1 \input texinfo @c -*- texinfo -*-
2 @setfilename gdbint.info
4 @dircategory Software development
6 * Gdb-Internals: (gdbint). The GNU debugger's internals.
10 This file documents the internals of the GNU debugger @value{GDBN}.
11 Copyright 1990,1991,1992,1993,1994,1996,1998,1999,2000,2001,2002,2003,2004
12 Free Software Foundation, Inc.
13 Contributed by Cygnus Solutions. Written by John Gilmore.
14 Second Edition by Stan Shebs.
16 Permission is granted to copy, distribute and/or modify this document
17 under the terms of the GNU Free Documentation License, Version 1.1 or
18 any later version published by the Free Software Foundation; with no
19 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
20 Texts. A copy of the license is included in the section entitled ``GNU
21 Free Documentation License''.
24 @setchapternewpage off
25 @settitle @value{GDBN} Internals
31 @title @value{GDBN} Internals
32 @subtitle{A guide to the internals of the GNU debugger}
34 @author Cygnus Solutions
35 @author Second Edition:
37 @author Cygnus Solutions
40 \def\$#1${{#1}} % Kluge: collect RCS revision info without $...$
41 \xdef\manvers{\$Revision$} % For use in headers, footers too
43 \hfill Cygnus Solutions\par
45 \hfill \TeX{}info \texinfoversion\par
49 @vskip 0pt plus 1filll
50 Copyright @copyright{} 1990,1991,1992,1993,1994,1996,1998,1999,2000,2001,
51 2002, 2003, 2004 Free Software Foundation, Inc.
53 Permission is granted to copy, distribute and/or modify this document
54 under the terms of the GNU Free Documentation License, Version 1.1 or
55 any later version published by the Free Software Foundation; with no
56 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
57 Texts. A copy of the license is included in the section entitled ``GNU
58 Free Documentation License''.
64 @c Perhaps this should be the title of the document (but only for info,
65 @c not for TeX). Existing GNU manuals seem inconsistent on this point.
66 @top Scope of this Document
68 This document documents the internals of the GNU debugger, @value{GDBN}. It
69 includes description of @value{GDBN}'s key algorithms and operations, as well
70 as the mechanisms that adapt @value{GDBN} to specific hosts and targets.
81 * Target Architecture Definition::
82 * Target Vector Definition::
91 * GDB Observers:: @value{GDBN} Currently available observers
92 * GNU Free Documentation License:: The license for this documentation
99 @cindex requirements for @value{GDBN}
101 Before diving into the internals, you should understand the formal
102 requirements and other expectations for @value{GDBN}. Although some
103 of these may seem obvious, there have been proposals for @value{GDBN}
104 that have run counter to these requirements.
106 First of all, @value{GDBN} is a debugger. It's not designed to be a
107 front panel for embedded systems. It's not a text editor. It's not a
108 shell. It's not a programming environment.
110 @value{GDBN} is an interactive tool. Although a batch mode is
111 available, @value{GDBN}'s primary role is to interact with a human
114 @value{GDBN} should be responsive to the user. A programmer hot on
115 the trail of a nasty bug, and operating under a looming deadline, is
116 going to be very impatient of everything, including the response time
117 to debugger commands.
119 @value{GDBN} should be relatively permissive, such as for expressions.
120 While the compiler should be picky (or have the option to be made
121 picky), since source code lives for a long time usually, the
122 programmer doing debugging shouldn't be spending time figuring out to
123 mollify the debugger.
125 @value{GDBN} will be called upon to deal with really large programs.
126 Executable sizes of 50 to 100 megabytes occur regularly, and we've
127 heard reports of programs approaching 1 gigabyte in size.
129 @value{GDBN} should be able to run everywhere. No other debugger is
130 available for even half as many configurations as @value{GDBN}
134 @node Overall Structure
136 @chapter Overall Structure
138 @value{GDBN} consists of three major subsystems: user interface,
139 symbol handling (the @dfn{symbol side}), and target system handling (the
142 The user interface consists of several actual interfaces, plus
145 The symbol side consists of object file readers, debugging info
146 interpreters, symbol table management, source language expression
147 parsing, type and value printing.
149 The target side consists of execution control, stack frame analysis, and
150 physical target manipulation.
152 The target side/symbol side division is not formal, and there are a
153 number of exceptions. For instance, core file support involves symbolic
154 elements (the basic core file reader is in BFD) and target elements (it
155 supplies the contents of memory and the values of registers). Instead,
156 this division is useful for understanding how the minor subsystems
159 @section The Symbol Side
161 The symbolic side of @value{GDBN} can be thought of as ``everything
162 you can do in @value{GDBN} without having a live program running''.
163 For instance, you can look at the types of variables, and evaluate
164 many kinds of expressions.
166 @section The Target Side
168 The target side of @value{GDBN} is the ``bits and bytes manipulator''.
169 Although it may make reference to symbolic info here and there, most
170 of the target side will run with only a stripped executable
171 available---or even no executable at all, in remote debugging cases.
173 Operations such as disassembly, stack frame crawls, and register
174 display, are able to work with no symbolic info at all. In some cases,
175 such as disassembly, @value{GDBN} will use symbolic info to present addresses
176 relative to symbols rather than as raw numbers, but it will work either
179 @section Configurations
183 @dfn{Host} refers to attributes of the system where @value{GDBN} runs.
184 @dfn{Target} refers to the system where the program being debugged
185 executes. In most cases they are the same machine, in which case a
186 third type of @dfn{Native} attributes come into play.
188 Defines and include files needed to build on the host are host support.
189 Examples are tty support, system defined types, host byte order, host
192 Defines and information needed to handle the target format are target
193 dependent. Examples are the stack frame format, instruction set,
194 breakpoint instruction, registers, and how to set up and tear down the stack
197 Information that is only needed when the host and target are the same,
198 is native dependent. One example is Unix child process support; if the
199 host and target are not the same, doing a fork to start the target
200 process is a bad idea. The various macros needed for finding the
201 registers in the @code{upage}, running @code{ptrace}, and such are all
202 in the native-dependent files.
204 Another example of native-dependent code is support for features that
205 are really part of the target environment, but which require
206 @code{#include} files that are only available on the host system. Core
207 file handling and @code{setjmp} handling are two common cases.
209 When you want to make @value{GDBN} work ``native'' on a particular machine, you
210 have to include all three kinds of information.
218 @value{GDBN} uses a number of debugging-specific algorithms. They are
219 often not very complicated, but get lost in the thicket of special
220 cases and real-world issues. This chapter describes the basic
221 algorithms and mentions some of the specific target definitions that
227 @cindex call stack frame
228 A frame is a construct that @value{GDBN} uses to keep track of calling
229 and called functions.
231 @findex create_new_frame
233 @code{FRAME_FP} in the machine description has no meaning to the
234 machine-independent part of @value{GDBN}, except that it is used when
235 setting up a new frame from scratch, as follows:
238 create_new_frame (read_register (DEPRECATED_FP_REGNUM), read_pc ()));
241 @cindex frame pointer register
242 Other than that, all the meaning imparted to @code{DEPRECATED_FP_REGNUM}
243 is imparted by the machine-dependent code. So,
244 @code{DEPRECATED_FP_REGNUM} can have any value that is convenient for
245 the code that creates new frames. (@code{create_new_frame} calls
246 @code{DEPRECATED_INIT_EXTRA_FRAME_INFO} if it is defined; that is where
247 you should use the @code{DEPRECATED_FP_REGNUM} value, if your frames are
251 Given a @value{GDBN} frame, define @code{DEPRECATED_FRAME_CHAIN} to
252 determine the address of the calling function's frame. This will be
253 used to create a new @value{GDBN} frame struct, and then
254 @code{DEPRECATED_INIT_EXTRA_FRAME_INFO} and
255 @code{DEPRECATED_INIT_FRAME_PC} will be called for the new frame.
257 @section Breakpoint Handling
260 In general, a breakpoint is a user-designated location in the program
261 where the user wants to regain control if program execution ever reaches
264 There are two main ways to implement breakpoints; either as ``hardware''
265 breakpoints or as ``software'' breakpoints.
267 @cindex hardware breakpoints
268 @cindex program counter
269 Hardware breakpoints are sometimes available as a builtin debugging
270 features with some chips. Typically these work by having dedicated
271 register into which the breakpoint address may be stored. If the PC
272 (shorthand for @dfn{program counter})
273 ever matches a value in a breakpoint registers, the CPU raises an
274 exception and reports it to @value{GDBN}.
276 Another possibility is when an emulator is in use; many emulators
277 include circuitry that watches the address lines coming out from the
278 processor, and force it to stop if the address matches a breakpoint's
281 A third possibility is that the target already has the ability to do
282 breakpoints somehow; for instance, a ROM monitor may do its own
283 software breakpoints. So although these are not literally ``hardware
284 breakpoints'', from @value{GDBN}'s point of view they work the same;
285 @value{GDBN} need not do anything more than set the breakpoint and wait
286 for something to happen.
288 Since they depend on hardware resources, hardware breakpoints may be
289 limited in number; when the user asks for more, @value{GDBN} will
290 start trying to set software breakpoints. (On some architectures,
291 notably the 32-bit x86 platforms, @value{GDBN} cannot always know
292 whether there's enough hardware resources to insert all the hardware
293 breakpoints and watchpoints. On those platforms, @value{GDBN} prints
294 an error message only when the program being debugged is continued.)
296 @cindex software breakpoints
297 Software breakpoints require @value{GDBN} to do somewhat more work.
298 The basic theory is that @value{GDBN} will replace a program
299 instruction with a trap, illegal divide, or some other instruction
300 that will cause an exception, and then when it's encountered,
301 @value{GDBN} will take the exception and stop the program. When the
302 user says to continue, @value{GDBN} will restore the original
303 instruction, single-step, re-insert the trap, and continue on.
305 Since it literally overwrites the program being tested, the program area
306 must be writable, so this technique won't work on programs in ROM. It
307 can also distort the behavior of programs that examine themselves,
308 although such a situation would be highly unusual.
310 Also, the software breakpoint instruction should be the smallest size of
311 instruction, so it doesn't overwrite an instruction that might be a jump
312 target, and cause disaster when the program jumps into the middle of the
313 breakpoint instruction. (Strictly speaking, the breakpoint must be no
314 larger than the smallest interval between instructions that may be jump
315 targets; perhaps there is an architecture where only even-numbered
316 instructions may jumped to.) Note that it's possible for an instruction
317 set not to have any instructions usable for a software breakpoint,
318 although in practice only the ARC has failed to define such an
322 The basic definition of the software breakpoint is the macro
325 Basic breakpoint object handling is in @file{breakpoint.c}. However,
326 much of the interesting breakpoint action is in @file{infrun.c}.
328 @section Single Stepping
330 @section Signal Handling
332 @section Thread Handling
334 @section Inferior Function Calls
336 @section Longjmp Support
338 @cindex @code{longjmp} debugging
339 @value{GDBN} has support for figuring out that the target is doing a
340 @code{longjmp} and for stopping at the target of the jump, if we are
341 stepping. This is done with a few specialized internal breakpoints,
342 which are visible in the output of the @samp{maint info breakpoint}
345 @findex GET_LONGJMP_TARGET
346 To make this work, you need to define a macro called
347 @code{GET_LONGJMP_TARGET}, which will examine the @code{jmp_buf}
348 structure and extract the longjmp target address. Since @code{jmp_buf}
349 is target specific, you will need to define it in the appropriate
350 @file{tm-@var{target}.h} file. Look in @file{tm-sun4os4.h} and
351 @file{sparc-tdep.c} for examples of how to do this.
356 Watchpoints are a special kind of breakpoints (@pxref{Algorithms,
357 breakpoints}) which break when data is accessed rather than when some
358 instruction is executed. When you have data which changes without
359 your knowing what code does that, watchpoints are the silver bullet to
360 hunt down and kill such bugs.
362 @cindex hardware watchpoints
363 @cindex software watchpoints
364 Watchpoints can be either hardware-assisted or not; the latter type is
365 known as ``software watchpoints.'' @value{GDBN} always uses
366 hardware-assisted watchpoints if they are available, and falls back on
367 software watchpoints otherwise. Typical situations where @value{GDBN}
368 will use software watchpoints are:
372 The watched memory region is too large for the underlying hardware
373 watchpoint support. For example, each x86 debug register can watch up
374 to 4 bytes of memory, so trying to watch data structures whose size is
375 more than 16 bytes will cause @value{GDBN} to use software
379 The value of the expression to be watched depends on data held in
380 registers (as opposed to memory).
383 Too many different watchpoints requested. (On some architectures,
384 this situation is impossible to detect until the debugged program is
385 resumed.) Note that x86 debug registers are used both for hardware
386 breakpoints and for watchpoints, so setting too many hardware
387 breakpoints might cause watchpoint insertion to fail.
390 No hardware-assisted watchpoints provided by the target
394 Software watchpoints are very slow, since @value{GDBN} needs to
395 single-step the program being debugged and test the value of the
396 watched expression(s) after each instruction. The rest of this
397 section is mostly irrelevant for software watchpoints.
399 @value{GDBN} uses several macros and primitives to support hardware
403 @findex TARGET_HAS_HARDWARE_WATCHPOINTS
404 @item TARGET_HAS_HARDWARE_WATCHPOINTS
405 If defined, the target supports hardware watchpoints.
407 @findex TARGET_CAN_USE_HARDWARE_WATCHPOINT
408 @item TARGET_CAN_USE_HARDWARE_WATCHPOINT (@var{type}, @var{count}, @var{other})
409 Return the number of hardware watchpoints of type @var{type} that are
410 possible to be set. The value is positive if @var{count} watchpoints
411 of this type can be set, zero if setting watchpoints of this type is
412 not supported, and negative if @var{count} is more than the maximum
413 number of watchpoints of type @var{type} that can be set. @var{other}
414 is non-zero if other types of watchpoints are currently enabled (there
415 are architectures which cannot set watchpoints of different types at
418 @findex TARGET_REGION_OK_FOR_HW_WATCHPOINT
419 @item TARGET_REGION_OK_FOR_HW_WATCHPOINT (@var{addr}, @var{len})
420 Return non-zero if hardware watchpoints can be used to watch a region
421 whose address is @var{addr} and whose length in bytes is @var{len}.
423 @findex TARGET_REGION_SIZE_OK_FOR_HW_WATCHPOINT
424 @item TARGET_REGION_SIZE_OK_FOR_HW_WATCHPOINT (@var{size})
425 Return non-zero if hardware watchpoints can be used to watch a region
426 whose size is @var{size}. @value{GDBN} only uses this macro as a
427 fall-back, in case @code{TARGET_REGION_OK_FOR_HW_WATCHPOINT} is not
430 @findex TARGET_DISABLE_HW_WATCHPOINTS
431 @item TARGET_DISABLE_HW_WATCHPOINTS (@var{pid})
432 Disables watchpoints in the process identified by @var{pid}. This is
433 used, e.g., on HP-UX which provides operations to disable and enable
434 the page-level memory protection that implements hardware watchpoints
437 @findex TARGET_ENABLE_HW_WATCHPOINTS
438 @item TARGET_ENABLE_HW_WATCHPOINTS (@var{pid})
439 Enables watchpoints in the process identified by @var{pid}. This is
440 used, e.g., on HP-UX which provides operations to disable and enable
441 the page-level memory protection that implements hardware watchpoints
444 @findex target_insert_watchpoint
445 @findex target_remove_watchpoint
446 @item target_insert_watchpoint (@var{addr}, @var{len}, @var{type})
447 @itemx target_remove_watchpoint (@var{addr}, @var{len}, @var{type})
448 Insert or remove a hardware watchpoint starting at @var{addr}, for
449 @var{len} bytes. @var{type} is the watchpoint type, one of the
450 possible values of the enumerated data type @code{target_hw_bp_type},
451 defined by @file{breakpoint.h} as follows:
454 enum target_hw_bp_type
456 hw_write = 0, /* Common (write) HW watchpoint */
457 hw_read = 1, /* Read HW watchpoint */
458 hw_access = 2, /* Access (read or write) HW watchpoint */
459 hw_execute = 3 /* Execute HW breakpoint */
464 These two macros should return 0 for success, non-zero for failure.
466 @cindex insert or remove hardware breakpoint
467 @findex target_remove_hw_breakpoint
468 @findex target_insert_hw_breakpoint
469 @item target_remove_hw_breakpoint (@var{addr}, @var{shadow})
470 @itemx target_insert_hw_breakpoint (@var{addr}, @var{shadow})
471 Insert or remove a hardware-assisted breakpoint at address @var{addr}.
472 Returns zero for success, non-zero for failure. @var{shadow} is the
473 real contents of the byte where the breakpoint has been inserted; it
474 is generally not valid when hardware breakpoints are used, but since
475 no other code touches these values, the implementations of the above
476 two macros can use them for their internal purposes.
478 @findex target_stopped_data_address
479 @item target_stopped_data_address ()
480 If the inferior has some watchpoint that triggered, return the address
481 associated with that watchpoint. Otherwise, return zero.
483 @findex HAVE_STEPPABLE_WATCHPOINT
484 @item HAVE_STEPPABLE_WATCHPOINT
485 If defined to a non-zero value, it is not necessary to disable a
486 watchpoint to step over it.
488 @findex HAVE_NONSTEPPABLE_WATCHPOINT
489 @item HAVE_NONSTEPPABLE_WATCHPOINT
490 If defined to a non-zero value, @value{GDBN} should disable a
491 watchpoint to step the inferior over it.
493 @findex HAVE_CONTINUABLE_WATCHPOINT
494 @item HAVE_CONTINUABLE_WATCHPOINT
495 If defined to a non-zero value, it is possible to continue the
496 inferior after a watchpoint has been hit.
498 @findex CANNOT_STEP_HW_WATCHPOINTS
499 @item CANNOT_STEP_HW_WATCHPOINTS
500 If this is defined to a non-zero value, @value{GDBN} will remove all
501 watchpoints before stepping the inferior.
503 @findex STOPPED_BY_WATCHPOINT
504 @item STOPPED_BY_WATCHPOINT (@var{wait_status})
505 Return non-zero if stopped by a watchpoint. @var{wait_status} is of
506 the type @code{struct target_waitstatus}, defined by @file{target.h}.
509 @subsection x86 Watchpoints
510 @cindex x86 debug registers
511 @cindex watchpoints, on x86
513 The 32-bit Intel x86 (a.k.a.@: ia32) processors feature special debug
514 registers designed to facilitate debugging. @value{GDBN} provides a
515 generic library of functions that x86-based ports can use to implement
516 support for watchpoints and hardware-assisted breakpoints. This
517 subsection documents the x86 watchpoint facilities in @value{GDBN}.
519 To use the generic x86 watchpoint support, a port should do the
523 @findex I386_USE_GENERIC_WATCHPOINTS
525 Define the macro @code{I386_USE_GENERIC_WATCHPOINTS} somewhere in the
526 target-dependent headers.
529 Include the @file{config/i386/nm-i386.h} header file @emph{after}
530 defining @code{I386_USE_GENERIC_WATCHPOINTS}.
533 Add @file{i386-nat.o} to the value of the Make variable
534 @code{NATDEPFILES} (@pxref{Native Debugging, NATDEPFILES}) or
535 @code{TDEPFILES} (@pxref{Target Architecture Definition, TDEPFILES}).
538 Provide implementations for the @code{I386_DR_LOW_*} macros described
539 below. Typically, each macro should call a target-specific function
540 which does the real work.
543 The x86 watchpoint support works by maintaining mirror images of the
544 debug registers. Values are copied between the mirror images and the
545 real debug registers via a set of macros which each target needs to
549 @findex I386_DR_LOW_SET_CONTROL
550 @item I386_DR_LOW_SET_CONTROL (@var{val})
551 Set the Debug Control (DR7) register to the value @var{val}.
553 @findex I386_DR_LOW_SET_ADDR
554 @item I386_DR_LOW_SET_ADDR (@var{idx}, @var{addr})
555 Put the address @var{addr} into the debug register number @var{idx}.
557 @findex I386_DR_LOW_RESET_ADDR
558 @item I386_DR_LOW_RESET_ADDR (@var{idx})
559 Reset (i.e.@: zero out) the address stored in the debug register
562 @findex I386_DR_LOW_GET_STATUS
563 @item I386_DR_LOW_GET_STATUS
564 Return the value of the Debug Status (DR6) register. This value is
565 used immediately after it is returned by
566 @code{I386_DR_LOW_GET_STATUS}, so as to support per-thread status
570 For each one of the 4 debug registers (whose indices are from 0 to 3)
571 that store addresses, a reference count is maintained by @value{GDBN},
572 to allow sharing of debug registers by several watchpoints. This
573 allows users to define several watchpoints that watch the same
574 expression, but with different conditions and/or commands, without
575 wasting debug registers which are in short supply. @value{GDBN}
576 maintains the reference counts internally, targets don't have to do
577 anything to use this feature.
579 The x86 debug registers can each watch a region that is 1, 2, or 4
580 bytes long. The ia32 architecture requires that each watched region
581 be appropriately aligned: 2-byte region on 2-byte boundary, 4-byte
582 region on 4-byte boundary. However, the x86 watchpoint support in
583 @value{GDBN} can watch unaligned regions and regions larger than 4
584 bytes (up to 16 bytes) by allocating several debug registers to watch
585 a single region. This allocation of several registers per a watched
586 region is also done automatically without target code intervention.
588 The generic x86 watchpoint support provides the following API for the
589 @value{GDBN}'s application code:
592 @findex i386_region_ok_for_watchpoint
593 @item i386_region_ok_for_watchpoint (@var{addr}, @var{len})
594 The macro @code{TARGET_REGION_OK_FOR_HW_WATCHPOINT} is set to call
595 this function. It counts the number of debug registers required to
596 watch a given region, and returns a non-zero value if that number is
597 less than 4, the number of debug registers available to x86
600 @findex i386_stopped_data_address
601 @item i386_stopped_data_address (void)
602 The macros @code{STOPPED_BY_WATCHPOINT} and
603 @code{target_stopped_data_address} are set to call this function. The
604 argument passed to @code{STOPPED_BY_WATCHPOINT} is ignored. This
605 function examines the breakpoint condition bits in the DR6 Debug
606 Status register, as returned by the @code{I386_DR_LOW_GET_STATUS}
607 macro, and returns the address associated with the first bit that is
610 @findex i386_insert_watchpoint
611 @findex i386_remove_watchpoint
612 @item i386_insert_watchpoint (@var{addr}, @var{len}, @var{type})
613 @itemx i386_remove_watchpoint (@var{addr}, @var{len}, @var{type})
614 Insert or remove a watchpoint. The macros
615 @code{target_insert_watchpoint} and @code{target_remove_watchpoint}
616 are set to call these functions. @code{i386_insert_watchpoint} first
617 looks for a debug register which is already set to watch the same
618 region for the same access types; if found, it just increments the
619 reference count of that debug register, thus implementing debug
620 register sharing between watchpoints. If no such register is found,
621 the function looks for a vacant debug register, sets its mirrored
622 value to @var{addr}, sets the mirrored value of DR7 Debug Control
623 register as appropriate for the @var{len} and @var{type} parameters,
624 and then passes the new values of the debug register and DR7 to the
625 inferior by calling @code{I386_DR_LOW_SET_ADDR} and
626 @code{I386_DR_LOW_SET_CONTROL}. If more than one debug register is
627 required to cover the given region, the above process is repeated for
630 @code{i386_remove_watchpoint} does the opposite: it resets the address
631 in the mirrored value of the debug register and its read/write and
632 length bits in the mirrored value of DR7, then passes these new
633 values to the inferior via @code{I386_DR_LOW_RESET_ADDR} and
634 @code{I386_DR_LOW_SET_CONTROL}. If a register is shared by several
635 watchpoints, each time a @code{i386_remove_watchpoint} is called, it
636 decrements the reference count, and only calls
637 @code{I386_DR_LOW_RESET_ADDR} and @code{I386_DR_LOW_SET_CONTROL} when
638 the count goes to zero.
640 @findex i386_insert_hw_breakpoint
641 @findex i386_remove_hw_breakpoint
642 @item i386_insert_hw_breakpoint (@var{addr}, @var{shadow}
643 @itemx i386_remove_hw_breakpoint (@var{addr}, @var{shadow})
644 These functions insert and remove hardware-assisted breakpoints. The
645 macros @code{target_insert_hw_breakpoint} and
646 @code{target_remove_hw_breakpoint} are set to call these functions.
647 These functions work like @code{i386_insert_watchpoint} and
648 @code{i386_remove_watchpoint}, respectively, except that they set up
649 the debug registers to watch instruction execution, and each
650 hardware-assisted breakpoint always requires exactly one debug
653 @findex i386_stopped_by_hwbp
654 @item i386_stopped_by_hwbp (void)
655 This function returns non-zero if the inferior has some watchpoint or
656 hardware breakpoint that triggered. It works like
657 @code{i386_stopped_data_address}, except that it doesn't return the
658 address whose watchpoint triggered.
660 @findex i386_cleanup_dregs
661 @item i386_cleanup_dregs (void)
662 This function clears all the reference counts, addresses, and control
663 bits in the mirror images of the debug registers. It doesn't affect
664 the actual debug registers in the inferior process.
671 x86 processors support setting watchpoints on I/O reads or writes.
672 However, since no target supports this (as of March 2001), and since
673 @code{enum target_hw_bp_type} doesn't even have an enumeration for I/O
674 watchpoints, this feature is not yet available to @value{GDBN} running
678 x86 processors can enable watchpoints locally, for the current task
679 only, or globally, for all the tasks. For each debug register,
680 there's a bit in the DR7 Debug Control register that determines
681 whether the associated address is watched locally or globally. The
682 current implementation of x86 watchpoint support in @value{GDBN}
683 always sets watchpoints to be locally enabled, since global
684 watchpoints might interfere with the underlying OS and are probably
685 unavailable in many platforms.
688 @section Observing changes in @value{GDBN} internals
689 @cindex observer pattern interface
690 @cindex notifications about changes in internals
692 In order to function properly, several modules need to be notified when
693 some changes occur in the @value{GDBN} internals. Traditionally, these
694 modules have relied on several paradigms, the most common ones being
695 hooks and gdb-events. Unfortunately, none of these paradigms was
696 versatile enough to become the standard notification mechanism in
697 @value{GDBN}. The fact that they only supported one ``client'' was also
700 A new paradigm, based on the Observer pattern of the @cite{Design
701 Patterns} book, has therefore been implemented. The goal was to provide
702 a new interface overcoming the issues with the notification mechanisms
703 previously available. This new interface needed to be strongly typed,
704 easy to extend, and versatile enough to be used as the standard
705 interface when adding new notifications.
707 See @ref{GDB Observers} for a brief description of the observers
708 currently implemented in GDB. The rationale for the current
709 implementation is also briefly discussed.
713 @chapter User Interface
715 @value{GDBN} has several user interfaces. Although the command-line interface
716 is the most common and most familiar, there are others.
718 @section Command Interpreter
720 @cindex command interpreter
722 The command interpreter in @value{GDBN} is fairly simple. It is designed to
723 allow for the set of commands to be augmented dynamically, and also
724 has a recursive subcommand capability, where the first argument to
725 a command may itself direct a lookup on a different command list.
727 For instance, the @samp{set} command just starts a lookup on the
728 @code{setlist} command list, while @samp{set thread} recurses
729 to the @code{set_thread_cmd_list}.
733 To add commands in general, use @code{add_cmd}. @code{add_com} adds to
734 the main command list, and should be used for those commands. The usual
735 place to add commands is in the @code{_initialize_@var{xyz}} routines at
736 the ends of most source files.
738 @findex add_setshow_cmd
739 @findex add_setshow_cmd_full
740 To add paired @samp{set} and @samp{show} commands, use
741 @code{add_setshow_cmd} or @code{add_setshow_cmd_full}. The former is
742 a slightly simpler interface which is useful when you don't need to
743 further modify the new command structures, while the latter returns
744 the new command structures for manipulation.
746 @cindex deprecating commands
747 @findex deprecate_cmd
748 Before removing commands from the command set it is a good idea to
749 deprecate them for some time. Use @code{deprecate_cmd} on commands or
750 aliases to set the deprecated flag. @code{deprecate_cmd} takes a
751 @code{struct cmd_list_element} as it's first argument. You can use the
752 return value from @code{add_com} or @code{add_cmd} to deprecate the
753 command immediately after it is created.
755 The first time a command is used the user will be warned and offered a
756 replacement (if one exists). Note that the replacement string passed to
757 @code{deprecate_cmd} should be the full name of the command, i.e. the
758 entire string the user should type at the command line.
760 @section UI-Independent Output---the @code{ui_out} Functions
761 @c This section is based on the documentation written by Fernando
762 @c Nasser <fnasser@redhat.com>.
764 @cindex @code{ui_out} functions
765 The @code{ui_out} functions present an abstraction level for the
766 @value{GDBN} output code. They hide the specifics of different user
767 interfaces supported by @value{GDBN}, and thus free the programmer
768 from the need to write several versions of the same code, one each for
769 every UI, to produce output.
771 @subsection Overview and Terminology
773 In general, execution of each @value{GDBN} command produces some sort
774 of output, and can even generate an input request.
776 Output can be generated for the following purposes:
780 to display a @emph{result} of an operation;
783 to convey @emph{info} or produce side-effects of a requested
787 to provide a @emph{notification} of an asynchronous event (including
788 progress indication of a prolonged asynchronous operation);
791 to display @emph{error messages} (including warnings);
794 to show @emph{debug data};
797 to @emph{query} or prompt a user for input (a special case).
801 This section mainly concentrates on how to build result output,
802 although some of it also applies to other kinds of output.
804 Generation of output that displays the results of an operation
805 involves one or more of the following:
809 output of the actual data
812 formatting the output as appropriate for console output, to make it
813 easily readable by humans
816 machine oriented formatting--a more terse formatting to allow for easy
817 parsing by programs which read @value{GDBN}'s output
820 annotation, whose purpose is to help legacy GUIs to identify interesting
824 The @code{ui_out} routines take care of the first three aspects.
825 Annotations are provided by separate annotation routines. Note that use
826 of annotations for an interface between a GUI and @value{GDBN} is
829 Output can be in the form of a single item, which we call a @dfn{field};
830 a @dfn{list} consisting of identical fields; a @dfn{tuple} consisting of
831 non-identical fields; or a @dfn{table}, which is a tuple consisting of a
832 header and a body. In a BNF-like form:
835 @item <table> @expansion{}
836 @code{<header> <body>}
837 @item <header> @expansion{}
838 @code{@{ <column> @}}
839 @item <column> @expansion{}
840 @code{<width> <alignment> <title>}
841 @item <body> @expansion{}
846 @subsection General Conventions
848 Most @code{ui_out} routines are of type @code{void}, the exceptions are
849 @code{ui_out_stream_new} (which returns a pointer to the newly created
850 object) and the @code{make_cleanup} routines.
852 The first parameter is always the @code{ui_out} vector object, a pointer
853 to a @code{struct ui_out}.
855 The @var{format} parameter is like in @code{printf} family of functions.
856 When it is present, there must also be a variable list of arguments
857 sufficient used to satisfy the @code{%} specifiers in the supplied
860 When a character string argument is not used in a @code{ui_out} function
861 call, a @code{NULL} pointer has to be supplied instead.
864 @subsection Table, Tuple and List Functions
866 @cindex list output functions
867 @cindex table output functions
868 @cindex tuple output functions
869 This section introduces @code{ui_out} routines for building lists,
870 tuples and tables. The routines to output the actual data items
871 (fields) are presented in the next section.
873 To recap: A @dfn{tuple} is a sequence of @dfn{fields}, each field
874 containing information about an object; a @dfn{list} is a sequence of
875 fields where each field describes an identical object.
877 Use the @dfn{table} functions when your output consists of a list of
878 rows (tuples) and the console output should include a heading. Use this
879 even when you are listing just one object but you still want the header.
881 @cindex nesting level in @code{ui_out} functions
882 Tables can not be nested. Tuples and lists can be nested up to a
883 maximum of five levels.
885 The overall structure of the table output code is something like this:
900 Here is the description of table-, tuple- and list-related @code{ui_out}
903 @deftypefun void ui_out_table_begin (struct ui_out *@var{uiout}, int @var{nbrofcols}, int @var{nr_rows}, const char *@var{tblid})
904 The function @code{ui_out_table_begin} marks the beginning of the output
905 of a table. It should always be called before any other @code{ui_out}
906 function for a given table. @var{nbrofcols} is the number of columns in
907 the table. @var{nr_rows} is the number of rows in the table.
908 @var{tblid} is an optional string identifying the table. The string
909 pointed to by @var{tblid} is copied by the implementation of
910 @code{ui_out_table_begin}, so the application can free the string if it
913 The companion function @code{ui_out_table_end}, described below, marks
914 the end of the table's output.
917 @deftypefun void ui_out_table_header (struct ui_out *@var{uiout}, int @var{width}, enum ui_align @var{alignment}, const char *@var{colhdr})
918 @code{ui_out_table_header} provides the header information for a single
919 table column. You call this function several times, one each for every
920 column of the table, after @code{ui_out_table_begin}, but before
921 @code{ui_out_table_body}.
923 The value of @var{width} gives the column width in characters. The
924 value of @var{alignment} is one of @code{left}, @code{center}, and
925 @code{right}, and it specifies how to align the header: left-justify,
926 center, or right-justify it. @var{colhdr} points to a string that
927 specifies the column header; the implementation copies that string, so
928 column header strings in @code{malloc}ed storage can be freed after the
932 @deftypefun void ui_out_table_body (struct ui_out *@var{uiout})
933 This function delimits the table header from the table body.
936 @deftypefun void ui_out_table_end (struct ui_out *@var{uiout})
937 This function signals the end of a table's output. It should be called
938 after the table body has been produced by the list and field output
941 There should be exactly one call to @code{ui_out_table_end} for each
942 call to @code{ui_out_table_begin}, otherwise the @code{ui_out} functions
943 will signal an internal error.
946 The output of the tuples that represent the table rows must follow the
947 call to @code{ui_out_table_body} and precede the call to
948 @code{ui_out_table_end}. You build a tuple by calling
949 @code{ui_out_tuple_begin} and @code{ui_out_tuple_end}, with suitable
950 calls to functions which actually output fields between them.
952 @deftypefun void ui_out_tuple_begin (struct ui_out *@var{uiout}, const char *@var{id})
953 This function marks the beginning of a tuple output. @var{id} points
954 to an optional string that identifies the tuple; it is copied by the
955 implementation, and so strings in @code{malloc}ed storage can be freed
959 @deftypefun void ui_out_tuple_end (struct ui_out *@var{uiout})
960 This function signals an end of a tuple output. There should be exactly
961 one call to @code{ui_out_tuple_end} for each call to
962 @code{ui_out_tuple_begin}, otherwise an internal @value{GDBN} error will
966 @deftypefun struct cleanup *make_cleanup_ui_out_tuple_begin_end (struct ui_out *@var{uiout}, const char *@var{id})
967 This function first opens the tuple and then establishes a cleanup
968 (@pxref{Coding, Cleanups}) to close the tuple. It provides a convenient
969 and correct implementation of the non-portable@footnote{The function
970 cast is not portable ISO C.} code sequence:
972 struct cleanup *old_cleanup;
973 ui_out_tuple_begin (uiout, "...");
974 old_cleanup = make_cleanup ((void(*)(void *)) ui_out_tuple_end,
979 @deftypefun void ui_out_list_begin (struct ui_out *@var{uiout}, const char *@var{id})
980 This function marks the beginning of a list output. @var{id} points to
981 an optional string that identifies the list; it is copied by the
982 implementation, and so strings in @code{malloc}ed storage can be freed
986 @deftypefun void ui_out_list_end (struct ui_out *@var{uiout})
987 This function signals an end of a list output. There should be exactly
988 one call to @code{ui_out_list_end} for each call to
989 @code{ui_out_list_begin}, otherwise an internal @value{GDBN} error will
993 @deftypefun struct cleanup *make_cleanup_ui_out_list_begin_end (struct ui_out *@var{uiout}, const char *@var{id})
994 Similar to @code{make_cleanup_ui_out_tuple_begin_end}, this function
995 opens a list and then establishes cleanup (@pxref{Coding, Cleanups})
996 that will close the list.list.
999 @subsection Item Output Functions
1001 @cindex item output functions
1002 @cindex field output functions
1004 The functions described below produce output for the actual data
1005 items, or fields, which contain information about the object.
1007 Choose the appropriate function accordingly to your particular needs.
1009 @deftypefun void ui_out_field_fmt (struct ui_out *@var{uiout}, char *@var{fldname}, char *@var{format}, ...)
1010 This is the most general output function. It produces the
1011 representation of the data in the variable-length argument list
1012 according to formatting specifications in @var{format}, a
1013 @code{printf}-like format string. The optional argument @var{fldname}
1014 supplies the name of the field. The data items themselves are
1015 supplied as additional arguments after @var{format}.
1017 This generic function should be used only when it is not possible to
1018 use one of the specialized versions (see below).
1021 @deftypefun void ui_out_field_int (struct ui_out *@var{uiout}, const char *@var{fldname}, int @var{value})
1022 This function outputs a value of an @code{int} variable. It uses the
1023 @code{"%d"} output conversion specification. @var{fldname} specifies
1024 the name of the field.
1027 @deftypefun void ui_out_field_fmt_int (struct ui_out *@var{uiout}, int @var{width}, enum ui_align @var{alignment}, const char *@var{fldname}, int @var{value})
1028 This function outputs a value of an @code{int} variable. It differs from
1029 @code{ui_out_field_int} in that the caller specifies the desired @var{width} and @var{alignment} of the output.
1030 @var{fldname} specifies
1031 the name of the field.
1034 @deftypefun void ui_out_field_core_addr (struct ui_out *@var{uiout}, const char *@var{fldname}, CORE_ADDR @var{address})
1035 This function outputs an address.
1038 @deftypefun void ui_out_field_string (struct ui_out *@var{uiout}, const char *@var{fldname}, const char *@var{string})
1039 This function outputs a string using the @code{"%s"} conversion
1043 Sometimes, there's a need to compose your output piece by piece using
1044 functions that operate on a stream, such as @code{value_print} or
1045 @code{fprintf_symbol_filtered}. These functions accept an argument of
1046 the type @code{struct ui_file *}, a pointer to a @code{ui_file} object
1047 used to store the data stream used for the output. When you use one
1048 of these functions, you need a way to pass their results stored in a
1049 @code{ui_file} object to the @code{ui_out} functions. To this end,
1050 you first create a @code{ui_stream} object by calling
1051 @code{ui_out_stream_new}, pass the @code{stream} member of that
1052 @code{ui_stream} object to @code{value_print} and similar functions,
1053 and finally call @code{ui_out_field_stream} to output the field you
1054 constructed. When the @code{ui_stream} object is no longer needed,
1055 you should destroy it and free its memory by calling
1056 @code{ui_out_stream_delete}.
1058 @deftypefun struct ui_stream *ui_out_stream_new (struct ui_out *@var{uiout})
1059 This function creates a new @code{ui_stream} object which uses the
1060 same output methods as the @code{ui_out} object whose pointer is
1061 passed in @var{uiout}. It returns a pointer to the newly created
1062 @code{ui_stream} object.
1065 @deftypefun void ui_out_stream_delete (struct ui_stream *@var{streambuf})
1066 This functions destroys a @code{ui_stream} object specified by
1070 @deftypefun void ui_out_field_stream (struct ui_out *@var{uiout}, const char *@var{fieldname}, struct ui_stream *@var{streambuf})
1071 This function consumes all the data accumulated in
1072 @code{streambuf->stream} and outputs it like
1073 @code{ui_out_field_string} does. After a call to
1074 @code{ui_out_field_stream}, the accumulated data no longer exists, but
1075 the stream is still valid and may be used for producing more fields.
1078 @strong{Important:} If there is any chance that your code could bail
1079 out before completing output generation and reaching the point where
1080 @code{ui_out_stream_delete} is called, it is necessary to set up a
1081 cleanup, to avoid leaking memory and other resources. Here's a
1082 skeleton code to do that:
1085 struct ui_stream *mybuf = ui_out_stream_new (uiout);
1086 struct cleanup *old = make_cleanup (ui_out_stream_delete, mybuf);
1091 If the function already has the old cleanup chain set (for other kinds
1092 of cleanups), you just have to add your cleanup to it:
1095 mybuf = ui_out_stream_new (uiout);
1096 make_cleanup (ui_out_stream_delete, mybuf);
1099 Note that with cleanups in place, you should not call
1100 @code{ui_out_stream_delete} directly, or you would attempt to free the
1103 @subsection Utility Output Functions
1105 @deftypefun void ui_out_field_skip (struct ui_out *@var{uiout}, const char *@var{fldname})
1106 This function skips a field in a table. Use it if you have to leave
1107 an empty field without disrupting the table alignment. The argument
1108 @var{fldname} specifies a name for the (missing) filed.
1111 @deftypefun void ui_out_text (struct ui_out *@var{uiout}, const char *@var{string})
1112 This function outputs the text in @var{string} in a way that makes it
1113 easy to be read by humans. For example, the console implementation of
1114 this method filters the text through a built-in pager, to prevent it
1115 from scrolling off the visible portion of the screen.
1117 Use this function for printing relatively long chunks of text around
1118 the actual field data: the text it produces is not aligned according
1119 to the table's format. Use @code{ui_out_field_string} to output a
1120 string field, and use @code{ui_out_message}, described below, to
1121 output short messages.
1124 @deftypefun void ui_out_spaces (struct ui_out *@var{uiout}, int @var{nspaces})
1125 This function outputs @var{nspaces} spaces. It is handy to align the
1126 text produced by @code{ui_out_text} with the rest of the table or
1130 @deftypefun void ui_out_message (struct ui_out *@var{uiout}, int @var{verbosity}, const char *@var{format}, ...)
1131 This function produces a formatted message, provided that the current
1132 verbosity level is at least as large as given by @var{verbosity}. The
1133 current verbosity level is specified by the user with the @samp{set
1134 verbositylevel} command.@footnote{As of this writing (April 2001),
1135 setting verbosity level is not yet implemented, and is always returned
1136 as zero. So calling @code{ui_out_message} with a @var{verbosity}
1137 argument more than zero will cause the message to never be printed.}
1140 @deftypefun void ui_out_wrap_hint (struct ui_out *@var{uiout}, char *@var{indent})
1141 This function gives the console output filter (a paging filter) a hint
1142 of where to break lines which are too long. Ignored for all other
1143 output consumers. @var{indent}, if non-@code{NULL}, is the string to
1144 be printed to indent the wrapped text on the next line; it must remain
1145 accessible until the next call to @code{ui_out_wrap_hint}, or until an
1146 explicit newline is produced by one of the other functions. If
1147 @var{indent} is @code{NULL}, the wrapped text will not be indented.
1150 @deftypefun void ui_out_flush (struct ui_out *@var{uiout})
1151 This function flushes whatever output has been accumulated so far, if
1152 the UI buffers output.
1156 @subsection Examples of Use of @code{ui_out} functions
1158 @cindex using @code{ui_out} functions
1159 @cindex @code{ui_out} functions, usage examples
1160 This section gives some practical examples of using the @code{ui_out}
1161 functions to generalize the old console-oriented code in
1162 @value{GDBN}. The examples all come from functions defined on the
1163 @file{breakpoints.c} file.
1165 This example, from the @code{breakpoint_1} function, shows how to
1168 The original code was:
1171 if (!found_a_breakpoint++)
1173 annotate_breakpoints_headers ();
1176 printf_filtered ("Num ");
1178 printf_filtered ("Type ");
1180 printf_filtered ("Disp ");
1182 printf_filtered ("Enb ");
1186 printf_filtered ("Address ");
1189 printf_filtered ("What\n");
1191 annotate_breakpoints_table ();
1195 Here's the new version:
1198 nr_printable_breakpoints = @dots{};
1201 ui_out_table_begin (ui, 6, nr_printable_breakpoints, "BreakpointTable");
1203 ui_out_table_begin (ui, 5, nr_printable_breakpoints, "BreakpointTable");
1205 if (nr_printable_breakpoints > 0)
1206 annotate_breakpoints_headers ();
1207 if (nr_printable_breakpoints > 0)
1209 ui_out_table_header (uiout, 3, ui_left, "number", "Num"); /* 1 */
1210 if (nr_printable_breakpoints > 0)
1212 ui_out_table_header (uiout, 14, ui_left, "type", "Type"); /* 2 */
1213 if (nr_printable_breakpoints > 0)
1215 ui_out_table_header (uiout, 4, ui_left, "disp", "Disp"); /* 3 */
1216 if (nr_printable_breakpoints > 0)
1218 ui_out_table_header (uiout, 3, ui_left, "enabled", "Enb"); /* 4 */
1221 if (nr_printable_breakpoints > 0)
1223 if (TARGET_ADDR_BIT <= 32)
1224 ui_out_table_header (uiout, 10, ui_left, "addr", "Address");/* 5 */
1226 ui_out_table_header (uiout, 18, ui_left, "addr", "Address");/* 5 */
1228 if (nr_printable_breakpoints > 0)
1230 ui_out_table_header (uiout, 40, ui_noalign, "what", "What"); /* 6 */
1231 ui_out_table_body (uiout);
1232 if (nr_printable_breakpoints > 0)
1233 annotate_breakpoints_table ();
1236 This example, from the @code{print_one_breakpoint} function, shows how
1237 to produce the actual data for the table whose structure was defined
1238 in the above example. The original code was:
1243 printf_filtered ("%-3d ", b->number);
1245 if ((int)b->type > (sizeof(bptypes)/sizeof(bptypes[0]))
1246 || ((int) b->type != bptypes[(int) b->type].type))
1247 internal_error ("bptypes table does not describe type #%d.",
1249 printf_filtered ("%-14s ", bptypes[(int)b->type].description);
1251 printf_filtered ("%-4s ", bpdisps[(int)b->disposition]);
1253 printf_filtered ("%-3c ", bpenables[(int)b->enable]);
1257 This is the new version:
1261 ui_out_tuple_begin (uiout, "bkpt");
1263 ui_out_field_int (uiout, "number", b->number);
1265 if (((int) b->type > (sizeof (bptypes) / sizeof (bptypes[0])))
1266 || ((int) b->type != bptypes[(int) b->type].type))
1267 internal_error ("bptypes table does not describe type #%d.",
1269 ui_out_field_string (uiout, "type", bptypes[(int)b->type].description);
1271 ui_out_field_string (uiout, "disp", bpdisps[(int)b->disposition]);
1273 ui_out_field_fmt (uiout, "enabled", "%c", bpenables[(int)b->enable]);
1277 This example, also from @code{print_one_breakpoint}, shows how to
1278 produce a complicated output field using the @code{print_expression}
1279 functions which requires a stream to be passed. It also shows how to
1280 automate stream destruction with cleanups. The original code was:
1284 print_expression (b->exp, gdb_stdout);
1290 struct ui_stream *stb = ui_out_stream_new (uiout);
1291 struct cleanup *old_chain = make_cleanup_ui_out_stream_delete (stb);
1294 print_expression (b->exp, stb->stream);
1295 ui_out_field_stream (uiout, "what", local_stream);
1298 This example, also from @code{print_one_breakpoint}, shows how to use
1299 @code{ui_out_text} and @code{ui_out_field_string}. The original code
1304 if (b->dll_pathname == NULL)
1305 printf_filtered ("<any library> ");
1307 printf_filtered ("library \"%s\" ", b->dll_pathname);
1314 if (b->dll_pathname == NULL)
1316 ui_out_field_string (uiout, "what", "<any library>");
1317 ui_out_spaces (uiout, 1);
1321 ui_out_text (uiout, "library \"");
1322 ui_out_field_string (uiout, "what", b->dll_pathname);
1323 ui_out_text (uiout, "\" ");
1327 The following example from @code{print_one_breakpoint} shows how to
1328 use @code{ui_out_field_int} and @code{ui_out_spaces}. The original
1333 if (b->forked_inferior_pid != 0)
1334 printf_filtered ("process %d ", b->forked_inferior_pid);
1341 if (b->forked_inferior_pid != 0)
1343 ui_out_text (uiout, "process ");
1344 ui_out_field_int (uiout, "what", b->forked_inferior_pid);
1345 ui_out_spaces (uiout, 1);
1349 Here's an example of using @code{ui_out_field_string}. The original
1354 if (b->exec_pathname != NULL)
1355 printf_filtered ("program \"%s\" ", b->exec_pathname);
1362 if (b->exec_pathname != NULL)
1364 ui_out_text (uiout, "program \"");
1365 ui_out_field_string (uiout, "what", b->exec_pathname);
1366 ui_out_text (uiout, "\" ");
1370 Finally, here's an example of printing an address. The original code:
1374 printf_filtered ("%s ",
1375 local_hex_string_custom ((unsigned long) b->address, "08l"));
1382 ui_out_field_core_addr (uiout, "Address", b->address);
1386 @section Console Printing
1395 @cindex @code{libgdb}
1396 @code{libgdb} 1.0 was an abortive project of years ago. The theory was
1397 to provide an API to @value{GDBN}'s functionality.
1400 @cindex @code{libgdb}
1401 @code{libgdb} 2.0 is an ongoing effort to update @value{GDBN} so that is
1402 better able to support graphical and other environments.
1404 Since @code{libgdb} development is on-going, its architecture is still
1405 evolving. The following components have so far been identified:
1409 Observer - @file{gdb-events.h}.
1411 Builder - @file{ui-out.h}
1413 Event Loop - @file{event-loop.h}
1415 Library - @file{gdb.h}
1418 The model that ties these components together is described below.
1420 @section The @code{libgdb} Model
1422 A client of @code{libgdb} interacts with the library in two ways.
1426 As an observer (using @file{gdb-events}) receiving notifications from
1427 @code{libgdb} of any internal state changes (break point changes, run
1430 As a client querying @code{libgdb} (using the @file{ui-out} builder) to
1431 obtain various status values from @value{GDBN}.
1434 Since @code{libgdb} could have multiple clients (e.g. a GUI supporting
1435 the existing @value{GDBN} CLI), those clients must co-operate when
1436 controlling @code{libgdb}. In particular, a client must ensure that
1437 @code{libgdb} is idle (i.e. no other client is using @code{libgdb})
1438 before responding to a @file{gdb-event} by making a query.
1440 @section CLI support
1442 At present @value{GDBN}'s CLI is very much entangled in with the core of
1443 @code{libgdb}. Consequently, a client wishing to include the CLI in
1444 their interface needs to carefully co-ordinate its own and the CLI's
1447 It is suggested that the client set @code{libgdb} up to be bi-modal
1448 (alternate between CLI and client query modes). The notes below sketch
1453 The client registers itself as an observer of @code{libgdb}.
1455 The client create and install @code{cli-out} builder using its own
1456 versions of the @code{ui-file} @code{gdb_stderr}, @code{gdb_stdtarg} and
1457 @code{gdb_stdout} streams.
1459 The client creates a separate custom @code{ui-out} builder that is only
1460 used while making direct queries to @code{libgdb}.
1463 When the client receives input intended for the CLI, it simply passes it
1464 along. Since the @code{cli-out} builder is installed by default, all
1465 the CLI output in response to that command is routed (pronounced rooted)
1466 through to the client controlled @code{gdb_stdout} et.@: al.@: streams.
1467 At the same time, the client is kept abreast of internal changes by
1468 virtue of being a @code{libgdb} observer.
1470 The only restriction on the client is that it must wait until
1471 @code{libgdb} becomes idle before initiating any queries (using the
1472 client's custom builder).
1474 @section @code{libgdb} components
1476 @subheading Observer - @file{gdb-events.h}
1477 @file{gdb-events} provides the client with a very raw mechanism that can
1478 be used to implement an observer. At present it only allows for one
1479 observer and that observer must, internally, handle the need to delay
1480 the processing of any event notifications until after @code{libgdb} has
1481 finished the current command.
1483 @subheading Builder - @file{ui-out.h}
1484 @file{ui-out} provides the infrastructure necessary for a client to
1485 create a builder. That builder is then passed down to @code{libgdb}
1486 when doing any queries.
1488 @subheading Event Loop - @file{event-loop.h}
1489 @c There could be an entire section on the event-loop
1490 @file{event-loop}, currently non-re-entrant, provides a simple event
1491 loop. A client would need to either plug its self into this loop or,
1492 implement a new event-loop that GDB would use.
1494 The event-loop will eventually be made re-entrant. This is so that
1495 @value{GDBN} can better handle the problem of some commands blocking
1496 instead of returning.
1498 @subheading Library - @file{gdb.h}
1499 @file{libgdb} is the most obvious component of this system. It provides
1500 the query interface. Each function is parameterized by a @code{ui-out}
1501 builder. The result of the query is constructed using that builder
1502 before the query function returns.
1504 @node Symbol Handling
1506 @chapter Symbol Handling
1508 Symbols are a key part of @value{GDBN}'s operation. Symbols include variables,
1509 functions, and types.
1511 @section Symbol Reading
1513 @cindex symbol reading
1514 @cindex reading of symbols
1515 @cindex symbol files
1516 @value{GDBN} reads symbols from @dfn{symbol files}. The usual symbol
1517 file is the file containing the program which @value{GDBN} is
1518 debugging. @value{GDBN} can be directed to use a different file for
1519 symbols (with the @samp{symbol-file} command), and it can also read
1520 more symbols via the @samp{add-file} and @samp{load} commands, or while
1521 reading symbols from shared libraries.
1523 @findex find_sym_fns
1524 Symbol files are initially opened by code in @file{symfile.c} using
1525 the BFD library (@pxref{Support Libraries}). BFD identifies the type
1526 of the file by examining its header. @code{find_sym_fns} then uses
1527 this identification to locate a set of symbol-reading functions.
1529 @findex add_symtab_fns
1530 @cindex @code{sym_fns} structure
1531 @cindex adding a symbol-reading module
1532 Symbol-reading modules identify themselves to @value{GDBN} by calling
1533 @code{add_symtab_fns} during their module initialization. The argument
1534 to @code{add_symtab_fns} is a @code{struct sym_fns} which contains the
1535 name (or name prefix) of the symbol format, the length of the prefix,
1536 and pointers to four functions. These functions are called at various
1537 times to process symbol files whose identification matches the specified
1540 The functions supplied by each module are:
1543 @item @var{xyz}_symfile_init(struct sym_fns *sf)
1545 @cindex secondary symbol file
1546 Called from @code{symbol_file_add} when we are about to read a new
1547 symbol file. This function should clean up any internal state (possibly
1548 resulting from half-read previous files, for example) and prepare to
1549 read a new symbol file. Note that the symbol file which we are reading
1550 might be a new ``main'' symbol file, or might be a secondary symbol file
1551 whose symbols are being added to the existing symbol table.
1553 The argument to @code{@var{xyz}_symfile_init} is a newly allocated
1554 @code{struct sym_fns} whose @code{bfd} field contains the BFD for the
1555 new symbol file being read. Its @code{private} field has been zeroed,
1556 and can be modified as desired. Typically, a struct of private
1557 information will be @code{malloc}'d, and a pointer to it will be placed
1558 in the @code{private} field.
1560 There is no result from @code{@var{xyz}_symfile_init}, but it can call
1561 @code{error} if it detects an unavoidable problem.
1563 @item @var{xyz}_new_init()
1565 Called from @code{symbol_file_add} when discarding existing symbols.
1566 This function needs only handle the symbol-reading module's internal
1567 state; the symbol table data structures visible to the rest of
1568 @value{GDBN} will be discarded by @code{symbol_file_add}. It has no
1569 arguments and no result. It may be called after
1570 @code{@var{xyz}_symfile_init}, if a new symbol table is being read, or
1571 may be called alone if all symbols are simply being discarded.
1573 @item @var{xyz}_symfile_read(struct sym_fns *sf, CORE_ADDR addr, int mainline)
1575 Called from @code{symbol_file_add} to actually read the symbols from a
1576 symbol-file into a set of psymtabs or symtabs.
1578 @code{sf} points to the @code{struct sym_fns} originally passed to
1579 @code{@var{xyz}_sym_init} for possible initialization. @code{addr} is
1580 the offset between the file's specified start address and its true
1581 address in memory. @code{mainline} is 1 if this is the main symbol
1582 table being read, and 0 if a secondary symbol file (e.g. shared library
1583 or dynamically loaded file) is being read.@refill
1586 In addition, if a symbol-reading module creates psymtabs when
1587 @var{xyz}_symfile_read is called, these psymtabs will contain a pointer
1588 to a function @code{@var{xyz}_psymtab_to_symtab}, which can be called
1589 from any point in the @value{GDBN} symbol-handling code.
1592 @item @var{xyz}_psymtab_to_symtab (struct partial_symtab *pst)
1594 Called from @code{psymtab_to_symtab} (or the @code{PSYMTAB_TO_SYMTAB} macro) if
1595 the psymtab has not already been read in and had its @code{pst->symtab}
1596 pointer set. The argument is the psymtab to be fleshed-out into a
1597 symtab. Upon return, @code{pst->readin} should have been set to 1, and
1598 @code{pst->symtab} should contain a pointer to the new corresponding symtab, or
1599 zero if there were no symbols in that part of the symbol file.
1602 @section Partial Symbol Tables
1604 @value{GDBN} has three types of symbol tables:
1607 @cindex full symbol table
1610 Full symbol tables (@dfn{symtabs}). These contain the main
1611 information about symbols and addresses.
1615 Partial symbol tables (@dfn{psymtabs}). These contain enough
1616 information to know when to read the corresponding part of the full
1619 @cindex minimal symbol table
1622 Minimal symbol tables (@dfn{msymtabs}). These contain information
1623 gleaned from non-debugging symbols.
1626 @cindex partial symbol table
1627 This section describes partial symbol tables.
1629 A psymtab is constructed by doing a very quick pass over an executable
1630 file's debugging information. Small amounts of information are
1631 extracted---enough to identify which parts of the symbol table will
1632 need to be re-read and fully digested later, when the user needs the
1633 information. The speed of this pass causes @value{GDBN} to start up very
1634 quickly. Later, as the detailed rereading occurs, it occurs in small
1635 pieces, at various times, and the delay therefrom is mostly invisible to
1637 @c (@xref{Symbol Reading}.)
1639 The symbols that show up in a file's psymtab should be, roughly, those
1640 visible to the debugger's user when the program is not running code from
1641 that file. These include external symbols and types, static symbols and
1642 types, and @code{enum} values declared at file scope.
1644 The psymtab also contains the range of instruction addresses that the
1645 full symbol table would represent.
1647 @cindex finding a symbol
1648 @cindex symbol lookup
1649 The idea is that there are only two ways for the user (or much of the
1650 code in the debugger) to reference a symbol:
1653 @findex find_pc_function
1654 @findex find_pc_line
1656 By its address (e.g. execution stops at some address which is inside a
1657 function in this file). The address will be noticed to be in the
1658 range of this psymtab, and the full symtab will be read in.
1659 @code{find_pc_function}, @code{find_pc_line}, and other
1660 @code{find_pc_@dots{}} functions handle this.
1662 @cindex lookup_symbol
1665 (e.g. the user asks to print a variable, or set a breakpoint on a
1666 function). Global names and file-scope names will be found in the
1667 psymtab, which will cause the symtab to be pulled in. Local names will
1668 have to be qualified by a global name, or a file-scope name, in which
1669 case we will have already read in the symtab as we evaluated the
1670 qualifier. Or, a local symbol can be referenced when we are ``in'' a
1671 local scope, in which case the first case applies. @code{lookup_symbol}
1672 does most of the work here.
1675 The only reason that psymtabs exist is to cause a symtab to be read in
1676 at the right moment. Any symbol that can be elided from a psymtab,
1677 while still causing that to happen, should not appear in it. Since
1678 psymtabs don't have the idea of scope, you can't put local symbols in
1679 them anyway. Psymtabs don't have the idea of the type of a symbol,
1680 either, so types need not appear, unless they will be referenced by
1683 It is a bug for @value{GDBN} to behave one way when only a psymtab has
1684 been read, and another way if the corresponding symtab has been read
1685 in. Such bugs are typically caused by a psymtab that does not contain
1686 all the visible symbols, or which has the wrong instruction address
1689 The psymtab for a particular section of a symbol file (objfile) could be
1690 thrown away after the symtab has been read in. The symtab should always
1691 be searched before the psymtab, so the psymtab will never be used (in a
1692 bug-free environment). Currently, psymtabs are allocated on an obstack,
1693 and all the psymbols themselves are allocated in a pair of large arrays
1694 on an obstack, so there is little to be gained by trying to free them
1695 unless you want to do a lot more work.
1699 @unnumberedsubsec Fundamental Types (e.g., @code{FT_VOID}, @code{FT_BOOLEAN}).
1701 @cindex fundamental types
1702 These are the fundamental types that @value{GDBN} uses internally. Fundamental
1703 types from the various debugging formats (stabs, ELF, etc) are mapped
1704 into one of these. They are basically a union of all fundamental types
1705 that @value{GDBN} knows about for all the languages that @value{GDBN}
1708 @unnumberedsubsec Type Codes (e.g., @code{TYPE_CODE_PTR}, @code{TYPE_CODE_ARRAY}).
1711 Each time @value{GDBN} builds an internal type, it marks it with one
1712 of these types. The type may be a fundamental type, such as
1713 @code{TYPE_CODE_INT}, or a derived type, such as @code{TYPE_CODE_PTR}
1714 which is a pointer to another type. Typically, several @code{FT_*}
1715 types map to one @code{TYPE_CODE_*} type, and are distinguished by
1716 other members of the type struct, such as whether the type is signed
1717 or unsigned, and how many bits it uses.
1719 @unnumberedsubsec Builtin Types (e.g., @code{builtin_type_void}, @code{builtin_type_char}).
1721 These are instances of type structs that roughly correspond to
1722 fundamental types and are created as global types for @value{GDBN} to
1723 use for various ugly historical reasons. We eventually want to
1724 eliminate these. Note for example that @code{builtin_type_int}
1725 initialized in @file{gdbtypes.c} is basically the same as a
1726 @code{TYPE_CODE_INT} type that is initialized in @file{c-lang.c} for
1727 an @code{FT_INTEGER} fundamental type. The difference is that the
1728 @code{builtin_type} is not associated with any particular objfile, and
1729 only one instance exists, while @file{c-lang.c} builds as many
1730 @code{TYPE_CODE_INT} types as needed, with each one associated with
1731 some particular objfile.
1733 @section Object File Formats
1734 @cindex object file formats
1738 @cindex @code{a.out} format
1739 The @code{a.out} format is the original file format for Unix. It
1740 consists of three sections: @code{text}, @code{data}, and @code{bss},
1741 which are for program code, initialized data, and uninitialized data,
1744 The @code{a.out} format is so simple that it doesn't have any reserved
1745 place for debugging information. (Hey, the original Unix hackers used
1746 @samp{adb}, which is a machine-language debugger!) The only debugging
1747 format for @code{a.out} is stabs, which is encoded as a set of normal
1748 symbols with distinctive attributes.
1750 The basic @code{a.out} reader is in @file{dbxread.c}.
1755 The COFF format was introduced with System V Release 3 (SVR3) Unix.
1756 COFF files may have multiple sections, each prefixed by a header. The
1757 number of sections is limited.
1759 The COFF specification includes support for debugging. Although this
1760 was a step forward, the debugging information was woefully limited. For
1761 instance, it was not possible to represent code that came from an
1764 The COFF reader is in @file{coffread.c}.
1768 @cindex ECOFF format
1769 ECOFF is an extended COFF originally introduced for Mips and Alpha
1772 The basic ECOFF reader is in @file{mipsread.c}.
1776 @cindex XCOFF format
1777 The IBM RS/6000 running AIX uses an object file format called XCOFF.
1778 The COFF sections, symbols, and line numbers are used, but debugging
1779 symbols are @code{dbx}-style stabs whose strings are located in the
1780 @code{.debug} section (rather than the string table). For more
1781 information, see @ref{Top,,,stabs,The Stabs Debugging Format}.
1783 The shared library scheme has a clean interface for figuring out what
1784 shared libraries are in use, but the catch is that everything which
1785 refers to addresses (symbol tables and breakpoints at least) needs to be
1786 relocated for both shared libraries and the main executable. At least
1787 using the standard mechanism this can only be done once the program has
1788 been run (or the core file has been read).
1792 @cindex PE-COFF format
1793 Windows 95 and NT use the PE (@dfn{Portable Executable}) format for their
1794 executables. PE is basically COFF with additional headers.
1796 While BFD includes special PE support, @value{GDBN} needs only the basic
1802 The ELF format came with System V Release 4 (SVR4) Unix. ELF is similar
1803 to COFF in being organized into a number of sections, but it removes
1804 many of COFF's limitations.
1806 The basic ELF reader is in @file{elfread.c}.
1811 SOM is HP's object file and debug format (not to be confused with IBM's
1812 SOM, which is a cross-language ABI).
1814 The SOM reader is in @file{hpread.c}.
1816 @subsection Other File Formats
1818 @cindex Netware Loadable Module format
1819 Other file formats that have been supported by @value{GDBN} include Netware
1820 Loadable Modules (@file{nlmread.c}).
1822 @section Debugging File Formats
1824 This section describes characteristics of debugging information that
1825 are independent of the object file format.
1829 @cindex stabs debugging info
1830 @code{stabs} started out as special symbols within the @code{a.out}
1831 format. Since then, it has been encapsulated into other file
1832 formats, such as COFF and ELF.
1834 While @file{dbxread.c} does some of the basic stab processing,
1835 including for encapsulated versions, @file{stabsread.c} does
1840 @cindex COFF debugging info
1841 The basic COFF definition includes debugging information. The level
1842 of support is minimal and non-extensible, and is not often used.
1844 @subsection Mips debug (Third Eye)
1846 @cindex ECOFF debugging info
1847 ECOFF includes a definition of a special debug format.
1849 The file @file{mdebugread.c} implements reading for this format.
1853 @cindex DWARF 1 debugging info
1854 DWARF 1 is a debugging format that was originally designed to be
1855 used with ELF in SVR4 systems.
1860 @c If defined, these are the producer strings in a DWARF 1 file. All of
1861 @c these have reasonable defaults already.
1863 The DWARF 1 reader is in @file{dwarfread.c}.
1867 @cindex DWARF 2 debugging info
1868 DWARF 2 is an improved but incompatible version of DWARF 1.
1870 The DWARF 2 reader is in @file{dwarf2read.c}.
1874 @cindex SOM debugging info
1875 Like COFF, the SOM definition includes debugging information.
1877 @section Adding a New Symbol Reader to @value{GDBN}
1879 @cindex adding debugging info reader
1880 If you are using an existing object file format (@code{a.out}, COFF, ELF, etc),
1881 there is probably little to be done.
1883 If you need to add a new object file format, you must first add it to
1884 BFD. This is beyond the scope of this document.
1886 You must then arrange for the BFD code to provide access to the
1887 debugging symbols. Generally @value{GDBN} will have to call swapping routines
1888 from BFD and a few other BFD internal routines to locate the debugging
1889 information. As much as possible, @value{GDBN} should not depend on the BFD
1890 internal data structures.
1892 For some targets (e.g., COFF), there is a special transfer vector used
1893 to call swapping routines, since the external data structures on various
1894 platforms have different sizes and layouts. Specialized routines that
1895 will only ever be implemented by one object file format may be called
1896 directly. This interface should be described in a file
1897 @file{bfd/lib@var{xyz}.h}, which is included by @value{GDBN}.
1900 @node Language Support
1902 @chapter Language Support
1904 @cindex language support
1905 @value{GDBN}'s language support is mainly driven by the symbol reader,
1906 although it is possible for the user to set the source language
1909 @value{GDBN} chooses the source language by looking at the extension
1910 of the file recorded in the debug info; @file{.c} means C, @file{.f}
1911 means Fortran, etc. It may also use a special-purpose language
1912 identifier if the debug format supports it, like with DWARF.
1914 @section Adding a Source Language to @value{GDBN}
1916 @cindex adding source language
1917 To add other languages to @value{GDBN}'s expression parser, follow the
1921 @item Create the expression parser.
1923 @cindex expression parser
1924 This should reside in a file @file{@var{lang}-exp.y}. Routines for
1925 building parsed expressions into a @code{union exp_element} list are in
1928 @cindex language parser
1929 Since we can't depend upon everyone having Bison, and YACC produces
1930 parsers that define a bunch of global names, the following lines
1931 @strong{must} be included at the top of the YACC parser, to prevent the
1932 various parsers from defining the same global names:
1935 #define yyparse @var{lang}_parse
1936 #define yylex @var{lang}_lex
1937 #define yyerror @var{lang}_error
1938 #define yylval @var{lang}_lval
1939 #define yychar @var{lang}_char
1940 #define yydebug @var{lang}_debug
1941 #define yypact @var{lang}_pact
1942 #define yyr1 @var{lang}_r1
1943 #define yyr2 @var{lang}_r2
1944 #define yydef @var{lang}_def
1945 #define yychk @var{lang}_chk
1946 #define yypgo @var{lang}_pgo
1947 #define yyact @var{lang}_act
1948 #define yyexca @var{lang}_exca
1949 #define yyerrflag @var{lang}_errflag
1950 #define yynerrs @var{lang}_nerrs
1953 At the bottom of your parser, define a @code{struct language_defn} and
1954 initialize it with the right values for your language. Define an
1955 @code{initialize_@var{lang}} routine and have it call
1956 @samp{add_language(@var{lang}_language_defn)} to tell the rest of @value{GDBN}
1957 that your language exists. You'll need some other supporting variables
1958 and functions, which will be used via pointers from your
1959 @code{@var{lang}_language_defn}. See the declaration of @code{struct
1960 language_defn} in @file{language.h}, and the other @file{*-exp.y} files,
1961 for more information.
1963 @item Add any evaluation routines, if necessary
1965 @cindex expression evaluation routines
1966 @findex evaluate_subexp
1967 @findex prefixify_subexp
1968 @findex length_of_subexp
1969 If you need new opcodes (that represent the operations of the language),
1970 add them to the enumerated type in @file{expression.h}. Add support
1971 code for these operations in the @code{evaluate_subexp} function
1972 defined in the file @file{eval.c}. Add cases
1973 for new opcodes in two functions from @file{parse.c}:
1974 @code{prefixify_subexp} and @code{length_of_subexp}. These compute
1975 the number of @code{exp_element}s that a given operation takes up.
1977 @item Update some existing code
1979 Add an enumerated identifier for your language to the enumerated type
1980 @code{enum language} in @file{defs.h}.
1982 Update the routines in @file{language.c} so your language is included.
1983 These routines include type predicates and such, which (in some cases)
1984 are language dependent. If your language does not appear in the switch
1985 statement, an error is reported.
1987 @vindex current_language
1988 Also included in @file{language.c} is the code that updates the variable
1989 @code{current_language}, and the routines that translate the
1990 @code{language_@var{lang}} enumerated identifier into a printable
1993 @findex _initialize_language
1994 Update the function @code{_initialize_language} to include your
1995 language. This function picks the default language upon startup, so is
1996 dependent upon which languages that @value{GDBN} is built for.
1998 @findex allocate_symtab
1999 Update @code{allocate_symtab} in @file{symfile.c} and/or symbol-reading
2000 code so that the language of each symtab (source file) is set properly.
2001 This is used to determine the language to use at each stack frame level.
2002 Currently, the language is set based upon the extension of the source
2003 file. If the language can be better inferred from the symbol
2004 information, please set the language of the symtab in the symbol-reading
2007 @findex print_subexp
2008 @findex op_print_tab
2009 Add helper code to @code{print_subexp} (in @file{expprint.c}) to handle any new
2010 expression opcodes you have added to @file{expression.h}. Also, add the
2011 printed representations of your operators to @code{op_print_tab}.
2013 @item Add a place of call
2016 Add a call to @code{@var{lang}_parse()} and @code{@var{lang}_error} in
2017 @code{parse_exp_1} (defined in @file{parse.c}).
2019 @item Use macros to trim code
2021 @cindex trimming language-dependent code
2022 The user has the option of building @value{GDBN} for some or all of the
2023 languages. If the user decides to build @value{GDBN} for the language
2024 @var{lang}, then every file dependent on @file{language.h} will have the
2025 macro @code{_LANG_@var{lang}} defined in it. Use @code{#ifdef}s to
2026 leave out large routines that the user won't need if he or she is not
2027 using your language.
2029 Note that you do not need to do this in your YACC parser, since if @value{GDBN}
2030 is not build for @var{lang}, then @file{@var{lang}-exp.tab.o} (the
2031 compiled form of your parser) is not linked into @value{GDBN} at all.
2033 See the file @file{configure.in} for how @value{GDBN} is configured
2034 for different languages.
2036 @item Edit @file{Makefile.in}
2038 Add dependencies in @file{Makefile.in}. Make sure you update the macro
2039 variables such as @code{HFILES} and @code{OBJS}, otherwise your code may
2040 not get linked in, or, worse yet, it may not get @code{tar}red into the
2045 @node Host Definition
2047 @chapter Host Definition
2049 With the advent of Autoconf, it's rarely necessary to have host
2050 definition machinery anymore. The following information is provided,
2051 mainly, as an historical reference.
2053 @section Adding a New Host
2055 @cindex adding a new host
2056 @cindex host, adding
2057 @value{GDBN}'s host configuration support normally happens via Autoconf.
2058 New host-specific definitions should not be needed. Older hosts
2059 @value{GDBN} still use the host-specific definitions and files listed
2060 below, but these mostly exist for historical reasons, and will
2061 eventually disappear.
2064 @item gdb/config/@var{arch}/@var{xyz}.mh
2065 This file once contained both host and native configuration information
2066 (@pxref{Native Debugging}) for the machine @var{xyz}. The host
2067 configuration information is now handed by Autoconf.
2069 Host configuration information included a definition of
2070 @code{XM_FILE=xm-@var{xyz}.h} and possibly definitions for @code{CC},
2071 @code{SYSV_DEFINE}, @code{XM_CFLAGS}, @code{XM_ADD_FILES},
2072 @code{XM_CLIBS}, @code{XM_CDEPS}, etc.; see @file{Makefile.in}.
2074 New host only configurations do not need this file.
2076 @item gdb/config/@var{arch}/xm-@var{xyz}.h
2077 This file once contained definitions and includes required when hosting
2078 gdb on machine @var{xyz}. Those definitions and includes are now
2079 handled by Autoconf.
2081 New host and native configurations do not need this file.
2083 @emph{Maintainer's note: Some hosts continue to use the @file{xm-xyz.h}
2084 file to define the macros @var{HOST_FLOAT_FORMAT},
2085 @var{HOST_DOUBLE_FORMAT} and @var{HOST_LONG_DOUBLE_FORMAT}. That code
2086 also needs to be replaced with either an Autoconf or run-time test.}
2090 @subheading Generic Host Support Files
2092 @cindex generic host support
2093 There are some ``generic'' versions of routines that can be used by
2094 various systems. These can be customized in various ways by macros
2095 defined in your @file{xm-@var{xyz}.h} file. If these routines work for
2096 the @var{xyz} host, you can just include the generic file's name (with
2097 @samp{.o}, not @samp{.c}) in @code{XDEPFILES}.
2099 Otherwise, if your machine needs custom support routines, you will need
2100 to write routines that perform the same functions as the generic file.
2101 Put them into @code{@var{xyz}-xdep.c}, and put @code{@var{xyz}-xdep.o}
2102 into @code{XDEPFILES}.
2105 @cindex remote debugging support
2106 @cindex serial line support
2108 This contains serial line support for Unix systems. This is always
2109 included, via the makefile variable @code{SER_HARDWIRE}; override this
2110 variable in the @file{.mh} file to avoid it.
2113 This contains serial line support for 32-bit programs running under DOS,
2114 using the DJGPP (a.k.a.@: GO32) execution environment.
2116 @cindex TCP remote support
2118 This contains generic TCP support using sockets.
2121 @section Host Conditionals
2123 When @value{GDBN} is configured and compiled, various macros are
2124 defined or left undefined, to control compilation based on the
2125 attributes of the host system. These macros and their meanings (or if
2126 the meaning is not documented here, then one of the source files where
2127 they are used is indicated) are:
2130 @item @value{GDBN}INIT_FILENAME
2131 The default name of @value{GDBN}'s initialization file (normally
2135 This macro is deprecated.
2137 @item SIGWINCH_HANDLER
2138 If your host defines @code{SIGWINCH}, you can define this to be the name
2139 of a function to be called if @code{SIGWINCH} is received.
2141 @item SIGWINCH_HANDLER_BODY
2142 Define this to expand into code that will define the function named by
2143 the expansion of @code{SIGWINCH_HANDLER}.
2145 @item ALIGN_STACK_ON_STARTUP
2146 @cindex stack alignment
2147 Define this if your system is of a sort that will crash in
2148 @code{tgetent} if the stack happens not to be longword-aligned when
2149 @code{main} is called. This is a rare situation, but is known to occur
2150 on several different types of systems.
2152 @item CRLF_SOURCE_FILES
2153 @cindex DOS text files
2154 Define this if host files use @code{\r\n} rather than @code{\n} as a
2155 line terminator. This will cause source file listings to omit @code{\r}
2156 characters when printing and it will allow @code{\r\n} line endings of files
2157 which are ``sourced'' by gdb. It must be possible to open files in binary
2158 mode using @code{O_BINARY} or, for fopen, @code{"rb"}.
2160 @item DEFAULT_PROMPT
2162 The default value of the prompt string (normally @code{"(gdb) "}).
2165 @cindex terminal device
2166 The name of the generic TTY device, defaults to @code{"/dev/tty"}.
2168 @item FCLOSE_PROVIDED
2169 Define this if the system declares @code{fclose} in the headers included
2170 in @code{defs.h}. This isn't needed unless your compiler is unusually
2174 Define this if binary files are opened the same way as text files.
2176 @item GETENV_PROVIDED
2177 Define this if the system declares @code{getenv} in its headers included
2178 in @code{defs.h}. This isn't needed unless your compiler is unusually
2183 In some cases, use the system call @code{mmap} for reading symbol
2184 tables. For some machines this allows for sharing and quick updates.
2187 Define this if the host system has @code{termio.h}.
2194 Values for host-side constants.
2197 Substitute for isatty, if not available.
2200 This is the longest integer type available on the host. If not defined,
2201 it will default to @code{long long} or @code{long}, depending on
2202 @code{CC_HAS_LONG_LONG}.
2204 @item CC_HAS_LONG_LONG
2205 @cindex @code{long long} data type
2206 Define this if the host C compiler supports @code{long long}. This is set
2207 by the @code{configure} script.
2209 @item PRINTF_HAS_LONG_LONG
2210 Define this if the host can handle printing of long long integers via
2211 the printf format conversion specifier @code{ll}. This is set by the
2212 @code{configure} script.
2214 @item HAVE_LONG_DOUBLE
2215 Define this if the host C compiler supports @code{long double}. This is
2216 set by the @code{configure} script.
2218 @item PRINTF_HAS_LONG_DOUBLE
2219 Define this if the host can handle printing of long double float-point
2220 numbers via the printf format conversion specifier @code{Lg}. This is
2221 set by the @code{configure} script.
2223 @item SCANF_HAS_LONG_DOUBLE
2224 Define this if the host can handle the parsing of long double
2225 float-point numbers via the scanf format conversion specifier
2226 @code{Lg}. This is set by the @code{configure} script.
2228 @item LSEEK_NOT_LINEAR
2229 Define this if @code{lseek (n)} does not necessarily move to byte number
2230 @code{n} in the file. This is only used when reading source files. It
2231 is normally faster to define @code{CRLF_SOURCE_FILES} when possible.
2234 This macro is used as the argument to @code{lseek} (or, most commonly,
2235 @code{bfd_seek}). FIXME, should be replaced by SEEK_SET instead,
2236 which is the POSIX equivalent.
2239 If defined, this should be one or more tokens, such as @code{volatile},
2240 that can be used in both the declaration and definition of functions to
2241 indicate that they never return. The default is already set correctly
2242 if compiling with GCC. This will almost never need to be defined.
2245 If defined, this should be one or more tokens, such as
2246 @code{__attribute__ ((noreturn))}, that can be used in the declarations
2247 of functions to indicate that they never return. The default is already
2248 set correctly if compiling with GCC. This will almost never need to be
2253 Define these to appropriate value for the system @code{lseek}, if not already
2257 This is the signal for stopping @value{GDBN}. Defaults to
2258 @code{SIGTSTP}. (Only redefined for the Convex.)
2261 Means that System V (prior to SVR4) include files are in use. (FIXME:
2262 This symbol is abused in @file{infrun.c}, @file{regex.c}, and
2263 @file{utils.c} for other things, at the moment.)
2266 Define this to help placate @code{lint} in some situations.
2269 Define this to override the defaults of @code{__volatile__} or
2274 @node Target Architecture Definition
2276 @chapter Target Architecture Definition
2278 @cindex target architecture definition
2279 @value{GDBN}'s target architecture defines what sort of
2280 machine-language programs @value{GDBN} can work with, and how it works
2283 The target architecture object is implemented as the C structure
2284 @code{struct gdbarch *}. The structure, and its methods, are generated
2285 using the Bourne shell script @file{gdbarch.sh}.
2287 @section Operating System ABI Variant Handling
2288 @cindex OS ABI variants
2290 @value{GDBN} provides a mechanism for handling variations in OS
2291 ABIs. An OS ABI variant may have influence over any number of
2292 variables in the target architecture definition. There are two major
2293 components in the OS ABI mechanism: sniffers and handlers.
2295 A @dfn{sniffer} examines a file matching a BFD architecture/flavour pair
2296 (the architecture may be wildcarded) in an attempt to determine the
2297 OS ABI of that file. Sniffers with a wildcarded architecture are considered
2298 to be @dfn{generic}, while sniffers for a specific architecture are
2299 considered to be @dfn{specific}. A match from a specific sniffer
2300 overrides a match from a generic sniffer. Multiple sniffers for an
2301 architecture/flavour may exist, in order to differentiate between two
2302 different operating systems which use the same basic file format. The
2303 OS ABI framework provides a generic sniffer for ELF-format files which
2304 examines the @code{EI_OSABI} field of the ELF header, as well as note
2305 sections known to be used by several operating systems.
2307 @cindex fine-tuning @code{gdbarch} structure
2308 A @dfn{handler} is used to fine-tune the @code{gdbarch} structure for the
2309 selected OS ABI. There may be only one handler for a given OS ABI
2310 for each BFD architecture.
2312 The following OS ABI variants are defined in @file{osabi.h}:
2316 @findex GDB_OSABI_UNKNOWN
2317 @item GDB_OSABI_UNKNOWN
2318 The ABI of the inferior is unknown. The default @code{gdbarch}
2319 settings for the architecture will be used.
2321 @findex GDB_OSABI_SVR4
2322 @item GDB_OSABI_SVR4
2323 UNIX System V Release 4
2325 @findex GDB_OSABI_HURD
2326 @item GDB_OSABI_HURD
2327 GNU using the Hurd kernel
2329 @findex GDB_OSABI_SOLARIS
2330 @item GDB_OSABI_SOLARIS
2333 @findex GDB_OSABI_OSF1
2334 @item GDB_OSABI_OSF1
2335 OSF/1, including Digital UNIX and Compaq Tru64 UNIX
2337 @findex GDB_OSABI_LINUX
2338 @item GDB_OSABI_LINUX
2339 GNU using the Linux kernel
2341 @findex GDB_OSABI_FREEBSD_AOUT
2342 @item GDB_OSABI_FREEBSD_AOUT
2343 FreeBSD using the a.out executable format
2345 @findex GDB_OSABI_FREEBSD_ELF
2346 @item GDB_OSABI_FREEBSD_ELF
2347 FreeBSD using the ELF executable format
2349 @findex GDB_OSABI_NETBSD_AOUT
2350 @item GDB_OSABI_NETBSD_AOUT
2351 NetBSD using the a.out executable format
2353 @findex GDB_OSABI_NETBSD_ELF
2354 @item GDB_OSABI_NETBSD_ELF
2355 NetBSD using the ELF executable format
2357 @findex GDB_OSABI_WINCE
2358 @item GDB_OSABI_WINCE
2361 @findex GDB_OSABI_GO32
2362 @item GDB_OSABI_GO32
2365 @findex GDB_OSABI_NETWARE
2366 @item GDB_OSABI_NETWARE
2369 @findex GDB_OSABI_ARM_EABI_V1
2370 @item GDB_OSABI_ARM_EABI_V1
2371 ARM Embedded ABI version 1
2373 @findex GDB_OSABI_ARM_EABI_V2
2374 @item GDB_OSABI_ARM_EABI_V2
2375 ARM Embedded ABI version 2
2377 @findex GDB_OSABI_ARM_APCS
2378 @item GDB_OSABI_ARM_APCS
2379 Generic ARM Procedure Call Standard
2383 Here are the functions that make up the OS ABI framework:
2385 @deftypefun const char *gdbarch_osabi_name (enum gdb_osabi @var{osabi})
2386 Return the name of the OS ABI corresponding to @var{osabi}.
2389 @deftypefun void gdbarch_register_osabi (enum bfd_architecture @var{arch}, unsigned long @var{machine}, enum gdb_osabi @var{osabi}, void (*@var{init_osabi})(struct gdbarch_info @var{info}, struct gdbarch *@var{gdbarch}))
2390 Register the OS ABI handler specified by @var{init_osabi} for the
2391 architecture, machine type and OS ABI specified by @var{arch},
2392 @var{machine} and @var{osabi}. In most cases, a value of zero for the
2393 machine type, which implies the architecture's default machine type,
2397 @deftypefun void gdbarch_register_osabi_sniffer (enum bfd_architecture @var{arch}, enum bfd_flavour @var{flavour}, enum gdb_osabi (*@var{sniffer})(bfd *@var{abfd}))
2398 Register the OS ABI file sniffer specified by @var{sniffer} for the
2399 BFD architecture/flavour pair specified by @var{arch} and @var{flavour}.
2400 If @var{arch} is @code{bfd_arch_unknown}, the sniffer is considered to
2401 be generic, and is allowed to examine @var{flavour}-flavoured files for
2405 @deftypefun enum gdb_osabi gdbarch_lookup_osabi (bfd *@var{abfd})
2406 Examine the file described by @var{abfd} to determine its OS ABI.
2407 The value @code{GDB_OSABI_UNKNOWN} is returned if the OS ABI cannot
2411 @deftypefun void gdbarch_init_osabi (struct gdbarch info @var{info}, struct gdbarch *@var{gdbarch}, enum gdb_osabi @var{osabi})
2412 Invoke the OS ABI handler corresponding to @var{osabi} to fine-tune the
2413 @code{gdbarch} structure specified by @var{gdbarch}. If a handler
2414 corresponding to @var{osabi} has not been registered for @var{gdbarch}'s
2415 architecture, a warning will be issued and the debugging session will continue
2416 with the defaults already established for @var{gdbarch}.
2419 @section Registers and Memory
2421 @value{GDBN}'s model of the target machine is rather simple.
2422 @value{GDBN} assumes the machine includes a bank of registers and a
2423 block of memory. Each register may have a different size.
2425 @value{GDBN} does not have a magical way to match up with the
2426 compiler's idea of which registers are which; however, it is critical
2427 that they do match up accurately. The only way to make this work is
2428 to get accurate information about the order that the compiler uses,
2429 and to reflect that in the @code{REGISTER_NAME} and related macros.
2431 @value{GDBN} can handle big-endian, little-endian, and bi-endian architectures.
2433 @section Pointers Are Not Always Addresses
2434 @cindex pointer representation
2435 @cindex address representation
2436 @cindex word-addressed machines
2437 @cindex separate data and code address spaces
2438 @cindex spaces, separate data and code address
2439 @cindex address spaces, separate data and code
2440 @cindex code pointers, word-addressed
2441 @cindex converting between pointers and addresses
2442 @cindex D10V addresses
2444 On almost all 32-bit architectures, the representation of a pointer is
2445 indistinguishable from the representation of some fixed-length number
2446 whose value is the byte address of the object pointed to. On such
2447 machines, the words ``pointer'' and ``address'' can be used interchangeably.
2448 However, architectures with smaller word sizes are often cramped for
2449 address space, so they may choose a pointer representation that breaks this
2450 identity, and allows a larger code address space.
2452 For example, the Renesas D10V is a 16-bit VLIW processor whose
2453 instructions are 32 bits long@footnote{Some D10V instructions are
2454 actually pairs of 16-bit sub-instructions. However, since you can't
2455 jump into the middle of such a pair, code addresses can only refer to
2456 full 32 bit instructions, which is what matters in this explanation.}.
2457 If the D10V used ordinary byte addresses to refer to code locations,
2458 then the processor would only be able to address 64kb of instructions.
2459 However, since instructions must be aligned on four-byte boundaries, the
2460 low two bits of any valid instruction's byte address are always
2461 zero---byte addresses waste two bits. So instead of byte addresses,
2462 the D10V uses word addresses---byte addresses shifted right two bits---to
2463 refer to code. Thus, the D10V can use 16-bit words to address 256kb of
2466 However, this means that code pointers and data pointers have different
2467 forms on the D10V. The 16-bit word @code{0xC020} refers to byte address
2468 @code{0xC020} when used as a data address, but refers to byte address
2469 @code{0x30080} when used as a code address.
2471 (The D10V also uses separate code and data address spaces, which also
2472 affects the correspondence between pointers and addresses, but we're
2473 going to ignore that here; this example is already too long.)
2475 To cope with architectures like this---the D10V is not the only
2476 one!---@value{GDBN} tries to distinguish between @dfn{addresses}, which are
2477 byte numbers, and @dfn{pointers}, which are the target's representation
2478 of an address of a particular type of data. In the example above,
2479 @code{0xC020} is the pointer, which refers to one of the addresses
2480 @code{0xC020} or @code{0x30080}, depending on the type imposed upon it.
2481 @value{GDBN} provides functions for turning a pointer into an address
2482 and vice versa, in the appropriate way for the current architecture.
2484 Unfortunately, since addresses and pointers are identical on almost all
2485 processors, this distinction tends to bit-rot pretty quickly. Thus,
2486 each time you port @value{GDBN} to an architecture which does
2487 distinguish between pointers and addresses, you'll probably need to
2488 clean up some architecture-independent code.
2490 Here are functions which convert between pointers and addresses:
2492 @deftypefun CORE_ADDR extract_typed_address (void *@var{buf}, struct type *@var{type})
2493 Treat the bytes at @var{buf} as a pointer or reference of type
2494 @var{type}, and return the address it represents, in a manner
2495 appropriate for the current architecture. This yields an address
2496 @value{GDBN} can use to read target memory, disassemble, etc. Note that
2497 @var{buf} refers to a buffer in @value{GDBN}'s memory, not the
2500 For example, if the current architecture is the Intel x86, this function
2501 extracts a little-endian integer of the appropriate length from
2502 @var{buf} and returns it. However, if the current architecture is the
2503 D10V, this function will return a 16-bit integer extracted from
2504 @var{buf}, multiplied by four if @var{type} is a pointer to a function.
2506 If @var{type} is not a pointer or reference type, then this function
2507 will signal an internal error.
2510 @deftypefun CORE_ADDR store_typed_address (void *@var{buf}, struct type *@var{type}, CORE_ADDR @var{addr})
2511 Store the address @var{addr} in @var{buf}, in the proper format for a
2512 pointer of type @var{type} in the current architecture. Note that
2513 @var{buf} refers to a buffer in @value{GDBN}'s memory, not the
2516 For example, if the current architecture is the Intel x86, this function
2517 stores @var{addr} unmodified as a little-endian integer of the
2518 appropriate length in @var{buf}. However, if the current architecture
2519 is the D10V, this function divides @var{addr} by four if @var{type} is
2520 a pointer to a function, and then stores it in @var{buf}.
2522 If @var{type} is not a pointer or reference type, then this function
2523 will signal an internal error.
2526 @deftypefun CORE_ADDR value_as_address (struct value *@var{val})
2527 Assuming that @var{val} is a pointer, return the address it represents,
2528 as appropriate for the current architecture.
2530 This function actually works on integral values, as well as pointers.
2531 For pointers, it performs architecture-specific conversions as
2532 described above for @code{extract_typed_address}.
2535 @deftypefun CORE_ADDR value_from_pointer (struct type *@var{type}, CORE_ADDR @var{addr})
2536 Create and return a value representing a pointer of type @var{type} to
2537 the address @var{addr}, as appropriate for the current architecture.
2538 This function performs architecture-specific conversions as described
2539 above for @code{store_typed_address}.
2542 Here are some macros which architectures can define to indicate the
2543 relationship between pointers and addresses. These have default
2544 definitions, appropriate for architectures on which all pointers are
2545 simple unsigned byte addresses.
2547 @deftypefn {Target Macro} CORE_ADDR POINTER_TO_ADDRESS (struct type *@var{type}, char *@var{buf})
2548 Assume that @var{buf} holds a pointer of type @var{type}, in the
2549 appropriate format for the current architecture. Return the byte
2550 address the pointer refers to.
2552 This function may safely assume that @var{type} is either a pointer or a
2553 C@t{++} reference type.
2556 @deftypefn {Target Macro} void ADDRESS_TO_POINTER (struct type *@var{type}, char *@var{buf}, CORE_ADDR @var{addr})
2557 Store in @var{buf} a pointer of type @var{type} representing the address
2558 @var{addr}, in the appropriate format for the current architecture.
2560 This function may safely assume that @var{type} is either a pointer or a
2561 C@t{++} reference type.
2564 @section Address Classes
2565 @cindex address classes
2566 @cindex DW_AT_byte_size
2567 @cindex DW_AT_address_class
2569 Sometimes information about different kinds of addresses is available
2570 via the debug information. For example, some programming environments
2571 define addresses of several different sizes. If the debug information
2572 distinguishes these kinds of address classes through either the size
2573 info (e.g, @code{DW_AT_byte_size} in @w{DWARF 2}) or through an explicit
2574 address class attribute (e.g, @code{DW_AT_address_class} in @w{DWARF 2}), the
2575 following macros should be defined in order to disambiguate these
2576 types within @value{GDBN} as well as provide the added information to
2577 a @value{GDBN} user when printing type expressions.
2579 @deftypefn {Target Macro} int ADDRESS_CLASS_TYPE_FLAGS (int @var{byte_size}, int @var{dwarf2_addr_class})
2580 Returns the type flags needed to construct a pointer type whose size
2581 is @var{byte_size} and whose address class is @var{dwarf2_addr_class}.
2582 This function is normally called from within a symbol reader. See
2583 @file{dwarf2read.c}.
2586 @deftypefn {Target Macro} char *ADDRESS_CLASS_TYPE_FLAGS_TO_NAME (int @var{type_flags})
2587 Given the type flags representing an address class qualifier, return
2590 @deftypefn {Target Macro} int ADDRESS_CLASS_NAME_to_TYPE_FLAGS (int @var{name}, int *var{type_flags_ptr})
2591 Given an address qualifier name, set the @code{int} refererenced by @var{type_flags_ptr} to the type flags
2592 for that address class qualifier.
2595 Since the need for address classes is rather rare, none of
2596 the address class macros defined by default. Predicate
2597 macros are provided to detect when they are defined.
2599 Consider a hypothetical architecture in which addresses are normally
2600 32-bits wide, but 16-bit addresses are also supported. Furthermore,
2601 suppose that the @w{DWARF 2} information for this architecture simply
2602 uses a @code{DW_AT_byte_size} value of 2 to indicate the use of one
2603 of these "short" pointers. The following functions could be defined
2604 to implement the address class macros:
2607 somearch_address_class_type_flags (int byte_size,
2608 int dwarf2_addr_class)
2611 return TYPE_FLAG_ADDRESS_CLASS_1;
2617 somearch_address_class_type_flags_to_name (int type_flags)
2619 if (type_flags & TYPE_FLAG_ADDRESS_CLASS_1)
2626 somearch_address_class_name_to_type_flags (char *name,
2627 int *type_flags_ptr)
2629 if (strcmp (name, "short") == 0)
2631 *type_flags_ptr = TYPE_FLAG_ADDRESS_CLASS_1;
2639 The qualifier @code{@@short} is used in @value{GDBN}'s type expressions
2640 to indicate the presence of one of these "short" pointers. E.g, if
2641 the debug information indicates that @code{short_ptr_var} is one of these
2642 short pointers, @value{GDBN} might show the following behavior:
2645 (gdb) ptype short_ptr_var
2646 type = int * @@short
2650 @section Raw and Virtual Register Representations
2651 @cindex raw register representation
2652 @cindex virtual register representation
2653 @cindex representations, raw and virtual registers
2655 @emph{Maintainer note: This section is pretty much obsolete. The
2656 functionality described here has largely been replaced by
2657 pseudo-registers and the mechanisms described in @ref{Target
2658 Architecture Definition, , Using Different Register and Memory Data
2659 Representations}. See also @uref{http://www.gnu.org/software/gdb/bugs/,
2660 Bug Tracking Database} and
2661 @uref{http://sources.redhat.com/gdb/current/ari/, ARI Index} for more
2662 up-to-date information.}
2664 Some architectures use one representation for a value when it lives in a
2665 register, but use a different representation when it lives in memory.
2666 In @value{GDBN}'s terminology, the @dfn{raw} representation is the one used in
2667 the target registers, and the @dfn{virtual} representation is the one
2668 used in memory, and within @value{GDBN} @code{struct value} objects.
2670 @emph{Maintainer note: Notice that the same mechanism is being used to
2671 both convert a register to a @code{struct value} and alternative
2674 For almost all data types on almost all architectures, the virtual and
2675 raw representations are identical, and no special handling is needed.
2676 However, they do occasionally differ. For example:
2680 The x86 architecture supports an 80-bit @code{long double} type. However, when
2681 we store those values in memory, they occupy twelve bytes: the
2682 floating-point number occupies the first ten, and the final two bytes
2683 are unused. This keeps the values aligned on four-byte boundaries,
2684 allowing more efficient access. Thus, the x86 80-bit floating-point
2685 type is the raw representation, and the twelve-byte loosely-packed
2686 arrangement is the virtual representation.
2689 Some 64-bit MIPS targets present 32-bit registers to @value{GDBN} as 64-bit
2690 registers, with garbage in their upper bits. @value{GDBN} ignores the top 32
2691 bits. Thus, the 64-bit form, with garbage in the upper 32 bits, is the
2692 raw representation, and the trimmed 32-bit representation is the
2693 virtual representation.
2696 In general, the raw representation is determined by the architecture, or
2697 @value{GDBN}'s interface to the architecture, while the virtual representation
2698 can be chosen for @value{GDBN}'s convenience. @value{GDBN}'s register file,
2699 @code{registers}, holds the register contents in raw format, and the
2700 @value{GDBN} remote protocol transmits register values in raw format.
2702 Your architecture may define the following macros to request
2703 conversions between the raw and virtual format:
2705 @deftypefn {Target Macro} int REGISTER_CONVERTIBLE (int @var{reg})
2706 Return non-zero if register number @var{reg}'s value needs different raw
2707 and virtual formats.
2709 You should not use @code{REGISTER_CONVERT_TO_VIRTUAL} for a register
2710 unless this macro returns a non-zero value for that register.
2713 @deftypefn {Target Macro} int DEPRECATED_REGISTER_RAW_SIZE (int @var{reg})
2714 The size of register number @var{reg}'s raw value. This is the number
2715 of bytes the register will occupy in @code{registers}, or in a @value{GDBN}
2716 remote protocol packet.
2719 @deftypefn {Target Macro} int DEPRECATED_REGISTER_VIRTUAL_SIZE (int @var{reg})
2720 The size of register number @var{reg}'s value, in its virtual format.
2721 This is the size a @code{struct value}'s buffer will have, holding that
2725 @deftypefn {Target Macro} struct type *DEPRECATED_REGISTER_VIRTUAL_TYPE (int @var{reg})
2726 This is the type of the virtual representation of register number
2727 @var{reg}. Note that there is no need for a macro giving a type for the
2728 register's raw form; once the register's value has been obtained, @value{GDBN}
2729 always uses the virtual form.
2732 @deftypefn {Target Macro} void REGISTER_CONVERT_TO_VIRTUAL (int @var{reg}, struct type *@var{type}, char *@var{from}, char *@var{to})
2733 Convert the value of register number @var{reg} to @var{type}, which
2734 should always be @code{DEPRECATED_REGISTER_VIRTUAL_TYPE (@var{reg})}. The buffer
2735 at @var{from} holds the register's value in raw format; the macro should
2736 convert the value to virtual format, and place it at @var{to}.
2738 Note that @code{REGISTER_CONVERT_TO_VIRTUAL} and
2739 @code{REGISTER_CONVERT_TO_RAW} take their @var{reg} and @var{type}
2740 arguments in different orders.
2742 You should only use @code{REGISTER_CONVERT_TO_VIRTUAL} with registers
2743 for which the @code{REGISTER_CONVERTIBLE} macro returns a non-zero
2747 @deftypefn {Target Macro} void REGISTER_CONVERT_TO_RAW (struct type *@var{type}, int @var{reg}, char *@var{from}, char *@var{to})
2748 Convert the value of register number @var{reg} to @var{type}, which
2749 should always be @code{DEPRECATED_REGISTER_VIRTUAL_TYPE (@var{reg})}. The buffer
2750 at @var{from} holds the register's value in raw format; the macro should
2751 convert the value to virtual format, and place it at @var{to}.
2753 Note that REGISTER_CONVERT_TO_VIRTUAL and REGISTER_CONVERT_TO_RAW take
2754 their @var{reg} and @var{type} arguments in different orders.
2758 @section Using Different Register and Memory Data Representations
2759 @cindex register representation
2760 @cindex memory representation
2761 @cindex representations, register and memory
2762 @cindex register data formats, converting
2763 @cindex @code{struct value}, converting register contents to
2765 @emph{Maintainer's note: The way GDB manipulates registers is undergoing
2766 significant change. Many of the macros and functions refered to in this
2767 section are likely to be subject to further revision. See
2768 @uref{http://sources.redhat.com/gdb/current/ari/, A.R. Index} and
2769 @uref{http://www.gnu.org/software/gdb/bugs, Bug Tracking Database} for
2770 further information. cagney/2002-05-06.}
2772 Some architectures can represent a data object in a register using a
2773 form that is different to the objects more normal memory representation.
2779 The Alpha architecture can represent 32 bit integer values in
2780 floating-point registers.
2783 The x86 architecture supports 80-bit floating-point registers. The
2784 @code{long double} data type occupies 96 bits in memory but only 80 bits
2785 when stored in a register.
2789 In general, the register representation of a data type is determined by
2790 the architecture, or @value{GDBN}'s interface to the architecture, while
2791 the memory representation is determined by the Application Binary
2794 For almost all data types on almost all architectures, the two
2795 representations are identical, and no special handling is needed.
2796 However, they do occasionally differ. Your architecture may define the
2797 following macros to request conversions between the register and memory
2798 representations of a data type:
2800 @deftypefn {Target Macro} int CONVERT_REGISTER_P (int @var{reg})
2801 Return non-zero if the representation of a data value stored in this
2802 register may be different to the representation of that same data value
2803 when stored in memory.
2805 When non-zero, the macros @code{REGISTER_TO_VALUE} and
2806 @code{VALUE_TO_REGISTER} are used to perform any necessary conversion.
2809 @deftypefn {Target Macro} void REGISTER_TO_VALUE (int @var{reg}, struct type *@var{type}, char *@var{from}, char *@var{to})
2810 Convert the value of register number @var{reg} to a data object of type
2811 @var{type}. The buffer at @var{from} holds the register's value in raw
2812 format; the converted value should be placed in the buffer at @var{to}.
2814 Note that @code{REGISTER_TO_VALUE} and @code{VALUE_TO_REGISTER} take
2815 their @var{reg} and @var{type} arguments in different orders.
2817 You should only use @code{REGISTER_TO_VALUE} with registers for which
2818 the @code{CONVERT_REGISTER_P} macro returns a non-zero value.
2821 @deftypefn {Target Macro} void VALUE_TO_REGISTER (struct type *@var{type}, int @var{reg}, char *@var{from}, char *@var{to})
2822 Convert a data value of type @var{type} to register number @var{reg}'
2825 Note that @code{REGISTER_TO_VALUE} and @code{VALUE_TO_REGISTER} take
2826 their @var{reg} and @var{type} arguments in different orders.
2828 You should only use @code{VALUE_TO_REGISTER} with registers for which
2829 the @code{CONVERT_REGISTER_P} macro returns a non-zero value.
2832 @deftypefn {Target Macro} void REGISTER_CONVERT_TO_TYPE (int @var{regnum}, struct type *@var{type}, char *@var{buf})
2833 See @file{mips-tdep.c}. It does not do what you want.
2837 @section Frame Interpretation
2839 @section Inferior Call Setup
2841 @section Compiler Characteristics
2843 @section Target Conditionals
2845 This section describes the macros that you can use to define the target
2850 @item ADDR_BITS_REMOVE (addr)
2851 @findex ADDR_BITS_REMOVE
2852 If a raw machine instruction address includes any bits that are not
2853 really part of the address, then define this macro to expand into an
2854 expression that zeroes those bits in @var{addr}. This is only used for
2855 addresses of instructions, and even then not in all contexts.
2857 For example, the two low-order bits of the PC on the Hewlett-Packard PA
2858 2.0 architecture contain the privilege level of the corresponding
2859 instruction. Since instructions must always be aligned on four-byte
2860 boundaries, the processor masks out these bits to generate the actual
2861 address of the instruction. ADDR_BITS_REMOVE should filter out these
2862 bits with an expression such as @code{((addr) & ~3)}.
2864 @item ADDRESS_CLASS_NAME_TO_TYPE_FLAGS (@var{name}, @var{type_flags_ptr})
2865 @findex ADDRESS_CLASS_NAME_TO_TYPE_FLAGS
2866 If @var{name} is a valid address class qualifier name, set the @code{int}
2867 referenced by @var{type_flags_ptr} to the mask representing the qualifier
2868 and return 1. If @var{name} is not a valid address class qualifier name,
2871 The value for @var{type_flags_ptr} should be one of
2872 @code{TYPE_FLAG_ADDRESS_CLASS_1}, @code{TYPE_FLAG_ADDRESS_CLASS_2}, or
2873 possibly some combination of these values or'd together.
2874 @xref{Target Architecture Definition, , Address Classes}.
2876 @item ADDRESS_CLASS_NAME_TO_TYPE_FLAGS_P ()
2877 @findex ADDRESS_CLASS_NAME_TO_TYPE_FLAGS_P
2878 Predicate which indicates whether @code{ADDRESS_CLASS_NAME_TO_TYPE_FLAGS}
2881 @item ADDRESS_CLASS_TYPE_FLAGS (@var{byte_size}, @var{dwarf2_addr_class})
2882 @findex ADDRESS_CLASS_TYPE_FLAGS (@var{byte_size}, @var{dwarf2_addr_class})
2883 Given a pointers byte size (as described by the debug information) and
2884 the possible @code{DW_AT_address_class} value, return the type flags
2885 used by @value{GDBN} to represent this address class. The value
2886 returned should be one of @code{TYPE_FLAG_ADDRESS_CLASS_1},
2887 @code{TYPE_FLAG_ADDRESS_CLASS_2}, or possibly some combination of these
2888 values or'd together.
2889 @xref{Target Architecture Definition, , Address Classes}.
2891 @item ADDRESS_CLASS_TYPE_FLAGS_P ()
2892 @findex ADDRESS_CLASS_TYPE_FLAGS_P
2893 Predicate which indicates whether @code{ADDRESS_CLASS_TYPE_FLAGS} has
2896 @item ADDRESS_CLASS_TYPE_FLAGS_TO_NAME (@var{type_flags})
2897 @findex ADDRESS_CLASS_TYPE_FLAGS_TO_NAME
2898 Return the name of the address class qualifier associated with the type
2899 flags given by @var{type_flags}.
2901 @item ADDRESS_CLASS_TYPE_FLAGS_TO_NAME_P ()
2902 @findex ADDRESS_CLASS_TYPE_FLAGS_TO_NAME_P
2903 Predicate which indicates whether @code{ADDRESS_CLASS_TYPE_FLAGS_TO_NAME} has
2905 @xref{Target Architecture Definition, , Address Classes}.
2907 @item ADDRESS_TO_POINTER (@var{type}, @var{buf}, @var{addr})
2908 @findex ADDRESS_TO_POINTER
2909 Store in @var{buf} a pointer of type @var{type} representing the address
2910 @var{addr}, in the appropriate format for the current architecture.
2911 This macro may safely assume that @var{type} is either a pointer or a
2912 C@t{++} reference type.
2913 @xref{Target Architecture Definition, , Pointers Are Not Always Addresses}.
2915 @item BELIEVE_PCC_PROMOTION
2916 @findex BELIEVE_PCC_PROMOTION
2917 Define if the compiler promotes a @code{short} or @code{char}
2918 parameter to an @code{int}, but still reports the parameter as its
2919 original type, rather than the promoted type.
2921 @item BITS_BIG_ENDIAN
2922 @findex BITS_BIG_ENDIAN
2923 Define this if the numbering of bits in the targets does @strong{not} match the
2924 endianness of the target byte order. A value of 1 means that the bits
2925 are numbered in a big-endian bit order, 0 means little-endian.
2929 This is the character array initializer for the bit pattern to put into
2930 memory where a breakpoint is set. Although it's common to use a trap
2931 instruction for a breakpoint, it's not required; for instance, the bit
2932 pattern could be an invalid instruction. The breakpoint must be no
2933 longer than the shortest instruction of the architecture.
2935 @code{BREAKPOINT} has been deprecated in favor of
2936 @code{BREAKPOINT_FROM_PC}.
2938 @item BIG_BREAKPOINT
2939 @itemx LITTLE_BREAKPOINT
2940 @findex LITTLE_BREAKPOINT
2941 @findex BIG_BREAKPOINT
2942 Similar to BREAKPOINT, but used for bi-endian targets.
2944 @code{BIG_BREAKPOINT} and @code{LITTLE_BREAKPOINT} have been deprecated in
2945 favor of @code{BREAKPOINT_FROM_PC}.
2947 @item DEPRECATED_REMOTE_BREAKPOINT
2948 @itemx DEPRECATED_LITTLE_REMOTE_BREAKPOINT
2949 @itemx DEPRECATED_BIG_REMOTE_BREAKPOINT
2950 @findex DEPRECATED_BIG_REMOTE_BREAKPOINT
2951 @findex DEPRECATED_LITTLE_REMOTE_BREAKPOINT
2952 @findex DEPRECATED_REMOTE_BREAKPOINT
2953 Specify the breakpoint instruction sequence for a remote target.
2954 @code{DEPRECATED_REMOTE_BREAKPOINT},
2955 @code{DEPRECATED_BIG_REMOTE_BREAKPOINT} and
2956 @code{DEPRECATED_LITTLE_REMOTE_BREAKPOINT} have been deprecated in
2957 favor of @code{BREAKPOINT_FROM_PC} (@pxref{BREAKPOINT_FROM_PC}).
2959 @item BREAKPOINT_FROM_PC (@var{pcptr}, @var{lenptr})
2960 @findex BREAKPOINT_FROM_PC
2961 @anchor{BREAKPOINT_FROM_PC} Use the program counter to determine the
2962 contents and size of a breakpoint instruction. It returns a pointer to
2963 a string of bytes that encode a breakpoint instruction, stores the
2964 length of the string to @code{*@var{lenptr}}, and adjusts the program
2965 counter (if necessary) to point to the actual memory location where the
2966 breakpoint should be inserted.
2968 Although it is common to use a trap instruction for a breakpoint, it's
2969 not required; for instance, the bit pattern could be an invalid
2970 instruction. The breakpoint must be no longer than the shortest
2971 instruction of the architecture.
2973 Replaces all the other @var{BREAKPOINT} macros.
2975 @item MEMORY_INSERT_BREAKPOINT (@var{addr}, @var{contents_cache})
2976 @itemx MEMORY_REMOVE_BREAKPOINT (@var{addr}, @var{contents_cache})
2977 @findex MEMORY_REMOVE_BREAKPOINT
2978 @findex MEMORY_INSERT_BREAKPOINT
2979 Insert or remove memory based breakpoints. Reasonable defaults
2980 (@code{default_memory_insert_breakpoint} and
2981 @code{default_memory_remove_breakpoint} respectively) have been
2982 provided so that it is not necessary to define these for most
2983 architectures. Architectures which may want to define
2984 @code{MEMORY_INSERT_BREAKPOINT} and @code{MEMORY_REMOVE_BREAKPOINT} will
2985 likely have instructions that are oddly sized or are not stored in a
2986 conventional manner.
2988 It may also be desirable (from an efficiency standpoint) to define
2989 custom breakpoint insertion and removal routines if
2990 @code{BREAKPOINT_FROM_PC} needs to read the target's memory for some
2993 @item ADJUST_BREAKPOINT_ADDRESS (@var{address})
2994 @findex ADJUST_BREAKPOINT_ADDRESS
2995 @cindex breakpoint address adjusted
2996 Given an address at which a breakpoint is desired, return a breakpoint
2997 address adjusted to account for architectural constraints on
2998 breakpoint placement. This method is not needed by most targets.
3000 The FR-V target (see @file{frv-tdep.c}) requires this method.
3001 The FR-V is a VLIW architecture in which a number of RISC-like
3002 instructions are grouped (packed) together into an aggregate
3003 instruction or instruction bundle. When the processor executes
3004 one of these bundles, the component instructions are executed
3007 In the course of optimization, the compiler may group instructions
3008 from distinct source statements into the same bundle. The line number
3009 information associated with one of the latter statements will likely
3010 refer to some instruction other than the first one in the bundle. So,
3011 if the user attempts to place a breakpoint on one of these latter
3012 statements, @value{GDBN} must be careful to @emph{not} place the break
3013 instruction on any instruction other than the first one in the bundle.
3014 (Remember though that the instructions within a bundle execute
3015 in parallel, so the @emph{first} instruction is the instruction
3016 at the lowest address and has nothing to do with execution order.)
3018 The FR-V's @code{ADJUST_BREAKPOINT_ADDRESS} method will adjust a
3019 breakpoint's address by scanning backwards for the beginning of
3020 the bundle, returning the address of the bundle.
3022 Since the adjustment of a breakpoint may significantly alter a user's
3023 expectation, @value{GDBN} prints a warning when an adjusted breakpoint
3024 is initially set and each time that that breakpoint is hit.
3026 @item CALL_DUMMY_LOCATION
3027 @findex CALL_DUMMY_LOCATION
3028 See the file @file{inferior.h}.
3030 This method has been replaced by @code{push_dummy_code}
3031 (@pxref{push_dummy_code}).
3033 @item CANNOT_FETCH_REGISTER (@var{regno})
3034 @findex CANNOT_FETCH_REGISTER
3035 A C expression that should be nonzero if @var{regno} cannot be fetched
3036 from an inferior process. This is only relevant if
3037 @code{FETCH_INFERIOR_REGISTERS} is not defined.
3039 @item CANNOT_STORE_REGISTER (@var{regno})
3040 @findex CANNOT_STORE_REGISTER
3041 A C expression that should be nonzero if @var{regno} should not be
3042 written to the target. This is often the case for program counters,
3043 status words, and other special registers. If this is not defined,
3044 @value{GDBN} will assume that all registers may be written.
3046 @item int CONVERT_REGISTER_P(@var{regnum})
3047 @findex CONVERT_REGISTER_P
3048 Return non-zero if register @var{regnum} can represent data values in a
3050 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
3052 @item DECR_PC_AFTER_BREAK
3053 @findex DECR_PC_AFTER_BREAK
3054 Define this to be the amount by which to decrement the PC after the
3055 program encounters a breakpoint. This is often the number of bytes in
3056 @code{BREAKPOINT}, though not always. For most targets this value will be 0.
3058 @item DISABLE_UNSETTABLE_BREAK (@var{addr})
3059 @findex DISABLE_UNSETTABLE_BREAK
3060 If defined, this should evaluate to 1 if @var{addr} is in a shared
3061 library in which breakpoints cannot be set and so should be disabled.
3063 @item PRINT_FLOAT_INFO()
3064 @findex PRINT_FLOAT_INFO
3065 If defined, then the @samp{info float} command will print information about
3066 the processor's floating point unit.
3068 @item print_registers_info (@var{gdbarch}, @var{frame}, @var{regnum}, @var{all})
3069 @findex print_registers_info
3070 If defined, pretty print the value of the register @var{regnum} for the
3071 specified @var{frame}. If the value of @var{regnum} is -1, pretty print
3072 either all registers (@var{all} is non zero) or a select subset of
3073 registers (@var{all} is zero).
3075 The default method prints one register per line, and if @var{all} is
3076 zero omits floating-point registers.
3078 @item PRINT_VECTOR_INFO()
3079 @findex PRINT_VECTOR_INFO
3080 If defined, then the @samp{info vector} command will call this function
3081 to print information about the processor's vector unit.
3083 By default, the @samp{info vector} command will print all vector
3084 registers (the register's type having the vector attribute).
3086 @item DWARF_REG_TO_REGNUM
3087 @findex DWARF_REG_TO_REGNUM
3088 Convert DWARF register number into @value{GDBN} regnum. If not defined,
3089 no conversion will be performed.
3091 @item DWARF2_REG_TO_REGNUM
3092 @findex DWARF2_REG_TO_REGNUM
3093 Convert DWARF2 register number into @value{GDBN} regnum. If not
3094 defined, no conversion will be performed.
3096 @item ECOFF_REG_TO_REGNUM
3097 @findex ECOFF_REG_TO_REGNUM
3098 Convert ECOFF register number into @value{GDBN} regnum. If not defined,
3099 no conversion will be performed.
3101 @item END_OF_TEXT_DEFAULT
3102 @findex END_OF_TEXT_DEFAULT
3103 This is an expression that should designate the end of the text section.
3106 @item EXTRACT_RETURN_VALUE(@var{type}, @var{regbuf}, @var{valbuf})
3107 @findex EXTRACT_RETURN_VALUE
3108 Define this to extract a function's return value of type @var{type} from
3109 the raw register state @var{regbuf} and copy that, in virtual format,
3112 This method has been deprecated in favour of @code{gdbarch_return_value}
3113 (@pxref{gdbarch_return_value}).
3115 @item DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS(@var{regbuf})
3116 @findex DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS
3117 @anchor{DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS}
3118 When defined, extract from the array @var{regbuf} (containing the raw
3119 register state) the @code{CORE_ADDR} at which a function should return
3120 its structure value.
3122 @xref{gdbarch_return_value}.
3124 @item DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS_P()
3125 @findex DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS_P
3126 Predicate for @code{DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS}.
3128 @item DEPRECATED_FP_REGNUM
3129 @findex DEPRECATED_FP_REGNUM
3130 If the virtual frame pointer is kept in a register, then define this
3131 macro to be the number (greater than or equal to zero) of that register.
3133 This should only need to be defined if @code{DEPRECATED_TARGET_READ_FP}
3136 @item DEPRECATED_FRAMELESS_FUNCTION_INVOCATION(@var{fi})
3137 @findex DEPRECATED_FRAMELESS_FUNCTION_INVOCATION
3138 Define this to an expression that returns 1 if the function invocation
3139 represented by @var{fi} does not have a stack frame associated with it.
3142 @item frame_align (@var{address})
3143 @anchor{frame_align}
3145 Define this to adjust @var{address} so that it meets the alignment
3146 requirements for the start of a new stack frame. A stack frame's
3147 alignment requirements are typically stronger than a target processors
3148 stack alignment requirements (@pxref{DEPRECATED_STACK_ALIGN}).
3150 This function is used to ensure that, when creating a dummy frame, both
3151 the initial stack pointer and (if needed) the address of the return
3152 value are correctly aligned.
3154 Unlike @code{DEPRECATED_STACK_ALIGN}, this function always adjusts the
3155 address in the direction of stack growth.
3157 By default, no frame based stack alignment is performed.
3159 @item int frame_red_zone_size
3161 The number of bytes, beyond the innermost-stack-address, reserved by the
3162 @sc{abi}. A function is permitted to use this scratch area (instead of
3163 allocating extra stack space).
3165 When performing an inferior function call, to ensure that it does not
3166 modify this area, @value{GDBN} adjusts the innermost-stack-address by
3167 @var{frame_red_zone_size} bytes before pushing parameters onto the
3170 By default, zero bytes are allocated. The value must be aligned
3171 (@pxref{frame_align}).
3173 The @sc{amd64} (nee x86-64) @sc{abi} documentation refers to the
3174 @emph{red zone} when describing this scratch area.
3177 @item DEPRECATED_FRAME_CHAIN(@var{frame})
3178 @findex DEPRECATED_FRAME_CHAIN
3179 Given @var{frame}, return a pointer to the calling frame.
3181 @item DEPRECATED_FRAME_CHAIN_VALID(@var{chain}, @var{thisframe})
3182 @findex DEPRECATED_FRAME_CHAIN_VALID
3183 Define this to be an expression that returns zero if the given frame is an
3184 outermost frame, with no caller, and nonzero otherwise. Most normal
3185 situations can be handled without defining this macro, including @code{NULL}
3186 chain pointers, dummy frames, and frames whose PC values are inside the
3187 startup file (e.g.@: @file{crt0.o}), inside @code{main}, or inside
3190 @item DEPRECATED_FRAME_INIT_SAVED_REGS(@var{frame})
3191 @findex DEPRECATED_FRAME_INIT_SAVED_REGS
3192 See @file{frame.h}. Determines the address of all registers in the
3193 current stack frame storing each in @code{frame->saved_regs}. Space for
3194 @code{frame->saved_regs} shall be allocated by
3195 @code{DEPRECATED_FRAME_INIT_SAVED_REGS} using
3196 @code{frame_saved_regs_zalloc}.
3198 @code{FRAME_FIND_SAVED_REGS} is deprecated.
3200 @item FRAME_NUM_ARGS (@var{fi})
3201 @findex FRAME_NUM_ARGS
3202 For the frame described by @var{fi} return the number of arguments that
3203 are being passed. If the number of arguments is not known, return
3206 @item DEPRECATED_FRAME_SAVED_PC(@var{frame})
3207 @findex DEPRECATED_FRAME_SAVED_PC
3208 @anchor{DEPRECATED_FRAME_SAVED_PC} Given @var{frame}, return the pc
3209 saved there. This is the return address.
3211 This method is deprecated. @xref{unwind_pc}.
3213 @item CORE_ADDR unwind_pc (struct frame_info *@var{this_frame})
3215 @anchor{unwind_pc} Return the instruction address, in @var{this_frame}'s
3216 caller, at which execution will resume after @var{this_frame} returns.
3217 This is commonly refered to as the return address.
3219 The implementation, which must be frame agnostic (work with any frame),
3220 is typically no more than:
3224 frame_unwind_unsigned_register (this_frame, D10V_PC_REGNUM, &pc);
3225 return d10v_make_iaddr (pc);
3229 @xref{DEPRECATED_FRAME_SAVED_PC}, which this method replaces.
3231 @item CORE_ADDR unwind_sp (struct frame_info *@var{this_frame})
3233 @anchor{unwind_sp} Return the frame's inner most stack address. This is
3234 commonly refered to as the frame's @dfn{stack pointer}.
3236 The implementation, which must be frame agnostic (work with any frame),
3237 is typically no more than:
3241 frame_unwind_unsigned_register (this_frame, D10V_SP_REGNUM, &sp);
3242 return d10v_make_daddr (sp);
3246 @xref{TARGET_READ_SP}, which this method replaces.
3248 @item FUNCTION_EPILOGUE_SIZE
3249 @findex FUNCTION_EPILOGUE_SIZE
3250 For some COFF targets, the @code{x_sym.x_misc.x_fsize} field of the
3251 function end symbol is 0. For such targets, you must define
3252 @code{FUNCTION_EPILOGUE_SIZE} to expand into the standard size of a
3253 function's epilogue.
3255 @item DEPRECATED_FUNCTION_START_OFFSET
3256 @findex DEPRECATED_FUNCTION_START_OFFSET
3257 An integer, giving the offset in bytes from a function's address (as
3258 used in the values of symbols, function pointers, etc.), and the
3259 function's first genuine instruction.
3261 This is zero on almost all machines: the function's address is usually
3262 the address of its first instruction. However, on the VAX, for
3263 example, each function starts with two bytes containing a bitmask
3264 indicating which registers to save upon entry to the function. The
3265 VAX @code{call} instructions check this value, and save the
3266 appropriate registers automatically. Thus, since the offset from the
3267 function's address to its first instruction is two bytes,
3268 @code{DEPRECATED_FUNCTION_START_OFFSET} would be 2 on the VAX.
3270 @item GCC_COMPILED_FLAG_SYMBOL
3271 @itemx GCC2_COMPILED_FLAG_SYMBOL
3272 @findex GCC2_COMPILED_FLAG_SYMBOL
3273 @findex GCC_COMPILED_FLAG_SYMBOL
3274 If defined, these are the names of the symbols that @value{GDBN} will
3275 look for to detect that GCC compiled the file. The default symbols
3276 are @code{gcc_compiled.} and @code{gcc2_compiled.},
3277 respectively. (Currently only defined for the Delta 68.)
3279 @item @value{GDBN}_MULTI_ARCH
3280 @findex @value{GDBN}_MULTI_ARCH
3281 If defined and non-zero, enables support for multiple architectures
3282 within @value{GDBN}.
3284 This support can be enabled at two levels. At level one, only
3285 definitions for previously undefined macros are provided; at level two,
3286 a multi-arch definition of all architecture dependent macros will be
3289 @item @value{GDBN}_TARGET_IS_HPPA
3290 @findex @value{GDBN}_TARGET_IS_HPPA
3291 This determines whether horrible kludge code in @file{dbxread.c} and
3292 @file{partial-stab.h} is used to mangle multiple-symbol-table files from
3293 HPPA's. This should all be ripped out, and a scheme like @file{elfread.c}
3296 @item GET_LONGJMP_TARGET
3297 @findex GET_LONGJMP_TARGET
3298 For most machines, this is a target-dependent parameter. On the
3299 DECstation and the Iris, this is a native-dependent parameter, since
3300 the header file @file{setjmp.h} is needed to define it.
3302 This macro determines the target PC address that @code{longjmp} will jump to,
3303 assuming that we have just stopped at a @code{longjmp} breakpoint. It takes a
3304 @code{CORE_ADDR *} as argument, and stores the target PC value through this
3305 pointer. It examines the current state of the machine as needed.
3307 @item DEPRECATED_GET_SAVED_REGISTER
3308 @findex DEPRECATED_GET_SAVED_REGISTER
3309 Define this if you need to supply your own definition for the function
3310 @code{DEPRECATED_GET_SAVED_REGISTER}.
3312 @item DEPRECATED_IBM6000_TARGET
3313 @findex DEPRECATED_IBM6000_TARGET
3314 Shows that we are configured for an IBM RS/6000 system. This
3315 conditional should be eliminated (FIXME) and replaced by
3316 feature-specific macros. It was introduced in a haste and we are
3317 repenting at leisure.
3319 @item I386_USE_GENERIC_WATCHPOINTS
3320 An x86-based target can define this to use the generic x86 watchpoint
3321 support; see @ref{Algorithms, I386_USE_GENERIC_WATCHPOINTS}.
3323 @item SYMBOLS_CAN_START_WITH_DOLLAR
3324 @findex SYMBOLS_CAN_START_WITH_DOLLAR
3325 Some systems have routines whose names start with @samp{$}. Giving this
3326 macro a non-zero value tells @value{GDBN}'s expression parser to check for such
3327 routines when parsing tokens that begin with @samp{$}.
3329 On HP-UX, certain system routines (millicode) have names beginning with
3330 @samp{$} or @samp{$$}. For example, @code{$$dyncall} is a millicode
3331 routine that handles inter-space procedure calls on PA-RISC.
3333 @item DEPRECATED_INIT_EXTRA_FRAME_INFO (@var{fromleaf}, @var{frame})
3334 @findex DEPRECATED_INIT_EXTRA_FRAME_INFO
3335 If additional information about the frame is required this should be
3336 stored in @code{frame->extra_info}. Space for @code{frame->extra_info}
3337 is allocated using @code{frame_extra_info_zalloc}.
3339 @item DEPRECATED_INIT_FRAME_PC (@var{fromleaf}, @var{prev})
3340 @findex DEPRECATED_INIT_FRAME_PC
3341 This is a C statement that sets the pc of the frame pointed to by
3342 @var{prev}. [By default...]
3344 @item INNER_THAN (@var{lhs}, @var{rhs})
3346 Returns non-zero if stack address @var{lhs} is inner than (nearer to the
3347 stack top) stack address @var{rhs}. Define this as @code{lhs < rhs} if
3348 the target's stack grows downward in memory, or @code{lhs > rsh} if the
3351 @item gdbarch_in_function_epilogue_p (@var{gdbarch}, @var{pc})
3352 @findex gdbarch_in_function_epilogue_p
3353 Returns non-zero if the given @var{pc} is in the epilogue of a function.
3354 The epilogue of a function is defined as the part of a function where
3355 the stack frame of the function already has been destroyed up to the
3356 final `return from function call' instruction.
3358 @item DEPRECATED_SIGTRAMP_START (@var{pc})
3359 @findex DEPRECATED_SIGTRAMP_START
3360 @itemx DEPRECATED_SIGTRAMP_END (@var{pc})
3361 @findex DEPRECATED_SIGTRAMP_END
3362 Define these to be the start and end address of the @code{sigtramp} for the
3363 given @var{pc}. On machines where the address is just a compile time
3364 constant, the macro expansion will typically just ignore the supplied
3367 @item IN_SOLIB_CALL_TRAMPOLINE (@var{pc}, @var{name})
3368 @findex IN_SOLIB_CALL_TRAMPOLINE
3369 Define this to evaluate to nonzero if the program is stopped in the
3370 trampoline that connects to a shared library.
3372 @item IN_SOLIB_RETURN_TRAMPOLINE (@var{pc}, @var{name})
3373 @findex IN_SOLIB_RETURN_TRAMPOLINE
3374 Define this to evaluate to nonzero if the program is stopped in the
3375 trampoline that returns from a shared library.
3377 @item IN_SOLIB_DYNSYM_RESOLVE_CODE (@var{pc})
3378 @findex IN_SOLIB_DYNSYM_RESOLVE_CODE
3379 Define this to evaluate to nonzero if the program is stopped in the
3382 @item SKIP_SOLIB_RESOLVER (@var{pc})
3383 @findex SKIP_SOLIB_RESOLVER
3384 Define this to evaluate to the (nonzero) address at which execution
3385 should continue to get past the dynamic linker's symbol resolution
3386 function. A zero value indicates that it is not important or necessary
3387 to set a breakpoint to get through the dynamic linker and that single
3388 stepping will suffice.
3390 @item INTEGER_TO_ADDRESS (@var{type}, @var{buf})
3391 @findex INTEGER_TO_ADDRESS
3392 @cindex converting integers to addresses
3393 Define this when the architecture needs to handle non-pointer to address
3394 conversions specially. Converts that value to an address according to
3395 the current architectures conventions.
3397 @emph{Pragmatics: When the user copies a well defined expression from
3398 their source code and passes it, as a parameter, to @value{GDBN}'s
3399 @code{print} command, they should get the same value as would have been
3400 computed by the target program. Any deviation from this rule can cause
3401 major confusion and annoyance, and needs to be justified carefully. In
3402 other words, @value{GDBN} doesn't really have the freedom to do these
3403 conversions in clever and useful ways. It has, however, been pointed
3404 out that users aren't complaining about how @value{GDBN} casts integers
3405 to pointers; they are complaining that they can't take an address from a
3406 disassembly listing and give it to @code{x/i}. Adding an architecture
3407 method like @code{INTEGER_TO_ADDRESS} certainly makes it possible for
3408 @value{GDBN} to ``get it right'' in all circumstances.}
3410 @xref{Target Architecture Definition, , Pointers Are Not Always
3413 @item NO_HIF_SUPPORT
3414 @findex NO_HIF_SUPPORT
3415 (Specific to the a29k.)
3417 @item POINTER_TO_ADDRESS (@var{type}, @var{buf})
3418 @findex POINTER_TO_ADDRESS
3419 Assume that @var{buf} holds a pointer of type @var{type}, in the
3420 appropriate format for the current architecture. Return the byte
3421 address the pointer refers to.
3422 @xref{Target Architecture Definition, , Pointers Are Not Always Addresses}.
3424 @item REGISTER_CONVERTIBLE (@var{reg})
3425 @findex REGISTER_CONVERTIBLE
3426 Return non-zero if @var{reg} uses different raw and virtual formats.
3427 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3429 @item REGISTER_TO_VALUE(@var{regnum}, @var{type}, @var{from}, @var{to})
3430 @findex REGISTER_TO_VALUE
3431 Convert the raw contents of register @var{regnum} into a value of type
3433 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
3435 @item DEPRECATED_REGISTER_RAW_SIZE (@var{reg})
3436 @findex DEPRECATED_REGISTER_RAW_SIZE
3437 Return the raw size of @var{reg}; defaults to the size of the register's
3439 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3441 @item register_reggroup_p (@var{gdbarch}, @var{regnum}, @var{reggroup})
3442 @findex register_reggroup_p
3443 @cindex register groups
3444 Return non-zero if register @var{regnum} is a member of the register
3445 group @var{reggroup}.
3447 By default, registers are grouped as follows:
3450 @item float_reggroup
3451 Any register with a valid name and a floating-point type.
3452 @item vector_reggroup
3453 Any register with a valid name and a vector type.
3454 @item general_reggroup
3455 Any register with a valid name and a type other than vector or
3456 floating-point. @samp{float_reggroup}.
3458 @itemx restore_reggroup
3460 Any register with a valid name.
3463 @item DEPRECATED_REGISTER_VIRTUAL_SIZE (@var{reg})
3464 @findex DEPRECATED_REGISTER_VIRTUAL_SIZE
3465 Return the virtual size of @var{reg}; defaults to the size of the
3466 register's virtual type.
3467 Return the virtual size of @var{reg}.
3468 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3470 @item DEPRECATED_REGISTER_VIRTUAL_TYPE (@var{reg})
3471 @findex REGISTER_VIRTUAL_TYPE
3472 Return the virtual type of @var{reg}.
3473 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3475 @item struct type *register_type (@var{gdbarch}, @var{reg})
3476 @findex register_type
3477 If defined, return the type of register @var{reg}. This function
3478 superseeds @code{DEPRECATED_REGISTER_VIRTUAL_TYPE}. @xref{Target Architecture
3479 Definition, , Raw and Virtual Register Representations}.
3481 @item REGISTER_CONVERT_TO_VIRTUAL(@var{reg}, @var{type}, @var{from}, @var{to})
3482 @findex REGISTER_CONVERT_TO_VIRTUAL
3483 Convert the value of register @var{reg} from its raw form to its virtual
3485 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3487 @item REGISTER_CONVERT_TO_RAW(@var{type}, @var{reg}, @var{from}, @var{to})
3488 @findex REGISTER_CONVERT_TO_RAW
3489 Convert the value of register @var{reg} from its virtual form to its raw
3491 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3493 @item const struct regset *regset_from_core_section (struct gdbarch * @var{gdbarch}, const char * @var{sect_name}, size_t @var{sect_size})
3494 @findex regset_from_core_section
3495 Return the appropriate register set for a core file section with name
3496 @var{sect_name} and size @var{sect_size}.
3498 @item SOFTWARE_SINGLE_STEP_P()
3499 @findex SOFTWARE_SINGLE_STEP_P
3500 Define this as 1 if the target does not have a hardware single-step
3501 mechanism. The macro @code{SOFTWARE_SINGLE_STEP} must also be defined.
3503 @item SOFTWARE_SINGLE_STEP(@var{signal}, @var{insert_breapoints_p})
3504 @findex SOFTWARE_SINGLE_STEP
3505 A function that inserts or removes (depending on
3506 @var{insert_breapoints_p}) breakpoints at each possible destinations of
3507 the next instruction. See @file{sparc-tdep.c} and @file{rs6000-tdep.c}
3510 @item SOFUN_ADDRESS_MAYBE_MISSING
3511 @findex SOFUN_ADDRESS_MAYBE_MISSING
3512 Somebody clever observed that, the more actual addresses you have in the
3513 debug information, the more time the linker has to spend relocating
3514 them. So whenever there's some other way the debugger could find the
3515 address it needs, you should omit it from the debug info, to make
3518 @code{SOFUN_ADDRESS_MAYBE_MISSING} indicates that a particular set of
3519 hacks of this sort are in use, affecting @code{N_SO} and @code{N_FUN}
3520 entries in stabs-format debugging information. @code{N_SO} stabs mark
3521 the beginning and ending addresses of compilation units in the text
3522 segment. @code{N_FUN} stabs mark the starts and ends of functions.
3524 @code{SOFUN_ADDRESS_MAYBE_MISSING} means two things:
3528 @code{N_FUN} stabs have an address of zero. Instead, you should find the
3529 addresses where the function starts by taking the function name from
3530 the stab, and then looking that up in the minsyms (the
3531 linker/assembler symbol table). In other words, the stab has the
3532 name, and the linker/assembler symbol table is the only place that carries
3536 @code{N_SO} stabs have an address of zero, too. You just look at the
3537 @code{N_FUN} stabs that appear before and after the @code{N_SO} stab,
3538 and guess the starting and ending addresses of the compilation unit from
3542 @item PC_LOAD_SEGMENT
3543 @findex PC_LOAD_SEGMENT
3544 If defined, print information about the load segment for the program
3545 counter. (Defined only for the RS/6000.)
3549 If the program counter is kept in a register, then define this macro to
3550 be the number (greater than or equal to zero) of that register.
3552 This should only need to be defined if @code{TARGET_READ_PC} and
3553 @code{TARGET_WRITE_PC} are not defined.
3556 @findex PARM_BOUNDARY
3557 If non-zero, round arguments to a boundary of this many bits before
3558 pushing them on the stack.
3560 @item stabs_argument_has_addr (@var{gdbarch}, @var{type})
3561 @findex stabs_argument_has_addr
3562 @findex DEPRECATED_REG_STRUCT_HAS_ADDR
3563 @anchor{stabs_argument_has_addr} Define this to return nonzero if a
3564 function argument of type @var{type} is passed by reference instead of
3567 This method replaces @code{DEPRECATED_REG_STRUCT_HAS_ADDR}
3568 (@pxref{DEPRECATED_REG_STRUCT_HAS_ADDR}).
3570 @item PROCESS_LINENUMBER_HOOK
3571 @findex PROCESS_LINENUMBER_HOOK
3572 A hook defined for XCOFF reading.
3574 @item PROLOGUE_FIRSTLINE_OVERLAP
3575 @findex PROLOGUE_FIRSTLINE_OVERLAP
3576 (Only used in unsupported Convex configuration.)
3580 If defined, this is the number of the processor status register. (This
3581 definition is only used in generic code when parsing "$ps".)
3583 @item DEPRECATED_POP_FRAME
3584 @findex DEPRECATED_POP_FRAME
3586 If defined, used by @code{frame_pop} to remove a stack frame. This
3587 method has been superseeded by generic code.
3589 @item push_dummy_call (@var{gdbarch}, @var{function}, @var{regcache}, @var{pc_addr}, @var{nargs}, @var{args}, @var{sp}, @var{struct_return}, @var{struct_addr})
3590 @findex push_dummy_call
3591 @findex DEPRECATED_PUSH_ARGUMENTS.
3592 @anchor{push_dummy_call} Define this to push the dummy frame's call to
3593 the inferior function onto the stack. In addition to pushing
3594 @var{nargs}, the code should push @var{struct_addr} (when
3595 @var{struct_return}), and the return address (@var{bp_addr}).
3597 @var{function} is a pointer to a @code{struct value}; on architectures that use
3598 function descriptors, this contains the function descriptor value.
3600 Returns the updated top-of-stack pointer.
3602 This method replaces @code{DEPRECATED_PUSH_ARGUMENTS}.
3604 @item CORE_ADDR push_dummy_code (@var{gdbarch}, @var{sp}, @var{funaddr}, @var{using_gcc}, @var{args}, @var{nargs}, @var{value_type}, @var{real_pc}, @var{bp_addr})
3605 @findex push_dummy_code
3606 @anchor{push_dummy_code} Given a stack based call dummy, push the
3607 instruction sequence (including space for a breakpoint) to which the
3608 called function should return.
3610 Set @var{bp_addr} to the address at which the breakpoint instruction
3611 should be inserted, @var{real_pc} to the resume address when starting
3612 the call sequence, and return the updated inner-most stack address.
3614 By default, the stack is grown sufficient to hold a frame-aligned
3615 (@pxref{frame_align}) breakpoint, @var{bp_addr} is set to the address
3616 reserved for that breakpoint, and @var{real_pc} set to @var{funaddr}.
3618 This method replaces @code{CALL_DUMMY_LOCATION},
3619 @code{DEPRECATED_REGISTER_SIZE}.
3621 @item REGISTER_NAME(@var{i})
3622 @findex REGISTER_NAME
3623 Return the name of register @var{i} as a string. May return @code{NULL}
3624 or @code{NUL} to indicate that register @var{i} is not valid.
3626 @item DEPRECATED_REG_STRUCT_HAS_ADDR (@var{gcc_p}, @var{type})
3627 @findex DEPRECATED_REG_STRUCT_HAS_ADDR
3628 @anchor{DEPRECATED_REG_STRUCT_HAS_ADDR}Define this to return 1 if the
3629 given type will be passed by pointer rather than directly.
3631 This method has been replaced by @code{stabs_argument_has_addr}
3632 (@pxref{stabs_argument_has_addr}).
3634 @item SAVE_DUMMY_FRAME_TOS (@var{sp})
3635 @findex SAVE_DUMMY_FRAME_TOS
3636 @anchor{SAVE_DUMMY_FRAME_TOS} Used in @samp{call_function_by_hand} to
3637 notify the target dependent code of the top-of-stack value that will be
3638 passed to the the inferior code. This is the value of the @code{SP}
3639 after both the dummy frame and space for parameters/results have been
3640 allocated on the stack. @xref{unwind_dummy_id}.
3642 @item SDB_REG_TO_REGNUM
3643 @findex SDB_REG_TO_REGNUM
3644 Define this to convert sdb register numbers into @value{GDBN} regnums. If not
3645 defined, no conversion will be done.
3647 @item enum return_value_convention gdbarch_return_value (struct gdbarch *@var{gdbarch}, struct type *@var{valtype}, struct regcache *@var{regcache}, void *@var{readbuf}, const void *@var{writebuf})
3648 @findex gdbarch_return_value
3649 @anchor{gdbarch_return_value} Given a function with a return-value of
3650 type @var{rettype}, return which return-value convention that function
3653 @value{GDBN} currently recognizes two function return-value conventions:
3654 @code{RETURN_VALUE_REGISTER_CONVENTION} where the return value is found
3655 in registers; and @code{RETURN_VALUE_STRUCT_CONVENTION} where the return
3656 value is found in memory and the address of that memory location is
3657 passed in as the function's first parameter.
3659 If the register convention is being used, and @var{writebuf} is
3660 non-@code{NULL}, also copy the return-value in @var{writebuf} into
3663 If the register convention is being used, and @var{readbuf} is
3664 non-@code{NULL}, also copy the return value from @var{regcache} into
3665 @var{readbuf} (@var{regcache} contains a copy of the registers from the
3666 just returned function).
3668 @xref{DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS}, for a description of how
3669 return-values that use the struct convention are handled.
3671 @emph{Maintainer note: This method replaces separate predicate, extract,
3672 store methods. By having only one method, the logic needed to determine
3673 the return-value convention need only be implemented in one place. If
3674 @value{GDBN} were written in an @sc{oo} language, this method would
3675 instead return an object that knew how to perform the register
3676 return-value extract and store.}
3678 @emph{Maintainer note: This method does not take a @var{gcc_p}
3679 parameter, and such a parameter should not be added. If an architecture
3680 that requires per-compiler or per-function information be identified,
3681 then the replacement of @var{rettype} with @code{struct value}
3682 @var{function} should be persued.}
3684 @emph{Maintainer note: The @var{regcache} parameter limits this methods
3685 to the inner most frame. While replacing @var{regcache} with a
3686 @code{struct frame_info} @var{frame} parameter would remove that
3687 limitation there has yet to be a demonstrated need for such a change.}
3689 @item SKIP_PERMANENT_BREAKPOINT
3690 @findex SKIP_PERMANENT_BREAKPOINT
3691 Advance the inferior's PC past a permanent breakpoint. @value{GDBN} normally
3692 steps over a breakpoint by removing it, stepping one instruction, and
3693 re-inserting the breakpoint. However, permanent breakpoints are
3694 hardwired into the inferior, and can't be removed, so this strategy
3695 doesn't work. Calling @code{SKIP_PERMANENT_BREAKPOINT} adjusts the processor's
3696 state so that execution will resume just after the breakpoint. This
3697 macro does the right thing even when the breakpoint is in the delay slot
3698 of a branch or jump.
3700 @item SKIP_PROLOGUE (@var{pc})
3701 @findex SKIP_PROLOGUE
3702 A C expression that returns the address of the ``real'' code beyond the
3703 function entry prologue found at @var{pc}.
3705 @item SKIP_TRAMPOLINE_CODE (@var{pc})
3706 @findex SKIP_TRAMPOLINE_CODE
3707 If the target machine has trampoline code that sits between callers and
3708 the functions being called, then define this macro to return a new PC
3709 that is at the start of the real function.
3713 If the stack-pointer is kept in a register, then define this macro to be
3714 the number (greater than or equal to zero) of that register, or -1 if
3715 there is no such register.
3717 @item STAB_REG_TO_REGNUM
3718 @findex STAB_REG_TO_REGNUM
3719 Define this to convert stab register numbers (as gotten from `r'
3720 declarations) into @value{GDBN} regnums. If not defined, no conversion will be
3723 @item DEPRECATED_STACK_ALIGN (@var{addr})
3724 @anchor{DEPRECATED_STACK_ALIGN}
3725 @findex DEPRECATED_STACK_ALIGN
3726 Define this to increase @var{addr} so that it meets the alignment
3727 requirements for the processor's stack.
3729 Unlike @ref{frame_align}, this function always adjusts @var{addr}
3732 By default, no stack alignment is performed.
3734 @item STEP_SKIPS_DELAY (@var{addr})
3735 @findex STEP_SKIPS_DELAY
3736 Define this to return true if the address is of an instruction with a
3737 delay slot. If a breakpoint has been placed in the instruction's delay
3738 slot, @value{GDBN} will single-step over that instruction before resuming
3739 normally. Currently only defined for the Mips.
3741 @item STORE_RETURN_VALUE (@var{type}, @var{regcache}, @var{valbuf})
3742 @findex STORE_RETURN_VALUE
3743 A C expression that writes the function return value, found in
3744 @var{valbuf}, into the @var{regcache}. @var{type} is the type of the
3745 value that is to be returned.
3747 This method has been deprecated in favour of @code{gdbarch_return_value}
3748 (@pxref{gdbarch_return_value}).
3750 @item SYMBOL_RELOADING_DEFAULT
3751 @findex SYMBOL_RELOADING_DEFAULT
3752 The default value of the ``symbol-reloading'' variable. (Never defined in
3755 @item TARGET_CHAR_BIT
3756 @findex TARGET_CHAR_BIT
3757 Number of bits in a char; defaults to 8.
3759 @item TARGET_CHAR_SIGNED
3760 @findex TARGET_CHAR_SIGNED
3761 Non-zero if @code{char} is normally signed on this architecture; zero if
3762 it should be unsigned.
3764 The ISO C standard requires the compiler to treat @code{char} as
3765 equivalent to either @code{signed char} or @code{unsigned char}; any
3766 character in the standard execution set is supposed to be positive.
3767 Most compilers treat @code{char} as signed, but @code{char} is unsigned
3768 on the IBM S/390, RS6000, and PowerPC targets.
3770 @item TARGET_COMPLEX_BIT
3771 @findex TARGET_COMPLEX_BIT
3772 Number of bits in a complex number; defaults to @code{2 * TARGET_FLOAT_BIT}.
3774 At present this macro is not used.
3776 @item TARGET_DOUBLE_BIT
3777 @findex TARGET_DOUBLE_BIT
3778 Number of bits in a double float; defaults to @code{8 * TARGET_CHAR_BIT}.
3780 @item TARGET_DOUBLE_COMPLEX_BIT
3781 @findex TARGET_DOUBLE_COMPLEX_BIT
3782 Number of bits in a double complex; defaults to @code{2 * TARGET_DOUBLE_BIT}.
3784 At present this macro is not used.
3786 @item TARGET_FLOAT_BIT
3787 @findex TARGET_FLOAT_BIT
3788 Number of bits in a float; defaults to @code{4 * TARGET_CHAR_BIT}.
3790 @item TARGET_INT_BIT
3791 @findex TARGET_INT_BIT
3792 Number of bits in an integer; defaults to @code{4 * TARGET_CHAR_BIT}.
3794 @item TARGET_LONG_BIT
3795 @findex TARGET_LONG_BIT
3796 Number of bits in a long integer; defaults to @code{4 * TARGET_CHAR_BIT}.
3798 @item TARGET_LONG_DOUBLE_BIT
3799 @findex TARGET_LONG_DOUBLE_BIT
3800 Number of bits in a long double float;
3801 defaults to @code{2 * TARGET_DOUBLE_BIT}.
3803 @item TARGET_LONG_LONG_BIT
3804 @findex TARGET_LONG_LONG_BIT
3805 Number of bits in a long long integer; defaults to @code{2 * TARGET_LONG_BIT}.
3807 @item TARGET_PTR_BIT
3808 @findex TARGET_PTR_BIT
3809 Number of bits in a pointer; defaults to @code{TARGET_INT_BIT}.
3811 @item TARGET_SHORT_BIT
3812 @findex TARGET_SHORT_BIT
3813 Number of bits in a short integer; defaults to @code{2 * TARGET_CHAR_BIT}.
3815 @item TARGET_READ_PC
3816 @findex TARGET_READ_PC
3817 @itemx TARGET_WRITE_PC (@var{val}, @var{pid})
3818 @findex TARGET_WRITE_PC
3819 @anchor{TARGET_WRITE_PC}
3820 @itemx TARGET_READ_SP
3821 @findex TARGET_READ_SP
3822 @itemx TARGET_READ_FP
3823 @findex TARGET_READ_FP
3828 @anchor{TARGET_READ_SP} These change the behavior of @code{read_pc},
3829 @code{write_pc} and @code{read_sp}. For most targets, these may be
3830 left undefined. @value{GDBN} will call the read and write register
3831 functions with the relevant @code{_REGNUM} argument.
3833 These macros are useful when a target keeps one of these registers in a
3834 hard to get at place; for example, part in a segment register and part
3835 in an ordinary register.
3837 @xref{unwind_sp}, which replaces @code{TARGET_READ_SP}.
3839 @item TARGET_VIRTUAL_FRAME_POINTER(@var{pc}, @var{regp}, @var{offsetp})
3840 @findex TARGET_VIRTUAL_FRAME_POINTER
3841 Returns a @code{(register, offset)} pair representing the virtual frame
3842 pointer in use at the code address @var{pc}. If virtual frame pointers
3843 are not used, a default definition simply returns
3844 @code{DEPRECATED_FP_REGNUM}, with an offset of zero.
3846 @item TARGET_HAS_HARDWARE_WATCHPOINTS
3847 If non-zero, the target has support for hardware-assisted
3848 watchpoints. @xref{Algorithms, watchpoints}, for more details and
3849 other related macros.
3851 @item TARGET_PRINT_INSN (@var{addr}, @var{info})
3852 @findex TARGET_PRINT_INSN
3853 This is the function used by @value{GDBN} to print an assembly
3854 instruction. It prints the instruction at address @var{addr} in
3855 debugged memory and returns the length of the instruction, in bytes. If
3856 a target doesn't define its own printing routine, it defaults to an
3857 accessor function for the global pointer
3858 @code{deprecated_tm_print_insn}. This usually points to a function in
3859 the @code{opcodes} library (@pxref{Support Libraries, ,Opcodes}).
3860 @var{info} is a structure (of type @code{disassemble_info}) defined in
3861 @file{include/dis-asm.h} used to pass information to the instruction
3864 @item struct frame_id unwind_dummy_id (struct frame_info *@var{frame})
3865 @findex unwind_dummy_id
3866 @anchor{unwind_dummy_id} Given @var{frame} return a @code{struct
3867 frame_id} that uniquely identifies an inferior function call's dummy
3868 frame. The value returned must match the dummy frame stack value
3869 previously saved using @code{SAVE_DUMMY_FRAME_TOS}.
3870 @xref{SAVE_DUMMY_FRAME_TOS}.
3872 @item DEPRECATED_USE_STRUCT_CONVENTION (@var{gcc_p}, @var{type})
3873 @findex DEPRECATED_USE_STRUCT_CONVENTION
3874 If defined, this must be an expression that is nonzero if a value of the
3875 given @var{type} being returned from a function must have space
3876 allocated for it on the stack. @var{gcc_p} is true if the function
3877 being considered is known to have been compiled by GCC; this is helpful
3878 for systems where GCC is known to use different calling convention than
3881 This method has been deprecated in favour of @code{gdbarch_return_value}
3882 (@pxref{gdbarch_return_value}).
3884 @item VALUE_TO_REGISTER(@var{type}, @var{regnum}, @var{from}, @var{to})
3885 @findex VALUE_TO_REGISTER
3886 Convert a value of type @var{type} into the raw contents of register
3888 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
3890 @item VARIABLES_INSIDE_BLOCK (@var{desc}, @var{gcc_p})
3891 @findex VARIABLES_INSIDE_BLOCK
3892 For dbx-style debugging information, if the compiler puts variable
3893 declarations inside LBRAC/RBRAC blocks, this should be defined to be
3894 nonzero. @var{desc} is the value of @code{n_desc} from the
3895 @code{N_RBRAC} symbol, and @var{gcc_p} is true if @value{GDBN} has noticed the
3896 presence of either the @code{GCC_COMPILED_SYMBOL} or the
3897 @code{GCC2_COMPILED_SYMBOL}. By default, this is 0.
3899 @item OS9K_VARIABLES_INSIDE_BLOCK (@var{desc}, @var{gcc_p})
3900 @findex OS9K_VARIABLES_INSIDE_BLOCK
3901 Similarly, for OS/9000. Defaults to 1.
3904 Motorola M68K target conditionals.
3908 Define this to be the 4-bit location of the breakpoint trap vector. If
3909 not defined, it will default to @code{0xf}.
3911 @item REMOTE_BPT_VECTOR
3912 Defaults to @code{1}.
3914 @item NAME_OF_MALLOC
3915 @findex NAME_OF_MALLOC
3916 A string containing the name of the function to call in order to
3917 allocate some memory in the inferior. The default value is "malloc".
3921 @section Adding a New Target
3923 @cindex adding a target
3924 The following files add a target to @value{GDBN}:
3928 @item gdb/config/@var{arch}/@var{ttt}.mt
3929 Contains a Makefile fragment specific to this target. Specifies what
3930 object files are needed for target @var{ttt}, by defining
3931 @samp{TDEPFILES=@dots{}} and @samp{TDEPLIBS=@dots{}}. Also specifies
3932 the header file which describes @var{ttt}, by defining @samp{TM_FILE=
3935 You can also define @samp{TM_CFLAGS}, @samp{TM_CLIBS}, @samp{TM_CDEPS},
3936 but these are now deprecated, replaced by autoconf, and may go away in
3937 future versions of @value{GDBN}.
3939 @item gdb/@var{ttt}-tdep.c
3940 Contains any miscellaneous code required for this target machine. On
3941 some machines it doesn't exist at all. Sometimes the macros in
3942 @file{tm-@var{ttt}.h} become very complicated, so they are implemented
3943 as functions here instead, and the macro is simply defined to call the
3944 function. This is vastly preferable, since it is easier to understand
3947 @item gdb/@var{arch}-tdep.c
3948 @itemx gdb/@var{arch}-tdep.h
3949 This often exists to describe the basic layout of the target machine's
3950 processor chip (registers, stack, etc.). If used, it is included by
3951 @file{@var{ttt}-tdep.h}. It can be shared among many targets that use
3954 @item gdb/config/@var{arch}/tm-@var{ttt}.h
3955 (@file{tm.h} is a link to this file, created by @code{configure}). Contains
3956 macro definitions about the target machine's registers, stack frame
3957 format and instructions.
3959 New targets do not need this file and should not create it.
3961 @item gdb/config/@var{arch}/tm-@var{arch}.h
3962 This often exists to describe the basic layout of the target machine's
3963 processor chip (registers, stack, etc.). If used, it is included by
3964 @file{tm-@var{ttt}.h}. It can be shared among many targets that use the
3967 New targets do not need this file and should not create it.
3971 If you are adding a new operating system for an existing CPU chip, add a
3972 @file{config/tm-@var{os}.h} file that describes the operating system
3973 facilities that are unusual (extra symbol table info; the breakpoint
3974 instruction needed; etc.). Then write a @file{@var{arch}/tm-@var{os}.h}
3975 that just @code{#include}s @file{tm-@var{arch}.h} and
3976 @file{config/tm-@var{os}.h}.
3979 @section Converting an existing Target Architecture to Multi-arch
3980 @cindex converting targets to multi-arch
3982 This section describes the current accepted best practice for converting
3983 an existing target architecture to the multi-arch framework.
3985 The process consists of generating, testing, posting and committing a
3986 sequence of patches. Each patch must contain a single change, for
3992 Directly convert a group of functions into macros (the conversion does
3993 not change the behavior of any of the functions).
3996 Replace a non-multi-arch with a multi-arch mechanism (e.g.,
4000 Enable multi-arch level one.
4003 Delete one or more files.
4008 There isn't a size limit on a patch, however, a developer is strongly
4009 encouraged to keep the patch size down.
4011 Since each patch is well defined, and since each change has been tested
4012 and shows no regressions, the patches are considered @emph{fairly}
4013 obvious. Such patches, when submitted by developers listed in the
4014 @file{MAINTAINERS} file, do not need approval. Occasional steps in the
4015 process may be more complicated and less clear. The developer is
4016 expected to use their judgment and is encouraged to seek advice as
4019 @subsection Preparation
4021 The first step is to establish control. Build (with @option{-Werror}
4022 enabled) and test the target so that there is a baseline against which
4023 the debugger can be compared.
4025 At no stage can the test results regress or @value{GDBN} stop compiling
4026 with @option{-Werror}.
4028 @subsection Add the multi-arch initialization code
4030 The objective of this step is to establish the basic multi-arch
4031 framework. It involves
4036 The addition of a @code{@var{arch}_gdbarch_init} function@footnote{The
4037 above is from the original example and uses K&R C. @value{GDBN}
4038 has since converted to ISO C but lets ignore that.} that creates
4041 static struct gdbarch *
4042 d10v_gdbarch_init (info, arches)
4043 struct gdbarch_info info;
4044 struct gdbarch_list *arches;
4046 struct gdbarch *gdbarch;
4047 /* there is only one d10v architecture */
4049 return arches->gdbarch;
4050 gdbarch = gdbarch_alloc (&info, NULL);
4058 A per-architecture dump function to print any architecture specific
4062 mips_dump_tdep (struct gdbarch *current_gdbarch,
4063 struct ui_file *file)
4065 @dots{} code to print architecture specific info @dots{}
4070 A change to @code{_initialize_@var{arch}_tdep} to register this new
4074 _initialize_mips_tdep (void)
4076 gdbarch_register (bfd_arch_mips, mips_gdbarch_init,
4081 Add the macro @code{GDB_MULTI_ARCH}, defined as 0 (zero), to the file@*
4082 @file{config/@var{arch}/tm-@var{arch}.h}.
4086 @subsection Update multi-arch incompatible mechanisms
4088 Some mechanisms do not work with multi-arch. They include:
4091 @item FRAME_FIND_SAVED_REGS
4092 Replaced with @code{DEPRECATED_FRAME_INIT_SAVED_REGS}
4096 At this stage you could also consider converting the macros into
4099 @subsection Prepare for multi-arch level to one
4101 Temporally set @code{GDB_MULTI_ARCH} to @code{GDB_MULTI_ARCH_PARTIAL}
4102 and then build and start @value{GDBN} (the change should not be
4103 committed). @value{GDBN} may not build, and once built, it may die with
4104 an internal error listing the architecture methods that must be
4107 Fix any build problems (patch(es)).
4109 Convert all the architecture methods listed, which are only macros, into
4110 functions (patch(es)).
4112 Update @code{@var{arch}_gdbarch_init} to set all the missing
4113 architecture methods and wrap the corresponding macros in @code{#if
4114 !GDB_MULTI_ARCH} (patch(es)).
4116 @subsection Set multi-arch level one
4118 Change the value of @code{GDB_MULTI_ARCH} to GDB_MULTI_ARCH_PARTIAL (a
4121 Any problems with throwing ``the switch'' should have been fixed
4124 @subsection Convert remaining macros
4126 Suggest converting macros into functions (and setting the corresponding
4127 architecture method) in small batches.
4129 @subsection Set multi-arch level to two
4131 This should go smoothly.
4133 @subsection Delete the TM file
4135 The @file{tm-@var{arch}.h} can be deleted. @file{@var{arch}.mt} and
4136 @file{configure.in} updated.
4139 @node Target Vector Definition
4141 @chapter Target Vector Definition
4142 @cindex target vector
4144 The target vector defines the interface between @value{GDBN}'s
4145 abstract handling of target systems, and the nitty-gritty code that
4146 actually exercises control over a process or a serial port.
4147 @value{GDBN} includes some 30-40 different target vectors; however,
4148 each configuration of @value{GDBN} includes only a few of them.
4150 @section File Targets
4152 Both executables and core files have target vectors.
4154 @section Standard Protocol and Remote Stubs
4156 @value{GDBN}'s file @file{remote.c} talks a serial protocol to code
4157 that runs in the target system. @value{GDBN} provides several sample
4158 @dfn{stubs} that can be integrated into target programs or operating
4159 systems for this purpose; they are named @file{*-stub.c}.
4161 The @value{GDBN} user's manual describes how to put such a stub into
4162 your target code. What follows is a discussion of integrating the
4163 SPARC stub into a complicated operating system (rather than a simple
4164 program), by Stu Grossman, the author of this stub.
4166 The trap handling code in the stub assumes the following upon entry to
4171 %l1 and %l2 contain pc and npc respectively at the time of the trap;
4177 you are in the correct trap window.
4180 As long as your trap handler can guarantee those conditions, then there
4181 is no reason why you shouldn't be able to ``share'' traps with the stub.
4182 The stub has no requirement that it be jumped to directly from the
4183 hardware trap vector. That is why it calls @code{exceptionHandler()},
4184 which is provided by the external environment. For instance, this could
4185 set up the hardware traps to actually execute code which calls the stub
4186 first, and then transfers to its own trap handler.
4188 For the most point, there probably won't be much of an issue with
4189 ``sharing'' traps, as the traps we use are usually not used by the kernel,
4190 and often indicate unrecoverable error conditions. Anyway, this is all
4191 controlled by a table, and is trivial to modify. The most important
4192 trap for us is for @code{ta 1}. Without that, we can't single step or
4193 do breakpoints. Everything else is unnecessary for the proper operation
4194 of the debugger/stub.
4196 From reading the stub, it's probably not obvious how breakpoints work.
4197 They are simply done by deposit/examine operations from @value{GDBN}.
4199 @section ROM Monitor Interface
4201 @section Custom Protocols
4203 @section Transport Layer
4205 @section Builtin Simulator
4208 @node Native Debugging
4210 @chapter Native Debugging
4211 @cindex native debugging
4213 Several files control @value{GDBN}'s configuration for native support:
4217 @item gdb/config/@var{arch}/@var{xyz}.mh
4218 Specifies Makefile fragments needed by a @emph{native} configuration on
4219 machine @var{xyz}. In particular, this lists the required
4220 native-dependent object files, by defining @samp{NATDEPFILES=@dots{}}.
4221 Also specifies the header file which describes native support on
4222 @var{xyz}, by defining @samp{NAT_FILE= nm-@var{xyz}.h}. You can also
4223 define @samp{NAT_CFLAGS}, @samp{NAT_ADD_FILES}, @samp{NAT_CLIBS},
4224 @samp{NAT_CDEPS}, etc.; see @file{Makefile.in}.
4226 @emph{Maintainer's note: The @file{.mh} suffix is because this file
4227 originally contained @file{Makefile} fragments for hosting @value{GDBN}
4228 on machine @var{xyz}. While the file is no longer used for this
4229 purpose, the @file{.mh} suffix remains. Perhaps someone will
4230 eventually rename these fragments so that they have a @file{.mn}
4233 @item gdb/config/@var{arch}/nm-@var{xyz}.h
4234 (@file{nm.h} is a link to this file, created by @code{configure}). Contains C
4235 macro definitions describing the native system environment, such as
4236 child process control and core file support.
4238 @item gdb/@var{xyz}-nat.c
4239 Contains any miscellaneous C code required for this native support of
4240 this machine. On some machines it doesn't exist at all.
4243 There are some ``generic'' versions of routines that can be used by
4244 various systems. These can be customized in various ways by macros
4245 defined in your @file{nm-@var{xyz}.h} file. If these routines work for
4246 the @var{xyz} host, you can just include the generic file's name (with
4247 @samp{.o}, not @samp{.c}) in @code{NATDEPFILES}.
4249 Otherwise, if your machine needs custom support routines, you will need
4250 to write routines that perform the same functions as the generic file.
4251 Put them into @file{@var{xyz}-nat.c}, and put @file{@var{xyz}-nat.o}
4252 into @code{NATDEPFILES}.
4256 This contains the @emph{target_ops vector} that supports Unix child
4257 processes on systems which use ptrace and wait to control the child.
4260 This contains the @emph{target_ops vector} that supports Unix child
4261 processes on systems which use /proc to control the child.
4264 This does the low-level grunge that uses Unix system calls to do a ``fork
4265 and exec'' to start up a child process.
4268 This is the low level interface to inferior processes for systems using
4269 the Unix @code{ptrace} call in a vanilla way.
4272 @section Native core file Support
4273 @cindex native core files
4276 @findex fetch_core_registers
4277 @item core-aout.c::fetch_core_registers()
4278 Support for reading registers out of a core file. This routine calls
4279 @code{register_addr()}, see below. Now that BFD is used to read core
4280 files, virtually all machines should use @code{core-aout.c}, and should
4281 just provide @code{fetch_core_registers} in @code{@var{xyz}-nat.c} (or
4282 @code{REGISTER_U_ADDR} in @code{nm-@var{xyz}.h}).
4284 @item core-aout.c::register_addr()
4285 If your @code{nm-@var{xyz}.h} file defines the macro
4286 @code{REGISTER_U_ADDR(addr, blockend, regno)}, it should be defined to
4287 set @code{addr} to the offset within the @samp{user} struct of @value{GDBN}
4288 register number @code{regno}. @code{blockend} is the offset within the
4289 ``upage'' of @code{u.u_ar0}. If @code{REGISTER_U_ADDR} is defined,
4290 @file{core-aout.c} will define the @code{register_addr()} function and
4291 use the macro in it. If you do not define @code{REGISTER_U_ADDR}, but
4292 you are using the standard @code{fetch_core_registers()}, you will need
4293 to define your own version of @code{register_addr()}, put it into your
4294 @code{@var{xyz}-nat.c} file, and be sure @code{@var{xyz}-nat.o} is in
4295 the @code{NATDEPFILES} list. If you have your own
4296 @code{fetch_core_registers()}, you may not need a separate
4297 @code{register_addr()}. Many custom @code{fetch_core_registers()}
4298 implementations simply locate the registers themselves.@refill
4301 When making @value{GDBN} run native on a new operating system, to make it
4302 possible to debug core files, you will need to either write specific
4303 code for parsing your OS's core files, or customize
4304 @file{bfd/trad-core.c}. First, use whatever @code{#include} files your
4305 machine uses to define the struct of registers that is accessible
4306 (possibly in the u-area) in a core file (rather than
4307 @file{machine/reg.h}), and an include file that defines whatever header
4308 exists on a core file (e.g. the u-area or a @code{struct core}). Then
4309 modify @code{trad_unix_core_file_p} to use these values to set up the
4310 section information for the data segment, stack segment, any other
4311 segments in the core file (perhaps shared library contents or control
4312 information), ``registers'' segment, and if there are two discontiguous
4313 sets of registers (e.g. integer and float), the ``reg2'' segment. This
4314 section information basically delimits areas in the core file in a
4315 standard way, which the section-reading routines in BFD know how to seek
4318 Then back in @value{GDBN}, you need a matching routine called
4319 @code{fetch_core_registers}. If you can use the generic one, it's in
4320 @file{core-aout.c}; if not, it's in your @file{@var{xyz}-nat.c} file.
4321 It will be passed a char pointer to the entire ``registers'' segment,
4322 its length, and a zero; or a char pointer to the entire ``regs2''
4323 segment, its length, and a 2. The routine should suck out the supplied
4324 register values and install them into @value{GDBN}'s ``registers'' array.
4326 If your system uses @file{/proc} to control processes, and uses ELF
4327 format core files, then you may be able to use the same routines for
4328 reading the registers out of processes and out of core files.
4336 @section shared libraries
4338 @section Native Conditionals
4339 @cindex native conditionals
4341 When @value{GDBN} is configured and compiled, various macros are
4342 defined or left undefined, to control compilation when the host and
4343 target systems are the same. These macros should be defined (or left
4344 undefined) in @file{nm-@var{system}.h}.
4348 @item CHILD_PREPARE_TO_STORE
4349 @findex CHILD_PREPARE_TO_STORE
4350 If the machine stores all registers at once in the child process, then
4351 define this to ensure that all values are correct. This usually entails
4352 a read from the child.
4354 [Note that this is incorrectly defined in @file{xm-@var{system}.h} files
4357 @item FETCH_INFERIOR_REGISTERS
4358 @findex FETCH_INFERIOR_REGISTERS
4359 Define this if the native-dependent code will provide its own routines
4360 @code{fetch_inferior_registers} and @code{store_inferior_registers} in
4361 @file{@var{host}-nat.c}. If this symbol is @emph{not} defined, and
4362 @file{infptrace.c} is included in this configuration, the default
4363 routines in @file{infptrace.c} are used for these functions.
4365 @item FILES_INFO_HOOK
4366 @findex FILES_INFO_HOOK
4367 (Only defined for Convex.)
4371 This macro is normally defined to be the number of the first floating
4372 point register, if the machine has such registers. As such, it would
4373 appear only in target-specific code. However, @file{/proc} support uses this
4374 to decide whether floats are in use on this target.
4376 @item GET_LONGJMP_TARGET
4377 @findex GET_LONGJMP_TARGET
4378 For most machines, this is a target-dependent parameter. On the
4379 DECstation and the Iris, this is a native-dependent parameter, since
4380 @file{setjmp.h} is needed to define it.
4382 This macro determines the target PC address that @code{longjmp} will jump to,
4383 assuming that we have just stopped at a longjmp breakpoint. It takes a
4384 @code{CORE_ADDR *} as argument, and stores the target PC value through this
4385 pointer. It examines the current state of the machine as needed.
4387 @item I386_USE_GENERIC_WATCHPOINTS
4388 An x86-based machine can define this to use the generic x86 watchpoint
4389 support; see @ref{Algorithms, I386_USE_GENERIC_WATCHPOINTS}.
4392 @findex KERNEL_U_ADDR
4393 Define this to the address of the @code{u} structure (the ``user
4394 struct'', also known as the ``u-page'') in kernel virtual memory. @value{GDBN}
4395 needs to know this so that it can subtract this address from absolute
4396 addresses in the upage, that are obtained via ptrace or from core files.
4397 On systems that don't need this value, set it to zero.
4399 @item KERNEL_U_ADDR_HPUX
4400 @findex KERNEL_U_ADDR_HPUX
4401 Define this to cause @value{GDBN} to determine the address of @code{u} at
4402 runtime, by using HP-style @code{nlist} on the kernel's image in the
4405 @item ONE_PROCESS_WRITETEXT
4406 @findex ONE_PROCESS_WRITETEXT
4407 Define this to be able to, when a breakpoint insertion fails, warn the
4408 user that another process may be running with the same executable.
4411 @findex PROC_NAME_FMT
4412 Defines the format for the name of a @file{/proc} device. Should be
4413 defined in @file{nm.h} @emph{only} in order to override the default
4414 definition in @file{procfs.c}.
4416 @item PTRACE_ARG3_TYPE
4417 @findex PTRACE_ARG3_TYPE
4418 The type of the third argument to the @code{ptrace} system call, if it
4419 exists and is different from @code{int}.
4421 @item REGISTER_U_ADDR
4422 @findex REGISTER_U_ADDR
4423 Defines the offset of the registers in the ``u area''.
4425 @item SHELL_COMMAND_CONCAT
4426 @findex SHELL_COMMAND_CONCAT
4427 If defined, is a string to prefix on the shell command used to start the
4432 If defined, this is the name of the shell to use to run the inferior.
4433 Defaults to @code{"/bin/sh"}.
4435 @item SOLIB_ADD (@var{filename}, @var{from_tty}, @var{targ}, @var{readsyms})
4437 Define this to expand into an expression that will cause the symbols in
4438 @var{filename} to be added to @value{GDBN}'s symbol table. If
4439 @var{readsyms} is zero symbols are not read but any necessary low level
4440 processing for @var{filename} is still done.
4442 @item SOLIB_CREATE_INFERIOR_HOOK
4443 @findex SOLIB_CREATE_INFERIOR_HOOK
4444 Define this to expand into any shared-library-relocation code that you
4445 want to be run just after the child process has been forked.
4447 @item START_INFERIOR_TRAPS_EXPECTED
4448 @findex START_INFERIOR_TRAPS_EXPECTED
4449 When starting an inferior, @value{GDBN} normally expects to trap
4451 the shell execs, and once when the program itself execs. If the actual
4452 number of traps is something other than 2, then define this macro to
4453 expand into the number expected.
4455 @item SVR4_SHARED_LIBS
4456 @findex SVR4_SHARED_LIBS
4457 Define this to indicate that SVR4-style shared libraries are in use.
4461 This determines whether small routines in @file{*-tdep.c}, which
4462 translate register values between @value{GDBN}'s internal
4463 representation and the @file{/proc} representation, are compiled.
4466 @findex U_REGS_OFFSET
4467 This is the offset of the registers in the upage. It need only be
4468 defined if the generic ptrace register access routines in
4469 @file{infptrace.c} are being used (that is, @file{infptrace.c} is
4470 configured in, and @code{FETCH_INFERIOR_REGISTERS} is not defined). If
4471 the default value from @file{infptrace.c} is good enough, leave it
4474 The default value means that u.u_ar0 @emph{points to} the location of
4475 the registers. I'm guessing that @code{#define U_REGS_OFFSET 0} means
4476 that @code{u.u_ar0} @emph{is} the location of the registers.
4480 See @file{objfiles.c}.
4483 @findex DEBUG_PTRACE
4484 Define this to debug @code{ptrace} calls.
4488 @node Support Libraries
4490 @chapter Support Libraries
4495 BFD provides support for @value{GDBN} in several ways:
4498 @item identifying executable and core files
4499 BFD will identify a variety of file types, including a.out, coff, and
4500 several variants thereof, as well as several kinds of core files.
4502 @item access to sections of files
4503 BFD parses the file headers to determine the names, virtual addresses,
4504 sizes, and file locations of all the various named sections in files
4505 (such as the text section or the data section). @value{GDBN} simply
4506 calls BFD to read or write section @var{x} at byte offset @var{y} for
4509 @item specialized core file support
4510 BFD provides routines to determine the failing command name stored in a
4511 core file, the signal with which the program failed, and whether a core
4512 file matches (i.e.@: could be a core dump of) a particular executable
4515 @item locating the symbol information
4516 @value{GDBN} uses an internal interface of BFD to determine where to find the
4517 symbol information in an executable file or symbol-file. @value{GDBN} itself
4518 handles the reading of symbols, since BFD does not ``understand'' debug
4519 symbols, but @value{GDBN} uses BFD's cached information to find the symbols,
4524 @cindex opcodes library
4526 The opcodes library provides @value{GDBN}'s disassembler. (It's a separate
4527 library because it's also used in binutils, for @file{objdump}).
4534 @cindex @code{libiberty} library
4536 The @code{libiberty} library provides a set of functions and features
4537 that integrate and improve on functionality found in modern operating
4538 systems. Broadly speaking, such features can be divided into three
4539 groups: supplemental functions (functions that may be missing in some
4540 environments and operating systems), replacement functions (providing
4541 a uniform and easier to use interface for commonly used standard
4542 functions), and extensions (which provide additional functionality
4543 beyond standard functions).
4545 @value{GDBN} uses various features provided by the @code{libiberty}
4546 library, for instance the C@t{++} demangler, the @acronym{IEEE}
4547 floating format support functions, the input options parser
4548 @samp{getopt}, the @samp{obstack} extension, and other functions.
4550 @subsection @code{obstacks} in @value{GDBN}
4551 @cindex @code{obstacks}
4553 The obstack mechanism provides a convenient way to allocate and free
4554 chunks of memory. Each obstack is a pool of memory that is managed
4555 like a stack. Objects (of any nature, size and alignment) are
4556 allocated and freed in a @acronym{LIFO} fashion on an obstack (see
4557 @code{libiberty}'s documenatation for a more detailed explanation of
4560 The most noticeable use of the @code{obstacks} in @value{GDBN} is in
4561 object files. There is an obstack associated with each internal
4562 representation of an object file. Lots of things get allocated on
4563 these @code{obstacks}: dictionary entries, blocks, blockvectors,
4564 symbols, minimal symbols, types, vectors of fundamental types, class
4565 fields of types, object files section lists, object files section
4566 offets lists, line tables, symbol tables, partial symbol tables,
4567 string tables, symbol table private data, macros tables, debug
4568 information sections and entries, import and export lists (som),
4569 unwind information (hppa), dwarf2 location expressions data. Plus
4570 various strings such as directory names strings, debug format strings,
4573 An essential and convenient property of all data on @code{obstacks} is
4574 that memory for it gets allocated (with @code{obstack_alloc}) at
4575 various times during a debugging sesssion, but it is released all at
4576 once using the @code{obstack_free} function. The @code{obstack_free}
4577 function takes a pointer to where in the stack it must start the
4578 deletion from (much like the cleanup chains have a pointer to where to
4579 start the cleanups). Because of the stack like structure of the
4580 @code{obstacks}, this allows to free only a top portion of the
4581 obstack. There are a few instances in @value{GDBN} where such thing
4582 happens. Calls to @code{obstack_free} are done after some local data
4583 is allocated to the obstack. Only the local data is deleted from the
4584 obstack. Of course this assumes that nothing between the
4585 @code{obstack_alloc} and the @code{obstack_free} allocates anything
4586 else on the same obstack. For this reason it is best and safest to
4587 use temporary @code{obstacks}.
4589 Releasing the whole obstack is also not safe per se. It is safe only
4590 under the condition that we know the @code{obstacks} memory is no
4591 longer needed. In @value{GDBN} we get rid of the @code{obstacks} only
4592 when we get rid of the whole objfile(s), for instance upon reading a
4596 @cindex regular expressions library
4607 @item SIGN_EXTEND_CHAR
4609 @item SWITCH_ENUM_BUG
4624 This chapter covers topics that are lower-level than the major
4625 algorithms of @value{GDBN}.
4630 Cleanups are a structured way to deal with things that need to be done
4633 When your code does something (e.g., @code{xmalloc} some memory, or
4634 @code{open} a file) that needs to be undone later (e.g., @code{xfree}
4635 the memory or @code{close} the file), it can make a cleanup. The
4636 cleanup will be done at some future point: when the command is finished
4637 and control returns to the top level; when an error occurs and the stack
4638 is unwound; or when your code decides it's time to explicitly perform
4639 cleanups. Alternatively you can elect to discard the cleanups you
4645 @item struct cleanup *@var{old_chain};
4646 Declare a variable which will hold a cleanup chain handle.
4648 @findex make_cleanup
4649 @item @var{old_chain} = make_cleanup (@var{function}, @var{arg});
4650 Make a cleanup which will cause @var{function} to be called with
4651 @var{arg} (a @code{char *}) later. The result, @var{old_chain}, is a
4652 handle that can later be passed to @code{do_cleanups} or
4653 @code{discard_cleanups}. Unless you are going to call
4654 @code{do_cleanups} or @code{discard_cleanups}, you can ignore the result
4655 from @code{make_cleanup}.
4658 @item do_cleanups (@var{old_chain});
4659 Do all cleanups added to the chain since the corresponding
4660 @code{make_cleanup} call was made.
4662 @findex discard_cleanups
4663 @item discard_cleanups (@var{old_chain});
4664 Same as @code{do_cleanups} except that it just removes the cleanups from
4665 the chain and does not call the specified functions.
4668 Cleanups are implemented as a chain. The handle returned by
4669 @code{make_cleanups} includes the cleanup passed to the call and any
4670 later cleanups appended to the chain (but not yet discarded or
4674 make_cleanup (a, 0);
4676 struct cleanup *old = make_cleanup (b, 0);
4684 will call @code{c()} and @code{b()} but will not call @code{a()}. The
4685 cleanup that calls @code{a()} will remain in the cleanup chain, and will
4686 be done later unless otherwise discarded.@refill
4688 Your function should explicitly do or discard the cleanups it creates.
4689 Failing to do this leads to non-deterministic behavior since the caller
4690 will arbitrarily do or discard your functions cleanups. This need leads
4691 to two common cleanup styles.
4693 The first style is try/finally. Before it exits, your code-block calls
4694 @code{do_cleanups} with the old cleanup chain and thus ensures that your
4695 code-block's cleanups are always performed. For instance, the following
4696 code-segment avoids a memory leak problem (even when @code{error} is
4697 called and a forced stack unwind occurs) by ensuring that the
4698 @code{xfree} will always be called:
4701 struct cleanup *old = make_cleanup (null_cleanup, 0);
4702 data = xmalloc (sizeof blah);
4703 make_cleanup (xfree, data);
4708 The second style is try/except. Before it exits, your code-block calls
4709 @code{discard_cleanups} with the old cleanup chain and thus ensures that
4710 any created cleanups are not performed. For instance, the following
4711 code segment, ensures that the file will be closed but only if there is
4715 FILE *file = fopen ("afile", "r");
4716 struct cleanup *old = make_cleanup (close_file, file);
4718 discard_cleanups (old);
4722 Some functions, e.g. @code{fputs_filtered()} or @code{error()}, specify
4723 that they ``should not be called when cleanups are not in place''. This
4724 means that any actions you need to reverse in the case of an error or
4725 interruption must be on the cleanup chain before you call these
4726 functions, since they might never return to your code (they
4727 @samp{longjmp} instead).
4729 @section Per-architecture module data
4730 @cindex per-architecture module data
4731 @cindex multi-arch data
4732 @cindex data-pointer, per-architecture/per-module
4734 The multi-arch framework includes a mechanism for adding module
4735 specific per-architecture data-pointers to the @code{struct gdbarch}
4736 architecture object.
4738 A module registers one or more per-architecture data-pointers using:
4740 @deftypefun struct gdbarch_data *gdbarch_data_register_pre_init (gdbarch_data_pre_init_ftype *@var{pre_init})
4741 @var{pre_init} is used to, on-demand, allocate an initial value for a
4742 per-architecture data-pointer using the architecture's obstack (passed
4743 in as a parameter). Since @var{pre_init} can be called during
4744 architecture creation, it is not parameterized with the architecture.
4745 and must not call modules that use per-architecture data.
4748 @deftypefun struct gdbarch_data *gdbarch_data_register_post_init (gdbarch_data_post_init_ftype *@var{post_init})
4749 @var{post_init} is used to obtain an initial value for a
4750 per-architecture data-pointer @emph{after}. Since @var{post_init} is
4751 always called after architecture creation, it both receives the fully
4752 initialized architecture and is free to call modules that use
4753 per-architecture data (care needs to be taken to ensure that those
4754 other modules do not try to call back to this module as that will
4755 create in cycles in the initialization call graph).
4758 These functions return a @code{struct gdbarch_data} that is used to
4759 identify the per-architecture data-pointer added for that module.
4761 The per-architecture data-pointer is accessed using the function:
4763 @deftypefun void *gdbarch_data (struct gdbarch *@var{gdbarch}, struct gdbarch_data *@var{data_handle})
4764 Given the architecture @var{arch} and module data handle
4765 @var{data_handle} (returned by @code{gdbarch_data_register_pre_init}
4766 or @code{gdbarch_data_register_post_init}), this function returns the
4767 current value of the per-architecture data-pointer. If the data
4768 pointer is @code{NULL}, it is first initialized by calling the
4769 corresponding @var{pre_init} or @var{post_init} method.
4772 The examples below assume the following definitions:
4775 struct nozel @{ int total; @};
4776 static struct gdbarch_data *nozel_handle;
4779 A module can extend the architecture vector, adding additional
4780 per-architecture data, using the @var{pre_init} method. The module's
4781 per-architecture data is then initialized during architecture
4784 In the below, the module's per-architecture @emph{nozel} is added. An
4785 architecture can specify its nozel by calling @code{set_gdbarch_nozel}
4786 from @code{gdbarch_init}.
4790 nozel_pre_init (struct obstack *obstack)
4792 struct nozel *data = OBSTACK_ZALLOC (obstack, struct nozel);
4799 set_gdbarch_nozel (struct gdbarch *gdbarch, int total)
4801 struct nozel *data = gdbarch_data (gdbarch, nozel_handle);
4802 data->total = nozel;
4806 A module can on-demand create architecture dependant data structures
4807 using @code{post_init}.
4809 In the below, the nozel's total is computed on-demand by
4810 @code{nozel_post_init} using information obtained from the
4815 nozel_post_init (struct gdbarch *gdbarch)
4817 struct nozel *data = GDBARCH_OBSTACK_ZALLOC (gdbarch, struct nozel);
4818 nozel->total = gdbarch@dots{} (gdbarch);
4825 nozel_total (struct gdbarch *gdbarch)
4827 struct nozel *data = gdbarch_data (gdbarch, nozel_handle);
4832 @section Wrapping Output Lines
4833 @cindex line wrap in output
4836 Output that goes through @code{printf_filtered} or @code{fputs_filtered}
4837 or @code{fputs_demangled} needs only to have calls to @code{wrap_here}
4838 added in places that would be good breaking points. The utility
4839 routines will take care of actually wrapping if the line width is
4842 The argument to @code{wrap_here} is an indentation string which is
4843 printed @emph{only} if the line breaks there. This argument is saved
4844 away and used later. It must remain valid until the next call to
4845 @code{wrap_here} or until a newline has been printed through the
4846 @code{*_filtered} functions. Don't pass in a local variable and then
4849 It is usually best to call @code{wrap_here} after printing a comma or
4850 space. If you call it before printing a space, make sure that your
4851 indentation properly accounts for the leading space that will print if
4852 the line wraps there.
4854 Any function or set of functions that produce filtered output must
4855 finish by printing a newline, to flush the wrap buffer, before switching
4856 to unfiltered (@code{printf}) output. Symbol reading routines that
4857 print warnings are a good example.
4859 @section @value{GDBN} Coding Standards
4860 @cindex coding standards
4862 @value{GDBN} follows the GNU coding standards, as described in
4863 @file{etc/standards.texi}. This file is also available for anonymous
4864 FTP from GNU archive sites. @value{GDBN} takes a strict interpretation
4865 of the standard; in general, when the GNU standard recommends a practice
4866 but does not require it, @value{GDBN} requires it.
4868 @value{GDBN} follows an additional set of coding standards specific to
4869 @value{GDBN}, as described in the following sections.
4874 @value{GDBN} assumes an ISO/IEC 9899:1990 (a.k.a.@: ISO C90) compliant
4877 @value{GDBN} does not assume an ISO C or POSIX compliant C library.
4880 @subsection Memory Management
4882 @value{GDBN} does not use the functions @code{malloc}, @code{realloc},
4883 @code{calloc}, @code{free} and @code{asprintf}.
4885 @value{GDBN} uses the functions @code{xmalloc}, @code{xrealloc} and
4886 @code{xcalloc} when allocating memory. Unlike @code{malloc} et.al.@:
4887 these functions do not return when the memory pool is empty. Instead,
4888 they unwind the stack using cleanups. These functions return
4889 @code{NULL} when requested to allocate a chunk of memory of size zero.
4891 @emph{Pragmatics: By using these functions, the need to check every
4892 memory allocation is removed. These functions provide portable
4895 @value{GDBN} does not use the function @code{free}.
4897 @value{GDBN} uses the function @code{xfree} to return memory to the
4898 memory pool. Consistent with ISO-C, this function ignores a request to
4899 free a @code{NULL} pointer.
4901 @emph{Pragmatics: On some systems @code{free} fails when passed a
4902 @code{NULL} pointer.}
4904 @value{GDBN} can use the non-portable function @code{alloca} for the
4905 allocation of small temporary values (such as strings).
4907 @emph{Pragmatics: This function is very non-portable. Some systems
4908 restrict the memory being allocated to no more than a few kilobytes.}
4910 @value{GDBN} uses the string function @code{xstrdup} and the print
4911 function @code{xstrprintf}.
4913 @emph{Pragmatics: @code{asprintf} and @code{strdup} can fail. Print
4914 functions such as @code{sprintf} are very prone to buffer overflow
4918 @subsection Compiler Warnings
4919 @cindex compiler warnings
4921 With few exceptions, developers should include the configuration option
4922 @samp{--enable-gdb-build-warnings=,-Werror} when building @value{GDBN}.
4923 The exceptions are listed in the file @file{gdb/MAINTAINERS}.
4925 This option causes @value{GDBN} (when built using GCC) to be compiled
4926 with a carefully selected list of compiler warning flags. Any warnings
4927 from those flags being treated as errors.
4929 The current list of warning flags includes:
4933 Since @value{GDBN} coding standard requires all functions to be declared
4934 using a prototype, the flag has the side effect of ensuring that
4935 prototyped functions are always visible with out resorting to
4936 @samp{-Wstrict-prototypes}.
4939 Such code often appears to work except on instruction set architectures
4940 that use register windows.
4947 @itemx -Wformat-nonliteral
4948 Since @value{GDBN} uses the @code{format printf} attribute on all
4949 @code{printf} like functions these check not just @code{printf} calls
4950 but also calls to functions such as @code{fprintf_unfiltered}.
4953 This warning includes uses of the assignment operator within an
4954 @code{if} statement.
4956 @item -Wpointer-arith
4958 @item -Wuninitialized
4960 @item -Wunused-label
4961 This warning has the additional benefit of detecting the absence of the
4962 @code{case} reserved word in a switch statement:
4964 enum @{ FD_SCHEDULED, NOTHING_SCHEDULED @} sched;
4977 @item -Wunused-function
4980 @emph{Pragmatics: Due to the way that @value{GDBN} is implemented most
4981 functions have unused parameters. Consequently the warning
4982 @samp{-Wunused-parameter} is precluded from the list. The macro
4983 @code{ATTRIBUTE_UNUSED} is not used as it leads to false negatives ---
4984 it is not an error to have @code{ATTRIBUTE_UNUSED} on a parameter that
4985 is being used. The options @samp{-Wall} and @samp{-Wunused} are also
4986 precluded because they both include @samp{-Wunused-parameter}.}
4988 @emph{Pragmatics: @value{GDBN} has not simply accepted the warnings
4989 enabled by @samp{-Wall -Werror -W...}. Instead it is selecting warnings
4990 when and where their benefits can be demonstrated.}
4992 @subsection Formatting
4994 @cindex source code formatting
4995 The standard GNU recommendations for formatting must be followed
4998 A function declaration should not have its name in column zero. A
4999 function definition should have its name in column zero.
5003 static void foo (void);
5011 @emph{Pragmatics: This simplifies scripting. Function definitions can
5012 be found using @samp{^function-name}.}
5014 There must be a space between a function or macro name and the opening
5015 parenthesis of its argument list (except for macro definitions, as
5016 required by C). There must not be a space after an open paren/bracket
5017 or before a close paren/bracket.
5019 While additional whitespace is generally helpful for reading, do not use
5020 more than one blank line to separate blocks, and avoid adding whitespace
5021 after the end of a program line (as of 1/99, some 600 lines had
5022 whitespace after the semicolon). Excess whitespace causes difficulties
5023 for @code{diff} and @code{patch} utilities.
5025 Pointers are declared using the traditional K&R C style:
5039 @subsection Comments
5041 @cindex comment formatting
5042 The standard GNU requirements on comments must be followed strictly.
5044 Block comments must appear in the following form, with no @code{/*}- or
5045 @code{*/}-only lines, and no leading @code{*}:
5048 /* Wait for control to return from inferior to debugger. If inferior
5049 gets a signal, we may decide to start it up again instead of
5050 returning. That is why there is a loop in this function. When
5051 this function actually returns it means the inferior should be left
5052 stopped and @value{GDBN} should read more commands. */
5055 (Note that this format is encouraged by Emacs; tabbing for a multi-line
5056 comment works correctly, and @kbd{M-q} fills the block consistently.)
5058 Put a blank line between the block comments preceding function or
5059 variable definitions, and the definition itself.
5061 In general, put function-body comments on lines by themselves, rather
5062 than trying to fit them into the 20 characters left at the end of a
5063 line, since either the comment or the code will inevitably get longer
5064 than will fit, and then somebody will have to move it anyhow.
5068 @cindex C data types
5069 Code must not depend on the sizes of C data types, the format of the
5070 host's floating point numbers, the alignment of anything, or the order
5071 of evaluation of expressions.
5073 @cindex function usage
5074 Use functions freely. There are only a handful of compute-bound areas
5075 in @value{GDBN} that might be affected by the overhead of a function
5076 call, mainly in symbol reading. Most of @value{GDBN}'s performance is
5077 limited by the target interface (whether serial line or system call).
5079 However, use functions with moderation. A thousand one-line functions
5080 are just as hard to understand as a single thousand-line function.
5082 @emph{Macros are bad, M'kay.}
5083 (But if you have to use a macro, make sure that the macro arguments are
5084 protected with parentheses.)
5088 Declarations like @samp{struct foo *} should be used in preference to
5089 declarations like @samp{typedef struct foo @{ @dots{} @} *foo_ptr}.
5092 @subsection Function Prototypes
5093 @cindex function prototypes
5095 Prototypes must be used when both @emph{declaring} and @emph{defining}
5096 a function. Prototypes for @value{GDBN} functions must include both the
5097 argument type and name, with the name matching that used in the actual
5098 function definition.
5100 All external functions should have a declaration in a header file that
5101 callers include, except for @code{_initialize_*} functions, which must
5102 be external so that @file{init.c} construction works, but shouldn't be
5103 visible to random source files.
5105 Where a source file needs a forward declaration of a static function,
5106 that declaration must appear in a block near the top of the source file.
5109 @subsection Internal Error Recovery
5111 During its execution, @value{GDBN} can encounter two types of errors.
5112 User errors and internal errors. User errors include not only a user
5113 entering an incorrect command but also problems arising from corrupt
5114 object files and system errors when interacting with the target.
5115 Internal errors include situations where @value{GDBN} has detected, at
5116 run time, a corrupt or erroneous situation.
5118 When reporting an internal error, @value{GDBN} uses
5119 @code{internal_error} and @code{gdb_assert}.
5121 @value{GDBN} must not call @code{abort} or @code{assert}.
5123 @emph{Pragmatics: There is no @code{internal_warning} function. Either
5124 the code detected a user error, recovered from it and issued a
5125 @code{warning} or the code failed to correctly recover from the user
5126 error and issued an @code{internal_error}.}
5128 @subsection File Names
5130 Any file used when building the core of @value{GDBN} must be in lower
5131 case. Any file used when building the core of @value{GDBN} must be 8.3
5132 unique. These requirements apply to both source and generated files.
5134 @emph{Pragmatics: The core of @value{GDBN} must be buildable on many
5135 platforms including DJGPP and MacOS/HFS. Every time an unfriendly file
5136 is introduced to the build process both @file{Makefile.in} and
5137 @file{configure.in} need to be modified accordingly. Compare the
5138 convoluted conversion process needed to transform @file{COPYING} into
5139 @file{copying.c} with the conversion needed to transform
5140 @file{version.in} into @file{version.c}.}
5142 Any file non 8.3 compliant file (that is not used when building the core
5143 of @value{GDBN}) must be added to @file{gdb/config/djgpp/fnchange.lst}.
5145 @emph{Pragmatics: This is clearly a compromise.}
5147 When @value{GDBN} has a local version of a system header file (ex
5148 @file{string.h}) the file name based on the POSIX header prefixed with
5149 @file{gdb_} (@file{gdb_string.h}). These headers should be relatively
5150 independent: they should use only macros defined by @file{configure},
5151 the compiler, or the host; they should include only system headers; they
5152 should refer only to system types. They may be shared between multiple
5153 programs, e.g.@: @value{GDBN} and @sc{gdbserver}.
5155 For other files @samp{-} is used as the separator.
5158 @subsection Include Files
5160 A @file{.c} file should include @file{defs.h} first.
5162 A @file{.c} file should directly include the @code{.h} file of every
5163 declaration and/or definition it directly refers to. It cannot rely on
5166 A @file{.h} file should directly include the @code{.h} file of every
5167 declaration and/or definition it directly refers to. It cannot rely on
5168 indirect inclusion. Exception: The file @file{defs.h} does not need to
5169 be directly included.
5171 An external declaration should only appear in one include file.
5173 An external declaration should never appear in a @code{.c} file.
5174 Exception: a declaration for the @code{_initialize} function that
5175 pacifies @option{-Wmissing-declaration}.
5177 A @code{typedef} definition should only appear in one include file.
5179 An opaque @code{struct} declaration can appear in multiple @file{.h}
5180 files. Where possible, a @file{.h} file should use an opaque
5181 @code{struct} declaration instead of an include.
5183 All @file{.h} files should be wrapped in:
5186 #ifndef INCLUDE_FILE_NAME_H
5187 #define INCLUDE_FILE_NAME_H
5193 @subsection Clean Design and Portable Implementation
5196 In addition to getting the syntax right, there's the little question of
5197 semantics. Some things are done in certain ways in @value{GDBN} because long
5198 experience has shown that the more obvious ways caused various kinds of
5201 @cindex assumptions about targets
5202 You can't assume the byte order of anything that comes from a target
5203 (including @var{value}s, object files, and instructions). Such things
5204 must be byte-swapped using @code{SWAP_TARGET_AND_HOST} in
5205 @value{GDBN}, or one of the swap routines defined in @file{bfd.h},
5206 such as @code{bfd_get_32}.
5208 You can't assume that you know what interface is being used to talk to
5209 the target system. All references to the target must go through the
5210 current @code{target_ops} vector.
5212 You can't assume that the host and target machines are the same machine
5213 (except in the ``native'' support modules). In particular, you can't
5214 assume that the target machine's header files will be available on the
5215 host machine. Target code must bring along its own header files --
5216 written from scratch or explicitly donated by their owner, to avoid
5220 Insertion of new @code{#ifdef}'s will be frowned upon. It's much better
5221 to write the code portably than to conditionalize it for various
5224 @cindex system dependencies
5225 New @code{#ifdef}'s which test for specific compilers or manufacturers
5226 or operating systems are unacceptable. All @code{#ifdef}'s should test
5227 for features. The information about which configurations contain which
5228 features should be segregated into the configuration files. Experience
5229 has proven far too often that a feature unique to one particular system
5230 often creeps into other systems; and that a conditional based on some
5231 predefined macro for your current system will become worthless over
5232 time, as new versions of your system come out that behave differently
5233 with regard to this feature.
5235 Adding code that handles specific architectures, operating systems,
5236 target interfaces, or hosts, is not acceptable in generic code.
5238 @cindex portable file name handling
5239 @cindex file names, portability
5240 One particularly notorious area where system dependencies tend to
5241 creep in is handling of file names. The mainline @value{GDBN} code
5242 assumes Posix semantics of file names: absolute file names begin with
5243 a forward slash @file{/}, slashes are used to separate leading
5244 directories, case-sensitive file names. These assumptions are not
5245 necessarily true on non-Posix systems such as MS-Windows. To avoid
5246 system-dependent code where you need to take apart or construct a file
5247 name, use the following portable macros:
5250 @findex HAVE_DOS_BASED_FILE_SYSTEM
5251 @item HAVE_DOS_BASED_FILE_SYSTEM
5252 This preprocessing symbol is defined to a non-zero value on hosts
5253 whose filesystems belong to the MS-DOS/MS-Windows family. Use this
5254 symbol to write conditional code which should only be compiled for
5257 @findex IS_DIR_SEPARATOR
5258 @item IS_DIR_SEPARATOR (@var{c})
5259 Evaluates to a non-zero value if @var{c} is a directory separator
5260 character. On Unix and GNU/Linux systems, only a slash @file{/} is
5261 such a character, but on Windows, both @file{/} and @file{\} will
5264 @findex IS_ABSOLUTE_PATH
5265 @item IS_ABSOLUTE_PATH (@var{file})
5266 Evaluates to a non-zero value if @var{file} is an absolute file name.
5267 For Unix and GNU/Linux hosts, a name which begins with a slash
5268 @file{/} is absolute. On DOS and Windows, @file{d:/foo} and
5269 @file{x:\bar} are also absolute file names.
5271 @findex FILENAME_CMP
5272 @item FILENAME_CMP (@var{f1}, @var{f2})
5273 Calls a function which compares file names @var{f1} and @var{f2} as
5274 appropriate for the underlying host filesystem. For Posix systems,
5275 this simply calls @code{strcmp}; on case-insensitive filesystems it
5276 will call @code{strcasecmp} instead.
5278 @findex DIRNAME_SEPARATOR
5279 @item DIRNAME_SEPARATOR
5280 Evaluates to a character which separates directories in
5281 @code{PATH}-style lists, typically held in environment variables.
5282 This character is @samp{:} on Unix, @samp{;} on DOS and Windows.
5284 @findex SLASH_STRING
5286 This evaluates to a constant string you should use to produce an
5287 absolute filename from leading directories and the file's basename.
5288 @code{SLASH_STRING} is @code{"/"} on most systems, but might be
5289 @code{"\\"} for some Windows-based ports.
5292 In addition to using these macros, be sure to use portable library
5293 functions whenever possible. For example, to extract a directory or a
5294 basename part from a file name, use the @code{dirname} and
5295 @code{basename} library functions (available in @code{libiberty} for
5296 platforms which don't provide them), instead of searching for a slash
5297 with @code{strrchr}.
5299 Another way to generalize @value{GDBN} along a particular interface is with an
5300 attribute struct. For example, @value{GDBN} has been generalized to handle
5301 multiple kinds of remote interfaces---not by @code{#ifdef}s everywhere, but
5302 by defining the @code{target_ops} structure and having a current target (as
5303 well as a stack of targets below it, for memory references). Whenever
5304 something needs to be done that depends on which remote interface we are
5305 using, a flag in the current target_ops structure is tested (e.g.,
5306 @code{target_has_stack}), or a function is called through a pointer in the
5307 current target_ops structure. In this way, when a new remote interface
5308 is added, only one module needs to be touched---the one that actually
5309 implements the new remote interface. Other examples of
5310 attribute-structs are BFD access to multiple kinds of object file
5311 formats, or @value{GDBN}'s access to multiple source languages.
5313 Please avoid duplicating code. For example, in @value{GDBN} 3.x all
5314 the code interfacing between @code{ptrace} and the rest of
5315 @value{GDBN} was duplicated in @file{*-dep.c}, and so changing
5316 something was very painful. In @value{GDBN} 4.x, these have all been
5317 consolidated into @file{infptrace.c}. @file{infptrace.c} can deal
5318 with variations between systems the same way any system-independent
5319 file would (hooks, @code{#if defined}, etc.), and machines which are
5320 radically different don't need to use @file{infptrace.c} at all.
5322 All debugging code must be controllable using the @samp{set debug
5323 @var{module}} command. Do not use @code{printf} to print trace
5324 messages. Use @code{fprintf_unfiltered(gdb_stdlog, ...}. Do not use
5325 @code{#ifdef DEBUG}.
5330 @chapter Porting @value{GDBN}
5331 @cindex porting to new machines
5333 Most of the work in making @value{GDBN} compile on a new machine is in
5334 specifying the configuration of the machine. This is done in a
5335 dizzying variety of header files and configuration scripts, which we
5336 hope to make more sensible soon. Let's say your new host is called an
5337 @var{xyz} (e.g., @samp{sun4}), and its full three-part configuration
5338 name is @code{@var{arch}-@var{xvend}-@var{xos}} (e.g.,
5339 @samp{sparc-sun-sunos4}). In particular:
5343 In the top level directory, edit @file{config.sub} and add @var{arch},
5344 @var{xvend}, and @var{xos} to the lists of supported architectures,
5345 vendors, and operating systems near the bottom of the file. Also, add
5346 @var{xyz} as an alias that maps to
5347 @code{@var{arch}-@var{xvend}-@var{xos}}. You can test your changes by
5351 ./config.sub @var{xyz}
5358 ./config.sub @code{@var{arch}-@var{xvend}-@var{xos}}
5362 which should both respond with @code{@var{arch}-@var{xvend}-@var{xos}}
5363 and no error messages.
5366 You need to port BFD, if that hasn't been done already. Porting BFD is
5367 beyond the scope of this manual.
5370 To configure @value{GDBN} itself, edit @file{gdb/configure.host} to recognize
5371 your system and set @code{gdb_host} to @var{xyz}, and (unless your
5372 desired target is already available) also edit @file{gdb/configure.tgt},
5373 setting @code{gdb_target} to something appropriate (for instance,
5376 @emph{Maintainer's note: Work in progress. The file
5377 @file{gdb/configure.host} originally needed to be modified when either a
5378 new native target or a new host machine was being added to @value{GDBN}.
5379 Recent changes have removed this requirement. The file now only needs
5380 to be modified when adding a new native configuration. This will likely
5381 changed again in the future.}
5384 Finally, you'll need to specify and define @value{GDBN}'s host-, native-, and
5385 target-dependent @file{.h} and @file{.c} files used for your
5391 @chapter Releasing @value{GDBN}
5392 @cindex making a new release of gdb
5394 @section Versions and Branches
5396 @subsection Version Identifiers
5398 @value{GDBN}'s version is determined by the file @file{gdb/version.in}.
5400 @value{GDBN}'s mainline uses ISO dates to differentiate between
5401 versions. The CVS repository uses @var{YYYY}-@var{MM}-@var{DD}-cvs
5402 while the corresponding snapshot uses @var{YYYYMMDD}.
5404 @value{GDBN}'s release branch uses a slightly more complicated scheme.
5405 When the branch is first cut, the mainline version identifier is
5406 prefixed with the @var{major}.@var{minor} from of the previous release
5407 series but with .90 appended. As draft releases are drawn from the
5408 branch, the minor minor number (.90) is incremented. Once the first
5409 release (@var{M}.@var{N}) has been made, the version prefix is updated
5410 to @var{M}.@var{N}.0.90 (dot zero, dot ninety). Follow on releases have
5411 an incremented minor minor version number (.0).
5413 Using 5.1 (previous) and 5.2 (current), the example below illustrates a
5414 typical sequence of version identifiers:
5418 final release from previous branch
5419 @item 2002-03-03-cvs
5420 main-line the day the branch is cut
5421 @item 5.1.90-2002-03-03-cvs
5422 corresponding branch version
5424 first draft release candidate
5425 @item 5.1.91-2002-03-17-cvs
5426 updated branch version
5428 second draft release candidate
5429 @item 5.1.92-2002-03-31-cvs
5430 updated branch version
5432 final release candidate (see below)
5435 @item 5.2.0.90-2002-04-07-cvs
5436 updated CVS branch version
5438 second official release
5445 Minor minor minor draft release candidates such as 5.2.0.91 have been
5446 omitted from the example. Such release candidates are, typically, never
5449 For 5.1.93 the bziped tar ball @file{gdb-5.1.93.tar.bz2} is just the
5450 official @file{gdb-5.2.tar} renamed and compressed.
5453 To avoid version conflicts, vendors are expected to modify the file
5454 @file{gdb/version.in} to include a vendor unique alphabetic identifier
5455 (an official @value{GDBN} release never uses alphabetic characters in
5456 its version identifer).
5458 Since @value{GDBN} does not make minor minor minor releases (e.g.,
5459 5.1.0.1) the conflict between that and a minor minor draft release
5460 identifier (e.g., 5.1.0.90) is avoided.
5463 @subsection Branches
5465 @value{GDBN} draws a release series (5.2, 5.2.1, @dots{}) from a single
5466 release branch (gdb_5_2-branch). Since minor minor minor releases
5467 (5.1.0.1) are not made, the need to branch the release branch is avoided
5468 (it also turns out that the effort required for such a a branch and
5469 release is significantly greater than the effort needed to create a new
5470 release from the head of the release branch).
5472 Releases 5.0 and 5.1 used branch and release tags of the form:
5475 gdb_N_M-YYYY-MM-DD-branchpoint
5476 gdb_N_M-YYYY-MM-DD-branch
5477 gdb_M_N-YYYY-MM-DD-release
5480 Release 5.2 is trialing the branch and release tags:
5483 gdb_N_M-YYYY-MM-DD-branchpoint
5485 gdb_M_N-YYYY-MM-DD-release
5488 @emph{Pragmatics: The branchpoint and release tags need to identify when
5489 a branch and release are made. The branch tag, denoting the head of the
5490 branch, does not have this criteria.}
5493 @section Branch Commit Policy
5495 The branch commit policy is pretty slack. @value{GDBN} releases 5.0,
5496 5.1 and 5.2 all used the below:
5500 The @file{gdb/MAINTAINERS} file still holds.
5502 Don't fix something on the branch unless/until it is also fixed in the
5503 trunk. If this isn't possible, mentioning it in the @file{gdb/PROBLEMS}
5504 file is better than committing a hack.
5506 When considering a patch for the branch, suggested criteria include:
5507 Does it fix a build? Does it fix the sequence @kbd{break main; run}
5508 when debugging a static binary?
5510 The further a change is from the core of @value{GDBN}, the less likely
5511 the change will worry anyone (e.g., target specific code).
5513 Only post a proposal to change the core of @value{GDBN} after you've
5514 sent individual bribes to all the people listed in the
5515 @file{MAINTAINERS} file @t{;-)}
5518 @emph{Pragmatics: Provided updates are restricted to non-core
5519 functionality there is little chance that a broken change will be fatal.
5520 This means that changes such as adding a new architectures or (within
5521 reason) support for a new host are considered acceptable.}
5524 @section Obsoleting code
5526 Before anything else, poke the other developers (and around the source
5527 code) to see if there is anything that can be removed from @value{GDBN}
5528 (an old target, an unused file).
5530 Obsolete code is identified by adding an @code{OBSOLETE} prefix to every
5531 line. Doing this means that it is easy to identify something that has
5532 been obsoleted when greping through the sources.
5534 The process is done in stages --- this is mainly to ensure that the
5535 wider @value{GDBN} community has a reasonable opportunity to respond.
5536 Remember, everything on the Internet takes a week.
5540 Post the proposal on @email{gdb@@sources.redhat.com, the GDB mailing
5541 list} Creating a bug report to track the task's state, is also highly
5546 Post the proposal on @email{gdb-announce@@sources.redhat.com, the GDB
5547 Announcement mailing list}.
5551 Go through and edit all relevant files and lines so that they are
5552 prefixed with the word @code{OBSOLETE}.
5554 Wait until the next GDB version, containing this obsolete code, has been
5557 Remove the obsolete code.
5561 @emph{Maintainer note: While removing old code is regrettable it is
5562 hopefully better for @value{GDBN}'s long term development. Firstly it
5563 helps the developers by removing code that is either no longer relevant
5564 or simply wrong. Secondly since it removes any history associated with
5565 the file (effectively clearing the slate) the developer has a much freer
5566 hand when it comes to fixing broken files.}
5570 @section Before the Branch
5572 The most important objective at this stage is to find and fix simple
5573 changes that become a pain to track once the branch is created. For
5574 instance, configuration problems that stop @value{GDBN} from even
5575 building. If you can't get the problem fixed, document it in the
5576 @file{gdb/PROBLEMS} file.
5578 @subheading Prompt for @file{gdb/NEWS}
5580 People always forget. Send a post reminding them but also if you know
5581 something interesting happened add it yourself. The @code{schedule}
5582 script will mention this in its e-mail.
5584 @subheading Review @file{gdb/README}
5586 Grab one of the nightly snapshots and then walk through the
5587 @file{gdb/README} looking for anything that can be improved. The
5588 @code{schedule} script will mention this in its e-mail.
5590 @subheading Refresh any imported files.
5592 A number of files are taken from external repositories. They include:
5596 @file{texinfo/texinfo.tex}
5598 @file{config.guess} et.@: al.@: (see the top-level @file{MAINTAINERS}
5601 @file{etc/standards.texi}, @file{etc/make-stds.texi}
5604 @subheading Check the ARI
5606 @uref{http://sources.redhat.com/gdb/ari,,A.R.I.} is an @code{awk} script
5607 (Awk Regression Index ;-) that checks for a number of errors and coding
5608 conventions. The checks include things like using @code{malloc} instead
5609 of @code{xmalloc} and file naming problems. There shouldn't be any
5612 @subsection Review the bug data base
5614 Close anything obviously fixed.
5616 @subsection Check all cross targets build
5618 The targets are listed in @file{gdb/MAINTAINERS}.
5621 @section Cut the Branch
5623 @subheading Create the branch
5628 $ V=`echo $v | sed 's/\./_/g'`
5629 $ D=`date -u +%Y-%m-%d`
5632 $ echo cvs -f -d :ext:sources.redhat.com:/cvs/src rtag \
5633 -D $D-gmt gdb_$V-$D-branchpoint insight+dejagnu
5634 cvs -f -d :ext:sources.redhat.com:/cvs/src rtag
5635 -D 2002-03-03-gmt gdb_5_2-2002-03-03-branchpoint insight+dejagnu
5638 $ echo cvs -f -d :ext:sources.redhat.com:/cvs/src rtag \
5639 -b -r gdb_$V-$D-branchpoint gdb_$V-branch insight+dejagnu
5640 cvs -f -d :ext:sources.redhat.com:/cvs/src rtag \
5641 -b -r gdb_5_2-2002-03-03-branchpoint gdb_5_2-branch insight+dejagnu
5649 by using @kbd{-D YYYY-MM-DD-gmt} the branch is forced to an exact
5652 the trunk is first taged so that the branch point can easily be found
5654 Insight (which includes GDB) and dejagnu are all tagged at the same time
5656 @file{version.in} gets bumped to avoid version number conflicts
5658 the reading of @file{.cvsrc} is disabled using @file{-f}
5661 @subheading Update @file{version.in}
5666 $ V=`echo $v | sed 's/\./_/g'`
5670 $ echo cvs -f -d :ext:sources.redhat.com:/cvs/src co \
5671 -r gdb_$V-branch src/gdb/version.in
5672 cvs -f -d :ext:sources.redhat.com:/cvs/src co
5673 -r gdb_5_2-branch src/gdb/version.in
5675 U src/gdb/version.in
5677 $ echo $u.90-0000-00-00-cvs > version.in
5679 5.1.90-0000-00-00-cvs
5680 $ cvs -f commit version.in
5685 @file{0000-00-00} is used as a date to pump prime the version.in update
5688 @file{.90} and the previous branch version are used as fairly arbitrary
5689 initial branch version number
5693 @subheading Update the web and news pages
5697 @subheading Tweak cron to track the new branch
5699 The file @file{gdbadmin/cron/crontab} contains gdbadmin's cron table.
5700 This file needs to be updated so that:
5704 a daily timestamp is added to the file @file{version.in}
5706 the new branch is included in the snapshot process
5710 See the file @file{gdbadmin/cron/README} for how to install the updated
5713 The file @file{gdbadmin/ss/README} should also be reviewed to reflect
5714 any changes. That file is copied to both the branch/ and current/
5715 snapshot directories.
5718 @subheading Update the NEWS and README files
5720 The @file{NEWS} file needs to be updated so that on the branch it refers
5721 to @emph{changes in the current release} while on the trunk it also
5722 refers to @emph{changes since the current release}.
5724 The @file{README} file needs to be updated so that it refers to the
5727 @subheading Post the branch info
5729 Send an announcement to the mailing lists:
5733 @email{gdb-announce@@sources.redhat.com, GDB Announcement mailing list}
5735 @email{gdb@@sources.redhat.com, GDB Discsussion mailing list} and
5736 @email{gdb-testers@@sources.redhat.com, GDB Discsussion mailing list}
5739 @emph{Pragmatics: The branch creation is sent to the announce list to
5740 ensure that people people not subscribed to the higher volume discussion
5743 The announcement should include:
5749 how to check out the branch using CVS
5751 the date/number of weeks until the release
5753 the branch commit policy
5757 @section Stabilize the branch
5759 Something goes here.
5761 @section Create a Release
5763 The process of creating and then making available a release is broken
5764 down into a number of stages. The first part addresses the technical
5765 process of creating a releasable tar ball. The later stages address the
5766 process of releasing that tar ball.
5768 When making a release candidate just the first section is needed.
5770 @subsection Create a release candidate
5772 The objective at this stage is to create a set of tar balls that can be
5773 made available as a formal release (or as a less formal release
5776 @subsubheading Freeze the branch
5778 Send out an e-mail notifying everyone that the branch is frozen to
5779 @email{gdb-patches@@sources.redhat.com}.
5781 @subsubheading Establish a few defaults.
5786 $ t=/sourceware/snapshot-tmp/gdbadmin-tmp
5788 /sourceware/snapshot-tmp/gdbadmin-tmp/gdb_5_2-branch/5.2
5792 /sourceware/snapshot-tmp/gdbadmin-tmp/gdb_5_2-branch/5.2
5794 /home/gdbadmin/bin/autoconf
5803 Check the @code{autoconf} version carefully. You want to be using the
5804 version taken from the @file{binutils} snapshot directory, which can be
5805 found at @uref{ftp://sources.redhat.com/pub/binutils/}. It is very
5806 unlikely that a system installed version of @code{autoconf} (e.g.,
5807 @file{/usr/bin/autoconf}) is correct.
5810 @subsubheading Check out the relevant modules:
5813 $ for m in gdb insight dejagnu
5815 ( mkdir -p $m && cd $m && cvs -q -f -d /cvs/src co -P -r $b $m )
5825 The reading of @file{.cvsrc} is disabled (@file{-f}) so that there isn't
5826 any confusion between what is written here and what your local
5827 @code{cvs} really does.
5830 @subsubheading Update relevant files.
5836 Major releases get their comments added as part of the mainline. Minor
5837 releases should probably mention any significant bugs that were fixed.
5839 Don't forget to include the @file{ChangeLog} entry.
5842 $ emacs gdb/src/gdb/NEWS
5847 $ cp gdb/src/gdb/NEWS insight/src/gdb/NEWS
5848 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
5853 You'll need to update:
5865 $ emacs gdb/src/gdb/README
5870 $ cp gdb/src/gdb/README insight/src/gdb/README
5871 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
5874 @emph{Maintainer note: Hopefully the @file{README} file was reviewed
5875 before the initial branch was cut so just a simple substitute is needed
5878 @emph{Maintainer note: Other projects generate @file{README} and
5879 @file{INSTALL} from the core documentation. This might be worth
5882 @item gdb/version.in
5885 $ echo $v > gdb/src/gdb/version.in
5886 $ cat gdb/src/gdb/version.in
5888 $ emacs gdb/src/gdb/version.in
5891 ... Bump to version ...
5893 $ cp gdb/src/gdb/version.in insight/src/gdb/version.in
5894 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
5897 @item dejagnu/src/dejagnu/configure.in
5899 Dejagnu is more complicated. The version number is a parameter to
5900 @code{AM_INIT_AUTOMAKE}. Tweak it to read something like gdb-5.1.91.
5902 Don't forget to re-generate @file{configure}.
5904 Don't forget to include a @file{ChangeLog} entry.
5907 $ emacs dejagnu/src/dejagnu/configure.in
5912 $ ( cd dejagnu/src/dejagnu && autoconf )
5917 @subsubheading Do the dirty work
5919 This is identical to the process used to create the daily snapshot.
5922 $ for m in gdb insight
5924 ( cd $m/src && gmake -f src-release $m.tar )
5926 $ ( m=dejagnu; cd $m/src && gmake -f src-release $m.tar.bz2 )
5929 If the top level source directory does not have @file{src-release}
5930 (@value{GDBN} version 5.3.1 or earlier), try these commands instead:
5933 $ for m in gdb insight
5935 ( cd $m/src && gmake -f Makefile.in $m.tar )
5937 $ ( m=dejagnu; cd $m/src && gmake -f Makefile.in $m.tar.bz2 )
5940 @subsubheading Check the source files
5942 You're looking for files that have mysteriously disappeared.
5943 @kbd{distclean} has the habit of deleting files it shouldn't. Watch out
5944 for the @file{version.in} update @kbd{cronjob}.
5947 $ ( cd gdb/src && cvs -f -q -n update )
5951 @dots{} lots of generated files @dots{}
5956 @dots{} lots of generated files @dots{}
5961 @emph{Don't worry about the @file{gdb.info-??} or
5962 @file{gdb/p-exp.tab.c}. They were generated (and yes @file{gdb.info-1}
5963 was also generated only something strange with CVS means that they
5964 didn't get supressed). Fixing it would be nice though.}
5966 @subsubheading Create compressed versions of the release
5972 dejagnu/ dejagnu-gdb-5.2.tar.bz2 gdb/ gdb-5.2.tar insight/ insight-5.2.tar
5973 $ for m in gdb insight
5975 bzip2 -v -9 -c $m-$v.tar > $m-$v.tar.bz2
5976 gzip -v -9 -c $m-$v.tar > $m-$v.tar.gz
5986 A pipe such as @kbd{bunzip2 < xxx.bz2 | gzip -9 > xxx.gz} is not since,
5987 in that mode, @code{gzip} does not know the name of the file and, hence,
5988 can not include it in the compressed file. This is also why the release
5989 process runs @code{tar} and @code{bzip2} as separate passes.
5992 @subsection Sanity check the tar ball
5994 Pick a popular machine (Solaris/PPC?) and try the build on that.
5997 $ bunzip2 < gdb-5.2.tar.bz2 | tar xpf -
6002 $ ./gdb/gdb ./gdb/gdb
6006 Breakpoint 1 at 0x80732bc: file main.c, line 734.
6008 Starting program: /tmp/gdb-5.2/gdb/gdb
6010 Breakpoint 1, main (argc=1, argv=0xbffff8b4) at main.c:734
6011 734 catch_errors (captured_main, &args, "", RETURN_MASK_ALL);
6013 $1 = @{argc = 136426532, argv = 0x821b7f0@}
6017 @subsection Make a release candidate available
6019 If this is a release candidate then the only remaining steps are:
6023 Commit @file{version.in} and @file{ChangeLog}
6025 Tweak @file{version.in} (and @file{ChangeLog} to read
6026 @var{L}.@var{M}.@var{N}-0000-00-00-cvs so that the version update
6027 process can restart.
6029 Make the release candidate available in
6030 @uref{ftp://sources.redhat.com/pub/gdb/snapshots/branch}
6032 Notify the relevant mailing lists ( @email{gdb@@sources.redhat.com} and
6033 @email{gdb-testers@@sources.redhat.com} that the candidate is available.
6036 @subsection Make a formal release available
6038 (And you thought all that was required was to post an e-mail.)
6040 @subsubheading Install on sware
6042 Copy the new files to both the release and the old release directory:
6045 $ cp *.bz2 *.gz ~ftp/pub/gdb/old-releases/
6046 $ cp *.bz2 *.gz ~ftp/pub/gdb/releases
6050 Clean up the releases directory so that only the most recent releases
6051 are available (e.g. keep 5.2 and 5.2.1 but remove 5.1):
6054 $ cd ~ftp/pub/gdb/releases
6059 Update the file @file{README} and @file{.message} in the releases
6066 $ ln README .message
6069 @subsubheading Update the web pages.
6073 @item htdocs/download/ANNOUNCEMENT
6074 This file, which is posted as the official announcement, includes:
6077 General announcement
6079 News. If making an @var{M}.@var{N}.1 release, retain the news from
6080 earlier @var{M}.@var{N} release.
6085 @item htdocs/index.html
6086 @itemx htdocs/news/index.html
6087 @itemx htdocs/download/index.html
6088 These files include:
6091 announcement of the most recent release
6093 news entry (remember to update both the top level and the news directory).
6095 These pages also need to be regenerate using @code{index.sh}.
6097 @item download/onlinedocs/
6098 You need to find the magic command that is used to generate the online
6099 docs from the @file{.tar.bz2}. The best way is to look in the output
6100 from one of the nightly @code{cron} jobs and then just edit accordingly.
6104 $ ~/ss/update-web-docs \
6105 ~ftp/pub/gdb/releases/gdb-5.2.tar.bz2 \
6107 /www/sourceware/htdocs/gdb/download/onlinedocs \
6112 Just like the online documentation. Something like:
6115 $ /bin/sh ~/ss/update-web-ari \
6116 ~ftp/pub/gdb/releases/gdb-5.2.tar.bz2 \
6118 /www/sourceware/htdocs/gdb/download/ari \
6124 @subsubheading Shadow the pages onto gnu
6126 Something goes here.
6129 @subsubheading Install the @value{GDBN} tar ball on GNU
6131 At the time of writing, the GNU machine was @kbd{gnudist.gnu.org} in
6132 @file{~ftp/gnu/gdb}.
6134 @subsubheading Make the @file{ANNOUNCEMENT}
6136 Post the @file{ANNOUNCEMENT} file you created above to:
6140 @email{gdb-announce@@sources.redhat.com, GDB Announcement mailing list}
6142 @email{info-gnu@@gnu.org, General GNU Announcement list} (but delay it a
6143 day or so to let things get out)
6145 @email{bug-gdb@@gnu.org, GDB Bug Report mailing list}
6150 The release is out but you're still not finished.
6152 @subsubheading Commit outstanding changes
6154 In particular you'll need to commit any changes to:
6158 @file{gdb/ChangeLog}
6160 @file{gdb/version.in}
6167 @subsubheading Tag the release
6172 $ d=`date -u +%Y-%m-%d`
6175 $ ( cd insight/src/gdb && cvs -f -q update )
6176 $ ( cd insight/src && cvs -f -q tag gdb_5_2-$d-release )
6179 Insight is used since that contains more of the release than
6180 @value{GDBN} (@code{dejagnu} doesn't get tagged but I think we can live
6183 @subsubheading Mention the release on the trunk
6185 Just put something in the @file{ChangeLog} so that the trunk also
6186 indicates when the release was made.
6188 @subsubheading Restart @file{gdb/version.in}
6190 If @file{gdb/version.in} does not contain an ISO date such as
6191 @kbd{2002-01-24} then the daily @code{cronjob} won't update it. Having
6192 committed all the release changes it can be set to
6193 @file{5.2.0_0000-00-00-cvs} which will restart things (yes the @kbd{_}
6194 is important - it affects the snapshot process).
6196 Don't forget the @file{ChangeLog}.
6198 @subsubheading Merge into trunk
6200 The files committed to the branch may also need changes merged into the
6203 @subsubheading Revise the release schedule
6205 Post a revised release schedule to @email{gdb@@sources.redhat.com, GDB
6206 Discussion List} with an updated announcement. The schedule can be
6207 generated by running:
6210 $ ~/ss/schedule `date +%s` schedule
6214 The first parameter is approximate date/time in seconds (from the epoch)
6215 of the most recent release.
6217 Also update the schedule @code{cronjob}.
6219 @section Post release
6221 Remove any @code{OBSOLETE} code.
6228 The testsuite is an important component of the @value{GDBN} package.
6229 While it is always worthwhile to encourage user testing, in practice
6230 this is rarely sufficient; users typically use only a small subset of
6231 the available commands, and it has proven all too common for a change
6232 to cause a significant regression that went unnoticed for some time.
6234 The @value{GDBN} testsuite uses the DejaGNU testing framework.
6235 DejaGNU is built using @code{Tcl} and @code{expect}. The tests
6236 themselves are calls to various @code{Tcl} procs; the framework runs all the
6237 procs and summarizes the passes and fails.
6239 @section Using the Testsuite
6241 @cindex running the test suite
6242 To run the testsuite, simply go to the @value{GDBN} object directory (or to the
6243 testsuite's objdir) and type @code{make check}. This just sets up some
6244 environment variables and invokes DejaGNU's @code{runtest} script. While
6245 the testsuite is running, you'll get mentions of which test file is in use,
6246 and a mention of any unexpected passes or fails. When the testsuite is
6247 finished, you'll get a summary that looks like this:
6252 # of expected passes 6016
6253 # of unexpected failures 58
6254 # of unexpected successes 5
6255 # of expected failures 183
6256 # of unresolved testcases 3
6257 # of untested testcases 5
6260 The ideal test run consists of expected passes only; however, reality
6261 conspires to keep us from this ideal. Unexpected failures indicate
6262 real problems, whether in @value{GDBN} or in the testsuite. Expected
6263 failures are still failures, but ones which have been decided are too
6264 hard to deal with at the time; for instance, a test case might work
6265 everywhere except on AIX, and there is no prospect of the AIX case
6266 being fixed in the near future. Expected failures should not be added
6267 lightly, since you may be masking serious bugs in @value{GDBN}.
6268 Unexpected successes are expected fails that are passing for some
6269 reason, while unresolved and untested cases often indicate some minor
6270 catastrophe, such as the compiler being unable to deal with a test
6273 When making any significant change to @value{GDBN}, you should run the
6274 testsuite before and after the change, to confirm that there are no
6275 regressions. Note that truly complete testing would require that you
6276 run the testsuite with all supported configurations and a variety of
6277 compilers; however this is more than really necessary. In many cases
6278 testing with a single configuration is sufficient. Other useful
6279 options are to test one big-endian (Sparc) and one little-endian (x86)
6280 host, a cross config with a builtin simulator (powerpc-eabi,
6281 mips-elf), or a 64-bit host (Alpha).
6283 If you add new functionality to @value{GDBN}, please consider adding
6284 tests for it as well; this way future @value{GDBN} hackers can detect
6285 and fix their changes that break the functionality you added.
6286 Similarly, if you fix a bug that was not previously reported as a test
6287 failure, please add a test case for it. Some cases are extremely
6288 difficult to test, such as code that handles host OS failures or bugs
6289 in particular versions of compilers, and it's OK not to try to write
6290 tests for all of those.
6292 @section Testsuite Organization
6294 @cindex test suite organization
6295 The testsuite is entirely contained in @file{gdb/testsuite}. While the
6296 testsuite includes some makefiles and configury, these are very minimal,
6297 and used for little besides cleaning up, since the tests themselves
6298 handle the compilation of the programs that @value{GDBN} will run. The file
6299 @file{testsuite/lib/gdb.exp} contains common utility procs useful for
6300 all @value{GDBN} tests, while the directory @file{testsuite/config} contains
6301 configuration-specific files, typically used for special-purpose
6302 definitions of procs like @code{gdb_load} and @code{gdb_start}.
6304 The tests themselves are to be found in @file{testsuite/gdb.*} and
6305 subdirectories of those. The names of the test files must always end
6306 with @file{.exp}. DejaGNU collects the test files by wildcarding
6307 in the test directories, so both subdirectories and individual files
6308 get chosen and run in alphabetical order.
6310 The following table lists the main types of subdirectories and what they
6311 are for. Since DejaGNU finds test files no matter where they are
6312 located, and since each test file sets up its own compilation and
6313 execution environment, this organization is simply for convenience and
6318 This is the base testsuite. The tests in it should apply to all
6319 configurations of @value{GDBN} (but generic native-only tests may live here).
6320 The test programs should be in the subset of C that is valid K&R,
6321 ANSI/ISO, and C@t{++} (@code{#ifdef}s are allowed if necessary, for instance
6324 @item gdb.@var{lang}
6325 Language-specific tests for any language @var{lang} besides C. Examples are
6326 @file{gdb.cp} and @file{gdb.java}.
6328 @item gdb.@var{platform}
6329 Non-portable tests. The tests are specific to a specific configuration
6330 (host or target), such as HP-UX or eCos. Example is @file{gdb.hp}, for
6333 @item gdb.@var{compiler}
6334 Tests specific to a particular compiler. As of this writing (June
6335 1999), there aren't currently any groups of tests in this category that
6336 couldn't just as sensibly be made platform-specific, but one could
6337 imagine a @file{gdb.gcc}, for tests of @value{GDBN}'s handling of GCC
6340 @item gdb.@var{subsystem}
6341 Tests that exercise a specific @value{GDBN} subsystem in more depth. For
6342 instance, @file{gdb.disasm} exercises various disassemblers, while
6343 @file{gdb.stabs} tests pathways through the stabs symbol reader.
6346 @section Writing Tests
6347 @cindex writing tests
6349 In many areas, the @value{GDBN} tests are already quite comprehensive; you
6350 should be able to copy existing tests to handle new cases.
6352 You should try to use @code{gdb_test} whenever possible, since it
6353 includes cases to handle all the unexpected errors that might happen.
6354 However, it doesn't cost anything to add new test procedures; for
6355 instance, @file{gdb.base/exprs.exp} defines a @code{test_expr} that
6356 calls @code{gdb_test} multiple times.
6358 Only use @code{send_gdb} and @code{gdb_expect} when absolutely
6359 necessary, such as when @value{GDBN} has several valid responses to a command.
6361 The source language programs do @emph{not} need to be in a consistent
6362 style. Since @value{GDBN} is used to debug programs written in many different
6363 styles, it's worth having a mix of styles in the testsuite; for
6364 instance, some @value{GDBN} bugs involving the display of source lines would
6365 never manifest themselves if the programs used GNU coding style
6372 Check the @file{README} file, it often has useful information that does not
6373 appear anywhere else in the directory.
6376 * Getting Started:: Getting started working on @value{GDBN}
6377 * Debugging GDB:: Debugging @value{GDBN} with itself
6380 @node Getting Started,,, Hints
6382 @section Getting Started
6384 @value{GDBN} is a large and complicated program, and if you first starting to
6385 work on it, it can be hard to know where to start. Fortunately, if you
6386 know how to go about it, there are ways to figure out what is going on.
6388 This manual, the @value{GDBN} Internals manual, has information which applies
6389 generally to many parts of @value{GDBN}.
6391 Information about particular functions or data structures are located in
6392 comments with those functions or data structures. If you run across a
6393 function or a global variable which does not have a comment correctly
6394 explaining what is does, this can be thought of as a bug in @value{GDBN}; feel
6395 free to submit a bug report, with a suggested comment if you can figure
6396 out what the comment should say. If you find a comment which is
6397 actually wrong, be especially sure to report that.
6399 Comments explaining the function of macros defined in host, target, or
6400 native dependent files can be in several places. Sometimes they are
6401 repeated every place the macro is defined. Sometimes they are where the
6402 macro is used. Sometimes there is a header file which supplies a
6403 default definition of the macro, and the comment is there. This manual
6404 also documents all the available macros.
6405 @c (@pxref{Host Conditionals}, @pxref{Target
6406 @c Conditionals}, @pxref{Native Conditionals}, and @pxref{Obsolete
6409 Start with the header files. Once you have some idea of how
6410 @value{GDBN}'s internal symbol tables are stored (see @file{symtab.h},
6411 @file{gdbtypes.h}), you will find it much easier to understand the
6412 code which uses and creates those symbol tables.
6414 You may wish to process the information you are getting somehow, to
6415 enhance your understanding of it. Summarize it, translate it to another
6416 language, add some (perhaps trivial or non-useful) feature to @value{GDBN}, use
6417 the code to predict what a test case would do and write the test case
6418 and verify your prediction, etc. If you are reading code and your eyes
6419 are starting to glaze over, this is a sign you need to use a more active
6422 Once you have a part of @value{GDBN} to start with, you can find more
6423 specifically the part you are looking for by stepping through each
6424 function with the @code{next} command. Do not use @code{step} or you
6425 will quickly get distracted; when the function you are stepping through
6426 calls another function try only to get a big-picture understanding
6427 (perhaps using the comment at the beginning of the function being
6428 called) of what it does. This way you can identify which of the
6429 functions being called by the function you are stepping through is the
6430 one which you are interested in. You may need to examine the data
6431 structures generated at each stage, with reference to the comments in
6432 the header files explaining what the data structures are supposed to
6435 Of course, this same technique can be used if you are just reading the
6436 code, rather than actually stepping through it. The same general
6437 principle applies---when the code you are looking at calls something
6438 else, just try to understand generally what the code being called does,
6439 rather than worrying about all its details.
6441 @cindex command implementation
6442 A good place to start when tracking down some particular area is with
6443 a command which invokes that feature. Suppose you want to know how
6444 single-stepping works. As a @value{GDBN} user, you know that the
6445 @code{step} command invokes single-stepping. The command is invoked
6446 via command tables (see @file{command.h}); by convention the function
6447 which actually performs the command is formed by taking the name of
6448 the command and adding @samp{_command}, or in the case of an
6449 @code{info} subcommand, @samp{_info}. For example, the @code{step}
6450 command invokes the @code{step_command} function and the @code{info
6451 display} command invokes @code{display_info}. When this convention is
6452 not followed, you might have to use @code{grep} or @kbd{M-x
6453 tags-search} in emacs, or run @value{GDBN} on itself and set a
6454 breakpoint in @code{execute_command}.
6456 @cindex @code{bug-gdb} mailing list
6457 If all of the above fail, it may be appropriate to ask for information
6458 on @code{bug-gdb}. But @emph{never} post a generic question like ``I was
6459 wondering if anyone could give me some tips about understanding
6460 @value{GDBN}''---if we had some magic secret we would put it in this manual.
6461 Suggestions for improving the manual are always welcome, of course.
6463 @node Debugging GDB,,,Hints
6465 @section Debugging @value{GDBN} with itself
6466 @cindex debugging @value{GDBN}
6468 If @value{GDBN} is limping on your machine, this is the preferred way to get it
6469 fully functional. Be warned that in some ancient Unix systems, like
6470 Ultrix 4.2, a program can't be running in one process while it is being
6471 debugged in another. Rather than typing the command @kbd{@w{./gdb
6472 ./gdb}}, which works on Suns and such, you can copy @file{gdb} to
6473 @file{gdb2} and then type @kbd{@w{./gdb ./gdb2}}.
6475 When you run @value{GDBN} in the @value{GDBN} source directory, it will read a
6476 @file{.gdbinit} file that sets up some simple things to make debugging
6477 gdb easier. The @code{info} command, when executed without a subcommand
6478 in a @value{GDBN} being debugged by gdb, will pop you back up to the top level
6479 gdb. See @file{.gdbinit} for details.
6481 If you use emacs, you will probably want to do a @code{make TAGS} after
6482 you configure your distribution; this will put the machine dependent
6483 routines for your local machine where they will be accessed first by
6486 Also, make sure that you've either compiled @value{GDBN} with your local cc, or
6487 have run @code{fixincludes} if you are compiling with gcc.
6489 @section Submitting Patches
6491 @cindex submitting patches
6492 Thanks for thinking of offering your changes back to the community of
6493 @value{GDBN} users. In general we like to get well designed enhancements.
6494 Thanks also for checking in advance about the best way to transfer the
6497 The @value{GDBN} maintainers will only install ``cleanly designed'' patches.
6498 This manual summarizes what we believe to be clean design for @value{GDBN}.
6500 If the maintainers don't have time to put the patch in when it arrives,
6501 or if there is any question about a patch, it goes into a large queue
6502 with everyone else's patches and bug reports.
6504 @cindex legal papers for code contributions
6505 The legal issue is that to incorporate substantial changes requires a
6506 copyright assignment from you and/or your employer, granting ownership
6507 of the changes to the Free Software Foundation. You can get the
6508 standard documents for doing this by sending mail to @code{gnu@@gnu.org}
6509 and asking for it. We recommend that people write in "All programs
6510 owned by the Free Software Foundation" as "NAME OF PROGRAM", so that
6511 changes in many programs (not just @value{GDBN}, but GAS, Emacs, GCC,
6513 contributed with only one piece of legalese pushed through the
6514 bureaucracy and filed with the FSF. We can't start merging changes until
6515 this paperwork is received by the FSF (their rules, which we follow
6516 since we maintain it for them).
6518 Technically, the easiest way to receive changes is to receive each
6519 feature as a small context diff or unidiff, suitable for @code{patch}.
6520 Each message sent to me should include the changes to C code and
6521 header files for a single feature, plus @file{ChangeLog} entries for
6522 each directory where files were modified, and diffs for any changes
6523 needed to the manuals (@file{gdb/doc/gdb.texinfo} or
6524 @file{gdb/doc/gdbint.texinfo}). If there are a lot of changes for a
6525 single feature, they can be split down into multiple messages.
6527 In this way, if we read and like the feature, we can add it to the
6528 sources with a single patch command, do some testing, and check it in.
6529 If you leave out the @file{ChangeLog}, we have to write one. If you leave
6530 out the doc, we have to puzzle out what needs documenting. Etc., etc.
6532 The reason to send each change in a separate message is that we will not
6533 install some of the changes. They'll be returned to you with questions
6534 or comments. If we're doing our job correctly, the message back to you
6535 will say what you have to fix in order to make the change acceptable.
6536 The reason to have separate messages for separate features is so that
6537 the acceptable changes can be installed while one or more changes are
6538 being reworked. If multiple features are sent in a single message, we
6539 tend to not put in the effort to sort out the acceptable changes from
6540 the unacceptable, so none of the features get installed until all are
6543 If this sounds painful or authoritarian, well, it is. But we get a lot
6544 of bug reports and a lot of patches, and many of them don't get
6545 installed because we don't have the time to finish the job that the bug
6546 reporter or the contributor could have done. Patches that arrive
6547 complete, working, and well designed, tend to get installed on the day
6548 they arrive. The others go into a queue and get installed as time
6549 permits, which, since the maintainers have many demands to meet, may not
6550 be for quite some time.
6552 Please send patches directly to
6553 @email{gdb-patches@@sources.redhat.com, the @value{GDBN} maintainers}.
6555 @section Obsolete Conditionals
6556 @cindex obsolete code
6558 Fragments of old code in @value{GDBN} sometimes reference or set the following
6559 configuration macros. They should not be used by new code, and old uses
6560 should be removed as those parts of the debugger are otherwise touched.
6563 @item STACK_END_ADDR
6564 This macro used to define where the end of the stack appeared, for use
6565 in interpreting core file formats that don't record this address in the
6566 core file itself. This information is now configured in BFD, and @value{GDBN}
6567 gets the info portably from there. The values in @value{GDBN}'s configuration
6568 files should be moved into BFD configuration files (if needed there),
6569 and deleted from all of @value{GDBN}'s config files.
6571 Any @file{@var{foo}-xdep.c} file that references STACK_END_ADDR
6572 is so old that it has never been converted to use BFD. Now that's old!
6576 @include observer.texi