscientious they will change the manual too---so they can
320 provide accurate and clear documentation for the modified program. A
321 manual that leaves you no choice but to write a new manual to document
322 a changed version of the program is not really available to our
325 Some kinds of limits on the way modification is handled are
326 acceptable. For example, requirements to preserve the original
327 author's copyright notice, the distribution terms, or the list of
328 authors, are ok. It is also no problem to require modified versions
329 to include notice that they were modified. Even entire sections that
330 may not be deleted or changed are acceptable, as long as they deal
331 with nontechnical topics (like this one). These kinds of restrictions
332 are acceptable because they don't obstruct the community's normal use
335 However, it must be possible to modify all the @emph{technical}
336 content of the manual, and then distribute the result in all the usual
337 media, through all the usual channels. Otherwise, the restrictions
338 obstruct the use of the manual, it is not free, and we need another
339 manual to replace it.
341 Please spread the word about this issue. Our community continues to
342 lose manuals to proprietary publishing. If we spread the word that
343 free software needs free reference manuals and free tutorials, perhaps
344 the next person who wants to contribute by writing documentation will
345 realize, before it is too late, that only free manuals contribute to
346 the free software community.
348 If you are writing documentation, please insist on publishing it under
349 the GNU Free Documentation License or another free documentation
350 license. Remember that this decision requires your approval---you
351 don't have to let the publisher decide. Some commercial publishers
352 will use a free license if you insist, but they will not propose the
353 option; it is up to you to raise the issue and say firmly that this is
354 what you want. If the publisher you are dealing with refuses, please
355 try other publishers. If you're not sure whether a proposed license
356 is free, write to @email{licensing@@gnu.org}.
358 You can encourage commercial publishers to sell more free, copylefted
359 manuals and tutorials by buying them, and particularly by buying
360 copies from the publishers that paid for their writing or for major
361 improvements. Meanwhile, try to avoid buying non-free documentation
362 at all. Check the distribution terms of a manual before you buy it,
363 and insist that whoever seeks your business must respect your freedom.
364 Check the history of the book, and try to reward the publishers that
365 have paid or pay the authors to work on it.
367 The Free Software Foundation maintains a list of free documentation
368 published by other publishers, at
369 @url{http://www.fsf.org/doc/other-free-books.html}.
372 @unnumberedsec Contributors to @value{GDBN}
374 Richard Stallman was the original author of @value{GDBN}, and of many
375 other @sc{gnu} programs. Many others have contributed to its
376 development. This section attempts to credit major contributors. One
377 of the virtues of free software is that everyone is free to contribute
378 to it; with regret, we cannot actually acknowledge everyone here. The
379 file @file{ChangeLog} in the @value{GDBN} distribution approximates a
380 blow-by-blow account.
382 Changes much prior to version 2.0 are lost in the mists of time.
385 @emph{Plea:} Additions to this section are particularly welcome. If you
386 or your friends (or enemies, to be evenhanded) have been unfairly
387 omitted from this list, we would like to add your names!
390 So that they may not regard their many labors as thankless, we
391 particularly thank those who shepherded @value{GDBN} through major
393 Andrew Cagney (releases 6.3, 6.2, 6.1, 6.0, 5.3, 5.2, 5.1 and 5.0);
394 Jim Blandy (release 4.18);
395 Jason Molenda (release 4.17);
396 Stan Shebs (release 4.14);
397 Fred Fish (releases 4.16, 4.15, 4.13, 4.12, 4.11, 4.10, and 4.9);
398 Stu Grossman and John Gilmore (releases 4.8, 4.7, 4.6, 4.5, and 4.4);
399 John Gilmore (releases 4.3, 4.2, 4.1, 4.0, and 3.9);
400 Jim Kingdon (releases 3.5, 3.4, and 3.3);
401 and Randy Smith (releases 3.2, 3.1, and 3.0).
403 Richard Stallman, assisted at various times by Peter TerMaat, Chris
404 Hanson, and Richard Mlynarik, handled releases through 2.8.
406 Michael Tiemann is the author of most of the @sc{gnu} C@t{++} support
407 in @value{GDBN}, with significant additional contributions from Per
408 Bothner and Daniel Berlin. James Clark wrote the @sc{gnu} C@t{++}
409 demangler. Early work on C@t{++} was by Peter TerMaat (who also did
410 much general update work leading to release 3.0).
412 @value{GDBN} uses the BFD subroutine library to examine multiple
413 object-file formats; BFD was a joint project of David V.
414 Henkel-Wallace, Rich Pixley, Steve Chamberlain, and John Gilmore.
416 David Johnson wrote the original COFF support; Pace Willison did
417 the original support for encapsulated COFF.
419 Brent Benson of Harris Computer Systems contributed DWARF 2 support.
421 Adam de Boor and Bradley Davis contributed the ISI Optimum V support.
422 Per Bothner, Noboyuki Hikichi, and Alessandro Forin contributed MIPS
424 Jean-Daniel Fekete contributed Sun 386i support.
425 Chris Hanson improved the HP9000 support.
426 Noboyuki Hikichi and Tomoyuki Hasei contributed Sony/News OS 3 support.
427 David Johnson contributed Encore Umax support.
428 Jyrki Kuoppala contributed Altos 3068 support.
429 Jeff Law contributed HP PA and SOM support.
430 Keith Packard contributed NS32K support.
431 Doug Rabson contributed Acorn Risc Machine support.
432 Bob Rusk contributed Harris Nighthawk CX-UX support.
433 Chris Smith contributed Convex support (and Fortran debugging).
434 Jonathan Stone contributed Pyramid support.
435 Michael Tiemann contributed SPARC support.
436 Tim Tucker contributed support for the Gould NP1 and Gould Powernode.
437 Pace Willison contributed Intel 386 support.
438 Jay Vosburgh contributed Symmetry support.
439 Marko Mlinar contributed OpenRISC 1000 support.
441 Andreas Schwab contributed M68K @sc{gnu}/Linux support.
443 Rich Schaefer and Peter Schauer helped with support of SunOS shared
446 Jay Fenlason and Roland McGrath ensured that @value{GDBN} and GAS agree
447 about several machine instruction sets.
449 Patrick Duval, Ted Goldstein, Vikram Koka and Glenn Engel helped develop
450 remote debugging. Intel Corporation, Wind River Systems, AMD, and ARM
451 contributed remote debugging modules for the i960, VxWorks, A29K UDI,
452 and RDI targets, respectively.
454 Brian Fox is the author of the readline libraries providing
455 command-line editing and command history.
457 Andrew Beers of SUNY Buffalo wrote the language-switching code, the
458 Modula-2 support, and contributed the Languages chapter of this manual.
460 Fred Fish wrote most of the support for Unix System Vr4.
461 He also enhanced the command-completion support to cover C@t{++} overloaded
464 Hitachi America (now Renesas America), Ltd. sponsored the support for
465 H8/300, H8/500, and Super-H processors.
467 NEC sponsored the support for the v850, Vr4xxx, and Vr5xxx processors.
469 Mitsubishi (now Renesas) sponsored the support for D10V, D30V, and M32R/D
472 Toshiba sponsored the support for the TX39 Mips processor.
474 Matsushita sponsored the support for the MN10200 and MN10300 processors.
476 Fujitsu sponsored the support for SPARClite and FR30 processors.
478 Kung Hsu, Jeff Law, and Rick Sladkey added support for hardware
481 Michael Snyder added support for tracepoints.
483 Stu Grossman wrote gdbserver.
485 Jim Kingdon, Peter Schauer, Ian Taylor, and Stu Grossman made
486 nearly innumerable bug fixes and cleanups throughout @value{GDBN}.
488 The following people at the Hewlett-Packard Company contributed
489 support for the PA-RISC 2.0 architecture, HP-UX 10.20, 10.30, and 11.0
490 (narrow mode), HP's implementation of kernel threads, HP's aC@t{++}
491 compiler, and the Text User Interface (nee Terminal User Interface):
492 Ben Krepp, Richard Title, John Bishop, Susan Macchia, Kathy Mann,
493 Satish Pai, India Paul, Steve Rehrauer, and Elena Zannoni. Kim Haase
494 provided HP-specific information in this manual.
496 DJ Delorie ported @value{GDBN} to MS-DOS, for the DJGPP project.
497 Robert Hoehne made significant contributions to the DJGPP port.
499 Cygnus Solutions has sponsored @value{GDBN} maintenance and much of its
500 development since 1991. Cygnus engineers who have worked on @value{GDBN}
501 fulltime include Mark Alexander, Jim Blandy, Per Bothner, Kevin
502 Buettner, Edith Epstein, Chris Faylor, Fred Fish, Martin Hunt, Jim
503 Ingham, John Gilmore, Stu Grossman, Kung Hsu, Jim Kingdon, John Metzler,
504 Fernando Nasser, Geoffrey Noer, Dawn Perchik, Rich Pixley, Zdenek
505 Radouch, Keith Seitz, Stan Shebs, David Taylor, and Elena Zannoni. In
506 addition, Dave Brolley, Ian Carmichael, Steve Chamberlain, Nick Clifton,
507 JT Conklin, Stan Cox, DJ Delorie, Ulrich Drepper, Frank Eigler, Doug
508 Evans, Sean Fagan, David Henkel-Wallace, Richard Henderson, Jeff
509 Holcomb, Jeff Law, Jim Lemke, Tom Lord, Bob Manson, Michael Meissner,
510 Jason Merrill, Catherine Moore, Drew Moseley, Ken Raeburn, Gavin
511 Romig-Koch, Rob Savoye, Jamie Smith, Mike Stump, Ian Taylor, Angela
512 Thomas, Michael Tiemann, Tom Tromey, Ron Unrau, Jim Wilson, and David
513 Zuhn have made contributions both large and small.
515 Andrew Cagney, Fernando Nasser, and Elena Zannoni, while working for
516 Cygnus Solutions, implemented the original @sc{gdb/mi} interface.
518 Jim Blandy added support for preprocessor macros, while working for Red
521 Andrew Cagney designed @value{GDBN}'s architecture vector. Many
522 people including Andrew Cagney, Stephane Carrez, Randolph Chung, Nick
523 Duffek, Richard Henderson, Mark Kettenis, Grace Sainsbury, Kei
524 Sakamoto, Yoshinori Sato, Michael Snyder, Andreas Schwab, Jason
525 Thorpe, Corinna Vinschen, Ulrich Weigand, and Elena Zannoni, helped
526 with the migration of old architectures to this new framework.
528 Andrew Cagney completely re-designed and re-implemented @value{GDBN}'s
529 unwinder framework, this consisting of a fresh new design featuring
530 frame IDs, independent frame sniffers, and the sentinel frame. Mark
531 Kettenis implemented the @sc{dwarf 2} unwinder, Jeff Johnston the
532 libunwind unwinder, and Andrew Cagney the dummy, sentinel, tramp, and
533 trad unwinders. The architecture-specific changes, each involving a
534 complete rewrite of the architecture's frame code, were carried out by
535 Jim Blandy, Joel Brobecker, Kevin Buettner, Andrew Cagney, Stephane
536 Carrez, Randolph Chung, Orjan Friberg, Richard Henderson, Daniel
537 Jacobowitz, Jeff Johnston, Mark Kettenis, Theodore A. Roth, Kei
538 Sakamoto, Yoshinori Sato, Michael Snyder, Corinna Vinschen, and Ulrich
541 Christian Zankel, Ross Morley, Bob Wilson, and Maxim Grigoriev from
542 Tensilica, Inc.@: contributed support for Xtensa processors. Others
543 who have worked on the Xtensa port of @value{GDBN} in the past include
544 Steve Tjiang, John Newlin, and Scott Foehner.
546 Michael Eager and staff of Xilinx, Inc., contributed support for the
547 Xilinx MicroBlaze architecture.
549 Initial support for the FreeBSD/mips target and native configuration
550 was developed by SRI International and the University of Cambridge
551 Computer Laboratory under DARPA/AFRL contract FA8750-10-C-0237
552 ("CTSRD"), as part of the DARPA CRASH research programme.
554 Initial support for the FreeBSD/riscv target and native configuration
555 was developed by SRI International and the University of Cambridge
556 Computer Laboratory (Department of Computer Science and Technology)
557 under DARPA contract HR0011-18-C-0016 ("ECATS"), as part of the DARPA
558 SSITH research programme.
560 The original port to the OpenRISC 1000 is believed to be due to
561 Alessandro Forin and Per Bothner. More recent ports have been the work
562 of Jeremy Bennett, Franck Jullien, Stefan Wallentowitz and
565 Initial support for heterogeneous program debugging and the
566 @acronym{AMD GPU} targets was developed by the following people at the
567 Advanced Micro Devices company: Scott Linder, Laurent Morichetti,
568 Qingchuan Shi, Tony Tye, and Zoran Zaric.
571 @chapter A Sample @value{GDBN} Session
573 You can use this manual at your leisure to read all about @value{GDBN}.
574 However, a handful of commands are enough to get started using the
575 debugger. This chapter illustrates those commands.
578 In this sample session, we emphasize user input like this: @b{input},
579 to make it easier to pick out from the surrounding output.
582 @c FIXME: this example may not be appropriate for some configs, where
583 @c FIXME...primary interest is in remote use.
585 One of the preliminary versions of @sc{gnu} @code{m4} (a generic macro
586 processor) exhibits the following bug: sometimes, when we change its
587 quote strings from the default, the commands used to capture one macro
588 definition within another stop working. In the following short @code{m4}
589 session, we define a macro @code{foo} which expands to @code{0000}; we
590 then use the @code{m4} built-in @code{defn} to define @code{bar} as the
591 same thing. However, when we change the open quote string to
592 @code{<QUOTE>} and the close quote string to @code{<UNQUOTE>}, the same
593 procedure fails to define a new synonym @code{baz}:
602 @b{define(bar,defn(`foo'))}
606 @b{changequote(<QUOTE>,<UNQUOTE>)}
608 @b{define(baz,defn(<QUOTE>foo<UNQUOTE>))}
611 m4: End of input: 0: fatal error: EOF in string
615 Let us use @value{GDBN} to try to see what is going on.
618 $ @b{@value{GDBP} m4}
619 @c FIXME: this falsifies the exact text played out, to permit smallbook
620 @c FIXME... format to come out better.
621 @value{GDBN} is free software and you are welcome to distribute copies
622 of it under certain conditions; type "show copying" to see
624 There is absolutely no warranty for @value{GDBN}; type "show warranty"
627 @value{GDBN} @value{GDBVN}, Copyright 1999 Free Software Foundation, Inc...
632 @value{GDBN} reads only enough symbol data to know where to find the
633 rest when needed; as a result, the first prompt comes up very quickly.
634 We now tell @value{GDBN} to use a narrower display width than usual, so
635 that examples fit in this manual.
638 (@value{GDBP}) @b{set width 70}
642 We need to see how the @code{m4} built-in @code{changequote} works.
643 Having looked at the source, we know the relevant subroutine is
644 @code{m4_changequote}, so we set a breakpoint there with the @value{GDBN}
645 @code{break} command.
648 (@value{GDBP}) @b{break m4_changequote}
649 Breakpoint 1 at 0x62f4: file builtin.c, line 879.
653 Using the @code{run} command, we start @code{m4} running under @value{GDBN}
654 control; as long as control does not reach the @code{m4_changequote}
655 subroutine, the program runs as usual:
658 (@value{GDBP}) @b{run}
659 Starting program: /work/Editorial/gdb/gnu/m4/m4
667 To trigger the breakpoint, we call @code{changequote}. @value{GDBN}
668 suspends execution of @code{m4}, displaying information about the
669 context where it stops.
672 @b{changequote(<QUOTE>,<UNQUOTE>)}
674 Breakpoint 1, m4_changequote (argc=3, argv=0x33c70)
676 879 if (bad_argc(TOKEN_DATA_TEXT(argv[0]),argc,1,3))
680 Now we use the command @code{n} (@code{next}) to advance execution to
681 the next line of the current function.
685 882 set_quotes((argc >= 2) ? TOKEN_DATA_TEXT(argv[1])\
690 @code{set_quotes} looks like a promising subroutine. We can go into it
691 by using the command @code{s} (@code{step}) instead of @code{next}.
692 @code{step} goes to the next line to be executed in @emph{any}
693 subroutine, so it steps into @code{set_quotes}.
697 set_quotes (lq=0x34c78 "<QUOTE>", rq=0x34c88 "<UNQUOTE>")
699 530 if (lquote != def_lquote)
703 The display that shows the subroutine where @code{m4} is now
704 suspended (and its arguments) is called a stack frame display. It
705 shows a summary of the stack. We can use the @code{backtrace}
706 command (which can also be spelled @code{bt}), to see where we are
707 in the stack as a whole: the @code{backtrace} command displays a
708 stack frame for each active subroutine.
711 (@value{GDBP}) @b{bt}
712 #0 set_quotes (lq=0x34c78 "<QUOTE>", rq=0x34c88 "<UNQUOTE>")
714 #1 0x6344 in m4_changequote (argc=3, argv=0x33c70)
716 #2 0x8174 in expand_macro (sym=0x33320) at macro.c:242
717 #3 0x7a88 in expand_token (obs=0x0, t=209696, td=0xf7fffa30)
719 #4 0x79dc in expand_input () at macro.c:40
720 #5 0x2930 in main (argc=0, argv=0xf7fffb20) at m4.c:195
724 We step through a few more lines to see what happens. The first two
725 times, we can use @samp{s}; the next two times we use @code{n} to avoid
726 falling into the @code{xstrdup} subroutine.
730 0x3b5c 532 if (rquote != def_rquote)
732 0x3b80 535 lquote = (lq == nil || *lq == '\0') ? \
733 def_lquote : xstrdup(lq);
735 536 rquote = (rq == nil || *rq == '\0') ? def_rquote\
738 538 len_lquote = strlen(rquote);
742 The last line displayed looks a little odd; we can examine the variables
743 @code{lquote} and @code{rquote} to see if they are in fact the new left
744 and right quotes we specified. We use the command @code{p}
745 (@code{print}) to see their values.
748 (@value{GDBP}) @b{p lquote}
749 $1 = 0x35d40 "<QUOTE>"
750 (@value{GDBP}) @b{p rquote}
751 $2 = 0x35d50 "<UNQUOTE>"
755 @code{lquote} and @code{rquote} are indeed the new left and right quotes.
756 To look at some context, we can display ten lines of source
757 surrounding the current line with the @code{l} (@code{list}) command.
763 535 lquote = (lq == nil || *lq == '\0') ? def_lquote\
765 536 rquote = (rq == nil || *rq == '\0') ? def_rquote\
768 538 len_lquote = strlen(rquote);
769 539 len_rquote = strlen(lquote);
776 Let us step past the two lines that set @code{len_lquote} and
777 @code{len_rquote}, and then examine the values of those variables.
781 539 len_rquote = strlen(lquote);
784 (@value{GDBP}) @b{p len_lquote}
786 (@value{GDBP}) @b{p len_rquote}
791 That certainly looks wrong, assuming @code{len_lquote} and
792 @code{len_rquote} are meant to be the lengths of @code{lquote} and
793 @code{rquote} respectively. We can set them to better values using
794 the @code{p} command, since it can print the value of
795 any expression---and that expression can include subroutine calls and
799 (@value{GDBP}) @b{p len_lquote=strlen(lquote)}
801 (@value{GDBP}) @b{p len_rquote=strlen(rquote)}
806 Is that enough to fix the problem of using the new quotes with the
807 @code{m4} built-in @code{defn}? We can allow @code{m4} to continue
808 executing with the @code{c} (@code{continue}) command, and then try the
809 example that caused trouble initially:
815 @b{define(baz,defn(<QUOTE>foo<UNQUOTE>))}
822 Success! The new quotes now work just as well as the default ones. The
823 problem seems to have been just the two typos defining the wrong
824 lengths. We allow @code{m4} exit by giving it an EOF as input:
828 Program exited normally.
832 The message @samp{Program exited normally.} is from @value{GDBN}; it
833 indicates @code{m4} has finished executing. We can end our @value{GDBN}
834 session with the @value{GDBN} @code{quit} command.
837 (@value{GDBP}) @b{quit}
841 @chapter Getting In and Out of @value{GDBN}
843 This chapter discusses how to start @value{GDBN}, and how to get out of it.
847 type @samp{@value{GDBP}} to start @value{GDBN}.
849 type @kbd{quit} or @kbd{Ctrl-d} to exit.
853 * Invoking GDB:: How to start @value{GDBN}
854 * Quitting GDB:: How to quit @value{GDBN}
855 * Shell Commands:: How to use shell commands inside @value{GDBN}
856 * Logging Output:: How to log @value{GDBN}'s output to a file
860 @section Invoking @value{GDBN}
862 Invoke @value{GDBN} by running the program @code{@value{GDBP}}. Once started,
863 @value{GDBN} reads commands from the terminal until you tell it to exit.
865 You can also run @code{@value{GDBP}} with a variety of arguments and options,
866 to specify more of your debugging environment at the outset.
868 The command-line options described here are designed
869 to cover a variety of situations; in some environments, some of these
870 options may effectively be unavailable.
872 The most usual way to start @value{GDBN} is with one argument,
873 specifying an executable program:
876 @value{GDBP} @var{program}
880 You can also start with both an executable program and a core file
884 @value{GDBP} @var{program} @var{core}
887 You can, instead, specify a process ID as a second argument or use option
888 @code{-p}, if you want to debug a running process:
891 @value{GDBP} @var{program} 1234
896 would attach @value{GDBN} to process @code{1234}. With option @option{-p} you
897 can omit the @var{program} filename.
899 Taking advantage of the second command-line argument requires a fairly
900 complete operating system; when you use @value{GDBN} as a remote
901 debugger attached to a bare board, there may not be any notion of
902 ``process'', and there is often no way to get a core dump. @value{GDBN}
903 will warn you if it is unable to attach or to read core dumps.
905 You can optionally have @code{@value{GDBP}} pass any arguments after the
906 executable file to the inferior using @code{--args}. This option stops
909 @value{GDBP} --args gcc -O2 -c foo.c
911 This will cause @code{@value{GDBP}} to debug @code{gcc}, and to set
912 @code{gcc}'s command-line arguments (@pxref{Arguments}) to @samp{-O2 -c foo.c}.
914 You can run @code{@value{GDBP}} without printing the front material, which describes
915 @value{GDBN}'s non-warranty, by specifying @code{--silent}
916 (or @code{-q}/@code{--quiet}):
919 @value{GDBP} --silent
923 You can further control how @value{GDBN} starts up by using command-line
924 options. @value{GDBN} itself can remind you of the options available.
934 to display all available options and briefly describe their use
935 (@samp{@value{GDBP} -h} is a shorter equivalent).
937 All options and command line arguments you give are processed
938 in sequential order. The order makes a difference when the
939 @samp{-x} option is used.
943 * File Options:: Choosing files
944 * Mode Options:: Choosing modes
945 * Startup:: What @value{GDBN} does during startup
949 @subsection Choosing Files
951 When @value{GDBN} starts, it reads any arguments other than options as
952 specifying an executable file and core file (or process ID). This is
953 the same as if the arguments were specified by the @samp{-se} and
954 @samp{-c} (or @samp{-p}) options respectively. (@value{GDBN} reads the
955 first argument that does not have an associated option flag as
956 equivalent to the @samp{-se} option followed by that argument; and the
957 second argument that does not have an associated option flag, if any, as
958 equivalent to the @samp{-c}/@samp{-p} option followed by that argument.)
959 If the second argument begins with a decimal digit, @value{GDBN} will
960 first attempt to attach to it as a process, and if that fails, attempt
961 to open it as a corefile. If you have a corefile whose name begins with
962 a digit, you can prevent @value{GDBN} from treating it as a pid by
963 prefixing it with @file{./}, e.g.@: @file{./12345}.
965 If @value{GDBN} has not been configured to included core file support,
966 such as for most embedded targets, then it will complain about a second
967 argument and ignore it.
969 Many options have both long and short forms; both are shown in the
970 following list. @value{GDBN} also recognizes the long forms if you truncate
971 them, so long as enough of the option is present to be unambiguous.
972 (If you prefer, you can flag option arguments with @samp{--} rather
973 than @samp{-}, though we illustrate the more usual convention.)
975 @c NOTE: the @cindex entries here use double dashes ON PURPOSE. This
976 @c way, both those who look for -foo and --foo in the index, will find
980 @item -symbols @var{file}
982 @cindex @code{--symbols}
984 Read symbol table from file @var{file}.
986 @item -exec @var{file}
988 @cindex @code{--exec}
990 Use file @var{file} as the executable file to execute when appropriate,
991 and for examining pure data in conjunction with a core dump.
995 Read symbol table from file @var{file} and use it as the executable
998 @item -core @var{file}
1000 @cindex @code{--core}
1002 Use file @var{file} as a core dump to examine.
1004 @item -pid @var{number}
1005 @itemx -p @var{number}
1006 @cindex @code{--pid}
1008 Connect to process ID @var{number}, as with the @code{attach} command.
1010 @item -command @var{file}
1011 @itemx -x @var{file}
1012 @cindex @code{--command}
1014 Execute commands from file @var{file}. The contents of this file is
1015 evaluated exactly as the @code{source} command would.
1016 @xref{Command Files,, Command files}.
1018 @item -eval-command @var{command}
1019 @itemx -ex @var{command}
1020 @cindex @code{--eval-command}
1022 Execute a single @value{GDBN} command.
1024 This option may be used multiple times to call multiple commands. It may
1025 also be interleaved with @samp{-command} as required.
1028 @value{GDBP} -ex 'target sim' -ex 'load' \
1029 -x setbreakpoints -ex 'run' a.out
1032 @item -init-command @var{file}
1033 @itemx -ix @var{file}
1034 @cindex @code{--init-command}
1036 Execute commands from file @var{file} before loading the inferior (but
1037 after loading gdbinit files).
1040 @item -init-eval-command @var{command}
1041 @itemx -iex @var{command}
1042 @cindex @code{--init-eval-command}
1044 Execute a single @value{GDBN} command before loading the inferior (but
1045 after loading gdbinit files).
1048 @item -directory @var{directory}
1049 @itemx -d @var{directory}
1050 @cindex @code{--directory}
1052 Add @var{directory} to the path to search for source and script files.
1056 @cindex @code{--readnow}
1058 Read each symbol file's entire symbol table immediately, rather than
1059 the default, which is to read it incrementally as it is needed.
1060 This makes startup slower, but makes future operations faster.
1063 @anchor{--readnever}
1064 @cindex @code{--readnever}, command-line option
1065 Do not read each symbol file's symbolic debug information. This makes
1066 startup faster but at the expense of not being able to perform
1067 symbolic debugging. DWARF unwind information is also not read,
1068 meaning backtraces may become incomplete or inaccurate. One use of
1069 this is when a user simply wants to do the following sequence: attach,
1070 dump core, detach. Loading the debugging information in this case is
1071 an unnecessary cause of delay.
1075 @subsection Choosing Modes
1077 You can run @value{GDBN} in various alternative modes---for example, in
1078 batch mode or quiet mode.
1086 Do not execute commands found in any initialization file.
1087 There are three init files, loaded in the following order:
1090 @item @file{system.gdbinit}
1091 This is the system-wide init file.
1092 Its location is specified with the @code{--with-system-gdbinit}
1093 configure option (@pxref{System-wide configuration}).
1094 It is loaded first when @value{GDBN} starts, before command line options
1095 have been processed.
1096 @item @file{system.gdbinit.d}
1097 This is the system-wide init directory.
1098 Its location is specified with the @code{--with-system-gdbinit-dir}
1099 configure option (@pxref{System-wide configuration}).
1100 Files in this directory are loaded in alphabetical order immediately after
1101 system.gdbinit (if enabled) when @value{GDBN} starts, before command line
1102 options have been processed. Files need to have a recognized scripting
1103 language extension (@file{.py}/@file{.scm}) or be named with a @file{.gdb}
1104 extension to be interpreted as regular @value{GDBN} commands. @value{GDBN}
1105 will not recurse into any subdirectories of this directory.
1106 @item @file{~/.gdbinit}
1107 This is the init file in your home directory.
1108 It is loaded next, after @file{system.gdbinit}, and before
1109 command options have been processed.
1110 @item @file{./.gdbinit}
1111 This is the init file in the current directory.
1112 It is loaded last, after command line options other than @code{-x} and
1113 @code{-ex} have been processed. Command line options @code{-x} and
1114 @code{-ex} are processed last, after @file{./.gdbinit} has been loaded.
1117 For further documentation on startup processing, @xref{Startup}.
1118 For documentation on how to write command files,
1119 @xref{Command Files,,Command Files}.
1124 Do not execute commands found in @file{~/.gdbinit}, the init file
1125 in your home directory.
1131 @cindex @code{--quiet}
1132 @cindex @code{--silent}
1134 ``Quiet''. Do not print the introductory and copyright messages. These
1135 messages are also suppressed in batch mode.
1138 @cindex @code{--batch}
1139 Run in batch mode. Exit with status @code{0} after processing all the
1140 command files specified with @samp{-x} (and all commands from
1141 initialization files, if not inhibited with @samp{-n}). Exit with
1142 nonzero status if an error occurs in executing the @value{GDBN} commands
1143 in the command files. Batch mode also disables pagination, sets unlimited
1144 terminal width and height @pxref{Screen Size}, and acts as if @kbd{set confirm
1145 off} were in effect (@pxref{Messages/Warnings}).
1147 Batch mode may be useful for running @value{GDBN} as a filter, for
1148 example to download and run a program on another computer; in order to
1149 make this more useful, the message
1152 Program exited normally.
1156 (which is ordinarily issued whenever a program running under
1157 @value{GDBN} control terminates) is not issued when running in batch
1161 @cindex @code{--batch-silent}
1162 Run in batch mode exactly like @samp{-batch}, but totally silently. All
1163 @value{GDBN} output to @code{stdout} is prevented (@code{stderr} is
1164 unaffected). This is much quieter than @samp{-silent} and would be useless
1165 for an interactive session.
1167 This is particularly useful when using targets that give @samp{Loading section}
1168 messages, for example.
1170 Note that targets that give their output via @value{GDBN}, as opposed to
1171 writing directly to @code{stdout}, will also be made silent.
1173 @item -return-child-result
1174 @cindex @code{--return-child-result}
1175 The return code from @value{GDBN} will be the return code from the child
1176 process (the process being debugged), with the following exceptions:
1180 @value{GDBN} exits abnormally. E.g., due to an incorrect argument or an
1181 internal error. In this case the exit code is the same as it would have been
1182 without @samp{-return-child-result}.
1184 The user quits with an explicit value. E.g., @samp{quit 1}.
1186 The child process never runs, or is not allowed to terminate, in which case
1187 the exit code will be -1.
1190 This option is useful in conjunction with @samp{-batch} or @samp{-batch-silent},
1191 when @value{GDBN} is being used as a remote program loader or simulator
1196 @cindex @code{--nowindows}
1198 ``No windows''. If @value{GDBN} comes with a graphical user interface
1199 (GUI) built in, then this option tells @value{GDBN} to only use the command-line
1200 interface. If no GUI is available, this option has no effect.
1204 @cindex @code{--windows}
1206 If @value{GDBN} includes a GUI, then this option requires it to be
1209 @item -cd @var{directory}
1211 Run @value{GDBN} using @var{directory} as its working directory,
1212 instead of the current directory.
1214 @item -data-directory @var{directory}
1215 @itemx -D @var{directory}
1216 @cindex @code{--data-directory}
1218 Run @value{GDBN} using @var{directory} as its data directory.
1219 The data directory is where @value{GDBN} searches for its
1220 auxiliary files. @xref{Data Files}.
1224 @cindex @code{--fullname}
1226 @sc{gnu} Emacs sets this option when it runs @value{GDBN} as a
1227 subprocess. It tells @value{GDBN} to output the full file name and line
1228 number in a standard, recognizable fashion each time a stack frame is
1229 displayed (which includes each time your program stops). This
1230 recognizable format looks like two @samp{\032} characters, followed by
1231 the file name, line number and character position separated by colons,
1232 and a newline. The Emacs-to-@value{GDBN} interface program uses the two
1233 @samp{\032} characters as a signal to display the source code for the
1236 @item -annotate @var{level}
1237 @cindex @code{--annotate}
1238 This option sets the @dfn{annotation level} inside @value{GDBN}. Its
1239 effect is identical to using @samp{set annotate @var{level}}
1240 (@pxref{Annotations}). The annotation @var{level} controls how much
1241 information @value{GDBN} prints together with its prompt, values of
1242 expressions, source lines, and other types of output. Level 0 is the
1243 normal, level 1 is for use when @value{GDBN} is run as a subprocess of
1244 @sc{gnu} Emacs, level 3 is the maximum annotation suitable for programs
1245 that control @value{GDBN}, and level 2 has been deprecated.
1247 The annotation mechanism has largely been superseded by @sc{gdb/mi}
1251 @cindex @code{--args}
1252 Change interpretation of command line so that arguments following the
1253 executable file are passed as command line arguments to the inferior.
1254 This option stops option processing.
1256 @item -baud @var{bps}
1258 @cindex @code{--baud}
1260 Set the line speed (baud rate or bits per second) of any serial
1261 interface used by @value{GDBN} for remote debugging.
1263 @item -l @var{timeout}
1265 Set the timeout (in seconds) of any communication used by @value{GDBN}
1266 for remote debugging.
1268 @item -tty @var{device}
1269 @itemx -t @var{device}
1270 @cindex @code{--tty}
1272 Run using @var{device} for your program's standard input and output.
1273 @c FIXME: kingdon thinks there is more to -tty. Investigate.
1275 @c resolve the situation of these eventually
1277 @cindex @code{--tui}
1278 Activate the @dfn{Text User Interface} when starting. The Text User
1279 Interface manages several text windows on the terminal, showing
1280 source, assembly, registers and @value{GDBN} command outputs
1281 (@pxref{TUI, ,@value{GDBN} Text User Interface}). Do not use this
1282 option if you run @value{GDBN} from Emacs (@pxref{Emacs, ,
1283 Using @value{GDBN} under @sc{gnu} Emacs}).
1285 @item -interpreter @var{interp}
1286 @cindex @code{--interpreter}
1287 Use the interpreter @var{interp} for interface with the controlling
1288 program or device. This option is meant to be set by programs which
1289 communicate with @value{GDBN} using it as a back end.
1290 @xref{Interpreters, , Command Interpreters}.
1292 @samp{--interpreter=mi} (or @samp{--interpreter=mi3}) causes
1293 @value{GDBN} to use the @dfn{@sc{gdb/mi} interface} version 3 (@pxref{GDB/MI, ,
1294 The @sc{gdb/mi} Interface}) included since @value{GDBN} version 9.1. @sc{gdb/mi}
1295 version 2 (@code{mi2}), included in @value{GDBN} 6.0 and version 1 (@code{mi1}),
1296 included in @value{GDBN} 5.3, are also available. Earlier @sc{gdb/mi}
1297 interfaces are no longer supported.
1300 @cindex @code{--write}
1301 Open the executable and core files for both reading and writing. This
1302 is equivalent to the @samp{set write on} command inside @value{GDBN}
1306 @cindex @code{--statistics}
1307 This option causes @value{GDBN} to print statistics about time and
1308 memory usage after it completes each command and returns to the prompt.
1311 @cindex @code{--version}
1312 This option causes @value{GDBN} to print its version number and
1313 no-warranty blurb, and exit.
1315 @item -configuration
1316 @cindex @code{--configuration}
1317 This option causes @value{GDBN} to print details about its build-time
1318 configuration parameters, and then exit. These details can be
1319 important when reporting @value{GDBN} bugs (@pxref{GDB Bugs}).
1324 @subsection What @value{GDBN} Does During Startup
1325 @cindex @value{GDBN} startup
1327 Here's the description of what @value{GDBN} does during session startup:
1331 Sets up the command interpreter as specified by the command line
1332 (@pxref{Mode Options, interpreter}).
1336 Reads the system-wide @dfn{init file} (if @option{--with-system-gdbinit} was
1337 used when building @value{GDBN}; @pxref{System-wide configuration,
1338 ,System-wide configuration and settings}) and the files in the system-wide
1339 gdbinit directory (if @option{--with-system-gdbinit-dir} was used) and executes
1340 all the commands in those files. The files need to be named with a @file{.gdb}
1341 extension to be interpreted as @value{GDBN} commands, or they can be written
1342 in a supported scripting language with an appropriate file extension.
1344 @anchor{Home Directory Init File}
1346 Reads the init file (if any) in your home directory@footnote{On
1347 DOS/Windows systems, the home directory is the one pointed to by the
1348 @code{HOME} environment variable.} and executes all the commands in
1351 @anchor{Option -init-eval-command}
1353 Executes commands and command files specified by the @samp{-iex} and
1354 @samp{-ix} options in their specified order. Usually you should use the
1355 @samp{-ex} and @samp{-x} options instead, but this way you can apply
1356 settings before @value{GDBN} init files get executed and before inferior
1360 Processes command line options and operands.
1362 @anchor{Init File in the Current Directory during Startup}
1364 Reads and executes the commands from init file (if any) in the current
1365 working directory as long as @samp{set auto-load local-gdbinit} is set to
1366 @samp{on} (@pxref{Init File in the Current Directory}).
1367 This is only done if the current directory is
1368 different from your home directory. Thus, you can have more than one
1369 init file, one generic in your home directory, and another, specific
1370 to the program you are debugging, in the directory where you invoke
1374 If the command line specified a program to debug, or a process to
1375 attach to, or a core file, @value{GDBN} loads any auto-loaded
1376 scripts provided for the program or for its loaded shared libraries.
1377 @xref{Auto-loading}.
1379 If you wish to disable the auto-loading during startup,
1380 you must do something like the following:
1383 $ gdb -iex "set auto-load python-scripts off" myprogram
1386 Option @samp{-ex} does not work because the auto-loading is then turned
1390 Executes commands and command files specified by the @samp{-ex} and
1391 @samp{-x} options in their specified order. @xref{Command Files}, for
1392 more details about @value{GDBN} command files.
1395 Reads the command history recorded in the @dfn{history file}.
1396 @xref{Command History}, for more details about the command history and the
1397 files where @value{GDBN} records it.
1400 Init files use the same syntax as @dfn{command files} (@pxref{Command
1401 Files}) and are processed by @value{GDBN} in the same way. The init
1402 file in your home directory can set options (such as @samp{set
1403 complaints}) that affect subsequent processing of command line options
1404 and operands. Init files are not executed if you use the @samp{-nx}
1405 option (@pxref{Mode Options, ,Choosing Modes}).
1407 To display the list of init files loaded by gdb at startup, you
1408 can use @kbd{gdb --help}.
1410 @cindex init file name
1411 @cindex @file{.gdbinit}
1412 @cindex @file{gdb.ini}
1413 The @value{GDBN} init files are normally called @file{.gdbinit}.
1414 The DJGPP port of @value{GDBN} uses the name @file{gdb.ini}, due to
1415 the limitations of file names imposed by DOS filesystems. The Windows
1416 port of @value{GDBN} uses the standard name, but if it finds a
1417 @file{gdb.ini} file in your home directory, it warns you about that
1418 and suggests to rename the file to the standard name.
1422 @section Quitting @value{GDBN}
1423 @cindex exiting @value{GDBN}
1424 @cindex leaving @value{GDBN}
1427 @kindex quit @r{[}@var{expression}@r{]}
1428 @kindex q @r{(@code{quit})}
1429 @item quit @r{[}@var{expression}@r{]}
1431 To exit @value{GDBN}, use the @code{quit} command (abbreviated
1432 @code{q}), or type an end-of-file character (usually @kbd{Ctrl-d}). If you
1433 do not supply @var{expression}, @value{GDBN} will terminate normally;
1434 otherwise it will terminate using the result of @var{expression} as the
1439 An interrupt (often @kbd{Ctrl-c}) does not exit from @value{GDBN}, but rather
1440 terminates the action of any @value{GDBN} command that is in progress and
1441 returns to @value{GDBN} command level. It is safe to type the interrupt
1442 character at any time because @value{GDBN} does not allow it to take effect
1443 until a time when it is safe.
1445 If you have been using @value{GDBN} to control an attached process or
1446 device, you can release it with the @code{detach} command
1447 (@pxref{Attach, ,Debugging an Already-running Process}).
1449 @node Shell Commands
1450 @section Shell Commands
1452 If you need to execute occasional shell commands during your
1453 debugging session, there is no need to leave or suspend @value{GDBN}; you can
1454 just use the @code{shell} command.
1459 @cindex shell escape
1460 @item shell @var{command-string}
1461 @itemx !@var{command-string}
1462 Invoke a standard shell to execute @var{command-string}.
1463 Note that no space is needed between @code{!} and @var{command-string}.
1464 If it exists, the environment variable @code{SHELL} determines which
1465 shell to run. Otherwise @value{GDBN} uses the default shell
1466 (@file{/bin/sh} on Unix systems, @file{COMMAND.COM} on MS-DOS, etc.).
1469 The utility @code{make} is often needed in development environments.
1470 You do not have to use the @code{shell} command for this purpose in
1475 @cindex calling make
1476 @item make @var{make-args}
1477 Execute the @code{make} program with the specified
1478 arguments. This is equivalent to @samp{shell make @var{make-args}}.
1484 @cindex send the output of a gdb command to a shell command
1486 @item pipe [@var{command}] | @var{shell_command}
1487 @itemx | [@var{command}] | @var{shell_command}
1488 @itemx pipe -d @var{delim} @var{command} @var{delim} @var{shell_command}
1489 @itemx | -d @var{delim} @var{command} @var{delim} @var{shell_command}
1490 Executes @var{command} and sends its output to @var{shell_command}.
1491 Note that no space is needed around @code{|}.
1492 If no @var{command} is provided, the last command executed is repeated.
1494 In case the @var{command} contains a @code{|}, the option @code{-d @var{delim}}
1495 can be used to specify an alternate delimiter string @var{delim} that separates
1496 the @var{command} from the @var{shell_command}.
1501 (@value{GDBP}) p var
1511 (@value{GDBP}) pipe p var|wc
1513 (@value{GDBP}) |p var|wc -l
1517 (@value{GDBP}) p /x var
1525 (@value{GDBP}) ||grep red
1529 (@value{GDBP}) | -d ! echo this contains a | char\n ! sed -e 's/|/PIPE/'
1530 this contains a PIPE char
1531 (@value{GDBP}) | -d xxx echo this contains a | char!\n xxx sed -e 's/|/PIPE/'
1532 this contains a PIPE char!
1538 The convenience variables @code{$_shell_exitcode} and @code{$_shell_exitsignal}
1539 can be used to examine the exit status of the last shell command launched
1540 by @code{shell}, @code{make}, @code{pipe} and @code{|}.
1541 @xref{Convenience Vars,, Convenience Variables}.
1543 @node Logging Output
1544 @section Logging Output
1545 @cindex logging @value{GDBN} output
1546 @cindex save @value{GDBN} output to a file
1548 You may want to save the output of @value{GDBN} commands to a file.
1549 There are several commands to control @value{GDBN}'s logging.
1553 @item set logging on
1555 @item set logging off
1557 @cindex logging file name
1558 @item set logging file @var{file}
1559 Change the name of the current logfile. The default logfile is @file{gdb.txt}.
1560 @item set logging overwrite [on|off]
1561 By default, @value{GDBN} will append to the logfile. Set @code{overwrite} if
1562 you want @code{set logging on} to overwrite the logfile instead.
1563 @item set logging redirect [on|off]
1564 By default, @value{GDBN} output will go to both the terminal and the logfile.
1565 Set @code{redirect} if you want output to go only to the log file.
1566 @item set logging debugredirect [on|off]
1567 By default, @value{GDBN} debug output will go to both the terminal and the logfile.
1568 Set @code{debugredirect} if you want debug output to go only to the log file.
1569 @kindex show logging
1571 Show the current values of the logging settings.
1574 You can also redirect the output of a @value{GDBN} command to a
1575 shell command. @xref{pipe}.
1577 @chapter @value{GDBN} Commands
1579 You can abbreviate a @value{GDBN} command to the first few letters of the command
1580 name, if that abbreviation is unambiguous; and you can repeat certain
1581 @value{GDBN} commands by typing just @key{RET}. You can also use the @key{TAB}
1582 key to get @value{GDBN} to fill out the rest of a word in a command (or to
1583 show you the alternatives available, if there is more than one possibility).
1586 * Command Syntax:: How to give commands to @value{GDBN}
1587 * Command Settings:: How to change default behavior of commands
1588 * Completion:: Command completion
1589 * Command Options:: Command options
1590 * Help:: How to ask @value{GDBN} for help
1593 @node Command Syntax
1594 @section Command Syntax
1596 A @value{GDBN} command is a single line of input. There is no limit on
1597 how long it can be. It starts with a command name, which is followed by
1598 arguments whose meaning depends on the command name. For example, the
1599 command @code{step} accepts an argument which is the number of times to
1600 step, as in @samp{step 5}. You can also use the @code{step} command
1601 with no arguments. Some commands do not allow any arguments.
1603 @cindex abbreviation
1604 @value{GDBN} command names may always be truncated if that abbreviation is
1605 unambiguous. Other possible command abbreviations are listed in the
1606 documentation for individual commands. In some cases, even ambiguous
1607 abbreviations are allowed; for example, @code{s} is specially defined as
1608 equivalent to @code{step} even though there are other commands whose
1609 names start with @code{s}. You can test abbreviations by using them as
1610 arguments to the @code{help} command.
1612 @cindex repeating commands
1613 @kindex RET @r{(repeat last command)}
1614 A blank line as input to @value{GDBN} (typing just @key{RET}) means to
1615 repeat the previous command. Certain commands (for example, @code{run})
1616 will not repeat this way; these are commands whose unintentional
1617 repetition might cause trouble and which you are unlikely to want to
1618 repeat. User-defined commands can disable this feature; see
1619 @ref{Define, dont-repeat}.
1621 The @code{list} and @code{x} commands, when you repeat them with
1622 @key{RET}, construct new arguments rather than repeating
1623 exactly as typed. This permits easy scanning of source or memory.
1625 @value{GDBN} can also use @key{RET} in another way: to partition lengthy
1626 output, in a way similar to the common utility @code{more}
1627 (@pxref{Screen Size,,Screen Size}). Since it is easy to press one
1628 @key{RET} too many in this situation, @value{GDBN} disables command
1629 repetition after any command that generates this sort of display.
1631 @kindex # @r{(a comment)}
1633 Any text from a @kbd{#} to the end of the line is a comment; it does
1634 nothing. This is useful mainly in command files (@pxref{Command
1635 Files,,Command Files}).
1637 @cindex repeating command sequences
1638 @kindex Ctrl-o @r{(operate-and-get-next)}
1639 The @kbd{Ctrl-o} binding is useful for repeating a complex sequence of
1640 commands. This command accepts the current line, like @key{RET}, and
1641 then fetches the next line relative to the current line from the history
1645 @node Command Settings
1646 @section Command Settings
1647 @cindex default behavior of commands, changing
1648 @cindex default settings, changing
1650 Many commands change their behavior according to command-specific
1651 variables or settings. These settings can be changed with the
1652 @code{set} subcommands. For example, the @code{print} command
1653 (@pxref{Data, ,Examining Data}) prints arrays differently depending on
1654 settings changeable with the commands @code{set print elements
1655 NUMBER-OF-ELEMENTS} and @code{set print array-indexes}, among others.
1657 You can change these settings to your preference in the gdbinit files
1658 loaded at @value{GDBN} startup. @xref{Startup}.
1660 The settings can also be changed interactively during the debugging
1661 session. For example, to change the limit of array elements to print,
1662 you can do the following:
1664 (@value{GDBP}) set print elements 10
1665 (@value{GDBP}) print some_array
1666 $1 = @{0, 10, 20, 30, 40, 50, 60, 70, 80, 90...@}
1669 The above @code{set print elements 10} command changes the number of
1670 elements to print from the default of 200 to 10. If you only intend
1671 this limit of 10 to be used for printing @code{some_array}, then you
1672 must restore the limit back to 200, with @code{set print elements
1675 Some commands allow overriding settings with command options. For
1676 example, the @code{print} command supports a number of options that
1677 allow overriding relevant global print settings as set by @code{set
1678 print} subcommands. @xref{print options}. The example above could be
1681 (@value{GDBP}) print -elements 10 -- some_array
1682 $1 = @{0, 10, 20, 30, 40, 50, 60, 70, 80, 90...@}
1685 Alternatively, you can use the @code{with} command to change a setting
1686 temporarily, for the duration of a command invocation.
1689 @kindex with command
1690 @kindex w @r{(@code{with})}
1692 @cindex temporarily change settings
1693 @item with @var{setting} [@var{value}] [-- @var{command}]
1694 @itemx w @var{setting} [@var{value}] [-- @var{command}]
1695 Temporarily set @var{setting} to @var{value} for the duration of
1698 @var{setting} is any setting you can change with the @code{set}
1699 subcommands. @var{value} is the value to assign to @code{setting}
1700 while running @code{command}.
1702 If no @var{command} is provided, the last command executed is
1705 If a @var{command} is provided, it must be preceded by a double dash
1706 (@code{--}) separator. This is required because some settings accept
1707 free-form arguments, such as expressions or filenames.
1709 For example, the command
1711 (@value{GDBP}) with print array on -- print some_array
1714 is equivalent to the following 3 commands:
1716 (@value{GDBP}) set print array on
1717 (@value{GDBP}) print some_array
1718 (@value{GDBP}) set print array off
1721 The @code{with} command is particularly useful when you want to
1722 override a setting while running user-defined commands, or commands
1723 defined in Python or Guile. @xref{Extending GDB,, Extending GDB}.
1726 (@value{GDBP}) with print pretty on -- my_complex_command
1729 To change several settings for the same command, you can nest
1730 @code{with} commands. For example, @code{with language ada -- with
1731 print elements 10} temporarily changes the language to Ada and sets a
1732 limit of 10 elements to print for arrays and strings.
1737 @section Command Completion
1740 @cindex word completion
1741 @value{GDBN} can fill in the rest of a word in a command for you, if there is
1742 only one possibility; it can also show you what the valid possibilities
1743 are for the next word in a command, at any time. This works for @value{GDBN}
1744 commands, @value{GDBN} subcommands, command options, and the names of symbols
1747 Press the @key{TAB} key whenever you want @value{GDBN} to fill out the rest
1748 of a word. If there is only one possibility, @value{GDBN} fills in the
1749 word, and waits for you to finish the command (or press @key{RET} to
1750 enter it). For example, if you type
1752 @c FIXME "@key" does not distinguish its argument sufficiently to permit
1753 @c complete accuracy in these examples; space introduced for clarity.
1754 @c If texinfo enhancements make it unnecessary, it would be nice to
1755 @c replace " @key" by "@key" in the following...
1757 (@value{GDBP}) info bre @key{TAB}
1761 @value{GDBN} fills in the rest of the word @samp{breakpoints}, since that is
1762 the only @code{info} subcommand beginning with @samp{bre}:
1765 (@value{GDBP}) info breakpoints
1769 You can either press @key{RET} at this point, to run the @code{info
1770 breakpoints} command, or backspace and enter something else, if
1771 @samp{breakpoints} does not look like the command you expected. (If you
1772 were sure you wanted @code{info breakpoints} in the first place, you
1773 might as well just type @key{RET} immediately after @samp{info bre},
1774 to exploit command abbreviations rather than command completion).
1776 If there is more than one possibility for the next word when you press
1777 @key{TAB}, @value{GDBN} sounds a bell. You can either supply more
1778 characters and try again, or just press @key{TAB} a second time;
1779 @value{GDBN} displays all the possible completions for that word. For
1780 example, you might want to set a breakpoint on a subroutine whose name
1781 begins with @samp{make_}, but when you type @kbd{b make_@key{TAB}} @value{GDBN}
1782 just sounds the bell. Typing @key{TAB} again displays all the
1783 function names in your program that begin with those characters, for
1787 (@value{GDBP}) b make_ @key{TAB}
1788 @exdent @value{GDBN} sounds bell; press @key{TAB} again, to see:
1789 make_a_section_from_file make_environ
1790 make_abs_section make_function_type
1791 make_blockvector make_pointer_type
1792 make_cleanup make_reference_type
1793 make_command make_symbol_completion_list
1794 (@value{GDBP}) b make_
1798 After displaying the available possibilities, @value{GDBN} copies your
1799 partial input (@samp{b make_} in the example) so you can finish the
1802 If you just want to see the list of alternatives in the first place, you
1803 can press @kbd{M-?} rather than pressing @key{TAB} twice. @kbd{M-?}
1804 means @kbd{@key{META} ?}. You can type this either by holding down a
1805 key designated as the @key{META} shift on your keyboard (if there is
1806 one) while typing @kbd{?}, or as @key{ESC} followed by @kbd{?}.
1808 If the number of possible completions is large, @value{GDBN} will
1809 print as much of the list as it has collected, as well as a message
1810 indicating that the list may be truncated.
1813 (@value{GDBP}) b m@key{TAB}@key{TAB}
1815 <... the rest of the possible completions ...>
1816 *** List may be truncated, max-completions reached. ***
1821 This behavior can be controlled with the following commands:
1824 @kindex set max-completions
1825 @item set max-completions @var{limit}
1826 @itemx set max-completions unlimited
1827 Set the maximum number of completion candidates. @value{GDBN} will
1828 stop looking for more completions once it collects this many candidates.
1829 This is useful when completing on things like function names as collecting
1830 all the possible candidates can be time consuming.
1831 The default value is 200. A value of zero disables tab-completion.
1832 Note that setting either no limit or a very large limit can make
1834 @kindex show max-completions
1835 @item show max-completions
1836 Show the maximum number of candidates that @value{GDBN} will collect and show
1840 @cindex quotes in commands
1841 @cindex completion of quoted strings
1842 Sometimes the string you need, while logically a ``word'', may contain
1843 parentheses or other characters that @value{GDBN} normally excludes from
1844 its notion of a word. To permit word completion to work in this
1845 situation, you may enclose words in @code{'} (single quote marks) in
1846 @value{GDBN} commands.
1848 A likely situation where you might need this is in typing an
1849 expression that involves a C@t{++} symbol name with template
1850 parameters. This is because when completing expressions, GDB treats
1851 the @samp{<} character as word delimiter, assuming that it's the
1852 less-than comparison operator (@pxref{C Operators, , C and C@t{++}
1855 For example, when you want to call a C@t{++} template function
1856 interactively using the @code{print} or @code{call} commands, you may
1857 need to distinguish whether you mean the version of @code{name} that
1858 was specialized for @code{int}, @code{name<int>()}, or the version
1859 that was specialized for @code{float}, @code{name<float>()}. To use
1860 the word-completion facilities in this situation, type a single quote
1861 @code{'} at the beginning of the function name. This alerts
1862 @value{GDBN} that it may need to consider more information than usual
1863 when you press @key{TAB} or @kbd{M-?} to request word completion:
1866 (@value{GDBP}) p 'func< @kbd{M-?}
1867 func<int>() func<float>()
1868 (@value{GDBP}) p 'func<
1871 When setting breakpoints however (@pxref{Specify Location}), you don't
1872 usually need to type a quote before the function name, because
1873 @value{GDBN} understands that you want to set a breakpoint on a
1877 (@value{GDBP}) b func< @kbd{M-?}
1878 func<int>() func<float>()
1879 (@value{GDBP}) b func<
1882 This is true even in the case of typing the name of C@t{++} overloaded
1883 functions (multiple definitions of the same function, distinguished by
1884 argument type). For example, when you want to set a breakpoint you
1885 don't need to distinguish whether you mean the version of @code{name}
1886 that takes an @code{int} parameter, @code{name(int)}, or the version
1887 that takes a @code{float} parameter, @code{name(float)}.
1890 (@value{GDBP}) b bubble( @kbd{M-?}
1891 bubble(int) bubble(double)
1892 (@value{GDBP}) b bubble(dou @kbd{M-?}
1896 See @ref{quoting names} for a description of other scenarios that
1899 For more information about overloaded functions, see @ref{C Plus Plus
1900 Expressions, ,C@t{++} Expressions}. You can use the command @code{set
1901 overload-resolution off} to disable overload resolution;
1902 see @ref{Debugging C Plus Plus, ,@value{GDBN} Features for C@t{++}}.
1904 @cindex completion of structure field names
1905 @cindex structure field name completion
1906 @cindex completion of union field names
1907 @cindex union field name completion
1908 When completing in an expression which looks up a field in a
1909 structure, @value{GDBN} also tries@footnote{The completer can be
1910 confused by certain kinds of invalid expressions. Also, it only
1911 examines the static type of the expression, not the dynamic type.} to
1912 limit completions to the field names available in the type of the
1916 (@value{GDBP}) p gdb_stdout.@kbd{M-?}
1917 magic to_fputs to_rewind
1918 to_data to_isatty to_write
1919 to_delete to_put to_write_async_safe
1924 This is because the @code{gdb_stdout} is a variable of the type
1925 @code{struct ui_file} that is defined in @value{GDBN} sources as
1932 ui_file_flush_ftype *to_flush;
1933 ui_file_write_ftype *to_write;
1934 ui_file_write_async_safe_ftype *to_write_async_safe;
1935 ui_file_fputs_ftype *to_fputs;
1936 ui_file_read_ftype *to_read;
1937 ui_file_delete_ftype *to_delete;
1938 ui_file_isatty_ftype *to_isatty;
1939 ui_file_rewind_ftype *to_rewind;
1940 ui_file_put_ftype *to_put;
1945 @node Command Options
1946 @section Command options
1948 @cindex command options
1949 Some commands accept options starting with a leading dash. For
1950 example, @code{print -pretty}. Similarly to command names, you can
1951 abbreviate a @value{GDBN} option to the first few letters of the
1952 option name, if that abbreviation is unambiguous, and you can also use
1953 the @key{TAB} key to get @value{GDBN} to fill out the rest of a word
1954 in an option (or to show you the alternatives available, if there is
1955 more than one possibility).
1957 @cindex command options, raw input
1958 Some commands take raw input as argument. For example, the print
1959 command processes arbitrary expressions in any of the languages
1960 supported by @value{GDBN}. With such commands, because raw input may
1961 start with a leading dash that would be confused with an option or any
1962 of its abbreviations, e.g.@: @code{print -p} (short for @code{print
1963 -pretty} or printing negative @code{p}?), if you specify any command
1964 option, then you must use a double-dash (@code{--}) delimiter to
1965 indicate the end of options.
1967 @cindex command options, boolean
1969 Some options are described as accepting an argument which can be
1970 either @code{on} or @code{off}. These are known as @dfn{boolean
1971 options}. Similarly to boolean settings commands---@code{on} and
1972 @code{off} are the typical values, but any of @code{1}, @code{yes} and
1973 @code{enable} can also be used as ``true'' value, and any of @code{0},
1974 @code{no} and @code{disable} can also be used as ``false'' value. You
1975 can also omit a ``true'' value, as it is implied by default.
1977 For example, these are equivalent:
1980 (@value{GDBP}) print -object on -pretty off -element unlimited -- *myptr
1981 (@value{GDBP}) p -o -p 0 -e u -- *myptr
1984 You can discover the set of options some command accepts by completing
1985 on @code{-} after the command name. For example:
1988 (@value{GDBP}) print -@key{TAB}@key{TAB}
1989 -address -max-depth -raw-values -union
1990 -array -null-stop -repeats -vtbl
1991 -array-indexes -object -static-members
1992 -elements -pretty -symbol
1995 Completion will in some cases guide you with a suggestion of what kind
1996 of argument an option expects. For example:
1999 (@value{GDBP}) print -elements @key{TAB}@key{TAB}
2003 Here, the option expects a number (e.g., @code{100}), not literal
2004 @code{NUMBER}. Such metasyntactical arguments are always presented in
2007 (For more on using the @code{print} command, see @ref{Data, ,Examining
2011 @section Getting Help
2012 @cindex online documentation
2015 You can always ask @value{GDBN} itself for information on its commands,
2016 using the command @code{help}.
2019 @kindex h @r{(@code{help})}
2022 You can use @code{help} (abbreviated @code{h}) with no arguments to
2023 display a short list of named classes of commands:
2027 List of classes of commands:
2029 aliases -- Aliases of other commands
2030 breakpoints -- Making program stop at certain points
2031 data -- Examining data
2032 files -- Specifying and examining files
2033 internals -- Maintenance commands
2034 obscure -- Obscure features
2035 running -- Running the program
2036 stack -- Examining the stack
2037 status -- Status inquiries
2038 support -- Support facilities
2039 tracepoints -- Tracing of program execution without
2040 stopping the program
2041 user-defined -- User-defined commands
2043 Type "help" followed by a class name for a list of
2044 commands in that class.
2045 Type "help" followed by command name for full
2047 Command name abbreviations are allowed if unambiguous.
2050 @c the above line break eliminates huge line overfull...
2052 @item help @var{class}
2053 Using one of the general help classes as an argument, you can get a
2054 list of the individual commands in that class. For example, here is the
2055 help display for the class @code{status}:
2058 (@value{GDBP}) help status
2063 @c Line break in "show" line falsifies real output, but needed
2064 @c to fit in smallbook page size.
2065 info -- Generic command for showing things
2066 about the program being debugged
2067 show -- Generic command for showing things
2070 Type "help" followed by command name for full
2072 Command name abbreviations are allowed if unambiguous.
2076 @item help @var{command}
2077 With a command name as @code{help} argument, @value{GDBN} displays a
2078 short paragraph on how to use that command.
2081 @item apropos [-v] @var{regexp}
2082 The @code{apropos} command searches through all of the @value{GDBN}
2083 commands, and their documentation, for the regular expression specified in
2084 @var{args}. It prints out all matches found. The optional flag @samp{-v},
2085 which stands for @samp{verbose}, indicates to output the full documentation
2086 of the matching commands and highlight the parts of the documentation
2087 matching @var{regexp}. For example:
2098 alias -- Define a new command that is an alias of an existing command
2099 aliases -- Aliases of other commands
2100 d -- Delete some breakpoints or auto-display expressions
2101 del -- Delete some breakpoints or auto-display expressions
2102 delete -- Delete some breakpoints or auto-display expressions
2110 apropos -v cut.*thread apply
2114 results in the below output, where @samp{cut for 'thread apply}
2115 is highlighted if styling is enabled.
2119 taas -- Apply a command to all threads (ignoring errors
2122 shortcut for 'thread apply all -s COMMAND'
2124 tfaas -- Apply a command to all frames of all threads
2125 (ignoring errors and empty output).
2126 Usage: tfaas COMMAND
2127 shortcut for 'thread apply all -s frame apply all -s COMMAND'
2132 @item complete @var{args}
2133 The @code{complete @var{args}} command lists all the possible completions
2134 for the beginning of a command. Use @var{args} to specify the beginning of the
2135 command you want completed. For example:
2141 @noindent results in:
2152 @noindent This is intended for use by @sc{gnu} Emacs.
2155 In addition to @code{help}, you can use the @value{GDBN} commands @code{info}
2156 and @code{show} to inquire about the state of your program, or the state
2157 of @value{GDBN} itself. Each command supports many topics of inquiry; this
2158 manual introduces each of them in the appropriate context. The listings
2159 under @code{info} and under @code{show} in the Command, Variable, and
2160 Function Index point to all the sub-commands. @xref{Command and Variable
2166 @kindex i @r{(@code{info})}
2168 This command (abbreviated @code{i}) is for describing the state of your
2169 program. For example, you can show the arguments passed to a function
2170 with @code{info args}, list the registers currently in use with @code{info
2171 registers}, or list the breakpoints you have set with @code{info breakpoints}.
2172 You can get a complete list of the @code{info} sub-commands with
2173 @w{@code{help info}}.
2177 You can assign the result of an expression to an environment variable with
2178 @code{set}. For example, you can set the @value{GDBN} prompt to a $-sign with
2179 @code{set prompt $}.
2183 In contrast to @code{info}, @code{show} is for describing the state of
2184 @value{GDBN} itself.
2185 You can change most of the things you can @code{show}, by using the
2186 related command @code{set}; for example, you can control what number
2187 system is used for displays with @code{set radix}, or simply inquire
2188 which is currently in use with @code{show radix}.
2191 To display all the settable parameters and their current
2192 values, you can use @code{show} with no arguments; you may also use
2193 @code{info set}. Both commands produce the same display.
2194 @c FIXME: "info set" violates the rule that "info" is for state of
2195 @c FIXME...program. Ck w/ GNU: "info set" to be called something else,
2196 @c FIXME...or change desc of rule---eg "state of prog and debugging session"?
2200 Here are several miscellaneous @code{show} subcommands, all of which are
2201 exceptional in lacking corresponding @code{set} commands:
2204 @kindex show version
2205 @cindex @value{GDBN} version number
2207 Show what version of @value{GDBN} is running. You should include this
2208 information in @value{GDBN} bug-reports. If multiple versions of
2209 @value{GDBN} are in use at your site, you may need to determine which
2210 version of @value{GDBN} you are running; as @value{GDBN} evolves, new
2211 commands are introduced, and old ones may wither away. Also, many
2212 system vendors ship variant versions of @value{GDBN}, and there are
2213 variant versions of @value{GDBN} in @sc{gnu}/Linux distributions as well.
2214 The version number is the same as the one announced when you start
2217 @kindex show copying
2218 @kindex info copying
2219 @cindex display @value{GDBN} copyright
2222 Display information about permission for copying @value{GDBN}.
2224 @kindex show warranty
2225 @kindex info warranty
2227 @itemx info warranty
2228 Display the @sc{gnu} ``NO WARRANTY'' statement, or a warranty,
2229 if your version of @value{GDBN} comes with one.
2231 @kindex show configuration
2232 @item show configuration
2233 Display detailed information about the way @value{GDBN} was configured
2234 when it was built. This displays the optional arguments passed to the
2235 @file{configure} script and also configuration parameters detected
2236 automatically by @command{configure}. When reporting a @value{GDBN}
2237 bug (@pxref{GDB Bugs}), it is important to include this information in
2243 @chapter Running Programs Under @value{GDBN}
2245 When you run a program under @value{GDBN}, you must first generate
2246 debugging information when you compile it.
2248 You may start @value{GDBN} with its arguments, if any, in an environment
2249 of your choice. If you are doing native debugging, you may redirect
2250 your program's input and output, debug an already running process, or
2251 kill a child process.
2254 * Compilation:: Compiling for debugging
2255 * Starting:: Starting your program
2256 * Arguments:: Your program's arguments
2257 * Environment:: Your program's environment
2259 * Working Directory:: Your program's working directory
2260 * Input/Output:: Your program's input and output
2261 * Attach:: Debugging an already-running process
2262 * Kill Process:: Killing the child process
2264 * Inferiors and Programs:: Debugging multiple inferiors and programs
2265 * Threads:: Debugging programs with multiple threads
2266 * Forks:: Debugging forks
2267 * Checkpoint/Restart:: Setting a @emph{bookmark} to return to later
2271 @section Compiling for Debugging
2273 In order to debug a program effectively, you need to generate
2274 debugging information when you compile it. This debugging information
2275 is stored in the object file; it describes the data type of each
2276 variable or function and the correspondence between source line numbers
2277 and addresses in the executable code.
2279 To request debugging information, specify the @option{-g} option when
2280 you run the compiler. However, to use the most expressive format
2281 available, including @value{GDBN} extensions if at all possible,
2282 @value{NGCC}, the @sc{gnu} C/C@t{++} compiler, supports
2283 @w{@option{-ggdb}} which produces debugging information for use by
2284 @value{GDBN}. We recommend that you @emph{always} use
2285 @w{@option{-ggdb}} instead of plain @option{-g} if it is supported by
2286 the compiler you are using.
2288 Programs that are to be shipped to your customers are compiled with
2289 optimizations, using the @option{-O} compiler option. However, some
2290 compilers are unable to handle the @option{-g} and @option{-O} options
2291 together. Using those compilers, you cannot generate optimized
2292 executables containing debugging information.
2294 @value{NGCC} supports @option{-g} with or without @option{-O}, making
2295 it possible to debug optimized code. We recommend that you
2296 @emph{always} use @option{-g} whenever you compile a program. You may
2297 think your program is correct, but there is no sense in pushing your
2298 luck. For more information, see @ref{Optimized Code}.
2300 Older versions of the @sc{gnu} C compiler permitted a variant option
2301 @w{@option{-gg}} for debugging information. @value{GDBN} no longer
2302 supports this format; if your @sc{gnu} C compiler has this option, do
2305 @value{GDBN} knows about preprocessor macros and can show you their
2306 expansion (@pxref{Macros}). Most compilers do not include information
2307 about preprocessor macros in the debugging information if you specify
2308 the @option{-g} flag alone. Version 3.1 and later of @value{NGCC},
2309 the @sc{gnu} C compiler, provides macro information if you are using
2310 the DWARF debugging format, and specify the option @w{@option{-g3}}.
2312 @xref{Debugging Options,,Options for Debugging Your Program or GCC,
2313 gcc, Using the @sc{gnu} Compiler Collection (GCC)}, for more
2314 information on @value{NGCC} options affecting debug information.
2316 You will have the best debugging experience if you use the latest
2317 version of the DWARF debugging format that your compiler supports.
2318 DWARF is currently the most expressive and best supported debugging
2319 format in @value{GDBN}.
2323 @section Starting your Program
2329 @kindex r @r{(@code{run})}
2332 Use the @code{run} command to start your program under @value{GDBN}.
2333 You must first specify the program name with an argument to
2334 @value{GDBN} (@pxref{Invocation, ,Getting In and Out of
2335 @value{GDBN}}), or by using the @code{file} or @code{exec-file}
2336 command (@pxref{Files, ,Commands to Specify Files}).
2340 If you are running your program in an execution environment that
2341 supports processes, @code{run} creates an inferior process and makes
2342 that process run your program. In some environments without processes,
2343 @code{run} jumps to the start of your program. Other targets,
2344 like @samp{remote}, are always running. If you get an error
2345 message like this one:
2348 The "remote" target does not support "run".
2349 Try "help target" or "continue".
2353 then use @code{continue} to run your program. You may need @code{load}
2354 first (@pxref{load}).
2356 The execution of a program is affected by certain information it
2357 receives from its superior. @value{GDBN} provides ways to specify this
2358 information, which you must do @emph{before} starting your program. (You
2359 can change it after starting your program, but such changes only affect
2360 your program the next time you start it.) This information may be
2361 divided into four categories:
2364 @item The @emph{arguments.}
2365 Specify the arguments to give your program as the arguments of the
2366 @code{run} command. If a shell is available on your target, the shell
2367 is used to pass the arguments, so that you may use normal conventions
2368 (such as wildcard expansion or variable substitution) in describing
2370 In Unix systems, you can control which shell is used with the
2371 @code{SHELL} environment variable. If you do not define @code{SHELL},
2372 @value{GDBN} uses the default shell (@file{/bin/sh}). You can disable
2373 use of any shell with the @code{set startup-with-shell} command (see
2376 @item The @emph{environment.}
2377 Your program normally inherits its environment from @value{GDBN}, but you can
2378 use the @value{GDBN} commands @code{set environment} and @code{unset
2379 environment} to change parts of the environment that affect
2380 your program. @xref{Environment, ,Your Program's Environment}.
2382 @item The @emph{working directory.}
2383 You can set your program's working directory with the command
2384 @kbd{set cwd}. If you do not set any working directory with this
2385 command, your program will inherit @value{GDBN}'s working directory if
2386 native debugging, or the remote server's working directory if remote
2387 debugging. @xref{Working Directory, ,Your Program's Working
2390 @item The @emph{standard input and output.}
2391 Your program normally uses the same device for standard input and
2392 standard output as @value{GDBN} is using. You can redirect input and output
2393 in the @code{run} command line, or you can use the @code{tty} command to
2394 set a different device for your program.
2395 @xref{Input/Output, ,Your Program's Input and Output}.
2398 @emph{Warning:} While input and output redirection work, you cannot use
2399 pipes to pass the output of the program you are debugging to another
2400 program; if you attempt this, @value{GDBN} is likely to wind up debugging the
2404 When you issue the @code{run} command, your program begins to execute
2405 immediately. @xref{Stopping, ,Stopping and Continuing}, for discussion
2406 of how to arrange for your program to stop. Once your program has
2407 stopped, you may call functions in your program, using the @code{print}
2408 or @code{call} commands. @xref{Data, ,Examining Data}.
2410 If the modification time of your symbol file has changed since the last
2411 time @value{GDBN} read its symbols, @value{GDBN} discards its symbol
2412 table, and reads it again. When it does this, @value{GDBN} tries to retain
2413 your current breakpoints.
2418 @cindex run to main procedure
2419 The name of the main procedure can vary from language to language.
2420 With C or C@t{++}, the main procedure name is always @code{main}, but
2421 other languages such as Ada do not require a specific name for their
2422 main procedure. The debugger provides a convenient way to start the
2423 execution of the program and to stop at the beginning of the main
2424 procedure, depending on the language used.
2426 The @samp{start} command does the equivalent of setting a temporary
2427 breakpoint at the beginning of the main procedure and then invoking
2428 the @samp{run} command.
2430 @cindex elaboration phase
2431 Some programs contain an @dfn{elaboration} phase where some startup code is
2432 executed before the main procedure is called. This depends on the
2433 languages used to write your program. In C@t{++}, for instance,
2434 constructors for static and global objects are executed before
2435 @code{main} is called. It is therefore possible that the debugger stops
2436 before reaching the main procedure. However, the temporary breakpoint
2437 will remain to halt execution.
2439 Specify the arguments to give to your program as arguments to the
2440 @samp{start} command. These arguments will be given verbatim to the
2441 underlying @samp{run} command. Note that the same arguments will be
2442 reused if no argument is provided during subsequent calls to
2443 @samp{start} or @samp{run}.
2445 It is sometimes necessary to debug the program during elaboration. In
2446 these cases, using the @code{start} command would stop the execution
2447 of your program too late, as the program would have already completed
2448 the elaboration phase. Under these circumstances, either insert
2449 breakpoints in your elaboration code before running your program or
2450 use the @code{starti} command.
2454 @cindex run to first instruction
2455 The @samp{starti} command does the equivalent of setting a temporary
2456 breakpoint at the first instruction of a program's execution and then
2457 invoking the @samp{run} command. For programs containing an
2458 elaboration phase, the @code{starti} command will stop execution at
2459 the start of the elaboration phase.
2461 @anchor{set exec-wrapper}
2462 @kindex set exec-wrapper
2463 @item set exec-wrapper @var{wrapper}
2464 @itemx show exec-wrapper
2465 @itemx unset exec-wrapper
2466 When @samp{exec-wrapper} is set, the specified wrapper is used to
2467 launch programs for debugging. @value{GDBN} starts your program
2468 with a shell command of the form @kbd{exec @var{wrapper}
2469 @var{program}}. Quoting is added to @var{program} and its
2470 arguments, but not to @var{wrapper}, so you should add quotes if
2471 appropriate for your shell. The wrapper runs until it executes
2472 your program, and then @value{GDBN} takes control.
2474 You can use any program that eventually calls @code{execve} with
2475 its arguments as a wrapper. Several standard Unix utilities do
2476 this, e.g.@: @code{env} and @code{nohup}. Any Unix shell script ending
2477 with @code{exec "$@@"} will also work.
2479 For example, you can use @code{env} to pass an environment variable to
2480 the debugged program, without setting the variable in your shell's
2484 (@value{GDBP}) set exec-wrapper env 'LD_PRELOAD=libtest.so'
2488 This command is available when debugging locally on most targets, excluding
2489 @sc{djgpp}, Cygwin, MS Windows, and QNX Neutrino.
2491 @kindex set startup-with-shell
2492 @anchor{set startup-with-shell}
2493 @item set startup-with-shell
2494 @itemx set startup-with-shell on
2495 @itemx set startup-with-shell off
2496 @itemx show startup-with-shell
2497 On Unix systems, by default, if a shell is available on your target,
2498 @value{GDBN}) uses it to start your program. Arguments of the
2499 @code{run} command are passed to the shell, which does variable
2500 substitution, expands wildcard characters and performs redirection of
2501 I/O. In some circumstances, it may be useful to disable such use of a
2502 shell, for example, when debugging the shell itself or diagnosing
2503 startup failures such as:
2507 Starting program: ./a.out
2508 During startup program terminated with signal SIGSEGV, Segmentation fault.
2512 which indicates the shell or the wrapper specified with
2513 @samp{exec-wrapper} crashed, not your program. Most often, this is
2514 caused by something odd in your shell's non-interactive mode
2515 initialization file---such as @file{.cshrc} for C-shell,
2516 $@file{.zshenv} for the Z shell, or the file specified in the
2517 @samp{BASH_ENV} environment variable for BASH.
2519 @anchor{set auto-connect-native-target}
2520 @kindex set auto-connect-native-target
2521 @item set auto-connect-native-target
2522 @itemx set auto-connect-native-target on
2523 @itemx set auto-connect-native-target off
2524 @itemx show auto-connect-native-target
2526 By default, if not connected to any target yet (e.g., with
2527 @code{target remote}), the @code{run} command starts your program as a
2528 native process under @value{GDBN}, on your local machine. If you're
2529 sure you don't want to debug programs on your local machine, you can
2530 tell @value{GDBN} to not connect to the native target automatically
2531 with the @code{set auto-connect-native-target off} command.
2533 If @code{on}, which is the default, and if @value{GDBN} is not
2534 connected to a target already, the @code{run} command automaticaly
2535 connects to the native target, if one is available.
2537 If @code{off}, and if @value{GDBN} is not connected to a target
2538 already, the @code{run} command fails with an error:
2542 Don't know how to run. Try "help target".
2545 If @value{GDBN} is already connected to a target, @value{GDBN} always
2546 uses it with the @code{run} command.
2548 In any case, you can explicitly connect to the native target with the
2549 @code{target native} command. For example,
2552 (@value{GDBP}) set auto-connect-native-target off
2554 Don't know how to run. Try "help target".
2555 (@value{GDBP}) target native
2557 Starting program: ./a.out
2558 [Inferior 1 (process 10421) exited normally]
2561 In case you connected explicitly to the @code{native} target,
2562 @value{GDBN} remains connected even if all inferiors exit, ready for
2563 the next @code{run} command. Use the @code{disconnect} command to
2566 Examples of other commands that likewise respect the
2567 @code{auto-connect-native-target} setting: @code{attach}, @code{info
2568 proc}, @code{info os}.
2570 @kindex set disable-randomization
2571 @item set disable-randomization
2572 @itemx set disable-randomization on
2573 This option (enabled by default in @value{GDBN}) will turn off the native
2574 randomization of the virtual address space of the started program. This option
2575 is useful for multiple debugging sessions to make the execution better
2576 reproducible and memory addresses reusable across debugging sessions.
2578 This feature is implemented only on certain targets, including @sc{gnu}/Linux.
2579 On @sc{gnu}/Linux you can get the same behavior using
2582 (@value{GDBP}) set exec-wrapper setarch `uname -m` -R
2585 @item set disable-randomization off
2586 Leave the behavior of the started executable unchanged. Some bugs rear their
2587 ugly heads only when the program is loaded at certain addresses. If your bug
2588 disappears when you run the program under @value{GDBN}, that might be because
2589 @value{GDBN} by default disables the address randomization on platforms, such
2590 as @sc{gnu}/Linux, which do that for stand-alone programs. Use @kbd{set
2591 disable-randomization off} to try to reproduce such elusive bugs.
2593 On targets where it is available, virtual address space randomization
2594 protects the programs against certain kinds of security attacks. In these
2595 cases the attacker needs to know the exact location of a concrete executable
2596 code. Randomizing its location makes it impossible to inject jumps misusing
2597 a code at its expected addresses.
2599 Prelinking shared libraries provides a startup performance advantage but it
2600 makes addresses in these libraries predictable for privileged processes by
2601 having just unprivileged access at the target system. Reading the shared
2602 library binary gives enough information for assembling the malicious code
2603 misusing it. Still even a prelinked shared library can get loaded at a new
2604 random address just requiring the regular relocation process during the
2605 startup. Shared libraries not already prelinked are always loaded at
2606 a randomly chosen address.
2608 Position independent executables (PIE) contain position independent code
2609 similar to the shared libraries and therefore such executables get loaded at
2610 a randomly chosen address upon startup. PIE executables always load even
2611 already prelinked shared libraries at a random address. You can build such
2612 executable using @command{gcc -fPIE -pie}.
2614 Heap (malloc storage), stack and custom mmap areas are always placed randomly
2615 (as long as the randomization is enabled).
2617 @item show disable-randomization
2618 Show the current setting of the explicit disable of the native randomization of
2619 the virtual address space of the started program.
2624 @section Your Program's Arguments
2626 @cindex arguments (to your program)
2627 The arguments to your program can be specified by the arguments of the
2629 They are passed to a shell, which expands wildcard characters and
2630 performs redirection of I/O, and thence to your program. Your
2631 @code{SHELL} environment variable (if it exists) specifies what shell
2632 @value{GDBN} uses. If you do not define @code{SHELL}, @value{GDBN} uses
2633 the default shell (@file{/bin/sh} on Unix).
2635 On non-Unix systems, the program is usually invoked directly by
2636 @value{GDBN}, which emulates I/O redirection via the appropriate system
2637 calls, and the wildcard characters are expanded by the startup code of
2638 the program, not by the shell.
2640 @code{run} with no arguments uses the same arguments used by the previous
2641 @code{run}, or those set by the @code{set args} command.
2646 Specify the arguments to be used the next time your program is run. If
2647 @code{set args} has no arguments, @code{run} executes your program
2648 with no arguments. Once you have run your program with arguments,
2649 using @code{set args} before the next @code{run} is the only way to run
2650 it again without arguments.
2654 Show the arguments to give your program when it is started.
2658 @section Your Program's Environment
2660 @cindex environment (of your program)
2661 The @dfn{environment} consists of a set of environment variables and
2662 their values. Environment variables conventionally record such things as
2663 your user name, your home directory, your terminal type, and your search
2664 path for programs to run. Usually you set up environment variables with
2665 the shell and they are inherited by all the other programs you run. When
2666 debugging, it can be useful to try running your program with a modified
2667 environment without having to start @value{GDBN} over again.
2671 @item path @var{directory}
2672 Add @var{directory} to the front of the @code{PATH} environment variable
2673 (the search path for executables) that will be passed to your program.
2674 The value of @code{PATH} used by @value{GDBN} does not change.
2675 You may specify several directory names, separated by whitespace or by a
2676 system-dependent separator character (@samp{:} on Unix, @samp{;} on
2677 MS-DOS and MS-Windows). If @var{directory} is already in the path, it
2678 is moved to the front, so it is searched sooner.
2680 You can use the string @samp{$cwd} to refer to whatever is the current
2681 working directory at the time @value{GDBN} searches the path. If you
2682 use @samp{.} instead, it refers to the directory where you executed the
2683 @code{path} command. @value{GDBN} replaces @samp{.} in the
2684 @var{directory} argument (with the current path) before adding
2685 @var{directory} to the search path.
2686 @c 'path' is explicitly nonrepeatable, but RMS points out it is silly to
2687 @c document that, since repeating it would be a no-op.
2691 Display the list of search paths for executables (the @code{PATH}
2692 environment variable).
2694 @kindex show environment
2695 @item show environment @r{[}@var{varname}@r{]}
2696 Print the value of environment variable @var{varname} to be given to
2697 your program when it starts. If you do not supply @var{varname},
2698 print the names and values of all environment variables to be given to
2699 your program. You can abbreviate @code{environment} as @code{env}.
2701 @kindex set environment
2702 @anchor{set environment}
2703 @item set environment @var{varname} @r{[}=@var{value}@r{]}
2704 Set environment variable @var{varname} to @var{value}. The value
2705 changes for your program (and the shell @value{GDBN} uses to launch
2706 it), not for @value{GDBN} itself. The @var{value} may be any string; the
2707 values of environment variables are just strings, and any
2708 interpretation is supplied by your program itself. The @var{value}
2709 parameter is optional; if it is eliminated, the variable is set to a
2711 @c "any string" here does not include leading, trailing
2712 @c blanks. Gnu asks: does anyone care?
2714 For example, this command:
2721 tells the debugged program, when subsequently run, that its user is named
2722 @samp{foo}. (The spaces around @samp{=} are used for clarity here; they
2723 are not actually required.)
2725 Note that on Unix systems, @value{GDBN} runs your program via a shell,
2726 which also inherits the environment set with @code{set environment}.
2727 If necessary, you can avoid that by using the @samp{env} program as a
2728 wrapper instead of using @code{set environment}. @xref{set
2729 exec-wrapper}, for an example doing just that.
2731 Environment variables that are set by the user are also transmitted to
2732 @command{gdbserver} to be used when starting the remote inferior.
2733 @pxref{QEnvironmentHexEncoded}.
2735 @kindex unset environment
2736 @anchor{unset environment}
2737 @item unset environment @var{varname}
2738 Remove variable @var{varname} from the environment to be passed to your
2739 program. This is different from @samp{set env @var{varname} =};
2740 @code{unset environment} removes the variable from the environment,
2741 rather than assigning it an empty value.
2743 Environment variables that are unset by the user are also unset on
2744 @command{gdbserver} when starting the remote inferior.
2745 @pxref{QEnvironmentUnset}.
2748 @emph{Warning:} On Unix systems, @value{GDBN} runs your program using
2749 the shell indicated by your @code{SHELL} environment variable if it
2750 exists (or @code{/bin/sh} if not). If your @code{SHELL} variable
2751 names a shell that runs an initialization file when started
2752 non-interactively---such as @file{.cshrc} for C-shell, $@file{.zshenv}
2753 for the Z shell, or the file specified in the @samp{BASH_ENV}
2754 environment variable for BASH---any variables you set in that file
2755 affect your program. You may wish to move setting of environment
2756 variables to files that are only run when you sign on, such as
2757 @file{.login} or @file{.profile}.
2759 @node Working Directory
2760 @section Your Program's Working Directory
2762 @cindex working directory (of your program)
2763 Each time you start your program with @code{run}, the inferior will be
2764 initialized with the current working directory specified by the
2765 @kbd{set cwd} command. If no directory has been specified by this
2766 command, then the inferior will inherit @value{GDBN}'s current working
2767 directory as its working directory if native debugging, or it will
2768 inherit the remote server's current working directory if remote
2773 @cindex change inferior's working directory
2774 @anchor{set cwd command}
2775 @item set cwd @r{[}@var{directory}@r{]}
2776 Set the inferior's working directory to @var{directory}, which will be
2777 @code{glob}-expanded in order to resolve tildes (@file{~}). If no
2778 argument has been specified, the command clears the setting and resets
2779 it to an empty state. This setting has no effect on @value{GDBN}'s
2780 working directory, and it only takes effect the next time you start
2781 the inferior. The @file{~} in @var{directory} is a short for the
2782 @dfn{home directory}, usually pointed to by the @env{HOME} environment
2783 variable. On MS-Windows, if @env{HOME} is not defined, @value{GDBN}
2784 uses the concatenation of @env{HOMEDRIVE} and @env{HOMEPATH} as
2787 You can also change @value{GDBN}'s current working directory by using
2788 the @code{cd} command.
2792 @cindex show inferior's working directory
2794 Show the inferior's working directory. If no directory has been
2795 specified by @kbd{set cwd}, then the default inferior's working
2796 directory is the same as @value{GDBN}'s working directory.
2799 @cindex change @value{GDBN}'s working directory
2801 @item cd @r{[}@var{directory}@r{]}
2802 Set the @value{GDBN} working directory to @var{directory}. If not
2803 given, @var{directory} uses @file{'~'}.
2805 The @value{GDBN} working directory serves as a default for the
2806 commands that specify files for @value{GDBN} to operate on.
2807 @xref{Files, ,Commands to Specify Files}.
2808 @xref{set cwd command}.
2812 Print the @value{GDBN} working directory.
2815 It is generally impossible to find the current working directory of
2816 the process being debugged (since a program can change its directory
2817 during its run). If you work on a system where @value{GDBN} supports
2818 the @code{info proc} command (@pxref{Process Information}), you can
2819 use the @code{info proc} command to find out the
2820 current working directory of the debuggee.
2823 @section Your Program's Input and Output
2828 By default, the program you run under @value{GDBN} does input and output to
2829 the same terminal that @value{GDBN} uses. @value{GDBN} switches the terminal
2830 to its own terminal modes to interact with you, but it records the terminal
2831 modes your program was using and switches back to them when you continue
2832 running your program.
2835 @kindex info terminal
2837 Displays information recorded by @value{GDBN} about the terminal modes your
2841 You can redirect your program's input and/or output using shell
2842 redirection with the @code{run} command. For example,
2849 starts your program, diverting its output to the file @file{outfile}.
2852 @cindex controlling terminal
2853 Another way to specify where your program should do input and output is
2854 with the @code{tty} command. This command accepts a file name as
2855 argument, and causes this file to be the default for future @code{run}
2856 commands. It also resets the controlling terminal for the child
2857 process, for future @code{run} commands. For example,
2864 directs that processes started with subsequent @code{run} commands
2865 default to do input and output on the terminal @file{/dev/ttyb} and have
2866 that as their controlling terminal.
2868 An explicit redirection in @code{run} overrides the @code{tty} command's
2869 effect on the input/output device, but not its effect on the controlling
2872 When you use the @code{tty} command or redirect input in the @code{run}
2873 command, only the input @emph{for your program} is affected. The input
2874 for @value{GDBN} still comes from your terminal. @code{tty} is an alias
2875 for @code{set inferior-tty}.
2877 @cindex inferior tty
2878 @cindex set inferior controlling terminal
2879 You can use the @code{show inferior-tty} command to tell @value{GDBN} to
2880 display the name of the terminal that will be used for future runs of your
2884 @item set inferior-tty [ @var{tty} ]
2885 @kindex set inferior-tty
2886 Set the tty for the program being debugged to @var{tty}. Omitting @var{tty}
2887 restores the default behavior, which is to use the same terminal as
2890 @item show inferior-tty
2891 @kindex show inferior-tty
2892 Show the current tty for the program being debugged.
2896 @section Debugging an Already-running Process
2901 @item attach @var{process-id}
2902 This command attaches to a running process---one that was started
2903 outside @value{GDBN}. (@code{info files} shows your active
2904 targets.) The command takes as argument a process ID. The usual way to
2905 find out the @var{process-id} of a Unix process is with the @code{ps} utility,
2906 or with the @samp{jobs -l} shell command.
2908 @code{attach} does not repeat if you press @key{RET} a second time after
2909 executing the command.
2912 To use @code{attach}, your program must be running in an environment
2913 which supports processes; for example, @code{attach} does not work for
2914 programs on bare-board targets that lack an operating system. You must
2915 also have permission to send the process a signal.
2917 When you use @code{attach}, the debugger finds the program running in
2918 the process first by looking in the current working directory, then (if
2919 the program is not found) by using the source file search path
2920 (@pxref{Source Path, ,Specifying Source Directories}). You can also use
2921 the @code{file} command to load the program. @xref{Files, ,Commands to
2924 The first thing @value{GDBN} does after arranging to debug the specified
2925 process is to stop it. You can examine and modify an attached process
2926 with all the @value{GDBN} commands that are ordinarily available when
2927 you start processes with @code{run}. You can insert breakpoints; you
2928 can step and continue; you can modify storage. If you would rather the
2929 process continue running, you may use the @code{continue} command after
2930 attaching @value{GDBN} to the process.
2935 When you have finished debugging the attached process, you can use the
2936 @code{detach} command to release it from @value{GDBN} control. Detaching
2937 the process continues its execution. After the @code{detach} command,
2938 that process and @value{GDBN} become completely independent once more, and you
2939 are ready to @code{attach} another process or start one with @code{run}.
2940 @code{detach} does not repeat if you press @key{RET} again after
2941 executing the command.
2944 If you exit @value{GDBN} while you have an attached process, you detach
2945 that process. If you use the @code{run} command, you kill that process.
2946 By default, @value{GDBN} asks for confirmation if you try to do either of these
2947 things; you can control whether or not you need to confirm by using the
2948 @code{set confirm} command (@pxref{Messages/Warnings, ,Optional Warnings and
2952 @section Killing the Child Process
2957 Kill the child process in which your program is running under @value{GDBN}.
2960 This command is useful if you wish to debug a core dump instead of a
2961 running process. @value{GDBN} ignores any core dump file while your program
2964 On some operating systems, a program cannot be executed outside @value{GDBN}
2965 while you have breakpoints set on it inside @value{GDBN}. You can use the
2966 @code{kill} command in this situation to permit running your program
2967 outside the debugger.
2969 The @code{kill} command is also useful if you wish to recompile and
2970 relink your program, since on many systems it is impossible to modify an
2971 executable file while it is running in a process. In this case, when you
2972 next type @code{run}, @value{GDBN} notices that the file has changed, and
2973 reads the symbol table again (while trying to preserve your current
2974 breakpoint settings).
2976 @node Inferiors and Programs
2977 @section Debugging Multiple Inferiors and Programs
2979 @value{GDBN} lets you run and debug multiple programs in a single
2980 session. In addition, @value{GDBN} on some systems may let you run
2981 several programs simultaneously (otherwise you have to exit from one
2982 before starting another). In the most general case, you can have
2983 multiple threads of execution in each of multiple processes, launched
2984 from multiple executables.
2987 @value{GDBN} represents the state of each program execution with an
2988 object called an @dfn{inferior}. An inferior typically corresponds to
2989 a process, but is more general and applies also to targets that do not
2990 have processes. Inferiors may be created before a process runs, and
2991 may be retained after a process exits. Inferiors have unique
2992 identifiers that are different from process ids. Usually each
2993 inferior will also have its own distinct address space, although some
2994 embedded targets may have several inferiors running in different parts
2995 of a single address space. Each inferior may in turn have multiple
2996 threads running in it.
2998 To find out what inferiors exist at any moment, use @w{@code{info
3002 @kindex info inferiors [ @var{id}@dots{} ]
3003 @item info inferiors
3004 Print a list of all inferiors currently being managed by @value{GDBN}.
3005 By default all inferiors are printed, but the argument @var{id}@dots{}
3006 -- a space separated list of inferior numbers -- can be used to limit
3007 the display to just the requested inferiors.
3009 @value{GDBN} displays for each inferior (in this order):
3013 the inferior number assigned by @value{GDBN}
3016 the target system's inferior identifier
3019 the name of the executable the inferior is running.
3024 An asterisk @samp{*} preceding the @value{GDBN} inferior number
3025 indicates the current inferior.
3029 @c end table here to get a little more width for example
3032 (@value{GDBP}) info inferiors
3033 Num Description Executable
3034 2 process 2307 hello
3035 * 1 process 3401 goodbye
3038 To switch focus between inferiors, use the @code{inferior} command:
3041 @kindex inferior @var{infno}
3042 @item inferior @var{infno}
3043 Make inferior number @var{infno} the current inferior. The argument
3044 @var{infno} is the inferior number assigned by @value{GDBN}, as shown
3045 in the first field of the @samp{info inferiors} display.
3048 @vindex $_inferior@r{, convenience variable}
3049 The debugger convenience variable @samp{$_inferior} contains the
3050 number of the current inferior. You may find this useful in writing
3051 breakpoint conditional expressions, command scripts, and so forth.
3052 @xref{Convenience Vars,, Convenience Variables}, for general
3053 information on convenience variables.
3055 You can get multiple executables into a debugging session via the
3056 @code{add-inferior} and @w{@code{clone-inferior}} commands. On some
3057 systems @value{GDBN} can add inferiors to the debug session
3058 automatically by following calls to @code{fork} and @code{exec}. To
3059 remove inferiors from the debugging session use the
3060 @w{@code{remove-inferiors}} command.
3063 @kindex add-inferior
3064 @item add-inferior [ -copies @var{n} ] [ -exec @var{executable} ]
3065 Adds @var{n} inferiors to be run using @var{executable} as the
3066 executable; @var{n} defaults to 1. If no executable is specified,
3067 the inferiors begins empty, with no program. You can still assign or
3068 change the program assigned to the inferior at any time by using the
3069 @code{file} command with the executable name as its argument.
3071 @kindex clone-inferior
3072 @item clone-inferior [ -copies @var{n} ] [ @var{infno} ]
3073 Adds @var{n} inferiors ready to execute the same program as inferior
3074 @var{infno}; @var{n} defaults to 1, and @var{infno} defaults to the
3075 number of the current inferior. This is a convenient command when you
3076 want to run another instance of the inferior you are debugging.
3079 (@value{GDBP}) info inferiors
3080 Num Description Executable
3081 * 1 process 29964 helloworld
3082 (@value{GDBP}) clone-inferior
3085 (@value{GDBP}) info inferiors
3086 Num Description Executable
3088 * 1 process 29964 helloworld
3091 You can now simply switch focus to inferior 2 and run it.
3093 @kindex remove-inferiors
3094 @item remove-inferiors @var{infno}@dots{}
3095 Removes the inferior or inferiors @var{infno}@dots{}. It is not
3096 possible to remove an inferior that is running with this command. For
3097 those, use the @code{kill} or @code{detach} command first.
3101 To quit debugging one of the running inferiors that is not the current
3102 inferior, you can either detach from it by using the @w{@code{detach
3103 inferior}} command (allowing it to run independently), or kill it
3104 using the @w{@code{kill inferiors}} command:
3107 @kindex detach inferiors @var{infno}@dots{}
3108 @item detach inferior @var{infno}@dots{}
3109 Detach from the inferior or inferiors identified by @value{GDBN}
3110 inferior number(s) @var{infno}@dots{}. Note that the inferior's entry
3111 still stays on the list of inferiors shown by @code{info inferiors},
3112 but its Description will show @samp{<null>}.
3114 @kindex kill inferiors @var{infno}@dots{}
3115 @item kill inferiors @var{infno}@dots{}
3116 Kill the inferior or inferiors identified by @value{GDBN} inferior
3117 number(s) @var{infno}@dots{}. Note that the inferior's entry still
3118 stays on the list of inferiors shown by @code{info inferiors}, but its
3119 Description will show @samp{<null>}.
3122 After the successful completion of a command such as @code{detach},
3123 @code{detach inferiors}, @code{kill} or @code{kill inferiors}, or after
3124 a normal process exit, the inferior is still valid and listed with
3125 @code{info inferiors}, ready to be restarted.
3128 To be notified when inferiors are started or exit under @value{GDBN}'s
3129 control use @w{@code{set print inferior-events}}:
3132 @kindex set print inferior-events
3133 @cindex print messages on inferior start and exit
3134 @item set print inferior-events
3135 @itemx set print inferior-events on
3136 @itemx set print inferior-events off
3137 The @code{set print inferior-events} command allows you to enable or
3138 disable printing of messages when @value{GDBN} notices that new
3139 inferiors have started or that inferiors have exited or have been
3140 detached. By default, these messages will not be printed.
3142 @kindex show print inferior-events
3143 @item show print inferior-events
3144 Show whether messages will be printed when @value{GDBN} detects that
3145 inferiors have started, exited or have been detached.
3148 Many commands will work the same with multiple programs as with a
3149 single program: e.g., @code{print myglobal} will simply display the
3150 value of @code{myglobal} in the current inferior.
3153 Occasionally, when debugging @value{GDBN} itself, it may be useful to
3154 get more info about the relationship of inferiors, programs, address
3155 spaces in a debug session. You can do that with the @w{@code{maint
3156 info program-spaces}} command.
3159 @kindex maint info program-spaces
3160 @item maint info program-spaces
3161 Print a list of all program spaces currently being managed by
3164 @value{GDBN} displays for each program space (in this order):
3168 the program space number assigned by @value{GDBN}
3171 the name of the executable loaded into the program space, with e.g.,
3172 the @code{file} command.
3177 An asterisk @samp{*} preceding the @value{GDBN} program space number
3178 indicates the current program space.
3180 In addition, below each program space line, @value{GDBN} prints extra
3181 information that isn't suitable to display in tabular form. For
3182 example, the list of inferiors bound to the program space.
3185 (@value{GDBP}) maint info program-spaces
3189 Bound inferiors: ID 1 (process 21561)
3192 Here we can see that no inferior is running the program @code{hello},
3193 while @code{process 21561} is running the program @code{goodbye}. On
3194 some targets, it is possible that multiple inferiors are bound to the
3195 same program space. The most common example is that of debugging both
3196 the parent and child processes of a @code{vfork} call. For example,
3199 (@value{GDBP}) maint info program-spaces
3202 Bound inferiors: ID 2 (process 18050), ID 1 (process 18045)
3205 Here, both inferior 2 and inferior 1 are running in the same program
3206 space as a result of inferior 1 having executed a @code{vfork} call.
3210 @section Debugging Programs with Multiple Threads
3212 @cindex threads of execution
3213 @cindex multiple threads
3214 @cindex switching threads
3215 In some operating systems, such as GNU/Linux and Solaris, a single program
3216 may have more than one @dfn{thread} of execution. The precise semantics
3217 of threads differ from one operating system to another, but in general
3218 the threads of a single program are akin to multiple processes---except
3219 that they share one address space (that is, they can all examine and
3220 modify the same variables). On the other hand, each thread has its own
3221 registers and execution stack, and perhaps private memory.
3223 @value{GDBN} provides these facilities for debugging multi-thread
3227 @item automatic notification of new threads
3228 @item @samp{thread @var{thread-id}}, a command to switch among threads
3229 @item @samp{info threads}, a command to inquire about existing threads
3230 @item @samp{thread apply @r{[}@var{thread-id-list} @r{|} all@r{]} @var{args}},
3231 a command to apply a command to a list of threads
3232 @item thread-specific breakpoints
3233 @item @samp{set print thread-events}, which controls printing of
3234 messages on thread start and exit.
3235 @item @samp{set libthread-db-search-path @var{path}}, which lets
3236 the user specify which @code{libthread_db} to use if the default choice
3237 isn't compatible with the program.
3240 @cindex focus of debugging
3241 @cindex current thread
3242 The @value{GDBN} thread debugging facility allows you to observe all
3243 threads while your program runs---but whenever @value{GDBN} takes
3244 control, one thread in particular is always the focus of debugging.
3245 This thread is called the @dfn{current thread}. Debugging commands show
3246 program information from the perspective of the current thread.
3248 @cindex @code{New} @var{systag} message
3249 @cindex thread identifier (system)
3250 @anchor{target system thread identifier}
3251 @c FIXME-implementors!! It would be more helpful if the [New...] message
3252 @c included GDB's numeric thread handle, so you could just go to that
3253 @c thread without first checking `info threads'.
3254 Whenever @value{GDBN} detects a new thread in your program, it displays
3255 the target system's identification for the thread with a message in the
3256 form @samp{[New @var{systag}]}, where @var{systag} is a thread identifier
3257 whose form varies depending on the particular system. For example, on
3258 @sc{gnu}/Linux, you might see
3261 [New Thread 0x41e02940 (LWP 25582)]
3265 when @value{GDBN} notices a new thread. In contrast, on other systems,
3266 the @var{systag} is simply something like @samp{process 368}, with no
3269 @c FIXME!! (1) Does the [New...] message appear even for the very first
3270 @c thread of a program, or does it only appear for the
3271 @c second---i.e.@: when it becomes obvious we have a multithread
3273 @c (2) *Is* there necessarily a first thread always? Or do some
3274 @c multithread systems permit starting a program with multiple
3275 @c threads ab initio?
3277 @anchor{thread numbers}
3278 @cindex thread number, per inferior
3279 @cindex thread identifier (GDB)
3280 For debugging purposes, @value{GDBN} associates its own thread number
3281 ---always a single integer---with each thread of an inferior. This
3282 number is unique between all threads of an inferior, but not unique
3283 between threads of different inferiors.
3285 @cindex qualified thread ID
3286 You can refer to a given thread in an inferior using the qualified
3287 @var{inferior-num}.@var{thread-num} syntax, also known as
3288 @dfn{qualified thread ID}, with @var{inferior-num} being the inferior
3289 number and @var{thread-num} being the thread number of the given
3290 inferior. For example, thread @code{2.3} refers to thread number 3 of
3291 inferior 2. If you omit @var{inferior-num} (e.g., @code{thread 3}),
3292 then @value{GDBN} infers you're referring to a thread of the current
3295 Until you create a second inferior, @value{GDBN} does not show the
3296 @var{inferior-num} part of thread IDs, even though you can always use
3297 the full @var{inferior-num}.@var{thread-num} form to refer to threads
3298 of inferior 1, the initial inferior.
3300 @anchor{thread ID list}
3301 @cindex thread ID list
3302 Some commands accept a space-separated @dfn{thread ID list} as
3303 argument. A list element can be:
3307 A thread ID as shown in the first field of the @samp{info threads}
3308 display, with or without an inferior qualifier. E.g., @samp{2.1} or
3312 A range of thread numbers, again with or without an inferior
3313 qualifier, as in @var{inf}.@var{thr1}-@var{thr2} or
3314 @var{thr1}-@var{thr2}. E.g., @samp{1.2-4} or @samp{2-4}.
3317 All threads of an inferior, specified with a star wildcard, with or
3318 without an inferior qualifier, as in @var{inf}.@code{*} (e.g.,
3319 @samp{1.*}) or @code{*}. The former refers to all threads of the
3320 given inferior, and the latter form without an inferior qualifier
3321 refers to all threads of the current inferior.
3325 For example, if the current inferior is 1, and inferior 7 has one
3326 thread with ID 7.1, the thread list @samp{1 2-3 4.5 6.7-9 7.*}
3327 includes threads 1 to 3 of inferior 1, thread 5 of inferior 4, threads
3328 7 to 9 of inferior 6 and all threads of inferior 7. That is, in
3329 expanded qualified form, the same as @samp{1.1 1.2 1.3 4.5 6.7 6.8 6.9
3333 @anchor{global thread number}
3334 @cindex global thread number
3335 @cindex global thread identifier (GDB)
3336 In addition to a @emph{per-inferior} number, each thread is also
3337 assigned a unique @emph{global} number, also known as @dfn{global
3338 thread ID}, a single integer. Unlike the thread number component of
3339 the thread ID, no two threads have the same global ID, even when
3340 you're debugging multiple inferiors.
3342 From @value{GDBN}'s perspective, a process always has at least one
3343 thread. In other words, @value{GDBN} assigns a thread number to the
3344 program's ``main thread'' even if the program is not multi-threaded.
3346 @vindex $_thread@r{, convenience variable}
3347 @vindex $_gthread@r{, convenience variable}
3348 @vindex $_thread_systag@r{, convenience variable}
3349 @vindex $_thread_name@r{, convenience variable}
3350 The debugger convenience variables @samp{$_thread}, @samp{$_gthread},
3351 @samp{$_thread_systag}, and @samp{$_thread_name} contain,
3352 respectively, the per-inferior thread number, the global thread
3353 number, the target system's thread identifier (@var{systag}) string,
3354 and the thread name string of the current thread. You may find this
3355 useful in writing breakpoint conditional expressions, command scripts,
3356 and so forth. @xref{Convenience Vars,, Convenience Variables}, for
3357 general information on convenience variables.
3359 If @value{GDBN} detects the program is multi-threaded, it augments the
3360 usual message about stopping at a breakpoint with the ID and name of
3361 the thread that hit the breakpoint.
3364 Thread 2 "client" hit Breakpoint 1, send_message () at client.c:68
3367 Likewise when the program receives a signal:
3370 Thread 1 "main" received signal SIGINT, Interrupt.
3374 @kindex info threads
3375 @item info threads @r{[}-gid@r{]} @r{[}@var{thread-id-list}@r{]}
3377 Display information about one or more threads. With no arguments
3378 displays information about all threads. You can specify the list of
3379 threads that you want to display using the thread ID list syntax
3380 (@pxref{thread ID list}).
3382 @value{GDBN} displays for each thread (in this order):
3386 the per-inferior thread number assigned by @value{GDBN}
3389 the global thread number assigned by @value{GDBN}, if the
3390 @w{@option{-gid}} option was specified
3393 the target system's thread identifier (@var{systag})
3396 the thread's name, if one is known. A thread can either be named by
3397 the user (see @code{thread name}, below), or, in some cases, by the
3401 the current stack frame summary for that thread
3405 An asterisk @samp{*} to the left of the @value{GDBN} thread number
3406 indicates the current thread.
3410 @c end table here to get a little more width for example
3413 (@value{GDBP}) info threads
3415 * 1 process 35 thread 13 main (argc=1, argv=0x7ffffff8)
3416 2 process 35 thread 23 0x34e5 in sigpause ()
3417 3 process 35 thread 27 0x34e5 in sigpause ()
3420 If you're debugging multiple inferiors, @value{GDBN} displays thread
3421 IDs using the qualified @var{inferior-num}.@var{thread-num} format.
3422 Otherwise, only @var{thread-num} is shown.
3424 If you specify the @w{@option{-gid}} option, @value{GDBN} displays a
3425 column indicating each thread's global thread ID:
3428 (@value{GDBP}) info threads -gid
3429 Id GId Target Id Frame
3430 1.1 1 process 35 thread 13 main (argc=1, argv=0x7ffffff8)
3431 1.2 3 process 35 thread 23 0x34e5 in sigpause ()
3432 1.3 4 process 35 thread 27 0x34e5 in sigpause ()
3433 * 2.1 2 process 65 thread 1 main (argc=1, argv=0x7ffffff8)
3436 On Solaris, you can display more information about user threads with a
3437 Solaris-specific command:
3440 @item maint info sol-threads
3441 @kindex maint info sol-threads
3442 @cindex thread info (Solaris)
3443 Display info on Solaris user threads.
3447 @kindex thread @r{[}-gid@r{]} @var{thread-id}
3448 @item thread @r{[}-gid@r{]} @var{thread-id}
3449 Make thread ID @var{thread-id} the current thread. The command
3450 argument @var{thread-id} is the @value{GDBN} thread ID: if the
3451 @w{@option{-gid}} option is given it is a global thread identifier, as
3452 shown in the second field of the @samp{info threads -gid} display;
3453 otherwise it is a per process thread identifier, with or without an
3454 inferior qualifier (e.g., @samp{2.1} or @samp{1}), as shown in the
3455 first field of the @samp{info threads} display.
3457 @value{GDBN} responds by displaying the system identifier of the
3458 thread you selected, and its current stack frame summary:
3461 (@value{GDBP}) thread 2
3462 [Switching to thread 2 (Thread 0xb7fdab70 (LWP 12747))]
3463 #0 some_function (ignore=0x0) at example.c:8
3464 8 printf ("hello\n");
3468 As with the @samp{[New @dots{}]} message, the form of the text after
3469 @samp{Switching to} depends on your system's conventions for identifying
3472 @anchor{thread apply all}
3473 @kindex thread apply
3474 @cindex apply command to several threads
3475 @item thread apply @r{[}@var{thread-id-list} @r{|} all @r{[}-ascending@r{]]} @r{[}@var{flag}@r{]@dots{}} @var{command}
3476 The @code{thread apply} command allows you to apply the named
3477 @var{command} to one or more threads. Specify the threads that you
3478 want affected using the thread ID list syntax (@pxref{thread ID
3479 list}), or specify @code{all} to apply to all threads. To apply a
3480 command to all threads in descending order, type @kbd{thread apply all
3481 @var{command}}. To apply a command to all threads in ascending order,
3482 type @kbd{thread apply all -ascending @var{command}}.
3484 The @var{flag} arguments control what output to produce and how to handle
3485 errors raised when applying @var{command} to a thread. @var{flag}
3486 must start with a @code{-} directly followed by one letter in
3487 @code{qcs}. If several flags are provided, they must be given
3488 individually, such as @code{-c -q}.
3490 By default, @value{GDBN} displays some thread information before the
3491 output produced by @var{command}, and an error raised during the
3492 execution of a @var{command} will abort @code{thread apply}. The
3493 following flags can be used to fine-tune this behavior:
3497 The flag @code{-c}, which stands for @samp{continue}, causes any
3498 errors in @var{command} to be displayed, and the execution of
3499 @code{thread apply} then continues.
3501 The flag @code{-s}, which stands for @samp{silent}, causes any errors
3502 or empty output produced by a @var{command} to be silently ignored.
3503 That is, the execution continues, but the thread information and errors
3506 The flag @code{-q} (@samp{quiet}) disables printing the thread
3510 Flags @code{-c} and @code{-s} cannot be used together.
3513 @cindex apply command to all threads (ignoring errors and empty output)
3514 @item taas [@var{option}]@dots{} @var{command}
3515 Shortcut for @code{thread apply all -s @r{[}@var{option}@r{]@dots{}} @var{command}}.
3516 Applies @var{command} on all threads, ignoring errors and empty output.
3518 The @code{taas} command accepts the same options as the @code{thread
3519 apply all} command. @xref{thread apply all}.
3522 @cindex apply a command to all frames of all threads (ignoring errors and empty output)
3523 @item tfaas [@var{option}]@dots{} @var{command}
3524 Shortcut for @code{thread apply all -s -- frame apply all -s @r{[}@var{option}@r{]@dots{}} @var{command}}.
3525 Applies @var{command} on all frames of all threads, ignoring errors
3526 and empty output. Note that the flag @code{-s} is specified twice:
3527 The first @code{-s} ensures that @code{thread apply} only shows the thread
3528 information of the threads for which @code{frame apply} produces
3529 some output. The second @code{-s} is needed to ensure that @code{frame
3530 apply} shows the frame information of a frame only if the
3531 @var{command} successfully produced some output.
3533 It can for example be used to print a local variable or a function
3534 argument without knowing the thread or frame where this variable or argument
3537 (@value{GDBP}) tfaas p some_local_var_i_do_not_remember_where_it_is
3540 The @code{tfaas} command accepts the same options as the @code{frame
3541 apply} command. @xref{frame apply}.
3544 @cindex name a thread
3545 @anchor{thread name}
3546 @item thread name [@var{name}]
3547 This command assigns a name to the current thread. If no argument is
3548 given, any existing user-specified name is removed. The thread name
3549 appears in the @samp{info threads} display.
3551 On some systems, such as @sc{gnu}/Linux, @value{GDBN} is able to
3552 determine the name of the thread as given by the OS. On these
3553 systems, a name specified with @samp{thread name} will override the
3554 system-give name, and removing the user-specified name will cause
3555 @value{GDBN} to once again display the system-specified name.
3558 @cindex search for a thread
3559 @anchor{thread find}
3560 @item thread find [@var{regexp}]
3561 Search for and display thread ids whose name or @var{systag}
3562 matches the supplied regular expression. The syntax of the regular
3563 expression is that specified by Python's regular expression support.
3565 As well as being the complement to the @samp{thread name} command,
3566 this command also allows you to identify a thread by its target
3567 @var{systag}. For instance, on @sc{gnu}/Linux, the target @var{systag}
3571 (@value{GDBP}) thread find 26688
3572 Thread 4 has target id 'Thread 0x41e02940 (LWP 26688)'
3573 (@value{GDBP}) info thread 4
3575 4 Thread 0x41e02940 (LWP 26688) 0x00000031ca6cd372 in select ()
3578 @kindex set print thread-events
3579 @cindex print messages on thread start and exit
3580 @item set print thread-events
3581 @itemx set print thread-events on
3582 @itemx set print thread-events off
3583 The @code{set print thread-events} command allows you to enable or
3584 disable printing of messages when @value{GDBN} notices that new threads have
3585 started or that threads have exited. By default, these messages will
3586 be printed if detection of these events is supported by the target.
3587 Note that these messages cannot be disabled on all targets.
3589 @kindex show print thread-events
3590 @item show print thread-events
3591 Show whether messages will be printed when @value{GDBN} detects that threads
3592 have started and exited.
3595 @xref{Thread Stops,,Stopping and Starting Multi-thread Programs}, for
3596 more information about how @value{GDBN} behaves when you stop and start
3597 programs with multiple threads.
3599 @xref{Set Watchpoints,,Setting Watchpoints}, for information about
3600 watchpoints in programs with multiple threads.
3602 @anchor{set libthread-db-search-path}
3604 @kindex set libthread-db-search-path
3605 @cindex search path for @code{libthread_db}
3606 @item set libthread-db-search-path @r{[}@var{path}@r{]}
3607 If this variable is set, @var{path} is a colon-separated list of
3608 directories @value{GDBN} will use to search for @code{libthread_db}.
3609 If you omit @var{path}, @samp{libthread-db-search-path} will be reset to
3610 its default value (@code{$sdir:$pdir} on @sc{gnu}/Linux and Solaris systems).
3611 Internally, the default value comes from the @code{LIBTHREAD_DB_SEARCH_PATH}
3614 On @sc{gnu}/Linux and Solaris systems, @value{GDBN} uses a ``helper''
3615 @code{libthread_db} library to obtain information about threads in the
3616 inferior process. @value{GDBN} will use @samp{libthread-db-search-path}
3617 to find @code{libthread_db}. @value{GDBN} also consults first if inferior
3618 specific thread debugging library loading is enabled
3619 by @samp{set auto-load libthread-db} (@pxref{libthread_db.so.1 file}).
3621 A special entry @samp{$sdir} for @samp{libthread-db-search-path}
3622 refers to the default system directories that are
3623 normally searched for loading shared libraries. The @samp{$sdir} entry
3624 is the only kind not needing to be enabled by @samp{set auto-load libthread-db}
3625 (@pxref{libthread_db.so.1 file}).
3627 A special entry @samp{$pdir} for @samp{libthread-db-search-path}
3628 refers to the directory from which @code{libpthread}
3629 was loaded in the inferior process.
3631 For any @code{libthread_db} library @value{GDBN} finds in above directories,
3632 @value{GDBN} attempts to initialize it with the current inferior process.
3633 If this initialization fails (which could happen because of a version
3634 mismatch between @code{libthread_db} and @code{libpthread}), @value{GDBN}
3635 will unload @code{libthread_db}, and continue with the next directory.
3636 If none of @code{libthread_db} libraries initialize successfully,
3637 @value{GDBN} will issue a warning and thread debugging will be disabled.
3639 Setting @code{libthread-db-search-path} is currently implemented
3640 only on some platforms.
3642 @kindex show libthread-db-search-path
3643 @item show libthread-db-search-path
3644 Display current libthread_db search path.
3646 @kindex set debug libthread-db
3647 @kindex show debug libthread-db
3648 @cindex debugging @code{libthread_db}
3649 @item set debug libthread-db
3650 @itemx show debug libthread-db
3651 Turns on or off display of @code{libthread_db}-related events.
3652 Use @code{1} to enable, @code{0} to disable.
3655 @xref{Heterogeneous Debugging} for additional information related to
3656 threads in heterogeneous systems.
3659 @section Debugging Forks
3661 @cindex fork, debugging programs which call
3662 @cindex multiple processes
3663 @cindex processes, multiple
3664 On most systems, @value{GDBN} has no special support for debugging
3665 programs which create additional processes using the @code{fork}
3666 function. When a program forks, @value{GDBN} will continue to debug the
3667 parent process and the child process will run unimpeded. If you have
3668 set a breakpoint in any code which the child then executes, the child
3669 will get a @code{SIGTRAP} signal which (unless it catches the signal)
3670 will cause it to terminate.
3672 However, if you want to debug the child process there is a workaround
3673 which isn't too painful. Put a call to @code{sleep} in the code which
3674 the child process executes after the fork. It may be useful to sleep
3675 only if a certain environment variable is set, or a certain file exists,
3676 so that the delay need not occur when you don't want to run @value{GDBN}
3677 on the child. While the child is sleeping, use the @code{ps} program to
3678 get its process ID. Then tell @value{GDBN} (a new invocation of
3679 @value{GDBN} if you are also debugging the parent process) to attach to
3680 the child process (@pxref{Attach}). From that point on you can debug
3681 the child process just like any other process which you attached to.
3683 On some systems, @value{GDBN} provides support for debugging programs
3684 that create additional processes using the @code{fork} or @code{vfork}
3685 functions. On @sc{gnu}/Linux platforms, this feature is supported
3686 with kernel version 2.5.46 and later.
3688 The fork debugging commands are supported in native mode and when
3689 connected to @code{gdbserver} in either @code{target remote} mode or
3690 @code{target extended-remote} mode.
3692 By default, when a program forks, @value{GDBN} will continue to debug
3693 the parent process and the child process will run unimpeded.
3695 If you want to follow the child process instead of the parent process,
3696 use the command @w{@code{set follow-fork-mode}}.
3699 @kindex set follow-fork-mode
3700 @item set follow-fork-mode @var{mode}
3701 Set the debugger response to a program call of @code{fork} or
3702 @code{vfork}. A call to @code{fork} or @code{vfork} creates a new
3703 process. The @var{mode} argument can be:
3707 The original process is debugged after a fork. The child process runs
3708 unimpeded. This is the default.
3711 The new process is debugged after a fork. The parent process runs
3716 @kindex show follow-fork-mode
3717 @item show follow-fork-mode
3718 Display the current debugger response to a @code{fork} or @code{vfork} call.
3721 @cindex debugging multiple processes
3722 On Linux, if you want to debug both the parent and child processes, use the
3723 command @w{@code{set detach-on-fork}}.
3726 @kindex set detach-on-fork
3727 @item set detach-on-fork @var{mode}
3728 Tells gdb whether to detach one of the processes after a fork, or
3729 retain debugger control over them both.
3733 The child process (or parent process, depending on the value of
3734 @code{follow-fork-mode}) will be detached and allowed to run
3735 independently. This is the default.
3738 Both processes will be held under the control of @value{GDBN}.
3739 One process (child or parent, depending on the value of
3740 @code{follow-fork-mode}) is debugged as usual, while the other
3745 @kindex show detach-on-fork
3746 @item show detach-on-fork
3747 Show whether detach-on-fork mode is on/off.
3750 If you choose to set @samp{detach-on-fork} mode off, then @value{GDBN}
3751 will retain control of all forked processes (including nested forks).
3752 You can list the forked processes under the control of @value{GDBN} by
3753 using the @w{@code{info inferiors}} command, and switch from one fork
3754 to another by using the @code{inferior} command (@pxref{Inferiors and
3755 Programs, ,Debugging Multiple Inferiors and Programs}).
3757 To quit debugging one of the forked processes, you can either detach
3758 from it by using the @w{@code{detach inferiors}} command (allowing it
3759 to run independently), or kill it using the @w{@code{kill inferiors}}
3760 command. @xref{Inferiors and Programs, ,Debugging Multiple Inferiors
3763 If you ask to debug a child process and a @code{vfork} is followed by an
3764 @code{exec}, @value{GDBN} executes the new target up to the first
3765 breakpoint in the new target. If you have a breakpoint set on
3766 @code{main} in your original program, the breakpoint will also be set on
3767 the child process's @code{main}.
3769 On some systems, when a child process is spawned by @code{vfork}, you
3770 cannot debug the child or parent until an @code{exec} call completes.
3772 If you issue a @code{run} command to @value{GDBN} after an @code{exec}
3773 call executes, the new target restarts. To restart the parent
3774 process, use the @code{file} command with the parent executable name
3775 as its argument. By default, after an @code{exec} call executes,
3776 @value{GDBN} discards the symbols of the previous executable image.
3777 You can change this behaviour with the @w{@code{set follow-exec-mode}}
3781 @kindex set follow-exec-mode
3782 @item set follow-exec-mode @var{mode}
3784 Set debugger response to a program call of @code{exec}. An
3785 @code{exec} call replaces the program image of a process.
3787 @code{follow-exec-mode} can be:
3791 @value{GDBN} creates a new inferior and rebinds the process to this
3792 new inferior. The program the process was running before the
3793 @code{exec} call can be restarted afterwards by restarting the
3799 (@value{GDBP}) info inferiors
3800 (@value{GDBP}) info inferior
3801 Id Description Executable
3804 process 12020 is executing new program: prog2
3805 Program exited normally.
3806 (@value{GDBP}) info inferiors
3807 Id Description Executable
3813 @value{GDBN} keeps the process bound to the same inferior. The new
3814 executable image replaces the previous executable loaded in the
3815 inferior. Restarting the inferior after the @code{exec} call, with
3816 e.g., the @code{run} command, restarts the executable the process was
3817 running after the @code{exec} call. This is the default mode.
3822 (@value{GDBP}) info inferiors
3823 Id Description Executable
3826 process 12020 is executing new program: prog2
3827 Program exited normally.
3828 (@value{GDBP}) info inferiors
3829 Id Description Executable
3836 @code{follow-exec-mode} is supported in native mode and
3837 @code{target extended-remote} mode.
3839 You can use the @code{catch} command to make @value{GDBN} stop whenever
3840 a @code{fork}, @code{vfork}, or @code{exec} call is made. @xref{Set
3841 Catchpoints, ,Setting Catchpoints}.
3843 @node Checkpoint/Restart
3844 @section Setting a @emph{Bookmark} to Return to Later
3849 @cindex snapshot of a process
3850 @cindex rewind program state
3852 On certain operating systems@footnote{Currently, only
3853 @sc{gnu}/Linux.}, @value{GDBN} is able to save a @dfn{snapshot} of a
3854 program's state, called a @dfn{checkpoint}, and come back to it
3857 Returning to a checkpoint effectively undoes everything that has
3858 happened in the program since the @code{checkpoint} was saved. This
3859 includes changes in memory, registers, and even (within some limits)
3860 system state. Effectively, it is like going back in time to the
3861 moment when the checkpoint was saved.
3863 Thus, if you're stepping thru a program and you think you're
3864 getting close to the point where things go wrong, you can save
3865 a checkpoint. Then, if you accidentally go too far and miss
3866 the critical statement, instead of having to restart your program
3867 from the beginning, you can just go back to the checkpoint and
3868 start again from there.
3870 This can be especially useful if it takes a lot of time or
3871 steps to reach the point where you think the bug occurs.
3873 To use the @code{checkpoint}/@code{restart} method of debugging:
3878 Save a snapshot of the debugged program's current execution state.
3879 The @code{checkpoint} command takes no arguments, but each checkpoint
3880 is assigned a small integer id, similar to a breakpoint id.
3882 @kindex info checkpoints
3883 @item info checkpoints
3884 List the checkpoints that have been saved in the current debugging
3885 session. For each checkpoint, the following information will be
3892 @item Source line, or label
3895 @kindex restart @var{checkpoint-id}
3896 @item restart @var{checkpoint-id}
3897 Restore the program state that was saved as checkpoint number
3898 @var{checkpoint-id}. All program variables, registers, stack frames
3899 etc.@: will be returned to the values that they had when the checkpoint
3900 was saved. In essence, gdb will ``wind back the clock'' to the point
3901 in time when the checkpoint was saved.
3903 Note that breakpoints, @value{GDBN} variables, command history etc.
3904 are not affected by restoring a checkpoint. In general, a checkpoint
3905 only restores things that reside in the program being debugged, not in
3908 @kindex delete checkpoint @var{checkpoint-id}
3909 @item delete checkpoint @var{checkpoint-id}
3910 Delete the previously-saved checkpoint identified by @var{checkpoint-id}.
3914 Returning to a previously saved checkpoint will restore the user state
3915 of the program being debugged, plus a significant subset of the system
3916 (OS) state, including file pointers. It won't ``un-write'' data from
3917 a file, but it will rewind the file pointer to the previous location,
3918 so that the previously written data can be overwritten. For files
3919 opened in read mode, the pointer will also be restored so that the
3920 previously read data can be read again.
3922 Of course, characters that have been sent to a printer (or other
3923 external device) cannot be ``snatched back'', and characters received
3924 from eg.@: a serial device can be removed from internal program buffers,
3925 but they cannot be ``pushed back'' into the serial pipeline, ready to
3926 be received again. Similarly, the actual contents of files that have
3927 been changed cannot be restored (at this time).
3929 However, within those constraints, you actually can ``rewind'' your
3930 program to a previously saved point in time, and begin debugging it
3931 again --- and you can change the course of events so as to debug a
3932 different execution path this time.
3934 @cindex checkpoints and process id
3935 Finally, there is one bit of internal program state that will be
3936 different when you return to a checkpoint --- the program's process
3937 id. Each checkpoint will have a unique process id (or @var{pid}),
3938 and each will be different from the program's original @var{pid}.
3939 If your program has saved a local copy of its process id, this could
3940 potentially pose a problem.
3942 @subsection A Non-obvious Benefit of Using Checkpoints
3944 On some systems such as @sc{gnu}/Linux, address space randomization
3945 is performed on new processes for security reasons. This makes it
3946 difficult or impossible to set a breakpoint, or watchpoint, on an
3947 absolute address if you have to restart the program, since the
3948 absolute location of a symbol will change from one execution to the
3951 A checkpoint, however, is an @emph{identical} copy of a process.
3952 Therefore if you create a checkpoint at (eg.@:) the start of main,
3953 and simply return to that checkpoint instead of restarting the
3954 process, you can avoid the effects of address randomization and
3955 your symbols will all stay in the same place.
3958 @chapter Stopping and Continuing
3960 The principal purposes of using a debugger are so that you can stop your
3961 program before it terminates; or so that, if your program runs into
3962 trouble, you can investigate and find out why.
3964 Inside @value{GDBN}, your program may stop for any of several reasons,
3965 such as a signal, a breakpoint, or reaching a new line after a
3966 @value{GDBN} command such as @code{step}. You may then examine and
3967 change variables, set new breakpoints or remove old ones, and then
3968 continue execution. Usually, the messages shown by @value{GDBN} provide
3969 ample explanation of the status of your program---but you can also
3970 explicitly request this information at any time.
3973 @kindex info program
3975 Display information about the status of your program: whether it is
3976 running or not, what process it is, and why it stopped.
3980 * Breakpoints:: Breakpoints, watchpoints, and catchpoints
3981 * Continuing and Stepping:: Resuming execution
3982 * Skipping Over Functions and Files::
3983 Skipping over functions and files
3985 * Thread Stops:: Stopping and starting multi-thread programs
3989 @section Breakpoints, Watchpoints, and Catchpoints
3992 A @dfn{breakpoint} makes your program stop whenever a certain point in
3993 the program is reached. For each breakpoint, you can add conditions to
3994 control in finer detail whether your program stops. You can set
3995 breakpoints with the @code{break} command and its variants (@pxref{Set
3996 Breaks, ,Setting Breakpoints}), to specify the place where your program
3997 should stop by line number, function name or exact address in the
4000 On some systems, you can set breakpoints in shared libraries before
4001 the executable is run.
4004 @cindex data breakpoints
4005 @cindex memory tracing
4006 @cindex breakpoint on memory address
4007 @cindex breakpoint on variable modification
4008 A @dfn{watchpoint} is a special breakpoint that stops your program
4009 when the value of an expression changes. The expression may be a value
4010 of a variable, or it could involve values of one or more variables
4011 combined by operators, such as @samp{a + b}. This is sometimes called
4012 @dfn{data breakpoints}. You must use a different command to set
4013 watchpoints (@pxref{Set Watchpoints, ,Setting Watchpoints}), but aside
4014 from that, you can manage a watchpoint like any other breakpoint: you
4015 enable, disable, and delete both breakpoints and watchpoints using the
4018 You can arrange to have values from your program displayed automatically
4019 whenever @value{GDBN} stops at a breakpoint. @xref{Auto Display,,
4023 @cindex breakpoint on events
4024 A @dfn{catchpoint} is another special breakpoint that stops your program
4025 when a certain kind of event occurs, such as the throwing of a C@t{++}
4026 exception or the loading of a library. As with watchpoints, you use a
4027 different command to set a catchpoint (@pxref{Set Catchpoints, ,Setting
4028 Catchpoints}), but aside from that, you can manage a catchpoint like any
4029 other breakpoint. (To stop when your program receives a signal, use the
4030 @code{handle} command; see @ref{Signals, ,Signals}.)
4032 @cindex breakpoint numbers
4033 @cindex numbers for breakpoints
4034 @value{GDBN} assigns a number to each breakpoint, watchpoint, or
4035 catchpoint when you create it; these numbers are successive integers
4036 starting with one. In many of the commands for controlling various
4037 features of breakpoints you use the breakpoint number to say which
4038 breakpoint you want to change. Each breakpoint may be @dfn{enabled} or
4039 @dfn{disabled}; if disabled, it has no effect on your program until you
4042 @cindex breakpoint ranges
4043 @cindex breakpoint lists
4044 @cindex ranges of breakpoints
4045 @cindex lists of breakpoints
4046 Some @value{GDBN} commands accept a space-separated list of breakpoints
4047 on which to operate. A list element can be either a single breakpoint number,
4048 like @samp{5}, or a range of such numbers, like @samp{5-7}.
4049 When a breakpoint list is given to a command, all breakpoints in that list
4053 * Set Breaks:: Setting breakpoints
4054 * Set Watchpoints:: Setting watchpoints
4055 * Set Catchpoints:: Setting catchpoints
4056 * Delete Breaks:: Deleting breakpoints
4057 * Disabling:: Disabling breakpoints
4058 * Conditions:: Break conditions
4059 * Break Commands:: Breakpoint command lists
4060 * Dynamic Printf:: Dynamic printf
4061 * Save Breakpoints:: How to save breakpoints in a file
4062 * Static Probe Points:: Listing static probe points
4063 * Error in Breakpoints:: ``Cannot insert breakpoints''
4064 * Breakpoint-related Warnings:: ``Breakpoint address adjusted...''
4068 @subsection Setting Breakpoints
4070 @c FIXME LMB what does GDB do if no code on line of breakpt?
4071 @c consider in particular declaration with/without initialization.
4073 @c FIXME 2 is there stuff on this already? break at fun start, already init?
4076 @kindex b @r{(@code{break})}
4077 @vindex $bpnum@r{, convenience variable}
4078 @cindex latest breakpoint
4079 Breakpoints are set with the @code{break} command (abbreviated
4080 @code{b}). The debugger convenience variable @samp{$bpnum} records the
4081 number of the breakpoint you've set most recently; see @ref{Convenience
4082 Vars,, Convenience Variables}, for a discussion of what you can do with
4083 convenience variables.
4086 @item break @var{location}
4087 Set a breakpoint at the given @var{location}, which can specify a
4088 function name, a line number, or an address of an instruction.
4089 (@xref{Specify Location}, for a list of all the possible ways to
4090 specify a @var{location}.) The breakpoint will stop your program just
4091 before it executes any of the code in the specified @var{location}.
4093 When using source languages that permit overloading of symbols, such as
4094 C@t{++}, a function name may refer to more than one possible place to break.
4095 @xref{Ambiguous Expressions,,Ambiguous Expressions}, for a discussion of
4098 It is also possible to insert a breakpoint that will stop the program
4099 only if a specific thread (@pxref{Thread-Specific Breakpoints})
4100 or a specific task (@pxref{Ada Tasks}) hits that breakpoint.
4103 When called without any arguments, @code{break} sets a breakpoint at
4104 the next instruction to be executed in the selected stack frame
4105 (@pxref{Stack, ,Examining the Stack}). In any selected frame but the
4106 innermost, this makes your program stop as soon as control
4107 returns to that frame. This is similar to the effect of a
4108 @code{finish} command in the frame inside the selected frame---except
4109 that @code{finish} does not leave an active breakpoint. If you use
4110 @code{break} without an argument in the innermost frame, @value{GDBN} stops
4111 the next time it reaches the current location; this may be useful
4114 @value{GDBN} normally ignores breakpoints when it resumes execution, until at
4115 least one instruction has been executed. If it did not do this, you
4116 would be unable to proceed past a breakpoint without first disabling the
4117 breakpoint. This rule applies whether or not the breakpoint already
4118 existed when your program stopped.
4120 @item break @dots{} if @var{cond}
4121 Set a breakpoint with condition @var{cond}; evaluate the expression
4122 @var{cond} each time the breakpoint is reached, and stop only if the
4123 value is nonzero---that is, if @var{cond} evaluates as true.
4124 @samp{@dots{}} stands for one of the possible arguments described
4125 above (or no argument) specifying where to break. @xref{Conditions,
4126 ,Break Conditions}, for more information on breakpoint conditions.
4129 @item tbreak @var{args}
4130 Set a breakpoint enabled only for one stop. The @var{args} are the
4131 same as for the @code{break} command, and the breakpoint is set in the same
4132 way, but the breakpoint is automatically deleted after the first time your
4133 program stops there. @xref{Disabling, ,Disabling Breakpoints}.
4136 @cindex hardware breakpoints
4137 @item hbreak @var{args}
4138 Set a hardware-assisted breakpoint. The @var{args} are the same as for the
4139 @code{break} command and the breakpoint is set in the same way, but the
4140 breakpoint requires hardware support and some target hardware may not
4141 have this support. The main purpose of this is EPROM/ROM code
4142 debugging, so you can set a breakpoint at an instruction without
4143 changing the instruction. This can be used with the new trap-generation
4144 provided by SPARClite DSU and most x86-based targets. These targets
4145 will generate traps when a program accesses some data or instruction
4146 address that is assigned to the debug registers. However the hardware
4147 breakpoint registers can take a limited number of breakpoints. For
4148 example, on the DSU, only two data breakpoints can be set at a time, and
4149 @value{GDBN} will reject this command if more than two are used. Delete
4150 or disable unused hardware breakpoints before setting new ones
4151 (@pxref{Disabling, ,Disabling Breakpoints}).
4152 @xref{Conditions, ,Break Conditions}.
4153 For remote targets, you can restrict the number of hardware
4154 breakpoints @value{GDBN} will use, see @ref{set remote
4155 hardware-breakpoint-limit}.
4158 @item thbreak @var{args}
4159 Set a hardware-assisted breakpoint enabled only for one stop. The @var{args}
4160 are the same as for the @code{hbreak} command and the breakpoint is set in
4161 the same way. However, like the @code{tbreak} command,
4162 the breakpoint is automatically deleted after the
4163 first time your program stops there. Also, like the @code{hbreak}
4164 command, the breakpoint requires hardware support and some target hardware
4165 may not have this support. @xref{Disabling, ,Disabling Breakpoints}.
4166 See also @ref{Conditions, ,Break Conditions}.
4169 @cindex regular expression
4170 @cindex breakpoints at functions matching a regexp
4171 @cindex set breakpoints in many functions
4172 @item rbreak @var{regex}
4173 Set breakpoints on all functions matching the regular expression
4174 @var{regex}. This command sets an unconditional breakpoint on all
4175 matches, printing a list of all breakpoints it set. Once these
4176 breakpoints are set, they are treated just like the breakpoints set with
4177 the @code{break} command. You can delete them, disable them, or make
4178 them conditional the same way as any other breakpoint.
4180 In programs using different languages, @value{GDBN} chooses the syntax
4181 to print the list of all breakpoints it sets according to the
4182 @samp{set language} value: using @samp{set language auto}
4183 (see @ref{Automatically, ,Set Language Automatically}) means to use the
4184 language of the breakpoint's function, other values mean to use
4185 the manually specified language (see @ref{Manually, ,Set Language Manually}).
4187 The syntax of the regular expression is the standard one used with tools
4188 like @file{grep}. Note that this is different from the syntax used by
4189 shells, so for instance @code{foo*} matches all functions that include
4190 an @code{fo} followed by zero or more @code{o}s. There is an implicit
4191 @code{.*} leading and trailing the regular expression you supply, so to
4192 match only functions that begin with @code{foo}, use @code{^foo}.
4194 @cindex non-member C@t{++} functions, set breakpoint in
4195 When debugging C@t{++} programs, @code{rbreak} is useful for setting
4196 breakpoints on overloaded functions that are not members of any special
4199 @cindex set breakpoints on all functions
4200 The @code{rbreak} command can be used to set breakpoints in
4201 @strong{all} the functions in a program, like this:
4204 (@value{GDBP}) rbreak .
4207 @item rbreak @var{file}:@var{regex}
4208 If @code{rbreak} is called with a filename qualification, it limits
4209 the search for functions matching the given regular expression to the
4210 specified @var{file}. This can be used, for example, to set breakpoints on
4211 every function in a given file:
4214 (@value{GDBP}) rbreak file.c:.
4217 The colon separating the filename qualifier from the regex may
4218 optionally be surrounded by spaces.
4220 @kindex info breakpoints
4221 @cindex @code{$_} and @code{info breakpoints}
4222 @item info breakpoints @r{[}@var{list}@dots{}@r{]}
4223 @itemx info break @r{[}@var{list}@dots{}@r{]}
4224 Print a table of all breakpoints, watchpoints, and catchpoints set and
4225 not deleted. Optional argument @var{n} means print information only
4226 about the specified breakpoint(s) (or watchpoint(s) or catchpoint(s)).
4227 For each breakpoint, following columns are printed:
4230 @item Breakpoint Numbers
4232 Breakpoint, watchpoint, or catchpoint.
4234 Whether the breakpoint is marked to be disabled or deleted when hit.
4235 @item Enabled or Disabled
4236 Enabled breakpoints are marked with @samp{y}. @samp{n} marks breakpoints
4237 that are not enabled.
4239 Where the breakpoint is in your program, as a memory address. For a
4240 pending breakpoint whose address is not yet known, this field will
4241 contain @samp{<PENDING>}. Such breakpoint won't fire until a shared
4242 library that has the symbol or line referred by breakpoint is loaded.
4243 See below for details. A breakpoint with several locations will
4244 have @samp{<MULTIPLE>} in this field---see below for details.
4246 Where the breakpoint is in the source for your program, as a file and
4247 line number. For a pending breakpoint, the original string passed to
4248 the breakpoint command will be listed as it cannot be resolved until
4249 the appropriate shared library is loaded in the future.
4253 If a breakpoint is conditional, there are two evaluation modes: ``host'' and
4254 ``target''. If mode is ``host'', breakpoint condition evaluation is done by
4255 @value{GDBN} on the host's side. If it is ``target'', then the condition
4256 is evaluated by the target. The @code{info break} command shows
4257 the condition on the line following the affected breakpoint, together with
4258 its condition evaluation mode in between parentheses.
4260 Breakpoint commands, if any, are listed after that. A pending breakpoint is
4261 allowed to have a condition specified for it. The condition is not parsed for
4262 validity until a shared library is loaded that allows the pending
4263 breakpoint to resolve to a valid location.
4266 @code{info break} with a breakpoint
4267 number @var{n} as argument lists only that breakpoint. The
4268 convenience variable @code{$_} and the default examining-address for
4269 the @code{x} command are set to the address of the last breakpoint
4270 listed (@pxref{Memory, ,Examining Memory}).
4273 @code{info break} displays a count of the number of times the breakpoint
4274 has been hit. This is especially useful in conjunction with the
4275 @code{ignore} command. You can ignore a large number of breakpoint
4276 hits, look at the breakpoint info to see how many times the breakpoint
4277 was hit, and then run again, ignoring one less than that number. This
4278 will get you quickly to the last hit of that breakpoint.
4281 For a breakpoints with an enable count (xref) greater than 1,
4282 @code{info break} also displays that count.
4286 @value{GDBN} allows you to set any number of breakpoints at the same place in
4287 your program. There is nothing silly or meaningless about this. When
4288 the breakpoints are conditional, this is even useful
4289 (@pxref{Conditions, ,Break Conditions}).
4291 @cindex multiple locations, breakpoints
4292 @cindex breakpoints, multiple locations
4293 It is possible that a breakpoint corresponds to several locations
4294 in your program. Examples of this situation are:
4298 Multiple functions in the program may have the same name.
4301 For a C@t{++} constructor, the @value{NGCC} compiler generates several
4302 instances of the function body, used in different cases.
4305 For a C@t{++} template function, a given line in the function can
4306 correspond to any number of instantiations.
4309 For an inlined function, a given source line can correspond to
4310 several places where that function is inlined.
4313 In all those cases, @value{GDBN} will insert a breakpoint at all
4314 the relevant locations.
4316 A breakpoint with multiple locations is displayed in the breakpoint
4317 table using several rows---one header row, followed by one row for
4318 each breakpoint location. The header row has @samp{<MULTIPLE>} in the
4319 address column. The rows for individual locations contain the actual
4320 addresses for locations, and show the functions to which those
4321 locations belong. The number column for a location is of the form
4322 @var{breakpoint-number}.@var{location-number}.
4327 Num Type Disp Enb Address What
4328 1 breakpoint keep y <MULTIPLE>
4330 breakpoint already hit 1 time
4331 1.1 y 0x080486a2 in void foo<int>() at t.cc:8
4332 1.2 y 0x080486ca in void foo<double>() at t.cc:8
4335 You cannot delete the individual locations from a breakpoint. However,
4336 each location can be individually enabled or disabled by passing
4337 @var{breakpoint-number}.@var{location-number} as argument to the
4338 @code{enable} and @code{disable} commands. It's also possible to
4339 @code{enable} and @code{disable} a range of @var{location-number}
4340 locations using a @var{breakpoint-number} and two @var{location-number}s,
4341 in increasing order, separated by a hyphen, like
4342 @kbd{@var{breakpoint-number}.@var{location-number1}-@var{location-number2}},
4343 in which case @value{GDBN} acts on all the locations in the range (inclusive).
4344 Disabling or enabling the parent breakpoint (@pxref{Disabling}) affects
4345 all of the locations that belong to that breakpoint.
4347 @cindex pending breakpoints
4348 It's quite common to have a breakpoint inside a shared library.
4349 Shared libraries can be loaded and unloaded explicitly,
4350 and possibly repeatedly, as the program is executed. To support
4351 this use case, @value{GDBN} updates breakpoint locations whenever
4352 any shared library is loaded or unloaded. Typically, you would
4353 set a breakpoint in a shared library at the beginning of your
4354 debugging session, when the library is not loaded, and when the
4355 symbols from the library are not available. When you try to set
4356 breakpoint, @value{GDBN} will ask you if you want to set
4357 a so called @dfn{pending breakpoint}---breakpoint whose address
4358 is not yet resolved.
4360 After the program is run, whenever a new shared library is loaded,
4361 @value{GDBN} reevaluates all the breakpoints. When a newly loaded
4362 shared library contains the symbol or line referred to by some
4363 pending breakpoint, that breakpoint is resolved and becomes an
4364 ordinary breakpoint. When a library is unloaded, all breakpoints
4365 that refer to its symbols or source lines become pending again.
4367 This logic works for breakpoints with multiple locations, too. For
4368 example, if you have a breakpoint in a C@t{++} template function, and
4369 a newly loaded shared library has an instantiation of that template,
4370 a new location is added to the list of locations for the breakpoint.
4372 Except for having unresolved address, pending breakpoints do not
4373 differ from regular breakpoints. You can set conditions or commands,
4374 enable and disable them and perform other breakpoint operations.
4376 @value{GDBN} provides some additional commands for controlling what
4377 happens when the @samp{break} command cannot resolve breakpoint
4378 address specification to an address:
4380 @kindex set breakpoint pending
4381 @kindex show breakpoint pending
4383 @item set breakpoint pending auto
4384 This is the default behavior. When @value{GDBN} cannot find the breakpoint
4385 location, it queries you whether a pending breakpoint should be created.
4387 @item set breakpoint pending on
4388 This indicates that an unrecognized breakpoint location should automatically
4389 result in a pending breakpoint being created.
4391 @item set breakpoint pending off
4392 This indicates that pending breakpoints are not to be created. Any
4393 unrecognized breakpoint location results in an error. This setting does
4394 not affect any pending breakpoints previously created.
4396 @item show breakpoint pending
4397 Show the current behavior setting for creating pending breakpoints.
4400 The settings above only affect the @code{break} command and its
4401 variants. Once breakpoint is set, it will be automatically updated
4402 as shared libraries are loaded and unloaded.
4404 @cindex automatic hardware breakpoints
4405 For some targets, @value{GDBN} can automatically decide if hardware or
4406 software breakpoints should be used, depending on whether the
4407 breakpoint address is read-only or read-write. This applies to
4408 breakpoints set with the @code{break} command as well as to internal
4409 breakpoints set by commands like @code{next} and @code{finish}. For
4410 breakpoints set with @code{hbreak}, @value{GDBN} will always use hardware
4413 You can control this automatic behaviour with the following commands:
4415 @kindex set breakpoint auto-hw
4416 @kindex show breakpoint auto-hw
4418 @item set breakpoint auto-hw on
4419 This is the default behavior. When @value{GDBN} sets a breakpoint, it
4420 will try to use the target memory map to decide if software or hardware
4421 breakpoint must be used.
4423 @item set breakpoint auto-hw off
4424 This indicates @value{GDBN} should not automatically select breakpoint
4425 type. If the target provides a memory map, @value{GDBN} will warn when
4426 trying to set software breakpoint at a read-only address.
4429 @value{GDBN} normally implements breakpoints by replacing the program code
4430 at the breakpoint address with a special instruction, which, when
4431 executed, given control to the debugger. By default, the program
4432 code is so modified only when the program is resumed. As soon as
4433 the program stops, @value{GDBN} restores the original instructions. This
4434 behaviour guards against leaving breakpoints inserted in the
4435 target should gdb abrubptly disconnect. However, with slow remote
4436 targets, inserting and removing breakpoint can reduce the performance.
4437 This behavior can be controlled with the following commands::
4439 @kindex set breakpoint always-inserted
4440 @kindex show breakpoint always-inserted
4442 @item set breakpoint always-inserted off
4443 All breakpoints, including newly added by the user, are inserted in
4444 the target only when the target is resumed. All breakpoints are
4445 removed from the target when it stops. This is the default mode.
4447 @item set breakpoint always-inserted on
4448 Causes all breakpoints to be inserted in the target at all times. If
4449 the user adds a new breakpoint, or changes an existing breakpoint, the
4450 breakpoints in the target are updated immediately. A breakpoint is
4451 removed from the target only when breakpoint itself is deleted.
4454 @value{GDBN} handles conditional breakpoints by evaluating these conditions
4455 when a breakpoint breaks. If the condition is true, then the process being
4456 debugged stops, otherwise the process is resumed.
4458 If the target supports evaluating conditions on its end, @value{GDBN} may
4459 download the breakpoint, together with its conditions, to it.
4461 This feature can be controlled via the following commands:
4463 @kindex set breakpoint condition-evaluation
4464 @kindex show breakpoint condition-evaluation
4466 @item set breakpoint condition-evaluation host
4467 This option commands @value{GDBN} to evaluate the breakpoint
4468 conditions on the host's side. Unconditional breakpoints are sent to
4469 the target which in turn receives the triggers and reports them back to GDB
4470 for condition evaluation. This is the standard evaluation mode.
4472 @item set breakpoint condition-evaluation target
4473 This option commands @value{GDBN} to download breakpoint conditions
4474 to the target at the moment of their insertion. The target
4475 is responsible for evaluating the conditional expression and reporting
4476 breakpoint stop events back to @value{GDBN} whenever the condition
4477 is true. Due to limitations of target-side evaluation, some conditions
4478 cannot be evaluated there, e.g., conditions that depend on local data
4479 that is only known to the host. Examples include
4480 conditional expressions involving convenience variables, complex types
4481 that cannot be handled by the agent expression parser and expressions
4482 that are too long to be sent over to the target, specially when the
4483 target is a remote system. In these cases, the conditions will be
4484 evaluated by @value{GDBN}.
4486 @item set breakpoint condition-evaluation auto
4487 This is the default mode. If the target supports evaluating breakpoint
4488 conditions on its end, @value{GDBN} will download breakpoint conditions to
4489 the target (limitations mentioned previously apply). If the target does
4490 not support breakpoint condition evaluation, then @value{GDBN} will fallback
4491 to evaluating all these conditions on the host's side.
4495 @cindex negative breakpoint numbers
4496 @cindex internal @value{GDBN} breakpoints
4497 @value{GDBN} itself sometimes sets breakpoints in your program for
4498 special purposes, such as proper handling of @code{longjmp} (in C
4499 programs). These internal breakpoints are assigned negative numbers,
4500 starting with @code{-1}; @samp{info breakpoints} does not display them.
4501 You can see these breakpoints with the @value{GDBN} maintenance command
4502 @samp{maint info breakpoints} (@pxref{maint info breakpoints}).
4505 @node Set Watchpoints
4506 @subsection Setting Watchpoints
4508 @cindex setting watchpoints
4509 You can use a watchpoint to stop execution whenever the value of an
4510 expression changes, without having to predict a particular place where
4511 this may happen. (This is sometimes called a @dfn{data breakpoint}.)
4512 The expression may be as simple as the value of a single variable, or
4513 as complex as many variables combined by operators. Examples include:
4517 A reference to the value of a single variable.
4520 An address cast to an appropriate data type. For example,
4521 @samp{*(int *)0x12345678} will watch a 4-byte region at the specified
4522 address (assuming an @code{int} occupies 4 bytes).
4525 An arbitrarily complex expression, such as @samp{a*b + c/d}. The
4526 expression can use any operators valid in the program's native
4527 language (@pxref{Languages}).
4530 You can set a watchpoint on an expression even if the expression can
4531 not be evaluated yet. For instance, you can set a watchpoint on
4532 @samp{*global_ptr} before @samp{global_ptr} is initialized.
4533 @value{GDBN} will stop when your program sets @samp{global_ptr} and
4534 the expression produces a valid value. If the expression becomes
4535 valid in some other way than changing a variable (e.g.@: if the memory
4536 pointed to by @samp{*global_ptr} becomes readable as the result of a
4537 @code{malloc} call), @value{GDBN} may not stop until the next time
4538 the expression changes.
4540 @cindex software watchpoints
4541 @cindex hardware watchpoints
4542 Depending on your system, watchpoints may be implemented in software or
4543 hardware. @value{GDBN} does software watchpointing by single-stepping your
4544 program and testing the variable's value each time, which is hundreds of
4545 times slower than normal execution. (But this may still be worth it, to
4546 catch errors where you have no clue what part of your program is the
4549 On some systems, such as most PowerPC or x86-based targets,
4550 @value{GDBN} includes support for hardware watchpoints, which do not
4551 slow down the running of your program.
4555 @item watch @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{thread-id}@r{]} @r{[}mask @var{maskvalue}@r{]}
4556 Set a watchpoint for an expression. @value{GDBN} will break when the
4557 expression @var{expr} is written into by the program and its value
4558 changes. The simplest (and the most popular) use of this command is
4559 to watch the value of a single variable:
4562 (@value{GDBP}) watch foo
4565 If the command includes a @code{@r{[}thread @var{thread-id}@r{]}}
4566 argument, @value{GDBN} breaks only when the thread identified by
4567 @var{thread-id} changes the value of @var{expr}. If any other threads
4568 change the value of @var{expr}, @value{GDBN} will not break. Note
4569 that watchpoints restricted to a single thread in this way only work
4570 with Hardware Watchpoints.
4572 Ordinarily a watchpoint respects the scope of variables in @var{expr}
4573 (see below). The @code{-location} argument tells @value{GDBN} to
4574 instead watch the memory referred to by @var{expr}. In this case,
4575 @value{GDBN} will evaluate @var{expr}, take the address of the result,
4576 and watch the memory at that address. The type of the result is used
4577 to determine the size of the watched memory. If the expression's
4578 result does not have an address, then @value{GDBN} will print an
4581 The @code{@r{[}mask @var{maskvalue}@r{]}} argument allows creation
4582 of masked watchpoints, if the current architecture supports this
4583 feature (e.g., PowerPC Embedded architecture, see @ref{PowerPC
4584 Embedded}.) A @dfn{masked watchpoint} specifies a mask in addition
4585 to an address to watch. The mask specifies that some bits of an address
4586 (the bits which are reset in the mask) should be ignored when matching
4587 the address accessed by the inferior against the watchpoint address.
4588 Thus, a masked watchpoint watches many addresses simultaneously---those
4589 addresses whose unmasked bits are identical to the unmasked bits in the
4590 watchpoint address. The @code{mask} argument implies @code{-location}.
4594 (@value{GDBP}) watch foo mask 0xffff00ff
4595 (@value{GDBP}) watch *0xdeadbeef mask 0xffffff00
4599 @item rwatch @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{thread-id}@r{]} @r{[}mask @var{maskvalue}@r{]}
4600 Set a watchpoint that will break when the value of @var{expr} is read
4604 @item awatch @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{thread-id}@r{]} @r{[}mask @var{maskvalue}@r{]}
4605 Set a watchpoint that will break when @var{expr} is either read from
4606 or written into by the program.
4608 @kindex info watchpoints @r{[}@var{list}@dots{}@r{]}
4609 @item info watchpoints @r{[}@var{list}@dots{}@r{]}
4610 This command prints a list of watchpoints, using the same format as
4611 @code{info break} (@pxref{Set Breaks}).
4614 If you watch for a change in a numerically entered address you need to
4615 dereference it, as the address itself is just a constant number which will
4616 never change. @value{GDBN} refuses to create a watchpoint that watches
4617 a never-changing value:
4620 (@value{GDBP}) watch 0x600850
4621 Cannot watch constant value 0x600850.
4622 (@value{GDBP}) watch *(int *) 0x600850
4623 Watchpoint 1: *(int *) 6293584
4626 @value{GDBN} sets a @dfn{hardware watchpoint} if possible. Hardware
4627 watchpoints execute very quickly, and the debugger reports a change in
4628 value at the exact instruction where the change occurs. If @value{GDBN}
4629 cannot set a hardware watchpoint, it sets a software watchpoint, which
4630 executes more slowly and reports the change in value at the next
4631 @emph{statement}, not the instruction, after the change occurs.
4633 @cindex use only software watchpoints
4634 You can force @value{GDBN} to use only software watchpoints with the
4635 @kbd{set can-use-hw-watchpoints 0} command. With this variable set to
4636 zero, @value{GDBN} will never try to use hardware watchpoints, even if
4637 the underlying system supports them. (Note that hardware-assisted
4638 watchpoints that were set @emph{before} setting
4639 @code{can-use-hw-watchpoints} to zero will still use the hardware
4640 mechanism of watching expression values.)
4643 @item set can-use-hw-watchpoints
4644 @kindex set can-use-hw-watchpoints
4645 Set whether or not to use hardware watchpoints.
4647 @item show can-use-hw-watchpoints
4648 @kindex show can-use-hw-watchpoints
4649 Show the current mode of using hardware watchpoints.
4652 For remote targets, you can restrict the number of hardware
4653 watchpoints @value{GDBN} will use, see @ref{set remote
4654 hardware-breakpoint-limit}.
4656 When you issue the @code{watch} command, @value{GDBN} reports
4659 Hardware watchpoint @var{num}: @var{expr}
4663 if it was able to set a hardware watchpoint.
4665 Currently, the @code{awatch} and @code{rwatch} commands can only set
4666 hardware watchpoints, because accesses to data that don't change the
4667 value of the watched expression cannot be detected without examining
4668 every instruction as it is being executed, and @value{GDBN} does not do
4669 that currently. If @value{GDBN} finds that it is unable to set a
4670 hardware breakpoint with the @code{awatch} or @code{rwatch} command, it
4671 will print a message like this:
4674 Expression cannot be implemented with read/access watchpoint.
4677 Sometimes, @value{GDBN} cannot set a hardware watchpoint because the
4678 data type of the watched expression is wider than what a hardware
4679 watchpoint on the target machine can handle. For example, some systems
4680 can only watch regions that are up to 4 bytes wide; on such systems you
4681 cannot set hardware watchpoints for an expression that yields a
4682 double-precision floating-point number (which is typically 8 bytes
4683 wide). As a work-around, it might be possible to break the large region
4684 into a series of smaller ones and watch them with separate watchpoints.
4686 If you set too many hardware watchpoints, @value{GDBN} might be unable
4687 to insert all of them when you resume the execution of your program.
4688 Since the precise number of active watchpoints is unknown until such
4689 time as the program is about to be resumed, @value{GDBN} might not be
4690 able to warn you about this when you set the watchpoints, and the
4691 warning will be printed only when the program is resumed:
4694 Hardware watchpoint @var{num}: Could not insert watchpoint
4698 If this happens, delete or disable some of the watchpoints.
4700 Watching complex expressions that reference many variables can also
4701 exhaust the resources available for hardware-assisted watchpoints.
4702 That's because @value{GDBN} needs to watch every variable in the
4703 expression with separately allocated resources.
4705 If you call a function interactively using @code{print} or @code{call},
4706 any watchpoints you have set will be inactive until @value{GDBN} reaches another
4707 kind of breakpoint or the call completes.
4709 @value{GDBN} automatically deletes watchpoints that watch local
4710 (automatic) variables, or expressions that involve such variables, when
4711 they go out of scope, that is, when the execution leaves the block in
4712 which these variables were defined. In particular, when the program
4713 being debugged terminates, @emph{all} local variables go out of scope,
4714 and so only watchpoints that watch global variables remain set. If you
4715 rerun the program, you will need to set all such watchpoints again. One
4716 way of doing that would be to set a code breakpoint at the entry to the
4717 @code{main} function and when it breaks, set all the watchpoints.
4719 @cindex watchpoints and threads
4720 @cindex threads and watchpoints
4721 In multi-threaded programs, watchpoints will detect changes to the
4722 watched expression from every thread.
4725 @emph{Warning:} In multi-threaded programs, software watchpoints
4726 have only limited usefulness. If @value{GDBN} creates a software
4727 watchpoint, it can only watch the value of an expression @emph{in a
4728 single thread}. If you are confident that the expression can only
4729 change due to the current thread's activity (and if you are also
4730 confident that no other thread can become current), then you can use
4731 software watchpoints as usual. However, @value{GDBN} may not notice
4732 when a non-current thread's activity changes the expression. (Hardware
4733 watchpoints, in contrast, watch an expression in all threads.)
4736 @xref{set remote hardware-watchpoint-limit}.
4738 @node Set Catchpoints
4739 @subsection Setting Catchpoints
4740 @cindex catchpoints, setting
4741 @cindex exception handlers
4742 @cindex event handling
4744 You can use @dfn{catchpoints} to cause the debugger to stop for certain
4745 kinds of program events, such as C@t{++} exceptions or the loading of a
4746 shared library. Use the @code{catch} command to set a catchpoint.
4750 @item catch @var{event}
4751 Stop when @var{event} occurs. The @var{event} can be any of the following:
4754 @item throw @r{[}@var{regexp}@r{]}
4755 @itemx rethrow @r{[}@var{regexp}@r{]}
4756 @itemx catch @r{[}@var{regexp}@r{]}
4758 @kindex catch rethrow
4760 @cindex stop on C@t{++} exceptions
4761 The throwing, re-throwing, or catching of a C@t{++} exception.
4763 If @var{regexp} is given, then only exceptions whose type matches the
4764 regular expression will be caught.
4766 @vindex $_exception@r{, convenience variable}
4767 The convenience variable @code{$_exception} is available at an
4768 exception-related catchpoint, on some systems. This holds the
4769 exception being thrown.
4771 There are currently some limitations to C@t{++} exception handling in
4776 The support for these commands is system-dependent. Currently, only
4777 systems using the @samp{gnu-v3} C@t{++} ABI (@pxref{ABI}) are
4781 The regular expression feature and the @code{$_exception} convenience
4782 variable rely on the presence of some SDT probes in @code{libstdc++}.
4783 If these probes are not present, then these features cannot be used.
4784 These probes were first available in the GCC 4.8 release, but whether
4785 or not they are available in your GCC also depends on how it was
4789 The @code{$_exception} convenience variable is only valid at the
4790 instruction at which an exception-related catchpoint is set.
4793 When an exception-related catchpoint is hit, @value{GDBN} stops at a
4794 location in the system library which implements runtime exception
4795 support for C@t{++}, usually @code{libstdc++}. You can use @code{up}
4796 (@pxref{Selection}) to get to your code.
4799 If you call a function interactively, @value{GDBN} normally returns
4800 control to you when the function has finished executing. If the call
4801 raises an exception, however, the call may bypass the mechanism that
4802 returns control to you and cause your program either to abort or to
4803 simply continue running until it hits a breakpoint, catches a signal
4804 that @value{GDBN} is listening for, or exits. This is the case even if
4805 you set a catchpoint for the exception; catchpoints on exceptions are
4806 disabled within interactive calls. @xref{Calling}, for information on
4807 controlling this with @code{set unwind-on-terminating-exception}.
4810 You cannot raise an exception interactively.
4813 You cannot install an exception handler interactively.
4816 @item exception @r{[}@var{name}@r{]}
4817 @kindex catch exception
4818 @cindex Ada exception catching
4819 @cindex catch Ada exceptions
4820 An Ada exception being raised. If an exception name is specified
4821 at the end of the command (eg @code{catch exception Program_Error}),
4822 the debugger will stop only when this specific exception is raised.
4823 Otherwise, the debugger stops execution when any Ada exception is raised.
4825 When inserting an exception catchpoint on a user-defined exception whose
4826 name is identical to one of the exceptions defined by the language, the
4827 fully qualified name must be used as the exception name. Otherwise,
4828 @value{GDBN} will assume that it should stop on the pre-defined exception
4829 rather than the user-defined one. For instance, assuming an exception
4830 called @code{Constraint_Error} is defined in package @code{Pck}, then
4831 the command to use to catch such exceptions is @kbd{catch exception
4832 Pck.Constraint_Error}.
4834 @vindex $_ada_exception@r{, convenience variable}
4835 The convenience variable @code{$_ada_exception} holds the address of
4836 the exception being thrown. This can be useful when setting a
4837 condition for such a catchpoint.
4839 @item exception unhandled
4840 @kindex catch exception unhandled
4841 An exception that was raised but is not handled by the program. The
4842 convenience variable @code{$_ada_exception} is set as for @code{catch
4845 @item handlers @r{[}@var{name}@r{]}
4846 @kindex catch handlers
4847 @cindex Ada exception handlers catching
4848 @cindex catch Ada exceptions when handled
4849 An Ada exception being handled. If an exception name is
4850 specified at the end of the command
4851 (eg @kbd{catch handlers Program_Error}), the debugger will stop
4852 only when this specific exception is handled.
4853 Otherwise, the debugger stops execution when any Ada exception is handled.
4855 When inserting a handlers catchpoint on a user-defined
4856 exception whose name is identical to one of the exceptions
4857 defined by the language, the fully qualified name must be used
4858 as the exception name. Otherwise, @value{GDBN} will assume that it
4859 should stop on the pre-defined exception rather than the
4860 user-defined one. For instance, assuming an exception called
4861 @code{Constraint_Error} is defined in package @code{Pck}, then the
4862 command to use to catch such exceptions handling is
4863 @kbd{catch handlers Pck.Constraint_Error}.
4865 The convenience variable @code{$_ada_exception} is set as for
4866 @code{catch exception}.
4869 @kindex catch assert
4870 A failed Ada assertion. Note that the convenience variable
4871 @code{$_ada_exception} is @emph{not} set by this catchpoint.
4875 @cindex break on fork/exec
4876 A call to @code{exec}.
4878 @anchor{catch syscall}
4880 @itemx syscall @r{[}@var{name} @r{|} @var{number} @r{|} @r{group:}@var{groupname} @r{|} @r{g:}@var{groupname}@r{]} @dots{}
4881 @kindex catch syscall
4882 @cindex break on a system call.
4883 A call to or return from a system call, a.k.a.@: @dfn{syscall}. A
4884 syscall is a mechanism for application programs to request a service
4885 from the operating system (OS) or one of the OS system services.
4886 @value{GDBN} can catch some or all of the syscalls issued by the
4887 debuggee, and show the related information for each syscall. If no
4888 argument is specified, calls to and returns from all system calls
4891 @var{name} can be any system call name that is valid for the
4892 underlying OS. Just what syscalls are valid depends on the OS. On
4893 GNU and Unix systems, you can find the full list of valid syscall
4894 names on @file{/usr/include/asm/unistd.h}.
4896 @c For MS-Windows, the syscall names and the corresponding numbers
4897 @c can be found, e.g., on this URL:
4898 @c http://www.metasploit.com/users/opcode/syscalls.html
4899 @c but we don't support Windows syscalls yet.
4901 Normally, @value{GDBN} knows in advance which syscalls are valid for
4902 each OS, so you can use the @value{GDBN} command-line completion
4903 facilities (@pxref{Completion,, command completion}) to list the
4906 You may also specify the system call numerically. A syscall's
4907 number is the value passed to the OS's syscall dispatcher to
4908 identify the requested service. When you specify the syscall by its
4909 name, @value{GDBN} uses its database of syscalls to convert the name
4910 into the corresponding numeric code, but using the number directly
4911 may be useful if @value{GDBN}'s database does not have the complete
4912 list of syscalls on your system (e.g., because @value{GDBN} lags
4913 behind the OS upgrades).
4915 You may specify a group of related syscalls to be caught at once using
4916 the @code{group:} syntax (@code{g:} is a shorter equivalent). For
4917 instance, on some platforms @value{GDBN} allows you to catch all
4918 network related syscalls, by passing the argument @code{group:network}
4919 to @code{catch syscall}. Note that not all syscall groups are
4920 available in every system. You can use the command completion
4921 facilities (@pxref{Completion,, command completion}) to list the
4922 syscall groups available on your environment.
4924 The example below illustrates how this command works if you don't provide
4928 (@value{GDBP}) catch syscall
4929 Catchpoint 1 (syscall)
4931 Starting program: /tmp/catch-syscall
4933 Catchpoint 1 (call to syscall 'close'), \
4934 0xffffe424 in __kernel_vsyscall ()
4938 Catchpoint 1 (returned from syscall 'close'), \
4939 0xffffe424 in __kernel_vsyscall ()
4943 Here is an example of catching a system call by name:
4946 (@value{GDBP}) catch syscall chroot
4947 Catchpoint 1 (syscall 'chroot' [61])
4949 Starting program: /tmp/catch-syscall
4951 Catchpoint 1 (call to syscall 'chroot'), \
4952 0xffffe424 in __kernel_vsyscall ()
4956 Catchpoint 1 (returned from syscall 'chroot'), \
4957 0xffffe424 in __kernel_vsyscall ()
4961 An example of specifying a system call numerically. In the case
4962 below, the syscall number has a corresponding entry in the XML
4963 file, so @value{GDBN} finds its name and prints it:
4966 (@value{GDBP}) catch syscall 252
4967 Catchpoint 1 (syscall(s) 'exit_group')
4969 Starting program: /tmp/catch-syscall
4971 Catchpoint 1 (call to syscall 'exit_group'), \
4972 0xffffe424 in __kernel_vsyscall ()
4976 Program exited normally.
4980 Here is an example of catching a syscall group:
4983 (@value{GDBP}) catch syscall group:process
4984 Catchpoint 1 (syscalls 'exit' [1] 'fork' [2] 'waitpid' [7]
4985 'execve' [11] 'wait4' [114] 'clone' [120] 'vfork' [190]
4986 'exit_group' [252] 'waitid' [284] 'unshare' [310])
4988 Starting program: /tmp/catch-syscall
4990 Catchpoint 1 (call to syscall fork), 0x00007ffff7df4e27 in open64 ()
4991 from /lib64/ld-linux-x86-64.so.2
4997 However, there can be situations when there is no corresponding name
4998 in XML file for that syscall number. In this case, @value{GDBN} prints
4999 a warning message saying that it was not able to find the syscall name,
5000 but the catchpoint will be set anyway. See the example below:
5003 (@value{GDBP}) catch syscall 764
5004 warning: The number '764' does not represent a known syscall.
5005 Catchpoint 2 (syscall 764)
5009 If you configure @value{GDBN} using the @samp{--without-expat} option,
5010 it will not be able to display syscall names. Also, if your
5011 architecture does not have an XML file describing its system calls,
5012 you will not be able to see the syscall names. It is important to
5013 notice that these two features are used for accessing the syscall
5014 name database. In either case, you will see a warning like this:
5017 (@value{GDBP}) catch syscall
5018 warning: Could not open "syscalls/i386-linux.xml"
5019 warning: Could not load the syscall XML file 'syscalls/i386-linux.xml'.
5020 GDB will not be able to display syscall names.
5021 Catchpoint 1 (syscall)
5025 Of course, the file name will change depending on your architecture and system.
5027 Still using the example above, you can also try to catch a syscall by its
5028 number. In this case, you would see something like:
5031 (@value{GDBP}) catch syscall 252
5032 Catchpoint 1 (syscall(s) 252)
5035 Again, in this case @value{GDBN} would not be able to display syscall's names.
5039 A call to @code{fork}.
5043 A call to @code{vfork}.
5045 @item load @r{[}@var{regexp}@r{]}
5046 @itemx unload @r{[}@var{regexp}@r{]}
5048 @kindex catch unload
5049 The loading or unloading of a shared library. If @var{regexp} is
5050 given, then the catchpoint will stop only if the regular expression
5051 matches one of the affected libraries.
5053 @item signal @r{[}@var{signal}@dots{} @r{|} @samp{all}@r{]}
5054 @kindex catch signal
5055 The delivery of a signal.
5057 With no arguments, this catchpoint will catch any signal that is not
5058 used internally by @value{GDBN}, specifically, all signals except
5059 @samp{SIGTRAP} and @samp{SIGINT}.
5061 With the argument @samp{all}, all signals, including those used by
5062 @value{GDBN}, will be caught. This argument cannot be used with other
5065 Otherwise, the arguments are a list of signal names as given to
5066 @code{handle} (@pxref{Signals}). Only signals specified in this list
5069 One reason that @code{catch signal} can be more useful than
5070 @code{handle} is that you can attach commands and conditions to the
5073 When a signal is caught by a catchpoint, the signal's @code{stop} and
5074 @code{print} settings, as specified by @code{handle}, are ignored.
5075 However, whether the signal is still delivered to the inferior depends
5076 on the @code{pass} setting; this can be changed in the catchpoint's
5081 @item tcatch @var{event}
5083 Set a catchpoint that is enabled only for one stop. The catchpoint is
5084 automatically deleted after the first time the event is caught.
5088 Use the @code{info break} command to list the current catchpoints.
5092 @subsection Deleting Breakpoints
5094 @cindex clearing breakpoints, watchpoints, catchpoints
5095 @cindex deleting breakpoints, watchpoints, catchpoints
5096 It is often necessary to eliminate a breakpoint, watchpoint, or
5097 catchpoint once it has done its job and you no longer want your program
5098 to stop there. This is called @dfn{deleting} the breakpoint. A
5099 breakpoint that has been deleted no longer exists; it is forgotten.
5101 With the @code{clear} command you can delete breakpoints according to
5102 where they are in your program. With the @code{delete} command you can
5103 delete individual breakpoints, watchpoints, or catchpoints by specifying
5104 their breakpoint numbers.
5106 It is not necessary to delete a breakpoint to proceed past it. @value{GDBN}
5107 automatically ignores breakpoints on the first instruction to be executed
5108 when you continue execution without changing the execution address.
5113 Delete any breakpoints at the next instruction to be executed in the
5114 selected stack frame (@pxref{Selection, ,Selecting a Frame}). When
5115 the innermost frame is selected, this is a good way to delete a
5116 breakpoint where your program just stopped.
5118 @item clear @var{location}
5119 Delete any breakpoints set at the specified @var{location}.
5120 @xref{Specify Location}, for the various forms of @var{location}; the
5121 most useful ones are listed below:
5124 @item clear @var{function}
5125 @itemx clear @var{filename}:@var{function}
5126 Delete any breakpoints set at entry to the named @var{function}.
5128 @item clear @var{linenum}
5129 @itemx clear @var{filename}:@var{linenum}
5130 Delete any breakpoints set at or within the code of the specified
5131 @var{linenum} of the specified @var{filename}.
5134 @cindex delete breakpoints
5136 @kindex d @r{(@code{delete})}
5137 @item delete @r{[}breakpoints@r{]} @r{[}@var{list}@dots{}@r{]}
5138 Delete the breakpoints, watchpoints, or catchpoints of the breakpoint
5139 list specified as argument. If no argument is specified, delete all
5140 breakpoints (@value{GDBN} asks confirmation, unless you have @code{set
5141 confirm off}). You can abbreviate this command as @code{d}.
5145 @subsection Disabling Breakpoints
5147 @cindex enable/disable a breakpoint
5148 Rather than deleting a breakpoint, watchpoint, or catchpoint, you might
5149 prefer to @dfn{disable} it. This makes the breakpoint inoperative as if
5150 it had been deleted, but remembers the information on the breakpoint so
5151 that you can @dfn{enable} it again later.
5153 You disable and enable breakpoints, watchpoints, and catchpoints with
5154 the @code{enable} and @code{disable} commands, optionally specifying
5155 one or more breakpoint numbers as arguments. Use @code{info break} to
5156 print a list of all breakpoints, watchpoints, and catchpoints if you
5157 do not know which numbers to use.
5159 Disabling and enabling a breakpoint that has multiple locations
5160 affects all of its locations.
5162 A breakpoint, watchpoint, or catchpoint can have any of several
5163 different states of enablement:
5167 Enabled. The breakpoint stops your program. A breakpoint set
5168 with the @code{break} command starts out in this state.
5170 Disabled. The breakpoint has no effect on your program.
5172 Enabled once. The breakpoint stops your program, but then becomes
5175 Enabled for a count. The breakpoint stops your program for the next
5176 N times, then becomes disabled.
5178 Enabled for deletion. The breakpoint stops your program, but
5179 immediately after it does so it is deleted permanently. A breakpoint
5180 set with the @code{tbreak} command starts out in this state.
5183 You can use the following commands to enable or disable breakpoints,
5184 watchpoints, and catchpoints:
5188 @kindex dis @r{(@code{disable})}
5189 @item disable @r{[}breakpoints@r{]} @r{[}@var{list}@dots{}@r{]}
5190 Disable the specified breakpoints---or all breakpoints, if none are
5191 listed. A disabled breakpoint has no effect but is not forgotten. All
5192 options such as ignore-counts, conditions and commands are remembered in
5193 case the breakpoint is enabled again later. You may abbreviate
5194 @code{disable} as @code{dis}.
5197 @item enable @r{[}breakpoints@r{]} @r{[}@var{list}@dots{}@r{]}
5198 Enable the specified breakpoints (or all defined breakpoints). They
5199 become effective once again in stopping your program.
5201 @item enable @r{[}breakpoints@r{]} once @var{list}@dots{}
5202 Enable the specified breakpoints temporarily. @value{GDBN} disables any
5203 of these breakpoints immediately after stopping your program.
5205 @item enable @r{[}breakpoints@r{]} count @var{count} @var{list}@dots{}
5206 Enable the specified breakpoints temporarily. @value{GDBN} records
5207 @var{count} with each of the specified breakpoints, and decrements a
5208 breakpoint's count when it is hit. When any count reaches 0,
5209 @value{GDBN} disables that breakpoint. If a breakpoint has an ignore
5210 count (@pxref{Conditions, ,Break Conditions}), that will be
5211 decremented to 0 before @var{count} is affected.
5213 @item enable @r{[}breakpoints@r{]} delete @var{list}@dots{}
5214 Enable the specified breakpoints to work once, then die. @value{GDBN}
5215 deletes any of these breakpoints as soon as your program stops there.
5216 Breakpoints set by the @code{tbreak} command start out in this state.
5219 @c FIXME: I think the following ``Except for [...] @code{tbreak}'' is
5220 @c confusing: tbreak is also initially enabled.
5221 Except for a breakpoint set with @code{tbreak} (@pxref{Set Breaks,
5222 ,Setting Breakpoints}), breakpoints that you set are initially enabled;
5223 subsequently, they become disabled or enabled only when you use one of
5224 the commands above. (The command @code{until} can set and delete a
5225 breakpoint of its own, but it does not change the state of your other
5226 breakpoints; see @ref{Continuing and Stepping, ,Continuing and
5230 @subsection Break Conditions
5231 @cindex conditional breakpoints
5232 @cindex breakpoint conditions
5234 @c FIXME what is scope of break condition expr? Context where wanted?
5235 @c in particular for a watchpoint?
5236 The simplest sort of breakpoint breaks every time your program reaches a
5237 specified place. You can also specify a @dfn{condition} for a
5238 breakpoint. A condition is just a Boolean expression in your
5239 programming language (@pxref{Expressions, ,Expressions}). A breakpoint with
5240 a condition evaluates the expression each time your program reaches it,
5241 and your program stops only if the condition is @emph{true}.
5243 This is the converse of using assertions for program validation; in that
5244 situation, you want to stop when the assertion is violated---that is,
5245 when the condition is false. In C, if you want to test an assertion expressed
5246 by the condition @var{assert}, you should set the condition
5247 @samp{! @var{assert}} on the appropriate breakpoint.
5249 Conditions are also accepted for watchpoints; you may not need them,
5250 since a watchpoint is inspecting the value of an expression anyhow---but
5251 it might be simpler, say, to just set a watchpoint on a variable name,
5252 and specify a condition that tests whether the new value is an interesting
5255 Break conditions can have side effects, and may even call functions in
5256 your program. This can be useful, for example, to activate functions
5257 that log program progress, or to use your own print functions to
5258 format special data structures. The effects are completely predictable
5259 unless there is another enabled breakpoint at the same address. (In
5260 that case, @value{GDBN} might see the other breakpoint first and stop your
5261 program without checking the condition of this one.) Note that
5262 breakpoint commands are usually more convenient and flexible than break
5264 purpose of performing side effects when a breakpoint is reached
5265 (@pxref{Break Commands, ,Breakpoint Command Lists}).
5267 Breakpoint conditions can also be evaluated on the target's side if
5268 the target supports it. Instead of evaluating the conditions locally,
5269 @value{GDBN} encodes the expression into an agent expression
5270 (@pxref{Agent Expressions}) suitable for execution on the target,
5271 independently of @value{GDBN}. Global variables become raw memory
5272 locations, locals become stack accesses, and so forth.
5274 In this case, @value{GDBN} will only be notified of a breakpoint trigger
5275 when its condition evaluates to true. This mechanism may provide faster
5276 response times depending on the performance characteristics of the target
5277 since it does not need to keep @value{GDBN} informed about
5278 every breakpoint trigger, even those with false conditions.
5280 Break conditions can be specified when a breakpoint is set, by using
5281 @samp{if} in the arguments to the @code{break} command. @xref{Set
5282 Breaks, ,Setting Breakpoints}. They can also be changed at any time
5283 with the @code{condition} command.
5285 You can also use the @code{if} keyword with the @code{watch} command.
5286 The @code{catch} command does not recognize the @code{if} keyword;
5287 @code{condition} is the only way to impose a further condition on a
5292 @item condition @var{bnum} @var{expression}
5293 Specify @var{expression} as the break condition for breakpoint,
5294 watchpoint, or catchpoint number @var{bnum}. After you set a condition,
5295 breakpoint @var{bnum} stops your program only if the value of
5296 @var{expression} is true (nonzero, in C). When you use
5297 @code{condition}, @value{GDBN} checks @var{expression} immediately for
5298 syntactic correctness, and to determine whether symbols in it have
5299 referents in the context of your breakpoint. If @var{expression} uses
5300 symbols not referenced in the context of the breakpoint, @value{GDBN}
5301 prints an error message:
5304 No symbol "foo" in current context.
5309 not actually evaluate @var{expression} at the time the @code{condition}
5310 command (or a command that sets a breakpoint with a condition, like
5311 @code{break if @dots{}}) is given, however. @xref{Expressions, ,Expressions}.
5313 @item condition @var{bnum}
5314 Remove the condition from breakpoint number @var{bnum}. It becomes
5315 an ordinary unconditional breakpoint.
5318 @cindex ignore count (of breakpoint)
5319 A special case of a breakpoint condition is to stop only when the
5320 breakpoint has been reached a certain number of times. This is so
5321 useful that there is a special way to do it, using the @dfn{ignore
5322 count} of the breakpoint. Every breakpoint has an ignore count, which
5323 is an integer. Most of the time, the ignore count is zero, and
5324 therefore has no effect. But if your program reaches a breakpoint whose
5325 ignore count is positive, then instead of stopping, it just decrements
5326 the ignore count by one and continues. As a result, if the ignore count
5327 value is @var{n}, the breakpoint does not stop the next @var{n} times
5328 your program reaches it.
5332 @item ignore @var{bnum} @var{count}
5333 Set the ignore count of breakpoint number @var{bnum} to @var{count}.
5334 The next @var{count} times the breakpoint is reached, your program's
5335 execution does not stop; other than to decrement the ignore count, @value{GDBN}
5338 To make the breakpoint stop the next time it is reached, specify
5341 When you use @code{continue} to resume execution of your program from a
5342 breakpoint, you can specify an ignore count directly as an argument to
5343 @code{continue}, rather than using @code{ignore}. @xref{Continuing and
5344 Stepping,,Continuing and Stepping}.
5346 If a breakpoint has a positive ignore count and a condition, the
5347 condition is not checked. Once the ignore count reaches zero,
5348 @value{GDBN} resumes checking the condition.
5350 You could achieve the effect of the ignore count with a condition such
5351 as @w{@samp{$foo-- <= 0}} using a debugger convenience variable that
5352 is decremented each time. @xref{Convenience Vars, ,Convenience
5356 Ignore counts apply to breakpoints, watchpoints, and catchpoints.
5359 @node Break Commands
5360 @subsection Breakpoint Command Lists
5362 @cindex breakpoint commands
5363 You can give any breakpoint (or watchpoint or catchpoint) a series of
5364 commands to execute when your program stops due to that breakpoint. For
5365 example, you might want to print the values of certain expressions, or
5366 enable other breakpoints.
5370 @kindex end@r{ (breakpoint commands)}
5371 @item commands @r{[}@var{list}@dots{}@r{]}
5372 @itemx @dots{} @var{command-list} @dots{}
5374 Specify a list of commands for the given breakpoints. The commands
5375 themselves appear on the following lines. Type a line containing just
5376 @code{end} to terminate the commands.
5378 To remove all commands from a breakpoint, type @code{commands} and
5379 follow it immediately with @code{end}; that is, give no commands.
5381 With no argument, @code{commands} refers to the last breakpoint,
5382 watchpoint, or catchpoint set (not to the breakpoint most recently
5383 encountered). If the most recent breakpoints were set with a single
5384 command, then the @code{commands} will apply to all the breakpoints
5385 set by that command. This applies to breakpoints set by
5386 @code{rbreak}, and also applies when a single @code{break} command
5387 creates multiple breakpoints (@pxref{Ambiguous Expressions,,Ambiguous
5391 Pressing @key{RET} as a means of repeating the last @value{GDBN} command is
5392 disabled within a @var{command-list}.
5394 You can use breakpoint commands to start your program up again. Simply
5395 use the @code{continue} command, or @code{step}, or any other command
5396 that resumes execution.
5398 Any other commands in the command list, after a command that resumes
5399 execution, are ignored. This is because any time you resume execution
5400 (even with a simple @code{next} or @code{step}), you may encounter
5401 another breakpoint---which could have its own command list, leading to
5402 ambiguities about which list to execute.
5405 If the first command you specify in a command list is @code{silent}, the
5406 usual message about stopping at a breakpoint is not printed. This may
5407 be desirable for breakpoints that are to print a specific message and
5408 then continue. If none of the remaining commands print anything, you
5409 see no sign that the breakpoint was reached. @code{silent} is
5410 meaningful only at the beginning of a breakpoint command list.
5412 The commands @code{echo}, @code{output}, and @code{printf} allow you to
5413 print precisely controlled output, and are often useful in silent
5414 breakpoints. @xref{Output, ,Commands for Controlled Output}.
5416 For example, here is how you could use breakpoint commands to print the
5417 value of @code{x} at entry to @code{foo} whenever @code{x} is positive.
5423 printf "x is %d\n",x
5428 One application for breakpoint commands is to compensate for one bug so
5429 you can test for another. Put a breakpoint just after the erroneous line
5430 of code, give it a condition to detect the case in which something
5431 erroneous has been done, and give it commands to assign correct values
5432 to any variables that need them. End with the @code{continue} command
5433 so that your program does not stop, and start with the @code{silent}
5434 command so that no output is produced. Here is an example:
5445 @node Dynamic Printf
5446 @subsection Dynamic Printf
5448 @cindex dynamic printf
5450 The dynamic printf command @code{dprintf} combines a breakpoint with
5451 formatted printing of your program's data to give you the effect of
5452 inserting @code{printf} calls into your program on-the-fly, without
5453 having to recompile it.
5455 In its most basic form, the output goes to the GDB console. However,
5456 you can set the variable @code{dprintf-style} for alternate handling.
5457 For instance, you can ask to format the output by calling your
5458 program's @code{printf} function. This has the advantage that the
5459 characters go to the program's output device, so they can recorded in
5460 redirects to files and so forth.
5462 If you are doing remote debugging with a stub or agent, you can also
5463 ask to have the printf handled by the remote agent. In addition to
5464 ensuring that the output goes to the remote program's device along
5465 with any other output the program might produce, you can also ask that
5466 the dprintf remain active even after disconnecting from the remote
5467 target. Using the stub/agent is also more efficient, as it can do
5468 everything without needing to communicate with @value{GDBN}.
5472 @item dprintf @var{location},@var{template},@var{expression}[,@var{expression}@dots{}]
5473 Whenever execution reaches @var{location}, print the values of one or
5474 more @var{expressions} under the control of the string @var{template}.
5475 To print several values, separate them with commas.
5477 @item set dprintf-style @var{style}
5478 Set the dprintf output to be handled in one of several different
5479 styles enumerated below. A change of style affects all existing
5480 dynamic printfs immediately. (If you need individual control over the
5481 print commands, simply define normal breakpoints with
5482 explicitly-supplied command lists.)
5486 @kindex dprintf-style gdb
5487 Handle the output using the @value{GDBN} @code{printf} command.
5490 @kindex dprintf-style call
5491 Handle the output by calling a function in your program (normally
5495 @kindex dprintf-style agent
5496 Have the remote debugging agent (such as @code{gdbserver}) handle
5497 the output itself. This style is only available for agents that
5498 support running commands on the target.
5501 @item set dprintf-function @var{function}
5502 Set the function to call if the dprintf style is @code{call}. By
5503 default its value is @code{printf}. You may set it to any expression.
5504 that @value{GDBN} can evaluate to a function, as per the @code{call}
5507 @item set dprintf-channel @var{channel}
5508 Set a ``channel'' for dprintf. If set to a non-empty value,
5509 @value{GDBN} will evaluate it as an expression and pass the result as
5510 a first argument to the @code{dprintf-function}, in the manner of
5511 @code{fprintf} and similar functions. Otherwise, the dprintf format
5512 string will be the first argument, in the manner of @code{printf}.
5514 As an example, if you wanted @code{dprintf} output to go to a logfile
5515 that is a standard I/O stream assigned to the variable @code{mylog},
5516 you could do the following:
5519 (@value{GDBP}) set dprintf-style call
5520 (@value{GDBP}) set dprintf-function fprintf
5521 (@value{GDBP}) set dprintf-channel mylog
5522 (@value{GDBP}) dprintf 25,"at line 25, glob=%d\n",glob
5523 Dprintf 1 at 0x123456: file main.c, line 25.
5524 (@value{GDBP}) info break
5525 1 dprintf keep y 0x00123456 in main at main.c:25
5526 call (void) fprintf (mylog,"at line 25, glob=%d\n",glob)
5531 Note that the @code{info break} displays the dynamic printf commands
5532 as normal breakpoint commands; you can thus easily see the effect of
5533 the variable settings.
5535 @item set disconnected-dprintf on
5536 @itemx set disconnected-dprintf off
5537 @kindex set disconnected-dprintf
5538 Choose whether @code{dprintf} commands should continue to run if
5539 @value{GDBN} has disconnected from the target. This only applies
5540 if the @code{dprintf-style} is @code{agent}.
5542 @item show disconnected-dprintf off
5543 @kindex show disconnected-dprintf
5544 Show the current choice for disconnected @code{dprintf}.
5548 @value{GDBN} does not check the validity of function and channel,
5549 relying on you to supply values that are meaningful for the contexts
5550 in which they are being used. For instance, the function and channel
5551 may be the values of local variables, but if that is the case, then
5552 all enabled dynamic prints must be at locations within the scope of
5553 those locals. If evaluation fails, @value{GDBN} will report an error.
5555 @node Save Breakpoints
5556 @subsection How to save breakpoints to a file
5558 To save breakpoint definitions to a file use the @w{@code{save
5559 breakpoints}} command.
5562 @kindex save breakpoints
5563 @cindex save breakpoints to a file for future sessions
5564 @item save breakpoints [@var{filename}]
5565 This command saves all current breakpoint definitions together with
5566 their commands and ignore counts, into a file @file{@var{filename}}
5567 suitable for use in a later debugging session. This includes all
5568 types of breakpoints (breakpoints, watchpoints, catchpoints,
5569 tracepoints). To read the saved breakpoint definitions, use the
5570 @code{source} command (@pxref{Command Files}). Note that watchpoints
5571 with expressions involving local variables may fail to be recreated
5572 because it may not be possible to access the context where the
5573 watchpoint is valid anymore. Because the saved breakpoint definitions
5574 are simply a sequence of @value{GDBN} commands that recreate the
5575 breakpoints, you can edit the file in your favorite editing program,
5576 and remove the breakpoint definitions you're not interested in, or
5577 that can no longer be recreated.
5580 @node Static Probe Points
5581 @subsection Static Probe Points
5583 @cindex static probe point, SystemTap
5584 @cindex static probe point, DTrace
5585 @value{GDBN} supports @dfn{SDT} probes in the code. @acronym{SDT} stands
5586 for Statically Defined Tracing, and the probes are designed to have a tiny
5587 runtime code and data footprint, and no dynamic relocations.
5589 Currently, the following types of probes are supported on
5590 ELF-compatible systems:
5594 @item @code{SystemTap} (@uref{http://sourceware.org/systemtap/})
5595 @acronym{SDT} probes@footnote{See
5596 @uref{http://sourceware.org/systemtap/wiki/AddingUserSpaceProbingToApps}
5597 for more information on how to add @code{SystemTap} @acronym{SDT}
5598 probes in your applications.}. @code{SystemTap} probes are usable
5599 from assembly, C and C@t{++} languages@footnote{See
5600 @uref{http://sourceware.org/systemtap/wiki/UserSpaceProbeImplementation}
5601 for a good reference on how the @acronym{SDT} probes are implemented.}.
5603 @item @code{DTrace} (@uref{http://oss.oracle.com/projects/DTrace})
5604 @acronym{USDT} probes. @code{DTrace} probes are usable from C and
5608 @cindex semaphores on static probe points
5609 Some @code{SystemTap} probes have an associated semaphore variable;
5610 for instance, this happens automatically if you defined your probe
5611 using a DTrace-style @file{.d} file. If your probe has a semaphore,
5612 @value{GDBN} will automatically enable it when you specify a
5613 breakpoint using the @samp{-probe-stap} notation. But, if you put a
5614 breakpoint at a probe's location by some other method (e.g.,
5615 @code{break file:line}), then @value{GDBN} will not automatically set
5616 the semaphore. @code{DTrace} probes do not support semaphores.
5618 You can examine the available static static probes using @code{info
5619 probes}, with optional arguments:
5623 @item info probes @r{[}@var{type}@r{]} @r{[}@var{provider} @r{[}@var{name} @r{[}@var{objfile}@r{]}@r{]}@r{]}
5624 If given, @var{type} is either @code{stap} for listing
5625 @code{SystemTap} probes or @code{dtrace} for listing @code{DTrace}
5626 probes. If omitted all probes are listed regardless of their types.
5628 If given, @var{provider} is a regular expression used to match against provider
5629 names when selecting which probes to list. If omitted, probes by all
5630 probes from all providers are listed.
5632 If given, @var{name} is a regular expression to match against probe names
5633 when selecting which probes to list. If omitted, probe names are not
5634 considered when deciding whether to display them.
5636 If given, @var{objfile} is a regular expression used to select which
5637 object files (executable or shared libraries) to examine. If not
5638 given, all object files are considered.
5640 @item info probes all
5641 List the available static probes, from all types.
5644 @cindex enabling and disabling probes
5645 Some probe points can be enabled and/or disabled. The effect of
5646 enabling or disabling a probe depends on the type of probe being
5647 handled. Some @code{DTrace} probes can be enabled or
5648 disabled, but @code{SystemTap} probes cannot be disabled.
5650 You can enable (or disable) one or more probes using the following
5651 commands, with optional arguments:
5654 @kindex enable probes
5655 @item enable probes @r{[}@var{provider} @r{[}@var{name} @r{[}@var{objfile}@r{]}@r{]}@r{]}
5656 If given, @var{provider} is a regular expression used to match against
5657 provider names when selecting which probes to enable. If omitted,
5658 all probes from all providers are enabled.
5660 If given, @var{name} is a regular expression to match against probe
5661 names when selecting which probes to enable. If omitted, probe names
5662 are not considered when deciding whether to enable them.
5664 If given, @var{objfile} is a regular expression used to select which
5665 object files (executable or shared libraries) to examine. If not
5666 given, all object files are considered.
5668 @kindex disable probes
5669 @item disable probes @r{[}@var{provider} @r{[}@var{name} @r{[}@var{objfile}@r{]}@r{]}@r{]}
5670 See the @code{enable probes} command above for a description of the
5671 optional arguments accepted by this command.
5674 @vindex $_probe_arg@r{, convenience variable}
5675 A probe may specify up to twelve arguments. These are available at the
5676 point at which the probe is defined---that is, when the current PC is
5677 at the probe's location. The arguments are available using the
5678 convenience variables (@pxref{Convenience Vars})
5679 @code{$_probe_arg0}@dots{}@code{$_probe_arg11}. In @code{SystemTap}
5680 probes each probe argument is an integer of the appropriate size;
5681 types are not preserved. In @code{DTrace} probes types are preserved
5682 provided that they are recognized as such by @value{GDBN}; otherwise
5683 the value of the probe argument will be a long integer. The
5684 convenience variable @code{$_probe_argc} holds the number of arguments
5685 at the current probe point.
5687 These variables are always available, but attempts to access them at
5688 any location other than a probe point will cause @value{GDBN} to give
5692 @c @ifclear BARETARGET
5693 @node Error in Breakpoints
5694 @subsection ``Cannot insert breakpoints''
5696 If you request too many active hardware-assisted breakpoints and
5697 watchpoints, you will see this error message:
5699 @c FIXME: the precise wording of this message may change; the relevant
5700 @c source change is not committed yet (Sep 3, 1999).
5702 Stopped; cannot insert breakpoints.
5703 You may have requested too many hardware breakpoints and watchpoints.
5707 This message is printed when you attempt to resume the program, since
5708 only then @value{GDBN} knows exactly how many hardware breakpoints and
5709 watchpoints it needs to insert.
5711 When this message is printed, you need to disable or remove some of the
5712 hardware-assisted breakpoints and watchpoints, and then continue.
5714 @node Breakpoint-related Warnings
5715 @subsection ``Breakpoint address adjusted...''
5716 @cindex breakpoint address adjusted
5718 Some processor architectures place constraints on the addresses at
5719 which breakpoints may be placed. For architectures thus constrained,
5720 @value{GDBN} will attempt to adjust the breakpoint's address to comply
5721 with the constraints dictated by the architecture.
5723 One example of such an architecture is the Fujitsu FR-V. The FR-V is
5724 a VLIW architecture in which a number of RISC-like instructions may be
5725 bundled together for parallel execution. The FR-V architecture
5726 constrains the location of a breakpoint instruction within such a
5727 bundle to the instruction with the lowest address. @value{GDBN}
5728 honors this constraint by adjusting a breakpoint's address to the
5729 first in the bundle.
5731 It is not uncommon for optimized code to have bundles which contain
5732 instructions from different source statements, thus it may happen that
5733 a breakpoint's address will be adjusted from one source statement to
5734 another. Since this adjustment may significantly alter @value{GDBN}'s
5735 breakpoint related behavior from what the user expects, a warning is
5736 printed when the breakpoint is first set and also when the breakpoint
5739 A warning like the one below is printed when setting a breakpoint
5740 that's been subject to address adjustment:
5743 warning: Breakpoint address adjusted from 0x00010414 to 0x00010410.
5746 Such warnings are printed both for user settable and @value{GDBN}'s
5747 internal breakpoints. If you see one of these warnings, you should
5748 verify that a breakpoint set at the adjusted address will have the
5749 desired affect. If not, the breakpoint in question may be removed and
5750 other breakpoints may be set which will have the desired behavior.
5751 E.g., it may be sufficient to place the breakpoint at a later
5752 instruction. A conditional breakpoint may also be useful in some
5753 cases to prevent the breakpoint from triggering too often.
5755 @value{GDBN} will also issue a warning when stopping at one of these
5756 adjusted breakpoints:
5759 warning: Breakpoint 1 address previously adjusted from 0x00010414
5763 When this warning is encountered, it may be too late to take remedial
5764 action except in cases where the breakpoint is hit earlier or more
5765 frequently than expected.
5767 @node Continuing and Stepping
5768 @section Continuing and Stepping
5772 @cindex resuming execution
5773 @dfn{Continuing} means resuming program execution until your program
5774 completes normally. In contrast, @dfn{stepping} means executing just
5775 one more ``step'' of your program, where ``step'' may mean either one
5776 line of source code, or one machine instruction (depending on what
5777 particular command you use). Either when continuing or when stepping,
5778 your program may stop even sooner, due to a breakpoint or a signal. (If
5779 it stops due to a signal, you may want to use @code{handle}, or use
5780 @samp{signal 0} to resume execution (@pxref{Signals, ,Signals}),
5781 or you may step into the signal's handler (@pxref{stepping and signal
5786 @kindex c @r{(@code{continue})}
5787 @kindex fg @r{(resume foreground execution)}
5788 @item continue @r{[}@var{ignore-count}@r{]}
5789 @itemx c @r{[}@var{ignore-count}@r{]}
5790 @itemx fg @r{[}@var{ignore-count}@r{]}
5791 Resume program execution, at the address where your program last stopped;
5792 any breakpoints set at that address are bypassed. The optional argument
5793 @var{ignore-count} allows you to specify a further number of times to
5794 ignore a breakpoint at this location; its effect is like that of
5795 @code{ignore} (@pxref{Conditions, ,Break Conditions}).
5797 The argument @var{ignore-count} is meaningful only when your program
5798 stopped due to a breakpoint. At other times, the argument to
5799 @code{continue} is ignored.
5801 The synonyms @code{c} and @code{fg} (for @dfn{foreground}, as the
5802 debugged program is deemed to be the foreground program) are provided
5803 purely for convenience, and have exactly the same behavior as
5807 To resume execution at a different place, you can use @code{return}
5808 (@pxref{Returning, ,Returning from a Function}) to go back to the
5809 calling function; or @code{jump} (@pxref{Jumping, ,Continuing at a
5810 Different Address}) to go to an arbitrary location in your program.
5812 A typical technique for using stepping is to set a breakpoint
5813 (@pxref{Breakpoints, ,Breakpoints; Watchpoints; and Catchpoints}) at the
5814 beginning of the function or the section of your program where a problem
5815 is believed to lie, run your program until it stops at that breakpoint,
5816 and then step through the suspect area, examining the variables that are
5817 interesting, until you see the problem happen.
5821 @kindex s @r{(@code{step})}
5823 Continue running your program until control reaches a different source
5824 line, then stop it and return control to @value{GDBN}. This command is
5825 abbreviated @code{s}.
5828 @c "without debugging information" is imprecise; actually "without line
5829 @c numbers in the debugging information". (gcc -g1 has debugging info but
5830 @c not line numbers). But it seems complex to try to make that
5831 @c distinction here.
5832 @emph{Warning:} If you use the @code{step} command while control is
5833 within a function that was compiled without debugging information,
5834 execution proceeds until control reaches a function that does have
5835 debugging information. Likewise, it will not step into a function which
5836 is compiled without debugging information. To step through functions
5837 without debugging information, use the @code{stepi} command, described
5841 The @code{step} command only stops at the first instruction of a source
5842 line. This prevents the multiple stops that could otherwise occur in
5843 @code{switch} statements, @code{for} loops, etc. @code{step} continues
5844 to stop if a function that has debugging information is called within
5845 the line. In other words, @code{step} @emph{steps inside} any functions
5846 called within the line.
5848 Also, the @code{step} command only enters a function if there is line
5849 number information for the function. Otherwise it acts like the
5850 @code{next} command. This avoids problems when using @code{cc -gl}
5851 on @acronym{MIPS} machines. Previously, @code{step} entered subroutines if there
5852 was any debugging information about the routine.
5854 @item step @var{count}
5855 Continue running as in @code{step}, but do so @var{count} times. If a
5856 breakpoint is reached, or a signal not related to stepping occurs before
5857 @var{count} steps, stepping stops right away.
5860 @kindex n @r{(@code{next})}
5861 @item next @r{[}@var{count}@r{]}
5862 Continue to the next source line in the current (innermost) stack frame.
5863 This is similar to @code{step}, but function calls that appear within
5864 the line of code are executed without stopping. Execution stops when
5865 control reaches a different line of code at the original stack level
5866 that was executing when you gave the @code{next} command. This command
5867 is abbreviated @code{n}.
5869 An argument @var{count} is a repeat count, as for @code{step}.
5872 @c FIX ME!! Do we delete this, or is there a way it fits in with
5873 @c the following paragraph? --- Vctoria
5875 @c @code{next} within a function that lacks debugging information acts like
5876 @c @code{step}, but any function calls appearing within the code of the
5877 @c function are executed without stopping.
5879 The @code{next} command only stops at the first instruction of a
5880 source line. This prevents multiple stops that could otherwise occur in
5881 @code{switch} statements, @code{for} loops, etc.
5883 @kindex set step-mode
5885 @cindex functions without line info, and stepping
5886 @cindex stepping into functions with no line info
5887 @itemx set step-mode on
5888 The @code{set step-mode on} command causes the @code{step} command to
5889 stop at the first instruction of a function which contains no debug line
5890 information rather than stepping over it.
5892 This is useful in cases where you may be interested in inspecting the
5893 machine instructions of a function which has no symbolic info and do not
5894 want @value{GDBN} to automatically skip over this function.
5896 @item set step-mode off
5897 Causes the @code{step} command to step over any functions which contains no
5898 debug information. This is the default.
5900 @item show step-mode
5901 Show whether @value{GDBN} will stop in or step over functions without
5902 source line debug information.
5905 @kindex fin @r{(@code{finish})}
5907 Continue running until just after function in the selected stack frame
5908 returns. Print the returned value (if any). This command can be
5909 abbreviated as @code{fin}.
5911 Contrast this with the @code{return} command (@pxref{Returning,
5912 ,Returning from a Function}).
5914 @kindex set print finish
5915 @kindex show print finish
5916 @item set print finish @r{[}on|off@r{]}
5917 @itemx show print finish
5918 By default the @code{finish} command will show the value that is
5919 returned by the function. This can be disabled using @code{set print
5920 finish off}. When disabled, the value is still entered into the value
5921 history (@pxref{Value History}), but not displayed.
5924 @kindex u @r{(@code{until})}
5925 @cindex run until specified location
5928 Continue running until a source line past the current line, in the
5929 current stack frame, is reached. This command is used to avoid single
5930 stepping through a loop more than once. It is like the @code{next}
5931 command, except that when @code{until} encounters a jump, it
5932 automatically continues execution until the program counter is greater
5933 than the address of the jump.
5935 This means that when you reach the end of a loop after single stepping
5936 though it, @code{until} makes your program continue execution until it
5937 exits the loop. In contrast, a @code{next} command at the end of a loop
5938 simply steps back to the beginning of the loop, which forces you to step
5939 through the next iteration.
5941 @code{until} always stops your program if it attempts to exit the current
5944 @code{until} may produce somewhat counterintuitive results if the order
5945 of machine code does not match the order of the source lines. For
5946 example, in the following excerpt from a debugging session, the @code{f}
5947 (@code{frame}) command shows that execution is stopped at line
5948 @code{206}; yet when we use @code{until}, we get to line @code{195}:
5952 #0 main (argc=4, argv=0xf7fffae8) at m4.c:206
5954 (@value{GDBP}) until
5955 195 for ( ; argc > 0; NEXTARG) @{
5958 This happened because, for execution efficiency, the compiler had
5959 generated code for the loop closure test at the end, rather than the
5960 start, of the loop---even though the test in a C @code{for}-loop is
5961 written before the body of the loop. The @code{until} command appeared
5962 to step back to the beginning of the loop when it advanced to this
5963 expression; however, it has not really gone to an earlier
5964 statement---not in terms of the actual machine code.
5966 @code{until} with no argument works by means of single
5967 instruction stepping, and hence is slower than @code{until} with an
5970 @item until @var{location}
5971 @itemx u @var{location}
5972 Continue running your program until either the specified @var{location} is
5973 reached, or the current stack frame returns. The location is any of
5974 the forms described in @ref{Specify Location}.
5975 This form of the command uses temporary breakpoints, and
5976 hence is quicker than @code{until} without an argument. The specified
5977 location is actually reached only if it is in the current frame. This
5978 implies that @code{until} can be used to skip over recursive function
5979 invocations. For instance in the code below, if the current location is
5980 line @code{96}, issuing @code{until 99} will execute the program up to
5981 line @code{99} in the same invocation of factorial, i.e., after the inner
5982 invocations have returned.
5985 94 int factorial (int value)
5987 96 if (value > 1) @{
5988 97 value *= factorial (value - 1);
5995 @kindex advance @var{location}
5996 @item advance @var{location}
5997 Continue running the program up to the given @var{location}. An argument is
5998 required, which should be of one of the forms described in
5999 @ref{Specify Location}.
6000 Execution will also stop upon exit from the current stack
6001 frame. This command is similar to @code{until}, but @code{advance} will
6002 not skip over recursive function calls, and the target location doesn't
6003 have to be in the same frame as the current one.
6007 @kindex si @r{(@code{stepi})}
6009 @itemx stepi @var{arg}
6011 Execute one machine instruction, then stop and return to the debugger.
6013 It is often useful to do @samp{display/i $pc} when stepping by machine
6014 instructions. This makes @value{GDBN} automatically display the next
6015 instruction to be executed, each time your program stops. @xref{Auto
6016 Display,, Automatic Display}.
6018 An argument is a repeat count, as in @code{step}.
6022 @kindex ni @r{(@code{nexti})}
6024 @itemx nexti @var{arg}
6026 Execute one machine instruction, but if it is a function call,
6027 proceed until the function returns.
6029 An argument is a repeat count, as in @code{next}.
6033 @anchor{range stepping}
6034 @cindex range stepping
6035 @cindex target-assisted range stepping
6036 By default, and if available, @value{GDBN} makes use of
6037 target-assisted @dfn{range stepping}. In other words, whenever you
6038 use a stepping command (e.g., @code{step}, @code{next}), @value{GDBN}
6039 tells the target to step the corresponding range of instruction
6040 addresses instead of issuing multiple single-steps. This speeds up
6041 line stepping, particularly for remote targets. Ideally, there should
6042 be no reason you would want to turn range stepping off. However, it's
6043 possible that a bug in the debug info, a bug in the remote stub (for
6044 remote targets), or even a bug in @value{GDBN} could make line
6045 stepping behave incorrectly when target-assisted range stepping is
6046 enabled. You can use the following command to turn off range stepping
6050 @kindex set range-stepping
6051 @kindex show range-stepping
6052 @item set range-stepping
6053 @itemx show range-stepping
6054 Control whether range stepping is enabled.
6056 If @code{on}, and the target supports it, @value{GDBN} tells the
6057 target to step a range of addresses itself, instead of issuing
6058 multiple single-steps. If @code{off}, @value{GDBN} always issues
6059 single-steps, even if range stepping is supported by the target. The
6060 default is @code{on}.
6064 @node Skipping Over Functions and Files
6065 @section Skipping Over Functions and Files
6066 @cindex skipping over functions and files
6068 The program you are debugging may contain some functions which are
6069 uninteresting to debug. The @code{skip} command lets you tell @value{GDBN} to
6070 skip a function, all functions in a file or a particular function in
6071 a particular file when stepping.
6073 For example, consider the following C function:
6084 Suppose you wish to step into the functions @code{foo} and @code{bar}, but you
6085 are not interested in stepping through @code{boring}. If you run @code{step}
6086 at line 103, you'll enter @code{boring()}, but if you run @code{next}, you'll
6087 step over both @code{foo} and @code{boring}!
6089 One solution is to @code{step} into @code{boring} and use the @code{finish}
6090 command to immediately exit it. But this can become tedious if @code{boring}
6091 is called from many places.
6093 A more flexible solution is to execute @kbd{skip boring}. This instructs
6094 @value{GDBN} never to step into @code{boring}. Now when you execute
6095 @code{step} at line 103, you'll step over @code{boring} and directly into
6098 Functions may be skipped by providing either a function name, linespec
6099 (@pxref{Specify Location}), regular expression that matches the function's
6100 name, file name or a @code{glob}-style pattern that matches the file name.
6102 On Posix systems the form of the regular expression is
6103 ``Extended Regular Expressions''. See for example @samp{man 7 regex}
6104 on @sc{gnu}/Linux systems. On non-Posix systems the form of the regular
6105 expression is whatever is provided by the @code{regcomp} function of
6106 the underlying system.
6107 See for example @samp{man 7 glob} on @sc{gnu}/Linux systems for a
6108 description of @code{glob}-style patterns.
6112 @item skip @r{[}@var{options}@r{]}
6113 The basic form of the @code{skip} command takes zero or more options
6114 that specify what to skip.
6115 The @var{options} argument is any useful combination of the following:
6118 @item -file @var{file}
6119 @itemx -fi @var{file}
6120 Functions in @var{file} will be skipped over when stepping.
6122 @item -gfile @var{file-glob-pattern}
6123 @itemx -gfi @var{file-glob-pattern}
6124 @cindex skipping over files via glob-style patterns
6125 Functions in files matching @var{file-glob-pattern} will be skipped
6129 (@value{GDBP}) skip -gfi utils/*.c
6132 @item -function @var{linespec}
6133 @itemx -fu @var{linespec}
6134 Functions named by @var{linespec} or the function containing the line
6135 named by @var{linespec} will be skipped over when stepping.
6136 @xref{Specify Location}.
6138 @item -rfunction @var{regexp}
6139 @itemx -rfu @var{regexp}
6140 @cindex skipping over functions via regular expressions
6141 Functions whose name matches @var{regexp} will be skipped over when stepping.
6143 This form is useful for complex function names.
6144 For example, there is generally no need to step into C@t{++} @code{std::string}
6145 constructors or destructors. Plus with C@t{++} templates it can be hard to
6146 write out the full name of the function, and often it doesn't matter what
6147 the template arguments are. Specifying the function to be skipped as a
6148 regular expression makes this easier.
6151 (@value{GDBP}) skip -rfu ^std::(allocator|basic_string)<.*>::~?\1 *\(
6154 If you want to skip every templated C@t{++} constructor and destructor
6155 in the @code{std} namespace you can do:
6158 (@value{GDBP}) skip -rfu ^std::([a-zA-z0-9_]+)<.*>::~?\1 *\(
6162 If no options are specified, the function you're currently debugging
6165 @kindex skip function
6166 @item skip function @r{[}@var{linespec}@r{]}
6167 After running this command, the function named by @var{linespec} or the
6168 function containing the line named by @var{linespec} will be skipped over when
6169 stepping. @xref{Specify Location}.
6171 If you do not specify @var{linespec}, the function you're currently debugging
6174 (If you have a function called @code{file} that you want to skip, use
6175 @kbd{skip function file}.)
6178 @item skip file @r{[}@var{filename}@r{]}
6179 After running this command, any function whose source lives in @var{filename}
6180 will be skipped over when stepping.
6183 (@value{GDBP}) skip file boring.c
6184 File boring.c will be skipped when stepping.
6187 If you do not specify @var{filename}, functions whose source lives in the file
6188 you're currently debugging will be skipped.
6191 Skips can be listed, deleted, disabled, and enabled, much like breakpoints.
6192 These are the commands for managing your list of skips:
6196 @item info skip @r{[}@var{range}@r{]}
6197 Print details about the specified skip(s). If @var{range} is not specified,
6198 print a table with details about all functions and files marked for skipping.
6199 @code{info skip} prints the following information about each skip:
6203 A number identifying this skip.
6204 @item Enabled or Disabled
6205 Enabled skips are marked with @samp{y}.
6206 Disabled skips are marked with @samp{n}.
6208 If the file name is a @samp{glob} pattern this is @samp{y}.
6209 Otherwise it is @samp{n}.
6211 The name or @samp{glob} pattern of the file to be skipped.
6212 If no file is specified this is @samp{<none>}.
6214 If the function name is a @samp{regular expression} this is @samp{y}.
6215 Otherwise it is @samp{n}.
6217 The name or regular expression of the function to skip.
6218 If no function is specified this is @samp{<none>}.
6222 @item skip delete @r{[}@var{range}@r{]}
6223 Delete the specified skip(s). If @var{range} is not specified, delete all
6227 @item skip enable @r{[}@var{range}@r{]}
6228 Enable the specified skip(s). If @var{range} is not specified, enable all
6231 @kindex skip disable
6232 @item skip disable @r{[}@var{range}@r{]}
6233 Disable the specified skip(s). If @var{range} is not specified, disable all
6236 @kindex set debug skip
6237 @item set debug skip @r{[}on|off@r{]}
6238 Set whether to print the debug output about skipping files and functions.
6240 @kindex show debug skip
6241 @item show debug skip
6242 Show whether the debug output about skipping files and functions is printed.
6250 A signal is an asynchronous event that can happen in a program. The
6251 operating system defines the possible kinds of signals, and gives each
6252 kind a name and a number. For example, in Unix @code{SIGINT} is the
6253 signal a program gets when you type an interrupt character (often @kbd{Ctrl-c});
6254 @code{SIGSEGV} is the signal a program gets from referencing a place in
6255 memory far away from all the areas in use; @code{SIGALRM} occurs when
6256 the alarm clock timer goes off (which happens only if your program has
6257 requested an alarm).
6259 @cindex fatal signals
6260 Some signals, including @code{SIGALRM}, are a normal part of the
6261 functioning of your program. Others, such as @code{SIGSEGV}, indicate
6262 errors; these signals are @dfn{fatal} (they kill your program immediately) if the
6263 program has not specified in advance some other way to handle the signal.
6264 @code{SIGINT} does not indicate an error in your program, but it is normally
6265 fatal so it can carry out the purpose of the interrupt: to kill the program.
6267 @value{GDBN} has the ability to detect any occurrence of a signal in your
6268 program. You can tell @value{GDBN} in advance what to do for each kind of
6271 @cindex handling signals
6272 Normally, @value{GDBN} is set up to let the non-erroneous signals like
6273 @code{SIGALRM} be silently passed to your program
6274 (so as not to interfere with their role in the program's functioning)
6275 but to stop your program immediately whenever an error signal happens.
6276 You can change these settings with the @code{handle} command.
6279 @kindex info signals
6283 Print a table of all the kinds of signals and how @value{GDBN} has been told to
6284 handle each one. You can use this to see the signal numbers of all
6285 the defined types of signals.
6287 @item info signals @var{sig}
6288 Similar, but print information only about the specified signal number.
6290 @code{info handle} is an alias for @code{info signals}.
6292 @item catch signal @r{[}@var{signal}@dots{} @r{|} @samp{all}@r{]}
6293 Set a catchpoint for the indicated signals. @xref{Set Catchpoints},
6294 for details about this command.
6297 @item handle @var{signal} @r{[}@var{keywords}@dots{}@r{]}
6298 Change the way @value{GDBN} handles signal @var{signal}. The @var{signal}
6299 can be the number of a signal or its name (with or without the
6300 @samp{SIG} at the beginning); a list of signal numbers of the form
6301 @samp{@var{low}-@var{high}}; or the word @samp{all}, meaning all the
6302 known signals. Optional arguments @var{keywords}, described below,
6303 say what change to make.
6307 The keywords allowed by the @code{handle} command can be abbreviated.
6308 Their full names are:
6312 @value{GDBN} should not stop your program when this signal happens. It may
6313 still print a message telling you that the signal has come in.
6316 @value{GDBN} should stop your program when this signal happens. This implies
6317 the @code{print} keyword as well.
6320 @value{GDBN} should print a message when this signal happens.
6323 @value{GDBN} should not mention the occurrence of the signal at all. This
6324 implies the @code{nostop} keyword as well.
6328 @value{GDBN} should allow your program to see this signal; your program
6329 can handle the signal, or else it may terminate if the signal is fatal
6330 and not handled. @code{pass} and @code{noignore} are synonyms.
6334 @value{GDBN} should not allow your program to see this signal.
6335 @code{nopass} and @code{ignore} are synonyms.
6339 When a signal stops your program, the signal is not visible to the
6341 continue. Your program sees the signal then, if @code{pass} is in
6342 effect for the signal in question @emph{at that time}. In other words,
6343 after @value{GDBN} reports a signal, you can use the @code{handle}
6344 command with @code{pass} or @code{nopass} to control whether your
6345 program sees that signal when you continue.
6347 The default is set to @code{nostop}, @code{noprint}, @code{pass} for
6348 non-erroneous signals such as @code{SIGALRM}, @code{SIGWINCH} and
6349 @code{SIGCHLD}, and to @code{stop}, @code{print}, @code{pass} for the
6352 You can also use the @code{signal} command to prevent your program from
6353 seeing a signal, or cause it to see a signal it normally would not see,
6354 or to give it any signal at any time. For example, if your program stopped
6355 due to some sort of memory reference error, you might store correct
6356 values into the erroneous variables and continue, hoping to see more
6357 execution; but your program would probably terminate immediately as
6358 a result of the fatal signal once it saw the signal. To prevent this,
6359 you can continue with @samp{signal 0}. @xref{Signaling, ,Giving your
6362 @cindex stepping and signal handlers
6363 @anchor{stepping and signal handlers}
6365 @value{GDBN} optimizes for stepping the mainline code. If a signal
6366 that has @code{handle nostop} and @code{handle pass} set arrives while
6367 a stepping command (e.g., @code{stepi}, @code{step}, @code{next}) is
6368 in progress, @value{GDBN} lets the signal handler run and then resumes
6369 stepping the mainline code once the signal handler returns. In other
6370 words, @value{GDBN} steps over the signal handler. This prevents
6371 signals that you've specified as not interesting (with @code{handle
6372 nostop}) from changing the focus of debugging unexpectedly. Note that
6373 the signal handler itself may still hit a breakpoint, stop for another
6374 signal that has @code{handle stop} in effect, or for any other event
6375 that normally results in stopping the stepping command sooner. Also
6376 note that @value{GDBN} still informs you that the program received a
6377 signal if @code{handle print} is set.
6379 @anchor{stepping into signal handlers}
6381 If you set @code{handle pass} for a signal, and your program sets up a
6382 handler for it, then issuing a stepping command, such as @code{step}
6383 or @code{stepi}, when your program is stopped due to the signal will
6384 step @emph{into} the signal handler (if the target supports that).
6386 Likewise, if you use the @code{queue-signal} command to queue a signal
6387 to be delivered to the current thread when execution of the thread
6388 resumes (@pxref{Signaling, ,Giving your Program a Signal}), then a
6389 stepping command will step into the signal handler.
6391 Here's an example, using @code{stepi} to step to the first instruction
6392 of @code{SIGUSR1}'s handler:
6395 (@value{GDBP}) handle SIGUSR1
6396 Signal Stop Print Pass to program Description
6397 SIGUSR1 Yes Yes Yes User defined signal 1
6401 Program received signal SIGUSR1, User defined signal 1.
6402 main () sigusr1.c:28
6405 sigusr1_handler () at sigusr1.c:9
6409 The same, but using @code{queue-signal} instead of waiting for the
6410 program to receive the signal first:
6415 (@value{GDBP}) queue-signal SIGUSR1
6417 sigusr1_handler () at sigusr1.c:9
6422 @cindex extra signal information
6423 @anchor{extra signal information}
6425 On some targets, @value{GDBN} can inspect extra signal information
6426 associated with the intercepted signal, before it is actually
6427 delivered to the program being debugged. This information is exported
6428 by the convenience variable @code{$_siginfo}, and consists of data
6429 that is passed by the kernel to the signal handler at the time of the
6430 receipt of a signal. The data type of the information itself is
6431 target dependent. You can see the data type using the @code{ptype
6432 $_siginfo} command. On Unix systems, it typically corresponds to the
6433 standard @code{siginfo_t} type, as defined in the @file{signal.h}
6436 Here's an example, on a @sc{gnu}/Linux system, printing the stray
6437 referenced address that raised a segmentation fault.
6441 (@value{GDBP}) continue
6442 Program received signal SIGSEGV, Segmentation fault.
6443 0x0000000000400766 in main ()
6445 (@value{GDBP}) ptype $_siginfo
6452 struct @{...@} _kill;
6453 struct @{...@} _timer;
6455 struct @{...@} _sigchld;
6456 struct @{...@} _sigfault;
6457 struct @{...@} _sigpoll;
6460 (@value{GDBP}) ptype $_siginfo._sifields._sigfault
6464 (@value{GDBP}) p $_siginfo._sifields._sigfault.si_addr
6465 $1 = (void *) 0x7ffff7ff7000
6469 Depending on target support, @code{$_siginfo} may also be writable.
6471 @cindex Intel MPX boundary violations
6472 @cindex boundary violations, Intel MPX
6473 On some targets, a @code{SIGSEGV} can be caused by a boundary
6474 violation, i.e., accessing an address outside of the allowed range.
6475 In those cases @value{GDBN} may displays additional information,
6476 depending on how @value{GDBN} has been told to handle the signal.
6477 With @code{handle stop SIGSEGV}, @value{GDBN} displays the violation
6478 kind: "Upper" or "Lower", the memory address accessed and the
6479 bounds, while with @code{handle nostop SIGSEGV} no additional
6480 information is displayed.
6482 The usual output of a segfault is:
6484 Program received signal SIGSEGV, Segmentation fault
6485 0x0000000000400d7c in upper () at i386-mpx-sigsegv.c:68
6486 68 value = *(p + len);
6489 While a bound violation is presented as:
6491 Program received signal SIGSEGV, Segmentation fault
6492 Upper bound violation while accessing address 0x7fffffffc3b3
6493 Bounds: [lower = 0x7fffffffc390, upper = 0x7fffffffc3a3]
6494 0x0000000000400d7c in upper () at i386-mpx-sigsegv.c:68
6495 68 value = *(p + len);
6499 @section Stopping and Starting Multi-thread Programs
6501 @cindex stopped threads
6502 @cindex threads, stopped
6504 @cindex continuing threads
6505 @cindex threads, continuing
6507 @value{GDBN} supports debugging programs with multiple threads
6508 (@pxref{Threads,, Debugging Programs with Multiple Threads}). There
6509 are two modes of controlling execution of your program within the
6510 debugger. In the default mode, referred to as @dfn{all-stop mode},
6511 when any thread in your program stops (for example, at a breakpoint
6512 or while being stepped), all other threads in the program are also stopped by
6513 @value{GDBN}. On some targets, @value{GDBN} also supports
6514 @dfn{non-stop mode}, in which other threads can continue to run freely while
6515 you examine the stopped thread in the debugger.
6518 * All-Stop Mode:: All threads stop when GDB takes control
6519 * Non-Stop Mode:: Other threads continue to execute
6520 * Background Execution:: Running your program asynchronously
6521 * Thread-Specific Breakpoints:: Controlling breakpoints
6522 * Interrupted System Calls:: GDB may interfere with system calls
6523 * Observer Mode:: GDB does not alter program behavior
6527 @subsection All-Stop Mode
6529 @cindex all-stop mode
6531 In all-stop mode, whenever your program stops under @value{GDBN} for any reason,
6532 @emph{all} threads of execution stop, not just the current thread. This
6533 allows you to examine the overall state of the program, including
6534 switching between threads, without worrying that things may change
6537 Conversely, whenever you restart the program, @emph{all} threads start
6538 executing. @emph{This is true even when single-stepping} with commands
6539 like @code{step} or @code{next}.
6541 In particular, @value{GDBN} cannot single-step all threads in lockstep.
6542 Since thread scheduling is up to your debugging target's operating
6543 system (not controlled by @value{GDBN}), other threads may
6544 execute more than one statement while the current thread completes a
6545 single step. Moreover, in general other threads stop in the middle of a
6546 statement, rather than at a clean statement boundary, when the program
6549 You might even find your program stopped in another thread after
6550 continuing or even single-stepping. This happens whenever some other
6551 thread runs into a breakpoint, a signal, or an exception before the
6552 first thread completes whatever you requested.
6554 @cindex automatic thread selection
6555 @cindex switching threads automatically
6556 @cindex threads, automatic switching
6557 Whenever @value{GDBN} stops your program, due to a breakpoint or a
6558 signal, it automatically selects the thread where that breakpoint or
6559 signal happened. @value{GDBN} alerts you to the context switch with a
6560 message such as @samp{[Switching to Thread @var{n}]} to identify the
6563 On some OSes, you can modify @value{GDBN}'s default behavior by
6564 locking the OS scheduler to allow only a single thread to run.
6567 @item set scheduler-locking @var{mode}
6568 @cindex scheduler locking mode
6569 @cindex lock scheduler
6570 Set the scheduler locking mode. It applies to normal execution,
6571 record mode, and replay mode. If it is @code{off}, then there is no
6572 locking and any thread may run at any time. If @code{on}, then only
6573 the current thread may run when the inferior is resumed. The
6574 @code{step} mode optimizes for single-stepping; it prevents other
6575 threads from preempting the current thread while you are stepping, so
6576 that the focus of debugging does not change unexpectedly. Other
6577 threads never get a chance to run when you step, and they are
6578 completely free to run when you use commands like @samp{continue},
6579 @samp{until}, or @samp{finish}. However, unless another thread hits a
6580 breakpoint during its timeslice, @value{GDBN} does not change the
6581 current thread away from the thread that you are debugging. The
6582 @code{replay} mode behaves like @code{off} in record mode and like
6583 @code{on} in replay mode.
6585 @item show scheduler-locking
6586 Display the current scheduler locking mode.
6589 @cindex resume threads of multiple processes simultaneously
6590 By default, when you issue one of the execution commands such as
6591 @code{continue}, @code{next} or @code{step}, @value{GDBN} allows only
6592 threads of the current inferior to run. For example, if @value{GDBN}
6593 is attached to two inferiors, each with two threads, the
6594 @code{continue} command resumes only the two threads of the current
6595 inferior. This is useful, for example, when you debug a program that
6596 forks and you want to hold the parent stopped (so that, for instance,
6597 it doesn't run to exit), while you debug the child. In other
6598 situations, you may not be interested in inspecting the current state
6599 of any of the processes @value{GDBN} is attached to, and you may want
6600 to resume them all until some breakpoint is hit. In the latter case,
6601 you can instruct @value{GDBN} to allow all threads of all the
6602 inferiors to run with the @w{@code{set schedule-multiple}} command.
6605 @kindex set schedule-multiple
6606 @item set schedule-multiple
6607 Set the mode for allowing threads of multiple processes to be resumed
6608 when an execution command is issued. When @code{on}, all threads of
6609 all processes are allowed to run. When @code{off}, only the threads
6610 of the current process are resumed. The default is @code{off}. The
6611 @code{scheduler-locking} mode takes precedence when set to @code{on},
6612 or while you are stepping and set to @code{step}.
6614 @item show schedule-multiple
6615 Display the current mode for resuming the execution of threads of
6620 @subsection Non-Stop Mode
6622 @cindex non-stop mode
6624 @c This section is really only a place-holder, and needs to be expanded
6625 @c with more details.
6627 For some multi-threaded targets, @value{GDBN} supports an optional
6628 mode of operation in which you can examine stopped program threads in
6629 the debugger while other threads continue to execute freely. This
6630 minimizes intrusion when debugging live systems, such as programs
6631 where some threads have real-time constraints or must continue to
6632 respond to external events. This is referred to as @dfn{non-stop} mode.
6634 In non-stop mode, when a thread stops to report a debugging event,
6635 @emph{only} that thread is stopped; @value{GDBN} does not stop other
6636 threads as well, in contrast to the all-stop mode behavior. Additionally,
6637 execution commands such as @code{continue} and @code{step} apply by default
6638 only to the current thread in non-stop mode, rather than all threads as
6639 in all-stop mode. This allows you to control threads explicitly in
6640 ways that are not possible in all-stop mode --- for example, stepping
6641 one thread while allowing others to run freely, stepping
6642 one thread while holding all others stopped, or stepping several threads
6643 independently and simultaneously.
6645 To enter non-stop mode, use this sequence of commands before you run
6646 or attach to your program:
6649 # If using the CLI, pagination breaks non-stop.
6652 # Finally, turn it on!
6656 You can use these commands to manipulate the non-stop mode setting:
6659 @kindex set non-stop
6660 @item set non-stop on
6661 Enable selection of non-stop mode.
6662 @item set non-stop off
6663 Disable selection of non-stop mode.
6664 @kindex show non-stop
6666 Show the current non-stop enablement setting.
6669 Note these commands only reflect whether non-stop mode is enabled,
6670 not whether the currently-executing program is being run in non-stop mode.
6671 In particular, the @code{set non-stop} preference is only consulted when
6672 @value{GDBN} starts or connects to the target program, and it is generally
6673 not possible to switch modes once debugging has started. Furthermore,
6674 since not all targets support non-stop mode, even when you have enabled
6675 non-stop mode, @value{GDBN} may still fall back to all-stop operation by
6678 In non-stop mode, all execution commands apply only to the current thread
6679 by default. That is, @code{continue} only continues one thread.
6680 To continue all threads, issue @code{continue -a} or @code{c -a}.
6682 You can use @value{GDBN}'s background execution commands
6683 (@pxref{Background Execution}) to run some threads in the background
6684 while you continue to examine or step others from @value{GDBN}.
6685 The MI execution commands (@pxref{GDB/MI Program Execution}) are
6686 always executed asynchronously in non-stop mode.
6688 Suspending execution is done with the @code{interrupt} command when
6689 running in the background, or @kbd{Ctrl-c} during foreground execution.
6690 In all-stop mode, this stops the whole process;
6691 but in non-stop mode the interrupt applies only to the current thread.
6692 To stop the whole program, use @code{interrupt -a}.
6694 Other execution commands do not currently support the @code{-a} option.
6696 In non-stop mode, when a thread stops, @value{GDBN} doesn't automatically make
6697 that thread current, as it does in all-stop mode. This is because the
6698 thread stop notifications are asynchronous with respect to @value{GDBN}'s
6699 command interpreter, and it would be confusing if @value{GDBN} unexpectedly
6700 changed to a different thread just as you entered a command to operate on the
6701 previously current thread.
6703 @node Background Execution
6704 @subsection Background Execution
6706 @cindex foreground execution
6707 @cindex background execution
6708 @cindex asynchronous execution
6709 @cindex execution, foreground, background and asynchronous
6711 @value{GDBN}'s execution commands have two variants: the normal
6712 foreground (synchronous) behavior, and a background
6713 (asynchronous) behavior. In foreground execution, @value{GDBN} waits for
6714 the program to report that some thread has stopped before prompting for
6715 another command. In background execution, @value{GDBN} immediately gives
6716 a command prompt so that you can issue other commands while your program runs.
6718 If the target doesn't support async mode, @value{GDBN} issues an error
6719 message if you attempt to use the background execution commands.
6721 @cindex @code{&}, background execution of commands
6722 To specify background execution, add a @code{&} to the command. For example,
6723 the background form of the @code{continue} command is @code{continue&}, or
6724 just @code{c&}. The execution commands that accept background execution
6730 @xref{Starting, , Starting your Program}.
6734 @xref{Attach, , Debugging an Already-running Process}.
6738 @xref{Continuing and Stepping, step}.
6742 @xref{Continuing and Stepping, stepi}.
6746 @xref{Continuing and Stepping, next}.
6750 @xref{Continuing and Stepping, nexti}.
6754 @xref{Continuing and Stepping, continue}.
6758 @xref{Continuing and Stepping, finish}.
6762 @xref{Continuing and Stepping, until}.
6766 Background execution is especially useful in conjunction with non-stop
6767 mode for debugging programs with multiple threads; see @ref{Non-Stop Mode}.
6768 However, you can also use these commands in the normal all-stop mode with
6769 the restriction that you cannot issue another execution command until the
6770 previous one finishes. Examples of commands that are valid in all-stop
6771 mode while the program is running include @code{help} and @code{info break}.
6773 You can interrupt your program while it is running in the background by
6774 using the @code{interrupt} command.
6781 Suspend execution of the running program. In all-stop mode,
6782 @code{interrupt} stops the whole process, but in non-stop mode, it stops
6783 only the current thread. To stop the whole program in non-stop mode,
6784 use @code{interrupt -a}.
6787 @node Thread-Specific Breakpoints
6788 @subsection Thread-Specific Breakpoints
6790 When your program has multiple threads (@pxref{Threads,, Debugging
6791 Programs with Multiple Threads}), you can choose whether to set
6792 breakpoints on all threads, or on a particular thread.
6795 @cindex breakpoints and threads
6796 @cindex thread breakpoints
6797 @kindex break @dots{} thread @var{thread-id}
6798 @item break @var{location} thread @var{thread-id}
6799 @itemx break @var{location} thread @var{thread-id} if @dots{}
6800 @var{location} specifies source lines; there are several ways of
6801 writing them (@pxref{Specify Location}), but the effect is always to
6802 specify some source line.
6804 Use the qualifier @samp{thread @var{thread-id}} with a breakpoint command
6805 to specify that you only want @value{GDBN} to stop the program when a
6806 particular thread reaches this breakpoint. The @var{thread-id} specifier
6807 is one of the thread identifiers assigned by @value{GDBN}, shown
6808 in the first column of the @samp{info threads} display.
6810 If you do not specify @samp{thread @var{thread-id}} when you set a
6811 breakpoint, the breakpoint applies to @emph{all} threads of your
6814 You can use the @code{thread} qualifier on conditional breakpoints as
6815 well; in this case, place @samp{thread @var{thread-id}} before or
6816 after the breakpoint condition, like this:
6819 (@value{GDBP}) break frik.c:13 thread 28 if bartab > lim
6824 Thread-specific breakpoints are automatically deleted when
6825 @value{GDBN} detects the corresponding thread is no longer in the
6826 thread list. For example:
6830 Thread-specific breakpoint 3 deleted - thread 28 no longer in the thread list.
6833 There are several ways for a thread to disappear, such as a regular
6834 thread exit, but also when you detach from the process with the
6835 @code{detach} command (@pxref{Attach, ,Debugging an Already-running
6836 Process}), or if @value{GDBN} loses the remote connection
6837 (@pxref{Remote Debugging}), etc. Note that with some targets,
6838 @value{GDBN} is only able to detect a thread has exited when the user
6839 explictly asks for the thread list with the @code{info threads}
6842 @node Interrupted System Calls
6843 @subsection Interrupted System Calls
6845 @cindex thread breakpoints and system calls
6846 @cindex system calls and thread breakpoints
6847 @cindex premature return from system calls
6848 There is an unfortunate side effect when using @value{GDBN} to debug
6849 multi-threaded programs. If one thread stops for a
6850 breakpoint, or for some other reason, and another thread is blocked in a
6851 system call, then the system call may return prematurely. This is a
6852 consequence of the interaction between multiple threads and the signals
6853 that @value{GDBN} uses to implement breakpoints and other events that
6856 To handle this problem, your program should check the return value of
6857 each system call and react appropriately. This is good programming
6860 For example, do not write code like this:
6866 The call to @code{sleep} will return early if a different thread stops
6867 at a breakpoint or for some other reason.
6869 Instead, write this:
6874 unslept = sleep (unslept);
6877 A system call is allowed to return early, so the system is still
6878 conforming to its specification. But @value{GDBN} does cause your
6879 multi-threaded program to behave differently than it would without
6882 Also, @value{GDBN} uses internal breakpoints in the thread library to
6883 monitor certain events such as thread creation and thread destruction.
6884 When such an event happens, a system call in another thread may return
6885 prematurely, even though your program does not appear to stop.
6888 @subsection Observer Mode
6890 If you want to build on non-stop mode and observe program behavior
6891 without any chance of disruption by @value{GDBN}, you can set
6892 variables to disable all of the debugger's attempts to modify state,
6893 whether by writing memory, inserting breakpoints, etc. These operate
6894 at a low level, intercepting operations from all commands.
6896 When all of these are set to @code{off}, then @value{GDBN} is said to
6897 be @dfn{observer mode}. As a convenience, the variable
6898 @code{observer} can be set to disable these, plus enable non-stop
6901 Note that @value{GDBN} will not prevent you from making nonsensical
6902 combinations of these settings. For instance, if you have enabled
6903 @code{may-insert-breakpoints} but disabled @code{may-write-memory},
6904 then breakpoints that work by writing trap instructions into the code
6905 stream will still not be able to be placed.
6910 @item set observer on
6911 @itemx set observer off
6912 When set to @code{on}, this disables all the permission variables
6913 below (except for @code{insert-fast-tracepoints}), plus enables
6914 non-stop debugging. Setting this to @code{off} switches back to
6915 normal debugging, though remaining in non-stop mode.
6918 Show whether observer mode is on or off.
6920 @kindex may-write-registers
6921 @item set may-write-registers on
6922 @itemx set may-write-registers off
6923 This controls whether @value{GDBN} will attempt to alter the values of
6924 registers, such as with assignment expressions in @code{print}, or the
6925 @code{jump} command. It defaults to @code{on}.
6927 @item show may-write-registers
6928 Show the current permission to write registers.
6930 @kindex may-write-memory
6931 @item set may-write-memory on
6932 @itemx set may-write-memory off
6933 This controls whether @value{GDBN} will attempt to alter the contents
6934 of memory, such as with assignment expressions in @code{print}. It
6935 defaults to @code{on}.
6937 @item show may-write-memory
6938 Show the current permission to write memory.
6940 @kindex may-insert-breakpoints
6941 @item set may-insert-breakpoints on
6942 @itemx set may-insert-breakpoints off
6943 This controls whether @value{GDBN} will attempt to insert breakpoints.
6944 This affects all breakpoints, including internal breakpoints defined
6945 by @value{GDBN}. It defaults to @code{on}.
6947 @item show may-insert-breakpoints
6948 Show the current permission to insert breakpoints.
6950 @kindex may-insert-tracepoints
6951 @item set may-insert-tracepoints on
6952 @itemx set may-insert-tracepoints off
6953 This controls whether @value{GDBN} will attempt to insert (regular)
6954 tracepoints at the beginning of a tracing experiment. It affects only
6955 non-fast tracepoints, fast tracepoints being under the control of
6956 @code{may-insert-fast-tracepoints}. It defaults to @code{on}.
6958 @item show may-insert-tracepoints
6959 Show the current permission to insert tracepoints.
6961 @kindex may-insert-fast-tracepoints
6962 @item set may-insert-fast-tracepoints on
6963 @itemx set may-insert-fast-tracepoints off
6964 This controls whether @value{GDBN} will attempt to insert fast
6965 tracepoints at the beginning of a tracing experiment. It affects only
6966 fast tracepoints, regular (non-fast) tracepoints being under the
6967 control of @code{may-insert-tracepoints}. It defaults to @code{on}.
6969 @item show may-insert-fast-tracepoints
6970 Show the current permission to insert fast tracepoints.
6972 @kindex may-interrupt
6973 @item set may-interrupt on
6974 @itemx set may-interrupt off
6975 This controls whether @value{GDBN} will attempt to interrupt or stop
6976 program execution. When this variable is @code{off}, the
6977 @code{interrupt} command will have no effect, nor will
6978 @kbd{Ctrl-c}. It defaults to @code{on}.
6980 @item show may-interrupt
6981 Show the current permission to interrupt or stop the program.
6985 @node Reverse Execution
6986 @chapter Running programs backward
6987 @cindex reverse execution
6988 @cindex running programs backward
6990 When you are debugging a program, it is not unusual to realize that
6991 you have gone too far, and some event of interest has already happened.
6992 If the target environment supports it, @value{GDBN} can allow you to
6993 ``rewind'' the program by running it backward.
6995 A target environment that supports reverse execution should be able
6996 to ``undo'' the changes in machine state that have taken place as the
6997 program was executing normally. Variables, registers etc.@: should
6998 revert to their previous values. Obviously this requires a great
6999 deal of sophistication on the part of the target environment; not
7000 all target environments can support reverse execution.
7002 When a program is executed in reverse, the instructions that
7003 have most recently been executed are ``un-executed'', in reverse
7004 order. The program counter runs backward, following the previous
7005 thread of execution in reverse. As each instruction is ``un-executed'',
7006 the values of memory and/or registers that were changed by that
7007 instruction are reverted to their previous states. After executing
7008 a piece of source code in reverse, all side effects of that code
7009 should be ``undone'', and all variables should be returned to their
7010 prior values@footnote{
7011 Note that some side effects are easier to undo than others. For instance,
7012 memory and registers are relatively easy, but device I/O is hard. Some
7013 targets may be able undo things like device I/O, and some may not.
7015 The contract between @value{GDBN} and the reverse executing target
7016 requires only that the target do something reasonable when
7017 @value{GDBN} tells it to execute backwards, and then report the
7018 results back to @value{GDBN}. Whatever the target reports back to
7019 @value{GDBN}, @value{GDBN} will report back to the user. @value{GDBN}
7020 assumes that the memory and registers that the target reports are in a
7021 consistent state, but @value{GDBN} accepts whatever it is given.
7024 On some platforms, @value{GDBN} has built-in support for reverse
7025 execution, activated with the @code{record} or @code{record btrace}
7026 commands. @xref{Process Record and Replay}. Some remote targets,
7027 typically full system emulators, support reverse execution directly
7028 without requiring any special command.
7030 If you are debugging in a target environment that supports
7031 reverse execution, @value{GDBN} provides the following commands.
7034 @kindex reverse-continue
7035 @kindex rc @r{(@code{reverse-continue})}
7036 @item reverse-continue @r{[}@var{ignore-count}@r{]}
7037 @itemx rc @r{[}@var{ignore-count}@r{]}
7038 Beginning at the point where your program last stopped, start executing
7039 in reverse. Reverse execution will stop for breakpoints and synchronous
7040 exceptions (signals), just like normal execution. Behavior of
7041 asynchronous signals depends on the target environment.
7043 @kindex reverse-step
7044 @kindex rs @r{(@code{step})}
7045 @item reverse-step @r{[}@var{count}@r{]}
7046 Run the program backward until control reaches the start of a
7047 different source line; then stop it, and return control to @value{GDBN}.
7049 Like the @code{step} command, @code{reverse-step} will only stop
7050 at the beginning of a source line. It ``un-executes'' the previously
7051 executed source line. If the previous source line included calls to
7052 debuggable functions, @code{reverse-step} will step (backward) into
7053 the called function, stopping at the beginning of the @emph{last}
7054 statement in the called function (typically a return statement).
7056 Also, as with the @code{step} command, if non-debuggable functions are
7057 called, @code{reverse-step} will run thru them backward without stopping.
7059 @kindex reverse-stepi
7060 @kindex rsi @r{(@code{reverse-stepi})}
7061 @item reverse-stepi @r{[}@var{count}@r{]}
7062 Reverse-execute one machine instruction. Note that the instruction
7063 to be reverse-executed is @emph{not} the one pointed to by the program
7064 counter, but the instruction executed prior to that one. For instance,
7065 if the last instruction was a jump, @code{reverse-stepi} will take you
7066 back from the destination of the jump to the jump instruction itself.
7068 @kindex reverse-next
7069 @kindex rn @r{(@code{reverse-next})}
7070 @item reverse-next @r{[}@var{count}@r{]}
7071 Run backward to the beginning of the previous line executed in
7072 the current (innermost) stack frame. If the line contains function
7073 calls, they will be ``un-executed'' without stopping. Starting from
7074 the first line of a function, @code{reverse-next} will take you back
7075 to the caller of that function, @emph{before} the function was called,
7076 just as the normal @code{next} command would take you from the last
7077 line of a function back to its return to its caller
7078 @footnote{Unless the code is too heavily optimized.}.
7080 @kindex reverse-nexti
7081 @kindex rni @r{(@code{reverse-nexti})}
7082 @item reverse-nexti @r{[}@var{count}@r{]}
7083 Like @code{nexti}, @code{reverse-nexti} executes a single instruction
7084 in reverse, except that called functions are ``un-executed'' atomically.
7085 That is, if the previously executed instruction was a return from
7086 another function, @code{reverse-nexti} will continue to execute
7087 in reverse until the call to that function (from the current stack
7090 @kindex reverse-finish
7091 @item reverse-finish
7092 Just as the @code{finish} command takes you to the point where the
7093 current function returns, @code{reverse-finish} takes you to the point
7094 where it was called. Instead of ending up at the end of the current
7095 function invocation, you end up at the beginning.
7097 @kindex set exec-direction
7098 @item set exec-direction
7099 Set the direction of target execution.
7100 @item set exec-direction reverse
7101 @cindex execute forward or backward in time
7102 @value{GDBN} will perform all execution commands in reverse, until the
7103 exec-direction mode is changed to ``forward''. Affected commands include
7104 @code{step, stepi, next, nexti, continue, and finish}. The @code{return}
7105 command cannot be used in reverse mode.
7106 @item set exec-direction forward
7107 @value{GDBN} will perform all execution commands in the normal fashion.
7108 This is the default.
7112 @node Process Record and Replay
7113 @chapter Recording Inferior's Execution and Replaying It
7114 @cindex process record and replay
7115 @cindex recording inferior's execution and replaying it
7117 On some platforms, @value{GDBN} provides a special @dfn{process record
7118 and replay} target that can record a log of the process execution, and
7119 replay it later with both forward and reverse execution commands.
7122 When this target is in use, if the execution log includes the record
7123 for the next instruction, @value{GDBN} will debug in @dfn{replay
7124 mode}. In the replay mode, the inferior does not really execute code
7125 instructions. Instead, all the events that normally happen during
7126 code execution are taken from the execution log. While code is not
7127 really executed in replay mode, the values of registers (including the
7128 program counter register) and the memory of the inferior are still
7129 changed as they normally would. Their contents are taken from the
7133 If the record for the next instruction is not in the execution log,
7134 @value{GDBN} will debug in @dfn{record mode}. In this mode, the
7135 inferior executes normally, and @value{GDBN} records the execution log
7138 The process record and replay target supports reverse execution
7139 (@pxref{Reverse Execution}), even if the platform on which the
7140 inferior runs does not. However, the reverse execution is limited in
7141 this case by the range of the instructions recorded in the execution
7142 log. In other words, reverse execution on platforms that don't
7143 support it directly can only be done in the replay mode.
7145 When debugging in the reverse direction, @value{GDBN} will work in
7146 replay mode as long as the execution log includes the record for the
7147 previous instruction; otherwise, it will work in record mode, if the
7148 platform supports reverse execution, or stop if not.
7150 Currently, process record and replay is supported on ARM, Aarch64,
7151 Moxie, PowerPC, PowerPC64, S/390, and x86 (i386/amd64) running
7152 GNU/Linux. Process record and replay can be used both when native
7153 debugging, and when remote debugging via @code{gdbserver}.
7155 For architecture environments that support process record and replay,
7156 @value{GDBN} provides the following commands:
7159 @kindex target record
7160 @kindex target record-full
7161 @kindex target record-btrace
7164 @kindex record btrace
7165 @kindex record btrace bts
7166 @kindex record btrace pt
7172 @kindex rec btrace bts
7173 @kindex rec btrace pt
7176 @item record @var{method}
7177 This command starts the process record and replay target. The
7178 recording method can be specified as parameter. Without a parameter
7179 the command uses the @code{full} recording method. The following
7180 recording methods are available:
7184 Full record/replay recording using @value{GDBN}'s software record and
7185 replay implementation. This method allows replaying and reverse
7188 @item btrace @var{format}
7189 Hardware-supported instruction recording, supported on Intel
7190 processors. This method does not record data. Further, the data is
7191 collected in a ring buffer so old data will be overwritten when the
7192 buffer is full. It allows limited reverse execution. Variables and
7193 registers are not available during reverse execution. In remote
7194 debugging, recording continues on disconnect. Recorded data can be
7195 inspected after reconnecting. The recording may be stopped using
7198 The recording format can be specified as parameter. Without a parameter
7199 the command chooses the recording format. The following recording
7200 formats are available:
7204 @cindex branch trace store
7205 Use the @dfn{Branch Trace Store} (@acronym{BTS}) recording format. In
7206 this format, the processor stores a from/to record for each executed
7207 branch in the btrace ring buffer.
7210 @cindex Intel Processor Trace
7211 Use the @dfn{Intel Processor Trace} recording format. In this
7212 format, the processor stores the execution trace in a compressed form
7213 that is afterwards decoded by @value{GDBN}.
7215 The trace can be recorded with very low overhead. The compressed
7216 trace format also allows small trace buffers to already contain a big
7217 number of instructions compared to @acronym{BTS}.
7219 Decoding the recorded execution trace, on the other hand, is more
7220 expensive than decoding @acronym{BTS} trace. This is mostly due to the
7221 increased number of instructions to process. You should increase the
7222 buffer-size with care.
7225 Not all recording formats may be available on all processors.
7228 The process record and replay target can only debug a process that is
7229 already running. Therefore, you need first to start the process with
7230 the @kbd{run} or @kbd{start} commands, and then start the recording
7231 with the @kbd{record @var{method}} command.
7233 @cindex displaced stepping, and process record and replay
7234 Displaced stepping (@pxref{Maintenance Commands,, displaced stepping})
7235 will be automatically disabled when process record and replay target
7236 is started. That's because the process record and replay target
7237 doesn't support displaced stepping.
7239 @cindex non-stop mode, and process record and replay
7240 @cindex asynchronous execution, and process record and replay
7241 If the inferior is in the non-stop mode (@pxref{Non-Stop Mode}) or in
7242 the asynchronous execution mode (@pxref{Background Execution}), not
7243 all recording methods are available. The @code{full} recording method
7244 does not support these two modes.
7249 Stop the process record and replay target. When process record and
7250 replay target stops, the entire execution log will be deleted and the
7251 inferior will either be terminated, or will remain in its final state.
7253 When you stop the process record and replay target in record mode (at
7254 the end of the execution log), the inferior will be stopped at the
7255 next instruction that would have been recorded. In other words, if
7256 you record for a while and then stop recording, the inferior process
7257 will be left in the same state as if the recording never happened.
7259 On the other hand, if the process record and replay target is stopped
7260 while in replay mode (that is, not at the end of the execution log,
7261 but at some earlier point), the inferior process will become ``live''
7262 at that earlier state, and it will then be possible to continue the
7263 usual ``live'' debugging of the process from that state.
7265 When the inferior process exits, or @value{GDBN} detaches from it,
7266 process record and replay target will automatically stop itself.
7270 Go to a specific location in the execution log. There are several
7271 ways to specify the location to go to:
7274 @item record goto begin
7275 @itemx record goto start
7276 Go to the beginning of the execution log.
7278 @item record goto end
7279 Go to the end of the execution log.
7281 @item record goto @var{n}
7282 Go to instruction number @var{n} in the execution log.
7286 @item record save @var{filename}
7287 Save the execution log to a file @file{@var{filename}}.
7288 Default filename is @file{gdb_record.@var{process_id}}, where
7289 @var{process_id} is the process ID of the inferior.
7291 This command may not be available for all recording methods.
7293 @kindex record restore
7294 @item record restore @var{filename}
7295 Restore the execution log from a file @file{@var{filename}}.
7296 File must have been created with @code{record save}.
7298 @kindex set record full
7299 @item set record full insn-number-max @var{limit}
7300 @itemx set record full insn-number-max unlimited
7301 Set the limit of instructions to be recorded for the @code{full}
7302 recording method. Default value is 200000.
7304 If @var{limit} is a positive number, then @value{GDBN} will start
7305 deleting instructions from the log once the number of the record
7306 instructions becomes greater than @var{limit}. For every new recorded
7307 instruction, @value{GDBN} will delete the earliest recorded
7308 instruction to keep the number of recorded instructions at the limit.
7309 (Since deleting recorded instructions loses information, @value{GDBN}
7310 lets you control what happens when the limit is reached, by means of
7311 the @code{stop-at-limit} option, described below.)
7313 If @var{limit} is @code{unlimited} or zero, @value{GDBN} will never
7314 delete recorded instructions from the execution log. The number of
7315 recorded instructions is limited only by the available memory.
7317 @kindex show record full
7318 @item show record full insn-number-max
7319 Show the limit of instructions to be recorded with the @code{full}
7322 @item set record full stop-at-limit
7323 Control the behavior of the @code{full} recording method when the
7324 number of recorded instructions reaches the limit. If ON (the
7325 default), @value{GDBN} will stop when the limit is reached for the
7326 first time and ask you whether you want to stop the inferior or
7327 continue running it and recording the execution log. If you decide
7328 to continue recording, each new recorded instruction will cause the
7329 oldest one to be deleted.
7331 If this option is OFF, @value{GDBN} will automatically delete the
7332 oldest record to make room for each new one, without asking.
7334 @item show record full stop-at-limit
7335 Show the current setting of @code{stop-at-limit}.
7337 @item set record full memory-query
7338 Control the behavior when @value{GDBN} is unable to record memory
7339 changes caused by an instruction for the @code{full} recording method.
7340 If ON, @value{GDBN} will query whether to stop the inferior in that
7343 If this option is OFF (the default), @value{GDBN} will automatically
7344 ignore the effect of such instructions on memory. Later, when
7345 @value{GDBN} replays this execution log, it will mark the log of this
7346 instruction as not accessible, and it will not affect the replay
7349 @item show record full memory-query
7350 Show the current setting of @code{memory-query}.
7352 @kindex set record btrace
7353 The @code{btrace} record target does not trace data. As a
7354 convenience, when replaying, @value{GDBN} reads read-only memory off
7355 the live program directly, assuming that the addresses of the
7356 read-only areas don't change. This for example makes it possible to
7357 disassemble code while replaying, but not to print variables.
7358 In some cases, being able to inspect variables might be useful.
7359 You can use the following command for that:
7361 @item set record btrace replay-memory-access
7362 Control the behavior of the @code{btrace} recording method when
7363 accessing memory during replay. If @code{read-only} (the default),
7364 @value{GDBN} will only allow accesses to read-only memory.
7365 If @code{read-write}, @value{GDBN} will allow accesses to read-only
7366 and to read-write memory. Beware that the accessed memory corresponds
7367 to the live target and not necessarily to the current replay
7370 @item set record btrace cpu @var{identifier}
7371 Set the processor to be used for enabling workarounds for processor
7372 errata when decoding the trace.
7374 Processor errata are defects in processor operation, caused by its
7375 design or manufacture. They can cause a trace not to match the
7376 specification. This, in turn, may cause trace decode to fail.
7377 @value{GDBN} can detect erroneous trace packets and correct them, thus
7378 avoiding the decoding failures. These corrections are known as
7379 @dfn{errata workarounds}, and are enabled based on the processor on
7380 which the trace was recorded.
7382 By default, @value{GDBN} attempts to detect the processor
7383 automatically, and apply the necessary workarounds for it. However,
7384 you may need to specify the processor if @value{GDBN} does not yet
7385 support it. This command allows you to do that, and also allows to
7386 disable the workarounds.
7388 The argument @var{identifier} identifies the @sc{cpu} and is of the
7389 form: @code{@var{vendor}:@var{processor identifier}}. In addition,
7390 there are two special identifiers, @code{none} and @code{auto}
7393 The following vendor identifiers and corresponding processor
7394 identifiers are currently supported:
7396 @multitable @columnfractions .1 .9
7399 @tab @var{family}/@var{model}[/@var{stepping}]
7403 On GNU/Linux systems, the processor @var{family}, @var{model}, and
7404 @var{stepping} can be obtained from @code{/proc/cpuinfo}.
7406 If @var{identifier} is @code{auto}, enable errata workarounds for the
7407 processor on which the trace was recorded. If @var{identifier} is
7408 @code{none}, errata workarounds are disabled.
7410 For example, when using an old @value{GDBN} on a new system, decode
7411 may fail because @value{GDBN} does not support the new processor. It
7412 often suffices to specify an older processor that @value{GDBN}
7416 (@value{GDBP}) info record
7417 Active record target: record-btrace
7418 Recording format: Intel Processor Trace.
7420 Failed to configure the Intel Processor Trace decoder: unknown cpu.
7421 (@value{GDBP}) set record btrace cpu intel:6/158
7422 (@value{GDBP}) info record
7423 Active record target: record-btrace
7424 Recording format: Intel Processor Trace.
7426 Recorded 84872 instructions in 3189 functions (0 gaps) for thread 1 (...).
7429 @kindex show record btrace
7430 @item show record btrace replay-memory-access
7431 Show the current setting of @code{replay-memory-access}.
7433 @item show record btrace cpu
7434 Show the processor to be used for enabling trace decode errata
7437 @kindex set record btrace bts
7438 @item set record btrace bts buffer-size @var{size}
7439 @itemx set record btrace bts buffer-size unlimited
7440 Set the requested ring buffer size for branch tracing in @acronym{BTS}
7441 format. Default is 64KB.
7443 If @var{size} is a positive number, then @value{GDBN} will try to
7444 allocate a buffer of at least @var{size} bytes for each new thread
7445 that uses the btrace recording method and the @acronym{BTS} format.
7446 The actually obtained buffer size may differ from the requested
7447 @var{size}. Use the @code{info record} command to see the actual
7448 buffer size for each thread that uses the btrace recording method and
7449 the @acronym{BTS} format.
7451 If @var{limit} is @code{unlimited} or zero, @value{GDBN} will try to
7452 allocate a buffer of 4MB.
7454 Bigger buffers mean longer traces. On the other hand, @value{GDBN} will
7455 also need longer to process the branch trace data before it can be used.
7457 @item show record btrace bts buffer-size @var{size}
7458 Show the current setting of the requested ring buffer size for branch
7459 tracing in @acronym{BTS} format.
7461 @kindex set record btrace pt
7462 @item set record btrace pt buffer-size @var{size}
7463 @itemx set record btrace pt buffer-size unlimited
7464 Set the requested ring buffer size for branch tracing in Intel
7465 Processor Trace format. Default is 16KB.
7467 If @var{size} is a positive number, then @value{GDBN} will try to
7468 allocate a buffer of at least @var{size} bytes for each new thread
7469 that uses the btrace recording method and the Intel Processor Trace
7470 format. The actually obtained buffer size may differ from the
7471 requested @var{size}. Use the @code{info record} command to see the
7472 actual buffer size for each thread.
7474 If @var{limit} is @code{unlimited} or zero, @value{GDBN} will try to
7475 allocate a buffer of 4MB.
7477 Bigger buffers mean longer traces. On the other hand, @value{GDBN} will
7478 also need longer to process the branch trace data before it can be used.
7480 @item show record btrace pt buffer-size @var{size}
7481 Show the current setting of the requested ring buffer size for branch
7482 tracing in Intel Processor Trace format.
7486 Show various statistics about the recording depending on the recording
7491 For the @code{full} recording method, it shows the state of process
7492 record and its in-memory execution log buffer, including:
7496 Whether in record mode or replay mode.
7498 Lowest recorded instruction number (counting from when the current execution log started recording instructions).
7500 Highest recorded instruction number.
7502 Current instruction about to be replayed (if in replay mode).
7504 Number of instructions contained in the execution log.
7506 Maximum number of instructions that may be contained in the execution log.
7510 For the @code{btrace} recording method, it shows:
7516 Number of instructions that have been recorded.
7518 Number of blocks of sequential control-flow formed by the recorded
7521 Whether in record mode or replay mode.
7524 For the @code{bts} recording format, it also shows:
7527 Size of the perf ring buffer.
7530 For the @code{pt} recording format, it also shows:
7533 Size of the perf ring buffer.
7537 @kindex record delete
7540 When record target runs in replay mode (``in the past''), delete the
7541 subsequent execution log and begin to record a new execution log starting
7542 from the current address. This means you will abandon the previously
7543 recorded ``future'' and begin recording a new ``future''.
7545 @kindex record instruction-history
7546 @kindex rec instruction-history
7547 @item record instruction-history
7548 Disassembles instructions from the recorded execution log. By
7549 default, ten instructions are disassembled. This can be changed using
7550 the @code{set record instruction-history-size} command. Instructions
7551 are printed in execution order.
7553 It can also print mixed source+disassembly if you specify the the
7554 @code{/m} or @code{/s} modifier, and print the raw instructions in hex
7555 as well as in symbolic form by specifying the @code{/r} modifier.
7557 The current position marker is printed for the instruction at the
7558 current program counter value. This instruction can appear multiple
7559 times in the trace and the current position marker will be printed
7560 every time. To omit the current position marker, specify the
7563 To better align the printed instructions when the trace contains
7564 instructions from more than one function, the function name may be
7565 omitted by specifying the @code{/f} modifier.
7567 Speculatively executed instructions are prefixed with @samp{?}. This
7568 feature is not available for all recording formats.
7570 There are several ways to specify what part of the execution log to
7574 @item record instruction-history @var{insn}
7575 Disassembles ten instructions starting from instruction number
7578 @item record instruction-history @var{insn}, +/-@var{n}
7579 Disassembles @var{n} instructions around instruction number
7580 @var{insn}. If @var{n} is preceded with @code{+}, disassembles
7581 @var{n} instructions after instruction number @var{insn}. If
7582 @var{n} is preceded with @code{-}, disassembles @var{n}
7583 instructions before instruction number @var{insn}.
7585 @item record instruction-history
7586 Disassembles ten more instructions after the last disassembly.
7588 @item record instruction-history -
7589 Disassembles ten more instructions before the last disassembly.
7591 @item record instruction-history @var{begin}, @var{end}
7592 Disassembles instructions beginning with instruction number
7593 @var{begin} until instruction number @var{end}. The instruction
7594 number @var{end} is included.
7597 This command may not be available for all recording methods.
7600 @item set record instruction-history-size @var{size}
7601 @itemx set record instruction-history-size unlimited
7602 Define how many instructions to disassemble in the @code{record
7603 instruction-history} command. The default value is 10.
7604 A @var{size} of @code{unlimited} means unlimited instructions.
7607 @item show record instruction-history-size
7608 Show how many instructions to disassemble in the @code{record
7609 instruction-history} command.
7611 @kindex record function-call-history
7612 @kindex rec function-call-history
7613 @item record function-call-history
7614 Prints the execution history at function granularity. It prints one
7615 line for each sequence of instructions that belong to the same
7616 function giving the name of that function, the source lines
7617 for this instruction sequence (if the @code{/l} modifier is
7618 specified), and the instructions numbers that form the sequence (if
7619 the @code{/i} modifier is specified). The function names are indented
7620 to reflect the call stack depth if the @code{/c} modifier is
7621 specified. The @code{/l}, @code{/i}, and @code{/c} modifiers can be
7625 (@value{GDBP}) @b{list 1, 10}
7636 (@value{GDBP}) @b{record function-call-history /ilc}
7637 1 bar inst 1,4 at foo.c:6,8
7638 2 foo inst 5,10 at foo.c:2,3
7639 3 bar inst 11,13 at foo.c:9,10
7642 By default, ten lines are printed. This can be changed using the
7643 @code{set record function-call-history-size} command. Functions are
7644 printed in execution order. There are several ways to specify what
7648 @item record function-call-history @var{func}
7649 Prints ten functions starting from function number @var{func}.
7651 @item record function-call-history @var{func}, +/-@var{n}
7652 Prints @var{n} functions around function number @var{func}. If
7653 @var{n} is preceded with @code{+}, prints @var{n} functions after
7654 function number @var{func}. If @var{n} is preceded with @code{-},
7655 prints @var{n} functions before function number @var{func}.
7657 @item record function-call-history
7658 Prints ten more functions after the last ten-line print.
7660 @item record function-call-history -
7661 Prints ten more functions before the last ten-line print.
7663 @item record function-call-history @var{begin}, @var{end}
7664 Prints functions beginning with function number @var{begin} until
7665 function number @var{end}. The function number @var{end} is included.
7668 This command may not be available for all recording methods.
7670 @item set record function-call-history-size @var{size}
7671 @itemx set record function-call-history-size unlimited
7672 Define how many lines to print in the
7673 @code{record function-call-history} command. The default value is 10.
7674 A size of @code{unlimited} means unlimited lines.
7676 @item show record function-call-history-size
7677 Show how many lines to print in the
7678 @code{record function-call-history} command.
7683 @chapter Examining the Stack
7685 When your program has stopped, the first thing you need to know is where it
7686 stopped and how it got there.
7689 Each time your program performs a function call, information about the call
7691 That information includes the location of the call in your program,
7692 the arguments of the call,
7693 and the local variables of the function being called.
7694 The information is saved in a block of data called a @dfn{stack frame}.
7695 The stack frames are allocated in a region of memory called the @dfn{call
7698 When your program stops, the @value{GDBN} commands for examining the
7699 stack allow you to see all of this information.
7701 @cindex selected frame
7702 One of the stack frames is @dfn{selected} by @value{GDBN} and many
7703 @value{GDBN} commands refer implicitly to the selected frame. In
7704 particular, whenever you ask @value{GDBN} for the value of a variable in
7705 your program, the value is found in the selected frame. There are
7706 special @value{GDBN} commands to select whichever frame you are
7707 interested in. @xref{Selection, ,Selecting a Frame}.
7709 When your program stops, @value{GDBN} automatically selects the
7710 currently executing frame and describes it briefly, similar to the
7711 @code{frame} command (@pxref{Frame Info, ,Information about a Frame}).
7714 * Frames:: Stack frames
7715 * Backtrace:: Backtraces
7716 * Selection:: Selecting a frame
7717 * Frame Info:: Information on a frame
7718 * Frame Apply:: Applying a command to several frames
7719 * Frame Filter Management:: Managing frame filters
7724 @section Stack Frames
7726 @cindex frame, definition
7728 The call stack is divided up into contiguous pieces called @dfn{stack
7729 frames}, or @dfn{frames} for short; each frame is the data associated
7730 with one call to one function. The frame contains the arguments given
7731 to the function, the function's local variables, and the address at
7732 which the function is executing.
7734 @cindex initial frame
7735 @cindex outermost frame
7736 @cindex innermost frame
7737 When your program is started, the stack has only one frame, that of the
7738 function @code{main}. This is called the @dfn{initial} frame or the
7739 @dfn{outermost} frame. Each time a function is called, a new frame is
7740 made. Each time a function returns, the frame for that function invocation
7741 is eliminated. If a function is recursive, there can be many frames for
7742 the same function. The frame for the function in which execution is
7743 actually occurring is called the @dfn{innermost} frame. This is the most
7744 recently created of all the stack frames that still exist.
7746 @cindex frame pointer
7747 Inside your program, stack frames are identified by their addresses. A
7748 stack frame consists of many bytes, each of which has its own address; each
7749 kind of computer has a convention for choosing one byte whose
7750 address serves as the address of the frame. Usually this address is kept
7751 in a register called the @dfn{frame pointer register}
7752 (@pxref{Registers, $fp}) while execution is going on in that frame.
7755 @cindex frame number
7756 @value{GDBN} labels each existing stack frame with a @dfn{level}, a
7757 number that is zero for the innermost frame, one for the frame that
7758 called it, and so on upward. These level numbers give you a way of
7759 designating stack frames in @value{GDBN} commands. The terms
7760 @dfn{frame number} and @dfn{frame level} can be used interchangeably to
7761 describe this number.
7763 @c The -fomit-frame-pointer below perennially causes hbox overflow
7764 @c underflow problems.
7765 @cindex frameless execution
7766 Some compilers provide a way to compile functions so that they operate
7767 without stack frames. (For example, the @value{NGCC} option
7769 @samp{-fomit-frame-pointer}
7771 generates functions without a frame.)
7772 This is occasionally done with heavily used library functions to save
7773 the frame setup time. @value{GDBN} has limited facilities for dealing
7774 with these function invocations. If the innermost function invocation
7775 has no stack frame, @value{GDBN} nevertheless regards it as though
7776 it had a separate frame, which is numbered zero as usual, allowing
7777 correct tracing of the function call chain. However, @value{GDBN} has
7778 no provision for frameless functions elsewhere in the stack.
7784 @cindex call stack traces
7785 A backtrace is a summary of how your program got where it is. It shows one
7786 line per frame, for many frames, starting with the currently executing
7787 frame (frame zero), followed by its caller (frame one), and on up the
7790 @anchor{backtrace-command}
7792 @kindex bt @r{(@code{backtrace})}
7793 To print a backtrace of the entire stack, use the @code{backtrace}
7794 command, or its alias @code{bt}. This command will print one line per
7795 frame for frames in the stack. By default, all stack frames are
7796 printed. You can stop the backtrace at any time by typing the system
7797 interrupt character, normally @kbd{Ctrl-c}.
7800 @item backtrace [@var{option}]@dots{} [@var{qualifier}]@dots{} [@var{count}]
7801 @itemx bt [@var{option}]@dots{} [@var{qualifier}]@dots{} [@var{count}]
7802 Print the backtrace of the entire stack.
7804 The optional @var{count} can be one of the following:
7809 Print only the innermost @var{n} frames, where @var{n} is a positive
7814 Print only the outermost @var{n} frames, where @var{n} is a positive
7822 Print the values of the local variables also. This can be combined
7823 with the optional @var{count} to limit the number of frames shown.
7826 Do not run Python frame filters on this backtrace. @xref{Frame
7827 Filter API}, for more information. Additionally use @ref{disable
7828 frame-filter all} to turn off all frame filters. This is only
7829 relevant when @value{GDBN} has been configured with @code{Python}
7833 A Python frame filter might decide to ``elide'' some frames. Normally
7834 such elided frames are still printed, but they are indented relative
7835 to the filtered frames that cause them to be elided. The @code{-hide}
7836 option causes elided frames to not be printed at all.
7839 The @code{backtrace} command also supports a number of options that
7840 allow overriding relevant global print settings as set by @code{set
7841 backtrace} and @code{set print} subcommands:
7844 @item -past-main [@code{on}|@code{off}]
7845 Set whether backtraces should continue past @code{main}. Related setting:
7846 @ref{set backtrace past-main}.
7848 @item -past-entry [@code{on}|@code{off}]
7849 Set whether backtraces should continue past the entry point of a program.
7850 Related setting: @ref{set backtrace past-entry}.
7852 @item -entry-values @code{no}|@code{only}|@code{preferred}|@code{if-needed}|@code{both}|@code{compact}|@code{default}
7853 Set printing of function arguments at function entry.
7854 Related setting: @ref{set print entry-values}.
7856 @item -frame-arguments @code{all}|@code{scalars}|@code{none}
7857 Set printing of non-scalar frame arguments.
7858 Related setting: @ref{set print frame-arguments}.
7860 @item -raw-frame-arguments [@code{on}|@code{off}]
7861 Set whether to print frame arguments in raw form.
7862 Related setting: @ref{set print raw-frame-arguments}.
7864 @item -frame-info @code{auto}|@code{source-line}|@code{location}|@code{source-and-location}|@code{location-and-address}|@code{short-location}
7865 Set printing of frame information.
7866 Related setting: @ref{set print frame-info}.
7869 The optional @var{qualifier} is maintained for backward compatibility.
7870 It can be one of the following:
7874 Equivalent to the @code{-full} option.
7877 Equivalent to the @code{-no-filters} option.
7880 Equivalent to the @code{-hide} option.
7887 The names @code{where} and @code{info stack} (abbreviated @code{info s})
7888 are additional aliases for @code{backtrace}.
7890 @cindex multiple threads, backtrace
7891 In a multi-threaded program, @value{GDBN} by default shows the
7892 backtrace only for the current thread. To display the backtrace for
7893 several or all of the threads, use the command @code{thread apply}
7894 (@pxref{Threads, thread apply}). For example, if you type @kbd{thread
7895 apply all backtrace}, @value{GDBN} will display the backtrace for all
7896 the threads; this is handy when you debug a core dump of a
7897 multi-threaded program.
7899 Each line in the backtrace shows the frame number and the function name.
7900 The program counter value is also shown---unless you use @code{set
7901 print address off}. The backtrace also shows the source file name and
7902 line number, as well as the arguments to the function. The program
7903 counter value is omitted if it is at the beginning of the code for that
7906 Here is an example of a backtrace. It was made with the command
7907 @samp{bt 3}, so it shows the innermost three frames.
7911 #0 m4_traceon (obs=0x24eb0, argc=1, argv=0x2b8c8)
7913 #1 0x6e38 in expand_macro (sym=0x2b600, data=...) at macro.c:242
7914 #2 0x6840 in expand_token (obs=0x0, t=177664, td=0xf7fffb08)
7916 (More stack frames follow...)
7921 The display for frame zero does not begin with a program counter
7922 value, indicating that your program has stopped at the beginning of the
7923 code for line @code{993} of @code{builtin.c}.
7926 The value of parameter @code{data} in frame 1 has been replaced by
7927 @code{@dots{}}. By default, @value{GDBN} prints the value of a parameter
7928 only if it is a scalar (integer, pointer, enumeration, etc). See command
7929 @kbd{set print frame-arguments} in @ref{Print Settings} for more details
7930 on how to configure the way function parameter values are printed.
7931 The command @kbd{set print frame-info} (@pxref{Print Settings}) controls
7932 what frame information is printed.
7934 @cindex optimized out, in backtrace
7935 @cindex function call arguments, optimized out
7936 If your program was compiled with optimizations, some compilers will
7937 optimize away arguments passed to functions if those arguments are
7938 never used after the call. Such optimizations generate code that
7939 passes arguments through registers, but doesn't store those arguments
7940 in the stack frame. @value{GDBN} has no way of displaying such
7941 arguments in stack frames other than the innermost one. Here's what
7942 such a backtrace might look like:
7946 #0 m4_traceon (obs=0x24eb0, argc=1, argv=0x2b8c8)
7948 #1 0x6e38 in expand_macro (sym=<optimized out>) at macro.c:242
7949 #2 0x6840 in expand_token (obs=0x0, t=<optimized out>, td=0xf7fffb08)
7951 (More stack frames follow...)
7956 The values of arguments that were not saved in their stack frames are
7957 shown as @samp{<optimized out>}.
7959 If you need to display the values of such optimized-out arguments,
7960 either deduce that from other variables whose values depend on the one
7961 you are interested in, or recompile without optimizations.
7963 @cindex backtrace beyond @code{main} function
7964 @cindex program entry point
7965 @cindex startup code, and backtrace
7966 Most programs have a standard user entry point---a place where system
7967 libraries and startup code transition into user code. For C this is
7968 @code{main}@footnote{
7969 Note that embedded programs (the so-called ``free-standing''
7970 environment) are not required to have a @code{main} function as the
7971 entry point. They could even have multiple entry points.}.
7972 When @value{GDBN} finds the entry function in a backtrace
7973 it will terminate the backtrace, to avoid tracing into highly
7974 system-specific (and generally uninteresting) code.
7976 If you need to examine the startup code, or limit the number of levels
7977 in a backtrace, you can change this behavior:
7980 @item set backtrace past-main
7981 @itemx set backtrace past-main on
7982 @anchor{set backtrace past-main}
7983 @kindex set backtrace
7984 Backtraces will continue past the user entry point.
7986 @item set backtrace past-main off
7987 Backtraces will stop when they encounter the user entry point. This is the
7990 @item show backtrace past-main
7991 @kindex show backtrace
7992 Display the current user entry point backtrace policy.
7994 @item set backtrace past-entry
7995 @itemx set backtrace past-entry on
7996 @anchor{set backtrace past-entry}
7997 Backtraces will continue past the internal entry point of an application.
7998 This entry point is encoded by the linker when the application is built,
7999 and is likely before the user entry point @code{main} (or equivalent) is called.
8001 @item set backtrace past-entry off
8002 Backtraces will stop when they encounter the internal entry point of an
8003 application. This is the default.
8005 @item show backtrace past-entry
8006 Display the current internal entry point backtrace policy.
8008 @item set backtrace limit @var{n}
8009 @itemx set backtrace limit 0
8010 @itemx set backtrace limit unlimited
8011 @anchor{set backtrace limit}
8012 @cindex backtrace limit
8013 Limit the backtrace to @var{n} levels. A value of @code{unlimited}
8014 or zero means unlimited levels.
8016 @item show backtrace limit
8017 Display the current limit on backtrace levels.
8020 You can control how file names are displayed.
8023 @item set filename-display
8024 @itemx set filename-display relative
8025 @cindex filename-display
8026 Display file names relative to the compilation directory. This is the default.
8028 @item set filename-display basename
8029 Display only basename of a filename.
8031 @item set filename-display absolute
8032 Display an absolute filename.
8034 @item show filename-display
8035 Show the current way to display filenames.
8039 @section Selecting a Frame
8041 Most commands for examining the stack and other data in your program work on
8042 whichever stack frame is selected at the moment. Here are the commands for
8043 selecting a stack frame; all of them finish by printing a brief description
8044 of the stack frame just selected.
8047 @kindex frame@r{, selecting}
8048 @kindex f @r{(@code{frame})}
8049 @item frame @r{[} @var{frame-selection-spec} @r{]}
8050 @item f @r{[} @var{frame-selection-spec} @r{]}
8051 The @command{frame} command allows different stack frames to be
8052 selected. The @var{frame-selection-spec} can be any of the following:
8057 @item level @var{num}
8058 Select frame level @var{num}. Recall that frame zero is the innermost
8059 (currently executing) frame, frame one is the frame that called the
8060 innermost one, and so on. The highest level frame is usually the one
8063 As this is the most common method of navigating the frame stack, the
8064 string @command{level} can be omitted. For example, the following two
8065 commands are equivalent:
8068 (@value{GDBP}) frame 3
8069 (@value{GDBP}) frame level 3
8072 @kindex frame address
8073 @item address @var{stack-address}
8074 Select the frame with stack address @var{stack-address}. The
8075 @var{stack-address} for a frame can be seen in the output of
8076 @command{info frame}, for example:
8079 (@value{GDBP}) info frame
8080 Stack level 1, frame at 0x7fffffffda30:
8081 rip = 0x40066d in b (amd64-entry-value.cc:59); saved rip 0x4004c5
8082 tail call frame, caller of frame at 0x7fffffffda30
8083 source language c++.
8084 Arglist at unknown address.
8085 Locals at unknown address, Previous frame's sp is 0x7fffffffda30
8088 The @var{stack-address} for this frame is @code{0x7fffffffda30} as
8089 indicated by the line:
8092 Stack level 1, frame at 0x7fffffffda30:
8095 @kindex frame function
8096 @item function @var{function-name}
8097 Select the stack frame for function @var{function-name}. If there are
8098 multiple stack frames for function @var{function-name} then the inner
8099 most stack frame is selected.
8102 @item view @var{stack-address} @r{[} @var{pc-addr} @r{]}
8103 View a frame that is not part of @value{GDBN}'s backtrace. The frame
8104 viewed has stack address @var{stack-addr}, and optionally, a program
8105 counter address of @var{pc-addr}.
8107 This is useful mainly if the chaining of stack frames has been
8108 damaged by a bug, making it impossible for @value{GDBN} to assign
8109 numbers properly to all frames. In addition, this can be useful
8110 when your program has multiple stacks and switches between them.
8112 When viewing a frame outside the current backtrace using
8113 @command{frame view} then you can always return to the original
8114 stack using one of the previous stack frame selection instructions,
8115 for example @command{frame level 0}.
8121 Move @var{n} frames up the stack; @var{n} defaults to 1. For positive
8122 numbers @var{n}, this advances toward the outermost frame, to higher
8123 frame numbers, to frames that have existed longer.
8126 @kindex do @r{(@code{down})}
8128 Move @var{n} frames down the stack; @var{n} defaults to 1. For
8129 positive numbers @var{n}, this advances toward the innermost frame, to
8130 lower frame numbers, to frames that were created more recently.
8131 You may abbreviate @code{down} as @code{do}.
8134 All of these commands end by printing two lines of output describing the
8135 frame. The first line shows the frame number, the function name, the
8136 arguments, and the source file and line number of execution in that
8137 frame. The second line shows the text of that source line.
8145 #1 0x22f0 in main (argc=1, argv=0xf7fffbf4, env=0xf7fffbfc)
8147 10 read_input_file (argv[i]);
8151 After such a printout, the @code{list} command with no arguments
8152 prints ten lines centered on the point of execution in the frame.
8153 You can also edit the program at the point of execution with your favorite
8154 editing program by typing @code{edit}.
8155 @xref{List, ,Printing Source Lines},
8159 @kindex select-frame
8160 @item select-frame @r{[} @var{frame-selection-spec} @r{]}
8161 The @code{select-frame} command is a variant of @code{frame} that does
8162 not display the new frame after selecting it. This command is
8163 intended primarily for use in @value{GDBN} command scripts, where the
8164 output might be unnecessary and distracting. The
8165 @var{frame-selection-spec} is as for the @command{frame} command
8166 described in @ref{Selection, ,Selecting a Frame}.
8168 @kindex down-silently
8170 @item up-silently @var{n}
8171 @itemx down-silently @var{n}
8172 These two commands are variants of @code{up} and @code{down},
8173 respectively; they differ in that they do their work silently, without
8174 causing display of the new frame. They are intended primarily for use
8175 in @value{GDBN} command scripts, where the output might be unnecessary and
8180 @section Information About a Frame
8182 There are several other commands to print information about the selected
8188 When used without any argument, this command does not change which
8189 frame is selected, but prints a brief description of the currently
8190 selected stack frame. It can be abbreviated @code{f}. With an
8191 argument, this command is used to select a stack frame.
8192 @xref{Selection, ,Selecting a Frame}.
8195 @kindex info f @r{(@code{info frame})}
8198 This command prints a verbose description of the selected stack frame,
8203 the address of the frame
8205 the address of the next frame down (called by this frame)
8207 the address of the next frame up (caller of this frame)
8209 the language in which the source code corresponding to this frame is written
8211 the address of the frame's arguments
8213 the address of the frame's local variables
8215 the program counter saved in it (the address of execution in the caller frame)
8217 which registers were saved in the frame
8220 @noindent The verbose description is useful when
8221 something has gone wrong that has made the stack format fail to fit
8222 the usual conventions.
8224 @item info frame @r{[} @var{frame-selection-spec} @r{]}
8225 @itemx info f @r{[} @var{frame-selection-spec} @r{]}
8226 Print a verbose description of the frame selected by
8227 @var{frame-selection-spec}. The @var{frame-selection-spec} is the
8228 same as for the @command{frame} command (@pxref{Selection, ,Selecting
8229 a Frame}). The selected frame remains unchanged by this command.
8232 @item info args [-q]
8233 Print the arguments of the selected frame, each on a separate line.
8235 The optional flag @samp{-q}, which stands for @samp{quiet}, disables
8236 printing header information and messages explaining why no argument
8239 @item info args [-q] [-t @var{type_regexp}] [@var{regexp}]
8240 Like @kbd{info args}, but only print the arguments selected
8241 with the provided regexp(s).
8243 If @var{regexp} is provided, print only the arguments whose names
8244 match the regular expression @var{regexp}.
8246 If @var{type_regexp} is provided, print only the arguments whose
8247 types, as printed by the @code{whatis} command, match
8248 the regular expression @var{type_regexp}.
8249 If @var{type_regexp} contains space(s), it should be enclosed in
8250 quote characters. If needed, use backslash to escape the meaning
8251 of special characters or quotes.
8253 If both @var{regexp} and @var{type_regexp} are provided, an argument
8254 is printed only if its name matches @var{regexp} and its type matches
8257 @item info locals [-q]
8259 Print the local variables of the selected frame, each on a separate
8260 line. These are all variables (declared either static or automatic)
8261 accessible at the point of execution of the selected frame.
8263 The optional flag @samp{-q}, which stands for @samp{quiet}, disables
8264 printing header information and messages explaining why no local variables
8267 @item info locals [-q] [-t @var{type_regexp}] [@var{regexp}]
8268 Like @kbd{info locals}, but only print the local variables selected
8269 with the provided regexp(s).
8271 If @var{regexp} is provided, print only the local variables whose names
8272 match the regular expression @var{regexp}.
8274 If @var{type_regexp} is provided, print only the local variables whose
8275 types, as printed by the @code{whatis} command, match
8276 the regular expression @var{type_regexp}.
8277 If @var{type_regexp} contains space(s), it should be enclosed in
8278 quote characters. If needed, use backslash to escape the meaning
8279 of special characters or quotes.
8281 If both @var{regexp} and @var{type_regexp} are provided, a local variable
8282 is printed only if its name matches @var{regexp} and its type matches
8285 The command @kbd{info locals -q -t @var{type_regexp}} can usefully be
8286 combined with the commands @kbd{frame apply} and @kbd{thread apply}.
8287 For example, your program might use Resource Acquisition Is
8288 Initialization types (RAII) such as @code{lock_something_t}: each
8289 local variable of type @code{lock_something_t} automatically places a
8290 lock that is destroyed when the variable goes out of scope. You can
8291 then list all acquired locks in your program by doing
8293 thread apply all -s frame apply all -s info locals -q -t lock_something_t
8296 or the equivalent shorter form
8298 tfaas i lo -q -t lock_something_t
8304 @section Applying a Command to Several Frames.
8305 @anchor{frame apply}
8307 @cindex apply command to several frames
8309 @item frame apply [all | @var{count} | @var{-count} | level @var{level}@dots{}] [@var{option}]@dots{} @var{command}
8310 The @code{frame apply} command allows you to apply the named
8311 @var{command} to one or more frames.
8315 Specify @code{all} to apply @var{command} to all frames.
8318 Use @var{count} to apply @var{command} to the innermost @var{count}
8319 frames, where @var{count} is a positive number.
8322 Use @var{-count} to apply @var{command} to the outermost @var{count}
8323 frames, where @var{count} is a positive number.
8326 Use @code{level} to apply @var{command} to the set of frames identified
8327 by the @var{level} list. @var{level} is a frame level or a range of frame
8328 levels as @var{level1}-@var{level2}. The frame level is the number shown
8329 in the first field of the @samp{backtrace} command output.
8330 E.g., @samp{2-4 6-8 3} indicates to apply @var{command} for the frames
8331 at levels 2, 3, 4, 6, 7, 8, and then again on frame at level 3.
8335 Note that the frames on which @code{frame apply} applies a command are
8336 also influenced by the @code{set backtrace} settings such as @code{set
8337 backtrace past-main} and @code{set backtrace limit N}.
8338 @xref{Backtrace,,Backtraces}.
8340 The @code{frame apply} command also supports a number of options that
8341 allow overriding relevant @code{set backtrace} settings:
8344 @item -past-main [@code{on}|@code{off}]
8345 Whether backtraces should continue past @code{main}.
8346 Related setting: @ref{set backtrace past-main}.
8348 @item -past-entry [@code{on}|@code{off}]
8349 Whether backtraces should continue past the entry point of a program.
8350 Related setting: @ref{set backtrace past-entry}.
8353 By default, @value{GDBN} displays some frame information before the
8354 output produced by @var{command}, and an error raised during the
8355 execution of a @var{command} will abort @code{frame apply}. The
8356 following options can be used to fine-tune these behaviors:
8360 The flag @code{-c}, which stands for @samp{continue}, causes any
8361 errors in @var{command} to be displayed, and the execution of
8362 @code{frame apply} then continues.
8364 The flag @code{-s}, which stands for @samp{silent}, causes any errors
8365 or empty output produced by a @var{command} to be silently ignored.
8366 That is, the execution continues, but the frame information and errors
8369 The flag @code{-q} (@samp{quiet}) disables printing the frame
8373 The following example shows how the flags @code{-c} and @code{-s} are
8374 working when applying the command @code{p j} to all frames, where
8375 variable @code{j} can only be successfully printed in the outermost
8376 @code{#1 main} frame.
8380 (@value{GDBP}) frame apply all p j
8381 #0 some_function (i=5) at fun.c:4
8382 No symbol "j" in current context.
8383 (@value{GDBP}) frame apply all -c p j
8384 #0 some_function (i=5) at fun.c:4
8385 No symbol "j" in current context.
8386 #1 0x565555fb in main (argc=1, argv=0xffffd2c4) at fun.c:11
8388 (@value{GDBP}) frame apply all -s p j
8389 #1 0x565555fb in main (argc=1, argv=0xffffd2c4) at fun.c:11
8395 By default, @samp{frame apply}, prints the frame location
8396 information before the command output:
8400 (@value{GDBP}) frame apply all p $sp
8401 #0 some_function (i=5) at fun.c:4
8402 $4 = (void *) 0xffffd1e0
8403 #1 0x565555fb in main (argc=1, argv=0xffffd2c4) at fun.c:11
8404 $5 = (void *) 0xffffd1f0
8409 If the flag @code{-q} is given, no frame information is printed:
8412 (@value{GDBP}) frame apply all -q p $sp
8413 $12 = (void *) 0xffffd1e0
8414 $13 = (void *) 0xffffd1f0
8424 @cindex apply a command to all frames (ignoring errors and empty output)
8425 @item faas @var{command}
8426 Shortcut for @code{frame apply all -s @var{command}}.
8427 Applies @var{command} on all frames, ignoring errors and empty output.
8429 It can for example be used to print a local variable or a function
8430 argument without knowing the frame where this variable or argument
8433 (@value{GDBP}) faas p some_local_var_i_do_not_remember_where_it_is
8436 The @code{faas} command accepts the same options as the @code{frame
8437 apply} command. @xref{frame apply}.
8439 Note that the command @code{tfaas @var{command}} applies @var{command}
8440 on all frames of all threads. See @xref{Threads,,Threads}.
8444 @node Frame Filter Management
8445 @section Management of Frame Filters.
8446 @cindex managing frame filters
8448 Frame filters are Python based utilities to manage and decorate the
8449 output of frames. @xref{Frame Filter API}, for further information.
8451 Managing frame filters is performed by several commands available
8452 within @value{GDBN}, detailed here.
8455 @kindex info frame-filter
8456 @item info frame-filter
8457 Print a list of installed frame filters from all dictionaries, showing
8458 their name, priority and enabled status.
8460 @kindex disable frame-filter
8461 @anchor{disable frame-filter all}
8462 @item disable frame-filter @var{filter-dictionary} @var{filter-name}
8463 Disable a frame filter in the dictionary matching
8464 @var{filter-dictionary} and @var{filter-name}. The
8465 @var{filter-dictionary} may be @code{all}, @code{global},
8466 @code{progspace}, or the name of the object file where the frame filter
8467 dictionary resides. When @code{all} is specified, all frame filters
8468 across all dictionaries are disabled. The @var{filter-name} is the name
8469 of the frame filter and is used when @code{all} is not the option for
8470 @var{filter-dictionary}. A disabled frame-filter is not deleted, it
8471 may be enabled again later.
8473 @kindex enable frame-filter
8474 @item enable frame-filter @var{filter-dictionary} @var{filter-name}
8475 Enable a frame filter in the dictionary matching
8476 @var{filter-dictionary} and @var{filter-name}. The
8477 @var{filter-dictionary} may be @code{all}, @code{global},
8478 @code{progspace} or the name of the object file where the frame filter
8479 dictionary resides. When @code{all} is specified, all frame filters across
8480 all dictionaries are enabled. The @var{filter-name} is the name of the frame
8481 filter and is used when @code{all} is not the option for
8482 @var{filter-dictionary}.
8487 (@value{GDBP}) info frame-filter
8489 global frame-filters:
8490 Priority Enabled Name
8491 1000 No PrimaryFunctionFilter
8494 progspace /build/test frame-filters:
8495 Priority Enabled Name
8496 100 Yes ProgspaceFilter
8498 objfile /build/test frame-filters:
8499 Priority Enabled Name
8500 999 Yes BuildProgramFilter
8502 (@value{GDBP}) disable frame-filter /build/test BuildProgramFilter
8503 (@value{GDBP}) info frame-filter
8505 global frame-filters:
8506 Priority Enabled Name
8507 1000 No PrimaryFunctionFilter
8510 progspace /build/test frame-filters:
8511 Priority Enabled Name
8512 100 Yes ProgspaceFilter
8514 objfile /build/test frame-filters:
8515 Priority Enabled Name
8516 999 No BuildProgramFilter
8518 (@value{GDBP}) enable frame-filter global PrimaryFunctionFilter
8519 (@value{GDBP}) info frame-filter
8521 global frame-filters:
8522 Priority Enabled Name
8523 1000 Yes PrimaryFunctionFilter
8526 progspace /build/test frame-filters:
8527 Priority Enabled Name
8528 100 Yes ProgspaceFilter
8530 objfile /build/test frame-filters:
8531 Priority Enabled Name
8532 999 No BuildProgramFilter
8535 @kindex set frame-filter priority
8536 @item set frame-filter priority @var{filter-dictionary} @var{filter-name} @var{priority}
8537 Set the @var{priority} of a frame filter in the dictionary matching
8538 @var{filter-dictionary}, and the frame filter name matching
8539 @var{filter-name}. The @var{filter-dictionary} may be @code{global},
8540 @code{progspace} or the name of the object file where the frame filter
8541 dictionary resides. The @var{priority} is an integer.
8543 @kindex show frame-filter priority
8544 @item show frame-filter priority @var{filter-dictionary} @var{filter-name}
8545 Show the @var{priority} of a frame filter in the dictionary matching
8546 @var{filter-dictionary}, and the frame filter name matching
8547 @var{filter-name}. The @var{filter-dictionary} may be @code{global},
8548 @code{progspace} or the name of the object file where the frame filter
8554 (@value{GDBP}) info frame-filter
8556 global frame-filters:
8557 Priority Enabled Name
8558 1000 Yes PrimaryFunctionFilter
8561 progspace /build/test frame-filters:
8562 Priority Enabled Name
8563 100 Yes ProgspaceFilter
8565 objfile /build/test frame-filters:
8566 Priority Enabled Name
8567 999 No BuildProgramFilter
8569 (@value{GDBP}) set frame-filter priority global Reverse 50
8570 (@value{GDBP}) info frame-filter
8572 global frame-filters:
8573 Priority Enabled Name
8574 1000 Yes PrimaryFunctionFilter
8577 progspace /build/test frame-filters:
8578 Priority Enabled Name
8579 100 Yes ProgspaceFilter
8581 objfile /build/test frame-filters:
8582 Priority Enabled Name
8583 999 No BuildProgramFilter
8588 @chapter Examining Source Files
8590 @value{GDBN} can print parts of your program's source, since the debugging
8591 information recorded in the program tells @value{GDBN} what source files were
8592 used to build it. When your program stops, @value{GDBN} spontaneously prints
8593 the line where it stopped. Likewise, when you select a stack frame
8594 (@pxref{Selection, ,Selecting a Frame}), @value{GDBN} prints the line where
8595 execution in that frame has stopped. You can print other portions of
8596 source files by explicit command.
8598 If you use @value{GDBN} through its @sc{gnu} Emacs interface, you may
8599 prefer to use Emacs facilities to view source; see @ref{Emacs, ,Using
8600 @value{GDBN} under @sc{gnu} Emacs}.
8603 * List:: Printing source lines
8604 * Specify Location:: How to specify code locations
8605 * Edit:: Editing source files
8606 * Search:: Searching source files
8607 * Source Path:: Specifying source directories
8608 * Machine Code:: Source and machine code
8612 @section Printing Source Lines
8615 @kindex l @r{(@code{list})}
8616 To print lines from a source file, use the @code{list} command
8617 (abbreviated @code{l}). By default, ten lines are printed.
8618 There are several ways to specify what part of the file you want to
8619 print; see @ref{Specify Location}, for the full list.
8621 Here are the forms of the @code{list} command most commonly used:
8624 @item list @var{linenum}
8625 Print lines centered around line number @var{linenum} in the
8626 current source file.
8628 @item list @var{function}
8629 Print lines centered around the beginning of function
8633 Print more lines. If the last lines printed were printed with a
8634 @code{list} command, this prints lines following the last lines
8635 printed; however, if the last line printed was a solitary line printed
8636 as part of displaying a stack frame (@pxref{Stack, ,Examining the
8637 Stack}), this prints lines centered around that line.
8640 Print lines just before the lines last printed.
8643 @cindex @code{list}, how many lines to display
8644 By default, @value{GDBN} prints ten source lines with any of these forms of
8645 the @code{list} command. You can change this using @code{set listsize}:
8648 @kindex set listsize
8649 @item set listsize @var{count}
8650 @itemx set listsize unlimited
8651 Make the @code{list} command display @var{count} source lines (unless
8652 the @code{list} argument explicitly specifies some other number).
8653 Setting @var{count} to @code{unlimited} or 0 means there's no limit.
8655 @kindex show listsize
8657 Display the number of lines that @code{list} prints.
8660 Repeating a @code{list} command with @key{RET} discards the argument,
8661 so it is equivalent to typing just @code{list}. This is more useful
8662 than listing the same lines again. An exception is made for an
8663 argument of @samp{-}; that argument is preserved in repetition so that
8664 each repetition moves up in the source file.
8666 In general, the @code{list} command expects you to supply zero, one or two
8667 @dfn{locations}. Locations specify source lines; there are several ways
8668 of writing them (@pxref{Specify Location}), but the effect is always
8669 to specify some source line.
8671 Here is a complete description of the possible arguments for @code{list}:
8674 @item list @var{location}
8675 Print lines centered around the line specified by @var{location}.
8677 @item list @var{first},@var{last}
8678 Print lines from @var{first} to @var{last}. Both arguments are
8679 locations. When a @code{list} command has two locations, and the
8680 source file of the second location is omitted, this refers to
8681 the same source file as the first location.
8683 @item list ,@var{last}
8684 Print lines ending with @var{last}.
8686 @item list @var{first},
8687 Print lines starting with @var{first}.
8690 Print lines just after the lines last printed.
8693 Print lines just before the lines last printed.
8696 As described in the preceding table.
8699 @node Specify Location
8700 @section Specifying a Location
8701 @cindex specifying location
8703 @cindex source location
8706 * Linespec Locations:: Linespec locations
8707 * Explicit Locations:: Explicit locations
8708 * Address Locations:: Address locations
8711 Several @value{GDBN} commands accept arguments that specify a location
8712 of your program's code. Since @value{GDBN} is a source-level
8713 debugger, a location usually specifies some line in the source code.
8714 Locations may be specified using three different formats:
8715 linespec locations, explicit locations, or address locations.
8717 @node Linespec Locations
8718 @subsection Linespec Locations
8719 @cindex linespec locations
8721 A @dfn{linespec} is a colon-separated list of source location parameters such
8722 as file name, function name, etc. Here are all the different ways of
8723 specifying a linespec:
8727 Specifies the line number @var{linenum} of the current source file.
8730 @itemx +@var{offset}
8731 Specifies the line @var{offset} lines before or after the @dfn{current
8732 line}. For the @code{list} command, the current line is the last one
8733 printed; for the breakpoint commands, this is the line at which
8734 execution stopped in the currently selected @dfn{stack frame}
8735 (@pxref{Frames, ,Frames}, for a description of stack frames.) When
8736 used as the second of the two linespecs in a @code{list} command,
8737 this specifies the line @var{offset} lines up or down from the first
8740 @item @var{filename}:@var{linenum}
8741 Specifies the line @var{linenum} in the source file @var{filename}.
8742 If @var{filename} is a relative file name, then it will match any
8743 source file name with the same trailing components. For example, if
8744 @var{filename} is @samp{gcc/expr.c}, then it will match source file
8745 name of @file{/build/trunk/gcc/expr.c}, but not
8746 @file{/build/trunk/libcpp/expr.c} or @file{/build/trunk/gcc/x-expr.c}.
8748 @item @var{function}
8749 Specifies the line that begins the body of the function @var{function}.
8750 For example, in C, this is the line with the open brace.
8752 By default, in C@t{++} and Ada, @var{function} is interpreted as
8753 specifying all functions named @var{function} in all scopes. For
8754 C@t{++}, this means in all namespaces and classes. For Ada, this
8755 means in all packages.
8757 For example, assuming a program with C@t{++} symbols named
8758 @code{A::B::func} and @code{B::func}, both commands @w{@kbd{break
8759 func}} and @w{@kbd{break B::func}} set a breakpoint on both symbols.
8761 Commands that accept a linespec let you override this with the
8762 @code{-qualified} option. For example, @w{@kbd{break -qualified
8763 func}} sets a breakpoint on a free-function named @code{func} ignoring
8764 any C@t{++} class methods and namespace functions called @code{func}.
8766 @xref{Explicit Locations}.
8768 @item @var{function}:@var{label}
8769 Specifies the line where @var{label} appears in @var{function}.
8771 @item @var{filename}:@var{function}
8772 Specifies the line that begins the body of the function @var{function}
8773 in the file @var{filename}. You only need the file name with a
8774 function name to avoid ambiguity when there are identically named
8775 functions in different source files.
8778 Specifies the line at which the label named @var{label} appears
8779 in the function corresponding to the currently selected stack frame.
8780 If there is no current selected stack frame (for instance, if the inferior
8781 is not running), then @value{GDBN} will not search for a label.
8783 @cindex breakpoint at static probe point
8784 @item -pstap|-probe-stap @r{[}@var{objfile}:@r{[}@var{provider}:@r{]}@r{]}@var{name}
8785 The @sc{gnu}/Linux tool @code{SystemTap} provides a way for
8786 applications to embed static probes. @xref{Static Probe Points}, for more
8787 information on finding and using static probes. This form of linespec
8788 specifies the location of such a static probe.
8790 If @var{objfile} is given, only probes coming from that shared library
8791 or executable matching @var{objfile} as a regular expression are considered.
8792 If @var{provider} is given, then only probes from that provider are considered.
8793 If several probes match the spec, @value{GDBN} will insert a breakpoint at
8794 each one of those probes.
8797 @node Explicit Locations
8798 @subsection Explicit Locations
8799 @cindex explicit locations
8801 @dfn{Explicit locations} allow the user to directly specify the source
8802 location's parameters using option-value pairs.
8804 Explicit locations are useful when several functions, labels, or
8805 file names have the same name (base name for files) in the program's
8806 sources. In these cases, explicit locations point to the source
8807 line you meant more accurately and unambiguously. Also, using
8808 explicit locations might be faster in large programs.
8810 For example, the linespec @samp{foo:bar} may refer to a function @code{bar}
8811 defined in the file named @file{foo} or the label @code{bar} in a function
8812 named @code{foo}. @value{GDBN} must search either the file system or
8813 the symbol table to know.
8815 The list of valid explicit location options is summarized in the
8819 @item -source @var{filename}
8820 The value specifies the source file name. To differentiate between
8821 files with the same base name, prepend as many directories as is necessary
8822 to uniquely identify the desired file, e.g., @file{foo/bar/baz.c}. Otherwise
8823 @value{GDBN} will use the first file it finds with the given base
8824 name. This option requires the use of either @code{-function} or @code{-line}.
8826 @item -function @var{function}
8827 The value specifies the name of a function. Operations
8828 on function locations unmodified by other options (such as @code{-label}
8829 or @code{-line}) refer to the line that begins the body of the function.
8830 In C, for example, this is the line with the open brace.
8832 By default, in C@t{++} and Ada, @var{function} is interpreted as
8833 specifying all functions named @var{function} in all scopes. For
8834 C@t{++}, this means in all namespaces and classes. For Ada, this
8835 means in all packages.
8837 For example, assuming a program with C@t{++} symbols named
8838 @code{A::B::func} and @code{B::func}, both commands @w{@kbd{break
8839 -function func}} and @w{@kbd{break -function B::func}} set a
8840 breakpoint on both symbols.
8842 You can use the @kbd{-qualified} flag to override this (see below).
8846 This flag makes @value{GDBN} interpret a function name specified with
8847 @kbd{-function} as a complete fully-qualified name.
8849 For example, assuming a C@t{++} program with symbols named
8850 @code{A::B::func} and @code{B::func}, the @w{@kbd{break -qualified
8851 -function B::func}} command sets a breakpoint on @code{B::func}, only.
8853 (Note: the @kbd{-qualified} option can precede a linespec as well
8854 (@pxref{Linespec Locations}), so the particular example above could be
8855 simplified as @w{@kbd{break -qualified B::func}}.)
8857 @item -label @var{label}
8858 The value specifies the name of a label. When the function
8859 name is not specified, the label is searched in the function of the currently
8860 selected stack frame.
8862 @item -line @var{number}
8863 The value specifies a line offset for the location. The offset may either
8864 be absolute (@code{-line 3}) or relative (@code{-line +3}), depending on
8865 the command. When specified without any other options, the line offset is
8866 relative to the current line.
8869 Explicit location options may be abbreviated by omitting any non-unique
8870 trailing characters from the option name, e.g., @w{@kbd{break -s main.c -li 3}}.
8872 @node Address Locations
8873 @subsection Address Locations
8874 @cindex address locations
8876 @dfn{Address locations} indicate a specific program address. They have
8877 the generalized form *@var{address}.
8879 For line-oriented commands, such as @code{list} and @code{edit}, this
8880 specifies a source line that contains @var{address}. For @code{break} and
8881 other breakpoint-oriented commands, this can be used to set breakpoints in
8882 parts of your program which do not have debugging information or
8885 Here @var{address} may be any expression valid in the current working
8886 language (@pxref{Languages, working language}) that specifies a code
8887 address. In addition, as a convenience, @value{GDBN} extends the
8888 semantics of expressions used in locations to cover several situations
8889 that frequently occur during debugging. Here are the various forms
8893 @item @var{expression}
8894 Any expression valid in the current working language.
8896 @item @var{funcaddr}
8897 An address of a function or procedure derived from its name. In C,
8898 C@t{++}, Objective-C, Fortran, minimal, and assembly, this is
8899 simply the function's name @var{function} (and actually a special case
8900 of a valid expression). In Pascal and Modula-2, this is
8901 @code{&@var{function}}. In Ada, this is @code{@var{function}'Address}
8902 (although the Pascal form also works).
8904 This form specifies the address of the function's first instruction,
8905 before the stack frame and arguments have been set up.
8907 @item '@var{filename}':@var{funcaddr}
8908 Like @var{funcaddr} above, but also specifies the name of the source
8909 file explicitly. This is useful if the name of the function does not
8910 specify the function unambiguously, e.g., if there are several
8911 functions with identical names in different source files.
8915 @section Editing Source Files
8916 @cindex editing source files
8919 @kindex e @r{(@code{edit})}
8920 To edit the lines in a source file, use the @code{edit} command.
8921 The editing program of your choice
8922 is invoked with the current line set to
8923 the active line in the program.
8924 Alternatively, there are several ways to specify what part of the file you
8925 want to print if you want to see other parts of the program:
8928 @item edit @var{location}
8929 Edit the source file specified by @code{location}. Editing starts at
8930 that @var{location}, e.g., at the specified source line of the
8931 specified file. @xref{Specify Location}, for all the possible forms
8932 of the @var{location} argument; here are the forms of the @code{edit}
8933 command most commonly used:
8936 @item edit @var{number}
8937 Edit the current source file with @var{number} as the active line number.
8939 @item edit @var{function}
8940 Edit the file containing @var{function} at the beginning of its definition.
8945 @subsection Choosing your Editor
8946 You can customize @value{GDBN} to use any editor you want
8948 The only restriction is that your editor (say @code{ex}), recognizes the
8949 following command-line syntax:
8951 ex +@var{number} file
8953 The optional numeric value +@var{number} specifies the number of the line in
8954 the file where to start editing.}.
8955 By default, it is @file{@value{EDITOR}}, but you can change this
8956 by setting the environment variable @code{EDITOR} before using
8957 @value{GDBN}. For example, to configure @value{GDBN} to use the
8958 @code{vi} editor, you could use these commands with the @code{sh} shell:
8964 or in the @code{csh} shell,
8966 setenv EDITOR /usr/bin/vi
8971 @section Searching Source Files
8972 @cindex searching source files
8974 There are two commands for searching through the current source file for a
8979 @kindex forward-search
8980 @kindex fo @r{(@code{forward-search})}
8981 @item forward-search @var{regexp}
8982 @itemx search @var{regexp}
8983 The command @samp{forward-search @var{regexp}} checks each line,
8984 starting with the one following the last line listed, for a match for
8985 @var{regexp}. It lists the line that is found. You can use the
8986 synonym @samp{search @var{regexp}} or abbreviate the command name as
8989 @kindex reverse-search
8990 @item reverse-search @var{regexp}
8991 The command @samp{reverse-search @var{regexp}} checks each line, starting
8992 with the one before the last line listed and going backward, for a match
8993 for @var{regexp}. It lists the line that is found. You can abbreviate
8994 this command as @code{rev}.
8998 @section Specifying Source Directories
9001 @cindex directories for source files
9002 Executable programs sometimes do not record the directories of the source
9003 files from which they were compiled, just the names. Even when they do,
9004 the directories could be moved between the compilation and your debugging
9005 session. @value{GDBN} has a list of directories to search for source files;
9006 this is called the @dfn{source path}. Each time @value{GDBN} wants a source file,
9007 it tries all the directories in the list, in the order they are present
9008 in the list, until it finds a file with the desired name.
9010 For example, suppose an executable references the file
9011 @file{/usr/src/foo-1.0/lib/foo.c}, does not record a compilation
9012 directory, and the @dfn{source path} is @file{/mnt/cross}.
9013 @value{GDBN} would look for the source file in the following
9018 @item @file{/usr/src/foo-1.0/lib/foo.c}
9019 @item @file{/mnt/cross/usr/src/foo-1.0/lib/foo.c}
9020 @item @file{/mnt/cross/foo.c}
9024 If the source file is not present at any of the above locations then
9025 an error is printed. @value{GDBN} does not look up the parts of the
9026 source file name, such as @file{/mnt/cross/src/foo-1.0/lib/foo.c}.
9027 Likewise, the subdirectories of the source path are not searched: if
9028 the source path is @file{/mnt/cross}, and the binary refers to
9029 @file{foo.c}, @value{GDBN} would not find it under
9030 @file{/mnt/cross/usr/src/foo-1.0/lib}.
9032 Plain file names, relative file names with leading directories, file
9033 names containing dots, etc.@: are all treated as described above,
9034 except that non-absolute file names are not looked up literally. If
9035 the @dfn{source path} is @file{/mnt/cross}, the source file is
9036 recorded as @file{../lib/foo.c}, and no compilation directory is
9037 recorded, then @value{GDBN} will search in the following locations:
9041 @item @file{/mnt/cross/../lib/foo.c}
9042 @item @file{/mnt/cross/foo.c}
9048 @vindex $cdir@r{, convenience variable}
9049 @vindex $cwd@r{, convenience variable}
9050 @cindex compilation directory
9051 @cindex current directory
9052 @cindex working directory
9053 @cindex directory, current
9054 @cindex directory, compilation
9055 The @dfn{source path} will always include two special entries
9056 @samp{$cdir} and @samp{$cwd}, these refer to the compilation directory
9057 (if one is recorded) and the current working directory respectively.
9059 @samp{$cdir} causes @value{GDBN} to search within the compilation
9060 directory, if one is recorded in the debug information. If no
9061 compilation directory is recorded in the debug information then
9062 @samp{$cdir} is ignored.
9064 @samp{$cwd} is not the same as @samp{.}---the former tracks the
9065 current working directory as it changes during your @value{GDBN}
9066 session, while the latter is immediately expanded to the current
9067 directory at the time you add an entry to the source path.
9069 If a compilation directory is recorded in the debug information, and
9070 @value{GDBN} has not found the source file after the first search
9071 using @dfn{source path}, then @value{GDBN} will combine the
9072 compilation directory and the filename, and then search for the source
9073 file again using the @dfn{source path}.
9075 For example, if the executable records the source file as
9076 @file{/usr/src/foo-1.0/lib/foo.c}, the compilation directory is
9077 recorded as @file{/project/build}, and the @dfn{source path} is
9078 @file{/mnt/cross:$cdir:$cwd} while the current working directory of
9079 the @value{GDBN} session is @file{/home/user}, then @value{GDBN} will
9080 search for the source file in the following locations:
9084 @item @file{/usr/src/foo-1.0/lib/foo.c}
9085 @item @file{/mnt/cross/usr/src/foo-1.0/lib/foo.c}
9086 @item @file{/project/build/usr/src/foo-1.0/lib/foo.c}
9087 @item @file{/home/user/usr/src/foo-1.0/lib/foo.c}
9088 @item @file{/mnt/cross/project/build/usr/src/foo-1.0/lib/foo.c}
9089 @item @file{/project/build/project/build/usr/src/foo-1.0/lib/foo.c}
9090 @item @file{/home/user/project/build/usr/src/foo-1.0/lib/foo.c}
9091 @item @file{/mnt/cross/foo.c}
9092 @item @file{/project/build/foo.c}
9093 @item @file{/home/user/foo.c}
9097 If the file name in the previous example had been recorded in the
9098 executable as a relative path rather than an absolute path, then the
9099 first look up would not have occurred, but all of the remaining steps
9102 When searching for source files on MS-DOS and MS-Windows, where
9103 absolute paths start with a drive letter (e.g.
9104 @file{C:/project/foo.c}), @value{GDBN} will remove the drive letter
9105 from the file name before appending it to a search directory from
9106 @dfn{source path}; for instance if the executable references the
9107 source file @file{C:/project/foo.c} and @dfn{source path} is set to
9108 @file{D:/mnt/cross}, then @value{GDBN} will search in the following
9109 locations for the source file:
9113 @item @file{C:/project/foo.c}
9114 @item @file{D:/mnt/cross/project/foo.c}
9115 @item @file{D:/mnt/cross/foo.c}
9119 Note that the executable search path is @emph{not} used to locate the
9122 Whenever you reset or rearrange the source path, @value{GDBN} clears out
9123 any information it has cached about where source files are found and where
9124 each line is in the file.
9128 When you start @value{GDBN}, its source path includes only @samp{$cdir}
9129 and @samp{$cwd}, in that order.
9130 To add other directories, use the @code{directory} command.
9132 The search path is used to find both program source files and @value{GDBN}
9133 script files (read using the @samp{-command} option and @samp{source} command).
9135 In addition to the source path, @value{GDBN} provides a set of commands
9136 that manage a list of source path substitution rules. A @dfn{substitution
9137 rule} specifies how to rewrite source directories stored in the program's
9138 debug information in case the sources were moved to a different
9139 directory between compilation and debugging. A rule is made of
9140 two strings, the first specifying what needs to be rewritten in
9141 the path, and the second specifying how it should be rewritten.
9142 In @ref{set substitute-path}, we name these two parts @var{from} and
9143 @var{to} respectively. @value{GDBN} does a simple string replacement
9144 of @var{from} with @var{to} at the start of the directory part of the
9145 source file name, and uses that result instead of the original file
9146 name to look up the sources.
9148 Using the previous example, suppose the @file{foo-1.0} tree has been
9149 moved from @file{/usr/src} to @file{/mnt/cross}, then you can tell
9150 @value{GDBN} to replace @file{/usr/src} in all source path names with
9151 @file{/mnt/cross}. The first lookup will then be
9152 @file{/mnt/cross/foo-1.0/lib/foo.c} in place of the original location
9153 of @file{/usr/src/foo-1.0/lib/foo.c}. To define a source path
9154 substitution rule, use the @code{set substitute-path} command
9155 (@pxref{set substitute-path}).
9157 To avoid unexpected substitution results, a rule is applied only if the
9158 @var{from} part of the directory name ends at a directory separator.
9159 For instance, a rule substituting @file{/usr/source} into
9160 @file{/mnt/cross} will be applied to @file{/usr/source/foo-1.0} but
9161 not to @file{/usr/sourceware/foo-2.0}. And because the substitution
9162 is applied only at the beginning of the directory name, this rule will
9163 not be applied to @file{/root/usr/source/baz.c} either.
9165 In many cases, you can achieve the same result using the @code{directory}
9166 command. However, @code{set substitute-path} can be more efficient in
9167 the case where the sources are organized in a complex tree with multiple
9168 subdirectories. With the @code{directory} command, you need to add each
9169 subdirectory of your project. If you moved the entire tree while
9170 preserving its internal organization, then @code{set substitute-path}
9171 allows you to direct the debugger to all the sources with one single
9174 @code{set substitute-path} is also more than just a shortcut command.
9175 The source path is only used if the file at the original location no
9176 longer exists. On the other hand, @code{set substitute-path} modifies
9177 the debugger behavior to look at the rewritten location instead. So, if
9178 for any reason a source file that is not relevant to your executable is
9179 located at the original location, a substitution rule is the only
9180 method available to point @value{GDBN} at the new location.
9182 @cindex @samp{--with-relocated-sources}
9183 @cindex default source path substitution
9184 You can configure a default source path substitution rule by
9185 configuring @value{GDBN} with the
9186 @samp{--with-relocated-sources=@var{dir}} option. The @var{dir}
9187 should be the name of a directory under @value{GDBN}'s configured
9188 prefix (set with @samp{--prefix} or @samp{--exec-prefix}), and
9189 directory names in debug information under @var{dir} will be adjusted
9190 automatically if the installed @value{GDBN} is moved to a new
9191 location. This is useful if @value{GDBN}, libraries or executables
9192 with debug information and corresponding source code are being moved
9196 @item directory @var{dirname} @dots{}
9197 @item dir @var{dirname} @dots{}
9198 Add directory @var{dirname} to the front of the source path. Several
9199 directory names may be given to this command, separated by @samp{:}
9200 (@samp{;} on MS-DOS and MS-Windows, where @samp{:} usually appears as
9201 part of absolute file names) or
9202 whitespace. You may specify a directory that is already in the source
9203 path; this moves it forward, so @value{GDBN} searches it sooner.
9205 The special strings @samp{$cdir} (to refer to the compilation
9206 directory, if one is recorded), and @samp{$cwd} (to refer to the
9207 current working directory) can also be included in the list of
9208 directories @var{dirname}. Though these will already be in the source
9209 path they will be moved forward in the list so @value{GDBN} searches
9213 Reset the source path to its default value (@samp{$cdir:$cwd} on Unix systems). This requires confirmation.
9215 @c RET-repeat for @code{directory} is explicitly disabled, but since
9216 @c repeating it would be a no-op we do not say that. (thanks to RMS)
9218 @item set directories @var{path-list}
9219 @kindex set directories
9220 Set the source path to @var{path-list}.
9221 @samp{$cdir:$cwd} are added if missing.
9223 @item show directories
9224 @kindex show directories
9225 Print the source path: show which directories it contains.
9227 @anchor{set substitute-path}
9228 @item set substitute-path @var{from} @var{to}
9229 @kindex set substitute-path
9230 Define a source path substitution rule, and add it at the end of the
9231 current list of existing substitution rules. If a rule with the same
9232 @var{from} was already defined, then the old rule is also deleted.
9234 For example, if the file @file{/foo/bar/baz.c} was moved to
9235 @file{/mnt/cross/baz.c}, then the command
9238 (@value{GDBP}) set substitute-path /foo/bar /mnt/cross
9242 will tell @value{GDBN} to replace @samp{/foo/bar} with
9243 @samp{/mnt/cross}, which will allow @value{GDBN} to find the file
9244 @file{baz.c} even though it was moved.
9246 In the case when more than one substitution rule have been defined,
9247 the rules are evaluated one by one in the order where they have been
9248 defined. The first one matching, if any, is selected to perform
9251 For instance, if we had entered the following commands:
9254 (@value{GDBP}) set substitute-path /usr/src/include /mnt/include
9255 (@value{GDBP}) set substitute-path /usr/src /mnt/src
9259 @value{GDBN} would then rewrite @file{/usr/src/include/defs.h} into
9260 @file{/mnt/include/defs.h} by using the first rule. However, it would
9261 use the second rule to rewrite @file{/usr/src/lib/foo.c} into
9262 @file{/mnt/src/lib/foo.c}.
9265 @item unset substitute-path [path]
9266 @kindex unset substitute-path
9267 If a path is specified, search the current list of substitution rules
9268 for a rule that would rewrite that path. Delete that rule if found.
9269 A warning is emitted by the debugger if no rule could be found.
9271 If no path is specified, then all substitution rules are deleted.
9273 @item show substitute-path [path]
9274 @kindex show substitute-path
9275 If a path is specified, then print the source path substitution rule
9276 which would rewrite that path, if any.
9278 If no path is specified, then print all existing source path substitution
9283 If your source path is cluttered with directories that are no longer of
9284 interest, @value{GDBN} may sometimes cause confusion by finding the wrong
9285 versions of source. You can correct the situation as follows:
9289 Use @code{directory} with no argument to reset the source path to its default value.
9292 Use @code{directory} with suitable arguments to reinstall the
9293 directories you want in the source path. You can add all the
9294 directories in one command.
9298 @section Source and Machine Code
9299 @cindex source line and its code address
9301 You can use the command @code{info line} to map source lines to program
9302 addresses (and vice versa), and the command @code{disassemble} to display
9303 a range of addresses as machine instructions. You can use the command
9304 @code{set disassemble-next-line} to set whether to disassemble next
9305 source line when execution stops. When run under @sc{gnu} Emacs
9306 mode, the @code{info line} command causes the arrow to point to the
9307 line specified. Also, @code{info line} prints addresses in symbolic form as
9313 @itemx info line @var{location}
9314 Print the starting and ending addresses of the compiled code for
9315 source line @var{location}. You can specify source lines in any of
9316 the ways documented in @ref{Specify Location}. With no @var{location}
9317 information about the current source line is printed.
9320 For example, we can use @code{info line} to discover the location of
9321 the object code for the first line of function
9322 @code{m4_changequote}:
9325 (@value{GDBP}) info line m4_changequote
9326 Line 895 of "builtin.c" starts at pc 0x634c <m4_changequote> and \
9327 ends at 0x6350 <m4_changequote+4>.
9331 @cindex code address and its source line
9332 We can also inquire (using @code{*@var{addr}} as the form for
9333 @var{location}) what source line covers a particular address:
9335 (@value{GDBP}) info line *0x63ff
9336 Line 926 of "builtin.c" starts at pc 0x63e4 <m4_changequote+152> and \
9337 ends at 0x6404 <m4_changequote+184>.
9340 @cindex @code{$_} and @code{info line}
9341 @cindex @code{x} command, default address
9342 @kindex x@r{(examine), and} info line
9343 After @code{info line}, the default address for the @code{x} command
9344 is changed to the starting address of the line, so that @samp{x/i} is
9345 sufficient to begin examining the machine code (@pxref{Memory,
9346 ,Examining Memory}). Also, this address is saved as the value of the
9347 convenience variable @code{$_} (@pxref{Convenience Vars, ,Convenience
9350 @cindex info line, repeated calls
9351 After @code{info line}, using @code{info line} again without
9352 specifying a location will display information about the next source
9357 @cindex assembly instructions
9358 @cindex instructions, assembly
9359 @cindex machine instructions
9360 @cindex listing machine instructions
9362 @itemx disassemble /m
9363 @itemx disassemble /s
9364 @itemx disassemble /r
9365 This specialized command dumps a range of memory as machine
9366 instructions. It can also print mixed source+disassembly by specifying
9367 the @code{/m} or @code{/s} modifier and print the raw instructions in hex
9368 as well as in symbolic form by specifying the @code{/r} modifier.
9369 The default memory range is the function surrounding the
9370 program counter of the selected frame. A single argument to this
9371 command is a program counter value; @value{GDBN} dumps the function
9372 surrounding this value. When two arguments are given, they should
9373 be separated by a comma, possibly surrounded by whitespace. The
9374 arguments specify a range of addresses to dump, in one of two forms:
9377 @item @var{start},@var{end}
9378 the addresses from @var{start} (inclusive) to @var{end} (exclusive)
9379 @item @var{start},+@var{length}
9380 the addresses from @var{start} (inclusive) to
9381 @code{@var{start}+@var{length}} (exclusive).
9385 When 2 arguments are specified, the name of the function is also
9386 printed (since there could be several functions in the given range).
9388 The argument(s) can be any expression yielding a numeric value, such as
9389 @samp{0x32c4}, @samp{&main+10} or @samp{$pc - 8}.
9391 If the range of memory being disassembled contains current program counter,
9392 the instruction at that location is shown with a @code{=>} marker.
9395 The following example shows the disassembly of a range of addresses of
9396 HP PA-RISC 2.0 code:
9399 (@value{GDBP}) disas 0x32c4, 0x32e4
9400 Dump of assembler code from 0x32c4 to 0x32e4:
9401 0x32c4 <main+204>: addil 0,dp
9402 0x32c8 <main+208>: ldw 0x22c(sr0,r1),r26
9403 0x32cc <main+212>: ldil 0x3000,r31
9404 0x32d0 <main+216>: ble 0x3f8(sr4,r31)
9405 0x32d4 <main+220>: ldo 0(r31),rp
9406 0x32d8 <main+224>: addil -0x800,dp
9407 0x32dc <main+228>: ldo 0x588(r1),r26
9408 0x32e0 <main+232>: ldil 0x3000,r31
9409 End of assembler dump.
9412 Here is an example showing mixed source+assembly for Intel x86
9413 with @code{/m} or @code{/s}, when the program is stopped just after
9414 function prologue in a non-optimized function with no inline code.
9417 (@value{GDBP}) disas /m main
9418 Dump of assembler code for function main:
9420 0x08048330 <+0>: push %ebp
9421 0x08048331 <+1>: mov %esp,%ebp
9422 0x08048333 <+3>: sub $0x8,%esp
9423 0x08048336 <+6>: and $0xfffffff0,%esp
9424 0x08048339 <+9>: sub $0x10,%esp
9426 6 printf ("Hello.\n");
9427 => 0x0804833c <+12>: movl $0x8048440,(%esp)
9428 0x08048343 <+19>: call 0x8048284 <puts@@plt>
9432 0x08048348 <+24>: mov $0x0,%eax
9433 0x0804834d <+29>: leave
9434 0x0804834e <+30>: ret
9436 End of assembler dump.
9439 The @code{/m} option is deprecated as its output is not useful when
9440 there is either inlined code or re-ordered code.
9441 The @code{/s} option is the preferred choice.
9442 Here is an example for AMD x86-64 showing the difference between
9443 @code{/m} output and @code{/s} output.
9444 This example has one inline function defined in a header file,
9445 and the code is compiled with @samp{-O2} optimization.
9446 Note how the @code{/m} output is missing the disassembly of
9447 several instructions that are present in the @code{/s} output.
9477 (@value{GDBP}) disas /m main
9478 Dump of assembler code for function main:
9482 0x0000000000400400 <+0>: mov 0x200c2e(%rip),%eax # 0x601034 <y>
9483 0x0000000000400417 <+23>: mov %eax,0x200c13(%rip) # 0x601030 <x>
9487 0x000000000040041d <+29>: xor %eax,%eax
9488 0x000000000040041f <+31>: retq
9489 0x0000000000400420 <+32>: add %eax,%eax
9490 0x0000000000400422 <+34>: jmp 0x400417 <main+23>
9492 End of assembler dump.
9493 (@value{GDBP}) disas /s main
9494 Dump of assembler code for function main:
9498 0x0000000000400400 <+0>: mov 0x200c2e(%rip),%eax # 0x601034 <y>
9502 0x0000000000400406 <+6>: test %eax,%eax
9503 0x0000000000400408 <+8>: js 0x400420 <main+32>
9508 0x000000000040040a <+10>: lea 0xa(%rax),%edx
9509 0x000000000040040d <+13>: test %eax,%eax
9510 0x000000000040040f <+15>: mov $0x1,%eax
9511 0x0000000000400414 <+20>: cmovne %edx,%eax
9515 0x0000000000400417 <+23>: mov %eax,0x200c13(%rip) # 0x601030 <x>
9519 0x000000000040041d <+29>: xor %eax,%eax
9520 0x000000000040041f <+31>: retq
9524 0x0000000000400420 <+32>: add %eax,%eax
9525 0x0000000000400422 <+34>: jmp 0x400417 <main+23>
9526 End of assembler dump.
9529 Here is another example showing raw instructions in hex for AMD x86-64,
9532 (@value{GDBP}) disas /r 0x400281,+10
9533 Dump of assembler code from 0x400281 to 0x40028b:
9534 0x0000000000400281: 38 36 cmp %dh,(%rsi)
9535 0x0000000000400283: 2d 36 34 2e 73 sub $0x732e3436,%eax
9536 0x0000000000400288: 6f outsl %ds:(%rsi),(%dx)
9537 0x0000000000400289: 2e 32 00 xor %cs:(%rax),%al
9538 End of assembler dump.
9541 Addresses cannot be specified as a location (@pxref{Specify Location}).
9542 So, for example, if you want to disassemble function @code{bar}
9543 in file @file{foo.c}, you must type @samp{disassemble 'foo.c'::bar}
9544 and not @samp{disassemble foo.c:bar}.
9546 Some architectures have more than one commonly-used set of instruction
9547 mnemonics or other syntax.
9549 For programs that were dynamically linked and use shared libraries,
9550 instructions that call functions or branch to locations in the shared
9551 libraries might show a seemingly bogus location---it's actually a
9552 location of the relocation table. On some architectures, @value{GDBN}
9553 might be able to resolve these to actual function names.
9556 @kindex set disassembler-options
9557 @cindex disassembler options
9558 @item set disassembler-options @var{option1}[,@var{option2}@dots{}]
9559 This command controls the passing of target specific information to
9560 the disassembler. For a list of valid options, please refer to the
9561 @code{-M}/@code{--disassembler-options} section of the @samp{objdump}
9562 manual and/or the output of @kbd{objdump --help}
9563 (@pxref{objdump,,objdump,binutils,The GNU Binary Utilities}).
9564 The default value is the empty string.
9566 If it is necessary to specify more than one disassembler option, then
9567 multiple options can be placed together into a comma separated list.
9568 Currently this command is only supported on targets ARM, MIPS, PowerPC
9571 @kindex show disassembler-options
9572 @item show disassembler-options
9573 Show the current setting of the disassembler options.
9577 @kindex set disassembly-flavor
9578 @cindex Intel disassembly flavor
9579 @cindex AT&T disassembly flavor
9580 @item set disassembly-flavor @var{instruction-set}
9581 Select the instruction set to use when disassembling the
9582 program via the @code{disassemble} or @code{x/i} commands.
9584 Currently this command is only defined for the Intel x86 family. You
9585 can set @var{instruction-set} to either @code{intel} or @code{att}.
9586 The default is @code{att}, the AT&T flavor used by default by Unix
9587 assemblers for x86-based targets.
9589 @kindex show disassembly-flavor
9590 @item show disassembly-flavor
9591 Show the current setting of the disassembly flavor.
9595 @kindex set disassemble-next-line
9596 @kindex show disassemble-next-line
9597 @item set disassemble-next-line
9598 @itemx show disassemble-next-line
9599 Control whether or not @value{GDBN} will disassemble the next source
9600 line or instruction when execution stops. If ON, @value{GDBN} will
9601 display disassembly of the next source line when execution of the
9602 program being debugged stops. This is @emph{in addition} to
9603 displaying the source line itself, which @value{GDBN} always does if
9604 possible. If the next source line cannot be displayed for some reason
9605 (e.g., if @value{GDBN} cannot find the source file, or there's no line
9606 info in the debug info), @value{GDBN} will display disassembly of the
9607 next @emph{instruction} instead of showing the next source line. If
9608 AUTO, @value{GDBN} will display disassembly of next instruction only
9609 if the source line cannot be displayed. This setting causes
9610 @value{GDBN} to display some feedback when you step through a function
9611 with no line info or whose source file is unavailable. The default is
9612 OFF, which means never display the disassembly of the next line or
9618 @chapter Examining Data
9620 @cindex printing data
9621 @cindex examining data
9624 The usual way to examine data in your program is with the @code{print}
9625 command (abbreviated @code{p}), or its synonym @code{inspect}. It
9626 evaluates and prints the value of an expression of the language your
9627 program is written in (@pxref{Languages, ,Using @value{GDBN} with
9628 Different Languages}). It may also print the expression using a
9629 Python-based pretty-printer (@pxref{Pretty Printing}).
9632 @item print [[@var{options}] --] @var{expr}
9633 @itemx print [[@var{options}] --] /@var{f} @var{expr}
9634 @var{expr} is an expression (in the source language). By default the
9635 value of @var{expr} is printed in a format appropriate to its data type;
9636 you can choose a different format by specifying @samp{/@var{f}}, where
9637 @var{f} is a letter specifying the format; see @ref{Output Formats,,Output
9640 @anchor{print options}
9641 The @code{print} command supports a number of options that allow
9642 overriding relevant global print settings as set by @code{set print}
9646 @item -address [@code{on}|@code{off}]
9647 Set printing of addresses.
9648 Related setting: @ref{set print address}.
9650 @item -array [@code{on}|@code{off}]
9651 Pretty formatting of arrays.
9652 Related setting: @ref{set print array}.
9654 @item -array-indexes [@code{on}|@code{off}]
9655 Set printing of array indexes.
9656 Related setting: @ref{set print array-indexes}.
9658 @item -elements @var{number-of-elements}|@code{unlimited}
9659 Set limit on string chars or array elements to print. The value
9660 @code{unlimited} causes there to be no limit. Related setting:
9661 @ref{set print elements}.
9663 @item -max-depth @var{depth}|@code{unlimited}
9664 Set the threshold after which nested structures are replaced with
9665 ellipsis. Related setting: @ref{set print max-depth}.
9667 @item -null-stop [@code{on}|@code{off}]
9668 Set printing of char arrays to stop at first null char. Related
9669 setting: @ref{set print null-stop}.
9671 @item -object [@code{on}|@code{off}]
9672 Set printing C@t{++} virtual function tables. Related setting:
9673 @ref{set print object}.
9675 @item -pretty [@code{on}|@code{off}]
9676 Set pretty formatting of structures. Related setting: @ref{set print
9679 @item -raw-values [@code{on}|@code{off}]
9680 Set whether to print values in raw form, bypassing any
9681 pretty-printers for that value. Related setting: @ref{set print
9684 @item -repeats @var{number-of-repeats}|@code{unlimited}
9685 Set threshold for repeated print elements. @code{unlimited} causes
9686 all elements to be individually printed. Related setting: @ref{set
9689 @item -static-members [@code{on}|@code{off}]
9690 Set printing C@t{++} static members. Related setting: @ref{set print
9693 @item -symbol [@code{on}|@code{off}]
9694 Set printing of symbol names when printing pointers. Related setting:
9695 @ref{set print symbol}.
9697 @item -union [@code{on}|@code{off}]
9698 Set printing of unions interior to structures. Related setting:
9699 @ref{set print union}.
9701 @item -vtbl [@code{on}|@code{off}]
9702 Set printing of C++ virtual function tables. Related setting:
9703 @ref{set print vtbl}.
9706 Because the @code{print} command accepts arbitrary expressions which
9707 may look like options (including abbreviations), if you specify any
9708 command option, then you must use a double dash (@code{--}) to mark
9709 the end of option processing.
9711 For example, this prints the value of the @code{-p} expression:
9714 (@value{GDBP}) print -p
9717 While this repeats the last value in the value history (see below)
9718 with the @code{-pretty} option in effect:
9721 (@value{GDBP}) print -p --
9724 Here is an example including both on option and an expression:
9728 (@value{GDBP}) print -pretty -- *myptr
9740 @item print [@var{options}]
9741 @itemx print [@var{options}] /@var{f}
9742 @cindex reprint the last value
9743 If you omit @var{expr}, @value{GDBN} displays the last value again (from the
9744 @dfn{value history}; @pxref{Value History, ,Value History}). This allows you to
9745 conveniently inspect the same value in an alternative format.
9748 A more low-level way of examining data is with the @code{x} command.
9749 It examines data in memory at a specified address and prints it in a
9750 specified format. @xref{Memory, ,Examining Memory}.
9752 If you are interested in information about types, or about how the
9753 fields of a struct or a class are declared, use the @code{ptype @var{exp}}
9754 command rather than @code{print}. @xref{Symbols, ,Examining the Symbol
9757 @cindex exploring hierarchical data structures
9759 Another way of examining values of expressions and type information is
9760 through the Python extension command @code{explore} (available only if
9761 the @value{GDBN} build is configured with @code{--with-python}). It
9762 offers an interactive way to start at the highest level (or, the most
9763 abstract level) of the data type of an expression (or, the data type
9764 itself) and explore all the way down to leaf scalar values/fields
9765 embedded in the higher level data types.
9768 @item explore @var{arg}
9769 @var{arg} is either an expression (in the source language), or a type
9770 visible in the current context of the program being debugged.
9773 The working of the @code{explore} command can be illustrated with an
9774 example. If a data type @code{struct ComplexStruct} is defined in your
9784 struct ComplexStruct
9786 struct SimpleStruct *ss_p;
9792 followed by variable declarations as
9795 struct SimpleStruct ss = @{ 10, 1.11 @};
9796 struct ComplexStruct cs = @{ &ss, @{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 @} @};
9800 then, the value of the variable @code{cs} can be explored using the
9801 @code{explore} command as follows.
9804 (@value{GDBP}) explore cs
9805 The value of `cs' is a struct/class of type `struct ComplexStruct' with
9806 the following fields:
9808 ss_p = <Enter 0 to explore this field of type `struct SimpleStruct *'>
9809 arr = <Enter 1 to explore this field of type `int [10]'>
9811 Enter the field number of choice:
9815 Since the fields of @code{cs} are not scalar values, you are being
9816 prompted to chose the field you want to explore. Let's say you choose
9817 the field @code{ss_p} by entering @code{0}. Then, since this field is a
9818 pointer, you will be asked if it is pointing to a single value. From
9819 the declaration of @code{cs} above, it is indeed pointing to a single
9820 value, hence you enter @code{y}. If you enter @code{n}, then you will
9821 be asked if it were pointing to an array of values, in which case this
9822 field will be explored as if it were an array.
9825 `cs.ss_p' is a pointer to a value of type `struct SimpleStruct'
9826 Continue exploring it as a pointer to a single value [y/n]: y
9827 The value of `*(cs.ss_p)' is a struct/class of type `struct
9828 SimpleStruct' with the following fields:
9830 i = 10 .. (Value of type `int')
9831 d = 1.1100000000000001 .. (Value of type `double')
9833 Press enter to return to parent value:
9837 If the field @code{arr} of @code{cs} was chosen for exploration by
9838 entering @code{1} earlier, then since it is as array, you will be
9839 prompted to enter the index of the element in the array that you want
9843 `cs.arr' is an array of `int'.
9844 Enter the index of the element you want to explore in `cs.arr': 5
9846 `(cs.arr)[5]' is a scalar value of type `int'.
9850 Press enter to return to parent value:
9853 In general, at any stage of exploration, you can go deeper towards the
9854 leaf values by responding to the prompts appropriately, or hit the
9855 return key to return to the enclosing data structure (the @i{higher}
9856 level data structure).
9858 Similar to exploring values, you can use the @code{explore} command to
9859 explore types. Instead of specifying a value (which is typically a
9860 variable name or an expression valid in the current context of the
9861 program being debugged), you specify a type name. If you consider the
9862 same example as above, your can explore the type
9863 @code{struct ComplexStruct} by passing the argument
9864 @code{struct ComplexStruct} to the @code{explore} command.
9867 (@value{GDBP}) explore struct ComplexStruct
9871 By responding to the prompts appropriately in the subsequent interactive
9872 session, you can explore the type @code{struct ComplexStruct} in a
9873 manner similar to how the value @code{cs} was explored in the above
9876 The @code{explore} command also has two sub-commands,
9877 @code{explore value} and @code{explore type}. The former sub-command is
9878 a way to explicitly specify that value exploration of the argument is
9879 being invoked, while the latter is a way to explicitly specify that type
9880 exploration of the argument is being invoked.
9883 @item explore value @var{expr}
9884 @cindex explore value
9885 This sub-command of @code{explore} explores the value of the
9886 expression @var{expr} (if @var{expr} is an expression valid in the
9887 current context of the program being debugged). The behavior of this
9888 command is identical to that of the behavior of the @code{explore}
9889 command being passed the argument @var{expr}.
9891 @item explore type @var{arg}
9892 @cindex explore type
9893 This sub-command of @code{explore} explores the type of @var{arg} (if
9894 @var{arg} is a type visible in the current context of program being
9895 debugged), or the type of the value/expression @var{arg} (if @var{arg}
9896 is an expression valid in the current context of the program being
9897 debugged). If @var{arg} is a type, then the behavior of this command is
9898 identical to that of the @code{explore} command being passed the
9899 argument @var{arg}. If @var{arg} is an expression, then the behavior of
9900 this command will be identical to that of the @code{explore} command
9901 being passed the type of @var{arg} as the argument.
9905 * Expressions:: Expressions
9906 * Ambiguous Expressions:: Ambiguous Expressions
9907 * Variables:: Program variables
9908 * Arrays:: Artificial arrays
9909 * Output Formats:: Output formats
9910 * Memory:: Examining memory
9911 * Auto Display:: Automatic display
9912 * Print Settings:: Print settings
9913 * Pretty Printing:: Python pretty printing
9914 * Value History:: Value history
9915 * Convenience Vars:: Convenience variables
9916 * Convenience Funs:: Convenience functions
9917 * Registers:: Registers
9918 * Floating Point Hardware:: Floating point hardware
9919 * Vector Unit:: Vector Unit
9920 * OS Information:: Auxiliary data provided by operating system
9921 * Memory Region Attributes:: Memory region attributes
9922 * Dump/Restore Files:: Copy between memory and a file
9923 * Core File Generation:: Cause a program dump its core
9924 * Character Sets:: Debugging programs that use a different
9925 character set than GDB does
9926 * Caching Target Data:: Data caching for targets
9927 * Searching Memory:: Searching memory for a sequence of bytes
9928 * Value Sizes:: Managing memory allocated for values
9932 @section Expressions
9935 @code{print} and many other @value{GDBN} commands accept an expression and
9936 compute its value. Any kind of constant, variable or operator defined
9937 by the programming language you are using is valid in an expression in
9938 @value{GDBN}. This includes conditional expressions, function calls,
9939 casts, and string constants. It also includes preprocessor macros, if
9940 you compiled your program to include this information; see
9943 @cindex arrays in expressions
9944 @value{GDBN} supports array constants in expressions input by
9945 the user. The syntax is @{@var{element}, @var{element}@dots{}@}. For example,
9946 you can use the command @code{print @{1, 2, 3@}} to create an array
9947 of three integers. If you pass an array to a function or assign it
9948 to a program variable, @value{GDBN} copies the array to memory that
9949 is @code{malloc}ed in the target program.
9951 Because C is so widespread, most of the expressions shown in examples in
9952 this manual are in C. @xref{Languages, , Using @value{GDBN} with Different
9953 Languages}, for information on how to use expressions in other
9956 In this section, we discuss operators that you can use in @value{GDBN}
9957 expressions regardless of your programming language.
9959 @cindex casts, in expressions
9960 Casts are supported in all languages, not just in C, because it is so
9961 useful to cast a number into a pointer in order to examine a structure
9962 at that address in memory.
9963 @c FIXME: casts supported---Mod2 true?
9965 @value{GDBN} supports these operators, in addition to those common
9966 to programming languages:
9970 @samp{@@} is a binary operator for treating parts of memory as arrays.
9971 @xref{Arrays, ,Artificial Arrays}, for more information.
9974 @samp{::} allows you to specify a variable in terms of the file or
9975 function where it is defined. @xref{Variables, ,Program Variables}.
9977 @cindex @{@var{type}@}
9978 @cindex type casting memory
9979 @cindex memory, viewing as typed object
9980 @cindex casts, to view memory
9981 @item @{@var{type}@} @var{addr}
9982 Refers to an object of type @var{type} stored at address @var{addr} in
9983 memory. The address @var{addr} may be any expression whose value is
9984 an integer or pointer (but parentheses are required around binary
9985 operators, just as in a cast). This construct is allowed regardless
9986 of what kind of data is normally supposed to reside at @var{addr}.
9989 @node Ambiguous Expressions
9990 @section Ambiguous Expressions
9991 @cindex ambiguous expressions
9993 Expressions can sometimes contain some ambiguous elements. For instance,
9994 some programming languages (notably Ada, C@t{++} and Objective-C) permit
9995 a single function name to be defined several times, for application in
9996 different contexts. This is called @dfn{overloading}. Another example
9997 involving Ada is generics. A @dfn{generic package} is similar to C@t{++}
9998 templates and is typically instantiated several times, resulting in
9999 the same function name being defined in different contexts.
10001 In some cases and depending on the language, it is possible to adjust
10002 the expression to remove the ambiguity. For instance in C@t{++}, you
10003 can specify the signature of the function you want to break on, as in
10004 @kbd{break @var{function}(@var{types})}. In Ada, using the fully
10005 qualified name of your function often makes the expression unambiguous
10008 When an ambiguity that needs to be resolved is detected, the debugger
10009 has the capability to display a menu of numbered choices for each
10010 possibility, and then waits for the selection with the prompt @samp{>}.
10011 The first option is always @samp{[0] cancel}, and typing @kbd{0 @key{RET}}
10012 aborts the current command. If the command in which the expression was
10013 used allows more than one choice to be selected, the next option in the
10014 menu is @samp{[1] all}, and typing @kbd{1 @key{RET}} selects all possible
10017 For example, the following session excerpt shows an attempt to set a
10018 breakpoint at the overloaded symbol @code{String::after}.
10019 We choose three particular definitions of that function name:
10021 @c FIXME! This is likely to change to show arg type lists, at least
10024 (@value{GDBP}) b String::after
10027 [2] file:String.cc; line number:867
10028 [3] file:String.cc; line number:860
10029 [4] file:String.cc; line number:875
10030 [5] file:String.cc; line number:853
10031 [6] file:String.cc; line number:846
10032 [7] file:String.cc; line number:735
10034 Breakpoint 1 at 0xb26c: file String.cc, line 867.
10035 Breakpoint 2 at 0xb344: file String.cc, line 875.
10036 Breakpoint 3 at 0xafcc: file String.cc, line 846.
10037 Multiple breakpoints were set.
10038 Use the "delete" command to delete unwanted
10045 @kindex set multiple-symbols
10046 @item set multiple-symbols @var{mode}
10047 @cindex multiple-symbols menu
10049 This option allows you to adjust the debugger behavior when an expression
10052 By default, @var{mode} is set to @code{all}. If the command with which
10053 the expression is used allows more than one choice, then @value{GDBN}
10054 automatically selects all possible choices. For instance, inserting
10055 a breakpoint on a function using an ambiguous name results in a breakpoint
10056 inserted on each possible match. However, if a unique choice must be made,
10057 then @value{GDBN} uses the menu to help you disambiguate the expression.
10058 For instance, printing the address of an overloaded function will result
10059 in the use of the menu.
10061 When @var{mode} is set to @code{ask}, the debugger always uses the menu
10062 when an ambiguity is detected.
10064 Finally, when @var{mode} is set to @code{cancel}, the debugger reports
10065 an error due to the ambiguity and the command is aborted.
10067 @kindex show multiple-symbols
10068 @item show multiple-symbols
10069 Show the current value of the @code{multiple-symbols} setting.
10073 @section Program Variables
10075 The most common kind of expression to use is the name of a variable
10078 Variables in expressions are understood in the selected stack frame
10079 (@pxref{Selection, ,Selecting a Frame}); they must be either:
10083 global (or file-static)
10090 visible according to the scope rules of the
10091 programming language from the point of execution in that frame
10094 @noindent This means that in the function
10109 you can examine and use the variable @code{a} whenever your program is
10110 executing within the function @code{foo}, but you can only use or
10111 examine the variable @code{b} while your program is executing inside
10112 the block where @code{b} is declared.
10114 @cindex variable name conflict
10115 There is an exception: you can refer to a variable or function whose
10116 scope is a single source file even if the current execution point is not
10117 in this file. But it is possible to have more than one such variable or
10118 function with the same name (in different source files). If that
10119 happens, referring to that name has unpredictable effects. If you wish,
10120 you can specify a static variable in a particular function or file by
10121 using the colon-colon (@code{::}) notation:
10123 @cindex colon-colon, context for variables/functions
10125 @c info cannot cope with a :: index entry, but why deprive hard copy readers?
10126 @cindex @code{::}, context for variables/functions
10129 @var{file}::@var{variable}
10130 @var{function}::@var{variable}
10134 Here @var{file} or @var{function} is the name of the context for the
10135 static @var{variable}. In the case of file names, you can use quotes to
10136 make sure @value{GDBN} parses the file name as a single word---for example,
10137 to print a global value of @code{x} defined in @file{f2.c}:
10140 (@value{GDBP}) p 'f2.c'::x
10143 The @code{::} notation is normally used for referring to
10144 static variables, since you typically disambiguate uses of local variables
10145 in functions by selecting the appropriate frame and using the
10146 simple name of the variable. However, you may also use this notation
10147 to refer to local variables in frames enclosing the selected frame:
10156 process (a); /* Stop here */
10167 For example, if there is a breakpoint at the commented line,
10168 here is what you might see
10169 when the program stops after executing the call @code{bar(0)}:
10174 (@value{GDBP}) p bar::a
10176 (@value{GDBP}) up 2
10177 #2 0x080483d0 in foo (a=5) at foobar.c:12
10180 (@value{GDBP}) p bar::a
10184 @cindex C@t{++} scope resolution
10185 These uses of @samp{::} are very rarely in conflict with the very
10186 similar use of the same notation in C@t{++}. When they are in
10187 conflict, the C@t{++} meaning takes precedence; however, this can be
10188 overridden by quoting the file or function name with single quotes.
10190 For example, suppose the program is stopped in a method of a class
10191 that has a field named @code{includefile}, and there is also an
10192 include file named @file{includefile} that defines a variable,
10193 @code{some_global}.
10196 (@value{GDBP}) p includefile
10198 (@value{GDBP}) p includefile::some_global
10199 A syntax error in expression, near `'.
10200 (@value{GDBP}) p 'includefile'::some_global
10204 @cindex wrong values
10205 @cindex variable values, wrong
10206 @cindex function entry/exit, wrong values of variables
10207 @cindex optimized code, wrong values of variables
10209 @emph{Warning:} Occasionally, a local variable may appear to have the
10210 wrong value at certain points in a function---just after entry to a new
10211 scope, and just before exit.
10213 You may see this problem when you are stepping by machine instructions.
10214 This is because, on most machines, it takes more than one instruction to
10215 set up a stack frame (including local variable definitions); if you are
10216 stepping by machine instructions, variables may appear to have the wrong
10217 values until the stack frame is completely built. On exit, it usually
10218 also takes more than one machine instruction to destroy a stack frame;
10219 after you begin stepping through that group of instructions, local
10220 variable definitions may be gone.
10222 This may also happen when the compiler does significant optimizations.
10223 To be sure of always seeing accurate values, turn off all optimization
10226 @cindex ``No symbol "foo" in current context''
10227 Another possible effect of compiler optimizations is to optimize
10228 unused variables out of existence, or assign variables to registers (as
10229 opposed to memory addresses). Depending on the support for such cases
10230 offered by the debug info format used by the compiler, @value{GDBN}
10231 might not be able to display values for such local variables. If that
10232 happens, @value{GDBN} will print a message like this:
10235 No symbol "foo" in current context.
10238 To solve such problems, either recompile without optimizations, or use a
10239 different debug info format, if the compiler supports several such
10240 formats. @xref{Compilation}, for more information on choosing compiler
10241 options. @xref{C, ,C and C@t{++}}, for more information about debug
10242 info formats that are best suited to C@t{++} programs.
10244 If you ask to print an object whose contents are unknown to
10245 @value{GDBN}, e.g., because its data type is not completely specified
10246 by the debug information, @value{GDBN} will say @samp{<incomplete
10247 type>}. @xref{Symbols, incomplete type}, for more about this.
10249 @cindex no debug info variables
10250 If you try to examine or use the value of a (global) variable for
10251 which @value{GDBN} has no type information, e.g., because the program
10252 includes no debug information, @value{GDBN} displays an error message.
10253 @xref{Symbols, unknown type}, for more about unknown types. If you
10254 cast the variable to its declared type, @value{GDBN} gets the
10255 variable's value using the cast-to type as the variable's type. For
10256 example, in a C program:
10259 (@value{GDBP}) p var
10260 'var' has unknown type; cast it to its declared type
10261 (@value{GDBP}) p (float) var
10265 If you append @kbd{@@entry} string to a function parameter name you get its
10266 value at the time the function got called. If the value is not available an
10267 error message is printed. Entry values are available only with some compilers.
10268 Entry values are normally also printed at the function parameter list according
10269 to @ref{set print entry-values}.
10272 Breakpoint 1, d (i=30) at gdb.base/entry-value.c:29
10274 (@value{GDBP}) next
10276 (@value{GDBP}) print i
10278 (@value{GDBP}) print i@@entry
10282 Strings are identified as arrays of @code{char} values without specified
10283 signedness. Arrays of either @code{signed char} or @code{unsigned char} get
10284 printed as arrays of 1 byte sized integers. @code{-fsigned-char} or
10285 @code{-funsigned-char} @value{NGCC} options have no effect as @value{GDBN}
10286 defines literal string type @code{"char"} as @code{char} without a sign.
10291 signed char var1[] = "A";
10294 You get during debugging
10296 (@value{GDBP}) print var0
10298 (@value{GDBP}) print var1
10299 $2 = @{65 'A', 0 '\0'@}
10303 @section Artificial Arrays
10305 @cindex artificial array
10307 @kindex @@@r{, referencing memory as an array}
10308 It is often useful to print out several successive objects of the
10309 same type in memory; a section of an array, or an array of
10310 dynamically determined size for which only a pointer exists in the
10313 You can do this by referring to a contiguous span of memory as an
10314 @dfn{artificial array}, using the binary operator @samp{@@}. The left
10315 operand of @samp{@@} should be the first element of the desired array
10316 and be an individual object. The right operand should be the desired length
10317 of the array. The result is an array value whose elements are all of
10318 the type of the left argument. The first element is actually the left
10319 argument; the second element comes from bytes of memory immediately
10320 following those that hold the first element, and so on. Here is an
10321 example. If a program says
10324 int *array = (int *) malloc (len * sizeof (int));
10328 you can print the contents of @code{array} with
10334 The left operand of @samp{@@} must reside in memory. Array values made
10335 with @samp{@@} in this way behave just like other arrays in terms of
10336 subscripting, and are coerced to pointers when used in expressions.
10337 Artificial arrays most often appear in expressions via the value history
10338 (@pxref{Value History, ,Value History}), after printing one out.
10340 Another way to create an artificial array is to use a cast.
10341 This re-interprets a value as if it were an array.
10342 The value need not be in memory:
10344 (@value{GDBP}) p/x (short[2])0x12345678
10345 $1 = @{0x1234, 0x5678@}
10348 As a convenience, if you leave the array length out (as in
10349 @samp{(@var{type}[])@var{value}}) @value{GDBN} calculates the size to fill
10350 the value (as @samp{sizeof(@var{value})/sizeof(@var{type})}:
10352 (@value{GDBP}) p/x (short[])0x12345678
10353 $2 = @{0x1234, 0x5678@}
10356 Sometimes the artificial array mechanism is not quite enough; in
10357 moderately complex data structures, the elements of interest may not
10358 actually be adjacent---for example, if you are interested in the values
10359 of pointers in an array. One useful work-around in this situation is
10360 to use a convenience variable (@pxref{Convenience Vars, ,Convenience
10361 Variables}) as a counter in an expression that prints the first
10362 interesting value, and then repeat that expression via @key{RET}. For
10363 instance, suppose you have an array @code{dtab} of pointers to
10364 structures, and you are interested in the values of a field @code{fv}
10365 in each structure. Here is an example of what you might type:
10375 @node Output Formats
10376 @section Output Formats
10378 @cindex formatted output
10379 @cindex output formats
10380 By default, @value{GDBN} prints a value according to its data type. Sometimes
10381 this is not what you want. For example, you might want to print a number
10382 in hex, or a pointer in decimal. Or you might want to view data in memory
10383 at a certain address as a character string or as an instruction. To do
10384 these things, specify an @dfn{output format} when you print a value.
10386 The simplest use of output formats is to say how to print a value
10387 already computed. This is done by starting the arguments of the
10388 @code{print} command with a slash and a format letter. The format
10389 letters supported are:
10393 Regard the bits of the value as an integer, and print the integer in
10397 Print as integer in signed decimal.
10400 Print as integer in unsigned decimal.
10403 Print as integer in octal.
10406 Print as integer in binary. The letter @samp{t} stands for ``two''.
10407 @footnote{@samp{b} cannot be used because these format letters are also
10408 used with the @code{x} command, where @samp{b} stands for ``byte'';
10409 see @ref{Memory,,Examining Memory}.}
10412 @cindex unknown address, locating
10413 @cindex locate address
10414 Print as an address, both absolute in hexadecimal and as an offset from
10415 the nearest preceding symbol. You can use this format used to discover
10416 where (in what function) an unknown address is located:
10419 (@value{GDBP}) p/a 0x54320
10420 $3 = 0x54320 <_initialize_vx+396>
10424 The command @code{info symbol 0x54320} yields similar results.
10425 @xref{Symbols, info symbol}.
10428 Regard as an integer and print it as a character constant. This
10429 prints both the numerical value and its character representation. The
10430 character representation is replaced with the octal escape @samp{\nnn}
10431 for characters outside the 7-bit @sc{ascii} range.
10433 Without this format, @value{GDBN} displays @code{char},
10434 @w{@code{unsigned char}}, and @w{@code{signed char}} data as character
10435 constants. Single-byte members of vectors are displayed as integer
10439 Regard the bits of the value as a floating point number and print
10440 using typical floating point syntax.
10443 @cindex printing strings
10444 @cindex printing byte arrays
10445 Regard as a string, if possible. With this format, pointers to single-byte
10446 data are displayed as null-terminated strings and arrays of single-byte data
10447 are displayed as fixed-length strings. Other values are displayed in their
10450 Without this format, @value{GDBN} displays pointers to and arrays of
10451 @code{char}, @w{@code{unsigned char}}, and @w{@code{signed char}} as
10452 strings. Single-byte members of a vector are displayed as an integer
10456 Like @samp{x} formatting, the value is treated as an integer and
10457 printed as hexadecimal, but leading zeros are printed to pad the value
10458 to the size of the integer type.
10461 @cindex raw printing
10462 Print using the @samp{raw} formatting. By default, @value{GDBN} will
10463 use a Python-based pretty-printer, if one is available (@pxref{Pretty
10464 Printing}). This typically results in a higher-level display of the
10465 value's contents. The @samp{r} format bypasses any Python
10466 pretty-printer which might exist.
10469 For example, to print the program counter in hex (@pxref{Registers}), type
10476 Note that no space is required before the slash; this is because command
10477 names in @value{GDBN} cannot contain a slash.
10479 To reprint the last value in the value history with a different format,
10480 you can use the @code{print} command with just a format and no
10481 expression. For example, @samp{p/x} reprints the last value in hex.
10484 @section Examining Memory
10486 You can use the command @code{x} (for ``examine'') to examine memory in
10487 any of several formats, independently of your program's data types.
10489 @cindex examining memory
10491 @kindex x @r{(examine memory)}
10492 @item x/@var{nfu} @var{addr}
10493 @itemx x @var{addr}
10495 Use the @code{x} command to examine memory.
10498 @var{n}, @var{f}, and @var{u} are all optional parameters that specify how
10499 much memory to display and how to format it; @var{addr} is an
10500 expression giving the address where you want to start displaying memory.
10501 If you use defaults for @var{nfu}, you need not type the slash @samp{/}.
10502 Several commands set convenient defaults for @var{addr}.
10505 @item @var{n}, the repeat count
10506 The repeat count is a decimal integer; the default is 1. It specifies
10507 how much memory (counting by units @var{u}) to display. If a negative
10508 number is specified, memory is examined backward from @var{addr}.
10509 @c This really is **decimal**; unaffected by 'set radix' as of GDB
10512 @item @var{f}, the display format
10513 The display format is one of the formats used by @code{print}
10514 (@samp{x}, @samp{d}, @samp{u}, @samp{o}, @samp{t}, @samp{a}, @samp{c},
10515 @samp{f}, @samp{s}), and in addition @samp{i} (for machine instructions).
10516 The default is @samp{x} (hexadecimal) initially. The default changes
10517 each time you use either @code{x} or @code{print}.
10519 @item @var{u}, the unit size
10520 The unit size is any of
10526 Halfwords (two bytes).
10528 Words (four bytes). This is the initial default.
10530 Giant words (eight bytes).
10533 Each time you specify a unit size with @code{x}, that size becomes the
10534 default unit the next time you use @code{x}. For the @samp{i} format,
10535 the unit size is ignored and is normally not written. For the @samp{s} format,
10536 the unit size defaults to @samp{b}, unless it is explicitly given.
10537 Use @kbd{x /hs} to display 16-bit char strings and @kbd{x /ws} to display
10538 32-bit strings. The next use of @kbd{x /s} will again display 8-bit strings.
10539 Note that the results depend on the programming language of the
10540 current compilation unit. If the language is C, the @samp{s}
10541 modifier will use the UTF-16 encoding while @samp{w} will use
10542 UTF-32. The encoding is set by the programming language and cannot
10545 @item @var{addr}, starting display address
10546 @var{addr} is the address where you want @value{GDBN} to begin displaying
10547 memory. The expression need not have a pointer value (though it may);
10548 it is always interpreted as an integer address of a byte of memory.
10549 @xref{Expressions, ,Expressions}, for more information on expressions. The default for
10550 @var{addr} is usually just after the last address examined---but several
10551 other commands also set the default address: @code{info breakpoints} (to
10552 the address of the last breakpoint listed), @code{info line} (to the
10553 starting address of a line), and @code{print} (if you use it to display
10554 a value from memory).
10557 For example, @samp{x/3uh 0x54320} is a request to display three halfwords
10558 (@code{h}) of memory, formatted as unsigned decimal integers (@samp{u}),
10559 starting at address @code{0x54320}. @samp{x/4xw $sp} prints the four
10560 words (@samp{w}) of memory above the stack pointer (here, @samp{$sp};
10561 @pxref{Registers, ,Registers}) in hexadecimal (@samp{x}).
10563 You can also specify a negative repeat count to examine memory backward
10564 from the given address. For example, @samp{x/-3uh 0x54320} prints three
10565 halfwords (@code{h}) at @code{0x54314}, @code{0x54328}, and @code{0x5431c}.
10567 Since the letters indicating unit sizes are all distinct from the
10568 letters specifying output formats, you do not have to remember whether
10569 unit size or format comes first; either order works. The output
10570 specifications @samp{4xw} and @samp{4wx} mean exactly the same thing.
10571 (However, the count @var{n} must come first; @samp{wx4} does not work.)
10573 Even though the unit size @var{u} is ignored for the formats @samp{s}
10574 and @samp{i}, you might still want to use a count @var{n}; for example,
10575 @samp{3i} specifies that you want to see three machine instructions,
10576 including any operands. For convenience, especially when used with
10577 the @code{display} command, the @samp{i} format also prints branch delay
10578 slot instructions, if any, beyond the count specified, which immediately
10579 follow the last instruction that is within the count. The command
10580 @code{disassemble} gives an alternative way of inspecting machine
10581 instructions; see @ref{Machine Code,,Source and Machine Code}.
10583 If a negative repeat count is specified for the formats @samp{s} or @samp{i},
10584 the command displays null-terminated strings or instructions before the given
10585 address as many as the absolute value of the given number. For the @samp{i}
10586 format, we use line number information in the debug info to accurately locate
10587 instruction boundaries while disassembling backward. If line info is not
10588 available, the command stops examining memory with an error message.
10590 All the defaults for the arguments to @code{x} are designed to make it
10591 easy to continue scanning memory with minimal specifications each time
10592 you use @code{x}. For example, after you have inspected three machine
10593 instructions with @samp{x/3i @var{addr}}, you can inspect the next seven
10594 with just @samp{x/7}. If you use @key{RET} to repeat the @code{x} command,
10595 the repeat count @var{n} is used again; the other arguments default as
10596 for successive uses of @code{x}.
10598 When examining machine instructions, the instruction at current program
10599 counter is shown with a @code{=>} marker. For example:
10602 (@value{GDBP}) x/5i $pc-6
10603 0x804837f <main+11>: mov %esp,%ebp
10604 0x8048381 <main+13>: push %ecx
10605 0x8048382 <main+14>: sub $0x4,%esp
10606 => 0x8048385 <main+17>: movl $0x8048460,(%esp)
10607 0x804838c <main+24>: call 0x80482d4 <puts@@plt>
10610 @cindex @code{$_}, @code{$__}, and value history
10611 The addresses and contents printed by the @code{x} command are not saved
10612 in the value history because there is often too much of them and they
10613 would get in the way. Instead, @value{GDBN} makes these values available for
10614 subsequent use in expressions as values of the convenience variables
10615 @code{$_} and @code{$__}. After an @code{x} command, the last address
10616 examined is available for use in expressions in the convenience variable
10617 @code{$_}. The contents of that address, as examined, are available in
10618 the convenience variable @code{$__}.
10620 If the @code{x} command has a repeat count, the address and contents saved
10621 are from the last memory unit printed; this is not the same as the last
10622 address printed if several units were printed on the last line of output.
10624 @anchor{addressable memory unit}
10625 @cindex addressable memory unit
10626 Most targets have an addressable memory unit size of 8 bits. This means
10627 that to each memory address are associated 8 bits of data. Some
10628 targets, however, have other addressable memory unit sizes.
10629 Within @value{GDBN} and this document, the term
10630 @dfn{addressable memory unit} (or @dfn{memory unit} for short) is used
10631 when explicitly referring to a chunk of data of that size. The word
10632 @dfn{byte} is used to refer to a chunk of data of 8 bits, regardless of
10633 the addressable memory unit size of the target. For most systems,
10634 addressable memory unit is a synonym of byte.
10636 @cindex remote memory comparison
10637 @cindex target memory comparison
10638 @cindex verify remote memory image
10639 @cindex verify target memory image
10640 When you are debugging a program running on a remote target machine
10641 (@pxref{Remote Debugging}), you may wish to verify the program's image
10642 in the remote machine's memory against the executable file you
10643 downloaded to the target. Or, on any target, you may want to check
10644 whether the program has corrupted its own read-only sections. The
10645 @code{compare-sections} command is provided for such situations.
10648 @kindex compare-sections
10649 @item compare-sections @r{[}@var{section-name}@r{|}@code{-r}@r{]}
10650 Compare the data of a loadable section @var{section-name} in the
10651 executable file of the program being debugged with the same section in
10652 the target machine's memory, and report any mismatches. With no
10653 arguments, compares all loadable sections. With an argument of
10654 @code{-r}, compares all loadable read-only sections.
10656 Note: for remote targets, this command can be accelerated if the
10657 target supports computing the CRC checksum of a block of memory
10658 (@pxref{qCRC packet}).
10662 @section Automatic Display
10663 @cindex automatic display
10664 @cindex display of expressions
10666 If you find that you want to print the value of an expression frequently
10667 (to see how it changes), you might want to add it to the @dfn{automatic
10668 display list} so that @value{GDBN} prints its value each time your program stops.
10669 Each expression added to the list is given a number to identify it;
10670 to remove an expression from the list, you specify that number.
10671 The automatic display looks like this:
10675 3: bar[5] = (struct hack *) 0x3804
10679 This display shows item numbers, expressions and their current values. As with
10680 displays you request manually using @code{x} or @code{print}, you can
10681 specify the output format you prefer; in fact, @code{display} decides
10682 whether to use @code{print} or @code{x} depending your format
10683 specification---it uses @code{x} if you specify either the @samp{i}
10684 or @samp{s} format, or a unit size; otherwise it uses @code{print}.
10688 @item display @var{expr}
10689 Add the expression @var{expr} to the list of expressions to display
10690 each time your program stops. @xref{Expressions, ,Expressions}.
10692 @code{display} does not repeat if you press @key{RET} again after using it.
10694 @item display/@var{fmt} @var{expr}
10695 For @var{fmt} specifying only a display format and not a size or
10696 count, add the expression @var{expr} to the auto-display list but
10697 arrange to display it each time in the specified format @var{fmt}.
10698 @xref{Output Formats,,Output Formats}.
10700 @item display/@var{fmt} @var{addr}
10701 For @var{fmt} @samp{i} or @samp{s}, or including a unit-size or a
10702 number of units, add the expression @var{addr} as a memory address to
10703 be examined each time your program stops. Examining means in effect
10704 doing @samp{x/@var{fmt} @var{addr}}. @xref{Memory, ,Examining Memory}.
10707 For example, @samp{display/i $pc} can be helpful, to see the machine
10708 instruction about to be executed each time execution stops (@samp{$pc}
10709 is a common name for the program counter; @pxref{Registers, ,Registers}).
10712 @kindex delete display
10714 @item undisplay @var{dnums}@dots{}
10715 @itemx delete display @var{dnums}@dots{}
10716 Remove items from the list of expressions to display. Specify the
10717 numbers of the displays that you want affected with the command
10718 argument @var{dnums}. It can be a single display number, one of the
10719 numbers shown in the first field of the @samp{info display} display;
10720 or it could be a range of display numbers, as in @code{2-4}.
10722 @code{undisplay} does not repeat if you press @key{RET} after using it.
10723 (Otherwise you would just get the error @samp{No display number @dots{}}.)
10725 @kindex disable display
10726 @item disable display @var{dnums}@dots{}
10727 Disable the display of item numbers @var{dnums}. A disabled display
10728 item is not printed automatically, but is not forgotten. It may be
10729 enabled again later. Specify the numbers of the displays that you
10730 want affected with the command argument @var{dnums}. It can be a
10731 single display number, one of the numbers shown in the first field of
10732 the @samp{info display} display; or it could be a range of display
10733 numbers, as in @code{2-4}.
10735 @kindex enable display
10736 @item enable display @var{dnums}@dots{}
10737 Enable display of item numbers @var{dnums}. It becomes effective once
10738 again in auto display of its expression, until you specify otherwise.
10739 Specify the numbers of the displays that you want affected with the
10740 command argument @var{dnums}. It can be a single display number, one
10741 of the numbers shown in the first field of the @samp{info display}
10742 display; or it could be a range of display numbers, as in @code{2-4}.
10745 Display the current values of the expressions on the list, just as is
10746 done when your program stops.
10748 @kindex info display
10750 Print the list of expressions previously set up to display
10751 automatically, each one with its item number, but without showing the
10752 values. This includes disabled expressions, which are marked as such.
10753 It also includes expressions which would not be displayed right now
10754 because they refer to automatic variables not currently available.
10757 @cindex display disabled out of scope
10758 If a display expression refers to local variables, then it does not make
10759 sense outside the lexical context for which it was set up. Such an
10760 expression is disabled when execution enters a context where one of its
10761 variables is not defined. For example, if you give the command
10762 @code{display last_char} while inside a function with an argument
10763 @code{last_char}, @value{GDBN} displays this argument while your program
10764 continues to stop inside that function. When it stops elsewhere---where
10765 there is no variable @code{last_char}---the display is disabled
10766 automatically. The next time your program stops where @code{last_char}
10767 is meaningful, you can enable the display expression once again.
10769 @node Print Settings
10770 @section Print Settings
10772 @cindex format options
10773 @cindex print settings
10774 @value{GDBN} provides the following ways to control how arrays, structures,
10775 and symbols are printed.
10778 These settings are useful for debugging programs in any language:
10782 @anchor{set print address}
10783 @item set print address
10784 @itemx set print address on
10785 @cindex print/don't print memory addresses
10786 @value{GDBN} prints memory addresses showing the location of stack
10787 traces, structure values, pointer values, breakpoints, and so forth,
10788 even when it also displays the contents of those addresses. The default
10789 is @code{on}. For example, this is what a stack frame display looks like with
10790 @code{set print address on}:
10795 #0 set_quotes (lq=0x34c78 "<<", rq=0x34c88 ">>")
10797 530 if (lquote != def_lquote)
10801 @item set print address off
10802 Do not print addresses when displaying their contents. For example,
10803 this is the same stack frame displayed with @code{set print address off}:
10807 (@value{GDBP}) set print addr off
10809 #0 set_quotes (lq="<<", rq=">>") at input.c:530
10810 530 if (lquote != def_lquote)
10814 You can use @samp{set print address off} to eliminate all machine
10815 dependent displays from the @value{GDBN} interface. For example, with
10816 @code{print address off}, you should get the same text for backtraces on
10817 all machines---whether or not they involve pointer arguments.
10820 @item show print address
10821 Show whether or not addresses are to be printed.
10824 When @value{GDBN} prints a symbolic address, it normally prints the
10825 closest earlier symbol plus an offset. If that symbol does not uniquely
10826 identify the address (for example, it is a name whose scope is a single
10827 source file), you may need to clarify. One way to do this is with
10828 @code{info line}, for example @samp{info line *0x4537}. Alternately,
10829 you can set @value{GDBN} to print the source file and line number when
10830 it prints a symbolic address:
10833 @item set print symbol-filename on
10834 @cindex source file and line of a symbol
10835 @cindex symbol, source file and line
10836 Tell @value{GDBN} to print the source file name and line number of a
10837 symbol in the symbolic form of an address.
10839 @item set print symbol-filename off
10840 Do not print source file name and line number of a symbol. This is the
10843 @item show print symbol-filename
10844 Show whether or not @value{GDBN} will print the source file name and
10845 line number of a symbol in the symbolic form of an address.
10848 Another situation where it is helpful to show symbol filenames and line
10849 numbers is when disassembling code; @value{GDBN} shows you the line
10850 number and source file that corresponds to each instruction.
10852 Also, you may wish to see the symbolic form only if the address being
10853 printed is reasonably close to the closest earlier symbol:
10856 @item set print max-symbolic-offset @var{max-offset}
10857 @itemx set print max-symbolic-offset unlimited
10858 @cindex maximum value for offset of closest symbol
10859 Tell @value{GDBN} to only display the symbolic form of an address if the
10860 offset between the closest earlier symbol and the address is less than
10861 @var{max-offset}. The default is @code{unlimited}, which tells @value{GDBN}
10862 to always print the symbolic form of an address if any symbol precedes
10863 it. Zero is equivalent to @code{unlimited}.
10865 @item show print max-symbolic-offset
10866 Ask how large the maximum offset is that @value{GDBN} prints in a
10870 @cindex wild pointer, interpreting
10871 @cindex pointer, finding referent
10872 If you have a pointer and you are not sure where it points, try
10873 @samp{set print symbol-filename on}. Then you can determine the name
10874 and source file location of the variable where it points, using
10875 @samp{p/a @var{pointer}}. This interprets the address in symbolic form.
10876 For example, here @value{GDBN} shows that a variable @code{ptt} points
10877 at another variable @code{t}, defined in @file{hi2.c}:
10880 (@value{GDBP}) set print symbol-filename on
10881 (@value{GDBP}) p/a ptt
10882 $4 = 0xe008 <t in hi2.c>
10886 @emph{Warning:} For pointers that point to a local variable, @samp{p/a}
10887 does not show the symbol name and filename of the referent, even with
10888 the appropriate @code{set print} options turned on.
10891 You can also enable @samp{/a}-like formatting all the time using
10892 @samp{set print symbol on}:
10894 @anchor{set print symbol}
10896 @item set print symbol on
10897 Tell @value{GDBN} to print the symbol corresponding to an address, if
10900 @item set print symbol off
10901 Tell @value{GDBN} not to print the symbol corresponding to an
10902 address. In this mode, @value{GDBN} will still print the symbol
10903 corresponding to pointers to functions. This is the default.
10905 @item show print symbol
10906 Show whether @value{GDBN} will display the symbol corresponding to an
10910 Other settings control how different kinds of objects are printed:
10913 @anchor{set print array}
10914 @item set print array
10915 @itemx set print array on
10916 @cindex pretty print arrays
10917 Pretty print arrays. This format is more convenient to read,
10918 but uses more space. The default is off.
10920 @item set print array off
10921 Return to compressed format for arrays.
10923 @item show print array
10924 Show whether compressed or pretty format is selected for displaying
10927 @cindex print array indexes
10928 @anchor{set print array-indexes}
10929 @item set print array-indexes
10930 @itemx set print array-indexes on
10931 Print the index of each element when displaying arrays. May be more
10932 convenient to locate a given element in the array or quickly find the
10933 index of a given element in that printed array. The default is off.
10935 @item set print array-indexes off
10936 Stop printing element indexes when displaying arrays.
10938 @item show print array-indexes
10939 Show whether the index of each element is printed when displaying
10942 @anchor{set print elements}
10943 @item set print elements @var{number-of-elements}
10944 @itemx set print elements unlimited
10945 @cindex number of array elements to print
10946 @cindex limit on number of printed array elements
10947 Set a limit on how many elements of an array @value{GDBN} will print.
10948 If @value{GDBN} is printing a large array, it stops printing after it has
10949 printed the number of elements set by the @code{set print elements} command.
10950 This limit also applies to the display of strings.
10951 When @value{GDBN} starts, this limit is set to 200.
10952 Setting @var{number-of-elements} to @code{unlimited} or zero means
10953 that the number of elements to print is unlimited.
10955 @item show print elements
10956 Display the number of elements of a large array that @value{GDBN} will print.
10957 If the number is 0, then the printing is unlimited.
10959 @anchor{set print frame-arguments}
10960 @item set print frame-arguments @var{value}
10961 @kindex set print frame-arguments
10962 @cindex printing frame argument values
10963 @cindex print all frame argument values
10964 @cindex print frame argument values for scalars only
10965 @cindex do not print frame arguments
10966 This command allows to control how the values of arguments are printed
10967 when the debugger prints a frame (@pxref{Frames}). The possible
10972 The values of all arguments are printed.
10975 Print the value of an argument only if it is a scalar. The value of more
10976 complex arguments such as arrays, structures, unions, etc, is replaced
10977 by @code{@dots{}}. This is the default. Here is an example where
10978 only scalar arguments are shown:
10981 #1 0x08048361 in call_me (i=3, s=@dots{}, ss=0xbf8d508c, u=@dots{}, e=green)
10986 None of the argument values are printed. Instead, the value of each argument
10987 is replaced by @code{@dots{}}. In this case, the example above now becomes:
10990 #1 0x08048361 in call_me (i=@dots{}, s=@dots{}, ss=@dots{}, u=@dots{}, e=@dots{})
10995 Only the presence of arguments is indicated by @code{@dots{}}.
10996 The @code{@dots{}} are not printed for function without any arguments.
10997 None of the argument names and values are printed.
10998 In this case, the example above now becomes:
11001 #1 0x08048361 in call_me (@dots{}) at frame-args.c:23
11006 By default, only scalar arguments are printed. This command can be used
11007 to configure the debugger to print the value of all arguments, regardless
11008 of their type. However, it is often advantageous to not print the value
11009 of more complex parameters. For instance, it reduces the amount of
11010 information printed in each frame, making the backtrace more readable.
11011 Also, it improves performance when displaying Ada frames, because
11012 the computation of large arguments can sometimes be CPU-intensive,
11013 especially in large applications. Setting @code{print frame-arguments}
11014 to @code{scalars} (the default), @code{none} or @code{presence} avoids
11015 this computation, thus speeding up the display of each Ada frame.
11017 @item show print frame-arguments
11018 Show how the value of arguments should be displayed when printing a frame.
11020 @anchor{set print raw-frame-arguments}
11021 @item set print raw-frame-arguments on
11022 Print frame arguments in raw, non pretty-printed, form.
11024 @item set print raw-frame-arguments off
11025 Print frame arguments in pretty-printed form, if there is a pretty-printer
11026 for the value (@pxref{Pretty Printing}),
11027 otherwise print the value in raw form.
11028 This is the default.
11030 @item show print raw-frame-arguments
11031 Show whether to print frame arguments in raw form.
11033 @anchor{set print entry-values}
11034 @item set print entry-values @var{value}
11035 @kindex set print entry-values
11036 Set printing of frame argument values at function entry. In some cases
11037 @value{GDBN} can determine the value of function argument which was passed by
11038 the function caller, even if the value was modified inside the called function
11039 and therefore is different. With optimized code, the current value could be
11040 unavailable, but the entry value may still be known.
11042 The default value is @code{default} (see below for its description). Older
11043 @value{GDBN} behaved as with the setting @code{no}. Compilers not supporting
11044 this feature will behave in the @code{default} setting the same way as with the
11047 This functionality is currently supported only by DWARF 2 debugging format and
11048 the compiler has to produce @samp{DW_TAG_call_site} tags. With
11049 @value{NGCC}, you need to specify @option{-O -g} during compilation, to get
11052 The @var{value} parameter can be one of the following:
11056 Print only actual parameter values, never print values from function entry
11060 #0 different (val=6)
11061 #0 lost (val=<optimized out>)
11063 #0 invalid (val=<optimized out>)
11067 Print only parameter values from function entry point. The actual parameter
11068 values are never printed.
11070 #0 equal (val@@entry=5)
11071 #0 different (val@@entry=5)
11072 #0 lost (val@@entry=5)
11073 #0 born (val@@entry=<optimized out>)
11074 #0 invalid (val@@entry=<optimized out>)
11078 Print only parameter values from function entry point. If value from function
11079 entry point is not known while the actual value is known, print the actual
11080 value for such parameter.
11082 #0 equal (val@@entry=5)
11083 #0 different (val@@entry=5)
11084 #0 lost (val@@entry=5)
11086 #0 invalid (val@@entry=<optimized out>)
11090 Print actual parameter values. If actual parameter value is not known while
11091 value from function entry point is known, print the entry point value for such
11095 #0 different (val=6)
11096 #0 lost (val@@entry=5)
11098 #0 invalid (val=<optimized out>)
11102 Always print both the actual parameter value and its value from function entry
11103 point, even if values of one or both are not available due to compiler
11106 #0 equal (val=5, val@@entry=5)
11107 #0 different (val=6, val@@entry=5)
11108 #0 lost (val=<optimized out>, val@@entry=5)
11109 #0 born (val=10, val@@entry=<optimized out>)
11110 #0 invalid (val=<optimized out>, val@@entry=<optimized out>)
11114 Print the actual parameter value if it is known and also its value from
11115 function entry point if it is known. If neither is known, print for the actual
11116 value @code{<optimized out>}. If not in MI mode (@pxref{GDB/MI}) and if both
11117 values are known and identical, print the shortened
11118 @code{param=param@@entry=VALUE} notation.
11120 #0 equal (val=val@@entry=5)
11121 #0 different (val=6, val@@entry=5)
11122 #0 lost (val@@entry=5)
11124 #0 invalid (val=<optimized out>)
11128 Always print the actual parameter value. Print also its value from function
11129 entry point, but only if it is known. If not in MI mode (@pxref{GDB/MI}) and
11130 if both values are known and identical, print the shortened
11131 @code{param=param@@entry=VALUE} notation.
11133 #0 equal (val=val@@entry=5)
11134 #0 different (val=6, val@@entry=5)
11135 #0 lost (val=<optimized out>, val@@entry=5)
11137 #0 invalid (val=<optimized out>)
11141 For analysis messages on possible failures of frame argument values at function
11142 entry resolution see @ref{set debug entry-values}.
11144 @item show print entry-values
11145 Show the method being used for printing of frame argument values at function
11148 @anchor{set print frame-info}
11149 @item set print frame-info @var{value}
11150 @kindex set print frame-info
11151 @cindex printing frame information
11152 @cindex frame information, printing
11153 This command allows to control the information printed when
11154 the debugger prints a frame. See @ref{Frames}, @ref{Backtrace},
11155 for a general explanation about frames and frame information.
11156 Note that some other settings (such as @code{set print frame-arguments}
11157 and @code{set print address}) are also influencing if and how some frame
11158 information is displayed. In particular, the frame program counter is never
11159 printed if @code{set print address} is off.
11161 The possible values for @code{set print frame-info} are:
11163 @item short-location
11164 Print the frame level, the program counter (if not at the
11165 beginning of the location source line), the function, the function
11168 Same as @code{short-location} but also print the source file and source line
11170 @item location-and-address
11171 Same as @code{location} but print the program counter even if located at the
11172 beginning of the location source line.
11174 Print the program counter (if not at the beginning of the location
11175 source line), the line number and the source line.
11176 @item source-and-location
11177 Print what @code{location} and @code{source-line} are printing.
11179 The information printed for a frame is decided automatically
11180 by the @value{GDBN} command that prints a frame.
11181 For example, @code{frame} prints the information printed by
11182 @code{source-and-location} while @code{stepi} will switch between
11183 @code{source-line} and @code{source-and-location} depending on the program
11185 The default value is @code{auto}.
11188 @anchor{set print repeats}
11189 @item set print repeats @var{number-of-repeats}
11190 @itemx set print repeats unlimited
11191 @cindex repeated array elements
11192 Set the threshold for suppressing display of repeated array
11193 elements. When the number of consecutive identical elements of an
11194 array exceeds the threshold, @value{GDBN} prints the string
11195 @code{"<repeats @var{n} times>"}, where @var{n} is the number of
11196 identical repetitions, instead of displaying the identical elements
11197 themselves. Setting the threshold to @code{unlimited} or zero will
11198 cause all elements to be individually printed. The default threshold
11201 @item show print repeats
11202 Display the current threshold for printing repeated identical
11205 @anchor{set print max-depth}
11206 @item set print max-depth @var{depth}
11207 @item set print max-depth unlimited
11208 @cindex printing nested structures
11209 Set the threshold after which nested structures are replaced with
11210 ellipsis, this can make visualising deeply nested structures easier.
11212 For example, given this C code
11215 typedef struct s1 @{ int a; @} s1;
11216 typedef struct s2 @{ s1 b; @} s2;
11217 typedef struct s3 @{ s2 c; @} s3;
11218 typedef struct s4 @{ s3 d; @} s4;
11220 s4 var = @{ @{ @{ @{ 3 @} @} @} @};
11223 The following table shows how different values of @var{depth} will
11224 effect how @code{var} is printed by @value{GDBN}:
11226 @multitable @columnfractions .3 .7
11227 @headitem @var{depth} setting @tab Result of @samp{p var}
11229 @tab @code{$1 = @{d = @{c = @{b = @{a = 3@}@}@}@}}
11231 @tab @code{$1 = @{...@}}
11233 @tab @code{$1 = @{d = @{...@}@}}
11235 @tab @code{$1 = @{d = @{c = @{...@}@}@}}
11237 @tab @code{$1 = @{d = @{c = @{b = @{...@}@}@}@}}
11239 @tab @code{$1 = @{d = @{c = @{b = @{a = 3@}@}@}@}}
11242 To see the contents of structures that have been hidden the user can
11243 either increase the print max-depth, or they can print the elements of
11244 the structure that are visible, for example
11247 (@value{GDBP}) set print max-depth 2
11248 (@value{GDBP}) p var
11249 $1 = @{d = @{c = @{...@}@}@}
11250 (@value{GDBP}) p var.d
11251 $2 = @{c = @{b = @{...@}@}@}
11252 (@value{GDBP}) p var.d.c
11253 $3 = @{b = @{a = 3@}@}
11256 The pattern used to replace nested structures varies based on
11257 language, for most languages @code{@{...@}} is used, but Fortran uses
11260 @item show print max-depth
11261 Display the current threshold after which nested structures are
11262 replaces with ellipsis.
11264 @anchor{set print null-stop}
11265 @item set print null-stop
11266 @cindex @sc{null} elements in arrays
11267 Cause @value{GDBN} to stop printing the characters of an array when the first
11268 @sc{null} is encountered. This is useful when large arrays actually
11269 contain only short strings.
11270 The default is off.
11272 @item show print null-stop
11273 Show whether @value{GDBN} stops printing an array on the first
11274 @sc{null} character.
11276 @anchor{set print pretty}
11277 @item set print pretty on
11278 @cindex print structures in indented form
11279 @cindex indentation in structure display
11280 Cause @value{GDBN} to print structures in an indented format with one member
11281 per line, like this:
11296 @item set print pretty off
11297 Cause @value{GDBN} to print structures in a compact format, like this:
11301 $1 = @{next = 0x0, flags = @{sweet = 1, sour = 1@}, \
11302 meat = 0x54 "Pork"@}
11307 This is the default format.
11309 @item show print pretty
11310 Show which format @value{GDBN} is using to print structures.
11312 @anchor{set print raw-values}
11313 @item set print raw-values on
11314 Print values in raw form, without applying the pretty
11315 printers for the value.
11317 @item set print raw-values off
11318 Print values in pretty-printed form, if there is a pretty-printer
11319 for the value (@pxref{Pretty Printing}),
11320 otherwise print the value in raw form.
11322 The default setting is ``off''.
11324 @item show print raw-values
11325 Show whether to print values in raw form.
11327 @item set print sevenbit-strings on
11328 @cindex eight-bit characters in strings
11329 @cindex octal escapes in strings
11330 Print using only seven-bit characters; if this option is set,
11331 @value{GDBN} displays any eight-bit characters (in strings or
11332 character values) using the notation @code{\}@var{nnn}. This setting is
11333 best if you are working in English (@sc{ascii}) and you use the
11334 high-order bit of characters as a marker or ``meta'' bit.
11336 @item set print sevenbit-strings off
11337 Print full eight-bit characters. This allows the use of more
11338 international character sets, and is the default.
11340 @item show print sevenbit-strings
11341 Show whether or not @value{GDBN} is printing only seven-bit characters.
11343 @anchor{set print union}
11344 @item set print union on
11345 @cindex unions in structures, printing
11346 Tell @value{GDBN} to print unions which are contained in structures
11347 and other unions. This is the default setting.
11349 @item set print union off
11350 Tell @value{GDBN} not to print unions which are contained in
11351 structures and other unions. @value{GDBN} will print @code{"@{...@}"}
11354 @item show print union
11355 Ask @value{GDBN} whether or not it will print unions which are contained in
11356 structures and other unions.
11358 For example, given the declarations
11361 typedef enum @{Tree, Bug@} Species;
11362 typedef enum @{Big_tree, Acorn, Seedling@} Tree_forms;
11363 typedef enum @{Caterpillar, Cocoon, Butterfly@}
11374 struct thing foo = @{Tree, @{Acorn@}@};
11378 with @code{set print union on} in effect @samp{p foo} would print
11381 $1 = @{it = Tree, form = @{tree = Acorn, bug = Cocoon@}@}
11385 and with @code{set print union off} in effect it would print
11388 $1 = @{it = Tree, form = @{...@}@}
11392 @code{set print union} affects programs written in C-like languages
11398 These settings are of interest when debugging C@t{++} programs:
11401 @cindex demangling C@t{++} names
11402 @item set print demangle
11403 @itemx set print demangle on
11404 Print C@t{++} names in their source form rather than in the encoded
11405 (``mangled'') form passed to the assembler and linker for type-safe
11406 linkage. The default is on.
11408 @item show print demangle
11409 Show whether C@t{++} names are printed in mangled or demangled form.
11411 @item set print asm-demangle
11412 @itemx set print asm-demangle on
11413 Print C@t{++} names in their source form rather than their mangled form, even
11414 in assembler code printouts such as instruction disassemblies.
11415 The default is off.
11417 @item show print asm-demangle
11418 Show whether C@t{++} names in assembly listings are printed in mangled
11421 @cindex C@t{++} symbol decoding style
11422 @cindex symbol decoding style, C@t{++}
11423 @kindex set demangle-style
11424 @item set demangle-style @var{style}
11425 Choose among several encoding schemes used by different compilers to represent
11426 C@t{++} names. If you omit @var{style}, you will see a list of possible
11427 formats. The default value is @var{auto}, which lets @value{GDBN} choose a
11428 decoding style by inspecting your program.
11430 @item show demangle-style
11431 Display the encoding style currently in use for decoding C@t{++} symbols.
11433 @anchor{set print object}
11434 @item set print object
11435 @itemx set print object on
11436 @cindex derived type of an object, printing
11437 @cindex display derived types
11438 When displaying a pointer to an object, identify the @emph{actual}
11439 (derived) type of the object rather than the @emph{declared} type, using
11440 the virtual function table. Note that the virtual function table is
11441 required---this feature can only work for objects that have run-time
11442 type identification; a single virtual method in the object's declared
11443 type is sufficient. Note that this setting is also taken into account when
11444 working with variable objects via MI (@pxref{GDB/MI}).
11446 @item set print object off
11447 Display only the declared type of objects, without reference to the
11448 virtual function table. This is the default setting.
11450 @item show print object
11451 Show whether actual, or declared, object types are displayed.
11453 @anchor{set print static-members}
11454 @item set print static-members
11455 @itemx set print static-members on
11456 @cindex static members of C@t{++} objects
11457 Print static members when displaying a C@t{++} object. The default is on.
11459 @item set print static-members off
11460 Do not print static members when displaying a C@t{++} object.
11462 @item show print static-members
11463 Show whether C@t{++} static members are printed or not.
11465 @item set print pascal_static-members
11466 @itemx set print pascal_static-members on
11467 @cindex static members of Pascal objects
11468 @cindex Pascal objects, static members display
11469 Print static members when displaying a Pascal object. The default is on.
11471 @item set print pascal_static-members off
11472 Do not print static members when displaying a Pascal object.
11474 @item show print pascal_static-members
11475 Show whether Pascal static members are printed or not.
11477 @c These don't work with HP ANSI C++ yet.
11478 @anchor{set print vtbl}
11479 @item set print vtbl
11480 @itemx set print vtbl on
11481 @cindex pretty print C@t{++} virtual function tables
11482 @cindex virtual functions (C@t{++}) display
11483 @cindex VTBL display
11484 Pretty print C@t{++} virtual function tables. The default is off.
11485 (The @code{vtbl} commands do not work on programs compiled with the HP
11486 ANSI C@t{++} compiler (@code{aCC}).)
11488 @item set print vtbl off
11489 Do not pretty print C@t{++} virtual function tables.
11491 @item show print vtbl
11492 Show whether C@t{++} virtual function tables are pretty printed, or not.
11495 @node Pretty Printing
11496 @section Pretty Printing
11498 @value{GDBN} provides a mechanism to allow pretty-printing of values using
11499 Python code. It greatly simplifies the display of complex objects. This
11500 mechanism works for both MI and the CLI.
11503 * Pretty-Printer Introduction:: Introduction to pretty-printers
11504 * Pretty-Printer Example:: An example pretty-printer
11505 * Pretty-Printer Commands:: Pretty-printer commands
11508 @node Pretty-Printer Introduction
11509 @subsection Pretty-Printer Introduction
11511 When @value{GDBN} prints a value, it first sees if there is a pretty-printer
11512 registered for the value. If there is then @value{GDBN} invokes the
11513 pretty-printer to print the value. Otherwise the value is printed normally.
11515 Pretty-printers are normally named. This makes them easy to manage.
11516 The @samp{info pretty-printer} command will list all the installed
11517 pretty-printers with their names.
11518 If a pretty-printer can handle multiple data types, then its
11519 @dfn{subprinters} are the printers for the individual data types.
11520 Each such subprinter has its own name.
11521 The format of the name is @var{printer-name};@var{subprinter-name}.
11523 Pretty-printers are installed by @dfn{registering} them with @value{GDBN}.
11524 Typically they are automatically loaded and registered when the corresponding
11525 debug information is loaded, thus making them available without having to
11526 do anything special.
11528 There are three places where a pretty-printer can be registered.
11532 Pretty-printers registered globally are available when debugging
11536 Pretty-printers registered with a program space are available only
11537 when debugging that program.
11538 @xref{Progspaces In Python}, for more details on program spaces in Python.
11541 Pretty-printers registered with an objfile are loaded and unloaded
11542 with the corresponding objfile (e.g., shared library).
11543 @xref{Objfiles In Python}, for more details on objfiles in Python.
11546 @xref{Selecting Pretty-Printers}, for further information on how
11547 pretty-printers are selected,
11549 @xref{Writing a Pretty-Printer}, for implementing pretty printers
11552 @node Pretty-Printer Example
11553 @subsection Pretty-Printer Example
11555 Here is how a C@t{++} @code{std::string} looks without a pretty-printer:
11558 (@value{GDBP}) print s
11560 static npos = 4294967295,
11562 <std::allocator<char>> = @{
11563 <__gnu_cxx::new_allocator<char>> = @{
11564 <No data fields>@}, <No data fields>
11566 members of std::basic_string<char, std::char_traits<char>,
11567 std::allocator<char> >::_Alloc_hider:
11568 _M_p = 0x804a014 "abcd"
11573 With a pretty-printer for @code{std::string} only the contents are printed:
11576 (@value{GDBP}) print s
11580 @node Pretty-Printer Commands
11581 @subsection Pretty-Printer Commands
11582 @cindex pretty-printer commands
11585 @kindex info pretty-printer
11586 @item info pretty-printer [@var{object-regexp} [@var{name-regexp}]]
11587 Print the list of installed pretty-printers.
11588 This includes disabled pretty-printers, which are marked as such.
11590 @var{object-regexp} is a regular expression matching the objects
11591 whose pretty-printers to list.
11592 Objects can be @code{global}, the program space's file
11593 (@pxref{Progspaces In Python}),
11594 and the object files within that program space (@pxref{Objfiles In Python}).
11595 @xref{Selecting Pretty-Printers}, for details on how @value{GDBN}
11596 looks up a printer from these three objects.
11598 @var{name-regexp} is a regular expression matching the name of the printers
11601 @kindex disable pretty-printer
11602 @item disable pretty-printer [@var{object-regexp} [@var{name-regexp}]]
11603 Disable pretty-printers matching @var{object-regexp} and @var{name-regexp}.
11604 A disabled pretty-printer is not forgotten, it may be enabled again later.
11606 @kindex enable pretty-printer
11607 @item enable pretty-printer [@var{object-regexp} [@var{name-regexp}]]
11608 Enable pretty-printers matching @var{object-regexp} and @var{name-regexp}.
11613 Suppose we have three pretty-printers installed: one from library1.so
11614 named @code{foo} that prints objects of type @code{foo}, and
11615 another from library2.so named @code{bar} that prints two types of objects,
11616 @code{bar1} and @code{bar2}.
11619 (@value{GDBP}) info pretty-printer
11626 (@value{GDBP}) info pretty-printer library2
11631 (@value{GDBP}) disable pretty-printer library1
11633 2 of 3 printers enabled
11634 (@value{GDBP}) info pretty-printer
11641 (@value{GDBP}) disable pretty-printer library2 bar;bar1
11643 1 of 3 printers enabled
11644 (@value{GDBP}) info pretty-printer library2
11651 (@value{GDBP}) disable pretty-printer library2 bar
11653 0 of 3 printers enabled
11654 (@value{GDBP}) info pretty-printer library2
11663 Note that for @code{bar} the entire printer can be disabled,
11664 as can each individual subprinter.
11666 Printing values and frame arguments is done by default using
11667 the enabled pretty printers.
11669 The print option @code{-raw-values} and @value{GDBN} setting
11670 @code{set print raw-values} (@pxref{set print raw-values}) can be
11671 used to print values without applying the enabled pretty printers.
11673 Similarly, the backtrace option @code{-raw-frame-arguments} and
11674 @value{GDBN} setting @code{set print raw-frame-arguments}
11675 (@pxref{set print raw-frame-arguments}) can be used to ignore the
11676 enabled pretty printers when printing frame argument values.
11678 @node Value History
11679 @section Value History
11681 @cindex value history
11682 @cindex history of values printed by @value{GDBN}
11683 Values printed by the @code{print} command are saved in the @value{GDBN}
11684 @dfn{value history}. This allows you to refer to them in other expressions.
11685 Values are kept until the symbol table is re-read or discarded
11686 (for example with the @code{file} or @code{symbol-file} commands).
11687 When the symbol table changes, the value history is discarded,
11688 since the values may contain pointers back to the types defined in the
11693 @cindex history number
11694 The values printed are given @dfn{history numbers} by which you can
11695 refer to them. These are successive integers starting with one.
11696 @code{print} shows you the history number assigned to a value by
11697 printing @samp{$@var{num} = } before the value; here @var{num} is the
11700 To refer to any previous value, use @samp{$} followed by the value's
11701 history number. The way @code{print} labels its output is designed to
11702 remind you of this. Just @code{$} refers to the most recent value in
11703 the history, and @code{$$} refers to the value before that.
11704 @code{$$@var{n}} refers to the @var{n}th value from the end; @code{$$2}
11705 is the value just prior to @code{$$}, @code{$$1} is equivalent to
11706 @code{$$}, and @code{$$0} is equivalent to @code{$}.
11708 For example, suppose you have just printed a pointer to a structure and
11709 want to see the contents of the structure. It suffices to type
11715 If you have a chain of structures where the component @code{next} points
11716 to the next one, you can print the contents of the next one with this:
11723 You can print successive links in the chain by repeating this
11724 command---which you can do by just typing @key{RET}.
11726 Note that the history records values, not expressions. If the value of
11727 @code{x} is 4 and you type these commands:
11735 then the value recorded in the value history by the @code{print} command
11736 remains 4 even though the value of @code{x} has changed.
11739 @kindex show values
11741 Print the last ten values in the value history, with their item numbers.
11742 This is like @samp{p@ $$9} repeated ten times, except that @code{show
11743 values} does not change the history.
11745 @item show values @var{n}
11746 Print ten history values centered on history item number @var{n}.
11748 @item show values +
11749 Print ten history values just after the values last printed. If no more
11750 values are available, @code{show values +} produces no display.
11753 Pressing @key{RET} to repeat @code{show values @var{n}} has exactly the
11754 same effect as @samp{show values +}.
11756 @node Convenience Vars
11757 @section Convenience Variables
11759 @cindex convenience variables
11760 @cindex user-defined variables
11761 @value{GDBN} provides @dfn{convenience variables} that you can use within
11762 @value{GDBN} to hold on to a value and refer to it later. These variables
11763 exist entirely within @value{GDBN}; they are not part of your program, and
11764 setting a convenience variable has no direct effect on further execution
11765 of your program. That is why you can use them freely.
11767 Convenience variables are prefixed with @samp{$}. Any name preceded by
11768 @samp{$} can be used for a convenience variable, unless it is one of
11769 the predefined machine-specific register names (@pxref{Registers, ,Registers}).
11770 (Value history references, in contrast, are @emph{numbers} preceded
11771 by @samp{$}. @xref{Value History, ,Value History}.)
11773 You can save a value in a convenience variable with an assignment
11774 expression, just as you would set a variable in your program.
11778 set $foo = *object_ptr
11782 would save in @code{$foo} the value contained in the object pointed to by
11785 Using a convenience variable for the first time creates it, but its
11786 value is @code{void} until you assign a new value. You can alter the
11787 value with another assignment at any time.
11789 Convenience variables have no fixed types. You can assign a convenience
11790 variable any type of value, including structures and arrays, even if
11791 that variable already has a value of a different type. The convenience
11792 variable, when used as an expression, has the type of its current value.
11795 @kindex show convenience
11796 @cindex show all user variables and functions
11797 @item show convenience
11798 Print a list of convenience variables used so far, and their values,
11799 as well as a list of the convenience functions.
11800 Abbreviated @code{show conv}.
11802 @kindex init-if-undefined
11803 @cindex convenience variables, initializing
11804 @item init-if-undefined $@var{variable} = @var{expression}
11805 Set a convenience variable if it has not already been set. This is useful
11806 for user-defined commands that keep some state. It is similar, in concept,
11807 to using local static variables with initializers in C (except that
11808 convenience variables are global). It can also be used to allow users to
11809 override default values used in a command script.
11811 If the variable is already defined then the expression is not evaluated so
11812 any side-effects do not occur.
11815 One of the ways to use a convenience variable is as a counter to be
11816 incremented or a pointer to be advanced. For example, to print
11817 a field from successive elements of an array of structures:
11821 print bar[$i++]->contents
11825 Repeat that command by typing @key{RET}.
11827 Some convenience variables are created automatically by @value{GDBN} and given
11828 values likely to be useful.
11831 @vindex $_@r{, convenience variable}
11833 The variable @code{$_} is automatically set by the @code{x} command to
11834 the last address examined (@pxref{Memory, ,Examining Memory}). Other
11835 commands which provide a default address for @code{x} to examine also
11836 set @code{$_} to that address; these commands include @code{info line}
11837 and @code{info breakpoint}. The type of @code{$_} is @code{void *}
11838 except when set by the @code{x} command, in which case it is a pointer
11839 to the type of @code{$__}.
11841 @vindex $__@r{, convenience variable}
11843 The variable @code{$__} is automatically set by the @code{x} command
11844 to the value found in the last address examined. Its type is chosen
11845 to match the format in which the data was printed.
11848 @vindex $_exitcode@r{, convenience variable}
11849 When the program being debugged terminates normally, @value{GDBN}
11850 automatically sets this variable to the exit code of the program, and
11851 resets @code{$_exitsignal} to @code{void}.
11854 @vindex $_exitsignal@r{, convenience variable}
11855 When the program being debugged dies due to an uncaught signal,
11856 @value{GDBN} automatically sets this variable to that signal's number,
11857 and resets @code{$_exitcode} to @code{void}.
11859 To distinguish between whether the program being debugged has exited
11860 (i.e., @code{$_exitcode} is not @code{void}) or signalled (i.e.,
11861 @code{$_exitsignal} is not @code{void}), the convenience function
11862 @code{$_isvoid} can be used (@pxref{Convenience Funs,, Convenience
11863 Functions}). For example, considering the following source code:
11866 #include <signal.h>
11869 main (int argc, char *argv[])
11876 A valid way of telling whether the program being debugged has exited
11877 or signalled would be:
11880 (@value{GDBP}) define has_exited_or_signalled
11881 Type commands for definition of ``has_exited_or_signalled''.
11882 End with a line saying just ``end''.
11883 >if $_isvoid ($_exitsignal)
11884 >echo The program has exited\n
11886 >echo The program has signalled\n
11892 Program terminated with signal SIGALRM, Alarm clock.
11893 The program no longer exists.
11894 (@value{GDBP}) has_exited_or_signalled
11895 The program has signalled
11898 As can be seen, @value{GDBN} correctly informs that the program being
11899 debugged has signalled, since it calls @code{raise} and raises a
11900 @code{SIGALRM} signal. If the program being debugged had not called
11901 @code{raise}, then @value{GDBN} would report a normal exit:
11904 (@value{GDBP}) has_exited_or_signalled
11905 The program has exited
11909 The variable @code{$_exception} is set to the exception object being
11910 thrown at an exception-related catchpoint. @xref{Set Catchpoints}.
11912 @item $_ada_exception
11913 The variable @code{$_ada_exception} is set to the address of the
11914 exception being caught or thrown at an Ada exception-related
11915 catchpoint. @xref{Set Catchpoints}.
11918 @itemx $_probe_arg0@dots{}$_probe_arg11
11919 Arguments to a static probe. @xref{Static Probe Points}.
11922 @vindex $_sdata@r{, inspect, convenience variable}
11923 The variable @code{$_sdata} contains extra collected static tracepoint
11924 data. @xref{Tracepoint Actions,,Tracepoint Action Lists}. Note that
11925 @code{$_sdata} could be empty, if not inspecting a trace buffer, or
11926 if extra static tracepoint data has not been collected.
11929 @vindex $_siginfo@r{, convenience variable}
11930 The variable @code{$_siginfo} contains extra signal information
11931 (@pxref{extra signal information}). Note that @code{$_siginfo}
11932 could be empty, if the application has not yet received any signals.
11933 For example, it will be empty before you execute the @code{run} command.
11936 @vindex $_tlb@r{, convenience variable}
11937 The variable @code{$_tlb} is automatically set when debugging
11938 applications running on MS-Windows in native mode or connected to
11939 gdbserver that supports the @code{qGetTIBAddr} request.
11940 @xref{General Query Packets}.
11941 This variable contains the address of the thread information block.
11944 The number of the current inferior. @xref{Inferiors and
11945 Programs, ,Debugging Multiple Inferiors and Programs}.
11948 The thread number of the current thread. @xref{thread numbers}.
11951 The global number of the current thread. @xref{global thread number}.
11953 @item $_thread_systag
11954 The target system's thread identifier (@var{systag}) string of the
11955 current thread. @xref{target system thread identifier}.
11957 @item $_thread_name
11958 The thread name string of the current thread, or the empty string if
11959 no name has been assigned. @xref{thread name}.
11963 @vindex $_gdb_major@r{, convenience variable}
11964 @vindex $_gdb_minor@r{, convenience variable}
11965 The major and minor version numbers of the running @value{GDBN}.
11966 Development snapshots and pretest versions have their minor version
11967 incremented by one; thus, @value{GDBN} pretest 9.11.90 will produce
11968 the value 12 for @code{$_gdb_minor}. These variables allow you to
11969 write scripts that work with different versions of @value{GDBN}
11970 without errors caused by features unavailable in some of those
11973 @item $_shell_exitcode
11974 @itemx $_shell_exitsignal
11975 @vindex $_shell_exitcode@r{, convenience variable}
11976 @vindex $_shell_exitsignal@r{, convenience variable}
11977 @cindex shell command, exit code
11978 @cindex shell command, exit signal
11979 @cindex exit status of shell commands
11980 @value{GDBN} commands such as @code{shell} and @code{|} are launching
11981 shell commands. When a launched command terminates, @value{GDBN}
11982 automatically maintains the variables @code{$_shell_exitcode}
11983 and @code{$_shell_exitsignal} according to the exit status of the last
11984 launched command. These variables are set and used similarly to
11985 the variables @code{$_exitcode} and @code{$_exitsignal}.
11988 The per-inferior heterogeneous agent number of the current thread, or 0
11989 if not associated with a heterogeneous dispatch. @xref{Heterogeneous
11993 The global heterogeneous agent number of the current thread, or 0 if not
11994 associated with a heterogeneous dispatch. @xref{Heterogeneous
11998 The per-inferior heterogeneous queue number of the current thread, or 0
11999 if not associated with a heterogeneous dispatch. @xref{Heterogeneous
12003 The global heterogeneous queue number of the current thread, or 0 if not
12004 associated with a heterogeneous dispatch. @xref{Heterogeneous
12008 The per-inferior heterogeneous dispatch number of the current thread, or
12009 0 if not associated with a heterogeneous dispatch. @xref{Heterogeneous
12013 The global heterogeneous dispatch number of the current thread, or 0 if
12014 not associated with a heterogeneous dispatch. @xref{Heterogeneous
12018 The per-inferior heterogeneous lane number of the current
12019 heterogeneous lane of the current thread. @xref{Heterogeneous
12022 @c FIXME-implementors!! Should there be @code{$_lane_index},
12023 @c @code{$_lane_active} and @code{$_lane_count} convenience variables?
12024 @c Note that the lane count needs to take into account when a grid
12025 @c size is not a multiple of the work-group size (resulting in partial
12026 @c work-groups on the dimension edges of the grid), and the work-group
12027 @c size is not a multiple of the wavefront size.
12030 The global heterogeneous lane number of the current heterogeneous
12031 lane. @xref{Heterogeneous Debugging}.
12033 @item $_lane_systag
12034 The target system's heterogeneous lane identifier (@var{lane_systag})
12035 string of the current lane in the current thread. @xref{target system
12039 The heterogeneous lane name string of the current heterogeneous lane, or
12040 the empty string if no name has been assigned by the @samp{lane name}
12041 command. @xref{Heterogeneous Debugging}.
12043 @item $_dispatch_pos
12044 The heterogeneous dispatch position string of the current thread, or the
12045 empty string if not associated with a heterogeneous dispatch.
12046 @xref{Heterogeneous Debugging}.
12048 @item $_thread_workgroup_pos
12049 @itemx $_lane_workgroup_pos
12050 The heterogeneous work-group position string of the current thread or
12051 heterogeneous lane respectively, or the empty string if not associated
12052 with a heterogeneous dispatch. @xref{Heterogeneous Debugging}.
12056 @node Convenience Funs
12057 @section Convenience Functions
12059 @cindex convenience functions
12060 @value{GDBN} also supplies some @dfn{convenience functions}. These
12061 have a syntax similar to convenience variables. A convenience
12062 function can be used in an expression just like an ordinary function;
12063 however, a convenience function is implemented internally to
12066 These functions do not require @value{GDBN} to be configured with
12067 @code{Python} support, which means that they are always available.
12071 @item $_isvoid (@var{expr})
12072 @findex $_isvoid@r{, convenience function}
12073 Return one if the expression @var{expr} is @code{void}. Otherwise it
12076 A @code{void} expression is an expression where the type of the result
12077 is @code{void}. For example, you can examine a convenience variable
12078 (see @ref{Convenience Vars,, Convenience Variables}) to check whether
12082 (@value{GDBP}) print $_exitcode
12084 (@value{GDBP}) print $_isvoid ($_exitcode)
12087 Starting program: ./a.out
12088 [Inferior 1 (process 29572) exited normally]
12089 (@value{GDBP}) print $_exitcode
12091 (@value{GDBP}) print $_isvoid ($_exitcode)
12095 In the example above, we used @code{$_isvoid} to check whether
12096 @code{$_exitcode} is @code{void} before and after the execution of the
12097 program being debugged. Before the execution there is no exit code to
12098 be examined, therefore @code{$_exitcode} is @code{void}. After the
12099 execution the program being debugged returned zero, therefore
12100 @code{$_exitcode} is zero, which means that it is not @code{void}
12103 The @code{void} expression can also be a call of a function from the
12104 program being debugged. For example, given the following function:
12113 The result of calling it inside @value{GDBN} is @code{void}:
12116 (@value{GDBP}) print foo ()
12118 (@value{GDBP}) print $_isvoid (foo ())
12120 (@value{GDBP}) set $v = foo ()
12121 (@value{GDBP}) print $v
12123 (@value{GDBP}) print $_isvoid ($v)
12127 @item $_gdb_setting_str (@var{setting})
12128 @findex $_gdb_setting_str@r{, convenience function}
12129 Return the value of the @value{GDBN} @var{setting} as a string.
12130 @var{setting} is any setting that can be used in a @code{set} or
12131 @code{show} command (@pxref{Controlling GDB}).
12134 (@value{GDBP}) show print frame-arguments
12135 Printing of non-scalar frame arguments is "scalars".
12136 (@value{GDBP}) p $_gdb_setting_str("print frame-arguments")
12138 (@value{GDBP}) p $_gdb_setting_str("height")
12143 @item $_gdb_setting (@var{setting})
12144 @findex $_gdb_setting@r{, convenience function}
12145 Return the value of the @value{GDBN} @var{setting}.
12146 The type of the returned value depends on the setting.
12148 The value type for boolean and auto boolean settings is @code{int}.
12149 The boolean values @code{off} and @code{on} are converted to
12150 the integer values @code{0} and @code{1}. The value @code{auto} is
12151 converted to the value @code{-1}.
12153 The value type for integer settings is either @code{unsigned int}
12154 or @code{int}, depending on the setting.
12156 Some integer settings accept an @code{unlimited} value.
12157 Depending on the setting, the @code{set} command also accepts
12158 the value @code{0} or the value @code{@minus{}1} as a synonym for
12160 For example, @code{set height unlimited} is equivalent to
12161 @code{set height 0}.
12163 Some other settings that accept the @code{unlimited} value
12164 use the value @code{0} to literally mean zero.
12165 For example, @code{set history size 0} indicates to not
12166 record any @value{GDBN} commands in the command history.
12167 For such settings, @code{@minus{}1} is the synonym
12168 for @code{unlimited}.
12170 See the documentation of the corresponding @code{set} command for
12171 the numerical value equivalent to @code{unlimited}.
12173 The @code{$_gdb_setting} function converts the unlimited value
12174 to a @code{0} or a @code{@minus{}1} value according to what the
12175 @code{set} command uses.
12179 (@value{GDBP}) p $_gdb_setting_str("height")
12181 (@value{GDBP}) p $_gdb_setting("height")
12183 (@value{GDBP}) set height unlimited
12184 (@value{GDBP}) p $_gdb_setting_str("height")
12186 (@value{GDBP}) p $_gdb_setting("height")
12190 (@value{GDBP}) p $_gdb_setting_str("history size")
12192 (@value{GDBP}) p $_gdb_setting("history size")
12194 (@value{GDBP}) p $_gdb_setting_str("disassemble-next-line")
12196 (@value{GDBP}) p $_gdb_setting("disassemble-next-line")
12202 Other setting types (enum, filename, optional filename, string, string noescape)
12203 are returned as string values.
12206 @item $_gdb_maint_setting_str (@var{setting})
12207 @findex $_gdb_maint_setting_str@r{, convenience function}
12208 Like the @code{$_gdb_setting_str} function, but works with
12209 @code{maintenance set} variables.
12211 @item $_gdb_maint_setting (@var{setting})
12212 @findex $_gdb_maint_setting@r{, convenience function}
12213 Like the @code{$_gdb_setting} function, but works with
12214 @code{maintenance set} variables.
12218 The following functions require @value{GDBN} to be configured with
12219 @code{Python} support.
12223 @item $_memeq(@var{buf1}, @var{buf2}, @var{length})
12224 @findex $_memeq@r{, convenience function}
12225 Returns one if the @var{length} bytes at the addresses given by
12226 @var{buf1} and @var{buf2} are equal.
12227 Otherwise it returns zero.
12229 @item $_regex(@var{str}, @var{regex})
12230 @findex $_regex@r{, convenience function}
12231 Returns one if the string @var{str} matches the regular expression
12232 @var{regex}. Otherwise it returns zero.
12233 The syntax of the regular expression is that specified by Python's
12234 regular expression support.
12236 @item $_streq(@var{str1}, @var{str2})
12237 @findex $_streq@r{, convenience function}
12238 Returns one if the strings @var{str1} and @var{str2} are equal.
12239 Otherwise it returns zero.
12241 @item $_strlen(@var{str})
12242 @findex $_strlen@r{, convenience function}
12243 Returns the length of string @var{str}.
12245 @item $_caller_is(@var{name}@r{[}, @var{number_of_frames}@r{]})
12246 @findex $_caller_is@r{, convenience function}
12247 Returns one if the calling function's name is equal to @var{name}.
12248 Otherwise it returns zero.
12250 If the optional argument @var{number_of_frames} is provided,
12251 it is the number of frames up in the stack to look.
12257 (@value{GDBP}) backtrace
12259 at testsuite/gdb.python/py-caller-is.c:21
12260 #1 0x00000000004005a0 in middle_func ()
12261 at testsuite/gdb.python/py-caller-is.c:27
12262 #2 0x00000000004005ab in top_func ()
12263 at testsuite/gdb.python/py-caller-is.c:33
12264 #3 0x00000000004005b6 in main ()
12265 at testsuite/gdb.python/py-caller-is.c:39
12266 (@value{GDBP}) print $_caller_is ("middle_func")
12268 (@value{GDBP}) print $_caller_is ("top_func", 2)
12272 @item $_caller_matches(@var{regexp}@r{[}, @var{number_of_frames}@r{]})
12273 @findex $_caller_matches@r{, convenience function}
12274 Returns one if the calling function's name matches the regular expression
12275 @var{regexp}. Otherwise it returns zero.
12277 If the optional argument @var{number_of_frames} is provided,
12278 it is the number of frames up in the stack to look.
12281 @item $_any_caller_is(@var{name}@r{[}, @var{number_of_frames}@r{]})
12282 @findex $_any_caller_is@r{, convenience function}
12283 Returns one if any calling function's name is equal to @var{name}.
12284 Otherwise it returns zero.
12286 If the optional argument @var{number_of_frames} is provided,
12287 it is the number of frames up in the stack to look.
12290 This function differs from @code{$_caller_is} in that this function
12291 checks all stack frames from the immediate caller to the frame specified
12292 by @var{number_of_frames}, whereas @code{$_caller_is} only checks the
12293 frame specified by @var{number_of_frames}.
12295 @item $_any_caller_matches(@var{regexp}@r{[}, @var{number_of_frames}@r{]})
12296 @findex $_any_caller_matches@r{, convenience function}
12297 Returns one if any calling function's name matches the regular expression
12298 @var{regexp}. Otherwise it returns zero.
12300 If the optional argument @var{number_of_frames} is provided,
12301 it is the number of frames up in the stack to look.
12304 This function differs from @code{$_caller_matches} in that this function
12305 checks all stack frames from the immediate caller to the frame specified
12306 by @var{number_of_frames}, whereas @code{$_caller_matches} only checks the
12307 frame specified by @var{number_of_frames}.
12309 @item $_as_string(@var{value})
12310 @findex $_as_string@r{, convenience function}
12311 Return the string representation of @var{value}.
12313 This function is useful to obtain the textual label (enumerator) of an
12314 enumeration value. For example, assuming the variable @var{node} is of
12315 an enumerated type:
12318 (@value{GDBP}) printf "Visiting node of type %s\n", $_as_string(node)
12319 Visiting node of type NODE_INTEGER
12322 @item $_cimag(@var{value})
12323 @itemx $_creal(@var{value})
12324 @findex $_cimag@r{, convenience function}
12325 @findex $_creal@r{, convenience function}
12326 Return the imaginary (@code{$_cimag}) or real (@code{$_creal}) part of
12327 the complex number @var{value}.
12329 The type of the imaginary or real part depends on the type of the
12330 complex number, e.g., using @code{$_cimag} on a @code{float complex}
12331 will return an imaginary part of type @code{float}.
12333 @item $_thread_find(@var{regex})
12334 @findex $_thread_find@r{, convenience function}
12335 Searches for threads whose name or @var{systag} matches the supplied
12336 regular expression. The syntax of the regular expression is that
12337 specified by Python's regular expression support.
12339 Returns a string that is the space separated list of per-inferior
12340 thread numbers of the found threads. If debugging multiple inferiors,
12341 the thread numbers are qualified with the inferior number. If no
12342 threads are found, the empty string is returned. The string can be
12343 used in commands that accept a thread ID list. @xref{thread ID
12346 @c FIXME-implementors!! Should this convenience function return a
12347 @c tuple rather than a string?
12349 For example, the following command lists all threads that are part of
12350 the heterogeneous work-group with dispatch position @samp{(1,2,3)}
12351 (@pxref{Heterogeneous Debugging}):
12354 (@value{GDBP}) info threads $_thread_find ("work-group(1,2,3)")
12357 @item $_thread_find_first_gid(@var{regex})
12358 @findex $_thread_find_first_gid@r{, convenience function}
12359 Similar to the @code{$_thread_find} convenience function, except it
12360 returns a number that is the global thread number of one of the
12361 threads found, or 0 if no threads were found. The number can be used
12362 in commands that accept a global thread number. @xref{global thread
12365 @c FIXME-implementors!! If @code{$_thread_find} returns a tuple then
12366 @c this convenience function may not be necessary as one can simply
12367 @c add @samp{[0]} to access the first element of a tuple.
12369 For example, the following command sets the current thread to one of
12370 the threads that are part of the heterogeneous work-group with
12371 dispatch position @samp{(1,2,3)} (@pxref{Heterogeneous Debugging}):
12374 (@value{GDBP}) thread -gid $_thread_find_first_gid ("work-group(1,2,3)")
12377 @item $_lane_find(@var{regex})
12378 @itemx $_lane_find_first_gid(@var{regex})
12379 Similar to @samp{$_thread_find} and @samp{$_thread_find_first_gid}
12380 except for heterogeneous lanes. @xref{Heterogeneous Debugging}.
12384 @value{GDBN} provides the ability to list and get help on
12385 convenience functions.
12388 @item help function
12389 @kindex help function
12390 @cindex show all convenience functions
12391 Print a list of all convenience functions.
12398 You can refer to machine register contents, in expressions, as variables
12399 with names starting with @samp{$}. The names of registers are different
12400 for each machine; use @code{info registers} to see the names used on
12404 @kindex info registers
12405 @item info registers
12406 Print the names and values of all registers except floating-point
12407 and vector registers (in the selected stack frame).
12409 @kindex info all-registers
12410 @cindex floating point registers
12411 @item info all-registers
12412 Print the names and values of all registers, including floating-point
12413 and vector registers (in the selected stack frame).
12415 @item info registers @var{reggroup} @dots{}
12416 Print the name and value of the registers in each of the specified
12417 @var{reggroup}s. The @var{reggroup} can be any of those returned by
12418 @code{maint print reggroups} (@pxref{Maintenance Commands}).
12420 @item info registers @var{regname} @dots{}
12421 Print the @dfn{relativized} value of each specified register @var{regname}.
12422 As discussed in detail below, register values are normally relative to
12423 the selected stack frame. The @var{regname} may be any register name valid on
12424 the machine you are using, with or without the initial @samp{$}.
12427 @anchor{standard registers}
12428 @cindex stack pointer register
12429 @cindex program counter register
12430 @cindex process status register
12431 @cindex frame pointer register
12432 @cindex standard registers
12433 @value{GDBN} has four ``standard'' register names that are available (in
12434 expressions) on most machines---whenever they do not conflict with an
12435 architecture's canonical mnemonics for registers. The register names
12436 @code{$pc} and @code{$sp} are used for the program counter register and
12437 the stack pointer. @code{$fp} is used for a register that contains a
12438 pointer to the current stack frame, and @code{$ps} is used for a
12439 register that contains the processor status. For example,
12440 you could print the program counter in hex with
12447 or print the instruction to be executed next with
12454 or add four to the stack pointer@footnote{This is a way of removing
12455 one word from the stack, on machines where stacks grow downward in
12456 memory (most machines, nowadays). This assumes that the innermost
12457 stack frame is selected; setting @code{$sp} is not allowed when other
12458 stack frames are selected. To pop entire frames off the stack,
12459 regardless of machine architecture, use @code{return};
12460 see @ref{Returning, ,Returning from a Function}.} with
12466 Whenever possible, these four standard register names are available on
12467 your machine even though the machine has different canonical mnemonics,
12468 so long as there is no conflict. The @code{info registers} command
12469 shows the canonical names. For example, on the SPARC, @code{info
12470 registers} displays the processor status register as @code{$psr} but you
12471 can also refer to it as @code{$ps}; and on x86-based machines @code{$ps}
12472 is an alias for the @sc{eflags} register.
12474 @value{GDBN} always considers the contents of an ordinary register as an
12475 integer when the register is examined in this way. Some machines have
12476 special registers which can hold nothing but floating point; these
12477 registers are considered to have floating point values. There is no way
12478 to refer to the contents of an ordinary register as floating point value
12479 (although you can @emph{print} it as a floating point value with
12480 @samp{print/f $@var{regname}}).
12482 Some registers have distinct ``raw'' and ``virtual'' data formats. This
12483 means that the data format in which the register contents are saved by
12484 the operating system is not the same one that your program normally
12485 sees. For example, the registers of the 68881 floating point
12486 coprocessor are always saved in ``extended'' (raw) format, but all C
12487 programs expect to work with ``double'' (virtual) format. In such
12488 cases, @value{GDBN} normally works with the virtual format only (the format
12489 that makes sense for your program), but the @code{info registers} command
12490 prints the data in both formats.
12492 @cindex SSE registers (x86)
12493 @cindex MMX registers (x86)
12494 Some machines have special registers whose contents can be interpreted
12495 in several different ways. For example, modern x86-based machines
12496 have SSE and MMX registers that can hold several values packed
12497 together in several different formats. @value{GDBN} refers to such
12498 registers in @code{struct} notation:
12501 (@value{GDBP}) print $xmm1
12503 v4_float = @{0, 3.43859137e-038, 1.54142831e-044, 1.821688e-044@},
12504 v2_double = @{9.92129282474342e-303, 2.7585945287983262e-313@},
12505 v16_int8 = "\000\000\000\000\3706;\001\v\000\000\000\r\000\000",
12506 v8_int16 = @{0, 0, 14072, 315, 11, 0, 13, 0@},
12507 v4_int32 = @{0, 20657912, 11, 13@},
12508 v2_int64 = @{88725056443645952, 55834574859@},
12509 uint128 = 0x0000000d0000000b013b36f800000000
12514 To set values of such registers, you need to tell @value{GDBN} which
12515 view of the register you wish to change, as if you were assigning
12516 value to a @code{struct} member:
12519 (@value{GDBP}) set $xmm1.uint128 = 0x000000000000000000000000FFFFFFFF
12522 Normally, register values are relative to the selected stack frame
12523 (@pxref{Selection, ,Selecting a Frame}). This means that you get the
12524 value that the register would contain if all stack frames farther in
12525 were exited and their saved registers restored. In order to see the
12526 true contents of hardware registers, you must select the innermost
12527 frame (with @samp{frame 0}).
12529 @cindex caller-saved registers
12530 @cindex call-clobbered registers
12531 @cindex volatile registers
12532 @cindex <not saved> values
12533 Usually ABIs reserve some registers as not needed to be saved by the
12534 callee (a.k.a.: ``caller-saved'', ``call-clobbered'' or ``volatile''
12535 registers). It may therefore not be possible for @value{GDBN} to know
12536 the value a register had before the call (in other words, in the outer
12537 frame), if the register value has since been changed by the callee.
12538 @value{GDBN} tries to deduce where the inner frame saved
12539 (``callee-saved'') registers, from the debug info, unwind info, or the
12540 machine code generated by your compiler. If some register is not
12541 saved, and @value{GDBN} knows the register is ``caller-saved'' (via
12542 its own knowledge of the ABI, or because the debug/unwind info
12543 explicitly says the register's value is undefined), @value{GDBN}
12544 displays @w{@samp{<not saved>}} as the register's value. With targets
12545 that @value{GDBN} has no knowledge of the register saving convention,
12546 if a register was not saved by the callee, then its value and location
12547 in the outer frame are assumed to be the same of the inner frame.
12548 This is usually harmless, because if the register is call-clobbered,
12549 the caller either does not care what is in the register after the
12550 call, or has code to restore the value that it does care about. Note,
12551 however, that if you change such a register in the outer frame, you
12552 may also be affecting the inner frame. Also, the more ``outer'' the
12553 frame is you're looking at, the more likely a call-clobbered
12554 register's value is to be wrong, in the sense that it doesn't actually
12555 represent the value the register had just before the call.
12557 @node Floating Point Hardware
12558 @section Floating Point Hardware
12559 @cindex floating point
12561 Depending on the configuration, @value{GDBN} may be able to give
12562 you more information about the status of the floating point hardware.
12567 Display hardware-dependent information about the floating
12568 point unit. The exact contents and layout vary depending on the
12569 floating point chip. Currently, @samp{info float} is supported on
12570 the ARM and x86 machines.
12574 @section Vector Unit
12575 @cindex vector unit
12577 Depending on the configuration, @value{GDBN} may be able to give you
12578 more information about the status of the vector unit.
12581 @kindex info vector
12583 Display information about the vector unit. The exact contents and
12584 layout vary depending on the hardware.
12587 @node OS Information
12588 @section Operating System Auxiliary Information
12589 @cindex OS information
12591 @value{GDBN} provides interfaces to useful OS facilities that can help
12592 you debug your program.
12594 @cindex auxiliary vector
12595 @cindex vector, auxiliary
12596 Some operating systems supply an @dfn{auxiliary vector} to programs at
12597 startup. This is akin to the arguments and environment that you
12598 specify for a program, but contains a system-dependent variety of
12599 binary values that tell system libraries important details about the
12600 hardware, operating system, and process. Each value's purpose is
12601 identified by an integer tag; the meanings are well-known but system-specific.
12602 Depending on the configuration and operating system facilities,
12603 @value{GDBN} may be able to show you this information. For remote
12604 targets, this functionality may further depend on the remote stub's
12605 support of the @samp{qXfer:auxv:read} packet, see
12606 @ref{qXfer auxiliary vector read}.
12611 Display the auxiliary vector of the inferior, which can be either a
12612 live process or a core dump file. @value{GDBN} prints each tag value
12613 numerically, and also shows names and text descriptions for recognized
12614 tags. Some values in the vector are numbers, some bit masks, and some
12615 pointers to strings or other data. @value{GDBN} displays each value in the
12616 most appropriate form for a recognized tag, and in hexadecimal for
12617 an unrecognized tag.
12620 On some targets, @value{GDBN} can access operating system-specific
12621 information and show it to you. The types of information available
12622 will differ depending on the type of operating system running on the
12623 target. The mechanism used to fetch the data is described in
12624 @ref{Operating System Information}. For remote targets, this
12625 functionality depends on the remote stub's support of the
12626 @samp{qXfer:osdata:read} packet, see @ref{qXfer osdata read}.
12630 @item info os @var{infotype}
12632 Display OS information of the requested type.
12634 On @sc{gnu}/Linux, the following values of @var{infotype} are valid:
12636 @anchor{linux info os infotypes}
12638 @kindex info os cpus
12640 Display the list of all CPUs/cores. For each CPU/core, @value{GDBN} prints
12641 the available fields from /proc/cpuinfo. For each supported architecture
12642 different fields are available. Two common entries are processor which gives
12643 CPU number and bogomips; a system constant that is calculated during
12644 kernel initialization.
12646 @kindex info os files
12648 Display the list of open file descriptors on the target. For each
12649 file descriptor, @value{GDBN} prints the identifier of the process
12650 owning the descriptor, the command of the owning process, the value
12651 of the descriptor, and the target of the descriptor.
12653 @kindex info os modules
12655 Display the list of all loaded kernel modules on the target. For each
12656 module, @value{GDBN} prints the module name, the size of the module in
12657 bytes, the number of times the module is used, the dependencies of the
12658 module, the status of the module, and the address of the loaded module
12661 @kindex info os msg
12663 Display the list of all System V message queues on the target. For each
12664 message queue, @value{GDBN} prints the message queue key, the message
12665 queue identifier, the access permissions, the current number of bytes
12666 on the queue, the current number of messages on the queue, the processes
12667 that last sent and received a message on the queue, the user and group
12668 of the owner and creator of the message queue, the times at which a
12669 message was last sent and received on the queue, and the time at which
12670 the message queue was last changed.
12672 @kindex info os processes
12674 Display the list of processes on the target. For each process,
12675 @value{GDBN} prints the process identifier, the name of the user, the
12676 command corresponding to the process, and the list of processor cores
12677 that the process is currently running on. (To understand what these
12678 properties mean, for this and the following info types, please consult
12679 the general @sc{gnu}/Linux documentation.)
12681 @kindex info os procgroups
12683 Display the list of process groups on the target. For each process,
12684 @value{GDBN} prints the identifier of the process group that it belongs
12685 to, the command corresponding to the process group leader, the process
12686 identifier, and the command line of the process. The list is sorted
12687 first by the process group identifier, then by the process identifier,
12688 so that processes belonging to the same process group are grouped together
12689 and the process group leader is listed first.
12691 @kindex info os semaphores
12693 Display the list of all System V semaphore sets on the target. For each
12694 semaphore set, @value{GDBN} prints the semaphore set key, the semaphore
12695 set identifier, the access permissions, the number of semaphores in the
12696 set, the user and group of the owner and creator of the semaphore set,
12697 and the times at which the semaphore set was operated upon and changed.
12699 @kindex info os shm
12701 Display the list of all System V shared-memory regions on the target.
12702 For each shared-memory region, @value{GDBN} prints the region key,
12703 the shared-memory identifier, the access permissions, the size of the
12704 region, the process that created the region, the process that last
12705 attached to or detached from the region, the current number of live
12706 attaches to the region, and the times at which the region was last
12707 attached to, detach from, and changed.
12709 @kindex info os sockets
12711 Display the list of Internet-domain sockets on the target. For each
12712 socket, @value{GDBN} prints the address and port of the local and
12713 remote endpoints, the current state of the connection, the creator of
12714 the socket, the IP address family of the socket, and the type of the
12717 @kindex info os threads
12719 Display the list of threads running on the target. For each thread,
12720 @value{GDBN} prints the identifier of the process that the thread
12721 belongs to, the command of the process, the thread identifier, and the
12722 processor core that it is currently running on. The main thread of a
12723 process is not listed.
12727 If @var{infotype} is omitted, then list the possible values for
12728 @var{infotype} and the kind of OS information available for each
12729 @var{infotype}. If the target does not return a list of possible
12730 types, this command will report an error.
12733 @node Memory Region Attributes
12734 @section Memory Region Attributes
12735 @cindex memory region attributes
12737 @dfn{Memory region attributes} allow you to describe special handling
12738 required by regions of your target's memory. @value{GDBN} uses
12739 attributes to determine whether to allow certain types of memory
12740 accesses; whether to use specific width accesses; and whether to cache
12741 target memory. By default the description of memory regions is
12742 fetched from the target (if the current target supports this), but the
12743 user can override the fetched regions.
12745 Defined memory regions can be individually enabled and disabled. When a
12746 memory region is disabled, @value{GDBN} uses the default attributes when
12747 accessing memory in that region. Similarly, if no memory regions have
12748 been defined, @value{GDBN} uses the default attributes when accessing
12751 When a memory region is defined, it is given a number to identify it;
12752 to enable, disable, or remove a memory region, you specify that number.
12756 @item mem @var{lower} @var{upper} @var{attributes}@dots{}
12757 Define a memory region bounded by @var{lower} and @var{upper} with
12758 attributes @var{attributes}@dots{}, and add it to the list of regions
12759 monitored by @value{GDBN}. Note that @var{upper} == 0 is a special
12760 case: it is treated as the target's maximum memory address.
12761 (0xffff on 16 bit targets, 0xffffffff on 32 bit targets, etc.)
12764 Discard any user changes to the memory regions and use target-supplied
12765 regions, if available, or no regions if the target does not support.
12768 @item delete mem @var{nums}@dots{}
12769 Remove memory regions @var{nums}@dots{} from the list of regions
12770 monitored by @value{GDBN}.
12772 @kindex disable mem
12773 @item disable mem @var{nums}@dots{}
12774 Disable monitoring of memory regions @var{nums}@dots{}.
12775 A disabled memory region is not forgotten.
12776 It may be enabled again later.
12779 @item enable mem @var{nums}@dots{}
12780 Enable monitoring of memory regions @var{nums}@dots{}.
12784 Print a table of all defined memory regions, with the following columns
12788 @item Memory Region Number
12789 @item Enabled or Disabled.
12790 Enabled memory regions are marked with @samp{y}.
12791 Disabled memory regions are marked with @samp{n}.
12794 The address defining the inclusive lower bound of the memory region.
12797 The address defining the exclusive upper bound of the memory region.
12800 The list of attributes set for this memory region.
12805 @subsection Attributes
12807 @subsubsection Memory Access Mode
12808 The access mode attributes set whether @value{GDBN} may make read or
12809 write accesses to a memory region.
12811 While these attributes prevent @value{GDBN} from performing invalid
12812 memory accesses, they do nothing to prevent the target system, I/O DMA,
12813 etc.@: from accessing memory.
12817 Memory is read only.
12819 Memory is write only.
12821 Memory is read/write. This is the default.
12824 @subsubsection Memory Access Size
12825 The access size attribute tells @value{GDBN} to use specific sized
12826 accesses in the memory region. Often memory mapped device registers
12827 require specific sized accesses. If no access size attribute is
12828 specified, @value{GDBN} may use accesses of any size.
12832 Use 8 bit memory accesses.
12834 Use 16 bit memory accesses.
12836 Use 32 bit memory accesses.
12838 Use 64 bit memory accesses.
12841 @c @subsubsection Hardware/Software Breakpoints
12842 @c The hardware/software breakpoint attributes set whether @value{GDBN}
12843 @c will use hardware or software breakpoints for the internal breakpoints
12844 @c used by the step, next, finish, until, etc. commands.
12848 @c Always use hardware breakpoints
12849 @c @item swbreak (default)
12852 @subsubsection Data Cache
12853 The data cache attributes set whether @value{GDBN} will cache target
12854 memory. While this generally improves performance by reducing debug
12855 protocol overhead, it can lead to incorrect results because @value{GDBN}
12856 does not know about volatile variables or memory mapped device
12861 Enable @value{GDBN} to cache target memory.
12863 Disable @value{GDBN} from caching target memory. This is the default.
12866 @subsection Memory Access Checking
12867 @value{GDBN} can be instructed to refuse accesses to memory that is
12868 not explicitly described. This can be useful if accessing such
12869 regions has undesired effects for a specific target, or to provide
12870 better error checking. The following commands control this behaviour.
12873 @kindex set mem inaccessible-by-default
12874 @item set mem inaccessible-by-default [on|off]
12875 If @code{on} is specified, make @value{GDBN} treat memory not
12876 explicitly described by the memory ranges as non-existent and refuse accesses
12877 to such memory. The checks are only performed if there's at least one
12878 memory range defined. If @code{off} is specified, make @value{GDBN}
12879 treat the memory not explicitly described by the memory ranges as RAM.
12880 The default value is @code{on}.
12881 @kindex show mem inaccessible-by-default
12882 @item show mem inaccessible-by-default
12883 Show the current handling of accesses to unknown memory.
12887 @c @subsubsection Memory Write Verification
12888 @c The memory write verification attributes set whether @value{GDBN}
12889 @c will re-reads data after each write to verify the write was successful.
12893 @c @item noverify (default)
12896 @node Dump/Restore Files
12897 @section Copy Between Memory and a File
12898 @cindex dump/restore files
12899 @cindex append data to a file
12900 @cindex dump data to a file
12901 @cindex restore data from a file
12903 You can use the commands @code{dump}, @code{append}, and
12904 @code{restore} to copy data between target memory and a file. The
12905 @code{dump} and @code{append} commands write data to a file, and the
12906 @code{restore} command reads data from a file back into the inferior's
12907 memory. Files may be in binary, Motorola S-record, Intel hex,
12908 Tektronix Hex, or Verilog Hex format; however, @value{GDBN} can only
12909 append to binary files, and cannot read from Verilog Hex files.
12914 @item dump @r{[}@var{format}@r{]} memory @var{filename} @var{start_addr} @var{end_addr}
12915 @itemx dump @r{[}@var{format}@r{]} value @var{filename} @var{expr}
12916 Dump the contents of memory from @var{start_addr} to @var{end_addr},
12917 or the value of @var{expr}, to @var{filename} in the given format.
12919 The @var{format} parameter may be any one of:
12926 Motorola S-record format.
12928 Tektronix Hex format.
12930 Verilog Hex format.
12933 @value{GDBN} uses the same definitions of these formats as the
12934 @sc{gnu} binary utilities, like @samp{objdump} and @samp{objcopy}. If
12935 @var{format} is omitted, @value{GDBN} dumps the data in raw binary
12939 @item append @r{[}binary@r{]} memory @var{filename} @var{start_addr} @var{end_addr}
12940 @itemx append @r{[}binary@r{]} value @var{filename} @var{expr}
12941 Append the contents of memory from @var{start_addr} to @var{end_addr},
12942 or the value of @var{expr}, to the file @var{filename}, in raw binary form.
12943 (@value{GDBN} can only append data to files in raw binary form.)
12946 @item restore @var{filename} @r{[}binary@r{]} @var{bias} @var{start} @var{end}
12947 Restore the contents of file @var{filename} into memory. The
12948 @code{restore} command can automatically recognize any known @sc{bfd}
12949 file format, except for raw binary. To restore a raw binary file you
12950 must specify the optional keyword @code{binary} after the filename.
12952 If @var{bias} is non-zero, its value will be added to the addresses
12953 contained in the file. Binary files always start at address zero, so
12954 they will be restored at address @var{bias}. Other bfd files have
12955 a built-in location; they will be restored at offset @var{bias}
12956 from that location.
12958 If @var{start} and/or @var{end} are non-zero, then only data between
12959 file offset @var{start} and file offset @var{end} will be restored.
12960 These offsets are relative to the addresses in the file, before
12961 the @var{bias} argument is applied.
12965 @node Core File Generation
12966 @section How to Produce a Core File from Your Program
12967 @cindex dump core from inferior
12969 A @dfn{core file} or @dfn{core dump} is a file that records the memory
12970 image of a running process and its process status (register values
12971 etc.). Its primary use is post-mortem debugging of a program that
12972 crashed while it ran outside a debugger. A program that crashes
12973 automatically produces a core file, unless this feature is disabled by
12974 the user. @xref{Files}, for information on invoking @value{GDBN} in
12975 the post-mortem debugging mode.
12977 Occasionally, you may wish to produce a core file of the program you
12978 are debugging in order to preserve a snapshot of its state.
12979 @value{GDBN} has a special command for that.
12983 @kindex generate-core-file
12984 @item generate-core-file [@var{file}]
12985 @itemx gcore [@var{file}]
12986 Produce a core dump of the inferior process. The optional argument
12987 @var{file} specifies the file name where to put the core dump. If not
12988 specified, the file name defaults to @file{core.@var{pid}}, where
12989 @var{pid} is the inferior process ID.
12991 Note that this command is implemented only for some systems (as of
12992 this writing, @sc{gnu}/Linux, FreeBSD, Solaris, and S390).
12994 On @sc{gnu}/Linux, this command can take into account the value of the
12995 file @file{/proc/@var{pid}/coredump_filter} when generating the core
12996 dump (@pxref{set use-coredump-filter}), and by default honors the
12997 @code{VM_DONTDUMP} flag for mappings where it is present in the file
12998 @file{/proc/@var{pid}/smaps} (@pxref{set dump-excluded-mappings}).
13000 @kindex set use-coredump-filter
13001 @anchor{set use-coredump-filter}
13002 @item set use-coredump-filter on
13003 @itemx set use-coredump-filter off
13004 Enable or disable the use of the file
13005 @file{/proc/@var{pid}/coredump_filter} when generating core dump
13006 files. This file is used by the Linux kernel to decide what types of
13007 memory mappings will be dumped or ignored when generating a core dump
13008 file. @var{pid} is the process ID of a currently running process.
13010 To make use of this feature, you have to write in the
13011 @file{/proc/@var{pid}/coredump_filter} file a value, in hexadecimal,
13012 which is a bit mask representing the memory mapping types. If a bit
13013 is set in the bit mask, then the memory mappings of the corresponding
13014 types will be dumped; otherwise, they will be ignored. This
13015 configuration is inherited by child processes. For more information
13016 about the bits that can be set in the
13017 @file{/proc/@var{pid}/coredump_filter} file, please refer to the
13018 manpage of @code{core(5)}.
13020 By default, this option is @code{on}. If this option is turned
13021 @code{off}, @value{GDBN} does not read the @file{coredump_filter} file
13022 and instead uses the same default value as the Linux kernel in order
13023 to decide which pages will be dumped in the core dump file. This
13024 value is currently @code{0x33}, which means that bits @code{0}
13025 (anonymous private mappings), @code{1} (anonymous shared mappings),
13026 @code{4} (ELF headers) and @code{5} (private huge pages) are active.
13027 This will cause these memory mappings to be dumped automatically.
13029 @kindex set dump-excluded-mappings
13030 @anchor{set dump-excluded-mappings}
13031 @item set dump-excluded-mappings on
13032 @itemx set dump-excluded-mappings off
13033 If @code{on} is specified, @value{GDBN} will dump memory mappings
13034 marked with the @code{VM_DONTDUMP} flag. This flag is represented in
13035 the file @file{/proc/@var{pid}/smaps} with the acronym @code{dd}.
13037 The default value is @code{off}.
13040 @node Character Sets
13041 @section Character Sets
13042 @cindex character sets
13044 @cindex translating between character sets
13045 @cindex host character set
13046 @cindex target character set
13048 If the program you are debugging uses a different character set to
13049 represent characters and strings than the one @value{GDBN} uses itself,
13050 @value{GDBN} can automatically translate between the character sets for
13051 you. The character set @value{GDBN} uses we call the @dfn{host
13052 character set}; the one the inferior program uses we call the
13053 @dfn{target character set}.
13055 For example, if you are running @value{GDBN} on a @sc{gnu}/Linux system, which
13056 uses the ISO Latin 1 character set, but you are using @value{GDBN}'s
13057 remote protocol (@pxref{Remote Debugging}) to debug a program
13058 running on an IBM mainframe, which uses the @sc{ebcdic} character set,
13059 then the host character set is Latin-1, and the target character set is
13060 @sc{ebcdic}. If you give @value{GDBN} the command @code{set
13061 target-charset EBCDIC-US}, then @value{GDBN} translates between
13062 @sc{ebcdic} and Latin 1 as you print character or string values, or use
13063 character and string literals in expressions.
13065 @value{GDBN} has no way to automatically recognize which character set
13066 the inferior program uses; you must tell it, using the @code{set
13067 target-charset} command, described below.
13069 Here are the commands for controlling @value{GDBN}'s character set
13073 @item set target-charset @var{charset}
13074 @kindex set target-charset
13075 Set the current target character set to @var{charset}. To display the
13076 list of supported target character sets, type
13077 @kbd{@w{set target-charset @key{TAB}@key{TAB}}}.
13079 @item set host-charset @var{charset}
13080 @kindex set host-charset
13081 Set the current host character set to @var{charset}.
13083 By default, @value{GDBN} uses a host character set appropriate to the
13084 system it is running on; you can override that default using the
13085 @code{set host-charset} command. On some systems, @value{GDBN} cannot
13086 automatically determine the appropriate host character set. In this
13087 case, @value{GDBN} uses @samp{UTF-8}.
13089 @value{GDBN} can only use certain character sets as its host character
13090 set. If you type @kbd{@w{set host-charset @key{TAB}@key{TAB}}},
13091 @value{GDBN} will list the host character sets it supports.
13093 @item set charset @var{charset}
13094 @kindex set charset
13095 Set the current host and target character sets to @var{charset}. As
13096 above, if you type @kbd{@w{set charset @key{TAB}@key{TAB}}},
13097 @value{GDBN} will list the names of the character sets that can be used
13098 for both host and target.
13101 @kindex show charset
13102 Show the names of the current host and target character sets.
13104 @item show host-charset
13105 @kindex show host-charset
13106 Show the name of the current host character set.
13108 @item show target-charset
13109 @kindex show target-charset
13110 Show the name of the current target character set.
13112 @item set target-wide-charset @var{charset}
13113 @kindex set target-wide-charset
13114 Set the current target's wide character set to @var{charset}. This is
13115 the character set used by the target's @code{wchar_t} type. To
13116 display the list of supported wide character sets, type
13117 @kbd{@w{set target-wide-charset @key{TAB}@key{TAB}}}.
13119 @item show target-wide-charset
13120 @kindex show target-wide-charset
13121 Show the name of the current target's wide character set.
13124 Here is an example of @value{GDBN}'s character set support in action.
13125 Assume that the following source code has been placed in the file
13126 @file{charset-test.c}:
13132 = @{72, 101, 108, 108, 111, 44, 32, 119,
13133 111, 114, 108, 100, 33, 10, 0@};
13134 char ibm1047_hello[]
13135 = @{200, 133, 147, 147, 150, 107, 64, 166,
13136 150, 153, 147, 132, 90, 37, 0@};
13140 printf ("Hello, world!\n");
13144 In this program, @code{ascii_hello} and @code{ibm1047_hello} are arrays
13145 containing the string @samp{Hello, world!} followed by a newline,
13146 encoded in the @sc{ascii} and @sc{ibm1047} character sets.
13148 We compile the program, and invoke the debugger on it:
13151 $ gcc -g charset-test.c -o charset-test
13152 $ gdb -nw charset-test
13153 GNU gdb 2001-12-19-cvs
13154 Copyright 2001 Free Software Foundation, Inc.
13159 We can use the @code{show charset} command to see what character sets
13160 @value{GDBN} is currently using to interpret and display characters and
13164 (@value{GDBP}) show charset
13165 The current host and target character set is `ISO-8859-1'.
13169 For the sake of printing this manual, let's use @sc{ascii} as our
13170 initial character set:
13172 (@value{GDBP}) set charset ASCII
13173 (@value{GDBP}) show charset
13174 The current host and target character set is `ASCII'.
13178 Let's assume that @sc{ascii} is indeed the correct character set for our
13179 host system --- in other words, let's assume that if @value{GDBN} prints
13180 characters using the @sc{ascii} character set, our terminal will display
13181 them properly. Since our current target character set is also
13182 @sc{ascii}, the contents of @code{ascii_hello} print legibly:
13185 (@value{GDBP}) print ascii_hello
13186 $1 = 0x401698 "Hello, world!\n"
13187 (@value{GDBP}) print ascii_hello[0]
13192 @value{GDBN} uses the target character set for character and string
13193 literals you use in expressions:
13196 (@value{GDBP}) print '+'
13201 The @sc{ascii} character set uses the number 43 to encode the @samp{+}
13204 @value{GDBN} relies on the user to tell it which character set the
13205 target program uses. If we print @code{ibm1047_hello} while our target
13206 character set is still @sc{ascii}, we get jibberish:
13209 (@value{GDBP}) print ibm1047_hello
13210 $4 = 0x4016a8 "\310\205\223\223\226k@@\246\226\231\223\204Z%"
13211 (@value{GDBP}) print ibm1047_hello[0]
13216 If we invoke the @code{set target-charset} followed by @key{TAB}@key{TAB},
13217 @value{GDBN} tells us the character sets it supports:
13220 (@value{GDBP}) set target-charset
13221 ASCII EBCDIC-US IBM1047 ISO-8859-1
13222 (@value{GDBP}) set target-charset
13225 We can select @sc{ibm1047} as our target character set, and examine the
13226 program's strings again. Now the @sc{ascii} string is wrong, but
13227 @value{GDBN} translates the contents of @code{ibm1047_hello} from the
13228 target character set, @sc{ibm1047}, to the host character set,
13229 @sc{ascii}, and they display correctly:
13232 (@value{GDBP}) set target-charset IBM1047
13233 (@value{GDBP}) show charset
13234 The current host character set is `ASCII'.
13235 The current target character set is `IBM1047'.
13236 (@value{GDBP}) print ascii_hello
13237 $6 = 0x401698 "\110\145%%?\054\040\167?\162%\144\041\012"
13238 (@value{GDBP}) print ascii_hello[0]
13240 (@value{GDBP}) print ibm1047_hello
13241 $8 = 0x4016a8 "Hello, world!\n"
13242 (@value{GDBP}) print ibm1047_hello[0]
13247 As above, @value{GDBN} uses the target character set for character and
13248 string literals you use in expressions:
13251 (@value{GDBP}) print '+'
13256 The @sc{ibm1047} character set uses the number 78 to encode the @samp{+}
13259 @node Caching Target Data
13260 @section Caching Data of Targets
13261 @cindex caching data of targets
13263 @value{GDBN} caches data exchanged between the debugger and a target.
13264 Each cache is associated with the address space of the inferior.
13265 @xref{Inferiors and Programs}, about inferior and address space.
13266 Such caching generally improves performance in remote debugging
13267 (@pxref{Remote Debugging}), because it reduces the overhead of the
13268 remote protocol by bundling memory reads and writes into large chunks.
13269 Unfortunately, simply caching everything would lead to incorrect results,
13270 since @value{GDBN} does not necessarily know anything about volatile
13271 values, memory-mapped I/O addresses, etc. Furthermore, in non-stop mode
13272 (@pxref{Non-Stop Mode}) memory can be changed @emph{while} a gdb command
13274 Therefore, by default, @value{GDBN} only caches data
13275 known to be on the stack@footnote{In non-stop mode, it is moderately
13276 rare for a running thread to modify the stack of a stopped thread
13277 in a way that would interfere with a backtrace, and caching of
13278 stack reads provides a significant speed up of remote backtraces.} or
13279 in the code segment.
13280 Other regions of memory can be explicitly marked as
13281 cacheable; @pxref{Memory Region Attributes}.
13284 @kindex set remotecache
13285 @item set remotecache on
13286 @itemx set remotecache off
13287 This option no longer does anything; it exists for compatibility
13290 @kindex show remotecache
13291 @item show remotecache
13292 Show the current state of the obsolete remotecache flag.
13294 @kindex set stack-cache
13295 @item set stack-cache on
13296 @itemx set stack-cache off
13297 Enable or disable caching of stack accesses. When @code{on}, use
13298 caching. By default, this option is @code{on}.
13300 @kindex show stack-cache
13301 @item show stack-cache
13302 Show the current state of data caching for memory accesses.
13304 @kindex set code-cache
13305 @item set code-cache on
13306 @itemx set code-cache off
13307 Enable or disable caching of code segment accesses. When @code{on},
13308 use caching. By default, this option is @code{on}. This improves
13309 performance of disassembly in remote debugging.
13311 @kindex show code-cache
13312 @item show code-cache
13313 Show the current state of target memory cache for code segment
13316 @kindex info dcache
13317 @item info dcache @r{[}line@r{]}
13318 Print the information about the performance of data cache of the
13319 current inferior's address space. The information displayed
13320 includes the dcache width and depth, and for each cache line, its
13321 number, address, and how many times it was referenced. This
13322 command is useful for debugging the data cache operation.
13324 If a line number is specified, the contents of that line will be
13327 @item set dcache size @var{size}
13328 @cindex dcache size
13329 @kindex set dcache size
13330 Set maximum number of entries in dcache (dcache depth above).
13332 @item set dcache line-size @var{line-size}
13333 @cindex dcache line-size
13334 @kindex set dcache line-size
13335 Set number of bytes each dcache entry caches (dcache width above).
13336 Must be a power of 2.
13338 @item show dcache size
13339 @kindex show dcache size
13340 Show maximum number of dcache entries. @xref{Caching Target Data, info dcache}.
13342 @item show dcache line-size
13343 @kindex show dcache line-size
13344 Show default size of dcache lines.
13348 @node Searching Memory
13349 @section Search Memory
13350 @cindex searching memory
13352 Memory can be searched for a particular sequence of bytes with the
13353 @code{find} command.
13357 @item find @r{[}/@var{sn}@r{]} @var{start_addr}, +@var{len}, @var{val1} @r{[}, @var{val2}, @dots{}@r{]}
13358 @itemx find @r{[}/@var{sn}@r{]} @var{start_addr}, @var{end_addr}, @var{val1} @r{[}, @var{val2}, @dots{}@r{]}
13359 Search memory for the sequence of bytes specified by @var{val1}, @var{val2},
13360 etc. The search begins at address @var{start_addr} and continues for either
13361 @var{len} bytes or through to @var{end_addr} inclusive.
13364 @var{s} and @var{n} are optional parameters.
13365 They may be specified in either order, apart or together.
13368 @item @var{s}, search query size
13369 The size of each search query value.
13375 halfwords (two bytes)
13379 giant words (eight bytes)
13382 All values are interpreted in the current language.
13383 This means, for example, that if the current source language is C/C@t{++}
13384 then searching for the string ``hello'' includes the trailing '\0'.
13385 The null terminator can be removed from searching by using casts,
13386 e.g.: @samp{@{char[5]@}"hello"}.
13388 If the value size is not specified, it is taken from the
13389 value's type in the current language.
13390 This is useful when one wants to specify the search
13391 pattern as a mixture of types.
13392 Note that this means, for example, that in the case of C-like languages
13393 a search for an untyped 0x42 will search for @samp{(int) 0x42}
13394 which is typically four bytes.
13396 @item @var{n}, maximum number of finds
13397 The maximum number of matches to print. The default is to print all finds.
13400 You can use strings as search values. Quote them with double-quotes
13402 The string value is copied into the search pattern byte by byte,
13403 regardless of the endianness of the target and the size specification.
13405 The address of each match found is printed as well as a count of the
13406 number of matches found.
13408 The address of the last value found is stored in convenience variable
13410 A count of the number of matches is stored in @samp{$numfound}.
13412 For example, if stopped at the @code{printf} in this function:
13418 static char hello[] = "hello-hello";
13419 static struct @{ char c; short s; int i; @}
13420 __attribute__ ((packed)) mixed
13421 = @{ 'c', 0x1234, 0x87654321 @};
13422 printf ("%s\n", hello);
13427 you get during debugging:
13430 (@value{GDBP}) find &hello[0], +sizeof(hello), "hello"
13431 0x804956d <hello.1620+6>
13433 (@value{GDBP}) find &hello[0], +sizeof(hello), 'h', 'e', 'l', 'l', 'o'
13434 0x8049567 <hello.1620>
13435 0x804956d <hello.1620+6>
13437 (@value{GDBP}) find &hello[0], +sizeof(hello), @{char[5]@}"hello"
13438 0x8049567 <hello.1620>
13439 0x804956d <hello.1620+6>
13441 (@value{GDBP}) find /b1 &hello[0], +sizeof(hello), 'h', 0x65, 'l'
13442 0x8049567 <hello.1620>
13444 (@value{GDBP}) find &mixed, +sizeof(mixed), (char) 'c', (short) 0x1234, (int) 0x87654321
13445 0x8049560 <mixed.1625>
13447 (@value{GDBP}) print $numfound
13449 (@value{GDBP}) print $_
13450 $2 = (void *) 0x8049560
13454 @section Value Sizes
13456 Whenever @value{GDBN} prints a value memory will be allocated within
13457 @value{GDBN} to hold the contents of the value. It is possible in
13458 some languages with dynamic typing systems, that an invalid program
13459 may indicate a value that is incorrectly large, this in turn may cause
13460 @value{GDBN} to try and allocate an overly large amount of memory.
13463 @kindex set max-value-size
13464 @item set max-value-size @var{bytes}
13465 @itemx set max-value-size unlimited
13466 Set the maximum size of memory that @value{GDBN} will allocate for the
13467 contents of a value to @var{bytes}, trying to display a value that
13468 requires more memory than that will result in an error.
13470 Setting this variable does not effect values that have already been
13471 allocated within @value{GDBN}, only future allocations.
13473 There's a minimum size that @code{max-value-size} can be set to in
13474 order that @value{GDBN} can still operate correctly, this minimum is
13475 currently 16 bytes.
13477 The limit applies to the results of some subexpressions as well as to
13478 complete expressions. For example, an expression denoting a simple
13479 integer component, such as @code{x.y.z}, may fail if the size of
13480 @var{x.y} is dynamic and exceeds @var{bytes}. On the other hand,
13481 @value{GDBN} is sometimes clever; the expression @code{A[i]}, where
13482 @var{A} is an array variable with non-constant size, will generally
13483 succeed regardless of the bounds on @var{A}, as long as the component
13484 size is less than @var{bytes}.
13486 The default value of @code{max-value-size} is currently 64k.
13488 @kindex show max-value-size
13489 @item show max-value-size
13490 Show the maximum size of memory, in bytes, that @value{GDBN} will
13491 allocate for the contents of a value.
13494 @node Optimized Code
13495 @chapter Debugging Optimized Code
13496 @cindex optimized code, debugging
13497 @cindex debugging optimized code
13499 Almost all compilers support optimization. With optimization
13500 disabled, the compiler generates assembly code that corresponds
13501 directly to your source code, in a simplistic way. As the compiler
13502 applies more powerful optimizations, the generated assembly code
13503 diverges from your original source code. With help from debugging
13504 information generated by the compiler, @value{GDBN} can map from
13505 the running program back to constructs from your original source.
13507 @value{GDBN} is more accurate with optimization disabled. If you
13508 can recompile without optimization, it is easier to follow the
13509 progress of your program during debugging. But, there are many cases
13510 where you may need to debug an optimized version.
13512 When you debug a program compiled with @samp{-g -O}, remember that the
13513 optimizer has rearranged your code; the debugger shows you what is
13514 really there. Do not be too surprised when the execution path does not
13515 exactly match your source file! An extreme example: if you define a
13516 variable, but never use it, @value{GDBN} never sees that
13517 variable---because the compiler optimizes it out of existence.
13519 Some things do not work as well with @samp{-g -O} as with just
13520 @samp{-g}, particularly on machines with instruction scheduling. If in
13521 doubt, recompile with @samp{-g} alone, and if this fixes the problem,
13522 please report it to us as a bug (including a test case!).
13523 @xref{Variables}, for more information about debugging optimized code.
13526 * Inline Functions:: How @value{GDBN} presents inlining
13527 * Tail Call Frames:: @value{GDBN} analysis of jumps to functions
13530 @node Inline Functions
13531 @section Inline Functions
13532 @cindex inline functions, debugging
13534 @dfn{Inlining} is an optimization that inserts a copy of the function
13535 body directly at each call site, instead of jumping to a shared
13536 routine. @value{GDBN} displays inlined functions just like
13537 non-inlined functions. They appear in backtraces. You can view their
13538 arguments and local variables, step into them with @code{step}, skip
13539 them with @code{next}, and escape from them with @code{finish}.
13540 You can check whether a function was inlined by using the
13541 @code{info frame} command.
13543 For @value{GDBN} to support inlined functions, the compiler must
13544 record information about inlining in the debug information ---
13545 @value{NGCC} using the @sc{dwarf 2} format does this, and several
13546 other compilers do also. @value{GDBN} only supports inlined functions
13547 when using @sc{dwarf 2}. Versions of @value{NGCC} before 4.1
13548 do not emit two required attributes (@samp{DW_AT_call_file} and
13549 @samp{DW_AT_call_line}); @value{GDBN} does not display inlined
13550 function calls with earlier versions of @value{NGCC}. It instead
13551 displays the arguments and local variables of inlined functions as
13552 local variables in the caller.
13554 The body of an inlined function is directly included at its call site;
13555 unlike a non-inlined function, there are no instructions devoted to
13556 the call. @value{GDBN} still pretends that the call site and the
13557 start of the inlined function are different instructions. Stepping to
13558 the call site shows the call site, and then stepping again shows
13559 the first line of the inlined function, even though no additional
13560 instructions are executed.
13562 This makes source-level debugging much clearer; you can see both the
13563 context of the call and then the effect of the call. Only stepping by
13564 a single instruction using @code{stepi} or @code{nexti} does not do
13565 this; single instruction steps always show the inlined body.
13567 There are some ways that @value{GDBN} does not pretend that inlined
13568 function calls are the same as normal calls:
13572 Setting breakpoints at the call site of an inlined function may not
13573 work, because the call site does not contain any code. @value{GDBN}
13574 may incorrectly move the breakpoint to the next line of the enclosing
13575 function, after the call. This limitation will be removed in a future
13576 version of @value{GDBN}; until then, set a breakpoint on an earlier line
13577 or inside the inlined function instead.
13580 @value{GDBN} cannot locate the return value of inlined calls after
13581 using the @code{finish} command. This is a limitation of compiler-generated
13582 debugging information; after @code{finish}, you can step to the next line
13583 and print a variable where your program stored the return value.
13587 @node Tail Call Frames
13588 @section Tail Call Frames
13589 @cindex tail call frames, debugging
13591 Function @code{B} can call function @code{C} in its very last statement. In
13592 unoptimized compilation the call of @code{C} is immediately followed by return
13593 instruction at the end of @code{B} code. Optimizing compiler may replace the
13594 call and return in function @code{B} into one jump to function @code{C}
13595 instead. Such use of a jump instruction is called @dfn{tail call}.
13597 During execution of function @code{C}, there will be no indication in the
13598 function call stack frames that it was tail-called from @code{B}. If function
13599 @code{A} regularly calls function @code{B} which tail-calls function @code{C},
13600 then @value{GDBN} will see @code{A} as the caller of @code{C}. However, in
13601 some cases @value{GDBN} can determine that @code{C} was tail-called from
13602 @code{B}, and it will then create fictitious call frame for that, with the
13603 return address set up as if @code{B} called @code{C} normally.
13605 This functionality is currently supported only by DWARF 2 debugging format and
13606 the compiler has to produce @samp{DW_TAG_call_site} tags. With
13607 @value{NGCC}, you need to specify @option{-O -g} during compilation, to get
13610 @kbd{info frame} command (@pxref{Frame Info}) will indicate the tail call frame
13611 kind by text @code{tail call frame} such as in this sample @value{GDBN} output:
13614 (@value{GDBP}) x/i $pc - 2
13615 0x40066b <b(int, double)+11>: jmp 0x400640 <c(int, double)>
13616 (@value{GDBP}) info frame
13617 Stack level 1, frame at 0x7fffffffda30:
13618 rip = 0x40066d in b (amd64-entry-value.cc:59); saved rip 0x4004c5
13619 tail call frame, caller of frame at 0x7fffffffda30
13620 source language c++.
13621 Arglist at unknown address.
13622 Locals at unknown address, Previous frame's sp is 0x7fffffffda30
13625 The detection of all the possible code path executions can find them ambiguous.
13626 There is no execution history stored (possible @ref{Reverse Execution} is never
13627 used for this purpose) and the last known caller could have reached the known
13628 callee by multiple different jump sequences. In such case @value{GDBN} still
13629 tries to show at least all the unambiguous top tail callers and all the
13630 unambiguous bottom tail calees, if any.
13633 @anchor{set debug entry-values}
13634 @item set debug entry-values
13635 @kindex set debug entry-values
13636 When set to on, enables printing of analysis messages for both frame argument
13637 values at function entry and tail calls. It will show all the possible valid
13638 tail calls code paths it has considered. It will also print the intersection
13639 of them with the final unambiguous (possibly partial or even empty) code path
13642 @item show debug entry-values
13643 @kindex show debug entry-values
13644 Show the current state of analysis messages printing for both frame argument
13645 values at function entry and tail calls.
13648 The analysis messages for tail calls can for example show why the virtual tail
13649 call frame for function @code{c} has not been recognized (due to the indirect
13650 reference by variable @code{x}):
13653 static void __attribute__((noinline, noclone)) c (void);
13654 void (*x) (void) = c;
13655 static void __attribute__((noinline, noclone)) a (void) @{ x++; @}
13656 static void __attribute__((noinline, noclone)) c (void) @{ a (); @}
13657 int main (void) @{ x (); return 0; @}
13659 Breakpoint 1, DW_OP_entry_value resolving cannot find
13660 DW_TAG_call_site 0x40039a in main
13662 3 static void __attribute__((noinline, noclone)) a (void) @{ x++; @}
13665 #1 0x000000000040039a in main () at t.c:5
13668 Another possibility is an ambiguous virtual tail call frames resolution:
13672 static void __attribute__((noinline, noclone)) f (void) @{ i++; @}
13673 static void __attribute__((noinline, noclone)) e (void) @{ f (); @}
13674 static void __attribute__((noinline, noclone)) d (void) @{ f (); @}
13675 static void __attribute__((noinline, noclone)) c (void) @{ d (); @}
13676 static void __attribute__((noinline, noclone)) b (void)
13677 @{ if (i) c (); else e (); @}
13678 static void __attribute__((noinline, noclone)) a (void) @{ b (); @}
13679 int main (void) @{ a (); return 0; @}
13681 tailcall: initial: 0x4004d2(a) 0x4004ce(b) 0x4004b2(c) 0x4004a2(d)
13682 tailcall: compare: 0x4004d2(a) 0x4004cc(b) 0x400492(e)
13683 tailcall: reduced: 0x4004d2(a) |
13686 #1 0x00000000004004d2 in a () at t.c:8
13687 #2 0x0000000000400395 in main () at t.c:9
13690 @set CALLSEQ1A @code{main@value{ARROW}a@value{ARROW}b@value{ARROW}c@value{ARROW}d@value{ARROW}f}
13691 @set CALLSEQ2A @code{main@value{ARROW}a@value{ARROW}b@value{ARROW}e@value{ARROW}f}
13693 @c Convert CALLSEQ#A to CALLSEQ#B depending on HAVE_MAKEINFO_CLICK.
13694 @ifset HAVE_MAKEINFO_CLICK
13695 @set ARROW @click{}
13696 @set CALLSEQ1B @clicksequence{@value{CALLSEQ1A}}
13697 @set CALLSEQ2B @clicksequence{@value{CALLSEQ2A}}
13699 @ifclear HAVE_MAKEINFO_CLICK
13701 @set CALLSEQ1B @value{CALLSEQ1A}
13702 @set CALLSEQ2B @value{CALLSEQ2A}
13705 Frames #0 and #2 are real, #1 is a virtual tail call frame.
13706 The code can have possible execution paths @value{CALLSEQ1B} or
13707 @value{CALLSEQ2B}, @value{GDBN} cannot find which one from the inferior state.
13709 @code{initial:} state shows some random possible calling sequence @value{GDBN}
13710 has found. It then finds another possible calling sequence - that one is
13711 prefixed by @code{compare:}. The non-ambiguous intersection of these two is
13712 printed as the @code{reduced:} calling sequence. That one could have many
13713 further @code{compare:} and @code{reduced:} statements as long as there remain
13714 any non-ambiguous sequence entries.
13716 For the frame of function @code{b} in both cases there are different possible
13717 @code{$pc} values (@code{0x4004cc} or @code{0x4004ce}), therefore this frame is
13718 also ambiguous. The only non-ambiguous frame is the one for function @code{a},
13719 therefore this one is displayed to the user while the ambiguous frames are
13722 There can be also reasons why printing of frame argument values at function
13727 static void __attribute__((noinline, noclone)) c (int i) @{ v++; @}
13728 static void __attribute__((noinline, noclone)) a (int i);
13729 static void __attribute__((noinline, noclone)) b (int i) @{ a (i); @}
13730 static void __attribute__((noinline, noclone)) a (int i)
13731 @{ if (i) b (i - 1); else c (0); @}
13732 int main (void) @{ a (5); return 0; @}
13735 #0 c (i=i@@entry=0) at t.c:2
13736 #1 0x0000000000400428 in a (DW_OP_entry_value resolving has found
13737 function "a" at 0x400420 can call itself via tail calls
13738 i=<optimized out>) at t.c:6
13739 #2 0x000000000040036e in main () at t.c:7
13742 @value{GDBN} cannot find out from the inferior state if and how many times did
13743 function @code{a} call itself (via function @code{b}) as these calls would be
13744 tail calls. Such tail calls would modify the @code{i} variable, therefore
13745 @value{GDBN} cannot be sure the value it knows would be right - @value{GDBN}
13746 prints @code{<optimized out>} instead.
13749 @chapter C Preprocessor Macros
13751 Some languages, such as C and C@t{++}, provide a way to define and invoke
13752 ``preprocessor macros'' which expand into strings of tokens.
13753 @value{GDBN} can evaluate expressions containing macro invocations, show
13754 the result of macro expansion, and show a macro's definition, including
13755 where it was defined.
13757 You may need to compile your program specially to provide @value{GDBN}
13758 with information about preprocessor macros. Most compilers do not
13759 include macros in their debugging information, even when you compile
13760 with the @option{-g} flag. @xref{Compilation}.
13762 A program may define a macro at one point, remove that definition later,
13763 and then provide a different definition after that. Thus, at different
13764 points in the program, a macro may have different definitions, or have
13765 no definition at all. If there is a current stack frame, @value{GDBN}
13766 uses the macros in scope at that frame's source code line. Otherwise,
13767 @value{GDBN} uses the macros in scope at the current listing location;
13770 Whenever @value{GDBN} evaluates an expression, it always expands any
13771 macro invocations present in the expression. @value{GDBN} also provides
13772 the following commands for working with macros explicitly.
13776 @kindex macro expand
13777 @cindex macro expansion, showing the results of preprocessor
13778 @cindex preprocessor macro expansion, showing the results of
13779 @cindex expanding preprocessor macros
13780 @item macro expand @var{expression}
13781 @itemx macro exp @var{expression}
13782 Show the results of expanding all preprocessor macro invocations in
13783 @var{expression}. Since @value{GDBN} simply expands macros, but does
13784 not parse the result, @var{expression} need not be a valid expression;
13785 it can be any string of tokens.
13788 @item macro expand-once @var{expression}
13789 @itemx macro exp1 @var{expression}
13790 @cindex expand macro once
13791 @i{(This command is not yet implemented.)} Show the results of
13792 expanding those preprocessor macro invocations that appear explicitly in
13793 @var{expression}. Macro invocations appearing in that expansion are
13794 left unchanged. This command allows you to see the effect of a
13795 particular macro more clearly, without being confused by further
13796 expansions. Since @value{GDBN} simply expands macros, but does not
13797 parse the result, @var{expression} need not be a valid expression; it
13798 can be any string of tokens.
13801 @cindex macro definition, showing
13802 @cindex definition of a macro, showing
13803 @cindex macros, from debug info
13804 @item info macro [-a|-all] [--] @var{macro}
13805 Show the current definition or all definitions of the named @var{macro},
13806 and describe the source location or compiler command-line where that
13807 definition was established. The optional double dash is to signify the end of
13808 argument processing and the beginning of @var{macro} for non C-like macros where
13809 the macro may begin with a hyphen.
13811 @kindex info macros
13812 @item info macros @var{location}
13813 Show all macro definitions that are in effect at the location specified
13814 by @var{location}, and describe the source location or compiler
13815 command-line where those definitions were established.
13817 @kindex macro define
13818 @cindex user-defined macros
13819 @cindex defining macros interactively
13820 @cindex macros, user-defined
13821 @item macro define @var{macro} @var{replacement-list}
13822 @itemx macro define @var{macro}(@var{arglist}) @var{replacement-list}
13823 Introduce a definition for a preprocessor macro named @var{macro},
13824 invocations of which are replaced by the tokens given in
13825 @var{replacement-list}. The first form of this command defines an
13826 ``object-like'' macro, which takes no arguments; the second form
13827 defines a ``function-like'' macro, which takes the arguments given in
13830 A definition introduced by this command is in scope in every
13831 expression evaluated in @value{GDBN}, until it is removed with the
13832 @code{macro undef} command, described below. The definition overrides
13833 all definitions for @var{macro} present in the program being debugged,
13834 as well as any previous user-supplied definition.
13836 @kindex macro undef
13837 @item macro undef @var{macro}
13838 Remove any user-supplied definition for the macro named @var{macro}.
13839 This command only affects definitions provided with the @code{macro
13840 define} command, described above; it cannot remove definitions present
13841 in the program being debugged.
13845 List all the macros defined using the @code{macro define} command.
13848 @cindex macros, example of debugging with
13849 Here is a transcript showing the above commands in action. First, we
13850 show our source files:
13855 #include "sample.h"
13858 #define ADD(x) (M + x)
13863 printf ("Hello, world!\n");
13865 printf ("We're so creative.\n");
13867 printf ("Goodbye, world!\n");
13874 Now, we compile the program using the @sc{gnu} C compiler,
13875 @value{NGCC}. We pass the @option{-gdwarf-2}@footnote{This is the
13876 minimum. Recent versions of @value{NGCC} support @option{-gdwarf-3}
13877 and @option{-gdwarf-4}; we recommend always choosing the most recent
13878 version of DWARF.} @emph{and} @option{-g3} flags to ensure the compiler
13879 includes information about preprocessor macros in the debugging
13883 $ gcc -gdwarf-2 -g3 sample.c -o sample
13887 Now, we start @value{GDBN} on our sample program:
13891 GNU gdb 2002-05-06-cvs
13892 Copyright 2002 Free Software Foundation, Inc.
13893 GDB is free software, @dots{}
13897 We can expand macros and examine their definitions, even when the
13898 program is not running. @value{GDBN} uses the current listing position
13899 to decide which macro definitions are in scope:
13902 (@value{GDBP}) list main
13905 5 #define ADD(x) (M + x)
13910 10 printf ("Hello, world!\n");
13912 12 printf ("We're so creative.\n");
13913 (@value{GDBP}) info macro ADD
13914 Defined at /home/jimb/gdb/macros/play/sample.c:5
13915 #define ADD(x) (M + x)
13916 (@value{GDBP}) info macro Q
13917 Defined at /home/jimb/gdb/macros/play/sample.h:1
13918 included at /home/jimb/gdb/macros/play/sample.c:2
13920 (@value{GDBP}) macro expand ADD(1)
13921 expands to: (42 + 1)
13922 (@value{GDBP}) macro expand-once ADD(1)
13923 expands to: once (M + 1)
13927 In the example above, note that @code{macro expand-once} expands only
13928 the macro invocation explicit in the original text --- the invocation of
13929 @code{ADD} --- but does not expand the invocation of the macro @code{M},
13930 which was introduced by @code{ADD}.
13932 Once the program is running, @value{GDBN} uses the macro definitions in
13933 force at the source line of the current stack frame:
13936 (@value{GDBP}) break main
13937 Breakpoint 1 at 0x8048370: file sample.c, line 10.
13939 Starting program: /home/jimb/gdb/macros/play/sample
13941 Breakpoint 1, main () at sample.c:10
13942 10 printf ("Hello, world!\n");
13946 At line 10, the definition of the macro @code{N} at line 9 is in force:
13949 (@value{GDBP}) info macro N
13950 Defined at /home/jimb/gdb/macros/play/sample.c:9
13952 (@value{GDBP}) macro expand N Q M
13953 expands to: 28 < 42
13954 (@value{GDBP}) print N Q M
13959 As we step over directives that remove @code{N}'s definition, and then
13960 give it a new definition, @value{GDBN} finds the definition (or lack
13961 thereof) in force at each point:
13964 (@value{GDBP}) next
13966 12 printf ("We're so creative.\n");
13967 (@value{GDBP}) info macro N
13968 The symbol `N' has no definition as a C/C++ preprocessor macro
13969 at /home/jimb/gdb/macros/play/sample.c:12
13970 (@value{GDBP}) next
13972 14 printf ("Goodbye, world!\n");
13973 (@value{GDBP}) info macro N
13974 Defined at /home/jimb/gdb/macros/play/sample.c:13
13976 (@value{GDBP}) macro expand N Q M
13977 expands to: 1729 < 42
13978 (@value{GDBP}) print N Q M
13983 In addition to source files, macros can be defined on the compilation command
13984 line using the @option{-D@var{name}=@var{value}} syntax. For macros defined in
13985 such a way, @value{GDBN} displays the location of their definition as line zero
13986 of the source file submitted to the compiler.
13989 (@value{GDBP}) info macro __STDC__
13990 Defined at /home/jimb/gdb/macros/play/sample.c:0
13997 @chapter Tracepoints
13998 @c This chapter is based on the documentation written by Michael
13999 @c Snyder, David Taylor, Jim Blandy, and Elena Zannoni.
14001 @cindex tracepoints
14002 In some applications, it is not feasible for the debugger to interrupt
14003 the program's execution long enough for the developer to learn
14004 anything helpful about its behavior. If the program's correctness
14005 depends on its real-time behavior, delays introduced by a debugger
14006 might cause the program to change its behavior drastically, or perhaps
14007 fail, even when the code itself is correct. It is useful to be able
14008 to observe the program's behavior without interrupting it.
14010 Using @value{GDBN}'s @code{trace} and @code{collect} commands, you can
14011 specify locations in the program, called @dfn{tracepoints}, and
14012 arbitrary expressions to evaluate when those tracepoints are reached.
14013 Later, using the @code{tfind} command, you can examine the values
14014 those expressions had when the program hit the tracepoints. The
14015 expressions may also denote objects in memory---structures or arrays,
14016 for example---whose values @value{GDBN} should record; while visiting
14017 a particular tracepoint, you may inspect those objects as if they were
14018 in memory at that moment. However, because @value{GDBN} records these
14019 values without interacting with you, it can do so quickly and
14020 unobtrusively, hopefully not disturbing the program's behavior.
14022 The tracepoint facility is currently available only for remote
14023 targets. @xref{Targets}. In addition, your remote target must know
14024 how to collect trace data. This functionality is implemented in the
14025 remote stub; however, none of the stubs distributed with @value{GDBN}
14026 support tracepoints as of this writing. The format of the remote
14027 packets used to implement tracepoints are described in @ref{Tracepoint
14030 It is also possible to get trace data from a file, in a manner reminiscent
14031 of corefiles; you specify the filename, and use @code{tfind} to search
14032 through the file. @xref{Trace Files}, for more details.
14034 This chapter describes the tracepoint commands and features.
14037 * Set Tracepoints::
14038 * Analyze Collected Data::
14039 * Tracepoint Variables::
14043 @node Set Tracepoints
14044 @section Commands to Set Tracepoints
14046 Before running such a @dfn{trace experiment}, an arbitrary number of
14047 tracepoints can be set. A tracepoint is actually a special type of
14048 breakpoint (@pxref{Set Breaks}), so you can manipulate it using
14049 standard breakpoint commands. For instance, as with breakpoints,
14050 tracepoint numbers are successive integers starting from one, and many
14051 of the commands associated with tracepoints take the tracepoint number
14052 as their argument, to identify which tracepoint to work on.
14054 For each tracepoint, you can specify, in advance, some arbitrary set
14055 of data that you want the target to collect in the trace buffer when
14056 it hits that tracepoint. The collected data can include registers,
14057 local variables, or global data. Later, you can use @value{GDBN}
14058 commands to examine the values these data had at the time the
14059 tracepoint was hit.
14061 Tracepoints do not support every breakpoint feature. Ignore counts on
14062 tracepoints have no effect, and tracepoints cannot run @value{GDBN}
14063 commands when they are hit. Tracepoints may not be thread-specific
14066 @cindex fast tracepoints
14067 Some targets may support @dfn{fast tracepoints}, which are inserted in
14068 a different way (such as with a jump instead of a trap), that is
14069 faster but possibly restricted in where they may be installed.
14071 @cindex static tracepoints
14072 @cindex markers, static tracepoints
14073 @cindex probing markers, static tracepoints
14074 Regular and fast tracepoints are dynamic tracing facilities, meaning
14075 that they can be used to insert tracepoints at (almost) any location
14076 in the target. Some targets may also support controlling @dfn{static
14077 tracepoints} from @value{GDBN}. With static tracing, a set of
14078 instrumentation points, also known as @dfn{markers}, are embedded in
14079 the target program, and can be activated or deactivated by name or
14080 address. These are usually placed at locations which facilitate
14081 investigating what the target is actually doing. @value{GDBN}'s
14082 support for static tracing includes being able to list instrumentation
14083 points, and attach them with @value{GDBN} defined high level
14084 tracepoints that expose the whole range of convenience of
14085 @value{GDBN}'s tracepoints support. Namely, support for collecting
14086 registers values and values of global or local (to the instrumentation
14087 point) variables; tracepoint conditions and trace state variables.
14088 The act of installing a @value{GDBN} static tracepoint on an
14089 instrumentation point, or marker, is referred to as @dfn{probing} a
14090 static tracepoint marker.
14092 @code{gdbserver} supports tracepoints on some target systems.
14093 @xref{Server,,Tracepoints support in @code{gdbserver}}.
14095 This section describes commands to set tracepoints and associated
14096 conditions and actions.
14099 * Create and Delete Tracepoints::
14100 * Enable and Disable Tracepoints::
14101 * Tracepoint Passcounts::
14102 * Tracepoint Conditions::
14103 * Trace State Variables::
14104 * Tracepoint Actions::
14105 * Listing Tracepoints::
14106 * Listing Static Tracepoint Markers::
14107 * Starting and Stopping Trace Experiments::
14108 * Tracepoint Restrictions::
14111 @node Create and Delete Tracepoints
14112 @subsection Create and Delete Tracepoints
14115 @cindex set tracepoint
14117 @item trace @var{location}
14118 The @code{trace} command is very similar to the @code{break} command.
14119 Its argument @var{location} can be any valid location.
14120 @xref{Specify Location}. The @code{trace} command defines a tracepoint,
14121 which is a point in the target program where the debugger will briefly stop,
14122 collect some data, and then allow the program to continue. Setting a tracepoint
14123 or changing its actions takes effect immediately if the remote stub
14124 supports the @samp{InstallInTrace} feature (@pxref{install tracepoint
14126 If remote stub doesn't support the @samp{InstallInTrace} feature, all
14127 these changes don't take effect until the next @code{tstart}
14128 command, and once a trace experiment is running, further changes will
14129 not have any effect until the next trace experiment starts. In addition,
14130 @value{GDBN} supports @dfn{pending tracepoints}---tracepoints whose
14131 address is not yet resolved. (This is similar to pending breakpoints.)
14132 Pending tracepoints are not downloaded to the target and not installed
14133 until they are resolved. The resolution of pending tracepoints requires
14134 @value{GDBN} support---when debugging with the remote target, and
14135 @value{GDBN} disconnects from the remote stub (@pxref{disconnected
14136 tracing}), pending tracepoints can not be resolved (and downloaded to
14137 the remote stub) while @value{GDBN} is disconnected.
14139 Here are some examples of using the @code{trace} command:
14142 (@value{GDBP}) @b{trace foo.c:121} // a source file and line number
14144 (@value{GDBP}) @b{trace +2} // 2 lines forward
14146 (@value{GDBP}) @b{trace my_function} // first source line of function
14148 (@value{GDBP}) @b{trace *my_function} // EXACT start address of function
14150 (@value{GDBP}) @b{trace *0x2117c4} // an address
14154 You can abbreviate @code{trace} as @code{tr}.
14156 @item trace @var{location} if @var{cond}
14157 Set a tracepoint with condition @var{cond}; evaluate the expression
14158 @var{cond} each time the tracepoint is reached, and collect data only
14159 if the value is nonzero---that is, if @var{cond} evaluates as true.
14160 @xref{Tracepoint Conditions, ,Tracepoint Conditions}, for more
14161 information on tracepoint conditions.
14163 @item ftrace @var{location} [ if @var{cond} ]
14164 @cindex set fast tracepoint
14165 @cindex fast tracepoints, setting
14167 The @code{ftrace} command sets a fast tracepoint. For targets that
14168 support them, fast tracepoints will use a more efficient but possibly
14169 less general technique to trigger data collection, such as a jump
14170 instruction instead of a trap, or some sort of hardware support. It
14171 may not be possible to create a fast tracepoint at the desired
14172 location, in which case the command will exit with an explanatory
14175 @value{GDBN} handles arguments to @code{ftrace} exactly as for
14178 On 32-bit x86-architecture systems, fast tracepoints normally need to
14179 be placed at an instruction that is 5 bytes or longer, but can be
14180 placed at 4-byte instructions if the low 64K of memory of the target
14181 program is available to install trampolines. Some Unix-type systems,
14182 such as @sc{gnu}/Linux, exclude low addresses from the program's
14183 address space; but for instance with the Linux kernel it is possible
14184 to let @value{GDBN} use this area by doing a @command{sysctl} command
14185 to set the @code{mmap_min_addr} kernel parameter, as in
14188 sudo sysctl -w vm.mmap_min_addr=32768
14192 which sets the low address to 32K, which leaves plenty of room for
14193 trampolines. The minimum address should be set to a page boundary.
14195 @item strace @var{location} [ if @var{cond} ]
14196 @cindex set static tracepoint
14197 @cindex static tracepoints, setting
14198 @cindex probe static tracepoint marker
14200 The @code{strace} command sets a static tracepoint. For targets that
14201 support it, setting a static tracepoint probes a static
14202 instrumentation point, or marker, found at @var{location}. It may not
14203 be possible to set a static tracepoint at the desired location, in
14204 which case the command will exit with an explanatory message.
14206 @value{GDBN} handles arguments to @code{strace} exactly as for
14207 @code{trace}, with the addition that the user can also specify
14208 @code{-m @var{marker}} as @var{location}. This probes the marker
14209 identified by the @var{marker} string identifier. This identifier
14210 depends on the static tracepoint backend library your program is
14211 using. You can find all the marker identifiers in the @samp{ID} field
14212 of the @code{info static-tracepoint-markers} command output.
14213 @xref{Listing Static Tracepoint Markers,,Listing Static Tracepoint
14214 Markers}. For example, in the following small program using the UST
14220 trace_mark(ust, bar33, "str %s", "FOOBAZ");
14225 the marker id is composed of joining the first two arguments to the
14226 @code{trace_mark} call with a slash, which translates to:
14229 (@value{GDBP}) info static-tracepoint-markers
14230 Cnt Enb ID Address What
14231 1 n ust/bar33 0x0000000000400ddc in main at stexample.c:22
14237 so you may probe the marker above with:
14240 (@value{GDBP}) strace -m ust/bar33
14243 Static tracepoints accept an extra collect action --- @code{collect
14244 $_sdata}. This collects arbitrary user data passed in the probe point
14245 call to the tracing library. In the UST example above, you'll see
14246 that the third argument to @code{trace_mark} is a printf-like format
14247 string. The user data is then the result of running that formatting
14248 string against the following arguments. Note that @code{info
14249 static-tracepoint-markers} command output lists that format string in
14250 the @samp{Data:} field.
14252 You can inspect this data when analyzing the trace buffer, by printing
14253 the $_sdata variable like any other variable available to
14254 @value{GDBN}. @xref{Tracepoint Actions,,Tracepoint Action Lists}.
14257 @cindex last tracepoint number
14258 @cindex recent tracepoint number
14259 @cindex tracepoint number
14260 The convenience variable @code{$tpnum} records the tracepoint number
14261 of the most recently set tracepoint.
14263 @kindex delete tracepoint
14264 @cindex tracepoint deletion
14265 @item delete tracepoint @r{[}@var{num}@r{]}
14266 Permanently delete one or more tracepoints. With no argument, the
14267 default is to delete all tracepoints. Note that the regular
14268 @code{delete} command can remove tracepoints also.
14273 (@value{GDBP}) @b{delete trace 1 2 3} // remove three tracepoints
14275 (@value{GDBP}) @b{delete trace} // remove all tracepoints
14279 You can abbreviate this command as @code{del tr}.
14282 @node Enable and Disable Tracepoints
14283 @subsection Enable and Disable Tracepoints
14285 These commands are deprecated; they are equivalent to plain @code{disable} and @code{enable}.
14288 @kindex disable tracepoint
14289 @item disable tracepoint @r{[}@var{num}@r{]}
14290 Disable tracepoint @var{num}, or all tracepoints if no argument
14291 @var{num} is given. A disabled tracepoint will have no effect during
14292 a trace experiment, but it is not forgotten. You can re-enable
14293 a disabled tracepoint using the @code{enable tracepoint} command.
14294 If the command is issued during a trace experiment and the debug target
14295 has support for disabling tracepoints during a trace experiment, then the
14296 change will be effective immediately. Otherwise, it will be applied to the
14297 next trace experiment.
14299 @kindex enable tracepoint
14300 @item enable tracepoint @r{[}@var{num}@r{]}
14301 Enable tracepoint @var{num}, or all tracepoints. If this command is
14302 issued during a trace experiment and the debug target supports enabling
14303 tracepoints during a trace experiment, then the enabled tracepoints will
14304 become effective immediately. Otherwise, they will become effective the
14305 next time a trace experiment is run.
14308 @node Tracepoint Passcounts
14309 @subsection Tracepoint Passcounts
14313 @cindex tracepoint pass count
14314 @item passcount @r{[}@var{n} @r{[}@var{num}@r{]]}
14315 Set the @dfn{passcount} of a tracepoint. The passcount is a way to
14316 automatically stop a trace experiment. If a tracepoint's passcount is
14317 @var{n}, then the trace experiment will be automatically stopped on
14318 the @var{n}'th time that tracepoint is hit. If the tracepoint number
14319 @var{num} is not specified, the @code{passcount} command sets the
14320 passcount of the most recently defined tracepoint. If no passcount is
14321 given, the trace experiment will run until stopped explicitly by the
14327 (@value{GDBP}) @b{passcount 5 2} // Stop on the 5th execution of
14328 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// tracepoint 2}
14330 (@value{GDBP}) @b{passcount 12} // Stop on the 12th execution of the
14331 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// most recently defined tracepoint.}
14332 (@value{GDBP}) @b{trace foo}
14333 (@value{GDBP}) @b{pass 3}
14334 (@value{GDBP}) @b{trace bar}
14335 (@value{GDBP}) @b{pass 2}
14336 (@value{GDBP}) @b{trace baz}
14337 (@value{GDBP}) @b{pass 1} // Stop tracing when foo has been
14338 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// executed 3 times OR when bar has}
14339 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// been executed 2 times}
14340 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// OR when baz has been executed 1 time.}
14344 @node Tracepoint Conditions
14345 @subsection Tracepoint Conditions
14346 @cindex conditional tracepoints
14347 @cindex tracepoint conditions
14349 The simplest sort of tracepoint collects data every time your program
14350 reaches a specified place. You can also specify a @dfn{condition} for
14351 a tracepoint. A condition is just a Boolean expression in your
14352 programming language (@pxref{Expressions, ,Expressions}). A
14353 tracepoint with a condition evaluates the expression each time your
14354 program reaches it, and data collection happens only if the condition
14357 Tracepoint conditions can be specified when a tracepoint is set, by
14358 using @samp{if} in the arguments to the @code{trace} command.
14359 @xref{Create and Delete Tracepoints, ,Setting Tracepoints}. They can
14360 also be set or changed at any time with the @code{condition} command,
14361 just as with breakpoints.
14363 Unlike breakpoint conditions, @value{GDBN} does not actually evaluate
14364 the conditional expression itself. Instead, @value{GDBN} encodes the
14365 expression into an agent expression (@pxref{Agent Expressions})
14366 suitable for execution on the target, independently of @value{GDBN}.
14367 Global variables become raw memory locations, locals become stack
14368 accesses, and so forth.
14370 For instance, suppose you have a function that is usually called
14371 frequently, but should not be called after an error has occurred. You
14372 could use the following tracepoint command to collect data about calls
14373 of that function that happen while the error code is propagating
14374 through the program; an unconditional tracepoint could end up
14375 collecting thousands of useless trace frames that you would have to
14379 (@value{GDBP}) @kbd{trace normal_operation if errcode > 0}
14382 @node Trace State Variables
14383 @subsection Trace State Variables
14384 @cindex trace state variables
14386 A @dfn{trace state variable} is a special type of variable that is
14387 created and managed by target-side code. The syntax is the same as
14388 that for GDB's convenience variables (a string prefixed with ``$''),
14389 but they are stored on the target. They must be created explicitly,
14390 using a @code{tvariable} command. They are always 64-bit signed
14393 Trace state variables are remembered by @value{GDBN}, and downloaded
14394 to the target along with tracepoint information when the trace
14395 experiment starts. There are no intrinsic limits on the number of
14396 trace state variables, beyond memory limitations of the target.
14398 @cindex convenience variables, and trace state variables
14399 Although trace state variables are managed by the target, you can use
14400 them in print commands and expressions as if they were convenience
14401 variables; @value{GDBN} will get the current value from the target
14402 while the trace experiment is running. Trace state variables share
14403 the same namespace as other ``$'' variables, which means that you
14404 cannot have trace state variables with names like @code{$23} or
14405 @code{$pc}, nor can you have a trace state variable and a convenience
14406 variable with the same name.
14410 @item tvariable $@var{name} [ = @var{expression} ]
14412 The @code{tvariable} command creates a new trace state variable named
14413 @code{$@var{name}}, and optionally gives it an initial value of
14414 @var{expression}. The @var{expression} is evaluated when this command is
14415 entered; the result will be converted to an integer if possible,
14416 otherwise @value{GDBN} will report an error. A subsequent
14417 @code{tvariable} command specifying the same name does not create a
14418 variable, but instead assigns the supplied initial value to the
14419 existing variable of that name, overwriting any previous initial
14420 value. The default initial value is 0.
14422 @item info tvariables
14423 @kindex info tvariables
14424 List all the trace state variables along with their initial values.
14425 Their current values may also be displayed, if the trace experiment is
14428 @item delete tvariable @r{[} $@var{name} @dots{} @r{]}
14429 @kindex delete tvariable
14430 Delete the given trace state variables, or all of them if no arguments
14435 @node Tracepoint Actions
14436 @subsection Tracepoint Action Lists
14440 @cindex tracepoint actions
14441 @item actions @r{[}@var{num}@r{]}
14442 This command will prompt for a list of actions to be taken when the
14443 tracepoint is hit. If the tracepoint number @var{num} is not
14444 specified, this command sets the actions for the one that was most
14445 recently defined (so that you can define a tracepoint and then say
14446 @code{actions} without bothering about its number). You specify the
14447 actions themselves on the following lines, one action at a time, and
14448 terminate the actions list with a line containing just @code{end}. So
14449 far, the only defined actions are @code{collect}, @code{teval}, and
14450 @code{while-stepping}.
14452 @code{actions} is actually equivalent to @code{commands} (@pxref{Break
14453 Commands, ,Breakpoint Command Lists}), except that only the defined
14454 actions are allowed; any other @value{GDBN} command is rejected.
14456 @cindex remove actions from a tracepoint
14457 To remove all actions from a tracepoint, type @samp{actions @var{num}}
14458 and follow it immediately with @samp{end}.
14461 (@value{GDBP}) @b{collect @var{data}} // collect some data
14463 (@value{GDBP}) @b{while-stepping 5} // single-step 5 times, collect data
14465 (@value{GDBP}) @b{end} // signals the end of actions.
14468 In the following example, the action list begins with @code{collect}
14469 commands indicating the things to be collected when the tracepoint is
14470 hit. Then, in order to single-step and collect additional data
14471 following the tracepoint, a @code{while-stepping} command is used,
14472 followed by the list of things to be collected after each step in a
14473 sequence of single steps. The @code{while-stepping} command is
14474 terminated by its own separate @code{end} command. Lastly, the action
14475 list is terminated by an @code{end} command.
14478 (@value{GDBP}) @b{trace foo}
14479 (@value{GDBP}) @b{actions}
14480 Enter actions for tracepoint 1, one per line:
14483 > while-stepping 12
14484 > collect $pc, arr[i]
14489 @kindex collect @r{(tracepoints)}
14490 @item collect@r{[}/@var{mods}@r{]} @var{expr1}, @var{expr2}, @dots{}
14491 Collect values of the given expressions when the tracepoint is hit.
14492 This command accepts a comma-separated list of any valid expressions.
14493 In addition to global, static, or local variables, the following
14494 special arguments are supported:
14498 Collect all registers.
14501 Collect all function arguments.
14504 Collect all local variables.
14507 Collect the return address. This is helpful if you want to see more
14510 @emph{Note:} The return address location can not always be reliably
14511 determined up front, and the wrong address / registers may end up
14512 collected instead. On some architectures the reliability is higher
14513 for tracepoints at function entry, while on others it's the opposite.
14514 When this happens, backtracing will stop because the return address is
14515 found unavailable (unless another collect rule happened to match it).
14518 Collects the number of arguments from the static probe at which the
14519 tracepoint is located.
14520 @xref{Static Probe Points}.
14522 @item $_probe_arg@var{n}
14523 @var{n} is an integer between 0 and 11. Collects the @var{n}th argument
14524 from the static probe at which the tracepoint is located.
14525 @xref{Static Probe Points}.
14528 @vindex $_sdata@r{, collect}
14529 Collect static tracepoint marker specific data. Only available for
14530 static tracepoints. @xref{Tracepoint Actions,,Tracepoint Action
14531 Lists}. On the UST static tracepoints library backend, an
14532 instrumentation point resembles a @code{printf} function call. The
14533 tracing library is able to collect user specified data formatted to a
14534 character string using the format provided by the programmer that
14535 instrumented the program. Other backends have similar mechanisms.
14536 Here's an example of a UST marker call:
14539 const char master_name[] = "$your_name";
14540 trace_mark(channel1, marker1, "hello %s", master_name)
14543 In this case, collecting @code{$_sdata} collects the string
14544 @samp{hello $yourname}. When analyzing the trace buffer, you can
14545 inspect @samp{$_sdata} like any other variable available to
14549 You can give several consecutive @code{collect} commands, each one
14550 with a single argument, or one @code{collect} command with several
14551 arguments separated by commas; the effect is the same.
14553 The optional @var{mods} changes the usual handling of the arguments.
14554 @code{s} requests that pointers to chars be handled as strings, in
14555 particular collecting the contents of the memory being pointed at, up
14556 to the first zero. The upper bound is by default the value of the
14557 @code{print elements} variable; if @code{s} is followed by a decimal
14558 number, that is the upper bound instead. So for instance
14559 @samp{collect/s25 mystr} collects as many as 25 characters at
14562 The command @code{info scope} (@pxref{Symbols, info scope}) is
14563 particularly useful for figuring out what data to collect.
14565 @kindex teval @r{(tracepoints)}
14566 @item teval @var{expr1}, @var{expr2}, @dots{}
14567 Evaluate the given expressions when the tracepoint is hit. This
14568 command accepts a comma-separated list of expressions. The results
14569 are discarded, so this is mainly useful for assigning values to trace
14570 state variables (@pxref{Trace State Variables}) without adding those
14571 values to the trace buffer, as would be the case if the @code{collect}
14574 @kindex while-stepping @r{(tracepoints)}
14575 @item while-stepping @var{n}
14576 Perform @var{n} single-step instruction traces after the tracepoint,
14577 collecting new data after each step. The @code{while-stepping}
14578 command is followed by the list of what to collect while stepping
14579 (followed by its own @code{end} command):
14582 > while-stepping 12
14583 > collect $regs, myglobal
14589 Note that @code{$pc} is not automatically collected by
14590 @code{while-stepping}; you need to explicitly collect that register if
14591 you need it. You may abbreviate @code{while-stepping} as @code{ws} or
14594 @item set default-collect @var{expr1}, @var{expr2}, @dots{}
14595 @kindex set default-collect
14596 @cindex default collection action
14597 This variable is a list of expressions to collect at each tracepoint
14598 hit. It is effectively an additional @code{collect} action prepended
14599 to every tracepoint action list. The expressions are parsed
14600 individually for each tracepoint, so for instance a variable named
14601 @code{xyz} may be interpreted as a global for one tracepoint, and a
14602 local for another, as appropriate to the tracepoint's location.
14604 @item show default-collect
14605 @kindex show default-collect
14606 Show the list of expressions that are collected by default at each
14611 @node Listing Tracepoints
14612 @subsection Listing Tracepoints
14615 @kindex info tracepoints @r{[}@var{n}@dots{}@r{]}
14616 @kindex info tp @r{[}@var{n}@dots{}@r{]}
14617 @cindex information about tracepoints
14618 @item info tracepoints @r{[}@var{num}@dots{}@r{]}
14619 Display information about the tracepoint @var{num}. If you don't
14620 specify a tracepoint number, displays information about all the
14621 tracepoints defined so far. The format is similar to that used for
14622 @code{info breakpoints}; in fact, @code{info tracepoints} is the same
14623 command, simply restricting itself to tracepoints.
14625 A tracepoint's listing may include additional information specific to
14630 its passcount as given by the @code{passcount @var{n}} command
14633 the state about installed on target of each location
14637 (@value{GDBP}) @b{info trace}
14638 Num Type Disp Enb Address What
14639 1 tracepoint keep y 0x0804ab57 in foo() at main.cxx:7
14641 collect globfoo, $regs
14646 2 tracepoint keep y <MULTIPLE>
14648 2.1 y 0x0804859c in func4 at change-loc.h:35
14649 installed on target
14650 2.2 y 0xb7ffc480 in func4 at change-loc.h:35
14651 installed on target
14652 2.3 y <PENDING> set_tracepoint
14653 3 tracepoint keep y 0x080485b1 in foo at change-loc.c:29
14654 not installed on target
14659 This command can be abbreviated @code{info tp}.
14662 @node Listing Static Tracepoint Markers
14663 @subsection Listing Static Tracepoint Markers
14666 @kindex info static-tracepoint-markers
14667 @cindex information about static tracepoint markers
14668 @item info static-tracepoint-markers
14669 Display information about all static tracepoint markers defined in the
14672 For each marker, the following columns are printed:
14676 An incrementing counter, output to help readability. This is not a
14679 The marker ID, as reported by the target.
14680 @item Enabled or Disabled
14681 Probed markers are tagged with @samp{y}. @samp{n} identifies marks
14682 that are not enabled.
14684 Where the marker is in your program, as a memory address.
14686 Where the marker is in the source for your program, as a file and line
14687 number. If the debug information included in the program does not
14688 allow @value{GDBN} to locate the source of the marker, this column
14689 will be left blank.
14693 In addition, the following information may be printed for each marker:
14697 User data passed to the tracing library by the marker call. In the
14698 UST backend, this is the format string passed as argument to the
14700 @item Static tracepoints probing the marker
14701 The list of static tracepoints attached to the marker.
14705 (@value{GDBP}) info static-tracepoint-markers
14706 Cnt ID Enb Address What
14707 1 ust/bar2 y 0x0000000000400e1a in main at stexample.c:25
14708 Data: number1 %d number2 %d
14709 Probed by static tracepoints: #2
14710 2 ust/bar33 n 0x0000000000400c87 in main at stexample.c:24
14716 @node Starting and Stopping Trace Experiments
14717 @subsection Starting and Stopping Trace Experiments
14720 @kindex tstart [ @var{notes} ]
14721 @cindex start a new trace experiment
14722 @cindex collected data discarded
14724 This command starts the trace experiment, and begins collecting data.
14725 It has the side effect of discarding all the data collected in the
14726 trace buffer during the previous trace experiment. If any arguments
14727 are supplied, they are taken as a note and stored with the trace
14728 experiment's state. The notes may be arbitrary text, and are
14729 especially useful with disconnected tracing in a multi-user context;
14730 the notes can explain what the trace is doing, supply user contact
14731 information, and so forth.
14733 @kindex tstop [ @var{notes} ]
14734 @cindex stop a running trace experiment
14736 This command stops the trace experiment. If any arguments are
14737 supplied, they are recorded with the experiment as a note. This is
14738 useful if you are stopping a trace started by someone else, for
14739 instance if the trace is interfering with the system's behavior and
14740 needs to be stopped quickly.
14742 @strong{Note}: a trace experiment and data collection may stop
14743 automatically if any tracepoint's passcount is reached
14744 (@pxref{Tracepoint Passcounts}), or if the trace buffer becomes full.
14747 @cindex status of trace data collection
14748 @cindex trace experiment, status of
14750 This command displays the status of the current trace data
14754 Here is an example of the commands we described so far:
14757 (@value{GDBP}) @b{trace gdb_c_test}
14758 (@value{GDBP}) @b{actions}
14759 Enter actions for tracepoint #1, one per line.
14760 > collect $regs,$locals,$args
14761 > while-stepping 11
14765 (@value{GDBP}) @b{tstart}
14766 [time passes @dots{}]
14767 (@value{GDBP}) @b{tstop}
14770 @anchor{disconnected tracing}
14771 @cindex disconnected tracing
14772 You can choose to continue running the trace experiment even if
14773 @value{GDBN} disconnects from the target, voluntarily or
14774 involuntarily. For commands such as @code{detach}, the debugger will
14775 ask what you want to do with the trace. But for unexpected
14776 terminations (@value{GDBN} crash, network outage), it would be
14777 unfortunate to lose hard-won trace data, so the variable
14778 @code{disconnected-tracing} lets you decide whether the trace should
14779 continue running without @value{GDBN}.
14782 @item set disconnected-tracing on
14783 @itemx set disconnected-tracing off
14784 @kindex set disconnected-tracing
14785 Choose whether a tracing run should continue to run if @value{GDBN}
14786 has disconnected from the target. Note that @code{detach} or
14787 @code{quit} will ask you directly what to do about a running trace no
14788 matter what this variable's setting, so the variable is mainly useful
14789 for handling unexpected situations, such as loss of the network.
14791 @item show disconnected-tracing
14792 @kindex show disconnected-tracing
14793 Show the current choice for disconnected tracing.
14797 When you reconnect to the target, the trace experiment may or may not
14798 still be running; it might have filled the trace buffer in the
14799 meantime, or stopped for one of the other reasons. If it is running,
14800 it will continue after reconnection.
14802 Upon reconnection, the target will upload information about the
14803 tracepoints in effect. @value{GDBN} will then compare that
14804 information to the set of tracepoints currently defined, and attempt
14805 to match them up, allowing for the possibility that the numbers may
14806 have changed due to creation and deletion in the meantime. If one of
14807 the target's tracepoints does not match any in @value{GDBN}, the
14808 debugger will create a new tracepoint, so that you have a number with
14809 which to specify that tracepoint. This matching-up process is
14810 necessarily heuristic, and it may result in useless tracepoints being
14811 created; you may simply delete them if they are of no use.
14813 @cindex circular trace buffer
14814 If your target agent supports a @dfn{circular trace buffer}, then you
14815 can run a trace experiment indefinitely without filling the trace
14816 buffer; when space runs out, the agent deletes already-collected trace
14817 frames, oldest first, until there is enough room to continue
14818 collecting. This is especially useful if your tracepoints are being
14819 hit too often, and your trace gets terminated prematurely because the
14820 buffer is full. To ask for a circular trace buffer, simply set
14821 @samp{circular-trace-buffer} to on. You can set this at any time,
14822 including during tracing; if the agent can do it, it will change
14823 buffer handling on the fly, otherwise it will not take effect until
14827 @item set circular-trace-buffer on
14828 @itemx set circular-trace-buffer off
14829 @kindex set circular-trace-buffer
14830 Choose whether a tracing run should use a linear or circular buffer
14831 for trace data. A linear buffer will not lose any trace data, but may
14832 fill up prematurely, while a circular buffer will discard old trace
14833 data, but it will have always room for the latest tracepoint hits.
14835 @item show circular-trace-buffer
14836 @kindex show circular-trace-buffer
14837 Show the current choice for the trace buffer. Note that this may not
14838 match the agent's current buffer handling, nor is it guaranteed to
14839 match the setting that might have been in effect during a past run,
14840 for instance if you are looking at frames from a trace file.
14845 @item set trace-buffer-size @var{n}
14846 @itemx set trace-buffer-size unlimited
14847 @kindex set trace-buffer-size
14848 Request that the target use a trace buffer of @var{n} bytes. Not all
14849 targets will honor the request; they may have a compiled-in size for
14850 the trace buffer, or some other limitation. Set to a value of
14851 @code{unlimited} or @code{-1} to let the target use whatever size it
14852 likes. This is also the default.
14854 @item show trace-buffer-size
14855 @kindex show trace-buffer-size
14856 Show the current requested size for the trace buffer. Note that this
14857 will only match the actual size if the target supports size-setting,
14858 and was able to handle the requested size. For instance, if the
14859 target can only change buffer size between runs, this variable will
14860 not reflect the change until the next run starts. Use @code{tstatus}
14861 to get a report of the actual buffer size.
14865 @item set trace-user @var{text}
14866 @kindex set trace-user
14868 @item show trace-user
14869 @kindex show trace-user
14871 @item set trace-notes @var{text}
14872 @kindex set trace-notes
14873 Set the trace run's notes.
14875 @item show trace-notes
14876 @kindex show trace-notes
14877 Show the trace run's notes.
14879 @item set trace-stop-notes @var{text}
14880 @kindex set trace-stop-notes
14881 Set the trace run's stop notes. The handling of the note is as for
14882 @code{tstop} arguments; the set command is convenient way to fix a
14883 stop note that is mistaken or incomplete.
14885 @item show trace-stop-notes
14886 @kindex show trace-stop-notes
14887 Show the trace run's stop notes.
14891 @node Tracepoint Restrictions
14892 @subsection Tracepoint Restrictions
14894 @cindex tracepoint restrictions
14895 There are a number of restrictions on the use of tracepoints. As
14896 described above, tracepoint data gathering occurs on the target
14897 without interaction from @value{GDBN}. Thus the full capabilities of
14898 the debugger are not available during data gathering, and then at data
14899 examination time, you will be limited by only having what was
14900 collected. The following items describe some common problems, but it
14901 is not exhaustive, and you may run into additional difficulties not
14907 Tracepoint expressions are intended to gather objects (lvalues). Thus
14908 the full flexibility of GDB's expression evaluator is not available.
14909 You cannot call functions, cast objects to aggregate types, access
14910 convenience variables or modify values (except by assignment to trace
14911 state variables). Some language features may implicitly call
14912 functions (for instance Objective-C fields with accessors), and therefore
14913 cannot be collected either.
14916 Collection of local variables, either individually or in bulk with
14917 @code{$locals} or @code{$args}, during @code{while-stepping} may
14918 behave erratically. The stepping action may enter a new scope (for
14919 instance by stepping into a function), or the location of the variable
14920 may change (for instance it is loaded into a register). The
14921 tracepoint data recorded uses the location information for the
14922 variables that is correct for the tracepoint location. When the
14923 tracepoint is created, it is not possible, in general, to determine
14924 where the steps of a @code{while-stepping} sequence will advance the
14925 program---particularly if a conditional branch is stepped.
14928 Collection of an incompletely-initialized or partially-destroyed object
14929 may result in something that @value{GDBN} cannot display, or displays
14930 in a misleading way.
14933 When @value{GDBN} displays a pointer to character it automatically
14934 dereferences the pointer to also display characters of the string
14935 being pointed to. However, collecting the pointer during tracing does
14936 not automatically collect the string. You need to explicitly
14937 dereference the pointer and provide size information if you want to
14938 collect not only the pointer, but the memory pointed to. For example,
14939 @code{*ptr@@50} can be used to collect the 50 element array pointed to
14943 It is not possible to collect a complete stack backtrace at a
14944 tracepoint. Instead, you may collect the registers and a few hundred
14945 bytes from the stack pointer with something like @code{*(unsigned char *)$esp@@300}
14946 (adjust to use the name of the actual stack pointer register on your
14947 target architecture, and the amount of stack you wish to capture).
14948 Then the @code{backtrace} command will show a partial backtrace when
14949 using a trace frame. The number of stack frames that can be examined
14950 depends on the sizes of the frames in the collected stack. Note that
14951 if you ask for a block so large that it goes past the bottom of the
14952 stack, the target agent may report an error trying to read from an
14956 If you do not collect registers at a tracepoint, @value{GDBN} can
14957 infer that the value of @code{$pc} must be the same as the address of
14958 the tracepoint and use that when you are looking at a trace frame
14959 for that tracepoint. However, this cannot work if the tracepoint has
14960 multiple locations (for instance if it was set in a function that was
14961 inlined), or if it has a @code{while-stepping} loop. In those cases
14962 @value{GDBN} will warn you that it can't infer @code{$pc}, and default
14967 @node Analyze Collected Data
14968 @section Using the Collected Data
14970 After the tracepoint experiment ends, you use @value{GDBN} commands
14971 for examining the trace data. The basic idea is that each tracepoint
14972 collects a trace @dfn{snapshot} every time it is hit and another
14973 snapshot every time it single-steps. All these snapshots are
14974 consecutively numbered from zero and go into a buffer, and you can
14975 examine them later. The way you examine them is to @dfn{focus} on a
14976 specific trace snapshot. When the remote stub is focused on a trace
14977 snapshot, it will respond to all @value{GDBN} requests for memory and
14978 registers by reading from the buffer which belongs to that snapshot,
14979 rather than from @emph{real} memory or registers of the program being
14980 debugged. This means that @strong{all} @value{GDBN} commands
14981 (@code{print}, @code{info registers}, @code{backtrace}, etc.) will
14982 behave as if we were currently debugging the program state as it was
14983 when the tracepoint occurred. Any requests for data that are not in
14984 the buffer will fail.
14987 * tfind:: How to select a trace snapshot
14988 * tdump:: How to display all data for a snapshot
14989 * save tracepoints:: How to save tracepoints for a future run
14993 @subsection @code{tfind @var{n}}
14996 @cindex select trace snapshot
14997 @cindex find trace snapshot
14998 The basic command for selecting a trace snapshot from the buffer is
14999 @code{tfind @var{n}}, which finds trace snapshot number @var{n},
15000 counting from zero. If no argument @var{n} is given, the next
15001 snapshot is selected.
15003 Here are the various forms of using the @code{tfind} command.
15007 Find the first snapshot in the buffer. This is a synonym for
15008 @code{tfind 0} (since 0 is the number of the first snapshot).
15011 Stop debugging trace snapshots, resume @emph{live} debugging.
15014 Same as @samp{tfind none}.
15017 No argument means find the next trace snapshot or find the first
15018 one if no trace snapshot is selected.
15021 Find the previous trace snapshot before the current one. This permits
15022 retracing earlier steps.
15024 @item tfind tracepoint @var{num}
15025 Find the next snapshot associated with tracepoint @var{num}. Search
15026 proceeds forward from the last examined trace snapshot. If no
15027 argument @var{num} is given, it means find the next snapshot collected
15028 for the same tracepoint as the current snapshot.
15030 @item tfind pc @var{addr}
15031 Find the next snapshot associated with the value @var{addr} of the
15032 program counter. Search proceeds forward from the last examined trace
15033 snapshot. If no argument @var{addr} is given, it means find the next
15034 snapshot with the same value of PC as the current snapshot.
15036 @item tfind outside @var{addr1}, @var{addr2}
15037 Find the next snapshot whose PC is outside the given range of
15038 addresses (exclusive).
15040 @item tfind range @var{addr1}, @var{addr2}
15041 Find the next snapshot whose PC is between @var{addr1} and
15042 @var{addr2} (inclusive).
15044 @item tfind line @r{[}@var{file}:@r{]}@var{n}
15045 Find the next snapshot associated with the source line @var{n}. If
15046 the optional argument @var{file} is given, refer to line @var{n} in
15047 that source file. Search proceeds forward from the last examined
15048 trace snapshot. If no argument @var{n} is given, it means find the
15049 next line other than the one currently being examined; thus saying
15050 @code{tfind line} repeatedly can appear to have the same effect as
15051 stepping from line to line in a @emph{live} debugging session.
15054 The default arguments for the @code{tfind} commands are specifically
15055 designed to make it easy to scan through the trace buffer. For
15056 instance, @code{tfind} with no argument selects the next trace
15057 snapshot, and @code{tfind -} with no argument selects the previous
15058 trace snapshot. So, by giving one @code{tfind} command, and then
15059 simply hitting @key{RET} repeatedly you can examine all the trace
15060 snapshots in order. Or, by saying @code{tfind -} and then hitting
15061 @key{RET} repeatedly you can examine the snapshots in reverse order.
15062 The @code{tfind line} command with no argument selects the snapshot
15063 for the next source line executed. The @code{tfind pc} command with
15064 no argument selects the next snapshot with the same program counter
15065 (PC) as the current frame. The @code{tfind tracepoint} command with
15066 no argument selects the next trace snapshot collected by the same
15067 tracepoint as the current one.
15069 In addition to letting you scan through the trace buffer manually,
15070 these commands make it easy to construct @value{GDBN} scripts that
15071 scan through the trace buffer and print out whatever collected data
15072 you are interested in. Thus, if we want to examine the PC, FP, and SP
15073 registers from each trace frame in the buffer, we can say this:
15076 (@value{GDBP}) @b{tfind start}
15077 (@value{GDBP}) @b{while ($trace_frame != -1)}
15078 > printf "Frame %d, PC = %08X, SP = %08X, FP = %08X\n", \
15079 $trace_frame, $pc, $sp, $fp
15083 Frame 0, PC = 0020DC64, SP = 0030BF3C, FP = 0030BF44
15084 Frame 1, PC = 0020DC6C, SP = 0030BF38, FP = 0030BF44
15085 Frame 2, PC = 0020DC70, SP = 0030BF34, FP = 0030BF44
15086 Frame 3, PC = 0020DC74, SP = 0030BF30, FP = 0030BF44
15087 Frame 4, PC = 0020DC78, SP = 0030BF2C, FP = 0030BF44
15088 Frame 5, PC = 0020DC7C, SP = 0030BF28, FP = 0030BF44
15089 Frame 6, PC = 0020DC80, SP = 0030BF24, FP = 0030BF44
15090 Frame 7, PC = 0020DC84, SP = 0030BF20, FP = 0030BF44
15091 Frame 8, PC = 0020DC88, SP = 0030BF1C, FP = 0030BF44
15092 Frame 9, PC = 0020DC8E, SP = 0030BF18, FP = 0030BF44
15093 Frame 10, PC = 00203F6C, SP = 0030BE3C, FP = 0030BF14
15096 Or, if we want to examine the variable @code{X} at each source line in
15100 (@value{GDBP}) @b{tfind start}
15101 (@value{GDBP}) @b{while ($trace_frame != -1)}
15102 > printf "Frame %d, X == %d\n", $trace_frame, X
15112 @subsection @code{tdump}
15114 @cindex dump all data collected at tracepoint
15115 @cindex tracepoint data, display
15117 This command takes no arguments. It prints all the data collected at
15118 the current trace snapshot.
15121 (@value{GDBP}) @b{trace 444}
15122 (@value{GDBP}) @b{actions}
15123 Enter actions for tracepoint #2, one per line:
15124 > collect $regs, $locals, $args, gdb_long_test
15127 (@value{GDBP}) @b{tstart}
15129 (@value{GDBP}) @b{tfind line 444}
15130 #0 gdb_test (p1=0x11, p2=0x22, p3=0x33, p4=0x44, p5=0x55, p6=0x66)
15132 444 printp( "%s: arguments = 0x%X 0x%X 0x%X 0x%X 0x%X 0x%X\n", )
15134 (@value{GDBP}) @b{tdump}
15135 Data collected at tracepoint 2, trace frame 1:
15136 d0 0xc4aa0085 -995491707
15140 d4 0x71aea3d 119204413
15143 d7 0x380035 3670069
15144 a0 0x19e24a 1696330
15145 a1 0x3000668 50333288
15147 a3 0x322000 3284992
15148 a4 0x3000698 50333336
15149 a5 0x1ad3cc 1758156
15150 fp 0x30bf3c 0x30bf3c
15151 sp 0x30bf34 0x30bf34
15153 pc 0x20b2c8 0x20b2c8
15157 p = 0x20e5b4 "gdb-test"
15164 gdb_long_test = 17 '\021'
15169 @code{tdump} works by scanning the tracepoint's current collection
15170 actions and printing the value of each expression listed. So
15171 @code{tdump} can fail, if after a run, you change the tracepoint's
15172 actions to mention variables that were not collected during the run.
15174 Also, for tracepoints with @code{while-stepping} loops, @code{tdump}
15175 uses the collected value of @code{$pc} to distinguish between trace
15176 frames that were collected at the tracepoint hit, and frames that were
15177 collected while stepping. This allows it to correctly choose whether
15178 to display the basic list of collections, or the collections from the
15179 body of the while-stepping loop. However, if @code{$pc} was not collected,
15180 then @code{tdump} will always attempt to dump using the basic collection
15181 list, and may fail if a while-stepping frame does not include all the
15182 same data that is collected at the tracepoint hit.
15183 @c This is getting pretty arcane, example would be good.
15185 @node save tracepoints
15186 @subsection @code{save tracepoints @var{filename}}
15187 @kindex save tracepoints
15188 @kindex save-tracepoints
15189 @cindex save tracepoints for future sessions
15191 This command saves all current tracepoint definitions together with
15192 their actions and passcounts, into a file @file{@var{filename}}
15193 suitable for use in a later debugging session. To read the saved
15194 tracepoint definitions, use the @code{source} command (@pxref{Command
15195 Files}). The @w{@code{save-tracepoints}} command is a deprecated
15196 alias for @w{@code{save tracepoints}}
15198 @node Tracepoint Variables
15199 @section Convenience Variables for Tracepoints
15200 @cindex tracepoint variables
15201 @cindex convenience variables for tracepoints
15204 @vindex $trace_frame
15205 @item (int) $trace_frame
15206 The current trace snapshot (a.k.a.@: @dfn{frame}) number, or -1 if no
15207 snapshot is selected.
15209 @vindex $tracepoint
15210 @item (int) $tracepoint
15211 The tracepoint for the current trace snapshot.
15213 @vindex $trace_line
15214 @item (int) $trace_line
15215 The line number for the current trace snapshot.
15217 @vindex $trace_file
15218 @item (char []) $trace_file
15219 The source file for the current trace snapshot.
15221 @vindex $trace_func
15222 @item (char []) $trace_func
15223 The name of the function containing @code{$tracepoint}.
15226 Note: @code{$trace_file} is not suitable for use in @code{printf},
15227 use @code{output} instead.
15229 Here's a simple example of using these convenience variables for
15230 stepping through all the trace snapshots and printing some of their
15231 data. Note that these are not the same as trace state variables,
15232 which are managed by the target.
15235 (@value{GDBP}) @b{tfind start}
15237 (@value{GDBP}) @b{while $trace_frame != -1}
15238 > output $trace_file
15239 > printf ", line %d (tracepoint #%d)\n", $trace_line, $tracepoint
15245 @section Using Trace Files
15246 @cindex trace files
15248 In some situations, the target running a trace experiment may no
15249 longer be available; perhaps it crashed, or the hardware was needed
15250 for a different activity. To handle these cases, you can arrange to
15251 dump the trace data into a file, and later use that file as a source
15252 of trace data, via the @code{target tfile} command.
15257 @item tsave [ -r ] @var{filename}
15258 @itemx tsave [-ctf] @var{dirname}
15259 Save the trace data to @var{filename}. By default, this command
15260 assumes that @var{filename} refers to the host filesystem, so if
15261 necessary @value{GDBN} will copy raw trace data up from the target and
15262 then save it. If the target supports it, you can also supply the
15263 optional argument @code{-r} (``remote'') to direct the target to save
15264 the data directly into @var{filename} in its own filesystem, which may be
15265 more efficient if the trace buffer is very large. (Note, however, that
15266 @code{target tfile} can only read from files accessible to the host.)
15267 By default, this command will save trace frame in tfile format.
15268 You can supply the optional argument @code{-ctf} to save data in CTF
15269 format. The @dfn{Common Trace Format} (CTF) is proposed as a trace format
15270 that can be shared by multiple debugging and tracing tools. Please go to
15271 @indicateurl{http://www.efficios.com/ctf} to get more information.
15273 @kindex target tfile
15277 @item target tfile @var{filename}
15278 @itemx target ctf @var{dirname}
15279 Use the file named @var{filename} or directory named @var{dirname} as
15280 a source of trace data. Commands that examine data work as they do with
15281 a live target, but it is not possible to run any new trace experiments.
15282 @code{tstatus} will report the state of the trace run at the moment
15283 the data was saved, as well as the current trace frame you are examining.
15284 Both @var{filename} and @var{dirname} must be on a filesystem accessible to
15288 (@value{GDBP}) target ctf ctf.ctf
15289 (@value{GDBP}) tfind
15290 Found trace frame 0, tracepoint 2
15291 39 ++a; /* set tracepoint 1 here */
15292 (@value{GDBP}) tdump
15293 Data collected at tracepoint 2, trace frame 0:
15297 c = @{"123", "456", "789", "123", "456", "789"@}
15298 d = @{@{@{a = 1, b = 2@}, @{a = 3, b = 4@}@}, @{@{a = 5, b = 6@}, @{a = 7, b = 8@}@}@}
15306 @chapter Debugging Programs That Use Overlays
15309 If your program is too large to fit completely in your target system's
15310 memory, you can sometimes use @dfn{overlays} to work around this
15311 problem. @value{GDBN} provides some support for debugging programs that
15315 * How Overlays Work:: A general explanation of overlays.
15316 * Overlay Commands:: Managing overlays in @value{GDBN}.
15317 * Automatic Overlay Debugging:: @value{GDBN} can find out which overlays are
15318 mapped by asking the inferior.
15319 * Overlay Sample Program:: A sample program using overlays.
15322 @node How Overlays Work
15323 @section How Overlays Work
15324 @cindex mapped overlays
15325 @cindex unmapped overlays
15326 @cindex load address, overlay's
15327 @cindex mapped address
15328 @cindex overlay area
15330 Suppose you have a computer whose instruction address space is only 64
15331 kilobytes long, but which has much more memory which can be accessed by
15332 other means: special instructions, segment registers, or memory
15333 management hardware, for example. Suppose further that you want to
15334 adapt a program which is larger than 64 kilobytes to run on this system.
15336 One solution is to identify modules of your program which are relatively
15337 independent, and need not call each other directly; call these modules
15338 @dfn{overlays}. Separate the overlays from the main program, and place
15339 their machine code in the larger memory. Place your main program in
15340 instruction memory, but leave at least enough space there to hold the
15341 largest overlay as well.
15343 Now, to call a function located in an overlay, you must first copy that
15344 overlay's machine code from the large memory into the space set aside
15345 for it in the instruction memory, and then jump to its entry point
15348 @c NB: In the below the mapped area's size is greater or equal to the
15349 @c size of all overlays. This is intentional to remind the developer
15350 @c that overlays don't necessarily need to be the same size.
15354 Data Instruction Larger
15355 Address Space Address Space Address Space
15356 +-----------+ +-----------+ +-----------+
15358 +-----------+ +-----------+ +-----------+<-- overlay 1
15359 | program | | main | .----| overlay 1 | load address
15360 | variables | | program | | +-----------+
15361 | and heap | | | | | |
15362 +-----------+ | | | +-----------+<-- overlay 2
15363 | | +-----------+ | | | load address
15364 +-----------+ | | | .-| overlay 2 |
15366 mapped --->+-----------+ | | +-----------+
15367 address | | | | | |
15368 | overlay | <-' | | |
15369 | area | <---' +-----------+<-- overlay 3
15370 | | <---. | | load address
15371 +-----------+ `--| overlay 3 |
15378 @anchor{A code overlay}A code overlay
15382 The diagram (@pxref{A code overlay}) shows a system with separate data
15383 and instruction address spaces. To map an overlay, the program copies
15384 its code from the larger address space to the instruction address space.
15385 Since the overlays shown here all use the same mapped address, only one
15386 may be mapped at a time. For a system with a single address space for
15387 data and instructions, the diagram would be similar, except that the
15388 program variables and heap would share an address space with the main
15389 program and the overlay area.
15391 An overlay loaded into instruction memory and ready for use is called a
15392 @dfn{mapped} overlay; its @dfn{mapped address} is its address in the
15393 instruction memory. An overlay not present (or only partially present)
15394 in instruction memory is called @dfn{unmapped}; its @dfn{load address}
15395 is its address in the larger memory. The mapped address is also called
15396 the @dfn{virtual memory address}, or @dfn{VMA}; the load address is also
15397 called the @dfn{load memory address}, or @dfn{LMA}.
15399 Unfortunately, overlays are not a completely transparent way to adapt a
15400 program to limited instruction memory. They introduce a new set of
15401 global constraints you must keep in mind as you design your program:
15406 Before calling or returning to a function in an overlay, your program
15407 must make sure that overlay is actually mapped. Otherwise, the call or
15408 return will transfer control to the right address, but in the wrong
15409 overlay, and your program will probably crash.
15412 If the process of mapping an overlay is expensive on your system, you
15413 will need to choose your overlays carefully to minimize their effect on
15414 your program's performance.
15417 The executable file you load onto your system must contain each
15418 overlay's instructions, appearing at the overlay's load address, not its
15419 mapped address. However, each overlay's instructions must be relocated
15420 and its symbols defined as if the overlay were at its mapped address.
15421 You can use GNU linker scripts to specify different load and relocation
15422 addresses for pieces of your program; see @ref{Overlay Description,,,
15423 ld.info, Using ld: the GNU linker}.
15426 The procedure for loading executable files onto your system must be able
15427 to load their contents into the larger address space as well as the
15428 instruction and data spaces.
15432 The overlay system described above is rather simple, and could be
15433 improved in many ways:
15438 If your system has suitable bank switch registers or memory management
15439 hardware, you could use those facilities to make an overlay's load area
15440 contents simply appear at their mapped address in instruction space.
15441 This would probably be faster than copying the overlay to its mapped
15442 area in the usual way.
15445 If your overlays are small enough, you could set aside more than one
15446 overlay area, and have more than one overlay mapped at a time.
15449 You can use overlays to manage data, as well as instructions. In
15450 general, data overlays are even less transparent to your design than
15451 code overlays: whereas code overlays only require care when you call or
15452 return to functions, data overlays require care every time you access
15453 the data. Also, if you change the contents of a data overlay, you
15454 must copy its contents back out to its load address before you can copy a
15455 different data overlay into the same mapped area.
15460 @node Overlay Commands
15461 @section Overlay Commands
15463 To use @value{GDBN}'s overlay support, each overlay in your program must
15464 correspond to a separate section of the executable file. The section's
15465 virtual memory address and load memory address must be the overlay's
15466 mapped and load addresses. Identifying overlays with sections allows
15467 @value{GDBN} to determine the appropriate address of a function or
15468 variable, depending on whether the overlay is mapped or not.
15470 @value{GDBN}'s overlay commands all start with the word @code{overlay};
15471 you can abbreviate this as @code{ov} or @code{ovly}. The commands are:
15476 Disable @value{GDBN}'s overlay support. When overlay support is
15477 disabled, @value{GDBN} assumes that all functions and variables are
15478 always present at their mapped addresses. By default, @value{GDBN}'s
15479 overlay support is disabled.
15481 @item overlay manual
15482 @cindex manual overlay debugging
15483 Enable @dfn{manual} overlay debugging. In this mode, @value{GDBN}
15484 relies on you to tell it which overlays are mapped, and which are not,
15485 using the @code{overlay map-overlay} and @code{overlay unmap-overlay}
15486 commands described below.
15488 @item overlay map-overlay @var{overlay}
15489 @itemx overlay map @var{overlay}
15490 @cindex map an overlay
15491 Tell @value{GDBN} that @var{overlay} is now mapped; @var{overlay} must
15492 be the name of the object file section containing the overlay. When an
15493 overlay is mapped, @value{GDBN} assumes it can find the overlay's
15494 functions and variables at their mapped addresses. @value{GDBN} assumes
15495 that any other overlays whose mapped ranges overlap that of
15496 @var{overlay} are now unmapped.
15498 @item overlay unmap-overlay @var{overlay}
15499 @itemx overlay unmap @var{overlay}
15500 @cindex unmap an overlay
15501 Tell @value{GDBN} that @var{overlay} is no longer mapped; @var{overlay}
15502 must be the name of the object file section containing the overlay.
15503 When an overlay is unmapped, @value{GDBN} assumes it can find the
15504 overlay's functions and variables at their load addresses.
15507 Enable @dfn{automatic} overlay debugging. In this mode, @value{GDBN}
15508 consults a data structure the overlay manager maintains in the inferior
15509 to see which overlays are mapped. For details, see @ref{Automatic
15510 Overlay Debugging}.
15512 @item overlay load-target
15513 @itemx overlay load
15514 @cindex reloading the overlay table
15515 Re-read the overlay table from the inferior. Normally, @value{GDBN}
15516 re-reads the table @value{GDBN} automatically each time the inferior
15517 stops, so this command should only be necessary if you have changed the
15518 overlay mapping yourself using @value{GDBN}. This command is only
15519 useful when using automatic overlay debugging.
15521 @item overlay list-overlays
15522 @itemx overlay list
15523 @cindex listing mapped overlays
15524 Display a list of the overlays currently mapped, along with their mapped
15525 addresses, load addresses, and sizes.
15529 Normally, when @value{GDBN} prints a code address, it includes the name
15530 of the function the address falls in:
15533 (@value{GDBP}) print main
15534 $3 = @{int ()@} 0x11a0 <main>
15537 When overlay debugging is enabled, @value{GDBN} recognizes code in
15538 unmapped overlays, and prints the names of unmapped functions with
15539 asterisks around them. For example, if @code{foo} is a function in an
15540 unmapped overlay, @value{GDBN} prints it this way:
15543 (@value{GDBP}) overlay list
15544 No sections are mapped.
15545 (@value{GDBP}) print foo
15546 $5 = @{int (int)@} 0x100000 <*foo*>
15549 When @code{foo}'s overlay is mapped, @value{GDBN} prints the function's
15553 (@value{GDBP}) overlay list
15554 Section .ov.foo.text, loaded at 0x100000 - 0x100034,
15555 mapped at 0x1016 - 0x104a
15556 (@value{GDBP}) print foo
15557 $6 = @{int (int)@} 0x1016 <foo>
15560 When overlay debugging is enabled, @value{GDBN} can find the correct
15561 address for functions and variables in an overlay, whether or not the
15562 overlay is mapped. This allows most @value{GDBN} commands, like
15563 @code{break} and @code{disassemble}, to work normally, even on unmapped
15564 code. However, @value{GDBN}'s breakpoint support has some limitations:
15568 @cindex breakpoints in overlays
15569 @cindex overlays, setting breakpoints in
15570 You can set breakpoints in functions in unmapped overlays, as long as
15571 @value{GDBN} can write to the overlay at its load address.
15573 @value{GDBN} can not set hardware or simulator-based breakpoints in
15574 unmapped overlays. However, if you set a breakpoint at the end of your
15575 overlay manager (and tell @value{GDBN} which overlays are now mapped, if
15576 you are using manual overlay management), @value{GDBN} will re-set its
15577 breakpoints properly.
15581 @node Automatic Overlay Debugging
15582 @section Automatic Overlay Debugging
15583 @cindex automatic overlay debugging
15585 @value{GDBN} can automatically track which overlays are mapped and which
15586 are not, given some simple co-operation from the overlay manager in the
15587 inferior. If you enable automatic overlay debugging with the
15588 @code{overlay auto} command (@pxref{Overlay Commands}), @value{GDBN}
15589 looks in the inferior's memory for certain variables describing the
15590 current state of the overlays.
15592 Here are the variables your overlay manager must define to support
15593 @value{GDBN}'s automatic overlay debugging:
15597 @item @code{_ovly_table}:
15598 This variable must be an array of the following structures:
15603 /* The overlay's mapped address. */
15606 /* The size of the overlay, in bytes. */
15607 unsigned long size;
15609 /* The overlay's load address. */
15612 /* Non-zero if the overlay is currently mapped;
15614 unsigned long mapped;
15618 @item @code{_novlys}:
15619 This variable must be a four-byte signed integer, holding the total
15620 number of elements in @code{_ovly_table}.
15624 To decide whether a particular overlay is mapped or not, @value{GDBN}
15625 looks for an entry in @w{@code{_ovly_table}} whose @code{vma} and
15626 @code{lma} members equal the VMA and LMA of the overlay's section in the
15627 executable file. When @value{GDBN} finds a matching entry, it consults
15628 the entry's @code{mapped} member to determine whether the overlay is
15631 In addition, your overlay manager may define a function called
15632 @code{_ovly_debug_event}. If this function is defined, @value{GDBN}
15633 will silently set a breakpoint there. If the overlay manager then
15634 calls this function whenever it has changed the overlay table, this
15635 will enable @value{GDBN} to accurately keep track of which overlays
15636 are in program memory, and update any breakpoints that may be set
15637 in overlays. This will allow breakpoints to work even if the
15638 overlays are kept in ROM or other non-writable memory while they
15639 are not being executed.
15641 @node Overlay Sample Program
15642 @section Overlay Sample Program
15643 @cindex overlay example program
15645 When linking a program which uses overlays, you must place the overlays
15646 at their load addresses, while relocating them to run at their mapped
15647 addresses. To do this, you must write a linker script (@pxref{Overlay
15648 Description,,, ld.info, Using ld: the GNU linker}). Unfortunately,
15649 since linker scripts are specific to a particular host system, target
15650 architecture, and target memory layout, this manual cannot provide
15651 portable sample code demonstrating @value{GDBN}'s overlay support.
15653 However, the @value{GDBN} source distribution does contain an overlaid
15654 program, with linker scripts for a few systems, as part of its test
15655 suite. The program consists of the following files from
15656 @file{gdb/testsuite/gdb.base}:
15660 The main program file.
15662 A simple overlay manager, used by @file{overlays.c}.
15667 Overlay modules, loaded and used by @file{overlays.c}.
15670 Linker scripts for linking the test program on the @code{d10v-elf}
15671 and @code{m32r-elf} targets.
15674 You can build the test program using the @code{d10v-elf} GCC
15675 cross-compiler like this:
15678 $ d10v-elf-gcc -g -c overlays.c
15679 $ d10v-elf-gcc -g -c ovlymgr.c
15680 $ d10v-elf-gcc -g -c foo.c
15681 $ d10v-elf-gcc -g -c bar.c
15682 $ d10v-elf-gcc -g -c baz.c
15683 $ d10v-elf-gcc -g -c grbx.c
15684 $ d10v-elf-gcc -g overlays.o ovlymgr.o foo.o bar.o \
15685 baz.o grbx.o -Wl,-Td10v.ld -o overlays
15688 The build process is identical for any other architecture, except that
15689 you must substitute the appropriate compiler and linker script for the
15690 target system for @code{d10v-elf-gcc} and @code{d10v.ld}.
15694 @chapter Using @value{GDBN} with Different Languages
15697 Although programming languages generally have common aspects, they are
15698 rarely expressed in the same manner. For instance, in ANSI C,
15699 dereferencing a pointer @code{p} is accomplished by @code{*p}, but in
15700 Modula-2, it is accomplished by @code{p^}. Values can also be
15701 represented (and displayed) differently. Hex numbers in C appear as
15702 @samp{0x1ae}, while in Modula-2 they appear as @samp{1AEH}.
15704 @cindex working language
15705 Language-specific information is built into @value{GDBN} for some languages,
15706 allowing you to express operations like the above in your program's
15707 native language, and allowing @value{GDBN} to output values in a manner
15708 consistent with the syntax of your program's native language. The
15709 language you use to build expressions is called the @dfn{working
15713 * Setting:: Switching between source languages
15714 * Show:: Displaying the language
15715 * Checks:: Type and range checks
15716 * Supported Languages:: Supported languages
15717 * Unsupported Languages:: Unsupported languages
15721 @section Switching Between Source Languages
15723 There are two ways to control the working language---either have @value{GDBN}
15724 set it automatically, or select it manually yourself. You can use the
15725 @code{set language} command for either purpose. On startup, @value{GDBN}
15726 defaults to setting the language automatically. The working language is
15727 used to determine how expressions you type are interpreted, how values
15730 In addition to the working language, every source file that
15731 @value{GDBN} knows about has its own working language. For some object
15732 file formats, the compiler might indicate which language a particular
15733 source file is in. However, most of the time @value{GDBN} infers the
15734 language from the name of the file. The language of a source file
15735 controls whether C@t{++} names are demangled---this way @code{backtrace} can
15736 show each frame appropriately for its own language. There is no way to
15737 set the language of a source file from within @value{GDBN}, but you can
15738 set the language associated with a filename extension. @xref{Show, ,
15739 Displaying the Language}.
15741 This is most commonly a problem when you use a program, such
15742 as @code{cfront} or @code{f2c}, that generates C but is written in
15743 another language. In that case, make the
15744 program use @code{#line} directives in its C output; that way
15745 @value{GDBN} will know the correct language of the source code of the original
15746 program, and will display that source code, not the generated C code.
15749 * Filenames:: Filename extensions and languages.
15750 * Manually:: Setting the working language manually
15751 * Automatically:: Having @value{GDBN} infer the source language
15755 @subsection List of Filename Extensions and Languages
15757 If a source file name ends in one of the following extensions, then
15758 @value{GDBN} infers that its language is the one indicated.
15776 C@t{++} source file
15782 Objective-C source file
15786 Fortran source file
15789 Modula-2 source file
15793 Assembler source file. This actually behaves almost like C, but
15794 @value{GDBN} does not skip over function prologues when stepping.
15797 In addition, you may set the language associated with a filename
15798 extension. @xref{Show, , Displaying the Language}.
15801 @subsection Setting the Working Language
15803 If you allow @value{GDBN} to set the language automatically,
15804 expressions are interpreted the same way in your debugging session and
15807 @kindex set language
15808 If you wish, you may set the language manually. To do this, issue the
15809 command @samp{set language @var{lang}}, where @var{lang} is the name of
15810 a language, such as
15811 @code{c} or @code{modula-2}.
15812 For a list of the supported languages, type @samp{set language}.
15814 Setting the language manually prevents @value{GDBN} from updating the working
15815 language automatically. This can lead to confusion if you try
15816 to debug a program when the working language is not the same as the
15817 source language, when an expression is acceptable to both
15818 languages---but means different things. For instance, if the current
15819 source file were written in C, and @value{GDBN} was parsing Modula-2, a
15827 might not have the effect you intended. In C, this means to add
15828 @code{b} and @code{c} and place the result in @code{a}. The result
15829 printed would be the value of @code{a}. In Modula-2, this means to compare
15830 @code{a} to the result of @code{b+c}, yielding a @code{BOOLEAN} value.
15832 @node Automatically
15833 @subsection Having @value{GDBN} Infer the Source Language
15835 To have @value{GDBN} set the working language automatically, use
15836 @samp{set language local} or @samp{set language auto}. @value{GDBN}
15837 then infers the working language. That is, when your program stops in a
15838 frame (usually by encountering a breakpoint), @value{GDBN} sets the
15839 working language to the language recorded for the function in that
15840 frame. If the language for a frame is unknown (that is, if the function
15841 or block corresponding to the frame was defined in a source file that
15842 does not have a recognized extension), the current working language is
15843 not changed, and @value{GDBN} issues a warning.
15845 This may not seem necessary for most programs, which are written
15846 entirely in one source language. However, program modules and libraries
15847 written in one source language can be used by a main program written in
15848 a different source language. Using @samp{set language auto} in this
15849 case frees you from having to set the working language manually.
15852 @section Displaying the Language
15854 The following commands help you find out which language is the
15855 working language, and also what language source files were written in.
15858 @item show language
15859 @anchor{show language}
15860 @kindex show language
15861 Display the current working language. This is the
15862 language you can use with commands such as @code{print} to
15863 build and compute expressions that may involve variables in your program.
15866 @kindex info frame@r{, show the source language}
15867 Display the source language for this frame. This language becomes the
15868 working language if you use an identifier from this frame.
15869 @xref{Frame Info, ,Information about a Frame}, to identify the other
15870 information listed here.
15873 @kindex info source@r{, show the source language}
15874 Display the source language of this source file.
15875 @xref{Symbols, ,Examining the Symbol Table}, to identify the other
15876 information listed here.
15879 In unusual circumstances, you may have source files with extensions
15880 not in the standard list. You can then set the extension associated
15881 with a language explicitly:
15884 @item set extension-language @var{ext} @var{language}
15885 @kindex set extension-language
15886 Tell @value{GDBN} that source files with extension @var{ext} are to be
15887 assumed as written in the source language @var{language}.
15889 @item info extensions
15890 @kindex info extensions
15891 List all the filename extensions and the associated languages.
15895 @section Type and Range Checking
15897 Some languages are designed to guard you against making seemingly common
15898 errors through a series of compile- and run-time checks. These include
15899 checking the type of arguments to functions and operators and making
15900 sure mathematical overflows are caught at run time. Checks such as
15901 these help to ensure a program's correctness once it has been compiled
15902 by eliminating type mismatches and providing active checks for range
15903 errors when your program is running.
15905 By default @value{GDBN} checks for these errors according to the
15906 rules of the current source language. Although @value{GDBN} does not check
15907 the statements in your program, it can check expressions entered directly
15908 into @value{GDBN} for evaluation via the @code{print} command, for example.
15911 * Type Checking:: An overview of type checking
15912 * Range Checking:: An overview of range checking
15915 @cindex type checking
15916 @cindex checks, type
15917 @node Type Checking
15918 @subsection An Overview of Type Checking
15920 Some languages, such as C and C@t{++}, are strongly typed, meaning that the
15921 arguments to operators and functions have to be of the correct type,
15922 otherwise an error occurs. These checks prevent type mismatch
15923 errors from ever causing any run-time problems. For example,
15926 int klass::my_method(char *b) @{ return b ? 1 : 2; @}
15928 (@value{GDBP}) print obj.my_method (0)
15931 (@value{GDBP}) print obj.my_method (0x1234)
15932 Cannot resolve method klass::my_method to any overloaded instance
15935 The second example fails because in C@t{++} the integer constant
15936 @samp{0x1234} is not type-compatible with the pointer parameter type.
15938 For the expressions you use in @value{GDBN} commands, you can tell
15939 @value{GDBN} to not enforce strict type checking or
15940 to treat any mismatches as errors and abandon the expression;
15941 When type checking is disabled, @value{GDBN} successfully evaluates
15942 expressions like the second example above.
15944 Even if type checking is off, there may be other reasons
15945 related to type that prevent @value{GDBN} from evaluating an expression.
15946 For instance, @value{GDBN} does not know how to add an @code{int} and
15947 a @code{struct foo}. These particular type errors have nothing to do
15948 with the language in use and usually arise from expressions which make
15949 little sense to evaluate anyway.
15951 @value{GDBN} provides some additional commands for controlling type checking:
15953 @kindex set check type
15954 @kindex show check type
15956 @item set check type on
15957 @itemx set check type off
15958 Set strict type checking on or off. If any type mismatches occur in
15959 evaluating an expression while type checking is on, @value{GDBN} prints a
15960 message and aborts evaluation of the expression.
15962 @item show check type
15963 Show the current setting of type checking and whether @value{GDBN}
15964 is enforcing strict type checking rules.
15967 @cindex range checking
15968 @cindex checks, range
15969 @node Range Checking
15970 @subsection An Overview of Range Checking
15972 In some languages (such as Modula-2), it is an error to exceed the
15973 bounds of a type; this is enforced with run-time checks. Such range
15974 checking is meant to ensure program correctness by making sure
15975 computations do not overflow, or indices on an array element access do
15976 not exceed the bounds of the array.
15978 For expressions you use in @value{GDBN} commands, you can tell
15979 @value{GDBN} to treat range errors in one of three ways: ignore them,
15980 always treat them as errors and abandon the expression, or issue
15981 warnings but evaluate the expression anyway.
15983 A range error can result from numerical overflow, from exceeding an
15984 array index bound, or when you type a constant that is not a member
15985 of any type. Some languages, however, do not treat overflows as an
15986 error. In many implementations of C, mathematical overflow causes the
15987 result to ``wrap around'' to lower values---for example, if @var{m} is
15988 the largest integer value, and @var{s} is the smallest, then
15991 @var{m} + 1 @result{} @var{s}
15994 This, too, is specific to individual languages, and in some cases
15995 specific to individual compilers or machines. @xref{Supported Languages, ,
15996 Supported Languages}, for further details on specific languages.
15998 @value{GDBN} provides some additional commands for controlling the range checker:
16000 @kindex set check range
16001 @kindex show check range
16003 @item set check range auto
16004 Set range checking on or off based on the current working language.
16005 @xref{Supported Languages, ,Supported Languages}, for the default settings for
16008 @item set check range on
16009 @itemx set check range off
16010 Set range checking on or off, overriding the default setting for the
16011 current working language. A warning is issued if the setting does not
16012 match the language default. If a range error occurs and range checking is on,
16013 then a message is printed and evaluation of the expression is aborted.
16015 @item set check range warn
16016 Output messages when the @value{GDBN} range checker detects a range error,
16017 but attempt to evaluate the expression anyway. Evaluating the
16018 expression may still be impossible for other reasons, such as accessing
16019 memory that the process does not own (a typical example from many Unix
16023 Show the current setting of the range checker, and whether or not it is
16024 being set automatically by @value{GDBN}.
16027 @node Supported Languages
16028 @section Supported Languages
16030 @value{GDBN} supports C, C@t{++}, D, Go, Objective-C, Fortran,
16031 OpenCL C, Pascal, Rust, assembly, Modula-2, and Ada.
16032 @c This is false ...
16033 Some @value{GDBN} features may be used in expressions regardless of the
16034 language you use: the @value{GDBN} @code{@@} and @code{::} operators,
16035 and the @samp{@{type@}addr} construct (@pxref{Expressions,
16036 ,Expressions}) can be used with the constructs of any supported
16039 The following sections detail to what degree each source language is
16040 supported by @value{GDBN}. These sections are not meant to be language
16041 tutorials or references, but serve only as a reference guide to what the
16042 @value{GDBN} expression parser accepts, and what input and output
16043 formats should look like for different languages. There are many good
16044 books written on each of these languages; please look to these for a
16045 language reference or tutorial.
16048 * C:: C and C@t{++}
16051 * Objective-C:: Objective-C
16052 * OpenCL C:: OpenCL C
16053 * Fortran:: Fortran
16056 * Modula-2:: Modula-2
16062 @subsection C and C@t{++}
16064 @cindex C and C@t{++}
16065 @cindex expressions in C or C@t{++}
16067 Since C and C@t{++} are so closely related, many features of @value{GDBN} apply
16068 to both languages. Whenever this is the case, we discuss those languages
16072 @cindex @code{g++}, @sc{gnu} C@t{++} compiler
16073 @cindex @sc{gnu} C@t{++}
16074 The C@t{++} debugging facilities are jointly implemented by the C@t{++}
16075 compiler and @value{GDBN}. Therefore, to debug your C@t{++} code
16076 effectively, you must compile your C@t{++} programs with a supported
16077 C@t{++} compiler, such as @sc{gnu} @code{g++}, or the HP ANSI C@t{++}
16078 compiler (@code{aCC}).
16081 * C Operators:: C and C@t{++} operators
16082 * C Constants:: C and C@t{++} constants
16083 * C Plus Plus Expressions:: C@t{++} expressions
16084 * C Defaults:: Default settings for C and C@t{++}
16085 * C Checks:: C and C@t{++} type and range checks
16086 * Debugging C:: @value{GDBN} and C
16087 * Debugging C Plus Plus:: @value{GDBN} features for C@t{++}
16088 * Decimal Floating Point:: Numbers in Decimal Floating Point format
16092 @subsubsection C and C@t{++} Operators
16094 @cindex C and C@t{++} operators
16096 Operators must be defined on values of specific types. For instance,
16097 @code{+} is defined on numbers, but not on structures. Operators are
16098 often defined on groups of types.
16100 For the purposes of C and C@t{++}, the following definitions hold:
16105 @emph{Integral types} include @code{int} with any of its storage-class
16106 specifiers; @code{char}; @code{enum}; and, for C@t{++}, @code{bool}.
16109 @emph{Floating-point types} include @code{float}, @code{double}, and
16110 @code{long double} (if supported by the target platform).
16113 @emph{Pointer types} include all types defined as @code{(@var{type} *)}.
16116 @emph{Scalar types} include all of the above.
16121 The following operators are supported. They are listed here
16122 in order of increasing precedence:
16126 The comma or sequencing operator. Expressions in a comma-separated list
16127 are evaluated from left to right, with the result of the entire
16128 expression being the last expression evaluated.
16131 Assignment. The value of an assignment expression is the value
16132 assigned. Defined on scalar types.
16135 Used in an expression of the form @w{@code{@var{a} @var{op}= @var{b}}},
16136 and translated to @w{@code{@var{a} = @var{a op b}}}.
16137 @w{@code{@var{op}=}} and @code{=} have the same precedence. The operator
16138 @var{op} is any one of the operators @code{|}, @code{^}, @code{&},
16139 @code{<<}, @code{>>}, @code{+}, @code{-}, @code{*}, @code{/}, @code{%}.
16142 The ternary operator. @code{@var{a} ? @var{b} : @var{c}} can be thought
16143 of as: if @var{a} then @var{b} else @var{c}. The argument @var{a}
16144 should be of an integral type.
16147 Logical @sc{or}. Defined on integral types.
16150 Logical @sc{and}. Defined on integral types.
16153 Bitwise @sc{or}. Defined on integral types.
16156 Bitwise exclusive-@sc{or}. Defined on integral types.
16159 Bitwise @sc{and}. Defined on integral types.
16162 Equality and inequality. Defined on scalar types. The value of these
16163 expressions is 0 for false and non-zero for true.
16165 @item <@r{, }>@r{, }<=@r{, }>=
16166 Less than, greater than, less than or equal, greater than or equal.
16167 Defined on scalar types. The value of these expressions is 0 for false
16168 and non-zero for true.
16171 left shift, and right shift. Defined on integral types.
16174 The @value{GDBN} ``artificial array'' operator (@pxref{Expressions, ,Expressions}).
16177 Addition and subtraction. Defined on integral types, floating-point types and
16180 @item *@r{, }/@r{, }%
16181 Multiplication, division, and modulus. Multiplication and division are
16182 defined on integral and floating-point types. Modulus is defined on
16186 Increment and decrement. When appearing before a variable, the
16187 operation is performed before the variable is used in an expression;
16188 when appearing after it, the variable's value is used before the
16189 operation takes place.
16192 Pointer dereferencing. Defined on pointer types. Same precedence as
16196 Address operator. Defined on variables. Same precedence as @code{++}.
16198 For debugging C@t{++}, @value{GDBN} implements a use of @samp{&} beyond what is
16199 allowed in the C@t{++} language itself: you can use @samp{&(&@var{ref})}
16200 to examine the address
16201 where a C@t{++} reference variable (declared with @samp{&@var{ref}}) is
16205 Negative. Defined on integral and floating-point types. Same
16206 precedence as @code{++}.
16209 Logical negation. Defined on integral types. Same precedence as
16213 Bitwise complement operator. Defined on integral types. Same precedence as
16218 Structure member, and pointer-to-structure member. For convenience,
16219 @value{GDBN} regards the two as equivalent, choosing whether to dereference a
16220 pointer based on the stored type information.
16221 Defined on @code{struct} and @code{union} data.
16224 Dereferences of pointers to members.
16227 Array indexing. @code{@var{a}[@var{i}]} is defined as
16228 @code{*(@var{a}+@var{i})}. Same precedence as @code{->}.
16231 Function parameter list. Same precedence as @code{->}.
16234 C@t{++} scope resolution operator. Defined on @code{struct}, @code{union},
16235 and @code{class} types.
16238 Doubled colons also represent the @value{GDBN} scope operator
16239 (@pxref{Expressions, ,Expressions}). Same precedence as @code{::},
16243 If an operator is redefined in the user code, @value{GDBN} usually
16244 attempts to invoke the redefined version instead of using the operator's
16245 predefined meaning.
16248 @subsubsection C and C@t{++} Constants
16250 @cindex C and C@t{++} constants
16252 @value{GDBN} allows you to express the constants of C and C@t{++} in the
16257 Integer constants are a sequence of digits. Octal constants are
16258 specified by a leading @samp{0} (i.e.@: zero), and hexadecimal constants
16259 by a leading @samp{0x} or @samp{0X}. Constants may also end with a letter
16260 @samp{l}, specifying that the constant should be treated as a
16264 Floating point constants are a sequence of digits, followed by a decimal
16265 point, followed by a sequence of digits, and optionally followed by an
16266 exponent. An exponent is of the form:
16267 @samp{@w{e@r{[[}+@r{]|}-@r{]}@var{nnn}}}, where @var{nnn} is another
16268 sequence of digits. The @samp{+} is optional for positive exponents.
16269 A floating-point constant may also end with a letter @samp{f} or
16270 @samp{F}, specifying that the constant should be treated as being of
16271 the @code{float} (as opposed to the default @code{double}) type; or with
16272 a letter @samp{l} or @samp{L}, which specifies a @code{long double}
16276 Enumerated constants consist of enumerated identifiers, or their
16277 integral equivalents.
16280 Character constants are a single character surrounded by single quotes
16281 (@code{'}), or a number---the ordinal value of the corresponding character
16282 (usually its @sc{ascii} value). Within quotes, the single character may
16283 be represented by a letter or by @dfn{escape sequences}, which are of
16284 the form @samp{\@var{nnn}}, where @var{nnn} is the octal representation
16285 of the character's ordinal value; or of the form @samp{\@var{x}}, where
16286 @samp{@var{x}} is a predefined special character---for example,
16287 @samp{\n} for newline.
16289 Wide character constants can be written by prefixing a character
16290 constant with @samp{L}, as in C. For example, @samp{L'x'} is the wide
16291 form of @samp{x}. The target wide character set is used when
16292 computing the value of this constant (@pxref{Character Sets}).
16295 String constants are a sequence of character constants surrounded by
16296 double quotes (@code{"}). Any valid character constant (as described
16297 above) may appear. Double quotes within the string must be preceded by
16298 a backslash, so for instance @samp{"a\"b'c"} is a string of five
16301 Wide string constants can be written by prefixing a string constant
16302 with @samp{L}, as in C. The target wide character set is used when
16303 computing the value of this constant (@pxref{Character Sets}).
16306 Pointer constants are an integral value. You can also write pointers
16307 to constants using the C operator @samp{&}.
16310 Array constants are comma-separated lists surrounded by braces @samp{@{}
16311 and @samp{@}}; for example, @samp{@{1,2,3@}} is a three-element array of
16312 integers, @samp{@{@{1,2@}, @{3,4@}, @{5,6@}@}} is a three-by-two array,
16313 and @samp{@{&"hi", &"there", &"fred"@}} is a three-element array of pointers.
16316 @node C Plus Plus Expressions
16317 @subsubsection C@t{++} Expressions
16319 @cindex expressions in C@t{++}
16320 @value{GDBN} expression handling can interpret most C@t{++} expressions.
16322 @cindex debugging C@t{++} programs
16323 @cindex C@t{++} compilers
16324 @cindex debug formats and C@t{++}
16325 @cindex @value{NGCC} and C@t{++}
16327 @emph{Warning:} @value{GDBN} can only debug C@t{++} code if you use
16328 the proper compiler and the proper debug format. Currently,
16329 @value{GDBN} works best when debugging C@t{++} code that is compiled
16330 with the most recent version of @value{NGCC} possible. The DWARF
16331 debugging format is preferred; @value{NGCC} defaults to this on most
16332 popular platforms. Other compilers and/or debug formats are likely to
16333 work badly or not at all when using @value{GDBN} to debug C@t{++}
16334 code. @xref{Compilation}.
16339 @cindex member functions
16341 Member function calls are allowed; you can use expressions like
16344 count = aml->GetOriginal(x, y)
16347 @vindex this@r{, inside C@t{++} member functions}
16348 @cindex namespace in C@t{++}
16350 While a member function is active (in the selected stack frame), your
16351 expressions have the same namespace available as the member function;
16352 that is, @value{GDBN} allows implicit references to the class instance
16353 pointer @code{this} following the same rules as C@t{++}. @code{using}
16354 declarations in the current scope are also respected by @value{GDBN}.
16356 @cindex call overloaded functions
16357 @cindex overloaded functions, calling
16358 @cindex type conversions in C@t{++}
16360 You can call overloaded functions; @value{GDBN} resolves the function
16361 call to the right definition, with some restrictions. @value{GDBN} does not
16362 perform overload resolution involving user-defined type conversions,
16363 calls to constructors, or instantiations of templates that do not exist
16364 in the program. It also cannot handle ellipsis argument lists or
16367 It does perform integral conversions and promotions, floating-point
16368 promotions, arithmetic conversions, pointer conversions, conversions of
16369 class objects to base classes, and standard conversions such as those of
16370 functions or arrays to pointers; it requires an exact match on the
16371 number of function arguments.
16373 Overload resolution is always performed, unless you have specified
16374 @code{set overload-resolution off}. @xref{Debugging C Plus Plus,
16375 ,@value{GDBN} Features for C@t{++}}.
16377 You must specify @code{set overload-resolution off} in order to use an
16378 explicit function signature to call an overloaded function, as in
16380 p 'foo(char,int)'('x', 13)
16383 The @value{GDBN} command-completion facility can simplify this;
16384 see @ref{Completion, ,Command Completion}.
16386 @cindex reference declarations
16388 @value{GDBN} understands variables declared as C@t{++} lvalue or rvalue
16389 references; you can use them in expressions just as you do in C@t{++}
16390 source---they are automatically dereferenced.
16392 In the parameter list shown when @value{GDBN} displays a frame, the values of
16393 reference variables are not displayed (unlike other variables); this
16394 avoids clutter, since references are often used for large structures.
16395 The @emph{address} of a reference variable is always shown, unless
16396 you have specified @samp{set print address off}.
16399 @value{GDBN} supports the C@t{++} name resolution operator @code{::}---your
16400 expressions can use it just as expressions in your program do. Since
16401 one scope may be defined in another, you can use @code{::} repeatedly if
16402 necessary, for example in an expression like
16403 @samp{@var{scope1}::@var{scope2}::@var{name}}. @value{GDBN} also allows
16404 resolving name scope by reference to source files, in both C and C@t{++}
16405 debugging (@pxref{Variables, ,Program Variables}).
16408 @value{GDBN} performs argument-dependent lookup, following the C@t{++}
16413 @subsubsection C and C@t{++} Defaults
16415 @cindex C and C@t{++} defaults
16417 If you allow @value{GDBN} to set range checking automatically, it
16418 defaults to @code{off} whenever the working language changes to
16419 C or C@t{++}. This happens regardless of whether you or @value{GDBN}
16420 selects the working language.
16422 If you allow @value{GDBN} to set the language automatically, it
16423 recognizes source files whose names end with @file{.c}, @file{.C}, or
16424 @file{.cc}, etc, and when @value{GDBN} enters code compiled from one of
16425 these files, it sets the working language to C or C@t{++}.
16426 @xref{Automatically, ,Having @value{GDBN} Infer the Source Language},
16427 for further details.
16430 @subsubsection C and C@t{++} Type and Range Checks
16432 @cindex C and C@t{++} checks
16434 By default, when @value{GDBN} parses C or C@t{++} expressions, strict type
16435 checking is used. However, if you turn type checking off, @value{GDBN}
16436 will allow certain non-standard conversions, such as promoting integer
16437 constants to pointers.
16439 Range checking, if turned on, is done on mathematical operations. Array
16440 indices are not checked, since they are often used to index a pointer
16441 that is not itself an array.
16444 @subsubsection @value{GDBN} and C
16446 The @code{set print union} and @code{show print union} commands apply to
16447 the @code{union} type. When set to @samp{on}, any @code{union} that is
16448 inside a @code{struct} or @code{class} is also printed. Otherwise, it
16449 appears as @samp{@{...@}}.
16451 The @code{@@} operator aids in the debugging of dynamic arrays, formed
16452 with pointers and a memory allocation function. @xref{Expressions,
16455 @node Debugging C Plus Plus
16456 @subsubsection @value{GDBN} Features for C@t{++}
16458 @cindex commands for C@t{++}
16460 Some @value{GDBN} commands are particularly useful with C@t{++}, and some are
16461 designed specifically for use with C@t{++}. Here is a summary:
16464 @cindex break in overloaded functions
16465 @item @r{breakpoint menus}
16466 When you want a breakpoint in a function whose name is overloaded,
16467 @value{GDBN} has the capability to display a menu of possible breakpoint
16468 locations to help you specify which function definition you want.
16469 @xref{Ambiguous Expressions,,Ambiguous Expressions}.
16471 @cindex overloading in C@t{++}
16472 @item rbreak @var{regex}
16473 Setting breakpoints using regular expressions is helpful for setting
16474 breakpoints on overloaded functions that are not members of any special
16476 @xref{Set Breaks, ,Setting Breakpoints}.
16478 @cindex C@t{++} exception handling
16480 @itemx catch rethrow
16482 Debug C@t{++} exception handling using these commands. @xref{Set
16483 Catchpoints, , Setting Catchpoints}.
16485 @cindex inheritance
16486 @item ptype @var{typename}
16487 Print inheritance relationships as well as other information for type
16489 @xref{Symbols, ,Examining the Symbol Table}.
16491 @item info vtbl @var{expression}.
16492 The @code{info vtbl} command can be used to display the virtual
16493 method tables of the object computed by @var{expression}. This shows
16494 one entry per virtual table; there may be multiple virtual tables when
16495 multiple inheritance is in use.
16497 @cindex C@t{++} demangling
16498 @item demangle @var{name}
16499 Demangle @var{name}.
16500 @xref{Symbols}, for a more complete description of the @code{demangle} command.
16502 @cindex C@t{++} symbol display
16503 @item set print demangle
16504 @itemx show print demangle
16505 @itemx set print asm-demangle
16506 @itemx show print asm-demangle
16507 Control whether C@t{++} symbols display in their source form, both when
16508 displaying code as C@t{++} source and when displaying disassemblies.
16509 @xref{Print Settings, ,Print Settings}.
16511 @item set print object
16512 @itemx show print object
16513 Choose whether to print derived (actual) or declared types of objects.
16514 @xref{Print Settings, ,Print Settings}.
16516 @item set print vtbl
16517 @itemx show print vtbl
16518 Control the format for printing virtual function tables.
16519 @xref{Print Settings, ,Print Settings}.
16520 (The @code{vtbl} commands do not work on programs compiled with the HP
16521 ANSI C@t{++} compiler (@code{aCC}).)
16523 @kindex set overload-resolution
16524 @cindex overloaded functions, overload resolution
16525 @item set overload-resolution on
16526 Enable overload resolution for C@t{++} expression evaluation. The default
16527 is on. For overloaded functions, @value{GDBN} evaluates the arguments
16528 and searches for a function whose signature matches the argument types,
16529 using the standard C@t{++} conversion rules (see @ref{C Plus Plus
16530 Expressions, ,C@t{++} Expressions}, for details).
16531 If it cannot find a match, it emits a message.
16533 @item set overload-resolution off
16534 Disable overload resolution for C@t{++} expression evaluation. For
16535 overloaded functions that are not class member functions, @value{GDBN}
16536 chooses the first function of the specified name that it finds in the
16537 symbol table, whether or not its arguments are of the correct type. For
16538 overloaded functions that are class member functions, @value{GDBN}
16539 searches for a function whose signature @emph{exactly} matches the
16542 @kindex show overload-resolution
16543 @item show overload-resolution
16544 Show the current setting of overload resolution.
16546 @item @r{Overloaded symbol names}
16547 You can specify a particular definition of an overloaded symbol, using
16548 the same notation that is used to declare such symbols in C@t{++}: type
16549 @code{@var{symbol}(@var{types})} rather than just @var{symbol}. You can
16550 also use the @value{GDBN} command-line word completion facilities to list the
16551 available choices, or to finish the type list for you.
16552 @xref{Completion,, Command Completion}, for details on how to do this.
16554 @item @r{Breakpoints in functions with ABI tags}
16556 The GNU C@t{++} compiler introduced the notion of ABI ``tags'', which
16557 correspond to changes in the ABI of a type, function, or variable that
16558 would not otherwise be reflected in a mangled name. See
16559 @url{https://developers.redhat.com/blog/2015/02/05/gcc5-and-the-c11-abi/}
16562 The ABI tags are visible in C@t{++} demangled names. For example, a
16563 function that returns a std::string:
16566 std::string function(int);
16570 when compiled for the C++11 ABI is marked with the @code{cxx11} ABI
16571 tag, and @value{GDBN} displays the symbol like this:
16574 function[abi:cxx11](int)
16577 You can set a breakpoint on such functions simply as if they had no
16581 (@value{GDBP}) b function(int)
16582 Breakpoint 2 at 0x40060d: file main.cc, line 10.
16583 (@value{GDBP}) info breakpoints
16584 Num Type Disp Enb Address What
16585 1 breakpoint keep y 0x0040060d in function[abi:cxx11](int)
16589 On the rare occasion you need to disambiguate between different ABI
16590 tags, you can do so by simply including the ABI tag in the function
16594 (@value{GDBP}) b ambiguous[abi:other_tag](int)
16598 @node Decimal Floating Point
16599 @subsubsection Decimal Floating Point format
16600 @cindex decimal floating point format
16602 @value{GDBN} can examine, set and perform computations with numbers in
16603 decimal floating point format, which in the C language correspond to the
16604 @code{_Decimal32}, @code{_Decimal64} and @code{_Decimal128} types as
16605 specified by the extension to support decimal floating-point arithmetic.
16607 There are two encodings in use, depending on the architecture: BID (Binary
16608 Integer Decimal) for x86 and x86-64, and DPD (Densely Packed Decimal) for
16609 PowerPC and S/390. @value{GDBN} will use the appropriate encoding for the
16612 Because of a limitation in @file{libdecnumber}, the library used by @value{GDBN}
16613 to manipulate decimal floating point numbers, it is not possible to convert
16614 (using a cast, for example) integers wider than 32-bit to decimal float.
16616 In addition, in order to imitate @value{GDBN}'s behaviour with binary floating
16617 point computations, error checking in decimal float operations ignores
16618 underflow, overflow and divide by zero exceptions.
16620 In the PowerPC architecture, @value{GDBN} provides a set of pseudo-registers
16621 to inspect @code{_Decimal128} values stored in floating point registers.
16622 See @ref{PowerPC,,PowerPC} for more details.
16628 @value{GDBN} can be used to debug programs written in D and compiled with
16629 GDC, LDC or DMD compilers. Currently @value{GDBN} supports only one D
16630 specific feature --- dynamic arrays.
16635 @cindex Go (programming language)
16636 @value{GDBN} can be used to debug programs written in Go and compiled with
16637 @file{gccgo} or @file{6g} compilers.
16639 Here is a summary of the Go-specific features and restrictions:
16642 @cindex current Go package
16643 @item The current Go package
16644 The name of the current package does not need to be specified when
16645 specifying global variables and functions.
16647 For example, given the program:
16651 var myglob = "Shall we?"
16657 When stopped inside @code{main} either of these work:
16660 (@value{GDBP}) p myglob
16661 (@value{GDBP}) p main.myglob
16664 @cindex builtin Go types
16665 @item Builtin Go types
16666 The @code{string} type is recognized by @value{GDBN} and is printed
16669 @cindex builtin Go functions
16670 @item Builtin Go functions
16671 The @value{GDBN} expression parser recognizes the @code{unsafe.Sizeof}
16672 function and handles it internally.
16674 @cindex restrictions on Go expressions
16675 @item Restrictions on Go expressions
16676 All Go operators are supported except @code{&^}.
16677 The Go @code{_} ``blank identifier'' is not supported.
16678 Automatic dereferencing of pointers is not supported.
16682 @subsection Objective-C
16684 @cindex Objective-C
16685 This section provides information about some commands and command
16686 options that are useful for debugging Objective-C code. See also
16687 @ref{Symbols, info classes}, and @ref{Symbols, info selectors}, for a
16688 few more commands specific to Objective-C support.
16691 * Method Names in Commands::
16692 * The Print Command with Objective-C::
16695 @node Method Names in Commands
16696 @subsubsection Method Names in Commands
16698 The following commands have been extended to accept Objective-C method
16699 names as line specifications:
16701 @kindex clear@r{, and Objective-C}
16702 @kindex break@r{, and Objective-C}
16703 @kindex info line@r{, and Objective-C}
16704 @kindex jump@r{, and Objective-C}
16705 @kindex list@r{, and Objective-C}
16709 @item @code{info line}
16714 A fully qualified Objective-C method name is specified as
16717 -[@var{Class} @var{methodName}]
16720 where the minus sign is used to indicate an instance method and a
16721 plus sign (not shown) is used to indicate a class method. The class
16722 name @var{Class} and method name @var{methodName} are enclosed in
16723 brackets, similar to the way messages are specified in Objective-C
16724 source code. For example, to set a breakpoint at the @code{create}
16725 instance method of class @code{Fruit} in the program currently being
16729 break -[Fruit create]
16732 To list ten program lines around the @code{initialize} class method,
16736 list +[NSText initialize]
16739 In the current version of @value{GDBN}, the plus or minus sign is
16740 required. In future versions of @value{GDBN}, the plus or minus
16741 sign will be optional, but you can use it to narrow the search. It
16742 is also possible to specify just a method name:
16748 You must specify the complete method name, including any colons. If
16749 your program's source files contain more than one @code{create} method,
16750 you'll be presented with a numbered list of classes that implement that
16751 method. Indicate your choice by number, or type @samp{0} to exit if
16754 As another example, to clear a breakpoint established at the
16755 @code{makeKeyAndOrderFront:} method of the @code{NSWindow} class, enter:
16758 clear -[NSWindow makeKeyAndOrderFront:]
16761 @node The Print Command with Objective-C
16762 @subsubsection The Print Command With Objective-C
16763 @cindex Objective-C, print objects
16764 @kindex print-object
16765 @kindex po @r{(@code{print-object})}
16767 The print command has also been extended to accept methods. For example:
16770 print -[@var{object} hash]
16773 @cindex print an Objective-C object description
16774 @cindex @code{_NSPrintForDebugger}, and printing Objective-C objects
16776 will tell @value{GDBN} to send the @code{hash} message to @var{object}
16777 and print the result. Also, an additional command has been added,
16778 @code{print-object} or @code{po} for short, which is meant to print
16779 the description of an object. However, this command may only work
16780 with certain Objective-C libraries that have a particular hook
16781 function, @code{_NSPrintForDebugger}, defined.
16784 @subsection OpenCL C
16787 This section provides information about @value{GDBN}s OpenCL C support.
16790 * OpenCL C Datatypes::
16791 * OpenCL C Expressions::
16792 * OpenCL C Operators::
16795 @node OpenCL C Datatypes
16796 @subsubsection OpenCL C Datatypes
16798 @cindex OpenCL C Datatypes
16799 @value{GDBN} supports the builtin scalar and vector datatypes specified
16800 by OpenCL 1.1. In addition the half- and double-precision floating point
16801 data types of the @code{cl_khr_fp16} and @code{cl_khr_fp64} OpenCL
16802 extensions are also known to @value{GDBN}.
16804 @node OpenCL C Expressions
16805 @subsubsection OpenCL C Expressions
16807 @cindex OpenCL C Expressions
16808 @value{GDBN} supports accesses to vector components including the access as
16809 lvalue where possible. Since OpenCL C is based on C99 most C expressions
16810 supported by @value{GDBN} can be used as well.
16812 @node OpenCL C Operators
16813 @subsubsection OpenCL C Operators
16815 @cindex OpenCL C Operators
16816 @value{GDBN} supports the operators specified by OpenCL 1.1 for scalar and
16820 @subsection Fortran
16821 @cindex Fortran-specific support in @value{GDBN}
16823 @value{GDBN} can be used to debug programs written in Fortran, but it
16824 currently supports only the features of Fortran 77 language.
16826 @cindex trailing underscore, in Fortran symbols
16827 Some Fortran compilers (@sc{gnu} Fortran 77 and Fortran 95 compilers
16828 among them) append an underscore to the names of variables and
16829 functions. When you debug programs compiled by those compilers, you
16830 will need to refer to variables and functions with a trailing
16834 * Fortran Operators:: Fortran operators and expressions
16835 * Fortran Defaults:: Default settings for Fortran
16836 * Special Fortran Commands:: Special @value{GDBN} commands for Fortran
16839 @node Fortran Operators
16840 @subsubsection Fortran Operators and Expressions
16842 @cindex Fortran operators and expressions
16844 Operators must be defined on values of specific types. For instance,
16845 @code{+} is defined on numbers, but not on characters or other non-
16846 arithmetic types. Operators are often defined on groups of types.
16850 The exponentiation operator. It raises the first operand to the power
16854 The range operator. Normally used in the form of array(low:high) to
16855 represent a section of array.
16858 The access component operator. Normally used to access elements in derived
16859 types. Also suitable for unions. As unions aren't part of regular Fortran,
16860 this can only happen when accessing a register that uses a gdbarch-defined
16863 The scope operator. Normally used to access variables in modules or
16864 to set breakpoints on subroutines nested in modules or in other
16865 subroutines (internal subroutines).
16868 @node Fortran Defaults
16869 @subsubsection Fortran Defaults
16871 @cindex Fortran Defaults
16873 Fortran symbols are usually case-insensitive, so @value{GDBN} by
16874 default uses case-insensitive matches for Fortran symbols. You can
16875 change that with the @samp{set case-insensitive} command, see
16876 @ref{Symbols}, for the details.
16878 @node Special Fortran Commands
16879 @subsubsection Special Fortran Commands
16881 @cindex Special Fortran commands
16883 @value{GDBN} has some commands to support Fortran-specific features,
16884 such as displaying common blocks.
16887 @cindex @code{COMMON} blocks, Fortran
16888 @kindex info common
16889 @item info common @r{[}@var{common-name}@r{]}
16890 This command prints the values contained in the Fortran @code{COMMON}
16891 block whose name is @var{common-name}. With no argument, the names of
16892 all @code{COMMON} blocks visible at the current program location are
16899 @cindex Pascal support in @value{GDBN}, limitations
16900 Debugging Pascal programs which use sets, subranges, file variables, or
16901 nested functions does not currently work. @value{GDBN} does not support
16902 entering expressions, printing values, or similar features using Pascal
16905 The Pascal-specific command @code{set print pascal_static-members}
16906 controls whether static members of Pascal objects are displayed.
16907 @xref{Print Settings, pascal_static-members}.
16912 @value{GDBN} supports the @url{https://www.rust-lang.org/, Rust
16913 Programming Language}. Type- and value-printing, and expression
16914 parsing, are reasonably complete. However, there are a few
16915 peculiarities and holes to be aware of.
16919 Linespecs (@pxref{Specify Location}) are never relative to the current
16920 crate. Instead, they act as if there were a global namespace of
16921 crates, somewhat similar to the way @code{extern crate} behaves.
16923 That is, if @value{GDBN} is stopped at a breakpoint in a function in
16924 crate @samp{A}, module @samp{B}, then @code{break B::f} will attempt
16925 to set a breakpoint in a function named @samp{f} in a crate named
16928 As a consequence of this approach, linespecs also cannot refer to
16929 items using @samp{self::} or @samp{super::}.
16932 Because @value{GDBN} implements Rust name-lookup semantics in
16933 expressions, it will sometimes prepend the current crate to a name.
16934 For example, if @value{GDBN} is stopped at a breakpoint in the crate
16935 @samp{K}, then @code{print ::x::y} will try to find the symbol
16938 However, since it is useful to be able to refer to other crates when
16939 debugging, @value{GDBN} provides the @code{extern} extension to
16940 circumvent this. To use the extension, just put @code{extern} before
16941 a path expression to refer to the otherwise unavailable ``global''
16944 In the above example, if you wanted to refer to the symbol @samp{y} in
16945 the crate @samp{x}, you would use @code{print extern x::y}.
16948 The Rust expression evaluator does not support ``statement-like''
16949 expressions such as @code{if} or @code{match}, or lambda expressions.
16952 Tuple expressions are not implemented.
16955 The Rust expression evaluator does not currently implement the
16956 @code{Drop} trait. Objects that may be created by the evaluator will
16957 never be destroyed.
16960 @value{GDBN} does not implement type inference for generics. In order
16961 to call generic functions or otherwise refer to generic items, you
16962 will have to specify the type parameters manually.
16965 @value{GDBN} currently uses the C@t{++} demangler for Rust. In most
16966 cases this does not cause any problems. However, in an expression
16967 context, completing a generic function name will give syntactically
16968 invalid results. This happens because Rust requires the @samp{::}
16969 operator between the function name and its generic arguments. For
16970 example, @value{GDBN} might provide a completion like
16971 @code{crate::f<u32>}, where the parser would require
16972 @code{crate::f::<u32>}.
16975 As of this writing, the Rust compiler (version 1.8) has a few holes in
16976 the debugging information it generates. These holes prevent certain
16977 features from being implemented by @value{GDBN}:
16981 Method calls cannot be made via traits.
16984 Operator overloading is not implemented.
16987 When debugging in a monomorphized function, you cannot use the generic
16991 The type @code{Self} is not available.
16994 @code{use} statements are not available, so some names may not be
16995 available in the crate.
17000 @subsection Modula-2
17002 @cindex Modula-2, @value{GDBN} support
17004 The extensions made to @value{GDBN} to support Modula-2 only support
17005 output from the @sc{gnu} Modula-2 compiler (which is currently being
17006 developed). Other Modula-2 compilers are not currently supported, and
17007 attempting to debug executables produced by them is most likely
17008 to give an error as @value{GDBN} reads in the executable's symbol
17011 @cindex expressions in Modula-2
17013 * M2 Operators:: Built-in operators
17014 * Built-In Func/Proc:: Built-in functions and procedures
17015 * M2 Constants:: Modula-2 constants
17016 * M2 Types:: Modula-2 types
17017 * M2 Defaults:: Default settings for Modula-2
17018 * Deviations:: Deviations from standard Modula-2
17019 * M2 Checks:: Modula-2 type and range checks
17020 * M2 Scope:: The scope operators @code{::} and @code{.}
17021 * GDB/M2:: @value{GDBN} and Modula-2
17025 @subsubsection Operators
17026 @cindex Modula-2 operators
17028 Operators must be defined on values of specific types. For instance,
17029 @code{+} is defined on numbers, but not on structures. Operators are
17030 often defined on groups of types. For the purposes of Modula-2, the
17031 following definitions hold:
17036 @emph{Integral types} consist of @code{INTEGER}, @code{CARDINAL}, and
17040 @emph{Character types} consist of @code{CHAR} and its subranges.
17043 @emph{Floating-point types} consist of @code{REAL}.
17046 @emph{Pointer types} consist of anything declared as @code{POINTER TO
17050 @emph{Scalar types} consist of all of the above.
17053 @emph{Set types} consist of @code{SET} and @code{BITSET} types.
17056 @emph{Boolean types} consist of @code{BOOLEAN}.
17060 The following operators are supported, and appear in order of
17061 increasing precedence:
17065 Function argument or array index separator.
17068 Assignment. The value of @var{var} @code{:=} @var{value} is
17072 Less than, greater than on integral, floating-point, or enumerated
17076 Less than or equal to, greater than or equal to
17077 on integral, floating-point and enumerated types, or set inclusion on
17078 set types. Same precedence as @code{<}.
17080 @item =@r{, }<>@r{, }#
17081 Equality and two ways of expressing inequality, valid on scalar types.
17082 Same precedence as @code{<}. In @value{GDBN} scripts, only @code{<>} is
17083 available for inequality, since @code{#} conflicts with the script
17087 Set membership. Defined on set types and the types of their members.
17088 Same precedence as @code{<}.
17091 Boolean disjunction. Defined on boolean types.
17094 Boolean conjunction. Defined on boolean types.
17097 The @value{GDBN} ``artificial array'' operator (@pxref{Expressions, ,Expressions}).
17100 Addition and subtraction on integral and floating-point types, or union
17101 and difference on set types.
17104 Multiplication on integral and floating-point types, or set intersection
17108 Division on floating-point types, or symmetric set difference on set
17109 types. Same precedence as @code{*}.
17112 Integer division and remainder. Defined on integral types. Same
17113 precedence as @code{*}.
17116 Negative. Defined on @code{INTEGER} and @code{REAL} data.
17119 Pointer dereferencing. Defined on pointer types.
17122 Boolean negation. Defined on boolean types. Same precedence as
17126 @code{RECORD} field selector. Defined on @code{RECORD} data. Same
17127 precedence as @code{^}.
17130 Array indexing. Defined on @code{ARRAY} data. Same precedence as @code{^}.
17133 Procedure argument list. Defined on @code{PROCEDURE} objects. Same precedence
17137 @value{GDBN} and Modula-2 scope operators.
17141 @emph{Warning:} Set expressions and their operations are not yet supported, so @value{GDBN}
17142 treats the use of the operator @code{IN}, or the use of operators
17143 @code{+}, @code{-}, @code{*}, @code{/}, @code{=}, , @code{<>}, @code{#},
17144 @code{<=}, and @code{>=} on sets as an error.
17148 @node Built-In Func/Proc
17149 @subsubsection Built-in Functions and Procedures
17150 @cindex Modula-2 built-ins
17152 Modula-2 also makes available several built-in procedures and functions.
17153 In describing these, the following metavariables are used:
17158 represents an @code{ARRAY} variable.
17161 represents a @code{CHAR} constant or variable.
17164 represents a variable or constant of integral type.
17167 represents an identifier that belongs to a set. Generally used in the
17168 same function with the metavariable @var{s}. The type of @var{s} should
17169 be @code{SET OF @var{mtype}} (where @var{mtype} is the type of @var{m}).
17172 represents a variable or constant of integral or floating-point type.
17175 represents a variable or constant of floating-point type.
17181 represents a variable.
17184 represents a variable or constant of one of many types. See the
17185 explanation of the function for details.
17188 All Modula-2 built-in procedures also return a result, described below.
17192 Returns the absolute value of @var{n}.
17195 If @var{c} is a lower case letter, it returns its upper case
17196 equivalent, otherwise it returns its argument.
17199 Returns the character whose ordinal value is @var{i}.
17202 Decrements the value in the variable @var{v} by one. Returns the new value.
17204 @item DEC(@var{v},@var{i})
17205 Decrements the value in the variable @var{v} by @var{i}. Returns the
17208 @item EXCL(@var{m},@var{s})
17209 Removes the element @var{m} from the set @var{s}. Returns the new
17212 @item FLOAT(@var{i})
17213 Returns the floating point equivalent of the integer @var{i}.
17215 @item HIGH(@var{a})
17216 Returns the index of the last member of @var{a}.
17219 Increments the value in the variable @var{v} by one. Returns the new value.
17221 @item INC(@var{v},@var{i})
17222 Increments the value in the variable @var{v} by @var{i}. Returns the
17225 @item INCL(@var{m},@var{s})
17226 Adds the element @var{m} to the set @var{s} if it is not already
17227 there. Returns the new set.
17230 Returns the maximum value of the type @var{t}.
17233 Returns the minimum value of the type @var{t}.
17236 Returns boolean TRUE if @var{i} is an odd number.
17239 Returns the ordinal value of its argument. For example, the ordinal
17240 value of a character is its @sc{ascii} value (on machines supporting
17241 the @sc{ascii} character set). The argument @var{x} must be of an
17242 ordered type, which include integral, character and enumerated types.
17244 @item SIZE(@var{x})
17245 Returns the size of its argument. The argument @var{x} can be a
17246 variable or a type.
17248 @item TRUNC(@var{r})
17249 Returns the integral part of @var{r}.
17251 @item TSIZE(@var{x})
17252 Returns the size of its argument. The argument @var{x} can be a
17253 variable or a type.
17255 @item VAL(@var{t},@var{i})
17256 Returns the member of the type @var{t} whose ordinal value is @var{i}.
17260 @emph{Warning:} Sets and their operations are not yet supported, so
17261 @value{GDBN} treats the use of procedures @code{INCL} and @code{EXCL} as
17265 @cindex Modula-2 constants
17267 @subsubsection Constants
17269 @value{GDBN} allows you to express the constants of Modula-2 in the following
17275 Integer constants are simply a sequence of digits. When used in an
17276 expression, a constant is interpreted to be type-compatible with the
17277 rest of the expression. Hexadecimal integers are specified by a
17278 trailing @samp{H}, and octal integers by a trailing @samp{B}.
17281 Floating point constants appear as a sequence of digits, followed by a
17282 decimal point and another sequence of digits. An optional exponent can
17283 then be specified, in the form @samp{E@r{[}+@r{|}-@r{]}@var{nnn}}, where
17284 @samp{@r{[}+@r{|}-@r{]}@var{nnn}} is the desired exponent. All of the
17285 digits of the floating point constant must be valid decimal (base 10)
17289 Character constants consist of a single character enclosed by a pair of
17290 like quotes, either single (@code{'}) or double (@code{"}). They may
17291 also be expressed by their ordinal value (their @sc{ascii} value, usually)
17292 followed by a @samp{C}.
17295 String constants consist of a sequence of characters enclosed by a
17296 pair of like quotes, either single (@code{'}) or double (@code{"}).
17297 Escape sequences in the style of C are also allowed. @xref{C
17298 Constants, ,C and C@t{++} Constants}, for a brief explanation of escape
17302 Enumerated constants consist of an enumerated identifier.
17305 Boolean constants consist of the identifiers @code{TRUE} and
17309 Pointer constants consist of integral values only.
17312 Set constants are not yet supported.
17316 @subsubsection Modula-2 Types
17317 @cindex Modula-2 types
17319 Currently @value{GDBN} can print the following data types in Modula-2
17320 syntax: array types, record types, set types, pointer types, procedure
17321 types, enumerated types, subrange types and base types. You can also
17322 print the contents of variables declared using these type.
17323 This section gives a number of simple source code examples together with
17324 sample @value{GDBN} sessions.
17326 The first example contains the following section of code:
17335 and you can request @value{GDBN} to interrogate the type and value of
17336 @code{r} and @code{s}.
17339 (@value{GDBP}) print s
17341 (@value{GDBP}) ptype s
17343 (@value{GDBP}) print r
17345 (@value{GDBP}) ptype r
17350 Likewise if your source code declares @code{s} as:
17354 s: SET ['A'..'Z'] ;
17358 then you may query the type of @code{s} by:
17361 (@value{GDBP}) ptype s
17362 type = SET ['A'..'Z']
17366 Note that at present you cannot interactively manipulate set
17367 expressions using the debugger.
17369 The following example shows how you might declare an array in Modula-2
17370 and how you can interact with @value{GDBN} to print its type and contents:
17374 s: ARRAY [-10..10] OF CHAR ;
17378 (@value{GDBP}) ptype s
17379 ARRAY [-10..10] OF CHAR
17382 Note that the array handling is not yet complete and although the type
17383 is printed correctly, expression handling still assumes that all
17384 arrays have a lower bound of zero and not @code{-10} as in the example
17387 Here are some more type related Modula-2 examples:
17391 colour = (blue, red, yellow, green) ;
17392 t = [blue..yellow] ;
17400 The @value{GDBN} interaction shows how you can query the data type
17401 and value of a variable.
17404 (@value{GDBP}) print s
17406 (@value{GDBP}) ptype t
17407 type = [blue..yellow]
17411 In this example a Modula-2 array is declared and its contents
17412 displayed. Observe that the contents are written in the same way as
17413 their @code{C} counterparts.
17417 s: ARRAY [1..5] OF CARDINAL ;
17423 (@value{GDBP}) print s
17424 $1 = @{1, 0, 0, 0, 0@}
17425 (@value{GDBP}) ptype s
17426 type = ARRAY [1..5] OF CARDINAL
17429 The Modula-2 language interface to @value{GDBN} also understands
17430 pointer types as shown in this example:
17434 s: POINTER TO ARRAY [1..5] OF CARDINAL ;
17441 and you can request that @value{GDBN} describes the type of @code{s}.
17444 (@value{GDBP}) ptype s
17445 type = POINTER TO ARRAY [1..5] OF CARDINAL
17448 @value{GDBN} handles compound types as we can see in this example.
17449 Here we combine array types, record types, pointer types and subrange
17460 myarray = ARRAY myrange OF CARDINAL ;
17461 myrange = [-2..2] ;
17463 s: POINTER TO ARRAY myrange OF foo ;
17467 and you can ask @value{GDBN} to describe the type of @code{s} as shown
17471 (@value{GDBP}) ptype s
17472 type = POINTER TO ARRAY [-2..2] OF foo = RECORD
17475 f3 : ARRAY [-2..2] OF CARDINAL;
17480 @subsubsection Modula-2 Defaults
17481 @cindex Modula-2 defaults
17483 If type and range checking are set automatically by @value{GDBN}, they
17484 both default to @code{on} whenever the working language changes to
17485 Modula-2. This happens regardless of whether you or @value{GDBN}
17486 selected the working language.
17488 If you allow @value{GDBN} to set the language automatically, then entering
17489 code compiled from a file whose name ends with @file{.mod} sets the
17490 working language to Modula-2. @xref{Automatically, ,Having @value{GDBN}
17491 Infer the Source Language}, for further details.
17494 @subsubsection Deviations from Standard Modula-2
17495 @cindex Modula-2, deviations from
17497 A few changes have been made to make Modula-2 programs easier to debug.
17498 This is done primarily via loosening its type strictness:
17502 Unlike in standard Modula-2, pointer constants can be formed by
17503 integers. This allows you to modify pointer variables during
17504 debugging. (In standard Modula-2, the actual address contained in a
17505 pointer variable is hidden from you; it can only be modified
17506 through direct assignment to another pointer variable or expression that
17507 returned a pointer.)
17510 C escape sequences can be used in strings and characters to represent
17511 non-printable characters. @value{GDBN} prints out strings with these
17512 escape sequences embedded. Single non-printable characters are
17513 printed using the @samp{CHR(@var{nnn})} format.
17516 The assignment operator (@code{:=}) returns the value of its right-hand
17520 All built-in procedures both modify @emph{and} return their argument.
17524 @subsubsection Modula-2 Type and Range Checks
17525 @cindex Modula-2 checks
17528 @emph{Warning:} in this release, @value{GDBN} does not yet perform type or
17531 @c FIXME remove warning when type/range checks added
17533 @value{GDBN} considers two Modula-2 variables type equivalent if:
17537 They are of types that have been declared equivalent via a @code{TYPE
17538 @var{t1} = @var{t2}} statement
17541 They have been declared on the same line. (Note: This is true of the
17542 @sc{gnu} Modula-2 compiler, but it may not be true of other compilers.)
17545 As long as type checking is enabled, any attempt to combine variables
17546 whose types are not equivalent is an error.
17548 Range checking is done on all mathematical operations, assignment, array
17549 index bounds, and all built-in functions and procedures.
17552 @subsubsection The Scope Operators @code{::} and @code{.}
17554 @cindex @code{.}, Modula-2 scope operator
17555 @cindex colon, doubled as scope operator
17557 @vindex colon-colon@r{, in Modula-2}
17558 @c Info cannot handle :: but TeX can.
17561 @vindex ::@r{, in Modula-2}
17564 There are a few subtle differences between the Modula-2 scope operator
17565 (@code{.}) and the @value{GDBN} scope operator (@code{::}). The two have
17570 @var{module} . @var{id}
17571 @var{scope} :: @var{id}
17575 where @var{scope} is the name of a module or a procedure,
17576 @var{module} the name of a module, and @var{id} is any declared
17577 identifier within your program, except another module.
17579 Using the @code{::} operator makes @value{GDBN} search the scope
17580 specified by @var{scope} for the identifier @var{id}. If it is not
17581 found in the specified scope, then @value{GDBN} searches all scopes
17582 enclosing the one specified by @var{scope}.
17584 Using the @code{.} operator makes @value{GDBN} search the current scope for
17585 the identifier specified by @var{id} that was imported from the
17586 definition module specified by @var{module}. With this operator, it is
17587 an error if the identifier @var{id} was not imported from definition
17588 module @var{module}, or if @var{id} is not an identifier in
17592 @subsubsection @value{GDBN} and Modula-2
17594 Some @value{GDBN} commands have little use when debugging Modula-2 programs.
17595 Five subcommands of @code{set print} and @code{show print} apply
17596 specifically to C and C@t{++}: @samp{vtbl}, @samp{demangle},
17597 @samp{asm-demangle}, @samp{object}, and @samp{union}. The first four
17598 apply to C@t{++}, and the last to the C @code{union} type, which has no direct
17599 analogue in Modula-2.
17601 The @code{@@} operator (@pxref{Expressions, ,Expressions}), while available
17602 with any language, is not useful with Modula-2. Its
17603 intent is to aid the debugging of @dfn{dynamic arrays}, which cannot be
17604 created in Modula-2 as they can in C or C@t{++}. However, because an
17605 address can be specified by an integral constant, the construct
17606 @samp{@{@var{type}@}@var{adrexp}} is still useful.
17608 @cindex @code{#} in Modula-2
17609 In @value{GDBN} scripts, the Modula-2 inequality operator @code{#} is
17610 interpreted as the beginning of a comment. Use @code{<>} instead.
17616 The extensions made to @value{GDBN} for Ada only support
17617 output from the @sc{gnu} Ada (GNAT) compiler.
17618 Other Ada compilers are not currently supported, and
17619 attempting to debug executables produced by them is most likely
17623 @cindex expressions in Ada
17625 * Ada Mode Intro:: General remarks on the Ada syntax
17626 and semantics supported by Ada mode
17628 * Omissions from Ada:: Restrictions on the Ada expression syntax.
17629 * Additions to Ada:: Extensions of the Ada expression syntax.
17630 * Overloading support for Ada:: Support for expressions involving overloaded
17632 * Stopping Before Main Program:: Debugging the program during elaboration.
17633 * Ada Exceptions:: Ada Exceptions
17634 * Ada Tasks:: Listing and setting breakpoints in tasks.
17635 * Ada Tasks and Core Files:: Tasking Support when Debugging Core Files
17636 * Ravenscar Profile:: Tasking Support when using the Ravenscar
17638 * Ada Settings:: New settable GDB parameters for Ada.
17639 * Ada Glitches:: Known peculiarities of Ada mode.
17642 @node Ada Mode Intro
17643 @subsubsection Introduction
17644 @cindex Ada mode, general
17646 The Ada mode of @value{GDBN} supports a fairly large subset of Ada expression
17647 syntax, with some extensions.
17648 The philosophy behind the design of this subset is
17652 That @value{GDBN} should provide basic literals and access to operations for
17653 arithmetic, dereferencing, field selection, indexing, and subprogram calls,
17654 leaving more sophisticated computations to subprograms written into the
17655 program (which therefore may be called from @value{GDBN}).
17658 That type safety and strict adherence to Ada language restrictions
17659 are not particularly important to the @value{GDBN} user.
17662 That brevity is important to the @value{GDBN} user.
17665 Thus, for brevity, the debugger acts as if all names declared in
17666 user-written packages are directly visible, even if they are not visible
17667 according to Ada rules, thus making it unnecessary to fully qualify most
17668 names with their packages, regardless of context. Where this causes
17669 ambiguity, @value{GDBN} asks the user's intent.
17671 The debugger will start in Ada mode if it detects an Ada main program.
17672 As for other languages, it will enter Ada mode when stopped in a program that
17673 was translated from an Ada source file.
17675 While in Ada mode, you may use `@t{--}' for comments. This is useful
17676 mostly for documenting command files. The standard @value{GDBN} comment
17677 (@samp{#}) still works at the beginning of a line in Ada mode, but not in the
17678 middle (to allow based literals).
17680 @node Omissions from Ada
17681 @subsubsection Omissions from Ada
17682 @cindex Ada, omissions from
17684 Here are the notable omissions from the subset:
17688 Only a subset of the attributes are supported:
17692 @t{'First}, @t{'Last}, and @t{'Length}
17693 on array objects (not on types and subtypes).
17696 @t{'Min} and @t{'Max}.
17699 @t{'Pos} and @t{'Val}.
17705 @t{'Range} on array objects (not subtypes), but only as the right
17706 operand of the membership (@code{in}) operator.
17709 @t{'Access}, @t{'Unchecked_Access}, and
17710 @t{'Unrestricted_Access} (a GNAT extension).
17718 @code{Characters.Latin_1} are not available and
17719 concatenation is not implemented. Thus, escape characters in strings are
17720 not currently available.
17723 Equality tests (@samp{=} and @samp{/=}) on arrays test for bitwise
17724 equality of representations. They will generally work correctly
17725 for strings and arrays whose elements have integer or enumeration types.
17726 They may not work correctly for arrays whose element
17727 types have user-defined equality, for arrays of real values
17728 (in particular, IEEE-conformant floating point, because of negative
17729 zeroes and NaNs), and for arrays whose elements contain unused bits with
17730 indeterminate values.
17733 The other component-by-component array operations (@code{and}, @code{or},
17734 @code{xor}, @code{not}, and relational tests other than equality)
17735 are not implemented.
17738 @cindex array aggregates (Ada)
17739 @cindex record aggregates (Ada)
17740 @cindex aggregates (Ada)
17741 There is limited support for array and record aggregates. They are
17742 permitted only on the right sides of assignments, as in these examples:
17745 (@value{GDBP}) set An_Array := (1, 2, 3, 4, 5, 6)
17746 (@value{GDBP}) set An_Array := (1, others => 0)
17747 (@value{GDBP}) set An_Array := (0|4 => 1, 1..3 => 2, 5 => 6)
17748 (@value{GDBP}) set A_2D_Array := ((1, 2, 3), (4, 5, 6), (7, 8, 9))
17749 (@value{GDBP}) set A_Record := (1, "Peter", True);
17750 (@value{GDBP}) set A_Record := (Name => "Peter", Id => 1, Alive => True)
17754 discriminant's value by assigning an aggregate has an
17755 undefined effect if that discriminant is used within the record.
17756 However, you can first modify discriminants by directly assigning to
17757 them (which normally would not be allowed in Ada), and then performing an
17758 aggregate assignment. For example, given a variable @code{A_Rec}
17759 declared to have a type such as:
17762 type Rec (Len : Small_Integer := 0) is record
17764 Vals : IntArray (1 .. Len);
17768 you can assign a value with a different size of @code{Vals} with two
17772 (@value{GDBP}) set A_Rec.Len := 4
17773 (@value{GDBP}) set A_Rec := (Id => 42, Vals => (1, 2, 3, 4))
17776 As this example also illustrates, @value{GDBN} is very loose about the usual
17777 rules concerning aggregates. You may leave out some of the
17778 components of an array or record aggregate (such as the @code{Len}
17779 component in the assignment to @code{A_Rec} above); they will retain their
17780 original values upon assignment. You may freely use dynamic values as
17781 indices in component associations. You may even use overlapping or
17782 redundant component associations, although which component values are
17783 assigned in such cases is not defined.
17786 Calls to dispatching subprograms are not implemented.
17789 The overloading algorithm is much more limited (i.e., less selective)
17790 than that of real Ada. It makes only limited use of the context in
17791 which a subexpression appears to resolve its meaning, and it is much
17792 looser in its rules for allowing type matches. As a result, some
17793 function calls will be ambiguous, and the user will be asked to choose
17794 the proper resolution.
17797 The @code{new} operator is not implemented.
17800 Entry calls are not implemented.
17803 Aside from printing, arithmetic operations on the native VAX floating-point
17804 formats are not supported.
17807 It is not possible to slice a packed array.
17810 The names @code{True} and @code{False}, when not part of a qualified name,
17811 are interpreted as if implicitly prefixed by @code{Standard}, regardless of
17813 Should your program
17814 redefine these names in a package or procedure (at best a dubious practice),
17815 you will have to use fully qualified names to access their new definitions.
17818 @node Additions to Ada
17819 @subsubsection Additions to Ada
17820 @cindex Ada, deviations from
17822 As it does for other languages, @value{GDBN} makes certain generic
17823 extensions to Ada (@pxref{Expressions}):
17827 If the expression @var{E} is a variable residing in memory (typically
17828 a local variable or array element) and @var{N} is a positive integer,
17829 then @code{@var{E}@@@var{N}} displays the values of @var{E} and the
17830 @var{N}-1 adjacent variables following it in memory as an array. In
17831 Ada, this operator is generally not necessary, since its prime use is
17832 in displaying parts of an array, and slicing will usually do this in
17833 Ada. However, there are occasional uses when debugging programs in
17834 which certain debugging information has been optimized away.
17837 @code{@var{B}::@var{var}} means ``the variable named @var{var} that
17838 appears in function or file @var{B}.'' When @var{B} is a file name,
17839 you must typically surround it in single quotes.
17842 The expression @code{@{@var{type}@} @var{addr}} means ``the variable of type
17843 @var{type} that appears at address @var{addr}.''
17846 A name starting with @samp{$} is a convenience variable
17847 (@pxref{Convenience Vars}) or a machine register (@pxref{Registers}).
17850 In addition, @value{GDBN} provides a few other shortcuts and outright
17851 additions specific to Ada:
17855 The assignment statement is allowed as an expression, returning
17856 its right-hand operand as its value. Thus, you may enter
17859 (@value{GDBP}) set x := y + 3
17860 (@value{GDBP}) print A(tmp := y + 1)
17864 The semicolon is allowed as an ``operator,'' returning as its value
17865 the value of its right-hand operand.
17866 This allows, for example,
17867 complex conditional breaks:
17870 (@value{GDBP}) break f
17871 (@value{GDBP}) condition 1 (report(i); k += 1; A(k) > 100)
17875 Rather than use catenation and symbolic character names to introduce special
17876 characters into strings, one may instead use a special bracket notation,
17877 which is also used to print strings. A sequence of characters of the form
17878 @samp{["@var{XX}"]} within a string or character literal denotes the
17879 (single) character whose numeric encoding is @var{XX} in hexadecimal. The
17880 sequence of characters @samp{["""]} also denotes a single quotation mark
17881 in strings. For example,
17883 "One line.["0a"]Next line.["0a"]"
17886 contains an ASCII newline character (@code{Ada.Characters.Latin_1.LF})
17890 The subtype used as a prefix for the attributes @t{'Pos}, @t{'Min}, and
17891 @t{'Max} is optional (and is ignored in any case). For example, it is valid
17895 (@value{GDBP}) print 'max(x, y)
17899 When printing arrays, @value{GDBN} uses positional notation when the
17900 array has a lower bound of 1, and uses a modified named notation otherwise.
17901 For example, a one-dimensional array of three integers with a lower bound
17902 of 3 might print as
17909 That is, in contrast to valid Ada, only the first component has a @code{=>}
17913 You may abbreviate attributes in expressions with any unique,
17914 multi-character subsequence of
17915 their names (an exact match gets preference).
17916 For example, you may use @t{a'len}, @t{a'gth}, or @t{a'lh}
17917 in place of @t{a'length}.
17920 @cindex quoting Ada internal identifiers
17921 Since Ada is case-insensitive, the debugger normally maps identifiers you type
17922 to lower case. The GNAT compiler uses upper-case characters for
17923 some of its internal identifiers, which are normally of no interest to users.
17924 For the rare occasions when you actually have to look at them,
17925 enclose them in angle brackets to avoid the lower-case mapping.
17928 (@value{GDBP}) print <JMPBUF_SAVE>[0]
17932 Printing an object of class-wide type or dereferencing an
17933 access-to-class-wide value will display all the components of the object's
17934 specific type (as indicated by its run-time tag). Likewise, component
17935 selection on such a value will operate on the specific type of the
17940 @node Overloading support for Ada
17941 @subsubsection Overloading support for Ada
17942 @cindex overloading, Ada
17944 The debugger supports limited overloading. Given a subprogram call in which
17945 the function symbol has multiple definitions, it will use the number of
17946 actual parameters and some information about their types to attempt to narrow
17947 the set of definitions. It also makes very limited use of context, preferring
17948 procedures to functions in the context of the @code{call} command, and
17949 functions to procedures elsewhere.
17951 If, after narrowing, the set of matching definitions still contains more than
17952 one definition, @value{GDBN} will display a menu to query which one it should
17956 (@value{GDBP}) print f(1)
17957 Multiple matches for f
17959 [1] foo.f (integer) return boolean at foo.adb:23
17960 [2] foo.f (foo.new_integer) return boolean at foo.adb:28
17964 In this case, just select one menu entry either to cancel expression evaluation
17965 (type @kbd{0} and press @key{RET}) or to continue evaluation with a specific
17966 instance (type the corresponding number and press @key{RET}).
17968 Here are a couple of commands to customize @value{GDBN}'s behavior in this
17973 @kindex set ada print-signatures
17974 @item set ada print-signatures
17975 Control whether parameter types and return types are displayed in overloads
17976 selection menus. It is @code{on} by default.
17977 @xref{Overloading support for Ada}.
17979 @kindex show ada print-signatures
17980 @item show ada print-signatures
17981 Show the current setting for displaying parameter types and return types in
17982 overloads selection menu.
17983 @xref{Overloading support for Ada}.
17987 @node Stopping Before Main Program
17988 @subsubsection Stopping at the Very Beginning
17990 @cindex breakpointing Ada elaboration code
17991 It is sometimes necessary to debug the program during elaboration, and
17992 before reaching the main procedure.
17993 As defined in the Ada Reference
17994 Manual, the elaboration code is invoked from a procedure called
17995 @code{adainit}. To run your program up to the beginning of
17996 elaboration, simply use the following two commands:
17997 @code{tbreak adainit} and @code{run}.
17999 @node Ada Exceptions
18000 @subsubsection Ada Exceptions
18002 A command is provided to list all Ada exceptions:
18005 @kindex info exceptions
18006 @item info exceptions
18007 @itemx info exceptions @var{regexp}
18008 The @code{info exceptions} command allows you to list all Ada exceptions
18009 defined within the program being debugged, as well as their addresses.
18010 With a regular expression, @var{regexp}, as argument, only those exceptions
18011 whose names match @var{regexp} are listed.
18014 Below is a small example, showing how the command can be used, first
18015 without argument, and next with a regular expression passed as an
18019 (@value{GDBP}) info exceptions
18020 All defined Ada exceptions:
18021 constraint_error: 0x613da0
18022 program_error: 0x613d20
18023 storage_error: 0x613ce0
18024 tasking_error: 0x613ca0
18025 const.aint_global_e: 0x613b00
18026 (@value{GDBP}) info exceptions const.aint
18027 All Ada exceptions matching regular expression "const.aint":
18028 constraint_error: 0x613da0
18029 const.aint_global_e: 0x613b00
18032 It is also possible to ask @value{GDBN} to stop your program's execution
18033 when an exception is raised. For more details, see @ref{Set Catchpoints}.
18036 @subsubsection Extensions for Ada Tasks
18037 @cindex Ada, tasking
18039 Support for Ada tasks is analogous to that for threads (@pxref{Threads}).
18040 @value{GDBN} provides the following task-related commands:
18045 This command shows a list of current Ada tasks, as in the following example:
18052 (@value{GDBP}) info tasks
18053 ID TID P-ID Pri State Name
18054 1 8088000 0 15 Child Activation Wait main_task
18055 2 80a4000 1 15 Accept Statement b
18056 3 809a800 1 15 Child Activation Wait a
18057 * 4 80ae800 3 15 Runnable c
18062 In this listing, the asterisk before the last task indicates it to be the
18063 task currently being inspected.
18067 Represents @value{GDBN}'s internal task number.
18073 The parent's task ID (@value{GDBN}'s internal task number).
18076 The base priority of the task.
18079 Current state of the task.
18083 The task has been created but has not been activated. It cannot be
18087 The task is not blocked for any reason known to Ada. (It may be waiting
18088 for a mutex, though.) It is conceptually "executing" in normal mode.
18091 The task is terminated, in the sense of ARM 9.3 (5). Any dependents
18092 that were waiting on terminate alternatives have been awakened and have
18093 terminated themselves.
18095 @item Child Activation Wait
18096 The task is waiting for created tasks to complete activation.
18098 @item Accept Statement
18099 The task is waiting on an accept or selective wait statement.
18101 @item Waiting on entry call
18102 The task is waiting on an entry call.
18104 @item Async Select Wait
18105 The task is waiting to start the abortable part of an asynchronous
18109 The task is waiting on a select statement with only a delay
18112 @item Child Termination Wait
18113 The task is sleeping having completed a master within itself, and is
18114 waiting for the tasks dependent on that master to become terminated or
18115 waiting on a terminate Phase.
18117 @item Wait Child in Term Alt
18118 The task is sleeping waiting for tasks on terminate alternatives to
18119 finish terminating.
18121 @item Accepting RV with @var{taskno}
18122 The task is accepting a rendez-vous with the task @var{taskno}.
18126 Name of the task in the program.
18130 @kindex info task @var{taskno}
18131 @item info task @var{taskno}
18132 This command shows detailed informations on the specified task, as in
18133 the following example:
18138 (@value{GDBP}) info tasks
18139 ID TID P-ID Pri State Name
18140 1 8077880 0 15 Child Activation Wait main_task
18141 * 2 807c468 1 15 Runnable task_1
18142 (@value{GDBP}) info task 2
18143 Ada Task: 0x807c468
18147 Parent: 1 ("main_task")
18153 @kindex task@r{ (Ada)}
18154 @cindex current Ada task ID
18155 This command prints the ID and name of the current task.
18161 (@value{GDBP}) info tasks
18162 ID TID P-ID Pri State Name
18163 1 8077870 0 15 Child Activation Wait main_task
18164 * 2 807c458 1 15 Runnable some_task
18165 (@value{GDBP}) task
18166 [Current task is 2 "some_task"]
18169 @item task @var{taskno}
18170 @cindex Ada task switching
18171 This command is like the @code{thread @var{thread-id}}
18172 command (@pxref{Threads}). It switches the context of debugging
18173 from the current task to the given task.
18179 (@value{GDBP}) info tasks
18180 ID TID P-ID Pri State Name
18181 1 8077870 0 15 Child Activation Wait main_task
18182 * 2 807c458 1 15 Runnable some_task
18183 (@value{GDBP}) task 1
18184 [Switching to task 1 "main_task"]
18185 #0 0x8067726 in pthread_cond_wait ()
18187 #0 0x8067726 in pthread_cond_wait ()
18188 #1 0x8056714 in system.os_interface.pthread_cond_wait ()
18189 #2 0x805cb63 in system.task_primitives.operations.sleep ()
18190 #3 0x806153e in system.tasking.stages.activate_tasks ()
18191 #4 0x804aacc in un () at un.adb:5
18194 @item break @var{location} task @var{taskno}
18195 @itemx break @var{location} task @var{taskno} if @dots{}
18196 @cindex breakpoints and tasks, in Ada
18197 @cindex task breakpoints, in Ada
18198 @kindex break @dots{} task @var{taskno}@r{ (Ada)}
18199 These commands are like the @code{break @dots{} thread @dots{}}
18200 command (@pxref{Thread Stops}). The
18201 @var{location} argument specifies source lines, as described
18202 in @ref{Specify Location}.
18204 Use the qualifier @samp{task @var{taskno}} with a breakpoint command
18205 to specify that you only want @value{GDBN} to stop the program when a
18206 particular Ada task reaches this breakpoint. The @var{taskno} is one of the
18207 numeric task identifiers assigned by @value{GDBN}, shown in the first
18208 column of the @samp{info tasks} display.
18210 If you do not specify @samp{task @var{taskno}} when you set a
18211 breakpoint, the breakpoint applies to @emph{all} tasks of your
18214 You can use the @code{task} qualifier on conditional breakpoints as
18215 well; in this case, place @samp{task @var{taskno}} before the
18216 breakpoint condition (before the @code{if}).
18224 (@value{GDBP}) info tasks
18225 ID TID P-ID Pri State Name
18226 1 140022020 0 15 Child Activation Wait main_task
18227 2 140045060 1 15 Accept/Select Wait t2
18228 3 140044840 1 15 Runnable t1
18229 * 4 140056040 1 15 Runnable t3
18230 (@value{GDBP}) b 15 task 2
18231 Breakpoint 5 at 0x120044cb0: file test_task_debug.adb, line 15.
18232 (@value{GDBP}) cont
18237 Breakpoint 5, test_task_debug () at test_task_debug.adb:15
18239 (@value{GDBP}) info tasks
18240 ID TID P-ID Pri State Name
18241 1 140022020 0 15 Child Activation Wait main_task
18242 * 2 140045060 1 15 Runnable t2
18243 3 140044840 1 15 Runnable t1
18244 4 140056040 1 15 Delay Sleep t3
18248 @node Ada Tasks and Core Files
18249 @subsubsection Tasking Support when Debugging Core Files
18250 @cindex Ada tasking and core file debugging
18252 When inspecting a core file, as opposed to debugging a live program,
18253 tasking support may be limited or even unavailable, depending on
18254 the platform being used.
18255 For instance, on x86-linux, the list of tasks is available, but task
18256 switching is not supported.
18258 On certain platforms, the debugger needs to perform some
18259 memory writes in order to provide Ada tasking support. When inspecting
18260 a core file, this means that the core file must be opened with read-write
18261 privileges, using the command @samp{"set write on"} (@pxref{Patching}).
18262 Under these circumstances, you should make a backup copy of the core
18263 file before inspecting it with @value{GDBN}.
18265 @node Ravenscar Profile
18266 @subsubsection Tasking Support when using the Ravenscar Profile
18267 @cindex Ravenscar Profile
18269 The @dfn{Ravenscar Profile} is a subset of the Ada tasking features,
18270 specifically designed for systems with safety-critical real-time
18274 @kindex set ravenscar task-switching on
18275 @cindex task switching with program using Ravenscar Profile
18276 @item set ravenscar task-switching on
18277 Allows task switching when debugging a program that uses the Ravenscar
18278 Profile. This is the default.
18280 @kindex set ravenscar task-switching off
18281 @item set ravenscar task-switching off
18282 Turn off task switching when debugging a program that uses the Ravenscar
18283 Profile. This is mostly intended to disable the code that adds support
18284 for the Ravenscar Profile, in case a bug in either @value{GDBN} or in
18285 the Ravenscar runtime is preventing @value{GDBN} from working properly.
18286 To be effective, this command should be run before the program is started.
18288 @kindex show ravenscar task-switching
18289 @item show ravenscar task-switching
18290 Show whether it is possible to switch from task to task in a program
18291 using the Ravenscar Profile.
18296 @subsubsection Ada Settings
18297 @cindex Ada settings
18300 @kindex set varsize-limit
18301 @item set varsize-limit @var{size}
18302 Prevent @value{GDBN} from attempting to evaluate objects whose size
18303 is above the given limit (@var{size}) when those sizes are computed
18304 from run-time quantities. This is typically the case when the object
18305 has a variable size, such as an array whose bounds are not known at
18306 compile time for example. Setting @var{size} to @code{unlimited}
18307 removes the size limitation. By default, the limit is about 65KB.
18309 The purpose of having such a limit is to prevent @value{GDBN} from
18310 trying to grab enormous chunks of virtual memory when asked to evaluate
18311 a quantity whose bounds have been corrupted or have not yet been fully
18312 initialized. The limit applies to the results of some subexpressions
18313 as well as to complete expressions. For example, an expression denoting
18314 a simple integer component, such as @code{x.y.z}, may fail if the size of
18315 @code{x.y} is variable and exceeds @code{size}. On the other hand,
18316 @value{GDBN} is sometimes clever; the expression @code{A(i)}, where
18317 @code{A} is an array variable with non-constant size, will generally
18318 succeed regardless of the bounds on @code{A}, as long as the component
18319 size is less than @var{size}.
18321 @kindex show varsize-limit
18322 @item show varsize-limit
18323 Show the limit on types whose size is determined by run-time quantities.
18327 @subsubsection Known Peculiarities of Ada Mode
18328 @cindex Ada, problems
18330 Besides the omissions listed previously (@pxref{Omissions from Ada}),
18331 we know of several problems with and limitations of Ada mode in
18333 some of which will be fixed with planned future releases of the debugger
18334 and the GNU Ada compiler.
18338 Static constants that the compiler chooses not to materialize as objects in
18339 storage are invisible to the debugger.
18342 Named parameter associations in function argument lists are ignored (the
18343 argument lists are treated as positional).
18346 Many useful library packages are currently invisible to the debugger.
18349 Fixed-point arithmetic, conversions, input, and output is carried out using
18350 floating-point arithmetic, and may give results that only approximate those on
18354 The GNAT compiler never generates the prefix @code{Standard} for any of
18355 the standard symbols defined by the Ada language. @value{GDBN} knows about
18356 this: it will strip the prefix from names when you use it, and will never
18357 look for a name you have so qualified among local symbols, nor match against
18358 symbols in other packages or subprograms. If you have
18359 defined entities anywhere in your program other than parameters and
18360 local variables whose simple names match names in @code{Standard},
18361 GNAT's lack of qualification here can cause confusion. When this happens,
18362 you can usually resolve the confusion
18363 by qualifying the problematic names with package
18364 @code{Standard} explicitly.
18367 Older versions of the compiler sometimes generate erroneous debugging
18368 information, resulting in the debugger incorrectly printing the value
18369 of affected entities. In some cases, the debugger is able to work
18370 around an issue automatically. In other cases, the debugger is able
18371 to work around the issue, but the work-around has to be specifically
18374 @kindex set ada trust-PAD-over-XVS
18375 @kindex show ada trust-PAD-over-XVS
18378 @item set ada trust-PAD-over-XVS on
18379 Configure GDB to strictly follow the GNAT encoding when computing the
18380 value of Ada entities, particularly when @code{PAD} and @code{PAD___XVS}
18381 types are involved (see @code{ada/exp_dbug.ads} in the GCC sources for
18382 a complete description of the encoding used by the GNAT compiler).
18383 This is the default.
18385 @item set ada trust-PAD-over-XVS off
18386 This is related to the encoding using by the GNAT compiler. If @value{GDBN}
18387 sometimes prints the wrong value for certain entities, changing @code{ada
18388 trust-PAD-over-XVS} to @code{off} activates a work-around which may fix
18389 the issue. It is always safe to set @code{ada trust-PAD-over-XVS} to
18390 @code{off}, but this incurs a slight performance penalty, so it is
18391 recommended to leave this setting to @code{on} unless necessary.
18395 @cindex GNAT descriptive types
18396 @cindex GNAT encoding
18397 Internally, the debugger also relies on the compiler following a number
18398 of conventions known as the @samp{GNAT Encoding}, all documented in
18399 @file{gcc/ada/exp_dbug.ads} in the GCC sources. This encoding describes
18400 how the debugging information should be generated for certain types.
18401 In particular, this convention makes use of @dfn{descriptive types},
18402 which are artificial types generated purely to help the debugger.
18404 These encodings were defined at a time when the debugging information
18405 format used was not powerful enough to describe some of the more complex
18406 types available in Ada. Since DWARF allows us to express nearly all
18407 Ada features, the long-term goal is to slowly replace these descriptive
18408 types by their pure DWARF equivalent. To facilitate that transition,
18409 a new maintenance option is available to force the debugger to ignore
18410 those descriptive types. It allows the user to quickly evaluate how
18411 well @value{GDBN} works without them.
18415 @kindex maint ada set ignore-descriptive-types
18416 @item maintenance ada set ignore-descriptive-types [on|off]
18417 Control whether the debugger should ignore descriptive types.
18418 The default is not to ignore descriptives types (@code{off}).
18420 @kindex maint ada show ignore-descriptive-types
18421 @item maintenance ada show ignore-descriptive-types
18422 Show if descriptive types are ignored by @value{GDBN}.
18430 @value{GDBN} supports the
18431 @url{https://github.com/ROCm-Developer-Tools/HIP/blob/master/docs/markdown/hip_kernel_language.md,
18432 HIP Programming Language}.
18434 @c TODO: Add any language specific differences.
18436 @node Unsupported Languages
18437 @section Unsupported Languages
18439 @cindex unsupported languages
18440 @cindex minimal language
18441 In addition to the other fully-supported programming languages,
18442 @value{GDBN} also provides a pseudo-language, called @code{minimal}.
18443 It does not represent a real programming language, but provides a set
18444 of capabilities close to what the C or assembly languages provide.
18445 This should allow most simple operations to be performed while debugging
18446 an application that uses a language currently not supported by @value{GDBN}.
18448 If the language is set to @code{auto}, @value{GDBN} will automatically
18449 select this language if the current frame corresponds to an unsupported
18453 @chapter Examining the Symbol Table
18455 The commands described in this chapter allow you to inquire about the
18456 symbols (names of variables, functions and types) defined in your
18457 program. This information is inherent in the text of your program and
18458 does not change as your program executes. @value{GDBN} finds it in your
18459 program's symbol table, in the file indicated when you started @value{GDBN}
18460 (@pxref{File Options, ,Choosing Files}), or by one of the
18461 file-management commands (@pxref{Files, ,Commands to Specify Files}).
18463 @cindex symbol names
18464 @cindex names of symbols
18465 @cindex quoting names
18466 @anchor{quoting names}
18467 Occasionally, you may need to refer to symbols that contain unusual
18468 characters, which @value{GDBN} ordinarily treats as word delimiters. The
18469 most frequent case is in referring to static variables in other
18470 source files (@pxref{Variables,,Program Variables}). File names
18471 are recorded in object files as debugging symbols, but @value{GDBN} would
18472 ordinarily parse a typical file name, like @file{foo.c}, as the three words
18473 @samp{foo} @samp{.} @samp{c}. To allow @value{GDBN} to recognize
18474 @samp{foo.c} as a single symbol, enclose it in single quotes; for example,
18481 looks up the value of @code{x} in the scope of the file @file{foo.c}.
18484 @cindex case-insensitive symbol names
18485 @cindex case sensitivity in symbol names
18486 @kindex set case-sensitive
18487 @item set case-sensitive on
18488 @itemx set case-sensitive off
18489 @itemx set case-sensitive auto
18490 Normally, when @value{GDBN} looks up symbols, it matches their names
18491 with case sensitivity determined by the current source language.
18492 Occasionally, you may wish to control that. The command @code{set
18493 case-sensitive} lets you do that by specifying @code{on} for
18494 case-sensitive matches or @code{off} for case-insensitive ones. If
18495 you specify @code{auto}, case sensitivity is reset to the default
18496 suitable for the source language. The default is case-sensitive
18497 matches for all languages except for Fortran, for which the default is
18498 case-insensitive matches.
18500 @kindex show case-sensitive
18501 @item show case-sensitive
18502 This command shows the current setting of case sensitivity for symbols
18505 @kindex set print type methods
18506 @item set print type methods
18507 @itemx set print type methods on
18508 @itemx set print type methods off
18509 Normally, when @value{GDBN} prints a class, it displays any methods
18510 declared in that class. You can control this behavior either by
18511 passing the appropriate flag to @code{ptype}, or using @command{set
18512 print type methods}. Specifying @code{on} will cause @value{GDBN} to
18513 display the methods; this is the default. Specifying @code{off} will
18514 cause @value{GDBN} to omit the methods.
18516 @kindex show print type methods
18517 @item show print type methods
18518 This command shows the current setting of method display when printing
18521 @kindex set print type nested-type-limit
18522 @item set print type nested-type-limit @var{limit}
18523 @itemx set print type nested-type-limit unlimited
18524 Set the limit of displayed nested types that the type printer will
18525 show. A @var{limit} of @code{unlimited} or @code{-1} will show all
18526 nested definitions. By default, the type printer will not show any nested
18527 types defined in classes.
18529 @kindex show print type nested-type-limit
18530 @item show print type nested-type-limit
18531 This command shows the current display limit of nested types when
18534 @kindex set print type typedefs
18535 @item set print type typedefs
18536 @itemx set print type typedefs on
18537 @itemx set print type typedefs off
18539 Normally, when @value{GDBN} prints a class, it displays any typedefs
18540 defined in that class. You can control this behavior either by
18541 passing the appropriate flag to @code{ptype}, or using @command{set
18542 print type typedefs}. Specifying @code{on} will cause @value{GDBN} to
18543 display the typedef definitions; this is the default. Specifying
18544 @code{off} will cause @value{GDBN} to omit the typedef definitions.
18545 Note that this controls whether the typedef definition itself is
18546 printed, not whether typedef names are substituted when printing other
18549 @kindex show print type typedefs
18550 @item show print type typedefs
18551 This command shows the current setting of typedef display when
18554 @kindex info address
18555 @cindex address of a symbol
18556 @item info address @var{symbol}
18557 Describe where the data for @var{symbol} is stored. For a register
18558 variable, this says which register it is kept in. For a non-register
18559 local variable, this prints the stack-frame offset at which the variable
18562 Note the contrast with @samp{print &@var{symbol}}, which does not work
18563 at all for a register variable, and for a stack local variable prints
18564 the exact address of the current instantiation of the variable.
18566 @kindex info symbol
18567 @cindex symbol from address
18568 @cindex closest symbol and offset for an address
18569 @item info symbol @var{addr}
18570 Print the name of a symbol which is stored at the address @var{addr}.
18571 If no symbol is stored exactly at @var{addr}, @value{GDBN} prints the
18572 nearest symbol and an offset from it:
18575 (@value{GDBP}) info symbol 0x54320
18576 _initialize_vx + 396 in section .text
18580 This is the opposite of the @code{info address} command. You can use
18581 it to find out the name of a variable or a function given its address.
18583 For dynamically linked executables, the name of executable or shared
18584 library containing the symbol is also printed:
18587 (@value{GDBP}) info symbol 0x400225
18588 _start + 5 in section .text of /tmp/a.out
18589 (@value{GDBP}) info symbol 0x2aaaac2811cf
18590 __read_nocancel + 6 in section .text of /usr/lib64/libc.so.6
18595 @item demangle @r{[}-l @var{language}@r{]} @r{[}@var{--}@r{]} @var{name}
18596 Demangle @var{name}.
18597 If @var{language} is provided it is the name of the language to demangle
18598 @var{name} in. Otherwise @var{name} is demangled in the current language.
18600 The @samp{--} option specifies the end of options,
18601 and is useful when @var{name} begins with a dash.
18603 The parameter @code{demangle-style} specifies how to interpret the kind
18604 of mangling used. @xref{Print Settings}.
18607 @item whatis[/@var{flags}] [@var{arg}]
18608 Print the data type of @var{arg}, which can be either an expression
18609 or a name of a data type. With no argument, print the data type of
18610 @code{$}, the last value in the value history.
18612 If @var{arg} is an expression (@pxref{Expressions, ,Expressions}), it
18613 is not actually evaluated, and any side-effecting operations (such as
18614 assignments or function calls) inside it do not take place.
18616 If @var{arg} is a variable or an expression, @code{whatis} prints its
18617 literal type as it is used in the source code. If the type was
18618 defined using a @code{typedef}, @code{whatis} will @emph{not} print
18619 the data type underlying the @code{typedef}. If the type of the
18620 variable or the expression is a compound data type, such as
18621 @code{struct} or @code{class}, @code{whatis} never prints their
18622 fields or methods. It just prints the @code{struct}/@code{class}
18623 name (a.k.a.@: its @dfn{tag}). If you want to see the members of
18624 such a compound data type, use @code{ptype}.
18626 If @var{arg} is a type name that was defined using @code{typedef},
18627 @code{whatis} @dfn{unrolls} only one level of that @code{typedef}.
18628 Unrolling means that @code{whatis} will show the underlying type used
18629 in the @code{typedef} declaration of @var{arg}. However, if that
18630 underlying type is also a @code{typedef}, @code{whatis} will not
18633 For C code, the type names may also have the form @samp{class
18634 @var{class-name}}, @samp{struct @var{struct-tag}}, @samp{union
18635 @var{union-tag}} or @samp{enum @var{enum-tag}}.
18637 @var{flags} can be used to modify how the type is displayed.
18638 Available flags are:
18642 Display in ``raw'' form. Normally, @value{GDBN} substitutes template
18643 parameters and typedefs defined in a class when printing the class'
18644 members. The @code{/r} flag disables this.
18647 Do not print methods defined in the class.
18650 Print methods defined in the class. This is the default, but the flag
18651 exists in case you change the default with @command{set print type methods}.
18654 Do not print typedefs defined in the class. Note that this controls
18655 whether the typedef definition itself is printed, not whether typedef
18656 names are substituted when printing other types.
18659 Print typedefs defined in the class. This is the default, but the flag
18660 exists in case you change the default with @command{set print type typedefs}.
18663 Print the offsets and sizes of fields in a struct, similar to what the
18664 @command{pahole} tool does. This option implies the @code{/tm} flags.
18666 For example, given the following declarations:
18702 Issuing a @kbd{ptype /o struct tuv} command would print:
18705 (@value{GDBP}) ptype /o struct tuv
18706 /* offset | size */ type = struct tuv @{
18707 /* 0 | 4 */ int a1;
18708 /* XXX 4-byte hole */
18709 /* 8 | 8 */ char *a2;
18710 /* 16 | 4 */ int a3;
18712 /* total size (bytes): 24 */
18716 Notice the format of the first column of comments. There, you can
18717 find two parts separated by the @samp{|} character: the @emph{offset},
18718 which indicates where the field is located inside the struct, in
18719 bytes, and the @emph{size} of the field. Another interesting line is
18720 the marker of a @emph{hole} in the struct, indicating that it may be
18721 possible to pack the struct and make it use less space by reorganizing
18724 It is also possible to print offsets inside an union:
18727 (@value{GDBP}) ptype /o union qwe
18728 /* offset | size */ type = union qwe @{
18729 /* 24 */ struct tuv @{
18730 /* 0 | 4 */ int a1;
18731 /* XXX 4-byte hole */
18732 /* 8 | 8 */ char *a2;
18733 /* 16 | 4 */ int a3;
18735 /* total size (bytes): 24 */
18737 /* 40 */ struct xyz @{
18738 /* 0 | 4 */ int f1;
18739 /* 4 | 1 */ char f2;
18740 /* XXX 3-byte hole */
18741 /* 8 | 8 */ void *f3;
18742 /* 16 | 24 */ struct tuv @{
18743 /* 16 | 4 */ int a1;
18744 /* XXX 4-byte hole */
18745 /* 24 | 8 */ char *a2;
18746 /* 32 | 4 */ int a3;
18748 /* total size (bytes): 24 */
18751 /* total size (bytes): 40 */
18754 /* total size (bytes): 40 */
18758 In this case, since @code{struct tuv} and @code{struct xyz} occupy the
18759 same space (because we are dealing with an union), the offset is not
18760 printed for them. However, you can still examine the offset of each
18761 of these structures' fields.
18763 Another useful scenario is printing the offsets of a struct containing
18767 (@value{GDBP}) ptype /o struct tyu
18768 /* offset | size */ type = struct tyu @{
18769 /* 0:31 | 4 */ int a1 : 1;
18770 /* 0:28 | 4 */ int a2 : 3;
18771 /* 0: 5 | 4 */ int a3 : 23;
18772 /* 3: 3 | 1 */ signed char a4 : 2;
18773 /* XXX 3-bit hole */
18774 /* XXX 4-byte hole */
18775 /* 8 | 8 */ int64_t a5;
18776 /* 16: 0 | 4 */ int a6 : 5;
18777 /* 16: 5 | 8 */ int64_t a7 : 3;
18778 "/* XXX 7-byte padding */
18780 /* total size (bytes): 24 */
18784 Note how the offset information is now extended to also include the
18785 first bit of the bitfield.
18789 @item ptype[/@var{flags}] [@var{arg}]
18790 @code{ptype} accepts the same arguments as @code{whatis}, but prints a
18791 detailed description of the type, instead of just the name of the type.
18792 @xref{Expressions, ,Expressions}.
18794 Contrary to @code{whatis}, @code{ptype} always unrolls any
18795 @code{typedef}s in its argument declaration, whether the argument is
18796 a variable, expression, or a data type. This means that @code{ptype}
18797 of a variable or an expression will not print literally its type as
18798 present in the source code---use @code{whatis} for that. @code{typedef}s at
18799 the pointer or reference targets are also unrolled. Only @code{typedef}s of
18800 fields, methods and inner @code{class typedef}s of @code{struct}s,
18801 @code{class}es and @code{union}s are not unrolled even with @code{ptype}.
18803 For example, for this variable declaration:
18806 typedef double real_t;
18807 struct complex @{ real_t real; double imag; @};
18808 typedef struct complex complex_t;
18810 real_t *real_pointer_var;
18814 the two commands give this output:
18818 (@value{GDBP}) whatis var
18820 (@value{GDBP}) ptype var
18821 type = struct complex @{
18825 (@value{GDBP}) whatis complex_t
18826 type = struct complex
18827 (@value{GDBP}) whatis struct complex
18828 type = struct complex
18829 (@value{GDBP}) ptype struct complex
18830 type = struct complex @{
18834 (@value{GDBP}) whatis real_pointer_var
18836 (@value{GDBP}) ptype real_pointer_var
18842 As with @code{whatis}, using @code{ptype} without an argument refers to
18843 the type of @code{$}, the last value in the value history.
18845 @cindex incomplete type
18846 Sometimes, programs use opaque data types or incomplete specifications
18847 of complex data structure. If the debug information included in the
18848 program does not allow @value{GDBN} to display a full declaration of
18849 the data type, it will say @samp{<incomplete type>}. For example,
18850 given these declarations:
18854 struct foo *fooptr;
18858 but no definition for @code{struct foo} itself, @value{GDBN} will say:
18861 (@value{GDBP}) ptype foo
18862 $1 = <incomplete type>
18866 ``Incomplete type'' is C terminology for data types that are not
18867 completely specified.
18869 @cindex unknown type
18870 Othertimes, information about a variable's type is completely absent
18871 from the debug information included in the program. This most often
18872 happens when the program or library where the variable is defined
18873 includes no debug information at all. @value{GDBN} knows the variable
18874 exists from inspecting the linker/loader symbol table (e.g., the ELF
18875 dynamic symbol table), but such symbols do not contain type
18876 information. Inspecting the type of a (global) variable for which
18877 @value{GDBN} has no type information shows:
18880 (@value{GDBP}) ptype var
18881 type = <data variable, no debug info>
18884 @xref{Variables, no debug info variables}, for how to print the values
18888 @item info types [-q] [@var{regexp}]
18889 Print a brief description of all types whose names match the regular
18890 expression @var{regexp} (or all types in your program, if you supply
18891 no argument). Each complete typename is matched as though it were a
18892 complete line; thus, @samp{i type value} gives information on all
18893 types in your program whose names include the string @code{value}, but
18894 @samp{i type ^value$} gives information only on types whose complete
18895 name is @code{value}.
18897 In programs using different languages, @value{GDBN} chooses the syntax
18898 to print the type description according to the
18899 @samp{set language} value: using @samp{set language auto}
18900 (see @ref{Automatically, ,Set Language Automatically}) means to use the
18901 language of the type, other values mean to use
18902 the manually specified language (see @ref{Manually, ,Set Language Manually}).
18904 This command differs from @code{ptype} in two ways: first, like
18905 @code{whatis}, it does not print a detailed description; second, it
18906 lists all source files and line numbers where a type is defined.
18908 The output from @samp{into types} is proceeded with a header line
18909 describing what types are being listed. The optional flag @samp{-q},
18910 which stands for @samp{quiet}, disables printing this header
18913 @kindex info type-printers
18914 @item info type-printers
18915 Versions of @value{GDBN} that ship with Python scripting enabled may
18916 have ``type printers'' available. When using @command{ptype} or
18917 @command{whatis}, these printers are consulted when the name of a type
18918 is needed. @xref{Type Printing API}, for more information on writing
18921 @code{info type-printers} displays all the available type printers.
18923 @kindex enable type-printer
18924 @kindex disable type-printer
18925 @item enable type-printer @var{name}@dots{}
18926 @item disable type-printer @var{name}@dots{}
18927 These commands can be used to enable or disable type printers.
18930 @cindex local variables
18931 @item info scope @var{location}
18932 List all the variables local to a particular scope. This command
18933 accepts a @var{location} argument---a function name, a source line, or
18934 an address preceded by a @samp{*}, and prints all the variables local
18935 to the scope defined by that location. (@xref{Specify Location}, for
18936 details about supported forms of @var{location}.) For example:
18939 (@value{GDBP}) @b{info scope command_line_handler}
18940 Scope for command_line_handler:
18941 Symbol rl is an argument at stack/frame offset 8, length 4.
18942 Symbol linebuffer is in static storage at address 0x150a18, length 4.
18943 Symbol linelength is in static storage at address 0x150a1c, length 4.
18944 Symbol p is a local variable in register $esi, length 4.
18945 Symbol p1 is a local variable in register $ebx, length 4.
18946 Symbol nline is a local variable in register $edx, length 4.
18947 Symbol repeat is a local variable at frame offset -8, length 4.
18951 This command is especially useful for determining what data to collect
18952 during a @dfn{trace experiment}, see @ref{Tracepoint Actions,
18955 @kindex info source
18957 Show information about the current source file---that is, the source file for
18958 the function containing the current point of execution:
18961 the name of the source file, and the directory containing it,
18963 the directory it was compiled in,
18965 its length, in lines,
18967 which programming language it is written in,
18969 if the debug information provides it, the program that compiled the file
18970 (which may include, e.g., the compiler version and command line arguments),
18972 whether the executable includes debugging information for that file, and
18973 if so, what format the information is in (e.g., STABS, Dwarf 2, etc.), and
18975 whether the debugging information includes information about
18976 preprocessor macros.
18980 @kindex info sources
18982 Print the names of all source files in your program for which there is
18983 debugging information, organized into two lists: files whose symbols
18984 have already been read, and files whose symbols will be read when needed.
18986 @item info sources [-dirname | -basename] [--] [@var{regexp}]
18987 Like @samp{info sources}, but only print the names of the files
18988 matching the provided @var{regexp}.
18989 By default, the @var{regexp} is used to match anywhere in the filename.
18990 If @code{-dirname}, only files having a dirname matching @var{regexp} are shown.
18991 If @code{-basename}, only files having a basename matching @var{regexp}
18993 The matching is case-sensitive, except on operating systems that
18994 have case-insensitive filesystem (e.g., MS-Windows).
18996 @kindex info functions
18997 @item info functions [-q] [-n]
18998 Print the names and data types of all defined functions.
18999 Similarly to @samp{info types}, this command groups its output by source
19000 files and annotates each function definition with its source line
19003 In programs using different languages, @value{GDBN} chooses the syntax
19004 to print the function name and type according to the
19005 @samp{set language} value: using @samp{set language auto}
19006 (see @ref{Automatically, ,Set Language Automatically}) means to use the
19007 language of the function, other values mean to use
19008 the manually specified language (see @ref{Manually, ,Set Language Manually}).
19010 The @samp{-n} flag excludes @dfn{non-debugging symbols} from the
19011 results. A non-debugging symbol is a symbol that comes from the
19012 executable's symbol table, not from the debug information (for
19013 example, DWARF) associated with the executable.
19015 The optional flag @samp{-q}, which stands for @samp{quiet}, disables
19016 printing header information and messages explaining why no functions
19019 @item info functions [-q] [-n] [-t @var{type_regexp}] [@var{regexp}]
19020 Like @samp{info functions}, but only print the names and data types
19021 of the functions selected with the provided regexp(s).
19023 If @var{regexp} is provided, print only the functions whose names
19024 match the regular expression @var{regexp}.
19025 Thus, @samp{info fun step} finds all functions whose
19026 names include @code{step}; @samp{info fun ^step} finds those whose names
19027 start with @code{step}. If a function name contains characters that
19028 conflict with the regular expression language (e.g.@:
19029 @samp{operator*()}), they may be quoted with a backslash.
19031 If @var{type_regexp} is provided, print only the functions whose
19032 types, as printed by the @code{whatis} command, match
19033 the regular expression @var{type_regexp}.
19034 If @var{type_regexp} contains space(s), it should be enclosed in
19035 quote characters. If needed, use backslash to escape the meaning
19036 of special characters or quotes.
19037 Thus, @samp{info fun -t '^int ('} finds the functions that return
19038 an integer; @samp{info fun -t '(.*int.*'} finds the functions that
19039 have an argument type containing int; @samp{info fun -t '^int (' ^step}
19040 finds the functions whose names start with @code{step} and that return
19043 If both @var{regexp} and @var{type_regexp} are provided, a function
19044 is printed only if its name matches @var{regexp} and its type matches
19048 @kindex info variables
19049 @item info variables [-q] [-n]
19050 Print the names and data types of all variables that are defined
19051 outside of functions (i.e.@: excluding local variables).
19052 The printed variables are grouped by source files and annotated with
19053 their respective source line numbers.
19055 In programs using different languages, @value{GDBN} chooses the syntax
19056 to print the variable name and type according to the
19057 @samp{set language} value: using @samp{set language auto}
19058 (see @ref{Automatically, ,Set Language Automatically}) means to use the
19059 language of the variable, other values mean to use
19060 the manually specified language (see @ref{Manually, ,Set Language Manually}).
19062 The @samp{-n} flag excludes non-debugging symbols from the results.
19064 The optional flag @samp{-q}, which stands for @samp{quiet}, disables
19065 printing header information and messages explaining why no variables
19068 @item info variables [-q] [-n] [-t @var{type_regexp}] [@var{regexp}]
19069 Like @kbd{info variables}, but only print the variables selected
19070 with the provided regexp(s).
19072 If @var{regexp} is provided, print only the variables whose names
19073 match the regular expression @var{regexp}.
19075 If @var{type_regexp} is provided, print only the variables whose
19076 types, as printed by the @code{whatis} command, match
19077 the regular expression @var{type_regexp}.
19078 If @var{type_regexp} contains space(s), it should be enclosed in
19079 quote characters. If needed, use backslash to escape the meaning
19080 of special characters or quotes.
19082 If both @var{regexp} and @var{type_regexp} are provided, an argument
19083 is printed only if its name matches @var{regexp} and its type matches
19086 @kindex info modules
19088 @item info modules @r{[}-q@r{]} @r{[}@var{regexp}@r{]}
19089 List all Fortran modules in the program, or all modules matching the
19090 optional regular expression @var{regexp}.
19092 The optional flag @samp{-q}, which stands for @samp{quiet}, disables
19093 printing header information and messages explaining why no modules
19096 @kindex info module
19097 @cindex Fortran modules, information about
19098 @cindex functions and variables by Fortran module
19099 @cindex module functions and variables
19100 @item info module functions @r{[}-q@r{]} @r{[}-m @var{module-regexp}@r{]} @r{[}-t @var{type-regexp}@r{]} @r{[}@var{regexp}@r{]}
19101 @itemx info module variables @r{[}-q@r{]} @r{[}-m @var{module-regexp}@r{]} @r{[}-t @var{type-regexp}@r{]} @r{[}@var{regexp}@r{]}
19102 List all functions or variables within all Fortran modules. The set
19103 of functions or variables listed can be limited by providing some or
19104 all of the optional regular expressions. If @var{module-regexp} is
19105 provided, then only Fortran modules matching @var{module-regexp} will
19106 be searched. Only functions or variables whose type matches the
19107 optional regular expression @var{type-regexp} will be listed. And
19108 only functions or variables whose name matches the optional regular
19109 expression @var{regexp} will be listed.
19111 The optional flag @samp{-q}, which stands for @samp{quiet}, disables
19112 printing header information and messages explaining why no functions
19113 or variables have been printed.
19115 @kindex info classes
19116 @cindex Objective-C, classes and selectors
19118 @itemx info classes @var{regexp}
19119 Display all Objective-C classes in your program, or
19120 (with the @var{regexp} argument) all those matching a particular regular
19123 @kindex info selectors
19124 @item info selectors
19125 @itemx info selectors @var{regexp}
19126 Display all Objective-C selectors in your program, or
19127 (with the @var{regexp} argument) all those matching a particular regular
19131 This was never implemented.
19132 @kindex info methods
19134 @itemx info methods @var{regexp}
19135 The @code{info methods} command permits the user to examine all defined
19136 methods within C@t{++} program, or (with the @var{regexp} argument) a
19137 specific set of methods found in the various C@t{++} classes. Many
19138 C@t{++} classes provide a large number of methods. Thus, the output
19139 from the @code{ptype} command can be overwhelming and hard to use. The
19140 @code{info-methods} command filters the methods, printing only those
19141 which match the regular-expression @var{regexp}.
19144 @cindex opaque data types
19145 @kindex set opaque-type-resolution
19146 @item set opaque-type-resolution on
19147 Tell @value{GDBN} to resolve opaque types. An opaque type is a type
19148 declared as a pointer to a @code{struct}, @code{class}, or
19149 @code{union}---for example, @code{struct MyType *}---that is used in one
19150 source file although the full declaration of @code{struct MyType} is in
19151 another source file. The default is on.
19153 A change in the setting of this subcommand will not take effect until
19154 the next time symbols for a file are loaded.
19156 @item set opaque-type-resolution off
19157 Tell @value{GDBN} not to resolve opaque types. In this case, the type
19158 is printed as follows:
19160 @{<no data fields>@}
19163 @kindex show opaque-type-resolution
19164 @item show opaque-type-resolution
19165 Show whether opaque types are resolved or not.
19167 @kindex set print symbol-loading
19168 @cindex print messages when symbols are loaded
19169 @item set print symbol-loading
19170 @itemx set print symbol-loading full
19171 @itemx set print symbol-loading brief
19172 @itemx set print symbol-loading off
19173 The @code{set print symbol-loading} command allows you to control the
19174 printing of messages when @value{GDBN} loads symbol information.
19175 By default a message is printed for the executable and one for each
19176 shared library, and normally this is what you want. However, when
19177 debugging apps with large numbers of shared libraries these messages
19179 When set to @code{brief} a message is printed for each executable,
19180 and when @value{GDBN} loads a collection of shared libraries at once
19181 it will only print one message regardless of the number of shared
19182 libraries. When set to @code{off} no messages are printed.
19184 @kindex show print symbol-loading
19185 @item show print symbol-loading
19186 Show whether messages will be printed when a @value{GDBN} command
19187 entered from the keyboard causes symbol information to be loaded.
19189 @kindex maint print symbols
19190 @cindex symbol dump
19191 @kindex maint print psymbols
19192 @cindex partial symbol dump
19193 @kindex maint print msymbols
19194 @cindex minimal symbol dump
19195 @item maint print symbols @r{[}-pc @var{address}@r{]} @r{[}@var{filename}@r{]}
19196 @itemx maint print symbols @r{[}-objfile @var{objfile}@r{]} @r{[}-source @var{source}@r{]} @r{[}--@r{]} @r{[}@var{filename}@r{]}
19197 @itemx maint print psymbols @r{[}-objfile @var{objfile}@r{]} @r{[}-pc @var{address}@r{]} @r{[}--@r{]} @r{[}@var{filename}@r{]}
19198 @itemx maint print psymbols @r{[}-objfile @var{objfile}@r{]} @r{[}-source @var{source}@r{]} @r{[}--@r{]} @r{[}@var{filename}@r{]}
19199 @itemx maint print msymbols @r{[}-objfile @var{objfile}@r{]} @r{[}--@r{]} @r{[}@var{filename}@r{]}
19200 Write a dump of debugging symbol data into the file @var{filename} or
19201 the terminal if @var{filename} is unspecified.
19202 If @code{-objfile @var{objfile}} is specified, only dump symbols for
19204 If @code{-pc @var{address}} is specified, only dump symbols for the file
19205 with code at that address. Note that @var{address} may be a symbol like
19207 If @code{-source @var{source}} is specified, only dump symbols for that
19210 These commands are used to debug the @value{GDBN} symbol-reading code.
19211 These commands do not modify internal @value{GDBN} state, therefore
19212 @samp{maint print symbols} will only print symbols for already expanded symbol
19214 You can use the command @code{info sources} to find out which files these are.
19215 If you use @samp{maint print psymbols} instead, the dump shows information
19216 about symbols that @value{GDBN} only knows partially---that is, symbols
19217 defined in files that @value{GDBN} has skimmed, but not yet read completely.
19218 Finally, @samp{maint print msymbols} just dumps ``minimal symbols'', e.g.,
19221 @xref{Files, ,Commands to Specify Files}, for a discussion of how
19222 @value{GDBN} reads symbols (in the description of @code{symbol-file}).
19224 @kindex maint info symtabs
19225 @kindex maint info psymtabs
19226 @cindex listing @value{GDBN}'s internal symbol tables
19227 @cindex symbol tables, listing @value{GDBN}'s internal
19228 @cindex full symbol tables, listing @value{GDBN}'s internal
19229 @cindex partial symbol tables, listing @value{GDBN}'s internal
19230 @item maint info symtabs @r{[} @var{regexp} @r{]}
19231 @itemx maint info psymtabs @r{[} @var{regexp} @r{]}
19233 List the @code{struct symtab} or @code{struct partial_symtab}
19234 structures whose names match @var{regexp}. If @var{regexp} is not
19235 given, list them all. The output includes expressions which you can
19236 copy into a @value{GDBN} debugging this one to examine a particular
19237 structure in more detail. For example:
19240 (@value{GDBP}) maint info psymtabs dwarf2read
19241 @{ objfile /home/gnu/build/gdb/gdb
19242 ((struct objfile *) 0x82e69d0)
19243 @{ psymtab /home/gnu/src/gdb/dwarf2read.c
19244 ((struct partial_symtab *) 0x8474b10)
19247 text addresses 0x814d3c8 -- 0x8158074
19248 globals (* (struct partial_symbol **) 0x8507a08 @@ 9)
19249 statics (* (struct partial_symbol **) 0x40e95b78 @@ 2882)
19250 dependencies (none)
19253 (@value{GDBP}) maint info symtabs
19257 We see that there is one partial symbol table whose filename contains
19258 the string @samp{dwarf2read}, belonging to the @samp{gdb} executable;
19259 and we see that @value{GDBN} has not read in any symtabs yet at all.
19260 If we set a breakpoint on a function, that will cause @value{GDBN} to
19261 read the symtab for the compilation unit containing that function:
19264 (@value{GDBP}) break dwarf2_psymtab_to_symtab
19265 Breakpoint 1 at 0x814e5da: file /home/gnu/src/gdb/dwarf2read.c,
19267 (@value{GDBP}) maint info symtabs
19268 @{ objfile /home/gnu/build/gdb/gdb
19269 ((struct objfile *) 0x82e69d0)
19270 @{ symtab /home/gnu/src/gdb/dwarf2read.c
19271 ((struct symtab *) 0x86c1f38)
19274 blockvector ((struct blockvector *) 0x86c1bd0) (primary)
19275 linetable ((struct linetable *) 0x8370fa0)
19276 debugformat DWARF 2
19282 @kindex maint info line-table
19283 @cindex listing @value{GDBN}'s internal line tables
19284 @cindex line tables, listing @value{GDBN}'s internal
19285 @item maint info line-table @r{[} @var{regexp} @r{]}
19287 List the @code{struct linetable} from all @code{struct symtab}
19288 instances whose name matches @var{regexp}. If @var{regexp} is not
19289 given, list the @code{struct linetable} from all @code{struct symtab}.
19291 @kindex maint set symbol-cache-size
19292 @cindex symbol cache size
19293 @item maint set symbol-cache-size @var{size}
19294 Set the size of the symbol cache to @var{size}.
19295 The default size is intended to be good enough for debugging
19296 most applications. This option exists to allow for experimenting
19297 with different sizes.
19299 @kindex maint show symbol-cache-size
19300 @item maint show symbol-cache-size
19301 Show the size of the symbol cache.
19303 @kindex maint print symbol-cache
19304 @cindex symbol cache, printing its contents
19305 @item maint print symbol-cache
19306 Print the contents of the symbol cache.
19307 This is useful when debugging symbol cache issues.
19309 @kindex maint print symbol-cache-statistics
19310 @cindex symbol cache, printing usage statistics
19311 @item maint print symbol-cache-statistics
19312 Print symbol cache usage statistics.
19313 This helps determine how well the cache is being utilized.
19315 @kindex maint flush-symbol-cache
19316 @cindex symbol cache, flushing
19317 @item maint flush-symbol-cache
19318 Flush the contents of the symbol cache, all entries are removed.
19319 This command is useful when debugging the symbol cache.
19320 It is also useful when collecting performance data.
19325 @chapter Altering Execution
19327 Once you think you have found an error in your program, you might want to
19328 find out for certain whether correcting the apparent error would lead to
19329 correct results in the rest of the run. You can find the answer by
19330 experiment, using the @value{GDBN} features for altering execution of the
19333 For example, you can store new values into variables or memory
19334 locations, give your program a signal, restart it at a different
19335 address, or even return prematurely from a function.
19338 * Assignment:: Assignment to variables
19339 * Jumping:: Continuing at a different address
19340 * Signaling:: Giving your program a signal
19341 * Returning:: Returning from a function
19342 * Calling:: Calling your program's functions
19343 * Patching:: Patching your program
19344 * Compiling and Injecting Code:: Compiling and injecting code in @value{GDBN}
19348 @section Assignment to Variables
19351 @cindex setting variables
19352 To alter the value of a variable, evaluate an assignment expression.
19353 @xref{Expressions, ,Expressions}. For example,
19360 stores the value 4 into the variable @code{x}, and then prints the
19361 value of the assignment expression (which is 4).
19362 @xref{Languages, ,Using @value{GDBN} with Different Languages}, for more
19363 information on operators in supported languages.
19365 @kindex set variable
19366 @cindex variables, setting
19367 If you are not interested in seeing the value of the assignment, use the
19368 @code{set} command instead of the @code{print} command. @code{set} is
19369 really the same as @code{print} except that the expression's value is
19370 not printed and is not put in the value history (@pxref{Value History,
19371 ,Value History}). The expression is evaluated only for its effects.
19373 If the beginning of the argument string of the @code{set} command
19374 appears identical to a @code{set} subcommand, use the @code{set
19375 variable} command instead of just @code{set}. This command is identical
19376 to @code{set} except for its lack of subcommands. For example, if your
19377 program has a variable @code{width}, you get an error if you try to set
19378 a new value with just @samp{set width=13}, because @value{GDBN} has the
19379 command @code{set width}:
19382 (@value{GDBP}) whatis width
19384 (@value{GDBP}) p width
19386 (@value{GDBP}) set width=47
19387 Invalid syntax in expression.
19391 The invalid expression, of course, is @samp{=47}. In
19392 order to actually set the program's variable @code{width}, use
19395 (@value{GDBP}) set var width=47
19398 Because the @code{set} command has many subcommands that can conflict
19399 with the names of program variables, it is a good idea to use the
19400 @code{set variable} command instead of just @code{set}. For example, if
19401 your program has a variable @code{g}, you run into problems if you try
19402 to set a new value with just @samp{set g=4}, because @value{GDBN} has
19403 the command @code{set gnutarget}, abbreviated @code{set g}:
19407 (@value{GDBP}) whatis g
19411 (@value{GDBP}) set g=4
19415 The program being debugged has been started already.
19416 Start it from the beginning? (y or n) y
19417 Starting program: /home/smith/cc_progs/a.out
19418 "/home/smith/cc_progs/a.out": can't open to read symbols:
19419 Invalid bfd target.
19420 (@value{GDBP}) show g
19421 The current BFD target is "=4".
19426 The program variable @code{g} did not change, and you silently set the
19427 @code{gnutarget} to an invalid value. In order to set the variable
19431 (@value{GDBP}) set var g=4
19434 @value{GDBN} allows more implicit conversions in assignments than C; you can
19435 freely store an integer value into a pointer variable or vice versa,
19436 and you can convert any structure to any other structure that is the
19437 same length or shorter.
19438 @comment FIXME: how do structs align/pad in these conversions?
19439 @comment /doc@cygnus.com 18dec1990
19441 To store values into arbitrary places in memory, use the @samp{@{@dots{}@}}
19442 construct to generate a value of specified type at a specified address
19443 (@pxref{Expressions, ,Expressions}). For example, @code{@{int@}0x83040} refers
19444 to memory location @code{0x83040} as an integer (which implies a certain size
19445 and representation in memory), and
19448 set @{int@}0x83040 = 4
19452 stores the value 4 into that memory location.
19455 @section Continuing at a Different Address
19457 Ordinarily, when you continue your program, you do so at the place where
19458 it stopped, with the @code{continue} command. You can instead continue at
19459 an address of your own choosing, with the following commands:
19463 @kindex j @r{(@code{jump})}
19464 @item jump @var{location}
19465 @itemx j @var{location}
19466 Resume execution at @var{location}. Execution stops again immediately
19467 if there is a breakpoint there. @xref{Specify Location}, for a description
19468 of the different forms of @var{location}. It is common
19469 practice to use the @code{tbreak} command in conjunction with
19470 @code{jump}. @xref{Set Breaks, ,Setting Breakpoints}.
19472 The @code{jump} command does not change the current stack frame, or
19473 the stack pointer, or the contents of any memory location or any
19474 register other than the program counter. If @var{location} is in
19475 a different function from the one currently executing, the results may
19476 be bizarre if the two functions expect different patterns of arguments or
19477 of local variables. For this reason, the @code{jump} command requests
19478 confirmation if the specified line is not in the function currently
19479 executing. However, even bizarre results are predictable if you are
19480 well acquainted with the machine-language code of your program.
19483 On many systems, you can get much the same effect as the @code{jump}
19484 command by storing a new value into the register @code{$pc}. The
19485 difference is that this does not start your program running; it only
19486 changes the address of where it @emph{will} run when you continue. For
19494 makes the next @code{continue} command or stepping command execute at
19495 address @code{0x485}, rather than at the address where your program stopped.
19496 @xref{Continuing and Stepping, ,Continuing and Stepping}.
19498 The most common occasion to use the @code{jump} command is to back
19499 up---perhaps with more breakpoints set---over a portion of a program
19500 that has already executed, in order to examine its execution in more
19505 @section Giving your Program a Signal
19506 @cindex deliver a signal to a program
19510 @item signal @var{signal}
19511 Resume execution where your program is stopped, but immediately give it the
19512 signal @var{signal}. The @var{signal} can be the name or the number of a
19513 signal. For example, on many systems @code{signal 2} and @code{signal
19514 SIGINT} are both ways of sending an interrupt signal.
19516 Alternatively, if @var{signal} is zero, continue execution without
19517 giving a signal. This is useful when your program stopped on account of
19518 a signal and would ordinarily see the signal when resumed with the
19519 @code{continue} command; @samp{signal 0} causes it to resume without a
19522 @emph{Note:} When resuming a multi-threaded program, @var{signal} is
19523 delivered to the currently selected thread, not the thread that last
19524 reported a stop. This includes the situation where a thread was
19525 stopped due to a signal. So if you want to continue execution
19526 suppressing the signal that stopped a thread, you should select that
19527 same thread before issuing the @samp{signal 0} command. If you issue
19528 the @samp{signal 0} command with another thread as the selected one,
19529 @value{GDBN} detects that and asks for confirmation.
19531 Invoking the @code{signal} command is not the same as invoking the
19532 @code{kill} utility from the shell. Sending a signal with @code{kill}
19533 causes @value{GDBN} to decide what to do with the signal depending on
19534 the signal handling tables (@pxref{Signals}). The @code{signal} command
19535 passes the signal directly to your program.
19537 @code{signal} does not repeat when you press @key{RET} a second time
19538 after executing the command.
19540 @kindex queue-signal
19541 @item queue-signal @var{signal}
19542 Queue @var{signal} to be delivered immediately to the current thread
19543 when execution of the thread resumes. The @var{signal} can be the name or
19544 the number of a signal. For example, on many systems @code{signal 2} and
19545 @code{signal SIGINT} are both ways of sending an interrupt signal.
19546 The handling of the signal must be set to pass the signal to the program,
19547 otherwise @value{GDBN} will report an error.
19548 You can control the handling of signals from @value{GDBN} with the
19549 @code{handle} command (@pxref{Signals}).
19551 Alternatively, if @var{signal} is zero, any currently queued signal
19552 for the current thread is discarded and when execution resumes no signal
19553 will be delivered. This is useful when your program stopped on account
19554 of a signal and would ordinarily see the signal when resumed with the
19555 @code{continue} command.
19557 This command differs from the @code{signal} command in that the signal
19558 is just queued, execution is not resumed. And @code{queue-signal} cannot
19559 be used to pass a signal whose handling state has been set to @code{nopass}
19564 @xref{stepping into signal handlers}, for information on how stepping
19565 commands behave when the thread has a signal queued.
19568 @section Returning from a Function
19571 @cindex returning from a function
19574 @itemx return @var{expression}
19575 You can cancel execution of a function call with the @code{return}
19576 command. If you give an
19577 @var{expression} argument, its value is used as the function's return
19581 When you use @code{return}, @value{GDBN} discards the selected stack frame
19582 (and all frames within it). You can think of this as making the
19583 discarded frame return prematurely. If you wish to specify a value to
19584 be returned, give that value as the argument to @code{return}.
19586 This pops the selected stack frame (@pxref{Selection, ,Selecting a
19587 Frame}), and any other frames inside of it, leaving its caller as the
19588 innermost remaining frame. That frame becomes selected. The
19589 specified value is stored in the registers used for returning values
19592 The @code{return} command does not resume execution; it leaves the
19593 program stopped in the state that would exist if the function had just
19594 returned. In contrast, the @code{finish} command (@pxref{Continuing
19595 and Stepping, ,Continuing and Stepping}) resumes execution until the
19596 selected stack frame returns naturally.
19598 @value{GDBN} needs to know how the @var{expression} argument should be set for
19599 the inferior. The concrete registers assignment depends on the OS ABI and the
19600 type being returned by the selected stack frame. For example it is common for
19601 OS ABI to return floating point values in FPU registers while integer values in
19602 CPU registers. Still some ABIs return even floating point values in CPU
19603 registers. Larger integer widths (such as @code{long long int}) also have
19604 specific placement rules. @value{GDBN} already knows the OS ABI from its
19605 current target so it needs to find out also the type being returned to make the
19606 assignment into the right register(s).
19608 Normally, the selected stack frame has debug info. @value{GDBN} will always
19609 use the debug info instead of the implicit type of @var{expression} when the
19610 debug info is available. For example, if you type @kbd{return -1}, and the
19611 function in the current stack frame is declared to return a @code{long long
19612 int}, @value{GDBN} transparently converts the implicit @code{int} value of -1
19613 into a @code{long long int}:
19616 Breakpoint 1, func () at gdb.base/return-nodebug.c:29
19618 (@value{GDBP}) return -1
19619 Make func return now? (y or n) y
19620 #0 0x004004f6 in main () at gdb.base/return-nodebug.c:43
19621 43 printf ("result=%lld\n", func ());
19625 However, if the selected stack frame does not have a debug info, e.g., if the
19626 function was compiled without debug info, @value{GDBN} has to find out the type
19627 to return from user. Specifying a different type by mistake may set the value
19628 in different inferior registers than the caller code expects. For example,
19629 typing @kbd{return -1} with its implicit type @code{int} would set only a part
19630 of a @code{long long int} result for a debug info less function (on 32-bit
19631 architectures). Therefore the user is required to specify the return type by
19632 an appropriate cast explicitly:
19635 Breakpoint 2, 0x0040050b in func ()
19636 (@value{GDBP}) return -1
19637 Return value type not available for selected stack frame.
19638 Please use an explicit cast of the value to return.
19639 (@value{GDBP}) return (long long int) -1
19640 Make selected stack frame return now? (y or n) y
19641 #0 0x00400526 in main ()
19646 @section Calling Program Functions
19649 @cindex calling functions
19650 @cindex inferior functions, calling
19651 @item print @var{expr}
19652 Evaluate the expression @var{expr} and display the resulting value.
19653 The expression may include calls to functions in the program being
19657 @item call @var{expr}
19658 Evaluate the expression @var{expr} without displaying @code{void}
19661 You can use this variant of the @code{print} command if you want to
19662 execute a function from your program that does not return anything
19663 (a.k.a.@: @dfn{a void function}), but without cluttering the output
19664 with @code{void} returned values that @value{GDBN} will otherwise
19665 print. If the result is not void, it is printed and saved in the
19669 It is possible for the function you call via the @code{print} or
19670 @code{call} command to generate a signal (e.g., if there's a bug in
19671 the function, or if you passed it incorrect arguments). What happens
19672 in that case is controlled by the @code{set unwindonsignal} command.
19674 Similarly, with a C@t{++} program it is possible for the function you
19675 call via the @code{print} or @code{call} command to generate an
19676 exception that is not handled due to the constraints of the dummy
19677 frame. In this case, any exception that is raised in the frame, but has
19678 an out-of-frame exception handler will not be found. GDB builds a
19679 dummy-frame for the inferior function call, and the unwinder cannot
19680 seek for exception handlers outside of this dummy-frame. What happens
19681 in that case is controlled by the
19682 @code{set unwind-on-terminating-exception} command.
19685 @item set unwindonsignal
19686 @kindex set unwindonsignal
19687 @cindex unwind stack in called functions
19688 @cindex call dummy stack unwinding
19689 Set unwinding of the stack if a signal is received while in a function
19690 that @value{GDBN} called in the program being debugged. If set to on,
19691 @value{GDBN} unwinds the stack it created for the call and restores
19692 the context to what it was before the call. If set to off (the
19693 default), @value{GDBN} stops in the frame where the signal was
19696 @item show unwindonsignal
19697 @kindex show unwindonsignal
19698 Show the current setting of stack unwinding in the functions called by
19701 @item set unwind-on-terminating-exception
19702 @kindex set unwind-on-terminating-exception
19703 @cindex unwind stack in called functions with unhandled exceptions
19704 @cindex call dummy stack unwinding on unhandled exception.
19705 Set unwinding of the stack if a C@t{++} exception is raised, but left
19706 unhandled while in a function that @value{GDBN} called in the program being
19707 debugged. If set to on (the default), @value{GDBN} unwinds the stack
19708 it created for the call and restores the context to what it was before
19709 the call. If set to off, @value{GDBN} the exception is delivered to
19710 the default C@t{++} exception handler and the inferior terminated.
19712 @item show unwind-on-terminating-exception
19713 @kindex show unwind-on-terminating-exception
19714 Show the current setting of stack unwinding in the functions called by
19717 @item set may-call-functions
19718 @kindex set may-call-functions
19719 @cindex disabling calling functions in the program
19720 @cindex calling functions in the program, disabling
19721 Set permission to call functions in the program.
19722 This controls whether @value{GDBN} will attempt to call functions in
19723 the program, such as with expressions in the @code{print} command. It
19724 defaults to @code{on}.
19726 To call a function in the program, @value{GDBN} has to temporarily
19727 modify the state of the inferior. This has potentially undesired side
19728 effects. Also, having @value{GDBN} call nested functions is likely to
19729 be erroneous and may even crash the program being debugged. You can
19730 avoid such hazards by forbidding @value{GDBN} from calling functions
19731 in the program being debugged. If calling functions in the program
19732 is forbidden, GDB will throw an error when a command (such as printing
19733 an expression) starts a function call in the program.
19735 @item show may-call-functions
19736 @kindex show may-call-functions
19737 Show permission to call functions in the program.
19741 @subsection Calling functions with no debug info
19743 @cindex no debug info functions
19744 Sometimes, a function you wish to call is missing debug information.
19745 In such case, @value{GDBN} does not know the type of the function,
19746 including the types of the function's parameters. To avoid calling
19747 the inferior function incorrectly, which could result in the called
19748 function functioning erroneously and even crash, @value{GDBN} refuses
19749 to call the function unless you tell it the type of the function.
19751 For prototyped (i.e.@: ANSI/ISO style) functions, there are two ways
19752 to do that. The simplest is to cast the call to the function's
19753 declared return type. For example:
19756 (@value{GDBP}) p getenv ("PATH")
19757 'getenv' has unknown return type; cast the call to its declared return type
19758 (@value{GDBP}) p (char *) getenv ("PATH")
19759 $1 = 0x7fffffffe7ba "/usr/local/bin:/"...
19762 Casting the return type of a no-debug function is equivalent to
19763 casting the function to a pointer to a prototyped function that has a
19764 prototype that matches the types of the passed-in arguments, and
19765 calling that. I.e., the call above is equivalent to:
19768 (@value{GDBP}) p ((char * (*) (const char *)) getenv) ("PATH")
19772 and given this prototyped C or C++ function with float parameters:
19775 float multiply (float v1, float v2) @{ return v1 * v2; @}
19779 these calls are equivalent:
19782 (@value{GDBP}) p (float) multiply (2.0f, 3.0f)
19783 (@value{GDBP}) p ((float (*) (float, float)) multiply) (2.0f, 3.0f)
19786 If the function you wish to call is declared as unprototyped (i.e.@:
19787 old K&R style), you must use the cast-to-function-pointer syntax, so
19788 that @value{GDBN} knows that it needs to apply default argument
19789 promotions (promote float arguments to double). @xref{ABI, float
19790 promotion}. For example, given this unprototyped C function with
19791 float parameters, and no debug info:
19795 multiply_noproto (v1, v2)
19803 you call it like this:
19806 (@value{GDBP}) p ((float (*) ()) multiply_noproto) (2.0f, 3.0f)
19810 @section Patching Programs
19812 @cindex patching binaries
19813 @cindex writing into executables
19814 @cindex writing into corefiles
19816 By default, @value{GDBN} opens the file containing your program's
19817 executable code (or the corefile) read-only. This prevents accidental
19818 alterations to machine code; but it also prevents you from intentionally
19819 patching your program's binary.
19821 If you'd like to be able to patch the binary, you can specify that
19822 explicitly with the @code{set write} command. For example, you might
19823 want to turn on internal debugging flags, or even to make emergency
19829 @itemx set write off
19830 If you specify @samp{set write on}, @value{GDBN} opens executable and
19831 core files for both reading and writing; if you specify @kbd{set write
19832 off} (the default), @value{GDBN} opens them read-only.
19834 If you have already loaded a file, you must load it again (using the
19835 @code{exec-file} or @code{core-file} command) after changing @code{set
19836 write}, for your new setting to take effect.
19840 Display whether executable files and core files are opened for writing
19841 as well as reading.
19844 @node Compiling and Injecting Code
19845 @section Compiling and injecting code in @value{GDBN}
19846 @cindex injecting code
19847 @cindex writing into executables
19848 @cindex compiling code
19850 @value{GDBN} supports on-demand compilation and code injection into
19851 programs running under @value{GDBN}. GCC 5.0 or higher built with
19852 @file{libcc1.so} must be installed for this functionality to be enabled.
19853 This functionality is implemented with the following commands.
19856 @kindex compile code
19857 @item compile code @var{source-code}
19858 @itemx compile code -raw @var{--} @var{source-code}
19859 Compile @var{source-code} with the compiler language found as the current
19860 language in @value{GDBN} (@pxref{Languages}). If compilation and
19861 injection is not supported with the current language specified in
19862 @value{GDBN}, or the compiler does not support this feature, an error
19863 message will be printed. If @var{source-code} compiles and links
19864 successfully, @value{GDBN} will load the object-code emitted,
19865 and execute it within the context of the currently selected inferior.
19866 It is important to note that the compiled code is executed immediately.
19867 After execution, the compiled code is removed from @value{GDBN} and any
19868 new types or variables you have defined will be deleted.
19870 The command allows you to specify @var{source-code} in two ways.
19871 The simplest method is to provide a single line of code to the command.
19875 compile code printf ("hello world\n");
19878 If you specify options on the command line as well as source code, they
19879 may conflict. The @samp{--} delimiter can be used to separate options
19880 from actual source code. E.g.:
19883 compile code -r -- printf ("hello world\n");
19886 Alternatively you can enter source code as multiple lines of text. To
19887 enter this mode, invoke the @samp{compile code} command without any text
19888 following the command. This will start the multiple-line editor and
19889 allow you to type as many lines of source code as required. When you
19890 have completed typing, enter @samp{end} on its own line to exit the
19895 >printf ("hello\n");
19896 >printf ("world\n");
19900 Specifying @samp{-raw}, prohibits @value{GDBN} from wrapping the
19901 provided @var{source-code} in a callable scope. In this case, you must
19902 specify the entry point of the code by defining a function named
19903 @code{_gdb_expr_}. The @samp{-raw} code cannot access variables of the
19904 inferior. Using @samp{-raw} option may be needed for example when
19905 @var{source-code} requires @samp{#include} lines which may conflict with
19906 inferior symbols otherwise.
19908 @kindex compile file
19909 @item compile file @var{filename}
19910 @itemx compile file -raw @var{filename}
19911 Like @code{compile code}, but take the source code from @var{filename}.
19914 compile file /home/user/example.c
19919 @item compile print [[@var{options}] --] @var{expr}
19920 @itemx compile print [[@var{options}] --] /@var{f} @var{expr}
19921 Compile and execute @var{expr} with the compiler language found as the
19922 current language in @value{GDBN} (@pxref{Languages}). By default the
19923 value of @var{expr} is printed in a format appropriate to its data type;
19924 you can choose a different format by specifying @samp{/@var{f}}, where
19925 @var{f} is a letter specifying the format; see @ref{Output Formats,,Output
19926 Formats}. The @code{compile print} command accepts the same options
19927 as the @code{print} command; see @ref{print options}.
19929 @item compile print [[@var{options}] --]
19930 @itemx compile print [[@var{options}] --] /@var{f}
19931 @cindex reprint the last value
19932 Alternatively you can enter the expression (source code producing it) as
19933 multiple lines of text. To enter this mode, invoke the @samp{compile print}
19934 command without any text following the command. This will start the
19935 multiple-line editor.
19939 The process of compiling and injecting the code can be inspected using:
19942 @anchor{set debug compile}
19943 @item set debug compile
19944 @cindex compile command debugging info
19945 Turns on or off display of @value{GDBN} process of compiling and
19946 injecting the code. The default is off.
19948 @item show debug compile
19949 Displays the current state of displaying @value{GDBN} process of
19950 compiling and injecting the code.
19952 @anchor{set debug compile-cplus-types}
19953 @item set debug compile-cplus-types
19954 @cindex compile C@t{++} type conversion
19955 Turns on or off the display of C@t{++} type conversion debugging information.
19956 The default is off.
19958 @item show debug compile-cplus-types
19959 Displays the current state of displaying debugging information for
19960 C@t{++} type conversion.
19963 @subsection Compilation options for the @code{compile} command
19965 @value{GDBN} needs to specify the right compilation options for the code
19966 to be injected, in part to make its ABI compatible with the inferior
19967 and in part to make the injected code compatible with @value{GDBN}'s
19971 The options used, in increasing precedence:
19974 @item target architecture and OS options (@code{gdbarch})
19975 These options depend on target processor type and target operating
19976 system, usually they specify at least 32-bit (@code{-m32}) or 64-bit
19977 (@code{-m64}) compilation option.
19979 @item compilation options recorded in the target
19980 @value{NGCC} (since version 4.7) stores the options used for compilation
19981 into @code{DW_AT_producer} part of DWARF debugging information according
19982 to the @value{NGCC} option @code{-grecord-gcc-switches}. One has to
19983 explicitly specify @code{-g} during inferior compilation otherwise
19984 @value{NGCC} produces no DWARF. This feature is only relevant for
19985 platforms where @code{-g} produces DWARF by default, otherwise one may
19986 try to enforce DWARF by using @code{-gdwarf-4}.
19988 @item compilation options set by @code{set compile-args}
19992 You can override compilation options using the following command:
19995 @item set compile-args
19996 @cindex compile command options override
19997 Set compilation options used for compiling and injecting code with the
19998 @code{compile} commands. These options override any conflicting ones
19999 from the target architecture and/or options stored during inferior
20002 @item show compile-args
20003 Displays the current state of compilation options override.
20004 This does not show all the options actually used during compilation,
20005 use @ref{set debug compile} for that.
20008 @subsection Caveats when using the @code{compile} command
20010 There are a few caveats to keep in mind when using the @code{compile}
20011 command. As the caveats are different per language, the table below
20012 highlights specific issues on a per language basis.
20015 @item C code examples and caveats
20016 When the language in @value{GDBN} is set to @samp{C}, the compiler will
20017 attempt to compile the source code with a @samp{C} compiler. The source
20018 code provided to the @code{compile} command will have much the same
20019 access to variables and types as it normally would if it were part of
20020 the program currently being debugged in @value{GDBN}.
20022 Below is a sample program that forms the basis of the examples that
20023 follow. This program has been compiled and loaded into @value{GDBN},
20024 much like any other normal debugging session.
20027 void function1 (void)
20030 printf ("function 1\n");
20033 void function2 (void)
20048 For the purposes of the examples in this section, the program above has
20049 been compiled, loaded into @value{GDBN}, stopped at the function
20050 @code{main}, and @value{GDBN} is awaiting input from the user.
20052 To access variables and types for any program in @value{GDBN}, the
20053 program must be compiled and packaged with debug information. The
20054 @code{compile} command is not an exception to this rule. Without debug
20055 information, you can still use the @code{compile} command, but you will
20056 be very limited in what variables and types you can access.
20058 So with that in mind, the example above has been compiled with debug
20059 information enabled. The @code{compile} command will have access to
20060 all variables and types (except those that may have been optimized
20061 out). Currently, as @value{GDBN} has stopped the program in the
20062 @code{main} function, the @code{compile} command would have access to
20063 the variable @code{k}. You could invoke the @code{compile} command
20064 and type some source code to set the value of @code{k}. You can also
20065 read it, or do anything with that variable you would normally do in
20066 @code{C}. Be aware that changes to inferior variables in the
20067 @code{compile} command are persistent. In the following example:
20070 compile code k = 3;
20074 the variable @code{k} is now 3. It will retain that value until
20075 something else in the example program changes it, or another
20076 @code{compile} command changes it.
20078 Normal scope and access rules apply to source code compiled and
20079 injected by the @code{compile} command. In the example, the variables
20080 @code{j} and @code{k} are not accessible yet, because the program is
20081 currently stopped in the @code{main} function, where these variables
20082 are not in scope. Therefore, the following command
20085 compile code j = 3;
20089 will result in a compilation error message.
20091 Once the program is continued, execution will bring these variables in
20092 scope, and they will become accessible; then the code you specify via
20093 the @code{compile} command will be able to access them.
20095 You can create variables and types with the @code{compile} command as
20096 part of your source code. Variables and types that are created as part
20097 of the @code{compile} command are not visible to the rest of the program for
20098 the duration of its run. This example is valid:
20101 compile code int ff = 5; printf ("ff is %d\n", ff);
20104 However, if you were to type the following into @value{GDBN} after that
20105 command has completed:
20108 compile code printf ("ff is %d\n'', ff);
20112 a compiler error would be raised as the variable @code{ff} no longer
20113 exists. Object code generated and injected by the @code{compile}
20114 command is removed when its execution ends. Caution is advised
20115 when assigning to program variables values of variables created by the
20116 code submitted to the @code{compile} command. This example is valid:
20119 compile code int ff = 5; k = ff;
20122 The value of the variable @code{ff} is assigned to @code{k}. The variable
20123 @code{k} does not require the existence of @code{ff} to maintain the value
20124 it has been assigned. However, pointers require particular care in
20125 assignment. If the source code compiled with the @code{compile} command
20126 changed the address of a pointer in the example program, perhaps to a
20127 variable created in the @code{compile} command, that pointer would point
20128 to an invalid location when the command exits. The following example
20129 would likely cause issues with your debugged program:
20132 compile code int ff = 5; p = &ff;
20135 In this example, @code{p} would point to @code{ff} when the
20136 @code{compile} command is executing the source code provided to it.
20137 However, as variables in the (example) program persist with their
20138 assigned values, the variable @code{p} would point to an invalid
20139 location when the command exists. A general rule should be followed
20140 in that you should either assign @code{NULL} to any assigned pointers,
20141 or restore a valid location to the pointer before the command exits.
20143 Similar caution must be exercised with any structs, unions, and typedefs
20144 defined in @code{compile} command. Types defined in the @code{compile}
20145 command will no longer be available in the next @code{compile} command.
20146 Therefore, if you cast a variable to a type defined in the
20147 @code{compile} command, care must be taken to ensure that any future
20148 need to resolve the type can be achieved.
20151 (@value{GDBP}) compile code static struct a @{ int a; @} v = @{ 42 @}; argv = &v;
20152 (@value{GDBP}) compile code printf ("%d\n", ((struct a *) argv)->a);
20153 gdb command line:1:36: error: dereferencing pointer to incomplete type ‘struct a’
20154 Compilation failed.
20155 (@value{GDBP}) compile code struct a @{ int a; @}; printf ("%d\n", ((struct a *) argv)->a);
20159 Variables that have been optimized away by the compiler are not
20160 accessible to the code submitted to the @code{compile} command.
20161 Access to those variables will generate a compiler error which @value{GDBN}
20162 will print to the console.
20165 @subsection Compiler search for the @code{compile} command
20167 @value{GDBN} needs to find @value{NGCC} for the inferior being debugged
20168 which may not be obvious for remote targets of different architecture
20169 than where @value{GDBN} is running. Environment variable @code{PATH} on
20170 @value{GDBN} host is searched for @value{NGCC} binary matching the
20171 target architecture and operating system. This search can be overriden
20172 by @code{set compile-gcc} @value{GDBN} command below. @code{PATH} is
20173 taken from shell that executed @value{GDBN}, it is not the value set by
20174 @value{GDBN} command @code{set environment}). @xref{Environment}.
20177 Specifically @code{PATH} is searched for binaries matching regular expression
20178 @code{@var{arch}(-[^-]*)?-@var{os}-gcc} according to the inferior target being
20179 debugged. @var{arch} is processor name --- multiarch is supported, so for
20180 example both @code{i386} and @code{x86_64} targets look for pattern
20181 @code{(x86_64|i.86)} and both @code{s390} and @code{s390x} targets look
20182 for pattern @code{s390x?}. @var{os} is currently supported only for
20183 pattern @code{linux(-gnu)?}.
20185 On Posix hosts the compiler driver @value{GDBN} needs to find also
20186 shared library @file{libcc1.so} from the compiler. It is searched in
20187 default shared library search path (overridable with usual environment
20188 variable @code{LD_LIBRARY_PATH}), unrelated to @code{PATH} or @code{set
20189 compile-gcc} settings. Contrary to it @file{libcc1plugin.so} is found
20190 according to the installation of the found compiler --- as possibly
20191 specified by the @code{set compile-gcc} command.
20194 @item set compile-gcc
20195 @cindex compile command driver filename override
20196 Set compilation command used for compiling and injecting code with the
20197 @code{compile} commands. If this option is not set (it is set to
20198 an empty string), the search described above will occur --- that is the
20201 @item show compile-gcc
20202 Displays the current compile command @value{NGCC} driver filename.
20203 If set, it is the main command @command{gcc}, found usually for example
20204 under name @file{x86_64-linux-gnu-gcc}.
20208 @chapter @value{GDBN} Files
20210 @value{GDBN} needs to know the file name of the program to be debugged,
20211 both in order to read its symbol table and in order to start your
20212 program. To debug a core dump of a previous run, you must also tell
20213 @value{GDBN} the name of the core dump file.
20216 * Files:: Commands to specify files
20217 * File Caching:: Information about @value{GDBN}'s file caching
20218 * Separate Debug Files:: Debugging information in separate files
20219 * MiniDebugInfo:: Debugging information in a special section
20220 * Index Files:: Index files speed up @value{GDBN}
20221 * Symbol Errors:: Errors reading symbol files
20222 * Data Files:: @value{GDBN} data files
20226 @section Commands to Specify Files
20228 @cindex symbol table
20229 @cindex core dump file
20231 You may want to specify executable and core dump file names. The usual
20232 way to do this is at start-up time, using the arguments to
20233 @value{GDBN}'s start-up commands (@pxref{Invocation, , Getting In and
20234 Out of @value{GDBN}}).
20236 Occasionally it is necessary to change to a different file during a
20237 @value{GDBN} session. Or you may run @value{GDBN} and forget to
20238 specify a file you want to use. Or you are debugging a remote target
20239 via @code{gdbserver} (@pxref{Server, file, Using the @code{gdbserver}
20240 Program}). In these situations the @value{GDBN} commands to specify
20241 new files are useful.
20244 @cindex executable file
20246 @item file @var{filename}
20247 Use @var{filename} as the program to be debugged. It is read for its
20248 symbols and for the contents of pure memory. It is also the program
20249 executed when you use the @code{run} command. If you do not specify a
20250 directory and the file is not found in the @value{GDBN} working directory,
20251 @value{GDBN} uses the environment variable @code{PATH} as a list of
20252 directories to search, just as the shell does when looking for a program
20253 to run. You can change the value of this variable, for both @value{GDBN}
20254 and your program, using the @code{path} command.
20256 @cindex unlinked object files
20257 @cindex patching object files
20258 You can load unlinked object @file{.o} files into @value{GDBN} using
20259 the @code{file} command. You will not be able to ``run'' an object
20260 file, but you can disassemble functions and inspect variables. Also,
20261 if the underlying BFD functionality supports it, you could use
20262 @kbd{gdb -write} to patch object files using this technique. Note
20263 that @value{GDBN} can neither interpret nor modify relocations in this
20264 case, so branches and some initialized variables will appear to go to
20265 the wrong place. But this feature is still handy from time to time.
20268 @code{file} with no argument makes @value{GDBN} discard any information it
20269 has on both executable file and the symbol table.
20272 @item exec-file @r{[} @var{filename} @r{]}
20273 Specify that the program to be run (but not the symbol table) is found
20274 in @var{filename}. @value{GDBN} searches the environment variable @code{PATH}
20275 if necessary to locate your program. Omitting @var{filename} means to
20276 discard information on the executable file.
20278 @kindex symbol-file
20279 @item symbol-file @r{[} @var{filename} @r{[} -o @var{offset} @r{]]}
20280 Read symbol table information from file @var{filename}. @code{PATH} is
20281 searched when necessary. Use the @code{file} command to get both symbol
20282 table and program to run from the same file.
20284 If an optional @var{offset} is specified, it is added to the start
20285 address of each section in the symbol file. This is useful if the
20286 program is relocated at runtime, such as the Linux kernel with kASLR
20289 @code{symbol-file} with no argument clears out @value{GDBN} information on your
20290 program's symbol table.
20292 The @code{symbol-file} command causes @value{GDBN} to forget the contents of
20293 some breakpoints and auto-display expressions. This is because they may
20294 contain pointers to the internal data recording symbols and data types,
20295 which are part of the old symbol table data being discarded inside
20298 @code{symbol-file} does not repeat if you press @key{RET} again after
20301 When @value{GDBN} is configured for a particular environment, it
20302 understands debugging information in whatever format is the standard
20303 generated for that environment; you may use either a @sc{gnu} compiler, or
20304 other compilers that adhere to the local conventions.
20305 Best results are usually obtained from @sc{gnu} compilers; for example,
20306 using @code{@value{NGCC}} you can generate debugging information for
20309 For most kinds of object files, with the exception of old SVR3 systems
20310 using COFF, the @code{symbol-file} command does not normally read the
20311 symbol table in full right away. Instead, it scans the symbol table
20312 quickly to find which source files and which symbols are present. The
20313 details are read later, one source file at a time, as they are needed.
20315 The purpose of this two-stage reading strategy is to make @value{GDBN}
20316 start up faster. For the most part, it is invisible except for
20317 occasional pauses while the symbol table details for a particular source
20318 file are being read. (The @code{set verbose} command can turn these
20319 pauses into messages if desired. @xref{Messages/Warnings, ,Optional
20320 Warnings and Messages}.)
20322 We have not implemented the two-stage strategy for COFF yet. When the
20323 symbol table is stored in COFF format, @code{symbol-file} reads the
20324 symbol table data in full right away. Note that ``stabs-in-COFF''
20325 still does the two-stage strategy, since the debug info is actually
20329 @cindex reading symbols immediately
20330 @cindex symbols, reading immediately
20331 @item symbol-file @r{[} -readnow @r{]} @var{filename}
20332 @itemx file @r{[} -readnow @r{]} @var{filename}
20333 You can override the @value{GDBN} two-stage strategy for reading symbol
20334 tables by using the @samp{-readnow} option with any of the commands that
20335 load symbol table information, if you want to be sure @value{GDBN} has the
20336 entire symbol table available.
20338 @cindex @code{-readnever}, option for symbol-file command
20339 @cindex never read symbols
20340 @cindex symbols, never read
20341 @item symbol-file @r{[} -readnever @r{]} @var{filename}
20342 @itemx file @r{[} -readnever @r{]} @var{filename}
20343 You can instruct @value{GDBN} to never read the symbolic information
20344 contained in @var{filename} by using the @samp{-readnever} option.
20345 @xref{--readnever}.
20347 @c FIXME: for now no mention of directories, since this seems to be in
20348 @c flux. 13mar1992 status is that in theory GDB would look either in
20349 @c current dir or in same dir as myprog; but issues like competing
20350 @c GDB's, or clutter in system dirs, mean that in practice right now
20351 @c only current dir is used. FFish says maybe a special GDB hierarchy
20352 @c (eg rooted in val of env var GDBSYMS) could exist for mappable symbol
20356 @item core-file @r{[}@var{filename}@r{]}
20358 Specify the whereabouts of a core dump file to be used as the ``contents
20359 of memory''. Traditionally, core files contain only some parts of the
20360 address space of the process that generated them; @value{GDBN} can access the
20361 executable file itself for other parts.
20363 @code{core-file} with no argument specifies that no core file is
20366 Note that the core file is ignored when your program is actually running
20367 under @value{GDBN}. So, if you have been running your program and you
20368 wish to debug a core file instead, you must kill the subprocess in which
20369 the program is running. To do this, use the @code{kill} command
20370 (@pxref{Kill Process, ,Killing the Child Process}).
20372 @kindex add-symbol-file
20373 @cindex dynamic linking
20374 @item add-symbol-file @var{filename} @r{[} -readnow @r{|} -readnever @r{]} @r{[} -o @var{offset} @r{]} @r{[} @var{textaddress} @r{]} @r{[} -s @var{section} @var{address} @dots{} @r{]}
20375 The @code{add-symbol-file} command reads additional symbol table
20376 information from the file @var{filename}. You would use this command
20377 when @var{filename} has been dynamically loaded (by some other means)
20378 into the program that is running. The @var{textaddress} parameter gives
20379 the memory address at which the file's text section has been loaded.
20380 You can additionally specify the base address of other sections using
20381 an arbitrary number of @samp{-s @var{section} @var{address}} pairs.
20382 If a section is omitted, @value{GDBN} will use its default addresses
20383 as found in @var{filename}. Any @var{address} or @var{textaddress}
20384 can be given as an expression.
20386 If an optional @var{offset} is specified, it is added to the start
20387 address of each section, except those for which the address was
20388 specified explicitly.
20390 The symbol table of the file @var{filename} is added to the symbol table
20391 originally read with the @code{symbol-file} command. You can use the
20392 @code{add-symbol-file} command any number of times; the new symbol data
20393 thus read is kept in addition to the old.
20395 Changes can be reverted using the command @code{remove-symbol-file}.
20397 @cindex relocatable object files, reading symbols from
20398 @cindex object files, relocatable, reading symbols from
20399 @cindex reading symbols from relocatable object files
20400 @cindex symbols, reading from relocatable object files
20401 @cindex @file{.o} files, reading symbols from
20402 Although @var{filename} is typically a shared library file, an
20403 executable file, or some other object file which has been fully
20404 relocated for loading into a process, you can also load symbolic
20405 information from relocatable @file{.o} files, as long as:
20409 the file's symbolic information refers only to linker symbols defined in
20410 that file, not to symbols defined by other object files,
20412 every section the file's symbolic information refers to has actually
20413 been loaded into the inferior, as it appears in the file, and
20415 you can determine the address at which every section was loaded, and
20416 provide these to the @code{add-symbol-file} command.
20420 Some embedded operating systems, like Sun Chorus and VxWorks, can load
20421 relocatable files into an already running program; such systems
20422 typically make the requirements above easy to meet. However, it's
20423 important to recognize that many native systems use complex link
20424 procedures (@code{.linkonce} section factoring and C@t{++} constructor table
20425 assembly, for example) that make the requirements difficult to meet. In
20426 general, one cannot assume that using @code{add-symbol-file} to read a
20427 relocatable object file's symbolic information will have the same effect
20428 as linking the relocatable object file into the program in the normal
20431 @code{add-symbol-file} does not repeat if you press @key{RET} after using it.
20433 @kindex remove-symbol-file
20434 @item remove-symbol-file @var{filename}
20435 @item remove-symbol-file -a @var{address}
20436 Remove a symbol file added via the @code{add-symbol-file} command. The
20437 file to remove can be identified by its @var{filename} or by an @var{address}
20438 that lies within the boundaries of this symbol file in memory. Example:
20441 (@value{GDBP}) add-symbol-file /home/user/gdb/mylib.so 0x7ffff7ff9480
20442 add symbol table from file "/home/user/gdb/mylib.so" at
20443 .text_addr = 0x7ffff7ff9480
20445 Reading symbols from /home/user/gdb/mylib.so...done.
20446 (@value{GDBP}) remove-symbol-file -a 0x7ffff7ff9480
20447 Remove symbol table from file "/home/user/gdb/mylib.so"? (y or n) y
20452 @code{remove-symbol-file} does not repeat if you press @key{RET} after using it.
20454 @kindex add-symbol-file-from-memory
20455 @cindex @code{syscall DSO}
20456 @cindex load symbols from memory
20457 @item add-symbol-file-from-memory @var{address}
20458 Load symbols from the given @var{address} in a dynamically loaded
20459 object file whose image is mapped directly into the inferior's memory.
20460 For example, the Linux kernel maps a @code{syscall DSO} into each
20461 process's address space; this DSO provides kernel-specific code for
20462 some system calls. The argument can be any expression whose
20463 evaluation yields the address of the file's shared object file header.
20464 For this command to work, you must have used @code{symbol-file} or
20465 @code{exec-file} commands in advance.
20468 @item section @var{section} @var{addr}
20469 The @code{section} command changes the base address of the named
20470 @var{section} of the exec file to @var{addr}. This can be used if the
20471 exec file does not contain section addresses, (such as in the
20472 @code{a.out} format), or when the addresses specified in the file
20473 itself are wrong. Each section must be changed separately. The
20474 @code{info files} command, described below, lists all the sections and
20478 @kindex info target
20481 @code{info files} and @code{info target} are synonymous; both print the
20482 current target (@pxref{Targets, ,Specifying a Debugging Target}),
20483 including the names of the executable and core dump files currently in
20484 use by @value{GDBN}, and the files from which symbols were loaded. The
20485 command @code{help target} lists all possible targets rather than
20488 @kindex maint info sections
20489 @item maint info sections
20490 Another command that can give you extra information about program sections
20491 is @code{maint info sections}. In addition to the section information
20492 displayed by @code{info files}, this command displays the flags and file
20493 offset of each section in the executable and core dump files. In addition,
20494 @code{maint info sections} provides the following command options (which
20495 may be arbitrarily combined):
20499 Display sections for all loaded object files, including shared libraries.
20500 @item @var{sections}
20501 Display info only for named @var{sections}.
20502 @item @var{section-flags}
20503 Display info only for sections for which @var{section-flags} are true.
20504 The section flags that @value{GDBN} currently knows about are:
20507 Section will have space allocated in the process when loaded.
20508 Set for all sections except those containing debug information.
20510 Section will be loaded from the file into the child process memory.
20511 Set for pre-initialized code and data, clear for @code{.bss} sections.
20513 Section needs to be relocated before loading.
20515 Section cannot be modified by the child process.
20517 Section contains executable code only.
20519 Section contains data only (no executable code).
20521 Section will reside in ROM.
20523 Section contains data for constructor/destructor lists.
20525 Section is not empty.
20527 An instruction to the linker to not output the section.
20528 @item COFF_SHARED_LIBRARY
20529 A notification to the linker that the section contains
20530 COFF shared library information.
20532 Section contains common symbols.
20535 @kindex set trust-readonly-sections
20536 @cindex read-only sections
20537 @item set trust-readonly-sections on
20538 Tell @value{GDBN} that readonly sections in your object file
20539 really are read-only (i.e.@: that their contents will not change).
20540 In that case, @value{GDBN} can fetch values from these sections
20541 out of the object file, rather than from the target program.
20542 For some targets (notably embedded ones), this can be a significant
20543 enhancement to debugging performance.
20545 The default is off.
20547 @item set trust-readonly-sections off
20548 Tell @value{GDBN} not to trust readonly sections. This means that
20549 the contents of the section might change while the program is running,
20550 and must therefore be fetched from the target when needed.
20552 @item show trust-readonly-sections
20553 Show the current setting of trusting readonly sections.
20556 All file-specifying commands allow both absolute and relative file names
20557 as arguments. @value{GDBN} always converts the file name to an absolute file
20558 name and remembers it that way.
20560 @cindex shared libraries
20561 @anchor{Shared Libraries}
20562 @value{GDBN} supports @sc{gnu}/Linux, MS-Windows, SunOS,
20563 Darwin/Mach-O, SVr4, IBM RS/6000 AIX, QNX Neutrino, FDPIC (FR-V), and
20564 DSBT (TIC6X) shared libraries.
20566 On MS-Windows @value{GDBN} must be linked with the Expat library to support
20567 shared libraries. @xref{Expat}.
20569 @value{GDBN} automatically loads symbol definitions from shared libraries
20570 when you use the @code{run} command, or when you examine a core file.
20571 (Before you issue the @code{run} command, @value{GDBN} does not understand
20572 references to a function in a shared library, however---unless you are
20573 debugging a core file).
20575 @c FIXME: some @value{GDBN} release may permit some refs to undef
20576 @c FIXME...symbols---eg in a break cmd---assuming they are from a shared
20577 @c FIXME...lib; check this from time to time when updating manual
20579 There are times, however, when you may wish to not automatically load
20580 symbol definitions from shared libraries, such as when they are
20581 particularly large or there are many of them.
20583 To control the automatic loading of shared library symbols, use the
20587 @kindex set auto-solib-add
20588 @item set auto-solib-add @var{mode}
20589 If @var{mode} is @code{on}, symbols from all shared object libraries
20590 will be loaded automatically when the inferior begins execution, you
20591 attach to an independently started inferior, or when the dynamic linker
20592 informs @value{GDBN} that a new library has been loaded. If @var{mode}
20593 is @code{off}, symbols must be loaded manually, using the
20594 @code{sharedlibrary} command. The default value is @code{on}.
20596 @cindex memory used for symbol tables
20597 If your program uses lots of shared libraries with debug info that
20598 takes large amounts of memory, you can decrease the @value{GDBN}
20599 memory footprint by preventing it from automatically loading the
20600 symbols from shared libraries. To that end, type @kbd{set
20601 auto-solib-add off} before running the inferior, then load each
20602 library whose debug symbols you do need with @kbd{sharedlibrary
20603 @var{regexp}}, where @var{regexp} is a regular expression that matches
20604 the libraries whose symbols you want to be loaded.
20606 @kindex show auto-solib-add
20607 @item show auto-solib-add
20608 Display the current autoloading mode.
20611 @cindex load shared library
20612 To explicitly load shared library symbols, use the @code{sharedlibrary}
20616 @kindex info sharedlibrary
20618 @item info share @var{regex}
20619 @itemx info sharedlibrary @var{regex}
20620 Print the names of the shared libraries which are currently loaded
20621 that match @var{regex}. If @var{regex} is omitted then print
20622 all shared libraries that are loaded.
20625 @item info dll @var{regex}
20626 This is an alias of @code{info sharedlibrary}.
20628 @kindex sharedlibrary
20630 @item sharedlibrary @var{regex}
20631 @itemx share @var{regex}
20632 Load shared object library symbols for files matching a
20633 Unix regular expression.
20634 As with files loaded automatically, it only loads shared libraries
20635 required by your program for a core file or after typing @code{run}. If
20636 @var{regex} is omitted all shared libraries required by your program are
20639 @item nosharedlibrary
20640 @kindex nosharedlibrary
20641 @cindex unload symbols from shared libraries
20642 Unload all shared object library symbols. This discards all symbols
20643 that have been loaded from all shared libraries. Symbols from shared
20644 libraries that were loaded by explicit user requests are not
20648 Sometimes you may wish that @value{GDBN} stops and gives you control
20649 when any of shared library events happen. The best way to do this is
20650 to use @code{catch load} and @code{catch unload} (@pxref{Set
20653 @value{GDBN} also supports the the @code{set stop-on-solib-events}
20654 command for this. This command exists for historical reasons. It is
20655 less useful than setting a catchpoint, because it does not allow for
20656 conditions or commands as a catchpoint does.
20659 @item set stop-on-solib-events
20660 @kindex set stop-on-solib-events
20661 This command controls whether @value{GDBN} should give you control
20662 when the dynamic linker notifies it about some shared library event.
20663 The most common event of interest is loading or unloading of a new
20666 @item show stop-on-solib-events
20667 @kindex show stop-on-solib-events
20668 Show whether @value{GDBN} stops and gives you control when shared
20669 library events happen.
20672 Shared libraries are also supported in many cross or remote debugging
20673 configurations. @value{GDBN} needs to have access to the target's libraries;
20674 this can be accomplished either by providing copies of the libraries
20675 on the host system, or by asking @value{GDBN} to automatically retrieve the
20676 libraries from the target. If copies of the target libraries are
20677 provided, they need to be the same as the target libraries, although the
20678 copies on the target can be stripped as long as the copies on the host are
20681 @cindex where to look for shared libraries
20682 For remote debugging, you need to tell @value{GDBN} where the target
20683 libraries are, so that it can load the correct copies---otherwise, it
20684 may try to load the host's libraries. @value{GDBN} has two variables
20685 to specify the search directories for target libraries.
20688 @cindex prefix for executable and shared library file names
20689 @cindex system root, alternate
20690 @kindex set solib-absolute-prefix
20691 @kindex set sysroot
20692 @item set sysroot @var{path}
20693 Use @var{path} as the system root for the program being debugged. Any
20694 absolute shared library paths will be prefixed with @var{path}; many
20695 runtime loaders store the absolute paths to the shared library in the
20696 target program's memory. When starting processes remotely, and when
20697 attaching to already-running processes (local or remote), their
20698 executable filenames will be prefixed with @var{path} if reported to
20699 @value{GDBN} as absolute by the operating system. If you use
20700 @code{set sysroot} to find executables and shared libraries, they need
20701 to be laid out in the same way that they are on the target, with
20702 e.g.@: a @file{/bin}, @file{/lib} and @file{/usr/lib} hierarchy under
20705 If @var{path} starts with the sequence @file{target:} and the target
20706 system is remote then @value{GDBN} will retrieve the target binaries
20707 from the remote system. This is only supported when using a remote
20708 target that supports the @code{remote get} command (@pxref{File
20709 Transfer,,Sending files to a remote system}). The part of @var{path}
20710 following the initial @file{target:} (if present) is used as system
20711 root prefix on the remote file system. If @var{path} starts with the
20712 sequence @file{remote:} this is converted to the sequence
20713 @file{target:} by @code{set sysroot}@footnote{Historically the
20714 functionality to retrieve binaries from the remote system was
20715 provided by prefixing @var{path} with @file{remote:}}. If you want
20716 to specify a local system root using a directory that happens to be
20717 named @file{target:} or @file{remote:}, you need to use some
20718 equivalent variant of the name like @file{./target:}.
20720 For targets with an MS-DOS based filesystem, such as MS-Windows and
20721 SymbianOS, @value{GDBN} tries prefixing a few variants of the target
20722 absolute file name with @var{path}. But first, on Unix hosts,
20723 @value{GDBN} converts all backslash directory separators into forward
20724 slashes, because the backslash is not a directory separator on Unix:
20727 c:\foo\bar.dll @result{} c:/foo/bar.dll
20730 Then, @value{GDBN} attempts prefixing the target file name with
20731 @var{path}, and looks for the resulting file name in the host file
20735 c:/foo/bar.dll @result{} /path/to/sysroot/c:/foo/bar.dll
20738 If that does not find the binary, @value{GDBN} tries removing
20739 the @samp{:} character from the drive spec, both for convenience, and,
20740 for the case of the host file system not supporting file names with
20744 c:/foo/bar.dll @result{} /path/to/sysroot/c/foo/bar.dll
20747 This makes it possible to have a system root that mirrors a target
20748 with more than one drive. E.g., you may want to setup your local
20749 copies of the target system shared libraries like so (note @samp{c} vs
20753 @file{/path/to/sysroot/c/sys/bin/foo.dll}
20754 @file{/path/to/sysroot/c/sys/bin/bar.dll}
20755 @file{/path/to/sysroot/z/sys/bin/bar.dll}
20759 and point the system root at @file{/path/to/sysroot}, so that
20760 @value{GDBN} can find the correct copies of both
20761 @file{c:\sys\bin\foo.dll}, and @file{z:\sys\bin\bar.dll}.
20763 If that still does not find the binary, @value{GDBN} tries
20764 removing the whole drive spec from the target file name:
20767 c:/foo/bar.dll @result{} /path/to/sysroot/foo/bar.dll
20770 This last lookup makes it possible to not care about the drive name,
20771 if you don't want or need to.
20773 The @code{set solib-absolute-prefix} command is an alias for @code{set
20776 @cindex default system root
20777 @cindex @samp{--with-sysroot}
20778 You can set the default system root by using the configure-time
20779 @samp{--with-sysroot} option. If the system root is inside
20780 @value{GDBN}'s configured binary prefix (set with @samp{--prefix} or
20781 @samp{--exec-prefix}), then the default system root will be updated
20782 automatically if the installed @value{GDBN} is moved to a new
20785 @kindex show sysroot
20787 Display the current executable and shared library prefix.
20789 @kindex set solib-search-path
20790 @item set solib-search-path @var{path}
20791 If this variable is set, @var{path} is a colon-separated list of
20792 directories to search for shared libraries. @samp{solib-search-path}
20793 is used after @samp{sysroot} fails to locate the library, or if the
20794 path to the library is relative instead of absolute. If you want to
20795 use @samp{solib-search-path} instead of @samp{sysroot}, be sure to set
20796 @samp{sysroot} to a nonexistent directory to prevent @value{GDBN} from
20797 finding your host's libraries. @samp{sysroot} is preferred; setting
20798 it to a nonexistent directory may interfere with automatic loading
20799 of shared library symbols.
20801 @kindex show solib-search-path
20802 @item show solib-search-path
20803 Display the current shared library search path.
20805 @cindex DOS file-name semantics of file names.
20806 @kindex set target-file-system-kind (unix|dos-based|auto)
20807 @kindex show target-file-system-kind
20808 @item set target-file-system-kind @var{kind}
20809 Set assumed file system kind for target reported file names.
20811 Shared library file names as reported by the target system may not
20812 make sense as is on the system @value{GDBN} is running on. For
20813 example, when remote debugging a target that has MS-DOS based file
20814 system semantics, from a Unix host, the target may be reporting to
20815 @value{GDBN} a list of loaded shared libraries with file names such as
20816 @file{c:\Windows\kernel32.dll}. On Unix hosts, there's no concept of
20817 drive letters, so the @samp{c:\} prefix is not normally understood as
20818 indicating an absolute file name, and neither is the backslash
20819 normally considered a directory separator character. In that case,
20820 the native file system would interpret this whole absolute file name
20821 as a relative file name with no directory components. This would make
20822 it impossible to point @value{GDBN} at a copy of the remote target's
20823 shared libraries on the host using @code{set sysroot}, and impractical
20824 with @code{set solib-search-path}. Setting
20825 @code{target-file-system-kind} to @code{dos-based} tells @value{GDBN}
20826 to interpret such file names similarly to how the target would, and to
20827 map them to file names valid on @value{GDBN}'s native file system
20828 semantics. The value of @var{kind} can be @code{"auto"}, in addition
20829 to one of the supported file system kinds. In that case, @value{GDBN}
20830 tries to determine the appropriate file system variant based on the
20831 current target's operating system (@pxref{ABI, ,Configuring the
20832 Current ABI}). The supported file system settings are:
20836 Instruct @value{GDBN} to assume the target file system is of Unix
20837 kind. Only file names starting the forward slash (@samp{/}) character
20838 are considered absolute, and the directory separator character is also
20842 Instruct @value{GDBN} to assume the target file system is DOS based.
20843 File names starting with either a forward slash, or a drive letter
20844 followed by a colon (e.g., @samp{c:}), are considered absolute, and
20845 both the slash (@samp{/}) and the backslash (@samp{\\}) characters are
20846 considered directory separators.
20849 Instruct @value{GDBN} to use the file system kind associated with the
20850 target operating system (@pxref{ABI, ,Configuring the Current ABI}).
20851 This is the default.
20855 @cindex file name canonicalization
20856 @cindex base name differences
20857 When processing file names provided by the user, @value{GDBN}
20858 frequently needs to compare them to the file names recorded in the
20859 program's debug info. Normally, @value{GDBN} compares just the
20860 @dfn{base names} of the files as strings, which is reasonably fast
20861 even for very large programs. (The base name of a file is the last
20862 portion of its name, after stripping all the leading directories.)
20863 This shortcut in comparison is based upon the assumption that files
20864 cannot have more than one base name. This is usually true, but
20865 references to files that use symlinks or similar filesystem
20866 facilities violate that assumption. If your program records files
20867 using such facilities, or if you provide file names to @value{GDBN}
20868 using symlinks etc., you can set @code{basenames-may-differ} to
20869 @code{true} to instruct @value{GDBN} to completely canonicalize each
20870 pair of file names it needs to compare. This will make file-name
20871 comparisons accurate, but at a price of a significant slowdown.
20874 @item set basenames-may-differ
20875 @kindex set basenames-may-differ
20876 Set whether a source file may have multiple base names.
20878 @item show basenames-may-differ
20879 @kindex show basenames-may-differ
20880 Show whether a source file may have multiple base names.
20884 @section File Caching
20885 @cindex caching of opened files
20886 @cindex caching of bfd objects
20888 To speed up file loading, and reduce memory usage, @value{GDBN} will
20889 reuse the @code{bfd} objects used to track open files. @xref{Top, ,
20890 BFD, bfd, The Binary File Descriptor Library}. The following commands
20891 allow visibility and control of the caching behavior.
20894 @kindex maint info bfds
20895 @item maint info bfds
20896 This prints information about each @code{bfd} object that is known to
20899 @kindex maint set bfd-sharing
20900 @kindex maint show bfd-sharing
20901 @kindex bfd caching
20902 @item maint set bfd-sharing
20903 @item maint show bfd-sharing
20904 Control whether @code{bfd} objects can be shared. When sharing is
20905 enabled @value{GDBN} reuses already open @code{bfd} objects rather
20906 than reopening the same file. Turning sharing off does not cause
20907 already shared @code{bfd} objects to be unshared, but all future files
20908 that are opened will create a new @code{bfd} object. Similarly,
20909 re-enabling sharing does not cause multiple existing @code{bfd}
20910 objects to be collapsed into a single shared @code{bfd} object.
20912 @kindex set debug bfd-cache @var{level}
20913 @kindex bfd caching
20914 @item set debug bfd-cache @var{level}
20915 Turns on debugging of the bfd cache, setting the level to @var{level}.
20917 @kindex show debug bfd-cache
20918 @kindex bfd caching
20919 @item show debug bfd-cache
20920 Show the current debugging level of the bfd cache.
20923 @node Separate Debug Files
20924 @section Debugging Information in Separate Files
20925 @cindex separate debugging information files
20926 @cindex debugging information in separate files
20927 @cindex @file{.debug} subdirectories
20928 @cindex debugging information directory, global
20929 @cindex global debugging information directories
20930 @cindex build ID, and separate debugging files
20931 @cindex @file{.build-id} directory
20933 @value{GDBN} allows you to put a program's debugging information in a
20934 file separate from the executable itself, in a way that allows
20935 @value{GDBN} to find and load the debugging information automatically.
20936 Since debugging information can be very large---sometimes larger
20937 than the executable code itself---some systems distribute debugging
20938 information for their executables in separate files, which users can
20939 install only when they need to debug a problem.
20941 @value{GDBN} supports two ways of specifying the separate debug info
20946 The executable contains a @dfn{debug link} that specifies the name of
20947 the separate debug info file. The separate debug file's name is
20948 usually @file{@var{executable}.debug}, where @var{executable} is the
20949 name of the corresponding executable file without leading directories
20950 (e.g., @file{ls.debug} for @file{/usr/bin/ls}). In addition, the
20951 debug link specifies a 32-bit @dfn{Cyclic Redundancy Check} (CRC)
20952 checksum for the debug file, which @value{GDBN} uses to validate that
20953 the executable and the debug file came from the same build.
20956 The executable contains a @dfn{build ID}, a unique bit string that is
20957 also present in the corresponding debug info file. (This is supported
20958 only on some operating systems, when using the ELF or PE file formats
20959 for binary files and the @sc{gnu} Binutils.) For more details about
20960 this feature, see the description of the @option{--build-id}
20961 command-line option in @ref{Options, , Command Line Options, ld,
20962 The GNU Linker}. The debug info file's name is not specified
20963 explicitly by the build ID, but can be computed from the build ID, see
20967 Depending on the way the debug info file is specified, @value{GDBN}
20968 uses two different methods of looking for the debug file:
20972 For the ``debug link'' method, @value{GDBN} looks up the named file in
20973 the directory of the executable file, then in a subdirectory of that
20974 directory named @file{.debug}, and finally under each one of the
20975 global debug directories, in a subdirectory whose name is identical to
20976 the leading directories of the executable's absolute file name. (On
20977 MS-Windows/MS-DOS, the drive letter of the executable's leading
20978 directories is converted to a one-letter subdirectory, i.e.@:
20979 @file{d:/usr/bin/} is converted to @file{/d/usr/bin/}, because Windows
20980 filesystems disallow colons in file names.)
20983 For the ``build ID'' method, @value{GDBN} looks in the
20984 @file{.build-id} subdirectory of each one of the global debug directories for
20985 a file named @file{@var{nn}/@var{nnnnnnnn}.debug}, where @var{nn} are the
20986 first 2 hex characters of the build ID bit string, and @var{nnnnnnnn}
20987 are the rest of the bit string. (Real build ID strings are 32 or more
20988 hex characters, not 10.)
20991 So, for example, suppose you ask @value{GDBN} to debug
20992 @file{/usr/bin/ls}, which has a debug link that specifies the
20993 file @file{ls.debug}, and a build ID whose value in hex is
20994 @code{abcdef1234}. If the list of the global debug directories includes
20995 @file{/usr/lib/debug}, then @value{GDBN} will look for the following
20996 debug information files, in the indicated order:
21000 @file{/usr/lib/debug/.build-id/ab/cdef1234.debug}
21002 @file{/usr/bin/ls.debug}
21004 @file{/usr/bin/.debug/ls.debug}
21006 @file{/usr/lib/debug/usr/bin/ls.debug}.
21009 @anchor{debug-file-directory}
21010 Global debugging info directories default to what is set by @value{GDBN}
21011 configure option @option{--with-separate-debug-dir}. During @value{GDBN} run
21012 you can also set the global debugging info directories, and view the list
21013 @value{GDBN} is currently using.
21017 @kindex set debug-file-directory
21018 @item set debug-file-directory @var{directories}
21019 Set the directories which @value{GDBN} searches for separate debugging
21020 information files to @var{directory}. Multiple path components can be set
21021 concatenating them by a path separator.
21023 @kindex show debug-file-directory
21024 @item show debug-file-directory
21025 Show the directories @value{GDBN} searches for separate debugging
21030 @cindex @code{.gnu_debuglink} sections
21031 @cindex debug link sections
21032 A debug link is a special section of the executable file named
21033 @code{.gnu_debuglink}. The section must contain:
21037 A filename, with any leading directory components removed, followed by
21040 zero to three bytes of padding, as needed to reach the next four-byte
21041 boundary within the section, and
21043 a four-byte CRC checksum, stored in the same endianness used for the
21044 executable file itself. The checksum is computed on the debugging
21045 information file's full contents by the function given below, passing
21046 zero as the @var{crc} argument.
21049 Any executable file format can carry a debug link, as long as it can
21050 contain a section named @code{.gnu_debuglink} with the contents
21053 @cindex @code{.note.gnu.build-id} sections
21054 @cindex build ID sections
21055 The build ID is a special section in the executable file (and in other
21056 ELF binary files that @value{GDBN} may consider). This section is
21057 often named @code{.note.gnu.build-id}, but that name is not mandatory.
21058 It contains unique identification for the built files---the ID remains
21059 the same across multiple builds of the same build tree. The default
21060 algorithm SHA1 produces 160 bits (40 hexadecimal characters) of the
21061 content for the build ID string. The same section with an identical
21062 value is present in the original built binary with symbols, in its
21063 stripped variant, and in the separate debugging information file.
21065 The debugging information file itself should be an ordinary
21066 executable, containing a full set of linker symbols, sections, and
21067 debugging information. The sections of the debugging information file
21068 should have the same names, addresses, and sizes as the original file,
21069 but they need not contain any data---much like a @code{.bss} section
21070 in an ordinary executable.
21072 The @sc{gnu} binary utilities (Binutils) package includes the
21073 @samp{objcopy} utility that can produce
21074 the separated executable / debugging information file pairs using the
21075 following commands:
21078 @kbd{objcopy --only-keep-debug foo foo.debug}
21083 These commands remove the debugging
21084 information from the executable file @file{foo} and place it in the file
21085 @file{foo.debug}. You can use the first, second or both methods to link the
21090 The debug link method needs the following additional command to also leave
21091 behind a debug link in @file{foo}:
21094 @kbd{objcopy --add-gnu-debuglink=foo.debug foo}
21097 Ulrich Drepper's @file{elfutils} package, starting with version 0.53, contains
21098 a version of the @code{strip} command such that the command @kbd{strip foo -f
21099 foo.debug} has the same functionality as the two @code{objcopy} commands and
21100 the @code{ln -s} command above, together.
21103 Build ID gets embedded into the main executable using @code{ld --build-id} or
21104 the @value{NGCC} counterpart @code{gcc -Wl,--build-id}. Build ID support plus
21105 compatibility fixes for debug files separation are present in @sc{gnu} binary
21106 utilities (Binutils) package since version 2.18.
21111 @cindex CRC algorithm definition
21112 The CRC used in @code{.gnu_debuglink} is the CRC-32 defined in
21113 IEEE 802.3 using the polynomial:
21115 @c TexInfo requires naked braces for multi-digit exponents for Tex
21116 @c output, but this causes HTML output to barf. HTML has to be set using
21117 @c raw commands. So we end up having to specify this equation in 2
21122 <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>
21123 + <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
21129 @math{x^{32} + x^{26} + x^{23} + x^{22} + x^{16} + x^{12} + x^{11}}
21130 @math{+ x^{10} + x^8 + x^7 + x^5 + x^4 + x^2 + x + 1}
21134 The function is computed byte at a time, taking the least
21135 significant bit of each byte first. The initial pattern
21136 @code{0xffffffff} is used, to ensure leading zeros affect the CRC and
21137 the final result is inverted to ensure trailing zeros also affect the
21140 @emph{Note:} This is the same CRC polynomial as used in handling the
21141 @dfn{Remote Serial Protocol} @code{qCRC} packet (@pxref{qCRC packet}).
21142 However in the case of the Remote Serial Protocol, the CRC is computed
21143 @emph{most} significant bit first, and the result is not inverted, so
21144 trailing zeros have no effect on the CRC value.
21146 To complete the description, we show below the code of the function
21147 which produces the CRC used in @code{.gnu_debuglink}. Inverting the
21148 initially supplied @code{crc} argument means that an initial call to
21149 this function passing in zero will start computing the CRC using
21152 @kindex gnu_debuglink_crc32
21155 gnu_debuglink_crc32 (unsigned long crc,
21156 unsigned char *buf, size_t len)
21158 static const unsigned long crc32_table[256] =
21160 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419,
21161 0x706af48f, 0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4,
21162 0xe0d5e91e, 0x97d2d988, 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07,
21163 0x90bf1d91, 0x1db71064, 0x6ab020f2, 0xf3b97148, 0x84be41de,
21164 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7, 0x136c9856,
21165 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9,
21166 0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4,
21167 0xa2677172, 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b,
21168 0x35b5a8fa, 0x42b2986c, 0xdbbbc9d6, 0xacbcf940, 0x32d86ce3,
21169 0x45df5c75, 0xdcd60dcf, 0xabd13d59, 0x26d930ac, 0x51de003a,
21170 0xc8d75180, 0xbfd06116, 0x21b4f4b5, 0x56b3c423, 0xcfba9599,
21171 0xb8bda50f, 0x2802b89e, 0x5f058808, 0xc60cd9b2, 0xb10be924,
21172 0x2f6f7c87, 0x58684c11, 0xc1611dab, 0xb6662d3d, 0x76dc4190,
21173 0x01db7106, 0x98d220bc, 0xefd5102a, 0x71b18589, 0x06b6b51f,
21174 0x9fbfe4a5, 0xe8b8d433, 0x7807c9a2, 0x0f00f934, 0x9609a88e,
21175 0xe10e9818, 0x7f6a0dbb, 0x086d3d2d, 0x91646c97, 0xe6635c01,
21176 0x6b6b51f4, 0x1c6c6162, 0x856530d8, 0xf262004e, 0x6c0695ed,
21177 0x1b01a57b, 0x8208f4c1, 0xf50fc457, 0x65b0d9c6, 0x12b7e950,
21178 0x8bbeb8ea, 0xfcb9887c, 0x62dd1ddf, 0x15da2d49, 0x8cd37cf3,
21179 0xfbd44c65, 0x4db26158, 0x3ab551ce, 0xa3bc0074, 0xd4bb30e2,
21180 0x4adfa541, 0x3dd895d7, 0xa4d1c46d, 0xd3d6f4fb, 0x4369e96a,
21181 0x346ed9fc, 0xad678846, 0xda60b8d0, 0x44042d73, 0x33031de5,
21182 0xaa0a4c5f, 0xdd0d7cc9, 0x5005713c, 0x270241aa, 0xbe0b1010,
21183 0xc90c2086, 0x5768b525, 0x206f85b3, 0xb966d409, 0xce61e49f,
21184 0x5edef90e, 0x29d9c998, 0xb0d09822, 0xc7d7a8b4, 0x59b33d17,
21185 0x2eb40d81, 0xb7bd5c3b, 0xc0ba6cad, 0xedb88320, 0x9abfb3b6,
21186 0x03b6e20c, 0x74b1d29a, 0xead54739, 0x9dd277af, 0x04db2615,
21187 0x73dc1683, 0xe3630b12, 0x94643b84, 0x0d6d6a3e, 0x7a6a5aa8,
21188 0xe40ecf0b, 0x9309ff9d, 0x0a00ae27, 0x7d079eb1, 0xf00f9344,
21189 0x8708a3d2, 0x1e01f268, 0x6906c2fe, 0xf762575d, 0x806567cb,
21190 0x196c3671, 0x6e6b06e7, 0xfed41b76, 0x89d32be0, 0x10da7a5a,
21191 0x67dd4acc, 0xf9b9df6f, 0x8ebeeff9, 0x17b7be43, 0x60b08ed5,
21192 0xd6d6a3e8, 0xa1d1937e, 0x38d8c2c4, 0x4fdff252, 0xd1bb67f1,
21193 0xa6bc5767, 0x3fb506dd, 0x48b2364b, 0xd80d2bda, 0xaf0a1b4c,
21194 0x36034af6, 0x41047a60, 0xdf60efc3, 0xa867df55, 0x316e8eef,
21195 0x4669be79, 0xcb61b38c, 0xbc66831a, 0x256fd2a0, 0x5268e236,
21196 0xcc0c7795, 0xbb0b4703, 0x220216b9, 0x5505262f, 0xc5ba3bbe,
21197 0xb2bd0b28, 0x2bb45a92, 0x5cb36a04, 0xc2d7ffa7, 0xb5d0cf31,
21198 0x2cd99e8b, 0x5bdeae1d, 0x9b64c2b0, 0xec63f226, 0x756aa39c,
21199 0x026d930a, 0x9c0906a9, 0xeb0e363f, 0x72076785, 0x05005713,
21200 0x95bf4a82, 0xe2b87a14, 0x7bb12bae, 0x0cb61b38, 0x92d28e9b,
21201 0xe5d5be0d, 0x7cdcefb7, 0x0bdbdf21, 0x86d3d2d4, 0xf1d4e242,
21202 0x68ddb3f8, 0x1fda836e, 0x81be16cd, 0xf6b9265b, 0x6fb077e1,
21203 0x18b74777, 0x88085ae6, 0xff0f6a70, 0x66063bca, 0x11010b5c,
21204 0x8f659eff, 0xf862ae69, 0x616bffd3, 0x166ccf45, 0xa00ae278,
21205 0xd70dd2ee, 0x4e048354, 0x3903b3c2, 0xa7672661, 0xd06016f7,
21206 0x4969474d, 0x3e6e77db, 0xaed16a4a, 0xd9d65adc, 0x40df0b66,
21207 0x37d83bf0, 0xa9bcae53, 0xdebb9ec5, 0x47b2cf7f, 0x30b5ffe9,
21208 0xbdbdf21c, 0xcabac28a, 0x53b39330, 0x24b4a3a6, 0xbad03605,
21209 0xcdd70693, 0x54de5729, 0x23d967bf, 0xb3667a2e, 0xc4614ab8,
21210 0x5d681b02, 0x2a6f2b94, 0xb40bbe37, 0xc30c8ea1, 0x5a05df1b,
21213 unsigned char *end;
21215 crc = ~crc & 0xffffffff;
21216 for (end = buf + len; buf < end; ++buf)
21217 crc = crc32_table[(crc ^ *buf) & 0xff] ^ (crc >> 8);
21218 return ~crc & 0xffffffff;
21223 This computation does not apply to the ``build ID'' method.
21225 @node MiniDebugInfo
21226 @section Debugging information in a special section
21227 @cindex separate debug sections
21228 @cindex @samp{.gnu_debugdata} section
21230 Some systems ship pre-built executables and libraries that have a
21231 special @samp{.gnu_debugdata} section. This feature is called
21232 @dfn{MiniDebugInfo}. This section holds an LZMA-compressed object and
21233 is used to supply extra symbols for backtraces.
21235 The intent of this section is to provide extra minimal debugging
21236 information for use in simple backtraces. It is not intended to be a
21237 replacement for full separate debugging information (@pxref{Separate
21238 Debug Files}). The example below shows the intended use; however,
21239 @value{GDBN} does not currently put restrictions on what sort of
21240 debugging information might be included in the section.
21242 @value{GDBN} has support for this extension. If the section exists,
21243 then it is used provided that no other source of debugging information
21244 can be found, and that @value{GDBN} was configured with LZMA support.
21246 This section can be easily created using @command{objcopy} and other
21247 standard utilities:
21250 # Extract the dynamic symbols from the main binary, there is no need
21251 # to also have these in the normal symbol table.
21252 nm -D @var{binary} --format=posix --defined-only \
21253 | awk '@{ print $1 @}' | sort > dynsyms
21255 # Extract all the text (i.e. function) symbols from the debuginfo.
21256 # (Note that we actually also accept "D" symbols, for the benefit
21257 # of platforms like PowerPC64 that use function descriptors.)
21258 nm @var{binary} --format=posix --defined-only \
21259 | awk '@{ if ($2 == "T" || $2 == "t" || $2 == "D") print $1 @}' \
21262 # Keep all the function symbols not already in the dynamic symbol
21264 comm -13 dynsyms funcsyms > keep_symbols
21266 # Separate full debug info into debug binary.
21267 objcopy --only-keep-debug @var{binary} debug
21269 # Copy the full debuginfo, keeping only a minimal set of symbols and
21270 # removing some unnecessary sections.
21271 objcopy -S --remove-section .gdb_index --remove-section .comment \
21272 --keep-symbols=keep_symbols debug mini_debuginfo
21274 # Drop the full debug info from the original binary.
21275 strip --strip-all -R .comment @var{binary}
21277 # Inject the compressed data into the .gnu_debugdata section of the
21280 objcopy --add-section .gnu_debugdata=mini_debuginfo.xz @var{binary}
21284 @section Index Files Speed Up @value{GDBN}
21285 @cindex index files
21286 @cindex @samp{.gdb_index} section
21288 When @value{GDBN} finds a symbol file, it scans the symbols in the
21289 file in order to construct an internal symbol table. This lets most
21290 @value{GDBN} operations work quickly---at the cost of a delay early
21291 on. For large programs, this delay can be quite lengthy, so
21292 @value{GDBN} provides a way to build an index, which speeds up
21295 For convenience, @value{GDBN} comes with a program,
21296 @command{gdb-add-index}, which can be used to add the index to a
21297 symbol file. It takes the symbol file as its only argument:
21300 $ gdb-add-index symfile
21303 @xref{gdb-add-index}.
21305 It is also possible to do the work manually. Here is what
21306 @command{gdb-add-index} does behind the curtains.
21308 The index is stored as a section in the symbol file. @value{GDBN} can
21309 write the index to a file, then you can put it into the symbol file
21310 using @command{objcopy}.
21312 To create an index file, use the @code{save gdb-index} command:
21315 @item save gdb-index [-dwarf-5] @var{directory}
21316 @kindex save gdb-index
21317 Create index files for all symbol files currently known by
21318 @value{GDBN}. For each known @var{symbol-file}, this command by
21319 default creates it produces a single file
21320 @file{@var{symbol-file}.gdb-index}. If you invoke this command with
21321 the @option{-dwarf-5} option, it produces 2 files:
21322 @file{@var{symbol-file}.debug_names} and
21323 @file{@var{symbol-file}.debug_str}. The files are created in the
21324 given @var{directory}.
21327 Once you have created an index file you can merge it into your symbol
21328 file, here named @file{symfile}, using @command{objcopy}:
21331 $ objcopy --add-section .gdb_index=symfile.gdb-index \
21332 --set-section-flags .gdb_index=readonly symfile symfile
21335 Or for @code{-dwarf-5}:
21338 $ objcopy --dump-section .debug_str=symfile.debug_str.new symfile
21339 $ cat symfile.debug_str >>symfile.debug_str.new
21340 $ objcopy --add-section .debug_names=symfile.gdb-index \
21341 --set-section-flags .debug_names=readonly \
21342 --update-section .debug_str=symfile.debug_str.new symfile symfile
21345 @value{GDBN} will normally ignore older versions of @file{.gdb_index}
21346 sections that have been deprecated. Usually they are deprecated because
21347 they are missing a new feature or have performance issues.
21348 To tell @value{GDBN} to use a deprecated index section anyway
21349 specify @code{set use-deprecated-index-sections on}.
21350 The default is @code{off}.
21351 This can speed up startup, but may result in some functionality being lost.
21352 @xref{Index Section Format}.
21354 @emph{Warning:} Setting @code{use-deprecated-index-sections} to @code{on}
21355 must be done before gdb reads the file. The following will not work:
21358 $ gdb -ex "set use-deprecated-index-sections on" <program>
21361 Instead you must do, for example,
21364 $ gdb -iex "set use-deprecated-index-sections on" <program>
21367 There are currently some limitation on indices. They only work when
21368 using DWARF debugging information, not stabs. And, only the
21369 @code{-dwarf-5} index works for programs using Ada.
21371 @subsection Automatic symbol index cache
21373 @cindex automatic symbol index cache
21374 It is possible for @value{GDBN} to automatically save a copy of this index in a
21375 cache on disk and retrieve it from there when loading the same binary in the
21376 future. This feature can be turned on with @kbd{set index-cache on}. The
21377 following commands can be used to tweak the behavior of the index cache.
21381 @kindex set index-cache
21382 @item set index-cache on
21383 @itemx set index-cache off
21384 Enable or disable the use of the symbol index cache.
21386 @item set index-cache directory @var{directory}
21387 @kindex show index-cache
21388 @itemx show index-cache directory
21389 Set/show the directory where index files will be saved.
21391 The default value for this directory depends on the host platform. On
21392 most systems, the index is cached in the @file{gdb} subdirectory of
21393 the directory pointed to by the @env{XDG_CACHE_HOME} environment
21394 variable, if it is defined, else in the @file{.cache/gdb} subdirectory
21395 of your home directory. However, on some systems, the default may
21396 differ according to local convention.
21398 There is no limit on the disk space used by index cache. It is perfectly safe
21399 to delete the content of that directory to free up disk space.
21401 @item show index-cache stats
21402 Print the number of cache hits and misses since the launch of @value{GDBN}.
21406 @node Symbol Errors
21407 @section Errors Reading Symbol Files
21409 While reading a symbol file, @value{GDBN} occasionally encounters problems,
21410 such as symbol types it does not recognize, or known bugs in compiler
21411 output. By default, @value{GDBN} does not notify you of such problems, since
21412 they are relatively common and primarily of interest to people
21413 debugging compilers. If you are interested in seeing information
21414 about ill-constructed symbol tables, you can either ask @value{GDBN} to print
21415 only one message about each such type of problem, no matter how many
21416 times the problem occurs; or you can ask @value{GDBN} to print more messages,
21417 to see how many times the problems occur, with the @code{set
21418 complaints} command (@pxref{Messages/Warnings, ,Optional Warnings and
21421 The messages currently printed, and their meanings, include:
21424 @item inner block not inside outer block in @var{symbol}
21426 The symbol information shows where symbol scopes begin and end
21427 (such as at the start of a function or a block of statements). This
21428 error indicates that an inner scope block is not fully contained
21429 in its outer scope blocks.
21431 @value{GDBN} circumvents the problem by treating the inner block as if it had
21432 the same scope as the outer block. In the error message, @var{symbol}
21433 may be shown as ``@code{(don't know)}'' if the outer block is not a
21436 @item block at @var{address} out of order
21438 The symbol information for symbol scope blocks should occur in
21439 order of increasing addresses. This error indicates that it does not
21442 @value{GDBN} does not circumvent this problem, and has trouble
21443 locating symbols in the source file whose symbols it is reading. (You
21444 can often determine what source file is affected by specifying
21445 @code{set verbose on}. @xref{Messages/Warnings, ,Optional Warnings and
21448 @item bad block start address patched
21450 The symbol information for a symbol scope block has a start address
21451 smaller than the address of the preceding source line. This is known
21452 to occur in the SunOS 4.1.1 (and earlier) C compiler.
21454 @value{GDBN} circumvents the problem by treating the symbol scope block as
21455 starting on the previous source line.
21457 @item bad string table offset in symbol @var{n}
21460 Symbol number @var{n} contains a pointer into the string table which is
21461 larger than the size of the string table.
21463 @value{GDBN} circumvents the problem by considering the symbol to have the
21464 name @code{foo}, which may cause other problems if many symbols end up
21467 @item unknown symbol type @code{0x@var{nn}}
21469 The symbol information contains new data types that @value{GDBN} does
21470 not yet know how to read. @code{0x@var{nn}} is the symbol type of the
21471 uncomprehended information, in hexadecimal.
21473 @value{GDBN} circumvents the error by ignoring this symbol information.
21474 This usually allows you to debug your program, though certain symbols
21475 are not accessible. If you encounter such a problem and feel like
21476 debugging it, you can debug @code{@value{GDBP}} with itself, breakpoint
21477 on @code{complain}, then go up to the function @code{read_dbx_symtab}
21478 and examine @code{*bufp} to see the symbol.
21480 @item stub type has NULL name
21482 @value{GDBN} could not find the full definition for a struct or class.
21484 @item const/volatile indicator missing (ok if using g++ v1.x), got@dots{}
21485 The symbol information for a C@t{++} member function is missing some
21486 information that recent versions of the compiler should have output for
21489 @item info mismatch between compiler and debugger
21491 @value{GDBN} could not parse a type specification output by the compiler.
21496 @section @value{GDBN} Data Files
21498 @cindex prefix for data files
21499 @value{GDBN} will sometimes read an auxiliary data file. These files
21500 are kept in a directory known as the @dfn{data directory}.
21502 You can set the data directory's name, and view the name @value{GDBN}
21503 is currently using.
21506 @kindex set data-directory
21507 @item set data-directory @var{directory}
21508 Set the directory which @value{GDBN} searches for auxiliary data files
21509 to @var{directory}.
21511 @kindex show data-directory
21512 @item show data-directory
21513 Show the directory @value{GDBN} searches for auxiliary data files.
21516 @cindex default data directory
21517 @cindex @samp{--with-gdb-datadir}
21518 You can set the default data directory by using the configure-time
21519 @samp{--with-gdb-datadir} option. If the data directory is inside
21520 @value{GDBN}'s configured binary prefix (set with @samp{--prefix} or
21521 @samp{--exec-prefix}), then the default data directory will be updated
21522 automatically if the installed @value{GDBN} is moved to a new
21525 The data directory may also be specified with the
21526 @code{--data-directory} command line option.
21527 @xref{Mode Options}.
21530 @chapter Specifying a Debugging Target
21532 @cindex debugging target
21533 A @dfn{target} is the execution environment occupied by your program.
21535 Often, @value{GDBN} runs in the same host environment as your program;
21536 in that case, the debugging target is specified as a side effect when
21537 you use the @code{file} or @code{core} commands. When you need more
21538 flexibility---for example, running @value{GDBN} on a physically separate
21539 host, or controlling a standalone system over a serial port or a
21540 realtime system over a TCP/IP connection---you can use the @code{target}
21541 command to specify one of the target types configured for @value{GDBN}
21542 (@pxref{Target Commands, ,Commands for Managing Targets}).
21544 @cindex target architecture
21545 It is possible to build @value{GDBN} for several different @dfn{target
21546 architectures}. When @value{GDBN} is built like that, you can choose
21547 one of the available architectures with the @kbd{set architecture}
21551 @kindex set architecture
21552 @kindex show architecture
21553 @item set architecture @var{arch}
21554 This command sets the current target architecture to @var{arch}. The
21555 value of @var{arch} can be @code{"auto"}, in addition to one of the
21556 supported architectures.
21558 @item show architecture
21559 Show the current target architecture.
21561 @item set processor
21563 @kindex set processor
21564 @kindex show processor
21565 These are alias commands for, respectively, @code{set architecture}
21566 and @code{show architecture}.
21570 * Active Targets:: Active targets
21571 * Target Commands:: Commands for managing targets
21572 * Byte Order:: Choosing target byte order
21575 @node Active Targets
21576 @section Active Targets
21578 @cindex stacking targets
21579 @cindex active targets
21580 @cindex multiple targets
21582 There are multiple classes of targets such as: processes, executable files or
21583 recording sessions. Core files belong to the process class, making core file
21584 and process mutually exclusive. Otherwise, @value{GDBN} can work concurrently
21585 on multiple active targets, one in each class. This allows you to (for
21586 example) start a process and inspect its activity, while still having access to
21587 the executable file after the process finishes. Or if you start process
21588 recording (@pxref{Reverse Execution}) and @code{reverse-step} there, you are
21589 presented a virtual layer of the recording target, while the process target
21590 remains stopped at the chronologically last point of the process execution.
21592 Use the @code{core-file} and @code{exec-file} commands to select a new core
21593 file or executable target (@pxref{Files, ,Commands to Specify Files}). To
21594 specify as a target a process that is already running, use the @code{attach}
21595 command (@pxref{Attach, ,Debugging an Already-running Process}).
21597 @node Target Commands
21598 @section Commands for Managing Targets
21601 @item target @var{type} @var{parameters}
21602 Connects the @value{GDBN} host environment to a target machine or
21603 process. A target is typically a protocol for talking to debugging
21604 facilities. You use the argument @var{type} to specify the type or
21605 protocol of the target machine.
21607 Further @var{parameters} are interpreted by the target protocol, but
21608 typically include things like device names or host names to connect
21609 with, process numbers, and baud rates.
21611 The @code{target} command does not repeat if you press @key{RET} again
21612 after executing the command.
21614 @kindex help target
21616 Displays the names of all targets available. To display targets
21617 currently selected, use either @code{info target} or @code{info files}
21618 (@pxref{Files, ,Commands to Specify Files}).
21620 @item help target @var{name}
21621 Describe a particular target, including any parameters necessary to
21624 @kindex set gnutarget
21625 @item set gnutarget @var{args}
21626 @value{GDBN} uses its own library BFD to read your files. @value{GDBN}
21627 knows whether it is reading an @dfn{executable},
21628 a @dfn{core}, or a @dfn{.o} file; however, you can specify the file format
21629 with the @code{set gnutarget} command. Unlike most @code{target} commands,
21630 with @code{gnutarget} the @code{target} refers to a program, not a machine.
21633 @emph{Warning:} To specify a file format with @code{set gnutarget},
21634 you must know the actual BFD name.
21638 @xref{Files, , Commands to Specify Files}.
21640 @kindex show gnutarget
21641 @item show gnutarget
21642 Use the @code{show gnutarget} command to display what file format
21643 @code{gnutarget} is set to read. If you have not set @code{gnutarget},
21644 @value{GDBN} will determine the file format for each file automatically,
21645 and @code{show gnutarget} displays @samp{The current BFD target is "auto"}.
21648 @cindex common targets
21649 Here are some common targets (available, or not, depending on the GDB
21654 @item target exec @var{program}
21655 @cindex executable file target
21656 An executable file. @samp{target exec @var{program}} is the same as
21657 @samp{exec-file @var{program}}.
21659 @item target core @var{filename}
21660 @cindex core dump file target
21661 A core dump file. @samp{target core @var{filename}} is the same as
21662 @samp{core-file @var{filename}}.
21664 @item target remote @var{medium}
21665 @cindex remote target
21666 A remote system connected to @value{GDBN} via a serial line or network
21667 connection. This command tells @value{GDBN} to use its own remote
21668 protocol over @var{medium} for debugging. @xref{Remote Debugging}.
21670 For example, if you have a board connected to @file{/dev/ttya} on the
21671 machine running @value{GDBN}, you could say:
21674 target remote /dev/ttya
21677 @code{target remote} supports the @code{load} command. This is only
21678 useful if you have some other way of getting the stub to the target
21679 system, and you can put it somewhere in memory where it won't get
21680 clobbered by the download.
21682 @item target sim @r{[}@var{simargs}@r{]} @dots{}
21683 @cindex built-in simulator target
21684 Builtin CPU simulator. @value{GDBN} includes simulators for most architectures.
21692 works; however, you cannot assume that a specific memory map, device
21693 drivers, or even basic I/O is available, although some simulators do
21694 provide these. For info about any processor-specific simulator details,
21695 see the appropriate section in @ref{Embedded Processors, ,Embedded
21698 @item target native
21699 @cindex native target
21700 Setup for local/native process debugging. Useful to make the
21701 @code{run} command spawn native processes (likewise @code{attach},
21702 etc.@:) even when @code{set auto-connect-native-target} is @code{off}
21703 (@pxref{set auto-connect-native-target}).
21707 Different targets are available on different configurations of @value{GDBN};
21708 your configuration may have more or fewer targets.
21710 Many remote targets require you to download the executable's code once
21711 you've successfully established a connection. You may wish to control
21712 various aspects of this process.
21717 @kindex set hash@r{, for remote monitors}
21718 @cindex hash mark while downloading
21719 This command controls whether a hash mark @samp{#} is displayed while
21720 downloading a file to the remote monitor. If on, a hash mark is
21721 displayed after each S-record is successfully downloaded to the
21725 @kindex show hash@r{, for remote monitors}
21726 Show the current status of displaying the hash mark.
21728 @item set debug monitor
21729 @kindex set debug monitor
21730 @cindex display remote monitor communications
21731 Enable or disable display of communications messages between
21732 @value{GDBN} and the remote monitor.
21734 @item show debug monitor
21735 @kindex show debug monitor
21736 Show the current status of displaying communications between
21737 @value{GDBN} and the remote monitor.
21742 @kindex load @var{filename} @var{offset}
21743 @item load @var{filename} @var{offset}
21745 Depending on what remote debugging facilities are configured into
21746 @value{GDBN}, the @code{load} command may be available. Where it exists, it
21747 is meant to make @var{filename} (an executable) available for debugging
21748 on the remote system---by downloading, or dynamic linking, for example.
21749 @code{load} also records the @var{filename} symbol table in @value{GDBN}, like
21750 the @code{add-symbol-file} command.
21752 If your @value{GDBN} does not have a @code{load} command, attempting to
21753 execute it gets the error message ``@code{You can't do that when your
21754 target is @dots{}}''
21756 The file is loaded at whatever address is specified in the executable.
21757 For some object file formats, you can specify the load address when you
21758 link the program; for other formats, like a.out, the object file format
21759 specifies a fixed address.
21760 @c FIXME! This would be a good place for an xref to the GNU linker doc.
21762 It is also possible to tell @value{GDBN} to load the executable file at a
21763 specific offset described by the optional argument @var{offset}. When
21764 @var{offset} is provided, @var{filename} must also be provided.
21766 Depending on the remote side capabilities, @value{GDBN} may be able to
21767 load programs into flash memory.
21769 @code{load} does not repeat if you press @key{RET} again after using it.
21774 @kindex flash-erase
21776 @anchor{flash-erase}
21778 Erases all known flash memory regions on the target.
21783 @section Choosing Target Byte Order
21785 @cindex choosing target byte order
21786 @cindex target byte order
21788 Some types of processors, such as the @acronym{MIPS}, PowerPC, and Renesas SH,
21789 offer the ability to run either big-endian or little-endian byte
21790 orders. Usually the executable or symbol will include a bit to
21791 designate the endian-ness, and you will not need to worry about
21792 which to use. However, you may still find it useful to adjust
21793 @value{GDBN}'s idea of processor endian-ness manually.
21797 @item set endian big
21798 Instruct @value{GDBN} to assume the target is big-endian.
21800 @item set endian little
21801 Instruct @value{GDBN} to assume the target is little-endian.
21803 @item set endian auto
21804 Instruct @value{GDBN} to use the byte order associated with the
21808 Display @value{GDBN}'s current idea of the target byte order.
21812 If the @code{set endian auto} mode is in effect and no executable has
21813 been selected, then the endianness used is the last one chosen either
21814 by one of the @code{set endian big} and @code{set endian little}
21815 commands or by inferring from the last executable used. If no
21816 endianness has been previously chosen, then the default for this mode
21817 is inferred from the target @value{GDBN} has been built for, and is
21818 @code{little} if the name of the target CPU has an @code{el} suffix
21819 and @code{big} otherwise.
21821 Note that these commands merely adjust interpretation of symbolic
21822 data on the host, and that they have absolutely no effect on the
21825 @node Heterogeneous Debugging
21826 @chapter Debugging Heterogeneous Programs
21827 @cindex heterogeneous debugging
21831 @emph{Note:} The commands presented in this chapter are not currently fully
21832 implemented. @xref{AMD GPU} for the current support available.
21836 @cindex heterogeneous system
21837 @cindex heterogeneous program
21838 In some operating systems, such as Linux with @acronym{AMD}'s
21839 @acronym{ROCm, Radeon Open Compute platforM} installed, a single
21840 program may have multiple threads in the same process, executing on
21841 different devices which may have different target architectures. Such
21842 a system is termed a @dfn{heterogeneous system} and a program that
21843 uses the multiple devices is termed a @dfn{heterogeneous program}.
21845 @cindex heterogeneous agent
21846 The multiple devices of a heterogeneous system are termed
21847 @dfn{heterogeneous agents}. They can include the following kinds of
21848 devices: @acronym{CPU, Central Processing Unit}, @acronym{GPU,
21849 Graphics Processing Unit}, @acronym{DSP, Digital Signal Processor},
21850 @acronym{FPGA, Field Programmable Gate Array}, as well as other
21851 specialized hardware.
21853 @cindex heterogeneous host agent
21854 The device of a heterogeneous system that starts the execution of the
21855 program is termed the @dfn{heterogeneous host agent}.
21857 The precise way threads are created on different heterogeneous agents
21858 may vary from one heterogeneous system to another, but in general the
21859 threads behave similarly no matter what heterogeneous agent is
21860 executing them, except that the target architecture may be different.
21862 @cindex heterogeneous queue
21863 @cindex heterogeneous packet
21864 A heterogeneous program can create @dfn{heterogeneous queues}
21865 associated with a heterogeneous agent. The heterogeneous program can
21866 then place @dfn{heterogeneous packets} on a heterogeneous queue to
21867 control the actions of the associated heterogeneous agent. A
21868 heterogeneous agent removes heterogeneous packets from the
21869 heterogeneous queues assocated with it and performs the requested
21870 actions. The packet actions and scheduling of packet processing
21871 varies depending on the heterogeneous system and the target
21872 architecture of the heterogeneous agent. @xref{Architectures}.
21874 @cindex heterogeneous dispatch packet
21875 @cindex heterogeneous dispatch
21876 A @dfn{heterogeneous dispatch packet} is used to initiate code
21877 execution on a heterogeneous agent. A single heterogeneous dispatch
21878 packet may specify that the heterogeneous agent create a set of
21879 threads that are all associated with a corresponding
21880 @dfn{heterogeneous dispatch}. Each thread typically has an associated
21881 position within the heterogeneous dispatch, possibly expressed as a
21882 multi-dimensional grid position. The heterogeneous agent typically
21883 can create multiple threads that execute concurrently. If a
21884 heterogeneous dispatch is larger than the number of concurrent threads
21885 that can be created, the heterogeneous agent creates threads of the
21886 heterogeneous dispatch as other threads complete. When all the
21887 threads of a heterogeneous dispatch have been created and have
21888 completed, the heterogeneous dispatch is considered complete.
21890 @cindex heterogeneous work-group
21891 The threads of a heterogeneous dispatch may be grouped into
21892 @dfn{heterogeneous work-groups}. The threads that belong to the same
21893 heterogeneous work-group may have special shared memory, and efficient
21894 execution synchronization abilities. A thread that is part of a
21895 heterogeneous work-group typically has an associated position within
21896 the heterogeneous work-group, possibly also expressed as a
21897 multi-dimensional grid position.
21899 Other heterogeneous packets may control heterogeneous packet
21900 scheduling, memory visibility between the threads of a heterogeneous
21901 dispatch and other threads, or other services supported by the
21902 heterogeneous system.
21904 @cindex heterogeneous lane
21905 On some heterogeneous systems there can be heterogeneous agents that
21906 support @acronym{SIMD, Single Instruction Multiple Data} or
21907 @acronym{SIMT, Single Instruction Multiple Threads} machine
21908 instructions. On these target achitectures, a single machine
21909 instruction can operate in parallel on multiple @dfn{heterogeneous
21912 @cindex divergent control flow
21913 Source languages used by heterogeneous programs can be implemented on
21914 target achitectures that support multiple heterogeneous lanes by
21915 mapping a source language thread of execution onto a heterogeneous
21916 lane of a single target architecture thread. Control flow in the
21917 source language may be implemented by controlling which heterogeneous
21918 lanes are active. If the source language control flow may result in
21919 some heterogeneous lanes becoming inactive while some remain active,
21920 the control flow is said to be @dfn{divergent}. Typically, the
21921 machine code may execute different divergent paths for different sets
21922 of heterogeneous lanes, before the control flow recoverges and all
21923 heterogeneous lanes become active.
21925 Just because a target architecture supports multiple lanes, does not
21926 mean that the source language is mapped to use them to implement
21927 source language threads of execution. Therefore, a thread is only
21928 considered to have multiple heterogeneous lanes if it's current frame
21929 corresponds to a source language that does do such a mapping.
21931 @anchor{Address Space}
21932 @cindex address space
21933 On some heterogeneous systems there can be heterogeneous agents with
21934 target achitectures that support multiple @dfn{address spaces}. In
21935 these target achitectures, there may be memory that is physically
21936 disjoint from regular global virtual memory. There can also be cases
21937 when the same underlying memory can be accessed using linear addresses
21938 that map to the underlying physical memory in an interleaved manner.
21939 In these target architectures there can be distinct machine
21940 instructions to access the distinct address spaces. For example,
21941 there may be physical hardware scratch pad memory that is allocated
21942 and accessible only to the threads that are associated with the same
21943 heterogeneous work-group. There may be hardware address swizzle logic
21944 that allows regular global virtual memory to be allocated per
21945 heterogeneous lane such that they have a linear address view, which in
21946 fact maps to an interleaved global virtual memory access to improve
21949 @value{GDBN} provides these facilities for debugging heterogeneous
21954 @item @code{info sharedlibrary}, command supports code objects for
21955 multiple architectures
21957 @item debugger convenience variables for heterogeneous entities
21959 @item @code{set architecture}, @code{show architecture}, @code{x/i},
21960 @code{disassemble}, commands to disassemble multiple architectures in
21963 @item @code{info threads}, @code{thread}, commands support threads
21964 executing on multiple heterogeneous agents
21966 @item @code{info agents}, @code{info queues}, @code{info packets},
21967 @code{info dispatches}, commands to inquire about the heterogeneous
21970 @item @code{info lanes}, @code{lane}, commands support source language
21971 threads of execution that are mapped to SIMD-like lanes of a thread
21973 @item @code{$_thread_find}, @code{$_thread_find_first_gid},
21974 @code{$_lane_find}, @code{$_lane_find_first_gid} debugger convenience
21975 functions can find threads and heterogeneous lanes associated with
21976 specific heterogeneous entities
21978 @item @code{maint print address-spaces}, command together with address
21979 qualifiers supports multiple address spaces
21983 A heterogeneous system may use separate code objects for the different
21984 target architectures of the heterogeneous agents. The @code{info
21985 sharedlibrary} command lists all the code objects currently loaded,
21986 regardless of their target architecture.
21988 The following rules apply in determining the target architecture used
21989 by commands when debugging heterogeneous programs:
21994 Typically the target architecture of the heterogeneous host agent is
21995 the target architecture of the program's code object. The @code{set
21996 architecture} command (@pxref{Targets,,Specifying a Debugging Target})
21997 can be used to change this target architecture. The target
21998 architecture of other heterogeneous agents is typically the target
21999 architecture of the associated device.
22002 The target architecture of a thread is the target architecture of the
22003 selected stack frame. Typically stack frames will have the same
22004 target architecture as the heterogeneous agent on which the thread was
22005 created, however, a target may assocociate different target
22006 architectures for different stack frames.
22009 The current target architecture is the target architecture of the
22010 selected thread, or the target architecture of the heterogeneous host
22011 agent if there are no threads.
22015 @value{GDBN} handles the heterogeneous agent, queue, and dispatch
22016 entities in a similar manner to threads (@pxref{Threads}):
22021 For debugging purposes, @value{GDBN} associates its own number
22022 ---always a single integer---with each heterogeneous entity of an
22023 inferior. This number is unique between all instances of
22024 heterogeneous entities of an inferior, but not unique between
22025 heterogeneous entities of different inferiors.
22028 You can refer to a given heterogeneous entity in an inferior using the
22029 qualified @var{inferior-num}.@var{heterogeneous-entity-num} syntax,
22030 also known as a @dfn{qualified heterogeneous entity ID}, with
22031 @var{inferior-num} being the inferior number and
22032 @var{heterogeneous-entity-num} being the heterogeneous entity number
22033 of the given inferior. If you omit @var{inferior-num}, then
22034 @value{GDBN} infers you're referring to a heterogeneous entity of the
22038 Until you create a second inferior, @value{GDBN} does not show the
22039 @var{inferior-num} part of heterogeneous entity IDs, even though you
22040 can always use the full
22041 @var{inferior-num}.@var{heterogeneous-entity-num} form to refer to
22042 heterogeneous entities of inferior 1, the initial inferior.
22045 @anchor{heterogeneous entity ID list}
22046 @cindex heterogeneous entity ID list
22047 Some commands accept a space-separated @dfn{heterogeneous entity ID
22048 list} as argument. The list element has the same forms as for thread
22049 ID lists. @xref{thread ID list}.
22052 @anchor{global heterogeneous entity numbers} In addition to a
22053 @emph{per-inferior} number, each heterogeneous entity is also assigned
22054 a unique @emph{global} number, also known as @dfn{global heterogeneous
22055 entity ID}, a single integer. Unlike the heterogeneous entity number
22056 component of the heterogeneous entity ID, no two threads have the same
22057 global heterogeneous entity ID, even when you're debugging multiple
22062 The following debugger convenience variables (@pxref{Convenience
22063 Vars,,Convenience Variables}) are related to heterogeneous debugging.
22064 You may find these useful in writing breakpoint conditional
22065 expressions, command scripts, and so forth.
22071 @itemx $_thread_systag
22072 @itemx $_thread_name
22073 @xref{Convenience Vars,,Convenience Variables}.
22075 @vindex $_agent@r{, convenience variable}
22076 @vindex $_gagent@r{, convenience variable}
22077 @vindex $_queue@r{, convenience variable}
22078 @vindex $_gqueue@r{, convenience variable}
22079 @vindex $_dispatch@r{, convenience variable}
22080 @vindex $_gdispatch@r{, convenience variable}
22081 @vindex $_lane@r{, convenience variable}
22082 @vindex $_glane@r{, convenience variable}
22091 There are debugger convenience variables that contain the number of
22092 each heterogeneous entity associated with the current thread if it was
22093 created by a heterogeneous dispatch, or 0 otherwise. @code{$_agent},
22094 @code{$_queue}, and @code{$_dispatch} contain the corresponding
22095 per-inferior heterogeneous entity number. While @code{$_gagent},
22096 @code{$_gqueue}, and @code{$_gdispatch}, contain the corresponding
22097 global heterogeneous entity number.
22099 @vindex $_lane@r{, convenience variable}
22100 @vindex $_glane@r{, convenience variable}
22103 The heterogeneous lane number of the current lane of the current thread.
22104 @code{$_lane} contains the corresponding per-inferior heterogeneous lane
22105 number. While @code{$_glane} contains the corresponding global
22106 heterogeneous lane number. If the current thread does not have multiple
22107 heterogeneous lanes, it is treated as if it has a single heterogeneous
22110 @vindex $_dispatch_pos@r{, convenience variable}
22111 @item $_dispatch_pos
22112 The heterogeneous dispatch position string of the current thread within
22113 its associated heterogeneous dispatch if it is was created by a
22114 heterogeneous dispatch, or the empty string otherwise. The format
22115 varies depending on the heterogeneous system and target architecture
22116 of the heterogeneous agent. @xref{Architectures}.
22118 @vindex $_lane_name@r{, convenience variable}
22120 The heterogeneous lane name string of the current heterogeneous lane, or
22121 the empty string if no name has been assigned by the @code{lane name}
22124 @vindex $_thread_workgroup_pos@r{, convenience variable}
22125 @vindex $_lane_workgroup_pos@r{, convenience variable}
22126 @item $_thread_workgroup_pos
22127 @item $_lane_workgroup_pos
22128 The heterogeneous work-group position string of the current thread or
22129 heterogeneous lane within its associated heterogeneous dispatch if it
22130 is was created by a heterogeneous dispatch, or the empty string
22131 otherwise. The format varies depending on the heterogeneous system
22132 and target architecture of the heterogeneous agent.
22133 @xref{Architectures}.
22135 @vindex $_lane_systag@r{, convenience variable}
22136 @item $_lane_systag
22137 The target system's heterogeneous lane identifier (@var{lane_systag})
22138 string of the current heterogeneous lane. @xref{target system lane
22143 The following debugger convenience functions (@pxref{Convenience
22144 Funs,,Convenience Functions}) are related to heterogeneous debugging.
22145 Given the very large number of threads on heterogeneous systems, these
22146 may be very useful. They allow threads or thread lists to be
22147 specified based on the target system's thread identifier
22148 (@var{systag}) or thread name, and allow heterogeneous lanes or
22149 heterogeneous lane lists to be specified based on the target system's
22150 heterogeneous lane identifier (@var{lane_systag}) or heterogeneous
22155 @item $_thread_find
22156 @itemx $_thread_find_first_gid
22157 @xref{Convenience Funs,,Convenience Functions}.
22159 @findex $_lane_find@r{, convenience function}
22160 @item $_lane_find(@var{regex})
22161 Searches for heterogeneous lanes whose name or @var{lane_systag}
22162 matches the supplied regular expression. The syntax of the regular
22163 expression is that specified by @code{Python}'s regular expression
22166 Returns a string that is the space separated list of per-inferior
22167 heterogeneous lane numbers of the found heterogeneous lanes. If
22168 debugging multiple inferiors, the heterogeneous lane numbers are
22169 qualified with the inferior number. If no heterogeneous lane are
22170 found, the empty string is returned. The string can be used in
22171 commands that accept a heterogeneous lane ID list.
22172 @xref{heterogeneous entity ID list}.
22174 For example, the following command lists all heterogeneous lanes that
22175 are part of a heterogeneous work-group with work-group position
22176 @samp{(1,2,3)} (@pxref{Heterogeneous Debugging}):
22179 (@value{GDBP}) info lanes $_thread_find ("work-item(1,2,3)")
22182 @item $_lane_find_first_gid(@var{regex})
22183 @findex $_lane_find_first_gid@r{, convenience function}
22184 Similar to the @code{$_lane_find} convenience function, except it
22185 returns a number that is the global heterogeneous lane number of one
22186 of the heterogeneous lanes found, or 0 if no heterogeneous lanes were
22187 found. The number can be used in commands that accept a global
22188 heterogeneous lane number. @xref{global heterogeneous entity
22191 For example, the following command sets the current heterogeneous lane
22192 to one of the heterogeneous lanes that are part of a heterogeneous
22193 work-group with work-item position @samp{(1,2,3)}:
22196 (@value{GDBP}) lane -gid $_lane_find_first_gid ("work-item(1,2,3)")
22201 The following commands are related to heterogeneous debugging:
22205 @item info agents @r{[}-gid@r{]} @r{[}@var{agent-id-list}@r{]}
22206 @itemx info queues @r{[}-gid@r{]} @r{[}@var{queue-id-list}@r{]}
22207 @itemx info dispatches @r{[}-gid@r{]} @r{[}@var{dispatch-id-list}@r{]}
22208 @kindex info agents
22209 @kindex info queues
22210 @kindex info dispatches
22211 @code{info agents}, @code{info queues} and @code{info dispatches}
22212 commands display information about one or more heterogeneous agents,
22213 heterogeneous queues and executing heterogeneous dispatches
22214 respectively. With no arguments displays information about all
22215 corresponding heterogeneous entities. You can specify the list of
22216 heterogeneous entities that you want to display using the
22217 heterogeneous entity ID list syntax (@pxref{heterogeneous entity ID
22220 @value{GDBN} displays for each heterogeneous entity (in this order):
22224 the per-inferior heterogeneous entity number assigned by @value{GDBN}
22227 the global heterogeneous entity number assigned by @value{GDBN}, if
22228 the @w{@option{-gid}} option was specified
22231 for the @code{info queues} and @code{info dispatches} commands, the
22232 associated heterogeneous agent number assigned by @value{GDBN},
22233 displayed as a global ID if the @w{@option{-gid}} option was
22234 specified, otherwise displayed as the per-inferior ID
22237 for the @code{info dispatches} command, the associated heterogeneous
22238 queue number assigned by @value{GDBN}, displayed as a global ID if the
22239 @w{@option{-gid}} option was specified, otherwise displayed as the
22243 additional information about the heterogeneous entity that varies
22244 depending on the heterogeneous system and may vary depending on the
22245 target architecture of the heterogeneous entity
22246 (@pxref{Architectures})
22250 Some heterogeneous agents may not be listed until the inferior has
22251 started execution of the program.
22253 @item info packets @r{[}-gid@r{]} @r{[}@var{queue-id-list}@r{]}
22254 @kindex info packets
22255 Display information about the heterogeneous packets on one or more
22256 heterogeneous queues. With no arguments displays information about
22257 all heterogeneous queues. You can specify the list of heterogeneous
22258 queues that you want to display using the heterogeneous queue ID list
22259 syntax (@pxref{heterogeneous entity ID list}).
22261 Since heterogeneous agents may be processing heterogeneous packets
22262 asynchronously, the display is at best a snapshot, and may be
22263 inconsistent due to the heterogeneous queues being updated while they
22264 are being inspected.
22266 The heterogeneous packets are listed contiguously for each
22267 heterogeneous agent, and for each heterogeneous queue of that
22268 heterogeneous agent, with the oldest packet first.
22270 @value{GDBN} displays for each heterogeneous packet (in this order):
22274 the associated heterogeneous agent number assigned by @value{GDBN},
22275 displayed as a global ID if the @w{@option{-gid}} option was
22276 specified, otherwise displayed as the per-inferior ID
22279 the associated heterogeneous queue number assigned by @value{GDBN},
22280 displayed as a global ID if the @w{@option{-gid}} option was
22281 specified, otherwise displayed as the per-inferior ID
22284 the packet position in the heterogeneous queue, with the oldest one
22288 additional information about the heterogeneous packet that varies
22289 depending on the heterogeneous system and may vary depending on the
22290 target architecture of the heterogeneous entity
22291 (@pxref{Architectures})
22295 @item info threads @r{[}-gid@r{]} @r{[}@var{thread-id-list}@r{]}
22296 The @code{info threads} command (@pxref{Threads}) lists the threads
22297 created on all the heterogeneous agents.
22299 If any of the threads listed have multiple heterogeneous lanes, then
22300 an additional @emph{Lanes} column is displayed before the target
22301 system's thread identifier (@var{systag}) column. For threads that
22302 have multiple heterogeneous lanes, the number of heterogeneous lanes
22303 that are active followed by a slash and the total number of
22304 heterogeneous lanes of the current frame of the thread is displayed.
22305 Otherwise, nothing is displayed.
22307 The target system's thread identifier (@var{systag}) (@pxref{target
22308 system thread identifier}) for threads associated with heterogeneous
22309 dispatches varies depending on the heterogeneous system and target
22310 architecture of the heterogeneous agent. However, it typically will
22311 include information about the heterogeneous agent, heterogeneous
22312 queue, heterogeneous dispatch, heterogeneous work-group position
22313 within the heterogeneous dispatch, and thread position within the
22314 heterogeneous work-group. @xref{Architectures}.
22316 The stack frame summary displayed is for the active lanes of the
22317 thread. This may differ from the stack frame information for the
22318 current lane if the focus is on an inactive lane. Use the @code{info
22319 lanes} command for information about individual lanes of a thread.
22324 @c end table here to get a little more width for example
22327 (@value{GDBP}) info threads
22328 Id Lanes Target Id Frame
22329 * 1 process 35 thread 13 main (argc=1, argv=0x7ffffff8)
22330 2 2/64 ROCm process 35 agent 1 queue 2 dispatch 3 work-group(2,3,4)/1 0x34e5 in saxpy ()
22331 3 64/64 ROCm process 65 agent 1 queue 2 dispatch 4 work-group(2,4,4)/2 0x34e5 in saxpy ()
22335 @cindex heterogeneous lane index
22336 @item thread @r{[}-gid@r{]} @var{thread-id} @r{[}@var{lane-index}@r{]}
22337 The @code{thread} command has an optional @var{lane-index} argument to
22338 specify the @dfn{heterogeneous lane index}. If the value is not
22339 between 1 and the number of heterogeneous lanes of the current frame
22340 of the thread, then @value{GDBN} will print an error. If omitted it
22343 The current thread is set to @var{thread-id} and the current
22344 heterogeneous lane is set to the heterogeneous lane corresponding to
22345 the specified heterogeneous lane index.
22347 If the thread has multiple heterogeneous lanes, @value{GDBN} responds
22348 by displaying the system identifier of the heterogeneous lane you
22349 selected, otherwise it responds with the system identifier of the
22350 thread you selected, followed by its current stack frame summary.
22352 @item thread apply @r{[}@var{thread-id-list} @r{|} all @r{[}-ascending@r{]]} @r{[}@var{flag}@r{]@dots{}} @var{command}
22353 @itemx taas [@var{option}]@dots{} @var{command}
22354 @itemx tfaas [@var{option}]@dots{} @var{command}
22357 These commands operate the same way for all threads, regardless of
22358 whether or not the thread is associated with a heterogeneous dispatch.
22360 If the thread's frame has multiple heterogeneous lanes then the
22361 heterogeneous lane index 1 is used. Use the heterogeneous lane
22362 counterpart commands if it is desired to perform the the @var{command}
22363 on each lane of a thread.
22367 @cindex lane identifier (system)
22368 @item info lanes @r{[}-gid@r{]} @var{lane-id}
22369 Display information about one or more heterogeneous lanes. With no
22370 arguments displays information about all heterogeneous lanes. You can
22371 specify the list of heterogeneous lanes that you want to display using
22372 the heterogeneous lane ID list syntax (@pxref{heterogeneous entity ID
22375 @value{GDBN} displays for each heterogeneous lane (in this order):
22379 The per-inferior heterogeneous lane number assigned by @value{GDBN}.
22382 The global heterogeneous lane number assigned by @value{GDBN}, if the
22383 @w{@option{-gid}} option was specified.
22386 The thread number assigned by @value{GDBN} for the thread that
22387 contains the heterogeneous lane. This is displayed as a global thread
22388 number if the @w{@option{-gid}} option was specified, otherwise as a
22389 per-inferior thread number. If the thread has multiple heterogeneous
22390 lanes then this is followed by a slash and the heterogeneous lane
22391 index of the heterogeneous lane within the thread with the first lane
22395 An indication of whether the heterogeneous lane is active or inactive.
22397 @anchor{target system lane identifier}
22399 The target system's heterogeneous lane identifier (@var{lane_systag}).
22400 This varies depending on the system and target architecture of the
22401 heterogeneous agent. However, for heterogeneous agents it typically
22402 will include information about the heterogeneous agent, heterogeneous
22403 queue, heterogeneous dispatch, heterogeneous work-group position
22404 within the heterogeneous dispatch, and position of the heterogeneous
22405 lane in the heterogeneous work-group. @xref{Architectures}.
22408 The heterogeneous lane's name, if one is assigned by the user (see
22409 @code{lane name}, below).
22412 The current stack frame summary for that heterogeneous lane. If the
22413 heterogeneous lane is inactive this is the source position at which the
22414 heterogeneous lane will resume.
22418 An asterisk @samp{*} to the left of the @value{GDBN} heterogeneous
22419 lane number indicates the current heterogeneous lane.
22423 @c end table here to get a little more width for example
22426 (@value{GDBP}) info lanes
22427 Id Thread Active Target Id Frame
22428 * 1 4 Y process 35 thread 13 main (argc=1, argv=0x7ffffff8)
22429 2 5/2 Y ROCm process 35 agent 1 queue 2 dispatch 3 work-group(2,3,4) work-item(1,2,4) 0x34e5 in saxpy ()
22430 3 6/12 N ROCm process 65 agent 1 queue 2 dispatch 4 work-group(2,4,4) work-item(1,2,3) 0x34e5 in saxpy ()
22433 If you're debugging multiple inferiors, @value{GDBN} displays
22434 heterogeneous lane IDs using the qualified
22435 @var{inferior-num}.@var{lane-num} format. Otherwise, only
22436 @var{lane-num} is shown.
22438 If you specify the @w{@option{-gid}} option, @value{GDBN} displays a
22439 column indicating each heterogeneous lane's global heterogeneous lane
22440 ID, and displays the thread's global thread number:
22443 (@value{GDBP}) info lanes -gid
22444 Id GId Thread Active Target Id Frame
22445 * 1.1 1 4 Y process 35 thread 13 main (argc=1, argv=0x7ffffff8)
22446 1.2 3 5/2 Y ROCm process 35 agent 1 queue 2 dispatch 3 work-group(2,3,4) work-item(1,2,4) 0x34e5 in saxpy ()
22447 2.1 1 4 Y process 65 thread 1 main (argc=1, argv=0x7ffffff8)
22448 2.2 4 6/12 N ROCm process 65 agent 1 queue 2 dispatch 4 work-group(2,4,4) work-item(1,2,3) 0x34e5 in saxpy ()
22453 @item lane @r{[}-gid@r{]} @var{lane-id}
22454 Make heterogeneous lane ID @var{lane-id} the current heterogeneous
22455 lane and the thread that contains the heterogeneous lane the current
22456 thread. The command argument @var{lane-id} is the @value{GDBN}
22457 heterogeneous lane ID: if the @w{@option{-gid}} option is given it is
22458 a global heterogeneous lane identifier, as shown in the second field
22459 of the @code{info lanes -gid} display; otherwise it is a per-inferior
22460 heterogeneous lane identifier, with or without an inferior qualifier
22461 (e.g., @samp{2.1} or @samp{1}), as shown in the first field of the
22462 @code{info lanes} display.
22464 @value{GDBN} responds by displaying the system identifier of the
22465 heterogeneous lane you selected, and its current stack frame summary:
22468 (@value{GDBP}) lane 2
22469 [Switching to lane 2 (Thread 0xb7fdab70 (LWP 12747))]
22470 #0 some_function (ignore=0x0) at example.c:8
22471 8 printf ("hello\n");
22475 As with the @samp{[New @dots{}]} message, the form of the text after
22476 @samp{Switching to} depends on your system's conventions for identifying
22477 heterogeneous lanes.
22480 @cindex name a heterogeneous lane
22481 @anchor{heterogeneous lane name}
22482 @item lane name [@var{name}]
22483 This command assigns a name to the current heterogeneous lane. If no
22484 argument is given, any existing user-specified name is removed. The
22485 heterogeneous lane name appears in the @code{info lanes} display.
22488 @cindex search for a heterogeneous lane
22490 @item lane find [@var{regexp}]
22491 Search for and display heterogeneous lane ids whose name or
22492 @var{lane_systag} matches the supplied regular expression. The syntax
22493 of the regular expression is that specified by @code{Python}'s regular
22494 expression support.
22496 As well as being the complement to the @code{lane name} command, this
22497 command also allows you to identify a heterogeneous lane by its target
22498 @var{lane_systag}. For instance, on @acronym{AMD ROCm}, the target
22499 @var{lane_systag} is the heterogeneous agent, heterogeneous queue,
22500 heterogeneous dispatch, heterogeneous work-group position and
22501 heterogeneous work-item position.
22504 (@value{GDBP}) lane find "work-group(2,3,4)"
22505 Lane 2 has lane id 'ROCm process 35 agent 1 queue 2 dispatch 3 work-group(2,3,4) work-item(1,2,4)'
22506 (@value{GDBP}) info lane 2
22507 Id Thread Active Target Id Frame
22508 2 5/2 Y ROCm process 35 agent 1 queue 2 dispatch 3 work-group(2,3,4) work-item(1,2,4) 0x34e5 in saxpy ()
22511 @c FIXME-implementors!! Perhaps better ways to find lanes and threads
22512 @c would be beneficial. If the @var{systag} and ${lane_systag} were
22513 @c considered as tuples and not a plain strings, structured queries
22514 @c could be used. Maybe that would also support the sort order of the
22515 @c returned list. SQL is an example to examine.
22517 @c User defined pretty printing functions could be allowed so that
22518 @c users can control how @var{systag} and ${lane_systag} values are
22519 @c displayed in commands that display them. This would allow cater to
22520 @c situations that benefit from full verbose output, and those where
22521 @c partial terse output is all that is needed. But the underlying
22522 @c @var{systag} and ${lane_systag} values always have the full
22525 @c Ways for commands that list lanes and threads to aggregate the
22526 @c output would be beneficial in heterogeneous systems that tend to
22527 @c have very large counts. For example, all lanes that have adjacent
22528 @c dispatch postions, and that are at the same source postion, could
22529 @c be displayed as a single row that specifies the range of postions.
22530 @c Perhaps target or user defined functions could be allowed to guide
22531 @c the aggregation, and return the aggregated range. That would allow
22532 @c different heterogeneous system to be supported that had different
22533 @c ways to represent dispatch positions. There may even be multiple
22534 @c ways to aggregate on some system.
22536 @item lane apply @r{[}@var{thread-id-list} @r{|} all @r{[}-ascending@r{]]} @r{[}@var{flag}@r{]@dots{}} @var{command}
22537 @itemx laas [@var{option}]@dots{} @var{command}
22538 @itemx lfaas [@var{option}]@dots{} @var{command}
22539 @code{lane apply}, @code{laas}, and @code{lfass} commands are simalar
22540 to their thread counterparts @code{thread apply}, @code{taas}, and
22541 @code{tfaas} respectively, except they operatate on heterogeneous
22542 lanes. @xref{Threads}.
22544 @item backtrace [@var{option}]@dots{} [@var{qualifier}]@dots{} [@var{count}]
22545 @itemx frame @r{[} @var{frame-selection-spec} @r{]}
22546 @itemx frame apply [all | @var{count} | @var{-count} | level @var{level}@dots{}] [@var{option}]@dots{} @var{command}
22547 @itemx select-frame @r{[} @var{frame-selection-spec} @r{]}
22548 @itemx up-silently @var{n}
22549 @itemx down-silently @var{n}
22551 @itemx info args [-q] [-t @var{type_regexp}] [@var{regexp}]
22552 @itemx info locals [-q] [-t @var{type_regexp}] [@var{regexp}]
22553 @itemx faas @var{command}
22554 The frame commands apply to the current heterogeneous lane.
22556 If the frame is switched from one that has multiple heterogeneous
22557 lanes to one with fewer (including only one) then the current lane is
22558 switched to the heterogeneous lane corresponding to the highest
22559 heterogeneous lane index of the new frame and @value{GDBN} responds by
22560 displaying the system identifier of the heterogeneous lane selected.
22562 @xref{Stack, ,Examining the Stack}.
22564 @item set libthread-db-search-path
22565 @itemx show libthread-db-search-path
22566 @itemx set debug libthread-db
22567 @itemx show debug libthread-db
22568 These commands only apply to threads created on the heterogeneous host
22569 agent that are not associated with a heterogeneous dispatch. There
22570 are no commands that support reporting of heterogeneous dispatch
22575 The @code{x/i} and @code{display/i} commands (@pxref{Memory,,Examining
22576 Memory}) can be used to disassemble machine instructions. They use
22577 the current target architecture.
22580 The @code{disassemble} command (@pxref{Machine Code,,Source and
22581 Machine Code}) can also be used to disassemble machine instructions.
22582 If the start address of the range is within a loaded code object, then
22583 the target architecture of the code object is used. Otherwise, the
22584 current target architecture is used.
22586 @c FIXME-implementors!! It would be more helpful if @code{set
22587 @c architecture} was an inferior setting used by both @code{x/i} and
22588 @c @code{disassemble} when not set to @code{auto}. When set to
22589 @c @code{auto} then the architecture of the code object containing the
22590 @c start address should be used by both commands. Otherwise, the
22591 @c thread target architecture should be used, or the heterogeneous host
22592 @c agent target architecture if there are no threads. That way a user
22593 @c can choose what architecture to disassemble in, and will get
22594 @c sensible behavior if they specify the default of @code{auto} even
22595 @c for heterogeneous systems.
22597 @item info registers
22598 @itemx info all-registers
22599 @itemx maint print reggroups
22600 The register commands display information about the current
22604 The @code{print} command evaluates the source language expression in
22605 the context of the current heterogeneous lane.
22613 If the current heterogeneous lane is set to an inactive heterogeneous
22614 lane, then the @code{step}, @code{next}, @code{finish} and
22615 @code{until} commands (@pxref{Continuing and Stepping, ,Continuing and
22616 Stepping}) may cause other heterogeneous lanes of the same thread to
22617 advance so that the current heterogeneous lane becomes active. This
22618 may result in other heterogeneous lanes completing whole functions.
22620 If the current heterogeneous lane is set to an inactive heterogeneous
22621 lane, then the @code{stepi} and @code{nexti} commands
22622 (@pxref{Continuing and Stepping, ,Continuing and Stepping}) may not
22623 cause the source position to appear to move until execution reaches a
22624 point that makes the current heterogeneous lane active. However,
22625 other heterogeneous lanes of the same thread will advance.
22627 @item break @r{[}-lane @var{lane-index}@r{]} @r{[}location@r{]} @r{[}if @var{cond}@r{]}
22628 @itemx tbreak @r{[}-lane @var{lane-index}@r{]} @r{[}location@r{]} @r{[}if @var{cond}@r{]}
22629 @itemx hbreak @r{[}-lane @var{lane-index}@r{]} @r{[}location@r{]} @r{[}if @var{cond}@r{]}
22630 @itemx thbreak @r{[}-lane @var{lane-index}@r{]} @r{[}location@r{]} @r{[}if @var{cond}@r{]}
22631 @itemx rbreak @r{[}-lane @var{lane-index}@r{]} @var{regex}
22632 @itemx info breakpoints @r{[}@var{list}@dots{}@r{]}
22633 @itemx watch @r{[}-lane @var{lane-index}@r{]} @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{thread-id}@r{]} @r{[}mask @var{maskvalue}@r{]}
22634 @itemx rwatch @r{[}-lane @var{lane-index}@r{]} @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{thread-id}@r{]} @r{[}mask @var{maskvalue}@r{]}
22635 @itemx awatch @r{[}-lane @var{lane-index}@r{]} @r{[}-l@r{|}-location@r{]} @var{expr} @r{[}thread @var{thread-id}@r{]} @r{[}mask @var{maskvalue}@r{]}
22636 @itemx info watchpoints @r{[}@var{list}@dots{}@r{]}
22637 @itemx catch @r{[}-lane @var{lane-index}@r{]} @var{event}
22638 @itemx tcatch @r{[}-lane @var{lane-index}@r{]} @var{event}
22639 When a breakpoint, watchpoint, or catchpoint (@pxref{Breakpoints,
22640 ,Breakpoints; Watchpoints; and Catchpoints}) is hit by a frame of a
22641 thread with multiple heterogeneous lanes, each active lane is treated
22647 The breakpoint condition, if present, is evaluated for each active
22648 heterogeneous lane.
22651 The breakpoint command, if present, is evaluated for each active
22652 heterogeneous lane that evaluates the breakpoint condition to true.
22655 If the breakpoint causes the heterogeneous lane to halt then he
22656 current heterogeneous lane is set to the halting heterogeneous lane
22657 and @value{GDBN} responds by displaying the system identifier of the
22658 heterogeneous lane selected.
22661 If the breakpoint is a temporary breakpoint, then it will be removed,
22662 and so any remaining heterogeneous lanes will not report the
22666 In non-stop mode all heterogeneous lanes that halt at the breakpoint
22670 In all-stop mode, continuing from the breakpoint will cause the next
22671 heterogeneous ative lane that hit the breakpoint to be processed.
22675 If a heterogeneous lane causes a thread to halt, then the other
22676 heterogeneous lanes of the thread will no longer execute even if in
22679 For @code{break}, @code{watch}, @code{catch}, and their variants, the
22680 @w{@option{-lane @var{lane-index}}} option can be specified. This
22681 limits @value{GDBN} to only process breakpoints if the heterogeneous
22682 lane has a heterogeneous lane index that matches @var{lane-index}.
22684 The @code{info break} and @code{info watch} commands add a @emph{Lane}
22685 column before the @emph{Address} column if any breakoint has a
22686 @var{lane-index} specified that displays the heterogeneous lane index.
22688 @c FIXME-implementors!! Should there be way to request all pending
22689 @c breakpoints to be processed? This may result in multiple
22690 @c lanes/threads being reported as halted. This would avoid the user
22691 @c having to continue a very large number of times to get all the
22692 @c threads/lanes that have unprocessed breakpoints to be processed.
22694 @c In addition, a way to list all the theards/lanes that are halted at
22695 @c a breakpoint. If this was avaiable as a conveniece function, then
22696 @c the @code{thread apply} and @code{lane apply} commands could be
22697 @c used to perform a command on all such threads in one action.
22699 @anchor{maint print address-spaces}
22700 @c FIXME-implementers!! This is not a maintenance command as it is
22701 @c displaying imformation about available address spaces that can be
22702 @c used. It has been defined as a @code{maint} command only to match
22703 @c the @code{maint print reggroups} command which also should not be a
22704 @c maintenace command for the same reason.
22705 @item maint print address-spaces @r{[}@var{file}@r{]}
22706 @code {maint print address-spaces} displays the address space names
22707 supported by each target achitecture. The optional argument
22708 @var{file} tells to what file to write the information.
22710 The address spaces info looks like this:
22713 (@value{GDBP}) @kbd{maint print address-spaces}
22721 The @var{global} address space corresponds to the default global
22722 virtual memory address space and is available for all target
22725 Every address entered or displayed can optionally specify the address
22726 space qualifier by appending an @samp{@@} followed by an address space
22727 name. @value{GDBN} will print an error if the address space name is
22728 not supported by the current architecture.
22733 (@value{GDBP}) x/x 0x10021608@@group
22734 0x10021608@@group: 0x0022fd98
22737 @c FIXME-implementors!! Perhaps the gdb internal types (such as used
22738 @c for register types) can be extended to support addresses in address
22741 If there is no current thread then the only address space that can be
22742 specified is @var{global}.
22744 If entering an address and no address space is specified, the
22745 @var{global} address space is used.
22747 If an address is displayed, the address space qualifier is omitted for
22748 the @var{global} address space.
22752 Heterogeneous systems often have very large numbers of threads.
22753 Breakpoint conditions can be used to limit the number of threads
22754 reporting breakpoint hits. For example,
22757 break kernel_foo if $_streq($_lane_workgroup_pos, "(0,0,0)")
22760 The @code{tbreak} command can be used so only one heterogeneous lane
22761 will report the breakpoint. Before continuing execution, the
22762 breakpoint will need to be set again if necessary.
22764 The @code{set scheduler-locking on} command (@pxref{Non-Stop Mode})
22765 together with the @w{@option{-lane}} breakpoint option can be used to
22766 lock @value{GDBN} to only resume the current thread, and only report
22767 breakoints for a fixed heterogeneous lane index. This avoids the
22768 overhead of resuming a large number of threads every time resuming
22769 from a breakpoint, and also avoids the focus being switched to other
22770 threads that hit the breakpoints. Note however that other threads
22771 will not be executed.
22773 The scheduler locking commands can also be helpful to prevent
22774 @value{GDBN} switching to other threads while concentrating on
22775 debugging one particular thread. The non-stop mode can be hepful to
22776 prevent the @code{continue} command from resuming other threads that
22777 are intentionally halted or from cancelling a single step command that
22778 is in progress by another thread and resuming it instead.
22779 @xref{Non-Stop Mode}.
22782 @c Change command parsing so convienence variable
22783 @c substitution will work as shown. Investigate using tuples and
22784 @c lists the result of convienence variables and convienence
22786 @c Update MI commands for heterogeneous commands.
22787 @c Update Python bindings for heterogeneous commands.
22788 @c Update gdbserver remote protocol for heterogeneous commands.
22790 @node Remote Debugging
22791 @chapter Debugging Remote Programs
22792 @cindex remote debugging
22794 If you are trying to debug a program running on a machine that cannot run
22795 @value{GDBN} in the usual way, it is often useful to use remote debugging.
22796 For example, you might use remote debugging on an operating system kernel,
22797 or on a small system which does not have a general purpose operating system
22798 powerful enough to run a full-featured debugger.
22800 Some configurations of @value{GDBN} have special serial or TCP/IP interfaces
22801 to make this work with particular debugging targets. In addition,
22802 @value{GDBN} comes with a generic serial protocol (specific to @value{GDBN},
22803 but not specific to any particular target system) which you can use if you
22804 write the remote stubs---the code that runs on the remote system to
22805 communicate with @value{GDBN}.
22807 Other remote targets may be available in your
22808 configuration of @value{GDBN}; use @code{help target} to list them.
22811 * Connecting:: Connecting to a remote target
22812 * File Transfer:: Sending files to a remote system
22813 * Server:: Using the gdbserver program
22814 * Remote Configuration:: Remote configuration
22815 * Remote Stub:: Implementing a remote stub
22819 @section Connecting to a Remote Target
22820 @cindex remote debugging, connecting
22821 @cindex @code{gdbserver}, connecting
22822 @cindex remote debugging, types of connections
22823 @cindex @code{gdbserver}, types of connections
22824 @cindex @code{gdbserver}, @code{target remote} mode
22825 @cindex @code{gdbserver}, @code{target extended-remote} mode
22827 This section describes how to connect to a remote target, including the
22828 types of connections and their differences, how to set up executable and
22829 symbol files on the host and target, and the commands used for
22830 connecting to and disconnecting from the remote target.
22832 @subsection Types of Remote Connections
22834 @value{GDBN} supports two types of remote connections, @code{target remote}
22835 mode and @code{target extended-remote} mode. Note that many remote targets
22836 support only @code{target remote} mode. There are several major
22837 differences between the two types of connections, enumerated here:
22841 @cindex remote debugging, detach and program exit
22842 @item Result of detach or program exit
22843 @strong{With target remote mode:} When the debugged program exits or you
22844 detach from it, @value{GDBN} disconnects from the target. When using
22845 @code{gdbserver}, @code{gdbserver} will exit.
22847 @strong{With target extended-remote mode:} When the debugged program exits or
22848 you detach from it, @value{GDBN} remains connected to the target, even
22849 though no program is running. You can rerun the program, attach to a
22850 running program, or use @code{monitor} commands specific to the target.
22852 When using @code{gdbserver} in this case, it does not exit unless it was
22853 invoked using the @option{--once} option. If the @option{--once} option
22854 was not used, you can ask @code{gdbserver} to exit using the
22855 @code{monitor exit} command (@pxref{Monitor Commands for gdbserver}).
22857 @item Specifying the program to debug
22858 For both connection types you use the @code{file} command to specify the
22859 program on the host system. If you are using @code{gdbserver} there are
22860 some differences in how to specify the location of the program on the
22863 @strong{With target remote mode:} You must either specify the program to debug
22864 on the @code{gdbserver} command line or use the @option{--attach} option
22865 (@pxref{Attaching to a program,,Attaching to a Running Program}).
22867 @cindex @option{--multi}, @code{gdbserver} option
22868 @strong{With target extended-remote mode:} You may specify the program to debug
22869 on the @code{gdbserver} command line, or you can load the program or attach
22870 to it using @value{GDBN} commands after connecting to @code{gdbserver}.
22872 @anchor{--multi Option in Types of Remote Connnections}
22873 You can start @code{gdbserver} without supplying an initial command to run
22874 or process ID to attach. To do this, use the @option{--multi} command line
22875 option. Then you can connect using @code{target extended-remote} and start
22876 the program you want to debug (see below for details on using the
22877 @code{run} command in this scenario). Note that the conditions under which
22878 @code{gdbserver} terminates depend on how @value{GDBN} connects to it
22879 (@code{target remote} or @code{target extended-remote}). The
22880 @option{--multi} option to @code{gdbserver} has no influence on that.
22882 @item The @code{run} command
22883 @strong{With target remote mode:} The @code{run} command is not
22884 supported. Once a connection has been established, you can use all
22885 the usual @value{GDBN} commands to examine and change data. The
22886 remote program is already running, so you can use commands like
22887 @kbd{step} and @kbd{continue}.
22889 @strong{With target extended-remote mode:} The @code{run} command is
22890 supported. The @code{run} command uses the value set by
22891 @code{set remote exec-file} (@pxref{set remote exec-file}) to select
22892 the program to run. Command line arguments are supported, except for
22893 wildcard expansion and I/O redirection (@pxref{Arguments}).
22895 If you specify the program to debug on the command line, then the
22896 @code{run} command is not required to start execution, and you can
22897 resume using commands like @kbd{step} and @kbd{continue} as with
22898 @code{target remote} mode.
22900 @anchor{Attaching in Types of Remote Connections}
22902 @strong{With target remote mode:} The @value{GDBN} command @code{attach} is
22903 not supported. To attach to a running program using @code{gdbserver}, you
22904 must use the @option{--attach} option (@pxref{Running gdbserver}).
22906 @strong{With target extended-remote mode:} To attach to a running program,
22907 you may use the @code{attach} command after the connection has been
22908 established. If you are using @code{gdbserver}, you may also invoke
22909 @code{gdbserver} using the @option{--attach} option
22910 (@pxref{Running gdbserver}).
22914 @anchor{Host and target files}
22915 @subsection Host and Target Files
22916 @cindex remote debugging, symbol files
22917 @cindex symbol files, remote debugging
22919 @value{GDBN}, running on the host, needs access to symbol and debugging
22920 information for your program running on the target. This requires
22921 access to an unstripped copy of your program, and possibly any associated
22922 symbol files. Note that this section applies equally to both @code{target
22923 remote} mode and @code{target extended-remote} mode.
22925 Some remote targets (@pxref{qXfer executable filename read}, and
22926 @pxref{Host I/O Packets}) allow @value{GDBN} to access program files over
22927 the same connection used to communicate with @value{GDBN}. With such a
22928 target, if the remote program is unstripped, the only command you need is
22929 @code{target remote} (or @code{target extended-remote}).
22931 If the remote program is stripped, or the target does not support remote
22932 program file access, start up @value{GDBN} using the name of the local
22933 unstripped copy of your program as the first argument, or use the
22934 @code{file} command. Use @code{set sysroot} to specify the location (on
22935 the host) of target libraries (unless your @value{GDBN} was compiled with
22936 the correct sysroot using @code{--with-sysroot}). Alternatively, you
22937 may use @code{set solib-search-path} to specify how @value{GDBN} locates
22940 The symbol file and target libraries must exactly match the executable
22941 and libraries on the target, with one exception: the files on the host
22942 system should not be stripped, even if the files on the target system
22943 are. Mismatched or missing files will lead to confusing results
22944 during debugging. On @sc{gnu}/Linux targets, mismatched or missing
22945 files may also prevent @code{gdbserver} from debugging multi-threaded
22948 @subsection Remote Connection Commands
22949 @cindex remote connection commands
22950 @value{GDBN} can communicate with the target over a serial line, a
22951 local Unix domain socket, or
22952 over an @acronym{IP} network using @acronym{TCP} or @acronym{UDP}. In
22953 each case, @value{GDBN} uses the same protocol for debugging your
22954 program; only the medium carrying the debugging packets varies. The
22955 @code{target remote} and @code{target extended-remote} commands
22956 establish a connection to the target. Both commands accept the same
22957 arguments, which indicate the medium to use:
22961 @item target remote @var{serial-device}
22962 @itemx target extended-remote @var{serial-device}
22963 @cindex serial line, @code{target remote}
22964 Use @var{serial-device} to communicate with the target. For example,
22965 to use a serial line connected to the device named @file{/dev/ttyb}:
22968 target remote /dev/ttyb
22971 If you're using a serial line, you may want to give @value{GDBN} the
22972 @samp{--baud} option, or use the @code{set serial baud} command
22973 (@pxref{Remote Configuration, set serial baud}) before the
22974 @code{target} command.
22976 @item target remote @var{local-socket}
22977 @itemx target extended-remote @var{local-socket}
22978 @cindex local socket, @code{target remote}
22979 @cindex Unix domain socket
22980 Use @var{local-socket} to communicate with the target. For example,
22981 to use a local Unix domain socket bound to the file system entry @file{/tmp/gdb-socket0}:
22984 target remote /tmp/gdb-socket0
22987 Note that this command has the same form as the command to connect
22988 to a serial line. @value{GDBN} will automatically determine which
22989 kind of file you have specified and will make the appropriate kind
22991 This feature is not available if the host system does not support
22992 Unix domain sockets.
22994 @item target remote @code{@var{host}:@var{port}}
22995 @itemx target remote @code{@var{[host]}:@var{port}}
22996 @itemx target remote @code{tcp:@var{host}:@var{port}}
22997 @itemx target remote @code{tcp:@var{[host]}:@var{port}}
22998 @itemx target remote @code{tcp4:@var{host}:@var{port}}
22999 @itemx target remote @code{tcp6:@var{host}:@var{port}}
23000 @itemx target remote @code{tcp6:@var{[host]}:@var{port}}
23001 @itemx target extended-remote @code{@var{host}:@var{port}}
23002 @itemx target extended-remote @code{@var{[host]}:@var{port}}
23003 @itemx target extended-remote @code{tcp:@var{host}:@var{port}}
23004 @itemx target extended-remote @code{tcp:@var{[host]}:@var{port}}
23005 @itemx target extended-remote @code{tcp4:@var{host}:@var{port}}
23006 @itemx target extended-remote @code{tcp6:@var{host}:@var{port}}
23007 @itemx target extended-remote @code{tcp6:@var{[host]}:@var{port}}
23008 @cindex @acronym{TCP} port, @code{target remote}
23009 Debug using a @acronym{TCP} connection to @var{port} on @var{host}.
23010 The @var{host} may be either a host name, a numeric @acronym{IPv4}
23011 address, or a numeric @acronym{IPv6} address (with or without the
23012 square brackets to separate the address from the port); @var{port}
23013 must be a decimal number. The @var{host} could be the target machine
23014 itself, if it is directly connected to the net, or it might be a
23015 terminal server which in turn has a serial line to the target.
23017 For example, to connect to port 2828 on a terminal server named
23021 target remote manyfarms:2828
23024 To connect to port 2828 on a terminal server whose address is
23025 @code{2001:0db8:85a3:0000:0000:8a2e:0370:7334}, you can either use the
23026 square bracket syntax:
23029 target remote [2001:0db8:85a3:0000:0000:8a2e:0370:7334]:2828
23033 or explicitly specify the @acronym{IPv6} protocol:
23036 target remote tcp6:2001:0db8:85a3:0000:0000:8a2e:0370:7334:2828
23039 This last example may be confusing to the reader, because there is no
23040 visible separation between the hostname and the port number.
23041 Therefore, we recommend the user to provide @acronym{IPv6} addresses
23042 using square brackets for clarity. However, it is important to
23043 mention that for @value{GDBN} there is no ambiguity: the number after
23044 the last colon is considered to be the port number.
23046 If your remote target is actually running on the same machine as your
23047 debugger session (e.g.@: a simulator for your target running on the
23048 same host), you can omit the hostname. For example, to connect to
23049 port 1234 on your local machine:
23052 target remote :1234
23056 Note that the colon is still required here.
23058 @item target remote @code{udp:@var{host}:@var{port}}
23059 @itemx target remote @code{udp:@var{[host]}:@var{port}}
23060 @itemx target remote @code{udp4:@var{host}:@var{port}}
23061 @itemx target remote @code{udp6:@var{[host]}:@var{port}}
23062 @itemx target extended-remote @code{udp:@var{host}:@var{port}}
23063 @itemx target extended-remote @code{udp:@var{host}:@var{port}}
23064 @itemx target extended-remote @code{udp:@var{[host]}:@var{port}}
23065 @itemx target extended-remote @code{udp4:@var{host}:@var{port}}
23066 @itemx target extended-remote @code{udp6:@var{host}:@var{port}}
23067 @itemx target extended-remote @code{udp6:@var{[host]}:@var{port}}
23068 @cindex @acronym{UDP} port, @code{target remote}
23069 Debug using @acronym{UDP} packets to @var{port} on @var{host}. For example, to
23070 connect to @acronym{UDP} port 2828 on a terminal server named @code{manyfarms}:
23073 target remote udp:manyfarms:2828
23076 When using a @acronym{UDP} connection for remote debugging, you should
23077 keep in mind that the `U' stands for ``Unreliable''. @acronym{UDP}
23078 can silently drop packets on busy or unreliable networks, which will
23079 cause havoc with your debugging session.
23081 @item target remote | @var{command}
23082 @itemx target extended-remote | @var{command}
23083 @cindex pipe, @code{target remote} to
23084 Run @var{command} in the background and communicate with it using a
23085 pipe. The @var{command} is a shell command, to be parsed and expanded
23086 by the system's command shell, @code{/bin/sh}; it should expect remote
23087 protocol packets on its standard input, and send replies on its
23088 standard output. You could use this to run a stand-alone simulator
23089 that speaks the remote debugging protocol, to make net connections
23090 using programs like @code{ssh}, or for other similar tricks.
23092 If @var{command} closes its standard output (perhaps by exiting),
23093 @value{GDBN} will try to send it a @code{SIGTERM} signal. (If the
23094 program has already exited, this will have no effect.)
23098 @cindex interrupting remote programs
23099 @cindex remote programs, interrupting
23100 Whenever @value{GDBN} is waiting for the remote program, if you type the
23101 interrupt character (often @kbd{Ctrl-c}), @value{GDBN} attempts to stop the
23102 program. This may or may not succeed, depending in part on the hardware
23103 and the serial drivers the remote system uses. If you type the
23104 interrupt character once again, @value{GDBN} displays this prompt:
23107 Interrupted while waiting for the program.
23108 Give up (and stop debugging it)? (y or n)
23111 In @code{target remote} mode, if you type @kbd{y}, @value{GDBN} abandons
23112 the remote debugging session. (If you decide you want to try again later,
23113 you can use @kbd{target remote} again to connect once more.) If you type
23114 @kbd{n}, @value{GDBN} goes back to waiting.
23116 In @code{target extended-remote} mode, typing @kbd{n} will leave
23117 @value{GDBN} connected to the target.
23120 @kindex detach (remote)
23122 When you have finished debugging the remote program, you can use the
23123 @code{detach} command to release it from @value{GDBN} control.
23124 Detaching from the target normally resumes its execution, but the results
23125 will depend on your particular remote stub. After the @code{detach}
23126 command in @code{target remote} mode, @value{GDBN} is free to connect to
23127 another target. In @code{target extended-remote} mode, @value{GDBN} is
23128 still connected to the target.
23132 The @code{disconnect} command closes the connection to the target, and
23133 the target is generally not resumed. It will wait for @value{GDBN}
23134 (this instance or another one) to connect and continue debugging. After
23135 the @code{disconnect} command, @value{GDBN} is again free to connect to
23138 @cindex send command to remote monitor
23139 @cindex extend @value{GDBN} for remote targets
23140 @cindex add new commands for external monitor
23142 @item monitor @var{cmd}
23143 This command allows you to send arbitrary commands directly to the
23144 remote monitor. Since @value{GDBN} doesn't care about the commands it
23145 sends like this, this command is the way to extend @value{GDBN}---you
23146 can add new commands that only the external monitor will understand
23150 @node File Transfer
23151 @section Sending files to a remote system
23152 @cindex remote target, file transfer
23153 @cindex file transfer
23154 @cindex sending files to remote systems
23156 Some remote targets offer the ability to transfer files over the same
23157 connection used to communicate with @value{GDBN}. This is convenient
23158 for targets accessible through other means, e.g.@: @sc{gnu}/Linux systems
23159 running @code{gdbserver} over a network interface. For other targets,
23160 e.g.@: embedded devices with only a single serial port, this may be
23161 the only way to upload or download files.
23163 Not all remote targets support these commands.
23167 @item remote put @var{hostfile} @var{targetfile}
23168 Copy file @var{hostfile} from the host system (the machine running
23169 @value{GDBN}) to @var{targetfile} on the target system.
23172 @item remote get @var{targetfile} @var{hostfile}
23173 Copy file @var{targetfile} from the target system to @var{hostfile}
23174 on the host system.
23176 @kindex remote delete
23177 @item remote delete @var{targetfile}
23178 Delete @var{targetfile} from the target system.
23183 @section Using the @code{gdbserver} Program
23186 @cindex remote connection without stubs
23187 @code{gdbserver} is a control program for Unix-like systems, which
23188 allows you to connect your program with a remote @value{GDBN} via
23189 @code{target remote} or @code{target extended-remote}---but without
23190 linking in the usual debugging stub.
23192 @code{gdbserver} is not a complete replacement for the debugging stubs,
23193 because it requires essentially the same operating-system facilities
23194 that @value{GDBN} itself does. In fact, a system that can run
23195 @code{gdbserver} to connect to a remote @value{GDBN} could also run
23196 @value{GDBN} locally! @code{gdbserver} is sometimes useful nevertheless,
23197 because it is a much smaller program than @value{GDBN} itself. It is
23198 also easier to port than all of @value{GDBN}, so you may be able to get
23199 started more quickly on a new system by using @code{gdbserver}.
23200 Finally, if you develop code for real-time systems, you may find that
23201 the tradeoffs involved in real-time operation make it more convenient to
23202 do as much development work as possible on another system, for example
23203 by cross-compiling. You can use @code{gdbserver} to make a similar
23204 choice for debugging.
23206 @value{GDBN} and @code{gdbserver} communicate via either a serial line
23207 or a TCP connection, using the standard @value{GDBN} remote serial
23211 @emph{Warning:} @code{gdbserver} does not have any built-in security.
23212 Do not run @code{gdbserver} connected to any public network; a
23213 @value{GDBN} connection to @code{gdbserver} provides access to the
23214 target system with the same privileges as the user running
23218 @anchor{Running gdbserver}
23219 @subsection Running @code{gdbserver}
23220 @cindex arguments, to @code{gdbserver}
23221 @cindex @code{gdbserver}, command-line arguments
23223 Run @code{gdbserver} on the target system. You need a copy of the
23224 program you want to debug, including any libraries it requires.
23225 @code{gdbserver} does not need your program's symbol table, so you can
23226 strip the program if necessary to save space. @value{GDBN} on the host
23227 system does all the symbol handling.
23229 To use the server, you must tell it how to communicate with @value{GDBN};
23230 the name of your program; and the arguments for your program. The usual
23234 target> gdbserver @var{comm} @var{program} [ @var{args} @dots{} ]
23237 @var{comm} is either a device name (to use a serial line), or a TCP
23238 hostname and portnumber, or @code{-} or @code{stdio} to use
23239 stdin/stdout of @code{gdbserver}.
23240 For example, to debug Emacs with the argument
23241 @samp{foo.txt} and communicate with @value{GDBN} over the serial port
23245 target> gdbserver /dev/com1 emacs foo.txt
23248 @code{gdbserver} waits passively for the host @value{GDBN} to communicate
23251 To use a TCP connection instead of a serial line:
23254 target> gdbserver host:2345 emacs foo.txt
23257 The only difference from the previous example is the first argument,
23258 specifying that you are communicating with the host @value{GDBN} via
23259 TCP. The @samp{host:2345} argument means that @code{gdbserver} is to
23260 expect a TCP connection from machine @samp{host} to local TCP port 2345.
23261 (Currently, the @samp{host} part is ignored.) You can choose any number
23262 you want for the port number as long as it does not conflict with any
23263 TCP ports already in use on the target system (for example, @code{23} is
23264 reserved for @code{telnet}).@footnote{If you choose a port number that
23265 conflicts with another service, @code{gdbserver} prints an error message
23266 and exits.} You must use the same port number with the host @value{GDBN}
23267 @code{target remote} command.
23269 The @code{stdio} connection is useful when starting @code{gdbserver}
23273 (@value{GDBP}) target remote | ssh -T hostname gdbserver - hello
23276 The @samp{-T} option to ssh is provided because we don't need a remote pty,
23277 and we don't want escape-character handling. Ssh does this by default when
23278 a command is provided, the flag is provided to make it explicit.
23279 You could elide it if you want to.
23281 Programs started with stdio-connected gdbserver have @file{/dev/null} for
23282 @code{stdin}, and @code{stdout},@code{stderr} are sent back to gdb for
23283 display through a pipe connected to gdbserver.
23284 Both @code{stdout} and @code{stderr} use the same pipe.
23286 @anchor{Attaching to a program}
23287 @subsubsection Attaching to a Running Program
23288 @cindex attach to a program, @code{gdbserver}
23289 @cindex @option{--attach}, @code{gdbserver} option
23291 On some targets, @code{gdbserver} can also attach to running programs.
23292 This is accomplished via the @code{--attach} argument. The syntax is:
23295 target> gdbserver --attach @var{comm} @var{pid}
23298 @var{pid} is the process ID of a currently running process. It isn't
23299 necessary to point @code{gdbserver} at a binary for the running process.
23301 In @code{target extended-remote} mode, you can also attach using the
23302 @value{GDBN} attach command
23303 (@pxref{Attaching in Types of Remote Connections}).
23306 You can debug processes by name instead of process ID if your target has the
23307 @code{pidof} utility:
23310 target> gdbserver --attach @var{comm} `pidof @var{program}`
23313 In case more than one copy of @var{program} is running, or @var{program}
23314 has multiple threads, most versions of @code{pidof} support the
23315 @code{-s} option to only return the first process ID.
23317 @subsubsection TCP port allocation lifecycle of @code{gdbserver}
23319 This section applies only when @code{gdbserver} is run to listen on a TCP
23322 @code{gdbserver} normally terminates after all of its debugged processes have
23323 terminated in @kbd{target remote} mode. On the other hand, for @kbd{target
23324 extended-remote}, @code{gdbserver} stays running even with no processes left.
23325 @value{GDBN} normally terminates the spawned debugged process on its exit,
23326 which normally also terminates @code{gdbserver} in the @kbd{target remote}
23327 mode. Therefore, when the connection drops unexpectedly, and @value{GDBN}
23328 cannot ask @code{gdbserver} to kill its debugged processes, @code{gdbserver}
23329 stays running even in the @kbd{target remote} mode.
23331 When @code{gdbserver} stays running, @value{GDBN} can connect to it again later.
23332 Such reconnecting is useful for features like @ref{disconnected tracing}. For
23333 completeness, at most one @value{GDBN} can be connected at a time.
23335 @cindex @option{--once}, @code{gdbserver} option
23336 By default, @code{gdbserver} keeps the listening TCP port open, so that
23337 subsequent connections are possible. However, if you start @code{gdbserver}
23338 with the @option{--once} option, it will stop listening for any further
23339 connection attempts after connecting to the first @value{GDBN} session. This
23340 means no further connections to @code{gdbserver} will be possible after the
23341 first one. It also means @code{gdbserver} will terminate after the first
23342 connection with remote @value{GDBN} has closed, even for unexpectedly closed
23343 connections and even in the @kbd{target extended-remote} mode. The
23344 @option{--once} option allows reusing the same port number for connecting to
23345 multiple instances of @code{gdbserver} running on the same host, since each
23346 instance closes its port after the first connection.
23348 @anchor{Other Command-Line Arguments for gdbserver}
23349 @subsubsection Other Command-Line Arguments for @code{gdbserver}
23351 You can use the @option{--multi} option to start @code{gdbserver} without
23352 specifying a program to debug or a process to attach to. Then you can
23353 attach in @code{target extended-remote} mode and run or attach to a
23354 program. For more information,
23355 @pxref{--multi Option in Types of Remote Connnections}.
23357 @cindex @option{--debug}, @code{gdbserver} option
23358 The @option{--debug} option tells @code{gdbserver} to display extra
23359 status information about the debugging process.
23360 @cindex @option{--remote-debug}, @code{gdbserver} option
23361 The @option{--remote-debug} option tells @code{gdbserver} to display
23362 remote protocol debug output.
23363 @cindex @option{--debug-file}, @code{gdbserver} option
23364 @cindex @code{gdbserver}, send all debug output to a single file
23365 The @option{--debug-file=@var{filename}} option tells @code{gdbserver} to
23366 write any debug output to the given @var{filename}. These options are intended
23367 for @code{gdbserver} development and for bug reports to the developers.
23369 @cindex @option{--debug-format}, @code{gdbserver} option
23370 The @option{--debug-format=option1[,option2,...]} option tells
23371 @code{gdbserver} to include additional information in each output.
23372 Possible options are:
23376 Turn off all extra information in debugging output.
23378 Turn on all extra information in debugging output.
23380 Include a timestamp in each line of debugging output.
23383 Options are processed in order. Thus, for example, if @option{none}
23384 appears last then no additional information is added to debugging output.
23386 @cindex @option{--wrapper}, @code{gdbserver} option
23387 The @option{--wrapper} option specifies a wrapper to launch programs
23388 for debugging. The option should be followed by the name of the
23389 wrapper, then any command-line arguments to pass to the wrapper, then
23390 @kbd{--} indicating the end of the wrapper arguments.
23392 @code{gdbserver} runs the specified wrapper program with a combined
23393 command line including the wrapper arguments, then the name of the
23394 program to debug, then any arguments to the program. The wrapper
23395 runs until it executes your program, and then @value{GDBN} gains control.
23397 You can use any program that eventually calls @code{execve} with
23398 its arguments as a wrapper. Several standard Unix utilities do
23399 this, e.g.@: @code{env} and @code{nohup}. Any Unix shell script ending
23400 with @code{exec "$@@"} will also work.
23402 For example, you can use @code{env} to pass an environment variable to
23403 the debugged program, without setting the variable in @code{gdbserver}'s
23407 $ gdbserver --wrapper env LD_PRELOAD=libtest.so -- :2222 ./testprog
23410 @cindex @option{--selftest}
23411 The @option{--selftest} option runs the self tests in @code{gdbserver}:
23414 $ gdbserver --selftest
23415 Ran 2 unit tests, 0 failed
23418 These tests are disabled in release.
23419 @subsection Connecting to @code{gdbserver}
23421 The basic procedure for connecting to the remote target is:
23425 Run @value{GDBN} on the host system.
23428 Make sure you have the necessary symbol files
23429 (@pxref{Host and target files}).
23430 Load symbols for your application using the @code{file} command before you
23431 connect. Use @code{set sysroot} to locate target libraries (unless your
23432 @value{GDBN} was compiled with the correct sysroot using
23433 @code{--with-sysroot}).
23436 Connect to your target (@pxref{Connecting,,Connecting to a Remote Target}).
23437 For TCP connections, you must start up @code{gdbserver} prior to using
23438 the @code{target} command. Otherwise you may get an error whose
23439 text depends on the host system, but which usually looks something like
23440 @samp{Connection refused}. Don't use the @code{load}
23441 command in @value{GDBN} when using @code{target remote} mode, since the
23442 program is already on the target.
23446 @anchor{Monitor Commands for gdbserver}
23447 @subsection Monitor Commands for @code{gdbserver}
23448 @cindex monitor commands, for @code{gdbserver}
23450 During a @value{GDBN} session using @code{gdbserver}, you can use the
23451 @code{monitor} command to send special requests to @code{gdbserver}.
23452 Here are the available commands.
23456 List the available monitor commands.
23458 @item monitor set debug 0
23459 @itemx monitor set debug 1
23460 Disable or enable general debugging messages.
23462 @item monitor set remote-debug 0
23463 @itemx monitor set remote-debug 1
23464 Disable or enable specific debugging messages associated with the remote
23465 protocol (@pxref{Remote Protocol}).
23467 @item monitor set debug-file filename
23468 @itemx monitor set debug-file
23469 Send any debug output to the given file, or to stderr.
23471 @item monitor set debug-format option1@r{[},option2,...@r{]}
23472 Specify additional text to add to debugging messages.
23473 Possible options are:
23477 Turn off all extra information in debugging output.
23479 Turn on all extra information in debugging output.
23481 Include a timestamp in each line of debugging output.
23484 Options are processed in order. Thus, for example, if @option{none}
23485 appears last then no additional information is added to debugging output.
23487 @item monitor set libthread-db-search-path [PATH]
23488 @cindex gdbserver, search path for @code{libthread_db}
23489 When this command is issued, @var{path} is a colon-separated list of
23490 directories to search for @code{libthread_db} (@pxref{Threads,,set
23491 libthread-db-search-path}). If you omit @var{path},
23492 @samp{libthread-db-search-path} will be reset to its default value.
23494 The special entry @samp{$pdir} for @samp{libthread-db-search-path} is
23495 not supported in @code{gdbserver}.
23498 Tell gdbserver to exit immediately. This command should be followed by
23499 @code{disconnect} to close the debugging session. @code{gdbserver} will
23500 detach from any attached processes and kill any processes it created.
23501 Use @code{monitor exit} to terminate @code{gdbserver} at the end
23502 of a multi-process mode debug session.
23506 @subsection Tracepoints support in @code{gdbserver}
23507 @cindex tracepoints support in @code{gdbserver}
23509 On some targets, @code{gdbserver} supports tracepoints, fast
23510 tracepoints and static tracepoints.
23512 For fast or static tracepoints to work, a special library called the
23513 @dfn{in-process agent} (IPA), must be loaded in the inferior process.
23514 This library is built and distributed as an integral part of
23515 @code{gdbserver}. In addition, support for static tracepoints
23516 requires building the in-process agent library with static tracepoints
23517 support. At present, the UST (LTTng Userspace Tracer,
23518 @url{http://lttng.org/ust}) tracing engine is supported. This support
23519 is automatically available if UST development headers are found in the
23520 standard include path when @code{gdbserver} is built, or if
23521 @code{gdbserver} was explicitly configured using @option{--with-ust}
23522 to point at such headers. You can explicitly disable the support
23523 using @option{--with-ust=no}.
23525 There are several ways to load the in-process agent in your program:
23528 @item Specifying it as dependency at link time
23530 You can link your program dynamically with the in-process agent
23531 library. On most systems, this is accomplished by adding
23532 @code{-linproctrace} to the link command.
23534 @item Using the system's preloading mechanisms
23536 You can force loading the in-process agent at startup time by using
23537 your system's support for preloading shared libraries. Many Unixes
23538 support the concept of preloading user defined libraries. In most
23539 cases, you do that by specifying @code{LD_PRELOAD=libinproctrace.so}
23540 in the environment. See also the description of @code{gdbserver}'s
23541 @option{--wrapper} command line option.
23543 @item Using @value{GDBN} to force loading the agent at run time
23545 On some systems, you can force the inferior to load a shared library,
23546 by calling a dynamic loader function in the inferior that takes care
23547 of dynamically looking up and loading a shared library. On most Unix
23548 systems, the function is @code{dlopen}. You'll use the @code{call}
23549 command for that. For example:
23552 (@value{GDBP}) call dlopen ("libinproctrace.so", ...)
23555 Note that on most Unix systems, for the @code{dlopen} function to be
23556 available, the program needs to be linked with @code{-ldl}.
23559 On systems that have a userspace dynamic loader, like most Unix
23560 systems, when you connect to @code{gdbserver} using @code{target
23561 remote}, you'll find that the program is stopped at the dynamic
23562 loader's entry point, and no shared library has been loaded in the
23563 program's address space yet, including the in-process agent. In that
23564 case, before being able to use any of the fast or static tracepoints
23565 features, you need to let the loader run and load the shared
23566 libraries. The simplest way to do that is to run the program to the
23567 main procedure. E.g., if debugging a C or C@t{++} program, start
23568 @code{gdbserver} like so:
23571 $ gdbserver :9999 myprogram
23574 Start GDB and connect to @code{gdbserver} like so, and run to main:
23578 (@value{GDBP}) target remote myhost:9999
23579 0x00007f215893ba60 in ?? () from /lib64/ld-linux-x86-64.so.2
23580 (@value{GDBP}) b main
23581 (@value{GDBP}) continue
23584 The in-process tracing agent library should now be loaded into the
23585 process; you can confirm it with the @code{info sharedlibrary}
23586 command, which will list @file{libinproctrace.so} as loaded in the
23587 process. You are now ready to install fast tracepoints, list static
23588 tracepoint markers, probe static tracepoints markers, and start
23591 @node Remote Configuration
23592 @section Remote Configuration
23595 @kindex show remote
23596 This section documents the configuration options available when
23597 debugging remote programs. For the options related to the File I/O
23598 extensions of the remote protocol, see @ref{system,
23599 system-call-allowed}.
23602 @item set remoteaddresssize @var{bits}
23603 @cindex address size for remote targets
23604 @cindex bits in remote address
23605 Set the maximum size of address in a memory packet to the specified
23606 number of bits. @value{GDBN} will mask off the address bits above
23607 that number, when it passes addresses to the remote target. The
23608 default value is the number of bits in the target's address.
23610 @item show remoteaddresssize
23611 Show the current value of remote address size in bits.
23613 @item set serial baud @var{n}
23614 @cindex baud rate for remote targets
23615 Set the baud rate for the remote serial I/O to @var{n} baud. The
23616 value is used to set the speed of the serial port used for debugging
23619 @item show serial baud
23620 Show the current speed of the remote connection.
23622 @item set serial parity @var{parity}
23623 Set the parity for the remote serial I/O. Supported values of @var{parity} are:
23624 @code{even}, @code{none}, and @code{odd}. The default is @code{none}.
23626 @item show serial parity
23627 Show the current parity of the serial port.
23629 @item set remotebreak
23630 @cindex interrupt remote programs
23631 @cindex BREAK signal instead of Ctrl-C
23632 @anchor{set remotebreak}
23633 If set to on, @value{GDBN} sends a @code{BREAK} signal to the remote
23634 when you type @kbd{Ctrl-c} to interrupt the program running
23635 on the remote. If set to off, @value{GDBN} sends the @samp{Ctrl-C}
23636 character instead. The default is off, since most remote systems
23637 expect to see @samp{Ctrl-C} as the interrupt signal.
23639 @item show remotebreak
23640 Show whether @value{GDBN} sends @code{BREAK} or @samp{Ctrl-C} to
23641 interrupt the remote program.
23643 @item set remoteflow on
23644 @itemx set remoteflow off
23645 @kindex set remoteflow
23646 Enable or disable hardware flow control (@code{RTS}/@code{CTS})
23647 on the serial port used to communicate to the remote target.
23649 @item show remoteflow
23650 @kindex show remoteflow
23651 Show the current setting of hardware flow control.
23653 @item set remotelogbase @var{base}
23654 Set the base (a.k.a.@: radix) of logging serial protocol
23655 communications to @var{base}. Supported values of @var{base} are:
23656 @code{ascii}, @code{octal}, and @code{hex}. The default is
23659 @item show remotelogbase
23660 Show the current setting of the radix for logging remote serial
23663 @item set remotelogfile @var{file}
23664 @cindex record serial communications on file
23665 Record remote serial communications on the named @var{file}. The
23666 default is not to record at all.
23668 @item show remotelogfile
23669 Show the current setting of the file name on which to record the
23670 serial communications.
23672 @item set remotetimeout @var{num}
23673 @cindex timeout for serial communications
23674 @cindex remote timeout
23675 Set the timeout limit to wait for the remote target to respond to
23676 @var{num} seconds. The default is 2 seconds.
23678 @item show remotetimeout
23679 Show the current number of seconds to wait for the remote target
23682 @cindex limit hardware breakpoints and watchpoints
23683 @cindex remote target, limit break- and watchpoints
23684 @anchor{set remote hardware-watchpoint-limit}
23685 @anchor{set remote hardware-breakpoint-limit}
23686 @item set remote hardware-watchpoint-limit @var{limit}
23687 @itemx set remote hardware-breakpoint-limit @var{limit}
23688 Restrict @value{GDBN} to using @var{limit} remote hardware watchpoints
23689 or breakpoints. The @var{limit} can be set to 0 to disable hardware
23690 watchpoints or breakpoints, and @code{unlimited} for unlimited
23691 watchpoints or breakpoints.
23693 @item show remote hardware-watchpoint-limit
23694 @itemx show remote hardware-breakpoint-limit
23695 Show the current limit for the number of hardware watchpoints or
23696 breakpoints that @value{GDBN} can use.
23698 @cindex limit hardware watchpoints length
23699 @cindex remote target, limit watchpoints length
23700 @anchor{set remote hardware-watchpoint-length-limit}
23701 @item set remote hardware-watchpoint-length-limit @var{limit}
23702 Restrict @value{GDBN} to using @var{limit} bytes for the maximum
23703 length of a remote hardware watchpoint. A @var{limit} of 0 disables
23704 hardware watchpoints and @code{unlimited} allows watchpoints of any
23707 @item show remote hardware-watchpoint-length-limit
23708 Show the current limit (in bytes) of the maximum length of
23709 a remote hardware watchpoint.
23711 @item set remote exec-file @var{filename}
23712 @itemx show remote exec-file
23713 @anchor{set remote exec-file}
23714 @cindex executable file, for remote target
23715 Select the file used for @code{run} with @code{target
23716 extended-remote}. This should be set to a filename valid on the
23717 target system. If it is not set, the target will use a default
23718 filename (e.g.@: the last program run).
23720 @item set remote interrupt-sequence
23721 @cindex interrupt remote programs
23722 @cindex select Ctrl-C, BREAK or BREAK-g
23723 Allow the user to select one of @samp{Ctrl-C}, a @code{BREAK} or
23724 @samp{BREAK-g} as the
23725 sequence to the remote target in order to interrupt the execution.
23726 @samp{Ctrl-C} is a default. Some system prefers @code{BREAK} which
23727 is high level of serial line for some certain time.
23728 Linux kernel prefers @samp{BREAK-g}, a.k.a Magic SysRq g.
23729 It is @code{BREAK} signal followed by character @code{g}.
23731 @item show interrupt-sequence
23732 Show which of @samp{Ctrl-C}, @code{BREAK} or @code{BREAK-g}
23733 is sent by @value{GDBN} to interrupt the remote program.
23734 @code{BREAK-g} is BREAK signal followed by @code{g} and
23735 also known as Magic SysRq g.
23737 @item set remote interrupt-on-connect
23738 @cindex send interrupt-sequence on start
23739 Specify whether interrupt-sequence is sent to remote target when
23740 @value{GDBN} connects to it. This is mostly needed when you debug
23741 Linux kernel. Linux kernel expects @code{BREAK} followed by @code{g}
23742 which is known as Magic SysRq g in order to connect @value{GDBN}.
23744 @item show interrupt-on-connect
23745 Show whether interrupt-sequence is sent
23746 to remote target when @value{GDBN} connects to it.
23750 @item set tcp auto-retry on
23751 @cindex auto-retry, for remote TCP target
23752 Enable auto-retry for remote TCP connections. This is useful if the remote
23753 debugging agent is launched in parallel with @value{GDBN}; there is a race
23754 condition because the agent may not become ready to accept the connection
23755 before @value{GDBN} attempts to connect. When auto-retry is
23756 enabled, if the initial attempt to connect fails, @value{GDBN} reattempts
23757 to establish the connection using the timeout specified by
23758 @code{set tcp connect-timeout}.
23760 @item set tcp auto-retry off
23761 Do not auto-retry failed TCP connections.
23763 @item show tcp auto-retry
23764 Show the current auto-retry setting.
23766 @item set tcp connect-timeout @var{seconds}
23767 @itemx set tcp connect-timeout unlimited
23768 @cindex connection timeout, for remote TCP target
23769 @cindex timeout, for remote target connection
23770 Set the timeout for establishing a TCP connection to the remote target to
23771 @var{seconds}. The timeout affects both polling to retry failed connections
23772 (enabled by @code{set tcp auto-retry on}) and waiting for connections
23773 that are merely slow to complete, and represents an approximate cumulative
23774 value. If @var{seconds} is @code{unlimited}, there is no timeout and
23775 @value{GDBN} will keep attempting to establish a connection forever,
23776 unless interrupted with @kbd{Ctrl-c}. The default is 15 seconds.
23778 @item show tcp connect-timeout
23779 Show the current connection timeout setting.
23782 @cindex remote packets, enabling and disabling
23783 The @value{GDBN} remote protocol autodetects the packets supported by
23784 your debugging stub. If you need to override the autodetection, you
23785 can use these commands to enable or disable individual packets. Each
23786 packet can be set to @samp{on} (the remote target supports this
23787 packet), @samp{off} (the remote target does not support this packet),
23788 or @samp{auto} (detect remote target support for this packet). They
23789 all default to @samp{auto}. For more information about each packet,
23790 see @ref{Remote Protocol}.
23792 During normal use, you should not have to use any of these commands.
23793 If you do, that may be a bug in your remote debugging stub, or a bug
23794 in @value{GDBN}. You may want to report the problem to the
23795 @value{GDBN} developers.
23797 For each packet @var{name}, the command to enable or disable the
23798 packet is @code{set remote @var{name}-packet}. The available settings
23801 @multitable @columnfractions 0.28 0.32 0.25
23804 @tab Related Features
23806 @item @code{fetch-register}
23808 @tab @code{info registers}
23810 @item @code{set-register}
23814 @item @code{binary-download}
23816 @tab @code{load}, @code{set}
23818 @item @code{read-aux-vector}
23819 @tab @code{qXfer:auxv:read}
23820 @tab @code{info auxv}
23822 @item @code{symbol-lookup}
23823 @tab @code{qSymbol}
23824 @tab Detecting multiple threads
23826 @item @code{attach}
23827 @tab @code{vAttach}
23830 @item @code{verbose-resume}
23832 @tab Stepping or resuming multiple threads
23838 @item @code{software-breakpoint}
23842 @item @code{hardware-breakpoint}
23846 @item @code{write-watchpoint}
23850 @item @code{read-watchpoint}
23854 @item @code{access-watchpoint}
23858 @item @code{pid-to-exec-file}
23859 @tab @code{qXfer:exec-file:read}
23860 @tab @code{attach}, @code{run}
23862 @item @code{target-features}
23863 @tab @code{qXfer:features:read}
23864 @tab @code{set architecture}
23866 @item @code{library-info}
23867 @tab @code{qXfer:libraries:read}
23868 @tab @code{info sharedlibrary}
23870 @item @code{memory-map}
23871 @tab @code{qXfer:memory-map:read}
23872 @tab @code{info mem}
23874 @item @code{read-sdata-object}
23875 @tab @code{qXfer:sdata:read}
23876 @tab @code{print $_sdata}
23878 @item @code{read-siginfo-object}
23879 @tab @code{qXfer:siginfo:read}
23880 @tab @code{print $_siginfo}
23882 @item @code{write-siginfo-object}
23883 @tab @code{qXfer:siginfo:write}
23884 @tab @code{set $_siginfo}
23886 @item @code{threads}
23887 @tab @code{qXfer:threads:read}
23888 @tab @code{info threads}
23890 @item @code{get-thread-local-@*storage-address}
23891 @tab @code{qGetTLSAddr}
23892 @tab Displaying @code{__thread} variables
23894 @item @code{get-thread-information-block-address}
23895 @tab @code{qGetTIBAddr}
23896 @tab Display MS-Windows Thread Information Block.
23898 @item @code{search-memory}
23899 @tab @code{qSearch:memory}
23902 @item @code{supported-packets}
23903 @tab @code{qSupported}
23904 @tab Remote communications parameters
23906 @item @code{catch-syscalls}
23907 @tab @code{QCatchSyscalls}
23908 @tab @code{catch syscall}
23910 @item @code{pass-signals}
23911 @tab @code{QPassSignals}
23912 @tab @code{handle @var{signal}}
23914 @item @code{program-signals}
23915 @tab @code{QProgramSignals}
23916 @tab @code{handle @var{signal}}
23918 @item @code{hostio-close-packet}
23919 @tab @code{vFile:close}
23920 @tab @code{remote get}, @code{remote put}
23922 @item @code{hostio-open-packet}
23923 @tab @code{vFile:open}
23924 @tab @code{remote get}, @code{remote put}
23926 @item @code{hostio-pread-packet}
23927 @tab @code{vFile:pread}
23928 @tab @code{remote get}, @code{remote put}
23930 @item @code{hostio-pwrite-packet}
23931 @tab @code{vFile:pwrite}
23932 @tab @code{remote get}, @code{remote put}
23934 @item @code{hostio-unlink-packet}
23935 @tab @code{vFile:unlink}
23936 @tab @code{remote delete}
23938 @item @code{hostio-readlink-packet}
23939 @tab @code{vFile:readlink}
23942 @item @code{hostio-fstat-packet}
23943 @tab @code{vFile:fstat}
23946 @item @code{hostio-setfs-packet}
23947 @tab @code{vFile:setfs}
23950 @item @code{noack-packet}
23951 @tab @code{QStartNoAckMode}
23952 @tab Packet acknowledgment
23954 @item @code{osdata}
23955 @tab @code{qXfer:osdata:read}
23956 @tab @code{info os}
23958 @item @code{query-attached}
23959 @tab @code{qAttached}
23960 @tab Querying remote process attach state.
23962 @item @code{trace-buffer-size}
23963 @tab @code{QTBuffer:size}
23964 @tab @code{set trace-buffer-size}
23966 @item @code{trace-status}
23967 @tab @code{qTStatus}
23968 @tab @code{tstatus}
23970 @item @code{traceframe-info}
23971 @tab @code{qXfer:traceframe-info:read}
23972 @tab Traceframe info
23974 @item @code{install-in-trace}
23975 @tab @code{InstallInTrace}
23976 @tab Install tracepoint in tracing
23978 @item @code{disable-randomization}
23979 @tab @code{QDisableRandomization}
23980 @tab @code{set disable-randomization}
23982 @item @code{startup-with-shell}
23983 @tab @code{QStartupWithShell}
23984 @tab @code{set startup-with-shell}
23986 @item @code{environment-hex-encoded}
23987 @tab @code{QEnvironmentHexEncoded}
23988 @tab @code{set environment}
23990 @item @code{environment-unset}
23991 @tab @code{QEnvironmentUnset}
23992 @tab @code{unset environment}
23994 @item @code{environment-reset}
23995 @tab @code{QEnvironmentReset}
23996 @tab @code{Reset the inferior environment (i.e., unset user-set variables)}
23998 @item @code{set-working-dir}
23999 @tab @code{QSetWorkingDir}
24000 @tab @code{set cwd}
24002 @item @code{conditional-breakpoints-packet}
24003 @tab @code{Z0 and Z1}
24004 @tab @code{Support for target-side breakpoint condition evaluation}
24006 @item @code{multiprocess-extensions}
24007 @tab @code{multiprocess extensions}
24008 @tab Debug multiple processes and remote process PID awareness
24010 @item @code{swbreak-feature}
24011 @tab @code{swbreak stop reason}
24014 @item @code{hwbreak-feature}
24015 @tab @code{hwbreak stop reason}
24018 @item @code{fork-event-feature}
24019 @tab @code{fork stop reason}
24022 @item @code{vfork-event-feature}
24023 @tab @code{vfork stop reason}
24026 @item @code{exec-event-feature}
24027 @tab @code{exec stop reason}
24030 @item @code{thread-events}
24031 @tab @code{QThreadEvents}
24032 @tab Tracking thread lifetime.
24034 @item @code{no-resumed-stop-reply}
24035 @tab @code{no resumed thread left stop reply}
24036 @tab Tracking thread lifetime.
24041 @section Implementing a Remote Stub
24043 @cindex debugging stub, example
24044 @cindex remote stub, example
24045 @cindex stub example, remote debugging
24046 The stub files provided with @value{GDBN} implement the target side of the
24047 communication protocol, and the @value{GDBN} side is implemented in the
24048 @value{GDBN} source file @file{remote.c}. Normally, you can simply allow
24049 these subroutines to communicate, and ignore the details. (If you're
24050 implementing your own stub file, you can still ignore the details: start
24051 with one of the existing stub files. @file{sparc-stub.c} is the best
24052 organized, and therefore the easiest to read.)
24054 @cindex remote serial debugging, overview
24055 To debug a program running on another machine (the debugging
24056 @dfn{target} machine), you must first arrange for all the usual
24057 prerequisites for the program to run by itself. For example, for a C
24062 A startup routine to set up the C runtime environment; these usually
24063 have a name like @file{crt0}. The startup routine may be supplied by
24064 your hardware supplier, or you may have to write your own.
24067 A C subroutine library to support your program's
24068 subroutine calls, notably managing input and output.
24071 A way of getting your program to the other machine---for example, a
24072 download program. These are often supplied by the hardware
24073 manufacturer, but you may have to write your own from hardware
24077 The next step is to arrange for your program to use a serial port to
24078 communicate with the machine where @value{GDBN} is running (the @dfn{host}
24079 machine). In general terms, the scheme looks like this:
24083 @value{GDBN} already understands how to use this protocol; when everything
24084 else is set up, you can simply use the @samp{target remote} command
24085 (@pxref{Targets,,Specifying a Debugging Target}).
24087 @item On the target,
24088 you must link with your program a few special-purpose subroutines that
24089 implement the @value{GDBN} remote serial protocol. The file containing these
24090 subroutines is called a @dfn{debugging stub}.
24092 On certain remote targets, you can use an auxiliary program
24093 @code{gdbserver} instead of linking a stub into your program.
24094 @xref{Server,,Using the @code{gdbserver} Program}, for details.
24097 The debugging stub is specific to the architecture of the remote
24098 machine; for example, use @file{sparc-stub.c} to debug programs on
24101 @cindex remote serial stub list
24102 These working remote stubs are distributed with @value{GDBN}:
24107 @cindex @file{i386-stub.c}
24110 For Intel 386 and compatible architectures.
24113 @cindex @file{m68k-stub.c}
24114 @cindex Motorola 680x0
24116 For Motorola 680x0 architectures.
24119 @cindex @file{sh-stub.c}
24122 For Renesas SH architectures.
24125 @cindex @file{sparc-stub.c}
24127 For @sc{sparc} architectures.
24129 @item sparcl-stub.c
24130 @cindex @file{sparcl-stub.c}
24133 For Fujitsu @sc{sparclite} architectures.
24137 The @file{README} file in the @value{GDBN} distribution may list other
24138 recently added stubs.
24141 * Stub Contents:: What the stub can do for you
24142 * Bootstrapping:: What you must do for the stub
24143 * Debug Session:: Putting it all together
24146 @node Stub Contents
24147 @subsection What the Stub Can Do for You
24149 @cindex remote serial stub
24150 The debugging stub for your architecture supplies these three
24154 @item set_debug_traps
24155 @findex set_debug_traps
24156 @cindex remote serial stub, initialization
24157 This routine arranges for @code{handle_exception} to run when your
24158 program stops. You must call this subroutine explicitly in your
24159 program's startup code.
24161 @item handle_exception
24162 @findex handle_exception
24163 @cindex remote serial stub, main routine
24164 This is the central workhorse, but your program never calls it
24165 explicitly---the setup code arranges for @code{handle_exception} to
24166 run when a trap is triggered.
24168 @code{handle_exception} takes control when your program stops during
24169 execution (for example, on a breakpoint), and mediates communications
24170 with @value{GDBN} on the host machine. This is where the communications
24171 protocol is implemented; @code{handle_exception} acts as the @value{GDBN}
24172 representative on the target machine. It begins by sending summary
24173 information on the state of your program, then continues to execute,
24174 retrieving and transmitting any information @value{GDBN} needs, until you
24175 execute a @value{GDBN} command that makes your program resume; at that point,
24176 @code{handle_exception} returns control to your own code on the target
24180 @cindex @code{breakpoint} subroutine, remote
24181 Use this auxiliary subroutine to make your program contain a
24182 breakpoint. Depending on the particular situation, this may be the only
24183 way for @value{GDBN} to get control. For instance, if your target
24184 machine has some sort of interrupt button, you won't need to call this;
24185 pressing the interrupt button transfers control to
24186 @code{handle_exception}---in effect, to @value{GDBN}. On some machines,
24187 simply receiving characters on the serial port may also trigger a trap;
24188 again, in that situation, you don't need to call @code{breakpoint} from
24189 your own program---simply running @samp{target remote} from the host
24190 @value{GDBN} session gets control.
24192 Call @code{breakpoint} if none of these is true, or if you simply want
24193 to make certain your program stops at a predetermined point for the
24194 start of your debugging session.
24197 @node Bootstrapping
24198 @subsection What You Must Do for the Stub
24200 @cindex remote stub, support routines
24201 The debugging stubs that come with @value{GDBN} are set up for a particular
24202 chip architecture, but they have no information about the rest of your
24203 debugging target machine.
24205 First of all you need to tell the stub how to communicate with the
24209 @item int getDebugChar()
24210 @findex getDebugChar
24211 Write this subroutine to read a single character from the serial port.
24212 It may be identical to @code{getchar} for your target system; a
24213 different name is used to allow you to distinguish the two if you wish.
24215 @item void putDebugChar(int)
24216 @findex putDebugChar
24217 Write this subroutine to write a single character to the serial port.
24218 It may be identical to @code{putchar} for your target system; a
24219 different name is used to allow you to distinguish the two if you wish.
24222 @cindex control C, and remote debugging
24223 @cindex interrupting remote targets
24224 If you want @value{GDBN} to be able to stop your program while it is
24225 running, you need to use an interrupt-driven serial driver, and arrange
24226 for it to stop when it receives a @code{^C} (@samp{\003}, the control-C
24227 character). That is the character which @value{GDBN} uses to tell the
24228 remote system to stop.
24230 Getting the debugging target to return the proper status to @value{GDBN}
24231 probably requires changes to the standard stub; one quick and dirty way
24232 is to just execute a breakpoint instruction (the ``dirty'' part is that
24233 @value{GDBN} reports a @code{SIGTRAP} instead of a @code{SIGINT}).
24235 Other routines you need to supply are:
24238 @item void exceptionHandler (int @var{exception_number}, void *@var{exception_address})
24239 @findex exceptionHandler
24240 Write this function to install @var{exception_address} in the exception
24241 handling tables. You need to do this because the stub does not have any
24242 way of knowing what the exception handling tables on your target system
24243 are like (for example, the processor's table might be in @sc{rom},
24244 containing entries which point to a table in @sc{ram}).
24245 The @var{exception_number} specifies the exception which should be changed;
24246 its meaning is architecture-dependent (for example, different numbers
24247 might represent divide by zero, misaligned access, etc). When this
24248 exception occurs, control should be transferred directly to
24249 @var{exception_address}, and the processor state (stack, registers,
24250 and so on) should be just as it is when a processor exception occurs. So if
24251 you want to use a jump instruction to reach @var{exception_address}, it
24252 should be a simple jump, not a jump to subroutine.
24254 For the 386, @var{exception_address} should be installed as an interrupt
24255 gate so that interrupts are masked while the handler runs. The gate
24256 should be at privilege level 0 (the most privileged level). The
24257 @sc{sparc} and 68k stubs are able to mask interrupts themselves without
24258 help from @code{exceptionHandler}.
24260 @item void flush_i_cache()
24261 @findex flush_i_cache
24262 On @sc{sparc} and @sc{sparclite} only, write this subroutine to flush the
24263 instruction cache, if any, on your target machine. If there is no
24264 instruction cache, this subroutine may be a no-op.
24266 On target machines that have instruction caches, @value{GDBN} requires this
24267 function to make certain that the state of your program is stable.
24271 You must also make sure this library routine is available:
24274 @item void *memset(void *, int, int)
24276 This is the standard library function @code{memset} that sets an area of
24277 memory to a known value. If you have one of the free versions of
24278 @code{libc.a}, @code{memset} can be found there; otherwise, you must
24279 either obtain it from your hardware manufacturer, or write your own.
24282 If you do not use the GNU C compiler, you may need other standard
24283 library subroutines as well; this varies from one stub to another,
24284 but in general the stubs are likely to use any of the common library
24285 subroutines which @code{@value{NGCC}} generates as inline code.
24288 @node Debug Session
24289 @subsection Putting it All Together
24291 @cindex remote serial debugging summary
24292 In summary, when your program is ready to debug, you must follow these
24297 Make sure you have defined the supporting low-level routines
24298 (@pxref{Bootstrapping,,What You Must Do for the Stub}):
24300 @code{getDebugChar}, @code{putDebugChar},
24301 @code{flush_i_cache}, @code{memset}, @code{exceptionHandler}.
24305 Insert these lines in your program's startup code, before the main
24306 procedure is called:
24313 On some machines, when a breakpoint trap is raised, the hardware
24314 automatically makes the PC point to the instruction after the
24315 breakpoint. If your machine doesn't do that, you may need to adjust
24316 @code{handle_exception} to arrange for it to return to the instruction
24317 after the breakpoint on this first invocation, so that your program
24318 doesn't keep hitting the initial breakpoint instead of making
24322 For the 680x0 stub only, you need to provide a variable called
24323 @code{exceptionHook}. Normally you just use:
24326 void (*exceptionHook)() = 0;
24330 but if before calling @code{set_debug_traps}, you set it to point to a
24331 function in your program, that function is called when
24332 @code{@value{GDBN}} continues after stopping on a trap (for example, bus
24333 error). The function indicated by @code{exceptionHook} is called with
24334 one parameter: an @code{int} which is the exception number.
24337 Compile and link together: your program, the @value{GDBN} debugging stub for
24338 your target architecture, and the supporting subroutines.
24341 Make sure you have a serial connection between your target machine and
24342 the @value{GDBN} host, and identify the serial port on the host.
24345 @c The "remote" target now provides a `load' command, so we should
24346 @c document that. FIXME.
24347 Download your program to your target machine (or get it there by
24348 whatever means the manufacturer provides), and start it.
24351 Start @value{GDBN} on the host, and connect to the target
24352 (@pxref{Connecting,,Connecting to a Remote Target}).
24356 @node Configurations
24357 @chapter Configuration-Specific Information
24359 While nearly all @value{GDBN} commands are available for all native and
24360 cross versions of the debugger, there are some exceptions. This chapter
24361 describes things that are only available in certain configurations.
24363 There are three major categories of configurations: native
24364 configurations, where the host and target are the same, embedded
24365 operating system configurations, which are usually the same for several
24366 different processor architectures, and bare embedded processors, which
24367 are quite different from each other.
24372 * Embedded Processors::
24379 This section describes details specific to particular native
24383 * BSD libkvm Interface:: Debugging BSD kernel memory images
24384 * Process Information:: Process information
24385 * DJGPP Native:: Features specific to the DJGPP port
24386 * Cygwin Native:: Features specific to the Cygwin port
24387 * Hurd Native:: Features specific to @sc{gnu} Hurd
24388 * Darwin:: Features specific to Darwin
24389 * FreeBSD:: Features specific to FreeBSD
24392 @node BSD libkvm Interface
24393 @subsection BSD libkvm Interface
24396 @cindex kernel memory image
24397 @cindex kernel crash dump
24399 BSD-derived systems (FreeBSD/NetBSD/OpenBSD) have a kernel memory
24400 interface that provides a uniform interface for accessing kernel virtual
24401 memory images, including live systems and crash dumps. @value{GDBN}
24402 uses this interface to allow you to debug live kernels and kernel crash
24403 dumps on many native BSD configurations. This is implemented as a
24404 special @code{kvm} debugging target. For debugging a live system, load
24405 the currently running kernel into @value{GDBN} and connect to the
24409 (@value{GDBP}) @b{target kvm}
24412 For debugging crash dumps, provide the file name of the crash dump as an
24416 (@value{GDBP}) @b{target kvm /var/crash/bsd.0}
24419 Once connected to the @code{kvm} target, the following commands are
24425 Set current context from the @dfn{Process Control Block} (PCB) address.
24428 Set current context from proc address. This command isn't available on
24429 modern FreeBSD systems.
24432 @node Process Information
24433 @subsection Process Information
24435 @cindex examine process image
24436 @cindex process info via @file{/proc}
24438 Some operating systems provide interfaces to fetch additional
24439 information about running processes beyond memory and per-thread
24440 register state. If @value{GDBN} is configured for an operating system
24441 with a supported interface, the command @code{info proc} is available
24442 to report information about the process running your program, or about
24443 any process running on your system.
24445 One supported interface is a facility called @samp{/proc} that can be
24446 used to examine the image of a running process using file-system
24447 subroutines. This facility is supported on @sc{gnu}/Linux and Solaris
24450 On FreeBSD systems, system control nodes are used to query process
24453 In addition, some systems may provide additional process information
24454 in core files. Note that a core file may include a subset of the
24455 information available from a live process. Process information is
24456 currently available from cores created on @sc{gnu}/Linux and FreeBSD
24463 @itemx info proc @var{process-id}
24464 Summarize available information about a process. If a
24465 process ID is specified by @var{process-id}, display information about
24466 that process; otherwise display information about the program being
24467 debugged. The summary includes the debugged process ID, the command
24468 line used to invoke it, its current working directory, and its
24469 executable file's absolute file name.
24471 On some systems, @var{process-id} can be of the form
24472 @samp{[@var{pid}]/@var{tid}} which specifies a certain thread ID
24473 within a process. If the optional @var{pid} part is missing, it means
24474 a thread from the process being debugged (the leading @samp{/} still
24475 needs to be present, or else @value{GDBN} will interpret the number as
24476 a process ID rather than a thread ID).
24478 @item info proc cmdline
24479 @cindex info proc cmdline
24480 Show the original command line of the process. This command is
24481 supported on @sc{gnu}/Linux and FreeBSD.
24483 @item info proc cwd
24484 @cindex info proc cwd
24485 Show the current working directory of the process. This command is
24486 supported on @sc{gnu}/Linux and FreeBSD.
24488 @item info proc exe
24489 @cindex info proc exe
24490 Show the name of executable of the process. This command is supported
24491 on @sc{gnu}/Linux and FreeBSD.
24493 @item info proc files
24494 @cindex info proc files
24495 Show the file descriptors open by the process. For each open file
24496 descriptor, @value{GDBN} shows its number, type (file, directory,
24497 character device, socket), file pointer offset, and the name of the
24498 resource open on the descriptor. The resource name can be a file name
24499 (for files, directories, and devices) or a protocol followed by socket
24500 address (for network connections). This command is supported on
24503 This example shows the open file descriptors for a process using a
24504 tty for standard input and output as well as two network sockets:
24507 (@value{GDBP}) info proc files 22136
24511 FD Type Offset Flags Name
24512 text file - r-------- /usr/bin/ssh
24513 ctty chr - rw------- /dev/pts/20
24514 cwd dir - r-------- /usr/home/john
24515 root dir - r-------- /
24516 0 chr 0x32933a4 rw------- /dev/pts/20
24517 1 chr 0x32933a4 rw------- /dev/pts/20
24518 2 chr 0x32933a4 rw------- /dev/pts/20
24519 3 socket 0x0 rw----n-- tcp4 10.0.1.2:53014 -> 10.0.1.10:22
24520 4 socket 0x0 rw------- unix stream:/tmp/ssh-FIt89oAzOn5f/agent.2456
24523 @item info proc mappings
24524 @cindex memory address space mappings
24525 Report the memory address space ranges accessible in a process. On
24526 Solaris and FreeBSD systems, each memory range includes information on
24527 whether the process has read, write, or execute access rights to each
24528 range. On @sc{gnu}/Linux and FreeBSD systems, each memory range
24529 includes the object file which is mapped to that range.
24531 @item info proc stat
24532 @itemx info proc status
24533 @cindex process detailed status information
24534 Show additional process-related information, including the user ID and
24535 group ID; virtual memory usage; the signals that are pending, blocked,
24536 and ignored; its TTY; its consumption of system and user time; its
24537 stack size; its @samp{nice} value; etc. These commands are supported
24538 on @sc{gnu}/Linux and FreeBSD.
24540 For @sc{gnu}/Linux systems, see the @samp{proc} man page for more
24541 information (type @kbd{man 5 proc} from your shell prompt).
24543 For FreeBSD systems, @code{info proc stat} is an alias for @code{info
24546 @item info proc all
24547 Show all the information about the process described under all of the
24548 above @code{info proc} subcommands.
24551 @comment These sub-options of 'info proc' were not included when
24552 @comment procfs.c was re-written. Keep their descriptions around
24553 @comment against the day when someone finds the time to put them back in.
24554 @kindex info proc times
24555 @item info proc times
24556 Starting time, user CPU time, and system CPU time for your program and
24559 @kindex info proc id
24561 Report on the process IDs related to your program: its own process ID,
24562 the ID of its parent, the process group ID, and the session ID.
24565 @item set procfs-trace
24566 @kindex set procfs-trace
24567 @cindex @code{procfs} API calls
24568 This command enables and disables tracing of @code{procfs} API calls.
24570 @item show procfs-trace
24571 @kindex show procfs-trace
24572 Show the current state of @code{procfs} API call tracing.
24574 @item set procfs-file @var{file}
24575 @kindex set procfs-file
24576 Tell @value{GDBN} to write @code{procfs} API trace to the named
24577 @var{file}. @value{GDBN} appends the trace info to the previous
24578 contents of the file. The default is to display the trace on the
24581 @item show procfs-file
24582 @kindex show procfs-file
24583 Show the file to which @code{procfs} API trace is written.
24585 @item proc-trace-entry
24586 @itemx proc-trace-exit
24587 @itemx proc-untrace-entry
24588 @itemx proc-untrace-exit
24589 @kindex proc-trace-entry
24590 @kindex proc-trace-exit
24591 @kindex proc-untrace-entry
24592 @kindex proc-untrace-exit
24593 These commands enable and disable tracing of entries into and exits
24594 from the @code{syscall} interface.
24597 @kindex info pidlist
24598 @cindex process list, QNX Neutrino
24599 For QNX Neutrino only, this command displays the list of all the
24600 processes and all the threads within each process.
24603 @kindex info meminfo
24604 @cindex mapinfo list, QNX Neutrino
24605 For QNX Neutrino only, this command displays the list of all mapinfos.
24609 @subsection Features for Debugging @sc{djgpp} Programs
24610 @cindex @sc{djgpp} debugging
24611 @cindex native @sc{djgpp} debugging
24612 @cindex MS-DOS-specific commands
24615 @sc{djgpp} is a port of the @sc{gnu} development tools to MS-DOS and
24616 MS-Windows. @sc{djgpp} programs are 32-bit protected-mode programs
24617 that use the @dfn{DPMI} (DOS Protected-Mode Interface) API to run on
24618 top of real-mode DOS systems and their emulations.
24620 @value{GDBN} supports native debugging of @sc{djgpp} programs, and
24621 defines a few commands specific to the @sc{djgpp} port. This
24622 subsection describes those commands.
24627 This is a prefix of @sc{djgpp}-specific commands which print
24628 information about the target system and important OS structures.
24631 @cindex MS-DOS system info
24632 @cindex free memory information (MS-DOS)
24633 @item info dos sysinfo
24634 This command displays assorted information about the underlying
24635 platform: the CPU type and features, the OS version and flavor, the
24636 DPMI version, and the available conventional and DPMI memory.
24641 @cindex segment descriptor tables
24642 @cindex descriptor tables display
24644 @itemx info dos ldt
24645 @itemx info dos idt
24646 These 3 commands display entries from, respectively, Global, Local,
24647 and Interrupt Descriptor Tables (GDT, LDT, and IDT). The descriptor
24648 tables are data structures which store a descriptor for each segment
24649 that is currently in use. The segment's selector is an index into a
24650 descriptor table; the table entry for that index holds the
24651 descriptor's base address and limit, and its attributes and access
24654 A typical @sc{djgpp} program uses 3 segments: a code segment, a data
24655 segment (used for both data and the stack), and a DOS segment (which
24656 allows access to DOS/BIOS data structures and absolute addresses in
24657 conventional memory). However, the DPMI host will usually define
24658 additional segments in order to support the DPMI environment.
24660 @cindex garbled pointers
24661 These commands allow to display entries from the descriptor tables.
24662 Without an argument, all entries from the specified table are
24663 displayed. An argument, which should be an integer expression, means
24664 display a single entry whose index is given by the argument. For
24665 example, here's a convenient way to display information about the
24666 debugged program's data segment:
24669 @exdent @code{(@value{GDBP}) info dos ldt $ds}
24670 @exdent @code{0x13f: base=0x11970000 limit=0x0009ffff 32-Bit Data (Read/Write, Exp-up)}
24674 This comes in handy when you want to see whether a pointer is outside
24675 the data segment's limit (i.e.@: @dfn{garbled}).
24677 @cindex page tables display (MS-DOS)
24679 @itemx info dos pte
24680 These two commands display entries from, respectively, the Page
24681 Directory and the Page Tables. Page Directories and Page Tables are
24682 data structures which control how virtual memory addresses are mapped
24683 into physical addresses. A Page Table includes an entry for every
24684 page of memory that is mapped into the program's address space; there
24685 may be several Page Tables, each one holding up to 4096 entries. A
24686 Page Directory has up to 4096 entries, one each for every Page Table
24687 that is currently in use.
24689 Without an argument, @kbd{info dos pde} displays the entire Page
24690 Directory, and @kbd{info dos pte} displays all the entries in all of
24691 the Page Tables. An argument, an integer expression, given to the
24692 @kbd{info dos pde} command means display only that entry from the Page
24693 Directory table. An argument given to the @kbd{info dos pte} command
24694 means display entries from a single Page Table, the one pointed to by
24695 the specified entry in the Page Directory.
24697 @cindex direct memory access (DMA) on MS-DOS
24698 These commands are useful when your program uses @dfn{DMA} (Direct
24699 Memory Access), which needs physical addresses to program the DMA
24702 These commands are supported only with some DPMI servers.
24704 @cindex physical address from linear address
24705 @item info dos address-pte @var{addr}
24706 This command displays the Page Table entry for a specified linear
24707 address. The argument @var{addr} is a linear address which should
24708 already have the appropriate segment's base address added to it,
24709 because this command accepts addresses which may belong to @emph{any}
24710 segment. For example, here's how to display the Page Table entry for
24711 the page where a variable @code{i} is stored:
24714 @exdent @code{(@value{GDBP}) info dos address-pte __djgpp_base_address + (char *)&i}
24715 @exdent @code{Page Table entry for address 0x11a00d30:}
24716 @exdent @code{Base=0x02698000 Dirty Acc. Not-Cached Write-Back Usr Read-Write +0xd30}
24720 This says that @code{i} is stored at offset @code{0xd30} from the page
24721 whose physical base address is @code{0x02698000}, and shows all the
24722 attributes of that page.
24724 Note that you must cast the addresses of variables to a @code{char *},
24725 since otherwise the value of @code{__djgpp_base_address}, the base
24726 address of all variables and functions in a @sc{djgpp} program, will
24727 be added using the rules of C pointer arithmetics: if @code{i} is
24728 declared an @code{int}, @value{GDBN} will add 4 times the value of
24729 @code{__djgpp_base_address} to the address of @code{i}.
24731 Here's another example, it displays the Page Table entry for the
24735 @exdent @code{(@value{GDBP}) info dos address-pte *((unsigned *)&_go32_info_block + 3)}
24736 @exdent @code{Page Table entry for address 0x29110:}
24737 @exdent @code{Base=0x00029000 Dirty Acc. Not-Cached Write-Back Usr Read-Write +0x110}
24741 (The @code{+ 3} offset is because the transfer buffer's address is the
24742 3rd member of the @code{_go32_info_block} structure.) The output
24743 clearly shows that this DPMI server maps the addresses in conventional
24744 memory 1:1, i.e.@: the physical (@code{0x00029000} + @code{0x110}) and
24745 linear (@code{0x29110}) addresses are identical.
24747 This command is supported only with some DPMI servers.
24750 @cindex DOS serial data link, remote debugging
24751 In addition to native debugging, the DJGPP port supports remote
24752 debugging via a serial data link. The following commands are specific
24753 to remote serial debugging in the DJGPP port of @value{GDBN}.
24756 @kindex set com1base
24757 @kindex set com1irq
24758 @kindex set com2base
24759 @kindex set com2irq
24760 @kindex set com3base
24761 @kindex set com3irq
24762 @kindex set com4base
24763 @kindex set com4irq
24764 @item set com1base @var{addr}
24765 This command sets the base I/O port address of the @file{COM1} serial
24768 @item set com1irq @var{irq}
24769 This command sets the @dfn{Interrupt Request} (@code{IRQ}) line to use
24770 for the @file{COM1} serial port.
24772 There are similar commands @samp{set com2base}, @samp{set com3irq},
24773 etc.@: for setting the port address and the @code{IRQ} lines for the
24776 @kindex show com1base
24777 @kindex show com1irq
24778 @kindex show com2base
24779 @kindex show com2irq
24780 @kindex show com3base
24781 @kindex show com3irq
24782 @kindex show com4base
24783 @kindex show com4irq
24784 The related commands @samp{show com1base}, @samp{show com1irq} etc.@:
24785 display the current settings of the base address and the @code{IRQ}
24786 lines used by the COM ports.
24789 @kindex info serial
24790 @cindex DOS serial port status
24791 This command prints the status of the 4 DOS serial ports. For each
24792 port, it prints whether it's active or not, its I/O base address and
24793 IRQ number, whether it uses a 16550-style FIFO, its baudrate, and the
24794 counts of various errors encountered so far.
24798 @node Cygwin Native
24799 @subsection Features for Debugging MS Windows PE Executables
24800 @cindex MS Windows debugging
24801 @cindex native Cygwin debugging
24802 @cindex Cygwin-specific commands
24804 @value{GDBN} supports native debugging of MS Windows programs, including
24805 DLLs with and without symbolic debugging information.
24807 @cindex Ctrl-BREAK, MS-Windows
24808 @cindex interrupt debuggee on MS-Windows
24809 MS-Windows programs that call @code{SetConsoleMode} to switch off the
24810 special meaning of the @samp{Ctrl-C} keystroke cannot be interrupted
24811 by typing @kbd{C-c}. For this reason, @value{GDBN} on MS-Windows
24812 supports @kbd{C-@key{BREAK}} as an alternative interrupt key
24813 sequence, which can be used to interrupt the debuggee even if it
24816 There are various additional Cygwin-specific commands, described in
24817 this section. Working with DLLs that have no debugging symbols is
24818 described in @ref{Non-debug DLL Symbols}.
24823 This is a prefix of MS Windows-specific commands which print
24824 information about the target system and important OS structures.
24826 @item info w32 selector
24827 This command displays information returned by
24828 the Win32 API @code{GetThreadSelectorEntry} function.
24829 It takes an optional argument that is evaluated to
24830 a long value to give the information about this given selector.
24831 Without argument, this command displays information
24832 about the six segment registers.
24834 @item info w32 thread-information-block
24835 This command displays thread specific information stored in the
24836 Thread Information Block (readable on the X86 CPU family using @code{$fs}
24837 selector for 32-bit programs and @code{$gs} for 64-bit programs).
24839 @kindex signal-event
24840 @item signal-event @var{id}
24841 This command signals an event with user-provided @var{id}. Used to resume
24842 crashing process when attached to it using MS-Windows JIT debugging (AeDebug).
24844 To use it, create or edit the following keys in
24845 @code{HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AeDebug} and/or
24846 @code{HKLM\SOFTWARE\Wow6432Node\Microsoft\Windows NT\CurrentVersion\AeDebug}
24847 (for x86_64 versions):
24851 @code{Debugger} (REG_SZ) --- a command to launch the debugger.
24852 Suggested command is: @code{@var{fully-qualified-path-to-gdb.exe} -ex
24853 "attach %ld" -ex "signal-event %ld" -ex "continue"}.
24855 The first @code{%ld} will be replaced by the process ID of the
24856 crashing process, the second @code{%ld} will be replaced by the ID of
24857 the event that blocks the crashing process, waiting for @value{GDBN}
24861 @code{Auto} (REG_SZ) --- either @code{1} or @code{0}. @code{1} will
24862 make the system run debugger specified by the Debugger key
24863 automatically, @code{0} will cause a dialog box with ``OK'' and
24864 ``Cancel'' buttons to appear, which allows the user to either
24865 terminate the crashing process (OK) or debug it (Cancel).
24868 @kindex set cygwin-exceptions
24869 @cindex debugging the Cygwin DLL
24870 @cindex Cygwin DLL, debugging
24871 @item set cygwin-exceptions @var{mode}
24872 If @var{mode} is @code{on}, @value{GDBN} will break on exceptions that
24873 happen inside the Cygwin DLL. If @var{mode} is @code{off},
24874 @value{GDBN} will delay recognition of exceptions, and may ignore some
24875 exceptions which seem to be caused by internal Cygwin DLL
24876 ``bookkeeping''. This option is meant primarily for debugging the
24877 Cygwin DLL itself; the default value is @code{off} to avoid annoying
24878 @value{GDBN} users with false @code{SIGSEGV} signals.
24880 @kindex show cygwin-exceptions
24881 @item show cygwin-exceptions
24882 Displays whether @value{GDBN} will break on exceptions that happen
24883 inside the Cygwin DLL itself.
24885 @kindex set new-console
24886 @item set new-console @var{mode}
24887 If @var{mode} is @code{on} the debuggee will
24888 be started in a new console on next start.
24889 If @var{mode} is @code{off}, the debuggee will
24890 be started in the same console as the debugger.
24892 @kindex show new-console
24893 @item show new-console
24894 Displays whether a new console is used
24895 when the debuggee is started.
24897 @kindex set new-group
24898 @item set new-group @var{mode}
24899 This boolean value controls whether the debuggee should
24900 start a new group or stay in the same group as the debugger.
24901 This affects the way the Windows OS handles
24904 @kindex show new-group
24905 @item show new-group
24906 Displays current value of new-group boolean.
24908 @kindex set debugevents
24909 @item set debugevents
24910 This boolean value adds debug output concerning kernel events related
24911 to the debuggee seen by the debugger. This includes events that
24912 signal thread and process creation and exit, DLL loading and
24913 unloading, console interrupts, and debugging messages produced by the
24914 Windows @code{OutputDebugString} API call.
24916 @kindex set debugexec
24917 @item set debugexec
24918 This boolean value adds debug output concerning execute events
24919 (such as resume thread) seen by the debugger.
24921 @kindex set debugexceptions
24922 @item set debugexceptions
24923 This boolean value adds debug output concerning exceptions in the
24924 debuggee seen by the debugger.
24926 @kindex set debugmemory
24927 @item set debugmemory
24928 This boolean value adds debug output concerning debuggee memory reads
24929 and writes by the debugger.
24933 This boolean values specifies whether the debuggee is called
24934 via a shell or directly (default value is on).
24938 Displays if the debuggee will be started with a shell.
24943 * Non-debug DLL Symbols:: Support for DLLs without debugging symbols
24946 @node Non-debug DLL Symbols
24947 @subsubsection Support for DLLs without Debugging Symbols
24948 @cindex DLLs with no debugging symbols
24949 @cindex Minimal symbols and DLLs
24951 Very often on windows, some of the DLLs that your program relies on do
24952 not include symbolic debugging information (for example,
24953 @file{kernel32.dll}). When @value{GDBN} doesn't recognize any debugging
24954 symbols in a DLL, it relies on the minimal amount of symbolic
24955 information contained in the DLL's export table. This section
24956 describes working with such symbols, known internally to @value{GDBN} as
24957 ``minimal symbols''.
24959 Note that before the debugged program has started execution, no DLLs
24960 will have been loaded. The easiest way around this problem is simply to
24961 start the program --- either by setting a breakpoint or letting the
24962 program run once to completion.
24964 @subsubsection DLL Name Prefixes
24966 In keeping with the naming conventions used by the Microsoft debugging
24967 tools, DLL export symbols are made available with a prefix based on the
24968 DLL name, for instance @code{KERNEL32!CreateFileA}. The plain name is
24969 also entered into the symbol table, so @code{CreateFileA} is often
24970 sufficient. In some cases there will be name clashes within a program
24971 (particularly if the executable itself includes full debugging symbols)
24972 necessitating the use of the fully qualified name when referring to the
24973 contents of the DLL. Use single-quotes around the name to avoid the
24974 exclamation mark (``!'') being interpreted as a language operator.
24976 Note that the internal name of the DLL may be all upper-case, even
24977 though the file name of the DLL is lower-case, or vice-versa. Since
24978 symbols within @value{GDBN} are @emph{case-sensitive} this may cause
24979 some confusion. If in doubt, try the @code{info functions} and
24980 @code{info variables} commands or even @code{maint print msymbols}
24981 (@pxref{Symbols}). Here's an example:
24984 (@value{GDBP}) info function CreateFileA
24985 All functions matching regular expression "CreateFileA":
24987 Non-debugging symbols:
24988 0x77e885f4 CreateFileA
24989 0x77e885f4 KERNEL32!CreateFileA
24993 (@value{GDBP}) info function !
24994 All functions matching regular expression "!":
24996 Non-debugging symbols:
24997 0x6100114c cygwin1!__assert
24998 0x61004034 cygwin1!_dll_crt0@@0
24999 0x61004240 cygwin1!dll_crt0(per_process *)
25003 @subsubsection Working with Minimal Symbols
25005 Symbols extracted from a DLL's export table do not contain very much
25006 type information. All that @value{GDBN} can do is guess whether a symbol
25007 refers to a function or variable depending on the linker section that
25008 contains the symbol. Also note that the actual contents of the memory
25009 contained in a DLL are not available unless the program is running. This
25010 means that you cannot examine the contents of a variable or disassemble
25011 a function within a DLL without a running program.
25013 Variables are generally treated as pointers and dereferenced
25014 automatically. For this reason, it is often necessary to prefix a
25015 variable name with the address-of operator (``&'') and provide explicit
25016 type information in the command. Here's an example of the type of
25020 (@value{GDBP}) print 'cygwin1!__argv'
25021 'cygwin1!__argv' has unknown type; cast it to its declared type
25025 (@value{GDBP}) x 'cygwin1!__argv'
25026 'cygwin1!__argv' has unknown type; cast it to its declared type
25029 And two possible solutions:
25032 (@value{GDBP}) print ((char **)'cygwin1!__argv')[0]
25033 $2 = 0x22fd98 "/cygdrive/c/mydirectory/myprogram"
25037 (@value{GDBP}) x/2x &'cygwin1!__argv'
25038 0x610c0aa8 <cygwin1!__argv>: 0x10021608 0x00000000
25039 (@value{GDBP}) x/x 0x10021608
25040 0x10021608: 0x0022fd98
25041 (@value{GDBP}) x/s 0x0022fd98
25042 0x22fd98: "/cygdrive/c/mydirectory/myprogram"
25045 Setting a break point within a DLL is possible even before the program
25046 starts execution. However, under these circumstances, @value{GDBN} can't
25047 examine the initial instructions of the function in order to skip the
25048 function's frame set-up code. You can work around this by using ``*&''
25049 to set the breakpoint at a raw memory address:
25052 (@value{GDBP}) break *&'python22!PyOS_Readline'
25053 Breakpoint 1 at 0x1e04eff0
25056 The author of these extensions is not entirely convinced that setting a
25057 break point within a shared DLL like @file{kernel32.dll} is completely
25061 @subsection Commands Specific to @sc{gnu} Hurd Systems
25062 @cindex @sc{gnu} Hurd debugging
25064 This subsection describes @value{GDBN} commands specific to the
25065 @sc{gnu} Hurd native debugging.
25070 @kindex set signals@r{, Hurd command}
25071 @kindex set sigs@r{, Hurd command}
25072 This command toggles the state of inferior signal interception by
25073 @value{GDBN}. Mach exceptions, such as breakpoint traps, are not
25074 affected by this command. @code{sigs} is a shorthand alias for
25079 @kindex show signals@r{, Hurd command}
25080 @kindex show sigs@r{, Hurd command}
25081 Show the current state of intercepting inferior's signals.
25083 @item set signal-thread
25084 @itemx set sigthread
25085 @kindex set signal-thread
25086 @kindex set sigthread
25087 This command tells @value{GDBN} which thread is the @code{libc} signal
25088 thread. That thread is run when a signal is delivered to a running
25089 process. @code{set sigthread} is the shorthand alias of @code{set
25092 @item show signal-thread
25093 @itemx show sigthread
25094 @kindex show signal-thread
25095 @kindex show sigthread
25096 These two commands show which thread will run when the inferior is
25097 delivered a signal.
25100 @kindex set stopped@r{, Hurd command}
25101 This commands tells @value{GDBN} that the inferior process is stopped,
25102 as with the @code{SIGSTOP} signal. The stopped process can be
25103 continued by delivering a signal to it.
25106 @kindex show stopped@r{, Hurd command}
25107 This command shows whether @value{GDBN} thinks the debuggee is
25110 @item set exceptions
25111 @kindex set exceptions@r{, Hurd command}
25112 Use this command to turn off trapping of exceptions in the inferior.
25113 When exception trapping is off, neither breakpoints nor
25114 single-stepping will work. To restore the default, set exception
25117 @item show exceptions
25118 @kindex show exceptions@r{, Hurd command}
25119 Show the current state of trapping exceptions in the inferior.
25121 @item set task pause
25122 @kindex set task@r{, Hurd commands}
25123 @cindex task attributes (@sc{gnu} Hurd)
25124 @cindex pause current task (@sc{gnu} Hurd)
25125 This command toggles task suspension when @value{GDBN} has control.
25126 Setting it to on takes effect immediately, and the task is suspended
25127 whenever @value{GDBN} gets control. Setting it to off will take
25128 effect the next time the inferior is continued. If this option is set
25129 to off, you can use @code{set thread default pause on} or @code{set
25130 thread pause on} (see below) to pause individual threads.
25132 @item show task pause
25133 @kindex show task@r{, Hurd commands}
25134 Show the current state of task suspension.
25136 @item set task detach-suspend-count
25137 @cindex task suspend count
25138 @cindex detach from task, @sc{gnu} Hurd
25139 This command sets the suspend count the task will be left with when
25140 @value{GDBN} detaches from it.
25142 @item show task detach-suspend-count
25143 Show the suspend count the task will be left with when detaching.
25145 @item set task exception-port
25146 @itemx set task excp
25147 @cindex task exception port, @sc{gnu} Hurd
25148 This command sets the task exception port to which @value{GDBN} will
25149 forward exceptions. The argument should be the value of the @dfn{send
25150 rights} of the task. @code{set task excp} is a shorthand alias.
25152 @item set noninvasive
25153 @cindex noninvasive task options
25154 This command switches @value{GDBN} to a mode that is the least
25155 invasive as far as interfering with the inferior is concerned. This
25156 is the same as using @code{set task pause}, @code{set exceptions}, and
25157 @code{set signals} to values opposite to the defaults.
25159 @item info send-rights
25160 @itemx info receive-rights
25161 @itemx info port-rights
25162 @itemx info port-sets
25163 @itemx info dead-names
25166 @cindex send rights, @sc{gnu} Hurd
25167 @cindex receive rights, @sc{gnu} Hurd
25168 @cindex port rights, @sc{gnu} Hurd
25169 @cindex port sets, @sc{gnu} Hurd
25170 @cindex dead names, @sc{gnu} Hurd
25171 These commands display information about, respectively, send rights,
25172 receive rights, port rights, port sets, and dead names of a task.
25173 There are also shorthand aliases: @code{info ports} for @code{info
25174 port-rights} and @code{info psets} for @code{info port-sets}.
25176 @item set thread pause
25177 @kindex set thread@r{, Hurd command}
25178 @cindex thread properties, @sc{gnu} Hurd
25179 @cindex pause current thread (@sc{gnu} Hurd)
25180 This command toggles current thread suspension when @value{GDBN} has
25181 control. Setting it to on takes effect immediately, and the current
25182 thread is suspended whenever @value{GDBN} gets control. Setting it to
25183 off will take effect the next time the inferior is continued.
25184 Normally, this command has no effect, since when @value{GDBN} has
25185 control, the whole task is suspended. However, if you used @code{set
25186 task pause off} (see above), this command comes in handy to suspend
25187 only the current thread.
25189 @item show thread pause
25190 @kindex show thread@r{, Hurd command}
25191 This command shows the state of current thread suspension.
25193 @item set thread run
25194 This command sets whether the current thread is allowed to run.
25196 @item show thread run
25197 Show whether the current thread is allowed to run.
25199 @item set thread detach-suspend-count
25200 @cindex thread suspend count, @sc{gnu} Hurd
25201 @cindex detach from thread, @sc{gnu} Hurd
25202 This command sets the suspend count @value{GDBN} will leave on a
25203 thread when detaching. This number is relative to the suspend count
25204 found by @value{GDBN} when it notices the thread; use @code{set thread
25205 takeover-suspend-count} to force it to an absolute value.
25207 @item show thread detach-suspend-count
25208 Show the suspend count @value{GDBN} will leave on the thread when
25211 @item set thread exception-port
25212 @itemx set thread excp
25213 Set the thread exception port to which to forward exceptions. This
25214 overrides the port set by @code{set task exception-port} (see above).
25215 @code{set thread excp} is the shorthand alias.
25217 @item set thread takeover-suspend-count
25218 Normally, @value{GDBN}'s thread suspend counts are relative to the
25219 value @value{GDBN} finds when it notices each thread. This command
25220 changes the suspend counts to be absolute instead.
25222 @item set thread default
25223 @itemx show thread default
25224 @cindex thread default settings, @sc{gnu} Hurd
25225 Each of the above @code{set thread} commands has a @code{set thread
25226 default} counterpart (e.g., @code{set thread default pause}, @code{set
25227 thread default exception-port}, etc.). The @code{thread default}
25228 variety of commands sets the default thread properties for all
25229 threads; you can then change the properties of individual threads with
25230 the non-default commands.
25237 @value{GDBN} provides the following commands specific to the Darwin target:
25240 @item set debug darwin @var{num}
25241 @kindex set debug darwin
25242 When set to a non zero value, enables debugging messages specific to
25243 the Darwin support. Higher values produce more verbose output.
25245 @item show debug darwin
25246 @kindex show debug darwin
25247 Show the current state of Darwin messages.
25249 @item set debug mach-o @var{num}
25250 @kindex set debug mach-o
25251 When set to a non zero value, enables debugging messages while
25252 @value{GDBN} is reading Darwin object files. (@dfn{Mach-O} is the
25253 file format used on Darwin for object and executable files.) Higher
25254 values produce more verbose output. This is a command to diagnose
25255 problems internal to @value{GDBN} and should not be needed in normal
25258 @item show debug mach-o
25259 @kindex show debug mach-o
25260 Show the current state of Mach-O file messages.
25262 @item set mach-exceptions on
25263 @itemx set mach-exceptions off
25264 @kindex set mach-exceptions
25265 On Darwin, faults are first reported as a Mach exception and are then
25266 mapped to a Posix signal. Use this command to turn on trapping of
25267 Mach exceptions in the inferior. This might be sometimes useful to
25268 better understand the cause of a fault. The default is off.
25270 @item show mach-exceptions
25271 @kindex show mach-exceptions
25272 Show the current state of exceptions trapping.
25276 @subsection FreeBSD
25279 When the ABI of a system call is changed in the FreeBSD kernel, this
25280 is implemented by leaving a compatibility system call using the old
25281 ABI at the existing number and allocating a new system call number for
25282 the version using the new ABI. As a convenience, when a system call
25283 is caught by name (@pxref{catch syscall}), compatibility system calls
25286 For example, FreeBSD 12 introduced a new variant of the @code{kevent}
25287 system call and catching the @code{kevent} system call by name catches
25291 (@value{GDBP}) catch syscall kevent
25292 Catchpoint 1 (syscalls 'freebsd11_kevent' [363] 'kevent' [560])
25298 @section Embedded Operating Systems
25300 This section describes configurations involving the debugging of
25301 embedded operating systems that are available for several different
25304 @value{GDBN} includes the ability to debug programs running on
25305 various real-time operating systems.
25307 @node Embedded Processors
25308 @section Embedded Processors
25310 This section goes into details specific to particular embedded
25313 @cindex send command to simulator
25314 Whenever a specific embedded processor has a simulator, @value{GDBN}
25315 allows to send an arbitrary command to the simulator.
25318 @item sim @var{command}
25319 @kindex sim@r{, a command}
25320 Send an arbitrary @var{command} string to the simulator. Consult the
25321 documentation for the specific simulator in use for information about
25322 acceptable commands.
25327 * ARC:: Synopsys ARC
25329 * M68K:: Motorola M68K
25330 * MicroBlaze:: Xilinx MicroBlaze
25331 * MIPS Embedded:: MIPS Embedded
25332 * OpenRISC 1000:: OpenRISC 1000 (or1k)
25333 * PowerPC Embedded:: PowerPC Embedded
25336 * Super-H:: Renesas Super-H
25340 @subsection Synopsys ARC
25341 @cindex Synopsys ARC
25342 @cindex ARC specific commands
25348 @value{GDBN} provides the following ARC-specific commands:
25351 @item set debug arc
25352 @kindex set debug arc
25353 Control the level of ARC specific debug messages. Use 0 for no messages (the
25354 default), 1 for debug messages, and 2 for even more debug messages.
25356 @item show debug arc
25357 @kindex show debug arc
25358 Show the level of ARC specific debugging in operation.
25360 @item maint print arc arc-instruction @var{address}
25361 @kindex maint print arc arc-instruction
25362 Print internal disassembler information about instruction at a given address.
25369 @value{GDBN} provides the following ARM-specific commands:
25372 @item set arm disassembler
25374 This commands selects from a list of disassembly styles. The
25375 @code{"std"} style is the standard style.
25377 @item show arm disassembler
25379 Show the current disassembly style.
25381 @item set arm apcs32
25382 @cindex ARM 32-bit mode
25383 This command toggles ARM operation mode between 32-bit and 26-bit.
25385 @item show arm apcs32
25386 Display the current usage of the ARM 32-bit mode.
25388 @item set arm fpu @var{fputype}
25389 This command sets the ARM floating-point unit (FPU) type. The
25390 argument @var{fputype} can be one of these:
25394 Determine the FPU type by querying the OS ABI.
25396 Software FPU, with mixed-endian doubles on little-endian ARM
25399 GCC-compiled FPA co-processor.
25401 Software FPU with pure-endian doubles.
25407 Show the current type of the FPU.
25410 This command forces @value{GDBN} to use the specified ABI.
25413 Show the currently used ABI.
25415 @item set arm fallback-mode (arm|thumb|auto)
25416 @value{GDBN} uses the symbol table, when available, to determine
25417 whether instructions are ARM or Thumb. This command controls
25418 @value{GDBN}'s default behavior when the symbol table is not
25419 available. The default is @samp{auto}, which causes @value{GDBN} to
25420 use the current execution mode (from the @code{T} bit in the @code{CPSR}
25423 @item show arm fallback-mode
25424 Show the current fallback instruction mode.
25426 @item set arm force-mode (arm|thumb|auto)
25427 This command overrides use of the symbol table to determine whether
25428 instructions are ARM or Thumb. The default is @samp{auto}, which
25429 causes @value{GDBN} to use the symbol table and then the setting
25430 of @samp{set arm fallback-mode}.
25432 @item show arm force-mode
25433 Show the current forced instruction mode.
25435 @item set debug arm
25436 Toggle whether to display ARM-specific debugging messages from the ARM
25437 target support subsystem.
25439 @item show debug arm
25440 Show whether ARM-specific debugging messages are enabled.
25444 @item target sim @r{[}@var{simargs}@r{]} @dots{}
25445 The @value{GDBN} ARM simulator accepts the following optional arguments.
25448 @item --swi-support=@var{type}
25449 Tell the simulator which SWI interfaces to support. The argument
25450 @var{type} may be a comma separated list of the following values.
25451 The default value is @code{all}.
25466 The Motorola m68k configuration includes ColdFire support.
25469 @subsection MicroBlaze
25470 @cindex Xilinx MicroBlaze
25471 @cindex XMD, Xilinx Microprocessor Debugger
25473 The MicroBlaze is a soft-core processor supported on various Xilinx
25474 FPGAs, such as Spartan or Virtex series. Boards with these processors
25475 usually have JTAG ports which connect to a host system running the Xilinx
25476 Embedded Development Kit (EDK) or Software Development Kit (SDK).
25477 This host system is used to download the configuration bitstream to
25478 the target FPGA. The Xilinx Microprocessor Debugger (XMD) program
25479 communicates with the target board using the JTAG interface and
25480 presents a @code{gdbserver} interface to the board. By default
25481 @code{xmd} uses port @code{1234}. (While it is possible to change
25482 this default port, it requires the use of undocumented @code{xmd}
25483 commands. Contact Xilinx support if you need to do this.)
25485 Use these GDB commands to connect to the MicroBlaze target processor.
25488 @item target remote :1234
25489 Use this command to connect to the target if you are running @value{GDBN}
25490 on the same system as @code{xmd}.
25492 @item target remote @var{xmd-host}:1234
25493 Use this command to connect to the target if it is connected to @code{xmd}
25494 running on a different system named @var{xmd-host}.
25497 Use this command to download a program to the MicroBlaze target.
25499 @item set debug microblaze @var{n}
25500 Enable MicroBlaze-specific debugging messages if non-zero.
25502 @item show debug microblaze @var{n}
25503 Show MicroBlaze-specific debugging level.
25506 @node MIPS Embedded
25507 @subsection @acronym{MIPS} Embedded
25510 @value{GDBN} supports these special commands for @acronym{MIPS} targets:
25513 @item set mipsfpu double
25514 @itemx set mipsfpu single
25515 @itemx set mipsfpu none
25516 @itemx set mipsfpu auto
25517 @itemx show mipsfpu
25518 @kindex set mipsfpu
25519 @kindex show mipsfpu
25520 @cindex @acronym{MIPS} remote floating point
25521 @cindex floating point, @acronym{MIPS} remote
25522 If your target board does not support the @acronym{MIPS} floating point
25523 coprocessor, you should use the command @samp{set mipsfpu none} (if you
25524 need this, you may wish to put the command in your @value{GDBN} init
25525 file). This tells @value{GDBN} how to find the return value of
25526 functions which return floating point values. It also allows
25527 @value{GDBN} to avoid saving the floating point registers when calling
25528 functions on the board. If you are using a floating point coprocessor
25529 with only single precision floating point support, as on the @sc{r4650}
25530 processor, use the command @samp{set mipsfpu single}. The default
25531 double precision floating point coprocessor may be selected using
25532 @samp{set mipsfpu double}.
25534 In previous versions the only choices were double precision or no
25535 floating point, so @samp{set mipsfpu on} will select double precision
25536 and @samp{set mipsfpu off} will select no floating point.
25538 As usual, you can inquire about the @code{mipsfpu} variable with
25539 @samp{show mipsfpu}.
25542 @node OpenRISC 1000
25543 @subsection OpenRISC 1000
25544 @cindex OpenRISC 1000
25547 The OpenRISC 1000 provides a free RISC instruction set architecture. It is
25548 mainly provided as a soft-core which can run on Xilinx, Altera and other
25551 @value{GDBN} for OpenRISC supports the below commands when connecting to
25559 Runs the builtin CPU simulator which can run very basic
25560 programs but does not support most hardware functions like MMU.
25561 For more complex use cases the user is advised to run an external
25562 target, and connect using @samp{target remote}.
25564 Example: @code{target sim}
25566 @item set debug or1k
25567 Toggle whether to display OpenRISC-specific debugging messages from the
25568 OpenRISC target support subsystem.
25570 @item show debug or1k
25571 Show whether OpenRISC-specific debugging messages are enabled.
25574 @node PowerPC Embedded
25575 @subsection PowerPC Embedded
25577 @cindex DVC register
25578 @value{GDBN} supports using the DVC (Data Value Compare) register to
25579 implement in hardware simple hardware watchpoint conditions of the form:
25582 (@value{GDBP}) watch @var{ADDRESS|VARIABLE} \
25583 if @var{ADDRESS|VARIABLE} == @var{CONSTANT EXPRESSION}
25586 The DVC register will be automatically used when @value{GDBN} detects
25587 such pattern in a condition expression, and the created watchpoint uses one
25588 debug register (either the @code{exact-watchpoints} option is on and the
25589 variable is scalar, or the variable has a length of one byte). This feature
25590 is available in native @value{GDBN} running on a Linux kernel version 2.6.34
25593 When running on PowerPC embedded processors, @value{GDBN} automatically uses
25594 ranged hardware watchpoints, unless the @code{exact-watchpoints} option is on,
25595 in which case watchpoints using only one debug register are created when
25596 watching variables of scalar types.
25598 You can create an artificial array to watch an arbitrary memory
25599 region using one of the following commands (@pxref{Expressions}):
25602 (@value{GDBP}) watch *((char *) @var{address})@@@var{length}
25603 (@value{GDBP}) watch @{char[@var{length}]@} @var{address}
25606 PowerPC embedded processors support masked watchpoints. See the discussion
25607 about the @code{mask} argument in @ref{Set Watchpoints}.
25609 @cindex ranged breakpoint
25610 PowerPC embedded processors support hardware accelerated
25611 @dfn{ranged breakpoints}. A ranged breakpoint stops execution of
25612 the inferior whenever it executes an instruction at any address within
25613 the range it specifies. To set a ranged breakpoint in @value{GDBN},
25614 use the @code{break-range} command.
25616 @value{GDBN} provides the following PowerPC-specific commands:
25619 @kindex break-range
25620 @item break-range @var{start-location}, @var{end-location}
25621 Set a breakpoint for an address range given by
25622 @var{start-location} and @var{end-location}, which can specify a function name,
25623 a line number, an offset of lines from the current line or from the start
25624 location, or an address of an instruction (see @ref{Specify Location},
25625 for a list of all the possible ways to specify a @var{location}.)
25626 The breakpoint will stop execution of the inferior whenever it
25627 executes an instruction at any address within the specified range,
25628 (including @var{start-location} and @var{end-location}.)
25630 @kindex set powerpc
25631 @item set powerpc soft-float
25632 @itemx show powerpc soft-float
25633 Force @value{GDBN} to use (or not use) a software floating point calling
25634 convention. By default, @value{GDBN} selects the calling convention based
25635 on the selected architecture and the provided executable file.
25637 @item set powerpc vector-abi
25638 @itemx show powerpc vector-abi
25639 Force @value{GDBN} to use the specified calling convention for vector
25640 arguments and return values. The valid options are @samp{auto};
25641 @samp{generic}, to avoid vector registers even if they are present;
25642 @samp{altivec}, to use AltiVec registers; and @samp{spe} to use SPE
25643 registers. By default, @value{GDBN} selects the calling convention
25644 based on the selected architecture and the provided executable file.
25646 @item set powerpc exact-watchpoints
25647 @itemx show powerpc exact-watchpoints
25648 Allow @value{GDBN} to use only one debug register when watching a variable
25649 of scalar type, thus assuming that the variable is accessed through the
25650 address of its first byte.
25655 @subsection Atmel AVR
25658 When configured for debugging the Atmel AVR, @value{GDBN} supports the
25659 following AVR-specific commands:
25662 @item info io_registers
25663 @kindex info io_registers@r{, AVR}
25664 @cindex I/O registers (Atmel AVR)
25665 This command displays information about the AVR I/O registers. For
25666 each register, @value{GDBN} prints its number and value.
25673 When configured for debugging CRIS, @value{GDBN} provides the
25674 following CRIS-specific commands:
25677 @item set cris-version @var{ver}
25678 @cindex CRIS version
25679 Set the current CRIS version to @var{ver}, either @samp{10} or @samp{32}.
25680 The CRIS version affects register names and sizes. This command is useful in
25681 case autodetection of the CRIS version fails.
25683 @item show cris-version
25684 Show the current CRIS version.
25686 @item set cris-dwarf2-cfi
25687 @cindex DWARF-2 CFI and CRIS
25688 Set the usage of DWARF-2 CFI for CRIS debugging. The default is @samp{on}.
25689 Change to @samp{off} when using @code{gcc-cris} whose version is below
25692 @item show cris-dwarf2-cfi
25693 Show the current state of using DWARF-2 CFI.
25695 @item set cris-mode @var{mode}
25697 Set the current CRIS mode to @var{mode}. It should only be changed when
25698 debugging in guru mode, in which case it should be set to
25699 @samp{guru} (the default is @samp{normal}).
25701 @item show cris-mode
25702 Show the current CRIS mode.
25706 @subsection Renesas Super-H
25709 For the Renesas Super-H processor, @value{GDBN} provides these
25713 @item set sh calling-convention @var{convention}
25714 @kindex set sh calling-convention
25715 Set the calling-convention used when calling functions from @value{GDBN}.
25716 Allowed values are @samp{gcc}, which is the default setting, and @samp{renesas}.
25717 With the @samp{gcc} setting, functions are called using the @value{NGCC} calling
25718 convention. If the DWARF-2 information of the called function specifies
25719 that the function follows the Renesas calling convention, the function
25720 is called using the Renesas calling convention. If the calling convention
25721 is set to @samp{renesas}, the Renesas calling convention is always used,
25722 regardless of the DWARF-2 information. This can be used to override the
25723 default of @samp{gcc} if debug information is missing, or the compiler
25724 does not emit the DWARF-2 calling convention entry for a function.
25726 @item show sh calling-convention
25727 @kindex show sh calling-convention
25728 Show the current calling convention setting.
25733 @node Architectures
25734 @section Architectures
25736 This section describes characteristics of architectures that affect
25737 all uses of @value{GDBN} with the architecture, both native and cross.
25744 * HPPA:: HP PA architecture
25749 * AMD GPU:: @acronym{AMD GPU} architectures
25753 @subsection AArch64
25754 @cindex AArch64 support
25756 When @value{GDBN} is debugging the AArch64 architecture, it provides the
25757 following special commands:
25760 @item set debug aarch64
25761 @kindex set debug aarch64
25762 This command determines whether AArch64 architecture-specific debugging
25763 messages are to be displayed.
25765 @item show debug aarch64
25766 Show whether AArch64 debugging messages are displayed.
25770 @subsubsection AArch64 SVE.
25771 @cindex AArch64 SVE.
25773 When @value{GDBN} is debugging the AArch64 architecture, if the Scalable Vector
25774 Extension (SVE) is present, then @value{GDBN} will provide the vector registers
25775 @code{$z0} through @code{$z31}, vector predicate registers @code{$p0} through
25776 @code{$p15}, and the @code{$ffr} register. In addition, the pseudo register
25777 @code{$vg} will be provided. This is the vector granule for the current thread
25778 and represents the number of 64-bit chunks in an SVE @code{z} register.
25780 If the vector length changes, then the @code{$vg} register will be updated,
25781 but the lengths of the @code{z} and @code{p} registers will not change. This
25782 is a known limitation of @value{GDBN} and does not affect the execution of the
25785 @subsubsection AArch64 Pointer Authentication.
25786 @cindex AArch64 Pointer Authentication.
25788 When @value{GDBN} is debugging the AArch64 architecture, and the program is
25789 using the v8.3-A feature Pointer Authentication (PAC), then whenever the link
25790 register @code{$lr} is pointing to an PAC function its value will be masked.
25791 When GDB prints a backtrace, any addresses that required unmasking will be
25792 postfixed with the marker [PAC]. When using the MI, this is printed as part
25793 of the @code{addr_flags} field.
25796 @subsection x86 Architecture-specific Issues
25799 @item set struct-convention @var{mode}
25800 @kindex set struct-convention
25801 @cindex struct return convention
25802 @cindex struct/union returned in registers
25803 Set the convention used by the inferior to return @code{struct}s and
25804 @code{union}s from functions to @var{mode}. Possible values of
25805 @var{mode} are @code{"pcc"}, @code{"reg"}, and @code{"default"} (the
25806 default). @code{"default"} or @code{"pcc"} means that @code{struct}s
25807 are returned on the stack, while @code{"reg"} means that a
25808 @code{struct} or a @code{union} whose size is 1, 2, 4, or 8 bytes will
25809 be returned in a register.
25811 @item show struct-convention
25812 @kindex show struct-convention
25813 Show the current setting of the convention to return @code{struct}s
25818 @subsubsection Intel @dfn{Memory Protection Extensions} (MPX).
25819 @cindex Intel Memory Protection Extensions (MPX).
25821 Memory Protection Extension (MPX) adds the bound registers @samp{BND0}
25822 @footnote{The register named with capital letters represent the architecture
25823 registers.} through @samp{BND3}. Bound registers store a pair of 64-bit values
25824 which are the lower bound and upper bound. Bounds are effective addresses or
25825 memory locations. The upper bounds are architecturally represented in 1's
25826 complement form. A bound having lower bound = 0, and upper bound = 0
25827 (1's complement of all bits set) will allow access to the entire address space.
25829 @samp{BND0} through @samp{BND3} are represented in @value{GDBN} as @samp{bnd0raw}
25830 through @samp{bnd3raw}. Pseudo registers @samp{bnd0} through @samp{bnd3}
25831 display the upper bound performing the complement of one operation on the
25832 upper bound value, i.e.@ when upper bound in @samp{bnd0raw} is 0 in the
25833 @value{GDBN} @samp{bnd0} it will be @code{0xfff@dots{}}. In this sense it
25834 can also be noted that the upper bounds are inclusive.
25836 As an example, assume that the register BND0 holds bounds for a pointer having
25837 access allowed for the range between 0x32 and 0x71. The values present on
25838 bnd0raw and bnd registers are presented as follows:
25841 bnd0raw = @{0x32, 0xffffffff8e@}
25842 bnd0 = @{lbound = 0x32, ubound = 0x71@} : size 64
25845 This way the raw value can be accessed via bnd0raw@dots{}bnd3raw. Any
25846 change on bnd0@dots{}bnd3 or bnd0raw@dots{}bnd3raw is reflect on its
25847 counterpart. When the bnd0@dots{}bnd3 registers are displayed via
25848 Python, the display includes the memory size, in bits, accessible to
25851 Bounds can also be stored in bounds tables, which are stored in
25852 application memory. These tables store bounds for pointers by specifying
25853 the bounds pointer's value along with its bounds. Evaluating and changing
25854 bounds located in bound tables is therefore interesting while investigating
25855 bugs on MPX context. @value{GDBN} provides commands for this purpose:
25858 @item show mpx bound @var{pointer}
25859 @kindex show mpx bound
25860 Display bounds of the given @var{pointer}.
25862 @item set mpx bound @var{pointer}, @var{lbound}, @var{ubound}
25863 @kindex set mpx bound
25864 Set the bounds of a pointer in the bound table.
25865 This command takes three parameters: @var{pointer} is the pointers
25866 whose bounds are to be changed, @var{lbound} and @var{ubound} are new values
25867 for lower and upper bounds respectively.
25870 When you call an inferior function on an Intel MPX enabled program,
25871 GDB sets the inferior's bound registers to the init (disabled) state
25872 before calling the function. As a consequence, bounds checks for the
25873 pointer arguments passed to the function will always pass.
25875 This is necessary because when you call an inferior function, the
25876 program is usually in the middle of the execution of other function.
25877 Since at that point bound registers are in an arbitrary state, not
25878 clearing them would lead to random bound violations in the called
25881 You can still examine the influence of the bound registers on the
25882 execution of the called function by stopping the execution of the
25883 called function at its prologue, setting bound registers, and
25884 continuing the execution. For example:
25888 Breakpoint 2 at 0x4009de: file i386-mpx-call.c, line 47.
25889 $ print upper (a, b, c, d, 1)
25890 Breakpoint 2, upper (a=0x0, b=0x6e0000005b, c=0x0, d=0x0, len=48)....
25892 @{lbound = 0x0, ubound = ffffffff@} : size -1
25895 At this last step the value of bnd0 can be changed for investigation of bound
25896 violations caused along the execution of the call. In order to know how to
25897 set the bound registers or bound table for the call consult the ABI.
25902 See the following section.
25905 @subsection @acronym{MIPS}
25907 @cindex stack on Alpha
25908 @cindex stack on @acronym{MIPS}
25909 @cindex Alpha stack
25910 @cindex @acronym{MIPS} stack
25911 Alpha- and @acronym{MIPS}-based computers use an unusual stack frame, which
25912 sometimes requires @value{GDBN} to search backward in the object code to
25913 find the beginning of a function.
25915 @cindex response time, @acronym{MIPS} debugging
25916 To improve response time (especially for embedded applications, where
25917 @value{GDBN} may be restricted to a slow serial line for this search)
25918 you may want to limit the size of this search, using one of these
25922 @cindex @code{heuristic-fence-post} (Alpha, @acronym{MIPS})
25923 @item set heuristic-fence-post @var{limit}
25924 Restrict @value{GDBN} to examining at most @var{limit} bytes in its
25925 search for the beginning of a function. A value of @var{0} (the
25926 default) means there is no limit. However, except for @var{0}, the
25927 larger the limit the more bytes @code{heuristic-fence-post} must search
25928 and therefore the longer it takes to run. You should only need to use
25929 this command when debugging a stripped executable.
25931 @item show heuristic-fence-post
25932 Display the current limit.
25936 These commands are available @emph{only} when @value{GDBN} is configured
25937 for debugging programs on Alpha or @acronym{MIPS} processors.
25939 Several @acronym{MIPS}-specific commands are available when debugging @acronym{MIPS}
25943 @item set mips abi @var{arg}
25944 @kindex set mips abi
25945 @cindex set ABI for @acronym{MIPS}
25946 Tell @value{GDBN} which @acronym{MIPS} ABI is used by the inferior. Possible
25947 values of @var{arg} are:
25951 The default ABI associated with the current binary (this is the
25961 @item show mips abi
25962 @kindex show mips abi
25963 Show the @acronym{MIPS} ABI used by @value{GDBN} to debug the inferior.
25965 @item set mips compression @var{arg}
25966 @kindex set mips compression
25967 @cindex code compression, @acronym{MIPS}
25968 Tell @value{GDBN} which @acronym{MIPS} compressed
25969 @acronym{ISA, Instruction Set Architecture} encoding is used by the
25970 inferior. @value{GDBN} uses this for code disassembly and other
25971 internal interpretation purposes. This setting is only referred to
25972 when no executable has been associated with the debugging session or
25973 the executable does not provide information about the encoding it uses.
25974 Otherwise this setting is automatically updated from information
25975 provided by the executable.
25977 Possible values of @var{arg} are @samp{mips16} and @samp{micromips}.
25978 The default compressed @acronym{ISA} encoding is @samp{mips16}, as
25979 executables containing @acronym{MIPS16} code frequently are not
25980 identified as such.
25982 This setting is ``sticky''; that is, it retains its value across
25983 debugging sessions until reset either explicitly with this command or
25984 implicitly from an executable.
25986 The compiler and/or assembler typically add symbol table annotations to
25987 identify functions compiled for the @acronym{MIPS16} or
25988 @acronym{microMIPS} @acronym{ISA}s. If these function-scope annotations
25989 are present, @value{GDBN} uses them in preference to the global
25990 compressed @acronym{ISA} encoding setting.
25992 @item show mips compression
25993 @kindex show mips compression
25994 Show the @acronym{MIPS} compressed @acronym{ISA} encoding used by
25995 @value{GDBN} to debug the inferior.
25998 @itemx show mipsfpu
25999 @xref{MIPS Embedded, set mipsfpu}.
26001 @item set mips mask-address @var{arg}
26002 @kindex set mips mask-address
26003 @cindex @acronym{MIPS} addresses, masking
26004 This command determines whether the most-significant 32 bits of 64-bit
26005 @acronym{MIPS} addresses are masked off. The argument @var{arg} can be
26006 @samp{on}, @samp{off}, or @samp{auto}. The latter is the default
26007 setting, which lets @value{GDBN} determine the correct value.
26009 @item show mips mask-address
26010 @kindex show mips mask-address
26011 Show whether the upper 32 bits of @acronym{MIPS} addresses are masked off or
26014 @item set remote-mips64-transfers-32bit-regs
26015 @kindex set remote-mips64-transfers-32bit-regs
26016 This command controls compatibility with 64-bit @acronym{MIPS} targets that
26017 transfer data in 32-bit quantities. If you have an old @acronym{MIPS} 64 target
26018 that transfers 32 bits for some registers, like @sc{sr} and @sc{fsr},
26019 and 64 bits for other registers, set this option to @samp{on}.
26021 @item show remote-mips64-transfers-32bit-regs
26022 @kindex show remote-mips64-transfers-32bit-regs
26023 Show the current setting of compatibility with older @acronym{MIPS} 64 targets.
26025 @item set debug mips
26026 @kindex set debug mips
26027 This command turns on and off debugging messages for the @acronym{MIPS}-specific
26028 target code in @value{GDBN}.
26030 @item show debug mips
26031 @kindex show debug mips
26032 Show the current setting of @acronym{MIPS} debugging messages.
26038 @cindex HPPA support
26040 When @value{GDBN} is debugging the HP PA architecture, it provides the
26041 following special commands:
26044 @item set debug hppa
26045 @kindex set debug hppa
26046 This command determines whether HPPA architecture-specific debugging
26047 messages are to be displayed.
26049 @item show debug hppa
26050 Show whether HPPA debugging messages are displayed.
26052 @item maint print unwind @var{address}
26053 @kindex maint print unwind@r{, HPPA}
26054 This command displays the contents of the unwind table entry at the
26055 given @var{address}.
26061 @subsection PowerPC
26062 @cindex PowerPC architecture
26064 When @value{GDBN} is debugging the PowerPC architecture, it provides a set of
26065 pseudo-registers to enable inspection of 128-bit wide Decimal Floating Point
26066 numbers stored in the floating point registers. These values must be stored
26067 in two consecutive registers, always starting at an even register like
26068 @code{f0} or @code{f2}.
26070 The pseudo-registers go from @code{$dl0} through @code{$dl15}, and are formed
26071 by joining the even/odd register pairs @code{f0} and @code{f1} for @code{$dl0},
26072 @code{f2} and @code{f3} for @code{$dl1} and so on.
26074 For POWER7 processors, @value{GDBN} provides a set of pseudo-registers, the 64-bit
26075 wide Extended Floating Point Registers (@samp{f32} through @samp{f63}).
26078 @subsection Nios II
26079 @cindex Nios II architecture
26081 When @value{GDBN} is debugging the Nios II architecture,
26082 it provides the following special commands:
26086 @item set debug nios2
26087 @kindex set debug nios2
26088 This command turns on and off debugging messages for the Nios II
26089 target code in @value{GDBN}.
26091 @item show debug nios2
26092 @kindex show debug nios2
26093 Show the current setting of Nios II debugging messages.
26097 @subsection Sparc64
26098 @cindex Sparc64 support
26099 @cindex Application Data Integrity
26100 @subsubsection ADI Support
26102 The M7 processor supports an Application Data Integrity (ADI) feature that
26103 detects invalid data accesses. When software allocates memory and enables
26104 ADI on the allocated memory, it chooses a 4-bit version number, sets the
26105 version in the upper 4 bits of the 64-bit pointer to that data, and stores
26106 the 4-bit version in every cacheline of that data. Hardware saves the latter
26107 in spare bits in the cache and memory hierarchy. On each load and store,
26108 the processor compares the upper 4 VA (virtual address) bits to the
26109 cacheline's version. If there is a mismatch, the processor generates a
26110 version mismatch trap which can be either precise or disrupting. The trap
26111 is an error condition which the kernel delivers to the process as a SIGSEGV
26114 Note that only 64-bit applications can use ADI and need to be built with
26117 Values of the ADI version tags, which are in granularity of a
26118 cacheline (64 bytes), can be viewed or modified.
26122 @kindex adi examine
26123 @item adi (examine | x) [ / @var{n} ] @var{addr}
26125 The @code{adi examine} command displays the value of one ADI version tag per
26128 @var{n} is a decimal integer specifying the number in bytes; the default
26129 is 1. It specifies how much ADI version information, at the ratio of 1:ADI
26130 block size, to display.
26132 @var{addr} is the address in user address space where you want @value{GDBN}
26133 to begin displaying the ADI version tags.
26135 Below is an example of displaying ADI versions of variable "shmaddr".
26138 (@value{GDBP}) adi x/100 shmaddr
26139 0xfff800010002c000: 0 0
26143 @item adi (assign | a) [ / @var{n} ] @var{addr} = @var{tag}
26145 The @code{adi assign} command is used to assign new ADI version tag
26148 @var{n} is a decimal integer specifying the number in bytes;
26149 the default is 1. It specifies how much ADI version information, at the
26150 ratio of 1:ADI block size, to modify.
26152 @var{addr} is the address in user address space where you want @value{GDBN}
26153 to begin modifying the ADI version tags.
26155 @var{tag} is the new ADI version tag.
26157 For example, do the following to modify then verify ADI versions of
26158 variable "shmaddr":
26161 (@value{GDBP}) adi a/100 shmaddr = 7
26162 (@value{GDBP}) adi x/100 shmaddr
26163 0xfff800010002c000: 7 7
26170 @cindex S12Z support
26172 When @value{GDBN} is debugging the S12Z architecture,
26173 it provides the following special command:
26176 @item maint info bdccsr
26177 @kindex maint info bdccsr@r{, S12Z}
26178 This command displays the current value of the microprocessor's
26183 @subsection @acronym{AMD GPU}
26184 @cindex @acronym{AMD GPU} support
26186 @value{GDBN} provides support for systems that have heterogeneous
26187 agents associated with @acronym{AMD GPU} devices (@pxref{Heterogeneous
26188 Debugging}) when @acronym{AMD}'s
26189 @url{https://rocm-documentation.readthedocs.io/, @acronym{ROCm, Radeon
26190 Open Compute platforM}} for HIP-Clang is installed.
26192 The following AMD GPU chips are supported:
26197 ``Vega 10'' which is displayed as @samp{vega10} by @value{GDBN} and
26198 denoted as @samp{gfx900} by the compiler.
26201 ``Vega 7nm'' which is displayed as @samp{vega20} by @value{GDBN} and
26202 denoted as @samp{gfx906} by the compiler.
26205 ``Arcturus'' which is displayed as @samp{arcturus} by @value{GDBN} and
26206 denoted as @samp{gfx908} by the compiler.
26210 @value{GDBN} supports the following source languages:
26216 @url{https://github.com/ROCm-Developer-Tools/HIP/blob/master/docs/markdown/hip_kernel_language.md,
26217 HIP Programming Language} is supported.
26219 When compiling, the @w{@option{-ggdb}} option should be used to
26220 produce debugging information suitable for use by @value{GDBN}. The
26221 @w{@option{--amdgpu-target}} option is used to specify the AMD GPUs
26222 that the executable is required to support. For example, to compile a
26223 HIP program that can utilize ``Vega 10'', ``Vega 7nm'', and
26224 ``Arcturus'' AMD GPU chips, with no optimization:
26227 hipcc -O0 -ggdb --amdgpu-target=gfx900 --amdgpu-target=gfx906 \
26228 --amdgpu-target=gfx908 bit_extract.cpp -o bit_extract
26231 The AMD GPU ROCm compiler maps HIP source language work-items to the
26232 lanes of an AMD GPU wavefront, which are represented in @value{GDBN}
26233 as heterogeneous lanes.
26235 @item Assembly Code
26236 Assembly code kernels are supported.
26238 @item Other Languages
26239 Other languages, including OpenCL and Fortran, are currently supported
26240 as the minimal pseudo-language, provided they are compiled specifying
26241 at least the AMD GPU Code Object V3 and DWARF 4 formats.
26242 @xref{Unsupported Languages}.
26246 The @code{info agents} command (@pxref{Heterogeneous Debugging}) lists
26247 the following information for each @acronym{AMD GPU} heterogeneous
26248 agent (in this order):
26252 the per-inferior heterogeneous agent number assigned by @value{GDBN}
26255 the global heterogeneous agent number assigned by @value{GDBN}, if the
26256 @w{@option{-gid}} option was specified
26259 the @acronym{PCIe} slot number in @acronym{BDF, Bus:Device.Function}
26266 the number of shader engines
26269 the number of @acronym{CU, Compute Unit}
26272 the number of @acronym{SIMD, Single Instruction Multiple Data} units per @acronym{CU}
26275 the number of wavefronts per @acronym{SIMD}
26282 (@value{GDBP}) info agents
26283 Id PCI Slot Device Name Shader Engines Compute Units SIMD/CU Wavefronts/SIMD
26284 1 43:00.0 vega10 4 56 4 10
26287 @acronym{AMD GPU} heterogeneous agents are not listed until the
26288 inferior has started executing the program.
26290 The @code{info queues}, @code{info dispatches}, and @code{info
26291 packets} commands are not yet supported by @acronym{AMD GPU}.
26293 An AMD GPU wavefront is represented in @value{GDBN} as a thread.
26295 @acronym{AMD GPU} supports the following @var{reggroup} values for the
26296 @samp{info registers @var{reggroup} @dots{}} command:
26314 The number of scalar and vector registers is configured when a
26315 wavefront is created. Only allocated registers are displayed. Scalar
26316 registers are reported as 32-bit signed integer values. Vector
26317 registers are reported as a wavefront size vector of signed 32-bit
26318 values. The @code{pc} is reported as a function pointer value. The
26319 @code{exec} register is reported as a wavefront size-bit unsigned
26320 integer value. The @code{vcc} and @code{xnack_mask} pseudo registers
26321 are reported as a wavefront size-bit unsigned integer value. The
26322 @code{flat_scratch} pseudo register is reported as a 64-bit unsigned
26325 AMD GPU code objects are loaded into each AMD GPU device separately.
26326 The @code{info sharedlibrary} command will therefore show the same
26327 code object loaded multiple times. As a consequence, setting a
26328 breakpoint in AMD GPU code will result in multiple breakpoints if
26329 there are multiple AMD GPU devices.
26331 If the source language runtime defers loading code objects until
26332 kernels are launched, then setting breakpoints may result in pending
26333 breakpoints that will be set when the code object is finally loaded.
26335 Threads created on @acronym{AMD GPU} heterogeneous agents have the
26336 following identifier formats:
26341 The target system's thread identifier (@var{systag}) string has the
26345 ROCm process @var{process-num} agent @var{agent-num} queue @var{queue-num} dispatch @var{dispatch-num} work-group(@var{work-group-x},@var{work-group-y},@var{work-group-z})/@var{work-group-thread-index}
26348 @c TODO: What order should coordinates be: x,y,z or z,y,x?
26350 @item @var{lane_systag}
26351 The target system's heterogeneous lane identifier (@var{lane_systag})
26352 string has the following format:
26355 ROCm process @var{process-num} agent @var{agent-num} queue @var{queue-num} dispatch @var{dispatch-num} work-group(@var{work-group-x},@var{work-group-y},@var{work-group-z}) work-item(@var{work-item-x},@var{work-item-y},@var{work-item-z})
26358 @item @code{$_dispatch_pos}
26359 The string returned by the @code{$_dispatch_pos} debugger convenience
26360 variable has the following format:
26363 (@var{work-group-x},@var{work-group-y},@var{work-group-z})/@var{work-group-thread-index}
26366 @item @code{$_thread_workgroup_pos}
26367 The string returned by the @code{$_thread_workgroup_pos} debugger
26368 convenience variable has the following format:
26371 @var{work-group-thread-index}
26374 @item @code{$_lane_workgroup_pos}
26375 The string returned by the @code{$_lane_workgroup_pos} debugger
26376 convenience variable has the following format:
26379 (@var{work-item-x},@var{work-item-y},@var{work-item-z})
26390 the inferior process LWP number
26394 @itemx dispatch-num
26395 the per-inferior heterogeneous agent number, the per-inferior
26396 heterogeneous queue number, and the per-inferior heterogeneous
26397 dispatch number associated with the thread respectively
26400 @itemx work-group-y
26401 @itemx work-group-z
26402 the grid position of the thread's work-group within the heterogeneous
26405 @item work-group-thread-index
26406 the threads's number within the heterogeneous work-group
26411 the position of the heterogeneous lane's work-item within the
26412 heterogeneous work-group
26416 @acronym{AMD GPU} heterogeneous agents support the following address
26422 the default global virtual address space
26425 the per heterogeneous work-group shared address space (@acronym{LDS,
26429 the per heterogeneous lane private address space (Scratch)
26432 the generic address space that can access the @var{global},
26433 @var{group}, or @var{private} address spaces (Flat)
26437 The @code{set debug amd-dbgapi log-level @var{level}} command can be
26438 used to enable diagnostic messages for the AMD GPU target, where
26439 @var{level} can be:
26444 no logging is enabled
26447 fatal errors are reported
26450 fatal errors and warnings are reported
26453 fatal errors, warnings, and info messages are reported
26456 all messages are reported
26460 The @code{show debug amd-dbgapi log-level} command displays the
26461 current AMD GPU target log level.
26463 For example, the following will enable information messages and send
26464 the log to a new file:
26467 (@value{GDBP}) set debug amd-dbgapi log-level info
26468 (@value{GDBP}) set logging overwrite
26469 (@value{GDBP}) set logging file log.out
26470 (@value{GDBP}) set logging debugredirect on
26471 (@value{GDBP}) set logging on
26474 If you want to print the log to both the console and a file, ommit the
26475 @code{set the logging debugredirect} command. @xref{Logging Output}.
26477 @c TODO: Add when support available:
26479 @c The @var{AMD_???} environment variable can be set to disable the kernel
26480 @c driver from ensuring that all AMD GPU wavefronts created will fully
26481 @c support the @value{GDBN} if it attached. If AMD GPU wavefronts are
26482 @c created when support is disabled, @value{GDBN} will be unable to
26483 @c report the heterogeneous dispatch associated with the wavefront, or the
26484 @c wavefront's heterogeneous work-group position. The default is enabled.
26485 @c Disabling may very marginally improve wavefront launch latency.
26487 @value{GDBN} @acronym{AMD GPU} support is currently a prototype and
26488 has the following restrictions. Future releases aim to address these
26494 The debugger convenience variables, convenience functions, and
26495 commands described in @ref{Heterogeneous Debugging} are not yet
26496 implemented. The exception is the @code{info agents} command, which
26497 only currently supports the textual MI interface and does not have a
26500 However, the debugger convenience variable @code{$_wave_id} is
26501 available which returns a string that has the format:
26504 (@var{work-group-z},@var{work-group-y},@var{work-group-x})/@var{work-group-thread-index}
26512 @itemx work-group-y
26513 @itemx work-group-z
26514 the grid position of the thread's work-group within the heterogeneous
26517 @item work-group-thread-index
26518 the threads's number within the heterogeneous work-group
26522 The AMD GPU system's thread identifier (@var{systag}) string format
26523 differs from that described above, and currently has the following
26527 AMDGPU Thread @var{dispatch-num}.@var{wave-num} (@var{work-group-z},@var{work-group-y},@var{work-group-x})/@var{work-group-thread-index}
26535 the thread's per-heterogeneous queue ROCm AQL packet number of the
26536 associated dispatch packet
26539 the thread's per-inferior AMD GPU target wavefront number
26542 @itemx work-group-y
26543 @itemx work-group-z
26544 the grid position of the thread's work-group within the heterogeneous
26547 @item work-group-thread-index
26548 the threads's number within the heterogeneous work-group
26552 The address space qualification of addresses described in
26553 @ref{Heterogeneous Debugging} is not implemented. However, the
26554 default address space for AMD GPU threads is @code{generic}. This
26555 allows a generic address to be used to read or write in the
26556 @code{global}, @code{group}, or @code{private} address spaces. For
26557 the ROCm release the AMD GPU generic address value for @code{global}
26558 addresses is the same, for @code{group} addresses it has the most
26559 significant 32-bits of the address set to 0x00010000, and for
26560 @code{private} addresses is has the host significant 32-bits of the
26561 address set to 0x00020000. A generic private address only accesses
26562 lane 0 of the currently focused wavefront. A group address accesses
26563 the @code{group} segment memory shared by all wavefronts that are
26564 members of the same work-group as the currently focused wavefront.
26567 The AMD GPU ROCm release compiler currently does not yet support
26568 generating valid DWARF information for symbolic variables and call
26569 frame information. As a consequence:
26574 Source variables or expressions cannot be specified in any command,
26575 such as the @code{print} command and breakpoint conditions. This
26576 includes static variables, local variables, function arguments, and
26577 any language types. However, global symbols for functions and
26578 variables can be specified, and source line information is available.
26581 The @code{backtrace} command can only show the current frame and
26582 parent frames that are fully inlined. Function or kernel arguments
26583 will not be displayed and instead an empty formal argument list may be
26587 The @code{next} command may not step over function calls, but instead
26588 stop at the first statement of the called function.
26591 Breakpoints are only reported for wavefronts. There is no support for
26592 HIP work-items that are mapped to heterogeneous lanes. The HIP
26593 work-item ID of a heterogeneous lane is not available.
26597 The AMD GPU ROCm compiler currently adds the
26598 @w{@option{-gline-tables-only}} @w{@option{-disable-O0-noinline}}
26599 @w{@option{-disable-O0-optnone}}
26600 @w{@option{-amdgpu-spill-cfi-saved-regs}} options when the
26601 @w{@option{-ggdb}} option is specified. These ensure source line
26602 information is generated, but not invalid DWARF, full inlining is
26603 performed, even at @w{@option{-O0}}, and registers not currently
26604 supported by the CFI generation are saved so the CFI information is
26605 correct. If these options are not used the invalid DWARF may cause
26606 @value{GDBN} to report that it is unable to read memory (such as when
26607 reading arguments in a backtrace), and may limit the backtrace to only
26610 @value{GDBN} does not currently support the AMD GPU compiler
26611 genenerated CFI information. The options to force full inlining allow
26612 the backtrace to be available even without the CFI support. Note that
26613 even with @w{@option{-ggdb}}, functions marked @code{noinline} may
26614 result in function call frames which will prevent a full backtrace.
26615 If function calls are not inlined, the @code{next} command may report
26616 errors inserting breakpoints when stepping over calls due to the
26617 missing CFI support.
26620 Only AMD GPU Code Object V3 and above is supported. This is the
26621 default for the AMD GPU ROCm release compiler. The following error
26622 will be reported for incompatible code objects:
26625 Error while mapping shared library sections:
26626 `file:///rocm/bit_extract#offset=6751&size=3136': ELF file ABI version (0) is not supported.
26630 DWARF 5 is not yet supported. There is no support for compressed or split
26633 DWARF 4 is the default for the AMD GPU ROCm release compiler.
26636 No support yet for AMD GPU core dumps.
26639 The @code{watch} command is not yet support on AMD GPU devices.
26642 When in all-stop mode, AMD GPU does not currently prevent new
26643 wavefronts from being created, which may report breakpoints being hit.
26644 However, @value{GDBN} is configured by default to not remove
26645 breakpoints when at the command line in all-stop mode. This prevents
26646 breakpoints being missed by wavefronts created after at the command
26647 line in all-stop mode. The @code{set breakpoint always-inserted on}
26648 command can be used to change the default to remove breakpoints when
26649 at the command line in all-stop mode, but this may result in new
26650 wavefronts missing breakpoints.
26653 The performance of resuming from a breakpoint when a large number of
26654 threads have hit a breakpoint can currently take up to 10 seconds on a
26655 fully occupied single AMD GPU device. The techniques described in
26656 @xref{Heterogeneous Debugging} can be used to mitigate this. Once
26657 continued from the first breakpoint hit, the responsiveness of
26658 commands normally is better. Other techniques that can improve
26659 responsiveness are:
26664 Try to avoid having a lot of threads stopping at a breakpoint. For
26665 example, by placing breakpoints in conditional paths only executed by
26669 Use of @code{tbreak} so only one thread reports the breakpoint and the
26670 other threads hitting the breakpoint will be continued. A similar
26671 effect can be achieved by deleting the breakpoint manually when it is
26675 Reduce the number of wavefronts when debugging if practical.
26680 Currently each AMD GPU device can only be in use by one process that
26681 is being debugged by @value{GDBN}. The Linux @emph{cgroups} facility
26682 can be used to limit which AMD GPU devices are used by a process. In
26683 order for a @value{GDBN} process to access the AMD GPU devices of the
26684 process it is debugging, the AMD GPU devices must be included in the
26685 @value{GDBN} process @emph{cgroup}.
26687 Therefore, multiple @value{GDBN} processes can each debug a process
26688 provided the @emph{cgroups} specify disjoint sets of AMD GPU devices.
26689 However, a single @value{GDBN} process cannot debug multiple inferiors
26690 that use AMD GPU devices even if those inferiors have @emph{cgroups}
26691 that specify disjoint AMD GPU devices. This is because the
26692 @value{GDBN} process must have all the AMD GPU devices in its
26693 @emph{cgroups} and so will attempt to enable debugging for all AMD GPU
26694 devices for all inferiors it is debugging.
26696 The @code{HIP_VISIBLE_DEVICES} environment variable can also be used
26697 to limit the visible GPUs used by the HIP runtime. For example,
26700 export HIP_VISIBLE_DEVICES=0
26704 Currently the @code{flat_scratch} and @code{xnack_mask} special scalar
26705 registers are only accessible using their scalar register numbers and
26706 not by their register names. This will not match the assembly source
26707 text which uses register names.
26710 The @code{until} command does not work when multiple AMD GPUs are
26711 present as @value{GDBN} has limitations when there are multiple code
26712 objects that have the same breakpoint set. The work around is to use
26713 @samp{tbreak @var{line}; continue}.
26716 The HIP runtime currently performs deferred code object loading by
26717 default. AMD GPU code objects are not loaded until the first kernel
26718 is launched. Before then, all breakpoints have to be set as pending
26719 breakpoints using source line positions.
26721 The @code{HIP_DISABLE_LAZY_KERNEL_LOADING} environment variable can be
26722 used to disable deferred code object loading by the HIP runtime. This
26723 allows breakpoints to be set in AMD GPU code as soon as the inferior
26724 reaches the @code{main} funtion.
26729 export HIP_DISABLE_LAZY_KERNEL_LOADING=1
26733 Memory violations are reported to the wavefronts that cause them.
26734 However, the program location at which they are reported by be after
26735 the source statement that caused them. The ROCm runtime can currently
26736 cause the inferior to terminate before the memory violation is
26737 reported. This can be avoided by setting a breakpoint in @code{abort}
26738 and using the non-stop mode (@pxref{Non-Stop Mode}). This will
26739 prevent the ROCm runtime from terminating the inferior, while allowing
26740 @value{GDBN} to report the memory violation.
26743 @value{GDBN} does not support following a forked process.
26746 The @code{gdbserver} is not supported.
26749 No language specific support for Fortran or OpenCL. No OpenMP
26750 language extension support for C, C++, or Fortran.
26753 Does not support the AMD GPU ROCm for HIP-HCC release compiler or
26754 runtime available as part of releases before ROCm 3.5.
26757 AMD GPU does not currently support the compiler address, memory, or
26761 AMD GPU does not currently support calling inferior functions.
26764 @value{GDBN} support for AMD GPU is not currently available under
26769 @node Controlling GDB
26770 @chapter Controlling @value{GDBN}
26772 You can alter the way @value{GDBN} interacts with you by using the
26773 @code{set} command. For commands controlling how @value{GDBN} displays
26774 data, see @ref{Print Settings, ,Print Settings}. Other settings are
26779 * Editing:: Command editing
26780 * Command History:: Command history
26781 * Screen Size:: Screen size
26782 * Output Styling:: Output styling
26783 * Numbers:: Numbers
26784 * ABI:: Configuring the current ABI
26785 * Auto-loading:: Automatically loading associated files
26786 * Messages/Warnings:: Optional warnings and messages
26787 * Debugging Output:: Optional messages about internal happenings
26788 * Other Misc Settings:: Other Miscellaneous Settings
26796 @value{GDBN} indicates its readiness to read a command by printing a string
26797 called the @dfn{prompt}. This string is normally @samp{(@value{GDBP})}. You
26798 can change the prompt string with the @code{set prompt} command. For
26799 instance, when debugging @value{GDBN} with @value{GDBN}, it is useful to change
26800 the prompt in one of the @value{GDBN} sessions so that you can always tell
26801 which one you are talking to.
26803 @emph{Note:} @code{set prompt} does not add a space for you after the
26804 prompt you set. This allows you to set a prompt which ends in a space
26805 or a prompt that does not.
26809 @item set prompt @var{newprompt}
26810 Directs @value{GDBN} to use @var{newprompt} as its prompt string henceforth.
26812 @kindex show prompt
26814 Prints a line of the form: @samp{Gdb's prompt is: @var{your-prompt}}
26817 Versions of @value{GDBN} that ship with Python scripting enabled have
26818 prompt extensions. The commands for interacting with these extensions
26822 @kindex set extended-prompt
26823 @item set extended-prompt @var{prompt}
26824 Set an extended prompt that allows for substitutions.
26825 @xref{gdb.prompt}, for a list of escape sequences that can be used for
26826 substitution. Any escape sequences specified as part of the prompt
26827 string are replaced with the corresponding strings each time the prompt
26833 set extended-prompt Current working directory: \w (@value{GDBP})
26836 Note that when an extended-prompt is set, it takes control of the
26837 @var{prompt_hook} hook. @xref{prompt_hook}, for further information.
26839 @kindex show extended-prompt
26840 @item show extended-prompt
26841 Prints the extended prompt. Any escape sequences specified as part of
26842 the prompt string with @code{set extended-prompt}, are replaced with the
26843 corresponding strings each time the prompt is displayed.
26847 @section Command Editing
26849 @cindex command line editing
26851 @value{GDBN} reads its input commands via the @dfn{Readline} interface. This
26852 @sc{gnu} library provides consistent behavior for programs which provide a
26853 command line interface to the user. Advantages are @sc{gnu} Emacs-style
26854 or @dfn{vi}-style inline editing of commands, @code{csh}-like history
26855 substitution, and a storage and recall of command history across
26856 debugging sessions.
26858 You may control the behavior of command line editing in @value{GDBN} with the
26859 command @code{set}.
26862 @kindex set editing
26865 @itemx set editing on
26866 Enable command line editing (enabled by default).
26868 @item set editing off
26869 Disable command line editing.
26871 @kindex show editing
26873 Show whether command line editing is enabled.
26876 @ifset SYSTEM_READLINE
26877 @xref{Command Line Editing, , , rluserman, GNU Readline Library},
26879 @ifclear SYSTEM_READLINE
26880 @xref{Command Line Editing},
26882 for more details about the Readline
26883 interface. Users unfamiliar with @sc{gnu} Emacs or @code{vi} are
26884 encouraged to read that chapter.
26886 @cindex Readline application name
26887 @value{GDBN} sets the Readline application name to @samp{gdb}. This
26888 is useful for conditions in @file{.inputrc}.
26890 @cindex operate-and-get-next
26891 @value{GDBN} defines a bindable Readline command,
26892 @code{operate-and-get-next}. This is bound to @kbd{C-o} by default.
26893 This command accepts the current line for execution and fetches the
26894 next line relative to the current line from the history for editing.
26895 Any argument is ignored.
26897 @node Command History
26898 @section Command History
26899 @cindex command history
26901 @value{GDBN} can keep track of the commands you type during your
26902 debugging sessions, so that you can be certain of precisely what
26903 happened. Use these commands to manage the @value{GDBN} command
26906 @value{GDBN} uses the @sc{gnu} History library, a part of the Readline
26907 package, to provide the history facility.
26908 @ifset SYSTEM_READLINE
26909 @xref{Using History Interactively, , , history, GNU History Library},
26911 @ifclear SYSTEM_READLINE
26912 @xref{Using History Interactively},
26914 for the detailed description of the History library.
26916 To issue a command to @value{GDBN} without affecting certain aspects of
26917 the state which is seen by users, prefix it with @samp{server }
26918 (@pxref{Server Prefix}). This
26919 means that this command will not affect the command history, nor will it
26920 affect @value{GDBN}'s notion of which command to repeat if @key{RET} is
26921 pressed on a line by itself.
26923 @cindex @code{server}, command prefix
26924 The server prefix does not affect the recording of values into the value
26925 history; to print a value without recording it into the value history,
26926 use the @code{output} command instead of the @code{print} command.
26928 Here is the description of @value{GDBN} commands related to command
26932 @cindex history substitution
26933 @cindex history file
26934 @kindex set history filename
26935 @cindex @env{GDBHISTFILE}, environment variable
26936 @item set history filename @var{fname}
26937 Set the name of the @value{GDBN} command history file to @var{fname}.
26938 This is the file where @value{GDBN} reads an initial command history
26939 list, and where it writes the command history from this session when it
26940 exits. You can access this list through history expansion or through
26941 the history command editing characters listed below. This file defaults
26942 to the value of the environment variable @code{GDBHISTFILE}, or to
26943 @file{./.gdb_history} (@file{./_gdb_history} on MS-DOS) if this variable
26946 @cindex save command history
26947 @kindex set history save
26948 @item set history save
26949 @itemx set history save on
26950 Record command history in a file, whose name may be specified with the
26951 @code{set history filename} command. By default, this option is disabled.
26953 @item set history save off
26954 Stop recording command history in a file.
26956 @cindex history size
26957 @kindex set history size
26958 @cindex @env{GDBHISTSIZE}, environment variable
26959 @item set history size @var{size}
26960 @itemx set history size unlimited
26961 Set the number of commands which @value{GDBN} keeps in its history list.
26962 This defaults to the value of the environment variable @env{GDBHISTSIZE}, or
26963 to 256 if this variable is not set. Non-numeric values of @env{GDBHISTSIZE}
26964 are ignored. If @var{size} is @code{unlimited} or if @env{GDBHISTSIZE} is
26965 either a negative number or the empty string, then the number of commands
26966 @value{GDBN} keeps in the history list is unlimited.
26968 @cindex remove duplicate history
26969 @kindex set history remove-duplicates
26970 @item set history remove-duplicates @var{count}
26971 @itemx set history remove-duplicates unlimited
26972 Control the removal of duplicate history entries in the command history list.
26973 If @var{count} is non-zero, @value{GDBN} will look back at the last @var{count}
26974 history entries and remove the first entry that is a duplicate of the current
26975 entry being added to the command history list. If @var{count} is
26976 @code{unlimited} then this lookbehind is unbounded. If @var{count} is 0, then
26977 removal of duplicate history entries is disabled.
26979 Only history entries added during the current session are considered for
26980 removal. This option is set to 0 by default.
26984 History expansion assigns special meaning to the character @kbd{!}.
26985 @ifset SYSTEM_READLINE
26986 @xref{Event Designators, , , history, GNU History Library},
26988 @ifclear SYSTEM_READLINE
26989 @xref{Event Designators},
26993 @cindex history expansion, turn on/off
26994 Since @kbd{!} is also the logical not operator in C, history expansion
26995 is off by default. If you decide to enable history expansion with the
26996 @code{set history expansion on} command, you may sometimes need to
26997 follow @kbd{!} (when it is used as logical not, in an expression) with
26998 a space or a tab to prevent it from being expanded. The readline
26999 history facilities do not attempt substitution on the strings
27000 @kbd{!=} and @kbd{!(}, even when history expansion is enabled.
27002 The commands to control history expansion are:
27005 @item set history expansion on
27006 @itemx set history expansion
27007 @kindex set history expansion
27008 Enable history expansion. History expansion is off by default.
27010 @item set history expansion off
27011 Disable history expansion.
27014 @kindex show history
27016 @itemx show history filename
27017 @itemx show history save
27018 @itemx show history size
27019 @itemx show history expansion
27020 These commands display the state of the @value{GDBN} history parameters.
27021 @code{show history} by itself displays all four states.
27026 @kindex show commands
27027 @cindex show last commands
27028 @cindex display command history
27029 @item show commands
27030 Display the last ten commands in the command history.
27032 @item show commands @var{n}
27033 Print ten commands centered on command number @var{n}.
27035 @item show commands +
27036 Print ten commands just after the commands last printed.
27040 @section Screen Size
27041 @cindex size of screen
27042 @cindex screen size
27045 @cindex pauses in output
27047 Certain commands to @value{GDBN} may produce large amounts of
27048 information output to the screen. To help you read all of it,
27049 @value{GDBN} pauses and asks you for input at the end of each page of
27050 output. Type @key{RET} when you want to see one more page of output,
27051 @kbd{q} to discard the remaining output, or @kbd{c} to continue
27052 without paging for the rest of the current command. Also, the screen
27053 width setting determines when to wrap lines of output. Depending on
27054 what is being printed, @value{GDBN} tries to break the line at a
27055 readable place, rather than simply letting it overflow onto the
27058 Normally @value{GDBN} knows the size of the screen from the terminal
27059 driver software. For example, on Unix @value{GDBN} uses the termcap data base
27060 together with the value of the @code{TERM} environment variable and the
27061 @code{stty rows} and @code{stty cols} settings. If this is not correct,
27062 you can override it with the @code{set height} and @code{set
27069 @kindex show height
27070 @item set height @var{lpp}
27071 @itemx set height unlimited
27073 @itemx set width @var{cpl}
27074 @itemx set width unlimited
27076 These @code{set} commands specify a screen height of @var{lpp} lines and
27077 a screen width of @var{cpl} characters. The associated @code{show}
27078 commands display the current settings.
27080 If you specify a height of either @code{unlimited} or zero lines,
27081 @value{GDBN} does not pause during output no matter how long the
27082 output is. This is useful if output is to a file or to an editor
27085 Likewise, you can specify @samp{set width unlimited} or @samp{set
27086 width 0} to prevent @value{GDBN} from wrapping its output.
27088 @item set pagination on
27089 @itemx set pagination off
27090 @kindex set pagination
27091 Turn the output pagination on or off; the default is on. Turning
27092 pagination off is the alternative to @code{set height unlimited}. Note that
27093 running @value{GDBN} with the @option{--batch} option (@pxref{Mode
27094 Options, -batch}) also automatically disables pagination.
27096 @item show pagination
27097 @kindex show pagination
27098 Show the current pagination mode.
27101 @node Output Styling
27102 @section Output Styling
27108 @value{GDBN} can style its output on a capable terminal. This is
27109 enabled by default on most systems, but disabled by default when in
27110 batch mode (@pxref{Mode Options}). Various style settings are available;
27111 and styles can also be disabled entirely.
27114 @item set style enabled @samp{on|off}
27115 Enable or disable all styling. The default is host-dependent, with
27116 most hosts defaulting to @samp{on}.
27118 @item show style enabled
27119 Show the current state of styling.
27121 @item set style sources @samp{on|off}
27122 Enable or disable source code styling. This affects whether source
27123 code, such as the output of the @code{list} command, is styled. Note
27124 that source styling only works if styling in general is enabled, and
27125 if @value{GDBN} was linked with the GNU Source Highlight library. The
27126 default is @samp{on}.
27128 @item show style sources
27129 Show the current state of source code styling.
27132 Subcommands of @code{set style} control specific forms of styling.
27133 These subcommands all follow the same pattern: each style-able object
27134 can be styled with a foreground color, a background color, and an
27137 For example, the style of file names can be controlled using the
27138 @code{set style filename} group of commands:
27141 @item set style filename background @var{color}
27142 Set the background to @var{color}. Valid colors are @samp{none}
27143 (meaning the terminal's default color), @samp{black}, @samp{red},
27144 @samp{green}, @samp{yellow}, @samp{blue}, @samp{magenta}, @samp{cyan},
27147 @item set style filename foreground @var{color}
27148 Set the foreground to @var{color}. Valid colors are @samp{none}
27149 (meaning the terminal's default color), @samp{black}, @samp{red},
27150 @samp{green}, @samp{yellow}, @samp{blue}, @samp{magenta}, @samp{cyan},
27153 @item set style filename intensity @var{value}
27154 Set the intensity to @var{value}. Valid intensities are @samp{normal}
27155 (the default), @samp{bold}, and @samp{dim}.
27158 The @code{show style} command and its subcommands are styling
27159 a style name in their output using its own style.
27160 So, use @command{show style} to see the complete list of styles,
27161 their characteristics and the visual aspect of each style.
27163 The style-able objects are:
27166 Control the styling of file names. By default, this style's
27167 foreground color is green.
27170 Control the styling of function names. These are managed with the
27171 @code{set style function} family of commands. By default, this
27172 style's foreground color is yellow.
27175 Control the styling of variable names. These are managed with the
27176 @code{set style variable} family of commands. By default, this style's
27177 foreground color is cyan.
27180 Control the styling of addresses. These are managed with the
27181 @code{set style address} family of commands. By default, this style's
27182 foreground color is blue.
27185 Control the styling of titles. These are managed with the
27186 @code{set style title} family of commands. By default, this style's
27187 intensity is bold. Commands are using the title style to improve
27188 the readability of large output. For example, the commands
27189 @command{apropos} and @command{help} are using the title style
27190 for the command names.
27193 Control the styling of highlightings. These are managed with the
27194 @code{set style highlight} family of commands. By default, this style's
27195 foreground color is red. Commands are using the highlight style to draw
27196 the user attention to some specific parts of their output. For example,
27197 the command @command{apropos -v REGEXP} uses the highlight style to
27198 mark the documentation parts matching @var{regexp}.
27201 Control the styling of the TUI border. Note that, unlike other
27202 styling options, only the color of the border can be controlled via
27203 @code{set style}. This was done for compatibility reasons, as TUI
27204 controls to set the border's intensity predated the addition of
27205 general styling to @value{GDBN}. @xref{TUI Configuration}.
27207 @item tui-active-border
27208 Control the styling of the active TUI border; that is, the TUI window
27209 that has the focus.
27215 @cindex number representation
27216 @cindex entering numbers
27218 You can always enter numbers in octal, decimal, or hexadecimal in
27219 @value{GDBN} by the usual conventions: octal numbers begin with
27220 @samp{0}, decimal numbers end with @samp{.}, and hexadecimal numbers
27221 begin with @samp{0x}. Numbers that neither begin with @samp{0} or
27222 @samp{0x}, nor end with a @samp{.} are, by default, entered in base
27223 10; likewise, the default display for numbers---when no particular
27224 format is specified---is base 10. You can change the default base for
27225 both input and output with the commands described below.
27228 @kindex set input-radix
27229 @item set input-radix @var{base}
27230 Set the default base for numeric input. Supported choices
27231 for @var{base} are decimal 8, 10, or 16. The base must itself be
27232 specified either unambiguously or using the current input radix; for
27236 set input-radix 012
27237 set input-radix 10.
27238 set input-radix 0xa
27242 sets the input base to decimal. On the other hand, @samp{set input-radix 10}
27243 leaves the input radix unchanged, no matter what it was, since
27244 @samp{10}, being without any leading or trailing signs of its base, is
27245 interpreted in the current radix. Thus, if the current radix is 16,
27246 @samp{10} is interpreted in hex, i.e.@: as 16 decimal, which doesn't
27249 @kindex set output-radix
27250 @item set output-radix @var{base}
27251 Set the default base for numeric display. Supported choices
27252 for @var{base} are decimal 8, 10, or 16. The base must itself be
27253 specified either unambiguously or using the current input radix.
27255 @kindex show input-radix
27256 @item show input-radix
27257 Display the current default base for numeric input.
27259 @kindex show output-radix
27260 @item show output-radix
27261 Display the current default base for numeric display.
27263 @item set radix @r{[}@var{base}@r{]}
27267 These commands set and show the default base for both input and output
27268 of numbers. @code{set radix} sets the radix of input and output to
27269 the same base; without an argument, it resets the radix back to its
27270 default value of 10.
27275 @section Configuring the Current ABI
27277 @value{GDBN} can determine the @dfn{ABI} (Application Binary Interface) of your
27278 application automatically. However, sometimes you need to override its
27279 conclusions. Use these commands to manage @value{GDBN}'s view of the
27285 @cindex Newlib OS ABI and its influence on the longjmp handling
27287 One @value{GDBN} configuration can debug binaries for multiple operating
27288 system targets, either via remote debugging or native emulation.
27289 @value{GDBN} will autodetect the @dfn{OS ABI} (Operating System ABI) in use,
27290 but you can override its conclusion using the @code{set osabi} command.
27291 One example where this is useful is in debugging of binaries which use
27292 an alternate C library (e.g.@: @sc{uClibc} for @sc{gnu}/Linux) which does
27293 not have the same identifying marks that the standard C library for your
27296 When @value{GDBN} is debugging the AArch64 architecture, it provides a
27297 ``Newlib'' OS ABI. This is useful for handling @code{setjmp} and
27298 @code{longjmp} when debugging binaries that use the @sc{newlib} C library.
27299 The ``Newlib'' OS ABI can be selected by @code{set osabi Newlib}.
27303 Show the OS ABI currently in use.
27306 With no argument, show the list of registered available OS ABI's.
27308 @item set osabi @var{abi}
27309 Set the current OS ABI to @var{abi}.
27312 @cindex float promotion
27314 Generally, the way that an argument of type @code{float} is passed to a
27315 function depends on whether the function is prototyped. For a prototyped
27316 (i.e.@: ANSI/ISO style) function, @code{float} arguments are passed unchanged,
27317 according to the architecture's convention for @code{float}. For unprototyped
27318 (i.e.@: K&R style) functions, @code{float} arguments are first promoted to type
27319 @code{double} and then passed.
27321 Unfortunately, some forms of debug information do not reliably indicate whether
27322 a function is prototyped. If @value{GDBN} calls a function that is not marked
27323 as prototyped, it consults @kbd{set coerce-float-to-double}.
27326 @kindex set coerce-float-to-double
27327 @item set coerce-float-to-double
27328 @itemx set coerce-float-to-double on
27329 Arguments of type @code{float} will be promoted to @code{double} when passed
27330 to an unprototyped function. This is the default setting.
27332 @item set coerce-float-to-double off
27333 Arguments of type @code{float} will be passed directly to unprototyped
27336 @kindex show coerce-float-to-double
27337 @item show coerce-float-to-double
27338 Show the current setting of promoting @code{float} to @code{double}.
27342 @kindex show cp-abi
27343 @value{GDBN} needs to know the ABI used for your program's C@t{++}
27344 objects. The correct C@t{++} ABI depends on which C@t{++} compiler was
27345 used to build your application. @value{GDBN} only fully supports
27346 programs with a single C@t{++} ABI; if your program contains code using
27347 multiple C@t{++} ABI's or if @value{GDBN} can not identify your
27348 program's ABI correctly, you can tell @value{GDBN} which ABI to use.
27349 Currently supported ABI's include ``gnu-v2'', for @code{g++} versions
27350 before 3.0, ``gnu-v3'', for @code{g++} versions 3.0 and later, and
27351 ``hpaCC'' for the HP ANSI C@t{++} compiler. Other C@t{++} compilers may
27352 use the ``gnu-v2'' or ``gnu-v3'' ABI's as well. The default setting is
27357 Show the C@t{++} ABI currently in use.
27360 With no argument, show the list of supported C@t{++} ABI's.
27362 @item set cp-abi @var{abi}
27363 @itemx set cp-abi auto
27364 Set the current C@t{++} ABI to @var{abi}, or return to automatic detection.
27368 @section Automatically loading associated files
27369 @cindex auto-loading
27371 @value{GDBN} sometimes reads files with commands and settings automatically,
27372 without being explicitly told so by the user. We call this feature
27373 @dfn{auto-loading}. While auto-loading is useful for automatically adapting
27374 @value{GDBN} to the needs of your project, it can sometimes produce unexpected
27375 results or introduce security risks (e.g., if the file comes from untrusted
27379 * Init File in the Current Directory:: @samp{set/show/info auto-load local-gdbinit}
27380 * libthread_db.so.1 file:: @samp{set/show/info auto-load libthread-db}
27382 * Auto-loading safe path:: @samp{set/show/info auto-load safe-path}
27383 * Auto-loading verbose mode:: @samp{set/show debug auto-load}
27386 There are various kinds of files @value{GDBN} can automatically load.
27387 In addition to these files, @value{GDBN} supports auto-loading code written
27388 in various extension languages. @xref{Auto-loading extensions}.
27390 Note that loading of these associated files (including the local @file{.gdbinit}
27391 file) requires accordingly configured @code{auto-load safe-path}
27392 (@pxref{Auto-loading safe path}).
27394 For these reasons, @value{GDBN} includes commands and options to let you
27395 control when to auto-load files and which files should be auto-loaded.
27398 @anchor{set auto-load off}
27399 @kindex set auto-load off
27400 @item set auto-load off
27401 Globally disable loading of all auto-loaded files.
27402 You may want to use this command with the @samp{-iex} option
27403 (@pxref{Option -init-eval-command}) such as:
27405 $ @kbd{gdb -iex "set auto-load off" untrusted-executable corefile}
27408 Be aware that system init file (@pxref{System-wide configuration})
27409 and init files from your home directory (@pxref{Home Directory Init File})
27410 still get read (as they come from generally trusted directories).
27411 To prevent @value{GDBN} from auto-loading even those init files, use the
27412 @option{-nx} option (@pxref{Mode Options}), in addition to
27413 @code{set auto-load no}.
27415 @anchor{show auto-load}
27416 @kindex show auto-load
27417 @item show auto-load
27418 Show whether auto-loading of each specific @samp{auto-load} file(s) is enabled
27422 (@value{GDBP}) show auto-load
27423 gdb-scripts: Auto-loading of canned sequences of commands scripts is on.
27424 libthread-db: Auto-loading of inferior specific libthread_db is on.
27425 local-gdbinit: Auto-loading of .gdbinit script from current directory
27427 python-scripts: Auto-loading of Python scripts is on.
27428 safe-path: List of directories from which it is safe to auto-load files
27429 is $debugdir:$datadir/auto-load.
27430 scripts-directory: List of directories from which to load auto-loaded scripts
27431 is $debugdir:$datadir/auto-load.
27434 @anchor{info auto-load}
27435 @kindex info auto-load
27436 @item info auto-load
27437 Print whether each specific @samp{auto-load} file(s) have been auto-loaded or
27441 (@value{GDBP}) info auto-load
27444 Yes /home/user/gdb/gdb-gdb.gdb
27445 libthread-db: No auto-loaded libthread-db.
27446 local-gdbinit: Local .gdbinit file "/home/user/gdb/.gdbinit" has been
27450 Yes /home/user/gdb/gdb-gdb.py
27454 These are @value{GDBN} control commands for the auto-loading:
27456 @multitable @columnfractions .5 .5
27457 @item @xref{set auto-load off}.
27458 @tab Disable auto-loading globally.
27459 @item @xref{show auto-load}.
27460 @tab Show setting of all kinds of files.
27461 @item @xref{info auto-load}.
27462 @tab Show state of all kinds of files.
27463 @item @xref{set auto-load gdb-scripts}.
27464 @tab Control for @value{GDBN} command scripts.
27465 @item @xref{show auto-load gdb-scripts}.
27466 @tab Show setting of @value{GDBN} command scripts.
27467 @item @xref{info auto-load gdb-scripts}.
27468 @tab Show state of @value{GDBN} command scripts.
27469 @item @xref{set auto-load python-scripts}.
27470 @tab Control for @value{GDBN} Python scripts.
27471 @item @xref{show auto-load python-scripts}.
27472 @tab Show setting of @value{GDBN} Python scripts.
27473 @item @xref{info auto-load python-scripts}.
27474 @tab Show state of @value{GDBN} Python scripts.
27475 @item @xref{set auto-load guile-scripts}.
27476 @tab Control for @value{GDBN} Guile scripts.
27477 @item @xref{show auto-load guile-scripts}.
27478 @tab Show setting of @value{GDBN} Guile scripts.
27479 @item @xref{info auto-load guile-scripts}.
27480 @tab Show state of @value{GDBN} Guile scripts.
27481 @item @xref{set auto-load scripts-directory}.
27482 @tab Control for @value{GDBN} auto-loaded scripts location.
27483 @item @xref{show auto-load scripts-directory}.
27484 @tab Show @value{GDBN} auto-loaded scripts location.
27485 @item @xref{add-auto-load-scripts-directory}.
27486 @tab Add directory for auto-loaded scripts location list.
27487 @item @xref{set auto-load local-gdbinit}.
27488 @tab Control for init file in the current directory.
27489 @item @xref{show auto-load local-gdbinit}.
27490 @tab Show setting of init file in the current directory.
27491 @item @xref{info auto-load local-gdbinit}.
27492 @tab Show state of init file in the current directory.
27493 @item @xref{set auto-load libthread-db}.
27494 @tab Control for thread debugging library.
27495 @item @xref{show auto-load libthread-db}.
27496 @tab Show setting of thread debugging library.
27497 @item @xref{info auto-load libthread-db}.
27498 @tab Show state of thread debugging library.
27499 @item @xref{set auto-load safe-path}.
27500 @tab Control directories trusted for automatic loading.
27501 @item @xref{show auto-load safe-path}.
27502 @tab Show directories trusted for automatic loading.
27503 @item @xref{add-auto-load-safe-path}.
27504 @tab Add directory trusted for automatic loading.
27507 @node Init File in the Current Directory
27508 @subsection Automatically loading init file in the current directory
27509 @cindex auto-loading init file in the current directory
27511 By default, @value{GDBN} reads and executes the canned sequences of commands
27512 from init file (if any) in the current working directory,
27513 see @ref{Init File in the Current Directory during Startup}.
27515 Note that loading of this local @file{.gdbinit} file also requires accordingly
27516 configured @code{auto-load safe-path} (@pxref{Auto-loading safe path}).
27519 @anchor{set auto-load local-gdbinit}
27520 @kindex set auto-load local-gdbinit
27521 @item set auto-load local-gdbinit [on|off]
27522 Enable or disable the auto-loading of canned sequences of commands
27523 (@pxref{Sequences}) found in init file in the current directory.
27525 @anchor{show auto-load local-gdbinit}
27526 @kindex show auto-load local-gdbinit
27527 @item show auto-load local-gdbinit
27528 Show whether auto-loading of canned sequences of commands from init file in the
27529 current directory is enabled or disabled.
27531 @anchor{info auto-load local-gdbinit}
27532 @kindex info auto-load local-gdbinit
27533 @item info auto-load local-gdbinit
27534 Print whether canned sequences of commands from init file in the
27535 current directory have been auto-loaded.
27538 @node libthread_db.so.1 file
27539 @subsection Automatically loading thread debugging library
27540 @cindex auto-loading libthread_db.so.1
27542 This feature is currently present only on @sc{gnu}/Linux native hosts.
27544 @value{GDBN} reads in some cases thread debugging library from places specific
27545 to the inferior (@pxref{set libthread-db-search-path}).
27547 The special @samp{libthread-db-search-path} entry @samp{$sdir} is processed
27548 without checking this @samp{set auto-load libthread-db} switch as system
27549 libraries have to be trusted in general. In all other cases of
27550 @samp{libthread-db-search-path} entries @value{GDBN} checks first if @samp{set
27551 auto-load libthread-db} is enabled before trying to open such thread debugging
27554 Note that loading of this debugging library also requires accordingly configured
27555 @code{auto-load safe-path} (@pxref{Auto-loading safe path}).
27558 @anchor{set auto-load libthread-db}
27559 @kindex set auto-load libthread-db
27560 @item set auto-load libthread-db [on|off]
27561 Enable or disable the auto-loading of inferior specific thread debugging library.
27563 @anchor{show auto-load libthread-db}
27564 @kindex show auto-load libthread-db
27565 @item show auto-load libthread-db
27566 Show whether auto-loading of inferior specific thread debugging library is
27567 enabled or disabled.
27569 @anchor{info auto-load libthread-db}
27570 @kindex info auto-load libthread-db
27571 @item info auto-load libthread-db
27572 Print the list of all loaded inferior specific thread debugging libraries and
27573 for each such library print list of inferior @var{pid}s using it.
27576 @node Auto-loading safe path
27577 @subsection Security restriction for auto-loading
27578 @cindex auto-loading safe-path
27580 As the files of inferior can come from untrusted source (such as submitted by
27581 an application user) @value{GDBN} does not always load any files automatically.
27582 @value{GDBN} provides the @samp{set auto-load safe-path} setting to list
27583 directories trusted for loading files not explicitly requested by user.
27584 Each directory can also be a shell wildcard pattern.
27586 If the path is not set properly you will see a warning and the file will not
27591 Reading symbols from /home/user/gdb/gdb...done.
27592 warning: File "/home/user/gdb/gdb-gdb.gdb" auto-loading has been
27593 declined by your `auto-load safe-path' set
27594 to "$debugdir:$datadir/auto-load".
27595 warning: File "/home/user/gdb/gdb-gdb.py" auto-loading has been
27596 declined by your `auto-load safe-path' set
27597 to "$debugdir:$datadir/auto-load".
27601 To instruct @value{GDBN} to go ahead and use the init files anyway,
27602 invoke @value{GDBN} like this:
27605 $ gdb -q -iex "set auto-load safe-path /home/user/gdb" ./gdb
27608 The list of trusted directories is controlled by the following commands:
27611 @anchor{set auto-load safe-path}
27612 @kindex set auto-load safe-path
27613 @item set auto-load safe-path @r{[}@var{directories}@r{]}
27614 Set the list of directories (and their subdirectories) trusted for automatic
27615 loading and execution of scripts. You can also enter a specific trusted file.
27616 Each directory can also be a shell wildcard pattern; wildcards do not match
27617 directory separator - see @code{FNM_PATHNAME} for system function @code{fnmatch}
27618 (@pxref{Wildcard Matching, fnmatch, , libc, GNU C Library Reference Manual}).
27619 If you omit @var{directories}, @samp{auto-load safe-path} will be reset to
27620 its default value as specified during @value{GDBN} compilation.
27622 The list of directories uses path separator (@samp{:} on GNU and Unix
27623 systems, @samp{;} on MS-Windows and MS-DOS) to separate directories, similarly
27624 to the @env{PATH} environment variable.
27626 @anchor{show auto-load safe-path}
27627 @kindex show auto-load safe-path
27628 @item show auto-load safe-path
27629 Show the list of directories trusted for automatic loading and execution of
27632 @anchor{add-auto-load-safe-path}
27633 @kindex add-auto-load-safe-path
27634 @item add-auto-load-safe-path
27635 Add an entry (or list of entries) to the list of directories trusted for
27636 automatic loading and execution of scripts. Multiple entries may be delimited
27637 by the host platform path separator in use.
27640 This variable defaults to what @code{--with-auto-load-dir} has been configured
27641 to (@pxref{with-auto-load-dir}). @file{$debugdir} and @file{$datadir}
27642 substitution applies the same as for @ref{set auto-load scripts-directory}.
27643 The default @code{set auto-load safe-path} value can be also overriden by
27644 @value{GDBN} configuration option @option{--with-auto-load-safe-path}.
27646 Setting this variable to @file{/} disables this security protection,
27647 corresponding @value{GDBN} configuration option is
27648 @option{--without-auto-load-safe-path}.
27649 This variable is supposed to be set to the system directories writable by the
27650 system superuser only. Users can add their source directories in init files in
27651 their home directories (@pxref{Home Directory Init File}). See also deprecated
27652 init file in the current directory
27653 (@pxref{Init File in the Current Directory during Startup}).
27655 To force @value{GDBN} to load the files it declined to load in the previous
27656 example, you could use one of the following ways:
27659 @item @file{~/.gdbinit}: @samp{add-auto-load-safe-path ~/src/gdb}
27660 Specify this trusted directory (or a file) as additional component of the list.
27661 You have to specify also any existing directories displayed by
27662 by @samp{show auto-load safe-path} (such as @samp{/usr:/bin} in this example).
27664 @item @kbd{gdb -iex "set auto-load safe-path /usr:/bin:~/src/gdb" @dots{}}
27665 Specify this directory as in the previous case but just for a single
27666 @value{GDBN} session.
27668 @item @kbd{gdb -iex "set auto-load safe-path /" @dots{}}
27669 Disable auto-loading safety for a single @value{GDBN} session.
27670 This assumes all the files you debug during this @value{GDBN} session will come
27671 from trusted sources.
27673 @item @kbd{./configure --without-auto-load-safe-path}
27674 During compilation of @value{GDBN} you may disable any auto-loading safety.
27675 This assumes all the files you will ever debug with this @value{GDBN} come from
27679 On the other hand you can also explicitly forbid automatic files loading which
27680 also suppresses any such warning messages:
27683 @item @kbd{gdb -iex "set auto-load no" @dots{}}
27684 You can use @value{GDBN} command-line option for a single @value{GDBN} session.
27686 @item @file{~/.gdbinit}: @samp{set auto-load no}
27687 Disable auto-loading globally for the user
27688 (@pxref{Home Directory Init File}). While it is improbable, you could also
27689 use system init file instead (@pxref{System-wide configuration}).
27692 This setting applies to the file names as entered by user. If no entry matches
27693 @value{GDBN} tries as a last resort to also resolve all the file names into
27694 their canonical form (typically resolving symbolic links) and compare the
27695 entries again. @value{GDBN} already canonicalizes most of the filenames on its
27696 own before starting the comparison so a canonical form of directories is
27697 recommended to be entered.
27699 @node Auto-loading verbose mode
27700 @subsection Displaying files tried for auto-load
27701 @cindex auto-loading verbose mode
27703 For better visibility of all the file locations where you can place scripts to
27704 be auto-loaded with inferior --- or to protect yourself against accidental
27705 execution of untrusted scripts --- @value{GDBN} provides a feature for printing
27706 all the files attempted to be loaded. Both existing and non-existing files may
27709 For example the list of directories from which it is safe to auto-load files
27710 (@pxref{Auto-loading safe path}) applies also to canonicalized filenames which
27711 may not be too obvious while setting it up.
27714 (@value{GDBP}) set debug auto-load on
27715 (@value{GDBP}) file ~/src/t/true
27716 auto-load: Loading canned sequences of commands script "/tmp/true-gdb.gdb"
27717 for objfile "/tmp/true".
27718 auto-load: Updating directories of "/usr:/opt".
27719 auto-load: Using directory "/usr".
27720 auto-load: Using directory "/opt".
27721 warning: File "/tmp/true-gdb.gdb" auto-loading has been declined
27722 by your `auto-load safe-path' set to "/usr:/opt".
27726 @anchor{set debug auto-load}
27727 @kindex set debug auto-load
27728 @item set debug auto-load [on|off]
27729 Set whether to print the filenames attempted to be auto-loaded.
27731 @anchor{show debug auto-load}
27732 @kindex show debug auto-load
27733 @item show debug auto-load
27734 Show whether printing of the filenames attempted to be auto-loaded is turned
27738 @node Messages/Warnings
27739 @section Optional Warnings and Messages
27741 @cindex verbose operation
27742 @cindex optional warnings
27743 By default, @value{GDBN} is silent about its inner workings. If you are
27744 running on a slow machine, you may want to use the @code{set verbose}
27745 command. This makes @value{GDBN} tell you when it does a lengthy
27746 internal operation, so you will not think it has crashed.
27748 Currently, the messages controlled by @code{set verbose} are those
27749 which announce that the symbol table for a source file is being read;
27750 see @code{symbol-file} in @ref{Files, ,Commands to Specify Files}.
27753 @kindex set verbose
27754 @item set verbose on
27755 Enables @value{GDBN} output of certain informational messages.
27757 @item set verbose off
27758 Disables @value{GDBN} output of certain informational messages.
27760 @kindex show verbose
27762 Displays whether @code{set verbose} is on or off.
27765 By default, if @value{GDBN} encounters bugs in the symbol table of an
27766 object file, it is silent; but if you are debugging a compiler, you may
27767 find this information useful (@pxref{Symbol Errors, ,Errors Reading
27772 @kindex set complaints
27773 @item set complaints @var{limit}
27774 Permits @value{GDBN} to output @var{limit} complaints about each type of
27775 unusual symbols before becoming silent about the problem. Set
27776 @var{limit} to zero to suppress all complaints; set it to a large number
27777 to prevent complaints from being suppressed.
27779 @kindex show complaints
27780 @item show complaints
27781 Displays how many symbol complaints @value{GDBN} is permitted to produce.
27785 @anchor{confirmation requests}
27786 By default, @value{GDBN} is cautious, and asks what sometimes seems to be a
27787 lot of stupid questions to confirm certain commands. For example, if
27788 you try to run a program which is already running:
27792 The program being debugged has been started already.
27793 Start it from the beginning? (y or n)
27796 If you are willing to unflinchingly face the consequences of your own
27797 commands, you can disable this ``feature'':
27801 @kindex set confirm
27803 @cindex confirmation
27804 @cindex stupid questions
27805 @item set confirm off
27806 Disables confirmation requests. Note that running @value{GDBN} with
27807 the @option{--batch} option (@pxref{Mode Options, -batch}) also
27808 automatically disables confirmation requests.
27810 @item set confirm on
27811 Enables confirmation requests (the default).
27813 @kindex show confirm
27815 Displays state of confirmation requests.
27819 @cindex command tracing
27820 If you need to debug user-defined commands or sourced files you may find it
27821 useful to enable @dfn{command tracing}. In this mode each command will be
27822 printed as it is executed, prefixed with one or more @samp{+} symbols, the
27823 quantity denoting the call depth of each command.
27826 @kindex set trace-commands
27827 @cindex command scripts, debugging
27828 @item set trace-commands on
27829 Enable command tracing.
27830 @item set trace-commands off
27831 Disable command tracing.
27832 @item show trace-commands
27833 Display the current state of command tracing.
27836 @node Debugging Output
27837 @section Optional Messages about Internal Happenings
27838 @cindex optional debugging messages
27840 @value{GDBN} has commands that enable optional debugging messages from
27841 various @value{GDBN} subsystems; normally these commands are of
27842 interest to @value{GDBN} maintainers, or when reporting a bug. This
27843 section documents those commands.
27846 @kindex set exec-done-display
27847 @item set exec-done-display
27848 Turns on or off the notification of asynchronous commands'
27849 completion. When on, @value{GDBN} will print a message when an
27850 asynchronous command finishes its execution. The default is off.
27851 @kindex show exec-done-display
27852 @item show exec-done-display
27853 Displays the current setting of asynchronous command completion
27856 @cindex ARM AArch64
27857 @item set debug aarch64
27858 Turns on or off display of debugging messages related to ARM AArch64.
27859 The default is off.
27861 @item show debug aarch64
27862 Displays the current state of displaying debugging messages related to
27864 @cindex gdbarch debugging info
27865 @cindex architecture debugging info
27866 @item set debug arch
27867 Turns on or off display of gdbarch debugging info. The default is off
27868 @item show debug arch
27869 Displays the current state of displaying gdbarch debugging info.
27870 @item set debug aix-solib
27871 @cindex AIX shared library debugging
27872 Control display of debugging messages from the AIX shared library
27873 support module. The default is off.
27874 @item show debug aix-thread
27875 Show the current state of displaying AIX shared library debugging messages.
27876 @item set debug aix-thread
27877 @cindex AIX threads
27878 Display debugging messages about inner workings of the AIX thread
27880 @item show debug aix-thread
27881 Show the current state of AIX thread debugging info display.
27882 @item set debug check-physname
27884 Check the results of the ``physname'' computation. When reading DWARF
27885 debugging information for C@t{++}, @value{GDBN} attempts to compute
27886 each entity's name. @value{GDBN} can do this computation in two
27887 different ways, depending on exactly what information is present.
27888 When enabled, this setting causes @value{GDBN} to compute the names
27889 both ways and display any discrepancies.
27890 @item show debug check-physname
27891 Show the current state of ``physname'' checking.
27892 @item set debug coff-pe-read
27893 @cindex COFF/PE exported symbols
27894 Control display of debugging messages related to reading of COFF/PE
27895 exported symbols. The default is off.
27896 @item show debug coff-pe-read
27897 Displays the current state of displaying debugging messages related to
27898 reading of COFF/PE exported symbols.
27899 @item set debug dwarf-die
27901 Dump DWARF DIEs after they are read in.
27902 The value is the number of nesting levels to print.
27903 A value of zero turns off the display.
27904 @item show debug dwarf-die
27905 Show the current state of DWARF DIE debugging.
27906 @item set debug dwarf-line
27907 @cindex DWARF Line Tables
27908 Turns on or off display of debugging messages related to reading
27909 DWARF line tables. The default is 0 (off).
27910 A value of 1 provides basic information.
27911 A value greater than 1 provides more verbose information.
27912 @item show debug dwarf-line
27913 Show the current state of DWARF line table debugging.
27914 @item set debug dwarf-read
27915 @cindex DWARF Reading
27916 Turns on or off display of debugging messages related to reading
27917 DWARF debug info. The default is 0 (off).
27918 A value of 1 provides basic information.
27919 A value greater than 1 provides more verbose information.
27920 @item show debug dwarf-read
27921 Show the current state of DWARF reader debugging.
27922 @item set debug displaced
27923 @cindex displaced stepping debugging info
27924 Turns on or off display of @value{GDBN} debugging info for the
27925 displaced stepping support. The default is off.
27926 @item show debug displaced
27927 Displays the current state of displaying @value{GDBN} debugging info
27928 related to displaced stepping.
27929 @item set debug event
27930 @cindex event debugging info
27931 Turns on or off display of @value{GDBN} event debugging info. The
27933 @item show debug event
27934 Displays the current state of displaying @value{GDBN} event debugging
27936 @item set debug expression
27937 @cindex expression debugging info
27938 Turns on or off display of debugging info about @value{GDBN}
27939 expression parsing. The default is off.
27940 @item show debug expression
27941 Displays the current state of displaying debugging info about
27942 @value{GDBN} expression parsing.
27943 @item set debug fbsd-lwp
27944 @cindex FreeBSD LWP debug messages
27945 Turns on or off debugging messages from the FreeBSD LWP debug support.
27946 @item show debug fbsd-lwp
27947 Show the current state of FreeBSD LWP debugging messages.
27948 @item set debug fbsd-nat
27949 @cindex FreeBSD native target debug messages
27950 Turns on or off debugging messages from the FreeBSD native target.
27951 @item show debug fbsd-nat
27952 Show the current state of FreeBSD native target debugging messages.
27953 @item set debug frame
27954 @cindex frame debugging info
27955 Turns on or off display of @value{GDBN} frame debugging info. The
27957 @item show debug frame
27958 Displays the current state of displaying @value{GDBN} frame debugging
27960 @item set debug gnu-nat
27961 @cindex @sc{gnu}/Hurd debug messages
27962 Turn on or off debugging messages from the @sc{gnu}/Hurd debug support.
27963 @item show debug gnu-nat
27964 Show the current state of @sc{gnu}/Hurd debugging messages.
27965 @item set debug infrun
27966 @cindex inferior debugging info
27967 Turns on or off display of @value{GDBN} debugging info for running the inferior.
27968 The default is off. @file{infrun.c} contains GDB's runtime state machine used
27969 for implementing operations such as single-stepping the inferior.
27970 @item show debug infrun
27971 Displays the current state of @value{GDBN} inferior debugging.
27972 @item set debug jit
27973 @cindex just-in-time compilation, debugging messages
27974 Turn on or off debugging messages from JIT debug support.
27975 @item show debug jit
27976 Displays the current state of @value{GDBN} JIT debugging.
27977 @item set debug lin-lwp
27978 @cindex @sc{gnu}/Linux LWP debug messages
27979 @cindex Linux lightweight processes
27980 Turn on or off debugging messages from the Linux LWP debug support.
27981 @item show debug lin-lwp
27982 Show the current state of Linux LWP debugging messages.
27983 @item set debug linux-namespaces
27984 @cindex @sc{gnu}/Linux namespaces debug messages
27985 Turn on or off debugging messages from the Linux namespaces debug support.
27986 @item show debug linux-namespaces
27987 Show the current state of Linux namespaces debugging messages.
27988 @item set debug mach-o
27989 @cindex Mach-O symbols processing
27990 Control display of debugging messages related to Mach-O symbols
27991 processing. The default is off.
27992 @item show debug mach-o
27993 Displays the current state of displaying debugging messages related to
27994 reading of COFF/PE exported symbols.
27995 @item set debug notification
27996 @cindex remote async notification debugging info
27997 Turn on or off debugging messages about remote async notification.
27998 The default is off.
27999 @item show debug notification
28000 Displays the current state of remote async notification debugging messages.
28001 @item set debug observer
28002 @cindex observer debugging info
28003 Turns on or off display of @value{GDBN} observer debugging. This
28004 includes info such as the notification of observable events.
28005 @item show debug observer
28006 Displays the current state of observer debugging.
28007 @item set debug overload
28008 @cindex C@t{++} overload debugging info
28009 Turns on or off display of @value{GDBN} C@t{++} overload debugging
28010 info. This includes info such as ranking of functions, etc. The default
28012 @item show debug overload
28013 Displays the current state of displaying @value{GDBN} C@t{++} overload
28015 @cindex expression parser, debugging info
28016 @cindex debug expression parser
28017 @item set debug parser
28018 Turns on or off the display of expression parser debugging output.
28019 Internally, this sets the @code{yydebug} variable in the expression
28020 parser. @xref{Tracing, , Tracing Your Parser, bison, Bison}, for
28021 details. The default is off.
28022 @item show debug parser
28023 Show the current state of expression parser debugging.
28024 @cindex packets, reporting on stdout
28025 @cindex serial connections, debugging
28026 @cindex debug remote protocol
28027 @cindex remote protocol debugging
28028 @cindex display remote packets
28029 @item set debug remote
28030 Turns on or off display of reports on all packets sent back and forth across
28031 the serial line to the remote machine. The info is printed on the
28032 @value{GDBN} standard output stream. The default is off.
28033 @item show debug remote
28034 Displays the state of display of remote packets.
28036 @item set debug remote-packet-max-chars
28037 Sets the maximum number of characters to display for each remote packet when
28038 @code{set debug remote} is on. This is useful to prevent @value{GDBN} from
28039 displaying lengthy remote packets and polluting the console.
28041 The default value is @code{512}, which means @value{GDBN} will truncate each
28042 remote packet after 512 bytes.
28044 Setting this option to @code{unlimited} will disable truncation and will output
28045 the full length of the remote packets.
28046 @item show debug remote-packet-max-chars
28047 Displays the number of bytes to output for remote packet debugging.
28049 @item set debug separate-debug-file
28050 Turns on or off display of debug output about separate debug file search.
28051 @item show debug separate-debug-file
28052 Displays the state of separate debug file search debug output.
28054 @item set debug serial
28055 Turns on or off display of @value{GDBN} serial debugging info. The
28057 @item show debug serial
28058 Displays the current state of displaying @value{GDBN} serial debugging
28060 @item set debug solib-frv
28061 @cindex FR-V shared-library debugging
28062 Turn on or off debugging messages for FR-V shared-library code.
28063 @item show debug solib-frv
28064 Display the current state of FR-V shared-library code debugging
28066 @item set debug symbol-lookup
28067 @cindex symbol lookup
28068 Turns on or off display of debugging messages related to symbol lookup.
28069 The default is 0 (off).
28070 A value of 1 provides basic information.
28071 A value greater than 1 provides more verbose information.
28072 @item show debug symbol-lookup
28073 Show the current state of symbol lookup debugging messages.
28074 @item set debug symfile
28075 @cindex symbol file functions
28076 Turns on or off display of debugging messages related to symbol file functions.
28077 The default is off. @xref{Files}.
28078 @item show debug symfile
28079 Show the current state of symbol file debugging messages.
28080 @item set debug symtab-create
28081 @cindex symbol table creation
28082 Turns on or off display of debugging messages related to symbol table creation.
28083 The default is 0 (off).
28084 A value of 1 provides basic information.
28085 A value greater than 1 provides more verbose information.
28086 @item show debug symtab-create
28087 Show the current state of symbol table creation debugging.
28088 @item set debug target
28089 @cindex target debugging info
28090 Turns on or off display of @value{GDBN} target debugging info. This info
28091 includes what is going on at the target level of GDB, as it happens. The
28092 default is 0. Set it to 1 to track events, and to 2 to also track the
28093 value of large memory transfers.
28094 @item show debug target
28095 Displays the current state of displaying @value{GDBN} target debugging
28097 @item set debug timestamp
28098 @cindex timestamping debugging info
28099 Turns on or off display of timestamps with @value{GDBN} debugging info.
28100 When enabled, seconds and microseconds are displayed before each debugging
28102 @item show debug timestamp
28103 Displays the current state of displaying timestamps with @value{GDBN}
28105 @item set debug varobj
28106 @cindex variable object debugging info
28107 Turns on or off display of @value{GDBN} variable object debugging
28108 info. The default is off.
28109 @item show debug varobj
28110 Displays the current state of displaying @value{GDBN} variable object
28112 @item set debug xml
28113 @cindex XML parser debugging
28114 Turn on or off debugging messages for built-in XML parsers.
28115 @item show debug xml
28116 Displays the current state of XML debugging messages.
28119 @node Other Misc Settings
28120 @section Other Miscellaneous Settings
28121 @cindex miscellaneous settings
28124 @kindex set interactive-mode
28125 @item set interactive-mode
28126 If @code{on}, forces @value{GDBN} to assume that GDB was started
28127 in a terminal. In practice, this means that @value{GDBN} should wait
28128 for the user to answer queries generated by commands entered at
28129 the command prompt. If @code{off}, forces @value{GDBN} to operate
28130 in the opposite mode, and it uses the default answers to all queries.
28131 If @code{auto} (the default), @value{GDBN} tries to determine whether
28132 its standard input is a terminal, and works in interactive-mode if it
28133 is, non-interactively otherwise.
28135 In the vast majority of cases, the debugger should be able to guess
28136 correctly which mode should be used. But this setting can be useful
28137 in certain specific cases, such as running a MinGW @value{GDBN}
28138 inside a cygwin window.
28140 @kindex show interactive-mode
28141 @item show interactive-mode
28142 Displays whether the debugger is operating in interactive mode or not.
28145 @node Extending GDB
28146 @chapter Extending @value{GDBN}
28147 @cindex extending GDB
28149 @value{GDBN} provides several mechanisms for extension.
28150 @value{GDBN} also provides the ability to automatically load
28151 extensions when it reads a file for debugging. This allows the
28152 user to automatically customize @value{GDBN} for the program
28156 * Sequences:: Canned Sequences of @value{GDBN} Commands
28157 * Python:: Extending @value{GDBN} using Python
28158 * Guile:: Extending @value{GDBN} using Guile
28159 * Auto-loading extensions:: Automatically loading extensions
28160 * Multiple Extension Languages:: Working with multiple extension languages
28161 * Aliases:: Creating new spellings of existing commands
28164 To facilitate the use of extension languages, @value{GDBN} is capable
28165 of evaluating the contents of a file. When doing so, @value{GDBN}
28166 can recognize which extension language is being used by looking at
28167 the filename extension. Files with an unrecognized filename extension
28168 are always treated as a @value{GDBN} Command Files.
28169 @xref{Command Files,, Command files}.
28171 You can control how @value{GDBN} evaluates these files with the following
28175 @kindex set script-extension
28176 @kindex show script-extension
28177 @item set script-extension off
28178 All scripts are always evaluated as @value{GDBN} Command Files.
28180 @item set script-extension soft
28181 The debugger determines the scripting language based on filename
28182 extension. If this scripting language is supported, @value{GDBN}
28183 evaluates the script using that language. Otherwise, it evaluates
28184 the file as a @value{GDBN} Command File.
28186 @item set script-extension strict
28187 The debugger determines the scripting language based on filename
28188 extension, and evaluates the script using that language. If the
28189 language is not supported, then the evaluation fails.
28191 @item show script-extension
28192 Display the current value of the @code{script-extension} option.
28196 @ifset SYSTEM_GDBINIT_DIR
28197 This setting is not used for files in the system-wide gdbinit directory.
28198 Files in that directory must have an extension matching their language,
28199 or have a @file{.gdb} extension to be interpreted as regular @value{GDBN}
28200 commands. @xref{Startup}.
28204 @section Canned Sequences of Commands
28206 Aside from breakpoint commands (@pxref{Break Commands, ,Breakpoint
28207 Command Lists}), @value{GDBN} provides two ways to store sequences of
28208 commands for execution as a unit: user-defined commands and command
28212 * Define:: How to define your own commands
28213 * Hooks:: Hooks for user-defined commands
28214 * Command Files:: How to write scripts of commands to be stored in a file
28215 * Output:: Commands for controlled output
28216 * Auto-loading sequences:: Controlling auto-loaded command files
28220 @subsection User-defined Commands
28222 @cindex user-defined command
28223 @cindex arguments, to user-defined commands
28224 A @dfn{user-defined command} is a sequence of @value{GDBN} commands to
28225 which you assign a new name as a command. This is done with the
28226 @code{define} command. User commands may accept an unlimited number of arguments
28227 separated by whitespace. Arguments are accessed within the user command
28228 via @code{$arg0@dots{}$argN}. A trivial example:
28232 print $arg0 + $arg1 + $arg2
28237 To execute the command use:
28244 This defines the command @code{adder}, which prints the sum of
28245 its three arguments. Note the arguments are text substitutions, so they may
28246 reference variables, use complex expressions, or even perform inferior
28249 @cindex argument count in user-defined commands
28250 @cindex how many arguments (user-defined commands)
28251 In addition, @code{$argc} may be used to find out how many arguments have
28257 print $arg0 + $arg1
28260 print $arg0 + $arg1 + $arg2
28265 Combining with the @code{eval} command (@pxref{eval}) makes it easier
28266 to process a variable number of arguments:
28273 eval "set $sum = $sum + $arg%d", $i
28283 @item define @var{commandname}
28284 Define a command named @var{commandname}. If there is already a command
28285 by that name, you are asked to confirm that you want to redefine it.
28286 The argument @var{commandname} may be a bare command name consisting of letters,
28287 numbers, dashes, dots, and underscores. It may also start with any
28288 predefined or user-defined prefix command.
28289 For example, @samp{define target my-target} creates
28290 a user-defined @samp{target my-target} command.
28292 The definition of the command is made up of other @value{GDBN} command lines,
28293 which are given following the @code{define} command. The end of these
28294 commands is marked by a line containing @code{end}.
28297 @kindex end@r{ (user-defined commands)}
28298 @item document @var{commandname}
28299 Document the user-defined command @var{commandname}, so that it can be
28300 accessed by @code{help}. The command @var{commandname} must already be
28301 defined. This command reads lines of documentation just as @code{define}
28302 reads the lines of the command definition, ending with @code{end}.
28303 After the @code{document} command is finished, @code{help} on command
28304 @var{commandname} displays the documentation you have written.
28306 You may use the @code{document} command again to change the
28307 documentation of a command. Redefining the command with @code{define}
28308 does not change the documentation.
28310 @kindex define-prefix
28311 @item define-prefix @var{commandname}
28312 Define or mark the command @var{commandname} as a user-defined prefix
28313 command. Once marked, @var{commandname} can be used as prefix command
28314 by the @code{define} command.
28315 Note that @code{define-prefix} can be used with a not yet defined
28316 @var{commandname}. In such a case, @var{commandname} is defined as
28317 an empty user-defined command.
28318 In case you redefine a command that was marked as a user-defined
28319 prefix command, the subcommands of the redefined command are kept
28320 (and @value{GDBN} indicates so to the user).
28324 (@value{GDBP}) define-prefix abc
28325 (@value{GDBP}) define-prefix abc def
28326 (@value{GDBP}) define abc def
28327 Type commands for definition of "abc def".
28328 End with a line saying just "end".
28329 >echo command initial def\n
28331 (@value{GDBP}) define abc def ghi
28332 Type commands for definition of "abc def ghi".
28333 End with a line saying just "end".
28334 >echo command ghi\n
28336 (@value{GDBP}) define abc def
28337 Keeping subcommands of prefix command "def".
28338 Redefine command "def"? (y or n) y
28339 Type commands for definition of "abc def".
28340 End with a line saying just "end".
28341 >echo command def\n
28343 (@value{GDBP}) abc def ghi
28345 (@value{GDBP}) abc def
28350 @kindex dont-repeat
28351 @cindex don't repeat command
28353 Used inside a user-defined command, this tells @value{GDBN} that this
28354 command should not be repeated when the user hits @key{RET}
28355 (@pxref{Command Syntax, repeat last command}).
28357 @kindex help user-defined
28358 @item help user-defined
28359 List all user-defined commands and all python commands defined in class
28360 COMMAND_USER. The first line of the documentation or docstring is
28365 @itemx show user @var{commandname}
28366 Display the @value{GDBN} commands used to define @var{commandname} (but
28367 not its documentation). If no @var{commandname} is given, display the
28368 definitions for all user-defined commands.
28369 This does not work for user-defined python commands.
28371 @cindex infinite recursion in user-defined commands
28372 @kindex show max-user-call-depth
28373 @kindex set max-user-call-depth
28374 @item show max-user-call-depth
28375 @itemx set max-user-call-depth
28376 The value of @code{max-user-call-depth} controls how many recursion
28377 levels are allowed in user-defined commands before @value{GDBN} suspects an
28378 infinite recursion and aborts the command.
28379 This does not apply to user-defined python commands.
28382 In addition to the above commands, user-defined commands frequently
28383 use control flow commands, described in @ref{Command Files}.
28385 When user-defined commands are executed, the
28386 commands of the definition are not printed. An error in any command
28387 stops execution of the user-defined command.
28389 If used interactively, commands that would ask for confirmation proceed
28390 without asking when used inside a user-defined command. Many @value{GDBN}
28391 commands that normally print messages to say what they are doing omit the
28392 messages when used in a user-defined command.
28395 @subsection User-defined Command Hooks
28396 @cindex command hooks
28397 @cindex hooks, for commands
28398 @cindex hooks, pre-command
28401 You may define @dfn{hooks}, which are a special kind of user-defined
28402 command. Whenever you run the command @samp{foo}, if the user-defined
28403 command @samp{hook-foo} exists, it is executed (with no arguments)
28404 before that command.
28406 @cindex hooks, post-command
28408 A hook may also be defined which is run after the command you executed.
28409 Whenever you run the command @samp{foo}, if the user-defined command
28410 @samp{hookpost-foo} exists, it is executed (with no arguments) after
28411 that command. Post-execution hooks may exist simultaneously with
28412 pre-execution hooks, for the same command.
28414 It is valid for a hook to call the command which it hooks. If this
28415 occurs, the hook is not re-executed, thereby avoiding infinite recursion.
28417 @c It would be nice if hookpost could be passed a parameter indicating
28418 @c if the command it hooks executed properly or not. FIXME!
28420 @kindex stop@r{, a pseudo-command}
28421 In addition, a pseudo-command, @samp{stop} exists. Defining
28422 (@samp{hook-stop}) makes the associated commands execute every time
28423 execution stops in your program: before breakpoint commands are run,
28424 displays are printed, or the stack frame is printed.
28426 For example, to ignore @code{SIGALRM} signals while
28427 single-stepping, but treat them normally during normal execution,
28432 handle SIGALRM nopass
28436 handle SIGALRM pass
28439 define hook-continue
28440 handle SIGALRM pass
28444 As a further example, to hook at the beginning and end of the @code{echo}
28445 command, and to add extra text to the beginning and end of the message,
28453 define hookpost-echo
28457 (@value{GDBP}) echo Hello World
28458 <<<---Hello World--->>>
28463 You can define a hook for any single-word command in @value{GDBN}, but
28464 not for command aliases; you should define a hook for the basic command
28465 name, e.g.@: @code{backtrace} rather than @code{bt}.
28466 @c FIXME! So how does Joe User discover whether a command is an alias
28468 You can hook a multi-word command by adding @code{hook-} or
28469 @code{hookpost-} to the last word of the command, e.g.@:
28470 @samp{define target hook-remote} to add a hook to @samp{target remote}.
28472 If an error occurs during the execution of your hook, execution of
28473 @value{GDBN} commands stops and @value{GDBN} issues a prompt
28474 (before the command that you actually typed had a chance to run).
28476 If you try to define a hook which does not match any known command, you
28477 get a warning from the @code{define} command.
28479 @node Command Files
28480 @subsection Command Files
28482 @cindex command files
28483 @cindex scripting commands
28484 A command file for @value{GDBN} is a text file made of lines that are
28485 @value{GDBN} commands. Comments (lines starting with @kbd{#}) may
28486 also be included. An empty line in a command file does nothing; it
28487 does not mean to repeat the last command, as it would from the
28490 You can request the execution of a command file with the @code{source}
28491 command. Note that the @code{source} command is also used to evaluate
28492 scripts that are not Command Files. The exact behavior can be configured
28493 using the @code{script-extension} setting.
28494 @xref{Extending GDB,, Extending GDB}.
28498 @cindex execute commands from a file
28499 @item source [-s] [-v] @var{filename}
28500 Execute the command file @var{filename}.
28503 The lines in a command file are generally executed sequentially,
28504 unless the order of execution is changed by one of the
28505 @emph{flow-control commands} described below. The commands are not
28506 printed as they are executed. An error in any command terminates
28507 execution of the command file and control is returned to the console.
28509 @value{GDBN} first searches for @var{filename} in the current directory.
28510 If the file is not found there, and @var{filename} does not specify a
28511 directory, then @value{GDBN} also looks for the file on the source search path
28512 (specified with the @samp{directory} command);
28513 except that @file{$cdir} is not searched because the compilation directory
28514 is not relevant to scripts.
28516 If @code{-s} is specified, then @value{GDBN} searches for @var{filename}
28517 on the search path even if @var{filename} specifies a directory.
28518 The search is done by appending @var{filename} to each element of the
28519 search path. So, for example, if @var{filename} is @file{mylib/myscript}
28520 and the search path contains @file{/home/user} then @value{GDBN} will
28521 look for the script @file{/home/user/mylib/myscript}.
28522 The search is also done if @var{filename} is an absolute path.
28523 For example, if @var{filename} is @file{/tmp/myscript} and
28524 the search path contains @file{/home/user} then @value{GDBN} will
28525 look for the script @file{/home/user/tmp/myscript}.
28526 For DOS-like systems, if @var{filename} contains a drive specification,
28527 it is stripped before concatenation. For example, if @var{filename} is
28528 @file{d:myscript} and the search path contains @file{c:/tmp} then @value{GDBN}
28529 will look for the script @file{c:/tmp/myscript}.
28531 If @code{-v}, for verbose mode, is given then @value{GDBN} displays
28532 each command as it is executed. The option must be given before
28533 @var{filename}, and is interpreted as part of the filename anywhere else.
28535 Commands that would ask for confirmation if used interactively proceed
28536 without asking when used in a command file. Many @value{GDBN} commands that
28537 normally print messages to say what they are doing omit the messages
28538 when called from command files.
28540 @value{GDBN} also accepts command input from standard input. In this
28541 mode, normal output goes to standard output and error output goes to
28542 standard error. Errors in a command file supplied on standard input do
28543 not terminate execution of the command file---execution continues with
28547 gdb < cmds > log 2>&1
28550 (The syntax above will vary depending on the shell used.) This example
28551 will execute commands from the file @file{cmds}. All output and errors
28552 would be directed to @file{log}.
28554 Since commands stored on command files tend to be more general than
28555 commands typed interactively, they frequently need to deal with
28556 complicated situations, such as different or unexpected values of
28557 variables and symbols, changes in how the program being debugged is
28558 built, etc. @value{GDBN} provides a set of flow-control commands to
28559 deal with these complexities. Using these commands, you can write
28560 complex scripts that loop over data structures, execute commands
28561 conditionally, etc.
28568 This command allows to include in your script conditionally executed
28569 commands. The @code{if} command takes a single argument, which is an
28570 expression to evaluate. It is followed by a series of commands that
28571 are executed only if the expression is true (its value is nonzero).
28572 There can then optionally be an @code{else} line, followed by a series
28573 of commands that are only executed if the expression was false. The
28574 end of the list is marked by a line containing @code{end}.
28578 This command allows to write loops. Its syntax is similar to
28579 @code{if}: the command takes a single argument, which is an expression
28580 to evaluate, and must be followed by the commands to execute, one per
28581 line, terminated by an @code{end}. These commands are called the
28582 @dfn{body} of the loop. The commands in the body of @code{while} are
28583 executed repeatedly as long as the expression evaluates to true.
28587 This command exits the @code{while} loop in whose body it is included.
28588 Execution of the script continues after that @code{while}s @code{end}
28591 @kindex loop_continue
28592 @item loop_continue
28593 This command skips the execution of the rest of the body of commands
28594 in the @code{while} loop in whose body it is included. Execution
28595 branches to the beginning of the @code{while} loop, where it evaluates
28596 the controlling expression.
28598 @kindex end@r{ (if/else/while commands)}
28600 Terminate the block of commands that are the body of @code{if},
28601 @code{else}, or @code{while} flow-control commands.
28606 @subsection Commands for Controlled Output
28608 During the execution of a command file or a user-defined command, normal
28609 @value{GDBN} output is suppressed; the only output that appears is what is
28610 explicitly printed by the commands in the definition. This section
28611 describes three commands useful for generating exactly the output you
28616 @item echo @var{text}
28617 @c I do not consider backslash-space a standard C escape sequence
28618 @c because it is not in ANSI.
28619 Print @var{text}. Nonprinting characters can be included in
28620 @var{text} using C escape sequences, such as @samp{\n} to print a
28621 newline. @strong{No newline is printed unless you specify one.}
28622 In addition to the standard C escape sequences, a backslash followed
28623 by a space stands for a space. This is useful for displaying a
28624 string with spaces at the beginning or the end, since leading and
28625 trailing spaces are otherwise trimmed from all arguments.
28626 To print @samp{@w{ }and foo =@w{ }}, use the command
28627 @samp{echo \@w{ }and foo = \@w{ }}.
28629 A backslash at the end of @var{text} can be used, as in C, to continue
28630 the command onto subsequent lines. For example,
28633 echo This is some text\n\
28634 which is continued\n\
28635 onto several lines.\n
28638 produces the same output as
28641 echo This is some text\n
28642 echo which is continued\n
28643 echo onto several lines.\n
28647 @item output @var{expression}
28648 Print the value of @var{expression} and nothing but that value: no
28649 newlines, no @samp{$@var{nn} = }. The value is not entered in the
28650 value history either. @xref{Expressions, ,Expressions}, for more information
28653 @item output/@var{fmt} @var{expression}
28654 Print the value of @var{expression} in format @var{fmt}. You can use
28655 the same formats as for @code{print}. @xref{Output Formats,,Output
28656 Formats}, for more information.
28659 @item printf @var{template}, @var{expressions}@dots{}
28660 Print the values of one or more @var{expressions} under the control of
28661 the string @var{template}. To print several values, make
28662 @var{expressions} be a comma-separated list of individual expressions,
28663 which may be either numbers or pointers. Their values are printed as
28664 specified by @var{template}, exactly as a C program would do by
28665 executing the code below:
28668 printf (@var{template}, @var{expressions}@dots{});
28671 As in @code{C} @code{printf}, ordinary characters in @var{template}
28672 are printed verbatim, while @dfn{conversion specification} introduced
28673 by the @samp{%} character cause subsequent @var{expressions} to be
28674 evaluated, their values converted and formatted according to type and
28675 style information encoded in the conversion specifications, and then
28678 For example, you can print two values in hex like this:
28681 printf "foo, bar-foo = 0x%x, 0x%x\n", foo, bar-foo
28684 @code{printf} supports all the standard @code{C} conversion
28685 specifications, including the flags and modifiers between the @samp{%}
28686 character and the conversion letter, with the following exceptions:
28690 The argument-ordering modifiers, such as @samp{2$}, are not supported.
28693 The modifier @samp{*} is not supported for specifying precision or
28697 The @samp{'} flag (for separation of digits into groups according to
28698 @code{LC_NUMERIC'}) is not supported.
28701 The type modifiers @samp{hh}, @samp{j}, @samp{t}, and @samp{z} are not
28705 The conversion letter @samp{n} (as in @samp{%n}) is not supported.
28708 The conversion letters @samp{a} and @samp{A} are not supported.
28712 Note that the @samp{ll} type modifier is supported only if the
28713 underlying @code{C} implementation used to build @value{GDBN} supports
28714 the @code{long long int} type, and the @samp{L} type modifier is
28715 supported only if @code{long double} type is available.
28717 As in @code{C}, @code{printf} supports simple backslash-escape
28718 sequences, such as @code{\n}, @samp{\t}, @samp{\\}, @samp{\"},
28719 @samp{\a}, and @samp{\f}, that consist of backslash followed by a
28720 single character. Octal and hexadecimal escape sequences are not
28723 Additionally, @code{printf} supports conversion specifications for DFP
28724 (@dfn{Decimal Floating Point}) types using the following length modifiers
28725 together with a floating point specifier.
28730 @samp{H} for printing @code{Decimal32} types.
28733 @samp{D} for printing @code{Decimal64} types.
28736 @samp{DD} for printing @code{Decimal128} types.
28739 If the underlying @code{C} implementation used to build @value{GDBN} has
28740 support for the three length modifiers for DFP types, other modifiers
28741 such as width and precision will also be available for @value{GDBN} to use.
28743 In case there is no such @code{C} support, no additional modifiers will be
28744 available and the value will be printed in the standard way.
28746 Here's an example of printing DFP types using the above conversion letters:
28748 printf "D32: %Hf - D64: %Df - D128: %DDf\n",1.2345df,1.2E10dd,1.2E1dl
28753 @item eval @var{template}, @var{expressions}@dots{}
28754 Convert the values of one or more @var{expressions} under the control of
28755 the string @var{template} to a command line, and call it.
28759 @node Auto-loading sequences
28760 @subsection Controlling auto-loading native @value{GDBN} scripts
28761 @cindex native script auto-loading
28763 When a new object file is read (for example, due to the @code{file}
28764 command, or because the inferior has loaded a shared library),
28765 @value{GDBN} will look for the command file @file{@var{objfile}-gdb.gdb}.
28766 @xref{Auto-loading extensions}.
28768 Auto-loading can be enabled or disabled,
28769 and the list of auto-loaded scripts can be printed.
28772 @anchor{set auto-load gdb-scripts}
28773 @kindex set auto-load gdb-scripts
28774 @item set auto-load gdb-scripts [on|off]
28775 Enable or disable the auto-loading of canned sequences of commands scripts.
28777 @anchor{show auto-load gdb-scripts}
28778 @kindex show auto-load gdb-scripts
28779 @item show auto-load gdb-scripts
28780 Show whether auto-loading of canned sequences of commands scripts is enabled or
28783 @anchor{info auto-load gdb-scripts}
28784 @kindex info auto-load gdb-scripts
28785 @cindex print list of auto-loaded canned sequences of commands scripts
28786 @item info auto-load gdb-scripts [@var{regexp}]
28787 Print the list of all canned sequences of commands scripts that @value{GDBN}
28791 If @var{regexp} is supplied only canned sequences of commands scripts with
28792 matching names are printed.
28794 @c Python docs live in a separate file.
28795 @include python.texi
28797 @c Guile docs live in a separate file.
28798 @include guile.texi
28800 @node Auto-loading extensions
28801 @section Auto-loading extensions
28802 @cindex auto-loading extensions
28804 @value{GDBN} provides two mechanisms for automatically loading extensions
28805 when a new object file is read (for example, due to the @code{file}
28806 command, or because the inferior has loaded a shared library):
28807 @file{@var{objfile}-gdb.@var{ext}} and the @code{.debug_gdb_scripts}
28808 section of modern file formats like ELF.
28811 * objfile-gdb.ext file: objfile-gdbdotext file. The @file{@var{objfile}-gdb.@var{ext}} file
28812 * .debug_gdb_scripts section: dotdebug_gdb_scripts section. The @code{.debug_gdb_scripts} section
28813 * Which flavor to choose?::
28816 The auto-loading feature is useful for supplying application-specific
28817 debugging commands and features.
28819 Auto-loading can be enabled or disabled,
28820 and the list of auto-loaded scripts can be printed.
28821 See the @samp{auto-loading} section of each extension language
28822 for more information.
28823 For @value{GDBN} command files see @ref{Auto-loading sequences}.
28824 For Python files see @ref{Python Auto-loading}.
28826 Note that loading of this script file also requires accordingly configured
28827 @code{auto-load safe-path} (@pxref{Auto-loading safe path}).
28829 @node objfile-gdbdotext file
28830 @subsection The @file{@var{objfile}-gdb.@var{ext}} file
28831 @cindex @file{@var{objfile}-gdb.gdb}
28832 @cindex @file{@var{objfile}-gdb.py}
28833 @cindex @file{@var{objfile}-gdb.scm}
28835 When a new object file is read, @value{GDBN} looks for a file named
28836 @file{@var{objfile}-gdb.@var{ext}} (we call it @var{script-name} below),
28837 where @var{objfile} is the object file's name and
28838 where @var{ext} is the file extension for the extension language:
28841 @item @file{@var{objfile}-gdb.gdb}
28842 GDB's own command language
28843 @item @file{@var{objfile}-gdb.py}
28845 @item @file{@var{objfile}-gdb.scm}
28849 @var{script-name} is formed by ensuring that the file name of @var{objfile}
28850 is absolute, following all symlinks, and resolving @code{.} and @code{..}
28851 components, and appending the @file{-gdb.@var{ext}} suffix.
28852 If this file exists and is readable, @value{GDBN} will evaluate it as a
28853 script in the specified extension language.
28855 If this file does not exist, then @value{GDBN} will look for
28856 @var{script-name} file in all of the directories as specified below.
28858 Note that loading of these files requires an accordingly configured
28859 @code{auto-load safe-path} (@pxref{Auto-loading safe path}).
28861 For object files using @file{.exe} suffix @value{GDBN} tries to load first the
28862 scripts normally according to its @file{.exe} filename. But if no scripts are
28863 found @value{GDBN} also tries script filenames matching the object file without
28864 its @file{.exe} suffix. This @file{.exe} stripping is case insensitive and it
28865 is attempted on any platform. This makes the script filenames compatible
28866 between Unix and MS-Windows hosts.
28869 @anchor{set auto-load scripts-directory}
28870 @kindex set auto-load scripts-directory
28871 @item set auto-load scripts-directory @r{[}@var{directories}@r{]}
28872 Control @value{GDBN} auto-loaded scripts location. Multiple directory entries
28873 may be delimited by the host platform path separator in use
28874 (@samp{:} on Unix, @samp{;} on MS-Windows and MS-DOS).
28876 Each entry here needs to be covered also by the security setting
28877 @code{set auto-load safe-path} (@pxref{set auto-load safe-path}).
28879 @anchor{with-auto-load-dir}
28880 This variable defaults to @file{$debugdir:$datadir/auto-load}. The default
28881 @code{set auto-load safe-path} value can be also overriden by @value{GDBN}
28882 configuration option @option{--with-auto-load-dir}.
28884 Any reference to @file{$debugdir} will get replaced by
28885 @var{debug-file-directory} value (@pxref{Separate Debug Files}) and any
28886 reference to @file{$datadir} will get replaced by @var{data-directory} which is
28887 determined at @value{GDBN} startup (@pxref{Data Files}). @file{$debugdir} and
28888 @file{$datadir} must be placed as a directory component --- either alone or
28889 delimited by @file{/} or @file{\} directory separators, depending on the host
28892 The list of directories uses path separator (@samp{:} on GNU and Unix
28893 systems, @samp{;} on MS-Windows and MS-DOS) to separate directories, similarly
28894 to the @env{PATH} environment variable.
28896 @anchor{show auto-load scripts-directory}
28897 @kindex show auto-load scripts-directory
28898 @item show auto-load scripts-directory
28899 Show @value{GDBN} auto-loaded scripts location.
28901 @anchor{add-auto-load-scripts-directory}
28902 @kindex add-auto-load-scripts-directory
28903 @item add-auto-load-scripts-directory @r{[}@var{directories}@dots{}@r{]}
28904 Add an entry (or list of entries) to the list of auto-loaded scripts locations.
28905 Multiple entries may be delimited by the host platform path separator in use.
28908 @value{GDBN} does not track which files it has already auto-loaded this way.
28909 @value{GDBN} will load the associated script every time the corresponding
28910 @var{objfile} is opened.
28911 So your @file{-gdb.@var{ext}} file should be careful to avoid errors if it
28912 is evaluated more than once.
28914 @node dotdebug_gdb_scripts section
28915 @subsection The @code{.debug_gdb_scripts} section
28916 @cindex @code{.debug_gdb_scripts} section
28918 For systems using file formats like ELF and COFF,
28919 when @value{GDBN} loads a new object file
28920 it will look for a special section named @code{.debug_gdb_scripts}.
28921 If this section exists, its contents is a list of null-terminated entries
28922 specifying scripts to load. Each entry begins with a non-null prefix byte that
28923 specifies the kind of entry, typically the extension language and whether the
28924 script is in a file or inlined in @code{.debug_gdb_scripts}.
28926 The following entries are supported:
28929 @item SECTION_SCRIPT_ID_PYTHON_FILE = 1
28930 @item SECTION_SCRIPT_ID_SCHEME_FILE = 3
28931 @item SECTION_SCRIPT_ID_PYTHON_TEXT = 4
28932 @item SECTION_SCRIPT_ID_SCHEME_TEXT = 6
28935 @subsubsection Script File Entries
28937 If the entry specifies a file, @value{GDBN} will look for the file first
28938 in the current directory and then along the source search path
28939 (@pxref{Source Path, ,Specifying Source Directories}),
28940 except that @file{$cdir} is not searched, since the compilation
28941 directory is not relevant to scripts.
28943 File entries can be placed in section @code{.debug_gdb_scripts} with,
28944 for example, this GCC macro for Python scripts.
28947 /* Note: The "MS" section flags are to remove duplicates. */
28948 #define DEFINE_GDB_PY_SCRIPT(script_name) \
28950 .pushsection \".debug_gdb_scripts\", \"MS\",@@progbits,1\n\
28951 .byte 1 /* Python */\n\
28952 .asciz \"" script_name "\"\n\
28958 For Guile scripts, replace @code{.byte 1} with @code{.byte 3}.
28959 Then one can reference the macro in a header or source file like this:
28962 DEFINE_GDB_PY_SCRIPT ("my-app-scripts.py")
28965 The script name may include directories if desired.
28967 Note that loading of this script file also requires accordingly configured
28968 @code{auto-load safe-path} (@pxref{Auto-loading safe path}).
28970 If the macro invocation is put in a header, any application or library
28971 using this header will get a reference to the specified script,
28972 and with the use of @code{"MS"} attributes on the section, the linker
28973 will remove duplicates.
28975 @subsubsection Script Text Entries
28977 Script text entries allow to put the executable script in the entry
28978 itself instead of loading it from a file.
28979 The first line of the entry, everything after the prefix byte and up to
28980 the first newline (@code{0xa}) character, is the script name, and must not
28981 contain any kind of space character, e.g., spaces or tabs.
28982 The rest of the entry, up to the trailing null byte, is the script to
28983 execute in the specified language. The name needs to be unique among
28984 all script names, as @value{GDBN} executes each script only once based
28987 Here is an example from file @file{py-section-script.c} in the @value{GDBN}
28991 #include "symcat.h"
28992 #include "gdb/section-scripts.h"
28994 ".pushsection \".debug_gdb_scripts\", \"MS\",@@progbits,1\n"
28995 ".byte " XSTRING (SECTION_SCRIPT_ID_PYTHON_TEXT) "\n"
28996 ".ascii \"gdb.inlined-script\\n\"\n"
28997 ".ascii \"class test_cmd (gdb.Command):\\n\"\n"
28998 ".ascii \" def __init__ (self):\\n\"\n"
28999 ".ascii \" super (test_cmd, self).__init__ ("
29000 "\\\"test-cmd\\\", gdb.COMMAND_OBSCURE)\\n\"\n"
29001 ".ascii \" def invoke (self, arg, from_tty):\\n\"\n"
29002 ".ascii \" print (\\\"test-cmd output, arg = %s\\\" % arg)\\n\"\n"
29003 ".ascii \"test_cmd ()\\n\"\n"
29009 Loading of inlined scripts requires a properly configured
29010 @code{auto-load safe-path} (@pxref{Auto-loading safe path}).
29011 The path to specify in @code{auto-load safe-path} is the path of the file
29012 containing the @code{.debug_gdb_scripts} section.
29014 @node Which flavor to choose?
29015 @subsection Which flavor to choose?
29017 Given the multiple ways of auto-loading extensions, it might not always
29018 be clear which one to choose. This section provides some guidance.
29021 Benefits of the @file{-gdb.@var{ext}} way:
29025 Can be used with file formats that don't support multiple sections.
29028 Ease of finding scripts for public libraries.
29030 Scripts specified in the @code{.debug_gdb_scripts} section are searched for
29031 in the source search path.
29032 For publicly installed libraries, e.g., @file{libstdc++}, there typically
29033 isn't a source directory in which to find the script.
29036 Doesn't require source code additions.
29040 Benefits of the @code{.debug_gdb_scripts} way:
29044 Works with static linking.
29046 Scripts for libraries done the @file{-gdb.@var{ext}} way require an objfile to
29047 trigger their loading. When an application is statically linked the only
29048 objfile available is the executable, and it is cumbersome to attach all the
29049 scripts from all the input libraries to the executable's
29050 @file{-gdb.@var{ext}} script.
29053 Works with classes that are entirely inlined.
29055 Some classes can be entirely inlined, and thus there may not be an associated
29056 shared library to attach a @file{-gdb.@var{ext}} script to.
29059 Scripts needn't be copied out of the source tree.
29061 In some circumstances, apps can be built out of large collections of internal
29062 libraries, and the build infrastructure necessary to install the
29063 @file{-gdb.@var{ext}} scripts in a place where @value{GDBN} can find them is
29064 cumbersome. It may be easier to specify the scripts in the
29065 @code{.debug_gdb_scripts} section as relative paths, and add a path to the
29066 top of the source tree to the source search path.
29069 @node Multiple Extension Languages
29070 @section Multiple Extension Languages
29072 The Guile and Python extension languages do not share any state,
29073 and generally do not interfere with each other.
29074 There are some things to be aware of, however.
29076 @subsection Python comes first
29078 Python was @value{GDBN}'s first extension language, and to avoid breaking
29079 existing behaviour Python comes first. This is generally solved by the
29080 ``first one wins'' principle. @value{GDBN} maintains a list of enabled
29081 extension languages, and when it makes a call to an extension language,
29082 (say to pretty-print a value), it tries each in turn until an extension
29083 language indicates it has performed the request (e.g., has returned the
29084 pretty-printed form of a value).
29085 This extends to errors while performing such requests: If an error happens
29086 while, for example, trying to pretty-print an object then the error is
29087 reported and any following extension languages are not tried.
29090 @section Creating new spellings of existing commands
29091 @cindex aliases for commands
29093 It is often useful to define alternate spellings of existing commands.
29094 For example, if a new @value{GDBN} command defined in Python has
29095 a long name to type, it is handy to have an abbreviated version of it
29096 that involves less typing.
29098 @value{GDBN} itself uses aliases. For example @samp{s} is an alias
29099 of the @samp{step} command even though it is otherwise an ambiguous
29100 abbreviation of other commands like @samp{set} and @samp{show}.
29102 Aliases are also used to provide shortened or more common versions
29103 of multi-word commands. For example, @value{GDBN} provides the
29104 @samp{tty} alias of the @samp{set inferior-tty} command.
29106 You can define a new alias with the @samp{alias} command.
29111 @item alias [-a] [--] @var{ALIAS} = @var{COMMAND}
29115 @var{ALIAS} specifies the name of the new alias.
29116 Each word of @var{ALIAS} must consist of letters, numbers, dashes and
29119 @var{COMMAND} specifies the name of an existing command
29120 that is being aliased.
29122 The @samp{-a} option specifies that the new alias is an abbreviation
29123 of the command. Abbreviations are not shown in command
29124 lists displayed by the @samp{help} command.
29126 The @samp{--} option specifies the end of options,
29127 and is useful when @var{ALIAS} begins with a dash.
29129 Here is a simple example showing how to make an abbreviation
29130 of a command so that there is less to type.
29131 Suppose you were tired of typing @samp{disas}, the current
29132 shortest unambiguous abbreviation of the @samp{disassemble} command
29133 and you wanted an even shorter version named @samp{di}.
29134 The following will accomplish this.
29137 (@value{GDBP}) alias -a di = disas
29140 Note that aliases are different from user-defined commands.
29141 With a user-defined command, you also need to write documentation
29142 for it with the @samp{document} command.
29143 An alias automatically picks up the documentation of the existing command.
29145 Here is an example where we make @samp{elms} an abbreviation of
29146 @samp{elements} in the @samp{set print elements} command.
29147 This is to show that you can make an abbreviation of any part
29151 (@value{GDBP}) alias -a set print elms = set print elements
29152 (@value{GDBP}) alias -a show print elms = show print elements
29153 (@value{GDBP}) set p elms 20
29154 (@value{GDBP}) show p elms
29155 Limit on string chars or array elements to print is 200.
29158 Note that if you are defining an alias of a @samp{set} command,
29159 and you want to have an alias for the corresponding @samp{show}
29160 command, then you need to define the latter separately.
29162 Unambiguously abbreviated commands are allowed in @var{COMMAND} and
29163 @var{ALIAS}, just as they are normally.
29166 (@value{GDBP}) alias -a set pr elms = set p ele
29169 Finally, here is an example showing the creation of a one word
29170 alias for a more complex command.
29171 This creates alias @samp{spe} of the command @samp{set print elements}.
29174 (@value{GDBP}) alias spe = set print elements
29175 (@value{GDBP}) spe 20
29179 @chapter Command Interpreters
29180 @cindex command interpreters
29182 @value{GDBN} supports multiple command interpreters, and some command
29183 infrastructure to allow users or user interface writers to switch
29184 between interpreters or run commands in other interpreters.
29186 @value{GDBN} currently supports two command interpreters, the console
29187 interpreter (sometimes called the command-line interpreter or @sc{cli})
29188 and the machine interface interpreter (or @sc{gdb/mi}). This manual
29189 describes both of these interfaces in great detail.
29191 By default, @value{GDBN} will start with the console interpreter.
29192 However, the user may choose to start @value{GDBN} with another
29193 interpreter by specifying the @option{-i} or @option{--interpreter}
29194 startup options. Defined interpreters include:
29198 @cindex console interpreter
29199 The traditional console or command-line interpreter. This is the most often
29200 used interpreter with @value{GDBN}. With no interpreter specified at runtime,
29201 @value{GDBN} will use this interpreter.
29204 @cindex mi interpreter
29205 The newest @sc{gdb/mi} interface (currently @code{mi3}). Used primarily
29206 by programs wishing to use @value{GDBN} as a backend for a debugger GUI
29207 or an IDE. For more information, see @ref{GDB/MI, ,The @sc{gdb/mi}
29211 @cindex mi3 interpreter
29212 The @sc{gdb/mi} interface introduced in @value{GDBN} 9.1.
29215 @cindex mi2 interpreter
29216 The @sc{gdb/mi} interface introduced in @value{GDBN} 6.0.
29219 @cindex mi1 interpreter
29220 The @sc{gdb/mi} interface introduced in @value{GDBN} 5.1.
29224 @cindex invoke another interpreter
29226 @kindex interpreter-exec
29227 You may execute commands in any interpreter from the current
29228 interpreter using the appropriate command. If you are running the
29229 console interpreter, simply use the @code{interpreter-exec} command:
29232 interpreter-exec mi "-data-list-register-names"
29235 @sc{gdb/mi} has a similar command, although it is only available in versions of
29236 @value{GDBN} which support @sc{gdb/mi} version 2 (or greater).
29238 Note that @code{interpreter-exec} only changes the interpreter for the
29239 duration of the specified command. It does not change the interpreter
29242 @cindex start a new independent interpreter
29244 Although you may only choose a single interpreter at startup, it is
29245 possible to run an independent interpreter on a specified input/output
29246 device (usually a tty).
29248 For example, consider a debugger GUI or IDE that wants to provide a
29249 @value{GDBN} console view. It may do so by embedding a terminal
29250 emulator widget in its GUI, starting @value{GDBN} in the traditional
29251 command-line mode with stdin/stdout/stderr redirected to that
29252 terminal, and then creating an MI interpreter running on a specified
29253 input/output device. The console interpreter created by @value{GDBN}
29254 at startup handles commands the user types in the terminal widget,
29255 while the GUI controls and synchronizes state with @value{GDBN} using
29256 the separate MI interpreter.
29258 To start a new secondary @dfn{user interface} running MI, use the
29259 @code{new-ui} command:
29262 @cindex new user interface
29264 new-ui @var{interpreter} @var{tty}
29267 The @var{interpreter} parameter specifies the interpreter to run.
29268 This accepts the same values as the @code{interpreter-exec} command.
29269 For example, @samp{console}, @samp{mi}, @samp{mi2}, etc. The
29270 @var{tty} parameter specifies the name of the bidirectional file the
29271 interpreter uses for input/output, usually the name of a
29272 pseudoterminal slave on Unix systems. For example:
29275 (@value{GDBP}) new-ui mi /dev/pts/9
29279 runs an MI interpreter on @file{/dev/pts/9}.
29282 @chapter @value{GDBN} Text User Interface
29284 @cindex Text User Interface
29287 * TUI Overview:: TUI overview
29288 * TUI Keys:: TUI key bindings
29289 * TUI Single Key Mode:: TUI single key mode
29290 * TUI Commands:: TUI-specific commands
29291 * TUI Configuration:: TUI configuration variables
29294 The @value{GDBN} Text User Interface (TUI) is a terminal
29295 interface which uses the @code{curses} library to show the source
29296 file, the assembly output, the program registers and @value{GDBN}
29297 commands in separate text windows. The TUI mode is supported only
29298 on platforms where a suitable version of the @code{curses} library
29301 The TUI mode is enabled by default when you invoke @value{GDBN} as
29302 @samp{@value{GDBP} -tui}.
29303 You can also switch in and out of TUI mode while @value{GDBN} runs by
29304 using various TUI commands and key bindings, such as @command{tui
29305 enable} or @kbd{C-x C-a}. @xref{TUI Commands, ,TUI Commands}, and
29306 @ref{TUI Keys, ,TUI Key Bindings}.
29309 @section TUI Overview
29311 In TUI mode, @value{GDBN} can display several text windows:
29315 This window is the @value{GDBN} command window with the @value{GDBN}
29316 prompt and the @value{GDBN} output. The @value{GDBN} input is still
29317 managed using readline.
29320 The source window shows the source file of the program. The current
29321 line and active breakpoints are displayed in this window.
29324 The assembly window shows the disassembly output of the program.
29327 This window shows the processor registers. Registers are highlighted
29328 when their values change.
29331 The source and assembly windows show the current program position
29332 by highlighting the current line and marking it with a @samp{>} marker.
29333 Breakpoints are indicated with two markers. The first marker
29334 indicates the breakpoint type:
29338 Breakpoint which was hit at least once.
29341 Breakpoint which was never hit.
29344 Hardware breakpoint which was hit at least once.
29347 Hardware breakpoint which was never hit.
29350 The second marker indicates whether the breakpoint is enabled or not:
29354 Breakpoint is enabled.
29357 Breakpoint is disabled.
29360 The source, assembly and register windows are updated when the current
29361 thread changes, when the frame changes, or when the program counter
29364 These windows are not all visible at the same time. The command
29365 window is always visible. The others can be arranged in several
29376 source and assembly,
29379 source and registers, or
29382 assembly and registers.
29385 A status line above the command window shows the following information:
29389 Indicates the current @value{GDBN} target.
29390 (@pxref{Targets, ,Specifying a Debugging Target}).
29393 Gives the current process or thread number.
29394 When no process is being debugged, this field is set to @code{No process}.
29397 Gives the current function name for the selected frame.
29398 The name is demangled if demangling is turned on (@pxref{Print Settings}).
29399 When there is no symbol corresponding to the current program counter,
29400 the string @code{??} is displayed.
29403 Indicates the current line number for the selected frame.
29404 When the current line number is not known, the string @code{??} is displayed.
29407 Indicates the current program counter address.
29411 @section TUI Key Bindings
29412 @cindex TUI key bindings
29414 The TUI installs several key bindings in the readline keymaps
29415 @ifset SYSTEM_READLINE
29416 (@pxref{Command Line Editing, , , rluserman, GNU Readline Library}).
29418 @ifclear SYSTEM_READLINE
29419 (@pxref{Command Line Editing}).
29421 The following key bindings are installed for both TUI mode and the
29422 @value{GDBN} standard mode.
29431 Enter or leave the TUI mode. When leaving the TUI mode,
29432 the curses window management stops and @value{GDBN} operates using
29433 its standard mode, writing on the terminal directly. When reentering
29434 the TUI mode, control is given back to the curses windows.
29435 The screen is then refreshed.
29437 This key binding uses the bindable Readline function
29438 @code{tui-switch-mode}.
29442 Use a TUI layout with only one window. The layout will
29443 either be @samp{source} or @samp{assembly}. When the TUI mode
29444 is not active, it will switch to the TUI mode.
29446 Think of this key binding as the Emacs @kbd{C-x 1} binding.
29448 This key binding uses the bindable Readline function
29449 @code{tui-delete-other-windows}.
29453 Use a TUI layout with at least two windows. When the current
29454 layout already has two windows, the next layout with two windows is used.
29455 When a new layout is chosen, one window will always be common to the
29456 previous layout and the new one.
29458 Think of it as the Emacs @kbd{C-x 2} binding.
29460 This key binding uses the bindable Readline function
29461 @code{tui-change-windows}.
29465 Change the active window. The TUI associates several key bindings
29466 (like scrolling and arrow keys) with the active window. This command
29467 gives the focus to the next TUI window.
29469 Think of it as the Emacs @kbd{C-x o} binding.
29471 This key binding uses the bindable Readline function
29472 @code{tui-other-window}.
29476 Switch in and out of the TUI SingleKey mode that binds single
29477 keys to @value{GDBN} commands (@pxref{TUI Single Key Mode}).
29479 This key binding uses the bindable Readline function
29480 @code{next-keymap}.
29483 The following key bindings only work in the TUI mode:
29488 Scroll the active window one page up.
29492 Scroll the active window one page down.
29496 Scroll the active window one line up.
29500 Scroll the active window one line down.
29504 Scroll the active window one column left.
29508 Scroll the active window one column right.
29512 Refresh the screen.
29515 Because the arrow keys scroll the active window in the TUI mode, they
29516 are not available for their normal use by readline unless the command
29517 window has the focus. When another window is active, you must use
29518 other readline key bindings such as @kbd{C-p}, @kbd{C-n}, @kbd{C-b}
29519 and @kbd{C-f} to control the command window.
29521 @node TUI Single Key Mode
29522 @section TUI Single Key Mode
29523 @cindex TUI single key mode
29525 The TUI also provides a @dfn{SingleKey} mode, which binds several
29526 frequently used @value{GDBN} commands to single keys. Type @kbd{C-x s} to
29527 switch into this mode, where the following key bindings are used:
29530 @kindex c @r{(SingleKey TUI key)}
29534 @kindex d @r{(SingleKey TUI key)}
29538 @kindex f @r{(SingleKey TUI key)}
29542 @kindex n @r{(SingleKey TUI key)}
29546 @kindex o @r{(SingleKey TUI key)}
29548 nexti. The shortcut letter @samp{o} stands for ``step Over''.
29550 @kindex q @r{(SingleKey TUI key)}
29552 exit the SingleKey mode.
29554 @kindex r @r{(SingleKey TUI key)}
29558 @kindex s @r{(SingleKey TUI key)}
29562 @kindex i @r{(SingleKey TUI key)}
29564 stepi. The shortcut letter @samp{i} stands for ``step Into''.
29566 @kindex u @r{(SingleKey TUI key)}
29570 @kindex v @r{(SingleKey TUI key)}
29574 @kindex w @r{(SingleKey TUI key)}
29579 Other keys temporarily switch to the @value{GDBN} command prompt.
29580 The key that was pressed is inserted in the editing buffer so that
29581 it is possible to type most @value{GDBN} commands without interaction
29582 with the TUI SingleKey mode. Once the command is entered the TUI
29583 SingleKey mode is restored. The only way to permanently leave
29584 this mode is by typing @kbd{q} or @kbd{C-x s}.
29586 @cindex SingleKey keymap name
29587 If @value{GDBN} was built with Readline 8.0 or later, the TUI
29588 SingleKey keymap will be named @samp{SingleKey}. This can be used in
29589 @file{.inputrc} to add additional bindings to this keymap.
29592 @section TUI-specific Commands
29593 @cindex TUI commands
29595 The TUI has specific commands to control the text windows.
29596 These commands are always available, even when @value{GDBN} is not in
29597 the TUI mode. When @value{GDBN} is in the standard mode, most
29598 of these commands will automatically switch to the TUI mode.
29600 Note that if @value{GDBN}'s @code{stdout} is not connected to a
29601 terminal, or @value{GDBN} has been started with the machine interface
29602 interpreter (@pxref{GDB/MI, ,The @sc{gdb/mi} Interface}), most of
29603 these commands will fail with an error, because it would not be
29604 possible or desirable to enable curses window management.
29609 Activate TUI mode. The last active TUI window layout will be used if
29610 TUI mode has previously been used in the current debugging session,
29611 otherwise a default layout is used.
29614 @kindex tui disable
29615 Disable TUI mode, returning to the console interpreter.
29619 List and give the size of all displayed windows.
29621 @item layout @var{name}
29623 Changes which TUI windows are displayed. In each layout the command
29624 window is always displayed, the @var{name} parameter controls which
29625 additional windows are displayed, and can be any of the following:
29629 Display the next layout.
29632 Display the previous layout.
29635 Display the source and command windows.
29638 Display the assembly and command windows.
29641 Display the source, assembly, and command windows.
29644 When in @code{src} layout display the register, source, and command
29645 windows. When in @code{asm} or @code{split} layout display the
29646 register, assembler, and command windows.
29649 @item focus @var{name}
29651 Changes which TUI window is currently active for scrolling. The
29652 @var{name} parameter can be any of the following:
29656 Make the next window active for scrolling.
29659 Make the previous window active for scrolling.
29662 Make the source window active for scrolling.
29665 Make the assembly window active for scrolling.
29668 Make the register window active for scrolling.
29671 Make the command window active for scrolling.
29676 Refresh the screen. This is similar to typing @kbd{C-L}.
29678 @item tui reg @var{group}
29680 Changes the register group displayed in the tui register window to
29681 @var{group}. If the register window is not currently displayed this
29682 command will cause the register window to be displayed. The list of
29683 register groups, as well as their order is target specific. The
29684 following groups are available on most targets:
29687 Repeatedly selecting this group will cause the display to cycle
29688 through all of the available register groups.
29691 Repeatedly selecting this group will cause the display to cycle
29692 through all of the available register groups in the reverse order to
29696 Display the general registers.
29698 Display the floating point registers.
29700 Display the system registers.
29702 Display the vector registers.
29704 Display all registers.
29709 Update the source window and the current execution point.
29711 @item winheight @var{name} +@var{count}
29712 @itemx winheight @var{name} -@var{count}
29714 Change the height of the window @var{name} by @var{count}
29715 lines. Positive counts increase the height, while negative counts
29716 decrease it. The @var{name} parameter can be one of @code{src} (the
29717 source window), @code{cmd} (the command window), @code{asm} (the
29718 disassembly window), or @code{regs} (the register display window).
29721 @node TUI Configuration
29722 @section TUI Configuration Variables
29723 @cindex TUI configuration variables
29725 Several configuration variables control the appearance of TUI windows.
29728 @item set tui border-kind @var{kind}
29729 @kindex set tui border-kind
29730 Select the border appearance for the source, assembly and register windows.
29731 The possible values are the following:
29734 Use a space character to draw the border.
29737 Use @sc{ascii} characters @samp{+}, @samp{-} and @samp{|} to draw the border.
29740 Use the Alternate Character Set to draw the border. The border is
29741 drawn using character line graphics if the terminal supports them.
29744 @item set tui border-mode @var{mode}
29745 @kindex set tui border-mode
29746 @itemx set tui active-border-mode @var{mode}
29747 @kindex set tui active-border-mode
29748 Select the display attributes for the borders of the inactive windows
29749 or the active window. The @var{mode} can be one of the following:
29752 Use normal attributes to display the border.
29758 Use reverse video mode.
29761 Use half bright mode.
29763 @item half-standout
29764 Use half bright and standout mode.
29767 Use extra bright or bold mode.
29769 @item bold-standout
29770 Use extra bright or bold and standout mode.
29773 @item set tui tab-width @var{nchars}
29774 @kindex set tui tab-width
29776 Set the width of tab stops to be @var{nchars} characters. This
29777 setting affects the display of TAB characters in the source and
29780 @item set tui compact-source @r{[}on@r{|}off@r{]}
29781 @kindex set tui compact-source
29782 Set whether the TUI source window is displayed in ``compact'' form.
29783 The default display uses more space for line numbers and starts the
29784 source text at the next tab stop; the compact display uses only as
29785 much space as is needed for the line numbers in the current file, and
29786 only a single space to separate the line numbers from the source.
29789 Note that the colors of the TUI borders can be controlled using the
29790 appropriate @code{set style} commands. @xref{Output Styling}.
29793 @chapter Using @value{GDBN} under @sc{gnu} Emacs
29796 @cindex @sc{gnu} Emacs
29797 A special interface allows you to use @sc{gnu} Emacs to view (and
29798 edit) the source files for the program you are debugging with
29801 To use this interface, use the command @kbd{M-x gdb} in Emacs. Give the
29802 executable file you want to debug as an argument. This command starts
29803 @value{GDBN} as a subprocess of Emacs, with input and output through a newly
29804 created Emacs buffer.
29805 @c (Do not use the @code{-tui} option to run @value{GDBN} from Emacs.)
29807 Running @value{GDBN} under Emacs can be just like running @value{GDBN} normally except for two
29812 All ``terminal'' input and output goes through an Emacs buffer, called
29815 This applies both to @value{GDBN} commands and their output, and to the input
29816 and output done by the program you are debugging.
29818 This is useful because it means that you can copy the text of previous
29819 commands and input them again; you can even use parts of the output
29822 All the facilities of Emacs' Shell mode are available for interacting
29823 with your program. In particular, you can send signals the usual
29824 way---for example, @kbd{C-c C-c} for an interrupt, @kbd{C-c C-z} for a
29828 @value{GDBN} displays source code through Emacs.
29830 Each time @value{GDBN} displays a stack frame, Emacs automatically finds the
29831 source file for that frame and puts an arrow (@samp{=>}) at the
29832 left margin of the current line. Emacs uses a separate buffer for
29833 source display, and splits the screen to show both your @value{GDBN} session
29836 Explicit @value{GDBN} @code{list} or search commands still produce output as
29837 usual, but you probably have no reason to use them from Emacs.
29840 We call this @dfn{text command mode}. Emacs 22.1, and later, also uses
29841 a graphical mode, enabled by default, which provides further buffers
29842 that can control the execution and describe the state of your program.
29843 @xref{GDB Graphical Interface,,, Emacs, The @sc{gnu} Emacs Manual}.
29845 If you specify an absolute file name when prompted for the @kbd{M-x
29846 gdb} argument, then Emacs sets your current working directory to where
29847 your program resides. If you only specify the file name, then Emacs
29848 sets your current working directory to the directory associated
29849 with the previous buffer. In this case, @value{GDBN} may find your
29850 program by searching your environment's @code{PATH} variable, but on
29851 some operating systems it might not find the source. So, although the
29852 @value{GDBN} input and output session proceeds normally, the auxiliary
29853 buffer does not display the current source and line of execution.
29855 The initial working directory of @value{GDBN} is printed on the top
29856 line of the GUD buffer and this serves as a default for the commands
29857 that specify files for @value{GDBN} to operate on. @xref{Files,
29858 ,Commands to Specify Files}.
29860 By default, @kbd{M-x gdb} calls the program called @file{gdb}. If you
29861 need to call @value{GDBN} by a different name (for example, if you
29862 keep several configurations around, with different names) you can
29863 customize the Emacs variable @code{gud-gdb-command-name} to run the
29866 In the GUD buffer, you can use these special Emacs commands in
29867 addition to the standard Shell mode commands:
29871 Describe the features of Emacs' GUD Mode.
29874 Execute to another source line, like the @value{GDBN} @code{step} command; also
29875 update the display window to show the current file and location.
29878 Execute to next source line in this function, skipping all function
29879 calls, like the @value{GDBN} @code{next} command. Then update the display window
29880 to show the current file and location.
29883 Execute one instruction, like the @value{GDBN} @code{stepi} command; update
29884 display window accordingly.
29887 Execute until exit from the selected stack frame, like the @value{GDBN}
29888 @code{finish} command.
29891 Continue execution of your program, like the @value{GDBN} @code{continue}
29895 Go up the number of frames indicated by the numeric argument
29896 (@pxref{Arguments, , Numeric Arguments, Emacs, The @sc{gnu} Emacs Manual}),
29897 like the @value{GDBN} @code{up} command.
29900 Go down the number of frames indicated by the numeric argument, like the
29901 @value{GDBN} @code{down} command.
29904 In any source file, the Emacs command @kbd{C-x @key{SPC}} (@code{gud-break})
29905 tells @value{GDBN} to set a breakpoint on the source line point is on.
29907 In text command mode, if you type @kbd{M-x speedbar}, Emacs displays a
29908 separate frame which shows a backtrace when the GUD buffer is current.
29909 Move point to any frame in the stack and type @key{RET} to make it
29910 become the current frame and display the associated source in the
29911 source buffer. Alternatively, click @kbd{Mouse-2} to make the
29912 selected frame become the current one. In graphical mode, the
29913 speedbar displays watch expressions.
29915 If you accidentally delete the source-display buffer, an easy way to get
29916 it back is to type the command @code{f} in the @value{GDBN} buffer, to
29917 request a frame display; when you run under Emacs, this recreates
29918 the source buffer if necessary to show you the context of the current
29921 The source files displayed in Emacs are in ordinary Emacs buffers
29922 which are visiting the source files in the usual way. You can edit
29923 the files with these buffers if you wish; but keep in mind that @value{GDBN}
29924 communicates with Emacs in terms of line numbers. If you add or
29925 delete lines from the text, the line numbers that @value{GDBN} knows cease
29926 to correspond properly with the code.
29928 A more detailed description of Emacs' interaction with @value{GDBN} is
29929 given in the Emacs manual (@pxref{Debuggers,,, Emacs, The @sc{gnu}
29933 @chapter The @sc{gdb/mi} Interface
29935 @unnumberedsec Function and Purpose
29937 @cindex @sc{gdb/mi}, its purpose
29938 @sc{gdb/mi} is a line based machine oriented text interface to
29939 @value{GDBN} and is activated by specifying using the
29940 @option{--interpreter} command line option (@pxref{Mode Options}). It
29941 is specifically intended to support the development of systems which
29942 use the debugger as just one small component of a larger system.
29944 This chapter is a specification of the @sc{gdb/mi} interface. It is written
29945 in the form of a reference manual.
29947 Note that @sc{gdb/mi} is still under construction, so some of the
29948 features described below are incomplete and subject to change
29949 (@pxref{GDB/MI Development and Front Ends, , @sc{gdb/mi} Development and Front Ends}).
29951 @unnumberedsec Notation and Terminology
29953 @cindex notational conventions, for @sc{gdb/mi}
29954 This chapter uses the following notation:
29958 @code{|} separates two alternatives.
29961 @code{[ @var{something} ]} indicates that @var{something} is optional:
29962 it may or may not be given.
29965 @code{( @var{group} )*} means that @var{group} inside the parentheses
29966 may repeat zero or more times.
29969 @code{( @var{group} )+} means that @var{group} inside the parentheses
29970 may repeat one or more times.
29973 @code{"@var{string}"} means a literal @var{string}.
29977 @heading Dependencies
29981 * GDB/MI General Design::
29982 * GDB/MI Command Syntax::
29983 * GDB/MI Compatibility with CLI::
29984 * GDB/MI Development and Front Ends::
29985 * GDB/MI Output Records::
29986 * GDB/MI Simple Examples::
29987 * GDB/MI Command Description Format::
29988 * GDB/MI Breakpoint Commands::
29989 * GDB/MI Catchpoint Commands::
29990 * GDB/MI Program Context::
29991 * GDB/MI Thread Commands::
29992 * GDB/MI Ada Tasking Commands::
29993 * GDB/MI Program Execution::
29994 * GDB/MI Stack Manipulation::
29995 * GDB/MI Variable Objects::
29996 * GDB/MI Data Manipulation::
29997 * GDB/MI Tracepoint Commands::
29998 * GDB/MI Symbol Query::
29999 * GDB/MI File Commands::
30001 * GDB/MI Kod Commands::
30002 * GDB/MI Memory Overlay Commands::
30003 * GDB/MI Signal Handling Commands::
30005 * GDB/MI Target Manipulation::
30006 * GDB/MI File Transfer Commands::
30007 * GDB/MI Ada Exceptions Commands::
30008 * GDB/MI Support Commands::
30009 * GDB/MI Miscellaneous Commands::
30012 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30013 @node GDB/MI General Design
30014 @section @sc{gdb/mi} General Design
30015 @cindex GDB/MI General Design
30017 Interaction of a @sc{GDB/MI} frontend with @value{GDBN} involves three
30018 parts---commands sent to @value{GDBN}, responses to those commands
30019 and notifications. Each command results in exactly one response,
30020 indicating either successful completion of the command, or an error.
30021 For the commands that do not resume the target, the response contains the
30022 requested information. For the commands that resume the target, the
30023 response only indicates whether the target was successfully resumed.
30024 Notifications is the mechanism for reporting changes in the state of the
30025 target, or in @value{GDBN} state, that cannot conveniently be associated with
30026 a command and reported as part of that command response.
30028 The important examples of notifications are:
30032 Exec notifications. These are used to report changes in
30033 target state---when a target is resumed, or stopped. It would not
30034 be feasible to include this information in response of resuming
30035 commands, because one resume commands can result in multiple events in
30036 different threads. Also, quite some time may pass before any event
30037 happens in the target, while a frontend needs to know whether the resuming
30038 command itself was successfully executed.
30041 Console output, and status notifications. Console output
30042 notifications are used to report output of CLI commands, as well as
30043 diagnostics for other commands. Status notifications are used to
30044 report the progress of a long-running operation. Naturally, including
30045 this information in command response would mean no output is produced
30046 until the command is finished, which is undesirable.
30049 General notifications. Commands may have various side effects on
30050 the @value{GDBN} or target state beyond their official purpose. For example,
30051 a command may change the selected thread. Although such changes can
30052 be included in command response, using notification allows for more
30053 orthogonal frontend design.
30057 There's no guarantee that whenever an MI command reports an error,
30058 @value{GDBN} or the target are in any specific state, and especially,
30059 the state is not reverted to the state before the MI command was
30060 processed. Therefore, whenever an MI command results in an error,
30061 we recommend that the frontend refreshes all the information shown in
30062 the user interface.
30066 * Context management::
30067 * Asynchronous and non-stop modes::
30071 @node Context management
30072 @subsection Context management
30074 @subsubsection Threads and Frames
30076 In most cases when @value{GDBN} accesses the target, this access is
30077 done in context of a specific thread and frame (@pxref{Frames}).
30078 Often, even when accessing global data, the target requires that a thread
30079 be specified. The CLI interface maintains the selected thread and frame,
30080 and supplies them to target on each command. This is convenient,
30081 because a command line user would not want to specify that information
30082 explicitly on each command, and because user interacts with
30083 @value{GDBN} via a single terminal, so no confusion is possible as
30084 to what thread and frame are the current ones.
30086 In the case of MI, the concept of selected thread and frame is less
30087 useful. First, a frontend can easily remember this information
30088 itself. Second, a graphical frontend can have more than one window,
30089 each one used for debugging a different thread, and the frontend might
30090 want to access additional threads for internal purposes. This
30091 increases the risk that by relying on implicitly selected thread, the
30092 frontend may be operating on a wrong one. Therefore, each MI command
30093 should explicitly specify which thread and frame to operate on. To
30094 make it possible, each MI command accepts the @samp{--thread} and
30095 @samp{--frame} options, the value to each is @value{GDBN} global
30096 identifier for thread and frame to operate on.
30098 Usually, each top-level window in a frontend allows the user to select
30099 a thread and a frame, and remembers the user selection for further
30100 operations. However, in some cases @value{GDBN} may suggest that the
30101 current thread or frame be changed. For example, when stopping on a
30102 breakpoint it is reasonable to switch to the thread where breakpoint is
30103 hit. For another example, if the user issues the CLI @samp{thread} or
30104 @samp{frame} commands via the frontend, it is desirable to change the
30105 frontend's selection to the one specified by user. @value{GDBN}
30106 communicates the suggestion to change current thread and frame using the
30107 @samp{=thread-selected} notification.
30109 Note that historically, MI shares the selected thread with CLI, so
30110 frontends used the @code{-thread-select} to execute commands in the
30111 right context. However, getting this to work right is cumbersome. The
30112 simplest way is for frontend to emit @code{-thread-select} command
30113 before every command. This doubles the number of commands that need
30114 to be sent. The alternative approach is to suppress @code{-thread-select}
30115 if the selected thread in @value{GDBN} is supposed to be identical to the
30116 thread the frontend wants to operate on. However, getting this
30117 optimization right can be tricky. In particular, if the frontend
30118 sends several commands to @value{GDBN}, and one of the commands changes the
30119 selected thread, then the behaviour of subsequent commands will
30120 change. So, a frontend should either wait for response from such
30121 problematic commands, or explicitly add @code{-thread-select} for
30122 all subsequent commands. No frontend is known to do this exactly
30123 right, so it is suggested to just always pass the @samp{--thread} and
30124 @samp{--frame} options.
30126 @subsubsection Language
30128 The execution of several commands depends on which language is selected.
30129 By default, the current language (@pxref{show language}) is used.
30130 But for commands known to be language-sensitive, it is recommended
30131 to use the @samp{--language} option. This option takes one argument,
30132 which is the name of the language to use while executing the command.
30136 -data-evaluate-expression --language c "sizeof (void*)"
30141 The valid language names are the same names accepted by the
30142 @samp{set language} command (@pxref{Manually}), excluding @samp{auto},
30143 @samp{local} or @samp{unknown}.
30145 @node Asynchronous and non-stop modes
30146 @subsection Asynchronous command execution and non-stop mode
30148 On some targets, @value{GDBN} is capable of processing MI commands
30149 even while the target is running. This is called @dfn{asynchronous
30150 command execution} (@pxref{Background Execution}). The frontend may
30151 specify a preference for asynchronous execution using the
30152 @code{-gdb-set mi-async 1} command, which should be emitted before
30153 either running the executable or attaching to the target. After the
30154 frontend has started the executable or attached to the target, it can
30155 find if asynchronous execution is enabled using the
30156 @code{-list-target-features} command.
30159 @item -gdb-set mi-async on
30160 @item -gdb-set mi-async off
30161 Set whether MI is in asynchronous mode.
30163 When @code{off}, which is the default, MI execution commands (e.g.,
30164 @code{-exec-continue}) are foreground commands, and @value{GDBN} waits
30165 for the program to stop before processing further commands.
30167 When @code{on}, MI execution commands are background execution
30168 commands (e.g., @code{-exec-continue} becomes the equivalent of the
30169 @code{c&} CLI command), and so @value{GDBN} is capable of processing
30170 MI commands even while the target is running.
30172 @item -gdb-show mi-async
30173 Show whether MI asynchronous mode is enabled.
30176 Note: In @value{GDBN} version 7.7 and earlier, this option was called
30177 @code{target-async} instead of @code{mi-async}, and it had the effect
30178 of both putting MI in asynchronous mode and making CLI background
30179 commands possible. CLI background commands are now always possible
30180 ``out of the box'' if the target supports them. The old spelling is
30181 kept as a deprecated alias for backwards compatibility.
30183 Even if @value{GDBN} can accept a command while target is running,
30184 many commands that access the target do not work when the target is
30185 running. Therefore, asynchronous command execution is most useful
30186 when combined with non-stop mode (@pxref{Non-Stop Mode}). Then,
30187 it is possible to examine the state of one thread, while other threads
30190 When a given thread is running, MI commands that try to access the
30191 target in the context of that thread may not work, or may work only on
30192 some targets. In particular, commands that try to operate on thread's
30193 stack will not work, on any target. Commands that read memory, or
30194 modify breakpoints, may work or not work, depending on the target. Note
30195 that even commands that operate on global state, such as @code{print},
30196 @code{set}, and breakpoint commands, still access the target in the
30197 context of a specific thread, so frontend should try to find a
30198 stopped thread and perform the operation on that thread (using the
30199 @samp{--thread} option).
30201 Which commands will work in the context of a running thread is
30202 highly target dependent. However, the two commands
30203 @code{-exec-interrupt}, to stop a thread, and @code{-thread-info},
30204 to find the state of a thread, will always work.
30206 @node Thread groups
30207 @subsection Thread groups
30208 @value{GDBN} may be used to debug several processes at the same time.
30209 On some platforms, @value{GDBN} may support debugging of several
30210 hardware systems, each one having several cores with several different
30211 processes running on each core. This section describes the MI
30212 mechanism to support such debugging scenarios.
30214 The key observation is that regardless of the structure of the
30215 target, MI can have a global list of threads, because most commands that
30216 accept the @samp{--thread} option do not need to know what process that
30217 thread belongs to. Therefore, it is not necessary to introduce
30218 neither additional @samp{--process} option, nor an notion of the
30219 current process in the MI interface. The only strictly new feature
30220 that is required is the ability to find how the threads are grouped
30223 To allow the user to discover such grouping, and to support arbitrary
30224 hierarchy of machines/cores/processes, MI introduces the concept of a
30225 @dfn{thread group}. Thread group is a collection of threads and other
30226 thread groups. A thread group always has a string identifier, a type,
30227 and may have additional attributes specific to the type. A new
30228 command, @code{-list-thread-groups}, returns the list of top-level
30229 thread groups, which correspond to processes that @value{GDBN} is
30230 debugging at the moment. By passing an identifier of a thread group
30231 to the @code{-list-thread-groups} command, it is possible to obtain
30232 the members of specific thread group.
30234 To allow the user to easily discover processes, and other objects, he
30235 wishes to debug, a concept of @dfn{available thread group} is
30236 introduced. Available thread group is an thread group that
30237 @value{GDBN} is not debugging, but that can be attached to, using the
30238 @code{-target-attach} command. The list of available top-level thread
30239 groups can be obtained using @samp{-list-thread-groups --available}.
30240 In general, the content of a thread group may be only retrieved only
30241 after attaching to that thread group.
30243 Thread groups are related to inferiors (@pxref{Inferiors and
30244 Programs}). Each inferior corresponds to a thread group of a special
30245 type @samp{process}, and some additional operations are permitted on
30246 such thread groups.
30248 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30249 @node GDB/MI Command Syntax
30250 @section @sc{gdb/mi} Command Syntax
30253 * GDB/MI Input Syntax::
30254 * GDB/MI Output Syntax::
30257 @node GDB/MI Input Syntax
30258 @subsection @sc{gdb/mi} Input Syntax
30260 @cindex input syntax for @sc{gdb/mi}
30261 @cindex @sc{gdb/mi}, input syntax
30263 @item @var{command} @expansion{}
30264 @code{@var{cli-command} | @var{mi-command}}
30266 @item @var{cli-command} @expansion{}
30267 @code{[ @var{token} ] @var{cli-command} @var{nl}}, where
30268 @var{cli-command} is any existing @value{GDBN} CLI command.
30270 @item @var{mi-command} @expansion{}
30271 @code{[ @var{token} ] "-" @var{operation} ( " " @var{option} )*
30272 @code{[} " --" @code{]} ( " " @var{parameter} )* @var{nl}}
30274 @item @var{token} @expansion{}
30275 "any sequence of digits"
30277 @item @var{option} @expansion{}
30278 @code{"-" @var{parameter} [ " " @var{parameter} ]}
30280 @item @var{parameter} @expansion{}
30281 @code{@var{non-blank-sequence} | @var{c-string}}
30283 @item @var{operation} @expansion{}
30284 @emph{any of the operations described in this chapter}
30286 @item @var{non-blank-sequence} @expansion{}
30287 @emph{anything, provided it doesn't contain special characters such as
30288 "-", @var{nl}, """ and of course " "}
30290 @item @var{c-string} @expansion{}
30291 @code{""" @var{seven-bit-iso-c-string-content} """}
30293 @item @var{nl} @expansion{}
30302 The CLI commands are still handled by the @sc{mi} interpreter; their
30303 output is described below.
30306 The @code{@var{token}}, when present, is passed back when the command
30310 Some @sc{mi} commands accept optional arguments as part of the parameter
30311 list. Each option is identified by a leading @samp{-} (dash) and may be
30312 followed by an optional argument parameter. Options occur first in the
30313 parameter list and can be delimited from normal parameters using
30314 @samp{--} (this is useful when some parameters begin with a dash).
30321 We want easy access to the existing CLI syntax (for debugging).
30324 We want it to be easy to spot a @sc{mi} operation.
30327 @node GDB/MI Output Syntax
30328 @subsection @sc{gdb/mi} Output Syntax
30330 @cindex output syntax of @sc{gdb/mi}
30331 @cindex @sc{gdb/mi}, output syntax
30332 The output from @sc{gdb/mi} consists of zero or more out-of-band records
30333 followed, optionally, by a single result record. This result record
30334 is for the most recent command. The sequence of output records is
30335 terminated by @samp{(gdb)}.
30337 If an input command was prefixed with a @code{@var{token}} then the
30338 corresponding output for that command will also be prefixed by that same
30342 @item @var{output} @expansion{}
30343 @code{( @var{out-of-band-record} )* [ @var{result-record} ] "(gdb)" @var{nl}}
30345 @item @var{result-record} @expansion{}
30346 @code{ [ @var{token} ] "^" @var{result-class} ( "," @var{result} )* @var{nl}}
30348 @item @var{out-of-band-record} @expansion{}
30349 @code{@var{async-record} | @var{stream-record}}
30351 @item @var{async-record} @expansion{}
30352 @code{@var{exec-async-output} | @var{status-async-output} | @var{notify-async-output}}
30354 @item @var{exec-async-output} @expansion{}
30355 @code{[ @var{token} ] "*" @var{async-output nl}}
30357 @item @var{status-async-output} @expansion{}
30358 @code{[ @var{token} ] "+" @var{async-output nl}}
30360 @item @var{notify-async-output} @expansion{}
30361 @code{[ @var{token} ] "=" @var{async-output nl}}
30363 @item @var{async-output} @expansion{}
30364 @code{@var{async-class} ( "," @var{result} )*}
30366 @item @var{result-class} @expansion{}
30367 @code{"done" | "running" | "connected" | "error" | "exit"}
30369 @item @var{async-class} @expansion{}
30370 @code{"stopped" | @var{others}} (where @var{others} will be added
30371 depending on the needs---this is still in development).
30373 @item @var{result} @expansion{}
30374 @code{ @var{variable} "=" @var{value}}
30376 @item @var{variable} @expansion{}
30377 @code{ @var{string} }
30379 @item @var{value} @expansion{}
30380 @code{ @var{const} | @var{tuple} | @var{list} }
30382 @item @var{const} @expansion{}
30383 @code{@var{c-string}}
30385 @item @var{tuple} @expansion{}
30386 @code{ "@{@}" | "@{" @var{result} ( "," @var{result} )* "@}" }
30388 @item @var{list} @expansion{}
30389 @code{ "[]" | "[" @var{value} ( "," @var{value} )* "]" | "["
30390 @var{result} ( "," @var{result} )* "]" }
30392 @item @var{stream-record} @expansion{}
30393 @code{@var{console-stream-output} | @var{target-stream-output} | @var{log-stream-output}}
30395 @item @var{console-stream-output} @expansion{}
30396 @code{"~" @var{c-string nl}}
30398 @item @var{target-stream-output} @expansion{}
30399 @code{"@@" @var{c-string nl}}
30401 @item @var{log-stream-output} @expansion{}
30402 @code{"&" @var{c-string nl}}
30404 @item @var{nl} @expansion{}
30407 @item @var{token} @expansion{}
30408 @emph{any sequence of digits}.
30416 All output sequences end in a single line containing a period.
30419 The @code{@var{token}} is from the corresponding request. Note that
30420 for all async output, while the token is allowed by the grammar and
30421 may be output by future versions of @value{GDBN} for select async
30422 output messages, it is generally omitted. Frontends should treat
30423 all async output as reporting general changes in the state of the
30424 target and there should be no need to associate async output to any
30428 @cindex status output in @sc{gdb/mi}
30429 @var{status-async-output} contains on-going status information about the
30430 progress of a slow operation. It can be discarded. All status output is
30431 prefixed by @samp{+}.
30434 @cindex async output in @sc{gdb/mi}
30435 @var{exec-async-output} contains asynchronous state change on the target
30436 (stopped, started, disappeared). All async output is prefixed by
30440 @cindex notify output in @sc{gdb/mi}
30441 @var{notify-async-output} contains supplementary information that the
30442 client should handle (e.g., a new breakpoint information). All notify
30443 output is prefixed by @samp{=}.
30446 @cindex console output in @sc{gdb/mi}
30447 @var{console-stream-output} is output that should be displayed as is in the
30448 console. It is the textual response to a CLI command. All the console
30449 output is prefixed by @samp{~}.
30452 @cindex target output in @sc{gdb/mi}
30453 @var{target-stream-output} is the output produced by the target program.
30454 All the target output is prefixed by @samp{@@}.
30457 @cindex log output in @sc{gdb/mi}
30458 @var{log-stream-output} is output text coming from @value{GDBN}'s internals, for
30459 instance messages that should be displayed as part of an error log. All
30460 the log output is prefixed by @samp{&}.
30463 @cindex list output in @sc{gdb/mi}
30464 New @sc{gdb/mi} commands should only output @var{lists} containing
30470 @xref{GDB/MI Stream Records, , @sc{gdb/mi} Stream Records}, for more
30471 details about the various output records.
30473 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30474 @node GDB/MI Compatibility with CLI
30475 @section @sc{gdb/mi} Compatibility with CLI
30477 @cindex compatibility, @sc{gdb/mi} and CLI
30478 @cindex @sc{gdb/mi}, compatibility with CLI
30480 For the developers convenience CLI commands can be entered directly,
30481 but there may be some unexpected behaviour. For example, commands
30482 that query the user will behave as if the user replied yes, breakpoint
30483 command lists are not executed and some CLI commands, such as
30484 @code{if}, @code{when} and @code{define}, prompt for further input with
30485 @samp{>}, which is not valid MI output.
30487 This feature may be removed at some stage in the future and it is
30488 recommended that front ends use the @code{-interpreter-exec} command
30489 (@pxref{-interpreter-exec}).
30491 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30492 @node GDB/MI Development and Front Ends
30493 @section @sc{gdb/mi} Development and Front Ends
30494 @cindex @sc{gdb/mi} development
30496 The application which takes the MI output and presents the state of the
30497 program being debugged to the user is called a @dfn{front end}.
30499 Since @sc{gdb/mi} is used by a variety of front ends to @value{GDBN}, changes
30500 to the MI interface may break existing usage. This section describes how the
30501 protocol changes and how to request previous version of the protocol when it
30504 Some changes in MI need not break a carefully designed front end, and
30505 for these the MI version will remain unchanged. The following is a
30506 list of changes that may occur within one level, so front ends should
30507 parse MI output in a way that can handle them:
30511 New MI commands may be added.
30514 New fields may be added to the output of any MI command.
30517 The range of values for fields with specified values, e.g.,
30518 @code{in_scope} (@pxref{-var-update}) may be extended.
30520 @c The format of field's content e.g type prefix, may change so parse it
30521 @c at your own risk. Yes, in general?
30523 @c The order of fields may change? Shouldn't really matter but it might
30524 @c resolve inconsistencies.
30527 If the changes are likely to break front ends, the MI version level
30528 will be increased by one. The new versions of the MI protocol are not compatible
30529 with the old versions. Old versions of MI remain available, allowing front ends
30530 to keep using them until they are modified to use the latest MI version.
30532 Since @code{--interpreter=mi} always points to the latest MI version, it is
30533 recommended that front ends request a specific version of MI when launching
30534 @value{GDBN} (e.g. @code{--interpreter=mi2}) to make sure they get an
30535 interpreter with the MI version they expect.
30537 The following table gives a summary of the the released versions of the MI
30538 interface: the version number, the version of GDB in which it first appeared
30539 and the breaking changes compared to the previous version.
30541 @multitable @columnfractions .05 .05 .9
30542 @headitem MI version @tab GDB version @tab Breaking changes
30559 The @code{-environment-pwd}, @code{-environment-directory} and
30560 @code{-environment-path} commands now returns values using the MI output
30561 syntax, rather than CLI output syntax.
30564 @code{-var-list-children}'s @code{children} result field is now a list, rather
30568 @code{-var-update}'s @code{changelist} result field is now a list, rather than
30580 The output of information about multi-location breakpoints has changed in the
30581 responses to the @code{-break-insert} and @code{-break-info} commands, as well
30582 as in the @code{=breakpoint-created} and @code{=breakpoint-modified} events.
30583 The multiple locations are now placed in a @code{locations} field, whose value
30589 If your front end cannot yet migrate to a more recent version of the
30590 MI protocol, you can nevertheless selectively enable specific features
30591 available in those recent MI versions, using the following commands:
30595 @item -fix-multi-location-breakpoint-output
30596 Use the output for multi-location breakpoints which was introduced by
30597 MI 3, even when using MI versions 2 or 1. This command has no
30598 effect when using MI version 3 or later.
30602 The best way to avoid unexpected changes in MI that might break your front
30603 end is to make your project known to @value{GDBN} developers and
30604 follow development on @email{gdb@@sourceware.org} and
30605 @email{gdb-patches@@sourceware.org}.
30606 @cindex mailing lists
30608 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30609 @node GDB/MI Output Records
30610 @section @sc{gdb/mi} Output Records
30613 * GDB/MI Result Records::
30614 * GDB/MI Stream Records::
30615 * GDB/MI Async Records::
30616 * GDB/MI Breakpoint Information::
30617 * GDB/MI Frame Information::
30618 * GDB/MI Thread Information::
30619 * GDB/MI Ada Exception Information::
30622 @node GDB/MI Result Records
30623 @subsection @sc{gdb/mi} Result Records
30625 @cindex result records in @sc{gdb/mi}
30626 @cindex @sc{gdb/mi}, result records
30627 In addition to a number of out-of-band notifications, the response to a
30628 @sc{gdb/mi} command includes one of the following result indications:
30632 @item "^done" [ "," @var{results} ]
30633 The synchronous operation was successful, @code{@var{results}} are the return
30638 This result record is equivalent to @samp{^done}. Historically, it
30639 was output instead of @samp{^done} if the command has resumed the
30640 target. This behaviour is maintained for backward compatibility, but
30641 all frontends should treat @samp{^done} and @samp{^running}
30642 identically and rely on the @samp{*running} output record to determine
30643 which threads are resumed.
30647 @value{GDBN} has connected to a remote target.
30649 @item "^error" "," "msg=" @var{c-string} [ "," "code=" @var{c-string} ]
30651 The operation failed. The @code{msg=@var{c-string}} variable contains
30652 the corresponding error message.
30654 If present, the @code{code=@var{c-string}} variable provides an error
30655 code on which consumers can rely on to detect the corresponding
30656 error condition. At present, only one error code is defined:
30659 @item "undefined-command"
30660 Indicates that the command causing the error does not exist.
30665 @value{GDBN} has terminated.
30669 @node GDB/MI Stream Records
30670 @subsection @sc{gdb/mi} Stream Records
30672 @cindex @sc{gdb/mi}, stream records
30673 @cindex stream records in @sc{gdb/mi}
30674 @value{GDBN} internally maintains a number of output streams: the console, the
30675 target, and the log. The output intended for each of these streams is
30676 funneled through the @sc{gdb/mi} interface using @dfn{stream records}.
30678 Each stream record begins with a unique @dfn{prefix character} which
30679 identifies its stream (@pxref{GDB/MI Output Syntax, , @sc{gdb/mi} Output
30680 Syntax}). In addition to the prefix, each stream record contains a
30681 @code{@var{string-output}}. This is either raw text (with an implicit new
30682 line) or a quoted C string (which does not contain an implicit newline).
30685 @item "~" @var{string-output}
30686 The console output stream contains text that should be displayed in the
30687 CLI console window. It contains the textual responses to CLI commands.
30689 @item "@@" @var{string-output}
30690 The target output stream contains any textual output from the running
30691 target. This is only present when GDB's event loop is truly
30692 asynchronous, which is currently only the case for remote targets.
30694 @item "&" @var{string-output}
30695 The log stream contains debugging messages being produced by @value{GDBN}'s
30699 @node GDB/MI Async Records
30700 @subsection @sc{gdb/mi} Async Records
30702 @cindex async records in @sc{gdb/mi}
30703 @cindex @sc{gdb/mi}, async records
30704 @dfn{Async} records are used to notify the @sc{gdb/mi} client of
30705 additional changes that have occurred. Those changes can either be a
30706 consequence of @sc{gdb/mi} commands (e.g., a breakpoint modified) or a result of
30707 target activity (e.g., target stopped).
30709 The following is the list of possible async records:
30713 @item *running,thread-id="@var{thread}"
30714 The target is now running. The @var{thread} field can be the global
30715 thread ID of the the thread that is now running, and it can be
30716 @samp{all} if all threads are running. The frontend should assume
30717 that no interaction with a running thread is possible after this
30718 notification is produced. The frontend should not assume that this
30719 notification is output only once for any command. @value{GDBN} may
30720 emit this notification several times, either for different threads,
30721 because it cannot resume all threads together, or even for a single
30722 thread, if the thread must be stepped though some code before letting
30725 @item *stopped,reason="@var{reason}",thread-id="@var{id}",stopped-threads="@var{stopped}",core="@var{core}"
30726 The target has stopped. The @var{reason} field can have one of the
30730 @item breakpoint-hit
30731 A breakpoint was reached.
30732 @item watchpoint-trigger
30733 A watchpoint was triggered.
30734 @item read-watchpoint-trigger
30735 A read watchpoint was triggered.
30736 @item access-watchpoint-trigger
30737 An access watchpoint was triggered.
30738 @item function-finished
30739 An -exec-finish or similar CLI command was accomplished.
30740 @item location-reached
30741 An -exec-until or similar CLI command was accomplished.
30742 @item watchpoint-scope
30743 A watchpoint has gone out of scope.
30744 @item end-stepping-range
30745 An -exec-next, -exec-next-instruction, -exec-step, -exec-step-instruction or
30746 similar CLI command was accomplished.
30747 @item exited-signalled
30748 The inferior exited because of a signal.
30750 The inferior exited.
30751 @item exited-normally
30752 The inferior exited normally.
30753 @item signal-received
30754 A signal was received by the inferior.
30756 The inferior has stopped due to a library being loaded or unloaded.
30757 This can happen when @code{stop-on-solib-events} (@pxref{Files}) is
30758 set or when a @code{catch load} or @code{catch unload} catchpoint is
30759 in use (@pxref{Set Catchpoints}).
30761 The inferior has forked. This is reported when @code{catch fork}
30762 (@pxref{Set Catchpoints}) has been used.
30764 The inferior has vforked. This is reported in when @code{catch vfork}
30765 (@pxref{Set Catchpoints}) has been used.
30766 @item syscall-entry
30767 The inferior entered a system call. This is reported when @code{catch
30768 syscall} (@pxref{Set Catchpoints}) has been used.
30769 @item syscall-return
30770 The inferior returned from a system call. This is reported when
30771 @code{catch syscall} (@pxref{Set Catchpoints}) has been used.
30773 The inferior called @code{exec}. This is reported when @code{catch exec}
30774 (@pxref{Set Catchpoints}) has been used.
30777 The @var{id} field identifies the global thread ID of the thread
30778 that directly caused the stop -- for example by hitting a breakpoint.
30779 Depending on whether all-stop
30780 mode is in effect (@pxref{All-Stop Mode}), @value{GDBN} may either
30781 stop all threads, or only the thread that directly triggered the stop.
30782 If all threads are stopped, the @var{stopped} field will have the
30783 value of @code{"all"}. Otherwise, the value of the @var{stopped}
30784 field will be a list of thread identifiers. Presently, this list will
30785 always include a single thread, but frontend should be prepared to see
30786 several threads in the list. The @var{core} field reports the
30787 processor core on which the stop event has happened. This field may be absent
30788 if such information is not available.
30790 @item =thread-group-added,id="@var{id}"
30791 @itemx =thread-group-removed,id="@var{id}"
30792 A thread group was either added or removed. The @var{id} field
30793 contains the @value{GDBN} identifier of the thread group. When a thread
30794 group is added, it generally might not be associated with a running
30795 process. When a thread group is removed, its id becomes invalid and
30796 cannot be used in any way.
30798 @item =thread-group-started,id="@var{id}",pid="@var{pid}"
30799 A thread group became associated with a running program,
30800 either because the program was just started or the thread group
30801 was attached to a program. The @var{id} field contains the
30802 @value{GDBN} identifier of the thread group. The @var{pid} field
30803 contains process identifier, specific to the operating system.
30805 @item =thread-group-exited,id="@var{id}"[,exit-code="@var{code}"]
30806 A thread group is no longer associated with a running program,
30807 either because the program has exited, or because it was detached
30808 from. The @var{id} field contains the @value{GDBN} identifier of the
30809 thread group. The @var{code} field is the exit code of the inferior; it exists
30810 only when the inferior exited with some code.
30812 @item =thread-created,id="@var{id}",group-id="@var{gid}"
30813 @itemx =thread-exited,id="@var{id}",group-id="@var{gid}"
30814 A thread either was created, or has exited. The @var{id} field
30815 contains the global @value{GDBN} identifier of the thread. The @var{gid}
30816 field identifies the thread group this thread belongs to.
30818 @item =thread-selected,id="@var{id}"[,frame="@var{frame}"]
30819 Informs that the selected thread or frame were changed. This notification
30820 is not emitted as result of the @code{-thread-select} or
30821 @code{-stack-select-frame} commands, but is emitted whenever an MI command
30822 that is not documented to change the selected thread and frame actually
30823 changes them. In particular, invoking, directly or indirectly
30824 (via user-defined command), the CLI @code{thread} or @code{frame} commands,
30825 will generate this notification. Changing the thread or frame from another
30826 user interface (see @ref{Interpreters}) will also generate this notification.
30828 The @var{frame} field is only present if the newly selected thread is
30829 stopped. See @ref{GDB/MI Frame Information} for the format of its value.
30831 We suggest that in response to this notification, front ends
30832 highlight the selected thread and cause subsequent commands to apply to
30835 @item =library-loaded,...
30836 Reports that a new library file was loaded by the program. This
30837 notification has 5 fields---@var{id}, @var{target-name},
30838 @var{host-name}, @var{symbols-loaded} and @var{ranges}. The @var{id} field is an
30839 opaque identifier of the library. For remote debugging case,
30840 @var{target-name} and @var{host-name} fields give the name of the
30841 library file on the target, and on the host respectively. For native
30842 debugging, both those fields have the same value. The
30843 @var{symbols-loaded} field is emitted only for backward compatibility
30844 and should not be relied on to convey any useful information. The
30845 @var{thread-group} field, if present, specifies the id of the thread
30846 group in whose context the library was loaded. If the field is
30847 absent, it means the library was loaded in the context of all present
30848 thread groups. The @var{ranges} field specifies the ranges of addresses belonging
30851 @item =library-unloaded,...
30852 Reports that a library was unloaded by the program. This notification
30853 has 3 fields---@var{id}, @var{target-name} and @var{host-name} with
30854 the same meaning as for the @code{=library-loaded} notification.
30855 The @var{thread-group} field, if present, specifies the id of the
30856 thread group in whose context the library was unloaded. If the field is
30857 absent, it means the library was unloaded in the context of all present
30860 @item =traceframe-changed,num=@var{tfnum},tracepoint=@var{tpnum}
30861 @itemx =traceframe-changed,end
30862 Reports that the trace frame was changed and its new number is
30863 @var{tfnum}. The number of the tracepoint associated with this trace
30864 frame is @var{tpnum}.
30866 @item =tsv-created,name=@var{name},initial=@var{initial}
30867 Reports that the new trace state variable @var{name} is created with
30868 initial value @var{initial}.
30870 @item =tsv-deleted,name=@var{name}
30871 @itemx =tsv-deleted
30872 Reports that the trace state variable @var{name} is deleted or all
30873 trace state variables are deleted.
30875 @item =tsv-modified,name=@var{name},initial=@var{initial}[,current=@var{current}]
30876 Reports that the trace state variable @var{name} is modified with
30877 the initial value @var{initial}. The current value @var{current} of
30878 trace state variable is optional and is reported if the current
30879 value of trace state variable is known.
30881 @item =breakpoint-created,bkpt=@{...@}
30882 @itemx =breakpoint-modified,bkpt=@{...@}
30883 @itemx =breakpoint-deleted,id=@var{number}
30884 Reports that a breakpoint was created, modified, or deleted,
30885 respectively. Only user-visible breakpoints are reported to the MI
30888 The @var{bkpt} argument is of the same form as returned by the various
30889 breakpoint commands; @xref{GDB/MI Breakpoint Commands}. The
30890 @var{number} is the ordinal number of the breakpoint.
30892 Note that if a breakpoint is emitted in the result record of a
30893 command, then it will not also be emitted in an async record.
30895 @item =record-started,thread-group="@var{id}",method="@var{method}"[,format="@var{format}"]
30896 @itemx =record-stopped,thread-group="@var{id}"
30897 Execution log recording was either started or stopped on an
30898 inferior. The @var{id} is the @value{GDBN} identifier of the thread
30899 group corresponding to the affected inferior.
30901 The @var{method} field indicates the method used to record execution. If the
30902 method in use supports multiple recording formats, @var{format} will be present
30903 and contain the currently used format. @xref{Process Record and Replay},
30904 for existing method and format values.
30906 @item =cmd-param-changed,param=@var{param},value=@var{value}
30907 Reports that a parameter of the command @code{set @var{param}} is
30908 changed to @var{value}. In the multi-word @code{set} command,
30909 the @var{param} is the whole parameter list to @code{set} command.
30910 For example, In command @code{set check type on}, @var{param}
30911 is @code{check type} and @var{value} is @code{on}.
30913 @item =memory-changed,thread-group=@var{id},addr=@var{addr},len=@var{len}[,type="code"]
30914 Reports that bytes from @var{addr} to @var{data} + @var{len} were
30915 written in an inferior. The @var{id} is the identifier of the
30916 thread group corresponding to the affected inferior. The optional
30917 @code{type="code"} part is reported if the memory written to holds
30921 @node GDB/MI Breakpoint Information
30922 @subsection @sc{gdb/mi} Breakpoint Information
30924 When @value{GDBN} reports information about a breakpoint, a
30925 tracepoint, a watchpoint, or a catchpoint, it uses a tuple with the
30930 The breakpoint number.
30933 The type of the breakpoint. For ordinary breakpoints this will be
30934 @samp{breakpoint}, but many values are possible.
30937 If the type of the breakpoint is @samp{catchpoint}, then this
30938 indicates the exact type of catchpoint.
30941 This is the breakpoint disposition---either @samp{del}, meaning that
30942 the breakpoint will be deleted at the next stop, or @samp{keep},
30943 meaning that the breakpoint will not be deleted.
30946 This indicates whether the breakpoint is enabled, in which case the
30947 value is @samp{y}, or disabled, in which case the value is @samp{n}.
30948 Note that this is not the same as the field @code{enable}.
30951 The address of the breakpoint. This may be a hexidecimal number,
30952 giving the address; or the string @samp{<PENDING>}, for a pending
30953 breakpoint; or the string @samp{<MULTIPLE>}, for a breakpoint with
30954 multiple locations. This field will not be present if no address can
30955 be determined. For example, a watchpoint does not have an address.
30958 Optional field containing any flags related to the address. These flags are
30959 architecture-dependent; see @ref{Architectures} for their meaning for a
30963 If known, the function in which the breakpoint appears.
30964 If not known, this field is not present.
30967 The name of the source file which contains this function, if known.
30968 If not known, this field is not present.
30971 The full file name of the source file which contains this function, if
30972 known. If not known, this field is not present.
30975 The line number at which this breakpoint appears, if known.
30976 If not known, this field is not present.
30979 If the source file is not known, this field may be provided. If
30980 provided, this holds the address of the breakpoint, possibly followed
30984 If this breakpoint is pending, this field is present and holds the
30985 text used to set the breakpoint, as entered by the user.
30988 Where this breakpoint's condition is evaluated, either @samp{host} or
30992 If this is a thread-specific breakpoint, then this identifies the
30993 thread in which the breakpoint can trigger.
30996 If this breakpoint is restricted to a particular Ada task, then this
30997 field will hold the task identifier.
31000 If the breakpoint is conditional, this is the condition expression.
31003 The ignore count of the breakpoint.
31006 The enable count of the breakpoint.
31008 @item traceframe-usage
31011 @item static-tracepoint-marker-string-id
31012 For a static tracepoint, the name of the static tracepoint marker.
31015 For a masked watchpoint, this is the mask.
31018 A tracepoint's pass count.
31020 @item original-location
31021 The location of the breakpoint as originally specified by the user.
31022 This field is optional.
31025 The number of times the breakpoint has been hit.
31028 This field is only given for tracepoints. This is either @samp{y},
31029 meaning that the tracepoint is installed, or @samp{n}, meaning that it
31033 Some extra data, the exact contents of which are type-dependent.
31036 This field is present if the breakpoint has multiple locations. It is also
31037 exceptionally present if the breakpoint is enabled and has a single, disabled
31040 The value is a list of locations. The format of a location is described below.
31044 A location in a multi-location breakpoint is represented as a tuple with the
31050 The location number as a dotted pair, like @samp{1.2}. The first digit is the
31051 number of the parent breakpoint. The second digit is the number of the
31052 location within that breakpoint.
31055 This indicates whether the location is enabled, in which case the
31056 value is @samp{y}, or disabled, in which case the value is @samp{n}.
31057 Note that this is not the same as the field @code{enable}.
31060 The address of this location as an hexidecimal number.
31063 Optional field containing any flags related to the address. These flags are
31064 architecture-dependent; see @ref{Architectures} for their meaning for a
31068 If known, the function in which the location appears.
31069 If not known, this field is not present.
31072 The name of the source file which contains this location, if known.
31073 If not known, this field is not present.
31076 The full file name of the source file which contains this location, if
31077 known. If not known, this field is not present.
31080 The line number at which this location appears, if known.
31081 If not known, this field is not present.
31083 @item thread-groups
31084 The thread groups this location is in.
31088 For example, here is what the output of @code{-break-insert}
31089 (@pxref{GDB/MI Breakpoint Commands}) might be:
31092 -> -break-insert main
31093 <- ^done,bkpt=@{number="1",type="breakpoint",disp="keep",
31094 enabled="y",addr="0x08048564",func="main",file="myprog.c",
31095 fullname="/home/nickrob/myprog.c",line="68",thread-groups=["i1"],
31100 @node GDB/MI Frame Information
31101 @subsection @sc{gdb/mi} Frame Information
31103 Response from many MI commands includes an information about stack
31104 frame. This information is a tuple that may have the following
31109 The level of the stack frame. The innermost frame has the level of
31110 zero. This field is always present.
31113 The name of the function corresponding to the frame. This field may
31114 be absent if @value{GDBN} is unable to determine the function name.
31117 The code address for the frame. This field is always present.
31120 Optional field containing any flags related to the address. These flags are
31121 architecture-dependent; see @ref{Architectures} for their meaning for a
31125 The name of the source files that correspond to the frame's code
31126 address. This field may be absent.
31129 The source line corresponding to the frames' code address. This field
31133 The name of the binary file (either executable or shared library) the
31134 corresponds to the frame's code address. This field may be absent.
31138 @node GDB/MI Thread Information
31139 @subsection @sc{gdb/mi} Thread Information
31141 Whenever @value{GDBN} has to report an information about a thread, it
31142 uses a tuple with the following fields. The fields are always present unless
31147 The global numeric id assigned to the thread by @value{GDBN}.
31150 The target-specific string identifying the thread.
31153 Additional information about the thread provided by the target.
31154 It is supposed to be human-readable and not interpreted by the
31155 frontend. This field is optional.
31158 The name of the thread. If the user specified a name using the
31159 @code{thread name} command, then this name is given. Otherwise, if
31160 @value{GDBN} can extract the thread name from the target, then that
31161 name is given. If @value{GDBN} cannot find the thread name, then this
31165 The execution state of the thread, either @samp{stopped} or @samp{running},
31166 depending on whether the thread is presently running.
31169 The stack frame currently executing in the thread. This field is only present
31170 if the thread is stopped. Its format is documented in
31171 @ref{GDB/MI Frame Information}.
31174 The value of this field is an integer number of the processor core the
31175 thread was last seen on. This field is optional.
31178 @node GDB/MI Ada Exception Information
31179 @subsection @sc{gdb/mi} Ada Exception Information
31181 Whenever a @code{*stopped} record is emitted because the program
31182 stopped after hitting an exception catchpoint (@pxref{Set Catchpoints}),
31183 @value{GDBN} provides the name of the exception that was raised via
31184 the @code{exception-name} field. Also, for exceptions that were raised
31185 with an exception message, @value{GDBN} provides that message via
31186 the @code{exception-message} field.
31188 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
31189 @node GDB/MI Simple Examples
31190 @section Simple Examples of @sc{gdb/mi} Interaction
31191 @cindex @sc{gdb/mi}, simple examples
31193 This subsection presents several simple examples of interaction using
31194 the @sc{gdb/mi} interface. In these examples, @samp{->} means that the
31195 following line is passed to @sc{gdb/mi} as input, while @samp{<-} means
31196 the output received from @sc{gdb/mi}.
31198 Note the line breaks shown in the examples are here only for
31199 readability, they don't appear in the real output.
31201 @subheading Setting a Breakpoint
31203 Setting a breakpoint generates synchronous output which contains detailed
31204 information of the breakpoint.
31207 -> -break-insert main
31208 <- ^done,bkpt=@{number="1",type="breakpoint",disp="keep",
31209 enabled="y",addr="0x08048564",func="main",file="myprog.c",
31210 fullname="/home/nickrob/myprog.c",line="68",thread-groups=["i1"],
31215 @subheading Program Execution
31217 Program execution generates asynchronous records and MI gives the
31218 reason that execution stopped.
31224 <- *stopped,reason="breakpoint-hit",disp="keep",bkptno="1",thread-id="0",
31225 frame=@{addr="0x08048564",func="main",
31226 args=[@{name="argc",value="1"@},@{name="argv",value="0xbfc4d4d4"@}],
31227 file="myprog.c",fullname="/home/nickrob/myprog.c",line="68",
31228 arch="i386:x86_64"@}
31233 <- *stopped,reason="exited-normally"
31237 @subheading Quitting @value{GDBN}
31239 Quitting @value{GDBN} just prints the result class @samp{^exit}.
31247 Please note that @samp{^exit} is printed immediately, but it might
31248 take some time for @value{GDBN} to actually exit. During that time, @value{GDBN}
31249 performs necessary cleanups, including killing programs being debugged
31250 or disconnecting from debug hardware, so the frontend should wait till
31251 @value{GDBN} exits and should only forcibly kill @value{GDBN} if it
31252 fails to exit in reasonable time.
31254 @subheading A Bad Command
31256 Here's what happens if you pass a non-existent command:
31260 <- ^error,msg="Undefined MI command: rubbish"
31265 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
31266 @node GDB/MI Command Description Format
31267 @section @sc{gdb/mi} Command Description Format
31269 The remaining sections describe blocks of commands. Each block of
31270 commands is laid out in a fashion similar to this section.
31272 @subheading Motivation
31274 The motivation for this collection of commands.
31276 @subheading Introduction
31278 A brief introduction to this collection of commands as a whole.
31280 @subheading Commands
31282 For each command in the block, the following is described:
31284 @subsubheading Synopsis
31287 -command @var{args}@dots{}
31290 @subsubheading Result
31292 @subsubheading @value{GDBN} Command
31294 The corresponding @value{GDBN} CLI command(s), if any.
31296 @subsubheading Example
31298 Example(s) formatted for readability. Some of the described commands have
31299 not been implemented yet and these are labeled N.A.@: (not available).
31302 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
31303 @node GDB/MI Breakpoint Commands
31304 @section @sc{gdb/mi} Breakpoint Commands
31306 @cindex breakpoint commands for @sc{gdb/mi}
31307 @cindex @sc{gdb/mi}, breakpoint commands
31308 This section documents @sc{gdb/mi} commands for manipulating
31311 @subheading The @code{-break-after} Command
31312 @findex -break-after
31314 @subsubheading Synopsis
31317 -break-after @var{number} @var{count}
31320 The breakpoint number @var{number} is not in effect until it has been
31321 hit @var{count} times. To see how this is reflected in the output of
31322 the @samp{-break-list} command, see the description of the
31323 @samp{-break-list} command below.
31325 @subsubheading @value{GDBN} Command
31327 The corresponding @value{GDBN} command is @samp{ignore}.
31329 @subsubheading Example
31334 ^done,bkpt=@{number="1",type="breakpoint",disp="keep",
31335 enabled="y",addr="0x000100d0",func="main",file="hello.c",
31336 fullname="/home/foo/hello.c",line="5",thread-groups=["i1"],
31344 ^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
31345 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31346 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31347 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31348 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31349 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31350 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31351 body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
31352 addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
31353 line="5",thread-groups=["i1"],times="0",ignore="3"@}]@}
31358 @subheading The @code{-break-catch} Command
31359 @findex -break-catch
31362 @subheading The @code{-break-commands} Command
31363 @findex -break-commands
31365 @subsubheading Synopsis
31368 -break-commands @var{number} [ @var{command1} ... @var{commandN} ]
31371 Specifies the CLI commands that should be executed when breakpoint
31372 @var{number} is hit. The parameters @var{command1} to @var{commandN}
31373 are the commands. If no command is specified, any previously-set
31374 commands are cleared. @xref{Break Commands}. Typical use of this
31375 functionality is tracing a program, that is, printing of values of
31376 some variables whenever breakpoint is hit and then continuing.
31378 @subsubheading @value{GDBN} Command
31380 The corresponding @value{GDBN} command is @samp{commands}.
31382 @subsubheading Example
31387 ^done,bkpt=@{number="1",type="breakpoint",disp="keep",
31388 enabled="y",addr="0x000100d0",func="main",file="hello.c",
31389 fullname="/home/foo/hello.c",line="5",thread-groups=["i1"],
31392 -break-commands 1 "print v" "continue"
31397 @subheading The @code{-break-condition} Command
31398 @findex -break-condition
31400 @subsubheading Synopsis
31403 -break-condition @var{number} @var{expr}
31406 Breakpoint @var{number} will stop the program only if the condition in
31407 @var{expr} is true. The condition becomes part of the
31408 @samp{-break-list} output (see the description of the @samp{-break-list}
31411 @subsubheading @value{GDBN} Command
31413 The corresponding @value{GDBN} command is @samp{condition}.
31415 @subsubheading Example
31419 -break-condition 1 1
31423 ^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
31424 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31425 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31426 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31427 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31428 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31429 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31430 body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
31431 addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
31432 line="5",cond="1",thread-groups=["i1"],times="0",ignore="3"@}]@}
31436 @subheading The @code{-break-delete} Command
31437 @findex -break-delete
31439 @subsubheading Synopsis
31442 -break-delete ( @var{breakpoint} )+
31445 Delete the breakpoint(s) whose number(s) are specified in the argument
31446 list. This is obviously reflected in the breakpoint list.
31448 @subsubheading @value{GDBN} Command
31450 The corresponding @value{GDBN} command is @samp{delete}.
31452 @subsubheading Example
31460 ^done,BreakpointTable=@{nr_rows="0",nr_cols="6",
31461 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31462 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31463 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31464 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31465 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31466 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31471 @subheading The @code{-break-disable} Command
31472 @findex -break-disable
31474 @subsubheading Synopsis
31477 -break-disable ( @var{breakpoint} )+
31480 Disable the named @var{breakpoint}(s). The field @samp{enabled} in the
31481 break list is now set to @samp{n} for the named @var{breakpoint}(s).
31483 @subsubheading @value{GDBN} Command
31485 The corresponding @value{GDBN} command is @samp{disable}.
31487 @subsubheading Example
31495 ^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
31496 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31497 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31498 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31499 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31500 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31501 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31502 body=[bkpt=@{number="2",type="breakpoint",disp="keep",enabled="n",
31503 addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
31504 line="5",thread-groups=["i1"],times="0"@}]@}
31508 @subheading The @code{-break-enable} Command
31509 @findex -break-enable
31511 @subsubheading Synopsis
31514 -break-enable ( @var{breakpoint} )+
31517 Enable (previously disabled) @var{breakpoint}(s).
31519 @subsubheading @value{GDBN} Command
31521 The corresponding @value{GDBN} command is @samp{enable}.
31523 @subsubheading Example
31531 ^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
31532 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31533 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31534 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31535 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31536 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31537 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31538 body=[bkpt=@{number="2",type="breakpoint",disp="keep",enabled="y",
31539 addr="0x000100d0",func="main",file="hello.c",fullname="/home/foo/hello.c",
31540 line="5",thread-groups=["i1"],times="0"@}]@}
31544 @subheading The @code{-break-info} Command
31545 @findex -break-info
31547 @subsubheading Synopsis
31550 -break-info @var{breakpoint}
31554 Get information about a single breakpoint.
31556 The result is a table of breakpoints. @xref{GDB/MI Breakpoint
31557 Information}, for details on the format of each breakpoint in the
31560 @subsubheading @value{GDBN} Command
31562 The corresponding @value{GDBN} command is @samp{info break @var{breakpoint}}.
31564 @subsubheading Example
31567 @subheading The @code{-break-insert} Command
31568 @findex -break-insert
31569 @anchor{-break-insert}
31571 @subsubheading Synopsis
31574 -break-insert [ -t ] [ -h ] [ -f ] [ -d ] [ -a ]
31575 [ -c @var{condition} ] [ -i @var{ignore-count} ]
31576 [ -p @var{thread-id} ] [ @var{location} ]
31580 If specified, @var{location}, can be one of:
31583 @item linespec location
31584 A linespec location. @xref{Linespec Locations}.
31586 @item explicit location
31587 An explicit location. @sc{gdb/mi} explicit locations are
31588 analogous to the CLI's explicit locations using the option names
31589 listed below. @xref{Explicit Locations}.
31592 @item --source @var{filename}
31593 The source file name of the location. This option requires the use
31594 of either @samp{--function} or @samp{--line}.
31596 @item --function @var{function}
31597 The name of a function or method.
31599 @item --label @var{label}
31600 The name of a label.
31602 @item --line @var{lineoffset}
31603 An absolute or relative line offset from the start of the location.
31606 @item address location
31607 An address location, *@var{address}. @xref{Address Locations}.
31611 The possible optional parameters of this command are:
31615 Insert a temporary breakpoint.
31617 Insert a hardware breakpoint.
31619 If @var{location} cannot be parsed (for example if it
31620 refers to unknown files or functions), create a pending
31621 breakpoint. Without this flag, @value{GDBN} will report
31622 an error, and won't create a breakpoint, if @var{location}
31625 Create a disabled breakpoint.
31627 Create a tracepoint. @xref{Tracepoints}. When this parameter
31628 is used together with @samp{-h}, a fast tracepoint is created.
31629 @item -c @var{condition}
31630 Make the breakpoint conditional on @var{condition}.
31631 @item -i @var{ignore-count}
31632 Initialize the @var{ignore-count}.
31633 @item -p @var{thread-id}
31634 Restrict the breakpoint to the thread with the specified global
31638 @subsubheading Result
31640 @xref{GDB/MI Breakpoint Information}, for details on the format of the
31641 resulting breakpoint.
31643 Note: this format is open to change.
31644 @c An out-of-band breakpoint instead of part of the result?
31646 @subsubheading @value{GDBN} Command
31648 The corresponding @value{GDBN} commands are @samp{break}, @samp{tbreak},
31649 @samp{hbreak}, and @samp{thbreak}. @c and @samp{rbreak}.
31651 @subsubheading Example
31656 ^done,bkpt=@{number="1",addr="0x0001072c",file="recursive2.c",
31657 fullname="/home/foo/recursive2.c,line="4",thread-groups=["i1"],
31660 -break-insert -t foo
31661 ^done,bkpt=@{number="2",addr="0x00010774",file="recursive2.c",
31662 fullname="/home/foo/recursive2.c,line="11",thread-groups=["i1"],
31666 ^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
31667 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31668 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31669 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31670 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31671 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31672 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31673 body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
31674 addr="0x0001072c", func="main",file="recursive2.c",
31675 fullname="/home/foo/recursive2.c,"line="4",thread-groups=["i1"],
31677 bkpt=@{number="2",type="breakpoint",disp="del",enabled="y",
31678 addr="0x00010774",func="foo",file="recursive2.c",
31679 fullname="/home/foo/recursive2.c",line="11",thread-groups=["i1"],
31682 @c -break-insert -r foo.*
31683 @c ~int foo(int, int);
31684 @c ^done,bkpt=@{number="3",addr="0x00010774",file="recursive2.c,
31685 @c "fullname="/home/foo/recursive2.c",line="11",thread-groups=["i1"],
31690 @subheading The @code{-dprintf-insert} Command
31691 @findex -dprintf-insert
31693 @subsubheading Synopsis
31696 -dprintf-insert [ -t ] [ -f ] [ -d ]
31697 [ -c @var{condition} ] [ -i @var{ignore-count} ]
31698 [ -p @var{thread-id} ] [ @var{location} ] [ @var{format} ]
31703 If supplied, @var{location} may be specified the same way as for
31704 the @code{-break-insert} command. @xref{-break-insert}.
31706 The possible optional parameters of this command are:
31710 Insert a temporary breakpoint.
31712 If @var{location} cannot be parsed (for example, if it
31713 refers to unknown files or functions), create a pending
31714 breakpoint. Without this flag, @value{GDBN} will report
31715 an error, and won't create a breakpoint, if @var{location}
31718 Create a disabled breakpoint.
31719 @item -c @var{condition}
31720 Make the breakpoint conditional on @var{condition}.
31721 @item -i @var{ignore-count}
31722 Set the ignore count of the breakpoint (@pxref{Conditions, ignore count})
31723 to @var{ignore-count}.
31724 @item -p @var{thread-id}
31725 Restrict the breakpoint to the thread with the specified global
31729 @subsubheading Result
31731 @xref{GDB/MI Breakpoint Information}, for details on the format of the
31732 resulting breakpoint.
31734 @c An out-of-band breakpoint instead of part of the result?
31736 @subsubheading @value{GDBN} Command
31738 The corresponding @value{GDBN} command is @samp{dprintf}.
31740 @subsubheading Example
31744 4-dprintf-insert foo "At foo entry\n"
31745 4^done,bkpt=@{number="1",type="dprintf",disp="keep",enabled="y",
31746 addr="0x000000000040061b",func="foo",file="mi-dprintf.c",
31747 fullname="mi-dprintf.c",line="25",thread-groups=["i1"],
31748 times="0",script=@{"printf \"At foo entry\\n\"","continue"@},
31749 original-location="foo"@}
31751 5-dprintf-insert 26 "arg=%d, g=%d\n" arg g
31752 5^done,bkpt=@{number="2",type="dprintf",disp="keep",enabled="y",
31753 addr="0x000000000040062a",func="foo",file="mi-dprintf.c",
31754 fullname="mi-dprintf.c",line="26",thread-groups=["i1"],
31755 times="0",script=@{"printf \"arg=%d, g=%d\\n\", arg, g","continue"@},
31756 original-location="mi-dprintf.c:26"@}
31760 @subheading The @code{-break-list} Command
31761 @findex -break-list
31763 @subsubheading Synopsis
31769 Displays the list of inserted breakpoints, showing the following fields:
31773 number of the breakpoint
31775 type of the breakpoint: @samp{breakpoint} or @samp{watchpoint}
31777 should the breakpoint be deleted or disabled when it is hit: @samp{keep}
31780 is the breakpoint enabled or no: @samp{y} or @samp{n}
31782 memory location at which the breakpoint is set
31784 logical location of the breakpoint, expressed by function name, file
31786 @item Thread-groups
31787 list of thread groups to which this breakpoint applies
31789 number of times the breakpoint has been hit
31792 If there are no breakpoints or watchpoints, the @code{BreakpointTable}
31793 @code{body} field is an empty list.
31795 @subsubheading @value{GDBN} Command
31797 The corresponding @value{GDBN} command is @samp{info break}.
31799 @subsubheading Example
31804 ^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
31805 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31806 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31807 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31808 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31809 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31810 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31811 body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
31812 addr="0x000100d0",func="main",file="hello.c",line="5",thread-groups=["i1"],
31814 bkpt=@{number="2",type="breakpoint",disp="keep",enabled="y",
31815 addr="0x00010114",func="foo",file="hello.c",fullname="/home/foo/hello.c",
31816 line="13",thread-groups=["i1"],times="0"@}]@}
31820 Here's an example of the result when there are no breakpoints:
31825 ^done,BreakpointTable=@{nr_rows="0",nr_cols="6",
31826 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31827 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31828 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31829 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31830 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31831 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31836 @subheading The @code{-break-passcount} Command
31837 @findex -break-passcount
31839 @subsubheading Synopsis
31842 -break-passcount @var{tracepoint-number} @var{passcount}
31845 Set the passcount for tracepoint @var{tracepoint-number} to
31846 @var{passcount}. If the breakpoint referred to by @var{tracepoint-number}
31847 is not a tracepoint, error is emitted. This corresponds to CLI
31848 command @samp{passcount}.
31850 @subheading The @code{-break-watch} Command
31851 @findex -break-watch
31853 @subsubheading Synopsis
31856 -break-watch [ -a | -r ]
31859 Create a watchpoint. With the @samp{-a} option it will create an
31860 @dfn{access} watchpoint, i.e., a watchpoint that triggers either on a
31861 read from or on a write to the memory location. With the @samp{-r}
31862 option, the watchpoint created is a @dfn{read} watchpoint, i.e., it will
31863 trigger only when the memory location is accessed for reading. Without
31864 either of the options, the watchpoint created is a regular watchpoint,
31865 i.e., it will trigger when the memory location is accessed for writing.
31866 @xref{Set Watchpoints, , Setting Watchpoints}.
31868 Note that @samp{-break-list} will report a single list of watchpoints and
31869 breakpoints inserted.
31871 @subsubheading @value{GDBN} Command
31873 The corresponding @value{GDBN} commands are @samp{watch}, @samp{awatch}, and
31876 @subsubheading Example
31878 Setting a watchpoint on a variable in the @code{main} function:
31883 ^done,wpt=@{number="2",exp="x"@}
31888 *stopped,reason="watchpoint-trigger",wpt=@{number="2",exp="x"@},
31889 value=@{old="-268439212",new="55"@},
31890 frame=@{func="main",args=[],file="recursive2.c",
31891 fullname="/home/foo/bar/recursive2.c",line="5",arch="i386:x86_64"@}
31895 Setting a watchpoint on a variable local to a function. @value{GDBN} will stop
31896 the program execution twice: first for the variable changing value, then
31897 for the watchpoint going out of scope.
31902 ^done,wpt=@{number="5",exp="C"@}
31907 *stopped,reason="watchpoint-trigger",
31908 wpt=@{number="5",exp="C"@},value=@{old="-276895068",new="3"@},
31909 frame=@{func="callee4",args=[],
31910 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31911 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="13",
31912 arch="i386:x86_64"@}
31917 *stopped,reason="watchpoint-scope",wpnum="5",
31918 frame=@{func="callee3",args=[@{name="strarg",
31919 value="0x11940 \"A string argument.\""@}],
31920 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31921 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="18",
31922 arch="i386:x86_64"@}
31926 Listing breakpoints and watchpoints, at different points in the program
31927 execution. Note that once the watchpoint goes out of scope, it is
31933 ^done,wpt=@{number="2",exp="C"@}
31936 ^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
31937 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31938 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31939 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31940 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31941 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31942 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31943 body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
31944 addr="0x00010734",func="callee4",
31945 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31946 fullname="/home/foo/devo/gdb/testsuite/gdb.mi/basics.c"line="8",thread-groups=["i1"],
31948 bkpt=@{number="2",type="watchpoint",disp="keep",
31949 enabled="y",addr="",what="C",thread-groups=["i1"],times="0"@}]@}
31954 *stopped,reason="watchpoint-trigger",wpt=@{number="2",exp="C"@},
31955 value=@{old="-276895068",new="3"@},
31956 frame=@{func="callee4",args=[],
31957 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31958 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="13",
31959 arch="i386:x86_64"@}
31962 ^done,BreakpointTable=@{nr_rows="2",nr_cols="6",
31963 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31964 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31965 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31966 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31967 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31968 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31969 body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
31970 addr="0x00010734",func="callee4",
31971 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31972 fullname="/home/foo/devo/gdb/testsuite/gdb.mi/basics.c",line="8",thread-groups=["i1"],
31974 bkpt=@{number="2",type="watchpoint",disp="keep",
31975 enabled="y",addr="",what="C",thread-groups=["i1"],times="-5"@}]@}
31979 ^done,reason="watchpoint-scope",wpnum="2",
31980 frame=@{func="callee3",args=[@{name="strarg",
31981 value="0x11940 \"A string argument.\""@}],
31982 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31983 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="18",
31984 arch="i386:x86_64"@}
31987 ^done,BreakpointTable=@{nr_rows="1",nr_cols="6",
31988 hdr=[@{width="3",alignment="-1",col_name="number",colhdr="Num"@},
31989 @{width="14",alignment="-1",col_name="type",colhdr="Type"@},
31990 @{width="4",alignment="-1",col_name="disp",colhdr="Disp"@},
31991 @{width="3",alignment="-1",col_name="enabled",colhdr="Enb"@},
31992 @{width="10",alignment="-1",col_name="addr",colhdr="Address"@},
31993 @{width="40",alignment="2",col_name="what",colhdr="What"@}],
31994 body=[bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
31995 addr="0x00010734",func="callee4",
31996 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
31997 fullname="/home/foo/devo/gdb/testsuite/gdb.mi/basics.c",line="8",
31998 thread-groups=["i1"],times="1"@}]@}
32003 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32004 @node GDB/MI Catchpoint Commands
32005 @section @sc{gdb/mi} Catchpoint Commands
32007 This section documents @sc{gdb/mi} commands for manipulating
32011 * Shared Library GDB/MI Catchpoint Commands::
32012 * Ada Exception GDB/MI Catchpoint Commands::
32013 * C++ Exception GDB/MI Catchpoint Commands::
32016 @node Shared Library GDB/MI Catchpoint Commands
32017 @subsection Shared Library @sc{gdb/mi} Catchpoints
32019 @subheading The @code{-catch-load} Command
32020 @findex -catch-load
32022 @subsubheading Synopsis
32025 -catch-load [ -t ] [ -d ] @var{regexp}
32028 Add a catchpoint for library load events. If the @samp{-t} option is used,
32029 the catchpoint is a temporary one (@pxref{Set Breaks, ,Setting
32030 Breakpoints}). If the @samp{-d} option is used, the catchpoint is created
32031 in a disabled state. The @samp{regexp} argument is a regular
32032 expression used to match the name of the loaded library.
32035 @subsubheading @value{GDBN} Command
32037 The corresponding @value{GDBN} command is @samp{catch load}.
32039 @subsubheading Example
32042 -catch-load -t foo.so
32043 ^done,bkpt=@{number="1",type="catchpoint",disp="del",enabled="y",
32044 what="load of library matching foo.so",catch-type="load",times="0"@}
32049 @subheading The @code{-catch-unload} Command
32050 @findex -catch-unload
32052 @subsubheading Synopsis
32055 -catch-unload [ -t ] [ -d ] @var{regexp}
32058 Add a catchpoint for library unload events. If the @samp{-t} option is
32059 used, the catchpoint is a temporary one (@pxref{Set Breaks, ,Setting
32060 Breakpoints}). If the @samp{-d} option is used, the catchpoint is
32061 created in a disabled state. The @samp{regexp} argument is a regular
32062 expression used to match the name of the unloaded library.
32064 @subsubheading @value{GDBN} Command
32066 The corresponding @value{GDBN} command is @samp{catch unload}.
32068 @subsubheading Example
32071 -catch-unload -d bar.so
32072 ^done,bkpt=@{number="2",type="catchpoint",disp="keep",enabled="n",
32073 what="load of library matching bar.so",catch-type="unload",times="0"@}
32077 @node Ada Exception GDB/MI Catchpoint Commands
32078 @subsection Ada Exception @sc{gdb/mi} Catchpoints
32080 The following @sc{gdb/mi} commands can be used to create catchpoints
32081 that stop the execution when Ada exceptions are being raised.
32083 @subheading The @code{-catch-assert} Command
32084 @findex -catch-assert
32086 @subsubheading Synopsis
32089 -catch-assert [ -c @var{condition}] [ -d ] [ -t ]
32092 Add a catchpoint for failed Ada assertions.
32094 The possible optional parameters for this command are:
32097 @item -c @var{condition}
32098 Make the catchpoint conditional on @var{condition}.
32100 Create a disabled catchpoint.
32102 Create a temporary catchpoint.
32105 @subsubheading @value{GDBN} Command
32107 The corresponding @value{GDBN} command is @samp{catch assert}.
32109 @subsubheading Example
32113 ^done,bkptno="5",bkpt=@{number="5",type="breakpoint",disp="keep",
32114 enabled="y",addr="0x0000000000404888",what="failed Ada assertions",
32115 thread-groups=["i1"],times="0",
32116 original-location="__gnat_debug_raise_assert_failure"@}
32120 @subheading The @code{-catch-exception} Command
32121 @findex -catch-exception
32123 @subsubheading Synopsis
32126 -catch-exception [ -c @var{condition}] [ -d ] [ -e @var{exception-name} ]
32130 Add a catchpoint stopping when Ada exceptions are raised.
32131 By default, the command stops the program when any Ada exception
32132 gets raised. But it is also possible, by using some of the
32133 optional parameters described below, to create more selective
32136 The possible optional parameters for this command are:
32139 @item -c @var{condition}
32140 Make the catchpoint conditional on @var{condition}.
32142 Create a disabled catchpoint.
32143 @item -e @var{exception-name}
32144 Only stop when @var{exception-name} is raised. This option cannot
32145 be used combined with @samp{-u}.
32147 Create a temporary catchpoint.
32149 Stop only when an unhandled exception gets raised. This option
32150 cannot be used combined with @samp{-e}.
32153 @subsubheading @value{GDBN} Command
32155 The corresponding @value{GDBN} commands are @samp{catch exception}
32156 and @samp{catch exception unhandled}.
32158 @subsubheading Example
32161 -catch-exception -e Program_Error
32162 ^done,bkptno="4",bkpt=@{number="4",type="breakpoint",disp="keep",
32163 enabled="y",addr="0x0000000000404874",
32164 what="`Program_Error' Ada exception", thread-groups=["i1"],
32165 times="0",original-location="__gnat_debug_raise_exception"@}
32169 @subheading The @code{-catch-handlers} Command
32170 @findex -catch-handlers
32172 @subsubheading Synopsis
32175 -catch-handlers [ -c @var{condition}] [ -d ] [ -e @var{exception-name} ]
32179 Add a catchpoint stopping when Ada exceptions are handled.
32180 By default, the command stops the program when any Ada exception
32181 gets handled. But it is also possible, by using some of the
32182 optional parameters described below, to create more selective
32185 The possible optional parameters for this command are:
32188 @item -c @var{condition}
32189 Make the catchpoint conditional on @var{condition}.
32191 Create a disabled catchpoint.
32192 @item -e @var{exception-name}
32193 Only stop when @var{exception-name} is handled.
32195 Create a temporary catchpoint.
32198 @subsubheading @value{GDBN} Command
32200 The corresponding @value{GDBN} command is @samp{catch handlers}.
32202 @subsubheading Example
32205 -catch-handlers -e Constraint_Error
32206 ^done,bkptno="4",bkpt=@{number="4",type="breakpoint",disp="keep",
32207 enabled="y",addr="0x0000000000402f68",
32208 what="`Constraint_Error' Ada exception handlers",thread-groups=["i1"],
32209 times="0",original-location="__gnat_begin_handler"@}
32213 @node C++ Exception GDB/MI Catchpoint Commands
32214 @subsection C@t{++} Exception @sc{gdb/mi} Catchpoints
32216 The following @sc{gdb/mi} commands can be used to create catchpoints
32217 that stop the execution when C@t{++} exceptions are being throw, rethrown,
32220 @subheading The @code{-catch-throw} Command
32221 @findex -catch-throw
32223 @subsubheading Synopsis
32226 -catch-throw [ -t ] [ -r @var{regexp}]
32229 Stop when the debuggee throws a C@t{++} exception. If @var{regexp} is
32230 given, then only exceptions whose type matches the regular expression
32233 If @samp{-t} is given, then the catchpoint is enabled only for one
32234 stop, the catchpoint is automatically deleted after stopping once for
32237 @subsubheading @value{GDBN} Command
32239 The corresponding @value{GDBN} commands are @samp{catch throw}
32240 and @samp{tcatch throw} (@pxref{Set Catchpoints}).
32242 @subsubheading Example
32245 -catch-throw -r exception_type
32246 ^done,bkpt=@{number="1",type="catchpoint",disp="keep",enabled="y",
32247 what="exception throw",catch-type="throw",
32248 thread-groups=["i1"],
32249 regexp="exception_type",times="0"@}
32255 ~"Catchpoint 1 (exception thrown), 0x00007ffff7ae00ed
32256 in __cxa_throw () from /lib64/libstdc++.so.6\n"
32257 *stopped,bkptno="1",reason="breakpoint-hit",disp="keep",
32258 frame=@{addr="0x00007ffff7ae00ed",func="__cxa_throw",
32259 args=[],from="/lib64/libstdc++.so.6",arch="i386:x86-64"@},
32260 thread-id="1",stopped-threads="all",core="6"
32264 @subheading The @code{-catch-rethrow} Command
32265 @findex -catch-rethrow
32267 @subsubheading Synopsis
32270 -catch-rethrow [ -t ] [ -r @var{regexp}]
32273 Stop when a C@t{++} exception is re-thrown. If @var{regexp} is given,
32274 then only exceptions whose type matches the regular expression will be
32277 If @samp{-t} is given, then the catchpoint is enabled only for one
32278 stop, the catchpoint is automatically deleted after the first event is
32281 @subsubheading @value{GDBN} Command
32283 The corresponding @value{GDBN} commands are @samp{catch rethrow}
32284 and @samp{tcatch rethrow} (@pxref{Set Catchpoints}).
32286 @subsubheading Example
32289 -catch-rethrow -r exception_type
32290 ^done,bkpt=@{number="1",type="catchpoint",disp="keep",enabled="y",
32291 what="exception rethrow",catch-type="rethrow",
32292 thread-groups=["i1"],
32293 regexp="exception_type",times="0"@}
32299 ~"Catchpoint 1 (exception rethrown), 0x00007ffff7ae00ed
32300 in __cxa_rethrow () from /lib64/libstdc++.so.6\n"
32301 *stopped,bkptno="1",reason="breakpoint-hit",disp="keep",
32302 frame=@{addr="0x00007ffff7ae00ed",func="__cxa_rethrow",
32303 args=[],from="/lib64/libstdc++.so.6",arch="i386:x86-64"@},
32304 thread-id="1",stopped-threads="all",core="6"
32308 @subheading The @code{-catch-catch} Command
32309 @findex -catch-catch
32311 @subsubheading Synopsis
32314 -catch-catch [ -t ] [ -r @var{regexp}]
32317 Stop when the debuggee catches a C@t{++} exception. If @var{regexp}
32318 is given, then only exceptions whose type matches the regular
32319 expression will be caught.
32321 If @samp{-t} is given, then the catchpoint is enabled only for one
32322 stop, the catchpoint is automatically deleted after the first event is
32325 @subsubheading @value{GDBN} Command
32327 The corresponding @value{GDBN} commands are @samp{catch catch}
32328 and @samp{tcatch catch} (@pxref{Set Catchpoints}).
32330 @subsubheading Example
32333 -catch-catch -r exception_type
32334 ^done,bkpt=@{number="1",type="catchpoint",disp="keep",enabled="y",
32335 what="exception catch",catch-type="catch",
32336 thread-groups=["i1"],
32337 regexp="exception_type",times="0"@}
32343 ~"Catchpoint 1 (exception caught), 0x00007ffff7ae00ed
32344 in __cxa_begin_catch () from /lib64/libstdc++.so.6\n"
32345 *stopped,bkptno="1",reason="breakpoint-hit",disp="keep",
32346 frame=@{addr="0x00007ffff7ae00ed",func="__cxa_begin_catch",
32347 args=[],from="/lib64/libstdc++.so.6",arch="i386:x86-64"@},
32348 thread-id="1",stopped-threads="all",core="6"
32352 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32353 @node GDB/MI Program Context
32354 @section @sc{gdb/mi} Program Context
32356 @subheading The @code{-exec-arguments} Command
32357 @findex -exec-arguments
32360 @subsubheading Synopsis
32363 -exec-arguments @var{args}
32366 Set the inferior program arguments, to be used in the next
32369 @subsubheading @value{GDBN} Command
32371 The corresponding @value{GDBN} command is @samp{set args}.
32373 @subsubheading Example
32377 -exec-arguments -v word
32384 @subheading The @code{-exec-show-arguments} Command
32385 @findex -exec-show-arguments
32387 @subsubheading Synopsis
32390 -exec-show-arguments
32393 Print the arguments of the program.
32395 @subsubheading @value{GDBN} Command
32397 The corresponding @value{GDBN} command is @samp{show args}.
32399 @subsubheading Example
32404 @subheading The @code{-environment-cd} Command
32405 @findex -environment-cd
32407 @subsubheading Synopsis
32410 -environment-cd @var{pathdir}
32413 Set @value{GDBN}'s working directory.
32415 @subsubheading @value{GDBN} Command
32417 The corresponding @value{GDBN} command is @samp{cd}.
32419 @subsubheading Example
32423 -environment-cd /kwikemart/marge/ezannoni/flathead-dev/devo/gdb
32429 @subheading The @code{-environment-directory} Command
32430 @findex -environment-directory
32432 @subsubheading Synopsis
32435 -environment-directory [ -r ] [ @var{pathdir} ]+
32438 Add directories @var{pathdir} to beginning of search path for source files.
32439 If the @samp{-r} option is used, the search path is reset to the default
32440 search path. If directories @var{pathdir} are supplied in addition to the
32441 @samp{-r} option, the search path is first reset and then addition
32443 Multiple directories may be specified, separated by blanks. Specifying
32444 multiple directories in a single command
32445 results in the directories added to the beginning of the
32446 search path in the same order they were presented in the command.
32447 If blanks are needed as
32448 part of a directory name, double-quotes should be used around
32449 the name. In the command output, the path will show up separated
32450 by the system directory-separator character. The directory-separator
32451 character must not be used
32452 in any directory name.
32453 If no directories are specified, the current search path is displayed.
32455 @subsubheading @value{GDBN} Command
32457 The corresponding @value{GDBN} command is @samp{dir}.
32459 @subsubheading Example
32463 -environment-directory /kwikemart/marge/ezannoni/flathead-dev/devo/gdb
32464 ^done,source-path="/kwikemart/marge/ezannoni/flathead-dev/devo/gdb:$cdir:$cwd"
32466 -environment-directory ""
32467 ^done,source-path="/kwikemart/marge/ezannoni/flathead-dev/devo/gdb:$cdir:$cwd"
32469 -environment-directory -r /home/jjohnstn/src/gdb /usr/src
32470 ^done,source-path="/home/jjohnstn/src/gdb:/usr/src:$cdir:$cwd"
32472 -environment-directory -r
32473 ^done,source-path="$cdir:$cwd"
32478 @subheading The @code{-environment-path} Command
32479 @findex -environment-path
32481 @subsubheading Synopsis
32484 -environment-path [ -r ] [ @var{pathdir} ]+
32487 Add directories @var{pathdir} to beginning of search path for object files.
32488 If the @samp{-r} option is used, the search path is reset to the original
32489 search path that existed at gdb start-up. If directories @var{pathdir} are
32490 supplied in addition to the
32491 @samp{-r} option, the search path is first reset and then addition
32493 Multiple directories may be specified, separated by blanks. Specifying
32494 multiple directories in a single command
32495 results in the directories added to the beginning of the
32496 search path in the same order they were presented in the command.
32497 If blanks are needed as
32498 part of a directory name, double-quotes should be used around
32499 the name. In the command output, the path will show up separated
32500 by the system directory-separator character. The directory-separator
32501 character must not be used
32502 in any directory name.
32503 If no directories are specified, the current path is displayed.
32506 @subsubheading @value{GDBN} Command
32508 The corresponding @value{GDBN} command is @samp{path}.
32510 @subsubheading Example
32515 ^done,path="/usr/bin"
32517 -environment-path /kwikemart/marge/ezannoni/flathead-dev/ppc-eabi/gdb /bin
32518 ^done,path="/kwikemart/marge/ezannoni/flathead-dev/ppc-eabi/gdb:/bin:/usr/bin"
32520 -environment-path -r /usr/local/bin
32521 ^done,path="/usr/local/bin:/usr/bin"
32526 @subheading The @code{-environment-pwd} Command
32527 @findex -environment-pwd
32529 @subsubheading Synopsis
32535 Show the current working directory.
32537 @subsubheading @value{GDBN} Command
32539 The corresponding @value{GDBN} command is @samp{pwd}.
32541 @subsubheading Example
32546 ^done,cwd="/kwikemart/marge/ezannoni/flathead-dev/devo/gdb"
32550 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32551 @node GDB/MI Thread Commands
32552 @section @sc{gdb/mi} Thread Commands
32555 @subheading The @code{-thread-info} Command
32556 @findex -thread-info
32558 @subsubheading Synopsis
32561 -thread-info [ @var{thread-id} ]
32564 Reports information about either a specific thread, if the
32565 @var{thread-id} parameter is present, or about all threads.
32566 @var{thread-id} is the thread's global thread ID. When printing
32567 information about all threads, also reports the global ID of the
32570 @subsubheading @value{GDBN} Command
32572 The @samp{info thread} command prints the same information
32575 @subsubheading Result
32577 The result contains the following attributes:
32581 A list of threads. The format of the elements of the list is described in
32582 @ref{GDB/MI Thread Information}.
32584 @item current-thread-id
32585 The global id of the currently selected thread. This field is omitted if there
32586 is no selected thread (for example, when the selected inferior is not running,
32587 and therefore has no threads) or if a @var{thread-id} argument was passed to
32592 @subsubheading Example
32597 @{id="2",target-id="Thread 0xb7e14b90 (LWP 21257)",
32598 frame=@{level="0",addr="0xffffe410",func="__kernel_vsyscall",
32599 args=[]@},state="running"@},
32600 @{id="1",target-id="Thread 0xb7e156b0 (LWP 21254)",
32601 frame=@{level="0",addr="0x0804891f",func="foo",
32602 args=[@{name="i",value="10"@}],
32603 file="/tmp/a.c",fullname="/tmp/a.c",line="158",arch="i386:x86_64"@},
32604 state="running"@}],
32605 current-thread-id="1"
32609 @subheading The @code{-thread-list-ids} Command
32610 @findex -thread-list-ids
32612 @subsubheading Synopsis
32618 Produces a list of the currently known global @value{GDBN} thread ids.
32619 At the end of the list it also prints the total number of such
32622 This command is retained for historical reasons, the
32623 @code{-thread-info} command should be used instead.
32625 @subsubheading @value{GDBN} Command
32627 Part of @samp{info threads} supplies the same information.
32629 @subsubheading Example
32634 ^done,thread-ids=@{thread-id="3",thread-id="2",thread-id="1"@},
32635 current-thread-id="1",number-of-threads="3"
32640 @subheading The @code{-thread-select} Command
32641 @findex -thread-select
32643 @subsubheading Synopsis
32646 -thread-select @var{thread-id}
32649 Make thread with global thread number @var{thread-id} the current
32650 thread. It prints the number of the new current thread, and the
32651 topmost frame for that thread.
32653 This command is deprecated in favor of explicitly using the
32654 @samp{--thread} option to each command.
32656 @subsubheading @value{GDBN} Command
32658 The corresponding @value{GDBN} command is @samp{thread}.
32660 @subsubheading Example
32667 *stopped,reason="end-stepping-range",thread-id="2",line="187",
32668 file="../../../devo/gdb/testsuite/gdb.threads/linux-dp.c"
32672 thread-ids=@{thread-id="3",thread-id="2",thread-id="1"@},
32673 number-of-threads="3"
32676 ^done,new-thread-id="3",
32677 frame=@{level="0",func="vprintf",
32678 args=[@{name="format",value="0x8048e9c \"%*s%c %d %c\\n\""@},
32679 @{name="arg",value="0x2"@}],file="vprintf.c",line="31",arch="i386:x86_64"@}
32683 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32684 @node GDB/MI Ada Tasking Commands
32685 @section @sc{gdb/mi} Ada Tasking Commands
32687 @subheading The @code{-ada-task-info} Command
32688 @findex -ada-task-info
32690 @subsubheading Synopsis
32693 -ada-task-info [ @var{task-id} ]
32696 Reports information about either a specific Ada task, if the
32697 @var{task-id} parameter is present, or about all Ada tasks.
32699 @subsubheading @value{GDBN} Command
32701 The @samp{info tasks} command prints the same information
32702 about all Ada tasks (@pxref{Ada Tasks}).
32704 @subsubheading Result
32706 The result is a table of Ada tasks. The following columns are
32707 defined for each Ada task:
32711 This field exists only for the current thread. It has the value @samp{*}.
32714 The identifier that @value{GDBN} uses to refer to the Ada task.
32717 The identifier that the target uses to refer to the Ada task.
32720 The global thread identifier of the thread corresponding to the Ada
32723 This field should always exist, as Ada tasks are always implemented
32724 on top of a thread. But if @value{GDBN} cannot find this corresponding
32725 thread for any reason, the field is omitted.
32728 This field exists only when the task was created by another task.
32729 In this case, it provides the ID of the parent task.
32732 The base priority of the task.
32735 The current state of the task. For a detailed description of the
32736 possible states, see @ref{Ada Tasks}.
32739 The name of the task.
32743 @subsubheading Example
32747 ^done,tasks=@{nr_rows="3",nr_cols="8",
32748 hdr=[@{width="1",alignment="-1",col_name="current",colhdr=""@},
32749 @{width="3",alignment="1",col_name="id",colhdr="ID"@},
32750 @{width="9",alignment="1",col_name="task-id",colhdr="TID"@},
32751 @{width="4",alignment="1",col_name="thread-id",colhdr=""@},
32752 @{width="4",alignment="1",col_name="parent-id",colhdr="P-ID"@},
32753 @{width="3",alignment="1",col_name="priority",colhdr="Pri"@},
32754 @{width="22",alignment="-1",col_name="state",colhdr="State"@},
32755 @{width="1",alignment="2",col_name="name",colhdr="Name"@}],
32756 body=[@{current="*",id="1",task-id=" 644010",thread-id="1",priority="48",
32757 state="Child Termination Wait",name="main_task"@}]@}
32761 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32762 @node GDB/MI Program Execution
32763 @section @sc{gdb/mi} Program Execution
32765 These are the asynchronous commands which generate the out-of-band
32766 record @samp{*stopped}. Currently @value{GDBN} only really executes
32767 asynchronously with remote targets and this interaction is mimicked in
32770 @subheading The @code{-exec-continue} Command
32771 @findex -exec-continue
32773 @subsubheading Synopsis
32776 -exec-continue [--reverse] [--all|--thread-group N]
32779 Resumes the execution of the inferior program, which will continue
32780 to execute until it reaches a debugger stop event. If the
32781 @samp{--reverse} option is specified, execution resumes in reverse until
32782 it reaches a stop event. Stop events may include
32785 breakpoints or watchpoints
32787 signals or exceptions
32789 the end of the process (or its beginning under @samp{--reverse})
32791 the end or beginning of a replay log if one is being used.
32793 In all-stop mode (@pxref{All-Stop
32794 Mode}), may resume only one thread, or all threads, depending on the
32795 value of the @samp{scheduler-locking} variable. If @samp{--all} is
32796 specified, all threads (in all inferiors) will be resumed. The @samp{--all} option is
32797 ignored in all-stop mode. If the @samp{--thread-group} options is
32798 specified, then all threads in that thread group are resumed.
32800 @subsubheading @value{GDBN} Command
32802 The corresponding @value{GDBN} corresponding is @samp{continue}.
32804 @subsubheading Example
32811 *stopped,reason="breakpoint-hit",disp="keep",bkptno="2",frame=@{
32812 func="foo",args=[],file="hello.c",fullname="/home/foo/bar/hello.c",
32813 line="13",arch="i386:x86_64"@}
32818 @subheading The @code{-exec-finish} Command
32819 @findex -exec-finish
32821 @subsubheading Synopsis
32824 -exec-finish [--reverse]
32827 Resumes the execution of the inferior program until the current
32828 function is exited. Displays the results returned by the function.
32829 If the @samp{--reverse} option is specified, resumes the reverse
32830 execution of the inferior program until the point where current
32831 function was called.
32833 @subsubheading @value{GDBN} Command
32835 The corresponding @value{GDBN} command is @samp{finish}.
32837 @subsubheading Example
32839 Function returning @code{void}.
32846 *stopped,reason="function-finished",frame=@{func="main",args=[],
32847 file="hello.c",fullname="/home/foo/bar/hello.c",line="7",arch="i386:x86_64"@}
32851 Function returning other than @code{void}. The name of the internal
32852 @value{GDBN} variable storing the result is printed, together with the
32859 *stopped,reason="function-finished",frame=@{addr="0x000107b0",func="foo",
32860 args=[@{name="a",value="1"],@{name="b",value="9"@}@},
32861 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
32862 arch="i386:x86_64"@},
32863 gdb-result-var="$1",return-value="0"
32868 @subheading The @code{-exec-interrupt} Command
32869 @findex -exec-interrupt
32871 @subsubheading Synopsis
32874 -exec-interrupt [--all|--thread-group N]
32877 Interrupts the background execution of the target. Note how the token
32878 associated with the stop message is the one for the execution command
32879 that has been interrupted. The token for the interrupt itself only
32880 appears in the @samp{^done} output. If the user is trying to
32881 interrupt a non-running program, an error message will be printed.
32883 Note that when asynchronous execution is enabled, this command is
32884 asynchronous just like other execution commands. That is, first the
32885 @samp{^done} response will be printed, and the target stop will be
32886 reported after that using the @samp{*stopped} notification.
32888 In non-stop mode, only the context thread is interrupted by default.
32889 All threads (in all inferiors) will be interrupted if the
32890 @samp{--all} option is specified. If the @samp{--thread-group}
32891 option is specified, all threads in that group will be interrupted.
32893 @subsubheading @value{GDBN} Command
32895 The corresponding @value{GDBN} command is @samp{interrupt}.
32897 @subsubheading Example
32908 111*stopped,signal-name="SIGINT",signal-meaning="Interrupt",
32909 frame=@{addr="0x00010140",func="foo",args=[],file="try.c",
32910 fullname="/home/foo/bar/try.c",line="13",arch="i386:x86_64"@}
32915 ^error,msg="mi_cmd_exec_interrupt: Inferior not executing."
32919 @subheading The @code{-exec-jump} Command
32922 @subsubheading Synopsis
32925 -exec-jump @var{location}
32928 Resumes execution of the inferior program at the location specified by
32929 parameter. @xref{Specify Location}, for a description of the
32930 different forms of @var{location}.
32932 @subsubheading @value{GDBN} Command
32934 The corresponding @value{GDBN} command is @samp{jump}.
32936 @subsubheading Example
32939 -exec-jump foo.c:10
32940 *running,thread-id="all"
32945 @subheading The @code{-exec-next} Command
32948 @subsubheading Synopsis
32951 -exec-next [--reverse]
32954 Resumes execution of the inferior program, stopping when the beginning
32955 of the next source line is reached.
32957 If the @samp{--reverse} option is specified, resumes reverse execution
32958 of the inferior program, stopping at the beginning of the previous
32959 source line. If you issue this command on the first line of a
32960 function, it will take you back to the caller of that function, to the
32961 source line where the function was called.
32964 @subsubheading @value{GDBN} Command
32966 The corresponding @value{GDBN} command is @samp{next}.
32968 @subsubheading Example
32974 *stopped,reason="end-stepping-range",line="8",file="hello.c"
32979 @subheading The @code{-exec-next-instruction} Command
32980 @findex -exec-next-instruction
32982 @subsubheading Synopsis
32985 -exec-next-instruction [--reverse]
32988 Executes one machine instruction. If the instruction is a function
32989 call, continues until the function returns. If the program stops at an
32990 instruction in the middle of a source line, the address will be
32993 If the @samp{--reverse} option is specified, resumes reverse execution
32994 of the inferior program, stopping at the previous instruction. If the
32995 previously executed instruction was a return from another function,
32996 it will continue to execute in reverse until the call to that function
32997 (from the current stack frame) is reached.
32999 @subsubheading @value{GDBN} Command
33001 The corresponding @value{GDBN} command is @samp{nexti}.
33003 @subsubheading Example
33007 -exec-next-instruction
33011 *stopped,reason="end-stepping-range",
33012 addr="0x000100d4",line="5",file="hello.c"
33017 @subheading The @code{-exec-return} Command
33018 @findex -exec-return
33020 @subsubheading Synopsis
33026 Makes current function return immediately. Doesn't execute the inferior.
33027 Displays the new current frame.
33029 @subsubheading @value{GDBN} Command
33031 The corresponding @value{GDBN} command is @samp{return}.
33033 @subsubheading Example
33037 200-break-insert callee4
33038 200^done,bkpt=@{number="1",addr="0x00010734",
33039 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",line="8"@}
33044 000*stopped,reason="breakpoint-hit",disp="keep",bkptno="1",
33045 frame=@{func="callee4",args=[],
33046 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33047 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="8",
33048 arch="i386:x86_64"@}
33054 111^done,frame=@{level="0",func="callee3",
33055 args=[@{name="strarg",
33056 value="0x11940 \"A string argument.\""@}],
33057 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33058 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="18",
33059 arch="i386:x86_64"@}
33064 @subheading The @code{-exec-run} Command
33067 @subsubheading Synopsis
33070 -exec-run [ --all | --thread-group N ] [ --start ]
33073 Starts execution of the inferior from the beginning. The inferior
33074 executes until either a breakpoint is encountered or the program
33075 exits. In the latter case the output will include an exit code, if
33076 the program has exited exceptionally.
33078 When neither the @samp{--all} nor the @samp{--thread-group} option
33079 is specified, the current inferior is started. If the
33080 @samp{--thread-group} option is specified, it should refer to a thread
33081 group of type @samp{process}, and that thread group will be started.
33082 If the @samp{--all} option is specified, then all inferiors will be started.
33084 Using the @samp{--start} option instructs the debugger to stop
33085 the execution at the start of the inferior's main subprogram,
33086 following the same behavior as the @code{start} command
33087 (@pxref{Starting}).
33089 @subsubheading @value{GDBN} Command
33091 The corresponding @value{GDBN} command is @samp{run}.
33093 @subsubheading Examples
33098 ^done,bkpt=@{number="1",addr="0x0001072c",file="recursive2.c",line="4"@}
33103 *stopped,reason="breakpoint-hit",disp="keep",bkptno="1",
33104 frame=@{func="main",args=[],file="recursive2.c",
33105 fullname="/home/foo/bar/recursive2.c",line="4",arch="i386:x86_64"@}
33110 Program exited normally:
33118 *stopped,reason="exited-normally"
33123 Program exited exceptionally:
33131 *stopped,reason="exited",exit-code="01"
33135 Another way the program can terminate is if it receives a signal such as
33136 @code{SIGINT}. In this case, @sc{gdb/mi} displays this:
33140 *stopped,reason="exited-signalled",signal-name="SIGINT",
33141 signal-meaning="Interrupt"
33145 @c @subheading -exec-signal
33148 @subheading The @code{-exec-step} Command
33151 @subsubheading Synopsis
33154 -exec-step [--reverse]
33157 Resumes execution of the inferior program, stopping when the beginning
33158 of the next source line is reached, if the next source line is not a
33159 function call. If it is, stop at the first instruction of the called
33160 function. If the @samp{--reverse} option is specified, resumes reverse
33161 execution of the inferior program, stopping at the beginning of the
33162 previously executed source line.
33164 @subsubheading @value{GDBN} Command
33166 The corresponding @value{GDBN} command is @samp{step}.
33168 @subsubheading Example
33170 Stepping into a function:
33176 *stopped,reason="end-stepping-range",
33177 frame=@{func="foo",args=[@{name="a",value="10"@},
33178 @{name="b",value="0"@}],file="recursive2.c",
33179 fullname="/home/foo/bar/recursive2.c",line="11",arch="i386:x86_64"@}
33189 *stopped,reason="end-stepping-range",line="14",file="recursive2.c"
33194 @subheading The @code{-exec-step-instruction} Command
33195 @findex -exec-step-instruction
33197 @subsubheading Synopsis
33200 -exec-step-instruction [--reverse]
33203 Resumes the inferior which executes one machine instruction. If the
33204 @samp{--reverse} option is specified, resumes reverse execution of the
33205 inferior program, stopping at the previously executed instruction.
33206 The output, once @value{GDBN} has stopped, will vary depending on
33207 whether we have stopped in the middle of a source line or not. In the
33208 former case, the address at which the program stopped will be printed
33211 @subsubheading @value{GDBN} Command
33213 The corresponding @value{GDBN} command is @samp{stepi}.
33215 @subsubheading Example
33219 -exec-step-instruction
33223 *stopped,reason="end-stepping-range",
33224 frame=@{func="foo",args=[],file="try.c",
33225 fullname="/home/foo/bar/try.c",line="10",arch="i386:x86_64"@}
33227 -exec-step-instruction
33231 *stopped,reason="end-stepping-range",
33232 frame=@{addr="0x000100f4",func="foo",args=[],file="try.c",
33233 fullname="/home/foo/bar/try.c",line="10",arch="i386:x86_64"@}
33238 @subheading The @code{-exec-until} Command
33239 @findex -exec-until
33241 @subsubheading Synopsis
33244 -exec-until [ @var{location} ]
33247 Executes the inferior until the @var{location} specified in the
33248 argument is reached. If there is no argument, the inferior executes
33249 until a source line greater than the current one is reached. The
33250 reason for stopping in this case will be @samp{location-reached}.
33252 @subsubheading @value{GDBN} Command
33254 The corresponding @value{GDBN} command is @samp{until}.
33256 @subsubheading Example
33260 -exec-until recursive2.c:6
33264 *stopped,reason="location-reached",frame=@{func="main",args=[],
33265 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="6",
33266 arch="i386:x86_64"@}
33271 @subheading -file-clear
33272 Is this going away????
33275 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
33276 @node GDB/MI Stack Manipulation
33277 @section @sc{gdb/mi} Stack Manipulation Commands
33279 @subheading The @code{-enable-frame-filters} Command
33280 @findex -enable-frame-filters
33283 -enable-frame-filters
33286 @value{GDBN} allows Python-based frame filters to affect the output of
33287 the MI commands relating to stack traces. As there is no way to
33288 implement this in a fully backward-compatible way, a front end must
33289 request that this functionality be enabled.
33291 Once enabled, this feature cannot be disabled.
33293 Note that if Python support has not been compiled into @value{GDBN},
33294 this command will still succeed (and do nothing).
33296 @subheading The @code{-stack-info-frame} Command
33297 @findex -stack-info-frame
33299 @subsubheading Synopsis
33305 Get info on the selected frame.
33307 @subsubheading @value{GDBN} Command
33309 The corresponding @value{GDBN} command is @samp{info frame} or @samp{frame}
33310 (without arguments).
33312 @subsubheading Example
33317 ^done,frame=@{level="1",addr="0x0001076c",func="callee3",
33318 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33319 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="17",
33320 arch="i386:x86_64"@}
33324 @subheading The @code{-stack-info-depth} Command
33325 @findex -stack-info-depth
33327 @subsubheading Synopsis
33330 -stack-info-depth [ @var{max-depth} ]
33333 Return the depth of the stack. If the integer argument @var{max-depth}
33334 is specified, do not count beyond @var{max-depth} frames.
33336 @subsubheading @value{GDBN} Command
33338 There's no equivalent @value{GDBN} command.
33340 @subsubheading Example
33342 For a stack with frame levels 0 through 11:
33349 -stack-info-depth 4
33352 -stack-info-depth 12
33355 -stack-info-depth 11
33358 -stack-info-depth 13
33363 @anchor{-stack-list-arguments}
33364 @subheading The @code{-stack-list-arguments} Command
33365 @findex -stack-list-arguments
33367 @subsubheading Synopsis
33370 -stack-list-arguments [ --no-frame-filters ] [ --skip-unavailable ] @var{print-values}
33371 [ @var{low-frame} @var{high-frame} ]
33374 Display a list of the arguments for the frames between @var{low-frame}
33375 and @var{high-frame} (inclusive). If @var{low-frame} and
33376 @var{high-frame} are not provided, list the arguments for the whole
33377 call stack. If the two arguments are equal, show the single frame
33378 at the corresponding level. It is an error if @var{low-frame} is
33379 larger than the actual number of frames. On the other hand,
33380 @var{high-frame} may be larger than the actual number of frames, in
33381 which case only existing frames will be returned.
33383 If @var{print-values} is 0 or @code{--no-values}, print only the names of
33384 the variables; if it is 1 or @code{--all-values}, print also their
33385 values; and if it is 2 or @code{--simple-values}, print the name,
33386 type and value for simple data types, and the name and type for arrays,
33387 structures and unions. If the option @code{--no-frame-filters} is
33388 supplied, then Python frame filters will not be executed.
33390 If the @code{--skip-unavailable} option is specified, arguments that
33391 are not available are not listed. Partially available arguments
33392 are still displayed, however.
33394 Use of this command to obtain arguments in a single frame is
33395 deprecated in favor of the @samp{-stack-list-variables} command.
33397 @subsubheading @value{GDBN} Command
33399 @value{GDBN} does not have an equivalent command. @code{gdbtk} has a
33400 @samp{gdb_get_args} command which partially overlaps with the
33401 functionality of @samp{-stack-list-arguments}.
33403 @subsubheading Example
33410 frame=@{level="0",addr="0x00010734",func="callee4",
33411 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33412 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="8",
33413 arch="i386:x86_64"@},
33414 frame=@{level="1",addr="0x0001076c",func="callee3",
33415 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33416 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="17",
33417 arch="i386:x86_64"@},
33418 frame=@{level="2",addr="0x0001078c",func="callee2",
33419 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33420 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="22",
33421 arch="i386:x86_64"@},
33422 frame=@{level="3",addr="0x000107b4",func="callee1",
33423 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33424 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="27",
33425 arch="i386:x86_64"@},
33426 frame=@{level="4",addr="0x000107e0",func="main",
33427 file="../../../devo/gdb/testsuite/gdb.mi/basics.c",
33428 fullname="/home/foo/bar/devo/gdb/testsuite/gdb.mi/basics.c",line="32",
33429 arch="i386:x86_64"@}]
33431 -stack-list-arguments 0
33434 frame=@{level="0",args=[]@},
33435 frame=@{level="1",args=[name="strarg"]@},
33436 frame=@{level="2",args=[name="intarg",name="strarg"]@},
33437 frame=@{level="3",args=[name="intarg",name="strarg",name="fltarg"]@},
33438 frame=@{level="4",args=[]@}]
33440 -stack-list-arguments 1
33443 frame=@{level="0",args=[]@},
33445 args=[@{name="strarg",value="0x11940 \"A string argument.\""@}]@},
33446 frame=@{level="2",args=[
33447 @{name="intarg",value="2"@},
33448 @{name="strarg",value="0x11940 \"A string argument.\""@}]@},
33449 @{frame=@{level="3",args=[
33450 @{name="intarg",value="2"@},
33451 @{name="strarg",value="0x11940 \"A string argument.\""@},
33452 @{name="fltarg",value="3.5"@}]@},
33453 frame=@{level="4",args=[]@}]
33455 -stack-list-arguments 0 2 2
33456 ^done,stack-args=[frame=@{level="2",args=[name="intarg",name="strarg"]@}]
33458 -stack-list-arguments 1 2 2
33459 ^done,stack-args=[frame=@{level="2",
33460 args=[@{name="intarg",value="2"@},
33461 @{name="strarg",value="0x11940 \"A string argument.\""@}]@}]
33465 @c @subheading -stack-list-exception-handlers
33468 @anchor{-stack-list-frames}
33469 @subheading The @code{-stack-list-frames} Command
33470 @findex -stack-list-frames
33472 @subsubheading Synopsis
33475 -stack-list-frames [ --no-frame-filters @var{low-frame} @var{high-frame} ]
33478 List the frames currently on the stack. For each frame it displays the
33483 The frame number, 0 being the topmost frame, i.e., the innermost function.
33485 The @code{$pc} value for that frame.
33489 File name of the source file where the function lives.
33490 @item @var{fullname}
33491 The full file name of the source file where the function lives.
33493 Line number corresponding to the @code{$pc}.
33495 The shared library where this function is defined. This is only given
33496 if the frame's function is not known.
33498 Frame's architecture.
33501 If invoked without arguments, this command prints a backtrace for the
33502 whole stack. If given two integer arguments, it shows the frames whose
33503 levels are between the two arguments (inclusive). If the two arguments
33504 are equal, it shows the single frame at the corresponding level. It is
33505 an error if @var{low-frame} is larger than the actual number of
33506 frames. On the other hand, @var{high-frame} may be larger than the
33507 actual number of frames, in which case only existing frames will be
33508 returned. If the option @code{--no-frame-filters} is supplied, then
33509 Python frame filters will not be executed.
33511 @subsubheading @value{GDBN} Command
33513 The corresponding @value{GDBN} commands are @samp{backtrace} and @samp{where}.
33515 @subsubheading Example
33517 Full stack backtrace:
33523 [frame=@{level="0",addr="0x0001076c",func="foo",
33524 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="11",
33525 arch="i386:x86_64"@},
33526 frame=@{level="1",addr="0x000107a4",func="foo",
33527 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33528 arch="i386:x86_64"@},
33529 frame=@{level="2",addr="0x000107a4",func="foo",
33530 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33531 arch="i386:x86_64"@},
33532 frame=@{level="3",addr="0x000107a4",func="foo",
33533 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33534 arch="i386:x86_64"@},
33535 frame=@{level="4",addr="0x000107a4",func="foo",
33536 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33537 arch="i386:x86_64"@},
33538 frame=@{level="5",addr="0x000107a4",func="foo",
33539 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33540 arch="i386:x86_64"@},
33541 frame=@{level="6",addr="0x000107a4",func="foo",
33542 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33543 arch="i386:x86_64"@},
33544 frame=@{level="7",addr="0x000107a4",func="foo",
33545 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33546 arch="i386:x86_64"@},
33547 frame=@{level="8",addr="0x000107a4",func="foo",
33548 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33549 arch="i386:x86_64"@},
33550 frame=@{level="9",addr="0x000107a4",func="foo",
33551 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33552 arch="i386:x86_64"@},
33553 frame=@{level="10",addr="0x000107a4",func="foo",
33554 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33555 arch="i386:x86_64"@},
33556 frame=@{level="11",addr="0x00010738",func="main",
33557 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="4",
33558 arch="i386:x86_64"@}]
33562 Show frames between @var{low_frame} and @var{high_frame}:
33566 -stack-list-frames 3 5
33568 [frame=@{level="3",addr="0x000107a4",func="foo",
33569 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33570 arch="i386:x86_64"@},
33571 frame=@{level="4",addr="0x000107a4",func="foo",
33572 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33573 arch="i386:x86_64"@},
33574 frame=@{level="5",addr="0x000107a4",func="foo",
33575 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33576 arch="i386:x86_64"@}]
33580 Show a single frame:
33584 -stack-list-frames 3 3
33586 [frame=@{level="3",addr="0x000107a4",func="foo",
33587 file="recursive2.c",fullname="/home/foo/bar/recursive2.c",line="14",
33588 arch="i386:x86_64"@}]
33593 @subheading The @code{-stack-list-locals} Command
33594 @findex -stack-list-locals
33595 @anchor{-stack-list-locals}
33597 @subsubheading Synopsis
33600 -stack-list-locals [ --no-frame-filters ] [ --skip-unavailable ] @var{print-values}
33603 Display the local variable names for the selected frame. If
33604 @var{print-values} is 0 or @code{--no-values}, print only the names of
33605 the variables; if it is 1 or @code{--all-values}, print also their
33606 values; and if it is 2 or @code{--simple-values}, print the name,
33607 type and value for simple data types, and the name and type for arrays,
33608 structures and unions. In this last case, a frontend can immediately
33609 display the value of simple data types and create variable objects for
33610 other data types when the user wishes to explore their values in
33611 more detail. If the option @code{--no-frame-filters} is supplied, then
33612 Python frame filters will not be executed.
33614 If the @code{--skip-unavailable} option is specified, local variables
33615 that are not available are not listed. Partially available local
33616 variables are still displayed, however.
33618 This command is deprecated in favor of the
33619 @samp{-stack-list-variables} command.
33621 @subsubheading @value{GDBN} Command
33623 @samp{info locals} in @value{GDBN}, @samp{gdb_get_locals} in @code{gdbtk}.
33625 @subsubheading Example
33629 -stack-list-locals 0
33630 ^done,locals=[name="A",name="B",name="C"]
33632 -stack-list-locals --all-values
33633 ^done,locals=[@{name="A",value="1"@},@{name="B",value="2"@},
33634 @{name="C",value="@{1, 2, 3@}"@}]
33635 -stack-list-locals --simple-values
33636 ^done,locals=[@{name="A",type="int",value="1"@},
33637 @{name="B",type="int",value="2"@},@{name="C",type="int [3]"@}]
33641 @anchor{-stack-list-variables}
33642 @subheading The @code{-stack-list-variables} Command
33643 @findex -stack-list-variables
33645 @subsubheading Synopsis
33648 -stack-list-variables [ --no-frame-filters ] [ --skip-unavailable ] @var{print-values}
33651 Display the names of local variables and function arguments for the selected frame. If
33652 @var{print-values} is 0 or @code{--no-values}, print only the names of
33653 the variables; if it is 1 or @code{--all-values}, print also their
33654 values; and if it is 2 or @code{--simple-values}, print the name,
33655 type and value for simple data types, and the name and type for arrays,
33656 structures and unions. If the option @code{--no-frame-filters} is
33657 supplied, then Python frame filters will not be executed.
33659 If the @code{--skip-unavailable} option is specified, local variables
33660 and arguments that are not available are not listed. Partially
33661 available arguments and local variables are still displayed, however.
33663 @subsubheading Example
33667 -stack-list-variables --thread 1 --frame 0 --all-values
33668 ^done,variables=[@{name="x",value="11"@},@{name="s",value="@{a = 1, b = 2@}"@}]
33673 @subheading The @code{-stack-select-frame} Command
33674 @findex -stack-select-frame
33676 @subsubheading Synopsis
33679 -stack-select-frame @var{framenum}
33682 Change the selected frame. Select a different frame @var{framenum} on
33685 This command in deprecated in favor of passing the @samp{--frame}
33686 option to every command.
33688 @subsubheading @value{GDBN} Command
33690 The corresponding @value{GDBN} commands are @samp{frame}, @samp{up},
33691 @samp{down}, @samp{select-frame}, @samp{up-silent}, and @samp{down-silent}.
33693 @subsubheading Example
33697 -stack-select-frame 2
33702 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
33703 @node GDB/MI Variable Objects
33704 @section @sc{gdb/mi} Variable Objects
33708 @subheading Motivation for Variable Objects in @sc{gdb/mi}
33710 For the implementation of a variable debugger window (locals, watched
33711 expressions, etc.), we are proposing the adaptation of the existing code
33712 used by @code{Insight}.
33714 The two main reasons for that are:
33718 It has been proven in practice (it is already on its second generation).
33721 It will shorten development time (needless to say how important it is
33725 The original interface was designed to be used by Tcl code, so it was
33726 slightly changed so it could be used through @sc{gdb/mi}. This section
33727 describes the @sc{gdb/mi} operations that will be available and gives some
33728 hints about their use.
33730 @emph{Note}: In addition to the set of operations described here, we
33731 expect the @sc{gui} implementation of a variable window to require, at
33732 least, the following operations:
33735 @item @code{-gdb-show} @code{output-radix}
33736 @item @code{-stack-list-arguments}
33737 @item @code{-stack-list-locals}
33738 @item @code{-stack-select-frame}
33743 @subheading Introduction to Variable Objects
33745 @cindex variable objects in @sc{gdb/mi}
33747 Variable objects are "object-oriented" MI interface for examining and
33748 changing values of expressions. Unlike some other MI interfaces that
33749 work with expressions, variable objects are specifically designed for
33750 simple and efficient presentation in the frontend. A variable object
33751 is identified by string name. When a variable object is created, the
33752 frontend specifies the expression for that variable object. The
33753 expression can be a simple variable, or it can be an arbitrary complex
33754 expression, and can even involve CPU registers. After creating a
33755 variable object, the frontend can invoke other variable object
33756 operations---for example to obtain or change the value of a variable
33757 object, or to change display format.
33759 Variable objects have hierarchical tree structure. Any variable object
33760 that corresponds to a composite type, such as structure in C, has
33761 a number of child variable objects, for example corresponding to each
33762 element of a structure. A child variable object can itself have
33763 children, recursively. Recursion ends when we reach
33764 leaf variable objects, which always have built-in types. Child variable
33765 objects are created only by explicit request, so if a frontend
33766 is not interested in the children of a particular variable object, no
33767 child will be created.
33769 For a leaf variable object it is possible to obtain its value as a
33770 string, or set the value from a string. String value can be also
33771 obtained for a non-leaf variable object, but it's generally a string
33772 that only indicates the type of the object, and does not list its
33773 contents. Assignment to a non-leaf variable object is not allowed.
33775 A frontend does not need to read the values of all variable objects each time
33776 the program stops. Instead, MI provides an update command that lists all
33777 variable objects whose values has changed since the last update
33778 operation. This considerably reduces the amount of data that must
33779 be transferred to the frontend. As noted above, children variable
33780 objects are created on demand, and only leaf variable objects have a
33781 real value. As result, gdb will read target memory only for leaf
33782 variables that frontend has created.
33784 The automatic update is not always desirable. For example, a frontend
33785 might want to keep a value of some expression for future reference,
33786 and never update it. For another example, fetching memory is
33787 relatively slow for embedded targets, so a frontend might want
33788 to disable automatic update for the variables that are either not
33789 visible on the screen, or ``closed''. This is possible using so
33790 called ``frozen variable objects''. Such variable objects are never
33791 implicitly updated.
33793 Variable objects can be either @dfn{fixed} or @dfn{floating}. For the
33794 fixed variable object, the expression is parsed when the variable
33795 object is created, including associating identifiers to specific
33796 variables. The meaning of expression never changes. For a floating
33797 variable object the values of variables whose names appear in the
33798 expressions are re-evaluated every time in the context of the current
33799 frame. Consider this example:
33804 struct work_state state;
33811 If a fixed variable object for the @code{state} variable is created in
33812 this function, and we enter the recursive call, the variable
33813 object will report the value of @code{state} in the top-level
33814 @code{do_work} invocation. On the other hand, a floating variable
33815 object will report the value of @code{state} in the current frame.
33817 If an expression specified when creating a fixed variable object
33818 refers to a local variable, the variable object becomes bound to the
33819 thread and frame in which the variable object is created. When such
33820 variable object is updated, @value{GDBN} makes sure that the
33821 thread/frame combination the variable object is bound to still exists,
33822 and re-evaluates the variable object in context of that thread/frame.
33824 The following is the complete set of @sc{gdb/mi} operations defined to
33825 access this functionality:
33827 @multitable @columnfractions .4 .6
33828 @item @strong{Operation}
33829 @tab @strong{Description}
33831 @item @code{-enable-pretty-printing}
33832 @tab enable Python-based pretty-printing
33833 @item @code{-var-create}
33834 @tab create a variable object
33835 @item @code{-var-delete}
33836 @tab delete the variable object and/or its children
33837 @item @code{-var-set-format}
33838 @tab set the display format of this variable
33839 @item @code{-var-show-format}
33840 @tab show the display format of this variable
33841 @item @code{-var-info-num-children}
33842 @tab tells how many children this object has
33843 @item @code{-var-list-children}
33844 @tab return a list of the object's children
33845 @item @code{-var-info-type}
33846 @tab show the type of this variable object
33847 @item @code{-var-info-expression}
33848 @tab print parent-relative expression that this variable object represents
33849 @item @code{-var-info-path-expression}
33850 @tab print full expression that this variable object represents
33851 @item @code{-var-show-attributes}
33852 @tab is this variable editable? does it exist here?
33853 @item @code{-var-evaluate-expression}
33854 @tab get the value of this variable
33855 @item @code{-var-assign}
33856 @tab set the value of this variable
33857 @item @code{-var-update}
33858 @tab update the variable and its children
33859 @item @code{-var-set-frozen}
33860 @tab set frozenness attribute
33861 @item @code{-var-set-update-range}
33862 @tab set range of children to display on update
33865 In the next subsection we describe each operation in detail and suggest
33866 how it can be used.
33868 @subheading Description And Use of Operations on Variable Objects
33870 @subheading The @code{-enable-pretty-printing} Command
33871 @findex -enable-pretty-printing
33874 -enable-pretty-printing
33877 @value{GDBN} allows Python-based visualizers to affect the output of the
33878 MI variable object commands. However, because there was no way to
33879 implement this in a fully backward-compatible way, a front end must
33880 request that this functionality be enabled.
33882 Once enabled, this feature cannot be disabled.
33884 Note that if Python support has not been compiled into @value{GDBN},
33885 this command will still succeed (and do nothing).
33887 This feature is currently (as of @value{GDBN} 7.0) experimental, and
33888 may work differently in future versions of @value{GDBN}.
33890 @subheading The @code{-var-create} Command
33891 @findex -var-create
33893 @subsubheading Synopsis
33896 -var-create @{@var{name} | "-"@}
33897 @{@var{frame-addr} | "*" | "@@"@} @var{expression}
33900 This operation creates a variable object, which allows the monitoring of
33901 a variable, the result of an expression, a memory cell or a CPU
33904 The @var{name} parameter is the string by which the object can be
33905 referenced. It must be unique. If @samp{-} is specified, the varobj
33906 system will generate a string ``varNNNNNN'' automatically. It will be
33907 unique provided that one does not specify @var{name} of that format.
33908 The command fails if a duplicate name is found.
33910 The frame under which the expression should be evaluated can be
33911 specified by @var{frame-addr}. A @samp{*} indicates that the current
33912 frame should be used. A @samp{@@} indicates that a floating variable
33913 object must be created.
33915 @var{expression} is any expression valid on the current language set (must not
33916 begin with a @samp{*}), or one of the following:
33920 @samp{*@var{addr}}, where @var{addr} is the address of a memory cell
33923 @samp{*@var{addr}-@var{addr}} --- a memory address range (TBD)
33926 @samp{$@var{regname}} --- a CPU register name
33929 @cindex dynamic varobj
33930 A varobj's contents may be provided by a Python-based pretty-printer. In this
33931 case the varobj is known as a @dfn{dynamic varobj}. Dynamic varobjs
33932 have slightly different semantics in some cases. If the
33933 @code{-enable-pretty-printing} command is not sent, then @value{GDBN}
33934 will never create a dynamic varobj. This ensures backward
33935 compatibility for existing clients.
33937 @subsubheading Result
33939 This operation returns attributes of the newly-created varobj. These
33944 The name of the varobj.
33947 The number of children of the varobj. This number is not necessarily
33948 reliable for a dynamic varobj. Instead, you must examine the
33949 @samp{has_more} attribute.
33952 The varobj's scalar value. For a varobj whose type is some sort of
33953 aggregate (e.g., a @code{struct}), or for a dynamic varobj, this value
33954 will not be interesting.
33957 The varobj's type. This is a string representation of the type, as
33958 would be printed by the @value{GDBN} CLI. If @samp{print object}
33959 (@pxref{Print Settings, set print object}) is set to @code{on}, the
33960 @emph{actual} (derived) type of the object is shown rather than the
33961 @emph{declared} one.
33964 If a variable object is bound to a specific thread, then this is the
33965 thread's global identifier.
33968 For a dynamic varobj, this indicates whether there appear to be any
33969 children available. For a non-dynamic varobj, this will be 0.
33972 This attribute will be present and have the value @samp{1} if the
33973 varobj is a dynamic varobj. If the varobj is not a dynamic varobj,
33974 then this attribute will not be present.
33977 A dynamic varobj can supply a display hint to the front end. The
33978 value comes directly from the Python pretty-printer object's
33979 @code{display_hint} method. @xref{Pretty Printing API}.
33982 Typical output will look like this:
33985 name="@var{name}",numchild="@var{N}",type="@var{type}",thread-id="@var{M}",
33986 has_more="@var{has_more}"
33990 @subheading The @code{-var-delete} Command
33991 @findex -var-delete
33993 @subsubheading Synopsis
33996 -var-delete [ -c ] @var{name}
33999 Deletes a previously created variable object and all of its children.
34000 With the @samp{-c} option, just deletes the children.
34002 Returns an error if the object @var{name} is not found.
34005 @subheading The @code{-var-set-format} Command
34006 @findex -var-set-format
34008 @subsubheading Synopsis
34011 -var-set-format @var{name} @var{format-spec}
34014 Sets the output format for the value of the object @var{name} to be
34017 @anchor{-var-set-format}
34018 The syntax for the @var{format-spec} is as follows:
34021 @var{format-spec} @expansion{}
34022 @{binary | decimal | hexadecimal | octal | natural | zero-hexadecimal@}
34025 The natural format is the default format choosen automatically
34026 based on the variable type (like decimal for an @code{int}, hex
34027 for pointers, etc.).
34029 The zero-hexadecimal format has a representation similar to hexadecimal
34030 but with padding zeroes to the left of the value. For example, a 32-bit
34031 hexadecimal value of 0x1234 would be represented as 0x00001234 in the
34032 zero-hexadecimal format.
34034 For a variable with children, the format is set only on the
34035 variable itself, and the children are not affected.
34037 @subheading The @code{-var-show-format} Command
34038 @findex -var-show-format
34040 @subsubheading Synopsis
34043 -var-show-format @var{name}
34046 Returns the format used to display the value of the object @var{name}.
34049 @var{format} @expansion{}
34054 @subheading The @code{-var-info-num-children} Command
34055 @findex -var-info-num-children
34057 @subsubheading Synopsis
34060 -var-info-num-children @var{name}
34063 Returns the number of children of a variable object @var{name}:
34069 Note that this number is not completely reliable for a dynamic varobj.
34070 It will return the current number of children, but more children may
34074 @subheading The @code{-var-list-children} Command
34075 @findex -var-list-children
34077 @subsubheading Synopsis
34080 -var-list-children [@var{print-values}] @var{name} [@var{from} @var{to}]
34082 @anchor{-var-list-children}
34084 Return a list of the children of the specified variable object and
34085 create variable objects for them, if they do not already exist. With
34086 a single argument or if @var{print-values} has a value of 0 or
34087 @code{--no-values}, print only the names of the variables; if
34088 @var{print-values} is 1 or @code{--all-values}, also print their
34089 values; and if it is 2 or @code{--simple-values} print the name and
34090 value for simple data types and just the name for arrays, structures
34093 @var{from} and @var{to}, if specified, indicate the range of children
34094 to report. If @var{from} or @var{to} is less than zero, the range is
34095 reset and all children will be reported. Otherwise, children starting
34096 at @var{from} (zero-based) and up to and excluding @var{to} will be
34099 If a child range is requested, it will only affect the current call to
34100 @code{-var-list-children}, but not future calls to @code{-var-update}.
34101 For this, you must instead use @code{-var-set-update-range}. The
34102 intent of this approach is to enable a front end to implement any
34103 update approach it likes; for example, scrolling a view may cause the
34104 front end to request more children with @code{-var-list-children}, and
34105 then the front end could call @code{-var-set-update-range} with a
34106 different range to ensure that future updates are restricted to just
34109 For each child the following results are returned:
34114 Name of the variable object created for this child.
34117 The expression to be shown to the user by the front end to designate this child.
34118 For example this may be the name of a structure member.
34120 For a dynamic varobj, this value cannot be used to form an
34121 expression. There is no way to do this at all with a dynamic varobj.
34123 For C/C@t{++} structures there are several pseudo children returned to
34124 designate access qualifiers. For these pseudo children @var{exp} is
34125 @samp{public}, @samp{private}, or @samp{protected}. In this case the
34126 type and value are not present.
34128 A dynamic varobj will not report the access qualifying
34129 pseudo-children, regardless of the language. This information is not
34130 available at all with a dynamic varobj.
34133 Number of children this child has. For a dynamic varobj, this will be
34137 The type of the child. If @samp{print object}
34138 (@pxref{Print Settings, set print object}) is set to @code{on}, the
34139 @emph{actual} (derived) type of the object is shown rather than the
34140 @emph{declared} one.
34143 If values were requested, this is the value.
34146 If this variable object is associated with a thread, this is the
34147 thread's global thread id. Otherwise this result is not present.
34150 If the variable object is frozen, this variable will be present with a value of 1.
34153 A dynamic varobj can supply a display hint to the front end. The
34154 value comes directly from the Python pretty-printer object's
34155 @code{display_hint} method. @xref{Pretty Printing API}.
34158 This attribute will be present and have the value @samp{1} if the
34159 varobj is a dynamic varobj. If the varobj is not a dynamic varobj,
34160 then this attribute will not be present.
34164 The result may have its own attributes:
34168 A dynamic varobj can supply a display hint to the front end. The
34169 value comes directly from the Python pretty-printer object's
34170 @code{display_hint} method. @xref{Pretty Printing API}.
34173 This is an integer attribute which is nonzero if there are children
34174 remaining after the end of the selected range.
34177 @subsubheading Example
34181 -var-list-children n
34182 ^done,numchild=@var{n},children=[child=@{name=@var{name},exp=@var{exp},
34183 numchild=@var{n},type=@var{type}@},@r{(repeats N times)}]
34185 -var-list-children --all-values n
34186 ^done,numchild=@var{n},children=[child=@{name=@var{name},exp=@var{exp},
34187 numchild=@var{n},value=@var{value},type=@var{type}@},@r{(repeats N times)}]
34191 @subheading The @code{-var-info-type} Command
34192 @findex -var-info-type
34194 @subsubheading Synopsis
34197 -var-info-type @var{name}
34200 Returns the type of the specified variable @var{name}. The type is
34201 returned as a string in the same format as it is output by the
34205 type=@var{typename}
34209 @subheading The @code{-var-info-expression} Command
34210 @findex -var-info-expression
34212 @subsubheading Synopsis
34215 -var-info-expression @var{name}
34218 Returns a string that is suitable for presenting this
34219 variable object in user interface. The string is generally
34220 not valid expression in the current language, and cannot be evaluated.
34222 For example, if @code{a} is an array, and variable object
34223 @code{A} was created for @code{a}, then we'll get this output:
34226 (gdb) -var-info-expression A.1
34227 ^done,lang="C",exp="1"
34231 Here, the value of @code{lang} is the language name, which can be
34232 found in @ref{Supported Languages}.
34234 Note that the output of the @code{-var-list-children} command also
34235 includes those expressions, so the @code{-var-info-expression} command
34238 @subheading The @code{-var-info-path-expression} Command
34239 @findex -var-info-path-expression
34241 @subsubheading Synopsis
34244 -var-info-path-expression @var{name}
34247 Returns an expression that can be evaluated in the current
34248 context and will yield the same value that a variable object has.
34249 Compare this with the @code{-var-info-expression} command, which
34250 result can be used only for UI presentation. Typical use of
34251 the @code{-var-info-path-expression} command is creating a
34252 watchpoint from a variable object.
34254 This command is currently not valid for children of a dynamic varobj,
34255 and will give an error when invoked on one.
34257 For example, suppose @code{C} is a C@t{++} class, derived from class
34258 @code{Base}, and that the @code{Base} class has a member called
34259 @code{m_size}. Assume a variable @code{c} is has the type of
34260 @code{C} and a variable object @code{C} was created for variable
34261 @code{c}. Then, we'll get this output:
34263 (gdb) -var-info-path-expression C.Base.public.m_size
34264 ^done,path_expr=((Base)c).m_size)
34267 @subheading The @code{-var-show-attributes} Command
34268 @findex -var-show-attributes
34270 @subsubheading Synopsis
34273 -var-show-attributes @var{name}
34276 List attributes of the specified variable object @var{name}:
34279 status=@var{attr} [ ( ,@var{attr} )* ]
34283 where @var{attr} is @code{@{ @{ editable | noneditable @} | TBD @}}.
34285 @subheading The @code{-var-evaluate-expression} Command
34286 @findex -var-evaluate-expression
34288 @subsubheading Synopsis
34291 -var-evaluate-expression [-f @var{format-spec}] @var{name}
34294 Evaluates the expression that is represented by the specified variable
34295 object and returns its value as a string. The format of the string
34296 can be specified with the @samp{-f} option. The possible values of
34297 this option are the same as for @code{-var-set-format}
34298 (@pxref{-var-set-format}). If the @samp{-f} option is not specified,
34299 the current display format will be used. The current display format
34300 can be changed using the @code{-var-set-format} command.
34306 Note that one must invoke @code{-var-list-children} for a variable
34307 before the value of a child variable can be evaluated.
34309 @subheading The @code{-var-assign} Command
34310 @findex -var-assign
34312 @subsubheading Synopsis
34315 -var-assign @var{name} @var{expression}
34318 Assigns the value of @var{expression} to the variable object specified
34319 by @var{name}. The object must be @samp{editable}. If the variable's
34320 value is altered by the assign, the variable will show up in any
34321 subsequent @code{-var-update} list.
34323 @subsubheading Example
34331 ^done,changelist=[@{name="var1",in_scope="true",type_changed="false"@}]
34335 @subheading The @code{-var-update} Command
34336 @findex -var-update
34338 @subsubheading Synopsis
34341 -var-update [@var{print-values}] @{@var{name} | "*"@}
34344 Reevaluate the expressions corresponding to the variable object
34345 @var{name} and all its direct and indirect children, and return the
34346 list of variable objects whose values have changed; @var{name} must
34347 be a root variable object. Here, ``changed'' means that the result of
34348 @code{-var-evaluate-expression} before and after the
34349 @code{-var-update} is different. If @samp{*} is used as the variable
34350 object names, all existing variable objects are updated, except
34351 for frozen ones (@pxref{-var-set-frozen}). The option
34352 @var{print-values} determines whether both names and values, or just
34353 names are printed. The possible values of this option are the same
34354 as for @code{-var-list-children} (@pxref{-var-list-children}). It is
34355 recommended to use the @samp{--all-values} option, to reduce the
34356 number of MI commands needed on each program stop.
34358 With the @samp{*} parameter, if a variable object is bound to a
34359 currently running thread, it will not be updated, without any
34362 If @code{-var-set-update-range} was previously used on a varobj, then
34363 only the selected range of children will be reported.
34365 @code{-var-update} reports all the changed varobjs in a tuple named
34368 Each item in the change list is itself a tuple holding:
34372 The name of the varobj.
34375 If values were requested for this update, then this field will be
34376 present and will hold the value of the varobj.
34379 @anchor{-var-update}
34380 This field is a string which may take one of three values:
34384 The variable object's current value is valid.
34387 The variable object does not currently hold a valid value but it may
34388 hold one in the future if its associated expression comes back into
34392 The variable object no longer holds a valid value.
34393 This can occur when the executable file being debugged has changed,
34394 either through recompilation or by using the @value{GDBN} @code{file}
34395 command. The front end should normally choose to delete these variable
34399 In the future new values may be added to this list so the front should
34400 be prepared for this possibility. @xref{GDB/MI Development and Front Ends, ,@sc{GDB/MI} Development and Front Ends}.
34403 This is only present if the varobj is still valid. If the type
34404 changed, then this will be the string @samp{true}; otherwise it will
34407 When a varobj's type changes, its children are also likely to have
34408 become incorrect. Therefore, the varobj's children are automatically
34409 deleted when this attribute is @samp{true}. Also, the varobj's update
34410 range, when set using the @code{-var-set-update-range} command, is
34414 If the varobj's type changed, then this field will be present and will
34417 @item new_num_children
34418 For a dynamic varobj, if the number of children changed, or if the
34419 type changed, this will be the new number of children.
34421 The @samp{numchild} field in other varobj responses is generally not
34422 valid for a dynamic varobj -- it will show the number of children that
34423 @value{GDBN} knows about, but because dynamic varobjs lazily
34424 instantiate their children, this will not reflect the number of
34425 children which may be available.
34427 The @samp{new_num_children} attribute only reports changes to the
34428 number of children known by @value{GDBN}. This is the only way to
34429 detect whether an update has removed children (which necessarily can
34430 only happen at the end of the update range).
34433 The display hint, if any.
34436 This is an integer value, which will be 1 if there are more children
34437 available outside the varobj's update range.
34440 This attribute will be present and have the value @samp{1} if the
34441 varobj is a dynamic varobj. If the varobj is not a dynamic varobj,
34442 then this attribute will not be present.
34445 If new children were added to a dynamic varobj within the selected
34446 update range (as set by @code{-var-set-update-range}), then they will
34447 be listed in this attribute.
34450 @subsubheading Example
34457 -var-update --all-values var1
34458 ^done,changelist=[@{name="var1",value="3",in_scope="true",
34459 type_changed="false"@}]
34463 @subheading The @code{-var-set-frozen} Command
34464 @findex -var-set-frozen
34465 @anchor{-var-set-frozen}
34467 @subsubheading Synopsis
34470 -var-set-frozen @var{name} @var{flag}
34473 Set the frozenness flag on the variable object @var{name}. The
34474 @var{flag} parameter should be either @samp{1} to make the variable
34475 frozen or @samp{0} to make it unfrozen. If a variable object is
34476 frozen, then neither itself, nor any of its children, are
34477 implicitly updated by @code{-var-update} of
34478 a parent variable or by @code{-var-update *}. Only
34479 @code{-var-update} of the variable itself will update its value and
34480 values of its children. After a variable object is unfrozen, it is
34481 implicitly updated by all subsequent @code{-var-update} operations.
34482 Unfreezing a variable does not update it, only subsequent
34483 @code{-var-update} does.
34485 @subsubheading Example
34489 -var-set-frozen V 1
34494 @subheading The @code{-var-set-update-range} command
34495 @findex -var-set-update-range
34496 @anchor{-var-set-update-range}
34498 @subsubheading Synopsis
34501 -var-set-update-range @var{name} @var{from} @var{to}
34504 Set the range of children to be returned by future invocations of
34505 @code{-var-update}.
34507 @var{from} and @var{to} indicate the range of children to report. If
34508 @var{from} or @var{to} is less than zero, the range is reset and all
34509 children will be reported. Otherwise, children starting at @var{from}
34510 (zero-based) and up to and excluding @var{to} will be reported.
34512 @subsubheading Example
34516 -var-set-update-range V 1 2
34520 @subheading The @code{-var-set-visualizer} command
34521 @findex -var-set-visualizer
34522 @anchor{-var-set-visualizer}
34524 @subsubheading Synopsis
34527 -var-set-visualizer @var{name} @var{visualizer}
34530 Set a visualizer for the variable object @var{name}.
34532 @var{visualizer} is the visualizer to use. The special value
34533 @samp{None} means to disable any visualizer in use.
34535 If not @samp{None}, @var{visualizer} must be a Python expression.
34536 This expression must evaluate to a callable object which accepts a
34537 single argument. @value{GDBN} will call this object with the value of
34538 the varobj @var{name} as an argument (this is done so that the same
34539 Python pretty-printing code can be used for both the CLI and MI).
34540 When called, this object must return an object which conforms to the
34541 pretty-printing interface (@pxref{Pretty Printing API}).
34543 The pre-defined function @code{gdb.default_visualizer} may be used to
34544 select a visualizer by following the built-in process
34545 (@pxref{Selecting Pretty-Printers}). This is done automatically when
34546 a varobj is created, and so ordinarily is not needed.
34548 This feature is only available if Python support is enabled. The MI
34549 command @code{-list-features} (@pxref{GDB/MI Support Commands})
34550 can be used to check this.
34552 @subsubheading Example
34554 Resetting the visualizer:
34558 -var-set-visualizer V None
34562 Reselecting the default (type-based) visualizer:
34566 -var-set-visualizer V gdb.default_visualizer
34570 Suppose @code{SomeClass} is a visualizer class. A lambda expression
34571 can be used to instantiate this class for a varobj:
34575 -var-set-visualizer V "lambda val: SomeClass()"
34579 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34580 @node GDB/MI Data Manipulation
34581 @section @sc{gdb/mi} Data Manipulation
34583 @cindex data manipulation, in @sc{gdb/mi}
34584 @cindex @sc{gdb/mi}, data manipulation
34585 This section describes the @sc{gdb/mi} commands that manipulate data:
34586 examine memory and registers, evaluate expressions, etc.
34588 For details about what an addressable memory unit is,
34589 @pxref{addressable memory unit}.
34591 @c REMOVED FROM THE INTERFACE.
34592 @c @subheading -data-assign
34593 @c Change the value of a program variable. Plenty of side effects.
34594 @c @subsubheading GDB Command
34596 @c @subsubheading Example
34599 @subheading The @code{-data-disassemble} Command
34600 @findex -data-disassemble
34602 @subsubheading Synopsis
34606 [ -s @var{start-addr} -e @var{end-addr} ]
34607 | [ -a @var{addr} ]
34608 | [ -f @var{filename} -l @var{linenum} [ -n @var{lines} ] ]
34616 @item @var{start-addr}
34617 is the beginning address (or @code{$pc})
34618 @item @var{end-addr}
34621 is an address anywhere within (or the name of) the function to
34622 disassemble. If an address is specified, the whole function
34623 surrounding that address will be disassembled. If a name is
34624 specified, the whole function with that name will be disassembled.
34625 @item @var{filename}
34626 is the name of the file to disassemble
34627 @item @var{linenum}
34628 is the line number to disassemble around
34630 is the number of disassembly lines to be produced. If it is -1,
34631 the whole function will be disassembled, in case no @var{end-addr} is
34632 specified. If @var{end-addr} is specified as a non-zero value, and
34633 @var{lines} is lower than the number of disassembly lines between
34634 @var{start-addr} and @var{end-addr}, only @var{lines} lines are
34635 displayed; if @var{lines} is higher than the number of lines between
34636 @var{start-addr} and @var{end-addr}, only the lines up to @var{end-addr}
34641 @item 0 disassembly only
34642 @item 1 mixed source and disassembly (deprecated)
34643 @item 2 disassembly with raw opcodes
34644 @item 3 mixed source and disassembly with raw opcodes (deprecated)
34645 @item 4 mixed source and disassembly
34646 @item 5 mixed source and disassembly with raw opcodes
34649 Modes 1 and 3 are deprecated. The output is ``source centric''
34650 which hasn't proved useful in practice.
34651 @xref{Machine Code}, for a discussion of the difference between
34652 @code{/m} and @code{/s} output of the @code{disassemble} command.
34655 @subsubheading Result
34657 The result of the @code{-data-disassemble} command will be a list named
34658 @samp{asm_insns}, the contents of this list depend on the @var{mode}
34659 used with the @code{-data-disassemble} command.
34661 For modes 0 and 2 the @samp{asm_insns} list contains tuples with the
34666 The address at which this instruction was disassembled.
34669 The name of the function this instruction is within.
34672 The decimal offset in bytes from the start of @samp{func-name}.
34675 The text disassembly for this @samp{address}.
34678 This field is only present for modes 2, 3 and 5. This contains the raw opcode
34679 bytes for the @samp{inst} field.
34683 For modes 1, 3, 4 and 5 the @samp{asm_insns} list contains tuples named
34684 @samp{src_and_asm_line}, each of which has the following fields:
34688 The line number within @samp{file}.
34691 The file name from the compilation unit. This might be an absolute
34692 file name or a relative file name depending on the compile command
34696 Absolute file name of @samp{file}. It is converted to a canonical form
34697 using the source file search path
34698 (@pxref{Source Path, ,Specifying Source Directories})
34699 and after resolving all the symbolic links.
34701 If the source file is not found this field will contain the path as
34702 present in the debug information.
34704 @item line_asm_insn
34705 This is a list of tuples containing the disassembly for @samp{line} in
34706 @samp{file}. The fields of each tuple are the same as for
34707 @code{-data-disassemble} in @var{mode} 0 and 2, so @samp{address},
34708 @samp{func-name}, @samp{offset}, @samp{inst}, and optionally
34713 Note that whatever included in the @samp{inst} field, is not
34714 manipulated directly by @sc{gdb/mi}, i.e., it is not possible to
34717 @subsubheading @value{GDBN} Command
34719 The corresponding @value{GDBN} command is @samp{disassemble}.
34721 @subsubheading Example
34723 Disassemble from the current value of @code{$pc} to @code{$pc + 20}:
34727 -data-disassemble -s $pc -e "$pc + 20" -- 0
34730 @{address="0x000107c0",func-name="main",offset="4",
34731 inst="mov 2, %o0"@},
34732 @{address="0x000107c4",func-name="main",offset="8",
34733 inst="sethi %hi(0x11800), %o2"@},
34734 @{address="0x000107c8",func-name="main",offset="12",
34735 inst="or %o2, 0x140, %o1\t! 0x11940 <_lib_version+8>"@},
34736 @{address="0x000107cc",func-name="main",offset="16",
34737 inst="sethi %hi(0x11800), %o2"@},
34738 @{address="0x000107d0",func-name="main",offset="20",
34739 inst="or %o2, 0x168, %o4\t! 0x11968 <_lib_version+48>"@}]
34743 Disassemble the whole @code{main} function. Line 32 is part of
34747 -data-disassemble -f basics.c -l 32 -- 0
34749 @{address="0x000107bc",func-name="main",offset="0",
34750 inst="save %sp, -112, %sp"@},
34751 @{address="0x000107c0",func-name="main",offset="4",
34752 inst="mov 2, %o0"@},
34753 @{address="0x000107c4",func-name="main",offset="8",
34754 inst="sethi %hi(0x11800), %o2"@},
34756 @{address="0x0001081c",func-name="main",offset="96",inst="ret "@},
34757 @{address="0x00010820",func-name="main",offset="100",inst="restore "@}]
34761 Disassemble 3 instructions from the start of @code{main}:
34765 -data-disassemble -f basics.c -l 32 -n 3 -- 0
34767 @{address="0x000107bc",func-name="main",offset="0",
34768 inst="save %sp, -112, %sp"@},
34769 @{address="0x000107c0",func-name="main",offset="4",
34770 inst="mov 2, %o0"@},
34771 @{address="0x000107c4",func-name="main",offset="8",
34772 inst="sethi %hi(0x11800), %o2"@}]
34776 Disassemble 3 instructions from the start of @code{main} in mixed mode:
34780 -data-disassemble -f basics.c -l 32 -n 3 -- 1
34782 src_and_asm_line=@{line="31",
34783 file="../../../src/gdb/testsuite/gdb.mi/basics.c",
34784 fullname="/absolute/path/to/src/gdb/testsuite/gdb.mi/basics.c",
34785 line_asm_insn=[@{address="0x000107bc",
34786 func-name="main",offset="0",inst="save %sp, -112, %sp"@}]@},
34787 src_and_asm_line=@{line="32",
34788 file="../../../src/gdb/testsuite/gdb.mi/basics.c",
34789 fullname="/absolute/path/to/src/gdb/testsuite/gdb.mi/basics.c",
34790 line_asm_insn=[@{address="0x000107c0",
34791 func-name="main",offset="4",inst="mov 2, %o0"@},
34792 @{address="0x000107c4",func-name="main",offset="8",
34793 inst="sethi %hi(0x11800), %o2"@}]@}]
34798 @subheading The @code{-data-evaluate-expression} Command
34799 @findex -data-evaluate-expression
34801 @subsubheading Synopsis
34804 -data-evaluate-expression @var{expr}
34807 Evaluate @var{expr} as an expression. The expression could contain an
34808 inferior function call. The function call will execute synchronously.
34809 If the expression contains spaces, it must be enclosed in double quotes.
34811 @subsubheading @value{GDBN} Command
34813 The corresponding @value{GDBN} commands are @samp{print}, @samp{output}, and
34814 @samp{call}. In @code{gdbtk} only, there's a corresponding
34815 @samp{gdb_eval} command.
34817 @subsubheading Example
34819 In the following example, the numbers that precede the commands are the
34820 @dfn{tokens} described in @ref{GDB/MI Command Syntax, ,@sc{gdb/mi}
34821 Command Syntax}. Notice how @sc{gdb/mi} returns the same tokens in its
34825 211-data-evaluate-expression A
34828 311-data-evaluate-expression &A
34829 311^done,value="0xefffeb7c"
34831 411-data-evaluate-expression A+3
34834 511-data-evaluate-expression "A + 3"
34840 @subheading The @code{-data-list-changed-registers} Command
34841 @findex -data-list-changed-registers
34843 @subsubheading Synopsis
34846 -data-list-changed-registers
34849 Display a list of the registers that have changed.
34851 @subsubheading @value{GDBN} Command
34853 @value{GDBN} doesn't have a direct analog for this command; @code{gdbtk}
34854 has the corresponding command @samp{gdb_changed_register_list}.
34856 @subsubheading Example
34858 On a PPC MBX board:
34866 *stopped,reason="breakpoint-hit",disp="keep",bkptno="1",frame=@{
34867 func="main",args=[],file="try.c",fullname="/home/foo/bar/try.c",
34868 line="5",arch="powerpc"@}
34870 -data-list-changed-registers
34871 ^done,changed-registers=["0","1","2","4","5","6","7","8","9",
34872 "10","11","13","14","15","16","17","18","19","20","21","22","23",
34873 "24","25","26","27","28","30","31","64","65","66","67","69"]
34878 @subheading The @code{-data-list-register-names} Command
34879 @findex -data-list-register-names
34881 @subsubheading Synopsis
34884 -data-list-register-names [ ( @var{regno} )+ ]
34887 Show a list of register names for the current target. If no arguments
34888 are given, it shows a list of the names of all the registers. If
34889 integer numbers are given as arguments, it will print a list of the
34890 names of the registers corresponding to the arguments. To ensure
34891 consistency between a register name and its number, the output list may
34892 include empty register names.
34894 @subsubheading @value{GDBN} Command
34896 @value{GDBN} does not have a command which corresponds to
34897 @samp{-data-list-register-names}. In @code{gdbtk} there is a
34898 corresponding command @samp{gdb_regnames}.
34900 @subsubheading Example
34902 For the PPC MBX board:
34905 -data-list-register-names
34906 ^done,register-names=["r0","r1","r2","r3","r4","r5","r6","r7",
34907 "r8","r9","r10","r11","r12","r13","r14","r15","r16","r17","r18",
34908 "r19","r20","r21","r22","r23","r24","r25","r26","r27","r28","r29",
34909 "r30","r31","f0","f1","f2","f3","f4","f5","f6","f7","f8","f9",
34910 "f10","f11","f12","f13","f14","f15","f16","f17","f18","f19","f20",
34911 "f21","f22","f23","f24","f25","f26","f27","f28","f29","f30","f31",
34912 "", "pc","ps","cr","lr","ctr","xer"]
34914 -data-list-register-names 1 2 3
34915 ^done,register-names=["r1","r2","r3"]
34919 @subheading The @code{-data-list-register-values} Command
34920 @findex -data-list-register-values
34922 @subsubheading Synopsis
34925 -data-list-register-values
34926 [ @code{--skip-unavailable} ] @var{fmt} [ ( @var{regno} )*]
34929 Display the registers' contents. The format according to which the
34930 registers' contents are to be returned is given by @var{fmt}, followed
34931 by an optional list of numbers specifying the registers to display. A
34932 missing list of numbers indicates that the contents of all the
34933 registers must be returned. The @code{--skip-unavailable} option
34934 indicates that only the available registers are to be returned.
34936 Allowed formats for @var{fmt} are:
34953 @subsubheading @value{GDBN} Command
34955 The corresponding @value{GDBN} commands are @samp{info reg}, @samp{info
34956 all-reg}, and (in @code{gdbtk}) @samp{gdb_fetch_registers}.
34958 @subsubheading Example
34960 For a PPC MBX board (note: line breaks are for readability only, they
34961 don't appear in the actual output):
34965 -data-list-register-values r 64 65
34966 ^done,register-values=[@{number="64",value="0xfe00a300"@},
34967 @{number="65",value="0x00029002"@}]
34969 -data-list-register-values x
34970 ^done,register-values=[@{number="0",value="0xfe0043c8"@},
34971 @{number="1",value="0x3fff88"@},@{number="2",value="0xfffffffe"@},
34972 @{number="3",value="0x0"@},@{number="4",value="0xa"@},
34973 @{number="5",value="0x3fff68"@},@{number="6",value="0x3fff58"@},
34974 @{number="7",value="0xfe011e98"@},@{number="8",value="0x2"@},
34975 @{number="9",value="0xfa202820"@},@{number="10",value="0xfa202808"@},
34976 @{number="11",value="0x1"@},@{number="12",value="0x0"@},
34977 @{number="13",value="0x4544"@},@{number="14",value="0xffdfffff"@},
34978 @{number="15",value="0xffffffff"@},@{number="16",value="0xfffffeff"@},
34979 @{number="17",value="0xefffffed"@},@{number="18",value="0xfffffffe"@},
34980 @{number="19",value="0xffffffff"@},@{number="20",value="0xffffffff"@},
34981 @{number="21",value="0xffffffff"@},@{number="22",value="0xfffffff7"@},
34982 @{number="23",value="0xffffffff"@},@{number="24",value="0xffffffff"@},
34983 @{number="25",value="0xffffffff"@},@{number="26",value="0xfffffffb"@},
34984 @{number="27",value="0xffffffff"@},@{number="28",value="0xf7bfffff"@},
34985 @{number="29",value="0x0"@},@{number="30",value="0xfe010000"@},
34986 @{number="31",value="0x0"@},@{number="32",value="0x0"@},
34987 @{number="33",value="0x0"@},@{number="34",value="0x0"@},
34988 @{number="35",value="0x0"@},@{number="36",value="0x0"@},
34989 @{number="37",value="0x0"@},@{number="38",value="0x0"@},
34990 @{number="39",value="0x0"@},@{number="40",value="0x0"@},
34991 @{number="41",value="0x0"@},@{number="42",value="0x0"@},
34992 @{number="43",value="0x0"@},@{number="44",value="0x0"@},
34993 @{number="45",value="0x0"@},@{number="46",value="0x0"@},
34994 @{number="47",value="0x0"@},@{number="48",value="0x0"@},
34995 @{number="49",value="0x0"@},@{number="50",value="0x0"@},
34996 @{number="51",value="0x0"@},@{number="52",value="0x0"@},
34997 @{number="53",value="0x0"@},@{number="54",value="0x0"@},
34998 @{number="55",value="0x0"@},@{number="56",value="0x0"@},
34999 @{number="57",value="0x0"@},@{number="58",value="0x0"@},
35000 @{number="59",value="0x0"@},@{number="60",value="0x0"@},
35001 @{number="61",value="0x0"@},@{number="62",value="0x0"@},
35002 @{number="63",value="0x0"@},@{number="64",value="0xfe00a300"@},
35003 @{number="65",value="0x29002"@},@{number="66",value="0x202f04b5"@},
35004 @{number="67",value="0xfe0043b0"@},@{number="68",value="0xfe00b3e4"@},
35005 @{number="69",value="0x20002b03"@}]
35010 @subheading The @code{-data-read-memory} Command
35011 @findex -data-read-memory
35013 This command is deprecated, use @code{-data-read-memory-bytes} instead.
35015 @subsubheading Synopsis
35018 -data-read-memory [ -o @var{byte-offset} ]
35019 @var{address} @var{word-format} @var{word-size}
35020 @var{nr-rows} @var{nr-cols} [ @var{aschar} ]
35027 @item @var{address}
35028 An expression specifying the address of the first memory word to be
35029 read. Complex expressions containing embedded white space should be
35030 quoted using the C convention.
35032 @item @var{word-format}
35033 The format to be used to print the memory words. The notation is the
35034 same as for @value{GDBN}'s @code{print} command (@pxref{Output Formats,
35037 @item @var{word-size}
35038 The size of each memory word in bytes.
35040 @item @var{nr-rows}
35041 The number of rows in the output table.
35043 @item @var{nr-cols}
35044 The number of columns in the output table.
35047 If present, indicates that each row should include an @sc{ascii} dump. The
35048 value of @var{aschar} is used as a padding character when a byte is not a
35049 member of the printable @sc{ascii} character set (printable @sc{ascii}
35050 characters are those whose code is between 32 and 126, inclusively).
35052 @item @var{byte-offset}
35053 An offset to add to the @var{address} before fetching memory.
35056 This command displays memory contents as a table of @var{nr-rows} by
35057 @var{nr-cols} words, each word being @var{word-size} bytes. In total,
35058 @code{@var{nr-rows} * @var{nr-cols} * @var{word-size}} bytes are read
35059 (returned as @samp{total-bytes}). Should less than the requested number
35060 of bytes be returned by the target, the missing words are identified
35061 using @samp{N/A}. The number of bytes read from the target is returned
35062 in @samp{nr-bytes} and the starting address used to read memory in
35065 The address of the next/previous row or page is available in
35066 @samp{next-row} and @samp{prev-row}, @samp{next-page} and
35069 @subsubheading @value{GDBN} Command
35071 The corresponding @value{GDBN} command is @samp{x}. @code{gdbtk} has
35072 @samp{gdb_get_mem} memory read command.
35074 @subsubheading Example
35076 Read six bytes of memory starting at @code{bytes+6} but then offset by
35077 @code{-6} bytes. Format as three rows of two columns. One byte per
35078 word. Display each word in hex.
35082 9-data-read-memory -o -6 -- bytes+6 x 1 3 2
35083 9^done,addr="0x00001390",nr-bytes="6",total-bytes="6",
35084 next-row="0x00001396",prev-row="0x0000138e",next-page="0x00001396",
35085 prev-page="0x0000138a",memory=[
35086 @{addr="0x00001390",data=["0x00","0x01"]@},
35087 @{addr="0x00001392",data=["0x02","0x03"]@},
35088 @{addr="0x00001394",data=["0x04","0x05"]@}]
35092 Read two bytes of memory starting at address @code{shorts + 64} and
35093 display as a single word formatted in decimal.
35097 5-data-read-memory shorts+64 d 2 1 1
35098 5^done,addr="0x00001510",nr-bytes="2",total-bytes="2",
35099 next-row="0x00001512",prev-row="0x0000150e",
35100 next-page="0x00001512",prev-page="0x0000150e",memory=[
35101 @{addr="0x00001510",data=["128"]@}]
35105 Read thirty two bytes of memory starting at @code{bytes+16} and format
35106 as eight rows of four columns. Include a string encoding with @samp{x}
35107 used as the non-printable character.
35111 4-data-read-memory bytes+16 x 1 8 4 x
35112 4^done,addr="0x000013a0",nr-bytes="32",total-bytes="32",
35113 next-row="0x000013c0",prev-row="0x0000139c",
35114 next-page="0x000013c0",prev-page="0x00001380",memory=[
35115 @{addr="0x000013a0",data=["0x10","0x11","0x12","0x13"],ascii="xxxx"@},
35116 @{addr="0x000013a4",data=["0x14","0x15","0x16","0x17"],ascii="xxxx"@},
35117 @{addr="0x000013a8",data=["0x18","0x19","0x1a","0x1b"],ascii="xxxx"@},
35118 @{addr="0x000013ac",data=["0x1c","0x1d","0x1e","0x1f"],ascii="xxxx"@},
35119 @{addr="0x000013b0",data=["0x20","0x21","0x22","0x23"],ascii=" !\"#"@},
35120 @{addr="0x000013b4",data=["0x24","0x25","0x26","0x27"],ascii="$%&'"@},
35121 @{addr="0x000013b8",data=["0x28","0x29","0x2a","0x2b"],ascii="()*+"@},
35122 @{addr="0x000013bc",data=["0x2c","0x2d","0x2e","0x2f"],ascii=",-./"@}]
35126 @subheading The @code{-data-read-memory-bytes} Command
35127 @findex -data-read-memory-bytes
35129 @subsubheading Synopsis
35132 -data-read-memory-bytes [ -o @var{offset} ]
35133 @var{address} @var{count}
35140 @item @var{address}
35141 An expression specifying the address of the first addressable memory unit
35142 to be read. Complex expressions containing embedded white space should be
35143 quoted using the C convention.
35146 The number of addressable memory units to read. This should be an integer
35150 The offset relative to @var{address} at which to start reading. This
35151 should be an integer literal. This option is provided so that a frontend
35152 is not required to first evaluate address and then perform address
35153 arithmetics itself.
35157 This command attempts to read all accessible memory regions in the
35158 specified range. First, all regions marked as unreadable in the memory
35159 map (if one is defined) will be skipped. @xref{Memory Region
35160 Attributes}. Second, @value{GDBN} will attempt to read the remaining
35161 regions. For each one, if reading full region results in an errors,
35162 @value{GDBN} will try to read a subset of the region.
35164 In general, every single memory unit in the region may be readable or not,
35165 and the only way to read every readable unit is to try a read at
35166 every address, which is not practical. Therefore, @value{GDBN} will
35167 attempt to read all accessible memory units at either beginning or the end
35168 of the region, using a binary division scheme. This heuristic works
35169 well for reading across a memory map boundary. Note that if a region
35170 has a readable range that is neither at the beginning or the end,
35171 @value{GDBN} will not read it.
35173 The result record (@pxref{GDB/MI Result Records}) that is output of
35174 the command includes a field named @samp{memory} whose content is a
35175 list of tuples. Each tuple represent a successfully read memory block
35176 and has the following fields:
35180 The start address of the memory block, as hexadecimal literal.
35183 The end address of the memory block, as hexadecimal literal.
35186 The offset of the memory block, as hexadecimal literal, relative to
35187 the start address passed to @code{-data-read-memory-bytes}.
35190 The contents of the memory block, in hex.
35196 @subsubheading @value{GDBN} Command
35198 The corresponding @value{GDBN} command is @samp{x}.
35200 @subsubheading Example
35204 -data-read-memory-bytes &a 10
35205 ^done,memory=[@{begin="0xbffff154",offset="0x00000000",
35207 contents="01000000020000000300"@}]
35212 @subheading The @code{-data-write-memory-bytes} Command
35213 @findex -data-write-memory-bytes
35215 @subsubheading Synopsis
35218 -data-write-memory-bytes @var{address} @var{contents}
35219 -data-write-memory-bytes @var{address} @var{contents} @r{[}@var{count}@r{]}
35226 @item @var{address}
35227 An expression specifying the address of the first addressable memory unit
35228 to be written. Complex expressions containing embedded white space should
35229 be quoted using the C convention.
35231 @item @var{contents}
35232 The hex-encoded data to write. It is an error if @var{contents} does
35233 not represent an integral number of addressable memory units.
35236 Optional argument indicating the number of addressable memory units to be
35237 written. If @var{count} is greater than @var{contents}' length,
35238 @value{GDBN} will repeatedly write @var{contents} until it fills
35239 @var{count} memory units.
35243 @subsubheading @value{GDBN} Command
35245 There's no corresponding @value{GDBN} command.
35247 @subsubheading Example
35251 -data-write-memory-bytes &a "aabbccdd"
35258 -data-write-memory-bytes &a "aabbccdd" 16e
35263 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
35264 @node GDB/MI Tracepoint Commands
35265 @section @sc{gdb/mi} Tracepoint Commands
35267 The commands defined in this section implement MI support for
35268 tracepoints. For detailed introduction, see @ref{Tracepoints}.
35270 @subheading The @code{-trace-find} Command
35271 @findex -trace-find
35273 @subsubheading Synopsis
35276 -trace-find @var{mode} [@var{parameters}@dots{}]
35279 Find a trace frame using criteria defined by @var{mode} and
35280 @var{parameters}. The following table lists permissible
35281 modes and their parameters. For details of operation, see @ref{tfind}.
35286 No parameters are required. Stops examining trace frames.
35289 An integer is required as parameter. Selects tracepoint frame with
35292 @item tracepoint-number
35293 An integer is required as parameter. Finds next
35294 trace frame that corresponds to tracepoint with the specified number.
35297 An address is required as parameter. Finds
35298 next trace frame that corresponds to any tracepoint at the specified
35301 @item pc-inside-range
35302 Two addresses are required as parameters. Finds next trace
35303 frame that corresponds to a tracepoint at an address inside the
35304 specified range. Both bounds are considered to be inside the range.
35306 @item pc-outside-range
35307 Two addresses are required as parameters. Finds
35308 next trace frame that corresponds to a tracepoint at an address outside
35309 the specified range. Both bounds are considered to be inside the range.
35312 Line specification is required as parameter. @xref{Specify Location}.
35313 Finds next trace frame that corresponds to a tracepoint at
35314 the specified location.
35318 If @samp{none} was passed as @var{mode}, the response does not
35319 have fields. Otherwise, the response may have the following fields:
35323 This field has either @samp{0} or @samp{1} as the value, depending
35324 on whether a matching tracepoint was found.
35327 The index of the found traceframe. This field is present iff
35328 the @samp{found} field has value of @samp{1}.
35331 The index of the found tracepoint. This field is present iff
35332 the @samp{found} field has value of @samp{1}.
35335 The information about the frame corresponding to the found trace
35336 frame. This field is present only if a trace frame was found.
35337 @xref{GDB/MI Frame Information}, for description of this field.
35341 @subsubheading @value{GDBN} Command
35343 The corresponding @value{GDBN} command is @samp{tfind}.
35345 @subheading -trace-define-variable
35346 @findex -trace-define-variable
35348 @subsubheading Synopsis
35351 -trace-define-variable @var{name} [ @var{value} ]
35354 Create trace variable @var{name} if it does not exist. If
35355 @var{value} is specified, sets the initial value of the specified
35356 trace variable to that value. Note that the @var{name} should start
35357 with the @samp{$} character.
35359 @subsubheading @value{GDBN} Command
35361 The corresponding @value{GDBN} command is @samp{tvariable}.
35363 @subheading The @code{-trace-frame-collected} Command
35364 @findex -trace-frame-collected
35366 @subsubheading Synopsis
35369 -trace-frame-collected
35370 [--var-print-values @var{var_pval}]
35371 [--comp-print-values @var{comp_pval}]
35372 [--registers-format @var{regformat}]
35373 [--memory-contents]
35376 This command returns the set of collected objects, register names,
35377 trace state variable names, memory ranges and computed expressions
35378 that have been collected at a particular trace frame. The optional
35379 parameters to the command affect the output format in different ways.
35380 See the output description table below for more details.
35382 The reported names can be used in the normal manner to create
35383 varobjs and inspect the objects themselves. The items returned by
35384 this command are categorized so that it is clear which is a variable,
35385 which is a register, which is a trace state variable, which is a
35386 memory range and which is a computed expression.
35388 For instance, if the actions were
35390 collect myVar, myArray[myIndex], myObj.field, myPtr->field, myCount + 2
35391 collect *(int*)0xaf02bef0@@40
35395 the object collected in its entirety would be @code{myVar}. The
35396 object @code{myArray} would be partially collected, because only the
35397 element at index @code{myIndex} would be collected. The remaining
35398 objects would be computed expressions.
35400 An example output would be:
35404 -trace-frame-collected
35406 explicit-variables=[@{name="myVar",value="1"@}],
35407 computed-expressions=[@{name="myArray[myIndex]",value="0"@},
35408 @{name="myObj.field",value="0"@},
35409 @{name="myPtr->field",value="1"@},
35410 @{name="myCount + 2",value="3"@},
35411 @{name="$tvar1 + 1",value="43970027"@}],
35412 registers=[@{number="0",value="0x7fe2c6e79ec8"@},
35413 @{number="1",value="0x0"@},
35414 @{number="2",value="0x4"@},
35416 @{number="125",value="0x0"@}],
35417 tvars=[@{name="$tvar1",current="43970026"@}],
35418 memory=[@{address="0x0000000000602264",length="4"@},
35419 @{address="0x0000000000615bc0",length="4"@}]
35426 @item explicit-variables
35427 The set of objects that have been collected in their entirety (as
35428 opposed to collecting just a few elements of an array or a few struct
35429 members). For each object, its name and value are printed.
35430 The @code{--var-print-values} option affects how or whether the value
35431 field is output. If @var{var_pval} is 0, then print only the names;
35432 if it is 1, print also their values; and if it is 2, print the name,
35433 type and value for simple data types, and the name and type for
35434 arrays, structures and unions.
35436 @item computed-expressions
35437 The set of computed expressions that have been collected at the
35438 current trace frame. The @code{--comp-print-values} option affects
35439 this set like the @code{--var-print-values} option affects the
35440 @code{explicit-variables} set. See above.
35443 The registers that have been collected at the current trace frame.
35444 For each register collected, the name and current value are returned.
35445 The value is formatted according to the @code{--registers-format}
35446 option. See the @command{-data-list-register-values} command for a
35447 list of the allowed formats. The default is @samp{x}.
35450 The trace state variables that have been collected at the current
35451 trace frame. For each trace state variable collected, the name and
35452 current value are returned.
35455 The set of memory ranges that have been collected at the current trace
35456 frame. Its content is a list of tuples. Each tuple represents a
35457 collected memory range and has the following fields:
35461 The start address of the memory range, as hexadecimal literal.
35464 The length of the memory range, as decimal literal.
35467 The contents of the memory block, in hex. This field is only present
35468 if the @code{--memory-contents} option is specified.
35474 @subsubheading @value{GDBN} Command
35476 There is no corresponding @value{GDBN} command.
35478 @subsubheading Example
35480 @subheading -trace-list-variables
35481 @findex -trace-list-variables
35483 @subsubheading Synopsis
35486 -trace-list-variables
35489 Return a table of all defined trace variables. Each element of the
35490 table has the following fields:
35494 The name of the trace variable. This field is always present.
35497 The initial value. This is a 64-bit signed integer. This
35498 field is always present.
35501 The value the trace variable has at the moment. This is a 64-bit
35502 signed integer. This field is absent iff current value is
35503 not defined, for example if the trace was never run, or is
35508 @subsubheading @value{GDBN} Command
35510 The corresponding @value{GDBN} command is @samp{tvariables}.
35512 @subsubheading Example
35516 -trace-list-variables
35517 ^done,trace-variables=@{nr_rows="1",nr_cols="3",
35518 hdr=[@{width="15",alignment="-1",col_name="name",colhdr="Name"@},
35519 @{width="11",alignment="-1",col_name="initial",colhdr="Initial"@},
35520 @{width="11",alignment="-1",col_name="current",colhdr="Current"@}],
35521 body=[variable=@{name="$trace_timestamp",initial="0"@}
35522 variable=@{name="$foo",initial="10",current="15"@}]@}
35526 @subheading -trace-save
35527 @findex -trace-save
35529 @subsubheading Synopsis
35532 -trace-save [ -r ] [ -ctf ] @var{filename}
35535 Saves the collected trace data to @var{filename}. Without the
35536 @samp{-r} option, the data is downloaded from the target and saved
35537 in a local file. With the @samp{-r} option the target is asked
35538 to perform the save.
35540 By default, this command will save the trace in the tfile format. You can
35541 supply the optional @samp{-ctf} argument to save it the CTF format. See
35542 @ref{Trace Files} for more information about CTF.
35544 @subsubheading @value{GDBN} Command
35546 The corresponding @value{GDBN} command is @samp{tsave}.
35549 @subheading -trace-start
35550 @findex -trace-start
35552 @subsubheading Synopsis
35558 Starts a tracing experiment. The result of this command does not
35561 @subsubheading @value{GDBN} Command
35563 The corresponding @value{GDBN} command is @samp{tstart}.
35565 @subheading -trace-status
35566 @findex -trace-status
35568 @subsubheading Synopsis
35574 Obtains the status of a tracing experiment. The result may include
35575 the following fields:
35580 May have a value of either @samp{0}, when no tracing operations are
35581 supported, @samp{1}, when all tracing operations are supported, or
35582 @samp{file} when examining trace file. In the latter case, examining
35583 of trace frame is possible but new tracing experiement cannot be
35584 started. This field is always present.
35587 May have a value of either @samp{0} or @samp{1} depending on whether
35588 tracing experiement is in progress on target. This field is present
35589 if @samp{supported} field is not @samp{0}.
35592 Report the reason why the tracing was stopped last time. This field
35593 may be absent iff tracing was never stopped on target yet. The
35594 value of @samp{request} means the tracing was stopped as result of
35595 the @code{-trace-stop} command. The value of @samp{overflow} means
35596 the tracing buffer is full. The value of @samp{disconnection} means
35597 tracing was automatically stopped when @value{GDBN} has disconnected.
35598 The value of @samp{passcount} means tracing was stopped when a
35599 tracepoint was passed a maximal number of times for that tracepoint.
35600 This field is present if @samp{supported} field is not @samp{0}.
35602 @item stopping-tracepoint
35603 The number of tracepoint whose passcount as exceeded. This field is
35604 present iff the @samp{stop-reason} field has the value of
35608 @itemx frames-created
35609 The @samp{frames} field is a count of the total number of trace frames
35610 in the trace buffer, while @samp{frames-created} is the total created
35611 during the run, including ones that were discarded, such as when a
35612 circular trace buffer filled up. Both fields are optional.
35616 These fields tell the current size of the tracing buffer and the
35617 remaining space. These fields are optional.
35620 The value of the circular trace buffer flag. @code{1} means that the
35621 trace buffer is circular and old trace frames will be discarded if
35622 necessary to make room, @code{0} means that the trace buffer is linear
35626 The value of the disconnected tracing flag. @code{1} means that
35627 tracing will continue after @value{GDBN} disconnects, @code{0} means
35628 that the trace run will stop.
35631 The filename of the trace file being examined. This field is
35632 optional, and only present when examining a trace file.
35636 @subsubheading @value{GDBN} Command
35638 The corresponding @value{GDBN} command is @samp{tstatus}.
35640 @subheading -trace-stop
35641 @findex -trace-stop
35643 @subsubheading Synopsis
35649 Stops a tracing experiment. The result of this command has the same
35650 fields as @code{-trace-status}, except that the @samp{supported} and
35651 @samp{running} fields are not output.
35653 @subsubheading @value{GDBN} Command
35655 The corresponding @value{GDBN} command is @samp{tstop}.
35658 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
35659 @node GDB/MI Symbol Query
35660 @section @sc{gdb/mi} Symbol Query Commands
35664 @subheading The @code{-symbol-info-address} Command
35665 @findex -symbol-info-address
35667 @subsubheading Synopsis
35670 -symbol-info-address @var{symbol}
35673 Describe where @var{symbol} is stored.
35675 @subsubheading @value{GDBN} Command
35677 The corresponding @value{GDBN} command is @samp{info address}.
35679 @subsubheading Example
35683 @subheading The @code{-symbol-info-file} Command
35684 @findex -symbol-info-file
35686 @subsubheading Synopsis
35692 Show the file for the symbol.
35694 @subsubheading @value{GDBN} Command
35696 There's no equivalent @value{GDBN} command. @code{gdbtk} has
35697 @samp{gdb_find_file}.
35699 @subsubheading Example
35703 @subheading The @code{-symbol-info-functions} Command
35704 @findex -symbol-info-functions
35705 @anchor{-symbol-info-functions}
35707 @subsubheading Synopsis
35710 -symbol-info-functions [--include-nondebug]
35711 [--type @var{type_regexp}]
35712 [--name @var{name_regexp}]
35713 [--max-results @var{limit}]
35717 Return a list containing the names and types for all global functions
35718 taken from the debug information. The functions are grouped by source
35719 file, and shown with the line number on which each function is
35722 The @code{--include-nondebug} option causes the output to include
35723 code symbols from the symbol table.
35725 The options @code{--type} and @code{--name} allow the symbols returned
35726 to be filtered based on either the name of the function, or the type
35727 signature of the function.
35729 The option @code{--max-results} restricts the command to return no
35730 more than @var{limit} results. If exactly @var{limit} results are
35731 returned then there might be additional results available if a higher
35734 @subsubheading @value{GDBN} Command
35736 The corresponding @value{GDBN} command is @samp{info functions}.
35738 @subsubheading Example
35742 -symbol-info-functions
35745 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35746 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35747 symbols=[@{line="36", name="f4", type="void (int *)",
35748 description="void f4(int *);"@},
35749 @{line="42", name="main", type="int ()",
35750 description="int main();"@},
35751 @{line="30", name="f1", type="my_int_t (int, int)",
35752 description="static my_int_t f1(int, int);"@}]@},
35753 @{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
35754 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
35755 symbols=[@{line="33", name="f2", type="float (another_float_t)",
35756 description="float f2(another_float_t);"@},
35757 @{line="39", name="f3", type="int (another_int_t)",
35758 description="int f3(another_int_t);"@},
35759 @{line="27", name="f1", type="another_float_t (int)",
35760 description="static another_float_t f1(int);"@}]@}]@}
35764 -symbol-info-functions --name f1
35767 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35768 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35769 symbols=[@{line="30", name="f1", type="my_int_t (int, int)",
35770 description="static my_int_t f1(int, int);"@}]@},
35771 @{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
35772 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
35773 symbols=[@{line="27", name="f1", type="another_float_t (int)",
35774 description="static another_float_t f1(int);"@}]@}]@}
35778 -symbol-info-functions --type void
35781 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35782 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35783 symbols=[@{line="36", name="f4", type="void (int *)",
35784 description="void f4(int *);"@}]@}]@}
35788 -symbol-info-functions --include-nondebug
35791 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35792 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
35793 symbols=[@{line="36", name="f4", type="void (int *)",
35794 description="void f4(int *);"@},
35795 @{line="42", name="main", type="int ()",
35796 description="int main();"@},
35797 @{line="30", name="f1", type="my_int_t (int, int)",
35798 description="static my_int_t f1(int, int);"@}]@},
35799 @{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
35800 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
35801 symbols=[@{line="33", name="f2", type="float (another_float_t)",
35802 description="float f2(another_float_t);"@},
35803 @{line="39", name="f3", type="int (another_int_t)",
35804 description="int f3(another_int_t);"@},
35805 @{line="27", name="f1", type="another_float_t (int)",
35806 description="static another_float_t f1(int);"@}]@}],
35808 [@{address="0x0000000000400398",name="_init"@},
35809 @{address="0x00000000004003b0",name="_start"@},
35815 @subheading The @code{-symbol-info-module-functions} Command
35816 @findex -symbol-info-module-functions
35817 @anchor{-symbol-info-module-functions}
35819 @subsubheading Synopsis
35822 -symbol-info-module-functions [--module @var{module_regexp}]
35823 [--name @var{name_regexp}]
35824 [--type @var{type_regexp}]
35828 Return a list containing the names of all known functions within all
35829 know Fortran modules. The functions are grouped by source file and
35830 containing module, and shown with the line number on which each
35831 function is defined.
35833 The option @code{--module} only returns results for modules matching
35834 @var{module_regexp}. The option @code{--name} only returns functions
35835 whose name matches @var{name_regexp}, and @code{--type} only returns
35836 functions whose type matches @var{type_regexp}.
35838 @subsubheading @value{GDBN} Command
35840 The corresponding @value{GDBN} command is @samp{info module functions}.
35842 @subsubheading Example
35847 -symbol-info-module-functions
35850 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35851 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35852 symbols=[@{line="21",name="mod1::check_all",type="void (void)",
35853 description="void mod1::check_all(void);"@}]@}]@},
35855 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35856 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35857 symbols=[@{line="30",name="mod2::check_var_i",type="void (void)",
35858 description="void mod2::check_var_i(void);"@}]@}]@},
35860 files=[@{filename="/projec/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35861 fullname="/projec/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35862 symbols=[@{line="21",name="mod3::check_all",type="void (void)",
35863 description="void mod3::check_all(void);"@},
35864 @{line="27",name="mod3::check_mod2",type="void (void)",
35865 description="void mod3::check_mod2(void);"@}]@}]@},
35866 @{module="modmany",
35867 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35868 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35869 symbols=[@{line="35",name="modmany::check_some",type="void (void)",
35870 description="void modmany::check_some(void);"@}]@}]@},
35872 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35873 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35874 symbols=[@{line="44",name="moduse::check_all",type="void (void)",
35875 description="void moduse::check_all(void);"@},
35876 @{line="49",name="moduse::check_var_x",type="void (void)",
35877 description="void moduse::check_var_x(void);"@}]@}]@}]
35881 @subheading The @code{-symbol-info-module-variables} Command
35882 @findex -symbol-info-module-variables
35883 @anchor{-symbol-info-module-variables}
35885 @subsubheading Synopsis
35888 -symbol-info-module-variables [--module @var{module_regexp}]
35889 [--name @var{name_regexp}]
35890 [--type @var{type_regexp}]
35894 Return a list containing the names of all known variables within all
35895 know Fortran modules. The variables are grouped by source file and
35896 containing module, and shown with the line number on which each
35897 variable is defined.
35899 The option @code{--module} only returns results for modules matching
35900 @var{module_regexp}. The option @code{--name} only returns variables
35901 whose name matches @var{name_regexp}, and @code{--type} only returns
35902 variables whose type matches @var{type_regexp}.
35904 @subsubheading @value{GDBN} Command
35906 The corresponding @value{GDBN} command is @samp{info module variables}.
35908 @subsubheading Example
35913 -symbol-info-module-variables
35916 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35917 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35918 symbols=[@{line="18",name="mod1::var_const",type="integer(kind=4)",
35919 description="integer(kind=4) mod1::var_const;"@},
35920 @{line="17",name="mod1::var_i",type="integer(kind=4)",
35921 description="integer(kind=4) mod1::var_i;"@}]@}]@},
35923 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35924 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35925 symbols=[@{line="28",name="mod2::var_i",type="integer(kind=4)",
35926 description="integer(kind=4) mod2::var_i;"@}]@}]@},
35928 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35929 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35930 symbols=[@{line="18",name="mod3::mod1",type="integer(kind=4)",
35931 description="integer(kind=4) mod3::mod1;"@},
35932 @{line="17",name="mod3::mod2",type="integer(kind=4)",
35933 description="integer(kind=4) mod3::mod2;"@},
35934 @{line="19",name="mod3::var_i",type="integer(kind=4)",
35935 description="integer(kind=4) mod3::var_i;"@}]@}]@},
35936 @{module="modmany",
35937 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35938 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35939 symbols=[@{line="33",name="modmany::var_a",type="integer(kind=4)",
35940 description="integer(kind=4) modmany::var_a;"@},
35941 @{line="33",name="modmany::var_b",type="integer(kind=4)",
35942 description="integer(kind=4) modmany::var_b;"@},
35943 @{line="33",name="modmany::var_c",type="integer(kind=4)",
35944 description="integer(kind=4) modmany::var_c;"@},
35945 @{line="33",name="modmany::var_i",type="integer(kind=4)",
35946 description="integer(kind=4) modmany::var_i;"@}]@}]@},
35948 files=[@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35949 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35950 symbols=[@{line="42",name="moduse::var_x",type="integer(kind=4)",
35951 description="integer(kind=4) moduse::var_x;"@},
35952 @{line="42",name="moduse::var_y",type="integer(kind=4)",
35953 description="integer(kind=4) moduse::var_y;"@}]@}]@}]
35957 @subheading The @code{-symbol-info-modules} Command
35958 @findex -symbol-info-modules
35959 @anchor{-symbol-info-modules}
35961 @subsubheading Synopsis
35964 -symbol-info-modules [--name @var{name_regexp}]
35965 [--max-results @var{limit}]
35970 Return a list containing the names of all known Fortran modules. The
35971 modules are grouped by source file, and shown with the line number on
35972 which each modules is defined.
35974 The option @code{--name} allows the modules returned to be filtered
35975 based the name of the module.
35977 The option @code{--max-results} restricts the command to return no
35978 more than @var{limit} results. If exactly @var{limit} results are
35979 returned then there might be additional results available if a higher
35982 @subsubheading @value{GDBN} Command
35984 The corresponding @value{GDBN} command is @samp{info modules}.
35986 @subsubheading Example
35990 -symbol-info-modules
35993 [@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35994 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
35995 symbols=[@{line="16",name="mod1"@},
35996 @{line="22",name="mod2"@}]@},
35997 @{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35998 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
35999 symbols=[@{line="16",name="mod3"@},
36000 @{line="22",name="modmany"@},
36001 @{line="26",name="moduse"@}]@}]@}
36005 -symbol-info-modules --name mod[123]
36008 [@{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
36009 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules-2.f90",
36010 symbols=[@{line="16",name="mod1"@},
36011 @{line="22",name="mod2"@}]@},
36012 @{filename="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
36013 fullname="/project/gdb/testsuite/gdb.mi/mi-fortran-modules.f90",
36014 symbols=[@{line="16",name="mod3"@}]@}]@}
36018 @subheading The @code{-symbol-info-types} Command
36019 @findex -symbol-info-types
36020 @anchor{-symbol-info-types}
36022 @subsubheading Synopsis
36025 -symbol-info-types [--name @var{name_regexp}]
36026 [--max-results @var{limit}]
36031 Return a list of all defined types. The types are grouped by source
36032 file, and shown with the line number on which each user defined type
36033 is defined. Some base types are not defined in the source code but
36034 are added to the debug information by the compiler, for example
36035 @code{int}, @code{float}, etc.; these types do not have an associated
36038 The option @code{--name} allows the list of types returned to be
36041 The option @code{--max-results} restricts the command to return no
36042 more than @var{limit} results. If exactly @var{limit} results are
36043 returned then there might be additional results available if a higher
36046 @subsubheading @value{GDBN} Command
36048 The corresponding @value{GDBN} command is @samp{info types}.
36050 @subsubheading Example
36057 [@{filename="gdb.mi/mi-sym-info-1.c",
36058 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36059 symbols=[@{name="float"@},
36061 @{line="27",name="typedef int my_int_t;"@}]@},
36062 @{filename="gdb.mi/mi-sym-info-2.c",
36063 fullname="/project/gdb.mi/mi-sym-info-2.c",
36064 symbols=[@{line="24",name="typedef float another_float_t;"@},
36065 @{line="23",name="typedef int another_int_t;"@},
36067 @{name="int"@}]@}]@}
36071 -symbol-info-types --name _int_
36074 [@{filename="gdb.mi/mi-sym-info-1.c",
36075 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36076 symbols=[@{line="27",name="typedef int my_int_t;"@}]@},
36077 @{filename="gdb.mi/mi-sym-info-2.c",
36078 fullname="/project/gdb.mi/mi-sym-info-2.c",
36079 symbols=[@{line="23",name="typedef int another_int_t;"@}]@}]@}
36083 @subheading The @code{-symbol-info-variables} Command
36084 @findex -symbol-info-variables
36085 @anchor{-symbol-info-variables}
36087 @subsubheading Synopsis
36090 -symbol-info-variables [--include-nondebug]
36091 [--type @var{type_regexp}]
36092 [--name @var{name_regexp}]
36093 [--max-results @var{limit}]
36098 Return a list containing the names and types for all global variables
36099 taken from the debug information. The variables are grouped by source
36100 file, and shown with the line number on which each variable is
36103 The @code{--include-nondebug} option causes the output to include
36104 data symbols from the symbol table.
36106 The options @code{--type} and @code{--name} allow the symbols returned
36107 to be filtered based on either the name of the variable, or the type
36110 The option @code{--max-results} restricts the command to return no
36111 more than @var{limit} results. If exactly @var{limit} results are
36112 returned then there might be additional results available if a higher
36115 @subsubheading @value{GDBN} Command
36117 The corresponding @value{GDBN} command is @samp{info variables}.
36119 @subsubheading Example
36123 -symbol-info-variables
36126 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36127 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36128 symbols=[@{line="25",name="global_f1",type="float",
36129 description="static float global_f1;"@},
36130 @{line="24",name="global_i1",type="int",
36131 description="static int global_i1;"@}]@},
36132 @{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36133 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36134 symbols=[@{line="21",name="global_f2",type="int",
36135 description="int global_f2;"@},
36136 @{line="20",name="global_i2",type="int",
36137 description="int global_i2;"@},
36138 @{line="19",name="global_f1",type="float",
36139 description="static float global_f1;"@},
36140 @{line="18",name="global_i1",type="int",
36141 description="static int global_i1;"@}]@}]@}
36145 -symbol-info-variables --name f1
36148 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36149 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36150 symbols=[@{line="25",name="global_f1",type="float",
36151 description="static float global_f1;"@}]@},
36152 @{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36153 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36154 symbols=[@{line="19",name="global_f1",type="float",
36155 description="static float global_f1;"@}]@}]@}
36159 -symbol-info-variables --type float
36162 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36163 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36164 symbols=[@{line="25",name="global_f1",type="float",
36165 description="static float global_f1;"@}]@},
36166 @{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36167 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36168 symbols=[@{line="19",name="global_f1",type="float",
36169 description="static float global_f1;"@}]@}]@}
36173 -symbol-info-variables --include-nondebug
36176 [@{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36177 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-1.c",
36178 symbols=[@{line="25",name="global_f1",type="float",
36179 description="static float global_f1;"@},
36180 @{line="24",name="global_i1",type="int",
36181 description="static int global_i1;"@}]@},
36182 @{filename="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36183 fullname="/project/gdb/testsuite/gdb.mi/mi-sym-info-2.c",
36184 symbols=[@{line="21",name="global_f2",type="int",
36185 description="int global_f2;"@},
36186 @{line="20",name="global_i2",type="int",
36187 description="int global_i2;"@},
36188 @{line="19",name="global_f1",type="float",
36189 description="static float global_f1;"@},
36190 @{line="18",name="global_i1",type="int",
36191 description="static int global_i1;"@}]@}],
36193 [@{address="0x00000000004005d0",name="_IO_stdin_used"@},
36194 @{address="0x00000000004005d8",name="__dso_handle"@}
36201 @subheading The @code{-symbol-info-line} Command
36202 @findex -symbol-info-line
36204 @subsubheading Synopsis
36210 Show the core addresses of the code for a source line.
36212 @subsubheading @value{GDBN} Command
36214 The corresponding @value{GDBN} command is @samp{info line}.
36215 @code{gdbtk} has the @samp{gdb_get_line} and @samp{gdb_get_file} commands.
36217 @subsubheading Example
36221 @subheading The @code{-symbol-info-symbol} Command
36222 @findex -symbol-info-symbol
36224 @subsubheading Synopsis
36227 -symbol-info-symbol @var{addr}
36230 Describe what symbol is at location @var{addr}.
36232 @subsubheading @value{GDBN} Command
36234 The corresponding @value{GDBN} command is @samp{info symbol}.
36236 @subsubheading Example
36240 @subheading The @code{-symbol-list-functions} Command
36241 @findex -symbol-list-functions
36243 @subsubheading Synopsis
36246 -symbol-list-functions
36249 List the functions in the executable.
36251 @subsubheading @value{GDBN} Command
36253 @samp{info functions} in @value{GDBN}, @samp{gdb_listfunc} and
36254 @samp{gdb_search} in @code{gdbtk}.
36256 @subsubheading Example
36261 @subheading The @code{-symbol-list-lines} Command
36262 @findex -symbol-list-lines
36264 @subsubheading Synopsis
36267 -symbol-list-lines @var{filename}
36270 Print the list of lines that contain code and their associated program
36271 addresses for the given source filename. The entries are sorted in
36272 ascending PC order.
36274 @subsubheading @value{GDBN} Command
36276 There is no corresponding @value{GDBN} command.
36278 @subsubheading Example
36281 -symbol-list-lines basics.c
36282 ^done,lines=[@{pc="0x08048554",line="7"@},@{pc="0x0804855a",line="8"@}]
36288 @subheading The @code{-symbol-list-types} Command
36289 @findex -symbol-list-types
36291 @subsubheading Synopsis
36297 List all the type names.
36299 @subsubheading @value{GDBN} Command
36301 The corresponding commands are @samp{info types} in @value{GDBN},
36302 @samp{gdb_search} in @code{gdbtk}.
36304 @subsubheading Example
36308 @subheading The @code{-symbol-list-variables} Command
36309 @findex -symbol-list-variables
36311 @subsubheading Synopsis
36314 -symbol-list-variables
36317 List all the global and static variable names.
36319 @subsubheading @value{GDBN} Command
36321 @samp{info variables} in @value{GDBN}, @samp{gdb_search} in @code{gdbtk}.
36323 @subsubheading Example
36327 @subheading The @code{-symbol-locate} Command
36328 @findex -symbol-locate
36330 @subsubheading Synopsis
36336 @subsubheading @value{GDBN} Command
36338 @samp{gdb_loc} in @code{gdbtk}.
36340 @subsubheading Example
36344 @subheading The @code{-symbol-type} Command
36345 @findex -symbol-type
36347 @subsubheading Synopsis
36350 -symbol-type @var{variable}
36353 Show type of @var{variable}.
36355 @subsubheading @value{GDBN} Command
36357 The corresponding @value{GDBN} command is @samp{ptype}, @code{gdbtk} has
36358 @samp{gdb_obj_variable}.
36360 @subsubheading Example
36365 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
36366 @node GDB/MI File Commands
36367 @section @sc{gdb/mi} File Commands
36369 This section describes the GDB/MI commands to specify executable file names
36370 and to read in and obtain symbol table information.
36372 @subheading The @code{-file-exec-and-symbols} Command
36373 @findex -file-exec-and-symbols
36375 @subsubheading Synopsis
36378 -file-exec-and-symbols @var{file}
36381 Specify the executable file to be debugged. This file is the one from
36382 which the symbol table is also read. If no file is specified, the
36383 command clears the executable and symbol information. If breakpoints
36384 are set when using this command with no arguments, @value{GDBN} will produce
36385 error messages. Otherwise, no output is produced, except a completion
36388 @subsubheading @value{GDBN} Command
36390 The corresponding @value{GDBN} command is @samp{file}.
36392 @subsubheading Example
36396 -file-exec-and-symbols /kwikemart/marge/ezannoni/TRUNK/mbx/hello.mbx
36402 @subheading The @code{-file-exec-file} Command
36403 @findex -file-exec-file
36405 @subsubheading Synopsis
36408 -file-exec-file @var{file}
36411 Specify the executable file to be debugged. Unlike
36412 @samp{-file-exec-and-symbols}, the symbol table is @emph{not} read
36413 from this file. If used without argument, @value{GDBN} clears the information
36414 about the executable file. No output is produced, except a completion
36417 @subsubheading @value{GDBN} Command
36419 The corresponding @value{GDBN} command is @samp{exec-file}.
36421 @subsubheading Example
36425 -file-exec-file /kwikemart/marge/ezannoni/TRUNK/mbx/hello.mbx
36432 @subheading The @code{-file-list-exec-sections} Command
36433 @findex -file-list-exec-sections
36435 @subsubheading Synopsis
36438 -file-list-exec-sections
36441 List the sections of the current executable file.
36443 @subsubheading @value{GDBN} Command
36445 The @value{GDBN} command @samp{info file} shows, among the rest, the same
36446 information as this command. @code{gdbtk} has a corresponding command
36447 @samp{gdb_load_info}.
36449 @subsubheading Example
36454 @subheading The @code{-file-list-exec-source-file} Command
36455 @findex -file-list-exec-source-file
36457 @subsubheading Synopsis
36460 -file-list-exec-source-file
36463 List the line number, the current source file, and the absolute path
36464 to the current source file for the current executable. The macro
36465 information field has a value of @samp{1} or @samp{0} depending on
36466 whether or not the file includes preprocessor macro information.
36468 @subsubheading @value{GDBN} Command
36470 The @value{GDBN} equivalent is @samp{info source}
36472 @subsubheading Example
36476 123-file-list-exec-source-file
36477 123^done,line="1",file="foo.c",fullname="/home/bar/foo.c,macro-info="1"
36482 @subheading The @code{-file-list-exec-source-files} Command
36483 @findex -file-list-exec-source-files
36485 @subsubheading Synopsis
36488 -file-list-exec-source-files
36491 List the source files for the current executable.
36493 It will always output both the filename and fullname (absolute file
36494 name) of a source file.
36496 @subsubheading @value{GDBN} Command
36498 The @value{GDBN} equivalent is @samp{info sources}.
36499 @code{gdbtk} has an analogous command @samp{gdb_listfiles}.
36501 @subsubheading Example
36504 -file-list-exec-source-files
36506 @{file=foo.c,fullname=/home/foo.c@},
36507 @{file=/home/bar.c,fullname=/home/bar.c@},
36508 @{file=gdb_could_not_find_fullpath.c@}]
36512 @subheading The @code{-file-list-shared-libraries} Command
36513 @findex -file-list-shared-libraries
36515 @subsubheading Synopsis
36518 -file-list-shared-libraries [ @var{regexp} ]
36521 List the shared libraries in the program.
36522 With a regular expression @var{regexp}, only those libraries whose
36523 names match @var{regexp} are listed.
36525 @subsubheading @value{GDBN} Command
36527 The corresponding @value{GDBN} command is @samp{info shared}. The fields
36528 have a similar meaning to the @code{=library-loaded} notification.
36529 The @code{ranges} field specifies the multiple segments belonging to this
36530 library. Each range has the following fields:
36534 The address defining the inclusive lower bound of the segment.
36536 The address defining the exclusive upper bound of the segment.
36539 @subsubheading Example
36542 -file-list-exec-source-files
36543 ^done,shared-libraries=[
36544 @{id="/lib/libfoo.so",target-name="/lib/libfoo.so",host-name="/lib/libfoo.so",symbols-loaded="1",thread-group="i1",ranges=[@{from="0x72815989",to="0x728162c0"@}]@},
36545 @{id="/lib/libbar.so",target-name="/lib/libbar.so",host-name="/lib/libbar.so",symbols-loaded="1",thread-group="i1",ranges=[@{from="0x76ee48c0",to="0x76ee9160"@}]@}]
36551 @subheading The @code{-file-list-symbol-files} Command
36552 @findex -file-list-symbol-files
36554 @subsubheading Synopsis
36557 -file-list-symbol-files
36562 @subsubheading @value{GDBN} Command
36564 The corresponding @value{GDBN} command is @samp{info file} (part of it).
36566 @subsubheading Example
36571 @subheading The @code{-file-symbol-file} Command
36572 @findex -file-symbol-file
36574 @subsubheading Synopsis
36577 -file-symbol-file @var{file}
36580 Read symbol table info from the specified @var{file} argument. When
36581 used without arguments, clears @value{GDBN}'s symbol table info. No output is
36582 produced, except for a completion notification.
36584 @subsubheading @value{GDBN} Command
36586 The corresponding @value{GDBN} command is @samp{symbol-file}.
36588 @subsubheading Example
36592 -file-symbol-file /kwikemart/marge/ezannoni/TRUNK/mbx/hello.mbx
36598 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
36599 @node GDB/MI Memory Overlay Commands
36600 @section @sc{gdb/mi} Memory Overlay Commands
36602 The memory overlay commands are not implemented.
36604 @c @subheading -overlay-auto
36606 @c @subheading -overlay-list-mapping-state
36608 @c @subheading -overlay-list-overlays
36610 @c @subheading -overlay-map
36612 @c @subheading -overlay-off
36614 @c @subheading -overlay-on
36616 @c @subheading -overlay-unmap
36618 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
36619 @node GDB/MI Signal Handling Commands
36620 @section @sc{gdb/mi} Signal Handling Commands
36622 Signal handling commands are not implemented.
36624 @c @subheading -signal-handle
36626 @c @subheading -signal-list-handle-actions
36628 @c @subheading -signal-list-signal-types
36632 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
36633 @node GDB/MI Target Manipulation
36634 @section @sc{gdb/mi} Target Manipulation Commands
36637 @subheading The @code{-target-attach} Command
36638 @findex -target-attach
36640 @subsubheading Synopsis
36643 -target-attach @var{pid} | @var{gid} | @var{file}
36646 Attach to a process @var{pid} or a file @var{file} outside of
36647 @value{GDBN}, or a thread group @var{gid}. If attaching to a thread
36648 group, the id previously returned by
36649 @samp{-list-thread-groups --available} must be used.
36651 @subsubheading @value{GDBN} Command
36653 The corresponding @value{GDBN} command is @samp{attach}.
36655 @subsubheading Example
36659 =thread-created,id="1"
36660 *stopped,thread-id="1",frame=@{addr="0xb7f7e410",func="bar",args=[]@}
36666 @subheading The @code{-target-compare-sections} Command
36667 @findex -target-compare-sections
36669 @subsubheading Synopsis
36672 -target-compare-sections [ @var{section} ]
36675 Compare data of section @var{section} on target to the exec file.
36676 Without the argument, all sections are compared.
36678 @subsubheading @value{GDBN} Command
36680 The @value{GDBN} equivalent is @samp{compare-sections}.
36682 @subsubheading Example
36687 @subheading The @code{-target-detach} Command
36688 @findex -target-detach
36690 @subsubheading Synopsis
36693 -target-detach [ @var{pid} | @var{gid} ]
36696 Detach from the remote target which normally resumes its execution.
36697 If either @var{pid} or @var{gid} is specified, detaches from either
36698 the specified process, or specified thread group. There's no output.
36700 @subsubheading @value{GDBN} Command
36702 The corresponding @value{GDBN} command is @samp{detach}.
36704 @subsubheading Example
36714 @subheading The @code{-target-disconnect} Command
36715 @findex -target-disconnect
36717 @subsubheading Synopsis
36723 Disconnect from the remote target. There's no output and the target is
36724 generally not resumed.
36726 @subsubheading @value{GDBN} Command
36728 The corresponding @value{GDBN} command is @samp{disconnect}.
36730 @subsubheading Example
36740 @subheading The @code{-target-download} Command
36741 @findex -target-download
36743 @subsubheading Synopsis
36749 Loads the executable onto the remote target.
36750 It prints out an update message every half second, which includes the fields:
36754 The name of the section.
36756 The size of what has been sent so far for that section.
36758 The size of the section.
36760 The total size of what was sent so far (the current and the previous sections).
36762 The size of the overall executable to download.
36766 Each message is sent as status record (@pxref{GDB/MI Output Syntax, ,
36767 @sc{gdb/mi} Output Syntax}).
36769 In addition, it prints the name and size of the sections, as they are
36770 downloaded. These messages include the following fields:
36774 The name of the section.
36776 The size of the section.
36778 The size of the overall executable to download.
36782 At the end, a summary is printed.
36784 @subsubheading @value{GDBN} Command
36786 The corresponding @value{GDBN} command is @samp{load}.
36788 @subsubheading Example
36790 Note: each status message appears on a single line. Here the messages
36791 have been broken down so that they can fit onto a page.
36796 +download,@{section=".text",section-size="6668",total-size="9880"@}
36797 +download,@{section=".text",section-sent="512",section-size="6668",
36798 total-sent="512",total-size="9880"@}
36799 +download,@{section=".text",section-sent="1024",section-size="6668",
36800 total-sent="1024",total-size="9880"@}
36801 +download,@{section=".text",section-sent="1536",section-size="6668",
36802 total-sent="1536",total-size="9880"@}
36803 +download,@{section=".text",section-sent="2048",section-size="6668",
36804 total-sent="2048",total-size="9880"@}
36805 +download,@{section=".text",section-sent="2560",section-size="6668",
36806 total-sent="2560",total-size="9880"@}
36807 +download,@{section=".text",section-sent="3072",section-size="6668",
36808 total-sent="3072",total-size="9880"@}
36809 +download,@{section=".text",section-sent="3584",section-size="6668",
36810 total-sent="3584",total-size="9880"@}
36811 +download,@{section=".text",section-sent="4096",section-size="6668",
36812 total-sent="4096",total-size="9880"@}
36813 +download,@{section=".text",section-sent="4608",section-size="6668",
36814 total-sent="4608",total-size="9880"@}
36815 +download,@{section=".text",section-sent="5120",section-size="6668",
36816 total-sent="5120",total-size="9880"@}
36817 +download,@{section=".text",section-sent="5632",section-size="6668",
36818 total-sent="5632",total-size="9880"@}
36819 +download,@{section=".text",section-sent="6144",section-size="6668",
36820 total-sent="6144",total-size="9880"@}
36821 +download,@{section=".text",section-sent="6656",section-size="6668",
36822 total-sent="6656",total-size="9880"@}
36823 +download,@{section=".init",section-size="28",total-size="9880"@}
36824 +download,@{section=".fini",section-size="28",total-size="9880"@}
36825 +download,@{section=".data",section-size="3156",total-size="9880"@}
36826 +download,@{section=".data",section-sent="512",section-size="3156",
36827 total-sent="7236",total-size="9880"@}
36828 +download,@{section=".data",section-sent="1024",section-size="3156",
36829 total-sent="7748",total-size="9880"@}
36830 +download,@{section=".data",section-sent="1536",section-size="3156",
36831 total-sent="8260",total-size="9880"@}
36832 +download,@{section=".data",section-sent="2048",section-size="3156",
36833 total-sent="8772",total-size="9880"@}
36834 +download,@{section=".data",section-sent="2560",section-size="3156",
36835 total-sent="9284",total-size="9880"@}
36836 +download,@{section=".data",section-sent="3072",section-size="3156",
36837 total-sent="9796",total-size="9880"@}
36838 ^done,address="0x10004",load-size="9880",transfer-rate="6586",
36845 @subheading The @code{-target-exec-status} Command
36846 @findex -target-exec-status
36848 @subsubheading Synopsis
36851 -target-exec-status
36854 Provide information on the state of the target (whether it is running or
36855 not, for instance).
36857 @subsubheading @value{GDBN} Command
36859 There's no equivalent @value{GDBN} command.
36861 @subsubheading Example
36865 @subheading The @code{-target-list-available-targets} Command
36866 @findex -target-list-available-targets
36868 @subsubheading Synopsis
36871 -target-list-available-targets
36874 List the possible targets to connect to.
36876 @subsubheading @value{GDBN} Command
36878 The corresponding @value{GDBN} command is @samp{help target}.
36880 @subsubheading Example
36884 @subheading The @code{-target-list-current-targets} Command
36885 @findex -target-list-current-targets
36887 @subsubheading Synopsis
36890 -target-list-current-targets
36893 Describe the current target.
36895 @subsubheading @value{GDBN} Command
36897 The corresponding information is printed by @samp{info file} (among
36900 @subsubheading Example
36904 @subheading The @code{-target-list-parameters} Command
36905 @findex -target-list-parameters
36907 @subsubheading Synopsis
36910 -target-list-parameters
36916 @subsubheading @value{GDBN} Command
36920 @subsubheading Example
36923 @subheading The @code{-target-flash-erase} Command
36924 @findex -target-flash-erase
36926 @subsubheading Synopsis
36929 -target-flash-erase
36932 Erases all known flash memory regions on the target.
36934 The corresponding @value{GDBN} command is @samp{flash-erase}.
36936 The output is a list of flash regions that have been erased, with starting
36937 addresses and memory region sizes.
36941 -target-flash-erase
36942 ^done,erased-regions=@{address="0x0",size="0x40000"@}
36946 @subheading The @code{-target-select} Command
36947 @findex -target-select
36949 @subsubheading Synopsis
36952 -target-select @var{type} @var{parameters @dots{}}
36955 Connect @value{GDBN} to the remote target. This command takes two args:
36959 The type of target, for instance @samp{remote}, etc.
36960 @item @var{parameters}
36961 Device names, host names and the like. @xref{Target Commands, ,
36962 Commands for Managing Targets}, for more details.
36965 The output is a connection notification, followed by the address at
36966 which the target program is, in the following form:
36969 ^connected,addr="@var{address}",func="@var{function name}",
36970 args=[@var{arg list}]
36973 @subsubheading @value{GDBN} Command
36975 The corresponding @value{GDBN} command is @samp{target}.
36977 @subsubheading Example
36981 -target-select remote /dev/ttya
36982 ^connected,addr="0xfe00a300",func="??",args=[]
36986 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
36987 @node GDB/MI File Transfer Commands
36988 @section @sc{gdb/mi} File Transfer Commands
36991 @subheading The @code{-target-file-put} Command
36992 @findex -target-file-put
36994 @subsubheading Synopsis
36997 -target-file-put @var{hostfile} @var{targetfile}
37000 Copy file @var{hostfile} from the host system (the machine running
37001 @value{GDBN}) to @var{targetfile} on the target system.
37003 @subsubheading @value{GDBN} Command
37005 The corresponding @value{GDBN} command is @samp{remote put}.
37007 @subsubheading Example
37011 -target-file-put localfile remotefile
37017 @subheading The @code{-target-file-get} Command
37018 @findex -target-file-get
37020 @subsubheading Synopsis
37023 -target-file-get @var{targetfile} @var{hostfile}
37026 Copy file @var{targetfile} from the target system to @var{hostfile}
37027 on the host system.
37029 @subsubheading @value{GDBN} Command
37031 The corresponding @value{GDBN} command is @samp{remote get}.
37033 @subsubheading Example
37037 -target-file-get remotefile localfile
37043 @subheading The @code{-target-file-delete} Command
37044 @findex -target-file-delete
37046 @subsubheading Synopsis
37049 -target-file-delete @var{targetfile}
37052 Delete @var{targetfile} from the target system.
37054 @subsubheading @value{GDBN} Command
37056 The corresponding @value{GDBN} command is @samp{remote delete}.
37058 @subsubheading Example
37062 -target-file-delete remotefile
37068 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
37069 @node GDB/MI Ada Exceptions Commands
37070 @section Ada Exceptions @sc{gdb/mi} Commands
37072 @subheading The @code{-info-ada-exceptions} Command
37073 @findex -info-ada-exceptions
37075 @subsubheading Synopsis
37078 -info-ada-exceptions [ @var{regexp}]
37081 List all Ada exceptions defined within the program being debugged.
37082 With a regular expression @var{regexp}, only those exceptions whose
37083 names match @var{regexp} are listed.
37085 @subsubheading @value{GDBN} Command
37087 The corresponding @value{GDBN} command is @samp{info exceptions}.
37089 @subsubheading Result
37091 The result is a table of Ada exceptions. The following columns are
37092 defined for each exception:
37096 The name of the exception.
37099 The address of the exception.
37103 @subsubheading Example
37106 -info-ada-exceptions aint
37107 ^done,ada-exceptions=@{nr_rows="2",nr_cols="2",
37108 hdr=[@{width="1",alignment="-1",col_name="name",colhdr="Name"@},
37109 @{width="1",alignment="-1",col_name="address",colhdr="Address"@}],
37110 body=[@{name="constraint_error",address="0x0000000000613da0"@},
37111 @{name="const.aint_global_e",address="0x0000000000613b00"@}]@}
37114 @subheading Catching Ada Exceptions
37116 The commands describing how to ask @value{GDBN} to stop when a program
37117 raises an exception are described at @ref{Ada Exception GDB/MI
37118 Catchpoint Commands}.
37121 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
37122 @node GDB/MI Support Commands
37123 @section @sc{gdb/mi} Support Commands
37125 Since new commands and features get regularly added to @sc{gdb/mi},
37126 some commands are available to help front-ends query the debugger
37127 about support for these capabilities. Similarly, it is also possible
37128 to query @value{GDBN} about target support of certain features.
37130 @subheading The @code{-info-gdb-mi-command} Command
37131 @cindex @code{-info-gdb-mi-command}
37132 @findex -info-gdb-mi-command
37134 @subsubheading Synopsis
37137 -info-gdb-mi-command @var{cmd_name}
37140 Query support for the @sc{gdb/mi} command named @var{cmd_name}.
37142 Note that the dash (@code{-}) starting all @sc{gdb/mi} commands
37143 is technically not part of the command name (@pxref{GDB/MI Input
37144 Syntax}), and thus should be omitted in @var{cmd_name}. However,
37145 for ease of use, this command also accepts the form with the leading
37148 @subsubheading @value{GDBN} Command
37150 There is no corresponding @value{GDBN} command.
37152 @subsubheading Result
37154 The result is a tuple. There is currently only one field:
37158 This field is equal to @code{"true"} if the @sc{gdb/mi} command exists,
37159 @code{"false"} otherwise.
37163 @subsubheading Example
37165 Here is an example where the @sc{gdb/mi} command does not exist:
37168 -info-gdb-mi-command unsupported-command
37169 ^done,command=@{exists="false"@}
37173 And here is an example where the @sc{gdb/mi} command is known
37177 -info-gdb-mi-command symbol-list-lines
37178 ^done,command=@{exists="true"@}
37181 @subheading The @code{-list-features} Command
37182 @findex -list-features
37183 @cindex supported @sc{gdb/mi} features, list
37185 Returns a list of particular features of the MI protocol that
37186 this version of gdb implements. A feature can be a command,
37187 or a new field in an output of some command, or even an
37188 important bugfix. While a frontend can sometimes detect presence
37189 of a feature at runtime, it is easier to perform detection at debugger
37192 The command returns a list of strings, with each string naming an
37193 available feature. Each returned string is just a name, it does not
37194 have any internal structure. The list of possible feature names
37200 (gdb) -list-features
37201 ^done,result=["feature1","feature2"]
37204 The current list of features is:
37207 @item frozen-varobjs
37208 Indicates support for the @code{-var-set-frozen} command, as well
37209 as possible presence of the @code{frozen} field in the output
37210 of @code{-varobj-create}.
37211 @item pending-breakpoints
37212 Indicates support for the @option{-f} option to the @code{-break-insert}
37215 Indicates Python scripting support, Python-based
37216 pretty-printing commands, and possible presence of the
37217 @samp{display_hint} field in the output of @code{-var-list-children}
37219 Indicates support for the @code{-thread-info} command.
37220 @item data-read-memory-bytes
37221 Indicates support for the @code{-data-read-memory-bytes} and the
37222 @code{-data-write-memory-bytes} commands.
37223 @item breakpoint-notifications
37224 Indicates that changes to breakpoints and breakpoints created via the
37225 CLI will be announced via async records.
37226 @item ada-task-info
37227 Indicates support for the @code{-ada-task-info} command.
37228 @item language-option
37229 Indicates that all @sc{gdb/mi} commands accept the @option{--language}
37230 option (@pxref{Context management}).
37231 @item info-gdb-mi-command
37232 Indicates support for the @code{-info-gdb-mi-command} command.
37233 @item undefined-command-error-code
37234 Indicates support for the "undefined-command" error code in error result
37235 records, produced when trying to execute an undefined @sc{gdb/mi} command
37236 (@pxref{GDB/MI Result Records}).
37237 @item exec-run-start-option
37238 Indicates that the @code{-exec-run} command supports the @option{--start}
37239 option (@pxref{GDB/MI Program Execution}).
37240 @item data-disassemble-a-option
37241 Indicates that the @code{-data-disassemble} command supports the @option{-a}
37242 option (@pxref{GDB/MI Data Manipulation}).
37245 @subheading The @code{-list-target-features} Command
37246 @findex -list-target-features
37248 Returns a list of particular features that are supported by the
37249 target. Those features affect the permitted MI commands, but
37250 unlike the features reported by the @code{-list-features} command, the
37251 features depend on which target GDB is using at the moment. Whenever
37252 a target can change, due to commands such as @code{-target-select},
37253 @code{-target-attach} or @code{-exec-run}, the list of target features
37254 may change, and the frontend should obtain it again.
37258 (gdb) -list-target-features
37259 ^done,result=["async"]
37262 The current list of features is:
37266 Indicates that the target is capable of asynchronous command
37267 execution, which means that @value{GDBN} will accept further commands
37268 while the target is running.
37271 Indicates that the target is capable of reverse execution.
37272 @xref{Reverse Execution}, for more information.
37276 @c %%%%%%%%%%%%%%%%%%%%%%%%%%%% SECTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
37277 @node GDB/MI Miscellaneous Commands
37278 @section Miscellaneous @sc{gdb/mi} Commands
37280 @c @subheading -gdb-complete
37282 @subheading The @code{-gdb-exit} Command
37285 @subsubheading Synopsis
37291 Exit @value{GDBN} immediately.
37293 @subsubheading @value{GDBN} Command
37295 Approximately corresponds to @samp{quit}.
37297 @subsubheading Example
37307 @subheading The @code{-exec-abort} Command
37308 @findex -exec-abort
37310 @subsubheading Synopsis
37316 Kill the inferior running program.
37318 @subsubheading @value{GDBN} Command
37320 The corresponding @value{GDBN} command is @samp{kill}.
37322 @subsubheading Example
37327 @subheading The @code{-gdb-set} Command
37330 @subsubheading Synopsis
37336 Set an internal @value{GDBN} variable.
37337 @c IS THIS A DOLLAR VARIABLE? OR SOMETHING LIKE ANNOTATE ?????
37339 @subsubheading @value{GDBN} Command
37341 The corresponding @value{GDBN} command is @samp{set}.
37343 @subsubheading Example
37353 @subheading The @code{-gdb-show} Command
37356 @subsubheading Synopsis
37362 Show the current value of a @value{GDBN} variable.
37364 @subsubheading @value{GDBN} Command
37366 The corresponding @value{GDBN} command is @samp{show}.
37368 @subsubheading Example
37377 @c @subheading -gdb-source
37380 @subheading The @code{-gdb-version} Command
37381 @findex -gdb-version
37383 @subsubheading Synopsis
37389 Show version information for @value{GDBN}. Used mostly in testing.
37391 @subsubheading @value{GDBN} Command
37393 The @value{GDBN} equivalent is @samp{show version}. @value{GDBN} by
37394 default shows this information when you start an interactive session.
37396 @subsubheading Example
37398 @c This example modifies the actual output from GDB to avoid overfull
37404 ~Copyright 2000 Free Software Foundation, Inc.
37405 ~GDB is free software, covered by the GNU General Public License, and
37406 ~you are welcome to change it and/or distribute copies of it under
37407 ~ certain conditions.
37408 ~Type "show copying" to see the conditions.
37409 ~There is absolutely no warranty for GDB. Type "show warranty" for
37411 ~This GDB was configured as
37412 "--host=sparc-sun-solaris2.5.1 --target=ppc-eabi".
37417 @subheading The @code{-list-thread-groups} Command
37418 @findex -list-thread-groups
37420 @subheading Synopsis
37423 -list-thread-groups [ --available ] [ --recurse 1 ] [ @var{group} ... ]
37426 Lists thread groups (@pxref{Thread groups}). When a single thread
37427 group is passed as the argument, lists the children of that group.
37428 When several thread group are passed, lists information about those
37429 thread groups. Without any parameters, lists information about all
37430 top-level thread groups.
37432 Normally, thread groups that are being debugged are reported.
37433 With the @samp{--available} option, @value{GDBN} reports thread groups
37434 available on the target.
37436 The output of this command may have either a @samp{threads} result or
37437 a @samp{groups} result. The @samp{thread} result has a list of tuples
37438 as value, with each tuple describing a thread (@pxref{GDB/MI Thread
37439 Information}). The @samp{groups} result has a list of tuples as value,
37440 each tuple describing a thread group. If top-level groups are
37441 requested (that is, no parameter is passed), or when several groups
37442 are passed, the output always has a @samp{groups} result. The format
37443 of the @samp{group} result is described below.
37445 To reduce the number of roundtrips it's possible to list thread groups
37446 together with their children, by passing the @samp{--recurse} option
37447 and the recursion depth. Presently, only recursion depth of 1 is
37448 permitted. If this option is present, then every reported thread group
37449 will also include its children, either as @samp{group} or
37450 @samp{threads} field.
37452 In general, any combination of option and parameters is permitted, with
37453 the following caveats:
37457 When a single thread group is passed, the output will typically
37458 be the @samp{threads} result. Because threads may not contain
37459 anything, the @samp{recurse} option will be ignored.
37462 When the @samp{--available} option is passed, limited information may
37463 be available. In particular, the list of threads of a process might
37464 be inaccessible. Further, specifying specific thread groups might
37465 not give any performance advantage over listing all thread groups.
37466 The frontend should assume that @samp{-list-thread-groups --available}
37467 is always an expensive operation and cache the results.
37471 The @samp{groups} result is a list of tuples, where each tuple may
37472 have the following fields:
37476 Identifier of the thread group. This field is always present.
37477 The identifier is an opaque string; frontends should not try to
37478 convert it to an integer, even though it might look like one.
37481 The type of the thread group. At present, only @samp{process} is a
37485 The target-specific process identifier. This field is only present
37486 for thread groups of type @samp{process} and only if the process exists.
37489 The exit code of this group's last exited thread, formatted in octal.
37490 This field is only present for thread groups of type @samp{process} and
37491 only if the process is not running.
37494 The number of children this thread group has. This field may be
37495 absent for an available thread group.
37498 This field has a list of tuples as value, each tuple describing a
37499 thread. It may be present if the @samp{--recurse} option is
37500 specified, and it's actually possible to obtain the threads.
37503 This field is a list of integers, each identifying a core that one
37504 thread of the group is running on. This field may be absent if
37505 such information is not available.
37508 The name of the executable file that corresponds to this thread group.
37509 The field is only present for thread groups of type @samp{process},
37510 and only if there is a corresponding executable file.
37514 @subheading Example
37518 -list-thread-groups
37519 ^done,groups=[@{id="17",type="process",pid="yyy",num_children="2"@}]
37520 -list-thread-groups 17
37521 ^done,threads=[@{id="2",target-id="Thread 0xb7e14b90 (LWP 21257)",
37522 frame=@{level="0",addr="0xffffe410",func="__kernel_vsyscall",args=[]@},state="running"@},
37523 @{id="1",target-id="Thread 0xb7e156b0 (LWP 21254)",
37524 frame=@{level="0",addr="0x0804891f",func="foo",args=[@{name="i",value="10"@}],
37525 file="/tmp/a.c",fullname="/tmp/a.c",line="158",arch="i386:x86_64"@},state="running"@}]]
37526 -list-thread-groups --available
37527 ^done,groups=[@{id="17",type="process",pid="yyy",num_children="2",cores=[1,2]@}]
37528 -list-thread-groups --available --recurse 1
37529 ^done,groups=[@{id="17", types="process",pid="yyy",num_children="2",cores=[1,2],
37530 threads=[@{id="1",target-id="Thread 0xb7e14b90",cores=[1]@},
37531 @{id="2",target-id="Thread 0xb7e14b90",cores=[2]@}]@},..]
37532 -list-thread-groups --available --recurse 1 17 18
37533 ^done,groups=[@{id="17", types="process",pid="yyy",num_children="2",cores=[1,2],
37534 threads=[@{id="1",target-id="Thread 0xb7e14b90",cores=[1]@},
37535 @{id="2",target-id="Thread 0xb7e14b90",cores=[2]@}]@},...]
37538 @subheading The @code{-info-os} Command
37541 @subsubheading Synopsis
37544 -info-os [ @var{type} ]
37547 If no argument is supplied, the command returns a table of available
37548 operating-system-specific information types. If one of these types is
37549 supplied as an argument @var{type}, then the command returns a table
37550 of data of that type.
37552 The types of information available depend on the target operating
37555 @subsubheading @value{GDBN} Command
37557 The corresponding @value{GDBN} command is @samp{info os}.
37559 @subsubheading Example
37561 When run on a @sc{gnu}/Linux system, the output will look something
37567 ^done,OSDataTable=@{nr_rows="10",nr_cols="3",
37568 hdr=[@{width="10",alignment="-1",col_name="col0",colhdr="Type"@},
37569 @{width="10",alignment="-1",col_name="col1",colhdr="Description"@},
37570 @{width="10",alignment="-1",col_name="col2",colhdr="Title"@}],
37571 body=[item=@{col0="cpus",col1="Listing of all cpus/cores on the system",
37573 item=@{col0="files",col1="Listing of all file descriptors",
37574 col2="File descriptors"@},
37575 item=@{col0="modules",col1="Listing of all loaded kernel modules",
37576 col2="Kernel modules"@},
37577 item=@{col0="msg",col1="Listing of all message queues",
37578 col2="Message queues"@},
37579 item=@{col0="processes",col1="Listing of all processes",
37580 col2="Processes"@},
37581 item=@{col0="procgroups",col1="Listing of all process groups",
37582 col2="Process groups"@},
37583 item=@{col0="semaphores",col1="Listing of all semaphores",
37584 col2="Semaphores"@},
37585 item=@{col0="shm",col1="Listing of all shared-memory regions",
37586 col2="Shared-memory regions"@},
37587 item=@{col0="sockets",col1="Listing of all internet-domain sockets",
37589 item=@{col0="threads",col1="Listing of all threads",
37593 ^done,OSDataTable=@{nr_rows="190",nr_cols="4",
37594 hdr=[@{width="10",alignment="-1",col_name="col0",colhdr="pid"@},
37595 @{width="10",alignment="-1",col_name="col1",colhdr="user"@},
37596 @{width="10",alignment="-1",col_name="col2",colhdr="command"@},
37597 @{width="10",alignment="-1",col_name="col3",colhdr="cores"@}],
37598 body=[item=@{col0="1",col1="root",col2="/sbin/init",col3="0"@},
37599 item=@{col0="2",col1="root",col2="[kthreadd]",col3="1"@},
37600 item=@{col0="3",col1="root",col2="[ksoftirqd/0]",col3="0"@},
37602 item=@{col0="26446",col1="stan",col2="bash",col3="0"@},
37603 item=@{col0="28152",col1="stan",col2="bash",col3="1"@}]@}
37607 (Note that the MI output here includes a @code{"Title"} column that
37608 does not appear in command-line @code{info os}; this column is useful
37609 for MI clients that want to enumerate the types of data, such as in a
37610 popup menu, but is needless clutter on the command line, and
37611 @code{info os} omits it.)
37613 @subheading The @code{-add-inferior} Command
37614 @findex -add-inferior
37616 @subheading Synopsis
37622 Creates a new inferior (@pxref{Inferiors and Programs}). The created
37623 inferior is not associated with any executable. Such association may
37624 be established with the @samp{-file-exec-and-symbols} command
37625 (@pxref{GDB/MI File Commands}). The command response has a single
37626 field, @samp{inferior}, whose value is the identifier of the
37627 thread group corresponding to the new inferior.
37629 @subheading Example
37634 ^done,inferior="i3"
37637 @subheading The @code{-interpreter-exec} Command
37638 @findex -interpreter-exec
37640 @subheading Synopsis
37643 -interpreter-exec @var{interpreter} @var{command}
37645 @anchor{-interpreter-exec}
37647 Execute the specified @var{command} in the given @var{interpreter}.
37649 @subheading @value{GDBN} Command
37651 The corresponding @value{GDBN} command is @samp{interpreter-exec}.
37653 @subheading Example
37657 -interpreter-exec console "break main"
37658 &"During symbol reading, couldn't parse type; debugger out of date?.\n"
37659 &"During symbol reading, bad structure-type format.\n"
37660 ~"Breakpoint 1 at 0x8074fc6: file ../../src/gdb/main.c, line 743.\n"
37665 @subheading The @code{-inferior-tty-set} Command
37666 @findex -inferior-tty-set
37668 @subheading Synopsis
37671 -inferior-tty-set /dev/pts/1
37674 Set terminal for future runs of the program being debugged.
37676 @subheading @value{GDBN} Command
37678 The corresponding @value{GDBN} command is @samp{set inferior-tty} /dev/pts/1.
37680 @subheading Example
37684 -inferior-tty-set /dev/pts/1
37689 @subheading The @code{-inferior-tty-show} Command
37690 @findex -inferior-tty-show
37692 @subheading Synopsis
37698 Show terminal for future runs of program being debugged.
37700 @subheading @value{GDBN} Command
37702 The corresponding @value{GDBN} command is @samp{show inferior-tty}.
37704 @subheading Example
37708 -inferior-tty-set /dev/pts/1
37712 ^done,inferior_tty_terminal="/dev/pts/1"
37716 @subheading The @code{-enable-timings} Command
37717 @findex -enable-timings
37719 @subheading Synopsis
37722 -enable-timings [yes | no]
37725 Toggle the printing of the wallclock, user and system times for an MI
37726 command as a field in its output. This command is to help frontend
37727 developers optimize the performance of their code. No argument is
37728 equivalent to @samp{yes}.
37730 @subheading @value{GDBN} Command
37734 @subheading Example
37742 ^done,bkpt=@{number="1",type="breakpoint",disp="keep",enabled="y",
37743 addr="0x080484ed",func="main",file="myprog.c",
37744 fullname="/home/nickrob/myprog.c",line="73",thread-groups=["i1"],
37746 time=@{wallclock="0.05185",user="0.00800",system="0.00000"@}
37754 *stopped,reason="breakpoint-hit",disp="keep",bkptno="1",thread-id="0",
37755 frame=@{addr="0x080484ed",func="main",args=[@{name="argc",value="1"@},
37756 @{name="argv",value="0xbfb60364"@}],file="myprog.c",
37757 fullname="/home/nickrob/myprog.c",line="73",arch="i386:x86_64"@}
37761 @subheading The @code{-complete} Command
37764 @subheading Synopsis
37767 -complete @var{command}
37770 Show a list of completions for partially typed CLI @var{command}.
37772 This command is intended for @sc{gdb/mi} frontends that cannot use two separate
37773 CLI and MI channels --- for example: because of lack of PTYs like on Windows or
37774 because @value{GDBN} is used remotely via a SSH connection.
37778 The result consists of two or three fields:
37782 This field contains the completed @var{command}. If @var{command}
37783 has no known completions, this field is omitted.
37786 This field contains a (possibly empty) array of matches. It is always present.
37788 @item max_completions_reached
37789 This field contains @code{1} if number of known completions is above
37790 @code{max-completions} limit (@pxref{Completion}), otherwise it contains
37791 @code{0}. It is always present.
37795 @subheading @value{GDBN} Command
37797 The corresponding @value{GDBN} command is @samp{complete}.
37799 @subheading Example
37804 ^done,completion="break",
37805 matches=["break","break-range"],
37806 max_completions_reached="0"
37809 ^done,completion="b ma",
37810 matches=["b madvise","b main"],max_completions_reached="0"
37812 -complete "b push_b"
37813 ^done,completion="b push_back(",
37815 "b A::push_back(void*)",
37816 "b std::string::push_back(char)",
37817 "b std::vector<int, std::allocator<int> >::push_back(int&&)"],
37818 max_completions_reached="0"
37820 -complete "nonexist"
37821 ^done,matches=[],max_completions_reached="0"
37827 @chapter @value{GDBN} Annotations
37829 This chapter describes annotations in @value{GDBN}. Annotations were
37830 designed to interface @value{GDBN} to graphical user interfaces or other
37831 similar programs which want to interact with @value{GDBN} at a
37832 relatively high level.
37834 The annotation mechanism has largely been superseded by @sc{gdb/mi}
37838 This is Edition @value{EDITION}, @value{DATE}.
37842 * Annotations Overview:: What annotations are; the general syntax.
37843 * Server Prefix:: Issuing a command without affecting user state.
37844 * Prompting:: Annotations marking @value{GDBN}'s need for input.
37845 * Errors:: Annotations for error messages.
37846 * Invalidation:: Some annotations describe things now invalid.
37847 * Annotations for Running::
37848 Whether the program is running, how it stopped, etc.
37849 * Source Annotations:: Annotations describing source code.
37852 @node Annotations Overview
37853 @section What is an Annotation?
37854 @cindex annotations
37856 Annotations start with a newline character, two @samp{control-z}
37857 characters, and the name of the annotation. If there is no additional
37858 information associated with this annotation, the name of the annotation
37859 is followed immediately by a newline. If there is additional
37860 information, the name of the annotation is followed by a space, the
37861 additional information, and a newline. The additional information
37862 cannot contain newline characters.
37864 Any output not beginning with a newline and two @samp{control-z}
37865 characters denotes literal output from @value{GDBN}. Currently there is
37866 no need for @value{GDBN} to output a newline followed by two
37867 @samp{control-z} characters, but if there was such a need, the
37868 annotations could be extended with an @samp{escape} annotation which
37869 means those three characters as output.
37871 The annotation @var{level}, which is specified using the
37872 @option{--annotate} command line option (@pxref{Mode Options}), controls
37873 how much information @value{GDBN} prints together with its prompt,
37874 values of expressions, source lines, and other types of output. Level 0
37875 is for no annotations, level 1 is for use when @value{GDBN} is run as a
37876 subprocess of @sc{gnu} Emacs, level 3 is the maximum annotation suitable
37877 for programs that control @value{GDBN}, and level 2 annotations have
37878 been made obsolete (@pxref{Limitations, , Limitations of the Annotation
37879 Interface, annotate, GDB's Obsolete Annotations}).
37882 @kindex set annotate
37883 @item set annotate @var{level}
37884 The @value{GDBN} command @code{set annotate} sets the level of
37885 annotations to the specified @var{level}.
37887 @item show annotate
37888 @kindex show annotate
37889 Show the current annotation level.
37892 This chapter describes level 3 annotations.
37894 A simple example of starting up @value{GDBN} with annotations is:
37897 $ @kbd{gdb --annotate=3}
37899 Copyright 2003 Free Software Foundation, Inc.
37900 GDB is free software, covered by the GNU General Public License,
37901 and you are welcome to change it and/or distribute copies of it
37902 under certain conditions.
37903 Type "show copying" to see the conditions.
37904 There is absolutely no warranty for GDB. Type "show warranty"
37906 This GDB was configured as "i386-pc-linux-gnu"
37917 Here @samp{quit} is input to @value{GDBN}; the rest is output from
37918 @value{GDBN}. The three lines beginning @samp{^Z^Z} (where @samp{^Z}
37919 denotes a @samp{control-z} character) are annotations; the rest is
37920 output from @value{GDBN}.
37922 @node Server Prefix
37923 @section The Server Prefix
37924 @cindex server prefix
37926 If you prefix a command with @samp{server } then it will not affect
37927 the command history, nor will it affect @value{GDBN}'s notion of which
37928 command to repeat if @key{RET} is pressed on a line by itself. This
37929 means that commands can be run behind a user's back by a front-end in
37930 a transparent manner.
37932 The @code{server } prefix does not affect the recording of values into
37933 the value history; to print a value without recording it into the
37934 value history, use the @code{output} command instead of the
37935 @code{print} command.
37937 Using this prefix also disables confirmation requests
37938 (@pxref{confirmation requests}).
37941 @section Annotation for @value{GDBN} Input
37943 @cindex annotations for prompts
37944 When @value{GDBN} prompts for input, it annotates this fact so it is possible
37945 to know when to send output, when the output from a given command is
37948 Different kinds of input each have a different @dfn{input type}. Each
37949 input type has three annotations: a @code{pre-} annotation, which
37950 denotes the beginning of any prompt which is being output, a plain
37951 annotation, which denotes the end of the prompt, and then a @code{post-}
37952 annotation which denotes the end of any echo which may (or may not) be
37953 associated with the input. For example, the @code{prompt} input type
37954 features the following annotations:
37962 The input types are
37965 @findex pre-prompt annotation
37966 @findex prompt annotation
37967 @findex post-prompt annotation
37969 When @value{GDBN} is prompting for a command (the main @value{GDBN} prompt).
37971 @findex pre-commands annotation
37972 @findex commands annotation
37973 @findex post-commands annotation
37975 When @value{GDBN} prompts for a set of commands, like in the @code{commands}
37976 command. The annotations are repeated for each command which is input.
37978 @findex pre-overload-choice annotation
37979 @findex overload-choice annotation
37980 @findex post-overload-choice annotation
37981 @item overload-choice
37982 When @value{GDBN} wants the user to select between various overloaded functions.
37984 @findex pre-query annotation
37985 @findex query annotation
37986 @findex post-query annotation
37988 When @value{GDBN} wants the user to confirm a potentially dangerous operation.
37990 @findex pre-prompt-for-continue annotation
37991 @findex prompt-for-continue annotation
37992 @findex post-prompt-for-continue annotation
37993 @item prompt-for-continue
37994 When @value{GDBN} is asking the user to press return to continue. Note: Don't
37995 expect this to work well; instead use @code{set height 0} to disable
37996 prompting. This is because the counting of lines is buggy in the
37997 presence of annotations.
38002 @cindex annotations for errors, warnings and interrupts
38004 @findex quit annotation
38009 This annotation occurs right before @value{GDBN} responds to an interrupt.
38011 @findex error annotation
38016 This annotation occurs right before @value{GDBN} responds to an error.
38018 Quit and error annotations indicate that any annotations which @value{GDBN} was
38019 in the middle of may end abruptly. For example, if a
38020 @code{value-history-begin} annotation is followed by a @code{error}, one
38021 cannot expect to receive the matching @code{value-history-end}. One
38022 cannot expect not to receive it either, however; an error annotation
38023 does not necessarily mean that @value{GDBN} is immediately returning all the way
38026 @findex error-begin annotation
38027 A quit or error annotation may be preceded by
38033 Any output between that and the quit or error annotation is the error
38036 Warning messages are not yet annotated.
38037 @c If we want to change that, need to fix warning(), type_error(),
38038 @c range_error(), and possibly other places.
38041 @section Invalidation Notices
38043 @cindex annotations for invalidation messages
38044 The following annotations say that certain pieces of state may have
38048 @findex frames-invalid annotation
38049 @item ^Z^Zframes-invalid
38051 The frames (for example, output from the @code{backtrace} command) may
38054 @findex breakpoints-invalid annotation
38055 @item ^Z^Zbreakpoints-invalid
38057 The breakpoints may have changed. For example, the user just added or
38058 deleted a breakpoint.
38061 @node Annotations for Running
38062 @section Running the Program
38063 @cindex annotations for running programs
38065 @findex starting annotation
38066 @findex stopping annotation
38067 When the program starts executing due to a @value{GDBN} command such as
38068 @code{step} or @code{continue},
38074 is output. When the program stops,
38080 is output. Before the @code{stopped} annotation, a variety of
38081 annotations describe how the program stopped.
38084 @findex exited annotation
38085 @item ^Z^Zexited @var{exit-status}
38086 The program exited, and @var{exit-status} is the exit status (zero for
38087 successful exit, otherwise nonzero).
38089 @findex signalled annotation
38090 @findex signal-name annotation
38091 @findex signal-name-end annotation
38092 @findex signal-string annotation
38093 @findex signal-string-end annotation
38094 @item ^Z^Zsignalled
38095 The program exited with a signal. After the @code{^Z^Zsignalled}, the
38096 annotation continues:
38102 ^Z^Zsignal-name-end
38106 ^Z^Zsignal-string-end
38111 where @var{name} is the name of the signal, such as @code{SIGILL} or
38112 @code{SIGSEGV}, and @var{string} is the explanation of the signal, such
38113 as @code{Illegal Instruction} or @code{Segmentation fault}. The arguments
38114 @var{intro-text}, @var{middle-text}, and @var{end-text} are for the
38115 user's benefit and have no particular format.
38117 @findex signal annotation
38119 The syntax of this annotation is just like @code{signalled}, but @value{GDBN} is
38120 just saying that the program received the signal, not that it was
38121 terminated with it.
38123 @findex breakpoint annotation
38124 @item ^Z^Zbreakpoint @var{number}
38125 The program hit breakpoint number @var{number}.
38127 @findex watchpoint annotation
38128 @item ^Z^Zwatchpoint @var{number}
38129 The program hit watchpoint number @var{number}.
38132 @node Source Annotations
38133 @section Displaying Source
38134 @cindex annotations for source display
38136 @findex source annotation
38137 The following annotation is used instead of displaying source code:
38140 ^Z^Zsource @var{filename}:@var{line}:@var{character}:@var{middle}:@var{addr}
38143 where @var{filename} is an absolute file name indicating which source
38144 file, @var{line} is the line number within that file (where 1 is the
38145 first line in the file), @var{character} is the character position
38146 within the file (where 0 is the first character in the file) (for most
38147 debug formats this will necessarily point to the beginning of a line),
38148 @var{middle} is @samp{middle} if @var{addr} is in the middle of the
38149 line, or @samp{beg} if @var{addr} is at the beginning of the line, and
38150 @var{addr} is the address in the target program associated with the
38151 source which is being displayed. The @var{addr} is in the form @samp{0x}
38152 followed by one or more lowercase hex digits (note that this does not
38153 depend on the language).
38155 @node JIT Interface
38156 @chapter JIT Compilation Interface
38157 @cindex just-in-time compilation
38158 @cindex JIT compilation interface
38160 This chapter documents @value{GDBN}'s @dfn{just-in-time} (JIT) compilation
38161 interface. A JIT compiler is a program or library that generates native
38162 executable code at runtime and executes it, usually in order to achieve good
38163 performance while maintaining platform independence.
38165 Programs that use JIT compilation are normally difficult to debug because
38166 portions of their code are generated at runtime, instead of being loaded from
38167 object files, which is where @value{GDBN} normally finds the program's symbols
38168 and debug information. In order to debug programs that use JIT compilation,
38169 @value{GDBN} has an interface that allows the program to register in-memory
38170 symbol files with @value{GDBN} at runtime.
38172 If you are using @value{GDBN} to debug a program that uses this interface, then
38173 it should work transparently so long as you have not stripped the binary. If
38174 you are developing a JIT compiler, then the interface is documented in the rest
38175 of this chapter. At this time, the only known client of this interface is the
38178 Broadly speaking, the JIT interface mirrors the dynamic loader interface. The
38179 JIT compiler communicates with @value{GDBN} by writing data into a global
38180 variable and calling a function at a well-known symbol. When @value{GDBN}
38181 attaches, it reads a linked list of symbol files from the global variable to
38182 find existing code, and puts a breakpoint in the function so that it can find
38183 out about additional code.
38186 * Declarations:: Relevant C struct declarations
38187 * Registering Code:: Steps to register code
38188 * Unregistering Code:: Steps to unregister code
38189 * Custom Debug Info:: Emit debug information in a custom format
38193 @section JIT Declarations
38195 These are the relevant struct declarations that a C program should include to
38196 implement the interface:
38206 struct jit_code_entry
38208 struct jit_code_entry *next_entry;
38209 struct jit_code_entry *prev_entry;
38210 const char *symfile_addr;
38211 uint64_t symfile_size;
38214 struct jit_descriptor
38217 /* This type should be jit_actions_t, but we use uint32_t
38218 to be explicit about the bitwidth. */
38219 uint32_t action_flag;
38220 struct jit_code_entry *relevant_entry;
38221 struct jit_code_entry *first_entry;
38224 /* GDB puts a breakpoint in this function. */
38225 void __attribute__((noinline)) __jit_debug_register_code() @{ @};
38227 /* Make sure to specify the version statically, because the
38228 debugger may check the version before we can set it. */
38229 struct jit_descriptor __jit_debug_descriptor = @{ 1, 0, 0, 0 @};
38232 If the JIT is multi-threaded, then it is important that the JIT synchronize any
38233 modifications to this global data properly, which can easily be done by putting
38234 a global mutex around modifications to these structures.
38236 @node Registering Code
38237 @section Registering Code
38239 To register code with @value{GDBN}, the JIT should follow this protocol:
38243 Generate an object file in memory with symbols and other desired debug
38244 information. The file must include the virtual addresses of the sections.
38247 Create a code entry for the file, which gives the start and size of the symbol
38251 Add it to the linked list in the JIT descriptor.
38254 Point the relevant_entry field of the descriptor at the entry.
38257 Set @code{action_flag} to @code{JIT_REGISTER} and call
38258 @code{__jit_debug_register_code}.
38261 When @value{GDBN} is attached and the breakpoint fires, @value{GDBN} uses the
38262 @code{relevant_entry} pointer so it doesn't have to walk the list looking for
38263 new code. However, the linked list must still be maintained in order to allow
38264 @value{GDBN} to attach to a running process and still find the symbol files.
38266 @node Unregistering Code
38267 @section Unregistering Code
38269 If code is freed, then the JIT should use the following protocol:
38273 Remove the code entry corresponding to the code from the linked list.
38276 Point the @code{relevant_entry} field of the descriptor at the code entry.
38279 Set @code{action_flag} to @code{JIT_UNREGISTER} and call
38280 @code{__jit_debug_register_code}.
38283 If the JIT frees or recompiles code without unregistering it, then @value{GDBN}
38284 and the JIT will leak the memory used for the associated symbol files.
38286 @node Custom Debug Info
38287 @section Custom Debug Info
38288 @cindex custom JIT debug info
38289 @cindex JIT debug info reader
38291 Generating debug information in platform-native file formats (like ELF
38292 or COFF) may be an overkill for JIT compilers; especially if all the
38293 debug info is used for is displaying a meaningful backtrace. The
38294 issue can be resolved by having the JIT writers decide on a debug info
38295 format and also provide a reader that parses the debug info generated
38296 by the JIT compiler. This section gives a brief overview on writing
38297 such a parser. More specific details can be found in the source file
38298 @file{gdb/jit-reader.in}, which is also installed as a header at
38299 @file{@var{includedir}/gdb/jit-reader.h} for easy inclusion.
38301 The reader is implemented as a shared object (so this functionality is
38302 not available on platforms which don't allow loading shared objects at
38303 runtime). Two @value{GDBN} commands, @code{jit-reader-load} and
38304 @code{jit-reader-unload} are provided, to be used to load and unload
38305 the readers from a preconfigured directory. Once loaded, the shared
38306 object is used the parse the debug information emitted by the JIT
38310 * Using JIT Debug Info Readers:: How to use supplied readers correctly
38311 * Writing JIT Debug Info Readers:: Creating a debug-info reader
38314 @node Using JIT Debug Info Readers
38315 @subsection Using JIT Debug Info Readers
38316 @kindex jit-reader-load
38317 @kindex jit-reader-unload
38319 Readers can be loaded and unloaded using the @code{jit-reader-load}
38320 and @code{jit-reader-unload} commands.
38323 @item jit-reader-load @var{reader}
38324 Load the JIT reader named @var{reader}, which is a shared
38325 object specified as either an absolute or a relative file name. In
38326 the latter case, @value{GDBN} will try to load the reader from a
38327 pre-configured directory, usually @file{@var{libdir}/gdb/} on a UNIX
38328 system (here @var{libdir} is the system library directory, often
38329 @file{/usr/local/lib}).
38331 Only one reader can be active at a time; trying to load a second
38332 reader when one is already loaded will result in @value{GDBN}
38333 reporting an error. A new JIT reader can be loaded by first unloading
38334 the current one using @code{jit-reader-unload} and then invoking
38335 @code{jit-reader-load}.
38337 @item jit-reader-unload
38338 Unload the currently loaded JIT reader.
38342 @node Writing JIT Debug Info Readers
38343 @subsection Writing JIT Debug Info Readers
38344 @cindex writing JIT debug info readers
38346 As mentioned, a reader is essentially a shared object conforming to a
38347 certain ABI. This ABI is described in @file{jit-reader.h}.
38349 @file{jit-reader.h} defines the structures, macros and functions
38350 required to write a reader. It is installed (along with
38351 @value{GDBN}), in @file{@var{includedir}/gdb} where @var{includedir} is
38352 the system include directory.
38354 Readers need to be released under a GPL compatible license. A reader
38355 can be declared as released under such a license by placing the macro
38356 @code{GDB_DECLARE_GPL_COMPATIBLE_READER} in a source file.
38358 The entry point for readers is the symbol @code{gdb_init_reader},
38359 which is expected to be a function with the prototype
38361 @findex gdb_init_reader
38363 extern struct gdb_reader_funcs *gdb_init_reader (void);
38366 @cindex @code{struct gdb_reader_funcs}
38368 @code{struct gdb_reader_funcs} contains a set of pointers to callback
38369 functions. These functions are executed to read the debug info
38370 generated by the JIT compiler (@code{read}), to unwind stack frames
38371 (@code{unwind}) and to create canonical frame IDs
38372 (@code{get_frame_id}). It also has a callback that is called when the
38373 reader is being unloaded (@code{destroy}). The struct looks like this
38376 struct gdb_reader_funcs
38378 /* Must be set to GDB_READER_INTERFACE_VERSION. */
38379 int reader_version;
38381 /* For use by the reader. */
38384 gdb_read_debug_info *read;
38385 gdb_unwind_frame *unwind;
38386 gdb_get_frame_id *get_frame_id;
38387 gdb_destroy_reader *destroy;
38391 @cindex @code{struct gdb_symbol_callbacks}
38392 @cindex @code{struct gdb_unwind_callbacks}
38394 The callbacks are provided with another set of callbacks by
38395 @value{GDBN} to do their job. For @code{read}, these callbacks are
38396 passed in a @code{struct gdb_symbol_callbacks} and for @code{unwind}
38397 and @code{get_frame_id}, in a @code{struct gdb_unwind_callbacks}.
38398 @code{struct gdb_symbol_callbacks} has callbacks to create new object
38399 files and new symbol tables inside those object files. @code{struct
38400 gdb_unwind_callbacks} has callbacks to read registers off the current
38401 frame and to write out the values of the registers in the previous
38402 frame. Both have a callback (@code{target_read}) to read bytes off the
38403 target's address space.
38405 @node In-Process Agent
38406 @chapter In-Process Agent
38407 @cindex debugging agent
38408 The traditional debugging model is conceptually low-speed, but works fine,
38409 because most bugs can be reproduced in debugging-mode execution. However,
38410 as multi-core or many-core processors are becoming mainstream, and
38411 multi-threaded programs become more and more popular, there should be more
38412 and more bugs that only manifest themselves at normal-mode execution, for
38413 example, thread races, because debugger's interference with the program's
38414 timing may conceal the bugs. On the other hand, in some applications,
38415 it is not feasible for the debugger to interrupt the program's execution
38416 long enough for the developer to learn anything helpful about its behavior.
38417 If the program's correctness depends on its real-time behavior, delays
38418 introduced by a debugger might cause the program to fail, even when the
38419 code itself is correct. It is useful to be able to observe the program's
38420 behavior without interrupting it.
38422 Therefore, traditional debugging model is too intrusive to reproduce
38423 some bugs. In order to reduce the interference with the program, we can
38424 reduce the number of operations performed by debugger. The
38425 @dfn{In-Process Agent}, a shared library, is running within the same
38426 process with inferior, and is able to perform some debugging operations
38427 itself. As a result, debugger is only involved when necessary, and
38428 performance of debugging can be improved accordingly. Note that
38429 interference with program can be reduced but can't be removed completely,
38430 because the in-process agent will still stop or slow down the program.
38432 The in-process agent can interpret and execute Agent Expressions
38433 (@pxref{Agent Expressions}) during performing debugging operations. The
38434 agent expressions can be used for different purposes, such as collecting
38435 data in tracepoints, and condition evaluation in breakpoints.
38437 @anchor{Control Agent}
38438 You can control whether the in-process agent is used as an aid for
38439 debugging with the following commands:
38442 @kindex set agent on
38444 Causes the in-process agent to perform some operations on behalf of the
38445 debugger. Just which operations requested by the user will be done
38446 by the in-process agent depends on the its capabilities. For example,
38447 if you request to evaluate breakpoint conditions in the in-process agent,
38448 and the in-process agent has such capability as well, then breakpoint
38449 conditions will be evaluated in the in-process agent.
38451 @kindex set agent off
38452 @item set agent off
38453 Disables execution of debugging operations by the in-process agent. All
38454 of the operations will be performed by @value{GDBN}.
38458 Display the current setting of execution of debugging operations by
38459 the in-process agent.
38463 * In-Process Agent Protocol::
38466 @node In-Process Agent Protocol
38467 @section In-Process Agent Protocol
38468 @cindex in-process agent protocol
38470 The in-process agent is able to communicate with both @value{GDBN} and
38471 GDBserver (@pxref{In-Process Agent}). This section documents the protocol
38472 used for communications between @value{GDBN} or GDBserver and the IPA.
38473 In general, @value{GDBN} or GDBserver sends commands
38474 (@pxref{IPA Protocol Commands}) and data to in-process agent, and then
38475 in-process agent replies back with the return result of the command, or
38476 some other information. The data sent to in-process agent is composed
38477 of primitive data types, such as 4-byte or 8-byte type, and composite
38478 types, which are called objects (@pxref{IPA Protocol Objects}).
38481 * IPA Protocol Objects::
38482 * IPA Protocol Commands::
38485 @node IPA Protocol Objects
38486 @subsection IPA Protocol Objects
38487 @cindex ipa protocol objects
38489 The commands sent to and results received from agent may contain some
38490 complex data types called @dfn{objects}.
38492 The in-process agent is running on the same machine with @value{GDBN}
38493 or GDBserver, so it doesn't have to handle as much differences between
38494 two ends as remote protocol (@pxref{Remote Protocol}) tries to handle.
38495 However, there are still some differences of two ends in two processes:
38499 word size. On some 64-bit machines, @value{GDBN} or GDBserver can be
38500 compiled as a 64-bit executable, while in-process agent is a 32-bit one.
38502 ABI. Some machines may have multiple types of ABI, @value{GDBN} or
38503 GDBserver is compiled with one, and in-process agent is compiled with
38507 Here are the IPA Protocol Objects:
38511 agent expression object. It represents an agent expression
38512 (@pxref{Agent Expressions}).
38513 @anchor{agent expression object}
38515 tracepoint action object. It represents a tracepoint action
38516 (@pxref{Tracepoint Actions,,Tracepoint Action Lists}) to collect registers,
38517 memory, static trace data and to evaluate expression.
38518 @anchor{tracepoint action object}
38520 tracepoint object. It represents a tracepoint (@pxref{Tracepoints}).
38521 @anchor{tracepoint object}
38525 The following table describes important attributes of each IPA protocol
38528 @multitable @columnfractions .30 .20 .50
38529 @headitem Name @tab Size @tab Description
38530 @item @emph{agent expression object} @tab @tab
38531 @item length @tab 4 @tab length of bytes code
38532 @item byte code @tab @var{length} @tab contents of byte code
38533 @item @emph{tracepoint action for collecting memory} @tab @tab
38534 @item 'M' @tab 1 @tab type of tracepoint action
38535 @item addr @tab 8 @tab if @var{basereg} is @samp{-1}, @var{addr} is the
38536 address of the lowest byte to collect, otherwise @var{addr} is the offset
38537 of @var{basereg} for memory collecting.
38538 @item len @tab 8 @tab length of memory for collecting
38539 @item basereg @tab 4 @tab the register number containing the starting
38540 memory address for collecting.
38541 @item @emph{tracepoint action for collecting registers} @tab @tab
38542 @item 'R' @tab 1 @tab type of tracepoint action
38543 @item @emph{tracepoint action for collecting static trace data} @tab @tab
38544 @item 'L' @tab 1 @tab type of tracepoint action
38545 @item @emph{tracepoint action for expression evaluation} @tab @tab
38546 @item 'X' @tab 1 @tab type of tracepoint action
38547 @item agent expression @tab length of @tab @ref{agent expression object}
38548 @item @emph{tracepoint object} @tab @tab
38549 @item number @tab 4 @tab number of tracepoint
38550 @item address @tab 8 @tab address of tracepoint inserted on
38551 @item type @tab 4 @tab type of tracepoint
38552 @item enabled @tab 1 @tab enable or disable of tracepoint
38553 @item step_count @tab 8 @tab step
38554 @item pass_count @tab 8 @tab pass
38555 @item numactions @tab 4 @tab number of tracepoint actions
38556 @item hit count @tab 8 @tab hit count
38557 @item trace frame usage @tab 8 @tab trace frame usage
38558 @item compiled_cond @tab 8 @tab compiled condition
38559 @item orig_size @tab 8 @tab orig size
38560 @item condition @tab 4 if condition is NULL otherwise length of
38561 @ref{agent expression object}
38562 @tab zero if condition is NULL, otherwise is
38563 @ref{agent expression object}
38564 @item actions @tab variable
38565 @tab numactions number of @ref{tracepoint action object}
38568 @node IPA Protocol Commands
38569 @subsection IPA Protocol Commands
38570 @cindex ipa protocol commands
38572 The spaces in each command are delimiters to ease reading this commands
38573 specification. They don't exist in real commands.
38577 @item FastTrace:@var{tracepoint_object} @var{gdb_jump_pad_head}
38578 Installs a new fast tracepoint described by @var{tracepoint_object}
38579 (@pxref{tracepoint object}). The @var{gdb_jump_pad_head}, 8-byte long, is the
38580 head of @dfn{jumppad}, which is used to jump to data collection routine
38585 @item OK @var{target_address} @var{gdb_jump_pad_head} @var{fjump_size} @var{fjump}
38586 @var{target_address} is address of tracepoint in the inferior.
38587 The @var{gdb_jump_pad_head} is updated head of jumppad. Both of
38588 @var{target_address} and @var{gdb_jump_pad_head} are 8-byte long.
38589 The @var{fjump} contains a sequence of instructions jump to jumppad entry.
38590 The @var{fjump_size}, 4-byte long, is the size of @var{fjump}.
38597 Closes the in-process agent. This command is sent when @value{GDBN} or GDBserver
38598 is about to kill inferiors.
38606 @item probe_marker_at:@var{address}
38607 Asks in-process agent to probe the marker at @var{address}.
38614 @item unprobe_marker_at:@var{address}
38615 Asks in-process agent to unprobe the marker at @var{address}.
38619 @chapter Reporting Bugs in @value{GDBN}
38620 @cindex bugs in @value{GDBN}
38621 @cindex reporting bugs in @value{GDBN}
38623 Your bug reports play an essential role in making @value{GDBN} reliable.
38625 Reporting a bug may help you by bringing a solution to your problem, or it
38626 may not. But in any case the principal function of a bug report is to help
38627 the entire community by making the next version of @value{GDBN} work better. Bug
38628 reports are your contribution to the maintenance of @value{GDBN}.
38630 In order for a bug report to serve its purpose, you must include the
38631 information that enables us to fix the bug.
38634 * Bug Criteria:: Have you found a bug?
38635 * Bug Reporting:: How to report bugs
38639 @section Have You Found a Bug?
38640 @cindex bug criteria
38642 If you are not sure whether you have found a bug, here are some guidelines:
38645 @cindex fatal signal
38646 @cindex debugger crash
38647 @cindex crash of debugger
38649 If the debugger gets a fatal signal, for any input whatever, that is a
38650 @value{GDBN} bug. Reliable debuggers never crash.
38652 @cindex error on valid input
38654 If @value{GDBN} produces an error message for valid input, that is a
38655 bug. (Note that if you're cross debugging, the problem may also be
38656 somewhere in the connection to the target.)
38658 @cindex invalid input
38660 If @value{GDBN} does not produce an error message for invalid input,
38661 that is a bug. However, you should note that your idea of
38662 ``invalid input'' might be our idea of ``an extension'' or ``support
38663 for traditional practice''.
38666 If you are an experienced user of debugging tools, your suggestions
38667 for improvement of @value{GDBN} are welcome in any case.
38670 @node Bug Reporting
38671 @section How to Report Bugs
38672 @cindex bug reports
38673 @cindex @value{GDBN} bugs, reporting
38675 A number of companies and individuals offer support for @sc{gnu} products.
38676 If you obtained @value{GDBN} from a support organization, we recommend you
38677 contact that organization first.
38679 You can find contact information for many support companies and
38680 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
38682 @c should add a web page ref...
38685 @ifset BUGURL_DEFAULT
38686 In any event, we also recommend that you submit bug reports for
38687 @value{GDBN}. The preferred method is to submit them directly using
38688 @uref{http://www.gnu.org/software/gdb/bugs/, @value{GDBN}'s Bugs web
38689 page}. Alternatively, the @email{bug-gdb@@gnu.org, e-mail gateway} can
38692 @strong{Do not send bug reports to @samp{info-gdb}, or to
38693 @samp{help-gdb}, or to any newsgroups.} Most users of @value{GDBN} do
38694 not want to receive bug reports. Those that do have arranged to receive
38697 The mailing list @samp{bug-gdb} has a newsgroup @samp{gnu.gdb.bug} which
38698 serves as a repeater. The mailing list and the newsgroup carry exactly
38699 the same messages. Often people think of posting bug reports to the
38700 newsgroup instead of mailing them. This appears to work, but it has one
38701 problem which can be crucial: a newsgroup posting often lacks a mail
38702 path back to the sender. Thus, if we need to ask for more information,
38703 we may be unable to reach you. For this reason, it is better to send
38704 bug reports to the mailing list.
38706 @ifclear BUGURL_DEFAULT
38707 In any event, we also recommend that you submit bug reports for
38708 @value{GDBN} to @value{BUGURL}.
38712 The fundamental principle of reporting bugs usefully is this:
38713 @strong{report all the facts}. If you are not sure whether to state a
38714 fact or leave it out, state it!
38716 Often people omit facts because they think they know what causes the
38717 problem and assume that some details do not matter. Thus, you might
38718 assume that the name of the variable you use in an example does not matter.
38719 Well, probably it does not, but one cannot be sure. Perhaps the bug is a
38720 stray memory reference which happens to fetch from the location where that
38721 name is stored in memory; perhaps, if the name were different, the contents
38722 of that location would fool the debugger into doing the right thing despite
38723 the bug. Play it safe and give a specific, complete example. That is the
38724 easiest thing for you to do, and the most helpful.
38726 Keep in mind that the purpose of a bug report is to enable us to fix the
38727 bug. It may be that the bug has been reported previously, but neither
38728 you nor we can know that unless your bug report is complete and
38731 Sometimes people give a few sketchy facts and ask, ``Does this ring a
38732 bell?'' Those bug reports are useless, and we urge everyone to
38733 @emph{refuse to respond to them} except to chide the sender to report
38736 To enable us to fix the bug, you should include all these things:
38740 The version of @value{GDBN}. @value{GDBN} announces it if you start
38741 with no arguments; you can also print it at any time using @code{show
38744 Without this, we will not know whether there is any point in looking for
38745 the bug in the current version of @value{GDBN}.
38748 The type of machine you are using, and the operating system name and
38752 The details of the @value{GDBN} build-time configuration.
38753 @value{GDBN} shows these details if you invoke it with the
38754 @option{--configuration} command-line option, or if you type
38755 @code{show configuration} at @value{GDBN}'s prompt.
38758 What compiler (and its version) was used to compile @value{GDBN}---e.g.@:
38759 ``@value{GCC}--2.8.1''.
38762 What compiler (and its version) was used to compile the program you are
38763 debugging---e.g.@: ``@value{GCC}--2.8.1'', or ``HP92453-01 A.10.32.03 HP
38764 C Compiler''. For @value{NGCC}, you can say @kbd{@value{GCC} --version}
38765 to get this information; for other compilers, see the documentation for
38769 The command arguments you gave the compiler to compile your example and
38770 observe the bug. For example, did you use @samp{-O}? To guarantee
38771 you will not omit something important, list them all. A copy of the
38772 Makefile (or the output from make) is sufficient.
38774 If we were to try to guess the arguments, we would probably guess wrong
38775 and then we might not encounter the bug.
38778 A complete input script, and all necessary source files, that will
38782 A description of what behavior you observe that you believe is
38783 incorrect. For example, ``It gets a fatal signal.''
38785 Of course, if the bug is that @value{GDBN} gets a fatal signal, then we
38786 will certainly notice it. But if the bug is incorrect output, we might
38787 not notice unless it is glaringly wrong. You might as well not give us
38788 a chance to make a mistake.
38790 Even if the problem you experience is a fatal signal, you should still
38791 say so explicitly. Suppose something strange is going on, such as, your
38792 copy of @value{GDBN} is out of synch, or you have encountered a bug in
38793 the C library on your system. (This has happened!) Your copy might
38794 crash and ours would not. If you told us to expect a crash, then when
38795 ours fails to crash, we would know that the bug was not happening for
38796 us. If you had not told us to expect a crash, then we would not be able
38797 to draw any conclusion from our observations.
38800 @cindex recording a session script
38801 To collect all this information, you can use a session recording program
38802 such as @command{script}, which is available on many Unix systems.
38803 Just run your @value{GDBN} session inside @command{script} and then
38804 include the @file{typescript} file with your bug report.
38806 Another way to record a @value{GDBN} session is to run @value{GDBN}
38807 inside Emacs and then save the entire buffer to a file.
38810 If you wish to suggest changes to the @value{GDBN} source, send us context
38811 diffs. If you even discuss something in the @value{GDBN} source, refer to
38812 it by context, not by line number.
38814 The line numbers in our development sources will not match those in your
38815 sources. Your line numbers would convey no useful information to us.
38819 Here are some things that are not necessary:
38823 A description of the envelope of the bug.
38825 Often people who encounter a bug spend a lot of time investigating
38826 which changes to the input file will make the bug go away and which
38827 changes will not affect it.
38829 This is often time consuming and not very useful, because the way we
38830 will find the bug is by running a single example under the debugger
38831 with breakpoints, not by pure deduction from a series of examples.
38832 We recommend that you save your time for something else.
38834 Of course, if you can find a simpler example to report @emph{instead}
38835 of the original one, that is a convenience for us. Errors in the
38836 output will be easier to spot, running under the debugger will take
38837 less time, and so on.
38839 However, simplification is not vital; if you do not want to do this,
38840 report the bug anyway and send us the entire test case you used.
38843 A patch for the bug.
38845 A patch for the bug does help us if it is a good one. But do not omit
38846 the necessary information, such as the test case, on the assumption that
38847 a patch is all we need. We might see problems with your patch and decide
38848 to fix the problem another way, or we might not understand it at all.
38850 Sometimes with a program as complicated as @value{GDBN} it is very hard to
38851 construct an example that will make the program follow a certain path
38852 through the code. If you do not send us the example, we will not be able
38853 to construct one, so we will not be able to verify that the bug is fixed.
38855 And if we cannot understand what bug you are trying to fix, or why your
38856 patch should be an improvement, we will not install it. A test case will
38857 help us to understand.
38860 A guess about what the bug is or what it depends on.
38862 Such guesses are usually wrong. Even we cannot guess right about such
38863 things without first using the debugger to find the facts.
38866 @c The readline documentation is distributed with the readline code
38867 @c and consists of the two following files:
38870 @c Use -I with makeinfo to point to the appropriate directory,
38871 @c environment var TEXINPUTS with TeX.
38872 @ifclear SYSTEM_READLINE
38873 @include rluser.texi
38874 @include hsuser.texi
38878 @appendix In Memoriam
38880 The @value{GDBN} project mourns the loss of the following long-time
38885 Fred was a long-standing contributor to @value{GDBN} (1991-2006), and
38886 to Free Software in general. Outside of @value{GDBN}, he was known in
38887 the Amiga world for his series of Fish Disks, and the GeekGadget project.
38889 @item Michael Snyder
38890 Michael was one of the Global Maintainers of the @value{GDBN} project,
38891 with contributions recorded as early as 1996, until 2011. In addition
38892 to his day to day participation, he was a large driving force behind
38893 adding Reverse Debugging to @value{GDBN}.
38896 Beyond their technical contributions to the project, they were also
38897 enjoyable members of the Free Software Community. We will miss them.
38899 @node Formatting Documentation
38900 @appendix Formatting Documentation
38902 @cindex @value{GDBN} reference card
38903 @cindex reference card
38904 The @value{GDBN} 4 release includes an already-formatted reference card, ready
38905 for printing with PostScript or Ghostscript, in the @file{gdb}
38906 subdirectory of the main source directory@footnote{In
38907 @file{gdb-@value{GDBVN}/gdb/refcard.ps} of the version @value{GDBVN}
38908 release.}. If you can use PostScript or Ghostscript with your printer,
38909 you can print the reference card immediately with @file{refcard.ps}.
38911 The release also includes the source for the reference card. You
38912 can format it, using @TeX{}, by typing:
38918 The @value{GDBN} reference card is designed to print in @dfn{landscape}
38919 mode on US ``letter'' size paper;
38920 that is, on a sheet 11 inches wide by 8.5 inches
38921 high. You will need to specify this form of printing as an option to
38922 your @sc{dvi} output program.
38924 @cindex documentation
38926 All the documentation for @value{GDBN} comes as part of the machine-readable
38927 distribution. The documentation is written in Texinfo format, which is
38928 a documentation system that uses a single source file to produce both
38929 on-line information and a printed manual. You can use one of the Info
38930 formatting commands to create the on-line version of the documentation
38931 and @TeX{} (or @code{texi2roff}) to typeset the printed version.
38933 @value{GDBN} includes an already formatted copy of the on-line Info
38934 version of this manual in the @file{gdb} subdirectory. The main Info
38935 file is @file{gdb-@value{GDBVN}/gdb/gdb.info}, and it refers to
38936 subordinate files matching @samp{gdb.info*} in the same directory. If
38937 necessary, you can print out these files, or read them with any editor;
38938 but they are easier to read using the @code{info} subsystem in @sc{gnu}
38939 Emacs or the standalone @code{info} program, available as part of the
38940 @sc{gnu} Texinfo distribution.
38942 If you want to format these Info files yourself, you need one of the
38943 Info formatting programs, such as @code{texinfo-format-buffer} or
38946 If you have @code{makeinfo} installed, and are in the top level
38947 @value{GDBN} source directory (@file{gdb-@value{GDBVN}}, in the case of
38948 version @value{GDBVN}), you can make the Info file by typing:
38955 If you want to typeset and print copies of this manual, you need @TeX{},
38956 a program to print its @sc{dvi} output files, and @file{texinfo.tex}, the
38957 Texinfo definitions file.
38959 @TeX{} is a typesetting program; it does not print files directly, but
38960 produces output files called @sc{dvi} files. To print a typeset
38961 document, you need a program to print @sc{dvi} files. If your system
38962 has @TeX{} installed, chances are it has such a program. The precise
38963 command to use depends on your system; @kbd{lpr -d} is common; another
38964 (for PostScript devices) is @kbd{dvips}. The @sc{dvi} print command may
38965 require a file name without any extension or a @samp{.dvi} extension.
38967 @TeX{} also requires a macro definitions file called
38968 @file{texinfo.tex}. This file tells @TeX{} how to typeset a document
38969 written in Texinfo format. On its own, @TeX{} cannot either read or
38970 typeset a Texinfo file. @file{texinfo.tex} is distributed with GDB
38971 and is located in the @file{gdb-@var{version-number}/texinfo}
38974 If you have @TeX{} and a @sc{dvi} printer program installed, you can
38975 typeset and print this manual. First switch to the @file{gdb}
38976 subdirectory of the main source directory (for example, to
38977 @file{gdb-@value{GDBVN}/gdb}) and type:
38983 Then give @file{gdb.dvi} to your @sc{dvi} printing program.
38985 @node Installing GDB
38986 @appendix Installing @value{GDBN}
38987 @cindex installation
38990 * Requirements:: Requirements for building @value{GDBN}
38991 * Running Configure:: Invoking the @value{GDBN} @file{configure} script
38992 * Separate Objdir:: Compiling @value{GDBN} in another directory
38993 * Config Names:: Specifying names for hosts and targets
38994 * Configure Options:: Summary of options for configure
38995 * System-wide configuration:: Having a system-wide init file
38999 @section Requirements for Building @value{GDBN}
39000 @cindex building @value{GDBN}, requirements for
39002 Building @value{GDBN} requires various tools and packages to be available.
39003 Other packages will be used only if they are found.
39005 @heading Tools/Packages Necessary for Building @value{GDBN}
39007 @item C@t{++}11 compiler
39008 @value{GDBN} is written in C@t{++}11. It should be buildable with any
39009 recent C@t{++}11 compiler, e.g.@: GCC.
39012 @value{GDBN}'s build system relies on features only found in the GNU
39013 make program. Other variants of @code{make} will not work.
39016 @heading Tools/Packages Optional for Building @value{GDBN}
39020 @value{GDBN} can use the Expat XML parsing library. This library may be
39021 included with your operating system distribution; if it is not, you
39022 can get the latest version from @url{http://expat.sourceforge.net}.
39023 The @file{configure} script will search for this library in several
39024 standard locations; if it is installed in an unusual path, you can
39025 use the @option{--with-libexpat-prefix} option to specify its location.
39031 Remote protocol memory maps (@pxref{Memory Map Format})
39033 Target descriptions (@pxref{Target Descriptions})
39035 Remote shared library lists (@xref{Library List Format},
39036 or alternatively @pxref{Library List Format for SVR4 Targets})
39038 MS-Windows shared libraries (@pxref{Shared Libraries})
39040 Traceframe info (@pxref{Traceframe Info Format})
39042 Branch trace (@pxref{Branch Trace Format},
39043 @pxref{Branch Trace Configuration Format})
39047 @value{GDBN} can be scripted using GNU Guile. @xref{Guile}. By
39048 default, @value{GDBN} will be compiled if the Guile libraries are
39049 installed and are found by @file{configure}. You can use the
39050 @code{--with-guile} option to request Guile, and pass either the Guile
39051 version number or the file name of the relevant @code{pkg-config}
39052 program to choose a particular version of Guile.
39055 @value{GDBN}'s features related to character sets (@pxref{Character
39056 Sets}) require a functioning @code{iconv} implementation. If you are
39057 on a GNU system, then this is provided by the GNU C Library. Some
39058 other systems also provide a working @code{iconv}.
39060 If @value{GDBN} is using the @code{iconv} program which is installed
39061 in a non-standard place, you will need to tell @value{GDBN} where to
39062 find it. This is done with @option{--with-iconv-bin} which specifies
39063 the directory that contains the @code{iconv} program. This program is
39064 run in order to make a list of the available character sets.
39066 On systems without @code{iconv}, you can install GNU Libiconv. If
39067 Libiconv is installed in a standard place, @value{GDBN} will
39068 automatically use it if it is needed. If you have previously
39069 installed Libiconv in a non-standard place, you can use the
39070 @option{--with-libiconv-prefix} option to @file{configure}.
39072 @value{GDBN}'s top-level @file{configure} and @file{Makefile} will
39073 arrange to build Libiconv if a directory named @file{libiconv} appears
39074 in the top-most source directory. If Libiconv is built this way, and
39075 if the operating system does not provide a suitable @code{iconv}
39076 implementation, then the just-built library will automatically be used
39077 by @value{GDBN}. One easy way to set this up is to download GNU
39078 Libiconv, unpack it inside the top-level directory of the @value{GDBN}
39079 source tree, and then rename the directory holding the Libiconv source
39080 code to @samp{libiconv}.
39083 @value{GDBN} can support debugging sections that are compressed with
39084 the LZMA library. @xref{MiniDebugInfo}. If this library is not
39085 included with your operating system, you can find it in the xz package
39086 at @url{http://tukaani.org/xz/}. If the LZMA library is available in
39087 the usual place, then the @file{configure} script will use it
39088 automatically. If it is installed in an unusual path, you can use the
39089 @option{--with-lzma-prefix} option to specify its location.
39093 @value{GDBN} can use the GNU MPFR multiple-precision floating-point
39094 library. This library may be included with your operating system
39095 distribution; if it is not, you can get the latest version from
39096 @url{http://www.mpfr.org}. The @file{configure} script will search
39097 for this library in several standard locations; if it is installed
39098 in an unusual path, you can use the @option{--with-libmpfr-prefix}
39099 option to specify its location.
39101 GNU MPFR is used to emulate target floating-point arithmetic during
39102 expression evaluation when the target uses different floating-point
39103 formats than the host. If GNU MPFR it is not available, @value{GDBN}
39104 will fall back to using host floating-point arithmetic.
39107 @value{GDBN} can be scripted using Python language. @xref{Python}.
39108 By default, @value{GDBN} will be compiled if the Python libraries are
39109 installed and are found by @file{configure}. You can use the
39110 @code{--with-python} option to request Python, and pass either the
39111 file name of the relevant @code{python} executable, or the name of the
39112 directory in which Python is installed, to choose a particular
39113 installation of Python.
39116 @cindex compressed debug sections
39117 @value{GDBN} will use the @samp{zlib} library, if available, to read
39118 compressed debug sections. Some linkers, such as GNU gold, are capable
39119 of producing binaries with compressed debug sections. If @value{GDBN}
39120 is compiled with @samp{zlib}, it will be able to read the debug
39121 information in such binaries.
39123 The @samp{zlib} library is likely included with your operating system
39124 distribution; if it is not, you can get the latest version from
39125 @url{http://zlib.net}.
39128 @node Running Configure
39129 @section Invoking the @value{GDBN} @file{configure} Script
39130 @cindex configuring @value{GDBN}
39131 @value{GDBN} comes with a @file{configure} script that automates the process
39132 of preparing @value{GDBN} for installation; you can then use @code{make} to
39133 build the @code{gdb} program.
39135 @c irrelevant in info file; it's as current as the code it lives with.
39136 @footnote{If you have a more recent version of @value{GDBN} than @value{GDBVN},
39137 look at the @file{README} file in the sources; we may have improved the
39138 installation procedures since publishing this manual.}
39141 The @value{GDBN} distribution includes all the source code you need for
39142 @value{GDBN} in a single directory, whose name is usually composed by
39143 appending the version number to @samp{gdb}.
39145 For example, the @value{GDBN} version @value{GDBVN} distribution is in the
39146 @file{gdb-@value{GDBVN}} directory. That directory contains:
39149 @item gdb-@value{GDBVN}/configure @r{(and supporting files)}
39150 script for configuring @value{GDBN} and all its supporting libraries
39152 @item gdb-@value{GDBVN}/gdb
39153 the source specific to @value{GDBN} itself
39155 @item gdb-@value{GDBVN}/bfd
39156 source for the Binary File Descriptor library
39158 @item gdb-@value{GDBVN}/include
39159 @sc{gnu} include files
39161 @item gdb-@value{GDBVN}/libiberty
39162 source for the @samp{-liberty} free software library
39164 @item gdb-@value{GDBVN}/opcodes
39165 source for the library of opcode tables and disassemblers
39167 @item gdb-@value{GDBVN}/readline
39168 source for the @sc{gnu} command-line interface
39171 There may be other subdirectories as well.
39173 The simplest way to configure and build @value{GDBN} is to run @file{configure}
39174 from the @file{gdb-@var{version-number}} source directory, which in
39175 this example is the @file{gdb-@value{GDBVN}} directory.
39177 First switch to the @file{gdb-@var{version-number}} source directory
39178 if you are not already in it; then run @file{configure}. Pass the
39179 identifier for the platform on which @value{GDBN} will run as an
39185 cd gdb-@value{GDBVN}
39190 Running @samp{configure} and then running @code{make} builds the
39191 included supporting libraries, then @code{gdb} itself. The configured
39192 source files, and the binaries, are left in the corresponding source
39196 @file{configure} is a Bourne-shell (@code{/bin/sh}) script; if your
39197 system does not recognize this automatically when you run a different
39198 shell, you may need to run @code{sh} on it explicitly:
39204 You should run the @file{configure} script from the top directory in the
39205 source tree, the @file{gdb-@var{version-number}} directory. If you run
39206 @file{configure} from one of the subdirectories, you will configure only
39207 that subdirectory. That is usually not what you want. In particular,
39208 if you run the first @file{configure} from the @file{gdb} subdirectory
39209 of the @file{gdb-@var{version-number}} directory, you will omit the
39210 configuration of @file{bfd}, @file{readline}, and other sibling
39211 directories of the @file{gdb} subdirectory. This leads to build errors
39212 about missing include files such as @file{bfd/bfd.h}.
39214 You can install @code{@value{GDBN}} anywhere. The best way to do this
39215 is to pass the @code{--prefix} option to @code{configure}, and then
39216 install it with @code{make install}.
39218 @node Separate Objdir
39219 @section Compiling @value{GDBN} in Another Directory
39221 If you want to run @value{GDBN} versions for several host or target machines,
39222 you need a different @code{gdb} compiled for each combination of
39223 host and target. @file{configure} is designed to make this easy by
39224 allowing you to generate each configuration in a separate subdirectory,
39225 rather than in the source directory. If your @code{make} program
39226 handles the @samp{VPATH} feature (@sc{gnu} @code{make} does), running
39227 @code{make} in each of these directories builds the @code{gdb}
39228 program specified there.
39230 To build @code{gdb} in a separate directory, run @file{configure}
39231 with the @samp{--srcdir} option to specify where to find the source.
39232 (You also need to specify a path to find @file{configure}
39233 itself from your working directory. If the path to @file{configure}
39234 would be the same as the argument to @samp{--srcdir}, you can leave out
39235 the @samp{--srcdir} option; it is assumed.)
39237 For example, with version @value{GDBVN}, you can build @value{GDBN} in a
39238 separate directory for a Sun 4 like this:
39242 cd gdb-@value{GDBVN}
39245 ../gdb-@value{GDBVN}/configure
39250 When @file{configure} builds a configuration using a remote source
39251 directory, it creates a tree for the binaries with the same structure
39252 (and using the same names) as the tree under the source directory. In
39253 the example, you'd find the Sun 4 library @file{libiberty.a} in the
39254 directory @file{gdb-sun4/libiberty}, and @value{GDBN} itself in
39255 @file{gdb-sun4/gdb}.
39257 Make sure that your path to the @file{configure} script has just one
39258 instance of @file{gdb} in it. If your path to @file{configure} looks
39259 like @file{../gdb-@value{GDBVN}/gdb/configure}, you are configuring only
39260 one subdirectory of @value{GDBN}, not the whole package. This leads to
39261 build errors about missing include files such as @file{bfd/bfd.h}.
39263 One popular reason to build several @value{GDBN} configurations in separate
39264 directories is to configure @value{GDBN} for cross-compiling (where
39265 @value{GDBN} runs on one machine---the @dfn{host}---while debugging
39266 programs that run on another machine---the @dfn{target}).
39267 You specify a cross-debugging target by
39268 giving the @samp{--target=@var{target}} option to @file{configure}.
39270 When you run @code{make} to build a program or library, you must run
39271 it in a configured directory---whatever directory you were in when you
39272 called @file{configure} (or one of its subdirectories).
39274 The @code{Makefile} that @file{configure} generates in each source
39275 directory also runs recursively. If you type @code{make} in a source
39276 directory such as @file{gdb-@value{GDBVN}} (or in a separate configured
39277 directory configured with @samp{--srcdir=@var{dirname}/gdb-@value{GDBVN}}), you
39278 will build all the required libraries, and then build GDB.
39280 When you have multiple hosts or targets configured in separate
39281 directories, you can run @code{make} on them in parallel (for example,
39282 if they are NFS-mounted on each of the hosts); they will not interfere
39286 @section Specifying Names for Hosts and Targets
39288 The specifications used for hosts and targets in the @file{configure}
39289 script are based on a three-part naming scheme, but some short predefined
39290 aliases are also supported. The full naming scheme encodes three pieces
39291 of information in the following pattern:
39294 @var{architecture}-@var{vendor}-@var{os}
39297 For example, you can use the alias @code{sun4} as a @var{host} argument,
39298 or as the value for @var{target} in a @code{--target=@var{target}}
39299 option. The equivalent full name is @samp{sparc-sun-sunos4}.
39301 The @file{configure} script accompanying @value{GDBN} does not provide
39302 any query facility to list all supported host and target names or
39303 aliases. @file{configure} calls the Bourne shell script
39304 @code{config.sub} to map abbreviations to full names; you can read the
39305 script, if you wish, or you can use it to test your guesses on
39306 abbreviations---for example:
39309 % sh config.sub i386-linux
39311 % sh config.sub alpha-linux
39312 alpha-unknown-linux-gnu
39313 % sh config.sub hp9k700
39315 % sh config.sub sun4
39316 sparc-sun-sunos4.1.1
39317 % sh config.sub sun3
39318 m68k-sun-sunos4.1.1
39319 % sh config.sub i986v
39320 Invalid configuration `i986v': machine `i986v' not recognized
39324 @code{config.sub} is also distributed in the @value{GDBN} source
39325 directory (@file{gdb-@value{GDBVN}}, for version @value{GDBVN}).
39327 @node Configure Options
39328 @section @file{configure} Options
39330 Here is a summary of the @file{configure} options and arguments that
39331 are most often useful for building @value{GDBN}. @file{configure}
39332 also has several other options not listed here. @inforef{Running
39333 configure scripts,,autoconf.info}, for a full
39334 explanation of @file{configure}.
39337 configure @r{[}--help@r{]}
39338 @r{[}--prefix=@var{dir}@r{]}
39339 @r{[}--exec-prefix=@var{dir}@r{]}
39340 @r{[}--srcdir=@var{dirname}@r{]}
39341 @r{[}--target=@var{target}@r{]}
39345 You may introduce options with a single @samp{-} rather than
39346 @samp{--} if you prefer; but you may abbreviate option names if you use
39351 Display a quick summary of how to invoke @file{configure}.
39353 @item --prefix=@var{dir}
39354 Configure the source to install programs and files under directory
39357 @item --exec-prefix=@var{dir}
39358 Configure the source to install programs under directory
39361 @c avoid splitting the warning from the explanation:
39363 @item --srcdir=@var{dirname}
39364 Use this option to make configurations in directories separate from the
39365 @value{GDBN} source directories. Among other things, you can use this to
39366 build (or maintain) several configurations simultaneously, in separate
39367 directories. @file{configure} writes configuration-specific files in
39368 the current directory, but arranges for them to use the source in the
39369 directory @var{dirname}. @file{configure} creates directories under
39370 the working directory in parallel to the source directories below
39373 @item --target=@var{target}
39374 Configure @value{GDBN} for cross-debugging programs running on the specified
39375 @var{target}. Without this option, @value{GDBN} is configured to debug
39376 programs that run on the same machine (@var{host}) as @value{GDBN} itself.
39378 There is no convenient way to generate a list of all available
39379 targets. Also see the @code{--enable-targets} option, below.
39382 There are many other options that are specific to @value{GDBN}. This
39383 lists just the most common ones; there are some very specialized
39384 options not described here.
39387 @item --enable-targets=@r{[}@var{target}@r{]}@dots{}
39388 @itemx --enable-targets=all
39389 Configure @value{GDBN} for cross-debugging programs running on the
39390 specified list of targets. The special value @samp{all} configures
39391 @value{GDBN} for debugging programs running on any target it supports.
39393 @item --with-gdb-datadir=@var{path}
39394 Set the @value{GDBN}-specific data directory. @value{GDBN} will look
39395 here for certain supporting files or scripts. This defaults to the
39396 @file{gdb} subdirectory of @samp{datadir} (which can be set using
39399 @item --with-relocated-sources=@var{dir}
39400 Sets up the default source path substitution rule so that directory
39401 names recorded in debug information will be automatically adjusted for
39402 any directory under @var{dir}. @var{dir} should be a subdirectory of
39403 @value{GDBN}'s configured prefix, the one mentioned in the
39404 @code{--prefix} or @code{--exec-prefix} options to configure. This
39405 option is useful if GDB is supposed to be moved to a different place
39408 @item --enable-64-bit-bfd
39409 Enable 64-bit support in BFD on 32-bit hosts.
39411 @item --disable-gdbmi
39412 Build @value{GDBN} without the GDB/MI machine interface
39416 Build @value{GDBN} with the text-mode full-screen user interface
39417 (TUI). Requires a curses library (ncurses and cursesX are also
39420 @item --with-curses
39421 Use the curses library instead of the termcap library, for text-mode
39422 terminal operations.
39424 @item --with-libunwind-ia64
39425 Use the libunwind library for unwinding function call stack on ia64
39426 target platforms. See http://www.nongnu.org/libunwind/index.html for
39429 @item --with-system-readline
39430 Use the readline library installed on the host, rather than the
39431 library supplied as part of @value{GDBN}. Readline 7 or newer is
39432 required; this is enforced by the build system.
39434 @item --with-system-zlib
39435 Use the zlib library installed on the host, rather than the library
39436 supplied as part of @value{GDBN}.
39439 Build @value{GDBN} with Expat, a library for XML parsing. (Done by
39440 default if libexpat is installed and found at configure time.) This
39441 library is used to read XML files supplied with @value{GDBN}. If it
39442 is unavailable, some features, such as remote protocol memory maps,
39443 target descriptions, and shared library lists, that are based on XML
39444 files, will not be available in @value{GDBN}. If your host does not
39445 have libexpat installed, you can get the latest version from
39446 `http://expat.sourceforge.net'.
39448 @item --with-libiconv-prefix@r{[}=@var{dir}@r{]}
39450 Build @value{GDBN} with GNU libiconv, a character set encoding
39451 conversion library. This is not done by default, as on GNU systems
39452 the @code{iconv} that is built in to the C library is sufficient. If
39453 your host does not have a working @code{iconv}, you can get the latest
39454 version of GNU iconv from `https://www.gnu.org/software/libiconv/'.
39456 @value{GDBN}'s build system also supports building GNU libiconv as
39457 part of the overall build. @xref{Requirements}.
39460 Build @value{GDBN} with LZMA, a compression library. (Done by default
39461 if liblzma is installed and found at configure time.) LZMA is used by
39462 @value{GDBN}'s "mini debuginfo" feature, which is only useful on
39463 platforms using the ELF object file format. If your host does not
39464 have liblzma installed, you can get the latest version from
39465 `https://tukaani.org/xz/'.
39468 Build @value{GDBN} with GNU MPFR, a library for multiple-precision
39469 floating-point computation with correct rounding. (Done by default if
39470 GNU MPFR is installed and found at configure time.) This library is
39471 used to emulate target floating-point arithmetic during expression
39472 evaluation when the target uses different floating-point formats than
39473 the host. If GNU MPFR is not available, @value{GDBN} will fall back
39474 to using host floating-point arithmetic. If your host does not have
39475 GNU MPFR installed, you can get the latest version from
39476 `http://www.mpfr.org'.
39478 @item --with-python@r{[}=@var{python}@r{]}
39479 Build @value{GDBN} with Python scripting support. (Done by default if
39480 libpython is present and found at configure time.) Python makes
39481 @value{GDBN} scripting much more powerful than the restricted CLI
39482 scripting language. If your host does not have Python installed, you
39483 can find it on `http://www.python.org/download/'. The oldest version
39484 of Python supported by GDB is 2.6. The optional argument @var{python}
39485 is used to find the Python headers and libraries. It can be either
39486 the name of a Python executable, or the name of the directory in which
39487 Python is installed.
39489 @item --with-guile[=GUILE]'
39490 Build @value{GDBN} with GNU Guile scripting support. (Done by default
39491 if libguile is present and found at configure time.) If your host
39492 does not have Guile installed, you can find it at
39493 `https://www.gnu.org/software/guile/'. The optional argument GUILE
39494 can be a version number, which will cause @code{configure} to try to
39495 use that version of Guile; or the file name of a @code{pkg-config}
39496 executable, which will be queried to find the information needed to
39497 compile and link against Guile.
39499 @item --without-included-regex
39500 Don't use the regex library included with @value{GDBN} (as part of the
39501 libiberty library). This is the default on hosts with version 2 of
39504 @item --with-sysroot=@var{dir}
39505 Use @var{dir} as the default system root directory for libraries whose
39506 file names begin with @file{/lib}' or @file{/usr/lib'}. (The value of
39507 @var{dir} can be modified at run time by using the @command{set
39508 sysroot} command.) If @var{dir} is under the @value{GDBN} configured
39509 prefix (set with @code{--prefix} or @code{--exec-prefix options}, the
39510 default system root will be automatically adjusted if and when
39511 @value{GDBN} is moved to a different location.
39513 @item --with-system-gdbinit=@var{file}
39514 Configure @value{GDBN} to automatically load a system-wide init file.
39515 @var{file} should be an absolute file name. If @var{file} is in a
39516 directory under the configured prefix, and @value{GDBN} is moved to
39517 another location after being built, the location of the system-wide
39518 init file will be adjusted accordingly.
39520 @item --with-system-gdbinit-dir=@var{directory}
39521 Configure @value{GDBN} to automatically load init files from a
39522 system-wide directory. @var{directory} should be an absolute directory
39523 name. If @var{directory} is in a directory under the configured
39524 prefix, and @value{GDBN} is moved to another location after being
39525 built, the location of the system-wide init directory will be
39526 adjusted accordingly.
39528 @item --enable-build-warnings
39529 When building the @value{GDBN} sources, ask the compiler to warn about
39530 any code which looks even vaguely suspicious. It passes many
39531 different warning flags, depending on the exact version of the
39532 compiler you are using.
39534 @item --enable-werror
39535 Treat compiler warnings as werrors. It adds the @code{-Werror} flag
39536 to the compiler, which will fail the compilation if the compiler
39537 outputs any warning messages.
39539 @item --enable-ubsan
39540 Enable the GCC undefined behavior sanitizer. This is disabled by
39541 default, but passing @code{--enable-ubsan=yes} or
39542 @code{--enable-ubsan=auto} to @code{configure} will enable it. The
39543 undefined behavior sanitizer checks for C@t{++} undefined behavior.
39544 It has a performance cost, so if you are looking at @value{GDBN}'s
39545 performance, you should disable it. The undefined behavior sanitizer
39546 was first introduced in GCC 4.9.
39549 @node System-wide configuration
39550 @section System-wide configuration and settings
39551 @cindex system-wide init file
39553 @value{GDBN} can be configured to have a system-wide init file and a
39554 system-wide init file directory; this file and files in that directory
39555 (if they have a recognized file extension) will be read and executed at
39556 startup (@pxref{Startup, , What @value{GDBN} does during startup}).
39558 Here are the corresponding configure options:
39561 @item --with-system-gdbinit=@var{file}
39562 Specify that the default location of the system-wide init file is
39564 @item --with-system-gdbinit-dir=@var{directory}
39565 Specify that the default location of the system-wide init file directory
39566 is @var{directory}.
39569 If @value{GDBN} has been configured with the option @option{--prefix=$prefix},
39570 they may be subject to relocation. Two possible cases:
39574 If the default location of this init file/directory contains @file{$prefix},
39575 it will be subject to relocation. Suppose that the configure options
39576 are @option{--prefix=$prefix --with-system-gdbinit=$prefix/etc/gdbinit};
39577 if @value{GDBN} is moved from @file{$prefix} to @file{$install}, the system
39578 init file is looked for as @file{$install/etc/gdbinit} instead of
39579 @file{$prefix/etc/gdbinit}.
39582 By contrast, if the default location does not contain the prefix,
39583 it will not be relocated. E.g.@: if @value{GDBN} has been configured with
39584 @option{--prefix=/usr/local --with-system-gdbinit=/usr/share/gdb/gdbinit},
39585 then @value{GDBN} will always look for @file{/usr/share/gdb/gdbinit},
39586 wherever @value{GDBN} is installed.
39589 If the configured location of the system-wide init file (as given by the
39590 @option{--with-system-gdbinit} option at configure time) is in the
39591 data-directory (as specified by @option{--with-gdb-datadir} at configure
39592 time) or in one of its subdirectories, then @value{GDBN} will look for the
39593 system-wide init file in the directory specified by the
39594 @option{--data-directory} command-line option.
39595 Note that the system-wide init file is only read once, during @value{GDBN}
39596 initialization. If the data-directory is changed after @value{GDBN} has
39597 started with the @code{set data-directory} command, the file will not be
39600 This applies similarly to the system-wide directory specified in
39601 @option{--with-system-gdbinit-dir}.
39603 Any supported scripting language can be used for these init files, as long
39604 as the file extension matches the scripting language. To be interpreted
39605 as regular @value{GDBN} commands, the files needs to have a @file{.gdb}
39609 * System-wide Configuration Scripts:: Installed System-wide Configuration Scripts
39612 @node System-wide Configuration Scripts
39613 @subsection Installed System-wide Configuration Scripts
39614 @cindex system-wide configuration scripts
39616 The @file{system-gdbinit} directory, located inside the data-directory
39617 (as specified by @option{--with-gdb-datadir} at configure time) contains
39618 a number of scripts which can be used as system-wide init files. To
39619 automatically source those scripts at startup, @value{GDBN} should be
39620 configured with @option{--with-system-gdbinit}. Otherwise, any user
39621 should be able to source them by hand as needed.
39623 The following scripts are currently available:
39626 @item @file{elinos.py}
39628 @cindex ELinOS system-wide configuration script
39629 This script is useful when debugging a program on an ELinOS target.
39630 It takes advantage of the environment variables defined in a standard
39631 ELinOS environment in order to determine the location of the system
39632 shared libraries, and then sets the @samp{solib-absolute-prefix}
39633 and @samp{solib-search-path} variables appropriately.
39635 @item @file{wrs-linux.py}
39636 @pindex wrs-linux.py
39637 @cindex Wind River Linux system-wide configuration script
39638 This script is useful when debugging a program on a target running
39639 Wind River Linux. It expects the @env{ENV_PREFIX} to be set to
39640 the host-side sysroot used by the target system.
39644 @node Maintenance Commands
39645 @appendix Maintenance Commands
39646 @cindex maintenance commands
39647 @cindex internal commands
39649 In addition to commands intended for @value{GDBN} users, @value{GDBN}
39650 includes a number of commands intended for @value{GDBN} developers,
39651 that are not documented elsewhere in this manual. These commands are
39652 provided here for reference. (For commands that turn on debugging
39653 messages, see @ref{Debugging Output}.)
39656 @kindex maint agent
39657 @kindex maint agent-eval
39658 @item maint agent @r{[}-at @var{location}@r{,}@r{]} @var{expression}
39659 @itemx maint agent-eval @r{[}-at @var{location}@r{,}@r{]} @var{expression}
39660 Translate the given @var{expression} into remote agent bytecodes.
39661 This command is useful for debugging the Agent Expression mechanism
39662 (@pxref{Agent Expressions}). The @samp{agent} version produces an
39663 expression useful for data collection, such as by tracepoints, while
39664 @samp{maint agent-eval} produces an expression that evaluates directly
39665 to a result. For instance, a collection expression for @code{globa +
39666 globb} will include bytecodes to record four bytes of memory at each
39667 of the addresses of @code{globa} and @code{globb}, while discarding
39668 the result of the addition, while an evaluation expression will do the
39669 addition and return the sum.
39670 If @code{-at} is given, generate remote agent bytecode for @var{location}.
39671 If not, generate remote agent bytecode for current frame PC address.
39673 @kindex maint agent-printf
39674 @item maint agent-printf @var{format},@var{expr},...
39675 Translate the given format string and list of argument expressions
39676 into remote agent bytecodes and display them as a disassembled list.
39677 This command is useful for debugging the agent version of dynamic
39678 printf (@pxref{Dynamic Printf}).
39680 @kindex maint info breakpoints
39681 @item @anchor{maint info breakpoints}maint info breakpoints
39682 Using the same format as @samp{info breakpoints}, display both the
39683 breakpoints you've set explicitly, and those @value{GDBN} is using for
39684 internal purposes. Internal breakpoints are shown with negative
39685 breakpoint numbers. The type column identifies what kind of breakpoint
39690 Normal, explicitly set breakpoint.
39693 Normal, explicitly set watchpoint.
39696 Internal breakpoint, used to handle correctly stepping through
39697 @code{longjmp} calls.
39699 @item longjmp resume
39700 Internal breakpoint at the target of a @code{longjmp}.
39703 Temporary internal breakpoint used by the @value{GDBN} @code{until} command.
39706 Temporary internal breakpoint used by the @value{GDBN} @code{finish} command.
39709 Shared library events.
39713 @kindex maint info btrace
39714 @item maint info btrace
39715 Pint information about raw branch tracing data.
39717 @kindex maint btrace packet-history
39718 @item maint btrace packet-history
39719 Print the raw branch trace packets that are used to compute the
39720 execution history for the @samp{record btrace} command. Both the
39721 information and the format in which it is printed depend on the btrace
39726 For the BTS recording format, print a list of blocks of sequential
39727 code. For each block, the following information is printed:
39731 Newer blocks have higher numbers. The oldest block has number zero.
39732 @item Lowest @samp{PC}
39733 @item Highest @samp{PC}
39737 For the Intel Processor Trace recording format, print a list of
39738 Intel Processor Trace packets. For each packet, the following
39739 information is printed:
39742 @item Packet number
39743 Newer packets have higher numbers. The oldest packet has number zero.
39745 The packet's offset in the trace stream.
39746 @item Packet opcode and payload
39750 @kindex maint btrace clear-packet-history
39751 @item maint btrace clear-packet-history
39752 Discards the cached packet history printed by the @samp{maint btrace
39753 packet-history} command. The history will be computed again when
39756 @kindex maint btrace clear
39757 @item maint btrace clear
39758 Discard the branch trace data. The data will be fetched anew and the
39759 branch trace will be recomputed when needed.
39761 This implicitly truncates the branch trace to a single branch trace
39762 buffer. When updating branch trace incrementally, the branch trace
39763 available to @value{GDBN} may be bigger than a single branch trace
39766 @kindex maint set btrace pt skip-pad
39767 @item maint set btrace pt skip-pad
39768 @kindex maint show btrace pt skip-pad
39769 @item maint show btrace pt skip-pad
39770 Control whether @value{GDBN} will skip PAD packets when computing the
39773 @kindex set displaced-stepping
39774 @kindex show displaced-stepping
39775 @cindex displaced stepping support
39776 @cindex out-of-line single-stepping
39777 @item set displaced-stepping
39778 @itemx show displaced-stepping
39779 Control whether or not @value{GDBN} will do @dfn{displaced stepping}
39780 if the target supports it. Displaced stepping is a way to single-step
39781 over breakpoints without removing them from the inferior, by executing
39782 an out-of-line copy of the instruction that was originally at the
39783 breakpoint location. It is also known as out-of-line single-stepping.
39786 @item set displaced-stepping on
39787 If the target architecture supports it, @value{GDBN} will use
39788 displaced stepping to step over breakpoints.
39790 @item set displaced-stepping off
39791 @value{GDBN} will not use displaced stepping to step over breakpoints,
39792 even if such is supported by the target architecture.
39794 @cindex non-stop mode, and @samp{set displaced-stepping}
39795 @item set displaced-stepping auto
39796 This is the default mode. @value{GDBN} will use displaced stepping
39797 only if non-stop mode is active (@pxref{Non-Stop Mode}) and the target
39798 architecture supports displaced stepping.
39801 @kindex maint check-psymtabs
39802 @item maint check-psymtabs
39803 Check the consistency of currently expanded psymtabs versus symtabs.
39804 Use this to check, for example, whether a symbol is in one but not the other.
39806 @kindex maint check-symtabs
39807 @item maint check-symtabs
39808 Check the consistency of currently expanded symtabs.
39810 @kindex maint expand-symtabs
39811 @item maint expand-symtabs [@var{regexp}]
39812 Expand symbol tables.
39813 If @var{regexp} is specified, only expand symbol tables for file
39814 names matching @var{regexp}.
39816 @kindex maint set catch-demangler-crashes
39817 @kindex maint show catch-demangler-crashes
39818 @cindex demangler crashes
39819 @item maint set catch-demangler-crashes [on|off]
39820 @itemx maint show catch-demangler-crashes
39821 Control whether @value{GDBN} should attempt to catch crashes in the
39822 symbol name demangler. The default is to attempt to catch crashes.
39823 If enabled, the first time a crash is caught, a core file is created,
39824 the offending symbol is displayed and the user is presented with the
39825 option to terminate the current session.
39827 @kindex maint cplus first_component
39828 @item maint cplus first_component @var{name}
39829 Print the first C@t{++} class/namespace component of @var{name}.
39831 @kindex maint cplus namespace
39832 @item maint cplus namespace
39833 Print the list of possible C@t{++} namespaces.
39835 @kindex maint deprecate
39836 @kindex maint undeprecate
39837 @cindex deprecated commands
39838 @item maint deprecate @var{command} @r{[}@var{replacement}@r{]}
39839 @itemx maint undeprecate @var{command}
39840 Deprecate or undeprecate the named @var{command}. Deprecated commands
39841 cause @value{GDBN} to issue a warning when you use them. The optional
39842 argument @var{replacement} says which newer command should be used in
39843 favor of the deprecated one; if it is given, @value{GDBN} will mention
39844 the replacement as part of the warning.
39846 @kindex maint dump-me
39847 @item maint dump-me
39848 @cindex @code{SIGQUIT} signal, dump core of @value{GDBN}
39849 Cause a fatal signal in the debugger and force it to dump its core.
39850 This is supported only on systems which support aborting a program
39851 with the @code{SIGQUIT} signal.
39853 @kindex maint internal-error
39854 @kindex maint internal-warning
39855 @kindex maint demangler-warning
39856 @cindex demangler crashes
39857 @item maint internal-error @r{[}@var{message-text}@r{]}
39858 @itemx maint internal-warning @r{[}@var{message-text}@r{]}
39859 @itemx maint demangler-warning @r{[}@var{message-text}@r{]}
39861 Cause @value{GDBN} to call the internal function @code{internal_error},
39862 @code{internal_warning} or @code{demangler_warning} and hence behave
39863 as though an internal problem has been detected. In addition to
39864 reporting the internal problem, these functions give the user the
39865 opportunity to either quit @value{GDBN} or (for @code{internal_error}
39866 and @code{internal_warning}) create a core file of the current
39867 @value{GDBN} session.
39869 These commands take an optional parameter @var{message-text} that is
39870 used as the text of the error or warning message.
39872 Here's an example of using @code{internal-error}:
39875 (@value{GDBP}) @kbd{maint internal-error testing, 1, 2}
39876 @dots{}/maint.c:121: internal-error: testing, 1, 2
39877 A problem internal to GDB has been detected. Further
39878 debugging may prove unreliable.
39879 Quit this debugging session? (y or n) @kbd{n}
39880 Create a core file? (y or n) @kbd{n}
39884 @cindex @value{GDBN} internal error
39885 @cindex internal errors, control of @value{GDBN} behavior
39886 @cindex demangler crashes
39888 @kindex maint set internal-error
39889 @kindex maint show internal-error
39890 @kindex maint set internal-warning
39891 @kindex maint show internal-warning
39892 @kindex maint set demangler-warning
39893 @kindex maint show demangler-warning
39894 @item maint set internal-error @var{action} [ask|yes|no]
39895 @itemx maint show internal-error @var{action}
39896 @itemx maint set internal-warning @var{action} [ask|yes|no]
39897 @itemx maint show internal-warning @var{action}
39898 @itemx maint set demangler-warning @var{action} [ask|yes|no]
39899 @itemx maint show demangler-warning @var{action}
39900 When @value{GDBN} reports an internal problem (error or warning) it
39901 gives the user the opportunity to both quit @value{GDBN} and create a
39902 core file of the current @value{GDBN} session. These commands let you
39903 override the default behaviour for each particular @var{action},
39904 described in the table below.
39908 You can specify that @value{GDBN} should always (yes) or never (no)
39909 quit. The default is to ask the user what to do.
39912 You can specify that @value{GDBN} should always (yes) or never (no)
39913 create a core file. The default is to ask the user what to do. Note
39914 that there is no @code{corefile} option for @code{demangler-warning}:
39915 demangler warnings always create a core file and this cannot be
39919 @kindex maint packet
39920 @item maint packet @var{text}
39921 If @value{GDBN} is talking to an inferior via the serial protocol,
39922 then this command sends the string @var{text} to the inferior, and
39923 displays the response packet. @value{GDBN} supplies the initial
39924 @samp{$} character, the terminating @samp{#} character, and the
39927 @kindex maint print architecture
39928 @item maint print architecture @r{[}@var{file}@r{]}
39929 Print the entire architecture configuration. The optional argument
39930 @var{file} names the file where the output goes.
39932 @kindex maint print c-tdesc @r{[}@var{file}@r{]}
39933 @item maint print c-tdesc
39934 Print the target description (@pxref{Target Descriptions}) as
39935 a C source file. By default, the target description is for the current
39936 target, but if the optional argument @var{file} is provided, that file
39937 is used to produce the description. The @var{file} should be an XML
39938 document, of the form described in @ref{Target Description Format}.
39939 The created source file is built into @value{GDBN} when @value{GDBN} is
39940 built again. This command is used by developers after they add or
39941 modify XML target descriptions.
39943 @kindex maint check xml-descriptions
39944 @item maint check xml-descriptions @var{dir}
39945 Check that the target descriptions dynamically created by @value{GDBN}
39946 equal the descriptions created from XML files found in @var{dir}.
39948 @anchor{maint check libthread-db}
39949 @kindex maint check libthread-db
39950 @item maint check libthread-db
39951 Run integrity checks on the current inferior's thread debugging
39952 library. This exercises all @code{libthread_db} functionality used by
39953 @value{GDBN} on GNU/Linux systems, and by extension also exercises the
39954 @code{proc_service} functions provided by @value{GDBN} that
39955 @code{libthread_db} uses. Note that parts of the test may be skipped
39956 on some platforms when debugging core files.
39958 @kindex maint print dummy-frames
39959 @item maint print dummy-frames
39960 Prints the contents of @value{GDBN}'s internal dummy-frame stack.
39963 (@value{GDBP}) @kbd{b add}
39965 (@value{GDBP}) @kbd{print add(2,3)}
39966 Breakpoint 2, add (a=2, b=3) at @dots{}
39968 The program being debugged stopped while in a function called from GDB.
39970 (@value{GDBP}) @kbd{maint print dummy-frames}
39971 0xa8206d8: id=@{stack=0xbfffe734,code=0xbfffe73f,!special@}, ptid=process 9353
39975 Takes an optional file parameter.
39977 @kindex maint print registers
39978 @kindex maint print raw-registers
39979 @kindex maint print cooked-registers
39980 @kindex maint print register-groups
39981 @kindex maint print remote-registers
39982 @item maint print registers @r{[}@var{file}@r{]}
39983 @itemx maint print raw-registers @r{[}@var{file}@r{]}
39984 @itemx maint print cooked-registers @r{[}@var{file}@r{]}
39985 @itemx maint print register-groups @r{[}@var{file}@r{]}
39986 @itemx maint print remote-registers @r{[}@var{file}@r{]}
39987 Print @value{GDBN}'s internal register data structures.
39989 The command @code{maint print raw-registers} includes the contents of
39990 the raw register cache; the command @code{maint print
39991 cooked-registers} includes the (cooked) value of all registers,
39992 including registers which aren't available on the target nor visible
39993 to user; the command @code{maint print register-groups} includes the
39994 groups that each register is a member of; and the command @code{maint
39995 print remote-registers} includes the remote target's register numbers
39996 and offsets in the `G' packets.
39998 These commands take an optional parameter, a file name to which to
39999 write the information.
40001 @kindex maint print reggroups
40002 @item maint print reggroups @r{[}@var{file}@r{]}
40003 Print @value{GDBN}'s internal register group data structures. The
40004 optional argument @var{file} tells to what file to write the
40007 The register groups info looks like this:
40010 (@value{GDBP}) @kbd{maint print reggroups}
40023 This command forces @value{GDBN} to flush its internal register cache.
40025 @kindex maint print address-spaces
40026 @item maint print address-spaces @r{[}@var{file}@r{]}
40027 Print @value{GDBN}'s internal address space data structures. The
40028 optional argument @var{file} tells to what file to write the
40029 information. @xref{maint print address-spaces,, @code{maint print
40032 @kindex maint print objfiles
40033 @cindex info for known object files
40034 @item maint print objfiles @r{[}@var{regexp}@r{]}
40035 Print a dump of all known object files.
40036 If @var{regexp} is specified, only print object files whose names
40037 match @var{regexp}. For each object file, this command prints its name,
40038 address in memory, and all of its psymtabs and symtabs.
40040 @kindex maint print user-registers
40041 @cindex user registers
40042 @item maint print user-registers
40043 List all currently available @dfn{user registers}. User registers
40044 typically provide alternate names for actual hardware registers. They
40045 include the four ``standard'' registers @code{$fp}, @code{$pc},
40046 @code{$sp}, and @code{$ps}. @xref{standard registers}. User
40047 registers can be used in expressions in the same way as the canonical
40048 register names, but only the latter are listed by the @code{info
40049 registers} and @code{maint print registers} commands.
40051 @kindex maint print section-scripts
40052 @cindex info for known .debug_gdb_scripts-loaded scripts
40053 @item maint print section-scripts [@var{regexp}]
40054 Print a dump of scripts specified in the @code{.debug_gdb_section} section.
40055 If @var{regexp} is specified, only print scripts loaded by object files
40056 matching @var{regexp}.
40057 For each script, this command prints its name as specified in the objfile,
40058 and the full path if known.
40059 @xref{dotdebug_gdb_scripts section}.
40061 @kindex maint print statistics
40062 @cindex bcache statistics
40063 @item maint print statistics
40064 This command prints, for each object file in the program, various data
40065 about that object file followed by the byte cache (@dfn{bcache})
40066 statistics for the object file. The objfile data includes the number
40067 of minimal, partial, full, and stabs symbols, the number of types
40068 defined by the objfile, the number of as yet unexpanded psym tables,
40069 the number of line tables and string tables, and the amount of memory
40070 used by the various tables. The bcache statistics include the counts,
40071 sizes, and counts of duplicates of all and unique objects, max,
40072 average, and median entry size, total memory used and its overhead and
40073 savings, and various measures of the hash table size and chain
40076 @kindex maint print target-stack
40077 @cindex target stack description
40078 @item maint print target-stack
40079 A @dfn{target} is an interface between the debugger and a particular
40080 kind of file or process. Targets can be stacked in @dfn{strata},
40081 so that more than one target can potentially respond to a request.
40082 In particular, memory accesses will walk down the stack of targets
40083 until they find a target that is interested in handling that particular
40086 This command prints a short description of each layer that was pushed on
40087 the @dfn{target stack}, starting from the top layer down to the bottom one.
40089 @kindex maint print type
40090 @cindex type chain of a data type
40091 @item maint print type @var{expr}
40092 Print the type chain for a type specified by @var{expr}. The argument
40093 can be either a type name or a symbol. If it is a symbol, the type of
40094 that symbol is described. The type chain produced by this command is
40095 a recursive definition of the data type as stored in @value{GDBN}'s
40096 data structures, including its flags and contained types.
40098 @kindex maint selftest
40100 @item maint selftest @r{[}@var{filter}@r{]}
40101 Run any self tests that were compiled in to @value{GDBN}. This will
40102 print a message showing how many tests were run, and how many failed.
40103 If a @var{filter} is passed, only the tests with @var{filter} in their
40106 @kindex maint info selftests
40108 @item maint info selftests
40109 List the selftests compiled in to @value{GDBN}.
40111 @kindex maint set dwarf always-disassemble
40112 @kindex maint show dwarf always-disassemble
40113 @item maint set dwarf always-disassemble
40114 @item maint show dwarf always-disassemble
40115 Control the behavior of @code{info address} when using DWARF debugging
40118 The default is @code{off}, which means that @value{GDBN} should try to
40119 describe a variable's location in an easily readable format. When
40120 @code{on}, @value{GDBN} will instead display the DWARF location
40121 expression in an assembly-like format. Note that some locations are
40122 too complex for @value{GDBN} to describe simply; in this case you will
40123 always see the disassembly form.
40125 Here is an example of the resulting disassembly:
40128 (@value{GDBP}) info addr argc
40129 Symbol "argc" is a complex DWARF expression:
40133 For more information on these expressions, see
40134 @uref{http://www.dwarfstd.org/, the DWARF standard}.
40136 @kindex maint set dwarf max-cache-age
40137 @kindex maint show dwarf max-cache-age
40138 @item maint set dwarf max-cache-age
40139 @itemx maint show dwarf max-cache-age
40140 Control the DWARF compilation unit cache.
40142 @cindex DWARF compilation units cache
40143 In object files with inter-compilation-unit references, such as those
40144 produced by the GCC option @samp{-feliminate-dwarf2-dups}, the DWARF
40145 reader needs to frequently refer to previously read compilation units.
40146 This setting controls how long a compilation unit will remain in the
40147 cache if it is not referenced. A higher limit means that cached
40148 compilation units will be stored in memory longer, and more total
40149 memory will be used. Setting it to zero disables caching, which will
40150 slow down @value{GDBN} startup, but reduce memory consumption.
40152 @kindex maint set dwarf unwinders
40153 @kindex maint show dwarf unwinders
40154 @item maint set dwarf unwinders
40155 @itemx maint show dwarf unwinders
40156 Control use of the DWARF frame unwinders.
40158 @cindex DWARF frame unwinders
40159 Many targets that support DWARF debugging use @value{GDBN}'s DWARF
40160 frame unwinders to build the backtrace. Many of these targets will
40161 also have a second mechanism for building the backtrace for use in
40162 cases where DWARF information is not available, this second mechanism
40163 is often an analysis of a function's prologue.
40165 In order to extend testing coverage of the second level stack
40166 unwinding mechanisms it is helpful to be able to disable the DWARF
40167 stack unwinders, this can be done with this switch.
40169 In normal use of @value{GDBN} disabling the DWARF unwinders is not
40170 advisable, there are cases that are better handled through DWARF than
40171 prologue analysis, and the debug experience is likely to be better
40172 with the DWARF frame unwinders enabled.
40174 If DWARF frame unwinders are not supported for a particular target
40175 architecture, then enabling this flag does not cause them to be used.
40177 @kindex maint set worker-threads
40178 @kindex maint show worker-threads
40179 @item maint set worker-threads
40180 @item maint show worker-threads
40181 Control the number of worker threads that may be used by @value{GDBN}.
40182 On capable hosts, @value{GDBN} may use multiple threads to speed up
40183 certain CPU-intensive operations, such as demangling symbol names.
40184 While the number of threads used by @value{GDBN} may vary, this
40185 command can be used to set an upper bound on this number. The default
40186 is @code{0} (disabled). A value of @code{unlimited} lets @value{GDBN} choose a
40187 reasonable number. Note that this only controls worker threads started by
40188 @value{GDBN} itself; libraries used by @value{GDBN} may start threads of their
40191 @kindex maint set profile
40192 @kindex maint show profile
40193 @cindex profiling GDB
40194 @item maint set profile
40195 @itemx maint show profile
40196 Control profiling of @value{GDBN}.
40198 Profiling will be disabled until you use the @samp{maint set profile}
40199 command to enable it. When you enable profiling, the system will begin
40200 collecting timing and execution count data; when you disable profiling or
40201 exit @value{GDBN}, the results will be written to a log file. Remember that
40202 if you use profiling, @value{GDBN} will overwrite the profiling log file
40203 (often called @file{gmon.out}). If you have a record of important profiling
40204 data in a @file{gmon.out} file, be sure to move it to a safe location.
40206 Configuring with @samp{--enable-profiling} arranges for @value{GDBN} to be
40207 compiled with the @samp{-pg} compiler option.
40209 @kindex maint set show-debug-regs
40210 @kindex maint show show-debug-regs
40211 @cindex hardware debug registers
40212 @item maint set show-debug-regs
40213 @itemx maint show show-debug-regs
40214 Control whether to show variables that mirror the hardware debug
40215 registers. Use @code{on} to enable, @code{off} to disable. If
40216 enabled, the debug registers values are shown when @value{GDBN} inserts or
40217 removes a hardware breakpoint or watchpoint, and when the inferior
40218 triggers a hardware-assisted breakpoint or watchpoint.
40220 @kindex maint set show-all-tib
40221 @kindex maint show show-all-tib
40222 @item maint set show-all-tib
40223 @itemx maint show show-all-tib
40224 Control whether to show all non zero areas within a 1k block starting
40225 at thread local base, when using the @samp{info w32 thread-information-block}
40228 @kindex maint set target-async
40229 @kindex maint show target-async
40230 @item maint set target-async
40231 @itemx maint show target-async
40232 This controls whether @value{GDBN} targets operate in synchronous or
40233 asynchronous mode (@pxref{Background Execution}). Normally the
40234 default is asynchronous, if it is available; but this can be changed
40235 to more easily debug problems occurring only in synchronous mode.
40237 @kindex maint set target-non-stop @var{mode} [on|off|auto]
40238 @kindex maint show target-non-stop
40239 @item maint set target-non-stop
40240 @itemx maint show target-non-stop
40242 This controls whether @value{GDBN} targets always operate in non-stop
40243 mode even if @code{set non-stop} is @code{off} (@pxref{Non-Stop
40244 Mode}). The default is @code{auto}, meaning non-stop mode is enabled
40245 if supported by the target.
40248 @item maint set target-non-stop auto
40249 This is the default mode. @value{GDBN} controls the target in
40250 non-stop mode if the target supports it.
40252 @item maint set target-non-stop on
40253 @value{GDBN} controls the target in non-stop mode even if the target
40254 does not indicate support.
40256 @item maint set target-non-stop off
40257 @value{GDBN} does not control the target in non-stop mode even if the
40258 target supports it.
40261 @kindex maint set tui-resize-message
40262 @kindex maint show tui-resize-message
40263 @item maint set tui-resize-message
40264 @item maint show tui-resize-message
40265 Control whether @value{GDBN} displays a message each time the terminal
40266 is resized when in TUI mode. The default is @code{off}, which means
40267 that @value{GDBN} is silent during resizes. When @code{on},
40268 @value{GDBN} will display a message after a resize is completed; the
40269 message will include a number indicating how many times the terminal
40270 has been resized. This setting is intended for use by the test suite,
40271 where it would otherwise be difficult to determine when a resize and
40272 refresh has been completed.
40274 @kindex maint set per-command
40275 @kindex maint show per-command
40276 @item maint set per-command
40277 @itemx maint show per-command
40278 @cindex resources used by commands
40280 @value{GDBN} can display the resources used by each command.
40281 This is useful in debugging performance problems.
40284 @item maint set per-command space [on|off]
40285 @itemx maint show per-command space
40286 Enable or disable the printing of the memory used by GDB for each command.
40287 If enabled, @value{GDBN} will display how much memory each command
40288 took, following the command's own output.
40289 This can also be requested by invoking @value{GDBN} with the
40290 @option{--statistics} command-line switch (@pxref{Mode Options}).
40292 @item maint set per-command time [on|off]
40293 @itemx maint show per-command time
40294 Enable or disable the printing of the execution time of @value{GDBN}
40296 If enabled, @value{GDBN} will display how much time it
40297 took to execute each command, following the command's own output.
40298 Both CPU time and wallclock time are printed.
40299 Printing both is useful when trying to determine whether the cost is
40300 CPU or, e.g., disk/network latency.
40301 Note that the CPU time printed is for @value{GDBN} only, it does not include
40302 the execution time of the inferior because there's no mechanism currently
40303 to compute how much time was spent by @value{GDBN} and how much time was
40304 spent by the program been debugged.
40305 This can also be requested by invoking @value{GDBN} with the
40306 @option{--statistics} command-line switch (@pxref{Mode Options}).
40308 @item maint set per-command symtab [on|off]
40309 @itemx maint show per-command symtab
40310 Enable or disable the printing of basic symbol table statistics
40312 If enabled, @value{GDBN} will display the following information:
40316 number of symbol tables
40318 number of primary symbol tables
40320 number of blocks in the blockvector
40324 @kindex maint set check-libthread-db
40325 @kindex maint show check-libthread-db
40326 @item maint set check-libthread-db [on|off]
40327 @itemx maint show check-libthread-db
40328 Control whether @value{GDBN} should run integrity checks on inferior
40329 specific thread debugging libraries as they are loaded. The default
40330 is not to perform such checks. If any check fails @value{GDBN} will
40331 unload the library and continue searching for a suitable candidate as
40332 described in @ref{set libthread-db-search-path}. For more information
40333 about the tests, see @ref{maint check libthread-db}.
40335 @kindex maint space
40336 @cindex memory used by commands
40337 @item maint space @var{value}
40338 An alias for @code{maint set per-command space}.
40339 A non-zero value enables it, zero disables it.
40342 @cindex time of command execution
40343 @item maint time @var{value}
40344 An alias for @code{maint set per-command time}.
40345 A non-zero value enables it, zero disables it.
40347 @kindex maint translate-address
40348 @item maint translate-address @r{[}@var{section}@r{]} @var{addr}
40349 Find the symbol stored at the location specified by the address
40350 @var{addr} and an optional section name @var{section}. If found,
40351 @value{GDBN} prints the name of the closest symbol and an offset from
40352 the symbol's location to the specified address. This is similar to
40353 the @code{info address} command (@pxref{Symbols}), except that this
40354 command also allows to find symbols in other sections.
40356 If section was not specified, the section in which the symbol was found
40357 is also printed. For dynamically linked executables, the name of
40358 executable or shared library containing the symbol is printed as well.
40360 @kindex maint test-options
40361 @item maint test-options require-delimiter
40362 @itemx maint test-options unknown-is-error
40363 @itemx maint test-options unknown-is-operand
40364 These commands are used by the testsuite to validate the command
40365 options framework. The @code{require-delimiter} variant requires a
40366 double-dash delimiter to indicate end of options. The
40367 @code{unknown-is-error} and @code{unknown-is-operand} do not. The
40368 @code{unknown-is-error} variant throws an error on unknown option,
40369 while @code{unknown-is-operand} treats unknown options as the start of
40370 the command's operands. When run, the commands output the result of
40371 the processed options. When completed, the commands store the
40372 internal result of completion in a variable exposed by the @code{maint
40373 show test-options-completion-result} command.
40375 @kindex maint show test-options-completion-result
40376 @item maint show test-options-completion-result
40377 Shows the result of completing the @code{maint test-options}
40378 subcommands. This is used by the testsuite to validate completion
40379 support in the command options framework.
40381 @kindex maint set test-settings
40382 @kindex maint show test-settings
40383 @item maint set test-settings @var{kind}
40384 @itemx maint show test-settings @var{kind}
40385 These are representative commands for each @var{kind} of setting type
40386 @value{GDBN} supports. They are used by the testsuite for exercising
40387 the settings infrastructure.
40390 @item maint with @var{setting} [@var{value}] [-- @var{command}]
40391 Like the @code{with} command, but works with @code{maintenance set}
40392 variables. This is used by the testsuite to exercise the @code{with}
40393 command's infrastructure.
40397 The following command is useful for non-interactive invocations of
40398 @value{GDBN}, such as in the test suite.
40401 @item set watchdog @var{nsec}
40402 @kindex set watchdog
40403 @cindex watchdog timer
40404 @cindex timeout for commands
40405 Set the maximum number of seconds @value{GDBN} will wait for the
40406 target operation to finish. If this time expires, @value{GDBN}
40407 reports and error and the command is aborted.
40409 @item show watchdog
40410 Show the current setting of the target wait timeout.
40413 @node Remote Protocol
40414 @appendix @value{GDBN} Remote Serial Protocol
40419 * Stop Reply Packets::
40420 * General Query Packets::
40421 * Architecture-Specific Protocol Details::
40422 * Tracepoint Packets::
40423 * Host I/O Packets::
40425 * Notification Packets::
40426 * Remote Non-Stop::
40427 * Packet Acknowledgment::
40429 * File-I/O Remote Protocol Extension::
40430 * Library List Format::
40431 * Library List Format for SVR4 Targets::
40432 * Memory Map Format::
40433 * Thread List Format::
40434 * Traceframe Info Format::
40435 * Branch Trace Format::
40436 * Branch Trace Configuration Format::
40442 There may be occasions when you need to know something about the
40443 protocol---for example, if there is only one serial port to your target
40444 machine, you might want your program to do something special if it
40445 recognizes a packet meant for @value{GDBN}.
40447 In the examples below, @samp{->} and @samp{<-} are used to indicate
40448 transmitted and received data, respectively.
40450 @cindex protocol, @value{GDBN} remote serial
40451 @cindex serial protocol, @value{GDBN} remote
40452 @cindex remote serial protocol
40453 All @value{GDBN} commands and responses (other than acknowledgments
40454 and notifications, see @ref{Notification Packets}) are sent as a
40455 @var{packet}. A @var{packet} is introduced with the character
40456 @samp{$}, the actual @var{packet-data}, and the terminating character
40457 @samp{#} followed by a two-digit @var{checksum}:
40460 @code{$}@var{packet-data}@code{#}@var{checksum}
40464 @cindex checksum, for @value{GDBN} remote
40466 The two-digit @var{checksum} is computed as the modulo 256 sum of all
40467 characters between the leading @samp{$} and the trailing @samp{#} (an
40468 eight bit unsigned checksum).
40470 Implementors should note that prior to @value{GDBN} 5.0 the protocol
40471 specification also included an optional two-digit @var{sequence-id}:
40474 @code{$}@var{sequence-id}@code{:}@var{packet-data}@code{#}@var{checksum}
40477 @cindex sequence-id, for @value{GDBN} remote
40479 That @var{sequence-id} was appended to the acknowledgment. @value{GDBN}
40480 has never output @var{sequence-id}s. Stubs that handle packets added
40481 since @value{GDBN} 5.0 must not accept @var{sequence-id}.
40483 When either the host or the target machine receives a packet, the first
40484 response expected is an acknowledgment: either @samp{+} (to indicate
40485 the package was received correctly) or @samp{-} (to request
40489 -> @code{$}@var{packet-data}@code{#}@var{checksum}
40494 The @samp{+}/@samp{-} acknowledgments can be disabled
40495 once a connection is established.
40496 @xref{Packet Acknowledgment}, for details.
40498 The host (@value{GDBN}) sends @var{command}s, and the target (the
40499 debugging stub incorporated in your program) sends a @var{response}. In
40500 the case of step and continue @var{command}s, the response is only sent
40501 when the operation has completed, and the target has again stopped all
40502 threads in all attached processes. This is the default all-stop mode
40503 behavior, but the remote protocol also supports @value{GDBN}'s non-stop
40504 execution mode; see @ref{Remote Non-Stop}, for details.
40506 @var{packet-data} consists of a sequence of characters with the
40507 exception of @samp{#} and @samp{$} (see @samp{X} packet for additional
40510 @cindex remote protocol, field separator
40511 Fields within the packet should be separated using @samp{,} @samp{;} or
40512 @samp{:}. Except where otherwise noted all numbers are represented in
40513 @sc{hex} with leading zeros suppressed.
40515 Implementors should note that prior to @value{GDBN} 5.0, the character
40516 @samp{:} could not appear as the third character in a packet (as it
40517 would potentially conflict with the @var{sequence-id}).
40519 @cindex remote protocol, binary data
40520 @anchor{Binary Data}
40521 Binary data in most packets is encoded either as two hexadecimal
40522 digits per byte of binary data. This allowed the traditional remote
40523 protocol to work over connections which were only seven-bit clean.
40524 Some packets designed more recently assume an eight-bit clean
40525 connection, and use a more efficient encoding to send and receive
40528 The binary data representation uses @code{7d} (@sc{ascii} @samp{@}})
40529 as an escape character. Any escaped byte is transmitted as the escape
40530 character followed by the original character XORed with @code{0x20}.
40531 For example, the byte @code{0x7d} would be transmitted as the two
40532 bytes @code{0x7d 0x5d}. The bytes @code{0x23} (@sc{ascii} @samp{#}),
40533 @code{0x24} (@sc{ascii} @samp{$}), and @code{0x7d} (@sc{ascii}
40534 @samp{@}}) must always be escaped. Responses sent by the stub
40535 must also escape @code{0x2a} (@sc{ascii} @samp{*}), so that it
40536 is not interpreted as the start of a run-length encoded sequence
40539 Response @var{data} can be run-length encoded to save space.
40540 Run-length encoding replaces runs of identical characters with one
40541 instance of the repeated character, followed by a @samp{*} and a
40542 repeat count. The repeat count is itself sent encoded, to avoid
40543 binary characters in @var{data}: a value of @var{n} is sent as
40544 @code{@var{n}+29}. For a repeat count greater or equal to 3, this
40545 produces a printable @sc{ascii} character, e.g.@: a space (@sc{ascii}
40546 code 32) for a repeat count of 3. (This is because run-length
40547 encoding starts to win for counts 3 or more.) Thus, for example,
40548 @samp{0* } is a run-length encoding of ``0000'': the space character
40549 after @samp{*} means repeat the leading @code{0} @w{@code{32 - 29 =
40552 The printable characters @samp{#} and @samp{$} or with a numeric value
40553 greater than 126 must not be used. Runs of six repeats (@samp{#}) or
40554 seven repeats (@samp{$}) can be expanded using a repeat count of only
40555 five (@samp{"}). For example, @samp{00000000} can be encoded as
40558 The error response returned for some packets includes a two character
40559 error number. That number is not well defined.
40561 @cindex empty response, for unsupported packets
40562 For any @var{command} not supported by the stub, an empty response
40563 (@samp{$#00}) should be returned. That way it is possible to extend the
40564 protocol. A newer @value{GDBN} can tell if a packet is supported based
40567 At a minimum, a stub is required to support the @samp{g} and @samp{G}
40568 commands for register access, and the @samp{m} and @samp{M} commands
40569 for memory access. Stubs that only control single-threaded targets
40570 can implement run control with the @samp{c} (continue), and @samp{s}
40571 (step) commands. Stubs that support multi-threading targets should
40572 support the @samp{vCont} command. All other commands are optional.
40577 The following table provides a complete list of all currently defined
40578 @var{command}s and their corresponding response @var{data}.
40579 @xref{File-I/O Remote Protocol Extension}, for details about the File
40580 I/O extension of the remote protocol.
40582 Each packet's description has a template showing the packet's overall
40583 syntax, followed by an explanation of the packet's meaning. We
40584 include spaces in some of the templates for clarity; these are not
40585 part of the packet's syntax. No @value{GDBN} packet uses spaces to
40586 separate its components. For example, a template like @samp{foo
40587 @var{bar} @var{baz}} describes a packet beginning with the three ASCII
40588 bytes @samp{foo}, followed by a @var{bar}, followed directly by a
40589 @var{baz}. @value{GDBN} does not transmit a space character between the
40590 @samp{foo} and the @var{bar}, or between the @var{bar} and the
40593 @cindex @var{thread-id}, in remote protocol
40594 @anchor{thread-id syntax}
40595 Several packets and replies include a @var{thread-id} field to identify
40596 a thread. Normally these are positive numbers with a target-specific
40597 interpretation, formatted as big-endian hex strings. A @var{thread-id}
40598 can also be a literal @samp{-1} to indicate all threads, or @samp{0} to
40601 In addition, the remote protocol supports a multiprocess feature in
40602 which the @var{thread-id} syntax is extended to optionally include both
40603 process and thread ID fields, as @samp{p@var{pid}.@var{tid}}.
40604 The @var{pid} (process) and @var{tid} (thread) components each have the
40605 format described above: a positive number with target-specific
40606 interpretation formatted as a big-endian hex string, literal @samp{-1}
40607 to indicate all processes or threads (respectively), or @samp{0} to
40608 indicate an arbitrary process or thread. Specifying just a process, as
40609 @samp{p@var{pid}}, is equivalent to @samp{p@var{pid}.-1}. It is an
40610 error to specify all processes but a specific thread, such as
40611 @samp{p-1.@var{tid}}. Note that the @samp{p} prefix is @emph{not} used
40612 for those packets and replies explicitly documented to include a process
40613 ID, rather than a @var{thread-id}.
40615 The multiprocess @var{thread-id} syntax extensions are only used if both
40616 @value{GDBN} and the stub report support for the @samp{multiprocess}
40617 feature using @samp{qSupported}. @xref{multiprocess extensions}, for
40620 Note that all packet forms beginning with an upper- or lower-case
40621 letter, other than those described here, are reserved for future use.
40623 Here are the packet descriptions.
40628 @cindex @samp{!} packet
40629 @anchor{extended mode}
40630 Enable extended mode. In extended mode, the remote server is made
40631 persistent. The @samp{R} packet is used to restart the program being
40637 The remote target both supports and has enabled extended mode.
40641 @cindex @samp{?} packet
40643 Indicate the reason the target halted. The reply is the same as for
40644 step and continue. This packet has a special interpretation when the
40645 target is in non-stop mode; see @ref{Remote Non-Stop}.
40648 @xref{Stop Reply Packets}, for the reply specifications.
40650 @item A @var{arglen},@var{argnum},@var{arg},@dots{}
40651 @cindex @samp{A} packet
40652 Initialized @code{argv[]} array passed into program. @var{arglen}
40653 specifies the number of bytes in the hex encoded byte stream
40654 @var{arg}. See @code{gdbserver} for more details.
40659 The arguments were set.
40665 @cindex @samp{b} packet
40666 (Don't use this packet; its behavior is not well-defined.)
40667 Change the serial line speed to @var{baud}.
40669 JTC: @emph{When does the transport layer state change? When it's
40670 received, or after the ACK is transmitted. In either case, there are
40671 problems if the command or the acknowledgment packet is dropped.}
40673 Stan: @emph{If people really wanted to add something like this, and get
40674 it working for the first time, they ought to modify ser-unix.c to send
40675 some kind of out-of-band message to a specially-setup stub and have the
40676 switch happen "in between" packets, so that from remote protocol's point
40677 of view, nothing actually happened.}
40679 @item B @var{addr},@var{mode}
40680 @cindex @samp{B} packet
40681 Set (@var{mode} is @samp{S}) or clear (@var{mode} is @samp{C}) a
40682 breakpoint at @var{addr}.
40684 Don't use this packet. Use the @samp{Z} and @samp{z} packets instead
40685 (@pxref{insert breakpoint or watchpoint packet}).
40687 @cindex @samp{bc} packet
40690 Backward continue. Execute the target system in reverse. No parameter.
40691 @xref{Reverse Execution}, for more information.
40694 @xref{Stop Reply Packets}, for the reply specifications.
40696 @cindex @samp{bs} packet
40699 Backward single step. Execute one instruction in reverse. No parameter.
40700 @xref{Reverse Execution}, for more information.
40703 @xref{Stop Reply Packets}, for the reply specifications.
40705 @item c @r{[}@var{addr}@r{]}
40706 @cindex @samp{c} packet
40707 Continue at @var{addr}, which is the address to resume. If @var{addr}
40708 is omitted, resume at current address.
40710 This packet is deprecated for multi-threading support. @xref{vCont
40714 @xref{Stop Reply Packets}, for the reply specifications.
40716 @item C @var{sig}@r{[};@var{addr}@r{]}
40717 @cindex @samp{C} packet
40718 Continue with signal @var{sig} (hex signal number). If
40719 @samp{;@var{addr}} is omitted, resume at same address.
40721 This packet is deprecated for multi-threading support. @xref{vCont
40725 @xref{Stop Reply Packets}, for the reply specifications.
40728 @cindex @samp{d} packet
40731 Don't use this packet; instead, define a general set packet
40732 (@pxref{General Query Packets}).
40736 @cindex @samp{D} packet
40737 The first form of the packet is used to detach @value{GDBN} from the
40738 remote system. It is sent to the remote target
40739 before @value{GDBN} disconnects via the @code{detach} command.
40741 The second form, including a process ID, is used when multiprocess
40742 protocol extensions are enabled (@pxref{multiprocess extensions}), to
40743 detach only a specific process. The @var{pid} is specified as a
40744 big-endian hex string.
40754 @item F @var{RC},@var{EE},@var{CF};@var{XX}
40755 @cindex @samp{F} packet
40756 A reply from @value{GDBN} to an @samp{F} packet sent by the target.
40757 This is part of the File-I/O protocol extension. @xref{File-I/O
40758 Remote Protocol Extension}, for the specification.
40761 @anchor{read registers packet}
40762 @cindex @samp{g} packet
40763 Read general registers.
40767 @item @var{XX@dots{}}
40768 Each byte of register data is described by two hex digits. The bytes
40769 with the register are transmitted in target byte order. The size of
40770 each register and their position within the @samp{g} packet are
40771 determined by the @value{GDBN} internal gdbarch functions
40772 @code{DEPRECATED_REGISTER_RAW_SIZE} and @code{gdbarch_register_name}.
40774 When reading registers from a trace frame (@pxref{Analyze Collected
40775 Data,,Using the Collected Data}), the stub may also return a string of
40776 literal @samp{x}'s in place of the register data digits, to indicate
40777 that the corresponding register has not been collected, thus its value
40778 is unavailable. For example, for an architecture with 4 registers of
40779 4 bytes each, the following reply indicates to @value{GDBN} that
40780 registers 0 and 2 have not been collected, while registers 1 and 3
40781 have been collected, and both have zero value:
40785 <- @code{xxxxxxxx00000000xxxxxxxx00000000}
40792 @item G @var{XX@dots{}}
40793 @cindex @samp{G} packet
40794 Write general registers. @xref{read registers packet}, for a
40795 description of the @var{XX@dots{}} data.
40805 @item H @var{op} @var{thread-id}
40806 @cindex @samp{H} packet
40807 Set thread for subsequent operations (@samp{m}, @samp{M}, @samp{g},
40808 @samp{G}, et.al.). Depending on the operation to be performed, @var{op}
40809 should be @samp{c} for step and continue operations (note that this
40810 is deprecated, supporting the @samp{vCont} command is a better
40811 option), and @samp{g} for other operations. The thread designator
40812 @var{thread-id} has the format and interpretation described in
40813 @ref{thread-id syntax}.
40824 @c 'H': How restrictive (or permissive) is the thread model. If a
40825 @c thread is selected and stopped, are other threads allowed
40826 @c to continue to execute? As I mentioned above, I think the
40827 @c semantics of each command when a thread is selected must be
40828 @c described. For example:
40830 @c 'g': If the stub supports threads and a specific thread is
40831 @c selected, returns the register block from that thread;
40832 @c otherwise returns current registers.
40834 @c 'G' If the stub supports threads and a specific thread is
40835 @c selected, sets the registers of the register block of
40836 @c that thread; otherwise sets current registers.
40838 @item i @r{[}@var{addr}@r{[},@var{nnn}@r{]]}
40839 @anchor{cycle step packet}
40840 @cindex @samp{i} packet
40841 Step the remote target by a single clock cycle. If @samp{,@var{nnn}} is
40842 present, cycle step @var{nnn} cycles. If @var{addr} is present, cycle
40843 step starting at that address.
40846 @cindex @samp{I} packet
40847 Signal, then cycle step. @xref{step with signal packet}. @xref{cycle
40851 @cindex @samp{k} packet
40854 The exact effect of this packet is not specified.
40856 For a bare-metal target, it may power cycle or reset the target
40857 system. For that reason, the @samp{k} packet has no reply.
40859 For a single-process target, it may kill that process if possible.
40861 A multiple-process target may choose to kill just one process, or all
40862 that are under @value{GDBN}'s control. For more precise control, use
40863 the vKill packet (@pxref{vKill packet}).
40865 If the target system immediately closes the connection in response to
40866 @samp{k}, @value{GDBN} does not consider the lack of packet
40867 acknowledgment to be an error, and assumes the kill was successful.
40869 If connected using @kbd{target extended-remote}, and the target does
40870 not close the connection in response to a kill request, @value{GDBN}
40871 probes the target state as if a new connection was opened
40872 (@pxref{? packet}).
40874 @item m @var{addr},@var{length}
40875 @cindex @samp{m} packet
40876 Read @var{length} addressable memory units starting at address @var{addr}
40877 (@pxref{addressable memory unit}). Note that @var{addr} may not be aligned to
40878 any particular boundary.
40880 The stub need not use any particular size or alignment when gathering
40881 data from memory for the response; even if @var{addr} is word-aligned
40882 and @var{length} is a multiple of the word size, the stub is free to
40883 use byte accesses, or not. For this reason, this packet may not be
40884 suitable for accessing memory-mapped I/O devices.
40885 @cindex alignment of remote memory accesses
40886 @cindex size of remote memory accesses
40887 @cindex memory, alignment and size of remote accesses
40891 @item @var{XX@dots{}}
40892 Memory contents; each byte is transmitted as a two-digit hexadecimal number.
40893 The reply may contain fewer addressable memory units than requested if the
40894 server was able to read only part of the region of memory.
40899 @item M @var{addr},@var{length}:@var{XX@dots{}}
40900 @cindex @samp{M} packet
40901 Write @var{length} addressable memory units starting at address @var{addr}
40902 (@pxref{addressable memory unit}). The data is given by @var{XX@dots{}}; each
40903 byte is transmitted as a two-digit hexadecimal number.
40910 for an error (this includes the case where only part of the data was
40915 @cindex @samp{p} packet
40916 Read the value of register @var{n}; @var{n} is in hex.
40917 @xref{read registers packet}, for a description of how the returned
40918 register value is encoded.
40922 @item @var{XX@dots{}}
40923 the register's value
40927 Indicating an unrecognized @var{query}.
40930 @item P @var{n@dots{}}=@var{r@dots{}}
40931 @anchor{write register packet}
40932 @cindex @samp{P} packet
40933 Write register @var{n@dots{}} with value @var{r@dots{}}. The register
40934 number @var{n} is in hexadecimal, and @var{r@dots{}} contains two hex
40935 digits for each byte in the register (target byte order).
40945 @item q @var{name} @var{params}@dots{}
40946 @itemx Q @var{name} @var{params}@dots{}
40947 @cindex @samp{q} packet
40948 @cindex @samp{Q} packet
40949 General query (@samp{q}) and set (@samp{Q}). These packets are
40950 described fully in @ref{General Query Packets}.
40953 @cindex @samp{r} packet
40954 Reset the entire system.
40956 Don't use this packet; use the @samp{R} packet instead.
40959 @cindex @samp{R} packet
40960 Restart the program being debugged. The @var{XX}, while needed, is ignored.
40961 This packet is only available in extended mode (@pxref{extended mode}).
40963 The @samp{R} packet has no reply.
40965 @item s @r{[}@var{addr}@r{]}
40966 @cindex @samp{s} packet
40967 Single step, resuming at @var{addr}. If
40968 @var{addr} is omitted, resume at same address.
40970 This packet is deprecated for multi-threading support. @xref{vCont
40974 @xref{Stop Reply Packets}, for the reply specifications.
40976 @item S @var{sig}@r{[};@var{addr}@r{]}
40977 @anchor{step with signal packet}
40978 @cindex @samp{S} packet
40979 Step with signal. This is analogous to the @samp{C} packet, but
40980 requests a single-step, rather than a normal resumption of execution.
40982 This packet is deprecated for multi-threading support. @xref{vCont
40986 @xref{Stop Reply Packets}, for the reply specifications.
40988 @item t @var{addr}:@var{PP},@var{MM}
40989 @cindex @samp{t} packet
40990 Search backwards starting at address @var{addr} for a match with pattern
40991 @var{PP} and mask @var{MM}, both of which are are 4 byte long.
40992 There must be at least 3 digits in @var{addr}.
40994 @item T @var{thread-id}
40995 @cindex @samp{T} packet
40996 Find out if the thread @var{thread-id} is alive. @xref{thread-id syntax}.
41001 thread is still alive
41007 Packets starting with @samp{v} are identified by a multi-letter name,
41008 up to the first @samp{;} or @samp{?} (or the end of the packet).
41010 @item vAttach;@var{pid}
41011 @cindex @samp{vAttach} packet
41012 Attach to a new process with the specified process ID @var{pid}.
41013 The process ID is a
41014 hexadecimal integer identifying the process. In all-stop mode, all
41015 threads in the attached process are stopped; in non-stop mode, it may be
41016 attached without being stopped if that is supported by the target.
41018 @c In non-stop mode, on a successful vAttach, the stub should set the
41019 @c current thread to a thread of the newly-attached process. After
41020 @c attaching, GDB queries for the attached process's thread ID with qC.
41021 @c Also note that, from a user perspective, whether or not the
41022 @c target is stopped on attach in non-stop mode depends on whether you
41023 @c use the foreground or background version of the attach command, not
41024 @c on what vAttach does; GDB does the right thing with respect to either
41025 @c stopping or restarting threads.
41027 This packet is only available in extended mode (@pxref{extended mode}).
41033 @item @r{Any stop packet}
41034 for success in all-stop mode (@pxref{Stop Reply Packets})
41036 for success in non-stop mode (@pxref{Remote Non-Stop})
41039 @item vCont@r{[};@var{action}@r{[}:@var{thread-id}@r{]]}@dots{}
41040 @cindex @samp{vCont} packet
41041 @anchor{vCont packet}
41042 Resume the inferior, specifying different actions for each thread.
41044 For each inferior thread, the leftmost action with a matching
41045 @var{thread-id} is applied. Threads that don't match any action
41046 remain in their current state. Thread IDs are specified using the
41047 syntax described in @ref{thread-id syntax}. If multiprocess
41048 extensions (@pxref{multiprocess extensions}) are supported, actions
41049 can be specified to match all threads in a process by using the
41050 @samp{p@var{pid}.-1} form of the @var{thread-id}. An action with no
41051 @var{thread-id} matches all threads. Specifying no actions is an
41054 Currently supported actions are:
41060 Continue with signal @var{sig}. The signal @var{sig} should be two hex digits.
41064 Step with signal @var{sig}. The signal @var{sig} should be two hex digits.
41067 @item r @var{start},@var{end}
41068 Step once, and then keep stepping as long as the thread stops at
41069 addresses between @var{start} (inclusive) and @var{end} (exclusive).
41070 The remote stub reports a stop reply when either the thread goes out
41071 of the range or is stopped due to an unrelated reason, such as hitting
41072 a breakpoint. @xref{range stepping}.
41074 If the range is empty (@var{start} == @var{end}), then the action
41075 becomes equivalent to the @samp{s} action. In other words,
41076 single-step once, and report the stop (even if the stepped instruction
41077 jumps to @var{start}).
41079 (A stop reply may be sent at any point even if the PC is still within
41080 the stepping range; for example, it is valid to implement this packet
41081 in a degenerate way as a single instruction step operation.)
41085 The optional argument @var{addr} normally associated with the
41086 @samp{c}, @samp{C}, @samp{s}, and @samp{S} packets is
41087 not supported in @samp{vCont}.
41089 The @samp{t} action is only relevant in non-stop mode
41090 (@pxref{Remote Non-Stop}) and may be ignored by the stub otherwise.
41091 A stop reply should be generated for any affected thread not already stopped.
41092 When a thread is stopped by means of a @samp{t} action,
41093 the corresponding stop reply should indicate that the thread has stopped with
41094 signal @samp{0}, regardless of whether the target uses some other signal
41095 as an implementation detail.
41097 The server must ignore @samp{c}, @samp{C}, @samp{s}, @samp{S}, and
41098 @samp{r} actions for threads that are already running. Conversely,
41099 the server must ignore @samp{t} actions for threads that are already
41102 @emph{Note:} In non-stop mode, a thread is considered running until
41103 @value{GDBN} acknowledges an asynchronous stop notification for it with
41104 the @samp{vStopped} packet (@pxref{Remote Non-Stop}).
41106 The stub must support @samp{vCont} if it reports support for
41107 multiprocess extensions (@pxref{multiprocess extensions}).
41110 @xref{Stop Reply Packets}, for the reply specifications.
41113 @cindex @samp{vCont?} packet
41114 Request a list of actions supported by the @samp{vCont} packet.
41118 @item vCont@r{[};@var{action}@dots{}@r{]}
41119 The @samp{vCont} packet is supported. Each @var{action} is a supported
41120 command in the @samp{vCont} packet.
41122 The @samp{vCont} packet is not supported.
41125 @anchor{vCtrlC packet}
41127 @cindex @samp{vCtrlC} packet
41128 Interrupt remote target as if a control-C was pressed on the remote
41129 terminal. This is the equivalent to reacting to the @code{^C}
41130 (@samp{\003}, the control-C character) character in all-stop mode
41131 while the target is running, except this works in non-stop mode.
41132 @xref{interrupting remote targets}, for more info on the all-stop
41143 @item vFile:@var{operation}:@var{parameter}@dots{}
41144 @cindex @samp{vFile} packet
41145 Perform a file operation on the target system. For details,
41146 see @ref{Host I/O Packets}.
41148 @item vFlashErase:@var{addr},@var{length}
41149 @cindex @samp{vFlashErase} packet
41150 Direct the stub to erase @var{length} bytes of flash starting at
41151 @var{addr}. The region may enclose any number of flash blocks, but
41152 its start and end must fall on block boundaries, as indicated by the
41153 flash block size appearing in the memory map (@pxref{Memory Map
41154 Format}). @value{GDBN} groups flash memory programming operations
41155 together, and sends a @samp{vFlashDone} request after each group; the
41156 stub is allowed to delay erase operation until the @samp{vFlashDone}
41157 packet is received.
41167 @item vFlashWrite:@var{addr}:@var{XX@dots{}}
41168 @cindex @samp{vFlashWrite} packet
41169 Direct the stub to write data to flash address @var{addr}. The data
41170 is passed in binary form using the same encoding as for the @samp{X}
41171 packet (@pxref{Binary Data}). The memory ranges specified by
41172 @samp{vFlashWrite} packets preceding a @samp{vFlashDone} packet must
41173 not overlap, and must appear in order of increasing addresses
41174 (although @samp{vFlashErase} packets for higher addresses may already
41175 have been received; the ordering is guaranteed only between
41176 @samp{vFlashWrite} packets). If a packet writes to an address that was
41177 neither erased by a preceding @samp{vFlashErase} packet nor by some other
41178 target-specific method, the results are unpredictable.
41186 for vFlashWrite addressing non-flash memory
41192 @cindex @samp{vFlashDone} packet
41193 Indicate to the stub that flash programming operation is finished.
41194 The stub is permitted to delay or batch the effects of a group of
41195 @samp{vFlashErase} and @samp{vFlashWrite} packets until a
41196 @samp{vFlashDone} packet is received. The contents of the affected
41197 regions of flash memory are unpredictable until the @samp{vFlashDone}
41198 request is completed.
41200 @item vKill;@var{pid}
41201 @cindex @samp{vKill} packet
41202 @anchor{vKill packet}
41203 Kill the process with the specified process ID @var{pid}, which is a
41204 hexadecimal integer identifying the process. This packet is used in
41205 preference to @samp{k} when multiprocess protocol extensions are
41206 supported; see @ref{multiprocess extensions}.
41216 @item vMustReplyEmpty
41217 @cindex @samp{vMustReplyEmpty} packet
41218 The correct reply to an unknown @samp{v} packet is to return the empty
41219 string, however, some older versions of @command{gdbserver} would
41220 incorrectly return @samp{OK} for unknown @samp{v} packets.
41222 The @samp{vMustReplyEmpty} is used as a feature test to check how
41223 @command{gdbserver} handles unknown packets, it is important that this
41224 packet be handled in the same way as other unknown @samp{v} packets.
41225 If this packet is handled differently to other unknown @samp{v}
41226 packets then it is possible that @value{GDBN} may run into problems in
41227 other areas, specifically around use of @samp{vFile:setfs:}.
41229 @item vRun;@var{filename}@r{[};@var{argument}@r{]}@dots{}
41230 @cindex @samp{vRun} packet
41231 Run the program @var{filename}, passing it each @var{argument} on its
41232 command line. The file and arguments are hex-encoded strings. If
41233 @var{filename} is an empty string, the stub may use a default program
41234 (e.g.@: the last program run). The program is created in the stopped
41237 @c FIXME: What about non-stop mode?
41239 This packet is only available in extended mode (@pxref{extended mode}).
41245 @item @r{Any stop packet}
41246 for success (@pxref{Stop Reply Packets})
41250 @cindex @samp{vStopped} packet
41251 @xref{Notification Packets}.
41253 @item X @var{addr},@var{length}:@var{XX@dots{}}
41255 @cindex @samp{X} packet
41256 Write data to memory, where the data is transmitted in binary.
41257 Memory is specified by its address @var{addr} and number of addressable memory
41258 units @var{length} (@pxref{addressable memory unit});
41259 @samp{@var{XX}@dots{}} is binary data (@pxref{Binary Data}).
41269 @item z @var{type},@var{addr},@var{kind}
41270 @itemx Z @var{type},@var{addr},@var{kind}
41271 @anchor{insert breakpoint or watchpoint packet}
41272 @cindex @samp{z} packet
41273 @cindex @samp{Z} packets
41274 Insert (@samp{Z}) or remove (@samp{z}) a @var{type} breakpoint or
41275 watchpoint starting at address @var{address} of kind @var{kind}.
41277 Each breakpoint and watchpoint packet @var{type} is documented
41280 @emph{Implementation notes: A remote target shall return an empty string
41281 for an unrecognized breakpoint or watchpoint packet @var{type}. A
41282 remote target shall support either both or neither of a given
41283 @samp{Z@var{type}@dots{}} and @samp{z@var{type}@dots{}} packet pair. To
41284 avoid potential problems with duplicate packets, the operations should
41285 be implemented in an idempotent way.}
41287 @item z0,@var{addr},@var{kind}
41288 @itemx Z0,@var{addr},@var{kind}@r{[};@var{cond_list}@dots{}@r{]}@r{[};cmds:@var{persist},@var{cmd_list}@dots{}@r{]}
41289 @cindex @samp{z0} packet
41290 @cindex @samp{Z0} packet
41291 Insert (@samp{Z0}) or remove (@samp{z0}) a software breakpoint at address
41292 @var{addr} of type @var{kind}.
41294 A software breakpoint is implemented by replacing the instruction at
41295 @var{addr} with a software breakpoint or trap instruction. The
41296 @var{kind} is target-specific and typically indicates the size of the
41297 breakpoint in bytes that should be inserted. E.g., the @sc{arm} and
41298 @sc{mips} can insert either a 2 or 4 byte breakpoint. Some
41299 architectures have additional meanings for @var{kind}
41300 (@pxref{Architecture-Specific Protocol Details}); if no
41301 architecture-specific value is being used, it should be @samp{0}.
41302 @var{kind} is hex-encoded. @var{cond_list} is an optional list of
41303 conditional expressions in bytecode form that should be evaluated on
41304 the target's side. These are the conditions that should be taken into
41305 consideration when deciding if the breakpoint trigger should be
41306 reported back to @value{GDBN}.
41308 See also the @samp{swbreak} stop reason (@pxref{swbreak stop reason})
41309 for how to best report a software breakpoint event to @value{GDBN}.
41311 The @var{cond_list} parameter is comprised of a series of expressions,
41312 concatenated without separators. Each expression has the following form:
41316 @item X @var{len},@var{expr}
41317 @var{len} is the length of the bytecode expression and @var{expr} is the
41318 actual conditional expression in bytecode form.
41322 The optional @var{cmd_list} parameter introduces commands that may be
41323 run on the target, rather than being reported back to @value{GDBN}.
41324 The parameter starts with a numeric flag @var{persist}; if the flag is
41325 nonzero, then the breakpoint may remain active and the commands
41326 continue to be run even when @value{GDBN} disconnects from the target.
41327 Following this flag is a series of expressions concatenated with no
41328 separators. Each expression has the following form:
41332 @item X @var{len},@var{expr}
41333 @var{len} is the length of the bytecode expression and @var{expr} is the
41334 actual commands expression in bytecode form.
41338 @emph{Implementation note: It is possible for a target to copy or move
41339 code that contains software breakpoints (e.g., when implementing
41340 overlays). The behavior of this packet, in the presence of such a
41341 target, is not defined.}
41353 @item z1,@var{addr},@var{kind}
41354 @itemx Z1,@var{addr},@var{kind}@r{[};@var{cond_list}@dots{}@r{]}@r{[};cmds:@var{persist},@var{cmd_list}@dots{}@r{]}
41355 @cindex @samp{z1} packet
41356 @cindex @samp{Z1} packet
41357 Insert (@samp{Z1}) or remove (@samp{z1}) a hardware breakpoint at
41358 address @var{addr}.
41360 A hardware breakpoint is implemented using a mechanism that is not
41361 dependent on being able to modify the target's memory. The
41362 @var{kind}, @var{cond_list}, and @var{cmd_list} arguments have the
41363 same meaning as in @samp{Z0} packets.
41365 @emph{Implementation note: A hardware breakpoint is not affected by code
41378 @item z2,@var{addr},@var{kind}
41379 @itemx Z2,@var{addr},@var{kind}
41380 @cindex @samp{z2} packet
41381 @cindex @samp{Z2} packet
41382 Insert (@samp{Z2}) or remove (@samp{z2}) a write watchpoint at @var{addr}.
41383 The number of bytes to watch is specified by @var{kind}.
41395 @item z3,@var{addr},@var{kind}
41396 @itemx Z3,@var{addr},@var{kind}
41397 @cindex @samp{z3} packet
41398 @cindex @samp{Z3} packet
41399 Insert (@samp{Z3}) or remove (@samp{z3}) a read watchpoint at @var{addr}.
41400 The number of bytes to watch is specified by @var{kind}.
41412 @item z4,@var{addr},@var{kind}
41413 @itemx Z4,@var{addr},@var{kind}
41414 @cindex @samp{z4} packet
41415 @cindex @samp{Z4} packet
41416 Insert (@samp{Z4}) or remove (@samp{z4}) an access watchpoint at @var{addr}.
41417 The number of bytes to watch is specified by @var{kind}.
41431 @node Stop Reply Packets
41432 @section Stop Reply Packets
41433 @cindex stop reply packets
41435 The @samp{C}, @samp{c}, @samp{S}, @samp{s}, @samp{vCont},
41436 @samp{vAttach}, @samp{vRun}, @samp{vStopped}, and @samp{?} packets can
41437 receive any of the below as a reply. Except for @samp{?}
41438 and @samp{vStopped}, that reply is only returned
41439 when the target halts. In the below the exact meaning of @dfn{signal
41440 number} is defined by the header @file{include/gdb/signals.h} in the
41441 @value{GDBN} source code.
41443 In non-stop mode, the server will simply reply @samp{OK} to commands
41444 such as @samp{vCont}; any stop will be the subject of a future
41445 notification. @xref{Remote Non-Stop}.
41447 As in the description of request packets, we include spaces in the
41448 reply templates for clarity; these are not part of the reply packet's
41449 syntax. No @value{GDBN} stop reply packet uses spaces to separate its
41455 The program received signal number @var{AA} (a two-digit hexadecimal
41456 number). This is equivalent to a @samp{T} response with no
41457 @var{n}:@var{r} pairs.
41459 @item T @var{AA} @var{n1}:@var{r1};@var{n2}:@var{r2};@dots{}
41460 @cindex @samp{T} packet reply
41461 The program received signal number @var{AA} (a two-digit hexadecimal
41462 number). This is equivalent to an @samp{S} response, except that the
41463 @samp{@var{n}:@var{r}} pairs can carry values of important registers
41464 and other information directly in the stop reply packet, reducing
41465 round-trip latency. Single-step and breakpoint traps are reported
41466 this way. Each @samp{@var{n}:@var{r}} pair is interpreted as follows:
41470 If @var{n} is a hexadecimal number, it is a register number, and the
41471 corresponding @var{r} gives that register's value. The data @var{r} is a
41472 series of bytes in target byte order, with each byte given by a
41473 two-digit hex number.
41476 If @var{n} is @samp{thread}, then @var{r} is the @var{thread-id} of
41477 the stopped thread, as specified in @ref{thread-id syntax}.
41480 If @var{n} is @samp{core}, then @var{r} is the hexadecimal number of
41481 the core on which the stop event was detected.
41484 If @var{n} is a recognized @dfn{stop reason}, it describes a more
41485 specific event that stopped the target. The currently defined stop
41486 reasons are listed below. The @var{aa} should be @samp{05}, the trap
41487 signal. At most one stop reason should be present.
41490 Otherwise, @value{GDBN} should ignore this @samp{@var{n}:@var{r}} pair
41491 and go on to the next; this allows us to extend the protocol in the
41495 The currently defined stop reasons are:
41501 The packet indicates a watchpoint hit, and @var{r} is the data address, in
41504 @item syscall_entry
41505 @itemx syscall_return
41506 The packet indicates a syscall entry or return, and @var{r} is the
41507 syscall number, in hex.
41509 @cindex shared library events, remote reply
41511 The packet indicates that the loaded libraries have changed.
41512 @value{GDBN} should use @samp{qXfer:libraries:read} to fetch a new
41513 list of loaded libraries. The @var{r} part is ignored.
41515 @cindex replay log events, remote reply
41517 The packet indicates that the target cannot continue replaying
41518 logged execution events, because it has reached the end (or the
41519 beginning when executing backward) of the log. The value of @var{r}
41520 will be either @samp{begin} or @samp{end}. @xref{Reverse Execution},
41521 for more information.
41524 @anchor{swbreak stop reason}
41525 The packet indicates a software breakpoint instruction was executed,
41526 irrespective of whether it was @value{GDBN} that planted the
41527 breakpoint or the breakpoint is hardcoded in the program. The @var{r}
41528 part must be left empty.
41530 On some architectures, such as x86, at the architecture level, when a
41531 breakpoint instruction executes the program counter points at the
41532 breakpoint address plus an offset. On such targets, the stub is
41533 responsible for adjusting the PC to point back at the breakpoint
41536 This packet should not be sent by default; older @value{GDBN} versions
41537 did not support it. @value{GDBN} requests it, by supplying an
41538 appropriate @samp{qSupported} feature (@pxref{qSupported}). The
41539 remote stub must also supply the appropriate @samp{qSupported} feature
41540 indicating support.
41542 This packet is required for correct non-stop mode operation.
41545 The packet indicates the target stopped for a hardware breakpoint.
41546 The @var{r} part must be left empty.
41548 The same remarks about @samp{qSupported} and non-stop mode above
41551 @cindex fork events, remote reply
41553 The packet indicates that @code{fork} was called, and @var{r}
41554 is the thread ID of the new child process. Refer to
41555 @ref{thread-id syntax} for the format of the @var{thread-id}
41556 field. This packet is only applicable to targets that support
41559 This packet should not be sent by default; older @value{GDBN} versions
41560 did not support it. @value{GDBN} requests it, by supplying an
41561 appropriate @samp{qSupported} feature (@pxref{qSupported}). The
41562 remote stub must also supply the appropriate @samp{qSupported} feature
41563 indicating support.
41565 @cindex vfork events, remote reply
41567 The packet indicates that @code{vfork} was called, and @var{r}
41568 is the thread ID of the new child process. Refer to
41569 @ref{thread-id syntax} for the format of the @var{thread-id}
41570 field. This packet is only applicable to targets that support
41573 This packet should not be sent by default; older @value{GDBN} versions
41574 did not support it. @value{GDBN} requests it, by supplying an
41575 appropriate @samp{qSupported} feature (@pxref{qSupported}). The
41576 remote stub must also supply the appropriate @samp{qSupported} feature
41577 indicating support.
41579 @cindex vforkdone events, remote reply
41581 The packet indicates that a child process created by a vfork
41582 has either called @code{exec} or terminated, so that the
41583 address spaces of the parent and child process are no longer
41584 shared. The @var{r} part is ignored. This packet is only
41585 applicable to targets that support vforkdone events.
41587 This packet should not be sent by default; older @value{GDBN} versions
41588 did not support it. @value{GDBN} requests it, by supplying an
41589 appropriate @samp{qSupported} feature (@pxref{qSupported}). The
41590 remote stub must also supply the appropriate @samp{qSupported} feature
41591 indicating support.
41593 @cindex exec events, remote reply
41595 The packet indicates that @code{execve} was called, and @var{r}
41596 is the absolute pathname of the file that was executed, in hex.
41597 This packet is only applicable to targets that support exec events.
41599 This packet should not be sent by default; older @value{GDBN} versions
41600 did not support it. @value{GDBN} requests it, by supplying an
41601 appropriate @samp{qSupported} feature (@pxref{qSupported}). The
41602 remote stub must also supply the appropriate @samp{qSupported} feature
41603 indicating support.
41605 @cindex thread create event, remote reply
41606 @anchor{thread create event}
41608 The packet indicates that the thread was just created. The new thread
41609 is stopped until @value{GDBN} sets it running with a resumption packet
41610 (@pxref{vCont packet}). This packet should not be sent by default;
41611 @value{GDBN} requests it with the @ref{QThreadEvents} packet. See
41612 also the @samp{w} (@pxref{thread exit event}) remote reply below. The
41613 @var{r} part is ignored.
41618 @itemx W @var{AA} ; process:@var{pid}
41619 The process exited, and @var{AA} is the exit status. This is only
41620 applicable to certain targets.
41622 The second form of the response, including the process ID of the
41623 exited process, can be used only when @value{GDBN} has reported
41624 support for multiprocess protocol extensions; see @ref{multiprocess
41625 extensions}. Both @var{AA} and @var{pid} are formatted as big-endian
41629 @itemx X @var{AA} ; process:@var{pid}
41630 The process terminated with signal @var{AA}.
41632 The second form of the response, including the process ID of the
41633 terminated process, can be used only when @value{GDBN} has reported
41634 support for multiprocess protocol extensions; see @ref{multiprocess
41635 extensions}. Both @var{AA} and @var{pid} are formatted as big-endian
41638 @anchor{thread exit event}
41639 @cindex thread exit event, remote reply
41640 @item w @var{AA} ; @var{tid}
41642 The thread exited, and @var{AA} is the exit status. This response
41643 should not be sent by default; @value{GDBN} requests it with the
41644 @ref{QThreadEvents} packet. See also @ref{thread create event} above.
41645 @var{AA} is formatted as a big-endian hex string.
41648 There are no resumed threads left in the target. In other words, even
41649 though the process is alive, the last resumed thread has exited. For
41650 example, say the target process has two threads: thread 1 and thread
41651 2. The client leaves thread 1 stopped, and resumes thread 2, which
41652 subsequently exits. At this point, even though the process is still
41653 alive, and thus no @samp{W} stop reply is sent, no thread is actually
41654 executing either. The @samp{N} stop reply thus informs the client
41655 that it can stop waiting for stop replies. This packet should not be
41656 sent by default; older @value{GDBN} versions did not support it.
41657 @value{GDBN} requests it, by supplying an appropriate
41658 @samp{qSupported} feature (@pxref{qSupported}). The remote stub must
41659 also supply the appropriate @samp{qSupported} feature indicating
41662 @item O @var{XX}@dots{}
41663 @samp{@var{XX}@dots{}} is hex encoding of @sc{ascii} data, to be
41664 written as the program's console output. This can happen at any time
41665 while the program is running and the debugger should continue to wait
41666 for @samp{W}, @samp{T}, etc. This reply is not permitted in non-stop mode.
41668 @item F @var{call-id},@var{parameter}@dots{}
41669 @var{call-id} is the identifier which says which host system call should
41670 be called. This is just the name of the function. Translation into the
41671 correct system call is only applicable as it's defined in @value{GDBN}.
41672 @xref{File-I/O Remote Protocol Extension}, for a list of implemented
41675 @samp{@var{parameter}@dots{}} is a list of parameters as defined for
41676 this very system call.
41678 The target replies with this packet when it expects @value{GDBN} to
41679 call a host system call on behalf of the target. @value{GDBN} replies
41680 with an appropriate @samp{F} packet and keeps up waiting for the next
41681 reply packet from the target. The latest @samp{C}, @samp{c}, @samp{S}
41682 or @samp{s} action is expected to be continued. @xref{File-I/O Remote
41683 Protocol Extension}, for more details.
41687 @node General Query Packets
41688 @section General Query Packets
41689 @cindex remote query requests
41691 Packets starting with @samp{q} are @dfn{general query packets};
41692 packets starting with @samp{Q} are @dfn{general set packets}. General
41693 query and set packets are a semi-unified form for retrieving and
41694 sending information to and from the stub.
41696 The initial letter of a query or set packet is followed by a name
41697 indicating what sort of thing the packet applies to. For example,
41698 @value{GDBN} may use a @samp{qSymbol} packet to exchange symbol
41699 definitions with the stub. These packet names follow some
41704 The name must not contain commas, colons or semicolons.
41706 Most @value{GDBN} query and set packets have a leading upper case
41709 The names of custom vendor packets should use a company prefix, in
41710 lower case, followed by a period. For example, packets designed at
41711 the Acme Corporation might begin with @samp{qacme.foo} (for querying
41712 foos) or @samp{Qacme.bar} (for setting bars).
41715 The name of a query or set packet should be separated from any
41716 parameters by a @samp{:}; the parameters themselves should be
41717 separated by @samp{,} or @samp{;}. Stubs must be careful to match the
41718 full packet name, and check for a separator or the end of the packet,
41719 in case two packet names share a common prefix. New packets should not begin
41720 with @samp{qC}, @samp{qP}, or @samp{qL}@footnote{The @samp{qP} and @samp{qL}
41721 packets predate these conventions, and have arguments without any terminator
41722 for the packet name; we suspect they are in widespread use in places that
41723 are difficult to upgrade. The @samp{qC} packet has no arguments, but some
41724 existing stubs (e.g.@: RedBoot) are known to not check for the end of the
41727 Like the descriptions of the other packets, each description here
41728 has a template showing the packet's overall syntax, followed by an
41729 explanation of the packet's meaning. We include spaces in some of the
41730 templates for clarity; these are not part of the packet's syntax. No
41731 @value{GDBN} packet uses spaces to separate its components.
41733 Here are the currently defined query and set packets:
41739 Turn on or off the agent as a helper to perform some debugging operations
41740 delegated from @value{GDBN} (@pxref{Control Agent}).
41742 @item QAllow:@var{op}:@var{val}@dots{}
41743 @cindex @samp{QAllow} packet
41744 Specify which operations @value{GDBN} expects to request of the
41745 target, as a semicolon-separated list of operation name and value
41746 pairs. Possible values for @var{op} include @samp{WriteReg},
41747 @samp{WriteMem}, @samp{InsertBreak}, @samp{InsertTrace},
41748 @samp{InsertFastTrace}, and @samp{Stop}. @var{val} is either 0,
41749 indicating that @value{GDBN} will not request the operation, or 1,
41750 indicating that it may. (The target can then use this to set up its
41751 own internals optimally, for instance if the debugger never expects to
41752 insert breakpoints, it may not need to install its own trap handler.)
41755 @cindex current thread, remote request
41756 @cindex @samp{qC} packet
41757 Return the current thread ID.
41761 @item QC @var{thread-id}
41762 Where @var{thread-id} is a thread ID as documented in
41763 @ref{thread-id syntax}.
41764 @item @r{(anything else)}
41765 Any other reply implies the old thread ID.
41768 @item qCRC:@var{addr},@var{length}
41769 @cindex CRC of memory block, remote request
41770 @cindex @samp{qCRC} packet
41771 @anchor{qCRC packet}
41772 Compute the CRC checksum of a block of memory using CRC-32 defined in
41773 IEEE 802.3. The CRC is computed byte at a time, taking the most
41774 significant bit of each byte first. The initial pattern code
41775 @code{0xffffffff} is used to ensure leading zeros affect the CRC.
41777 @emph{Note:} This is the same CRC used in validating separate debug
41778 files (@pxref{Separate Debug Files, , Debugging Information in Separate
41779 Files}). However the algorithm is slightly different. When validating
41780 separate debug files, the CRC is computed taking the @emph{least}
41781 significant bit of each byte first, and the final result is inverted to
41782 detect trailing zeros.
41787 An error (such as memory fault)
41788 @item C @var{crc32}
41789 The specified memory region's checksum is @var{crc32}.
41792 @item QDisableRandomization:@var{value}
41793 @cindex disable address space randomization, remote request
41794 @cindex @samp{QDisableRandomization} packet
41795 Some target operating systems will randomize the virtual address space
41796 of the inferior process as a security feature, but provide a feature
41797 to disable such randomization, e.g.@: to allow for a more deterministic
41798 debugging experience. On such systems, this packet with a @var{value}
41799 of 1 directs the target to disable address space randomization for
41800 processes subsequently started via @samp{vRun} packets, while a packet
41801 with a @var{value} of 0 tells the target to enable address space
41804 This packet is only available in extended mode (@pxref{extended mode}).
41809 The request succeeded.
41812 An error occurred. The error number @var{nn} is given as hex digits.
41815 An empty reply indicates that @samp{QDisableRandomization} is not supported
41819 This packet is not probed by default; the remote stub must request it,
41820 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
41821 This should only be done on targets that actually support disabling
41822 address space randomization.
41824 @item QStartupWithShell:@var{value}
41825 @cindex startup with shell, remote request
41826 @cindex @samp{QStartupWithShell} packet
41827 On UNIX-like targets, it is possible to start the inferior using a
41828 shell program. This is the default behavior on both @value{GDBN} and
41829 @command{gdbserver} (@pxref{set startup-with-shell}). This packet is
41830 used to inform @command{gdbserver} whether it should start the
41831 inferior using a shell or not.
41833 If @var{value} is @samp{0}, @command{gdbserver} will not use a shell
41834 to start the inferior. If @var{value} is @samp{1},
41835 @command{gdbserver} will use a shell to start the inferior. All other
41836 values are considered an error.
41838 This packet is only available in extended mode (@pxref{extended
41844 The request succeeded.
41847 An error occurred. The error number @var{nn} is given as hex digits.
41850 This packet is not probed by default; the remote stub must request it,
41851 by supplying an appropriate @samp{qSupported} response
41852 (@pxref{qSupported}). This should only be done on targets that
41853 actually support starting the inferior using a shell.
41855 Use of this packet is controlled by the @code{set startup-with-shell}
41856 command; @pxref{set startup-with-shell}.
41858 @item QEnvironmentHexEncoded:@var{hex-value}
41859 @anchor{QEnvironmentHexEncoded}
41860 @cindex set environment variable, remote request
41861 @cindex @samp{QEnvironmentHexEncoded} packet
41862 On UNIX-like targets, it is possible to set environment variables that
41863 will be passed to the inferior during the startup process. This
41864 packet is used to inform @command{gdbserver} of an environment
41865 variable that has been defined by the user on @value{GDBN} (@pxref{set
41868 The packet is composed by @var{hex-value}, an hex encoded
41869 representation of the @var{name=value} format representing an
41870 environment variable. The name of the environment variable is
41871 represented by @var{name}, and the value to be assigned to the
41872 environment variable is represented by @var{value}. If the variable
41873 has no value (i.e., the value is @code{null}), then @var{value} will
41876 This packet is only available in extended mode (@pxref{extended
41882 The request succeeded.
41885 This packet is not probed by default; the remote stub must request it,
41886 by supplying an appropriate @samp{qSupported} response
41887 (@pxref{qSupported}). This should only be done on targets that
41888 actually support passing environment variables to the starting
41891 This packet is related to the @code{set environment} command;
41892 @pxref{set environment}.
41894 @item QEnvironmentUnset:@var{hex-value}
41895 @anchor{QEnvironmentUnset}
41896 @cindex unset environment variable, remote request
41897 @cindex @samp{QEnvironmentUnset} packet
41898 On UNIX-like targets, it is possible to unset environment variables
41899 before starting the inferior in the remote target. This packet is
41900 used to inform @command{gdbserver} of an environment variable that has
41901 been unset by the user on @value{GDBN} (@pxref{unset environment}).
41903 The packet is composed by @var{hex-value}, an hex encoded
41904 representation of the name of the environment variable to be unset.
41906 This packet is only available in extended mode (@pxref{extended
41912 The request succeeded.
41915 This packet is not probed by default; the remote stub must request it,
41916 by supplying an appropriate @samp{qSupported} response
41917 (@pxref{qSupported}). This should only be done on targets that
41918 actually support passing environment variables to the starting
41921 This packet is related to the @code{unset environment} command;
41922 @pxref{unset environment}.
41924 @item QEnvironmentReset
41925 @anchor{QEnvironmentReset}
41926 @cindex reset environment, remote request
41927 @cindex @samp{QEnvironmentReset} packet
41928 On UNIX-like targets, this packet is used to reset the state of
41929 environment variables in the remote target before starting the
41930 inferior. In this context, reset means unsetting all environment
41931 variables that were previously set by the user (i.e., were not
41932 initially present in the environment). It is sent to
41933 @command{gdbserver} before the @samp{QEnvironmentHexEncoded}
41934 (@pxref{QEnvironmentHexEncoded}) and the @samp{QEnvironmentUnset}
41935 (@pxref{QEnvironmentUnset}) packets.
41937 This packet is only available in extended mode (@pxref{extended
41943 The request succeeded.
41946 This packet is not probed by default; the remote stub must request it,
41947 by supplying an appropriate @samp{qSupported} response
41948 (@pxref{qSupported}). This should only be done on targets that
41949 actually support passing environment variables to the starting
41952 @item QSetWorkingDir:@r{[}@var{directory}@r{]}
41953 @anchor{QSetWorkingDir packet}
41954 @cindex set working directory, remote request
41955 @cindex @samp{QSetWorkingDir} packet
41956 This packet is used to inform the remote server of the intended
41957 current working directory for programs that are going to be executed.
41959 The packet is composed by @var{directory}, an hex encoded
41960 representation of the directory that the remote inferior will use as
41961 its current working directory. If @var{directory} is an empty string,
41962 the remote server should reset the inferior's current working
41963 directory to its original, empty value.
41965 This packet is only available in extended mode (@pxref{extended
41971 The request succeeded.
41975 @itemx qsThreadInfo
41976 @cindex list active threads, remote request
41977 @cindex @samp{qfThreadInfo} packet
41978 @cindex @samp{qsThreadInfo} packet
41979 Obtain a list of all active thread IDs from the target (OS). Since there
41980 may be too many active threads to fit into one reply packet, this query
41981 works iteratively: it may require more than one query/reply sequence to
41982 obtain the entire list of threads. The first query of the sequence will
41983 be the @samp{qfThreadInfo} query; subsequent queries in the
41984 sequence will be the @samp{qsThreadInfo} query.
41986 NOTE: This packet replaces the @samp{qL} query (see below).
41990 @item m @var{thread-id}
41992 @item m @var{thread-id},@var{thread-id}@dots{}
41993 a comma-separated list of thread IDs
41995 (lower case letter @samp{L}) denotes end of list.
41998 In response to each query, the target will reply with a list of one or
41999 more thread IDs, separated by commas.
42000 @value{GDBN} will respond to each reply with a request for more thread
42001 ids (using the @samp{qs} form of the query), until the target responds
42002 with @samp{l} (lower-case ell, for @dfn{last}).
42003 Refer to @ref{thread-id syntax}, for the format of the @var{thread-id}
42006 @emph{Note: @value{GDBN} will send the @code{qfThreadInfo} query during the
42007 initial connection with the remote target, and the very first thread ID
42008 mentioned in the reply will be stopped by @value{GDBN} in a subsequent
42009 message. Therefore, the stub should ensure that the first thread ID in
42010 the @code{qfThreadInfo} reply is suitable for being stopped by @value{GDBN}.}
42012 @item qGetTLSAddr:@var{thread-id},@var{offset},@var{lm}
42013 @cindex get thread-local storage address, remote request
42014 @cindex @samp{qGetTLSAddr} packet
42015 Fetch the address associated with thread local storage specified
42016 by @var{thread-id}, @var{offset}, and @var{lm}.
42018 @var{thread-id} is the thread ID associated with the
42019 thread for which to fetch the TLS address. @xref{thread-id syntax}.
42021 @var{offset} is the (big endian, hex encoded) offset associated with the
42022 thread local variable. (This offset is obtained from the debug
42023 information associated with the variable.)
42025 @var{lm} is the (big endian, hex encoded) OS/ABI-specific encoding of the
42026 load module associated with the thread local storage. For example,
42027 a @sc{gnu}/Linux system will pass the link map address of the shared
42028 object associated with the thread local storage under consideration.
42029 Other operating environments may choose to represent the load module
42030 differently, so the precise meaning of this parameter will vary.
42034 @item @var{XX}@dots{}
42035 Hex encoded (big endian) bytes representing the address of the thread
42036 local storage requested.
42039 An error occurred. The error number @var{nn} is given as hex digits.
42042 An empty reply indicates that @samp{qGetTLSAddr} is not supported by the stub.
42045 @item qGetTIBAddr:@var{thread-id}
42046 @cindex get thread information block address
42047 @cindex @samp{qGetTIBAddr} packet
42048 Fetch address of the Windows OS specific Thread Information Block.
42050 @var{thread-id} is the thread ID associated with the thread.
42054 @item @var{XX}@dots{}
42055 Hex encoded (big endian) bytes representing the linear address of the
42056 thread information block.
42059 An error occured. This means that either the thread was not found, or the
42060 address could not be retrieved.
42063 An empty reply indicates that @samp{qGetTIBAddr} is not supported by the stub.
42066 @item qL @var{startflag} @var{threadcount} @var{nextthread}
42067 Obtain thread information from RTOS. Where: @var{startflag} (one hex
42068 digit) is one to indicate the first query and zero to indicate a
42069 subsequent query; @var{threadcount} (two hex digits) is the maximum
42070 number of threads the response packet can contain; and @var{nextthread}
42071 (eight hex digits), for subsequent queries (@var{startflag} is zero), is
42072 returned in the response as @var{argthread}.
42074 Don't use this packet; use the @samp{qfThreadInfo} query instead (see above).
42078 @item qM @var{count} @var{done} @var{argthread} @var{thread}@dots{}
42079 Where: @var{count} (two hex digits) is the number of threads being
42080 returned; @var{done} (one hex digit) is zero to indicate more threads
42081 and one indicates no further threads; @var{argthreadid} (eight hex
42082 digits) is @var{nextthread} from the request packet; @var{thread}@dots{}
42083 is a sequence of thread IDs, @var{threadid} (eight hex
42084 digits), from the target. See @code{remote.c:parse_threadlist_response()}.
42088 @cindex section offsets, remote request
42089 @cindex @samp{qOffsets} packet
42090 Get section offsets that the target used when relocating the downloaded
42095 @item Text=@var{xxx};Data=@var{yyy}@r{[};Bss=@var{zzz}@r{]}
42096 Relocate the @code{Text} section by @var{xxx} from its original address.
42097 Relocate the @code{Data} section by @var{yyy} from its original address.
42098 If the object file format provides segment information (e.g.@: @sc{elf}
42099 @samp{PT_LOAD} program headers), @value{GDBN} will relocate entire
42100 segments by the supplied offsets.
42102 @emph{Note: while a @code{Bss} offset may be included in the response,
42103 @value{GDBN} ignores this and instead applies the @code{Data} offset
42104 to the @code{Bss} section.}
42106 @item TextSeg=@var{xxx}@r{[};DataSeg=@var{yyy}@r{]}
42107 Relocate the first segment of the object file, which conventionally
42108 contains program code, to a starting address of @var{xxx}. If
42109 @samp{DataSeg} is specified, relocate the second segment, which
42110 conventionally contains modifiable data, to a starting address of
42111 @var{yyy}. @value{GDBN} will report an error if the object file
42112 does not contain segment information, or does not contain at least
42113 as many segments as mentioned in the reply. Extra segments are
42114 kept at fixed offsets relative to the last relocated segment.
42117 @item qP @var{mode} @var{thread-id}
42118 @cindex thread information, remote request
42119 @cindex @samp{qP} packet
42120 Returns information on @var{thread-id}. Where: @var{mode} is a hex
42121 encoded 32 bit mode; @var{thread-id} is a thread ID
42122 (@pxref{thread-id syntax}).
42124 Don't use this packet; use the @samp{qThreadExtraInfo} query instead
42127 Reply: see @code{remote.c:remote_unpack_thread_info_response()}.
42131 @cindex non-stop mode, remote request
42132 @cindex @samp{QNonStop} packet
42134 Enter non-stop (@samp{QNonStop:1}) or all-stop (@samp{QNonStop:0}) mode.
42135 @xref{Remote Non-Stop}, for more information.
42140 The request succeeded.
42143 An error occurred. The error number @var{nn} is given as hex digits.
42146 An empty reply indicates that @samp{QNonStop} is not supported by
42150 This packet is not probed by default; the remote stub must request it,
42151 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
42152 Use of this packet is controlled by the @code{set non-stop} command;
42153 @pxref{Non-Stop Mode}.
42155 @item QCatchSyscalls:1 @r{[};@var{sysno}@r{]}@dots{}
42156 @itemx QCatchSyscalls:0
42157 @cindex catch syscalls from inferior, remote request
42158 @cindex @samp{QCatchSyscalls} packet
42159 @anchor{QCatchSyscalls}
42160 Enable (@samp{QCatchSyscalls:1}) or disable (@samp{QCatchSyscalls:0})
42161 catching syscalls from the inferior process.
42163 For @samp{QCatchSyscalls:1}, each listed syscall @var{sysno} (encoded
42164 in hex) should be reported to @value{GDBN}. If no syscall @var{sysno}
42165 is listed, every system call should be reported.
42167 Note that if a syscall not in the list is reported, @value{GDBN} will
42168 still filter the event according to its own list from all corresponding
42169 @code{catch syscall} commands. However, it is more efficient to only
42170 report the requested syscalls.
42172 Multiple @samp{QCatchSyscalls:1} packets do not combine; any earlier
42173 @samp{QCatchSyscalls:1} list is completely replaced by the new list.
42175 If the inferior process execs, the state of @samp{QCatchSyscalls} is
42176 kept for the new process too. On targets where exec may affect syscall
42177 numbers, for example with exec between 32 and 64-bit processes, the
42178 client should send a new packet with the new syscall list.
42183 The request succeeded.
42186 An error occurred. @var{nn} are hex digits.
42189 An empty reply indicates that @samp{QCatchSyscalls} is not supported by
42193 Use of this packet is controlled by the @code{set remote catch-syscalls}
42194 command (@pxref{Remote Configuration, set remote catch-syscalls}).
42195 This packet is not probed by default; the remote stub must request it,
42196 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
42198 @item QPassSignals: @var{signal} @r{[};@var{signal}@r{]}@dots{}
42199 @cindex pass signals to inferior, remote request
42200 @cindex @samp{QPassSignals} packet
42201 @anchor{QPassSignals}
42202 Each listed @var{signal} should be passed directly to the inferior process.
42203 Signals are numbered identically to continue packets and stop replies
42204 (@pxref{Stop Reply Packets}). Each @var{signal} list item should be
42205 strictly greater than the previous item. These signals do not need to stop
42206 the inferior, or be reported to @value{GDBN}. All other signals should be
42207 reported to @value{GDBN}. Multiple @samp{QPassSignals} packets do not
42208 combine; any earlier @samp{QPassSignals} list is completely replaced by the
42209 new list. This packet improves performance when using @samp{handle
42210 @var{signal} nostop noprint pass}.
42215 The request succeeded.
42218 An error occurred. The error number @var{nn} is given as hex digits.
42221 An empty reply indicates that @samp{QPassSignals} is not supported by
42225 Use of this packet is controlled by the @code{set remote pass-signals}
42226 command (@pxref{Remote Configuration, set remote pass-signals}).
42227 This packet is not probed by default; the remote stub must request it,
42228 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
42230 @item QProgramSignals: @var{signal} @r{[};@var{signal}@r{]}@dots{}
42231 @cindex signals the inferior may see, remote request
42232 @cindex @samp{QProgramSignals} packet
42233 @anchor{QProgramSignals}
42234 Each listed @var{signal} may be delivered to the inferior process.
42235 Others should be silently discarded.
42237 In some cases, the remote stub may need to decide whether to deliver a
42238 signal to the program or not without @value{GDBN} involvement. One
42239 example of that is while detaching --- the program's threads may have
42240 stopped for signals that haven't yet had a chance of being reported to
42241 @value{GDBN}, and so the remote stub can use the signal list specified
42242 by this packet to know whether to deliver or ignore those pending
42245 This does not influence whether to deliver a signal as requested by a
42246 resumption packet (@pxref{vCont packet}).
42248 Signals are numbered identically to continue packets and stop replies
42249 (@pxref{Stop Reply Packets}). Each @var{signal} list item should be
42250 strictly greater than the previous item. Multiple
42251 @samp{QProgramSignals} packets do not combine; any earlier
42252 @samp{QProgramSignals} list is completely replaced by the new list.
42257 The request succeeded.
42260 An error occurred. The error number @var{nn} is given as hex digits.
42263 An empty reply indicates that @samp{QProgramSignals} is not supported
42267 Use of this packet is controlled by the @code{set remote program-signals}
42268 command (@pxref{Remote Configuration, set remote program-signals}).
42269 This packet is not probed by default; the remote stub must request it,
42270 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
42272 @anchor{QThreadEvents}
42273 @item QThreadEvents:1
42274 @itemx QThreadEvents:0
42275 @cindex thread create/exit events, remote request
42276 @cindex @samp{QThreadEvents} packet
42278 Enable (@samp{QThreadEvents:1}) or disable (@samp{QThreadEvents:0})
42279 reporting of thread create and exit events. @xref{thread create
42280 event}, for the reply specifications. For example, this is used in
42281 non-stop mode when @value{GDBN} stops a set of threads and
42282 synchronously waits for the their corresponding stop replies. Without
42283 exit events, if one of the threads exits, @value{GDBN} would hang
42284 forever not knowing that it should no longer expect a stop for that
42285 same thread. @value{GDBN} does not enable this feature unless the
42286 stub reports that it supports it by including @samp{QThreadEvents+} in
42287 its @samp{qSupported} reply.
42292 The request succeeded.
42295 An error occurred. The error number @var{nn} is given as hex digits.
42298 An empty reply indicates that @samp{QThreadEvents} is not supported by
42302 Use of this packet is controlled by the @code{set remote thread-events}
42303 command (@pxref{Remote Configuration, set remote thread-events}).
42305 @item qRcmd,@var{command}
42306 @cindex execute remote command, remote request
42307 @cindex @samp{qRcmd} packet
42308 @var{command} (hex encoded) is passed to the local interpreter for
42309 execution. Invalid commands should be reported using the output
42310 string. Before the final result packet, the target may also respond
42311 with a number of intermediate @samp{O@var{output}} console output
42312 packets. @emph{Implementors should note that providing access to a
42313 stubs's interpreter may have security implications}.
42318 A command response with no output.
42320 A command response with the hex encoded output string @var{OUTPUT}.
42322 Indicate a badly formed request.
42324 An empty reply indicates that @samp{qRcmd} is not recognized.
42327 (Note that the @code{qRcmd} packet's name is separated from the
42328 command by a @samp{,}, not a @samp{:}, contrary to the naming
42329 conventions above. Please don't use this packet as a model for new
42332 @item qSearch:memory:@var{address};@var{length};@var{search-pattern}
42333 @cindex searching memory, in remote debugging
42335 @cindex @samp{qSearch:memory} packet
42337 @cindex @samp{qSearch memory} packet
42338 @anchor{qSearch memory}
42339 Search @var{length} bytes at @var{address} for @var{search-pattern}.
42340 Both @var{address} and @var{length} are encoded in hex;
42341 @var{search-pattern} is a sequence of bytes, also hex encoded.
42346 The pattern was not found.
42348 The pattern was found at @var{address}.
42350 A badly formed request or an error was encountered while searching memory.
42352 An empty reply indicates that @samp{qSearch:memory} is not recognized.
42355 @item QStartNoAckMode
42356 @cindex @samp{QStartNoAckMode} packet
42357 @anchor{QStartNoAckMode}
42358 Request that the remote stub disable the normal @samp{+}/@samp{-}
42359 protocol acknowledgments (@pxref{Packet Acknowledgment}).
42364 The stub has switched to no-acknowledgment mode.
42365 @value{GDBN} acknowledges this response,
42366 but neither the stub nor @value{GDBN} shall send or expect further
42367 @samp{+}/@samp{-} acknowledgments in the current connection.
42369 An empty reply indicates that the stub does not support no-acknowledgment mode.
42372 @item qSupported @r{[}:@var{gdbfeature} @r{[};@var{gdbfeature}@r{]}@dots{} @r{]}
42373 @cindex supported packets, remote query
42374 @cindex features of the remote protocol
42375 @cindex @samp{qSupported} packet
42376 @anchor{qSupported}
42377 Tell the remote stub about features supported by @value{GDBN}, and
42378 query the stub for features it supports. This packet allows
42379 @value{GDBN} and the remote stub to take advantage of each others'
42380 features. @samp{qSupported} also consolidates multiple feature probes
42381 at startup, to improve @value{GDBN} performance---a single larger
42382 packet performs better than multiple smaller probe packets on
42383 high-latency links. Some features may enable behavior which must not
42384 be on by default, e.g.@: because it would confuse older clients or
42385 stubs. Other features may describe packets which could be
42386 automatically probed for, but are not. These features must be
42387 reported before @value{GDBN} will use them. This ``default
42388 unsupported'' behavior is not appropriate for all packets, but it
42389 helps to keep the initial connection time under control with new
42390 versions of @value{GDBN} which support increasing numbers of packets.
42394 @item @var{stubfeature} @r{[};@var{stubfeature}@r{]}@dots{}
42395 The stub supports or does not support each returned @var{stubfeature},
42396 depending on the form of each @var{stubfeature} (see below for the
42399 An empty reply indicates that @samp{qSupported} is not recognized,
42400 or that no features needed to be reported to @value{GDBN}.
42403 The allowed forms for each feature (either a @var{gdbfeature} in the
42404 @samp{qSupported} packet, or a @var{stubfeature} in the response)
42408 @item @var{name}=@var{value}
42409 The remote protocol feature @var{name} is supported, and associated
42410 with the specified @var{value}. The format of @var{value} depends
42411 on the feature, but it must not include a semicolon.
42413 The remote protocol feature @var{name} is supported, and does not
42414 need an associated value.
42416 The remote protocol feature @var{name} is not supported.
42418 The remote protocol feature @var{name} may be supported, and
42419 @value{GDBN} should auto-detect support in some other way when it is
42420 needed. This form will not be used for @var{gdbfeature} notifications,
42421 but may be used for @var{stubfeature} responses.
42424 Whenever the stub receives a @samp{qSupported} request, the
42425 supplied set of @value{GDBN} features should override any previous
42426 request. This allows @value{GDBN} to put the stub in a known
42427 state, even if the stub had previously been communicating with
42428 a different version of @value{GDBN}.
42430 The following values of @var{gdbfeature} (for the packet sent by @value{GDBN})
42435 This feature indicates whether @value{GDBN} supports multiprocess
42436 extensions to the remote protocol. @value{GDBN} does not use such
42437 extensions unless the stub also reports that it supports them by
42438 including @samp{multiprocess+} in its @samp{qSupported} reply.
42439 @xref{multiprocess extensions}, for details.
42442 This feature indicates that @value{GDBN} supports the XML target
42443 description. If the stub sees @samp{xmlRegisters=} with target
42444 specific strings separated by a comma, it will report register
42448 This feature indicates whether @value{GDBN} supports the
42449 @samp{qRelocInsn} packet (@pxref{Tracepoint Packets,,Relocate
42450 instruction reply packet}).
42453 This feature indicates whether @value{GDBN} supports the swbreak stop
42454 reason in stop replies. @xref{swbreak stop reason}, for details.
42457 This feature indicates whether @value{GDBN} supports the hwbreak stop
42458 reason in stop replies. @xref{swbreak stop reason}, for details.
42461 This feature indicates whether @value{GDBN} supports fork event
42462 extensions to the remote protocol. @value{GDBN} does not use such
42463 extensions unless the stub also reports that it supports them by
42464 including @samp{fork-events+} in its @samp{qSupported} reply.
42467 This feature indicates whether @value{GDBN} supports vfork event
42468 extensions to the remote protocol. @value{GDBN} does not use such
42469 extensions unless the stub also reports that it supports them by
42470 including @samp{vfork-events+} in its @samp{qSupported} reply.
42473 This feature indicates whether @value{GDBN} supports exec event
42474 extensions to the remote protocol. @value{GDBN} does not use such
42475 extensions unless the stub also reports that it supports them by
42476 including @samp{exec-events+} in its @samp{qSupported} reply.
42478 @item vContSupported
42479 This feature indicates whether @value{GDBN} wants to know the
42480 supported actions in the reply to @samp{vCont?} packet.
42483 Stubs should ignore any unknown values for
42484 @var{gdbfeature}. Any @value{GDBN} which sends a @samp{qSupported}
42485 packet supports receiving packets of unlimited length (earlier
42486 versions of @value{GDBN} may reject overly long responses). Additional values
42487 for @var{gdbfeature} may be defined in the future to let the stub take
42488 advantage of new features in @value{GDBN}, e.g.@: incompatible
42489 improvements in the remote protocol---the @samp{multiprocess} feature is
42490 an example of such a feature. The stub's reply should be independent
42491 of the @var{gdbfeature} entries sent by @value{GDBN}; first @value{GDBN}
42492 describes all the features it supports, and then the stub replies with
42493 all the features it supports.
42495 Similarly, @value{GDBN} will silently ignore unrecognized stub feature
42496 responses, as long as each response uses one of the standard forms.
42498 Some features are flags. A stub which supports a flag feature
42499 should respond with a @samp{+} form response. Other features
42500 require values, and the stub should respond with an @samp{=}
42503 Each feature has a default value, which @value{GDBN} will use if
42504 @samp{qSupported} is not available or if the feature is not mentioned
42505 in the @samp{qSupported} response. The default values are fixed; a
42506 stub is free to omit any feature responses that match the defaults.
42508 Not all features can be probed, but for those which can, the probing
42509 mechanism is useful: in some cases, a stub's internal
42510 architecture may not allow the protocol layer to know some information
42511 about the underlying target in advance. This is especially common in
42512 stubs which may be configured for multiple targets.
42514 These are the currently defined stub features and their properties:
42516 @multitable @columnfractions 0.35 0.2 0.12 0.2
42517 @c NOTE: The first row should be @headitem, but we do not yet require
42518 @c a new enough version of Texinfo (4.7) to use @headitem.
42520 @tab Value Required
42524 @item @samp{PacketSize}
42529 @item @samp{qXfer:auxv:read}
42534 @item @samp{qXfer:btrace:read}
42539 @item @samp{qXfer:btrace-conf:read}
42544 @item @samp{qXfer:exec-file:read}
42549 @item @samp{qXfer:features:read}
42554 @item @samp{qXfer:libraries:read}
42559 @item @samp{qXfer:libraries-svr4:read}
42564 @item @samp{augmented-libraries-svr4-read}
42569 @item @samp{qXfer:memory-map:read}
42574 @item @samp{qXfer:sdata:read}
42579 @item @samp{qXfer:siginfo:read}
42584 @item @samp{qXfer:siginfo:write}
42589 @item @samp{qXfer:threads:read}
42594 @item @samp{qXfer:traceframe-info:read}
42599 @item @samp{qXfer:uib:read}
42604 @item @samp{qXfer:fdpic:read}
42609 @item @samp{Qbtrace:off}
42614 @item @samp{Qbtrace:bts}
42619 @item @samp{Qbtrace:pt}
42624 @item @samp{Qbtrace-conf:bts:size}
42629 @item @samp{Qbtrace-conf:pt:size}
42634 @item @samp{QNonStop}
42639 @item @samp{QCatchSyscalls}
42644 @item @samp{QPassSignals}
42649 @item @samp{QStartNoAckMode}
42654 @item @samp{multiprocess}
42659 @item @samp{ConditionalBreakpoints}
42664 @item @samp{ConditionalTracepoints}
42669 @item @samp{ReverseContinue}
42674 @item @samp{ReverseStep}
42679 @item @samp{TracepointSource}
42684 @item @samp{QAgent}
42689 @item @samp{QAllow}
42694 @item @samp{QDisableRandomization}
42699 @item @samp{EnableDisableTracepoints}
42704 @item @samp{QTBuffer:size}
42709 @item @samp{tracenz}
42714 @item @samp{BreakpointCommands}
42719 @item @samp{swbreak}
42724 @item @samp{hwbreak}
42729 @item @samp{fork-events}
42734 @item @samp{vfork-events}
42739 @item @samp{exec-events}
42744 @item @samp{QThreadEvents}
42749 @item @samp{no-resumed}
42756 These are the currently defined stub features, in more detail:
42759 @cindex packet size, remote protocol
42760 @item PacketSize=@var{bytes}
42761 The remote stub can accept packets up to at least @var{bytes} in
42762 length. @value{GDBN} will send packets up to this size for bulk
42763 transfers, and will never send larger packets. This is a limit on the
42764 data characters in the packet, including the frame and checksum.
42765 There is no trailing NUL byte in a remote protocol packet; if the stub
42766 stores packets in a NUL-terminated format, it should allow an extra
42767 byte in its buffer for the NUL. If this stub feature is not supported,
42768 @value{GDBN} guesses based on the size of the @samp{g} packet response.
42770 @item qXfer:auxv:read
42771 The remote stub understands the @samp{qXfer:auxv:read} packet
42772 (@pxref{qXfer auxiliary vector read}).
42774 @item qXfer:btrace:read
42775 The remote stub understands the @samp{qXfer:btrace:read}
42776 packet (@pxref{qXfer btrace read}).
42778 @item qXfer:btrace-conf:read
42779 The remote stub understands the @samp{qXfer:btrace-conf:read}
42780 packet (@pxref{qXfer btrace-conf read}).
42782 @item qXfer:exec-file:read
42783 The remote stub understands the @samp{qXfer:exec-file:read} packet
42784 (@pxref{qXfer executable filename read}).
42786 @item qXfer:features:read
42787 The remote stub understands the @samp{qXfer:features:read} packet
42788 (@pxref{qXfer target description read}).
42790 @item qXfer:libraries:read
42791 The remote stub understands the @samp{qXfer:libraries:read} packet
42792 (@pxref{qXfer library list read}).
42794 @item qXfer:libraries-svr4:read
42795 The remote stub understands the @samp{qXfer:libraries-svr4:read} packet
42796 (@pxref{qXfer svr4 library list read}).
42798 @item augmented-libraries-svr4-read
42799 The remote stub understands the augmented form of the
42800 @samp{qXfer:libraries-svr4:read} packet
42801 (@pxref{qXfer svr4 library list read}).
42803 @item qXfer:memory-map:read
42804 The remote stub understands the @samp{qXfer:memory-map:read} packet
42805 (@pxref{qXfer memory map read}).
42807 @item qXfer:sdata:read
42808 The remote stub understands the @samp{qXfer:sdata:read} packet
42809 (@pxref{qXfer sdata read}).
42811 @item qXfer:siginfo:read
42812 The remote stub understands the @samp{qXfer:siginfo:read} packet
42813 (@pxref{qXfer siginfo read}).
42815 @item qXfer:siginfo:write
42816 The remote stub understands the @samp{qXfer:siginfo:write} packet
42817 (@pxref{qXfer siginfo write}).
42819 @item qXfer:threads:read
42820 The remote stub understands the @samp{qXfer:threads:read} packet
42821 (@pxref{qXfer threads read}).
42823 @item qXfer:traceframe-info:read
42824 The remote stub understands the @samp{qXfer:traceframe-info:read}
42825 packet (@pxref{qXfer traceframe info read}).
42827 @item qXfer:uib:read
42828 The remote stub understands the @samp{qXfer:uib:read}
42829 packet (@pxref{qXfer unwind info block}).
42831 @item qXfer:fdpic:read
42832 The remote stub understands the @samp{qXfer:fdpic:read}
42833 packet (@pxref{qXfer fdpic loadmap read}).
42836 The remote stub understands the @samp{QNonStop} packet
42837 (@pxref{QNonStop}).
42839 @item QCatchSyscalls
42840 The remote stub understands the @samp{QCatchSyscalls} packet
42841 (@pxref{QCatchSyscalls}).
42844 The remote stub understands the @samp{QPassSignals} packet
42845 (@pxref{QPassSignals}).
42847 @item QStartNoAckMode
42848 The remote stub understands the @samp{QStartNoAckMode} packet and
42849 prefers to operate in no-acknowledgment mode. @xref{Packet Acknowledgment}.
42852 @anchor{multiprocess extensions}
42853 @cindex multiprocess extensions, in remote protocol
42854 The remote stub understands the multiprocess extensions to the remote
42855 protocol syntax. The multiprocess extensions affect the syntax of
42856 thread IDs in both packets and replies (@pxref{thread-id syntax}), and
42857 add process IDs to the @samp{D} packet and @samp{W} and @samp{X}
42858 replies. Note that reporting this feature indicates support for the
42859 syntactic extensions only, not that the stub necessarily supports
42860 debugging of more than one process at a time. The stub must not use
42861 multiprocess extensions in packet replies unless @value{GDBN} has also
42862 indicated it supports them in its @samp{qSupported} request.
42864 @item qXfer:osdata:read
42865 The remote stub understands the @samp{qXfer:osdata:read} packet
42866 ((@pxref{qXfer osdata read}).
42868 @item ConditionalBreakpoints
42869 The target accepts and implements evaluation of conditional expressions
42870 defined for breakpoints. The target will only report breakpoint triggers
42871 when such conditions are true (@pxref{Conditions, ,Break Conditions}).
42873 @item ConditionalTracepoints
42874 The remote stub accepts and implements conditional expressions defined
42875 for tracepoints (@pxref{Tracepoint Conditions}).
42877 @item ReverseContinue
42878 The remote stub accepts and implements the reverse continue packet
42882 The remote stub accepts and implements the reverse step packet
42885 @item TracepointSource
42886 The remote stub understands the @samp{QTDPsrc} packet that supplies
42887 the source form of tracepoint definitions.
42890 The remote stub understands the @samp{QAgent} packet.
42893 The remote stub understands the @samp{QAllow} packet.
42895 @item QDisableRandomization
42896 The remote stub understands the @samp{QDisableRandomization} packet.
42898 @item StaticTracepoint
42899 @cindex static tracepoints, in remote protocol
42900 The remote stub supports static tracepoints.
42902 @item InstallInTrace
42903 @anchor{install tracepoint in tracing}
42904 The remote stub supports installing tracepoint in tracing.
42906 @item EnableDisableTracepoints
42907 The remote stub supports the @samp{QTEnable} (@pxref{QTEnable}) and
42908 @samp{QTDisable} (@pxref{QTDisable}) packets that allow tracepoints
42909 to be enabled and disabled while a trace experiment is running.
42911 @item QTBuffer:size
42912 The remote stub supports the @samp{QTBuffer:size} (@pxref{QTBuffer-size})
42913 packet that allows to change the size of the trace buffer.
42916 @cindex string tracing, in remote protocol
42917 The remote stub supports the @samp{tracenz} bytecode for collecting strings.
42918 See @ref{Bytecode Descriptions} for details about the bytecode.
42920 @item BreakpointCommands
42921 @cindex breakpoint commands, in remote protocol
42922 The remote stub supports running a breakpoint's command list itself,
42923 rather than reporting the hit to @value{GDBN}.
42926 The remote stub understands the @samp{Qbtrace:off} packet.
42929 The remote stub understands the @samp{Qbtrace:bts} packet.
42932 The remote stub understands the @samp{Qbtrace:pt} packet.
42934 @item Qbtrace-conf:bts:size
42935 The remote stub understands the @samp{Qbtrace-conf:bts:size} packet.
42937 @item Qbtrace-conf:pt:size
42938 The remote stub understands the @samp{Qbtrace-conf:pt:size} packet.
42941 The remote stub reports the @samp{swbreak} stop reason for memory
42945 The remote stub reports the @samp{hwbreak} stop reason for hardware
42949 The remote stub reports the @samp{fork} stop reason for fork events.
42952 The remote stub reports the @samp{vfork} stop reason for vfork events
42953 and vforkdone events.
42956 The remote stub reports the @samp{exec} stop reason for exec events.
42958 @item vContSupported
42959 The remote stub reports the supported actions in the reply to
42960 @samp{vCont?} packet.
42962 @item QThreadEvents
42963 The remote stub understands the @samp{QThreadEvents} packet.
42966 The remote stub reports the @samp{N} stop reply.
42971 @cindex symbol lookup, remote request
42972 @cindex @samp{qSymbol} packet
42973 Notify the target that @value{GDBN} is prepared to serve symbol lookup
42974 requests. Accept requests from the target for the values of symbols.
42979 The target does not need to look up any (more) symbols.
42980 @item qSymbol:@var{sym_name}
42981 The target requests the value of symbol @var{sym_name} (hex encoded).
42982 @value{GDBN} may provide the value by using the
42983 @samp{qSymbol:@var{sym_value}:@var{sym_name}} message, described
42987 @item qSymbol:@var{sym_value}:@var{sym_name}
42988 Set the value of @var{sym_name} to @var{sym_value}.
42990 @var{sym_name} (hex encoded) is the name of a symbol whose value the
42991 target has previously requested.
42993 @var{sym_value} (hex) is the value for symbol @var{sym_name}. If
42994 @value{GDBN} cannot supply a value for @var{sym_name}, then this field
43000 The target does not need to look up any (more) symbols.
43001 @item qSymbol:@var{sym_name}
43002 The target requests the value of a new symbol @var{sym_name} (hex
43003 encoded). @value{GDBN} will continue to supply the values of symbols
43004 (if available), until the target ceases to request them.
43009 @itemx QTDisconnected
43016 @itemx qTMinFTPILen
43018 @xref{Tracepoint Packets}.
43020 @item qThreadExtraInfo,@var{thread-id}
43021 @cindex thread attributes info, remote request
43022 @cindex @samp{qThreadExtraInfo} packet
43023 Obtain from the target OS a printable string description of thread
43024 attributes for the thread @var{thread-id}; see @ref{thread-id syntax},
43025 for the forms of @var{thread-id}. This
43026 string may contain anything that the target OS thinks is interesting
43027 for @value{GDBN} to tell the user about the thread. The string is
43028 displayed in @value{GDBN}'s @code{info threads} display. Some
43029 examples of possible thread extra info strings are @samp{Runnable}, or
43030 @samp{Blocked on Mutex}.
43034 @item @var{XX}@dots{}
43035 Where @samp{@var{XX}@dots{}} is a hex encoding of @sc{ascii} data,
43036 comprising the printable string containing the extra information about
43037 the thread's attributes.
43040 (Note that the @code{qThreadExtraInfo} packet's name is separated from
43041 the command by a @samp{,}, not a @samp{:}, contrary to the naming
43042 conventions above. Please don't use this packet as a model for new
43061 @xref{Tracepoint Packets}.
43063 @item qXfer:@var{object}:read:@var{annex}:@var{offset},@var{length}
43064 @cindex read special object, remote request
43065 @cindex @samp{qXfer} packet
43066 @anchor{qXfer read}
43067 Read uninterpreted bytes from the target's special data area
43068 identified by the keyword @var{object}. Request @var{length} bytes
43069 starting at @var{offset} bytes into the data. The content and
43070 encoding of @var{annex} is specific to @var{object}; it can supply
43071 additional details about what data to access.
43076 Data @var{data} (@pxref{Binary Data}) has been read from the
43077 target. There may be more data at a higher address (although
43078 it is permitted to return @samp{m} even for the last valid
43079 block of data, as long as at least one byte of data was read).
43080 It is possible for @var{data} to have fewer bytes than the @var{length} in the
43084 Data @var{data} (@pxref{Binary Data}) has been read from the target.
43085 There is no more data to be read. It is possible for @var{data} to
43086 have fewer bytes than the @var{length} in the request.
43089 The @var{offset} in the request is at the end of the data.
43090 There is no more data to be read.
43093 The request was malformed, or @var{annex} was invalid.
43096 The offset was invalid, or there was an error encountered reading the data.
43097 The @var{nn} part is a hex-encoded @code{errno} value.
43100 An empty reply indicates the @var{object} string was not recognized by
43101 the stub, or that the object does not support reading.
43104 Here are the specific requests of this form defined so far. All the
43105 @samp{qXfer:@var{object}:read:@dots{}} requests use the same reply
43106 formats, listed above.
43109 @item qXfer:auxv:read::@var{offset},@var{length}
43110 @anchor{qXfer auxiliary vector read}
43111 Access the target's @dfn{auxiliary vector}. @xref{OS Information,
43112 auxiliary vector}. Note @var{annex} must be empty.
43114 This packet is not probed by default; the remote stub must request it,
43115 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43117 @item qXfer:btrace:read:@var{annex}:@var{offset},@var{length}
43118 @anchor{qXfer btrace read}
43120 Return a description of the current branch trace.
43121 @xref{Branch Trace Format}. The annex part of the generic @samp{qXfer}
43122 packet may have one of the following values:
43126 Returns all available branch trace.
43129 Returns all available branch trace if the branch trace changed since
43130 the last read request.
43133 Returns the new branch trace since the last read request. Adds a new
43134 block to the end of the trace that begins at zero and ends at the source
43135 location of the first branch in the trace buffer. This extra block is
43136 used to stitch traces together.
43138 If the trace buffer overflowed, returns an error indicating the overflow.
43141 This packet is not probed by default; the remote stub must request it
43142 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43144 @item qXfer:btrace-conf:read::@var{offset},@var{length}
43145 @anchor{qXfer btrace-conf read}
43147 Return a description of the current branch trace configuration.
43148 @xref{Branch Trace Configuration Format}.
43150 This packet is not probed by default; the remote stub must request it
43151 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43153 @item qXfer:exec-file:read:@var{annex}:@var{offset},@var{length}
43154 @anchor{qXfer executable filename read}
43155 Return the full absolute name of the file that was executed to create
43156 a process running on the remote system. The annex specifies the
43157 numeric process ID of the process to query, encoded as a hexadecimal
43158 number. If the annex part is empty the remote stub should return the
43159 filename corresponding to the currently executing process.
43161 This packet is not probed by default; the remote stub must request it,
43162 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43164 @item qXfer:features:read:@var{annex}:@var{offset},@var{length}
43165 @anchor{qXfer target description read}
43166 Access the @dfn{target description}. @xref{Target Descriptions}. The
43167 annex specifies which XML document to access. The main description is
43168 always loaded from the @samp{target.xml} annex.
43170 This packet is not probed by default; the remote stub must request it,
43171 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43173 @item qXfer:libraries:read:@var{annex}:@var{offset},@var{length}
43174 @anchor{qXfer library list read}
43175 Access the target's list of loaded libraries. @xref{Library List Format}.
43176 The annex part of the generic @samp{qXfer} packet must be empty
43177 (@pxref{qXfer read}).
43179 Targets which maintain a list of libraries in the program's memory do
43180 not need to implement this packet; it is designed for platforms where
43181 the operating system manages the list of loaded libraries.
43183 This packet is not probed by default; the remote stub must request it,
43184 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43186 @item qXfer:libraries-svr4:read:@var{annex}:@var{offset},@var{length}
43187 @anchor{qXfer svr4 library list read}
43188 Access the target's list of loaded libraries when the target is an SVR4
43189 platform. @xref{Library List Format for SVR4 Targets}. The annex part
43190 of the generic @samp{qXfer} packet must be empty unless the remote
43191 stub indicated it supports the augmented form of this packet
43192 by supplying an appropriate @samp{qSupported} response
43193 (@pxref{qXfer read}, @ref{qSupported}).
43195 This packet is optional for better performance on SVR4 targets.
43196 @value{GDBN} uses memory read packets to read the SVR4 library list otherwise.
43198 This packet is not probed by default; the remote stub must request it,
43199 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43201 If the remote stub indicates it supports the augmented form of this
43202 packet then the annex part of the generic @samp{qXfer} packet may
43203 contain a semicolon-separated list of @samp{@var{name}=@var{value}}
43204 arguments. The currently supported arguments are:
43207 @item start=@var{address}
43208 A hexadecimal number specifying the address of the @samp{struct
43209 link_map} to start reading the library list from. If unset or zero
43210 then the first @samp{struct link_map} in the library list will be
43211 chosen as the starting point.
43213 @item prev=@var{address}
43214 A hexadecimal number specifying the address of the @samp{struct
43215 link_map} immediately preceding the @samp{struct link_map}
43216 specified by the @samp{start} argument. If unset or zero then
43217 the remote stub will expect that no @samp{struct link_map}
43218 exists prior to the starting point.
43222 Arguments that are not understood by the remote stub will be silently
43225 @item qXfer:memory-map:read::@var{offset},@var{length}
43226 @anchor{qXfer memory map read}
43227 Access the target's @dfn{memory-map}. @xref{Memory Map Format}. The
43228 annex part of the generic @samp{qXfer} packet must be empty
43229 (@pxref{qXfer read}).
43231 This packet is not probed by default; the remote stub must request it,
43232 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43234 @item qXfer:sdata:read::@var{offset},@var{length}
43235 @anchor{qXfer sdata read}
43237 Read contents of the extra collected static tracepoint marker
43238 information. The annex part of the generic @samp{qXfer} packet must
43239 be empty (@pxref{qXfer read}). @xref{Tracepoint Actions,,Tracepoint
43242 This packet is not probed by default; the remote stub must request it,
43243 by supplying an appropriate @samp{qSupported} response
43244 (@pxref{qSupported}).
43246 @item qXfer:siginfo:read::@var{offset},@var{length}
43247 @anchor{qXfer siginfo read}
43248 Read contents of the extra signal information on the target
43249 system. The annex part of the generic @samp{qXfer} packet must be
43250 empty (@pxref{qXfer read}).
43252 This packet is not probed by default; the remote stub must request it,
43253 by supplying an appropriate @samp{qSupported} response
43254 (@pxref{qSupported}).
43256 @item qXfer:threads:read::@var{offset},@var{length}
43257 @anchor{qXfer threads read}
43258 Access the list of threads on target. @xref{Thread List Format}. The
43259 annex part of the generic @samp{qXfer} packet must be empty
43260 (@pxref{qXfer read}).
43262 This packet is not probed by default; the remote stub must request it,
43263 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43265 @item qXfer:traceframe-info:read::@var{offset},@var{length}
43266 @anchor{qXfer traceframe info read}
43268 Return a description of the current traceframe's contents.
43269 @xref{Traceframe Info Format}. The annex part of the generic
43270 @samp{qXfer} packet must be empty (@pxref{qXfer read}).
43272 This packet is not probed by default; the remote stub must request it,
43273 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43275 @item qXfer:uib:read:@var{pc}:@var{offset},@var{length}
43276 @anchor{qXfer unwind info block}
43278 Return the unwind information block for @var{pc}. This packet is used
43279 on OpenVMS/ia64 to ask the kernel unwind information.
43281 This packet is not probed by default.
43283 @item qXfer:fdpic:read:@var{annex}:@var{offset},@var{length}
43284 @anchor{qXfer fdpic loadmap read}
43285 Read contents of @code{loadmap}s on the target system. The
43286 annex, either @samp{exec} or @samp{interp}, specifies which @code{loadmap},
43287 executable @code{loadmap} or interpreter @code{loadmap} to read.
43289 This packet is not probed by default; the remote stub must request it,
43290 by supplying an appropriate @samp{qSupported} response (@pxref{qSupported}).
43292 @item qXfer:osdata:read::@var{offset},@var{length}
43293 @anchor{qXfer osdata read}
43294 Access the target's @dfn{operating system information}.
43295 @xref{Operating System Information}.
43299 @item qXfer:@var{object}:write:@var{annex}:@var{offset}:@var{data}@dots{}
43300 @cindex write data into object, remote request
43301 @anchor{qXfer write}
43302 Write uninterpreted bytes into the target's special data area
43303 identified by the keyword @var{object}, starting at @var{offset} bytes
43304 into the data. The binary-encoded data (@pxref{Binary Data}) to be
43305 written is given by @var{data}@dots{}. The content and encoding of @var{annex}
43306 is specific to @var{object}; it can supply additional details about what data
43312 @var{nn} (hex encoded) is the number of bytes written.
43313 This may be fewer bytes than supplied in the request.
43316 The request was malformed, or @var{annex} was invalid.
43319 The offset was invalid, or there was an error encountered writing the data.
43320 The @var{nn} part is a hex-encoded @code{errno} value.
43323 An empty reply indicates the @var{object} string was not
43324 recognized by the stub, or that the object does not support writing.
43327 Here are the specific requests of this form defined so far. All the
43328 @samp{qXfer:@var{object}:write:@dots{}} requests use the same reply
43329 formats, listed above.
43332 @item qXfer:siginfo:write::@var{offset}:@var{data}@dots{}
43333 @anchor{qXfer siginfo write}
43334 Write @var{data} to the extra signal information on the target system.
43335 The annex part of the generic @samp{qXfer} packet must be
43336 empty (@pxref{qXfer write}).
43338 This packet is not probed by default; the remote stub must request it,
43339 by supplying an appropriate @samp{qSupported} response
43340 (@pxref{qSupported}).
43343 @item qXfer:@var{object}:@var{operation}:@dots{}
43344 Requests of this form may be added in the future. When a stub does
43345 not recognize the @var{object} keyword, or its support for
43346 @var{object} does not recognize the @var{operation} keyword, the stub
43347 must respond with an empty packet.
43349 @item qAttached:@var{pid}
43350 @cindex query attached, remote request
43351 @cindex @samp{qAttached} packet
43352 Return an indication of whether the remote server attached to an
43353 existing process or created a new process. When the multiprocess
43354 protocol extensions are supported (@pxref{multiprocess extensions}),
43355 @var{pid} is an integer in hexadecimal format identifying the target
43356 process. Otherwise, @value{GDBN} will omit the @var{pid} field and
43357 the query packet will be simplified as @samp{qAttached}.
43359 This query is used, for example, to know whether the remote process
43360 should be detached or killed when a @value{GDBN} session is ended with
43361 the @code{quit} command.
43366 The remote server attached to an existing process.
43368 The remote server created a new process.
43370 A badly formed request or an error was encountered.
43374 Enable branch tracing for the current thread using Branch Trace Store.
43379 Branch tracing has been enabled.
43381 A badly formed request or an error was encountered.
43385 Enable branch tracing for the current thread using Intel Processor Trace.
43390 Branch tracing has been enabled.
43392 A badly formed request or an error was encountered.
43396 Disable branch tracing for the current thread.
43401 Branch tracing has been disabled.
43403 A badly formed request or an error was encountered.
43406 @item Qbtrace-conf:bts:size=@var{value}
43407 Set the requested ring buffer size for new threads that use the
43408 btrace recording method in bts format.
43413 The ring buffer size has been set.
43415 A badly formed request or an error was encountered.
43418 @item Qbtrace-conf:pt:size=@var{value}
43419 Set the requested ring buffer size for new threads that use the
43420 btrace recording method in pt format.
43425 The ring buffer size has been set.
43427 A badly formed request or an error was encountered.
43432 @node Architecture-Specific Protocol Details
43433 @section Architecture-Specific Protocol Details
43435 This section describes how the remote protocol is applied to specific
43436 target architectures. Also see @ref{Standard Target Features}, for
43437 details of XML target descriptions for each architecture.
43440 * ARM-Specific Protocol Details::
43441 * MIPS-Specific Protocol Details::
43444 @node ARM-Specific Protocol Details
43445 @subsection @acronym{ARM}-specific Protocol Details
43448 * ARM Breakpoint Kinds::
43451 @node ARM Breakpoint Kinds
43452 @subsubsection @acronym{ARM} Breakpoint Kinds
43453 @cindex breakpoint kinds, @acronym{ARM}
43455 These breakpoint kinds are defined for the @samp{Z0} and @samp{Z1} packets.
43460 16-bit Thumb mode breakpoint.
43463 32-bit Thumb mode (Thumb-2) breakpoint.
43466 32-bit @acronym{ARM} mode breakpoint.
43470 @node MIPS-Specific Protocol Details
43471 @subsection @acronym{MIPS}-specific Protocol Details
43474 * MIPS Register packet Format::
43475 * MIPS Breakpoint Kinds::
43478 @node MIPS Register packet Format
43479 @subsubsection @acronym{MIPS} Register Packet Format
43480 @cindex register packet format, @acronym{MIPS}
43482 The following @code{g}/@code{G} packets have previously been defined.
43483 In the below, some thirty-two bit registers are transferred as
43484 sixty-four bits. Those registers should be zero/sign extended (which?)
43485 to fill the space allocated. Register bytes are transferred in target
43486 byte order. The two nibbles within a register byte are transferred
43487 most-significant -- least-significant.
43492 All registers are transferred as thirty-two bit quantities in the order:
43493 32 general-purpose; sr; lo; hi; bad; cause; pc; 32 floating-point
43494 registers; fsr; fir; fp.
43497 All registers are transferred as sixty-four bit quantities (including
43498 thirty-two bit registers such as @code{sr}). The ordering is the same
43503 @node MIPS Breakpoint Kinds
43504 @subsubsection @acronym{MIPS} Breakpoint Kinds
43505 @cindex breakpoint kinds, @acronym{MIPS}
43507 These breakpoint kinds are defined for the @samp{Z0} and @samp{Z1} packets.
43512 16-bit @acronym{MIPS16} mode breakpoint.
43515 16-bit @acronym{microMIPS} mode breakpoint.
43518 32-bit standard @acronym{MIPS} mode breakpoint.
43521 32-bit @acronym{microMIPS} mode breakpoint.
43525 @node Tracepoint Packets
43526 @section Tracepoint Packets
43527 @cindex tracepoint packets
43528 @cindex packets, tracepoint
43530 Here we describe the packets @value{GDBN} uses to implement
43531 tracepoints (@pxref{Tracepoints}).
43535 @item QTDP:@var{n}:@var{addr}:@var{ena}:@var{step}:@var{pass}[:F@var{flen}][:X@var{len},@var{bytes}]@r{[}-@r{]}
43536 @cindex @samp{QTDP} packet
43537 Create a new tracepoint, number @var{n}, at @var{addr}. If @var{ena}
43538 is @samp{E}, then the tracepoint is enabled; if it is @samp{D}, then
43539 the tracepoint is disabled. The @var{step} gives the tracepoint's step
43540 count, and @var{pass} gives its pass count. If an @samp{F} is present,
43541 then the tracepoint is to be a fast tracepoint, and the @var{flen} is
43542 the number of bytes that the target should copy elsewhere to make room
43543 for the tracepoint. If an @samp{X} is present, it introduces a
43544 tracepoint condition, which consists of a hexadecimal length, followed
43545 by a comma and hex-encoded bytes, in a manner similar to action
43546 encodings as described below. If the trailing @samp{-} is present,
43547 further @samp{QTDP} packets will follow to specify this tracepoint's
43553 The packet was understood and carried out.
43555 @xref{Tracepoint Packets,,Relocate instruction reply packet}.
43557 The packet was not recognized.
43560 @item QTDP:-@var{n}:@var{addr}:@r{[}S@r{]}@var{action}@dots{}@r{[}-@r{]}
43561 Define actions to be taken when a tracepoint is hit. The @var{n} and
43562 @var{addr} must be the same as in the initial @samp{QTDP} packet for
43563 this tracepoint. This packet may only be sent immediately after
43564 another @samp{QTDP} packet that ended with a @samp{-}. If the
43565 trailing @samp{-} is present, further @samp{QTDP} packets will follow,
43566 specifying more actions for this tracepoint.
43568 In the series of action packets for a given tracepoint, at most one
43569 can have an @samp{S} before its first @var{action}. If such a packet
43570 is sent, it and the following packets define ``while-stepping''
43571 actions. Any prior packets define ordinary actions --- that is, those
43572 taken when the tracepoint is first hit. If no action packet has an
43573 @samp{S}, then all the packets in the series specify ordinary
43574 tracepoint actions.
43576 The @samp{@var{action}@dots{}} portion of the packet is a series of
43577 actions, concatenated without separators. Each action has one of the
43583 Collect the registers whose bits are set in @var{mask},
43584 a hexadecimal number whose @var{i}'th bit is set if register number
43585 @var{i} should be collected. (The least significant bit is numbered
43586 zero.) Note that @var{mask} may be any number of digits long; it may
43587 not fit in a 32-bit word.
43589 @item M @var{basereg},@var{offset},@var{len}
43590 Collect @var{len} bytes of memory starting at the address in register
43591 number @var{basereg}, plus @var{offset}. If @var{basereg} is
43592 @samp{-1}, then the range has a fixed address: @var{offset} is the
43593 address of the lowest byte to collect. The @var{basereg},
43594 @var{offset}, and @var{len} parameters are all unsigned hexadecimal
43595 values (the @samp{-1} value for @var{basereg} is a special case).
43597 @item X @var{len},@var{expr}
43598 Evaluate @var{expr}, whose length is @var{len}, and collect memory as
43599 it directs. The agent expression @var{expr} is as described in
43600 @ref{Agent Expressions}. Each byte of the expression is encoded as a
43601 two-digit hex number in the packet; @var{len} is the number of bytes
43602 in the expression (and thus one-half the number of hex digits in the
43607 Any number of actions may be packed together in a single @samp{QTDP}
43608 packet, as long as the packet does not exceed the maximum packet
43609 length (400 bytes, for many stubs). There may be only one @samp{R}
43610 action per tracepoint, and it must precede any @samp{M} or @samp{X}
43611 actions. Any registers referred to by @samp{M} and @samp{X} actions
43612 must be collected by a preceding @samp{R} action. (The
43613 ``while-stepping'' actions are treated as if they were attached to a
43614 separate tracepoint, as far as these restrictions are concerned.)
43619 The packet was understood and carried out.
43621 @xref{Tracepoint Packets,,Relocate instruction reply packet}.
43623 The packet was not recognized.
43626 @item QTDPsrc:@var{n}:@var{addr}:@var{type}:@var{start}:@var{slen}:@var{bytes}
43627 @cindex @samp{QTDPsrc} packet
43628 Specify a source string of tracepoint @var{n} at address @var{addr}.
43629 This is useful to get accurate reproduction of the tracepoints
43630 originally downloaded at the beginning of the trace run. The @var{type}
43631 is the name of the tracepoint part, such as @samp{cond} for the
43632 tracepoint's conditional expression (see below for a list of types), while
43633 @var{bytes} is the string, encoded in hexadecimal.
43635 @var{start} is the offset of the @var{bytes} within the overall source
43636 string, while @var{slen} is the total length of the source string.
43637 This is intended for handling source strings that are longer than will
43638 fit in a single packet.
43639 @c Add detailed example when this info is moved into a dedicated
43640 @c tracepoint descriptions section.
43642 The available string types are @samp{at} for the location,
43643 @samp{cond} for the conditional, and @samp{cmd} for an action command.
43644 @value{GDBN} sends a separate packet for each command in the action
43645 list, in the same order in which the commands are stored in the list.
43647 The target does not need to do anything with source strings except
43648 report them back as part of the replies to the @samp{qTfP}/@samp{qTsP}
43651 Although this packet is optional, and @value{GDBN} will only send it
43652 if the target replies with @samp{TracepointSource} @xref{General
43653 Query Packets}, it makes both disconnected tracing and trace files
43654 much easier to use. Otherwise the user must be careful that the
43655 tracepoints in effect while looking at trace frames are identical to
43656 the ones in effect during the trace run; even a small discrepancy
43657 could cause @samp{tdump} not to work, or a particular trace frame not
43660 @item QTDV:@var{n}:@var{value}:@var{builtin}:@var{name}
43661 @cindex define trace state variable, remote request
43662 @cindex @samp{QTDV} packet
43663 Create a new trace state variable, number @var{n}, with an initial
43664 value of @var{value}, which is a 64-bit signed integer. Both @var{n}
43665 and @var{value} are encoded as hexadecimal values. @value{GDBN} has
43666 the option of not using this packet for initial values of zero; the
43667 target should simply create the trace state variables as they are
43668 mentioned in expressions. The value @var{builtin} should be 1 (one)
43669 if the trace state variable is builtin and 0 (zero) if it is not builtin.
43670 @value{GDBN} only sets @var{builtin} to 1 if a previous @samp{qTfV} or
43671 @samp{qTsV} packet had it set. The contents of @var{name} is the
43672 hex-encoded name (without the leading @samp{$}) of the trace state
43675 @item QTFrame:@var{n}
43676 @cindex @samp{QTFrame} packet
43677 Select the @var{n}'th tracepoint frame from the buffer, and use the
43678 register and memory contents recorded there to answer subsequent
43679 request packets from @value{GDBN}.
43681 A successful reply from the stub indicates that the stub has found the
43682 requested frame. The response is a series of parts, concatenated
43683 without separators, describing the frame we selected. Each part has
43684 one of the following forms:
43688 The selected frame is number @var{n} in the trace frame buffer;
43689 @var{f} is a hexadecimal number. If @var{f} is @samp{-1}, then there
43690 was no frame matching the criteria in the request packet.
43693 The selected trace frame records a hit of tracepoint number @var{t};
43694 @var{t} is a hexadecimal number.
43698 @item QTFrame:pc:@var{addr}
43699 Like @samp{QTFrame:@var{n}}, but select the first tracepoint frame after the
43700 currently selected frame whose PC is @var{addr};
43701 @var{addr} is a hexadecimal number.
43703 @item QTFrame:tdp:@var{t}
43704 Like @samp{QTFrame:@var{n}}, but select the first tracepoint frame after the
43705 currently selected frame that is a hit of tracepoint @var{t}; @var{t}
43706 is a hexadecimal number.
43708 @item QTFrame:range:@var{start}:@var{end}
43709 Like @samp{QTFrame:@var{n}}, but select the first tracepoint frame after the
43710 currently selected frame whose PC is between @var{start} (inclusive)
43711 and @var{end} (inclusive); @var{start} and @var{end} are hexadecimal
43714 @item QTFrame:outside:@var{start}:@var{end}
43715 Like @samp{QTFrame:range:@var{start}:@var{end}}, but select the first
43716 frame @emph{outside} the given range of addresses (exclusive).
43719 @cindex @samp{qTMinFTPILen} packet
43720 This packet requests the minimum length of instruction at which a fast
43721 tracepoint (@pxref{Set Tracepoints}) may be placed. For instance, on
43722 the 32-bit x86 architecture, it is possible to use a 4-byte jump, but
43723 it depends on the target system being able to create trampolines in
43724 the first 64K of memory, which might or might not be possible for that
43725 system. So the reply to this packet will be 4 if it is able to
43732 The minimum instruction length is currently unknown.
43734 The minimum instruction length is @var{length}, where @var{length}
43735 is a hexadecimal number greater or equal to 1. A reply
43736 of 1 means that a fast tracepoint may be placed on any instruction
43737 regardless of size.
43739 An error has occurred.
43741 An empty reply indicates that the request is not supported by the stub.
43745 @cindex @samp{QTStart} packet
43746 Begin the tracepoint experiment. Begin collecting data from
43747 tracepoint hits in the trace frame buffer. This packet supports the
43748 @samp{qRelocInsn} reply (@pxref{Tracepoint Packets,,Relocate
43749 instruction reply packet}).
43752 @cindex @samp{QTStop} packet
43753 End the tracepoint experiment. Stop collecting trace frames.
43755 @item QTEnable:@var{n}:@var{addr}
43757 @cindex @samp{QTEnable} packet
43758 Enable tracepoint @var{n} at address @var{addr} in a started tracepoint
43759 experiment. If the tracepoint was previously disabled, then collection
43760 of data from it will resume.
43762 @item QTDisable:@var{n}:@var{addr}
43764 @cindex @samp{QTDisable} packet
43765 Disable tracepoint @var{n} at address @var{addr} in a started tracepoint
43766 experiment. No more data will be collected from the tracepoint unless
43767 @samp{QTEnable:@var{n}:@var{addr}} is subsequently issued.
43770 @cindex @samp{QTinit} packet
43771 Clear the table of tracepoints, and empty the trace frame buffer.
43773 @item QTro:@var{start1},@var{end1}:@var{start2},@var{end2}:@dots{}
43774 @cindex @samp{QTro} packet
43775 Establish the given ranges of memory as ``transparent''. The stub
43776 will answer requests for these ranges from memory's current contents,
43777 if they were not collected as part of the tracepoint hit.
43779 @value{GDBN} uses this to mark read-only regions of memory, like those
43780 containing program code. Since these areas never change, they should
43781 still have the same contents they did when the tracepoint was hit, so
43782 there's no reason for the stub to refuse to provide their contents.
43784 @item QTDisconnected:@var{value}
43785 @cindex @samp{QTDisconnected} packet
43786 Set the choice to what to do with the tracing run when @value{GDBN}
43787 disconnects from the target. A @var{value} of 1 directs the target to
43788 continue the tracing run, while 0 tells the target to stop tracing if
43789 @value{GDBN} is no longer in the picture.
43792 @cindex @samp{qTStatus} packet
43793 Ask the stub if there is a trace experiment running right now.
43795 The reply has the form:
43799 @item T@var{running}@r{[};@var{field}@r{]}@dots{}
43800 @var{running} is a single digit @code{1} if the trace is presently
43801 running, or @code{0} if not. It is followed by semicolon-separated
43802 optional fields that an agent may use to report additional status.
43806 If the trace is not running, the agent may report any of several
43807 explanations as one of the optional fields:
43812 No trace has been run yet.
43814 @item tstop[:@var{text}]:0
43815 The trace was stopped by a user-originated stop command. The optional
43816 @var{text} field is a user-supplied string supplied as part of the
43817 stop command (for instance, an explanation of why the trace was
43818 stopped manually). It is hex-encoded.
43821 The trace stopped because the trace buffer filled up.
43823 @item tdisconnected:0
43824 The trace stopped because @value{GDBN} disconnected from the target.
43826 @item tpasscount:@var{tpnum}
43827 The trace stopped because tracepoint @var{tpnum} exceeded its pass count.
43829 @item terror:@var{text}:@var{tpnum}
43830 The trace stopped because tracepoint @var{tpnum} had an error. The
43831 string @var{text} is available to describe the nature of the error
43832 (for instance, a divide by zero in the condition expression); it
43836 The trace stopped for some other reason.
43840 Additional optional fields supply statistical and other information.
43841 Although not required, they are extremely useful for users monitoring
43842 the progress of a trace run. If a trace has stopped, and these
43843 numbers are reported, they must reflect the state of the just-stopped
43848 @item tframes:@var{n}
43849 The number of trace frames in the buffer.
43851 @item tcreated:@var{n}
43852 The total number of trace frames created during the run. This may
43853 be larger than the trace frame count, if the buffer is circular.
43855 @item tsize:@var{n}
43856 The total size of the trace buffer, in bytes.
43858 @item tfree:@var{n}
43859 The number of bytes still unused in the buffer.
43861 @item circular:@var{n}
43862 The value of the circular trace buffer flag. @code{1} means that the
43863 trace buffer is circular and old trace frames will be discarded if
43864 necessary to make room, @code{0} means that the trace buffer is linear
43867 @item disconn:@var{n}
43868 The value of the disconnected tracing flag. @code{1} means that
43869 tracing will continue after @value{GDBN} disconnects, @code{0} means
43870 that the trace run will stop.
43874 @item qTP:@var{tp}:@var{addr}
43875 @cindex tracepoint status, remote request
43876 @cindex @samp{qTP} packet
43877 Ask the stub for the current state of tracepoint number @var{tp} at
43878 address @var{addr}.
43882 @item V@var{hits}:@var{usage}
43883 The tracepoint has been hit @var{hits} times so far during the trace
43884 run, and accounts for @var{usage} in the trace buffer. Note that
43885 @code{while-stepping} steps are not counted as separate hits, but the
43886 steps' space consumption is added into the usage number.
43890 @item qTV:@var{var}
43891 @cindex trace state variable value, remote request
43892 @cindex @samp{qTV} packet
43893 Ask the stub for the value of the trace state variable number @var{var}.
43898 The value of the variable is @var{value}. This will be the current
43899 value of the variable if the user is examining a running target, or a
43900 saved value if the variable was collected in the trace frame that the
43901 user is looking at. Note that multiple requests may result in
43902 different reply values, such as when requesting values while the
43903 program is running.
43906 The value of the variable is unknown. This would occur, for example,
43907 if the user is examining a trace frame in which the requested variable
43912 @cindex @samp{qTfP} packet
43914 @cindex @samp{qTsP} packet
43915 These packets request data about tracepoints that are being used by
43916 the target. @value{GDBN} sends @code{qTfP} to get the first piece
43917 of data, and multiple @code{qTsP} to get additional pieces. Replies
43918 to these packets generally take the form of the @code{QTDP} packets
43919 that define tracepoints. (FIXME add detailed syntax)
43922 @cindex @samp{qTfV} packet
43924 @cindex @samp{qTsV} packet
43925 These packets request data about trace state variables that are on the
43926 target. @value{GDBN} sends @code{qTfV} to get the first vari of data,
43927 and multiple @code{qTsV} to get additional variables. Replies to
43928 these packets follow the syntax of the @code{QTDV} packets that define
43929 trace state variables.
43935 @cindex @samp{qTfSTM} packet
43936 @cindex @samp{qTsSTM} packet
43937 These packets request data about static tracepoint markers that exist
43938 in the target program. @value{GDBN} sends @code{qTfSTM} to get the
43939 first piece of data, and multiple @code{qTsSTM} to get additional
43940 pieces. Replies to these packets take the following form:
43944 @item m @var{address}:@var{id}:@var{extra}
43946 @item m @var{address}:@var{id}:@var{extra},@var{address}:@var{id}:@var{extra}@dots{}
43947 a comma-separated list of markers
43949 (lower case letter @samp{L}) denotes end of list.
43951 An error occurred. The error number @var{nn} is given as hex digits.
43953 An empty reply indicates that the request is not supported by the
43957 The @var{address} is encoded in hex;
43958 @var{id} and @var{extra} are strings encoded in hex.
43960 In response to each query, the target will reply with a list of one or
43961 more markers, separated by commas. @value{GDBN} will respond to each
43962 reply with a request for more markers (using the @samp{qs} form of the
43963 query), until the target responds with @samp{l} (lower-case ell, for
43966 @item qTSTMat:@var{address}
43968 @cindex @samp{qTSTMat} packet
43969 This packets requests data about static tracepoint markers in the
43970 target program at @var{address}. Replies to this packet follow the
43971 syntax of the @samp{qTfSTM} and @code{qTsSTM} packets that list static
43972 tracepoint markers.
43974 @item QTSave:@var{filename}
43975 @cindex @samp{QTSave} packet
43976 This packet directs the target to save trace data to the file name
43977 @var{filename} in the target's filesystem. The @var{filename} is encoded
43978 as a hex string; the interpretation of the file name (relative vs
43979 absolute, wild cards, etc) is up to the target.
43981 @item qTBuffer:@var{offset},@var{len}
43982 @cindex @samp{qTBuffer} packet
43983 Return up to @var{len} bytes of the current contents of trace buffer,
43984 starting at @var{offset}. The trace buffer is treated as if it were
43985 a contiguous collection of traceframes, as per the trace file format.
43986 The reply consists as many hex-encoded bytes as the target can deliver
43987 in a packet; it is not an error to return fewer than were asked for.
43988 A reply consisting of just @code{l} indicates that no bytes are
43991 @item QTBuffer:circular:@var{value}
43992 This packet directs the target to use a circular trace buffer if
43993 @var{value} is 1, or a linear buffer if the value is 0.
43995 @item QTBuffer:size:@var{size}
43996 @anchor{QTBuffer-size}
43997 @cindex @samp{QTBuffer size} packet
43998 This packet directs the target to make the trace buffer be of size
43999 @var{size} if possible. A value of @code{-1} tells the target to
44000 use whatever size it prefers.
44002 @item QTNotes:@r{[}@var{type}:@var{text}@r{]}@r{[};@var{type}:@var{text}@r{]}@dots{}
44003 @cindex @samp{QTNotes} packet
44004 This packet adds optional textual notes to the trace run. Allowable
44005 types include @code{user}, @code{notes}, and @code{tstop}, the
44006 @var{text} fields are arbitrary strings, hex-encoded.
44010 @subsection Relocate instruction reply packet
44011 When installing fast tracepoints in memory, the target may need to
44012 relocate the instruction currently at the tracepoint address to a
44013 different address in memory. For most instructions, a simple copy is
44014 enough, but, for example, call instructions that implicitly push the
44015 return address on the stack, and relative branches or other
44016 PC-relative instructions require offset adjustment, so that the effect
44017 of executing the instruction at a different address is the same as if
44018 it had executed in the original location.
44020 In response to several of the tracepoint packets, the target may also
44021 respond with a number of intermediate @samp{qRelocInsn} request
44022 packets before the final result packet, to have @value{GDBN} handle
44023 this relocation operation. If a packet supports this mechanism, its
44024 documentation will explicitly say so. See for example the above
44025 descriptions for the @samp{QTStart} and @samp{QTDP} packets. The
44026 format of the request is:
44029 @item qRelocInsn:@var{from};@var{to}
44031 This requests @value{GDBN} to copy instruction at address @var{from}
44032 to address @var{to}, possibly adjusted so that executing the
44033 instruction at @var{to} has the same effect as executing it at
44034 @var{from}. @value{GDBN} writes the adjusted instruction to target
44035 memory starting at @var{to}.
44040 @item qRelocInsn:@var{adjusted_size}
44041 Informs the stub the relocation is complete. The @var{adjusted_size} is
44042 the length in bytes of resulting relocated instruction sequence.
44044 A badly formed request was detected, or an error was encountered while
44045 relocating the instruction.
44048 @node Host I/O Packets
44049 @section Host I/O Packets
44050 @cindex Host I/O, remote protocol
44051 @cindex file transfer, remote protocol
44053 The @dfn{Host I/O} packets allow @value{GDBN} to perform I/O
44054 operations on the far side of a remote link. For example, Host I/O is
44055 used to upload and download files to a remote target with its own
44056 filesystem. Host I/O uses the same constant values and data structure
44057 layout as the target-initiated File-I/O protocol. However, the
44058 Host I/O packets are structured differently. The target-initiated
44059 protocol relies on target memory to store parameters and buffers.
44060 Host I/O requests are initiated by @value{GDBN}, and the
44061 target's memory is not involved. @xref{File-I/O Remote Protocol
44062 Extension}, for more details on the target-initiated protocol.
44064 The Host I/O request packets all encode a single operation along with
44065 its arguments. They have this format:
44069 @item vFile:@var{operation}: @var{parameter}@dots{}
44070 @var{operation} is the name of the particular request; the target
44071 should compare the entire packet name up to the second colon when checking
44072 for a supported operation. The format of @var{parameter} depends on
44073 the operation. Numbers are always passed in hexadecimal. Negative
44074 numbers have an explicit minus sign (i.e.@: two's complement is not
44075 used). Strings (e.g.@: filenames) are encoded as a series of
44076 hexadecimal bytes. The last argument to a system call may be a
44077 buffer of escaped binary data (@pxref{Binary Data}).
44081 The valid responses to Host I/O packets are:
44085 @item F @var{result} [, @var{errno}] [; @var{attachment}]
44086 @var{result} is the integer value returned by this operation, usually
44087 non-negative for success and -1 for errors. If an error has occured,
44088 @var{errno} will be included in the result specifying a
44089 value defined by the File-I/O protocol (@pxref{Errno Values}). For
44090 operations which return data, @var{attachment} supplies the data as a
44091 binary buffer. Binary buffers in response packets are escaped in the
44092 normal way (@pxref{Binary Data}). See the individual packet
44093 documentation for the interpretation of @var{result} and
44097 An empty response indicates that this operation is not recognized.
44101 These are the supported Host I/O operations:
44104 @item vFile:open: @var{filename}, @var{flags}, @var{mode}
44105 Open a file at @var{filename} and return a file descriptor for it, or
44106 return -1 if an error occurs. The @var{filename} is a string,
44107 @var{flags} is an integer indicating a mask of open flags
44108 (@pxref{Open Flags}), and @var{mode} is an integer indicating a mask
44109 of mode bits to use if the file is created (@pxref{mode_t Values}).
44110 @xref{open}, for details of the open flags and mode values.
44112 @item vFile:close: @var{fd}
44113 Close the open file corresponding to @var{fd} and return 0, or
44114 -1 if an error occurs.
44116 @item vFile:pread: @var{fd}, @var{count}, @var{offset}
44117 Read data from the open file corresponding to @var{fd}. Up to
44118 @var{count} bytes will be read from the file, starting at @var{offset}
44119 relative to the start of the file. The target may read fewer bytes;
44120 common reasons include packet size limits and an end-of-file
44121 condition. The number of bytes read is returned. Zero should only be
44122 returned for a successful read at the end of the file, or if
44123 @var{count} was zero.
44125 The data read should be returned as a binary attachment on success.
44126 If zero bytes were read, the response should include an empty binary
44127 attachment (i.e.@: a trailing semicolon). The return value is the
44128 number of target bytes read; the binary attachment may be longer if
44129 some characters were escaped.
44131 @item vFile:pwrite: @var{fd}, @var{offset}, @var{data}
44132 Write @var{data} (a binary buffer) to the open file corresponding
44133 to @var{fd}. Start the write at @var{offset} from the start of the
44134 file. Unlike many @code{write} system calls, there is no
44135 separate @var{count} argument; the length of @var{data} in the
44136 packet is used. @samp{vFile:pwrite} returns the number of bytes written,
44137 which may be shorter than the length of @var{data}, or -1 if an
44140 @item vFile:fstat: @var{fd}
44141 Get information about the open file corresponding to @var{fd}.
44142 On success the information is returned as a binary attachment
44143 and the return value is the size of this attachment in bytes.
44144 If an error occurs the return value is -1. The format of the
44145 returned binary attachment is as described in @ref{struct stat}.
44147 @item vFile:unlink: @var{filename}
44148 Delete the file at @var{filename} on the target. Return 0,
44149 or -1 if an error occurs. The @var{filename} is a string.
44151 @item vFile:readlink: @var{filename}
44152 Read value of symbolic link @var{filename} on the target. Return
44153 the number of bytes read, or -1 if an error occurs.
44155 The data read should be returned as a binary attachment on success.
44156 If zero bytes were read, the response should include an empty binary
44157 attachment (i.e.@: a trailing semicolon). The return value is the
44158 number of target bytes read; the binary attachment may be longer if
44159 some characters were escaped.
44161 @item vFile:setfs: @var{pid}
44162 Select the filesystem on which @code{vFile} operations with
44163 @var{filename} arguments will operate. This is required for
44164 @value{GDBN} to be able to access files on remote targets where
44165 the remote stub does not share a common filesystem with the
44168 If @var{pid} is nonzero, select the filesystem as seen by process
44169 @var{pid}. If @var{pid} is zero, select the filesystem as seen by
44170 the remote stub. Return 0 on success, or -1 if an error occurs.
44171 If @code{vFile:setfs:} indicates success, the selected filesystem
44172 remains selected until the next successful @code{vFile:setfs:}
44178 @section Interrupts
44179 @cindex interrupts (remote protocol)
44180 @anchor{interrupting remote targets}
44182 In all-stop mode, when a program on the remote target is running,
44183 @value{GDBN} may attempt to interrupt it by sending a @samp{Ctrl-C},
44184 @code{BREAK} or a @code{BREAK} followed by @code{g}, control of which
44185 is specified via @value{GDBN}'s @samp{interrupt-sequence}.
44187 The precise meaning of @code{BREAK} is defined by the transport
44188 mechanism and may, in fact, be undefined. @value{GDBN} does not
44189 currently define a @code{BREAK} mechanism for any of the network
44190 interfaces except for TCP, in which case @value{GDBN} sends the
44191 @code{telnet} BREAK sequence.
44193 @samp{Ctrl-C}, on the other hand, is defined and implemented for all
44194 transport mechanisms. It is represented by sending the single byte
44195 @code{0x03} without any of the usual packet overhead described in
44196 the Overview section (@pxref{Overview}). When a @code{0x03} byte is
44197 transmitted as part of a packet, it is considered to be packet data
44198 and does @emph{not} represent an interrupt. E.g., an @samp{X} packet
44199 (@pxref{X packet}), used for binary downloads, may include an unescaped
44200 @code{0x03} as part of its packet.
44202 @code{BREAK} followed by @code{g} is also known as Magic SysRq g.
44203 When Linux kernel receives this sequence from serial port,
44204 it stops execution and connects to gdb.
44206 In non-stop mode, because packet resumptions are asynchronous
44207 (@pxref{vCont packet}), @value{GDBN} is always free to send a remote
44208 command to the remote stub, even when the target is running. For that
44209 reason, @value{GDBN} instead sends a regular packet (@pxref{vCtrlC
44210 packet}) with the usual packet framing instead of the single byte
44213 Stubs are not required to recognize these interrupt mechanisms and the
44214 precise meaning associated with receipt of the interrupt is
44215 implementation defined. If the target supports debugging of multiple
44216 threads and/or processes, it should attempt to interrupt all
44217 currently-executing threads and processes.
44218 If the stub is successful at interrupting the
44219 running program, it should send one of the stop
44220 reply packets (@pxref{Stop Reply Packets}) to @value{GDBN} as a result
44221 of successfully stopping the program in all-stop mode, and a stop reply
44222 for each stopped thread in non-stop mode.
44223 Interrupts received while the
44224 program is stopped are queued and the program will be interrupted when
44225 it is resumed next time.
44227 @node Notification Packets
44228 @section Notification Packets
44229 @cindex notification packets
44230 @cindex packets, notification
44232 The @value{GDBN} remote serial protocol includes @dfn{notifications},
44233 packets that require no acknowledgment. Both the GDB and the stub
44234 may send notifications (although the only notifications defined at
44235 present are sent by the stub). Notifications carry information
44236 without incurring the round-trip latency of an acknowledgment, and so
44237 are useful for low-impact communications where occasional packet loss
44240 A notification packet has the form @samp{% @var{data} #
44241 @var{checksum}}, where @var{data} is the content of the notification,
44242 and @var{checksum} is a checksum of @var{data}, computed and formatted
44243 as for ordinary @value{GDBN} packets. A notification's @var{data}
44244 never contains @samp{$}, @samp{%} or @samp{#} characters. Upon
44245 receiving a notification, the recipient sends no @samp{+} or @samp{-}
44246 to acknowledge the notification's receipt or to report its corruption.
44248 Every notification's @var{data} begins with a name, which contains no
44249 colon characters, followed by a colon character.
44251 Recipients should silently ignore corrupted notifications and
44252 notifications they do not understand. Recipients should restart
44253 timeout periods on receipt of a well-formed notification, whether or
44254 not they understand it.
44256 Senders should only send the notifications described here when this
44257 protocol description specifies that they are permitted. In the
44258 future, we may extend the protocol to permit existing notifications in
44259 new contexts; this rule helps older senders avoid confusing newer
44262 (Older versions of @value{GDBN} ignore bytes received until they see
44263 the @samp{$} byte that begins an ordinary packet, so new stubs may
44264 transmit notifications without fear of confusing older clients. There
44265 are no notifications defined for @value{GDBN} to send at the moment, but we
44266 assume that most older stubs would ignore them, as well.)
44268 Each notification is comprised of three parts:
44270 @item @var{name}:@var{event}
44271 The notification packet is sent by the side that initiates the
44272 exchange (currently, only the stub does that), with @var{event}
44273 carrying the specific information about the notification, and
44274 @var{name} specifying the name of the notification.
44276 The acknowledge sent by the other side, usually @value{GDBN}, to
44277 acknowledge the exchange and request the event.
44280 The purpose of an asynchronous notification mechanism is to report to
44281 @value{GDBN} that something interesting happened in the remote stub.
44283 The remote stub may send notification @var{name}:@var{event}
44284 at any time, but @value{GDBN} acknowledges the notification when
44285 appropriate. The notification event is pending before @value{GDBN}
44286 acknowledges. Only one notification at a time may be pending; if
44287 additional events occur before @value{GDBN} has acknowledged the
44288 previous notification, they must be queued by the stub for later
44289 synchronous transmission in response to @var{ack} packets from
44290 @value{GDBN}. Because the notification mechanism is unreliable,
44291 the stub is permitted to resend a notification if it believes
44292 @value{GDBN} may not have received it.
44294 Specifically, notifications may appear when @value{GDBN} is not
44295 otherwise reading input from the stub, or when @value{GDBN} is
44296 expecting to read a normal synchronous response or a
44297 @samp{+}/@samp{-} acknowledgment to a packet it has sent.
44298 Notification packets are distinct from any other communication from
44299 the stub so there is no ambiguity.
44301 After receiving a notification, @value{GDBN} shall acknowledge it by
44302 sending a @var{ack} packet as a regular, synchronous request to the
44303 stub. Such acknowledgment is not required to happen immediately, as
44304 @value{GDBN} is permitted to send other, unrelated packets to the
44305 stub first, which the stub should process normally.
44307 Upon receiving a @var{ack} packet, if the stub has other queued
44308 events to report to @value{GDBN}, it shall respond by sending a
44309 normal @var{event}. @value{GDBN} shall then send another @var{ack}
44310 packet to solicit further responses; again, it is permitted to send
44311 other, unrelated packets as well which the stub should process
44314 If the stub receives a @var{ack} packet and there are no additional
44315 @var{event} to report, the stub shall return an @samp{OK} response.
44316 At this point, @value{GDBN} has finished processing a notification
44317 and the stub has completed sending any queued events. @value{GDBN}
44318 won't accept any new notifications until the final @samp{OK} is
44319 received . If further notification events occur, the stub shall send
44320 a new notification, @value{GDBN} shall accept the notification, and
44321 the process shall be repeated.
44323 The process of asynchronous notification can be illustrated by the
44326 <- @code{%Stop:T0505:98e7ffbf;04:4ce6ffbf;08:b1b6e54c;thread:p7526.7526;core:0;}
44329 <- @code{T0505:68f37db7;04:40f37db7;08:63850408;thread:p7526.7528;core:0;}
44331 <- @code{T0505:68e3fdb6;04:40e3fdb6;08:63850408;thread:p7526.7529;core:0;}
44336 The following notifications are defined:
44337 @multitable @columnfractions 0.12 0.12 0.38 0.38
44346 @tab @var{reply}. The @var{reply} has the form of a stop reply, as
44347 described in @ref{Stop Reply Packets}. Refer to @ref{Remote Non-Stop},
44348 for information on how these notifications are acknowledged by
44350 @tab Report an asynchronous stop event in non-stop mode.
44354 @node Remote Non-Stop
44355 @section Remote Protocol Support for Non-Stop Mode
44357 @value{GDBN}'s remote protocol supports non-stop debugging of
44358 multi-threaded programs, as described in @ref{Non-Stop Mode}. If the stub
44359 supports non-stop mode, it should report that to @value{GDBN} by including
44360 @samp{QNonStop+} in its @samp{qSupported} response (@pxref{qSupported}).
44362 @value{GDBN} typically sends a @samp{QNonStop} packet only when
44363 establishing a new connection with the stub. Entering non-stop mode
44364 does not alter the state of any currently-running threads, but targets
44365 must stop all threads in any already-attached processes when entering
44366 all-stop mode. @value{GDBN} uses the @samp{?} packet as necessary to
44367 probe the target state after a mode change.
44369 In non-stop mode, when an attached process encounters an event that
44370 would otherwise be reported with a stop reply, it uses the
44371 asynchronous notification mechanism (@pxref{Notification Packets}) to
44372 inform @value{GDBN}. In contrast to all-stop mode, where all threads
44373 in all processes are stopped when a stop reply is sent, in non-stop
44374 mode only the thread reporting the stop event is stopped. That is,
44375 when reporting a @samp{S} or @samp{T} response to indicate completion
44376 of a step operation, hitting a breakpoint, or a fault, only the
44377 affected thread is stopped; any other still-running threads continue
44378 to run. When reporting a @samp{W} or @samp{X} response, all running
44379 threads belonging to other attached processes continue to run.
44381 In non-stop mode, the target shall respond to the @samp{?} packet as
44382 follows. First, any incomplete stop reply notification/@samp{vStopped}
44383 sequence in progress is abandoned. The target must begin a new
44384 sequence reporting stop events for all stopped threads, whether or not
44385 it has previously reported those events to @value{GDBN}. The first
44386 stop reply is sent as a synchronous reply to the @samp{?} packet, and
44387 subsequent stop replies are sent as responses to @samp{vStopped} packets
44388 using the mechanism described above. The target must not send
44389 asynchronous stop reply notifications until the sequence is complete.
44390 If all threads are running when the target receives the @samp{?} packet,
44391 or if the target is not attached to any process, it shall respond
44394 If the stub supports non-stop mode, it should also support the
44395 @samp{swbreak} stop reason if software breakpoints are supported, and
44396 the @samp{hwbreak} stop reason if hardware breakpoints are supported
44397 (@pxref{swbreak stop reason}). This is because given the asynchronous
44398 nature of non-stop mode, between the time a thread hits a breakpoint
44399 and the time the event is finally processed by @value{GDBN}, the
44400 breakpoint may have already been removed from the target. Due to
44401 this, @value{GDBN} needs to be able to tell whether a trap stop was
44402 caused by a delayed breakpoint event, which should be ignored, as
44403 opposed to a random trap signal, which should be reported to the user.
44404 Note the @samp{swbreak} feature implies that the target is responsible
44405 for adjusting the PC when a software breakpoint triggers, if
44406 necessary, such as on the x86 architecture.
44408 @node Packet Acknowledgment
44409 @section Packet Acknowledgment
44411 @cindex acknowledgment, for @value{GDBN} remote
44412 @cindex packet acknowledgment, for @value{GDBN} remote
44413 By default, when either the host or the target machine receives a packet,
44414 the first response expected is an acknowledgment: either @samp{+} (to indicate
44415 the package was received correctly) or @samp{-} (to request retransmission).
44416 This mechanism allows the @value{GDBN} remote protocol to operate over
44417 unreliable transport mechanisms, such as a serial line.
44419 In cases where the transport mechanism is itself reliable (such as a pipe or
44420 TCP connection), the @samp{+}/@samp{-} acknowledgments are redundant.
44421 It may be desirable to disable them in that case to reduce communication
44422 overhead, or for other reasons. This can be accomplished by means of the
44423 @samp{QStartNoAckMode} packet; @pxref{QStartNoAckMode}.
44425 When in no-acknowledgment mode, neither the stub nor @value{GDBN} shall send or
44426 expect @samp{+}/@samp{-} protocol acknowledgments. The packet
44427 and response format still includes the normal checksum, as described in
44428 @ref{Overview}, but the checksum may be ignored by the receiver.
44430 If the stub supports @samp{QStartNoAckMode} and prefers to operate in
44431 no-acknowledgment mode, it should report that to @value{GDBN}
44432 by including @samp{QStartNoAckMode+} in its response to @samp{qSupported};
44433 @pxref{qSupported}.
44434 If @value{GDBN} also supports @samp{QStartNoAckMode} and it has not been
44435 disabled via the @code{set remote noack-packet off} command
44436 (@pxref{Remote Configuration}),
44437 @value{GDBN} may then send a @samp{QStartNoAckMode} packet to the stub.
44438 Only then may the stub actually turn off packet acknowledgments.
44439 @value{GDBN} sends a final @samp{+} acknowledgment of the stub's @samp{OK}
44440 response, which can be safely ignored by the stub.
44442 Note that @code{set remote noack-packet} command only affects negotiation
44443 between @value{GDBN} and the stub when subsequent connections are made;
44444 it does not affect the protocol acknowledgment state for any current
44446 Since @samp{+}/@samp{-} acknowledgments are enabled by default when a
44447 new connection is established,
44448 there is also no protocol request to re-enable the acknowledgments
44449 for the current connection, once disabled.
44454 Example sequence of a target being re-started. Notice how the restart
44455 does not get any direct output:
44460 @emph{target restarts}
44463 <- @code{T001:1234123412341234}
44467 Example sequence of a target being stepped by a single instruction:
44470 -> @code{G1445@dots{}}
44475 <- @code{T001:1234123412341234}
44479 <- @code{1455@dots{}}
44483 @node File-I/O Remote Protocol Extension
44484 @section File-I/O Remote Protocol Extension
44485 @cindex File-I/O remote protocol extension
44488 * File-I/O Overview::
44489 * Protocol Basics::
44490 * The F Request Packet::
44491 * The F Reply Packet::
44492 * The Ctrl-C Message::
44494 * List of Supported Calls::
44495 * Protocol-specific Representation of Datatypes::
44497 * File-I/O Examples::
44500 @node File-I/O Overview
44501 @subsection File-I/O Overview
44502 @cindex file-i/o overview
44504 The @dfn{File I/O remote protocol extension} (short: File-I/O) allows the
44505 target to use the host's file system and console I/O to perform various
44506 system calls. System calls on the target system are translated into a
44507 remote protocol packet to the host system, which then performs the needed
44508 actions and returns a response packet to the target system.
44509 This simulates file system operations even on targets that lack file systems.
44511 The protocol is defined to be independent of both the host and target systems.
44512 It uses its own internal representation of datatypes and values. Both
44513 @value{GDBN} and the target's @value{GDBN} stub are responsible for
44514 translating the system-dependent value representations into the internal
44515 protocol representations when data is transmitted.
44517 The communication is synchronous. A system call is possible only when
44518 @value{GDBN} is waiting for a response from the @samp{C}, @samp{c}, @samp{S}
44519 or @samp{s} packets. While @value{GDBN} handles the request for a system call,
44520 the target is stopped to allow deterministic access to the target's
44521 memory. Therefore File-I/O is not interruptible by target signals. On
44522 the other hand, it is possible to interrupt File-I/O by a user interrupt
44523 (@samp{Ctrl-C}) within @value{GDBN}.
44525 The target's request to perform a host system call does not finish
44526 the latest @samp{C}, @samp{c}, @samp{S} or @samp{s} action. That means,
44527 after finishing the system call, the target returns to continuing the
44528 previous activity (continue, step). No additional continue or step
44529 request from @value{GDBN} is required.
44532 (@value{GDBP}) continue
44533 <- target requests 'system call X'
44534 target is stopped, @value{GDBN} executes system call
44535 -> @value{GDBN} returns result
44536 ... target continues, @value{GDBN} returns to wait for the target
44537 <- target hits breakpoint and sends a Txx packet
44540 The protocol only supports I/O on the console and to regular files on
44541 the host file system. Character or block special devices, pipes,
44542 named pipes, sockets or any other communication method on the host
44543 system are not supported by this protocol.
44545 File I/O is not supported in non-stop mode.
44547 @node Protocol Basics
44548 @subsection Protocol Basics
44549 @cindex protocol basics, file-i/o
44551 The File-I/O protocol uses the @code{F} packet as the request as well
44552 as reply packet. Since a File-I/O system call can only occur when
44553 @value{GDBN} is waiting for a response from the continuing or stepping target,
44554 the File-I/O request is a reply that @value{GDBN} has to expect as a result
44555 of a previous @samp{C}, @samp{c}, @samp{S} or @samp{s} packet.
44556 This @code{F} packet contains all information needed to allow @value{GDBN}
44557 to call the appropriate host system call:
44561 A unique identifier for the requested system call.
44564 All parameters to the system call. Pointers are given as addresses
44565 in the target memory address space. Pointers to strings are given as
44566 pointer/length pair. Numerical values are given as they are.
44567 Numerical control flags are given in a protocol-specific representation.
44571 At this point, @value{GDBN} has to perform the following actions.
44575 If the parameters include pointer values to data needed as input to a
44576 system call, @value{GDBN} requests this data from the target with a
44577 standard @code{m} packet request. This additional communication has to be
44578 expected by the target implementation and is handled as any other @code{m}
44582 @value{GDBN} translates all value from protocol representation to host
44583 representation as needed. Datatypes are coerced into the host types.
44586 @value{GDBN} calls the system call.
44589 It then coerces datatypes back to protocol representation.
44592 If the system call is expected to return data in buffer space specified
44593 by pointer parameters to the call, the data is transmitted to the
44594 target using a @code{M} or @code{X} packet. This packet has to be expected
44595 by the target implementation and is handled as any other @code{M} or @code{X}
44600 Eventually @value{GDBN} replies with another @code{F} packet which contains all
44601 necessary information for the target to continue. This at least contains
44608 @code{errno}, if has been changed by the system call.
44615 After having done the needed type and value coercion, the target continues
44616 the latest continue or step action.
44618 @node The F Request Packet
44619 @subsection The @code{F} Request Packet
44620 @cindex file-i/o request packet
44621 @cindex @code{F} request packet
44623 The @code{F} request packet has the following format:
44626 @item F@var{call-id},@var{parameter@dots{}}
44628 @var{call-id} is the identifier to indicate the host system call to be called.
44629 This is just the name of the function.
44631 @var{parameter@dots{}} are the parameters to the system call.
44632 Parameters are hexadecimal integer values, either the actual values in case
44633 of scalar datatypes, pointers to target buffer space in case of compound
44634 datatypes and unspecified memory areas, or pointer/length pairs in case
44635 of string parameters. These are appended to the @var{call-id} as a
44636 comma-delimited list. All values are transmitted in ASCII
44637 string representation, pointer/length pairs separated by a slash.
44643 @node The F Reply Packet
44644 @subsection The @code{F} Reply Packet
44645 @cindex file-i/o reply packet
44646 @cindex @code{F} reply packet
44648 The @code{F} reply packet has the following format:
44652 @item F@var{retcode},@var{errno},@var{Ctrl-C flag};@var{call-specific attachment}
44654 @var{retcode} is the return code of the system call as hexadecimal value.
44656 @var{errno} is the @code{errno} set by the call, in protocol-specific
44658 This parameter can be omitted if the call was successful.
44660 @var{Ctrl-C flag} is only sent if the user requested a break. In this
44661 case, @var{errno} must be sent as well, even if the call was successful.
44662 The @var{Ctrl-C flag} itself consists of the character @samp{C}:
44669 or, if the call was interrupted before the host call has been performed:
44676 assuming 4 is the protocol-specific representation of @code{EINTR}.
44681 @node The Ctrl-C Message
44682 @subsection The @samp{Ctrl-C} Message
44683 @cindex ctrl-c message, in file-i/o protocol
44685 If the @samp{Ctrl-C} flag is set in the @value{GDBN}
44686 reply packet (@pxref{The F Reply Packet}),
44687 the target should behave as if it had
44688 gotten a break message. The meaning for the target is ``system call
44689 interrupted by @code{SIGINT}''. Consequentially, the target should actually stop
44690 (as with a break message) and return to @value{GDBN} with a @code{T02}
44693 It's important for the target to know in which
44694 state the system call was interrupted. There are two possible cases:
44698 The system call hasn't been performed on the host yet.
44701 The system call on the host has been finished.
44705 These two states can be distinguished by the target by the value of the
44706 returned @code{errno}. If it's the protocol representation of @code{EINTR}, the system
44707 call hasn't been performed. This is equivalent to the @code{EINTR} handling
44708 on POSIX systems. In any other case, the target may presume that the
44709 system call has been finished --- successfully or not --- and should behave
44710 as if the break message arrived right after the system call.
44712 @value{GDBN} must behave reliably. If the system call has not been called
44713 yet, @value{GDBN} may send the @code{F} reply immediately, setting @code{EINTR} as
44714 @code{errno} in the packet. If the system call on the host has been finished
44715 before the user requests a break, the full action must be finished by
44716 @value{GDBN}. This requires sending @code{M} or @code{X} packets as necessary.
44717 The @code{F} packet may only be sent when either nothing has happened
44718 or the full action has been completed.
44721 @subsection Console I/O
44722 @cindex console i/o as part of file-i/o
44724 By default and if not explicitly closed by the target system, the file
44725 descriptors 0, 1 and 2 are connected to the @value{GDBN} console. Output
44726 on the @value{GDBN} console is handled as any other file output operation
44727 (@code{write(1, @dots{})} or @code{write(2, @dots{})}). Console input is handled
44728 by @value{GDBN} so that after the target read request from file descriptor
44729 0 all following typing is buffered until either one of the following
44734 The user types @kbd{Ctrl-c}. The behaviour is as explained above, and the
44736 system call is treated as finished.
44739 The user presses @key{RET}. This is treated as end of input with a trailing
44743 The user types @kbd{Ctrl-d}. This is treated as end of input. No trailing
44744 character (neither newline nor @samp{Ctrl-D}) is appended to the input.
44748 If the user has typed more characters than fit in the buffer given to
44749 the @code{read} call, the trailing characters are buffered in @value{GDBN} until
44750 either another @code{read(0, @dots{})} is requested by the target, or debugging
44751 is stopped at the user's request.
44754 @node List of Supported Calls
44755 @subsection List of Supported Calls
44756 @cindex list of supported file-i/o calls
44773 @unnumberedsubsubsec open
44774 @cindex open, file-i/o system call
44779 int open(const char *pathname, int flags);
44780 int open(const char *pathname, int flags, mode_t mode);
44784 @samp{Fopen,@var{pathptr}/@var{len},@var{flags},@var{mode}}
44787 @var{flags} is the bitwise @code{OR} of the following values:
44791 If the file does not exist it will be created. The host
44792 rules apply as far as file ownership and time stamps
44796 When used with @code{O_CREAT}, if the file already exists it is
44797 an error and open() fails.
44800 If the file already exists and the open mode allows
44801 writing (@code{O_RDWR} or @code{O_WRONLY} is given) it will be
44802 truncated to zero length.
44805 The file is opened in append mode.
44808 The file is opened for reading only.
44811 The file is opened for writing only.
44814 The file is opened for reading and writing.
44818 Other bits are silently ignored.
44822 @var{mode} is the bitwise @code{OR} of the following values:
44826 User has read permission.
44829 User has write permission.
44832 Group has read permission.
44835 Group has write permission.
44838 Others have read permission.
44841 Others have write permission.
44845 Other bits are silently ignored.
44848 @item Return value:
44849 @code{open} returns the new file descriptor or -1 if an error
44856 @var{pathname} already exists and @code{O_CREAT} and @code{O_EXCL} were used.
44859 @var{pathname} refers to a directory.
44862 The requested access is not allowed.
44865 @var{pathname} was too long.
44868 A directory component in @var{pathname} does not exist.
44871 @var{pathname} refers to a device, pipe, named pipe or socket.
44874 @var{pathname} refers to a file on a read-only filesystem and
44875 write access was requested.
44878 @var{pathname} is an invalid pointer value.
44881 No space on device to create the file.
44884 The process already has the maximum number of files open.
44887 The limit on the total number of files open on the system
44891 The call was interrupted by the user.
44897 @unnumberedsubsubsec close
44898 @cindex close, file-i/o system call
44907 @samp{Fclose,@var{fd}}
44909 @item Return value:
44910 @code{close} returns zero on success, or -1 if an error occurred.
44916 @var{fd} isn't a valid open file descriptor.
44919 The call was interrupted by the user.
44925 @unnumberedsubsubsec read
44926 @cindex read, file-i/o system call
44931 int read(int fd, void *buf, unsigned int count);
44935 @samp{Fread,@var{fd},@var{bufptr},@var{count}}
44937 @item Return value:
44938 On success, the number of bytes read is returned.
44939 Zero indicates end of file. If count is zero, read
44940 returns zero as well. On error, -1 is returned.
44946 @var{fd} is not a valid file descriptor or is not open for
44950 @var{bufptr} is an invalid pointer value.
44953 The call was interrupted by the user.
44959 @unnumberedsubsubsec write
44960 @cindex write, file-i/o system call
44965 int write(int fd, const void *buf, unsigned int count);
44969 @samp{Fwrite,@var{fd},@var{bufptr},@var{count}}
44971 @item Return value:
44972 On success, the number of bytes written are returned.
44973 Zero indicates nothing was written. On error, -1
44980 @var{fd} is not a valid file descriptor or is not open for
44984 @var{bufptr} is an invalid pointer value.
44987 An attempt was made to write a file that exceeds the
44988 host-specific maximum file size allowed.
44991 No space on device to write the data.
44994 The call was interrupted by the user.
45000 @unnumberedsubsubsec lseek
45001 @cindex lseek, file-i/o system call
45006 long lseek (int fd, long offset, int flag);
45010 @samp{Flseek,@var{fd},@var{offset},@var{flag}}
45012 @var{flag} is one of:
45016 The offset is set to @var{offset} bytes.
45019 The offset is set to its current location plus @var{offset}
45023 The offset is set to the size of the file plus @var{offset}
45027 @item Return value:
45028 On success, the resulting unsigned offset in bytes from
45029 the beginning of the file is returned. Otherwise, a
45030 value of -1 is returned.
45036 @var{fd} is not a valid open file descriptor.
45039 @var{fd} is associated with the @value{GDBN} console.
45042 @var{flag} is not a proper value.
45045 The call was interrupted by the user.
45051 @unnumberedsubsubsec rename
45052 @cindex rename, file-i/o system call
45057 int rename(const char *oldpath, const char *newpath);
45061 @samp{Frename,@var{oldpathptr}/@var{len},@var{newpathptr}/@var{len}}
45063 @item Return value:
45064 On success, zero is returned. On error, -1 is returned.
45070 @var{newpath} is an existing directory, but @var{oldpath} is not a
45074 @var{newpath} is a non-empty directory.
45077 @var{oldpath} or @var{newpath} is a directory that is in use by some
45081 An attempt was made to make a directory a subdirectory
45085 A component used as a directory in @var{oldpath} or new
45086 path is not a directory. Or @var{oldpath} is a directory
45087 and @var{newpath} exists but is not a directory.
45090 @var{oldpathptr} or @var{newpathptr} are invalid pointer values.
45093 No access to the file or the path of the file.
45097 @var{oldpath} or @var{newpath} was too long.
45100 A directory component in @var{oldpath} or @var{newpath} does not exist.
45103 The file is on a read-only filesystem.
45106 The device containing the file has no room for the new
45110 The call was interrupted by the user.
45116 @unnumberedsubsubsec unlink
45117 @cindex unlink, file-i/o system call
45122 int unlink(const char *pathname);
45126 @samp{Funlink,@var{pathnameptr}/@var{len}}
45128 @item Return value:
45129 On success, zero is returned. On error, -1 is returned.
45135 No access to the file or the path of the file.
45138 The system does not allow unlinking of directories.
45141 The file @var{pathname} cannot be unlinked because it's
45142 being used by another process.
45145 @var{pathnameptr} is an invalid pointer value.
45148 @var{pathname} was too long.
45151 A directory component in @var{pathname} does not exist.
45154 A component of the path is not a directory.
45157 The file is on a read-only filesystem.
45160 The call was interrupted by the user.
45166 @unnumberedsubsubsec stat/fstat
45167 @cindex fstat, file-i/o system call
45168 @cindex stat, file-i/o system call
45173 int stat(const char *pathname, struct stat *buf);
45174 int fstat(int fd, struct stat *buf);
45178 @samp{Fstat,@var{pathnameptr}/@var{len},@var{bufptr}}@*
45179 @samp{Ffstat,@var{fd},@var{bufptr}}
45181 @item Return value:
45182 On success, zero is returned. On error, -1 is returned.
45188 @var{fd} is not a valid open file.
45191 A directory component in @var{pathname} does not exist or the
45192 path is an empty string.
45195 A component of the path is not a directory.
45198 @var{pathnameptr} is an invalid pointer value.
45201 No access to the file or the path of the file.
45204 @var{pathname} was too long.
45207 The call was interrupted by the user.
45213 @unnumberedsubsubsec gettimeofday
45214 @cindex gettimeofday, file-i/o system call
45219 int gettimeofday(struct timeval *tv, void *tz);
45223 @samp{Fgettimeofday,@var{tvptr},@var{tzptr}}
45225 @item Return value:
45226 On success, 0 is returned, -1 otherwise.
45232 @var{tz} is a non-NULL pointer.
45235 @var{tvptr} and/or @var{tzptr} is an invalid pointer value.
45241 @unnumberedsubsubsec isatty
45242 @cindex isatty, file-i/o system call
45247 int isatty(int fd);
45251 @samp{Fisatty,@var{fd}}
45253 @item Return value:
45254 Returns 1 if @var{fd} refers to the @value{GDBN} console, 0 otherwise.
45260 The call was interrupted by the user.
45265 Note that the @code{isatty} call is treated as a special case: it returns
45266 1 to the target if the file descriptor is attached
45267 to the @value{GDBN} console, 0 otherwise. Implementing through system calls
45268 would require implementing @code{ioctl} and would be more complex than
45273 @unnumberedsubsubsec system
45274 @cindex system, file-i/o system call
45279 int system(const char *command);
45283 @samp{Fsystem,@var{commandptr}/@var{len}}
45285 @item Return value:
45286 If @var{len} is zero, the return value indicates whether a shell is
45287 available. A zero return value indicates a shell is not available.
45288 For non-zero @var{len}, the value returned is -1 on error and the
45289 return status of the command otherwise. Only the exit status of the
45290 command is returned, which is extracted from the host's @code{system}
45291 return value by calling @code{WEXITSTATUS(retval)}. In case
45292 @file{/bin/sh} could not be executed, 127 is returned.
45298 The call was interrupted by the user.
45303 @value{GDBN} takes over the full task of calling the necessary host calls
45304 to perform the @code{system} call. The return value of @code{system} on
45305 the host is simplified before it's returned
45306 to the target. Any termination signal information from the child process
45307 is discarded, and the return value consists
45308 entirely of the exit status of the called command.
45310 Due to security concerns, the @code{system} call is by default refused
45311 by @value{GDBN}. The user has to allow this call explicitly with the
45312 @code{set remote system-call-allowed 1} command.
45315 @item set remote system-call-allowed
45316 @kindex set remote system-call-allowed
45317 Control whether to allow the @code{system} calls in the File I/O
45318 protocol for the remote target. The default is zero (disabled).
45320 @item show remote system-call-allowed
45321 @kindex show remote system-call-allowed
45322 Show whether the @code{system} calls are allowed in the File I/O
45326 @node Protocol-specific Representation of Datatypes
45327 @subsection Protocol-specific Representation of Datatypes
45328 @cindex protocol-specific representation of datatypes, in file-i/o protocol
45331 * Integral Datatypes::
45333 * Memory Transfer::
45338 @node Integral Datatypes
45339 @unnumberedsubsubsec Integral Datatypes
45340 @cindex integral datatypes, in file-i/o protocol
45342 The integral datatypes used in the system calls are @code{int},
45343 @code{unsigned int}, @code{long}, @code{unsigned long},
45344 @code{mode_t}, and @code{time_t}.
45346 @code{int}, @code{unsigned int}, @code{mode_t} and @code{time_t} are
45347 implemented as 32 bit values in this protocol.
45349 @code{long} and @code{unsigned long} are implemented as 64 bit types.
45351 @xref{Limits}, for corresponding MIN and MAX values (similar to those
45352 in @file{limits.h}) to allow range checking on host and target.
45354 @code{time_t} datatypes are defined as seconds since the Epoch.
45356 All integral datatypes transferred as part of a memory read or write of a
45357 structured datatype e.g.@: a @code{struct stat} have to be given in big endian
45360 @node Pointer Values
45361 @unnumberedsubsubsec Pointer Values
45362 @cindex pointer values, in file-i/o protocol
45364 Pointers to target data are transmitted as they are. An exception
45365 is made for pointers to buffers for which the length isn't
45366 transmitted as part of the function call, namely strings. Strings
45367 are transmitted as a pointer/length pair, both as hex values, e.g.@:
45374 which is a pointer to data of length 18 bytes at position 0x1aaf.
45375 The length is defined as the full string length in bytes, including
45376 the trailing null byte. For example, the string @code{"hello world"}
45377 at address 0x123456 is transmitted as
45383 @node Memory Transfer
45384 @unnumberedsubsubsec Memory Transfer
45385 @cindex memory transfer, in file-i/o protocol
45387 Structured data which is transferred using a memory read or write (for
45388 example, a @code{struct stat}) is expected to be in a protocol-specific format
45389 with all scalar multibyte datatypes being big endian. Translation to
45390 this representation needs to be done both by the target before the @code{F}
45391 packet is sent, and by @value{GDBN} before
45392 it transfers memory to the target. Transferred pointers to structured
45393 data should point to the already-coerced data at any time.
45397 @unnumberedsubsubsec struct stat
45398 @cindex struct stat, in file-i/o protocol
45400 The buffer of type @code{struct stat} used by the target and @value{GDBN}
45401 is defined as follows:
45405 unsigned int st_dev; /* device */
45406 unsigned int st_ino; /* inode */
45407 mode_t st_mode; /* protection */
45408 unsigned int st_nlink; /* number of hard links */
45409 unsigned int st_uid; /* user ID of owner */
45410 unsigned int st_gid; /* group ID of owner */
45411 unsigned int st_rdev; /* device type (if inode device) */
45412 unsigned long st_size; /* total size, in bytes */
45413 unsigned long st_blksize; /* blocksize for filesystem I/O */
45414 unsigned long st_blocks; /* number of blocks allocated */
45415 time_t st_atime; /* time of last access */
45416 time_t st_mtime; /* time of last modification */
45417 time_t st_ctime; /* time of last change */
45421 The integral datatypes conform to the definitions given in the
45422 appropriate section (see @ref{Integral Datatypes}, for details) so this
45423 structure is of size 64 bytes.
45425 The values of several fields have a restricted meaning and/or
45431 A value of 0 represents a file, 1 the console.
45434 No valid meaning for the target. Transmitted unchanged.
45437 Valid mode bits are described in @ref{Constants}. Any other
45438 bits have currently no meaning for the target.
45443 No valid meaning for the target. Transmitted unchanged.
45448 These values have a host and file system dependent
45449 accuracy. Especially on Windows hosts, the file system may not
45450 support exact timing values.
45453 The target gets a @code{struct stat} of the above representation and is
45454 responsible for coercing it to the target representation before
45457 Note that due to size differences between the host, target, and protocol
45458 representations of @code{struct stat} members, these members could eventually
45459 get truncated on the target.
45461 @node struct timeval
45462 @unnumberedsubsubsec struct timeval
45463 @cindex struct timeval, in file-i/o protocol
45465 The buffer of type @code{struct timeval} used by the File-I/O protocol
45466 is defined as follows:
45470 time_t tv_sec; /* second */
45471 long tv_usec; /* microsecond */
45475 The integral datatypes conform to the definitions given in the
45476 appropriate section (see @ref{Integral Datatypes}, for details) so this
45477 structure is of size 8 bytes.
45480 @subsection Constants
45481 @cindex constants, in file-i/o protocol
45483 The following values are used for the constants inside of the
45484 protocol. @value{GDBN} and target are responsible for translating these
45485 values before and after the call as needed.
45496 @unnumberedsubsubsec Open Flags
45497 @cindex open flags, in file-i/o protocol
45499 All values are given in hexadecimal representation.
45511 @node mode_t Values
45512 @unnumberedsubsubsec mode_t Values
45513 @cindex mode_t values, in file-i/o protocol
45515 All values are given in octal representation.
45532 @unnumberedsubsubsec Errno Values
45533 @cindex errno values, in file-i/o protocol
45535 All values are given in decimal representation.
45560 @code{EUNKNOWN} is used as a fallback error value if a host system returns
45561 any error value not in the list of supported error numbers.
45564 @unnumberedsubsubsec Lseek Flags
45565 @cindex lseek flags, in file-i/o protocol
45574 @unnumberedsubsubsec Limits
45575 @cindex limits, in file-i/o protocol
45577 All values are given in decimal representation.
45580 INT_MIN -2147483648
45582 UINT_MAX 4294967295
45583 LONG_MIN -9223372036854775808
45584 LONG_MAX 9223372036854775807
45585 ULONG_MAX 18446744073709551615
45588 @node File-I/O Examples
45589 @subsection File-I/O Examples
45590 @cindex file-i/o examples
45592 Example sequence of a write call, file descriptor 3, buffer is at target
45593 address 0x1234, 6 bytes should be written:
45596 <- @code{Fwrite,3,1234,6}
45597 @emph{request memory read from target}
45600 @emph{return "6 bytes written"}
45604 Example sequence of a read call, file descriptor 3, buffer is at target
45605 address 0x1234, 6 bytes should be read:
45608 <- @code{Fread,3,1234,6}
45609 @emph{request memory write to target}
45610 -> @code{X1234,6:XXXXXX}
45611 @emph{return "6 bytes read"}
45615 Example sequence of a read call, call fails on the host due to invalid
45616 file descriptor (@code{EBADF}):
45619 <- @code{Fread,3,1234,6}
45623 Example sequence of a read call, user presses @kbd{Ctrl-c} before syscall on
45627 <- @code{Fread,3,1234,6}
45632 Example sequence of a read call, user presses @kbd{Ctrl-c} after syscall on
45636 <- @code{Fread,3,1234,6}
45637 -> @code{X1234,6:XXXXXX}
45641 @node Library List Format
45642 @section Library List Format
45643 @cindex library list format, remote protocol
45645 On some platforms, a dynamic loader (e.g.@: @file{ld.so}) runs in the
45646 same process as your application to manage libraries. In this case,
45647 @value{GDBN} can use the loader's symbol table and normal memory
45648 operations to maintain a list of shared libraries. On other
45649 platforms, the operating system manages loaded libraries.
45650 @value{GDBN} can not retrieve the list of currently loaded libraries
45651 through memory operations, so it uses the @samp{qXfer:libraries:read}
45652 packet (@pxref{qXfer library list read}) instead. The remote stub
45653 queries the target's operating system and reports which libraries
45656 The @samp{qXfer:libraries:read} packet returns an XML document which
45657 lists loaded libraries and their offsets. Each library has an
45658 associated name and one or more segment or section base addresses,
45659 which report where the library was loaded in memory.
45661 For the common case of libraries that are fully linked binaries, the
45662 library should have a list of segments. If the target supports
45663 dynamic linking of a relocatable object file, its library XML element
45664 should instead include a list of allocated sections. The segment or
45665 section bases are start addresses, not relocation offsets; they do not
45666 depend on the library's link-time base addresses.
45668 @value{GDBN} must be linked with the Expat library to support XML
45669 library lists. @xref{Expat}.
45671 A simple memory map, with one loaded library relocated by a single
45672 offset, looks like this:
45676 <library name="/lib/libc.so.6">
45677 <segment address="0x10000000"/>
45682 Another simple memory map, with one loaded library with three
45683 allocated sections (.text, .data, .bss), looks like this:
45687 <library name="sharedlib.o">
45688 <section address="0x10000000"/>
45689 <section address="0x20000000"/>
45690 <section address="0x30000000"/>
45695 The format of a library list is described by this DTD:
45698 <!-- library-list: Root element with versioning -->
45699 <!ELEMENT library-list (library)*>
45700 <!ATTLIST library-list version CDATA #FIXED "1.0">
45701 <!ELEMENT library (segment*, section*)>
45702 <!ATTLIST library name CDATA #REQUIRED>
45703 <!ELEMENT segment EMPTY>
45704 <!ATTLIST segment address CDATA #REQUIRED>
45705 <!ELEMENT section EMPTY>
45706 <!ATTLIST section address CDATA #REQUIRED>
45709 In addition, segments and section descriptors cannot be mixed within a
45710 single library element, and you must supply at least one segment or
45711 section for each library.
45713 @node Library List Format for SVR4 Targets
45714 @section Library List Format for SVR4 Targets
45715 @cindex library list format, remote protocol
45717 On SVR4 platforms @value{GDBN} can use the symbol table of a dynamic loader
45718 (e.g.@: @file{ld.so}) and normal memory operations to maintain a list of
45719 shared libraries. Still a special library list provided by this packet is
45720 more efficient for the @value{GDBN} remote protocol.
45722 The @samp{qXfer:libraries-svr4:read} packet returns an XML document which lists
45723 loaded libraries and their SVR4 linker parameters. For each library on SVR4
45724 target, the following parameters are reported:
45728 @code{name}, the absolute file name from the @code{l_name} field of
45729 @code{struct link_map}.
45731 @code{lm} with address of @code{struct link_map} used for TLS
45732 (Thread Local Storage) access.
45734 @code{l_addr}, the displacement as read from the field @code{l_addr} of
45735 @code{struct link_map}. For prelinked libraries this is not an absolute
45736 memory address. It is a displacement of absolute memory address against
45737 address the file was prelinked to during the library load.
45739 @code{l_ld}, which is memory address of the @code{PT_DYNAMIC} segment
45742 Additionally the single @code{main-lm} attribute specifies address of
45743 @code{struct link_map} used for the main executable. This parameter is used
45744 for TLS access and its presence is optional.
45746 @value{GDBN} must be linked with the Expat library to support XML
45747 SVR4 library lists. @xref{Expat}.
45749 A simple memory map, with two loaded libraries (which do not use prelink),
45753 <library-list-svr4 version="1.0" main-lm="0xe4f8f8">
45754 <library name="/lib/ld-linux.so.2" lm="0xe4f51c" l_addr="0xe2d000"
45756 <library name="/lib/libc.so.6" lm="0xe4fbe8" l_addr="0x154000"
45758 </library-list-svr>
45761 The format of an SVR4 library list is described by this DTD:
45764 <!-- library-list-svr4: Root element with versioning -->
45765 <!ELEMENT library-list-svr4 (library)*>
45766 <!ATTLIST library-list-svr4 version CDATA #FIXED "1.0">
45767 <!ATTLIST library-list-svr4 main-lm CDATA #IMPLIED>
45768 <!ELEMENT library EMPTY>
45769 <!ATTLIST library name CDATA #REQUIRED>
45770 <!ATTLIST library lm CDATA #REQUIRED>
45771 <!ATTLIST library l_addr CDATA #REQUIRED>
45772 <!ATTLIST library l_ld CDATA #REQUIRED>
45775 @node Memory Map Format
45776 @section Memory Map Format
45777 @cindex memory map format
45779 To be able to write into flash memory, @value{GDBN} needs to obtain a
45780 memory map from the target. This section describes the format of the
45783 The memory map is obtained using the @samp{qXfer:memory-map:read}
45784 (@pxref{qXfer memory map read}) packet and is an XML document that
45785 lists memory regions.
45787 @value{GDBN} must be linked with the Expat library to support XML
45788 memory maps. @xref{Expat}.
45790 The top-level structure of the document is shown below:
45793 <?xml version="1.0"?>
45794 <!DOCTYPE memory-map
45795 PUBLIC "+//IDN gnu.org//DTD GDB Memory Map V1.0//EN"
45796 "http://sourceware.org/gdb/gdb-memory-map.dtd">
45802 Each region can be either:
45807 A region of RAM starting at @var{addr} and extending for @var{length}
45811 <memory type="ram" start="@var{addr}" length="@var{length}"/>
45816 A region of read-only memory:
45819 <memory type="rom" start="@var{addr}" length="@var{length}"/>
45824 A region of flash memory, with erasure blocks @var{blocksize}
45828 <memory type="flash" start="@var{addr}" length="@var{length}">
45829 <property name="blocksize">@var{blocksize}</property>
45835 Regions must not overlap. @value{GDBN} assumes that areas of memory not covered
45836 by the memory map are RAM, and uses the ordinary @samp{M} and @samp{X}
45837 packets to write to addresses in such ranges.
45839 The formal DTD for memory map format is given below:
45842 <!-- ................................................... -->
45843 <!-- Memory Map XML DTD ................................ -->
45844 <!-- File: memory-map.dtd .............................. -->
45845 <!-- .................................... .............. -->
45846 <!-- memory-map.dtd -->
45847 <!-- memory-map: Root element with versioning -->
45848 <!ELEMENT memory-map (memory)*>
45849 <!ATTLIST memory-map version CDATA #FIXED "1.0.0">
45850 <!ELEMENT memory (property)*>
45851 <!-- memory: Specifies a memory region,
45852 and its type, or device. -->
45853 <!ATTLIST memory type (ram|rom|flash) #REQUIRED
45854 start CDATA #REQUIRED
45855 length CDATA #REQUIRED>
45856 <!-- property: Generic attribute tag -->
45857 <!ELEMENT property (#PCDATA | property)*>
45858 <!ATTLIST property name (blocksize) #REQUIRED>
45861 @node Thread List Format
45862 @section Thread List Format
45863 @cindex thread list format
45865 To efficiently update the list of threads and their attributes,
45866 @value{GDBN} issues the @samp{qXfer:threads:read} packet
45867 (@pxref{qXfer threads read}) and obtains the XML document with
45868 the following structure:
45871 <?xml version="1.0"?>
45873 <thread id="id" core="0" name="name">
45874 ... description ...
45879 Each @samp{thread} element must have the @samp{id} attribute that
45880 identifies the thread (@pxref{thread-id syntax}). The
45881 @samp{core} attribute, if present, specifies which processor core
45882 the thread was last executing on. The @samp{name} attribute, if
45883 present, specifies the human-readable name of the thread. The content
45884 of the of @samp{thread} element is interpreted as human-readable
45885 auxiliary information. The @samp{handle} attribute, if present,
45886 is a hex encoded representation of the thread handle.
45889 @node Traceframe Info Format
45890 @section Traceframe Info Format
45891 @cindex traceframe info format
45893 To be able to know which objects in the inferior can be examined when
45894 inspecting a tracepoint hit, @value{GDBN} needs to obtain the list of
45895 memory ranges, registers and trace state variables that have been
45896 collected in a traceframe.
45898 This list is obtained using the @samp{qXfer:traceframe-info:read}
45899 (@pxref{qXfer traceframe info read}) packet and is an XML document.
45901 @value{GDBN} must be linked with the Expat library to support XML
45902 traceframe info discovery. @xref{Expat}.
45904 The top-level structure of the document is shown below:
45907 <?xml version="1.0"?>
45908 <!DOCTYPE traceframe-info
45909 PUBLIC "+//IDN gnu.org//DTD GDB Memory Map V1.0//EN"
45910 "http://sourceware.org/gdb/gdb-traceframe-info.dtd">
45916 Each traceframe block can be either:
45921 A region of collected memory starting at @var{addr} and extending for
45922 @var{length} bytes from there:
45925 <memory start="@var{addr}" length="@var{length}"/>
45929 A block indicating trace state variable numbered @var{number} has been
45933 <tvar id="@var{number}"/>
45938 The formal DTD for the traceframe info format is given below:
45941 <!ELEMENT traceframe-info (memory | tvar)* >
45942 <!ATTLIST traceframe-info version CDATA #FIXED "1.0">
45944 <!ELEMENT memory EMPTY>
45945 <!ATTLIST memory start CDATA #REQUIRED
45946 length CDATA #REQUIRED>
45948 <!ATTLIST tvar id CDATA #REQUIRED>
45951 @node Branch Trace Format
45952 @section Branch Trace Format
45953 @cindex branch trace format
45955 In order to display the branch trace of an inferior thread,
45956 @value{GDBN} needs to obtain the list of branches. This list is
45957 represented as list of sequential code blocks that are connected via
45958 branches. The code in each block has been executed sequentially.
45960 This list is obtained using the @samp{qXfer:btrace:read}
45961 (@pxref{qXfer btrace read}) packet and is an XML document.
45963 @value{GDBN} must be linked with the Expat library to support XML
45964 traceframe info discovery. @xref{Expat}.
45966 The top-level structure of the document is shown below:
45969 <?xml version="1.0"?>
45971 PUBLIC "+//IDN gnu.org//DTD GDB Branch Trace V1.0//EN"
45972 "http://sourceware.org/gdb/gdb-btrace.dtd">
45981 A block of sequentially executed instructions starting at @var{begin}
45982 and ending at @var{end}:
45985 <block begin="@var{begin}" end="@var{end}"/>
45990 The formal DTD for the branch trace format is given below:
45993 <!ELEMENT btrace (block* | pt) >
45994 <!ATTLIST btrace version CDATA #FIXED "1.0">
45996 <!ELEMENT block EMPTY>
45997 <!ATTLIST block begin CDATA #REQUIRED
45998 end CDATA #REQUIRED>
46000 <!ELEMENT pt (pt-config?, raw?)>
46002 <!ELEMENT pt-config (cpu?)>
46004 <!ELEMENT cpu EMPTY>
46005 <!ATTLIST cpu vendor CDATA #REQUIRED
46006 family CDATA #REQUIRED
46007 model CDATA #REQUIRED
46008 stepping CDATA #REQUIRED>
46010 <!ELEMENT raw (#PCDATA)>
46013 @node Branch Trace Configuration Format
46014 @section Branch Trace Configuration Format
46015 @cindex branch trace configuration format
46017 For each inferior thread, @value{GDBN} can obtain the branch trace
46018 configuration using the @samp{qXfer:btrace-conf:read}
46019 (@pxref{qXfer btrace-conf read}) packet.
46021 The configuration describes the branch trace format and configuration
46022 settings for that format. The following information is described:
46026 This thread uses the @dfn{Branch Trace Store} (@acronym{BTS}) format.
46029 The size of the @acronym{BTS} ring buffer in bytes.
46032 This thread uses the @dfn{Intel Processor Trace} (@acronym{Intel
46036 The size of the @acronym{Intel PT} ring buffer in bytes.
46040 @value{GDBN} must be linked with the Expat library to support XML
46041 branch trace configuration discovery. @xref{Expat}.
46043 The formal DTD for the branch trace configuration format is given below:
46046 <!ELEMENT btrace-conf (bts?, pt?)>
46047 <!ATTLIST btrace-conf version CDATA #FIXED "1.0">
46049 <!ELEMENT bts EMPTY>
46050 <!ATTLIST bts size CDATA #IMPLIED>
46052 <!ELEMENT pt EMPTY>
46053 <!ATTLIST pt size CDATA #IMPLIED>
46056 @include agentexpr.texi
46058 @node Target Descriptions
46059 @appendix Target Descriptions
46060 @cindex target descriptions
46062 One of the challenges of using @value{GDBN} to debug embedded systems
46063 is that there are so many minor variants of each processor
46064 architecture in use. It is common practice for vendors to start with
46065 a standard processor core --- ARM, PowerPC, or @acronym{MIPS}, for example ---
46066 and then make changes to adapt it to a particular market niche. Some
46067 architectures have hundreds of variants, available from dozens of
46068 vendors. This leads to a number of problems:
46072 With so many different customized processors, it is difficult for
46073 the @value{GDBN} maintainers to keep up with the changes.
46075 Since individual variants may have short lifetimes or limited
46076 audiences, it may not be worthwhile to carry information about every
46077 variant in the @value{GDBN} source tree.
46079 When @value{GDBN} does support the architecture of the embedded system
46080 at hand, the task of finding the correct architecture name to give the
46081 @command{set architecture} command can be error-prone.
46084 To address these problems, the @value{GDBN} remote protocol allows a
46085 target system to not only identify itself to @value{GDBN}, but to
46086 actually describe its own features. This lets @value{GDBN} support
46087 processor variants it has never seen before --- to the extent that the
46088 descriptions are accurate, and that @value{GDBN} understands them.
46090 @value{GDBN} must be linked with the Expat library to support XML
46091 target descriptions. @xref{Expat}.
46094 * Retrieving Descriptions:: How descriptions are fetched from a target.
46095 * Target Description Format:: The contents of a target description.
46096 * Predefined Target Types:: Standard types available for target
46098 * Enum Target Types:: How to define enum target types.
46099 * Standard Target Features:: Features @value{GDBN} knows about.
46102 @node Retrieving Descriptions
46103 @section Retrieving Descriptions
46105 Target descriptions can be read from the target automatically, or
46106 specified by the user manually. The default behavior is to read the
46107 description from the target. @value{GDBN} retrieves it via the remote
46108 protocol using @samp{qXfer} requests (@pxref{General Query Packets,
46109 qXfer}). The @var{annex} in the @samp{qXfer} packet will be
46110 @samp{target.xml}. The contents of the @samp{target.xml} annex are an
46111 XML document, of the form described in @ref{Target Description
46114 Alternatively, you can specify a file to read for the target description.
46115 If a file is set, the target will not be queried. The commands to
46116 specify a file are:
46119 @cindex set tdesc filename
46120 @item set tdesc filename @var{path}
46121 Read the target description from @var{path}.
46123 @cindex unset tdesc filename
46124 @item unset tdesc filename
46125 Do not read the XML target description from a file. @value{GDBN}
46126 will use the description supplied by the current target.
46128 @cindex show tdesc filename
46129 @item show tdesc filename
46130 Show the filename to read for a target description, if any.
46134 @node Target Description Format
46135 @section Target Description Format
46136 @cindex target descriptions, XML format
46138 A target description annex is an @uref{http://www.w3.org/XML/, XML}
46139 document which complies with the Document Type Definition provided in
46140 the @value{GDBN} sources in @file{gdb/features/gdb-target.dtd}. This
46141 means you can use generally available tools like @command{xmllint} to
46142 check that your feature descriptions are well-formed and valid.
46143 However, to help people unfamiliar with XML write descriptions for
46144 their targets, we also describe the grammar here.
46146 Target descriptions can identify the architecture of the remote target
46147 and (for some architectures) provide information about custom register
46148 sets. They can also identify the OS ABI of the remote target.
46149 @value{GDBN} can use this information to autoconfigure for your
46150 target, or to warn you if you connect to an unsupported target.
46152 Here is a simple target description:
46155 <target version="1.0">
46156 <architecture>i386:x86-64</architecture>
46161 This minimal description only says that the target uses
46162 the x86-64 architecture.
46164 A target description has the following overall form, with [ ] marking
46165 optional elements and @dots{} marking repeatable elements. The elements
46166 are explained further below.
46169 <?xml version="1.0"?>
46170 <!DOCTYPE target SYSTEM "gdb-target.dtd">
46171 <target version="1.0">
46172 @r{[}@var{architecture}@r{]}
46173 @r{[}@var{osabi}@r{]}
46174 @r{[}@var{compatible}@r{]}
46175 @r{[}@var{feature}@dots{}@r{]}
46180 The description is generally insensitive to whitespace and line
46181 breaks, under the usual common-sense rules. The XML version
46182 declaration and document type declaration can generally be omitted
46183 (@value{GDBN} does not require them), but specifying them may be
46184 useful for XML validation tools. The @samp{version} attribute for
46185 @samp{<target>} may also be omitted, but we recommend
46186 including it; if future versions of @value{GDBN} use an incompatible
46187 revision of @file{gdb-target.dtd}, they will detect and report
46188 the version mismatch.
46190 @subsection Inclusion
46191 @cindex target descriptions, inclusion
46194 @cindex <xi:include>
46197 It can sometimes be valuable to split a target description up into
46198 several different annexes, either for organizational purposes, or to
46199 share files between different possible target descriptions. You can
46200 divide a description into multiple files by replacing any element of
46201 the target description with an inclusion directive of the form:
46204 <xi:include href="@var{document}"/>
46208 When @value{GDBN} encounters an element of this form, it will retrieve
46209 the named XML @var{document}, and replace the inclusion directive with
46210 the contents of that document. If the current description was read
46211 using @samp{qXfer}, then so will be the included document;
46212 @var{document} will be interpreted as the name of an annex. If the
46213 current description was read from a file, @value{GDBN} will look for
46214 @var{document} as a file in the same directory where it found the
46215 original description.
46217 @subsection Architecture
46218 @cindex <architecture>
46220 An @samp{<architecture>} element has this form:
46223 <architecture>@var{arch}</architecture>
46226 @var{arch} is one of the architectures from the set accepted by
46227 @code{set architecture} (@pxref{Targets, ,Specifying a Debugging Target}).
46230 @cindex @code{<osabi>}
46232 This optional field was introduced in @value{GDBN} version 7.0.
46233 Previous versions of @value{GDBN} ignore it.
46235 An @samp{<osabi>} element has this form:
46238 <osabi>@var{abi-name}</osabi>
46241 @var{abi-name} is an OS ABI name from the same selection accepted by
46242 @w{@code{set osabi}} (@pxref{ABI, ,Configuring the Current ABI}).
46244 @subsection Compatible Architecture
46245 @cindex @code{<compatible>}
46247 This optional field was introduced in @value{GDBN} version 7.0.
46248 Previous versions of @value{GDBN} ignore it.
46250 A @samp{<compatible>} element has this form:
46253 <compatible>@var{arch}</compatible>
46256 @var{arch} is one of the architectures from the set accepted by
46257 @code{set architecture} (@pxref{Targets, ,Specifying a Debugging Target}).
46259 A @samp{<compatible>} element is used to specify that the target
46260 is able to run binaries in some other than the main target architecture
46261 given by the @samp{<architecture>} element. For example, on the
46262 Cell Broadband Engine, the main architecture is @code{powerpc:common}
46263 or @code{powerpc:common64}, but the system is able to run binaries
46264 in the @code{spu} architecture as well. The way to describe this
46265 capability with @samp{<compatible>} is as follows:
46268 <architecture>powerpc:common</architecture>
46269 <compatible>spu</compatible>
46272 @subsection Features
46275 Each @samp{<feature>} describes some logical portion of the target
46276 system. Features are currently used to describe available CPU
46277 registers and the types of their contents. A @samp{<feature>} element
46281 <feature name="@var{name}">
46282 @r{[}@var{type}@dots{}@r{]}
46288 Each feature's name should be unique within the description. The name
46289 of a feature does not matter unless @value{GDBN} has some special
46290 knowledge of the contents of that feature; if it does, the feature
46291 should have its standard name. @xref{Standard Target Features}.
46295 Any register's value is a collection of bits which @value{GDBN} must
46296 interpret. The default interpretation is a two's complement integer,
46297 but other types can be requested by name in the register description.
46298 Some predefined types are provided by @value{GDBN} (@pxref{Predefined
46299 Target Types}), and the description can define additional composite
46302 Each type element must have an @samp{id} attribute, which gives
46303 a unique (within the containing @samp{<feature>}) name to the type.
46304 Types must be defined before they are used.
46307 Some targets offer vector registers, which can be treated as arrays
46308 of scalar elements. These types are written as @samp{<vector>} elements,
46309 specifying the array element type, @var{type}, and the number of elements,
46313 <vector id="@var{id}" type="@var{type}" count="@var{count}"/>
46317 If a register's value is usefully viewed in multiple ways, define it
46318 with a union type containing the useful representations. The
46319 @samp{<union>} element contains one or more @samp{<field>} elements,
46320 each of which has a @var{name} and a @var{type}:
46323 <union id="@var{id}">
46324 <field name="@var{name}" type="@var{type}"/>
46331 If a register's value is composed from several separate values, define
46332 it with either a structure type or a flags type.
46333 A flags type may only contain bitfields.
46334 A structure type may either contain only bitfields or contain no bitfields.
46335 If the value contains only bitfields, its total size in bytes must be
46338 Non-bitfield values have a @var{name} and @var{type}.
46341 <struct id="@var{id}">
46342 <field name="@var{name}" type="@var{type}"/>
46347 Both @var{name} and @var{type} values are required.
46348 No implicit padding is added.
46350 Bitfield values have a @var{name}, @var{start}, @var{end} and @var{type}.
46353 <struct id="@var{id}" size="@var{size}">
46354 <field name="@var{name}" start="@var{start}" end="@var{end}" type="@var{type}"/>
46360 <flags id="@var{id}" size="@var{size}">
46361 <field name="@var{name}" start="@var{start}" end="@var{end}" type="@var{type}"/>
46366 The @var{name} value is required.
46367 Bitfield values may be named with the empty string, @samp{""},
46368 in which case the field is ``filler'' and its value is not printed.
46369 Not all bits need to be specified, so ``filler'' fields are optional.
46371 The @var{start} and @var{end} values are required, and @var{type}
46373 The field's @var{start} must be less than or equal to its @var{end},
46374 and zero represents the least significant bit.
46376 The default value of @var{type} is @code{bool} for single bit fields,
46377 and an unsigned integer otherwise.
46379 Which to choose? Structures or flags?
46381 Registers defined with @samp{flags} have these advantages over
46382 defining them with @samp{struct}:
46386 Arithmetic may be performed on them as if they were integers.
46388 They are printed in a more readable fashion.
46391 Registers defined with @samp{struct} have one advantage over
46392 defining them with @samp{flags}:
46396 One can fetch individual fields like in @samp{C}.
46399 (@value{GDBP}) print $my_struct_reg.field3
46405 @subsection Registers
46408 Each register is represented as an element with this form:
46411 <reg name="@var{name}"
46412 bitsize="@var{size}"
46413 @r{[}regnum="@var{num}"@r{]}
46414 @r{[}save-restore="@var{save-restore}"@r{]}
46415 @r{[}type="@var{type}"@r{]}
46416 @r{[}group="@var{group}"@r{]}/>
46420 The components are as follows:
46425 The register's name; it must be unique within the target description.
46428 The register's size, in bits.
46431 The register's number. If omitted, a register's number is one greater
46432 than that of the previous register (either in the current feature or in
46433 a preceding feature); the first register in the target description
46434 defaults to zero. This register number is used to read or write
46435 the register; e.g.@: it is used in the remote @code{p} and @code{P}
46436 packets, and registers appear in the @code{g} and @code{G} packets
46437 in order of increasing register number.
46440 Whether the register should be preserved across inferior function
46441 calls; this must be either @code{yes} or @code{no}. The default is
46442 @code{yes}, which is appropriate for most registers except for
46443 some system control registers; this is not related to the target's
46447 The type of the register. It may be a predefined type, a type
46448 defined in the current feature, or one of the special types @code{int}
46449 and @code{float}. @code{int} is an integer type of the correct size
46450 for @var{bitsize}, and @code{float} is a floating point type (in the
46451 architecture's normal floating point format) of the correct size for
46452 @var{bitsize}. The default is @code{int}.
46455 The register group to which this register belongs. It can be one of the
46456 standard register groups @code{general}, @code{float}, @code{vector} or an
46457 arbitrary string. Group names should be limited to alphanumeric characters.
46458 If a group name is made up of multiple words the words may be separated by
46459 hyphens; e.g.@: @code{special-group} or @code{ultra-special-group}. If no
46460 @var{group} is specified, @value{GDBN} will not display the register in
46461 @code{info registers}.
46465 @node Predefined Target Types
46466 @section Predefined Target Types
46467 @cindex target descriptions, predefined types
46469 Type definitions in the self-description can build up composite types
46470 from basic building blocks, but can not define fundamental types. Instead,
46471 standard identifiers are provided by @value{GDBN} for the fundamental
46472 types. The currently supported types are:
46477 Boolean type, occupying a single bit.
46485 Signed integer types holding the specified number of bits.
46493 Unsigned integer types holding the specified number of bits.
46497 Pointers to unspecified code and data. The program counter and
46498 any dedicated return address register may be marked as code
46499 pointers; printing a code pointer converts it into a symbolic
46500 address. The stack pointer and any dedicated address registers
46501 may be marked as data pointers.
46504 Single precision IEEE floating point.
46507 Double precision IEEE floating point.
46510 The 12-byte extended precision format used by ARM FPA registers.
46513 The 10-byte extended precision format used by x87 registers.
46516 32bit @sc{eflags} register used by x86.
46519 32bit @sc{mxcsr} register used by x86.
46523 @node Enum Target Types
46524 @section Enum Target Types
46525 @cindex target descriptions, enum types
46527 Enum target types are useful in @samp{struct} and @samp{flags}
46528 register descriptions. @xref{Target Description Format}.
46530 Enum types have a name, size and a list of name/value pairs.
46533 <enum id="@var{id}" size="@var{size}">
46534 <evalue name="@var{name}" value="@var{value}"/>
46539 Enums must be defined before they are used.
46542 <enum id="levels_type" size="4">
46543 <evalue name="low" value="0"/>
46544 <evalue name="high" value="1"/>
46546 <flags id="flags_type" size="4">
46547 <field name="X" start="0"/>
46548 <field name="LEVEL" start="1" end="1" type="levels_type"/>
46550 <reg name="flags" bitsize="32" type="flags_type"/>
46553 Given that description, a value of 3 for the @samp{flags} register
46554 would be printed as:
46557 (@value{GDBP}) info register flags
46558 flags 0x3 [ X LEVEL=high ]
46561 @node Standard Target Features
46562 @section Standard Target Features
46563 @cindex target descriptions, standard features
46565 A target description must contain either no registers or all the
46566 target's registers. If the description contains no registers, then
46567 @value{GDBN} will assume a default register layout, selected based on
46568 the architecture. If the description contains any registers, the
46569 default layout will not be used; the standard registers must be
46570 described in the target description, in such a way that @value{GDBN}
46571 can recognize them.
46573 This is accomplished by giving specific names to feature elements
46574 which contain standard registers. @value{GDBN} will look for features
46575 with those names and verify that they contain the expected registers;
46576 if any known feature is missing required registers, or if any required
46577 feature is missing, @value{GDBN} will reject the target
46578 description. You can add additional registers to any of the
46579 standard features --- @value{GDBN} will display them just as if
46580 they were added to an unrecognized feature.
46582 This section lists the known features and their expected contents.
46583 Sample XML documents for these features are included in the
46584 @value{GDBN} source tree, in the directory @file{gdb/features}.
46586 Names recognized by @value{GDBN} should include the name of the
46587 company or organization which selected the name, and the overall
46588 architecture to which the feature applies; so e.g.@: the feature
46589 containing ARM core registers is named @samp{org.gnu.gdb.arm.core}.
46591 The names of registers are not case sensitive for the purpose
46592 of recognizing standard features, but @value{GDBN} will only display
46593 registers using the capitalization used in the description.
46596 * AArch64 Features::
46600 * MicroBlaze Features::
46604 * Nios II Features::
46605 * OpenRISC 1000 Features::
46606 * PowerPC Features::
46607 * RISC-V Features::
46609 * S/390 and System z Features::
46615 @node AArch64 Features
46616 @subsection AArch64 Features
46617 @cindex target descriptions, AArch64 features
46619 The @samp{org.gnu.gdb.aarch64.core} feature is required for AArch64
46620 targets. It should contain registers @samp{x0} through @samp{x30},
46621 @samp{sp}, @samp{pc}, and @samp{cpsr}.
46623 The @samp{org.gnu.gdb.aarch64.fpu} feature is optional. If present,
46624 it should contain registers @samp{v0} through @samp{v31}, @samp{fpsr},
46627 The @samp{org.gnu.gdb.aarch64.sve} feature is optional. If present,
46628 it should contain registers @samp{z0} through @samp{z31}, @samp{p0}
46629 through @samp{p15}, @samp{ffr} and @samp{vg}.
46631 The @samp{org.gnu.gdb.aarch64.pauth} feature is optional. If present,
46632 it should contain registers @samp{pauth_dmask} and @samp{pauth_cmask}.
46635 @subsection ARC Features
46636 @cindex target descriptions, ARC Features
46638 ARC processors are highly configurable, so even core registers and their number
46639 are not completely predetermined. In addition flags and PC registers which are
46640 important to @value{GDBN} are not ``core'' registers in ARC. It is required
46641 that one of the core registers features is present.
46642 @samp{org.gnu.gdb.arc.aux-minimal} feature is mandatory.
46644 The @samp{org.gnu.gdb.arc.core.v2} feature is required for ARC EM and ARC HS
46645 targets with a normal register file. It should contain registers @samp{r0}
46646 through @samp{r25}, @samp{gp}, @samp{fp}, @samp{sp}, @samp{r30}, @samp{blink},
46647 @samp{lp_count} and @samp{pcl}. This feature may contain register @samp{ilink}
46648 and any of extension core registers @samp{r32} through @samp{r59/acch}.
46649 @samp{ilink} and extension core registers are not available to read/write, when
46650 debugging GNU/Linux applications, thus @samp{ilink} is made optional.
46652 The @samp{org.gnu.gdb.arc.core-reduced.v2} feature is required for ARC EM and
46653 ARC HS targets with a reduced register file. It should contain registers
46654 @samp{r0} through @samp{r3}, @samp{r10} through @samp{r15}, @samp{gp},
46655 @samp{fp}, @samp{sp}, @samp{r30}, @samp{blink}, @samp{lp_count} and @samp{pcl}.
46656 This feature may contain register @samp{ilink} and any of extension core
46657 registers @samp{r32} through @samp{r59/acch}.
46659 The @samp{org.gnu.gdb.arc.core.arcompact} feature is required for ARCompact
46660 targets with a normal register file. It should contain registers @samp{r0}
46661 through @samp{r25}, @samp{gp}, @samp{fp}, @samp{sp}, @samp{r30}, @samp{blink},
46662 @samp{lp_count} and @samp{pcl}. This feature may contain registers
46663 @samp{ilink1}, @samp{ilink2} and any of extension core registers @samp{r32}
46664 through @samp{r59/acch}. @samp{ilink1} and @samp{ilink2} and extension core
46665 registers are not available when debugging GNU/Linux applications. The only
46666 difference with @samp{org.gnu.gdb.arc.core.v2} feature is in the names of
46667 @samp{ilink1} and @samp{ilink2} registers and that @samp{r30} is mandatory in
46668 ARC v2, but @samp{ilink2} is optional on ARCompact.
46670 The @samp{org.gnu.gdb.arc.aux-minimal} feature is required for all ARC
46671 targets. It should contain registers @samp{pc} and @samp{status32}.
46674 @subsection ARM Features
46675 @cindex target descriptions, ARM features
46677 The @samp{org.gnu.gdb.arm.core} feature is required for non-M-profile
46679 It should contain registers @samp{r0} through @samp{r13}, @samp{sp},
46680 @samp{lr}, @samp{pc}, and @samp{cpsr}.
46682 For M-profile targets (e.g. Cortex-M3), the @samp{org.gnu.gdb.arm.core}
46683 feature is replaced by @samp{org.gnu.gdb.arm.m-profile}. It should contain
46684 registers @samp{r0} through @samp{r13}, @samp{sp}, @samp{lr}, @samp{pc},
46687 The @samp{org.gnu.gdb.arm.fpa} feature is optional. If present, it
46688 should contain registers @samp{f0} through @samp{f7} and @samp{fps}.
46690 The @samp{org.gnu.gdb.xscale.iwmmxt} feature is optional. If present,
46691 it should contain at least registers @samp{wR0} through @samp{wR15} and
46692 @samp{wCGR0} through @samp{wCGR3}. The @samp{wCID}, @samp{wCon},
46693 @samp{wCSSF}, and @samp{wCASF} registers are optional.
46695 The @samp{org.gnu.gdb.arm.vfp} feature is optional. If present, it
46696 should contain at least registers @samp{d0} through @samp{d15}. If
46697 they are present, @samp{d16} through @samp{d31} should also be included.
46698 @value{GDBN} will synthesize the single-precision registers from
46699 halves of the double-precision registers.
46701 The @samp{org.gnu.gdb.arm.neon} feature is optional. It does not
46702 need to contain registers; it instructs @value{GDBN} to display the
46703 VFP double-precision registers as vectors and to synthesize the
46704 quad-precision registers from pairs of double-precision registers.
46705 If this feature is present, @samp{org.gnu.gdb.arm.vfp} must also
46706 be present and include 32 double-precision registers.
46708 @node i386 Features
46709 @subsection i386 Features
46710 @cindex target descriptions, i386 features
46712 The @samp{org.gnu.gdb.i386.core} feature is required for i386/amd64
46713 targets. It should describe the following registers:
46717 @samp{eax} through @samp{edi} plus @samp{eip} for i386
46719 @samp{rax} through @samp{r15} plus @samp{rip} for amd64
46721 @samp{eflags}, @samp{cs}, @samp{ss}, @samp{ds}, @samp{es},
46722 @samp{fs}, @samp{gs}
46724 @samp{st0} through @samp{st7}
46726 @samp{fctrl}, @samp{fstat}, @samp{ftag}, @samp{fiseg}, @samp{fioff},
46727 @samp{foseg}, @samp{fooff} and @samp{fop}
46730 The register sets may be different, depending on the target.
46732 The @samp{org.gnu.gdb.i386.sse} feature is optional. It should
46733 describe registers:
46737 @samp{xmm0} through @samp{xmm7} for i386
46739 @samp{xmm0} through @samp{xmm15} for amd64
46744 The @samp{org.gnu.gdb.i386.avx} feature is optional and requires the
46745 @samp{org.gnu.gdb.i386.sse} feature. It should
46746 describe the upper 128 bits of @sc{ymm} registers:
46750 @samp{ymm0h} through @samp{ymm7h} for i386
46752 @samp{ymm0h} through @samp{ymm15h} for amd64
46755 The @samp{org.gnu.gdb.i386.mpx} is an optional feature representing Intel
46756 Memory Protection Extension (MPX). It should describe the following registers:
46760 @samp{bnd0raw} through @samp{bnd3raw} for i386 and amd64.
46762 @samp{bndcfgu} and @samp{bndstatus} for i386 and amd64.
46765 The @samp{org.gnu.gdb.i386.linux} feature is optional. It should
46766 describe a single register, @samp{orig_eax}.
46768 The @samp{org.gnu.gdb.i386.segments} feature is optional. It should
46769 describe two system registers: @samp{fs_base} and @samp{gs_base}.
46771 The @samp{org.gnu.gdb.i386.avx512} feature is optional and requires the
46772 @samp{org.gnu.gdb.i386.avx} feature. It should
46773 describe additional @sc{xmm} registers:
46777 @samp{xmm16h} through @samp{xmm31h}, only valid for amd64.
46780 It should describe the upper 128 bits of additional @sc{ymm} registers:
46784 @samp{ymm16h} through @samp{ymm31h}, only valid for amd64.
46788 describe the upper 256 bits of @sc{zmm} registers:
46792 @samp{zmm0h} through @samp{zmm7h} for i386.
46794 @samp{zmm0h} through @samp{zmm15h} for amd64.
46798 describe the additional @sc{zmm} registers:
46802 @samp{zmm16h} through @samp{zmm31h}, only valid for amd64.
46805 The @samp{org.gnu.gdb.i386.pkeys} feature is optional. It should
46806 describe a single register, @samp{pkru}. It is a 32-bit register
46807 valid for i386 and amd64.
46809 @node MicroBlaze Features
46810 @subsection MicroBlaze Features
46811 @cindex target descriptions, MicroBlaze features
46813 The @samp{org.gnu.gdb.microblaze.core} feature is required for MicroBlaze
46814 targets. It should contain registers @samp{r0} through @samp{r31},
46815 @samp{rpc}, @samp{rmsr}, @samp{rear}, @samp{resr}, @samp{rfsr}, @samp{rbtr},
46816 @samp{rpvr}, @samp{rpvr1} through @samp{rpvr11}, @samp{redr}, @samp{rpid},
46817 @samp{rzpr}, @samp{rtlbx}, @samp{rtlbsx}, @samp{rtlblo}, and @samp{rtlbhi}.
46819 The @samp{org.gnu.gdb.microblaze.stack-protect} feature is optional.
46820 If present, it should contain registers @samp{rshr} and @samp{rslr}
46822 @node MIPS Features
46823 @subsection @acronym{MIPS} Features
46824 @cindex target descriptions, @acronym{MIPS} features
46826 The @samp{org.gnu.gdb.mips.cpu} feature is required for @acronym{MIPS} targets.
46827 It should contain registers @samp{r0} through @samp{r31}, @samp{lo},
46828 @samp{hi}, and @samp{pc}. They may be 32-bit or 64-bit depending
46831 The @samp{org.gnu.gdb.mips.cp0} feature is also required. It should
46832 contain at least the @samp{status}, @samp{badvaddr}, and @samp{cause}
46833 registers. They may be 32-bit or 64-bit depending on the target.
46835 The @samp{org.gnu.gdb.mips.fpu} feature is currently required, though
46836 it may be optional in a future version of @value{GDBN}. It should
46837 contain registers @samp{f0} through @samp{f31}, @samp{fcsr}, and
46838 @samp{fir}. They may be 32-bit or 64-bit depending on the target.
46840 The @samp{org.gnu.gdb.mips.dsp} feature is optional. It should
46841 contain registers @samp{hi1} through @samp{hi3}, @samp{lo1} through
46842 @samp{lo3}, and @samp{dspctl}. The @samp{dspctl} register should
46843 be 32-bit and the rest may be 32-bit or 64-bit depending on the target.
46845 The @samp{org.gnu.gdb.mips.linux} feature is optional. It should
46846 contain a single register, @samp{restart}, which is used by the
46847 Linux kernel to control restartable syscalls.
46849 @node M68K Features
46850 @subsection M68K Features
46851 @cindex target descriptions, M68K features
46854 @item @samp{org.gnu.gdb.m68k.core}
46855 @itemx @samp{org.gnu.gdb.coldfire.core}
46856 @itemx @samp{org.gnu.gdb.fido.core}
46857 One of those features must be always present.
46858 The feature that is present determines which flavor of m68k is
46859 used. The feature that is present should contain registers
46860 @samp{d0} through @samp{d7}, @samp{a0} through @samp{a5}, @samp{fp},
46861 @samp{sp}, @samp{ps} and @samp{pc}.
46863 @item @samp{org.gnu.gdb.coldfire.fp}
46864 This feature is optional. If present, it should contain registers
46865 @samp{fp0} through @samp{fp7}, @samp{fpcontrol}, @samp{fpstatus} and
46869 @node NDS32 Features
46870 @subsection NDS32 Features
46871 @cindex target descriptions, NDS32 features
46873 The @samp{org.gnu.gdb.nds32.core} feature is required for NDS32
46874 targets. It should contain at least registers @samp{r0} through
46875 @samp{r10}, @samp{r15}, @samp{fp}, @samp{gp}, @samp{lp}, @samp{sp},
46878 The @samp{org.gnu.gdb.nds32.fpu} feature is optional. If present,
46879 it should contain 64-bit double-precision floating-point registers
46880 @samp{fd0} through @emph{fdN}, which should be @samp{fd3}, @samp{fd7},
46881 @samp{fd15}, or @samp{fd31} based on the FPU configuration implemented.
46883 @emph{Note:} The first sixteen 64-bit double-precision floating-point
46884 registers are overlapped with the thirty-two 32-bit single-precision
46885 floating-point registers. The 32-bit single-precision registers, if
46886 not being listed explicitly, will be synthesized from halves of the
46887 overlapping 64-bit double-precision registers. Listing 32-bit
46888 single-precision registers explicitly is deprecated, and the
46889 support to it could be totally removed some day.
46891 @node Nios II Features
46892 @subsection Nios II Features
46893 @cindex target descriptions, Nios II features
46895 The @samp{org.gnu.gdb.nios2.cpu} feature is required for Nios II
46896 targets. It should contain the 32 core registers (@samp{zero},
46897 @samp{at}, @samp{r2} through @samp{r23}, @samp{et} through @samp{ra}),
46898 @samp{pc}, and the 16 control registers (@samp{status} through
46901 @node OpenRISC 1000 Features
46902 @subsection Openrisc 1000 Features
46903 @cindex target descriptions, OpenRISC 1000 features
46905 The @samp{org.gnu.gdb.or1k.group0} feature is required for OpenRISC 1000
46906 targets. It should contain the 32 general purpose registers (@samp{r0}
46907 through @samp{r31}), @samp{ppc}, @samp{npc} and @samp{sr}.
46909 @node PowerPC Features
46910 @subsection PowerPC Features
46911 @cindex target descriptions, PowerPC features
46913 The @samp{org.gnu.gdb.power.core} feature is required for PowerPC
46914 targets. It should contain registers @samp{r0} through @samp{r31},
46915 @samp{pc}, @samp{msr}, @samp{cr}, @samp{lr}, @samp{ctr}, and
46916 @samp{xer}. They may be 32-bit or 64-bit depending on the target.
46918 The @samp{org.gnu.gdb.power.fpu} feature is optional. It should
46919 contain registers @samp{f0} through @samp{f31} and @samp{fpscr}.
46921 The @samp{org.gnu.gdb.power.altivec} feature is optional. It should
46922 contain registers @samp{vr0} through @samp{vr31}, @samp{vscr}, and
46923 @samp{vrsave}. @value{GDBN} will define pseudo-registers @samp{v0}
46924 through @samp{v31} as aliases for the corresponding @samp{vrX}
46927 The @samp{org.gnu.gdb.power.vsx} feature is optional. It should
46928 contain registers @samp{vs0h} through @samp{vs31h}. @value{GDBN} will
46929 combine these registers with the floating point registers (@samp{f0}
46930 through @samp{f31}) and the altivec registers (@samp{vr0} through
46931 @samp{vr31}) to present the 128-bit wide registers @samp{vs0} through
46932 @samp{vs63}, the set of vector-scalar registers for POWER7.
46933 Therefore, this feature requires both @samp{org.gnu.gdb.power.fpu} and
46934 @samp{org.gnu.gdb.power.altivec}.
46936 The @samp{org.gnu.gdb.power.spe} feature is optional. It should
46937 contain registers @samp{ev0h} through @samp{ev31h}, @samp{acc}, and
46938 @samp{spefscr}. SPE targets should provide 32-bit registers in
46939 @samp{org.gnu.gdb.power.core} and provide the upper halves in
46940 @samp{ev0h} through @samp{ev31h}. @value{GDBN} will combine
46941 these to present registers @samp{ev0} through @samp{ev31} to the
46944 The @samp{org.gnu.gdb.power.ppr} feature is optional. It should
46945 contain the 64-bit register @samp{ppr}.
46947 The @samp{org.gnu.gdb.power.dscr} feature is optional. It should
46948 contain the 64-bit register @samp{dscr}.
46950 The @samp{org.gnu.gdb.power.tar} feature is optional. It should
46951 contain the 64-bit register @samp{tar}.
46953 The @samp{org.gnu.gdb.power.ebb} feature is optional. It should
46954 contain registers @samp{bescr}, @samp{ebbhr} and @samp{ebbrr}, all
46957 The @samp{org.gnu.gdb.power.linux.pmu} feature is optional. It should
46958 contain registers @samp{mmcr0}, @samp{mmcr2}, @samp{siar}, @samp{sdar}
46959 and @samp{sier}, all 64-bit wide. This is the subset of the isa 2.07
46960 server PMU registers provided by @sc{gnu}/Linux.
46962 The @samp{org.gnu.gdb.power.htm.spr} feature is optional. It should
46963 contain registers @samp{tfhar}, @samp{texasr} and @samp{tfiar}, all
46966 The @samp{org.gnu.gdb.power.htm.core} feature is optional. It should
46967 contain the checkpointed general-purpose registers @samp{cr0} through
46968 @samp{cr31}, as well as the checkpointed registers @samp{clr} and
46969 @samp{cctr}. These registers may all be either 32-bit or 64-bit
46970 depending on the target. It should also contain the checkpointed
46971 registers @samp{ccr} and @samp{cxer}, which should both be 32-bit
46974 The @samp{org.gnu.gdb.power.htm.fpu} feature is optional. It should
46975 contain the checkpointed 64-bit floating-point registers @samp{cf0}
46976 through @samp{cf31}, as well as the checkpointed 64-bit register
46979 The @samp{org.gnu.gdb.power.htm.altivec} feature is optional. It
46980 should contain the checkpointed altivec registers @samp{cvr0} through
46981 @samp{cvr31}, all 128-bit wide. It should also contain the
46982 checkpointed registers @samp{cvscr} and @samp{cvrsave}, both 32-bit
46985 The @samp{org.gnu.gdb.power.htm.vsx} feature is optional. It should
46986 contain registers @samp{cvs0h} through @samp{cvs31h}. @value{GDBN}
46987 will combine these registers with the checkpointed floating point
46988 registers (@samp{cf0} through @samp{cf31}) and the checkpointed
46989 altivec registers (@samp{cvr0} through @samp{cvr31}) to present the
46990 128-bit wide checkpointed vector-scalar registers @samp{cvs0} through
46991 @samp{cvs63}. Therefore, this feature requires both
46992 @samp{org.gnu.gdb.power.htm.altivec} and
46993 @samp{org.gnu.gdb.power.htm.fpu}.
46995 The @samp{org.gnu.gdb.power.htm.ppr} feature is optional. It should
46996 contain the 64-bit checkpointed register @samp{cppr}.
46998 The @samp{org.gnu.gdb.power.htm.dscr} feature is optional. It should
46999 contain the 64-bit checkpointed register @samp{cdscr}.
47001 The @samp{org.gnu.gdb.power.htm.tar} feature is optional. It should
47002 contain the 64-bit checkpointed register @samp{ctar}.
47005 @node RISC-V Features
47006 @subsection RISC-V Features
47007 @cindex target descriptions, RISC-V Features
47009 The @samp{org.gnu.gdb.riscv.cpu} feature is required for RISC-V
47010 targets. It should contain the registers @samp{x0} through
47011 @samp{x31}, and @samp{pc}. Either the architectural names (@samp{x0},
47012 @samp{x1}, etc) can be used, or the ABI names (@samp{zero}, @samp{ra},
47015 The @samp{org.gnu.gdb.riscv.fpu} feature is optional. If present, it
47016 should contain registers @samp{f0} through @samp{f31}, @samp{fflags},
47017 @samp{frm}, and @samp{fcsr}. As with the cpu feature, either the
47018 architectural register names, or the ABI names can be used.
47020 The @samp{org.gnu.gdb.riscv.virtual} feature is optional. If present,
47021 it should contain registers that are not backed by real registers on
47022 the target, but are instead virtual, where the register value is
47023 derived from other target state. In many ways these are like
47024 @value{GDBN}s pseudo-registers, except implemented by the target.
47025 Currently the only register expected in this set is the one byte
47026 @samp{priv} register that contains the target's privilege level in the
47027 least significant two bits.
47029 The @samp{org.gnu.gdb.riscv.csr} feature is optional. If present, it
47030 should contain all of the target's standard CSRs. Standard CSRs are
47031 those defined in the RISC-V specification documents. There is some
47032 overlap between this feature and the fpu feature; the @samp{fflags},
47033 @samp{frm}, and @samp{fcsr} registers could be in either feature. The
47034 expectation is that these registers will be in the fpu feature if the
47035 target has floating point hardware, but can be moved into the csr
47036 feature if the target has the floating point control registers, but no
47037 other floating point hardware.
47040 @subsection RX Features
47041 @cindex target descriptions, RX Features
47043 The @samp{org.gnu.gdb.rx.core} feature is required for RX
47044 targets. It should contain the registers @samp{r0} through
47045 @samp{r15}, @samp{usp}, @samp{isp}, @samp{psw}, @samp{pc}, @samp{intb},
47046 @samp{bpsw}, @samp{bpc}, @samp{fintv}, @samp{fpsw}, and @samp{acc}.
47048 @node S/390 and System z Features
47049 @subsection S/390 and System z Features
47050 @cindex target descriptions, S/390 features
47051 @cindex target descriptions, System z features
47053 The @samp{org.gnu.gdb.s390.core} feature is required for S/390 and
47054 System z targets. It should contain the PSW and the 16 general
47055 registers. In particular, System z targets should provide the 64-bit
47056 registers @samp{pswm}, @samp{pswa}, and @samp{r0} through @samp{r15}.
47057 S/390 targets should provide the 32-bit versions of these registers.
47058 A System z target that runs in 31-bit addressing mode should provide
47059 32-bit versions of @samp{pswm} and @samp{pswa}, as well as the general
47060 register's upper halves @samp{r0h} through @samp{r15h}, and their
47061 lower halves @samp{r0l} through @samp{r15l}.
47063 The @samp{org.gnu.gdb.s390.fpr} feature is required. It should
47064 contain the 64-bit registers @samp{f0} through @samp{f15}, and
47067 The @samp{org.gnu.gdb.s390.acr} feature is required. It should
47068 contain the 32-bit registers @samp{acr0} through @samp{acr15}.
47070 The @samp{org.gnu.gdb.s390.linux} feature is optional. It should
47071 contain the register @samp{orig_r2}, which is 64-bit wide on System z
47072 targets and 32-bit otherwise. In addition, the feature may contain
47073 the @samp{last_break} register, whose width depends on the addressing
47074 mode, as well as the @samp{system_call} register, which is always
47077 The @samp{org.gnu.gdb.s390.tdb} feature is optional. It should
47078 contain the 64-bit registers @samp{tdb0}, @samp{tac}, @samp{tct},
47079 @samp{atia}, and @samp{tr0} through @samp{tr15}.
47081 The @samp{org.gnu.gdb.s390.vx} feature is optional. It should contain
47082 64-bit wide registers @samp{v0l} through @samp{v15l}, which will be
47083 combined by @value{GDBN} with the floating point registers @samp{f0}
47084 through @samp{f15} to present the 128-bit wide vector registers
47085 @samp{v0} through @samp{v15}. In addition, this feature should
47086 contain the 128-bit wide vector registers @samp{v16} through
47089 The @samp{org.gnu.gdb.s390.gs} feature is optional. It should contain
47090 the 64-bit wide guarded-storage-control registers @samp{gsd},
47091 @samp{gssm}, and @samp{gsepla}.
47093 The @samp{org.gnu.gdb.s390.gsbc} feature is optional. It should contain
47094 the 64-bit wide guarded-storage broadcast control registers
47095 @samp{bc_gsd}, @samp{bc_gssm}, and @samp{bc_gsepla}.
47097 @node Sparc Features
47098 @subsection Sparc Features
47099 @cindex target descriptions, sparc32 features
47100 @cindex target descriptions, sparc64 features
47101 The @samp{org.gnu.gdb.sparc.cpu} feature is required for sparc32/sparc64
47102 targets. It should describe the following registers:
47106 @samp{g0} through @samp{g7}
47108 @samp{o0} through @samp{o7}
47110 @samp{l0} through @samp{l7}
47112 @samp{i0} through @samp{i7}
47115 They may be 32-bit or 64-bit depending on the target.
47117 Also the @samp{org.gnu.gdb.sparc.fpu} feature is required for sparc32/sparc64
47118 targets. It should describe the following registers:
47122 @samp{f0} through @samp{f31}
47124 @samp{f32} through @samp{f62} for sparc64
47127 The @samp{org.gnu.gdb.sparc.cp0} feature is required for sparc32/sparc64
47128 targets. It should describe the following registers:
47132 @samp{y}, @samp{psr}, @samp{wim}, @samp{tbr}, @samp{pc}, @samp{npc},
47133 @samp{fsr}, and @samp{csr} for sparc32
47135 @samp{pc}, @samp{npc}, @samp{state}, @samp{fsr}, @samp{fprs}, and @samp{y}
47139 @node TIC6x Features
47140 @subsection TMS320C6x Features
47141 @cindex target descriptions, TIC6x features
47142 @cindex target descriptions, TMS320C6x features
47143 The @samp{org.gnu.gdb.tic6x.core} feature is required for TMS320C6x
47144 targets. It should contain registers @samp{A0} through @samp{A15},
47145 registers @samp{B0} through @samp{B15}, @samp{CSR} and @samp{PC}.
47147 The @samp{org.gnu.gdb.tic6x.gp} feature is optional. It should
47148 contain registers @samp{A16} through @samp{A31} and @samp{B16}
47149 through @samp{B31}.
47151 The @samp{org.gnu.gdb.tic6x.c6xp} feature is optional. It should
47152 contain registers @samp{TSR}, @samp{ILC} and @samp{RILC}.
47154 @node Operating System Information
47155 @appendix Operating System Information
47156 @cindex operating system information
47162 Users of @value{GDBN} often wish to obtain information about the state of
47163 the operating system running on the target---for example the list of
47164 processes, or the list of open files. This section describes the
47165 mechanism that makes it possible. This mechanism is similar to the
47166 target features mechanism (@pxref{Target Descriptions}), but focuses
47167 on a different aspect of target.
47169 Operating system information is retrieved from the target via the
47170 remote protocol, using @samp{qXfer} requests (@pxref{qXfer osdata
47171 read}). The object name in the request should be @samp{osdata}, and
47172 the @var{annex} identifies the data to be fetched.
47175 @appendixsection Process list
47176 @cindex operating system information, process list
47178 When requesting the process list, the @var{annex} field in the
47179 @samp{qXfer} request should be @samp{processes}. The returned data is
47180 an XML document. The formal syntax of this document is defined in
47181 @file{gdb/features/osdata.dtd}.
47183 An example document is:
47186 <?xml version="1.0"?>
47187 <!DOCTYPE target SYSTEM "osdata.dtd">
47188 <osdata type="processes">
47190 <column name="pid">1</column>
47191 <column name="user">root</column>
47192 <column name="command">/sbin/init</column>
47193 <column name="cores">1,2,3</column>
47198 Each item should include a column whose name is @samp{pid}. The value
47199 of that column should identify the process on the target. The
47200 @samp{user} and @samp{command} columns are optional, and will be
47201 displayed by @value{GDBN}. The @samp{cores} column, if present,
47202 should contain a comma-separated list of cores that this process
47203 is running on. Target may provide additional columns,
47204 which @value{GDBN} currently ignores.
47206 @node Trace File Format
47207 @appendix Trace File Format
47208 @cindex trace file format
47210 The trace file comes in three parts: a header, a textual description
47211 section, and a trace frame section with binary data.
47213 The header has the form @code{\x7fTRACE0\n}. The first byte is
47214 @code{0x7f} so as to indicate that the file contains binary data,
47215 while the @code{0} is a version number that may have different values
47218 The description section consists of multiple lines of @sc{ascii} text
47219 separated by newline characters (@code{0xa}). The lines may include a
47220 variety of optional descriptive or context-setting information, such
47221 as tracepoint definitions or register set size. @value{GDBN} will
47222 ignore any line that it does not recognize. An empty line marks the end
47227 Specifies the size of a register block in bytes. This is equal to the
47228 size of a @code{g} packet payload in the remote protocol. @var{size}
47229 is an ascii decimal number. There should be only one such line in
47230 a single trace file.
47232 @item status @var{status}
47233 Trace status. @var{status} has the same format as a @code{qTStatus}
47234 remote packet reply. There should be only one such line in a single trace
47237 @item tp @var{payload}
47238 Tracepoint definition. The @var{payload} has the same format as
47239 @code{qTfP}/@code{qTsP} remote packet reply payload. A single tracepoint
47240 may take multiple lines of definition, corresponding to the multiple
47243 @item tsv @var{payload}
47244 Trace state variable definition. The @var{payload} has the same format as
47245 @code{qTfV}/@code{qTsV} remote packet reply payload. A single variable
47246 may take multiple lines of definition, corresponding to the multiple
47249 @item tdesc @var{payload}
47250 Target description in XML format. The @var{payload} is a single line of
47251 the XML file. All such lines should be concatenated together to get
47252 the original XML file. This file is in the same format as @code{qXfer}
47253 @code{features} payload, and corresponds to the main @code{target.xml}
47254 file. Includes are not allowed.
47258 The trace frame section consists of a number of consecutive frames.
47259 Each frame begins with a two-byte tracepoint number, followed by a
47260 four-byte size giving the amount of data in the frame. The data in
47261 the frame consists of a number of blocks, each introduced by a
47262 character indicating its type (at least register, memory, and trace
47263 state variable). The data in this section is raw binary, not a
47264 hexadecimal or other encoding; its endianness matches the target's
47267 @c FIXME bi-arch may require endianness/arch info in description section
47270 @item R @var{bytes}
47271 Register block. The number and ordering of bytes matches that of a
47272 @code{g} packet in the remote protocol. Note that these are the
47273 actual bytes, in target order, not a hexadecimal encoding.
47275 @item M @var{address} @var{length} @var{bytes}...
47276 Memory block. This is a contiguous block of memory, at the 8-byte
47277 address @var{address}, with a 2-byte length @var{length}, followed by
47278 @var{length} bytes.
47280 @item V @var{number} @var{value}
47281 Trace state variable block. This records the 8-byte signed value
47282 @var{value} of trace state variable numbered @var{number}.
47286 Future enhancements of the trace file format may include additional types
47289 @node Index Section Format
47290 @appendix @code{.gdb_index} section format
47291 @cindex .gdb_index section format
47292 @cindex index section format
47294 This section documents the index section that is created by @code{save
47295 gdb-index} (@pxref{Index Files}). The index section is
47296 DWARF-specific; some knowledge of DWARF is assumed in this
47299 The mapped index file format is designed to be directly
47300 @code{mmap}able on any architecture. In most cases, a datum is
47301 represented using a little-endian 32-bit integer value, called an
47302 @code{offset_type}. Big endian machines must byte-swap the values
47303 before using them. Exceptions to this rule are noted. The data is
47304 laid out such that alignment is always respected.
47306 A mapped index consists of several areas, laid out in order.
47310 The file header. This is a sequence of values, of @code{offset_type}
47311 unless otherwise noted:
47315 The version number, currently 8. Versions 1, 2 and 3 are obsolete.
47316 Version 4 uses a different hashing function from versions 5 and 6.
47317 Version 6 includes symbols for inlined functions, whereas versions 4
47318 and 5 do not. Version 7 adds attributes to the CU indices in the
47319 symbol table. Version 8 specifies that symbols from DWARF type units
47320 (@samp{DW_TAG_type_unit}) refer to the type unit's symbol table and not the
47321 compilation unit (@samp{DW_TAG_comp_unit}) using the type.
47323 @value{GDBN} will only read version 4, 5, or 6 indices
47324 by specifying @code{set use-deprecated-index-sections on}.
47325 GDB has a workaround for potentially broken version 7 indices so it is
47326 currently not flagged as deprecated.
47329 The offset, from the start of the file, of the CU list.
47332 The offset, from the start of the file, of the types CU list. Note
47333 that this area can be empty, in which case this offset will be equal
47334 to the next offset.
47337 The offset, from the start of the file, of the address area.
47340 The offset, from the start of the file, of the symbol table.
47343 The offset, from the start of the file, of the constant pool.
47347 The CU list. This is a sequence of pairs of 64-bit little-endian
47348 values, sorted by the CU offset. The first element in each pair is
47349 the offset of a CU in the @code{.debug_info} section. The second
47350 element in each pair is the length of that CU. References to a CU
47351 elsewhere in the map are done using a CU index, which is just the
47352 0-based index into this table. Note that if there are type CUs, then
47353 conceptually CUs and type CUs form a single list for the purposes of
47357 The types CU list. This is a sequence of triplets of 64-bit
47358 little-endian values. In a triplet, the first value is the CU offset,
47359 the second value is the type offset in the CU, and the third value is
47360 the type signature. The types CU list is not sorted.
47363 The address area. The address area consists of a sequence of address
47364 entries. Each address entry has three elements:
47368 The low address. This is a 64-bit little-endian value.
47371 The high address. This is a 64-bit little-endian value. Like
47372 @code{DW_AT_high_pc}, the value is one byte beyond the end.
47375 The CU index. This is an @code{offset_type} value.
47379 The symbol table. This is an open-addressed hash table. The size of
47380 the hash table is always a power of 2.
47382 Each slot in the hash table consists of a pair of @code{offset_type}
47383 values. The first value is the offset of the symbol's name in the
47384 constant pool. The second value is the offset of the CU vector in the
47387 If both values are 0, then this slot in the hash table is empty. This
47388 is ok because while 0 is a valid constant pool index, it cannot be a
47389 valid index for both a string and a CU vector.
47391 The hash value for a table entry is computed by applying an
47392 iterative hash function to the symbol's name. Starting with an
47393 initial value of @code{r = 0}, each (unsigned) character @samp{c} in
47394 the string is incorporated into the hash using the formula depending on the
47399 The formula is @code{r = r * 67 + c - 113}.
47401 @item Versions 5 to 7
47402 The formula is @code{r = r * 67 + tolower (c) - 113}.
47405 The terminating @samp{\0} is not incorporated into the hash.
47407 The step size used in the hash table is computed via
47408 @code{((hash * 17) & (size - 1)) | 1}, where @samp{hash} is the hash
47409 value, and @samp{size} is the size of the hash table. The step size
47410 is used to find the next candidate slot when handling a hash
47413 The names of C@t{++} symbols in the hash table are canonicalized. We
47414 don't currently have a simple description of the canonicalization
47415 algorithm; if you intend to create new index sections, you must read
47419 The constant pool. This is simply a bunch of bytes. It is organized
47420 so that alignment is correct: CU vectors are stored first, followed by
47423 A CU vector in the constant pool is a sequence of @code{offset_type}
47424 values. The first value is the number of CU indices in the vector.
47425 Each subsequent value is the index and symbol attributes of a CU in
47426 the CU list. This element in the hash table is used to indicate which
47427 CUs define the symbol and how the symbol is used.
47428 See below for the format of each CU index+attributes entry.
47430 A string in the constant pool is zero-terminated.
47433 Attributes were added to CU index values in @code{.gdb_index} version 7.
47434 If a symbol has multiple uses within a CU then there is one
47435 CU index+attributes value for each use.
47437 The format of each CU index+attributes entry is as follows
47443 This is the index of the CU in the CU list.
47445 These bits are reserved for future purposes and must be zero.
47447 The kind of the symbol in the CU.
47451 This value is reserved and should not be used.
47452 By reserving zero the full @code{offset_type} value is backwards compatible
47453 with previous versions of the index.
47455 The symbol is a type.
47457 The symbol is a variable or an enum value.
47459 The symbol is a function.
47461 Any other kind of symbol.
47463 These values are reserved.
47467 This bit is zero if the value is global and one if it is static.
47469 The determination of whether a symbol is global or static is complicated.
47470 The authorative reference is the file @file{dwarf2read.c} in
47471 @value{GDBN} sources.
47475 This pseudo-code describes the computation of a symbol's kind and
47476 global/static attributes in the index.
47479 is_external = get_attribute (die, DW_AT_external);
47480 language = get_attribute (cu_die, DW_AT_language);
47483 case DW_TAG_typedef:
47484 case DW_TAG_base_type:
47485 case DW_TAG_subrange_type:
47489 case DW_TAG_enumerator:
47491 is_static = language != CPLUS;
47493 case DW_TAG_subprogram:
47495 is_static = ! (is_external || language == ADA);
47497 case DW_TAG_constant:
47499 is_static = ! is_external;
47501 case DW_TAG_variable:
47503 is_static = ! is_external;
47505 case DW_TAG_namespace:
47509 case DW_TAG_class_type:
47510 case DW_TAG_interface_type:
47511 case DW_TAG_structure_type:
47512 case DW_TAG_union_type:
47513 case DW_TAG_enumeration_type:
47515 is_static = language != CPLUS;
47523 @appendix Manual pages
47527 * gdb man:: The GNU Debugger man page
47528 * gdbserver man:: Remote Server for the GNU Debugger man page
47529 * gcore man:: Generate a core file of a running program
47530 * gdbinit man:: gdbinit scripts
47531 * gdb-add-index man:: Add index files to speed up GDB
47537 @c man title gdb The GNU Debugger
47539 @c man begin SYNOPSIS gdb
47540 gdb [@option{-help}] [@option{-nh}] [@option{-nx}] [@option{-q}]
47541 [@option{-batch}] [@option{-cd=}@var{dir}] [@option{-f}]
47542 [@option{-b}@w{ }@var{bps}]
47543 [@option{-tty=}@var{dev}] [@option{-s} @var{symfile}]
47544 [@option{-e}@w{ }@var{prog}] [@option{-se}@w{ }@var{prog}]
47545 [@option{-c}@w{ }@var{core}] [@option{-p}@w{ }@var{procID}]
47546 [@option{-x}@w{ }@var{cmds}] [@option{-d}@w{ }@var{dir}]
47547 [@var{prog}|@var{prog} @var{procID}|@var{prog} @var{core}]
47550 @c man begin DESCRIPTION gdb
47551 The purpose of a debugger such as @value{GDBN} is to allow you to see what is
47552 going on ``inside'' another program while it executes -- or what another
47553 program was doing at the moment it crashed.
47555 @value{GDBN} can do four main kinds of things (plus other things in support of
47556 these) to help you catch bugs in the act:
47560 Start your program, specifying anything that might affect its behavior.
47563 Make your program stop on specified conditions.
47566 Examine what has happened, when your program has stopped.
47569 Change things in your program, so you can experiment with correcting the
47570 effects of one bug and go on to learn about another.
47573 You can use @value{GDBN} to debug programs written in C, C@t{++}, Fortran and
47576 @value{GDBN} is invoked with the shell command @code{gdb}. Once started, it reads
47577 commands from the terminal until you tell it to exit with the @value{GDBN}
47578 command @code{quit}. You can get online help from @value{GDBN} itself
47579 by using the command @code{help}.
47581 You can run @code{gdb} with no arguments or options; but the most
47582 usual way to start @value{GDBN} is with one argument or two, specifying an
47583 executable program as the argument:
47589 You can also start with both an executable program and a core file specified:
47595 You can, instead, specify a process ID as a second argument or use option
47596 @code{-p}, if you want to debug a running process:
47604 would attach @value{GDBN} to process @code{1234}. With option @option{-p} you
47605 can omit the @var{program} filename.
47607 Here are some of the most frequently needed @value{GDBN} commands:
47609 @c pod2man highlights the right hand side of the @item lines.
47611 @item break [@var{file}:]@var{function}
47612 Set a breakpoint at @var{function} (in @var{file}).
47614 @item run [@var{arglist}]
47615 Start your program (with @var{arglist}, if specified).
47618 Backtrace: display the program stack.
47620 @item print @var{expr}
47621 Display the value of an expression.
47624 Continue running your program (after stopping, e.g. at a breakpoint).
47627 Execute next program line (after stopping); step @emph{over} any
47628 function calls in the line.
47630 @item edit [@var{file}:]@var{function}
47631 look at the program line where it is presently stopped.
47633 @item list [@var{file}:]@var{function}
47634 type the text of the program in the vicinity of where it is presently stopped.
47637 Execute next program line (after stopping); step @emph{into} any
47638 function calls in the line.
47640 @item help [@var{name}]
47641 Show information about @value{GDBN} command @var{name}, or general information
47642 about using @value{GDBN}.
47645 Exit from @value{GDBN}.
47649 For full details on @value{GDBN},
47650 see @cite{Using GDB: A Guide to the GNU Source-Level Debugger},
47651 by Richard M. Stallman and Roland H. Pesch. The same text is available online
47652 as the @code{gdb} entry in the @code{info} program.
47656 @c man begin OPTIONS gdb
47657 Any arguments other than options specify an executable
47658 file and core file (or process ID); that is, the first argument
47659 encountered with no
47660 associated option flag is equivalent to a @option{-se} option, and the second,
47661 if any, is equivalent to a @option{-c} option if it's the name of a file.
47663 both long and short forms; both are shown here. The long forms are also
47664 recognized if you truncate them, so long as enough of the option is
47665 present to be unambiguous. (If you prefer, you can flag option
47666 arguments with @option{+} rather than @option{-}, though we illustrate the
47667 more usual convention.)
47669 All the options and command line arguments you give are processed
47670 in sequential order. The order makes a difference when the @option{-x}
47676 List all options, with brief explanations.
47678 @item -symbols=@var{file}
47679 @itemx -s @var{file}
47680 Read symbol table from file @var{file}.
47683 Enable writing into executable and core files.
47685 @item -exec=@var{file}
47686 @itemx -e @var{file}
47687 Use file @var{file} as the executable file to execute when
47688 appropriate, and for examining pure data in conjunction with a core
47691 @item -se=@var{file}
47692 Read symbol table from file @var{file} and use it as the executable
47695 @item -core=@var{file}
47696 @itemx -c @var{file}
47697 Use file @var{file} as a core dump to examine.
47699 @item -command=@var{file}
47700 @itemx -x @var{file}
47701 Execute @value{GDBN} commands from file @var{file}.
47703 @item -ex @var{command}
47704 Execute given @value{GDBN} @var{command}.
47706 @item -directory=@var{directory}
47707 @itemx -d @var{directory}
47708 Add @var{directory} to the path to search for source files.
47711 Do not execute commands from @file{~/.gdbinit}.
47715 Do not execute commands from any @file{.gdbinit} initialization files.
47719 ``Quiet''. Do not print the introductory and copyright messages. These
47720 messages are also suppressed in batch mode.
47723 Run in batch mode. Exit with status @code{0} after processing all the command
47724 files specified with @option{-x} (and @file{.gdbinit}, if not inhibited).
47725 Exit with nonzero status if an error occurs in executing the @value{GDBN}
47726 commands in the command files.
47728 Batch mode may be useful for running @value{GDBN} as a filter, for example to
47729 download and run a program on another computer; in order to make this
47730 more useful, the message
47733 Program exited normally.
47737 (which is ordinarily issued whenever a program running under @value{GDBN} control
47738 terminates) is not issued when running in batch mode.
47740 @item -cd=@var{directory}
47741 Run @value{GDBN} using @var{directory} as its working directory,
47742 instead of the current directory.
47746 Emacs sets this option when it runs @value{GDBN} as a subprocess. It tells
47747 @value{GDBN} to output the full file name and line number in a standard,
47748 recognizable fashion each time a stack frame is displayed (which
47749 includes each time the program stops). This recognizable format looks
47750 like two @samp{\032} characters, followed by the file name, line number
47751 and character position separated by colons, and a newline. The
47752 Emacs-to-@value{GDBN} interface program uses the two @samp{\032}
47753 characters as a signal to display the source code for the frame.
47756 Set the line speed (baud rate or bits per second) of any serial
47757 interface used by @value{GDBN} for remote debugging.
47759 @item -tty=@var{device}
47760 Run using @var{device} for your program's standard input and output.
47764 @c man begin SEEALSO gdb
47766 The full documentation for @value{GDBN} is maintained as a Texinfo manual.
47767 If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
47768 documentation are properly installed at your site, the command
47775 should give you access to the complete manual.
47777 @cite{Using GDB: A Guide to the GNU Source-Level Debugger},
47778 Richard M. Stallman and Roland H. Pesch, July 1991.
47782 @node gdbserver man
47783 @heading gdbserver man
47785 @c man title gdbserver Remote Server for the GNU Debugger
47787 @c man begin SYNOPSIS gdbserver
47788 gdbserver @var{comm} @var{prog} [@var{args}@dots{}]
47790 gdbserver --attach @var{comm} @var{pid}
47792 gdbserver --multi @var{comm}
47796 @c man begin DESCRIPTION gdbserver
47797 @command{gdbserver} is a program that allows you to run @value{GDBN} on a different machine
47798 than the one which is running the program being debugged.
47801 @subheading Usage (server (target) side)
47804 Usage (server (target) side):
47807 First, you need to have a copy of the program you want to debug put onto
47808 the target system. The program can be stripped to save space if needed, as
47809 @command{gdbserver} doesn't care about symbols. All symbol handling is taken care of by
47810 the @value{GDBN} running on the host system.
47812 To use the server, you log on to the target system, and run the @command{gdbserver}
47813 program. You must tell it (a) how to communicate with @value{GDBN}, (b) the name of
47814 your program, and (c) its arguments. The general syntax is:
47817 target> gdbserver @var{comm} @var{program} [@var{args} ...]
47820 For example, using a serial port, you might say:
47824 @c @file would wrap it as F</dev/com1>.
47825 target> gdbserver /dev/com1 emacs foo.txt
47828 target> gdbserver @file{/dev/com1} emacs foo.txt
47832 This tells @command{gdbserver} to debug emacs with an argument of foo.txt, and
47833 to communicate with @value{GDBN} via @file{/dev/com1}. @command{gdbserver} now
47834 waits patiently for the host @value{GDBN} to communicate with it.
47836 To use a TCP connection, you could say:
47839 target> gdbserver host:2345 emacs foo.txt
47842 This says pretty much the same thing as the last example, except that we are
47843 going to communicate with the @code{host} @value{GDBN} via TCP. The @code{host:2345} argument means
47844 that we are expecting to see a TCP connection from @code{host} to local TCP port
47845 2345. (Currently, the @code{host} part is ignored.) You can choose any number you
47846 want for the port number as long as it does not conflict with any existing TCP
47847 ports on the target system. This same port number must be used in the host
47848 @value{GDBN}s @code{target remote} command, which will be described shortly. Note that if
47849 you chose a port number that conflicts with another service, @command{gdbserver} will
47850 print an error message and exit.
47852 @command{gdbserver} can also attach to running programs.
47853 This is accomplished via the @option{--attach} argument. The syntax is:
47856 target> gdbserver --attach @var{comm} @var{pid}
47859 @var{pid} is the process ID of a currently running process. It isn't
47860 necessary to point @command{gdbserver} at a binary for the running process.
47862 To start @code{gdbserver} without supplying an initial command to run
47863 or process ID to attach, use the @option{--multi} command line option.
47864 In such case you should connect using @kbd{target extended-remote} to start
47865 the program you want to debug.
47868 target> gdbserver --multi @var{comm}
47872 @subheading Usage (host side)
47878 You need an unstripped copy of the target program on your host system, since
47879 @value{GDBN} needs to examine its symbol tables and such. Start up @value{GDBN} as you normally
47880 would, with the target program as the first argument. (You may need to use the
47881 @option{--baud} option if the serial line is running at anything except 9600 baud.)
47882 That is @code{gdb TARGET-PROG}, or @code{gdb --baud BAUD TARGET-PROG}. After that, the only
47883 new command you need to know about is @code{target remote}
47884 (or @code{target extended-remote}). Its argument is either
47885 a device name (usually a serial device, like @file{/dev/ttyb}), or a @code{HOST:PORT}
47886 descriptor. For example:
47890 @c @file would wrap it as F</dev/ttyb>.
47891 (@value{GDBP}) target remote /dev/ttyb
47894 (@value{GDBP}) target remote @file{/dev/ttyb}
47899 communicates with the server via serial line @file{/dev/ttyb}, and:
47902 (@value{GDBP}) target remote the-target:2345
47906 communicates via a TCP connection to port 2345 on host `the-target', where
47907 you previously started up @command{gdbserver} with the same port number. Note that for
47908 TCP connections, you must start up @command{gdbserver} prior to using the `target remote'
47909 command, otherwise you may get an error that looks something like
47910 `Connection refused'.
47912 @command{gdbserver} can also debug multiple inferiors at once,
47915 the @value{GDBN} manual in node @code{Inferiors and Programs}
47916 -- shell command @code{info -f gdb -n 'Inferiors and Programs'}.
47919 @ref{Inferiors and Programs}.
47921 In such case use the @code{extended-remote} @value{GDBN} command variant:
47924 (@value{GDBP}) target extended-remote the-target:2345
47927 The @command{gdbserver} option @option{--multi} may or may not be used in such
47931 @c man begin OPTIONS gdbserver
47932 There are three different modes for invoking @command{gdbserver}:
47937 Debug a specific program specified by its program name:
47940 gdbserver @var{comm} @var{prog} [@var{args}@dots{}]
47943 The @var{comm} parameter specifies how should the server communicate
47944 with @value{GDBN}; it is either a device name (to use a serial line),
47945 a TCP port number (@code{:1234}), or @code{-} or @code{stdio} to use
47946 stdin/stdout of @code{gdbserver}. Specify the name of the program to
47947 debug in @var{prog}. Any remaining arguments will be passed to the
47948 program verbatim. When the program exits, @value{GDBN} will close the
47949 connection, and @code{gdbserver} will exit.
47952 Debug a specific program by specifying the process ID of a running
47956 gdbserver --attach @var{comm} @var{pid}
47959 The @var{comm} parameter is as described above. Supply the process ID
47960 of a running program in @var{pid}; @value{GDBN} will do everything
47961 else. Like with the previous mode, when the process @var{pid} exits,
47962 @value{GDBN} will close the connection, and @code{gdbserver} will exit.
47965 Multi-process mode -- debug more than one program/process:
47968 gdbserver --multi @var{comm}
47971 In this mode, @value{GDBN} can instruct @command{gdbserver} which
47972 command(s) to run. Unlike the other 2 modes, @value{GDBN} will not
47973 close the connection when a process being debugged exits, so you can
47974 debug several processes in the same session.
47977 In each of the modes you may specify these options:
47982 List all options, with brief explanations.
47985 This option causes @command{gdbserver} to print its version number and exit.
47988 @command{gdbserver} will attach to a running program. The syntax is:
47991 target> gdbserver --attach @var{comm} @var{pid}
47994 @var{pid} is the process ID of a currently running process. It isn't
47995 necessary to point @command{gdbserver} at a binary for the running process.
47998 To start @code{gdbserver} without supplying an initial command to run
47999 or process ID to attach, use this command line option.
48000 Then you can connect using @kbd{target extended-remote} and start
48001 the program you want to debug. The syntax is:
48004 target> gdbserver --multi @var{comm}
48008 Instruct @code{gdbserver} to display extra status information about the debugging
48010 This option is intended for @code{gdbserver} development and for bug reports to
48013 @item --remote-debug
48014 Instruct @code{gdbserver} to display remote protocol debug output.
48015 This option is intended for @code{gdbserver} development and for bug reports to
48018 @item --debug-file=@var{filename}
48019 Instruct @code{gdbserver} to send any debug output to the given @var{filename}.
48020 This option is intended for @code{gdbserver} development and for bug reports to
48023 @item --debug-format=option1@r{[},option2,...@r{]}
48024 Instruct @code{gdbserver} to include extra information in each line
48025 of debugging output.
48026 @xref{Other Command-Line Arguments for gdbserver}.
48029 Specify a wrapper to launch programs
48030 for debugging. The option should be followed by the name of the
48031 wrapper, then any command-line arguments to pass to the wrapper, then
48032 @kbd{--} indicating the end of the wrapper arguments.
48035 By default, @command{gdbserver} keeps the listening TCP port open, so that
48036 additional connections are possible. However, if you start @code{gdbserver}
48037 with the @option{--once} option, it will stop listening for any further
48038 connection attempts after connecting to the first @value{GDBN} session.
48040 @c --disable-packet is not documented for users.
48042 @c --disable-randomization and --no-disable-randomization are superseded by
48043 @c QDisableRandomization.
48048 @c man begin SEEALSO gdbserver
48050 The full documentation for @value{GDBN} is maintained as a Texinfo manual.
48051 If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
48052 documentation are properly installed at your site, the command
48058 should give you access to the complete manual.
48060 @cite{Using GDB: A Guide to the GNU Source-Level Debugger},
48061 Richard M. Stallman and Roland H. Pesch, July 1991.
48068 @c man title gcore Generate a core file of a running program
48071 @c man begin SYNOPSIS gcore
48072 gcore [-a] [-o @var{prefix}] @var{pid1} [@var{pid2}...@var{pidN}]
48076 @c man begin DESCRIPTION gcore
48077 Generate core dumps of one or more running programs with process IDs
48078 @var{pid1}, @var{pid2}, etc. A core file produced by @command{gcore}
48079 is equivalent to one produced by the kernel when the process crashes
48080 (and when @kbd{ulimit -c} was used to set up an appropriate core dump
48081 limit). However, unlike after a crash, after @command{gcore} finishes
48082 its job the program remains running without any change.
48085 @c man begin OPTIONS gcore
48088 Dump all memory mappings. The actual effect of this option depends on
48089 the Operating System. On @sc{gnu}/Linux, it will disable
48090 @code{use-coredump-filter} (@pxref{set use-coredump-filter}) and
48091 enable @code{dump-excluded-mappings} (@pxref{set
48092 dump-excluded-mappings}).
48094 @item -o @var{prefix}
48095 The optional argument @var{prefix} specifies the prefix to be used
48096 when composing the file names of the core dumps. The file name is
48097 composed as @file{@var{prefix}.@var{pid}}, where @var{pid} is the
48098 process ID of the running program being analyzed by @command{gcore}.
48099 If not specified, @var{prefix} defaults to @var{gcore}.
48103 @c man begin SEEALSO gcore
48105 The full documentation for @value{GDBN} is maintained as a Texinfo manual.
48106 If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
48107 documentation are properly installed at your site, the command
48114 should give you access to the complete manual.
48116 @cite{Using GDB: A Guide to the GNU Source-Level Debugger},
48117 Richard M. Stallman and Roland H. Pesch, July 1991.
48124 @c man title gdbinit GDB initialization scripts
48127 @c man begin SYNOPSIS gdbinit
48128 @ifset SYSTEM_GDBINIT
48129 @value{SYSTEM_GDBINIT}
48132 @ifset SYSTEM_GDBINIT_DIR
48133 @value{SYSTEM_GDBINIT_DIR}/*
48142 @c man begin DESCRIPTION gdbinit
48143 These files contain @value{GDBN} commands to automatically execute during
48144 @value{GDBN} startup. The lines of contents are canned sequences of commands,
48147 the @value{GDBN} manual in node @code{Sequences}
48148 -- shell command @code{info -f gdb -n Sequences}.
48154 Please read more in
48156 the @value{GDBN} manual in node @code{Startup}
48157 -- shell command @code{info -f gdb -n Startup}.
48164 @ifset SYSTEM_GDBINIT
48165 @item @value{SYSTEM_GDBINIT}
48167 @ifclear SYSTEM_GDBINIT
48168 @item (not enabled with @code{--with-system-gdbinit} during compilation)
48170 System-wide initialization file. It is executed unless user specified
48171 @value{GDBN} option @code{-nx} or @code{-n}.
48174 the @value{GDBN} manual in node @code{System-wide configuration}
48175 -- shell command @code{info -f gdb -n 'System-wide configuration'}.
48177 @ifset SYSTEM_GDBINIT_DIR
48178 @item @value{SYSTEM_GDBINIT_DIR}
48180 @ifclear SYSTEM_GDBINIT_DIR
48181 @item (not enabled with @code{--with-system-gdbinit-dir} during compilation)
48183 System-wide initialization directory. All files in this directory are
48184 executed on startup unless user specified @value{GDBN} option @code{-nx} or
48185 @code{-n}, as long as they have a recognized file extension.
48188 the @value{GDBN} manual in node @code{System-wide configuration}
48189 -- shell command @code{info -f gdb -n 'System-wide configuration'}.
48192 @ref{System-wide configuration}.
48196 User initialization file. It is executed unless user specified
48197 @value{GDBN} options @code{-nx}, @code{-n} or @code{-nh}.
48200 Initialization file for current directory. It may need to be enabled with
48201 @value{GDBN} security command @code{set auto-load local-gdbinit}.
48204 the @value{GDBN} manual in node @code{Init File in the Current Directory}
48205 -- shell command @code{info -f gdb -n 'Init File in the Current Directory'}.
48208 @ref{Init File in the Current Directory}.
48213 @c man begin SEEALSO gdbinit
48215 gdb(1), @code{info -f gdb -n Startup}
48217 The full documentation for @value{GDBN} is maintained as a Texinfo manual.
48218 If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
48219 documentation are properly installed at your site, the command
48225 should give you access to the complete manual.
48227 @cite{Using GDB: A Guide to the GNU Source-Level Debugger},
48228 Richard M. Stallman and Roland H. Pesch, July 1991.
48232 @node gdb-add-index man
48233 @heading gdb-add-index
48234 @pindex gdb-add-index
48235 @anchor{gdb-add-index}
48237 @c man title gdb-add-index Add index files to speed up GDB
48239 @c man begin SYNOPSIS gdb-add-index
48240 gdb-add-index @var{filename}
48243 @c man begin DESCRIPTION gdb-add-index
48244 When @value{GDBN} finds a symbol file, it scans the symbols in the
48245 file in order to construct an internal symbol table. This lets most
48246 @value{GDBN} operations work quickly--at the cost of a delay early on.
48247 For large programs, this delay can be quite lengthy, so @value{GDBN}
48248 provides a way to build an index, which speeds up startup.
48250 To determine whether a file contains such an index, use the command
48251 @kbd{readelf -S filename}: the index is stored in a section named
48252 @code{.gdb_index}. The index file can only be produced on systems
48253 which use ELF binaries and DWARF debug information (i.e., sections
48254 named @code{.debug_*}).
48256 @command{gdb-add-index} uses @value{GDBN} and @command{objdump} found
48257 in the @env{PATH} environment variable. If you want to use different
48258 versions of these programs, you can specify them through the
48259 @env{GDB} and @env{OBJDUMP} environment variables.
48263 the @value{GDBN} manual in node @code{Index Files}
48264 -- shell command @kbd{info -f gdb -n "Index Files"}.
48271 @c man begin SEEALSO gdb-add-index
48273 The full documentation for @value{GDBN} is maintained as a Texinfo manual.
48274 If the @code{info} and @code{gdb} programs and @value{GDBN}'s Texinfo
48275 documentation are properly installed at your site, the command
48281 should give you access to the complete manual.
48283 @cite{Using GDB: A Guide to the GNU Source-Level Debugger},
48284 Richard M. Stallman and Roland H. Pesch, July 1991.
48290 @node GNU Free Documentation License
48291 @appendix GNU Free Documentation License
48294 @node Concept Index
48295 @unnumbered Concept Index
48299 @node Command and Variable Index
48300 @unnumbered Command, Variable, and Function Index
48305 % I think something like @@colophon should be in texinfo. In the
48307 \long\def\colophon{\hbox to0pt{}\vfill
48308 \centerline{The body of this manual is set in}
48309 \centerline{\fontname\tenrm,}
48310 \centerline{with headings in {\bf\fontname\tenbf}}
48311 \centerline{and examples in {\tt\fontname\tentt}.}
48312 \centerline{{\it\fontname\tenit\/},}
48313 \centerline{{\bf\fontname\tenbf}, and}
48314 \centerline{{\sl\fontname\tensl\/}}
48315 \centerline{are used for emphasis.}\vfill}
48317 % Blame: doc@@cygnus.com, 1991.