1 \input texinfo @c -*-texinfo-*-
2 @c Copyright 1988, 1989, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1998,
4 @c Free Software Foundation, Inc.
7 @c makeinfo ignores cmds prev to setfilename, so its arg cannot make use
8 @c of @set vars. However, you can override filename with makeinfo -o.
13 @settitle Debugging with @value{GDBN}
14 @setchapternewpage odd
25 @c readline appendices use @vindex, @findex and @ftable,
26 @c annotate.texi and gdbmi use @findex.
30 @c !!set GDB manual's edition---not the same as GDB version!
33 @c !!set GDB manual's revision date
34 @set DATE December 2001
36 @c THIS MANUAL REQUIRES TEXINFO 3.12 OR LATER.
38 @c This is a dir.info fragment to support semi-automated addition of
39 @c manuals to an info tree.
40 @dircategory Programming & development tools.
42 * Gdb: (gdb). The @sc{gnu} debugger.
46 This file documents the @sc{gnu} debugger @value{GDBN}.
49 This is the @value{EDITION} Edition, @value{DATE},
50 of @cite{Debugging with @value{GDBN}: the @sc{gnu} Source-Level Debugger}
51 for @value{GDBN} Version @value{GDBVN}.
53 Copyright (C) 1988,1989,1990,1991,1992,1993,1994,1995,1996,1998,1999,2000,2001, 2002 Free Software Foundation, Inc.
55 Permission is granted to copy, distribute and/or modify this document
56 under the terms of the GNU Free Documentation License, Version 1.1 or
57 any later version published by the Free Software Foundation; with the
58 Invariant Sections being ``Free Software'' and ``Free Software Needs
59 Free Documentation'', with the Front-Cover Texts being ``A GNU Manual,''
60 and with the Back-Cover Texts as in (a) below.
62 (a) The Free Software Foundation's Back-Cover Text is: ``You have
63 freedom to copy and modify this GNU Manual, like GNU software. Copies
64 published by the Free Software Foundation raise funds for GNU
69 @title Debugging with @value{GDBN}
70 @subtitle The @sc{gnu} Source-Level Debugger
72 @subtitle @value{EDITION} Edition, for @value{GDBN} version @value{GDBVN}
73 @subtitle @value{DATE}
74 @author Richard Stallman, Roland Pesch, Stan Shebs, et al.
78 \hfill (Send bugs and comments on @value{GDBN} to bug-gdb\@gnu.org.)\par
79 \hfill {\it Debugging with @value{GDBN}}\par
80 \hfill \TeX{}info \texinfoversion\par
84 @vskip 0pt plus 1filll
85 Copyright @copyright{} 1988,1989,1990,1991,1992,1993,1994,1995,1996,@*
86 1998,1999,2000,2001,2002 Free Software Foundation, Inc.
88 Published by the Free Software Foundation @*
89 59 Temple Place - Suite 330, @*
90 Boston, MA 02111-1307 USA @*
93 Permission is granted to copy, distribute and/or modify this document
94 under the terms of the GNU Free Documentation License, Version 1.1 or
95 any later version published by the Free Software Foundation; with the
96 Invariant Sections being ``Free Software'' and ``Free Software Needs
97 Free Documentation'', with the Front-Cover Texts being ``A GNU Manual,''
98 and with the Back-Cover Texts as in (a) below.
100 (a) The Free Software Foundation's Back-Cover Text is: ``You have
101 freedom to copy and modify this GNU Manual, like GNU software. Copies
102 published by the Free Software Foundation raise funds for GNU
108 @node Top, Summary, (dir), (dir)
110 @top Debugging with @value{GDBN}
112 This file describes @value{GDBN}, the @sc{gnu} symbolic debugger.
114 This is the @value{EDITION} Edition, @value{DATE}, for @value{GDBN} Version
117 Copyright (C) 1988-2001 Free Software Foundation, Inc.
120 * Summary:: Summary of @value{GDBN}
121 * Sample Session:: A sample @value{GDBN} session
123 * Invocation:: Getting in and out of @value{GDBN}
124 * Commands:: @value{GDBN} commands
125 * Running:: Running programs under @value{GDBN}
126 * Stopping:: Stopping and continuing
127 * Stack:: Examining the stack
128 * Source:: Examining source files
129 * Data:: Examining data
130 * Tracepoints:: Debugging remote targets non-intrusively
131 * Overlays:: Debugging programs that use overlays
133 * Languages:: Using @value{GDBN} with different languages
135 * Symbols:: Examining the symbol table
136 * Altering:: Altering execution
137 * GDB Files:: @value{GDBN} files
138 * Targets:: Specifying a debugging target
139 * Configurations:: Configuration-specific information
140 * Controlling GDB:: Controlling @value{GDBN}
141 * Sequences:: Canned sequences of commands
142 * TUI:: @value{GDBN} Text User Interface
143 * Emacs:: Using @value{GDBN} under @sc{gnu} Emacs
144 * Annotations:: @value{GDBN}'s annotation interface.
145 * GDB/MI:: @value{GDBN}'s Machine Interface.
147 * GDB Bugs:: Reporting bugs in @value{GDBN}
148 * Formatting Documentation:: How to format and print @value{GDBN} documentation
150 * Command Line Editing:: Command Line Editing
151 * Using History Interactively:: Using History Interactively
152 * Installing GDB:: Installing GDB
153 * GNU Free Documentation License:: The license for this documentation
159 @c the replication sucks, but this avoids a texinfo 3.12 lameness
164 @top Debugging with @value{GDBN}
166 This file describes @value{GDBN}, the @sc{gnu} symbolic debugger.
168 This is the @value{EDITION} Edition, @value{DATE}, for @value{GDBN} Version
171 Copyright (C) 1988-2000 Free Software Foundation, Inc.
174 * Summary:: Summary of @value{GDBN}
175 * Sample Session:: A sample @value{GDBN} session
177 * Invocation:: Getting in and out of @value{GDBN}
178 * Commands:: @value{GDBN} commands
179 * Running:: Running programs under @value{GDBN}
180 * Stopping:: Stopping and continuing
181 * Stack:: Examining the stack
182 * Source:: Examining source files
183 * Data:: Examining data
184 * Tracepoints:: Debugging remote targets non-intrusively
185 * Overlays:: Debugging programs that use overlays
187 * Languages:: Using @value{GDBN} with different languages
189 * Symbols:: Examining the symbol table
190 * Altering:: Altering execution
191 * GDB Files:: @value{GDBN} files
192 * Targets:: Specifying a debugging target
193 * Configurations:: Configuration-specific information
194 * Controlling GDB:: Controlling @value{GDBN}
195 * Sequences:: Canned sequences of commands
196 * TUI:: @value{GDBN} Text User Interface
197 * Emacs:: Using @value{GDBN} under @sc{gnu} Emacs
198 * Annotations:: @value{GDBN}'s annotation interface.
199 * GDB/MI:: @value{GDBN}'s Machine Interface.
201 * GDB Bugs:: Reporting bugs in @value{GDBN}
202 * Formatting Documentation:: How to format and print @value{GDBN} documentation
204 * Command Line Editing:: Command Line Editing
205 * Using History Interactively:: Using History Interactively
206 * Installing GDB:: Installing GDB
207 * GNU Free Documentation License:: The license for this documentation
213 @c TeX can handle the contents at the start but makeinfo 3.12 can not
219 @unnumbered Summary of @value{GDBN}
221 The purpose of a debugger such as @value{GDBN} is to allow you to see what is
222 going on ``inside'' another program while it executes---or what another
223 program was doing at the moment it crashed.
225 @value{GDBN} can do four main kinds of things (plus other things in support of
226 these) to help you catch bugs in the act:
230 Start your program, specifying anything that might affect its behavior.
233 Make your program stop on specified conditions.
236 Examine what has happened, when your program has stopped.
239 Change things in your program, so you can experiment with correcting the
240 effects of one bug and go on to learn about another.
243 You can use @value{GDBN} to debug programs written in C and C++.
244 For more information, see @ref{Support,,Supported languages}.
245 For more information, see @ref{C,,C and C++}.
249 Support for Modula-2 and Chill is partial. For information on Modula-2,
250 see @ref{Modula-2,,Modula-2}. For information on Chill, see @ref{Chill}.
253 Debugging Pascal programs which use sets, subranges, file variables, or
254 nested functions does not currently work. @value{GDBN} does not support
255 entering expressions, printing values, or similar features using Pascal
259 @value{GDBN} can be used to debug programs written in Fortran, although
260 it may be necessary to refer to some variables with a trailing
264 * Free Software:: Freely redistributable software
265 * Contributors:: Contributors to GDB
269 @unnumberedsec Free software
271 @value{GDBN} is @dfn{free software}, protected by the @sc{gnu}
272 General Public License
273 (GPL). The GPL gives you the freedom to copy or adapt a licensed
274 program---but every person getting a copy also gets with it the
275 freedom to modify that copy (which means that they must get access to
276 the source code), and the freedom to distribute further copies.
277 Typical software companies use copyrights to limit your freedoms; the
278 Free Software Foundation uses the GPL to preserve these freedoms.
280 Fundamentally, the General Public License is a license which says that
281 you have these freedoms and that you cannot take these freedoms away
284 @unnumberedsec Free Software Needs Free Documentation
286 The biggest deficiency in the free software community today is not in
287 the software---it is the lack of good free documentation that we can
288 include with the free software. Many of our most important
289 programs do not come with free reference manuals and free introductory
290 texts. Documentation is an essential part of any software package;
291 when an important free software package does not come with a free
292 manual and a free tutorial, that is a major gap. We have many such
295 Consider Perl, for instance. The tutorial manuals that people
296 normally use are non-free. How did this come about? Because the
297 authors of those manuals published them with restrictive terms---no
298 copying, no modification, source files not available---which exclude
299 them from the free software world.
301 That wasn't the first time this sort of thing happened, and it was far
302 from the last. Many times we have heard a GNU user eagerly describe a
303 manual that he is writing, his intended contribution to the community,
304 only to learn that he had ruined everything by signing a publication
305 contract to make it non-free.
307 Free documentation, like free software, is a matter of freedom, not
308 price. The problem with the non-free manual is not that publishers
309 charge a price for printed copies---that in itself is fine. (The Free
310 Software Foundation sells printed copies of manuals, too.) The
311 problem is the restrictions on the use of the manual. Free manuals
312 are available in source code form, and give you permission to copy and
313 modify. Non-free manuals do not allow this.
315 The criteria of freedom for a free manual are roughly the same as for
316 free software. Redistribution (including the normal kinds of
317 commercial redistribution) must be permitted, so that the manual can
318 accompany every copy of the program, both on-line and on paper.
320 Permission for modification of the technical content is crucial too.
321 When people modify the software, adding or changing features, if they
322 are conscientious they will change the manual too---so they can
323 provide accurate and clear documentation for the modified program. A
324 manual that leaves you no choice but to write a new manual to document
325 a changed version of the program is not really available to our
328 Some kinds of limits on the way modification is handled are
329 acceptable. For example, requirements to preserve the original
330 author's copyright notice, the distribution terms, or the list of
331 authors, are ok. It is also no problem to require modified versions
332 to include notice that they were modified. Even entire sections that
333 may not be deleted or changed are acceptable, as long as they deal
334 with nontechnical topics (like this one). These kinds of restrictions
335 are acceptable because they don't obstruct the community's normal use
338 However, it must be possible to modify all the @emph{technical}
339 content of the manual, and then distribute the result in all the usual
340 media, through all the usual channels. Otherwise, the restrictions
341 obstruct the use of the manual, it is not free, and we need another
342 manual to replace it.
344 Please spread the word about this issue. Our community continues to
345 lose manuals to proprietary publishing. If we spread the word that
346 free software needs free reference manuals and free tutorials, perhaps
347 the next person who wants to contribute by writing documentation will
348 realize, before it is too late, that only free manuals contribute to
349 the free software community.
351 If you are writing documentation, please insist on publishing it under
352 the GNU Free Documentation License or another free documentation
353 license. Remember that this decision requires your approval---you
354 don't have to let the publisher decide. Some commercial publishers
355 will use a free license if you insist, but they will not propose the
356 option; it is up to you to raise the issue and say firmly that this is
357 what you want. If the publisher you are dealing with refuses, please
358 try other publishers. If you're not sure whether a proposed license
359 is free, write to @email{licensing@@gnu.org}.
361 You can encourage commercial publishers to sell more free, copylefted
362 manuals and tutorials by buying them, and particularly by buying
363 copies from the publishers that paid for their writing or for major
364 improvements. Meanwhile, try to avoid buying non-free documentation
365 at all. Check the distribution terms of a manual before you buy it,
366 and insist that whoever seeks your business must respect your freedom.
367 Check the history of the book, and try to reward the publishers that
368 have paid or pay the authors to work on it.
370 The Free Software Foundation maintains a list of free documentation
371 published by other publishers, at
372 @url{http://www.fsf.org/doc/other-free-books.html}.
375 @unnumberedsec Contributors to @value{GDBN}
377 Richard Stallman was the original author of @value{GDBN}, and of many
378 other @sc{gnu} programs. Many others have contributed to its
379 development. This section attempts to credit major contributors. One
380 of the virtues of free software is that everyone is free to contribute
381 to it; with regret, we cannot actually acknowledge everyone here. The
382 file @file{ChangeLog} in the @value{GDBN} distribution approximates a
383 blow-by-blow account.
385 Changes much prior to version 2.0 are lost in the mists of time.
388 @emph{Plea:} Additions to this section are particularly welcome. If you
389 or your friends (or enemies, to be evenhanded) have been unfairly
390 omitted from this list, we would like to add your names!
393 So that they may not regard their many labors as thankless, we
394 particularly thank those who shepherded @value{GDBN} through major
396 Andrew Cagney (releases 5.0 and 5.1);
397 Jim Blandy (release 4.18);
398 Jason Molenda (release 4.17);
399 Stan Shebs (release 4.14);
400 Fred Fish (releases 4.16, 4.15, 4.13, 4.12, 4.11, 4.10, and 4.9);
401 Stu Grossman and John Gilmore (releases 4.8, 4.7, 4.6, 4.5, and 4.4);
402 John Gilmore (releases 4.3, 4.2, 4.1, 4.0, and 3.9);
403 Jim Kingdon (releases 3.5, 3.4, and 3.3);
404 and Randy Smith (releases 3.2, 3.1, and 3.0).
406 Richard Stallman, assisted at various times by Peter TerMaat, Chris
407 Hanson, and Richard Mlynarik, handled releases through 2.8.
409 Michael Tiemann is the author of most of the @sc{gnu} C@t{++} support
410 in @value{GDBN}, with significant additional contributions from Per
411 Bothner and Daniel Berlin. James Clark wrote the @sc{gnu} C@t{++}
412 demangler. Early work on C@t{++} was by Peter TerMaat (who also did
413 much general update work leading to release 3.0).
415 @value{GDBN} uses the BFD subroutine library to examine multiple
416 object-file formats; BFD was a joint project of David V.
417 Henkel-Wallace, Rich Pixley, Steve Chamberlain, and John Gilmore.
419 David Johnson wrote the original COFF support; Pace Willison did
420 the original support for encapsulated COFF.
422 Brent Benson of Harris Computer Systems contributed DWARF2 support.
424 Adam de Boor and Bradley Davis contributed the ISI Optimum V support.
425 Per Bothner, Noboyuki Hikichi, and Alessandro Forin contributed MIPS
427 Jean-Daniel Fekete contributed Sun 386i support.
428 Chris Hanson improved the HP9000 support.
429 Noboyuki Hikichi and Tomoyuki Hasei contributed Sony/News OS 3 support.
430 David Johnson contributed Encore Umax support.
431 Jyrki Kuoppala contributed Altos 3068 support.
432 Jeff Law contributed HP PA and SOM support.
433 Keith Packard contributed NS32K support.
434 Doug Rabson contributed Acorn Risc Machine support.
435 Bob Rusk contributed Harris Nighthawk CX-UX support.
436 Chris Smith contributed Convex support (and Fortran debugging).
437 Jonathan Stone contributed Pyramid support.
438 Michael Tiemann contributed SPARC support.
439 Tim Tucker contributed support for the Gould NP1 and Gould Powernode.
440 Pace Willison contributed Intel 386 support.
441 Jay Vosburgh contributed Symmetry support.
443 Andreas Schwab contributed M68K Linux support.
445 Rich Schaefer and Peter Schauer helped with support of SunOS shared
448 Jay Fenlason and Roland McGrath ensured that @value{GDBN} and GAS agree
449 about several machine instruction sets.
451 Patrick Duval, Ted Goldstein, Vikram Koka and Glenn Engel helped develop
452 remote debugging. Intel Corporation, Wind River Systems, AMD, and ARM
453 contributed remote debugging modules for the i960, VxWorks, A29K UDI,
454 and RDI targets, respectively.
456 Brian Fox is the author of the readline libraries providing
457 command-line editing and command history.
459 Andrew Beers of SUNY Buffalo wrote the language-switching code, the
460 Modula-2 support, and contributed the Languages chapter of this manual.
462 Fred Fish wrote most of the support for Unix System Vr4.
463 He also enhanced the command-completion support to cover C@t{++} overloaded
466 Hitachi America, Ltd. sponsored the support for H8/300, H8/500, and
469 NEC sponsored the support for the v850, Vr4xxx, and Vr5xxx processors.
471 Mitsubishi sponsored the support for D10V, D30V, and M32R/D processors.
473 Toshiba sponsored the support for the TX39 Mips processor.
475 Matsushita sponsored the support for the MN10200 and MN10300 processors.
477 Fujitsu sponsored the support for SPARClite and FR30 processors.
479 Kung Hsu, Jeff Law, and Rick Sladkey added support for hardware
482 Michael Snyder added support for tracepoints.
484 Stu Grossman wrote gdbserver.
486 Jim Kingdon, Peter Schauer, Ian Taylor, and Stu Grossman made
487 nearly innumerable bug fixes and cleanups throughout @value{GDBN}.
489 The following people at the Hewlett-Packard Company contributed
490 support for the PA-RISC 2.0 architecture, HP-UX 10.20, 10.30, and 11.0
491 (narrow mode), HP's implementation of kernel threads, HP's aC@t{++}
492 compiler, and the terminal user interface: Ben Krepp, Richard Title,
493 John Bishop, Susan Macchia, Kathy Mann, Satish Pai, India Paul, Steve
494 Rehrauer, and Elena Zannoni. Kim Haase provided HP-specific
495 information in this manual.
497 DJ Delorie ported @value{GDBN} to MS-DOS, for the DJGPP project.
498 Robert Hoehne made significant contributions to the DJGPP port.
500 Cygnus Solutions has sponsored @value{GDBN} maintenance and much of its
501 development since 1991. Cygnus engineers who have worked on @value{GDBN}
502 fulltime include Mark Alexander, Jim Blandy, Per Bothner, Kevin
503 Buettner, Edith Epstein, Chris Faylor, Fred Fish, Martin Hunt, Jim
504 Ingham, John Gilmore, Stu Grossman, Kung Hsu, Jim Kingdon, John Metzler,
505 Fernando Nasser, Geoffrey Noer, Dawn Perchik, Rich Pixley, Zdenek
506 Radouch, Keith Seitz, Stan Shebs, David Taylor, and Elena Zannoni. In
507 addition, Dave Brolley, Ian Carmichael, Steve Chamberlain, Nick Clifton,
508 JT Conklin, Stan Cox, DJ Delorie, Ulrich Drepper, Frank Eigler, Doug
509 Evans, Sean Fagan, David Henkel-Wallace, Richard Henderson, Jeff
510 Holcomb, Jeff Law, Jim Lemke, Tom Lord, Bob Manson, Michael Meissner,
511 Jason Merrill, Catherine Moore, Drew Moseley, Ken Raeburn, Gavin
512 Romig-Koch, Rob Savoye, Jamie Smith, Mike Stump, Ian Taylor, Angela
513 Thomas, Michael Tiemann, Tom Tromey, Ron Unrau, Jim Wilson, and David
514 Zuhn have made contributions both large and small.
518 @chapter A Sample @value{GDBN} Session
520 You can use this manual at your leisure to read all about @value{GDBN}.
521 However, a handful of commands are enough to get started using the
522 debugger. This chapter illustrates those commands.
525 In this sample session, we emphasize user input like this: @b{input},
526 to make it easier to pick out from the surrounding output.
529 @c FIXME: this example may not be appropriate for some configs, where
530 @c FIXME...primary interest is in remote use.
532 One of the preliminary versions of @sc{gnu} @code{m4} (a generic macro
533 processor) exhibits the following bug: sometimes, when we change its
534 quote strings from the default, the commands used to capture one macro
535 definition within another stop working. In the following short @code{m4}
536 session, we define a macro @code{foo} which expands to @code{0000}; we
537 then use the @code{m4} built-in @code{defn} to define @code{bar} as the
538 same thing. However, when we change the open quote string to
539 @code{<QUOTE>} and the close quote string to @code{<UNQUOTE>}, the same
540 procedure fails to define a new synonym @code{baz}:
549 @b{define(bar,defn(`foo'))}
553 @b{changequote(<QUOTE>,<UNQUOTE>)}
555 @b{define(baz,defn(<QUOTE>foo<UNQUOTE>))}
558 m4: End of input: 0: fatal error: EOF in string
562 Let us use @value{GDBN} to try to see what is going on.
565 $ @b{@value{GDBP} m4}
566 @c FIXME: this falsifies the exact text played out, to permit smallbook
567 @c FIXME... format to come out better.
568 @value{GDBN} is free software and you are welcome to distribute copies
569 of it under certain conditions; type "show copying" to see
571 There is absolutely no warranty for @value{GDBN}; type "show warranty"
574 @value{GDBN} @value{GDBVN}, Copyright 1999 Free Software Foundation, Inc...
579 @value{GDBN} reads only enough symbol data to know where to find the
580 rest when needed; as a result, the first prompt comes up very quickly.
581 We now tell @value{GDBN} to use a narrower display width than usual, so
582 that examples fit in this manual.
585 (@value{GDBP}) @b{set width 70}
589 We need to see how the @code{m4} built-in @code{changequote} works.
590 Having looked at the source, we know the relevant subroutine is
591 @code{m4_changequote}, so we set a breakpoint there with the @value{GDBN}
592 @code{break} command.
595 (@value{GDBP}) @b{break m4_changequote}
596 Breakpoint 1 at 0x62f4: file builtin.c, line 879.
600 Using the @code{run} command, we start @code{m4} running under @value{GDBN}
601 control; as long as control does not reach the @code{m4_changequote}
602 subroutine, the program runs as usual:
605 (@value{GDBP}) @b{run}
606 Starting program: /work/Editorial/gdb/gnu/m4/m4
614 To trigger the breakpoint, we call @code{changequote}. @value{GDBN}
615 suspends execution of @code{m4}, displaying information about the
616 context where it stops.
619 @b{changequote(<QUOTE>,<UNQUOTE>)}
621 Breakpoint 1, m4_changequote (argc=3, argv=0x33c70)
623 879 if (bad_argc(TOKEN_DATA_TEXT(argv[0]),argc,1,3))
627 Now we use the command @code{n} (@code{next}) to advance execution to
628 the next line of the current function.
632 882 set_quotes((argc >= 2) ? TOKEN_DATA_TEXT(argv[1])\
637 @code{set_quotes} looks like a promising subroutine. We can go into it
638 by using the command @code{s} (@code{step}) instead of @code{next}.
639 @code{step} goes to the next line to be executed in @emph{any}
640 subroutine, so it steps into @code{set_quotes}.
644 set_quotes (lq=0x34c78 "<QUOTE>", rq=0x34c88 "<UNQUOTE>")
646 530 if (lquote != def_lquote)
650 The display that shows the subroutine where @code{m4} is now
651 suspended (and its arguments) is called a stack frame display. It
652 shows a summary of the stack. We can use the @code{backtrace}
653 command (which can also be spelled @code{bt}), to see where we are
654 in the stack as a whole: the @code{backtrace} command displays a
655 stack frame for each active subroutine.
658 (@value{GDBP}) @b{bt}
659 #0 set_quotes (lq=0x34c78 "<QUOTE>", rq=0x34c88 "<UNQUOTE>")
661 #1 0x6344 in m4_changequote (argc=3, argv=0x33c70)
663 #2 0x8174 in expand_macro (sym=0x33320) at macro.c:242
664 #3 0x7a88 in expand_token (obs=0x0, t=209696, td=0xf7fffa30)
666 #4 0x79dc in expand_input () at macro.c:40
667 #5 0x2930 in main (argc=0, argv=0xf7fffb20) at m4.c:195
671 We step through a few more lines to see what happens. The first two
672 times, we can use @samp{s}; the next two times we use @code{n} to avoid
673 falling into the @code{xstrdup} subroutine.
677 0x3b5c 532 if (rquote != def_rquote)
679 0x3b80 535 lquote = (lq == nil || *lq == '\0') ? \
680 def_lquote : xstrdup(lq);
682 536 rquote = (rq == nil || *rq == '\0') ? def_rquote\
685 538 len_lquote = strlen(rquote);
689 The last line displayed looks a little odd; we can examine the variables
690 @code{lquote} and @code{rquote} to see if they are in fact the new left
691 and right quotes we specified. We use the command @code{p}
692 (@code{print}) to see their values.
695 (@value{GDBP}) @b{p lquote}
696 $1 = 0x35d40 "<QUOTE>"
697 (@value{GDBP}) @b{p rquote}
698 $2 = 0x35d50 "<UNQUOTE>"
702 @code{lquote} and @code{rquote} are indeed the new left and right quotes.
703 To look at some context, we can display ten lines of source
704 surrounding the current line with the @code{l} (@code{list}) command.
710 535 lquote = (lq == nil || *lq == '\0') ? def_lquote\
712 536 rquote = (rq == nil || *rq == '\0') ? def_rquote\
715 538 len_lquote = strlen(rquote);
716 539 len_rquote = strlen(lquote);
723 Let us step past the two lines that set @code{len_lquote} and
724 @code{len_rquote}, and then examine the values of those variables.
728 539 len_rquote = strlen(lquote);
731 (@value{GDBP}) @b{p len_lquote}
733 (@value{GDBP}) @b{p len_rquote}
738 That certainly looks wrong, assuming @code{len_lquote} and
739 @code{len_rquote} are meant to be the lengths of @code{lquote} and
740 @code{rquote} respectively. We can set them to better values using
741 the @code{p} command, since it can print the value of
742 any expression---and that expression can include subroutine calls and
746 (@value{GDBP}) @b{p len_lquote=strlen(lquote)}
748 (@value{GDBP}) @b{p len_rquote=strlen(rquote)}
753 Is that enough to fix the problem of using the new quotes with the
754 @code{m4} built-in @code{defn}? We can allow @code{m4} to continue
755 executing with the @code{c} (@code{continue}) command, and then try the
756 example that caused trouble initially:
762 @b{define(baz,defn(<QUOTE>foo<UNQUOTE>))}
769 Success! The new quotes now work just as well as the default ones. The
770 problem seems to have been just the two typos defining the wrong
771 lengths. We allow @code{m4} exit by giving it an EOF as input:
775 Program exited normally.
779 The message @samp{Program exited normally.} is from @value{GDBN}; it
780 indicates @code{m4} has finished executing. We can end our @value{GDBN}
781 session with the @value{GDBN} @code{quit} command.
784 (@value{GDBP}) @b{quit}
788 @chapter Getting In and Out of @value{GDBN}
790 This chapter discusses how to start @value{GDBN}, and how to get out of it.
794 type @samp{@value{GDBP}} to start @value{GDBN}.
796 type @kbd{quit} or @kbd{C-d} to exit.
800 * Invoking GDB:: How to start @value{GDBN}
801 * Quitting GDB:: How to quit @value{GDBN}
802 * Shell Commands:: How to use shell commands inside @value{GDBN}
806 @section Invoking @value{GDBN}
808 Invoke @value{GDBN} by running the program @code{@value{GDBP}}. Once started,
809 @value{GDBN} reads commands from the terminal until you tell it to exit.
811 You can also run @code{@value{GDBP}} with a variety of arguments and options,
812 to specify more of your debugging environment at the outset.
814 The command-line options described here are designed
815 to cover a variety of situations; in some environments, some of these
816 options may effectively be unavailable.
818 The most usual way to start @value{GDBN} is with one argument,
819 specifying an executable program:
822 @value{GDBP} @var{program}
826 You can also start with both an executable program and a core file
830 @value{GDBP} @var{program} @var{core}
833 You can, instead, specify a process ID as a second argument, if you want
834 to debug a running process:
837 @value{GDBP} @var{program} 1234
841 would attach @value{GDBN} to process @code{1234} (unless you also have a file
842 named @file{1234}; @value{GDBN} does check for a core file first).
844 Taking advantage of the second command-line argument requires a fairly
845 complete operating system; when you use @value{GDBN} as a remote
846 debugger attached to a bare board, there may not be any notion of
847 ``process'', and there is often no way to get a core dump. @value{GDBN}
848 will warn you if it is unable to attach or to read core dumps.
850 You can optionally have @code{@value{GDBP}} pass any arguments after the
851 executable file to the inferior using @code{--args}. This option stops
854 gdb --args gcc -O2 -c foo.c
856 This will cause @code{@value{GDBP}} to debug @code{gcc}, and to set
857 @code{gcc}'s command-line arguments (@pxref{Arguments}) to @samp{-O2 -c foo.c}.
859 You can run @code{@value{GDBP}} without printing the front material, which describes
860 @value{GDBN}'s non-warranty, by specifying @code{-silent}:
867 You can further control how @value{GDBN} starts up by using command-line
868 options. @value{GDBN} itself can remind you of the options available.
878 to display all available options and briefly describe their use
879 (@samp{@value{GDBP} -h} is a shorter equivalent).
881 All options and command line arguments you give are processed
882 in sequential order. The order makes a difference when the
883 @samp{-x} option is used.
887 * File Options:: Choosing files
888 * Mode Options:: Choosing modes
892 @subsection Choosing files
894 When @value{GDBN} starts, it reads any arguments other than options as
895 specifying an executable file and core file (or process ID). This is
896 the same as if the arguments were specified by the @samp{-se} and
897 @samp{-c} (or @samp{-p} options respectively. (@value{GDBN} reads the
898 first argument that does not have an associated option flag as
899 equivalent to the @samp{-se} option followed by that argument; and the
900 second argument that does not have an associated option flag, if any, as
901 equivalent to the @samp{-c}/@samp{-p} option followed by that argument.)
902 If the second argument begins with a decimal digit, @value{GDBN} will
903 first attempt to attach to it as a process, and if that fails, attempt
904 to open it as a corefile. If you have a corefile whose name begins with
905 a digit, you can prevent @value{GDBN} from treating it as a pid by
906 prefixing it with @file{./}, eg. @file{./12345}.
908 If @value{GDBN} has not been configured to included core file support,
909 such as for most embedded targets, then it will complain about a second
910 argument and ignore it.
912 Many options have both long and short forms; both are shown in the
913 following list. @value{GDBN} also recognizes the long forms if you truncate
914 them, so long as enough of the option is present to be unambiguous.
915 (If you prefer, you can flag option arguments with @samp{--} rather
916 than @samp{-}, though we illustrate the more usual convention.)
918 @c NOTE: the @cindex entries here use double dashes ON PURPOSE. This
919 @c way, both those who look for -foo and --foo in the index, will find
923 @item -symbols @var{file}
925 @cindex @code{--symbols}
927 Read symbol table from file @var{file}.
929 @item -exec @var{file}
931 @cindex @code{--exec}
933 Use file @var{file} as the executable file to execute when appropriate,
934 and for examining pure data in conjunction with a core dump.
938 Read symbol table from file @var{file} and use it as the executable
941 @item -core @var{file}
943 @cindex @code{--core}
945 Use file @var{file} as a core dump to examine.
947 @item -c @var{number}
948 @item -pid @var{number}
949 @itemx -p @var{number}
952 Connect to process ID @var{number}, as with the @code{attach} command.
953 If there is no such process, @value{GDBN} will attempt to open a core
954 file named @var{number}.
956 @item -command @var{file}
958 @cindex @code{--command}
960 Execute @value{GDBN} commands from file @var{file}. @xref{Command
961 Files,, Command files}.
963 @item -directory @var{directory}
964 @itemx -d @var{directory}
965 @cindex @code{--directory}
967 Add @var{directory} to the path to search for source files.
971 @cindex @code{--mapped}
973 @emph{Warning: this option depends on operating system facilities that are not
974 supported on all systems.}@*
975 If memory-mapped files are available on your system through the @code{mmap}
976 system call, you can use this option
977 to have @value{GDBN} write the symbols from your
978 program into a reusable file in the current directory. If the program you are debugging is
979 called @file{/tmp/fred}, the mapped symbol file is @file{/tmp/fred.syms}.
980 Future @value{GDBN} debugging sessions notice the presence of this file,
981 and can quickly map in symbol information from it, rather than reading
982 the symbol table from the executable program.
984 The @file{.syms} file is specific to the host machine where @value{GDBN}
985 is run. It holds an exact image of the internal @value{GDBN} symbol
986 table. It cannot be shared across multiple host platforms.
990 @cindex @code{--readnow}
992 Read each symbol file's entire symbol table immediately, rather than
993 the default, which is to read it incrementally as it is needed.
994 This makes startup slower, but makes future operations faster.
998 You typically combine the @code{-mapped} and @code{-readnow} options in
999 order to build a @file{.syms} file that contains complete symbol
1000 information. (@xref{Files,,Commands to specify files}, for information
1001 on @file{.syms} files.) A simple @value{GDBN} invocation to do nothing
1002 but build a @file{.syms} file for future use is:
1005 gdb -batch -nx -mapped -readnow programname
1009 @subsection Choosing modes
1011 You can run @value{GDBN} in various alternative modes---for example, in
1012 batch mode or quiet mode.
1019 Do not execute commands found in any initialization files. Normally,
1020 @value{GDBN} executes the commands in these files after all the command
1021 options and arguments have been processed. @xref{Command Files,,Command
1027 @cindex @code{--quiet}
1028 @cindex @code{--silent}
1030 ``Quiet''. Do not print the introductory and copyright messages. These
1031 messages are also suppressed in batch mode.
1034 @cindex @code{--batch}
1035 Run in batch mode. Exit with status @code{0} after processing all the
1036 command files specified with @samp{-x} (and all commands from
1037 initialization files, if not inhibited with @samp{-n}). Exit with
1038 nonzero status if an error occurs in executing the @value{GDBN} commands
1039 in the command files.
1041 Batch mode may be useful for running @value{GDBN} as a filter, for
1042 example to download and run a program on another computer; in order to
1043 make this more useful, the message
1046 Program exited normally.
1050 (which is ordinarily issued whenever a program running under
1051 @value{GDBN} control terminates) is not issued when running in batch
1056 @cindex @code{--nowindows}
1058 ``No windows''. If @value{GDBN} comes with a graphical user interface
1059 (GUI) built in, then this option tells @value{GDBN} to only use the command-line
1060 interface. If no GUI is available, this option has no effect.
1064 @cindex @code{--windows}
1066 If @value{GDBN} includes a GUI, then this option requires it to be
1069 @item -cd @var{directory}
1071 Run @value{GDBN} using @var{directory} as its working directory,
1072 instead of the current directory.
1076 @cindex @code{--fullname}
1078 @sc{gnu} Emacs sets this option when it runs @value{GDBN} as a
1079 subprocess. It tells @value{GDBN} to output the full file name and line
1080 number in a standard, recognizable fashion each time a stack frame is
1081 displayed (which includes each time your program stops). This
1082 recognizable format looks like two @samp{\032} characters, followed by
1083 the file name, line number and character position separated by colons,
1084 and a newline. The Emacs-to-@value{GDBN} interface program uses the two
1085 @samp{\032} characters as a signal to display the source code for the
1089 @cindex @code{--epoch}
1090 The Epoch Emacs-@value{GDBN} interface sets this option when it runs
1091 @value{GDBN} as a subprocess. It tells @value{GDBN} to modify its print
1092 routines so as to allow Epoch to display values of expressions in a
1095 @item -annotate @var{level}
1096 @cindex @code{--annotate}
1097 This option sets the @dfn{annotation level} inside @value{GDBN}. Its
1098 effect is identical to using @samp{set annotate @var{level}}
1099 (@pxref{Annotations}).
1100 Annotation level controls how much information does @value{GDBN} print
1101 together with its prompt, values of expressions, source lines, and other
1102 types of output. Level 0 is the normal, level 1 is for use when
1103 @value{GDBN} is run as a subprocess of @sc{gnu} Emacs, level 2 is the
1104 maximum annotation suitable for programs that control @value{GDBN}.
1107 @cindex @code{--async}
1108 Use the asynchronous event loop for the command-line interface.
1109 @value{GDBN} processes all events, such as user keyboard input, via a
1110 special event loop. This allows @value{GDBN} to accept and process user
1111 commands in parallel with the debugged process being
1112 run@footnote{@value{GDBN} built with @sc{djgpp} tools for
1113 MS-DOS/MS-Windows supports this mode of operation, but the event loop is
1114 suspended when the debuggee runs.}, so you don't need to wait for
1115 control to return to @value{GDBN} before you type the next command.
1116 (@emph{Note:} as of version 5.1, the target side of the asynchronous
1117 operation is not yet in place, so @samp{-async} does not work fully
1119 @c FIXME: when the target side of the event loop is done, the above NOTE
1120 @c should be removed.
1122 When the standard input is connected to a terminal device, @value{GDBN}
1123 uses the asynchronous event loop by default, unless disabled by the
1124 @samp{-noasync} option.
1127 @cindex @code{--noasync}
1128 Disable the asynchronous event loop for the command-line interface.
1131 @cindex @code{--args}
1132 Change interpretation of command line so that arguments following the
1133 executable file are passed as command line arguments to the inferior.
1134 This option stops option processing.
1136 @item -baud @var{bps}
1138 @cindex @code{--baud}
1140 Set the line speed (baud rate or bits per second) of any serial
1141 interface used by @value{GDBN} for remote debugging.
1143 @item -tty @var{device}
1144 @itemx -t @var{device}
1145 @cindex @code{--tty}
1147 Run using @var{device} for your program's standard input and output.
1148 @c FIXME: kingdon thinks there is more to -tty. Investigate.
1150 @c resolve the situation of these eventually
1152 @cindex @code{--tui}
1153 Activate the Terminal User Interface when starting.
1154 The Terminal User Interface manages several text windows on the terminal,
1155 showing source, assembly, registers and @value{GDBN} command outputs
1156 (@pxref{TUI, ,@value{GDBN} Text User Interface}).
1157 Do not use this option if you run @value{GDBN} from Emacs
1158 (@pxref{Emacs, ,Using @value{GDBN} under @sc{gnu} Emacs}).
1161 @c @cindex @code{--xdb}
1162 @c Run in XDB compatibility mode, allowing the use of certain XDB commands.
1163 @c For information, see the file @file{xdb_trans.html}, which is usually
1164 @c installed in the directory @code{/opt/langtools/wdb/doc} on HP-UX
1167 @item -interpreter @var{interp}
1168 @cindex @code{--interpreter}
1169 Use the interpreter @var{interp} for interface with the controlling
1170 program or device. This option is meant to be set by programs which
1171 communicate with @value{GDBN} using it as a back end.
1173 @samp{--interpreter=mi} (or @samp{--interpreter=mi1}) causes
1174 @value{GDBN} to use the @dfn{gdb/mi interface} (@pxref{GDB/MI, , The
1175 @sc{gdb/mi} Interface}). The older @sc{gdb/mi} interface, included in
1176 @value{GDBN} version 5.0 can be selected with @samp{--interpreter=mi0}.
1179 @cindex @code{--write}
1180 Open the executable and core files for both reading and writing. This
1181 is equivalent to the @samp{set write on} command inside @value{GDBN}
1185 @cindex @code{--statistics}
1186 This option causes @value{GDBN} to print statistics about time and
1187 memory usage after it completes each command and returns to the prompt.
1190 @cindex @code{--version}
1191 This option causes @value{GDBN} to print its version number and
1192 no-warranty blurb, and exit.
1197 @section Quitting @value{GDBN}
1198 @cindex exiting @value{GDBN}
1199 @cindex leaving @value{GDBN}
1202 @kindex quit @r{[}@var{expression}@r{]}
1203 @kindex q @r{(@code{quit})}
1204 @item quit @r{[}@var{expression}@r{]}
1206 To exit @value{GDBN}, use the @code{quit} command (abbreviated
1207 @code{q}), or type an end-of-file character (usually @kbd{C-d}). If you
1208 do not supply @var{expression}, @value{GDBN} will terminate normally;
1209 otherwise it will terminate using the result of @var{expression} as the
1214 An interrupt (often @kbd{C-c}) does not exit from @value{GDBN}, but rather
1215 terminates the action of any @value{GDBN} command that is in progress and
1216 returns to @value{GDBN} command level. It is safe to type the interrupt
1217 character at any time because @value{GDBN} does not allow it to take effect
1218 until a time when it is safe.
1220 If you have been using @value{GDBN} to control an attached process or
1221 device, you can release it with the @code{detach} command
1222 (@pxref{Attach, ,Debugging an already-running process}).
1224 @node Shell Commands
1225 @section Shell commands
1227 If you need to execute occasional shell commands during your
1228 debugging session, there is no need to leave or suspend @value{GDBN}; you can
1229 just use the @code{shell} command.
1233 @cindex shell escape
1234 @item shell @var{command string}
1235 Invoke a standard shell to execute @var{command string}.
1236 If it exists, the environment variable @code{SHELL} determines which
1237 shell to run. Otherwise @value{GDBN} uses the default shell
1238 (@file{/bin/sh} on Unix systems, @file{COMMAND.COM} on MS-DOS, etc.).
1241 The utility @code{make} is often needed in development environments.
1242 You do not have to use the @code{shell} command for this purpose in
1247 @cindex calling make
1248 @item make @var{make-args}
1249 Execute the @code{make} program with the specified
1250 arguments. This is equivalent to @samp{shell make @var{make-args}}.
1254 @chapter @value{GDBN} Commands
1256 You can abbreviate a @value{GDBN} command to the first few letters of the command
1257 name, if that abbreviation is unambiguous; and you can repeat certain
1258 @value{GDBN} commands by typing just @key{RET}. You can also use the @key{TAB}
1259 key to get @value{GDBN} to fill out the rest of a word in a command (or to
1260 show you the alternatives available, if there is more than one possibility).
1263 * Command Syntax:: How to give commands to @value{GDBN}
1264 * Completion:: Command completion
1265 * Help:: How to ask @value{GDBN} for help
1268 @node Command Syntax
1269 @section Command syntax
1271 A @value{GDBN} command is a single line of input. There is no limit on
1272 how long it can be. It starts with a command name, which is followed by
1273 arguments whose meaning depends on the command name. For example, the
1274 command @code{step} accepts an argument which is the number of times to
1275 step, as in @samp{step 5}. You can also use the @code{step} command
1276 with no arguments. Some commands do not allow any arguments.
1278 @cindex abbreviation
1279 @value{GDBN} command names may always be truncated if that abbreviation is
1280 unambiguous. Other possible command abbreviations are listed in the
1281 documentation for individual commands. In some cases, even ambiguous
1282 abbreviations are allowed; for example, @code{s} is specially defined as
1283 equivalent to @code{step} even though there are other commands whose
1284 names start with @code{s}. You can test abbreviations by using them as
1285 arguments to the @code{help} command.
1287 @cindex repeating commands
1288 @kindex RET @r{(repeat last command)}
1289 A blank line as input to @value{GDBN} (typing just @key{RET}) means to
1290 repeat the previous command. Certain commands (for example, @code{run})
1291 will not repeat this way; these are commands whose unintentional
1292 repetition might cause trouble and which you are unlikely to want to
1295 The @code{list} and @code{x} commands, when you repeat them with
1296 @key{RET}, construct new arguments rather than repeating
1297 exactly as typed. This permits easy scanning of source or memory.
1299 @value{GDBN} can also use @key{RET} in another way: to partition lengthy
1300 output, in a way similar to the common utility @code{more}
1301 (@pxref{Screen Size,,Screen size}). Since it is easy to press one
1302 @key{RET} too many in this situation, @value{GDBN} disables command
1303 repetition after any command that generates this sort of display.
1305 @kindex # @r{(a comment)}
1307 Any text from a @kbd{#} to the end of the line is a comment; it does
1308 nothing. This is useful mainly in command files (@pxref{Command
1309 Files,,Command files}).
1311 @cindex repeating command sequences
1312 @kindex C-o @r{(operate-and-get-next)}
1313 The @kbd{C-o} binding is useful for repeating a complex sequence of
1314 commands. This command accepts the current line, like @kbd{RET}, and
1315 then fetches the next line relative to the current line from the history
1319 @section Command completion
1322 @cindex word completion
1323 @value{GDBN} can fill in the rest of a word in a command for you, if there is
1324 only one possibility; it can also show you what the valid possibilities
1325 are for the next word in a command, at any time. This works for @value{GDBN}
1326 commands, @value{GDBN} subcommands, and the names of symbols in your program.
1328 Press the @key{TAB} key whenever you want @value{GDBN} to fill out the rest
1329 of a word. If there is only one possibility, @value{GDBN} fills in the
1330 word, and waits for you to finish the command (or press @key{RET} to
1331 enter it). For example, if you type
1333 @c FIXME "@key" does not distinguish its argument sufficiently to permit
1334 @c complete accuracy in these examples; space introduced for clarity.
1335 @c If texinfo enhancements make it unnecessary, it would be nice to
1336 @c replace " @key" by "@key" in the following...
1338 (@value{GDBP}) info bre @key{TAB}
1342 @value{GDBN} fills in the rest of the word @samp{breakpoints}, since that is
1343 the only @code{info} subcommand beginning with @samp{bre}:
1346 (@value{GDBP}) info breakpoints
1350 You can either press @key{RET} at this point, to run the @code{info
1351 breakpoints} command, or backspace and enter something else, if
1352 @samp{breakpoints} does not look like the command you expected. (If you
1353 were sure you wanted @code{info breakpoints} in the first place, you
1354 might as well just type @key{RET} immediately after @samp{info bre},
1355 to exploit command abbreviations rather than command completion).
1357 If there is more than one possibility for the next word when you press
1358 @key{TAB}, @value{GDBN} sounds a bell. You can either supply more
1359 characters and try again, or just press @key{TAB} a second time;
1360 @value{GDBN} displays all the possible completions for that word. For
1361 example, you might want to set a breakpoint on a subroutine whose name
1362 begins with @samp{make_}, but when you type @kbd{b make_@key{TAB}} @value{GDBN}
1363 just sounds the bell. Typing @key{TAB} again displays all the
1364 function names in your program that begin with those characters, for
1368 (@value{GDBP}) b make_ @key{TAB}
1369 @exdent @value{GDBN} sounds bell; press @key{TAB} again, to see:
1370 make_a_section_from_file make_environ
1371 make_abs_section make_function_type
1372 make_blockvector make_pointer_type
1373 make_cleanup make_reference_type
1374 make_command make_symbol_completion_list
1375 (@value{GDBP}) b make_
1379 After displaying the available possibilities, @value{GDBN} copies your
1380 partial input (@samp{b make_} in the example) so you can finish the
1383 If you just want to see the list of alternatives in the first place, you
1384 can press @kbd{M-?} rather than pressing @key{TAB} twice. @kbd{M-?}
1385 means @kbd{@key{META} ?}. You can type this either by holding down a
1386 key designated as the @key{META} shift on your keyboard (if there is
1387 one) while typing @kbd{?}, or as @key{ESC} followed by @kbd{?}.
1389 @cindex quotes in commands
1390 @cindex completion of quoted strings
1391 Sometimes the string you need, while logically a ``word'', may contain
1392 parentheses or other characters that @value{GDBN} normally excludes from
1393 its notion of a word. To permit word completion to work in this
1394 situation, you may enclose words in @code{'} (single quote marks) in
1395 @value{GDBN} commands.
1397 The most likely situation where you might need this is in typing the
1398 name of a C@t{++} function. This is because C@t{++} allows function
1399 overloading (multiple definitions of the same function, distinguished
1400 by argument type). For example, when you want to set a breakpoint you
1401 may need to distinguish whether you mean the version of @code{name}
1402 that takes an @code{int} parameter, @code{name(int)}, or the version
1403 that takes a @code{float} parameter, @code{name(float)}. To use the
1404 word-completion facilities in this situation, type a single quote
1405 @code{'} at the beginning of the function name. This alerts
1406 @value{GDBN} that it may need to consider more information than usual
1407 when you press @key{TAB} or @kbd{M-?} to request word completion:
1410 (@value{GDBP}) b 'bubble( @kbd{M-?}
1411 bubble(double,double) bubble(int,int)
1412 (@value{GDBP}) b 'bubble(
1415 In some cases, @value{GDBN} can tell that completing a name requires using
1416 quotes. When this happens, @value{GDBN} inserts the quote for you (while
1417 completing as much as it can) if you do not type the quote in the first
1421 (@value{GDBP}) b bub @key{TAB}
1422 @exdent @value{GDBN} alters your input line to the following, and rings a bell:
1423 (@value{GDBP}) b 'bubble(
1427 In general, @value{GDBN} can tell that a quote is needed (and inserts it) if
1428 you have not yet started typing the argument list when you ask for
1429 completion on an overloaded symbol.
1431 For more information about overloaded functions, see @ref{C plus plus
1432 expressions, ,C@t{++} expressions}. You can use the command @code{set
1433 overload-resolution off} to disable overload resolution;
1434 see @ref{Debugging C plus plus, ,@value{GDBN} features for C@t{++}}.
1438 @section Getting help
1439 @cindex online documentation
1442 You can always ask @value{GDBN} itself for information on its commands,
1443 using the command @code{help}.
1446 @kindex h @r{(@code{help})}
1449 You can use @code{help} (abbreviated @code{h}) with no arguments to
1450 display a short list of named classes of commands:
1454 List of classes of commands:
1456 aliases -- Aliases of other commands
1457 breakpoints -- Making program stop at certain points
1458 data -- Examining data
1459 files -- Specifying and examining files
1460 internals -- Maintenance commands
1461 obscure -- Obscure features
1462 running -- Running the program
1463 stack -- Examining the stack
1464 status -- Status inquiries
1465 support -- Support facilities
1466 tracepoints -- Tracing of program execution without@*
1467 stopping the program
1468 user-defined -- User-defined commands
1470 Type "help" followed by a class name for a list of
1471 commands in that class.
1472 Type "help" followed by command name for full
1474 Command name abbreviations are allowed if unambiguous.
1477 @c the above line break eliminates huge line overfull...
1479 @item help @var{class}
1480 Using one of the general help classes as an argument, you can get a
1481 list of the individual commands in that class. For example, here is the
1482 help display for the class @code{status}:
1485 (@value{GDBP}) help status
1490 @c Line break in "show" line falsifies real output, but needed
1491 @c to fit in smallbook page size.
1492 info -- Generic command for showing things
1493 about the program being debugged
1494 show -- Generic command for showing things
1497 Type "help" followed by command name for full
1499 Command name abbreviations are allowed if unambiguous.
1503 @item help @var{command}
1504 With a command name as @code{help} argument, @value{GDBN} displays a
1505 short paragraph on how to use that command.
1508 @item apropos @var{args}
1509 The @code{apropos @var{args}} command searches through all of the @value{GDBN}
1510 commands, and their documentation, for the regular expression specified in
1511 @var{args}. It prints out all matches found. For example:
1522 set symbol-reloading -- Set dynamic symbol table reloading
1523 multiple times in one run
1524 show symbol-reloading -- Show dynamic symbol table reloading
1525 multiple times in one run
1530 @item complete @var{args}
1531 The @code{complete @var{args}} command lists all the possible completions
1532 for the beginning of a command. Use @var{args} to specify the beginning of the
1533 command you want completed. For example:
1539 @noindent results in:
1550 @noindent This is intended for use by @sc{gnu} Emacs.
1553 In addition to @code{help}, you can use the @value{GDBN} commands @code{info}
1554 and @code{show} to inquire about the state of your program, or the state
1555 of @value{GDBN} itself. Each command supports many topics of inquiry; this
1556 manual introduces each of them in the appropriate context. The listings
1557 under @code{info} and under @code{show} in the Index point to
1558 all the sub-commands. @xref{Index}.
1563 @kindex i @r{(@code{info})}
1565 This command (abbreviated @code{i}) is for describing the state of your
1566 program. For example, you can list the arguments given to your program
1567 with @code{info args}, list the registers currently in use with @code{info
1568 registers}, or list the breakpoints you have set with @code{info breakpoints}.
1569 You can get a complete list of the @code{info} sub-commands with
1570 @w{@code{help info}}.
1574 You can assign the result of an expression to an environment variable with
1575 @code{set}. For example, you can set the @value{GDBN} prompt to a $-sign with
1576 @code{set prompt $}.
1580 In contrast to @code{info}, @code{show} is for describing the state of
1581 @value{GDBN} itself.
1582 You can change most of the things you can @code{show}, by using the
1583 related command @code{set}; for example, you can control what number
1584 system is used for displays with @code{set radix}, or simply inquire
1585 which is currently in use with @code{show radix}.
1588 To display all the settable parameters and their current
1589 values, you can use @code{show} with no arguments; you may also use
1590 @code{info set}. Both commands produce the same display.
1591 @c FIXME: "info set" violates the rule that "info" is for state of
1592 @c FIXME...program. Ck w/ GNU: "info set" to be called something else,
1593 @c FIXME...or change desc of rule---eg "state of prog and debugging session"?
1597 Here are three miscellaneous @code{show} subcommands, all of which are
1598 exceptional in lacking corresponding @code{set} commands:
1601 @kindex show version
1602 @cindex version number
1604 Show what version of @value{GDBN} is running. You should include this
1605 information in @value{GDBN} bug-reports. If multiple versions of
1606 @value{GDBN} are in use at your site, you may need to determine which
1607 version of @value{GDBN} you are running; as @value{GDBN} evolves, new
1608 commands are introduced, and old ones may wither away. Also, many
1609 system vendors ship variant versions of @value{GDBN}, and there are
1610 variant versions of @value{GDBN} in @sc{gnu}/Linux distributions as well.
1611 The version number is the same as the one announced when you start
1614 @kindex show copying
1616 Display information about permission for copying @value{GDBN}.
1618 @kindex show warranty
1620 Display the @sc{gnu} ``NO WARRANTY'' statement, or a warranty,
1621 if your version of @value{GDBN} comes with one.
1626 @chapter Running Programs Under @value{GDBN}
1628 When you run a program under @value{GDBN}, you must first generate
1629 debugging information when you compile it.
1631 You may start @value{GDBN} with its arguments, if any, in an environment
1632 of your choice. If you are doing native debugging, you may redirect
1633 your program's input and output, debug an already running process, or
1634 kill a child process.
1637 * Compilation:: Compiling for debugging
1638 * Starting:: Starting your program
1639 * Arguments:: Your program's arguments
1640 * Environment:: Your program's environment
1642 * Working Directory:: Your program's working directory
1643 * Input/Output:: Your program's input and output
1644 * Attach:: Debugging an already-running process
1645 * Kill Process:: Killing the child process
1647 * Threads:: Debugging programs with multiple threads
1648 * Processes:: Debugging programs with multiple processes
1652 @section Compiling for debugging
1654 In order to debug a program effectively, you need to generate
1655 debugging information when you compile it. This debugging information
1656 is stored in the object file; it describes the data type of each
1657 variable or function and the correspondence between source line numbers
1658 and addresses in the executable code.
1660 To request debugging information, specify the @samp{-g} option when you run
1663 Many C compilers are unable to handle the @samp{-g} and @samp{-O}
1664 options together. Using those compilers, you cannot generate optimized
1665 executables containing debugging information.
1667 @value{NGCC}, the @sc{gnu} C compiler, supports @samp{-g} with or
1668 without @samp{-O}, making it possible to debug optimized code. We
1669 recommend that you @emph{always} use @samp{-g} whenever you compile a
1670 program. You may think your program is correct, but there is no sense
1671 in pushing your luck.
1673 @cindex optimized code, debugging
1674 @cindex debugging optimized code
1675 When you debug a program compiled with @samp{-g -O}, remember that the
1676 optimizer is rearranging your code; the debugger shows you what is
1677 really there. Do not be too surprised when the execution path does not
1678 exactly match your source file! An extreme example: if you define a
1679 variable, but never use it, @value{GDBN} never sees that
1680 variable---because the compiler optimizes it out of existence.
1682 Some things do not work as well with @samp{-g -O} as with just
1683 @samp{-g}, particularly on machines with instruction scheduling. If in
1684 doubt, recompile with @samp{-g} alone, and if this fixes the problem,
1685 please report it to us as a bug (including a test case!).
1687 Older versions of the @sc{gnu} C compiler permitted a variant option
1688 @w{@samp{-gg}} for debugging information. @value{GDBN} no longer supports this
1689 format; if your @sc{gnu} C compiler has this option, do not use it.
1693 @section Starting your program
1699 @kindex r @r{(@code{run})}
1702 Use the @code{run} command to start your program under @value{GDBN}.
1703 You must first specify the program name (except on VxWorks) with an
1704 argument to @value{GDBN} (@pxref{Invocation, ,Getting In and Out of
1705 @value{GDBN}}), or by using the @code{file} or @code{exec-file} command
1706 (@pxref{Files, ,Commands to specify files}).
1710 If you are running your program in an execution environment that
1711 supports processes, @code{run} creates an inferior process and makes
1712 that process run your program. (In environments without processes,
1713 @code{run} jumps to the start of your program.)
1715 The execution of a program is affected by certain information it
1716 receives from its superior. @value{GDBN} provides ways to specify this
1717 information, which you must do @emph{before} starting your program. (You
1718 can change it after starting your program, but such changes only affect
1719 your program the next time you start it.) This information may be
1720 divided into four categories:
1723 @item The @emph{arguments.}
1724 Specify the arguments to give your program as the arguments of the
1725 @code{run} command. If a shell is available on your target, the shell
1726 is used to pass the arguments, so that you may use normal conventions
1727 (such as wildcard expansion or variable substitution) in describing
1729 In Unix systems, you can control which shell is used with the
1730 @code{SHELL} environment variable.
1731 @xref{Arguments, ,Your program's arguments}.
1733 @item The @emph{environment.}
1734 Your program normally inherits its environment from @value{GDBN}, but you can
1735 use the @value{GDBN} commands @code{set environment} and @code{unset
1736 environment} to change parts of the environment that affect
1737 your program. @xref{Environment, ,Your program's environment}.
1739 @item The @emph{working directory.}
1740 Your program inherits its working directory from @value{GDBN}. You can set
1741 the @value{GDBN} working directory with the @code{cd} command in @value{GDBN}.
1742 @xref{Working Directory, ,Your program's working directory}.
1744 @item The @emph{standard input and output.}
1745 Your program normally uses the same device for standard input and
1746 standard output as @value{GDBN} is using. You can redirect input and output
1747 in the @code{run} command line, or you can use the @code{tty} command to
1748 set a different device for your program.
1749 @xref{Input/Output, ,Your program's input and output}.
1752 @emph{Warning:} While input and output redirection work, you cannot use
1753 pipes to pass the output of the program you are debugging to another
1754 program; if you attempt this, @value{GDBN} is likely to wind up debugging the
1758 When you issue the @code{run} command, your program begins to execute
1759 immediately. @xref{Stopping, ,Stopping and continuing}, for discussion
1760 of how to arrange for your program to stop. Once your program has
1761 stopped, you may call functions in your program, using the @code{print}
1762 or @code{call} commands. @xref{Data, ,Examining Data}.
1764 If the modification time of your symbol file has changed since the last
1765 time @value{GDBN} read its symbols, @value{GDBN} discards its symbol
1766 table, and reads it again. When it does this, @value{GDBN} tries to retain
1767 your current breakpoints.
1770 @section Your program's arguments
1772 @cindex arguments (to your program)
1773 The arguments to your program can be specified by the arguments of the
1775 They are passed to a shell, which expands wildcard characters and
1776 performs redirection of I/O, and thence to your program. Your
1777 @code{SHELL} environment variable (if it exists) specifies what shell
1778 @value{GDBN} uses. If you do not define @code{SHELL}, @value{GDBN} uses
1779 the default shell (@file{/bin/sh} on Unix).
1781 On non-Unix systems, the program is usually invoked directly by
1782 @value{GDBN}, which emulates I/O redirection via the appropriate system
1783 calls, and the wildcard characters are expanded by the startup code of
1784 the program, not by the shell.
1786 @code{run} with no arguments uses the same arguments used by the previous
1787 @code{run}, or those set by the @code{set args} command.
1792 Specify the arguments to be used the next time your program is run. If
1793 @code{set args} has no arguments, @code{run} executes your program
1794 with no arguments. Once you have run your program with arguments,
1795 using @code{set args} before the next @code{run} is the only way to run
1796 it again without arguments.
1800 Show the arguments to give your program when it is started.
1804 @section Your program's environment
1806 @cindex environment (of your program)
1807 The @dfn{environment} consists of a set of environment variables and
1808 their values. Environment variables conventionally record such things as
1809 your user name, your home directory, your terminal type, and your search
1810 path for programs to run. Usually you set up environment variables with
1811 the shell and they are inherited by all the other programs you run. When
1812 debugging, it can be useful to try running your program with a modified
1813 environment without having to start @value{GDBN} over again.
1817 @item path @var{directory}
1818 Add @var{directory} to the front of the @code{PATH} environment variable
1819 (the search path for executables) that will be passed to your program.
1820 The value of @code{PATH} used by @value{GDBN} does not change.
1821 You may specify several directory names, separated by whitespace or by a
1822 system-dependent separator character (@samp{:} on Unix, @samp{;} on
1823 MS-DOS and MS-Windows). If @var{directory} is already in the path, it
1824 is moved to the front, so it is searched sooner.
1826 You can use the string @samp{$cwd} to refer to whatever is the current
1827 working directory at the time @value{GDBN} searches the path. If you
1828 use @samp{.} instead, it refers to the directory where you executed the
1829 @code{path} command. @value{GDBN} replaces @samp{.} in the
1830 @var{directory} argument (with the current path) before adding
1831 @var{directory} to the search path.
1832 @c 'path' is explicitly nonrepeatable, but RMS points out it is silly to
1833 @c document that, since repeating it would be a no-op.
1837 Display the list of search paths for executables (the @code{PATH}
1838 environment variable).
1840 @kindex show environment
1841 @item show environment @r{[}@var{varname}@r{]}
1842 Print the value of environment variable @var{varname} to be given to
1843 your program when it starts. If you do not supply @var{varname},
1844 print the names and values of all environment variables to be given to
1845 your program. You can abbreviate @code{environment} as @code{env}.
1847 @kindex set environment
1848 @item set environment @var{varname} @r{[}=@var{value}@r{]}
1849 Set environment variable @var{varname} to @var{value}. The value
1850 changes for your program only, not for @value{GDBN} itself. @var{value} may
1851 be any string; the values of environment variables are just strings, and
1852 any interpretation is supplied by your program itself. The @var{value}
1853 parameter is optional; if it is eliminated, the variable is set to a
1855 @c "any string" here does not include leading, trailing
1856 @c blanks. Gnu asks: does anyone care?
1858 For example, this command:
1865 tells the debugged program, when subsequently run, that its user is named
1866 @samp{foo}. (The spaces around @samp{=} are used for clarity here; they
1867 are not actually required.)
1869 @kindex unset environment
1870 @item unset environment @var{varname}
1871 Remove variable @var{varname} from the environment to be passed to your
1872 program. This is different from @samp{set env @var{varname} =};
1873 @code{unset environment} removes the variable from the environment,
1874 rather than assigning it an empty value.
1877 @emph{Warning:} On Unix systems, @value{GDBN} runs your program using
1879 by your @code{SHELL} environment variable if it exists (or
1880 @code{/bin/sh} if not). If your @code{SHELL} variable names a shell
1881 that runs an initialization file---such as @file{.cshrc} for C-shell, or
1882 @file{.bashrc} for BASH---any variables you set in that file affect
1883 your program. You may wish to move setting of environment variables to
1884 files that are only run when you sign on, such as @file{.login} or
1887 @node Working Directory
1888 @section Your program's working directory
1890 @cindex working directory (of your program)
1891 Each time you start your program with @code{run}, it inherits its
1892 working directory from the current working directory of @value{GDBN}.
1893 The @value{GDBN} working directory is initially whatever it inherited
1894 from its parent process (typically the shell), but you can specify a new
1895 working directory in @value{GDBN} with the @code{cd} command.
1897 The @value{GDBN} working directory also serves as a default for the commands
1898 that specify files for @value{GDBN} to operate on. @xref{Files, ,Commands to
1903 @item cd @var{directory}
1904 Set the @value{GDBN} working directory to @var{directory}.
1908 Print the @value{GDBN} working directory.
1912 @section Your program's input and output
1917 By default, the program you run under @value{GDBN} does input and output to
1918 the same terminal that @value{GDBN} uses. @value{GDBN} switches the terminal
1919 to its own terminal modes to interact with you, but it records the terminal
1920 modes your program was using and switches back to them when you continue
1921 running your program.
1924 @kindex info terminal
1926 Displays information recorded by @value{GDBN} about the terminal modes your
1930 You can redirect your program's input and/or output using shell
1931 redirection with the @code{run} command. For example,
1938 starts your program, diverting its output to the file @file{outfile}.
1941 @cindex controlling terminal
1942 Another way to specify where your program should do input and output is
1943 with the @code{tty} command. This command accepts a file name as
1944 argument, and causes this file to be the default for future @code{run}
1945 commands. It also resets the controlling terminal for the child
1946 process, for future @code{run} commands. For example,
1953 directs that processes started with subsequent @code{run} commands
1954 default to do input and output on the terminal @file{/dev/ttyb} and have
1955 that as their controlling terminal.
1957 An explicit redirection in @code{run} overrides the @code{tty} command's
1958 effect on the input/output device, but not its effect on the controlling
1961 When you use the @code{tty} command or redirect input in the @code{run}
1962 command, only the input @emph{for your program} is affected. The input
1963 for @value{GDBN} still comes from your terminal.
1966 @section Debugging an already-running process
1971 @item attach @var{process-id}
1972 This command attaches to a running process---one that was started
1973 outside @value{GDBN}. (@code{info files} shows your active
1974 targets.) The command takes as argument a process ID. The usual way to
1975 find out the process-id of a Unix process is with the @code{ps} utility,
1976 or with the @samp{jobs -l} shell command.
1978 @code{attach} does not repeat if you press @key{RET} a second time after
1979 executing the command.
1982 To use @code{attach}, your program must be running in an environment
1983 which supports processes; for example, @code{attach} does not work for
1984 programs on bare-board targets that lack an operating system. You must
1985 also have permission to send the process a signal.
1987 When you use @code{attach}, the debugger finds the program running in
1988 the process first by looking in the current working directory, then (if
1989 the program is not found) by using the source file search path
1990 (@pxref{Source Path, ,Specifying source directories}). You can also use
1991 the @code{file} command to load the program. @xref{Files, ,Commands to
1994 The first thing @value{GDBN} does after arranging to debug the specified
1995 process is to stop it. You can examine and modify an attached process
1996 with all the @value{GDBN} commands that are ordinarily available when
1997 you start processes with @code{run}. You can insert breakpoints; you
1998 can step and continue; you can modify storage. If you would rather the
1999 process continue running, you may use the @code{continue} command after
2000 attaching @value{GDBN} to the process.
2005 When you have finished debugging the attached process, you can use the
2006 @code{detach} command to release it from @value{GDBN} control. Detaching
2007 the process continues its execution. After the @code{detach} command,
2008 that process and @value{GDBN} become completely independent once more, and you
2009 are ready to @code{attach} another process or start one with @code{run}.
2010 @code{detach} does not repeat if you press @key{RET} again after
2011 executing the command.
2014 If you exit @value{GDBN} or use the @code{run} command while you have an
2015 attached process, you kill that process. By default, @value{GDBN} asks
2016 for confirmation if you try to do either of these things; you can
2017 control whether or not you need to confirm by using the @code{set
2018 confirm} command (@pxref{Messages/Warnings, ,Optional warnings and
2022 @section Killing the child process
2027 Kill the child process in which your program is running under @value{GDBN}.
2030 This command is useful if you wish to debug a core dump instead of a
2031 running process. @value{GDBN} ignores any core dump file while your program
2034 On some operating systems, a program cannot be executed outside @value{GDBN}
2035 while you have breakpoints set on it inside @value{GDBN}. You can use the
2036 @code{kill} command in this situation to permit running your program
2037 outside the debugger.
2039 The @code{kill} command is also useful if you wish to recompile and
2040 relink your program, since on many systems it is impossible to modify an
2041 executable file while it is running in a process. In this case, when you
2042 next type @code{run}, @value{GDBN} notices that the file has changed, and
2043 reads the symbol table again (while trying to preserve your current
2044 breakpoint settings).
2047 @section Debugging programs with multiple threads
2049 @cindex threads of execution
2050 @cindex multiple threads
2051 @cindex switching threads
2052 In some operating systems, such as HP-UX and Solaris, a single program
2053 may have more than one @dfn{thread} of execution. The precise semantics
2054 of threads differ from one operating system to another, but in general
2055 the threads of a single program are akin to multiple processes---except
2056 that they share one address space (that is, they can all examine and
2057 modify the same variables). On the other hand, each thread has its own
2058 registers and execution stack, and perhaps private memory.
2060 @value{GDBN} provides these facilities for debugging multi-thread
2064 @item automatic notification of new threads
2065 @item @samp{thread @var{threadno}}, a command to switch among threads
2066 @item @samp{info threads}, a command to inquire about existing threads
2067 @item @samp{thread apply [@var{threadno}] [@var{all}] @var{args}},
2068 a command to apply a command to a list of threads
2069 @item thread-specific breakpoints
2073 @emph{Warning:} These facilities are not yet available on every
2074 @value{GDBN} configuration where the operating system supports threads.
2075 If your @value{GDBN} does not support threads, these commands have no
2076 effect. For example, a system without thread support shows no output
2077 from @samp{info threads}, and always rejects the @code{thread} command,
2081 (@value{GDBP}) info threads
2082 (@value{GDBP}) thread 1
2083 Thread ID 1 not known. Use the "info threads" command to
2084 see the IDs of currently known threads.
2086 @c FIXME to implementors: how hard would it be to say "sorry, this GDB
2087 @c doesn't support threads"?
2090 @cindex focus of debugging
2091 @cindex current thread
2092 The @value{GDBN} thread debugging facility allows you to observe all
2093 threads while your program runs---but whenever @value{GDBN} takes
2094 control, one thread in particular is always the focus of debugging.
2095 This thread is called the @dfn{current thread}. Debugging commands show
2096 program information from the perspective of the current thread.
2098 @cindex @code{New} @var{systag} message
2099 @cindex thread identifier (system)
2100 @c FIXME-implementors!! It would be more helpful if the [New...] message
2101 @c included GDB's numeric thread handle, so you could just go to that
2102 @c thread without first checking `info threads'.
2103 Whenever @value{GDBN} detects a new thread in your program, it displays
2104 the target system's identification for the thread with a message in the
2105 form @samp{[New @var{systag}]}. @var{systag} is a thread identifier
2106 whose form varies depending on the particular system. For example, on
2107 LynxOS, you might see
2110 [New process 35 thread 27]
2114 when @value{GDBN} notices a new thread. In contrast, on an SGI system,
2115 the @var{systag} is simply something like @samp{process 368}, with no
2118 @c FIXME!! (1) Does the [New...] message appear even for the very first
2119 @c thread of a program, or does it only appear for the
2120 @c second---i.e., when it becomes obvious we have a multithread
2122 @c (2) *Is* there necessarily a first thread always? Or do some
2123 @c multithread systems permit starting a program with multiple
2124 @c threads ab initio?
2126 @cindex thread number
2127 @cindex thread identifier (GDB)
2128 For debugging purposes, @value{GDBN} associates its own thread
2129 number---always a single integer---with each thread in your program.
2132 @kindex info threads
2134 Display a summary of all threads currently in your
2135 program. @value{GDBN} displays for each thread (in this order):
2138 @item the thread number assigned by @value{GDBN}
2140 @item the target system's thread identifier (@var{systag})
2142 @item the current stack frame summary for that thread
2146 An asterisk @samp{*} to the left of the @value{GDBN} thread number
2147 indicates the current thread.
2151 @c end table here to get a little more width for example
2154 (@value{GDBP}) info threads
2155 3 process 35 thread 27 0x34e5 in sigpause ()
2156 2 process 35 thread 23 0x34e5 in sigpause ()
2157 * 1 process 35 thread 13 main (argc=1, argv=0x7ffffff8)
2163 @cindex thread number
2164 @cindex thread identifier (GDB)
2165 For debugging purposes, @value{GDBN} associates its own thread
2166 number---a small integer assigned in thread-creation order---with each
2167 thread in your program.
2169 @cindex @code{New} @var{systag} message, on HP-UX
2170 @cindex thread identifier (system), on HP-UX
2171 @c FIXME-implementors!! It would be more helpful if the [New...] message
2172 @c included GDB's numeric thread handle, so you could just go to that
2173 @c thread without first checking `info threads'.
2174 Whenever @value{GDBN} detects a new thread in your program, it displays
2175 both @value{GDBN}'s thread number and the target system's identification for the thread with a message in the
2176 form @samp{[New @var{systag}]}. @var{systag} is a thread identifier
2177 whose form varies depending on the particular system. For example, on
2181 [New thread 2 (system thread 26594)]
2185 when @value{GDBN} notices a new thread.
2188 @kindex info threads
2190 Display a summary of all threads currently in your
2191 program. @value{GDBN} displays for each thread (in this order):
2194 @item the thread number assigned by @value{GDBN}
2196 @item the target system's thread identifier (@var{systag})
2198 @item the current stack frame summary for that thread
2202 An asterisk @samp{*} to the left of the @value{GDBN} thread number
2203 indicates the current thread.
2207 @c end table here to get a little more width for example
2210 (@value{GDBP}) info threads
2211 * 3 system thread 26607 worker (wptr=0x7b09c318 "@@") \@*
2213 2 system thread 26606 0x7b0030d8 in __ksleep () \@*
2214 from /usr/lib/libc.2
2215 1 system thread 27905 0x7b003498 in _brk () \@*
2216 from /usr/lib/libc.2
2220 @kindex thread @var{threadno}
2221 @item thread @var{threadno}
2222 Make thread number @var{threadno} the current thread. The command
2223 argument @var{threadno} is the internal @value{GDBN} thread number, as
2224 shown in the first field of the @samp{info threads} display.
2225 @value{GDBN} responds by displaying the system identifier of the thread
2226 you selected, and its current stack frame summary:
2229 @c FIXME!! This example made up; find a @value{GDBN} w/threads and get real one
2230 (@value{GDBP}) thread 2
2231 [Switching to process 35 thread 23]
2232 0x34e5 in sigpause ()
2236 As with the @samp{[New @dots{}]} message, the form of the text after
2237 @samp{Switching to} depends on your system's conventions for identifying
2240 @kindex thread apply
2241 @item thread apply [@var{threadno}] [@var{all}] @var{args}
2242 The @code{thread apply} command allows you to apply a command to one or
2243 more threads. Specify the numbers of the threads that you want affected
2244 with the command argument @var{threadno}. @var{threadno} is the internal
2245 @value{GDBN} thread number, as shown in the first field of the @samp{info
2246 threads} display. To apply a command to all threads, use
2247 @code{thread apply all} @var{args}.
2250 @cindex automatic thread selection
2251 @cindex switching threads automatically
2252 @cindex threads, automatic switching
2253 Whenever @value{GDBN} stops your program, due to a breakpoint or a
2254 signal, it automatically selects the thread where that breakpoint or
2255 signal happened. @value{GDBN} alerts you to the context switch with a
2256 message of the form @samp{[Switching to @var{systag}]} to identify the
2259 @xref{Thread Stops,,Stopping and starting multi-thread programs}, for
2260 more information about how @value{GDBN} behaves when you stop and start
2261 programs with multiple threads.
2263 @xref{Set Watchpoints,,Setting watchpoints}, for information about
2264 watchpoints in programs with multiple threads.
2267 @section Debugging programs with multiple processes
2269 @cindex fork, debugging programs which call
2270 @cindex multiple processes
2271 @cindex processes, multiple
2272 On most systems, @value{GDBN} has no special support for debugging
2273 programs which create additional processes using the @code{fork}
2274 function. When a program forks, @value{GDBN} will continue to debug the
2275 parent process and the child process will run unimpeded. If you have
2276 set a breakpoint in any code which the child then executes, the child
2277 will get a @code{SIGTRAP} signal which (unless it catches the signal)
2278 will cause it to terminate.
2280 However, if you want to debug the child process there is a workaround
2281 which isn't too painful. Put a call to @code{sleep} in the code which
2282 the child process executes after the fork. It may be useful to sleep
2283 only if a certain environment variable is set, or a certain file exists,
2284 so that the delay need not occur when you don't want to run @value{GDBN}
2285 on the child. While the child is sleeping, use the @code{ps} program to
2286 get its process ID. Then tell @value{GDBN} (a new invocation of
2287 @value{GDBN} if you are also debugging the parent process) to attach to
2288 the child process (@pxref{Attach}). From that point on you can debug
2289 the child process just like any other process which you attached to.
2291 On HP-UX (11.x and later only?), @value{GDBN} provides support for
2292 debugging programs that create additional processes using the
2293 @code{fork} or @code{vfork} function.
2295 By default, when a program forks, @value{GDBN} will continue to debug
2296 the parent process and the child process will run unimpeded.
2298 If you want to follow the child process instead of the parent process,
2299 use the command @w{@code{set follow-fork-mode}}.
2302 @kindex set follow-fork-mode
2303 @item set follow-fork-mode @var{mode}
2304 Set the debugger response to a program call of @code{fork} or
2305 @code{vfork}. A call to @code{fork} or @code{vfork} creates a new
2306 process. The @var{mode} can be:
2310 The original process is debugged after a fork. The child process runs
2311 unimpeded. This is the default.
2314 The new process is debugged after a fork. The parent process runs
2318 The debugger will ask for one of the above choices.
2321 @item show follow-fork-mode
2322 Display the current debugger response to a @code{fork} or @code{vfork} call.
2325 If you ask to debug a child process and a @code{vfork} is followed by an
2326 @code{exec}, @value{GDBN} executes the new target up to the first
2327 breakpoint in the new target. If you have a breakpoint set on
2328 @code{main} in your original program, the breakpoint will also be set on
2329 the child process's @code{main}.
2331 When a child process is spawned by @code{vfork}, you cannot debug the
2332 child or parent until an @code{exec} call completes.
2334 If you issue a @code{run} command to @value{GDBN} after an @code{exec}
2335 call executes, the new target restarts. To restart the parent process,
2336 use the @code{file} command with the parent executable name as its
2339 You can use the @code{catch} command to make @value{GDBN} stop whenever
2340 a @code{fork}, @code{vfork}, or @code{exec} call is made. @xref{Set
2341 Catchpoints, ,Setting catchpoints}.
2344 @chapter Stopping and Continuing
2346 The principal purposes of using a debugger are so that you can stop your
2347 program before it terminates; or so that, if your program runs into
2348 trouble, you can investigate and find out why.
2350 Inside @value{GDBN}, your program may stop for any of several reasons,
2351 such as a signal, a breakpoint, or reaching a new line after a
2352 @value{GDBN} command such as @code{step}. You may then examine and
2353 change variables, set new breakpoints or remove old ones, and then
2354 continue execution. Usually, the messages shown by @value{GDBN} provide
2355 ample explanation of the status of your program---but you can also
2356 explicitly request this information at any time.
2359 @kindex info program
2361 Display information about the status of your program: whether it is
2362 running or not, what process it is, and why it stopped.
2366 * Breakpoints:: Breakpoints, watchpoints, and catchpoints
2367 * Continuing and Stepping:: Resuming execution
2369 * Thread Stops:: Stopping and starting multi-thread programs
2373 @section Breakpoints, watchpoints, and catchpoints
2376 A @dfn{breakpoint} makes your program stop whenever a certain point in
2377 the program is reached. For each breakpoint, you can add conditions to
2378 control in finer detail whether your program stops. You can set
2379 breakpoints with the @code{break} command and its variants (@pxref{Set
2380 Breaks, ,Setting breakpoints}), to specify the place where your program
2381 should stop by line number, function name or exact address in the
2384 In HP-UX, SunOS 4.x, SVR4, and Alpha OSF/1 configurations, you can set
2385 breakpoints in shared libraries before the executable is run. There is
2386 a minor limitation on HP-UX systems: you must wait until the executable
2387 is run in order to set breakpoints in shared library routines that are
2388 not called directly by the program (for example, routines that are
2389 arguments in a @code{pthread_create} call).
2392 @cindex memory tracing
2393 @cindex breakpoint on memory address
2394 @cindex breakpoint on variable modification
2395 A @dfn{watchpoint} is a special breakpoint that stops your program
2396 when the value of an expression changes. You must use a different
2397 command to set watchpoints (@pxref{Set Watchpoints, ,Setting
2398 watchpoints}), but aside from that, you can manage a watchpoint like
2399 any other breakpoint: you enable, disable, and delete both breakpoints
2400 and watchpoints using the same commands.
2402 You can arrange to have values from your program displayed automatically
2403 whenever @value{GDBN} stops at a breakpoint. @xref{Auto Display,,
2407 @cindex breakpoint on events
2408 A @dfn{catchpoint} is another special breakpoint that stops your program
2409 when a certain kind of event occurs, such as the throwing of a C@t{++}
2410 exception or the loading of a library. As with watchpoints, you use a
2411 different command to set a catchpoint (@pxref{Set Catchpoints, ,Setting
2412 catchpoints}), but aside from that, you can manage a catchpoint like any
2413 other breakpoint. (To stop when your program receives a signal, use the
2414 @code{handle} command; see @ref{Signals, ,Signals}.)
2416 @cindex breakpoint numbers
2417 @cindex numbers for breakpoints
2418 @value{GDBN} assigns a number to each breakpoint, watchpoint, or
2419 catchpoint when you create it; these numbers are successive integers
2420 starting with one. In many of the commands for controlling various
2421 features of breakpoints you use the breakpoint number to say which
2422 breakpoint you want to change. Each breakpoint may be @dfn{enabled} or
2423 @dfn{disabled}; if disabled, it has no effect on your program until you
2426 @cindex breakpoint ranges
2427 @cindex ranges of breakpoints
2428 Some @value{GDBN} commands accept a range of breakpoints on which to
2429 operate. A breakpoint range is either a single breakpoint number, like
2430 @samp{5}, or two such numbers, in increasing order, separated by a
2431 hyphen, like @samp{5-7}. When a breakpoint range is given to a command,
2432 all breakpoint in that range are operated on.
2435 * Set Breaks:: Setting breakpoints
2436 * Set Watchpoints:: Setting watchpoints
2437 * Set Catchpoints:: Setting catchpoints
2438 * Delete Breaks:: Deleting breakpoints
2439 * Disabling:: Disabling breakpoints
2440 * Conditions:: Break conditions
2441 * Break Commands:: Breakpoint command lists
2442 * Breakpoint Menus:: Breakpoint menus
2443 * Error in Breakpoints:: ``Cannot insert breakpoints''
2447 @subsection Setting breakpoints
2449 @c FIXME LMB what does GDB do if no code on line of breakpt?
2450 @c consider in particular declaration with/without initialization.
2452 @c FIXME 2 is there stuff on this already? break at fun start, already init?
2455 @kindex b @r{(@code{break})}
2456 @vindex $bpnum@r{, convenience variable}
2457 @cindex latest breakpoint
2458 Breakpoints are set with the @code{break} command (abbreviated
2459 @code{b}). The debugger convenience variable @samp{$bpnum} records the
2460 number of the breakpoint you've set most recently; see @ref{Convenience
2461 Vars,, Convenience variables}, for a discussion of what you can do with
2462 convenience variables.
2464 You have several ways to say where the breakpoint should go.
2467 @item break @var{function}
2468 Set a breakpoint at entry to function @var{function}.
2469 When using source languages that permit overloading of symbols, such as
2470 C@t{++}, @var{function} may refer to more than one possible place to break.
2471 @xref{Breakpoint Menus,,Breakpoint menus}, for a discussion of that situation.
2473 @item break +@var{offset}
2474 @itemx break -@var{offset}
2475 Set a breakpoint some number of lines forward or back from the position
2476 at which execution stopped in the currently selected @dfn{stack frame}.
2477 (@xref{Frames, ,Frames}, for a description of stack frames.)
2479 @item break @var{linenum}
2480 Set a breakpoint at line @var{linenum} in the current source file.
2481 The current source file is the last file whose source text was printed.
2482 The breakpoint will stop your program just before it executes any of the
2485 @item break @var{filename}:@var{linenum}
2486 Set a breakpoint at line @var{linenum} in source file @var{filename}.
2488 @item break @var{filename}:@var{function}
2489 Set a breakpoint at entry to function @var{function} found in file
2490 @var{filename}. Specifying a file name as well as a function name is
2491 superfluous except when multiple files contain similarly named
2494 @item break *@var{address}
2495 Set a breakpoint at address @var{address}. You can use this to set
2496 breakpoints in parts of your program which do not have debugging
2497 information or source files.
2500 When called without any arguments, @code{break} sets a breakpoint at
2501 the next instruction to be executed in the selected stack frame
2502 (@pxref{Stack, ,Examining the Stack}). In any selected frame but the
2503 innermost, this makes your program stop as soon as control
2504 returns to that frame. This is similar to the effect of a
2505 @code{finish} command in the frame inside the selected frame---except
2506 that @code{finish} does not leave an active breakpoint. If you use
2507 @code{break} without an argument in the innermost frame, @value{GDBN} stops
2508 the next time it reaches the current location; this may be useful
2511 @value{GDBN} normally ignores breakpoints when it resumes execution, until at
2512 least one instruction has been executed. If it did not do this, you
2513 would be unable to proceed past a breakpoint without first disabling the
2514 breakpoint. This rule applies whether or not the breakpoint already
2515 existed when your program stopped.
2517 @item break @dots{} if @var{cond}
2518 Set a breakpoint with condition @var{cond}; evaluate the expression
2519 @var{cond} each time the breakpoint is reached, and stop only if the
2520 value is nonzero---that is, if @var{cond} evaluates as true.
2521 @samp{@dots{}} stands for one of the possible arguments described
2522 above (or no argument) specifying where to break. @xref{Conditions,
2523 ,Break conditions}, for more information on breakpoint conditions.
2526 @item tbreak @var{args}
2527 Set a breakpoint enabled only for one stop. @var{args} are the
2528 same as for the @code{break} command, and the breakpoint is set in the same
2529 way, but the breakpoint is automatically deleted after the first time your
2530 program stops there. @xref{Disabling, ,Disabling breakpoints}.
2533 @item hbreak @var{args}
2534 Set a hardware-assisted breakpoint. @var{args} are the same as for the
2535 @code{break} command and the breakpoint is set in the same way, but the
2536 breakpoint requires hardware support and some target hardware may not
2537 have this support. The main purpose of this is EPROM/ROM code
2538 debugging, so you can set a breakpoint at an instruction without
2539 changing the instruction. This can be used with the new trap-generation
2540 provided by SPARClite DSU and some x86-based targets. These targets
2541 will generate traps when a program accesses some data or instruction
2542 address that is assigned to the debug registers. However the hardware
2543 breakpoint registers can take a limited number of breakpoints. For
2544 example, on the DSU, only two data breakpoints can be set at a time, and
2545 @value{GDBN} will reject this command if more than two are used. Delete
2546 or disable unused hardware breakpoints before setting new ones
2547 (@pxref{Disabling, ,Disabling}). @xref{Conditions, ,Break conditions}.
2550 @item thbreak @var{args}
2551 Set a hardware-assisted breakpoint enabled only for one stop. @var{args}
2552 are the same as for the @code{hbreak} command and the breakpoint is set in
2553 the same way. However, like the @code{tbreak} command,
2554 the breakpoint is automatically deleted after the
2555 first time your program stops there. Also, like the @code{hbreak}
2556 command, the breakpoint requires hardware support and some target hardware
2557 may not have this support. @xref{Disabling, ,Disabling breakpoints}.
2558 See also @ref{Conditions, ,Break conditions}.
2561 @cindex regular expression
2562 @item rbreak @var{regex}
2563 Set breakpoints on all functions matching the regular expression
2564 @var{regex}. This command sets an unconditional breakpoint on all
2565 matches, printing a list of all breakpoints it set. Once these
2566 breakpoints are set, they are treated just like the breakpoints set with
2567 the @code{break} command. You can delete them, disable them, or make
2568 them conditional the same way as any other breakpoint.
2570 The syntax of the regular expression is the standard one used with tools
2571 like @file{grep}. Note that this is different from the syntax used by
2572 shells, so for instance @code{foo*} matches all functions that include
2573 an @code{fo} followed by zero or more @code{o}s. There is an implicit
2574 @code{.*} leading and trailing the regular expression you supply, so to
2575 match only functions that begin with @code{foo}, use @code{^foo}.
2577 When debugging C@t{++} programs, @code{rbreak} is useful for setting
2578 breakpoints on overloaded functions that are not members of any special
2581 @kindex info breakpoints
2582 @cindex @code{$_} and @code{info breakpoints}
2583 @item info breakpoints @r{[}@var{n}@r{]}
2584 @itemx info break @r{[}@var{n}@r{]}
2585 @itemx info watchpoints @r{[}@var{n}@r{]}
2586 Print a table of all breakpoints, watchpoints, and catchpoints set and
2587 not deleted, with the following columns for each breakpoint:
2590 @item Breakpoint Numbers
2592 Breakpoint, watchpoint, or catchpoint.
2594 Whether the breakpoint is marked to be disabled or deleted when hit.
2595 @item Enabled or Disabled
2596 Enabled breakpoints are marked with @samp{y}. @samp{n} marks breakpoints
2597 that are not enabled.
2599 Where the breakpoint is in your program, as a memory address.
2601 Where the breakpoint is in the source for your program, as a file and
2606 If a breakpoint is conditional, @code{info break} shows the condition on
2607 the line following the affected breakpoint; breakpoint commands, if any,
2608 are listed after that.
2611 @code{info break} with a breakpoint
2612 number @var{n} as argument lists only that breakpoint. The
2613 convenience variable @code{$_} and the default examining-address for
2614 the @code{x} command are set to the address of the last breakpoint
2615 listed (@pxref{Memory, ,Examining memory}).
2618 @code{info break} displays a count of the number of times the breakpoint
2619 has been hit. This is especially useful in conjunction with the
2620 @code{ignore} command. You can ignore a large number of breakpoint
2621 hits, look at the breakpoint info to see how many times the breakpoint
2622 was hit, and then run again, ignoring one less than that number. This
2623 will get you quickly to the last hit of that breakpoint.
2626 @value{GDBN} allows you to set any number of breakpoints at the same place in
2627 your program. There is nothing silly or meaningless about this. When
2628 the breakpoints are conditional, this is even useful
2629 (@pxref{Conditions, ,Break conditions}).
2631 @cindex negative breakpoint numbers
2632 @cindex internal @value{GDBN} breakpoints
2633 @value{GDBN} itself sometimes sets breakpoints in your program for special
2634 purposes, such as proper handling of @code{longjmp} (in C programs).
2635 These internal breakpoints are assigned negative numbers, starting with
2636 @code{-1}; @samp{info breakpoints} does not display them.
2638 You can see these breakpoints with the @value{GDBN} maintenance command
2639 @samp{maint info breakpoints}.
2642 @kindex maint info breakpoints
2643 @item maint info breakpoints
2644 Using the same format as @samp{info breakpoints}, display both the
2645 breakpoints you've set explicitly, and those @value{GDBN} is using for
2646 internal purposes. Internal breakpoints are shown with negative
2647 breakpoint numbers. The type column identifies what kind of breakpoint
2652 Normal, explicitly set breakpoint.
2655 Normal, explicitly set watchpoint.
2658 Internal breakpoint, used to handle correctly stepping through
2659 @code{longjmp} calls.
2661 @item longjmp resume
2662 Internal breakpoint at the target of a @code{longjmp}.
2665 Temporary internal breakpoint used by the @value{GDBN} @code{until} command.
2668 Temporary internal breakpoint used by the @value{GDBN} @code{finish} command.
2671 Shared library events.
2678 @node Set Watchpoints
2679 @subsection Setting watchpoints
2681 @cindex setting watchpoints
2682 @cindex software watchpoints
2683 @cindex hardware watchpoints
2684 You can use a watchpoint to stop execution whenever the value of an
2685 expression changes, without having to predict a particular place where
2688 Depending on your system, watchpoints may be implemented in software or
2689 hardware. @value{GDBN} does software watchpointing by single-stepping your
2690 program and testing the variable's value each time, which is hundreds of
2691 times slower than normal execution. (But this may still be worth it, to
2692 catch errors where you have no clue what part of your program is the
2695 On some systems, such as HP-UX, Linux and some other x86-based targets,
2696 @value{GDBN} includes support for
2697 hardware watchpoints, which do not slow down the running of your
2702 @item watch @var{expr}
2703 Set a watchpoint for an expression. @value{GDBN} will break when @var{expr}
2704 is written into by the program and its value changes.
2707 @item rwatch @var{expr}
2708 Set a watchpoint that will break when watch @var{expr} is read by the program.
2711 @item awatch @var{expr}
2712 Set a watchpoint that will break when @var{expr} is either read or written into
2715 @kindex info watchpoints
2716 @item info watchpoints
2717 This command prints a list of watchpoints, breakpoints, and catchpoints;
2718 it is the same as @code{info break}.
2721 @value{GDBN} sets a @dfn{hardware watchpoint} if possible. Hardware
2722 watchpoints execute very quickly, and the debugger reports a change in
2723 value at the exact instruction where the change occurs. If @value{GDBN}
2724 cannot set a hardware watchpoint, it sets a software watchpoint, which
2725 executes more slowly and reports the change in value at the next
2726 statement, not the instruction, after the change occurs.
2728 When you issue the @code{watch} command, @value{GDBN} reports
2731 Hardware watchpoint @var{num}: @var{expr}
2735 if it was able to set a hardware watchpoint.
2737 Currently, the @code{awatch} and @code{rwatch} commands can only set
2738 hardware watchpoints, because accesses to data that don't change the
2739 value of the watched expression cannot be detected without examining
2740 every instruction as it is being executed, and @value{GDBN} does not do
2741 that currently. If @value{GDBN} finds that it is unable to set a
2742 hardware breakpoint with the @code{awatch} or @code{rwatch} command, it
2743 will print a message like this:
2746 Expression cannot be implemented with read/access watchpoint.
2749 Sometimes, @value{GDBN} cannot set a hardware watchpoint because the
2750 data type of the watched expression is wider than what a hardware
2751 watchpoint on the target machine can handle. For example, some systems
2752 can only watch regions that are up to 4 bytes wide; on such systems you
2753 cannot set hardware watchpoints for an expression that yields a
2754 double-precision floating-point number (which is typically 8 bytes
2755 wide). As a work-around, it might be possible to break the large region
2756 into a series of smaller ones and watch them with separate watchpoints.
2758 If you set too many hardware watchpoints, @value{GDBN} might be unable
2759 to insert all of them when you resume the execution of your program.
2760 Since the precise number of active watchpoints is unknown until such
2761 time as the program is about to be resumed, @value{GDBN} might not be
2762 able to warn you about this when you set the watchpoints, and the
2763 warning will be printed only when the program is resumed:
2766 Hardware watchpoint @var{num}: Could not insert watchpoint
2770 If this happens, delete or disable some of the watchpoints.
2772 The SPARClite DSU will generate traps when a program accesses some data
2773 or instruction address that is assigned to the debug registers. For the
2774 data addresses, DSU facilitates the @code{watch} command. However the
2775 hardware breakpoint registers can only take two data watchpoints, and
2776 both watchpoints must be the same kind. For example, you can set two
2777 watchpoints with @code{watch} commands, two with @code{rwatch} commands,
2778 @strong{or} two with @code{awatch} commands, but you cannot set one
2779 watchpoint with one command and the other with a different command.
2780 @value{GDBN} will reject the command if you try to mix watchpoints.
2781 Delete or disable unused watchpoint commands before setting new ones.
2783 If you call a function interactively using @code{print} or @code{call},
2784 any watchpoints you have set will be inactive until @value{GDBN} reaches another
2785 kind of breakpoint or the call completes.
2787 @value{GDBN} automatically deletes watchpoints that watch local
2788 (automatic) variables, or expressions that involve such variables, when
2789 they go out of scope, that is, when the execution leaves the block in
2790 which these variables were defined. In particular, when the program
2791 being debugged terminates, @emph{all} local variables go out of scope,
2792 and so only watchpoints that watch global variables remain set. If you
2793 rerun the program, you will need to set all such watchpoints again. One
2794 way of doing that would be to set a code breakpoint at the entry to the
2795 @code{main} function and when it breaks, set all the watchpoints.
2798 @cindex watchpoints and threads
2799 @cindex threads and watchpoints
2800 @emph{Warning:} In multi-thread programs, watchpoints have only limited
2801 usefulness. With the current watchpoint implementation, @value{GDBN}
2802 can only watch the value of an expression @emph{in a single thread}. If
2803 you are confident that the expression can only change due to the current
2804 thread's activity (and if you are also confident that no other thread
2805 can become current), then you can use watchpoints as usual. However,
2806 @value{GDBN} may not notice when a non-current thread's activity changes
2809 @c FIXME: this is almost identical to the previous paragraph.
2810 @emph{HP-UX Warning:} In multi-thread programs, software watchpoints
2811 have only limited usefulness. If @value{GDBN} creates a software
2812 watchpoint, it can only watch the value of an expression @emph{in a
2813 single thread}. If you are confident that the expression can only
2814 change due to the current thread's activity (and if you are also
2815 confident that no other thread can become current), then you can use
2816 software watchpoints as usual. However, @value{GDBN} may not notice
2817 when a non-current thread's activity changes the expression. (Hardware
2818 watchpoints, in contrast, watch an expression in all threads.)
2821 @node Set Catchpoints
2822 @subsection Setting catchpoints
2823 @cindex catchpoints, setting
2824 @cindex exception handlers
2825 @cindex event handling
2827 You can use @dfn{catchpoints} to cause the debugger to stop for certain
2828 kinds of program events, such as C@t{++} exceptions or the loading of a
2829 shared library. Use the @code{catch} command to set a catchpoint.
2833 @item catch @var{event}
2834 Stop when @var{event} occurs. @var{event} can be any of the following:
2838 The throwing of a C@t{++} exception.
2842 The catching of a C@t{++} exception.
2846 A call to @code{exec}. This is currently only available for HP-UX.
2850 A call to @code{fork}. This is currently only available for HP-UX.
2854 A call to @code{vfork}. This is currently only available for HP-UX.
2857 @itemx load @var{libname}
2859 The dynamic loading of any shared library, or the loading of the library
2860 @var{libname}. This is currently only available for HP-UX.
2863 @itemx unload @var{libname}
2864 @kindex catch unload
2865 The unloading of any dynamically loaded shared library, or the unloading
2866 of the library @var{libname}. This is currently only available for HP-UX.
2869 @item tcatch @var{event}
2870 Set a catchpoint that is enabled only for one stop. The catchpoint is
2871 automatically deleted after the first time the event is caught.
2875 Use the @code{info break} command to list the current catchpoints.
2877 There are currently some limitations to C@t{++} exception handling
2878 (@code{catch throw} and @code{catch catch}) in @value{GDBN}:
2882 If you call a function interactively, @value{GDBN} normally returns
2883 control to you when the function has finished executing. If the call
2884 raises an exception, however, the call may bypass the mechanism that
2885 returns control to you and cause your program either to abort or to
2886 simply continue running until it hits a breakpoint, catches a signal
2887 that @value{GDBN} is listening for, or exits. This is the case even if
2888 you set a catchpoint for the exception; catchpoints on exceptions are
2889 disabled within interactive calls.
2892 You cannot raise an exception interactively.
2895 You cannot install an exception handler interactively.
2898 @cindex raise exceptions
2899 Sometimes @code{catch} is not the best way to debug exception handling:
2900 if you need to know exactly where an exception is raised, it is better to
2901 stop @emph{before} the exception handler is called, since that way you
2902 can see the stack before any unwinding takes place. If you set a
2903 breakpoint in an exception handler instead, it may not be easy to find
2904 out where the exception was raised.
2906 To stop just before an exception handler is called, you need some
2907 knowledge of the implementation. In the case of @sc{gnu} C@t{++}, exceptions are
2908 raised by calling a library function named @code{__raise_exception}
2909 which has the following ANSI C interface:
2912 /* @var{addr} is where the exception identifier is stored.
2913 @var{id} is the exception identifier. */
2914 void __raise_exception (void **addr, void *id);
2918 To make the debugger catch all exceptions before any stack
2919 unwinding takes place, set a breakpoint on @code{__raise_exception}
2920 (@pxref{Breakpoints, ,Breakpoints; watchpoints; and exceptions}).
2922 With a conditional breakpoint (@pxref{Conditions, ,Break conditions})
2923 that depends on the value of @var{id}, you can stop your program when
2924 a specific exception is raised. You can use multiple conditional
2925 breakpoints to stop your program when any of a number of exceptions are
2930 @subsection Deleting breakpoints
2932 @cindex clearing breakpoints, watchpoints, catchpoints
2933 @cindex deleting breakpoints, watchpoints, catchpoints
2934 It is often necessary to eliminate a breakpoint, watchpoint, or
2935 catchpoint once it has done its job and you no longer want your program
2936 to stop there. This is called @dfn{deleting} the breakpoint. A
2937 breakpoint that has been deleted no longer exists; it is forgotten.
2939 With the @code{clear} command you can delete breakpoints according to
2940 where they are in your program. With the @code{delete} command you can
2941 delete individual breakpoints, watchpoints, or catchpoints by specifying
2942 their breakpoint numbers.
2944 It is not necessary to delete a breakpoint to proceed past it. @value{GDBN}
2945 automatically ignores breakpoints on the first instruction to be executed
2946 when you continue execution without changing the execution address.
2951 Delete any breakpoints at the next instruction to be executed in the
2952 selected stack frame (@pxref{Selection, ,Selecting a frame}). When
2953 the innermost frame is selected, this is a good way to delete a
2954 breakpoint where your program just stopped.
2956 @item clear @var{function}
2957 @itemx clear @var{filename}:@var{function}
2958 Delete any breakpoints set at entry to the function @var{function}.
2960 @item clear @var{linenum}
2961 @itemx clear @var{filename}:@var{linenum}
2962 Delete any breakpoints set at or within the code of the specified line.
2964 @cindex delete breakpoints
2966 @kindex d @r{(@code{delete})}
2967 @item delete @r{[}breakpoints@r{]} @r{[}@var{range}@dots{}@r{]}
2968 Delete the breakpoints, watchpoints, or catchpoints of the breakpoint
2969 ranges specified as arguments. If no argument is specified, delete all
2970 breakpoints (@value{GDBN} asks confirmation, unless you have @code{set
2971 confirm off}). You can abbreviate this command as @code{d}.
2975 @subsection Disabling breakpoints
2977 @kindex disable breakpoints
2978 @kindex enable breakpoints
2979 Rather than deleting a breakpoint, watchpoint, or catchpoint, you might
2980 prefer to @dfn{disable} it. This makes the breakpoint inoperative as if
2981 it had been deleted, but remembers the information on the breakpoint so
2982 that you can @dfn{enable} it again later.
2984 You disable and enable breakpoints, watchpoints, and catchpoints with
2985 the @code{enable} and @code{disable} commands, optionally specifying one
2986 or more breakpoint numbers as arguments. Use @code{info break} or
2987 @code{info watch} to print a list of breakpoints, watchpoints, and
2988 catchpoints if you do not know which numbers to use.
2990 A breakpoint, watchpoint, or catchpoint can have any of four different
2991 states of enablement:
2995 Enabled. The breakpoint stops your program. A breakpoint set
2996 with the @code{break} command starts out in this state.
2998 Disabled. The breakpoint has no effect on your program.
3000 Enabled once. The breakpoint stops your program, but then becomes
3003 Enabled for deletion. The breakpoint stops your program, but
3004 immediately after it does so it is deleted permanently. A breakpoint
3005 set with the @code{tbreak} command starts out in this state.
3008 You can use the following commands to enable or disable breakpoints,
3009 watchpoints, and catchpoints:
3012 @kindex disable breakpoints
3014 @kindex dis @r{(@code{disable})}
3015 @item disable @r{[}breakpoints@r{]} @r{[}@var{range}@dots{}@r{]}
3016 Disable the specified breakpoints---or all breakpoints, if none are
3017 listed. A disabled breakpoint has no effect but is not forgotten. All
3018 options such as ignore-counts, conditions and commands are remembered in
3019 case the breakpoint is enabled again later. You may abbreviate
3020 @code{disable} as @code{dis}.
3022 @kindex enable breakpoints
3024 @item enable @r{[}breakpoints@r{]} @r{[}@var{range}@dots{}@r{]}
3025 Enable the specified breakpoints (or all defined breakpoints). They
3026 become effective once again in stopping your program.
3028 @item enable @r{[}breakpoints@r{]} once @var{range}@dots{}
3029 Enable the specified breakpoints temporarily. @value{GDBN} disables any
3030 of these breakpoints immediately after stopping your program.
3032 @item enable @r{[}breakpoints@r{]} delete @var{range}@dots{}
3033 Enable the specified breakpoints to work once, then die. @value{GDBN}
3034 deletes any of these breakpoints as soon as your program stops there.
3037 @c FIXME: I think the following ``Except for [...] @code{tbreak}'' is
3038 @c confusing: tbreak is also initially enabled.
3039 Except for a breakpoint set with @code{tbreak} (@pxref{Set Breaks,
3040 ,Setting breakpoints}), breakpoints that you set are initially enabled;
3041 subsequently, they become disabled or enabled only when you use one of
3042 the commands above. (The command @code{until} can set and delete a
3043 breakpoint of its own, but it does not change the state of your other
3044 breakpoints; see @ref{Continuing and Stepping, ,Continuing and
3048 @subsection Break conditions
3049 @cindex conditional breakpoints
3050 @cindex breakpoint conditions
3052 @c FIXME what is scope of break condition expr? Context where wanted?
3053 @c in particular for a watchpoint?
3054 The simplest sort of breakpoint breaks every time your program reaches a
3055 specified place. You can also specify a @dfn{condition} for a
3056 breakpoint. A condition is just a Boolean expression in your
3057 programming language (@pxref{Expressions, ,Expressions}). A breakpoint with
3058 a condition evaluates the expression each time your program reaches it,
3059 and your program stops only if the condition is @emph{true}.
3061 This is the converse of using assertions for program validation; in that
3062 situation, you want to stop when the assertion is violated---that is,
3063 when the condition is false. In C, if you want to test an assertion expressed
3064 by the condition @var{assert}, you should set the condition
3065 @samp{! @var{assert}} on the appropriate breakpoint.
3067 Conditions are also accepted for watchpoints; you may not need them,
3068 since a watchpoint is inspecting the value of an expression anyhow---but
3069 it might be simpler, say, to just set a watchpoint on a variable name,
3070 and specify a condition that tests whether the new value is an interesting
3073 Break conditions can have side effects, and may even call functions in
3074 your program. This can be useful, for example, to activate functions
3075 that log program progress, or to use your own print functions to
3076 format special data structures. The effects are completely predictable
3077 unless there is another enabled breakpoint at the same address. (In
3078 that case, @value{GDBN} might see the other breakpoint first and stop your
3079 program without checking the condition of this one.) Note that
3080 breakpoint commands are usually more convenient and flexible than break
3082 purpose of performing side effects when a breakpoint is reached
3083 (@pxref{Break Commands, ,Breakpoint command lists}).
3085 Break conditions can be specified when a breakpoint is set, by using
3086 @samp{if} in the arguments to the @code{break} command. @xref{Set
3087 Breaks, ,Setting breakpoints}. They can also be changed at any time
3088 with the @code{condition} command.
3090 You can also use the @code{if} keyword with the @code{watch} command.
3091 The @code{catch} command does not recognize the @code{if} keyword;
3092 @code{condition} is the only way to impose a further condition on a
3097 @item condition @var{bnum} @var{expression}
3098 Specify @var{expression} as the break condition for breakpoint,
3099 watchpoint, or catchpoint number @var{bnum}. After you set a condition,
3100 breakpoint @var{bnum} stops your program only if the value of
3101 @var{expression} is true (nonzero, in C). When you use
3102 @code{condition}, @value{GDBN} checks @var{expression} immediately for
3103 syntactic correctness, and to determine whether symbols in it have
3104 referents in the context of your breakpoint. If @var{expression} uses
3105 symbols not referenced in the context of the breakpoint, @value{GDBN}
3106 prints an error message:
3109 No symbol "foo" in current context.
3114 not actually evaluate @var{expression} at the time the @code{condition}
3115 command (or a command that sets a breakpoint with a condition, like
3116 @code{break if @dots{}}) is given, however. @xref{Expressions, ,Expressions}.
3118 @item condition @var{bnum}
3119 Remove the condition from breakpoint number @var{bnum}. It becomes
3120 an ordinary unconditional breakpoint.
3123 @cindex ignore count (of breakpoint)
3124 A special case of a breakpoint condition is to stop only when the
3125 breakpoint has been reached a certain number of times. This is so
3126 useful that there is a special way to do it, using the @dfn{ignore
3127 count} of the breakpoint. Every breakpoint has an ignore count, which
3128 is an integer. Most of the time, the ignore count is zero, and
3129 therefore has no effect. But if your program reaches a breakpoint whose
3130 ignore count is positive, then instead of stopping, it just decrements
3131 the ignore count by one and continues. As a result, if the ignore count
3132 value is @var{n}, the breakpoint does not stop the next @var{n} times
3133 your program reaches it.
3137 @item ignore @var{bnum} @var{count}
3138 Set the ignore count of breakpoint number @var{bnum} to @var{count}.
3139 The next @var{count} times the breakpoint is reached, your program's
3140 execution does not stop; other than to decrement the ignore count, @value{GDBN}
3143 To make the breakpoint stop the next time it is reached, specify
3146 When you use @code{continue} to resume execution of your program from a
3147 breakpoint, you can specify an ignore count directly as an argument to
3148 @code{continue}, rather than using @code{ignore}. @xref{Continuing and
3149 Stepping,,Continuing and stepping}.
3151 If a breakpoint has a positive ignore count and a condition, the
3152 condition is not checked. Once the ignore count reaches zero,
3153 @value{GDBN} resumes checking the condition.
3155 You could achieve the effect of the ignore count with a condition such
3156 as @w{@samp{$foo-- <= 0}} using a debugger convenience variable that
3157 is decremented each time. @xref{Convenience Vars, ,Convenience
3161 Ignore counts apply to breakpoints, watchpoints, and catchpoints.
3164 @node Break Commands
3165 @subsection Breakpoint command lists
3167 @cindex breakpoint commands
3168 You can give any breakpoint (or watchpoint or catchpoint) a series of
3169 commands to execute when your program stops due to that breakpoint. For
3170 example, you might want to print the values of certain expressions, or
3171 enable other breakpoints.
3176 @item commands @r{[}@var{bnum}@r{]}
3177 @itemx @dots{} @var{command-list} @dots{}
3179 Specify a list of commands for breakpoint number @var{bnum}. The commands
3180 themselves appear on the following lines. Type a line containing just
3181 @code{end} to terminate the commands.
3183 To remove all commands from a breakpoint, type @code{commands} and
3184 follow it immediately with @code{end}; that is, give no commands.
3186 With no @var{bnum} argument, @code{commands} refers to the last
3187 breakpoint, watchpoint, or catchpoint set (not to the breakpoint most
3188 recently encountered).
3191 Pressing @key{RET} as a means of repeating the last @value{GDBN} command is
3192 disabled within a @var{command-list}.
3194 You can use breakpoint commands to start your program up again. Simply
3195 use the @code{continue} command, or @code{step}, or any other command
3196 that resumes execution.
3198 Any other commands in the command list, after a command that resumes
3199 execution, are ignored. This is because any time you resume execution
3200 (even with a simple @code{next} or @code{step}), you may encounter
3201 another breakpoint---which could have its own command list, leading to
3202 ambiguities about which list to execute.
3205 If the first command you specify in a command list is @code{silent}, the
3206 usual message about stopping at a breakpoint is not printed. This may
3207 be desirable for breakpoints that are to print a specific message and
3208 then continue. If none of the remaining commands print anything, you
3209 see no sign that the breakpoint was reached. @code{silent} is
3210 meaningful only at the beginning of a breakpoint command list.
3212 The commands @code{echo}, @code{output}, and @code{printf} allow you to
3213 print precisely controlled output, and are often useful in silent
3214 breakpoints. @xref{Output, ,Commands for controlled output}.
3216 For example, here is how you could use breakpoint commands to print the
3217 value of @code{x} at entry to @code{foo} whenever @code{x} is positive.
3223 printf "x is %d\n",x
3228 One application for breakpoint commands is to compensate for one bug so
3229 you can test for another. Put a breakpoint just after the erroneous line
3230 of code, give it a condition to detect the case in which something
3231 erroneous has been done, and give it commands to assign correct values
3232 to any variables that need them. End with the @code{continue} command
3233 so that your program does not stop, and start with the @code{silent}
3234 command so that no output is produced. Here is an example:
3245 @node Breakpoint Menus
3246 @subsection Breakpoint menus
3248 @cindex symbol overloading
3250 Some programming languages (notably C@t{++}) permit a single function name
3251 to be defined several times, for application in different contexts.
3252 This is called @dfn{overloading}. When a function name is overloaded,
3253 @samp{break @var{function}} is not enough to tell @value{GDBN} where you want
3254 a breakpoint. If you realize this is a problem, you can use
3255 something like @samp{break @var{function}(@var{types})} to specify which
3256 particular version of the function you want. Otherwise, @value{GDBN} offers
3257 you a menu of numbered choices for different possible breakpoints, and
3258 waits for your selection with the prompt @samp{>}. The first two
3259 options are always @samp{[0] cancel} and @samp{[1] all}. Typing @kbd{1}
3260 sets a breakpoint at each definition of @var{function}, and typing
3261 @kbd{0} aborts the @code{break} command without setting any new
3264 For example, the following session excerpt shows an attempt to set a
3265 breakpoint at the overloaded symbol @code{String::after}.
3266 We choose three particular definitions of that function name:
3268 @c FIXME! This is likely to change to show arg type lists, at least
3271 (@value{GDBP}) b String::after
3274 [2] file:String.cc; line number:867
3275 [3] file:String.cc; line number:860
3276 [4] file:String.cc; line number:875
3277 [5] file:String.cc; line number:853
3278 [6] file:String.cc; line number:846
3279 [7] file:String.cc; line number:735
3281 Breakpoint 1 at 0xb26c: file String.cc, line 867.
3282 Breakpoint 2 at 0xb344: file String.cc, line 875.
3283 Breakpoint 3 at 0xafcc: file String.cc, line 846.
3284 Multiple breakpoints were set.
3285 Use the "delete" command to delete unwanted
3291 @c @ifclear BARETARGET
3292 @node Error in Breakpoints
3293 @subsection ``Cannot insert breakpoints''
3295 @c FIXME!! 14/6/95 Is there a real example of this? Let's use it.
3297 Under some operating systems, breakpoints cannot be used in a program if
3298 any other process is running that program. In this situation,
3299 attempting to run or continue a program with a breakpoint causes
3300 @value{GDBN} to print an error message:
3303 Cannot insert breakpoints.
3304 The same program may be running in another process.
3307 When this happens, you have three ways to proceed:
3311 Remove or disable the breakpoints, then continue.
3314 Suspend @value{GDBN}, and copy the file containing your program to a new
3315 name. Resume @value{GDBN} and use the @code{exec-file} command to specify
3316 that @value{GDBN} should run your program under that name.
3317 Then start your program again.
3320 Relink your program so that the text segment is nonsharable, using the
3321 linker option @samp{-N}. The operating system limitation may not apply
3322 to nonsharable executables.
3326 A similar message can be printed if you request too many active
3327 hardware-assisted breakpoints and watchpoints:
3329 @c FIXME: the precise wording of this message may change; the relevant
3330 @c source change is not committed yet (Sep 3, 1999).
3332 Stopped; cannot insert breakpoints.
3333 You may have requested too many hardware breakpoints and watchpoints.
3337 This message is printed when you attempt to resume the program, since
3338 only then @value{GDBN} knows exactly how many hardware breakpoints and
3339 watchpoints it needs to insert.
3341 When this message is printed, you need to disable or remove some of the
3342 hardware-assisted breakpoints and watchpoints, and then continue.
3345 @node Continuing and Stepping
3346 @section Continuing and stepping
3350 @cindex resuming execution
3351 @dfn{Continuing} means resuming program execution until your program
3352 completes normally. In contrast, @dfn{stepping} means executing just
3353 one more ``step'' of your program, where ``step'' may mean either one
3354 line of source code, or one machine instruction (depending on what
3355 particular command you use). Either when continuing or when stepping,
3356 your program may stop even sooner, due to a breakpoint or a signal. (If
3357 it stops due to a signal, you may want to use @code{handle}, or use
3358 @samp{signal 0} to resume execution. @xref{Signals, ,Signals}.)
3362 @kindex c @r{(@code{continue})}
3363 @kindex fg @r{(resume foreground execution)}
3364 @item continue @r{[}@var{ignore-count}@r{]}
3365 @itemx c @r{[}@var{ignore-count}@r{]}
3366 @itemx fg @r{[}@var{ignore-count}@r{]}
3367 Resume program execution, at the address where your program last stopped;
3368 any breakpoints set at that address are bypassed. The optional argument
3369 @var{ignore-count} allows you to specify a further number of times to
3370 ignore a breakpoint at this location; its effect is like that of
3371 @code{ignore} (@pxref{Conditions, ,Break conditions}).
3373 The argument @var{ignore-count} is meaningful only when your program
3374 stopped due to a breakpoint. At other times, the argument to
3375 @code{continue} is ignored.
3377 The synonyms @code{c} and @code{fg} (for @dfn{foreground}, as the
3378 debugged program is deemed to be the foreground program) are provided
3379 purely for convenience, and have exactly the same behavior as
3383 To resume execution at a different place, you can use @code{return}
3384 (@pxref{Returning, ,Returning from a function}) to go back to the
3385 calling function; or @code{jump} (@pxref{Jumping, ,Continuing at a
3386 different address}) to go to an arbitrary location in your program.
3388 A typical technique for using stepping is to set a breakpoint
3389 (@pxref{Breakpoints, ,Breakpoints; watchpoints; and catchpoints}) at the
3390 beginning of the function or the section of your program where a problem
3391 is believed to lie, run your program until it stops at that breakpoint,
3392 and then step through the suspect area, examining the variables that are
3393 interesting, until you see the problem happen.
3397 @kindex s @r{(@code{step})}
3399 Continue running your program until control reaches a different source
3400 line, then stop it and return control to @value{GDBN}. This command is
3401 abbreviated @code{s}.
3404 @c "without debugging information" is imprecise; actually "without line
3405 @c numbers in the debugging information". (gcc -g1 has debugging info but
3406 @c not line numbers). But it seems complex to try to make that
3407 @c distinction here.
3408 @emph{Warning:} If you use the @code{step} command while control is
3409 within a function that was compiled without debugging information,
3410 execution proceeds until control reaches a function that does have
3411 debugging information. Likewise, it will not step into a function which
3412 is compiled without debugging information. To step through functions
3413 without debugging information, use the @code{stepi} command, described
3417 The @code{step} command only stops at the first instruction of a source
3418 line. This prevents the multiple stops that could otherwise occur in
3419 @code{switch} statements, @code{for} loops, etc. @code{step} continues
3420 to stop if a function that has debugging information is called within
3421 the line. In other words, @code{step} @emph{steps inside} any functions
3422 called within the line.
3424 Also, the @code{step} command only enters a function if there is line
3425 number information for the function. Otherwise it acts like the
3426 @code{next} command. This avoids problems when using @code{cc -gl}
3427 on MIPS machines. Previously, @code{step} entered subroutines if there
3428 was any debugging information about the routine.
3430 @item step @var{count}
3431 Continue running as in @code{step}, but do so @var{count} times. If a
3432 breakpoint is reached, or a signal not related to stepping occurs before
3433 @var{count} steps, stepping stops right away.
3436 @kindex n @r{(@code{next})}
3437 @item next @r{[}@var{count}@r{]}
3438 Continue to the next source line in the current (innermost) stack frame.
3439 This is similar to @code{step}, but function calls that appear within
3440 the line of code are executed without stopping. Execution stops when
3441 control reaches a different line of code at the original stack level
3442 that was executing when you gave the @code{next} command. This command
3443 is abbreviated @code{n}.
3445 An argument @var{count} is a repeat count, as for @code{step}.
3448 @c FIX ME!! Do we delete this, or is there a way it fits in with
3449 @c the following paragraph? --- Vctoria
3451 @c @code{next} within a function that lacks debugging information acts like
3452 @c @code{step}, but any function calls appearing within the code of the
3453 @c function are executed without stopping.
3455 The @code{next} command only stops at the first instruction of a
3456 source line. This prevents multiple stops that could otherwise occur in
3457 @code{switch} statements, @code{for} loops, etc.
3459 @kindex set step-mode
3461 @cindex functions without line info, and stepping
3462 @cindex stepping into functions with no line info
3463 @itemx set step-mode on
3464 The @code{set step-mode on} command causes the @code{step} command to
3465 stop at the first instruction of a function which contains no debug line
3466 information rather than stepping over it.
3468 This is useful in cases where you may be interested in inspecting the
3469 machine instructions of a function which has no symbolic info and do not
3470 want @value{GDBN} to automatically skip over this function.
3472 @item set step-mode off
3473 Causes the @code{step} command to step over any functions which contains no
3474 debug information. This is the default.
3478 Continue running until just after function in the selected stack frame
3479 returns. Print the returned value (if any).
3481 Contrast this with the @code{return} command (@pxref{Returning,
3482 ,Returning from a function}).
3485 @kindex u @r{(@code{until})}
3488 Continue running until a source line past the current line, in the
3489 current stack frame, is reached. This command is used to avoid single
3490 stepping through a loop more than once. It is like the @code{next}
3491 command, except that when @code{until} encounters a jump, it
3492 automatically continues execution until the program counter is greater
3493 than the address of the jump.
3495 This means that when you reach the end of a loop after single stepping
3496 though it, @code{until} makes your program continue execution until it
3497 exits the loop. In contrast, a @code{next} command at the end of a loop
3498 simply steps back to the beginning of the loop, which forces you to step
3499 through the next iteration.
3501 @code{until} always stops your program if it attempts to exit the current
3504 @code{until} may produce somewhat counterintuitive results if the order
3505 of machine code does not match the order of the source lines. For
3506 example, in the following excerpt from a debugging session, the @code{f}
3507 (@code{frame}) command shows that execution is stopped at line
3508 @code{206}; yet when we use @code{until}, we get to line @code{195}:
3512 #0 main (argc=4, argv=0xf7fffae8) at m4.c:206
3514 (@value{GDBP}) until
3515 195 for ( ; argc > 0; NEXTARG) @{
3518 This happened because, for execution efficiency, the compiler had
3519 generated code for the loop closure test at the end, rather than the
3520 start, of the loop---even though the test in a C @code{for}-loop is
3521 written before the body of the loop. The @code{until} command appeared
3522 to step back to the beginning of the loop when it advanced to this
3523 expression; however, it has not really gone to an earlier
3524 statement---not in terms of the actual machine code.
3526 @code{until} with no argument works by means of single
3527 instruction stepping, and hence is slower than @code{until} with an
3530 @item until @var{location}
3531 @itemx u @var{location}
3532 Continue running your program until either the specified location is
3533 reached, or the current stack frame returns. @var{location} is any of
3534 the forms of argument acceptable to @code{break} (@pxref{Set Breaks,
3535 ,Setting breakpoints}). This form of the command uses breakpoints,
3536 and hence is quicker than @code{until} without an argument.
3539 @kindex si @r{(@code{stepi})}
3541 @itemx stepi @var{arg}
3543 Execute one machine instruction, then stop and return to the debugger.
3545 It is often useful to do @samp{display/i $pc} when stepping by machine
3546 instructions. This makes @value{GDBN} automatically display the next
3547 instruction to be executed, each time your program stops. @xref{Auto
3548 Display,, Automatic display}.
3550 An argument is a repeat count, as in @code{step}.
3554 @kindex ni @r{(@code{nexti})}
3556 @itemx nexti @var{arg}
3558 Execute one machine instruction, but if it is a function call,
3559 proceed until the function returns.
3561 An argument is a repeat count, as in @code{next}.
3568 A signal is an asynchronous event that can happen in a program. The
3569 operating system defines the possible kinds of signals, and gives each
3570 kind a name and a number. For example, in Unix @code{SIGINT} is the
3571 signal a program gets when you type an interrupt character (often @kbd{C-c});
3572 @code{SIGSEGV} is the signal a program gets from referencing a place in
3573 memory far away from all the areas in use; @code{SIGALRM} occurs when
3574 the alarm clock timer goes off (which happens only if your program has
3575 requested an alarm).
3577 @cindex fatal signals
3578 Some signals, including @code{SIGALRM}, are a normal part of the
3579 functioning of your program. Others, such as @code{SIGSEGV}, indicate
3580 errors; these signals are @dfn{fatal} (they kill your program immediately) if the
3581 program has not specified in advance some other way to handle the signal.
3582 @code{SIGINT} does not indicate an error in your program, but it is normally
3583 fatal so it can carry out the purpose of the interrupt: to kill the program.
3585 @value{GDBN} has the ability to detect any occurrence of a signal in your
3586 program. You can tell @value{GDBN} in advance what to do for each kind of
3589 @cindex handling signals
3590 Normally, @value{GDBN} is set up to let the non-erroneous signals like
3591 @code{SIGALRM} be silently passed to your program
3592 (so as not to interfere with their role in the program's functioning)
3593 but to stop your program immediately whenever an error signal happens.
3594 You can change these settings with the @code{handle} command.
3597 @kindex info signals
3600 Print a table of all the kinds of signals and how @value{GDBN} has been told to
3601 handle each one. You can use this to see the signal numbers of all
3602 the defined types of signals.
3604 @code{info handle} is an alias for @code{info signals}.
3607 @item handle @var{signal} @var{keywords}@dots{}
3608 Change the way @value{GDBN} handles signal @var{signal}. @var{signal}
3609 can be the number of a signal or its name (with or without the
3610 @samp{SIG} at the beginning); a list of signal numbers of the form
3611 @samp{@var{low}-@var{high}}; or the word @samp{all}, meaning all the
3612 known signals. The @var{keywords} say what change to make.
3616 The keywords allowed by the @code{handle} command can be abbreviated.
3617 Their full names are:
3621 @value{GDBN} should not stop your program when this signal happens. It may
3622 still print a message telling you that the signal has come in.
3625 @value{GDBN} should stop your program when this signal happens. This implies
3626 the @code{print} keyword as well.
3629 @value{GDBN} should print a message when this signal happens.
3632 @value{GDBN} should not mention the occurrence of the signal at all. This
3633 implies the @code{nostop} keyword as well.
3637 @value{GDBN} should allow your program to see this signal; your program
3638 can handle the signal, or else it may terminate if the signal is fatal
3639 and not handled. @code{pass} and @code{noignore} are synonyms.
3643 @value{GDBN} should not allow your program to see this signal.
3644 @code{nopass} and @code{ignore} are synonyms.
3648 When a signal stops your program, the signal is not visible to the
3650 continue. Your program sees the signal then, if @code{pass} is in
3651 effect for the signal in question @emph{at that time}. In other words,
3652 after @value{GDBN} reports a signal, you can use the @code{handle}
3653 command with @code{pass} or @code{nopass} to control whether your
3654 program sees that signal when you continue.
3656 The default is set to @code{nostop}, @code{noprint}, @code{pass} for
3657 non-erroneous signals such as @code{SIGALRM}, @code{SIGWINCH} and
3658 @code{SIGCHLD}, and to @code{stop}, @code{print}, @code{pass} for the
3661 You can also use the @code{signal} command to prevent your program from
3662 seeing a signal, or cause it to see a signal it normally would not see,
3663 or to give it any signal at any time. For example, if your program stopped
3664 due to some sort of memory reference error, you might store correct
3665 values into the erroneous variables and continue, hoping to see more
3666 execution; but your program would probably terminate immediately as
3667 a result of the fatal signal once it saw the signal. To prevent this,
3668 you can continue with @samp{signal 0}. @xref{Signaling, ,Giving your
3672 @section Stopping and starting multi-thread programs
3674 When your program has multiple threads (@pxref{Threads,, Debugging
3675 programs with multiple threads}), you can choose whether to set
3676 breakpoints on all threads, or on a particular thread.
3679 @cindex breakpoints and threads
3680 @cindex thread breakpoints
3681 @kindex break @dots{} thread @var{threadno}
3682 @item break @var{linespec} thread @var{threadno}
3683 @itemx break @var{linespec} thread @var{threadno} if @dots{}
3684 @var{linespec} specifies source lines; there are several ways of
3685 writing them, but the effect is always to specify some source line.
3687 Use the qualifier @samp{thread @var{threadno}} with a breakpoint command
3688 to specify that you only want @value{GDBN} to stop the program when a
3689 particular thread reaches this breakpoint. @var{threadno} is one of the
3690 numeric thread identifiers assigned by @value{GDBN}, shown in the first
3691 column of the @samp{info threads} display.
3693 If you do not specify @samp{thread @var{threadno}} when you set a
3694 breakpoint, the breakpoint applies to @emph{all} threads of your
3697 You can use the @code{thread} qualifier on conditional breakpoints as
3698 well; in this case, place @samp{thread @var{threadno}} before the
3699 breakpoint condition, like this:
3702 (@value{GDBP}) break frik.c:13 thread 28 if bartab > lim
3707 @cindex stopped threads
3708 @cindex threads, stopped
3709 Whenever your program stops under @value{GDBN} for any reason,
3710 @emph{all} threads of execution stop, not just the current thread. This
3711 allows you to examine the overall state of the program, including
3712 switching between threads, without worrying that things may change
3715 @cindex continuing threads
3716 @cindex threads, continuing
3717 Conversely, whenever you restart the program, @emph{all} threads start
3718 executing. @emph{This is true even when single-stepping} with commands
3719 like @code{step} or @code{next}.
3721 In particular, @value{GDBN} cannot single-step all threads in lockstep.
3722 Since thread scheduling is up to your debugging target's operating
3723 system (not controlled by @value{GDBN}), other threads may
3724 execute more than one statement while the current thread completes a
3725 single step. Moreover, in general other threads stop in the middle of a
3726 statement, rather than at a clean statement boundary, when the program
3729 You might even find your program stopped in another thread after
3730 continuing or even single-stepping. This happens whenever some other
3731 thread runs into a breakpoint, a signal, or an exception before the
3732 first thread completes whatever you requested.
3734 On some OSes, you can lock the OS scheduler and thus allow only a single
3738 @item set scheduler-locking @var{mode}
3739 Set the scheduler locking mode. If it is @code{off}, then there is no
3740 locking and any thread may run at any time. If @code{on}, then only the
3741 current thread may run when the inferior is resumed. The @code{step}
3742 mode optimizes for single-stepping. It stops other threads from
3743 ``seizing the prompt'' by preempting the current thread while you are
3744 stepping. Other threads will only rarely (or never) get a chance to run
3745 when you step. They are more likely to run when you @samp{next} over a
3746 function call, and they are completely free to run when you use commands
3747 like @samp{continue}, @samp{until}, or @samp{finish}. However, unless another
3748 thread hits a breakpoint during its timeslice, they will never steal the
3749 @value{GDBN} prompt away from the thread that you are debugging.
3751 @item show scheduler-locking
3752 Display the current scheduler locking mode.
3757 @chapter Examining the Stack
3759 When your program has stopped, the first thing you need to know is where it
3760 stopped and how it got there.
3763 Each time your program performs a function call, information about the call
3765 That information includes the location of the call in your program,
3766 the arguments of the call,
3767 and the local variables of the function being called.
3768 The information is saved in a block of data called a @dfn{stack frame}.
3769 The stack frames are allocated in a region of memory called the @dfn{call
3772 When your program stops, the @value{GDBN} commands for examining the
3773 stack allow you to see all of this information.
3775 @cindex selected frame
3776 One of the stack frames is @dfn{selected} by @value{GDBN} and many
3777 @value{GDBN} commands refer implicitly to the selected frame. In
3778 particular, whenever you ask @value{GDBN} for the value of a variable in
3779 your program, the value is found in the selected frame. There are
3780 special @value{GDBN} commands to select whichever frame you are
3781 interested in. @xref{Selection, ,Selecting a frame}.
3783 When your program stops, @value{GDBN} automatically selects the
3784 currently executing frame and describes it briefly, similar to the
3785 @code{frame} command (@pxref{Frame Info, ,Information about a frame}).
3788 * Frames:: Stack frames
3789 * Backtrace:: Backtraces
3790 * Selection:: Selecting a frame
3791 * Frame Info:: Information on a frame
3796 @section Stack frames
3798 @cindex frame, definition
3800 The call stack is divided up into contiguous pieces called @dfn{stack
3801 frames}, or @dfn{frames} for short; each frame is the data associated
3802 with one call to one function. The frame contains the arguments given
3803 to the function, the function's local variables, and the address at
3804 which the function is executing.
3806 @cindex initial frame
3807 @cindex outermost frame
3808 @cindex innermost frame
3809 When your program is started, the stack has only one frame, that of the
3810 function @code{main}. This is called the @dfn{initial} frame or the
3811 @dfn{outermost} frame. Each time a function is called, a new frame is
3812 made. Each time a function returns, the frame for that function invocation
3813 is eliminated. If a function is recursive, there can be many frames for
3814 the same function. The frame for the function in which execution is
3815 actually occurring is called the @dfn{innermost} frame. This is the most
3816 recently created of all the stack frames that still exist.
3818 @cindex frame pointer
3819 Inside your program, stack frames are identified by their addresses. A
3820 stack frame consists of many bytes, each of which has its own address; each
3821 kind of computer has a convention for choosing one byte whose
3822 address serves as the address of the frame. Usually this address is kept
3823 in a register called the @dfn{frame pointer register} while execution is
3824 going on in that frame.
3826 @cindex frame number
3827 @value{GDBN} assigns numbers to all existing stack frames, starting with
3828 zero for the innermost frame, one for the frame that called it,
3829 and so on upward. These numbers do not really exist in your program;
3830 they are assigned by @value{GDBN} to give you a way of designating stack
3831 frames in @value{GDBN} commands.
3833 @c The -fomit-frame-pointer below perennially causes hbox overflow
3834 @c underflow problems.
3835 @cindex frameless execution
3836 Some compilers provide a way to compile functions so that they operate
3837 without stack frames. (For example, the @value{GCC} option
3839 @samp{-fomit-frame-pointer}
3841 generates functions without a frame.)
3842 This is occasionally done with heavily used library functions to save
3843 the frame setup time. @value{GDBN} has limited facilities for dealing
3844 with these function invocations. If the innermost function invocation
3845 has no stack frame, @value{GDBN} nevertheless regards it as though
3846 it had a separate frame, which is numbered zero as usual, allowing
3847 correct tracing of the function call chain. However, @value{GDBN} has
3848 no provision for frameless functions elsewhere in the stack.
3851 @kindex frame@r{, command}
3852 @cindex current stack frame
3853 @item frame @var{args}
3854 The @code{frame} command allows you to move from one stack frame to another,
3855 and to print the stack frame you select. @var{args} may be either the
3856 address of the frame or the stack frame number. Without an argument,
3857 @code{frame} prints the current stack frame.
3859 @kindex select-frame
3860 @cindex selecting frame silently
3862 The @code{select-frame} command allows you to move from one stack frame
3863 to another without printing the frame. This is the silent version of
3872 @cindex stack traces
3873 A backtrace is a summary of how your program got where it is. It shows one
3874 line per frame, for many frames, starting with the currently executing
3875 frame (frame zero), followed by its caller (frame one), and on up the
3880 @kindex bt @r{(@code{backtrace})}
3883 Print a backtrace of the entire stack: one line per frame for all
3884 frames in the stack.
3886 You can stop the backtrace at any time by typing the system interrupt
3887 character, normally @kbd{C-c}.
3889 @item backtrace @var{n}
3891 Similar, but print only the innermost @var{n} frames.
3893 @item backtrace -@var{n}
3895 Similar, but print only the outermost @var{n} frames.
3900 @kindex info s @r{(@code{info stack})}
3901 The names @code{where} and @code{info stack} (abbreviated @code{info s})
3902 are additional aliases for @code{backtrace}.
3904 Each line in the backtrace shows the frame number and the function name.
3905 The program counter value is also shown---unless you use @code{set
3906 print address off}. The backtrace also shows the source file name and
3907 line number, as well as the arguments to the function. The program
3908 counter value is omitted if it is at the beginning of the code for that
3911 Here is an example of a backtrace. It was made with the command
3912 @samp{bt 3}, so it shows the innermost three frames.
3916 #0 m4_traceon (obs=0x24eb0, argc=1, argv=0x2b8c8)
3918 #1 0x6e38 in expand_macro (sym=0x2b600) at macro.c:242
3919 #2 0x6840 in expand_token (obs=0x0, t=177664, td=0xf7fffb08)
3921 (More stack frames follow...)
3926 The display for frame zero does not begin with a program counter
3927 value, indicating that your program has stopped at the beginning of the
3928 code for line @code{993} of @code{builtin.c}.
3931 @section Selecting a frame
3933 Most commands for examining the stack and other data in your program work on
3934 whichever stack frame is selected at the moment. Here are the commands for
3935 selecting a stack frame; all of them finish by printing a brief description
3936 of the stack frame just selected.
3939 @kindex frame@r{, selecting}
3940 @kindex f @r{(@code{frame})}
3943 Select frame number @var{n}. Recall that frame zero is the innermost
3944 (currently executing) frame, frame one is the frame that called the
3945 innermost one, and so on. The highest-numbered frame is the one for
3948 @item frame @var{addr}
3950 Select the frame at address @var{addr}. This is useful mainly if the
3951 chaining of stack frames has been damaged by a bug, making it
3952 impossible for @value{GDBN} to assign numbers properly to all frames. In
3953 addition, this can be useful when your program has multiple stacks and
3954 switches between them.
3956 On the SPARC architecture, @code{frame} needs two addresses to
3957 select an arbitrary frame: a frame pointer and a stack pointer.
3959 On the MIPS and Alpha architecture, it needs two addresses: a stack
3960 pointer and a program counter.
3962 On the 29k architecture, it needs three addresses: a register stack
3963 pointer, a program counter, and a memory stack pointer.
3964 @c note to future updaters: this is conditioned on a flag
3965 @c SETUP_ARBITRARY_FRAME in the tm-*.h files. The above is up to date
3966 @c as of 27 Jan 1994.
3970 Move @var{n} frames up the stack. For positive numbers @var{n}, this
3971 advances toward the outermost frame, to higher frame numbers, to frames
3972 that have existed longer. @var{n} defaults to one.
3975 @kindex do @r{(@code{down})}
3977 Move @var{n} frames down the stack. For positive numbers @var{n}, this
3978 advances toward the innermost frame, to lower frame numbers, to frames
3979 that were created more recently. @var{n} defaults to one. You may
3980 abbreviate @code{down} as @code{do}.
3983 All of these commands end by printing two lines of output describing the
3984 frame. The first line shows the frame number, the function name, the
3985 arguments, and the source file and line number of execution in that
3986 frame. The second line shows the text of that source line.
3994 #1 0x22f0 in main (argc=1, argv=0xf7fffbf4, env=0xf7fffbfc)
3996 10 read_input_file (argv[i]);
4000 After such a printout, the @code{list} command with no arguments
4001 prints ten lines centered on the point of execution in the frame.
4002 @xref{List, ,Printing source lines}.
4005 @kindex down-silently
4007 @item up-silently @var{n}
4008 @itemx down-silently @var{n}
4009 These two commands are variants of @code{up} and @code{down},
4010 respectively; they differ in that they do their work silently, without
4011 causing display of the new frame. They are intended primarily for use
4012 in @value{GDBN} command scripts, where the output might be unnecessary and
4017 @section Information about a frame
4019 There are several other commands to print information about the selected
4025 When used without any argument, this command does not change which
4026 frame is selected, but prints a brief description of the currently
4027 selected stack frame. It can be abbreviated @code{f}. With an
4028 argument, this command is used to select a stack frame.
4029 @xref{Selection, ,Selecting a frame}.
4032 @kindex info f @r{(@code{info frame})}
4035 This command prints a verbose description of the selected stack frame,
4040 the address of the frame
4042 the address of the next frame down (called by this frame)
4044 the address of the next frame up (caller of this frame)
4046 the language in which the source code corresponding to this frame is written
4048 the address of the frame's arguments
4050 the address of the frame's local variables
4052 the program counter saved in it (the address of execution in the caller frame)
4054 which registers were saved in the frame
4057 @noindent The verbose description is useful when
4058 something has gone wrong that has made the stack format fail to fit
4059 the usual conventions.
4061 @item info frame @var{addr}
4062 @itemx info f @var{addr}
4063 Print a verbose description of the frame at address @var{addr}, without
4064 selecting that frame. The selected frame remains unchanged by this
4065 command. This requires the same kind of address (more than one for some
4066 architectures) that you specify in the @code{frame} command.
4067 @xref{Selection, ,Selecting a frame}.
4071 Print the arguments of the selected frame, each on a separate line.
4075 Print the local variables of the selected frame, each on a separate
4076 line. These are all variables (declared either static or automatic)
4077 accessible at the point of execution of the selected frame.
4080 @cindex catch exceptions, list active handlers
4081 @cindex exception handlers, how to list
4083 Print a list of all the exception handlers that are active in the
4084 current stack frame at the current point of execution. To see other
4085 exception handlers, visit the associated frame (using the @code{up},
4086 @code{down}, or @code{frame} commands); then type @code{info catch}.
4087 @xref{Set Catchpoints, , Setting catchpoints}.
4093 @chapter Examining Source Files
4095 @value{GDBN} can print parts of your program's source, since the debugging
4096 information recorded in the program tells @value{GDBN} what source files were
4097 used to build it. When your program stops, @value{GDBN} spontaneously prints
4098 the line where it stopped. Likewise, when you select a stack frame
4099 (@pxref{Selection, ,Selecting a frame}), @value{GDBN} prints the line where
4100 execution in that frame has stopped. You can print other portions of
4101 source files by explicit command.
4103 If you use @value{GDBN} through its @sc{gnu} Emacs interface, you may
4104 prefer to use Emacs facilities to view source; see @ref{Emacs, ,Using
4105 @value{GDBN} under @sc{gnu} Emacs}.
4108 * List:: Printing source lines
4109 * Search:: Searching source files
4110 * Source Path:: Specifying source directories
4111 * Machine Code:: Source and machine code
4115 @section Printing source lines
4118 @kindex l @r{(@code{list})}
4119 To print lines from a source file, use the @code{list} command
4120 (abbreviated @code{l}). By default, ten lines are printed.
4121 There are several ways to specify what part of the file you want to print.
4123 Here are the forms of the @code{list} command most commonly used:
4126 @item list @var{linenum}
4127 Print lines centered around line number @var{linenum} in the
4128 current source file.
4130 @item list @var{function}
4131 Print lines centered around the beginning of function
4135 Print more lines. If the last lines printed were printed with a
4136 @code{list} command, this prints lines following the last lines
4137 printed; however, if the last line printed was a solitary line printed
4138 as part of displaying a stack frame (@pxref{Stack, ,Examining the
4139 Stack}), this prints lines centered around that line.
4142 Print lines just before the lines last printed.
4145 By default, @value{GDBN} prints ten source lines with any of these forms of
4146 the @code{list} command. You can change this using @code{set listsize}:
4149 @kindex set listsize
4150 @item set listsize @var{count}
4151 Make the @code{list} command display @var{count} source lines (unless
4152 the @code{list} argument explicitly specifies some other number).
4154 @kindex show listsize
4156 Display the number of lines that @code{list} prints.
4159 Repeating a @code{list} command with @key{RET} discards the argument,
4160 so it is equivalent to typing just @code{list}. This is more useful
4161 than listing the same lines again. An exception is made for an
4162 argument of @samp{-}; that argument is preserved in repetition so that
4163 each repetition moves up in the source file.
4166 In general, the @code{list} command expects you to supply zero, one or two
4167 @dfn{linespecs}. Linespecs specify source lines; there are several ways
4168 of writing them, but the effect is always to specify some source line.
4169 Here is a complete description of the possible arguments for @code{list}:
4172 @item list @var{linespec}
4173 Print lines centered around the line specified by @var{linespec}.
4175 @item list @var{first},@var{last}
4176 Print lines from @var{first} to @var{last}. Both arguments are
4179 @item list ,@var{last}
4180 Print lines ending with @var{last}.
4182 @item list @var{first},
4183 Print lines starting with @var{first}.
4186 Print lines just after the lines last printed.
4189 Print lines just before the lines last printed.
4192 As described in the preceding table.
4195 Here are the ways of specifying a single source line---all the
4200 Specifies line @var{number} of the current source file.
4201 When a @code{list} command has two linespecs, this refers to
4202 the same source file as the first linespec.
4205 Specifies the line @var{offset} lines after the last line printed.
4206 When used as the second linespec in a @code{list} command that has
4207 two, this specifies the line @var{offset} lines down from the
4211 Specifies the line @var{offset} lines before the last line printed.
4213 @item @var{filename}:@var{number}
4214 Specifies line @var{number} in the source file @var{filename}.
4216 @item @var{function}
4217 Specifies the line that begins the body of the function @var{function}.
4218 For example: in C, this is the line with the open brace.
4220 @item @var{filename}:@var{function}
4221 Specifies the line of the open-brace that begins the body of the
4222 function @var{function} in the file @var{filename}. You only need the
4223 file name with a function name to avoid ambiguity when there are
4224 identically named functions in different source files.
4226 @item *@var{address}
4227 Specifies the line containing the program address @var{address}.
4228 @var{address} may be any expression.
4232 @section Searching source files
4234 @kindex reverse-search
4236 There are two commands for searching through the current source file for a
4241 @kindex forward-search
4242 @item forward-search @var{regexp}
4243 @itemx search @var{regexp}
4244 The command @samp{forward-search @var{regexp}} checks each line,
4245 starting with the one following the last line listed, for a match for
4246 @var{regexp}. It lists the line that is found. You can use the
4247 synonym @samp{search @var{regexp}} or abbreviate the command name as
4250 @item reverse-search @var{regexp}
4251 The command @samp{reverse-search @var{regexp}} checks each line, starting
4252 with the one before the last line listed and going backward, for a match
4253 for @var{regexp}. It lists the line that is found. You can abbreviate
4254 this command as @code{rev}.
4258 @section Specifying source directories
4261 @cindex directories for source files
4262 Executable programs sometimes do not record the directories of the source
4263 files from which they were compiled, just the names. Even when they do,
4264 the directories could be moved between the compilation and your debugging
4265 session. @value{GDBN} has a list of directories to search for source files;
4266 this is called the @dfn{source path}. Each time @value{GDBN} wants a source file,
4267 it tries all the directories in the list, in the order they are present
4268 in the list, until it finds a file with the desired name. Note that
4269 the executable search path is @emph{not} used for this purpose. Neither is
4270 the current working directory, unless it happens to be in the source
4273 If @value{GDBN} cannot find a source file in the source path, and the
4274 object program records a directory, @value{GDBN} tries that directory
4275 too. If the source path is empty, and there is no record of the
4276 compilation directory, @value{GDBN} looks in the current directory as a
4279 Whenever you reset or rearrange the source path, @value{GDBN} clears out
4280 any information it has cached about where source files are found and where
4281 each line is in the file.
4285 When you start @value{GDBN}, its source path includes only @samp{cdir}
4286 and @samp{cwd}, in that order.
4287 To add other directories, use the @code{directory} command.
4290 @item directory @var{dirname} @dots{}
4291 @item dir @var{dirname} @dots{}
4292 Add directory @var{dirname} to the front of the source path. Several
4293 directory names may be given to this command, separated by @samp{:}
4294 (@samp{;} on MS-DOS and MS-Windows, where @samp{:} usually appears as
4295 part of absolute file names) or
4296 whitespace. You may specify a directory that is already in the source
4297 path; this moves it forward, so @value{GDBN} searches it sooner.
4301 @vindex $cdir@r{, convenience variable}
4302 @vindex $cwdr@r{, convenience variable}
4303 @cindex compilation directory
4304 @cindex current directory
4305 @cindex working directory
4306 @cindex directory, current
4307 @cindex directory, compilation
4308 You can use the string @samp{$cdir} to refer to the compilation
4309 directory (if one is recorded), and @samp{$cwd} to refer to the current
4310 working directory. @samp{$cwd} is not the same as @samp{.}---the former
4311 tracks the current working directory as it changes during your @value{GDBN}
4312 session, while the latter is immediately expanded to the current
4313 directory at the time you add an entry to the source path.
4316 Reset the source path to empty again. This requires confirmation.
4318 @c RET-repeat for @code{directory} is explicitly disabled, but since
4319 @c repeating it would be a no-op we do not say that. (thanks to RMS)
4321 @item show directories
4322 @kindex show directories
4323 Print the source path: show which directories it contains.
4326 If your source path is cluttered with directories that are no longer of
4327 interest, @value{GDBN} may sometimes cause confusion by finding the wrong
4328 versions of source. You can correct the situation as follows:
4332 Use @code{directory} with no argument to reset the source path to empty.
4335 Use @code{directory} with suitable arguments to reinstall the
4336 directories you want in the source path. You can add all the
4337 directories in one command.
4341 @section Source and machine code
4343 You can use the command @code{info line} to map source lines to program
4344 addresses (and vice versa), and the command @code{disassemble} to display
4345 a range of addresses as machine instructions. When run under @sc{gnu} Emacs
4346 mode, the @code{info line} command causes the arrow to point to the
4347 line specified. Also, @code{info line} prints addresses in symbolic form as
4352 @item info line @var{linespec}
4353 Print the starting and ending addresses of the compiled code for
4354 source line @var{linespec}. You can specify source lines in any of
4355 the ways understood by the @code{list} command (@pxref{List, ,Printing
4359 For example, we can use @code{info line} to discover the location of
4360 the object code for the first line of function
4361 @code{m4_changequote}:
4363 @c FIXME: I think this example should also show the addresses in
4364 @c symbolic form, as they usually would be displayed.
4366 (@value{GDBP}) info line m4_changequote
4367 Line 895 of "builtin.c" starts at pc 0x634c and ends at 0x6350.
4371 We can also inquire (using @code{*@var{addr}} as the form for
4372 @var{linespec}) what source line covers a particular address:
4374 (@value{GDBP}) info line *0x63ff
4375 Line 926 of "builtin.c" starts at pc 0x63e4 and ends at 0x6404.
4378 @cindex @code{$_} and @code{info line}
4379 @kindex x@r{(examine), and} info line
4380 After @code{info line}, the default address for the @code{x} command
4381 is changed to the starting address of the line, so that @samp{x/i} is
4382 sufficient to begin examining the machine code (@pxref{Memory,
4383 ,Examining memory}). Also, this address is saved as the value of the
4384 convenience variable @code{$_} (@pxref{Convenience Vars, ,Convenience
4389 @cindex assembly instructions
4390 @cindex instructions, assembly
4391 @cindex machine instructions
4392 @cindex listing machine instructions
4394 This specialized command dumps a range of memory as machine
4395 instructions. The default memory range is the function surrounding the
4396 program counter of the selected frame. A single argument to this
4397 command is a program counter value; @value{GDBN} dumps the function
4398 surrounding this value. Two arguments specify a range of addresses
4399 (first inclusive, second exclusive) to dump.
4402 The following example shows the disassembly of a range of addresses of
4403 HP PA-RISC 2.0 code:
4406 (@value{GDBP}) disas 0x32c4 0x32e4
4407 Dump of assembler code from 0x32c4 to 0x32e4:
4408 0x32c4 <main+204>: addil 0,dp
4409 0x32c8 <main+208>: ldw 0x22c(sr0,r1),r26
4410 0x32cc <main+212>: ldil 0x3000,r31
4411 0x32d0 <main+216>: ble 0x3f8(sr4,r31)
4412 0x32d4 <main+220>: ldo 0(r31),rp
4413 0x32d8 <main+224>: addil -0x800,dp
4414 0x32dc <main+228>: ldo 0x588(r1),r26
4415 0x32e0 <main+232>: ldil 0x3000,r31
4416 End of assembler dump.
4419 Some architectures have more than one commonly-used set of instruction
4420 mnemonics or other syntax.
4423 @kindex set disassembly-flavor
4424 @cindex assembly instructions
4425 @cindex instructions, assembly
4426 @cindex machine instructions
4427 @cindex listing machine instructions
4428 @cindex Intel disassembly flavor
4429 @cindex AT&T disassembly flavor
4430 @item set disassembly-flavor @var{instruction-set}
4431 Select the instruction set to use when disassembling the
4432 program via the @code{disassemble} or @code{x/i} commands.
4434 Currently this command is only defined for the Intel x86 family. You
4435 can set @var{instruction-set} to either @code{intel} or @code{att}.
4436 The default is @code{att}, the AT&T flavor used by default by Unix
4437 assemblers for x86-based targets.
4442 @chapter Examining Data
4444 @cindex printing data
4445 @cindex examining data
4448 @c "inspect" is not quite a synonym if you are using Epoch, which we do not
4449 @c document because it is nonstandard... Under Epoch it displays in a
4450 @c different window or something like that.
4451 The usual way to examine data in your program is with the @code{print}
4452 command (abbreviated @code{p}), or its synonym @code{inspect}. It
4453 evaluates and prints the value of an expression of the language your
4454 program is written in (@pxref{Languages, ,Using @value{GDBN} with
4455 Different Languages}).
4458 @item print @var{expr}
4459 @itemx print /@var{f} @var{expr}
4460 @var{expr} is an expression (in the source language). By default the
4461 value of @var{expr} is printed in a format appropriate to its data type;
4462 you can choose a different format by specifying @samp{/@var{f}}, where
4463 @var{f} is a letter specifying the format; see @ref{Output Formats,,Output
4467 @itemx print /@var{f}
4468 If you omit @var{expr}, @value{GDBN} displays the last value again (from the
4469 @dfn{value history}; @pxref{Value History, ,Value history}). This allows you to
4470 conveniently inspect the same value in an alternative format.
4473 A more low-level way of examining data is with the @code{x} command.
4474 It examines data in memory at a specified address and prints it in a
4475 specified format. @xref{Memory, ,Examining memory}.
4477 If you are interested in information about types, or about how the
4478 fields of a struct or a class are declared, use the @code{ptype @var{exp}}
4479 command rather than @code{print}. @xref{Symbols, ,Examining the Symbol
4483 * Expressions:: Expressions
4484 * Variables:: Program variables
4485 * Arrays:: Artificial arrays
4486 * Output Formats:: Output formats
4487 * Memory:: Examining memory
4488 * Auto Display:: Automatic display
4489 * Print Settings:: Print settings
4490 * Value History:: Value history
4491 * Convenience Vars:: Convenience variables
4492 * Registers:: Registers
4493 * Floating Point Hardware:: Floating point hardware
4494 * Memory Region Attributes:: Memory region attributes
4498 @section Expressions
4501 @code{print} and many other @value{GDBN} commands accept an expression and
4502 compute its value. Any kind of constant, variable or operator defined
4503 by the programming language you are using is valid in an expression in
4504 @value{GDBN}. This includes conditional expressions, function calls, casts
4505 and string constants. It unfortunately does not include symbols defined
4506 by preprocessor @code{#define} commands.
4508 @value{GDBN} supports array constants in expressions input by
4509 the user. The syntax is @{@var{element}, @var{element}@dots{}@}. For example,
4510 you can use the command @code{print @{1, 2, 3@}} to build up an array in
4511 memory that is @code{malloc}ed in the target program.
4513 Because C is so widespread, most of the expressions shown in examples in
4514 this manual are in C. @xref{Languages, , Using @value{GDBN} with Different
4515 Languages}, for information on how to use expressions in other
4518 In this section, we discuss operators that you can use in @value{GDBN}
4519 expressions regardless of your programming language.
4521 Casts are supported in all languages, not just in C, because it is so
4522 useful to cast a number into a pointer in order to examine a structure
4523 at that address in memory.
4524 @c FIXME: casts supported---Mod2 true?
4526 @value{GDBN} supports these operators, in addition to those common
4527 to programming languages:
4531 @samp{@@} is a binary operator for treating parts of memory as arrays.
4532 @xref{Arrays, ,Artificial arrays}, for more information.
4535 @samp{::} allows you to specify a variable in terms of the file or
4536 function where it is defined. @xref{Variables, ,Program variables}.
4538 @cindex @{@var{type}@}
4539 @cindex type casting memory
4540 @cindex memory, viewing as typed object
4541 @cindex casts, to view memory
4542 @item @{@var{type}@} @var{addr}
4543 Refers to an object of type @var{type} stored at address @var{addr} in
4544 memory. @var{addr} may be any expression whose value is an integer or
4545 pointer (but parentheses are required around binary operators, just as in
4546 a cast). This construct is allowed regardless of what kind of data is
4547 normally supposed to reside at @var{addr}.
4551 @section Program variables
4553 The most common kind of expression to use is the name of a variable
4556 Variables in expressions are understood in the selected stack frame
4557 (@pxref{Selection, ,Selecting a frame}); they must be either:
4561 global (or file-static)
4568 visible according to the scope rules of the
4569 programming language from the point of execution in that frame
4572 @noindent This means that in the function
4587 you can examine and use the variable @code{a} whenever your program is
4588 executing within the function @code{foo}, but you can only use or
4589 examine the variable @code{b} while your program is executing inside
4590 the block where @code{b} is declared.
4592 @cindex variable name conflict
4593 There is an exception: you can refer to a variable or function whose
4594 scope is a single source file even if the current execution point is not
4595 in this file. But it is possible to have more than one such variable or
4596 function with the same name (in different source files). If that
4597 happens, referring to that name has unpredictable effects. If you wish,
4598 you can specify a static variable in a particular function or file,
4599 using the colon-colon notation:
4601 @cindex colon-colon, context for variables/functions
4603 @c info cannot cope with a :: index entry, but why deprive hard copy readers?
4604 @cindex @code{::}, context for variables/functions
4607 @var{file}::@var{variable}
4608 @var{function}::@var{variable}
4612 Here @var{file} or @var{function} is the name of the context for the
4613 static @var{variable}. In the case of file names, you can use quotes to
4614 make sure @value{GDBN} parses the file name as a single word---for example,
4615 to print a global value of @code{x} defined in @file{f2.c}:
4618 (@value{GDBP}) p 'f2.c'::x
4621 @cindex C@t{++} scope resolution
4622 This use of @samp{::} is very rarely in conflict with the very similar
4623 use of the same notation in C@t{++}. @value{GDBN} also supports use of the C@t{++}
4624 scope resolution operator in @value{GDBN} expressions.
4625 @c FIXME: Um, so what happens in one of those rare cases where it's in
4628 @cindex wrong values
4629 @cindex variable values, wrong
4631 @emph{Warning:} Occasionally, a local variable may appear to have the
4632 wrong value at certain points in a function---just after entry to a new
4633 scope, and just before exit.
4635 You may see this problem when you are stepping by machine instructions.
4636 This is because, on most machines, it takes more than one instruction to
4637 set up a stack frame (including local variable definitions); if you are
4638 stepping by machine instructions, variables may appear to have the wrong
4639 values until the stack frame is completely built. On exit, it usually
4640 also takes more than one machine instruction to destroy a stack frame;
4641 after you begin stepping through that group of instructions, local
4642 variable definitions may be gone.
4644 This may also happen when the compiler does significant optimizations.
4645 To be sure of always seeing accurate values, turn off all optimization
4648 @cindex ``No symbol "foo" in current context''
4649 Another possible effect of compiler optimizations is to optimize
4650 unused variables out of existence, or assign variables to registers (as
4651 opposed to memory addresses). Depending on the support for such cases
4652 offered by the debug info format used by the compiler, @value{GDBN}
4653 might not be able to display values for such local variables. If that
4654 happens, @value{GDBN} will print a message like this:
4657 No symbol "foo" in current context.
4660 To solve such problems, either recompile without optimizations, or use a
4661 different debug info format, if the compiler supports several such
4662 formats. For example, @value{NGCC}, the @sc{gnu} C/C@t{++} compiler usually
4663 supports the @samp{-gstabs} option. @samp{-gstabs} produces debug info
4664 in a format that is superior to formats such as COFF. You may be able
4665 to use DWARF2 (@samp{-gdwarf-2}), which is also an effective form for
4666 debug info. See @ref{Debugging Options,,Options for Debugging Your
4667 Program or @sc{gnu} CC, gcc.info, Using @sc{gnu} CC}, for more
4672 @section Artificial arrays
4674 @cindex artificial array
4675 @kindex @@@r{, referencing memory as an array}
4676 It is often useful to print out several successive objects of the
4677 same type in memory; a section of an array, or an array of
4678 dynamically determined size for which only a pointer exists in the
4681 You can do this by referring to a contiguous span of memory as an
4682 @dfn{artificial array}, using the binary operator @samp{@@}. The left
4683 operand of @samp{@@} should be the first element of the desired array
4684 and be an individual object. The right operand should be the desired length
4685 of the array. The result is an array value whose elements are all of
4686 the type of the left argument. The first element is actually the left
4687 argument; the second element comes from bytes of memory immediately
4688 following those that hold the first element, and so on. Here is an
4689 example. If a program says
4692 int *array = (int *) malloc (len * sizeof (int));
4696 you can print the contents of @code{array} with
4702 The left operand of @samp{@@} must reside in memory. Array values made
4703 with @samp{@@} in this way behave just like other arrays in terms of
4704 subscripting, and are coerced to pointers when used in expressions.
4705 Artificial arrays most often appear in expressions via the value history
4706 (@pxref{Value History, ,Value history}), after printing one out.
4708 Another way to create an artificial array is to use a cast.
4709 This re-interprets a value as if it were an array.
4710 The value need not be in memory:
4712 (@value{GDBP}) p/x (short[2])0x12345678
4713 $1 = @{0x1234, 0x5678@}
4716 As a convenience, if you leave the array length out (as in
4717 @samp{(@var{type}[])@var{value}}) @value{GDBN} calculates the size to fill
4718 the value (as @samp{sizeof(@var{value})/sizeof(@var{type})}:
4720 (@value{GDBP}) p/x (short[])0x12345678
4721 $2 = @{0x1234, 0x5678@}
4724 Sometimes the artificial array mechanism is not quite enough; in
4725 moderately complex data structures, the elements of interest may not
4726 actually be adjacent---for example, if you are interested in the values
4727 of pointers in an array. One useful work-around in this situation is
4728 to use a convenience variable (@pxref{Convenience Vars, ,Convenience
4729 variables}) as a counter in an expression that prints the first
4730 interesting value, and then repeat that expression via @key{RET}. For
4731 instance, suppose you have an array @code{dtab} of pointers to
4732 structures, and you are interested in the values of a field @code{fv}
4733 in each structure. Here is an example of what you might type:
4743 @node Output Formats
4744 @section Output formats
4746 @cindex formatted output
4747 @cindex output formats
4748 By default, @value{GDBN} prints a value according to its data type. Sometimes
4749 this is not what you want. For example, you might want to print a number
4750 in hex, or a pointer in decimal. Or you might want to view data in memory
4751 at a certain address as a character string or as an instruction. To do
4752 these things, specify an @dfn{output format} when you print a value.
4754 The simplest use of output formats is to say how to print a value
4755 already computed. This is done by starting the arguments of the
4756 @code{print} command with a slash and a format letter. The format
4757 letters supported are:
4761 Regard the bits of the value as an integer, and print the integer in
4765 Print as integer in signed decimal.
4768 Print as integer in unsigned decimal.
4771 Print as integer in octal.
4774 Print as integer in binary. The letter @samp{t} stands for ``two''.
4775 @footnote{@samp{b} cannot be used because these format letters are also
4776 used with the @code{x} command, where @samp{b} stands for ``byte'';
4777 see @ref{Memory,,Examining memory}.}
4780 @cindex unknown address, locating
4781 @cindex locate address
4782 Print as an address, both absolute in hexadecimal and as an offset from
4783 the nearest preceding symbol. You can use this format used to discover
4784 where (in what function) an unknown address is located:
4787 (@value{GDBP}) p/a 0x54320
4788 $3 = 0x54320 <_initialize_vx+396>
4792 The command @code{info symbol 0x54320} yields similar results.
4793 @xref{Symbols, info symbol}.
4796 Regard as an integer and print it as a character constant.
4799 Regard the bits of the value as a floating point number and print
4800 using typical floating point syntax.
4803 For example, to print the program counter in hex (@pxref{Registers}), type
4810 Note that no space is required before the slash; this is because command
4811 names in @value{GDBN} cannot contain a slash.
4813 To reprint the last value in the value history with a different format,
4814 you can use the @code{print} command with just a format and no
4815 expression. For example, @samp{p/x} reprints the last value in hex.
4818 @section Examining memory
4820 You can use the command @code{x} (for ``examine'') to examine memory in
4821 any of several formats, independently of your program's data types.
4823 @cindex examining memory
4825 @kindex x @r{(examine memory)}
4826 @item x/@var{nfu} @var{addr}
4829 Use the @code{x} command to examine memory.
4832 @var{n}, @var{f}, and @var{u} are all optional parameters that specify how
4833 much memory to display and how to format it; @var{addr} is an
4834 expression giving the address where you want to start displaying memory.
4835 If you use defaults for @var{nfu}, you need not type the slash @samp{/}.
4836 Several commands set convenient defaults for @var{addr}.
4839 @item @var{n}, the repeat count
4840 The repeat count is a decimal integer; the default is 1. It specifies
4841 how much memory (counting by units @var{u}) to display.
4842 @c This really is **decimal**; unaffected by 'set radix' as of GDB
4845 @item @var{f}, the display format
4846 The display format is one of the formats used by @code{print},
4847 @samp{s} (null-terminated string), or @samp{i} (machine instruction).
4848 The default is @samp{x} (hexadecimal) initially.
4849 The default changes each time you use either @code{x} or @code{print}.
4851 @item @var{u}, the unit size
4852 The unit size is any of
4858 Halfwords (two bytes).
4860 Words (four bytes). This is the initial default.
4862 Giant words (eight bytes).
4865 Each time you specify a unit size with @code{x}, that size becomes the
4866 default unit the next time you use @code{x}. (For the @samp{s} and
4867 @samp{i} formats, the unit size is ignored and is normally not written.)
4869 @item @var{addr}, starting display address
4870 @var{addr} is the address where you want @value{GDBN} to begin displaying
4871 memory. The expression need not have a pointer value (though it may);
4872 it is always interpreted as an integer address of a byte of memory.
4873 @xref{Expressions, ,Expressions}, for more information on expressions. The default for
4874 @var{addr} is usually just after the last address examined---but several
4875 other commands also set the default address: @code{info breakpoints} (to
4876 the address of the last breakpoint listed), @code{info line} (to the
4877 starting address of a line), and @code{print} (if you use it to display
4878 a value from memory).
4881 For example, @samp{x/3uh 0x54320} is a request to display three halfwords
4882 (@code{h}) of memory, formatted as unsigned decimal integers (@samp{u}),
4883 starting at address @code{0x54320}. @samp{x/4xw $sp} prints the four
4884 words (@samp{w}) of memory above the stack pointer (here, @samp{$sp};
4885 @pxref{Registers, ,Registers}) in hexadecimal (@samp{x}).
4887 Since the letters indicating unit sizes are all distinct from the
4888 letters specifying output formats, you do not have to remember whether
4889 unit size or format comes first; either order works. The output
4890 specifications @samp{4xw} and @samp{4wx} mean exactly the same thing.
4891 (However, the count @var{n} must come first; @samp{wx4} does not work.)
4893 Even though the unit size @var{u} is ignored for the formats @samp{s}
4894 and @samp{i}, you might still want to use a count @var{n}; for example,
4895 @samp{3i} specifies that you want to see three machine instructions,
4896 including any operands. The command @code{disassemble} gives an
4897 alternative way of inspecting machine instructions; see @ref{Machine
4898 Code,,Source and machine code}.
4900 All the defaults for the arguments to @code{x} are designed to make it
4901 easy to continue scanning memory with minimal specifications each time
4902 you use @code{x}. For example, after you have inspected three machine
4903 instructions with @samp{x/3i @var{addr}}, you can inspect the next seven
4904 with just @samp{x/7}. If you use @key{RET} to repeat the @code{x} command,
4905 the repeat count @var{n} is used again; the other arguments default as
4906 for successive uses of @code{x}.
4908 @cindex @code{$_}, @code{$__}, and value history
4909 The addresses and contents printed by the @code{x} command are not saved
4910 in the value history because there is often too much of them and they
4911 would get in the way. Instead, @value{GDBN} makes these values available for
4912 subsequent use in expressions as values of the convenience variables
4913 @code{$_} and @code{$__}. After an @code{x} command, the last address
4914 examined is available for use in expressions in the convenience variable
4915 @code{$_}. The contents of that address, as examined, are available in
4916 the convenience variable @code{$__}.
4918 If the @code{x} command has a repeat count, the address and contents saved
4919 are from the last memory unit printed; this is not the same as the last
4920 address printed if several units were printed on the last line of output.
4923 @section Automatic display
4924 @cindex automatic display
4925 @cindex display of expressions
4927 If you find that you want to print the value of an expression frequently
4928 (to see how it changes), you might want to add it to the @dfn{automatic
4929 display list} so that @value{GDBN} prints its value each time your program stops.
4930 Each expression added to the list is given a number to identify it;
4931 to remove an expression from the list, you specify that number.
4932 The automatic display looks like this:
4936 3: bar[5] = (struct hack *) 0x3804
4940 This display shows item numbers, expressions and their current values. As with
4941 displays you request manually using @code{x} or @code{print}, you can
4942 specify the output format you prefer; in fact, @code{display} decides
4943 whether to use @code{print} or @code{x} depending on how elaborate your
4944 format specification is---it uses @code{x} if you specify a unit size,
4945 or one of the two formats (@samp{i} and @samp{s}) that are only
4946 supported by @code{x}; otherwise it uses @code{print}.
4950 @item display @var{expr}
4951 Add the expression @var{expr} to the list of expressions to display
4952 each time your program stops. @xref{Expressions, ,Expressions}.
4954 @code{display} does not repeat if you press @key{RET} again after using it.
4956 @item display/@var{fmt} @var{expr}
4957 For @var{fmt} specifying only a display format and not a size or
4958 count, add the expression @var{expr} to the auto-display list but
4959 arrange to display it each time in the specified format @var{fmt}.
4960 @xref{Output Formats,,Output formats}.
4962 @item display/@var{fmt} @var{addr}
4963 For @var{fmt} @samp{i} or @samp{s}, or including a unit-size or a
4964 number of units, add the expression @var{addr} as a memory address to
4965 be examined each time your program stops. Examining means in effect
4966 doing @samp{x/@var{fmt} @var{addr}}. @xref{Memory, ,Examining memory}.
4969 For example, @samp{display/i $pc} can be helpful, to see the machine
4970 instruction about to be executed each time execution stops (@samp{$pc}
4971 is a common name for the program counter; @pxref{Registers, ,Registers}).
4974 @kindex delete display
4976 @item undisplay @var{dnums}@dots{}
4977 @itemx delete display @var{dnums}@dots{}
4978 Remove item numbers @var{dnums} from the list of expressions to display.
4980 @code{undisplay} does not repeat if you press @key{RET} after using it.
4981 (Otherwise you would just get the error @samp{No display number @dots{}}.)
4983 @kindex disable display
4984 @item disable display @var{dnums}@dots{}
4985 Disable the display of item numbers @var{dnums}. A disabled display
4986 item is not printed automatically, but is not forgotten. It may be
4987 enabled again later.
4989 @kindex enable display
4990 @item enable display @var{dnums}@dots{}
4991 Enable display of item numbers @var{dnums}. It becomes effective once
4992 again in auto display of its expression, until you specify otherwise.
4995 Display the current values of the expressions on the list, just as is
4996 done when your program stops.
4998 @kindex info display
5000 Print the list of expressions previously set up to display
5001 automatically, each one with its item number, but without showing the
5002 values. This includes disabled expressions, which are marked as such.
5003 It also includes expressions which would not be displayed right now
5004 because they refer to automatic variables not currently available.
5007 If a display expression refers to local variables, then it does not make
5008 sense outside the lexical context for which it was set up. Such an
5009 expression is disabled when execution enters a context where one of its
5010 variables is not defined. For example, if you give the command
5011 @code{display last_char} while inside a function with an argument
5012 @code{last_char}, @value{GDBN} displays this argument while your program
5013 continues to stop inside that function. When it stops elsewhere---where
5014 there is no variable @code{last_char}---the display is disabled
5015 automatically. The next time your program stops where @code{last_char}
5016 is meaningful, you can enable the display expression once again.
5018 @node Print Settings
5019 @section Print settings
5021 @cindex format options
5022 @cindex print settings
5023 @value{GDBN} provides the following ways to control how arrays, structures,
5024 and symbols are printed.
5027 These settings are useful for debugging programs in any language:
5030 @kindex set print address
5031 @item set print address
5032 @itemx set print address on
5033 @value{GDBN} prints memory addresses showing the location of stack
5034 traces, structure values, pointer values, breakpoints, and so forth,
5035 even when it also displays the contents of those addresses. The default
5036 is @code{on}. For example, this is what a stack frame display looks like with
5037 @code{set print address on}:
5042 #0 set_quotes (lq=0x34c78 "<<", rq=0x34c88 ">>")
5044 530 if (lquote != def_lquote)
5048 @item set print address off
5049 Do not print addresses when displaying their contents. For example,
5050 this is the same stack frame displayed with @code{set print address off}:
5054 (@value{GDBP}) set print addr off
5056 #0 set_quotes (lq="<<", rq=">>") at input.c:530
5057 530 if (lquote != def_lquote)
5061 You can use @samp{set print address off} to eliminate all machine
5062 dependent displays from the @value{GDBN} interface. For example, with
5063 @code{print address off}, you should get the same text for backtraces on
5064 all machines---whether or not they involve pointer arguments.
5066 @kindex show print address
5067 @item show print address
5068 Show whether or not addresses are to be printed.
5071 When @value{GDBN} prints a symbolic address, it normally prints the
5072 closest earlier symbol plus an offset. If that symbol does not uniquely
5073 identify the address (for example, it is a name whose scope is a single
5074 source file), you may need to clarify. One way to do this is with
5075 @code{info line}, for example @samp{info line *0x4537}. Alternately,
5076 you can set @value{GDBN} to print the source file and line number when
5077 it prints a symbolic address:
5080 @kindex set print symbol-filename
5081 @item set print symbol-filename on
5082 Tell @value{GDBN} to print the source file name and line number of a
5083 symbol in the symbolic form of an address.
5085 @item set print symbol-filename off
5086 Do not print source file name and line number of a symbol. This is the
5089 @kindex show print symbol-filename
5090 @item show print symbol-filename
5091 Show whether or not @value{GDBN} will print the source file name and
5092 line number of a symbol in the symbolic form of an address.
5095 Another situation where it is helpful to show symbol filenames and line
5096 numbers is when disassembling code; @value{GDBN} shows you the line
5097 number and source file that corresponds to each instruction.
5099 Also, you may wish to see the symbolic form only if the address being
5100 printed is reasonably close to the closest earlier symbol:
5103 @kindex set print max-symbolic-offset
5104 @item set print max-symbolic-offset @var{max-offset}
5105 Tell @value{GDBN} to only display the symbolic form of an address if the
5106 offset between the closest earlier symbol and the address is less than
5107 @var{max-offset}. The default is 0, which tells @value{GDBN}
5108 to always print the symbolic form of an address if any symbol precedes it.
5110 @kindex show print max-symbolic-offset
5111 @item show print max-symbolic-offset
5112 Ask how large the maximum offset is that @value{GDBN} prints in a
5116 @cindex wild pointer, interpreting
5117 @cindex pointer, finding referent
5118 If you have a pointer and you are not sure where it points, try
5119 @samp{set print symbol-filename on}. Then you can determine the name
5120 and source file location of the variable where it points, using
5121 @samp{p/a @var{pointer}}. This interprets the address in symbolic form.
5122 For example, here @value{GDBN} shows that a variable @code{ptt} points
5123 at another variable @code{t}, defined in @file{hi2.c}:
5126 (@value{GDBP}) set print symbol-filename on
5127 (@value{GDBP}) p/a ptt
5128 $4 = 0xe008 <t in hi2.c>
5132 @emph{Warning:} For pointers that point to a local variable, @samp{p/a}
5133 does not show the symbol name and filename of the referent, even with
5134 the appropriate @code{set print} options turned on.
5137 Other settings control how different kinds of objects are printed:
5140 @kindex set print array
5141 @item set print array
5142 @itemx set print array on
5143 Pretty print arrays. This format is more convenient to read,
5144 but uses more space. The default is off.
5146 @item set print array off
5147 Return to compressed format for arrays.
5149 @kindex show print array
5150 @item show print array
5151 Show whether compressed or pretty format is selected for displaying
5154 @kindex set print elements
5155 @item set print elements @var{number-of-elements}
5156 Set a limit on how many elements of an array @value{GDBN} will print.
5157 If @value{GDBN} is printing a large array, it stops printing after it has
5158 printed the number of elements set by the @code{set print elements} command.
5159 This limit also applies to the display of strings.
5160 When @value{GDBN} starts, this limit is set to 200.
5161 Setting @var{number-of-elements} to zero means that the printing is unlimited.
5163 @kindex show print elements
5164 @item show print elements
5165 Display the number of elements of a large array that @value{GDBN} will print.
5166 If the number is 0, then the printing is unlimited.
5168 @kindex set print null-stop
5169 @item set print null-stop
5170 Cause @value{GDBN} to stop printing the characters of an array when the first
5171 @sc{null} is encountered. This is useful when large arrays actually
5172 contain only short strings.
5175 @kindex set print pretty
5176 @item set print pretty on
5177 Cause @value{GDBN} to print structures in an indented format with one member
5178 per line, like this:
5193 @item set print pretty off
5194 Cause @value{GDBN} to print structures in a compact format, like this:
5198 $1 = @{next = 0x0, flags = @{sweet = 1, sour = 1@}, \
5199 meat = 0x54 "Pork"@}
5204 This is the default format.
5206 @kindex show print pretty
5207 @item show print pretty
5208 Show which format @value{GDBN} is using to print structures.
5210 @kindex set print sevenbit-strings
5211 @item set print sevenbit-strings on
5212 Print using only seven-bit characters; if this option is set,
5213 @value{GDBN} displays any eight-bit characters (in strings or
5214 character values) using the notation @code{\}@var{nnn}. This setting is
5215 best if you are working in English (@sc{ascii}) and you use the
5216 high-order bit of characters as a marker or ``meta'' bit.
5218 @item set print sevenbit-strings off
5219 Print full eight-bit characters. This allows the use of more
5220 international character sets, and is the default.
5222 @kindex show print sevenbit-strings
5223 @item show print sevenbit-strings
5224 Show whether or not @value{GDBN} is printing only seven-bit characters.
5226 @kindex set print union
5227 @item set print union on
5228 Tell @value{GDBN} to print unions which are contained in structures. This
5229 is the default setting.
5231 @item set print union off
5232 Tell @value{GDBN} not to print unions which are contained in structures.
5234 @kindex show print union
5235 @item show print union
5236 Ask @value{GDBN} whether or not it will print unions which are contained in
5239 For example, given the declarations
5242 typedef enum @{Tree, Bug@} Species;
5243 typedef enum @{Big_tree, Acorn, Seedling@} Tree_forms;
5244 typedef enum @{Caterpillar, Cocoon, Butterfly@}
5255 struct thing foo = @{Tree, @{Acorn@}@};
5259 with @code{set print union on} in effect @samp{p foo} would print
5262 $1 = @{it = Tree, form = @{tree = Acorn, bug = Cocoon@}@}
5266 and with @code{set print union off} in effect it would print
5269 $1 = @{it = Tree, form = @{...@}@}
5275 These settings are of interest when debugging C@t{++} programs:
5279 @kindex set print demangle
5280 @item set print demangle
5281 @itemx set print demangle on
5282 Print C@t{++} names in their source form rather than in the encoded
5283 (``mangled'') form passed to the assembler and linker for type-safe
5284 linkage. The default is on.
5286 @kindex show print demangle
5287 @item show print demangle
5288 Show whether C@t{++} names are printed in mangled or demangled form.
5290 @kindex set print asm-demangle
5291 @item set print asm-demangle
5292 @itemx set print asm-demangle on
5293 Print C@t{++} names in their source form rather than their mangled form, even
5294 in assembler code printouts such as instruction disassemblies.
5297 @kindex show print asm-demangle
5298 @item show print asm-demangle
5299 Show whether C@t{++} names in assembly listings are printed in mangled
5302 @kindex set demangle-style
5303 @cindex C@t{++} symbol decoding style
5304 @cindex symbol decoding style, C@t{++}
5305 @item set demangle-style @var{style}
5306 Choose among several encoding schemes used by different compilers to
5307 represent C@t{++} names. The choices for @var{style} are currently:
5311 Allow @value{GDBN} to choose a decoding style by inspecting your program.
5314 Decode based on the @sc{gnu} C@t{++} compiler (@code{g++}) encoding algorithm.
5315 This is the default.
5318 Decode based on the HP ANSI C@t{++} (@code{aCC}) encoding algorithm.
5321 Decode based on the Lucid C@t{++} compiler (@code{lcc}) encoding algorithm.
5324 Decode using the algorithm in the @cite{C@t{++} Annotated Reference Manual}.
5325 @strong{Warning:} this setting alone is not sufficient to allow
5326 debugging @code{cfront}-generated executables. @value{GDBN} would
5327 require further enhancement to permit that.
5330 If you omit @var{style}, you will see a list of possible formats.
5332 @kindex show demangle-style
5333 @item show demangle-style
5334 Display the encoding style currently in use for decoding C@t{++} symbols.
5336 @kindex set print object
5337 @item set print object
5338 @itemx set print object on
5339 When displaying a pointer to an object, identify the @emph{actual}
5340 (derived) type of the object rather than the @emph{declared} type, using
5341 the virtual function table.
5343 @item set print object off
5344 Display only the declared type of objects, without reference to the
5345 virtual function table. This is the default setting.
5347 @kindex show print object
5348 @item show print object
5349 Show whether actual, or declared, object types are displayed.
5351 @kindex set print static-members
5352 @item set print static-members
5353 @itemx set print static-members on
5354 Print static members when displaying a C@t{++} object. The default is on.
5356 @item set print static-members off
5357 Do not print static members when displaying a C@t{++} object.
5359 @kindex show print static-members
5360 @item show print static-members
5361 Show whether C@t{++} static members are printed, or not.
5363 @c These don't work with HP ANSI C++ yet.
5364 @kindex set print vtbl
5365 @item set print vtbl
5366 @itemx set print vtbl on
5367 Pretty print C@t{++} virtual function tables. The default is off.
5368 (The @code{vtbl} commands do not work on programs compiled with the HP
5369 ANSI C@t{++} compiler (@code{aCC}).)
5371 @item set print vtbl off
5372 Do not pretty print C@t{++} virtual function tables.
5374 @kindex show print vtbl
5375 @item show print vtbl
5376 Show whether C@t{++} virtual function tables are pretty printed, or not.
5380 @section Value history
5382 @cindex value history
5383 Values printed by the @code{print} command are saved in the @value{GDBN}
5384 @dfn{value history}. This allows you to refer to them in other expressions.
5385 Values are kept until the symbol table is re-read or discarded
5386 (for example with the @code{file} or @code{symbol-file} commands).
5387 When the symbol table changes, the value history is discarded,
5388 since the values may contain pointers back to the types defined in the
5393 @cindex history number
5394 The values printed are given @dfn{history numbers} by which you can
5395 refer to them. These are successive integers starting with one.
5396 @code{print} shows you the history number assigned to a value by
5397 printing @samp{$@var{num} = } before the value; here @var{num} is the
5400 To refer to any previous value, use @samp{$} followed by the value's
5401 history number. The way @code{print} labels its output is designed to
5402 remind you of this. Just @code{$} refers to the most recent value in
5403 the history, and @code{$$} refers to the value before that.
5404 @code{$$@var{n}} refers to the @var{n}th value from the end; @code{$$2}
5405 is the value just prior to @code{$$}, @code{$$1} is equivalent to
5406 @code{$$}, and @code{$$0} is equivalent to @code{$}.
5408 For example, suppose you have just printed a pointer to a structure and
5409 want to see the contents of the structure. It suffices to type
5415 If you have a chain of structures where the component @code{next} points
5416 to the next one, you can print the contents of the next one with this:
5423 You can print successive links in the chain by repeating this
5424 command---which you can do by just typing @key{RET}.
5426 Note that the history records values, not expressions. If the value of
5427 @code{x} is 4 and you type these commands:
5435 then the value recorded in the value history by the @code{print} command
5436 remains 4 even though the value of @code{x} has changed.
5441 Print the last ten values in the value history, with their item numbers.
5442 This is like @samp{p@ $$9} repeated ten times, except that @code{show
5443 values} does not change the history.
5445 @item show values @var{n}
5446 Print ten history values centered on history item number @var{n}.
5449 Print ten history values just after the values last printed. If no more
5450 values are available, @code{show values +} produces no display.
5453 Pressing @key{RET} to repeat @code{show values @var{n}} has exactly the
5454 same effect as @samp{show values +}.
5456 @node Convenience Vars
5457 @section Convenience variables
5459 @cindex convenience variables
5460 @value{GDBN} provides @dfn{convenience variables} that you can use within
5461 @value{GDBN} to hold on to a value and refer to it later. These variables
5462 exist entirely within @value{GDBN}; they are not part of your program, and
5463 setting a convenience variable has no direct effect on further execution
5464 of your program. That is why you can use them freely.
5466 Convenience variables are prefixed with @samp{$}. Any name preceded by
5467 @samp{$} can be used for a convenience variable, unless it is one of
5468 the predefined machine-specific register names (@pxref{Registers, ,Registers}).
5469 (Value history references, in contrast, are @emph{numbers} preceded
5470 by @samp{$}. @xref{Value History, ,Value history}.)
5472 You can save a value in a convenience variable with an assignment
5473 expression, just as you would set a variable in your program.
5477 set $foo = *object_ptr
5481 would save in @code{$foo} the value contained in the object pointed to by
5484 Using a convenience variable for the first time creates it, but its
5485 value is @code{void} until you assign a new value. You can alter the
5486 value with another assignment at any time.
5488 Convenience variables have no fixed types. You can assign a convenience
5489 variable any type of value, including structures and arrays, even if
5490 that variable already has a value of a different type. The convenience
5491 variable, when used as an expression, has the type of its current value.
5494 @kindex show convenience
5495 @item show convenience
5496 Print a list of convenience variables used so far, and their values.
5497 Abbreviated @code{show conv}.
5500 One of the ways to use a convenience variable is as a counter to be
5501 incremented or a pointer to be advanced. For example, to print
5502 a field from successive elements of an array of structures:
5506 print bar[$i++]->contents
5510 Repeat that command by typing @key{RET}.
5512 Some convenience variables are created automatically by @value{GDBN} and given
5513 values likely to be useful.
5516 @vindex $_@r{, convenience variable}
5518 The variable @code{$_} is automatically set by the @code{x} command to
5519 the last address examined (@pxref{Memory, ,Examining memory}). Other
5520 commands which provide a default address for @code{x} to examine also
5521 set @code{$_} to that address; these commands include @code{info line}
5522 and @code{info breakpoint}. The type of @code{$_} is @code{void *}
5523 except when set by the @code{x} command, in which case it is a pointer
5524 to the type of @code{$__}.
5526 @vindex $__@r{, convenience variable}
5528 The variable @code{$__} is automatically set by the @code{x} command
5529 to the value found in the last address examined. Its type is chosen
5530 to match the format in which the data was printed.
5533 @vindex $_exitcode@r{, convenience variable}
5534 The variable @code{$_exitcode} is automatically set to the exit code when
5535 the program being debugged terminates.
5538 On HP-UX systems, if you refer to a function or variable name that
5539 begins with a dollar sign, @value{GDBN} searches for a user or system
5540 name first, before it searches for a convenience variable.
5546 You can refer to machine register contents, in expressions, as variables
5547 with names starting with @samp{$}. The names of registers are different
5548 for each machine; use @code{info registers} to see the names used on
5552 @kindex info registers
5553 @item info registers
5554 Print the names and values of all registers except floating-point
5555 registers (in the selected stack frame).
5557 @kindex info all-registers
5558 @cindex floating point registers
5559 @item info all-registers
5560 Print the names and values of all registers, including floating-point
5563 @item info registers @var{regname} @dots{}
5564 Print the @dfn{relativized} value of each specified register @var{regname}.
5565 As discussed in detail below, register values are normally relative to
5566 the selected stack frame. @var{regname} may be any register name valid on
5567 the machine you are using, with or without the initial @samp{$}.
5570 @value{GDBN} has four ``standard'' register names that are available (in
5571 expressions) on most machines---whenever they do not conflict with an
5572 architecture's canonical mnemonics for registers. The register names
5573 @code{$pc} and @code{$sp} are used for the program counter register and
5574 the stack pointer. @code{$fp} is used for a register that contains a
5575 pointer to the current stack frame, and @code{$ps} is used for a
5576 register that contains the processor status. For example,
5577 you could print the program counter in hex with
5584 or print the instruction to be executed next with
5591 or add four to the stack pointer@footnote{This is a way of removing
5592 one word from the stack, on machines where stacks grow downward in
5593 memory (most machines, nowadays). This assumes that the innermost
5594 stack frame is selected; setting @code{$sp} is not allowed when other
5595 stack frames are selected. To pop entire frames off the stack,
5596 regardless of machine architecture, use @code{return};
5597 see @ref{Returning, ,Returning from a function}.} with
5603 Whenever possible, these four standard register names are available on
5604 your machine even though the machine has different canonical mnemonics,
5605 so long as there is no conflict. The @code{info registers} command
5606 shows the canonical names. For example, on the SPARC, @code{info
5607 registers} displays the processor status register as @code{$psr} but you
5608 can also refer to it as @code{$ps}; and on x86-based machines @code{$ps}
5609 is an alias for the @sc{eflags} register.
5611 @value{GDBN} always considers the contents of an ordinary register as an
5612 integer when the register is examined in this way. Some machines have
5613 special registers which can hold nothing but floating point; these
5614 registers are considered to have floating point values. There is no way
5615 to refer to the contents of an ordinary register as floating point value
5616 (although you can @emph{print} it as a floating point value with
5617 @samp{print/f $@var{regname}}).
5619 Some registers have distinct ``raw'' and ``virtual'' data formats. This
5620 means that the data format in which the register contents are saved by
5621 the operating system is not the same one that your program normally
5622 sees. For example, the registers of the 68881 floating point
5623 coprocessor are always saved in ``extended'' (raw) format, but all C
5624 programs expect to work with ``double'' (virtual) format. In such
5625 cases, @value{GDBN} normally works with the virtual format only (the format
5626 that makes sense for your program), but the @code{info registers} command
5627 prints the data in both formats.
5629 Normally, register values are relative to the selected stack frame
5630 (@pxref{Selection, ,Selecting a frame}). This means that you get the
5631 value that the register would contain if all stack frames farther in
5632 were exited and their saved registers restored. In order to see the
5633 true contents of hardware registers, you must select the innermost
5634 frame (with @samp{frame 0}).
5636 However, @value{GDBN} must deduce where registers are saved, from the machine
5637 code generated by your compiler. If some registers are not saved, or if
5638 @value{GDBN} is unable to locate the saved registers, the selected stack
5639 frame makes no difference.
5641 @node Floating Point Hardware
5642 @section Floating point hardware
5643 @cindex floating point
5645 Depending on the configuration, @value{GDBN} may be able to give
5646 you more information about the status of the floating point hardware.
5651 Display hardware-dependent information about the floating
5652 point unit. The exact contents and layout vary depending on the
5653 floating point chip. Currently, @samp{info float} is supported on
5654 the ARM and x86 machines.
5657 @node Memory Region Attributes
5658 @section Memory Region Attributes
5659 @cindex memory region attributes
5661 @dfn{Memory region attributes} allow you to describe special handling
5662 required by regions of your target's memory. @value{GDBN} uses attributes
5663 to determine whether to allow certain types of memory accesses; whether to
5664 use specific width accesses; and whether to cache target memory.
5666 Defined memory regions can be individually enabled and disabled. When a
5667 memory region is disabled, @value{GDBN} uses the default attributes when
5668 accessing memory in that region. Similarly, if no memory regions have
5669 been defined, @value{GDBN} uses the default attributes when accessing
5672 When a memory region is defined, it is given a number to identify it;
5673 to enable, disable, or remove a memory region, you specify that number.
5677 @item mem @var{address1} @var{address1} @var{attributes}@dots{}
5678 Define memory region bounded by @var{address1} and @var{address2}
5679 with attributes @var{attributes}@dots{}.
5682 @item delete mem @var{nums}@dots{}
5683 Remove memory region numbers @var{nums}.
5686 @item disable mem @var{nums}@dots{}
5687 Disable memory region numbers @var{nums}.
5688 A disabled memory region is not forgotten.
5689 It may be enabled again later.
5692 @item enable mem @var{nums}@dots{}
5693 Enable memory region numbers @var{nums}.
5697 Print a table of all defined memory regions, with the following columns
5701 @item Memory Region Number
5702 @item Enabled or Disabled.
5703 Enabled memory regions are marked with @samp{y}.
5704 Disabled memory regions are marked with @samp{n}.
5707 The address defining the inclusive lower bound of the memory region.
5710 The address defining the exclusive upper bound of the memory region.
5713 The list of attributes set for this memory region.
5718 @subsection Attributes
5720 @subsubsection Memory Access Mode
5721 The access mode attributes set whether @value{GDBN} may make read or
5722 write accesses to a memory region.
5724 While these attributes prevent @value{GDBN} from performing invalid
5725 memory accesses, they do nothing to prevent the target system, I/O DMA,
5726 etc. from accessing memory.
5730 Memory is read only.
5732 Memory is write only.
5734 Memory is read/write (default).
5737 @subsubsection Memory Access Size
5738 The acccess size attributes tells @value{GDBN} to use specific sized
5739 accesses in the memory region. Often memory mapped device registers
5740 require specific sized accesses. If no access size attribute is
5741 specified, @value{GDBN} may use accesses of any size.
5745 Use 8 bit memory accesses.
5747 Use 16 bit memory accesses.
5749 Use 32 bit memory accesses.
5751 Use 64 bit memory accesses.
5754 @c @subsubsection Hardware/Software Breakpoints
5755 @c The hardware/software breakpoint attributes set whether @value{GDBN}
5756 @c will use hardware or software breakpoints for the internal breakpoints
5757 @c used by the step, next, finish, until, etc. commands.
5761 @c Always use hardware breakpoints
5762 @c @item swbreak (default)
5765 @subsubsection Data Cache
5766 The data cache attributes set whether @value{GDBN} will cache target
5767 memory. While this generally improves performance by reducing debug
5768 protocol overhead, it can lead to incorrect results because @value{GDBN}
5769 does not know about volatile variables or memory mapped device
5774 Enable @value{GDBN} to cache target memory.
5775 @item nocache (default)
5776 Disable @value{GDBN} from caching target memory.
5779 @c @subsubsection Memory Write Verification
5780 @c The memory write verification attributes set whether @value{GDBN}
5781 @c will re-reads data after each write to verify the write was successful.
5785 @c @item noverify (default)
5789 @chapter Tracepoints
5790 @c This chapter is based on the documentation written by Michael
5791 @c Snyder, David Taylor, Jim Blandy, and Elena Zannoni.
5794 In some applications, it is not feasible for the debugger to interrupt
5795 the program's execution long enough for the developer to learn
5796 anything helpful about its behavior. If the program's correctness
5797 depends on its real-time behavior, delays introduced by a debugger
5798 might cause the program to change its behavior drastically, or perhaps
5799 fail, even when the code itself is correct. It is useful to be able
5800 to observe the program's behavior without interrupting it.
5802 Using @value{GDBN}'s @code{trace} and @code{collect} commands, you can
5803 specify locations in the program, called @dfn{tracepoints}, and
5804 arbitrary expressions to evaluate when those tracepoints are reached.
5805 Later, using the @code{tfind} command, you can examine the values
5806 those expressions had when the program hit the tracepoints. The
5807 expressions may also denote objects in memory---structures or arrays,
5808 for example---whose values @value{GDBN} should record; while visiting
5809 a particular tracepoint, you may inspect those objects as if they were
5810 in memory at that moment. However, because @value{GDBN} records these
5811 values without interacting with you, it can do so quickly and
5812 unobtrusively, hopefully not disturbing the program's behavior.
5814 The tracepoint facility is currently available only for remote
5815 targets. @xref{Targets}. In addition, your remote target must know how
5816 to collect trace data. This functionality is implemented in the remote
5817 stub; however, none of the stubs distributed with @value{GDBN} support
5818 tracepoints as of this writing.
5820 This chapter describes the tracepoint commands and features.
5824 * Analyze Collected Data::
5825 * Tracepoint Variables::
5828 @node Set Tracepoints
5829 @section Commands to Set Tracepoints
5831 Before running such a @dfn{trace experiment}, an arbitrary number of
5832 tracepoints can be set. Like a breakpoint (@pxref{Set Breaks}), a
5833 tracepoint has a number assigned to it by @value{GDBN}. Like with
5834 breakpoints, tracepoint numbers are successive integers starting from
5835 one. Many of the commands associated with tracepoints take the
5836 tracepoint number as their argument, to identify which tracepoint to
5839 For each tracepoint, you can specify, in advance, some arbitrary set
5840 of data that you want the target to collect in the trace buffer when
5841 it hits that tracepoint. The collected data can include registers,
5842 local variables, or global data. Later, you can use @value{GDBN}
5843 commands to examine the values these data had at the time the
5846 This section describes commands to set tracepoints and associated
5847 conditions and actions.
5850 * Create and Delete Tracepoints::
5851 * Enable and Disable Tracepoints::
5852 * Tracepoint Passcounts::
5853 * Tracepoint Actions::
5854 * Listing Tracepoints::
5855 * Starting and Stopping Trace Experiment::
5858 @node Create and Delete Tracepoints
5859 @subsection Create and Delete Tracepoints
5862 @cindex set tracepoint
5865 The @code{trace} command is very similar to the @code{break} command.
5866 Its argument can be a source line, a function name, or an address in
5867 the target program. @xref{Set Breaks}. The @code{trace} command
5868 defines a tracepoint, which is a point in the target program where the
5869 debugger will briefly stop, collect some data, and then allow the
5870 program to continue. Setting a tracepoint or changing its commands
5871 doesn't take effect until the next @code{tstart} command; thus, you
5872 cannot change the tracepoint attributes once a trace experiment is
5875 Here are some examples of using the @code{trace} command:
5878 (@value{GDBP}) @b{trace foo.c:121} // a source file and line number
5880 (@value{GDBP}) @b{trace +2} // 2 lines forward
5882 (@value{GDBP}) @b{trace my_function} // first source line of function
5884 (@value{GDBP}) @b{trace *my_function} // EXACT start address of function
5886 (@value{GDBP}) @b{trace *0x2117c4} // an address
5890 You can abbreviate @code{trace} as @code{tr}.
5893 @cindex last tracepoint number
5894 @cindex recent tracepoint number
5895 @cindex tracepoint number
5896 The convenience variable @code{$tpnum} records the tracepoint number
5897 of the most recently set tracepoint.
5899 @kindex delete tracepoint
5900 @cindex tracepoint deletion
5901 @item delete tracepoint @r{[}@var{num}@r{]}
5902 Permanently delete one or more tracepoints. With no argument, the
5903 default is to delete all tracepoints.
5908 (@value{GDBP}) @b{delete trace 1 2 3} // remove three tracepoints
5910 (@value{GDBP}) @b{delete trace} // remove all tracepoints
5914 You can abbreviate this command as @code{del tr}.
5917 @node Enable and Disable Tracepoints
5918 @subsection Enable and Disable Tracepoints
5921 @kindex disable tracepoint
5922 @item disable tracepoint @r{[}@var{num}@r{]}
5923 Disable tracepoint @var{num}, or all tracepoints if no argument
5924 @var{num} is given. A disabled tracepoint will have no effect during
5925 the next trace experiment, but it is not forgotten. You can re-enable
5926 a disabled tracepoint using the @code{enable tracepoint} command.
5928 @kindex enable tracepoint
5929 @item enable tracepoint @r{[}@var{num}@r{]}
5930 Enable tracepoint @var{num}, or all tracepoints. The enabled
5931 tracepoints will become effective the next time a trace experiment is
5935 @node Tracepoint Passcounts
5936 @subsection Tracepoint Passcounts
5940 @cindex tracepoint pass count
5941 @item passcount @r{[}@var{n} @r{[}@var{num}@r{]]}
5942 Set the @dfn{passcount} of a tracepoint. The passcount is a way to
5943 automatically stop a trace experiment. If a tracepoint's passcount is
5944 @var{n}, then the trace experiment will be automatically stopped on
5945 the @var{n}'th time that tracepoint is hit. If the tracepoint number
5946 @var{num} is not specified, the @code{passcount} command sets the
5947 passcount of the most recently defined tracepoint. If no passcount is
5948 given, the trace experiment will run until stopped explicitly by the
5954 (@value{GDBP}) @b{passcount 5 2} // Stop on the 5th execution of
5955 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// tracepoint 2}
5957 (@value{GDBP}) @b{passcount 12} // Stop on the 12th execution of the
5958 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// most recently defined tracepoint.}
5959 (@value{GDBP}) @b{trace foo}
5960 (@value{GDBP}) @b{pass 3}
5961 (@value{GDBP}) @b{trace bar}
5962 (@value{GDBP}) @b{pass 2}
5963 (@value{GDBP}) @b{trace baz}
5964 (@value{GDBP}) @b{pass 1} // Stop tracing when foo has been
5965 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// executed 3 times OR when bar has}
5966 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// been executed 2 times}
5967 @exdent @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @code{// OR when baz has been executed 1 time.}
5971 @node Tracepoint Actions
5972 @subsection Tracepoint Action Lists
5976 @cindex tracepoint actions
5977 @item actions @r{[}@var{num}@r{]}
5978 This command will prompt for a list of actions to be taken when the
5979 tracepoint is hit. If the tracepoint number @var{num} is not
5980 specified, this command sets the actions for the one that was most
5981 recently defined (so that you can define a tracepoint and then say
5982 @code{actions} without bothering about its number). You specify the
5983 actions themselves on the following lines, one action at a time, and
5984 terminate the actions list with a line containing just @code{end}. So
5985 far, the only defined actions are @code{collect} and
5986 @code{while-stepping}.
5988 @cindex remove actions from a tracepoint
5989 To remove all actions from a tracepoint, type @samp{actions @var{num}}
5990 and follow it immediately with @samp{end}.
5993 (@value{GDBP}) @b{collect @var{data}} // collect some data
5995 (@value{GDBP}) @b{while-stepping 5} // single-step 5 times, collect data
5997 (@value{GDBP}) @b{end} // signals the end of actions.
6000 In the following example, the action list begins with @code{collect}
6001 commands indicating the things to be collected when the tracepoint is
6002 hit. Then, in order to single-step and collect additional data
6003 following the tracepoint, a @code{while-stepping} command is used,
6004 followed by the list of things to be collected while stepping. The
6005 @code{while-stepping} command is terminated by its own separate
6006 @code{end} command. Lastly, the action list is terminated by an
6010 (@value{GDBP}) @b{trace foo}
6011 (@value{GDBP}) @b{actions}
6012 Enter actions for tracepoint 1, one per line:
6021 @kindex collect @r{(tracepoints)}
6022 @item collect @var{expr1}, @var{expr2}, @dots{}
6023 Collect values of the given expressions when the tracepoint is hit.
6024 This command accepts a comma-separated list of any valid expressions.
6025 In addition to global, static, or local variables, the following
6026 special arguments are supported:
6030 collect all registers
6033 collect all function arguments
6036 collect all local variables.
6039 You can give several consecutive @code{collect} commands, each one
6040 with a single argument, or one @code{collect} command with several
6041 arguments separated by commas: the effect is the same.
6043 The command @code{info scope} (@pxref{Symbols, info scope}) is
6044 particularly useful for figuring out what data to collect.
6046 @kindex while-stepping @r{(tracepoints)}
6047 @item while-stepping @var{n}
6048 Perform @var{n} single-step traces after the tracepoint, collecting
6049 new data at each step. The @code{while-stepping} command is
6050 followed by the list of what to collect while stepping (followed by
6051 its own @code{end} command):
6055 > collect $regs, myglobal
6061 You may abbreviate @code{while-stepping} as @code{ws} or
6065 @node Listing Tracepoints
6066 @subsection Listing Tracepoints
6069 @kindex info tracepoints
6070 @cindex information about tracepoints
6071 @item info tracepoints @r{[}@var{num}@r{]}
6072 @c FIXME: Shouldn't there be an "at", "on", or "for" in the next
6073 @c sentence? - Brian Y., FSF office staff
6074 Display information the tracepoint @var{num}. If you don't specify a
6075 tracepoint number displays information about all the tracepoints
6076 defined so far. For each tracepoint, the following information is
6083 whether it is enabled or disabled
6087 its passcount as given by the @code{passcount @var{n}} command
6089 its step count as given by the @code{while-stepping @var{n}} command
6091 where in the source files is the tracepoint set
6093 its action list as given by the @code{actions} command
6097 (@value{GDBP}) @b{info trace}
6098 Num Enb Address PassC StepC What
6099 1 y 0x002117c4 0 0 <gdb_asm>
6100 2 y 0x0020dc64 0 0 in g_test at g_test.c:1375
6101 3 y 0x0020b1f4 0 0 in get_data at ../foo.c:41
6106 This command can be abbreviated @code{info tp}.
6109 @node Starting and Stopping Trace Experiment
6110 @subsection Starting and Stopping Trace Experiment
6114 @cindex start a new trace experiment
6115 @cindex collected data discarded
6117 This command takes no arguments. It starts the trace experiment, and
6118 begins collecting data. This has the side effect of discarding all
6119 the data collected in the trace buffer during the previous trace
6123 @cindex stop a running trace experiment
6125 This command takes no arguments. It ends the trace experiment, and
6126 stops collecting data.
6128 @strong{Note:} a trace experiment and data collection may stop
6129 automatically if any tracepoint's passcount is reached
6130 (@pxref{Tracepoint Passcounts}), or if the trace buffer becomes full.
6133 @cindex status of trace data collection
6134 @cindex trace experiment, status of
6136 This command displays the status of the current trace data
6140 Here is an example of the commands we described so far:
6143 (@value{GDBP}) @b{trace gdb_c_test}
6144 (@value{GDBP}) @b{actions}
6145 Enter actions for tracepoint #1, one per line.
6146 > collect $regs,$locals,$args
6151 (@value{GDBP}) @b{tstart}
6152 [time passes @dots{}]
6153 (@value{GDBP}) @b{tstop}
6157 @node Analyze Collected Data
6158 @section Using the collected data
6160 After the tracepoint experiment ends, you use @value{GDBN} commands
6161 for examining the trace data. The basic idea is that each tracepoint
6162 collects a trace @dfn{snapshot} every time it is hit and another
6163 snapshot every time it single-steps. All these snapshots are
6164 consecutively numbered from zero and go into a buffer, and you can
6165 examine them later. The way you examine them is to @dfn{focus} on a
6166 specific trace snapshot. When the remote stub is focused on a trace
6167 snapshot, it will respond to all @value{GDBN} requests for memory and
6168 registers by reading from the buffer which belongs to that snapshot,
6169 rather than from @emph{real} memory or registers of the program being
6170 debugged. This means that @strong{all} @value{GDBN} commands
6171 (@code{print}, @code{info registers}, @code{backtrace}, etc.) will
6172 behave as if we were currently debugging the program state as it was
6173 when the tracepoint occurred. Any requests for data that are not in
6174 the buffer will fail.
6177 * tfind:: How to select a trace snapshot
6178 * tdump:: How to display all data for a snapshot
6179 * save-tracepoints:: How to save tracepoints for a future run
6183 @subsection @code{tfind @var{n}}
6186 @cindex select trace snapshot
6187 @cindex find trace snapshot
6188 The basic command for selecting a trace snapshot from the buffer is
6189 @code{tfind @var{n}}, which finds trace snapshot number @var{n},
6190 counting from zero. If no argument @var{n} is given, the next
6191 snapshot is selected.
6193 Here are the various forms of using the @code{tfind} command.
6197 Find the first snapshot in the buffer. This is a synonym for
6198 @code{tfind 0} (since 0 is the number of the first snapshot).
6201 Stop debugging trace snapshots, resume @emph{live} debugging.
6204 Same as @samp{tfind none}.
6207 No argument means find the next trace snapshot.
6210 Find the previous trace snapshot before the current one. This permits
6211 retracing earlier steps.
6213 @item tfind tracepoint @var{num}
6214 Find the next snapshot associated with tracepoint @var{num}. Search
6215 proceeds forward from the last examined trace snapshot. If no
6216 argument @var{num} is given, it means find the next snapshot collected
6217 for the same tracepoint as the current snapshot.
6219 @item tfind pc @var{addr}
6220 Find the next snapshot associated with the value @var{addr} of the
6221 program counter. Search proceeds forward from the last examined trace
6222 snapshot. If no argument @var{addr} is given, it means find the next
6223 snapshot with the same value of PC as the current snapshot.
6225 @item tfind outside @var{addr1}, @var{addr2}
6226 Find the next snapshot whose PC is outside the given range of
6229 @item tfind range @var{addr1}, @var{addr2}
6230 Find the next snapshot whose PC is between @var{addr1} and
6231 @var{addr2}. @c FIXME: Is the range inclusive or exclusive?
6233 @item tfind line @r{[}@var{file}:@r{]}@var{n}
6234 Find the next snapshot associated with the source line @var{n}. If
6235 the optional argument @var{file} is given, refer to line @var{n} in
6236 that source file. Search proceeds forward from the last examined
6237 trace snapshot. If no argument @var{n} is given, it means find the
6238 next line other than the one currently being examined; thus saying
6239 @code{tfind line} repeatedly can appear to have the same effect as
6240 stepping from line to line in a @emph{live} debugging session.
6243 The default arguments for the @code{tfind} commands are specifically
6244 designed to make it easy to scan through the trace buffer. For
6245 instance, @code{tfind} with no argument selects the next trace
6246 snapshot, and @code{tfind -} with no argument selects the previous
6247 trace snapshot. So, by giving one @code{tfind} command, and then
6248 simply hitting @key{RET} repeatedly you can examine all the trace
6249 snapshots in order. Or, by saying @code{tfind -} and then hitting
6250 @key{RET} repeatedly you can examine the snapshots in reverse order.
6251 The @code{tfind line} command with no argument selects the snapshot
6252 for the next source line executed. The @code{tfind pc} command with
6253 no argument selects the next snapshot with the same program counter
6254 (PC) as the current frame. The @code{tfind tracepoint} command with
6255 no argument selects the next trace snapshot collected by the same
6256 tracepoint as the current one.
6258 In addition to letting you scan through the trace buffer manually,
6259 these commands make it easy to construct @value{GDBN} scripts that
6260 scan through the trace buffer and print out whatever collected data
6261 you are interested in. Thus, if we want to examine the PC, FP, and SP
6262 registers from each trace frame in the buffer, we can say this:
6265 (@value{GDBP}) @b{tfind start}
6266 (@value{GDBP}) @b{while ($trace_frame != -1)}
6267 > printf "Frame %d, PC = %08X, SP = %08X, FP = %08X\n", \
6268 $trace_frame, $pc, $sp, $fp
6272 Frame 0, PC = 0020DC64, SP = 0030BF3C, FP = 0030BF44
6273 Frame 1, PC = 0020DC6C, SP = 0030BF38, FP = 0030BF44
6274 Frame 2, PC = 0020DC70, SP = 0030BF34, FP = 0030BF44
6275 Frame 3, PC = 0020DC74, SP = 0030BF30, FP = 0030BF44
6276 Frame 4, PC = 0020DC78, SP = 0030BF2C, FP = 0030BF44
6277 Frame 5, PC = 0020DC7C, SP = 0030BF28, FP = 0030BF44
6278 Frame 6, PC = 0020DC80, SP = 0030BF24, FP = 0030BF44
6279 Frame 7, PC = 0020DC84, SP = 0030BF20, FP = 0030BF44
6280 Frame 8, PC = 0020DC88, SP = 0030BF1C, FP = 0030BF44
6281 Frame 9, PC = 0020DC8E, SP = 0030BF18, FP = 0030BF44
6282 Frame 10, PC = 00203F6C, SP = 0030BE3C, FP = 0030BF14
6285 Or, if we want to examine the variable @code{X} at each source line in
6289 (@value{GDBP}) @b{tfind start}
6290 (@value{GDBP}) @b{while ($trace_frame != -1)}
6291 > printf "Frame %d, X == %d\n", $trace_frame, X
6301 @subsection @code{tdump}
6303 @cindex dump all data collected at tracepoint
6304 @cindex tracepoint data, display
6306 This command takes no arguments. It prints all the data collected at
6307 the current trace snapshot.
6310 (@value{GDBP}) @b{trace 444}
6311 (@value{GDBP}) @b{actions}
6312 Enter actions for tracepoint #2, one per line:
6313 > collect $regs, $locals, $args, gdb_long_test
6316 (@value{GDBP}) @b{tstart}
6318 (@value{GDBP}) @b{tfind line 444}
6319 #0 gdb_test (p1=0x11, p2=0x22, p3=0x33, p4=0x44, p5=0x55, p6=0x66)
6321 444 printp( "%s: arguments = 0x%X 0x%X 0x%X 0x%X 0x%X 0x%X\n", )
6323 (@value{GDBP}) @b{tdump}
6324 Data collected at tracepoint 2, trace frame 1:
6325 d0 0xc4aa0085 -995491707
6329 d4 0x71aea3d 119204413
6334 a1 0x3000668 50333288
6337 a4 0x3000698 50333336
6339 fp 0x30bf3c 0x30bf3c
6340 sp 0x30bf34 0x30bf34
6342 pc 0x20b2c8 0x20b2c8
6346 p = 0x20e5b4 "gdb-test"
6353 gdb_long_test = 17 '\021'
6358 @node save-tracepoints
6359 @subsection @code{save-tracepoints @var{filename}}
6360 @kindex save-tracepoints
6361 @cindex save tracepoints for future sessions
6363 This command saves all current tracepoint definitions together with
6364 their actions and passcounts, into a file @file{@var{filename}}
6365 suitable for use in a later debugging session. To read the saved
6366 tracepoint definitions, use the @code{source} command (@pxref{Command
6369 @node Tracepoint Variables
6370 @section Convenience Variables for Tracepoints
6371 @cindex tracepoint variables
6372 @cindex convenience variables for tracepoints
6375 @vindex $trace_frame
6376 @item (int) $trace_frame
6377 The current trace snapshot (a.k.a.@: @dfn{frame}) number, or -1 if no
6378 snapshot is selected.
6381 @item (int) $tracepoint
6382 The tracepoint for the current trace snapshot.
6385 @item (int) $trace_line
6386 The line number for the current trace snapshot.
6389 @item (char []) $trace_file
6390 The source file for the current trace snapshot.
6393 @item (char []) $trace_func
6394 The name of the function containing @code{$tracepoint}.
6397 Note: @code{$trace_file} is not suitable for use in @code{printf},
6398 use @code{output} instead.
6400 Here's a simple example of using these convenience variables for
6401 stepping through all the trace snapshots and printing some of their
6405 (@value{GDBP}) @b{tfind start}
6407 (@value{GDBP}) @b{while $trace_frame != -1}
6408 > output $trace_file
6409 > printf ", line %d (tracepoint #%d)\n", $trace_line, $tracepoint
6415 @chapter Debugging Programs That Use Overlays
6418 If your program is too large to fit completely in your target system's
6419 memory, you can sometimes use @dfn{overlays} to work around this
6420 problem. @value{GDBN} provides some support for debugging programs that
6424 * How Overlays Work:: A general explanation of overlays.
6425 * Overlay Commands:: Managing overlays in @value{GDBN}.
6426 * Automatic Overlay Debugging:: @value{GDBN} can find out which overlays are
6427 mapped by asking the inferior.
6428 * Overlay Sample Program:: A sample program using overlays.
6431 @node How Overlays Work
6432 @section How Overlays Work
6433 @cindex mapped overlays
6434 @cindex unmapped overlays
6435 @cindex load address, overlay's
6436 @cindex mapped address
6437 @cindex overlay area
6439 Suppose you have a computer whose instruction address space is only 64
6440 kilobytes long, but which has much more memory which can be accessed by
6441 other means: special instructions, segment registers, or memory
6442 management hardware, for example. Suppose further that you want to
6443 adapt a program which is larger than 64 kilobytes to run on this system.
6445 One solution is to identify modules of your program which are relatively
6446 independent, and need not call each other directly; call these modules
6447 @dfn{overlays}. Separate the overlays from the main program, and place
6448 their machine code in the larger memory. Place your main program in
6449 instruction memory, but leave at least enough space there to hold the
6450 largest overlay as well.
6452 Now, to call a function located in an overlay, you must first copy that
6453 overlay's machine code from the large memory into the space set aside
6454 for it in the instruction memory, and then jump to its entry point
6459 Data Instruction Larger
6460 Address Space Address Space Address Space
6461 +-----------+ +-----------+ +-----------+
6463 +-----------+ +-----------+ +-----------+<-- overlay 1
6464 | program | | main | | | load address
6465 | variables | | program | | overlay 1 |
6466 | and heap | | | ,---| |
6467 +-----------+ | | | | |
6468 | | +-----------+ | +-----------+
6469 +-----------+ | | | | |
6470 mapped --->+-----------+ / +-----------+<-- overlay 2
6471 address | overlay | <-' | overlay 2 | load address
6473 | | <---. +-----------+
6476 | | | +-----------+<-- overlay 3
6477 +-----------+ `--| | load address
6484 To map an overlay, copy its code from the larger address space
6485 to the instruction address space. Since the overlays shown here
6486 all use the same mapped address, only one may be mapped at a time.
6490 This diagram shows a system with separate data and instruction address
6491 spaces. For a system with a single address space for data and
6492 instructions, the diagram would be similar, except that the program
6493 variables and heap would share an address space with the main program
6494 and the overlay area.
6496 An overlay loaded into instruction memory and ready for use is called a
6497 @dfn{mapped} overlay; its @dfn{mapped address} is its address in the
6498 instruction memory. An overlay not present (or only partially present)
6499 in instruction memory is called @dfn{unmapped}; its @dfn{load address}
6500 is its address in the larger memory. The mapped address is also called
6501 the @dfn{virtual memory address}, or @dfn{VMA}; the load address is also
6502 called the @dfn{load memory address}, or @dfn{LMA}.
6504 Unfortunately, overlays are not a completely transparent way to adapt a
6505 program to limited instruction memory. They introduce a new set of
6506 global constraints you must keep in mind as you design your program:
6511 Before calling or returning to a function in an overlay, your program
6512 must make sure that overlay is actually mapped. Otherwise, the call or
6513 return will transfer control to the right address, but in the wrong
6514 overlay, and your program will probably crash.
6517 If the process of mapping an overlay is expensive on your system, you
6518 will need to choose your overlays carefully to minimize their effect on
6519 your program's performance.
6522 The executable file you load onto your system must contain each
6523 overlay's instructions, appearing at the overlay's load address, not its
6524 mapped address. However, each overlay's instructions must be relocated
6525 and its symbols defined as if the overlay were at its mapped address.
6526 You can use GNU linker scripts to specify different load and relocation
6527 addresses for pieces of your program; see @ref{Overlay Description,,,
6528 ld.info, Using ld: the GNU linker}.
6531 The procedure for loading executable files onto your system must be able
6532 to load their contents into the larger address space as well as the
6533 instruction and data spaces.
6537 The overlay system described above is rather simple, and could be
6538 improved in many ways:
6543 If your system has suitable bank switch registers or memory management
6544 hardware, you could use those facilities to make an overlay's load area
6545 contents simply appear at their mapped address in instruction space.
6546 This would probably be faster than copying the overlay to its mapped
6547 area in the usual way.
6550 If your overlays are small enough, you could set aside more than one
6551 overlay area, and have more than one overlay mapped at a time.
6554 You can use overlays to manage data, as well as instructions. In
6555 general, data overlays are even less transparent to your design than
6556 code overlays: whereas code overlays only require care when you call or
6557 return to functions, data overlays require care every time you access
6558 the data. Also, if you change the contents of a data overlay, you
6559 must copy its contents back out to its load address before you can copy a
6560 different data overlay into the same mapped area.
6565 @node Overlay Commands
6566 @section Overlay Commands
6568 To use @value{GDBN}'s overlay support, each overlay in your program must
6569 correspond to a separate section of the executable file. The section's
6570 virtual memory address and load memory address must be the overlay's
6571 mapped and load addresses. Identifying overlays with sections allows
6572 @value{GDBN} to determine the appropriate address of a function or
6573 variable, depending on whether the overlay is mapped or not.
6575 @value{GDBN}'s overlay commands all start with the word @code{overlay};
6576 you can abbreviate this as @code{ov} or @code{ovly}. The commands are:
6581 Disable @value{GDBN}'s overlay support. When overlay support is
6582 disabled, @value{GDBN} assumes that all functions and variables are
6583 always present at their mapped addresses. By default, @value{GDBN}'s
6584 overlay support is disabled.
6586 @item overlay manual
6587 @kindex overlay manual
6588 @cindex manual overlay debugging
6589 Enable @dfn{manual} overlay debugging. In this mode, @value{GDBN}
6590 relies on you to tell it which overlays are mapped, and which are not,
6591 using the @code{overlay map-overlay} and @code{overlay unmap-overlay}
6592 commands described below.
6594 @item overlay map-overlay @var{overlay}
6595 @itemx overlay map @var{overlay}
6596 @kindex overlay map-overlay
6597 @cindex map an overlay
6598 Tell @value{GDBN} that @var{overlay} is now mapped; @var{overlay} must
6599 be the name of the object file section containing the overlay. When an
6600 overlay is mapped, @value{GDBN} assumes it can find the overlay's
6601 functions and variables at their mapped addresses. @value{GDBN} assumes
6602 that any other overlays whose mapped ranges overlap that of
6603 @var{overlay} are now unmapped.
6605 @item overlay unmap-overlay @var{overlay}
6606 @itemx overlay unmap @var{overlay}
6607 @kindex overlay unmap-overlay
6608 @cindex unmap an overlay
6609 Tell @value{GDBN} that @var{overlay} is no longer mapped; @var{overlay}
6610 must be the name of the object file section containing the overlay.
6611 When an overlay is unmapped, @value{GDBN} assumes it can find the
6612 overlay's functions and variables at their load addresses.
6615 @kindex overlay auto
6616 Enable @dfn{automatic} overlay debugging. In this mode, @value{GDBN}
6617 consults a data structure the overlay manager maintains in the inferior
6618 to see which overlays are mapped. For details, see @ref{Automatic
6621 @item overlay load-target
6623 @kindex overlay load-target
6624 @cindex reloading the overlay table
6625 Re-read the overlay table from the inferior. Normally, @value{GDBN}
6626 re-reads the table @value{GDBN} automatically each time the inferior
6627 stops, so this command should only be necessary if you have changed the
6628 overlay mapping yourself using @value{GDBN}. This command is only
6629 useful when using automatic overlay debugging.
6631 @item overlay list-overlays
6633 @cindex listing mapped overlays
6634 Display a list of the overlays currently mapped, along with their mapped
6635 addresses, load addresses, and sizes.
6639 Normally, when @value{GDBN} prints a code address, it includes the name
6640 of the function the address falls in:
6644 $3 = @{int ()@} 0x11a0 <main>
6647 When overlay debugging is enabled, @value{GDBN} recognizes code in
6648 unmapped overlays, and prints the names of unmapped functions with
6649 asterisks around them. For example, if @code{foo} is a function in an
6650 unmapped overlay, @value{GDBN} prints it this way:
6654 No sections are mapped.
6656 $5 = @{int (int)@} 0x100000 <*foo*>
6659 When @code{foo}'s overlay is mapped, @value{GDBN} prints the function's
6664 Section .ov.foo.text, loaded at 0x100000 - 0x100034,
6665 mapped at 0x1016 - 0x104a
6667 $6 = @{int (int)@} 0x1016 <foo>
6670 When overlay debugging is enabled, @value{GDBN} can find the correct
6671 address for functions and variables in an overlay, whether or not the
6672 overlay is mapped. This allows most @value{GDBN} commands, like
6673 @code{break} and @code{disassemble}, to work normally, even on unmapped
6674 code. However, @value{GDBN}'s breakpoint support has some limitations:
6678 @cindex breakpoints in overlays
6679 @cindex overlays, setting breakpoints in
6680 You can set breakpoints in functions in unmapped overlays, as long as
6681 @value{GDBN} can write to the overlay at its load address.
6683 @value{GDBN} can not set hardware or simulator-based breakpoints in
6684 unmapped overlays. However, if you set a breakpoint at the end of your
6685 overlay manager (and tell @value{GDBN} which overlays are now mapped, if
6686 you are using manual overlay management), @value{GDBN} will re-set its
6687 breakpoints properly.
6691 @node Automatic Overlay Debugging
6692 @section Automatic Overlay Debugging
6693 @cindex automatic overlay debugging
6695 @value{GDBN} can automatically track which overlays are mapped and which
6696 are not, given some simple co-operation from the overlay manager in the
6697 inferior. If you enable automatic overlay debugging with the
6698 @code{overlay auto} command (@pxref{Overlay Commands}), @value{GDBN}
6699 looks in the inferior's memory for certain variables describing the
6700 current state of the overlays.
6702 Here are the variables your overlay manager must define to support
6703 @value{GDBN}'s automatic overlay debugging:
6707 @item @code{_ovly_table}:
6708 This variable must be an array of the following structures:
6713 /* The overlay's mapped address. */
6716 /* The size of the overlay, in bytes. */
6719 /* The overlay's load address. */
6722 /* Non-zero if the overlay is currently mapped;
6724 unsigned long mapped;
6728 @item @code{_novlys}:
6729 This variable must be a four-byte signed integer, holding the total
6730 number of elements in @code{_ovly_table}.
6734 To decide whether a particular overlay is mapped or not, @value{GDBN}
6735 looks for an entry in @w{@code{_ovly_table}} whose @code{vma} and
6736 @code{lma} members equal the VMA and LMA of the overlay's section in the
6737 executable file. When @value{GDBN} finds a matching entry, it consults
6738 the entry's @code{mapped} member to determine whether the overlay is
6742 @node Overlay Sample Program
6743 @section Overlay Sample Program
6744 @cindex overlay example program
6746 When linking a program which uses overlays, you must place the overlays
6747 at their load addresses, while relocating them to run at their mapped
6748 addresses. To do this, you must write a linker script (@pxref{Overlay
6749 Description,,, ld.info, Using ld: the GNU linker}). Unfortunately,
6750 since linker scripts are specific to a particular host system, target
6751 architecture, and target memory layout, this manual cannot provide
6752 portable sample code demonstrating @value{GDBN}'s overlay support.
6754 However, the @value{GDBN} source distribution does contain an overlaid
6755 program, with linker scripts for a few systems, as part of its test
6756 suite. The program consists of the following files from
6757 @file{gdb/testsuite/gdb.base}:
6761 The main program file.
6763 A simple overlay manager, used by @file{overlays.c}.
6768 Overlay modules, loaded and used by @file{overlays.c}.
6771 Linker scripts for linking the test program on the @code{d10v-elf}
6772 and @code{m32r-elf} targets.
6775 You can build the test program using the @code{d10v-elf} GCC
6776 cross-compiler like this:
6779 $ d10v-elf-gcc -g -c overlays.c
6780 $ d10v-elf-gcc -g -c ovlymgr.c
6781 $ d10v-elf-gcc -g -c foo.c
6782 $ d10v-elf-gcc -g -c bar.c
6783 $ d10v-elf-gcc -g -c baz.c
6784 $ d10v-elf-gcc -g -c grbx.c
6785 $ d10v-elf-gcc -g overlays.o ovlymgr.o foo.o bar.o \
6786 baz.o grbx.o -Wl,-Td10v.ld -o overlays
6789 The build process is identical for any other architecture, except that
6790 you must substitute the appropriate compiler and linker script for the
6791 target system for @code{d10v-elf-gcc} and @code{d10v.ld}.
6795 @chapter Using @value{GDBN} with Different Languages
6798 Although programming languages generally have common aspects, they are
6799 rarely expressed in the same manner. For instance, in ANSI C,
6800 dereferencing a pointer @code{p} is accomplished by @code{*p}, but in
6801 Modula-2, it is accomplished by @code{p^}. Values can also be
6802 represented (and displayed) differently. Hex numbers in C appear as
6803 @samp{0x1ae}, while in Modula-2 they appear as @samp{1AEH}.
6805 @cindex working language
6806 Language-specific information is built into @value{GDBN} for some languages,
6807 allowing you to express operations like the above in your program's
6808 native language, and allowing @value{GDBN} to output values in a manner
6809 consistent with the syntax of your program's native language. The
6810 language you use to build expressions is called the @dfn{working
6814 * Setting:: Switching between source languages
6815 * Show:: Displaying the language
6816 * Checks:: Type and range checks
6817 * Support:: Supported languages
6821 @section Switching between source languages
6823 There are two ways to control the working language---either have @value{GDBN}
6824 set it automatically, or select it manually yourself. You can use the
6825 @code{set language} command for either purpose. On startup, @value{GDBN}
6826 defaults to setting the language automatically. The working language is
6827 used to determine how expressions you type are interpreted, how values
6830 In addition to the working language, every source file that
6831 @value{GDBN} knows about has its own working language. For some object
6832 file formats, the compiler might indicate which language a particular
6833 source file is in. However, most of the time @value{GDBN} infers the
6834 language from the name of the file. The language of a source file
6835 controls whether C@t{++} names are demangled---this way @code{backtrace} can
6836 show each frame appropriately for its own language. There is no way to
6837 set the language of a source file from within @value{GDBN}, but you can
6838 set the language associated with a filename extension. @xref{Show, ,
6839 Displaying the language}.
6841 This is most commonly a problem when you use a program, such
6842 as @code{cfront} or @code{f2c}, that generates C but is written in
6843 another language. In that case, make the
6844 program use @code{#line} directives in its C output; that way
6845 @value{GDBN} will know the correct language of the source code of the original
6846 program, and will display that source code, not the generated C code.
6849 * Filenames:: Filename extensions and languages.
6850 * Manually:: Setting the working language manually
6851 * Automatically:: Having @value{GDBN} infer the source language
6855 @subsection List of filename extensions and languages
6857 If a source file name ends in one of the following extensions, then
6858 @value{GDBN} infers that its language is the one indicated.
6883 Modula-2 source file
6887 Assembler source file. This actually behaves almost like C, but
6888 @value{GDBN} does not skip over function prologues when stepping.
6891 In addition, you may set the language associated with a filename
6892 extension. @xref{Show, , Displaying the language}.
6895 @subsection Setting the working language
6897 If you allow @value{GDBN} to set the language automatically,
6898 expressions are interpreted the same way in your debugging session and
6901 @kindex set language
6902 If you wish, you may set the language manually. To do this, issue the
6903 command @samp{set language @var{lang}}, where @var{lang} is the name of
6905 @code{c} or @code{modula-2}.
6906 For a list of the supported languages, type @samp{set language}.
6908 Setting the language manually prevents @value{GDBN} from updating the working
6909 language automatically. This can lead to confusion if you try
6910 to debug a program when the working language is not the same as the
6911 source language, when an expression is acceptable to both
6912 languages---but means different things. For instance, if the current
6913 source file were written in C, and @value{GDBN} was parsing Modula-2, a
6921 might not have the effect you intended. In C, this means to add
6922 @code{b} and @code{c} and place the result in @code{a}. The result
6923 printed would be the value of @code{a}. In Modula-2, this means to compare
6924 @code{a} to the result of @code{b+c}, yielding a @code{BOOLEAN} value.
6927 @subsection Having @value{GDBN} infer the source language
6929 To have @value{GDBN} set the working language automatically, use
6930 @samp{set language local} or @samp{set language auto}. @value{GDBN}
6931 then infers the working language. That is, when your program stops in a
6932 frame (usually by encountering a breakpoint), @value{GDBN} sets the
6933 working language to the language recorded for the function in that
6934 frame. If the language for a frame is unknown (that is, if the function
6935 or block corresponding to the frame was defined in a source file that
6936 does not have a recognized extension), the current working language is
6937 not changed, and @value{GDBN} issues a warning.
6939 This may not seem necessary for most programs, which are written
6940 entirely in one source language. However, program modules and libraries
6941 written in one source language can be used by a main program written in
6942 a different source language. Using @samp{set language auto} in this
6943 case frees you from having to set the working language manually.
6946 @section Displaying the language
6948 The following commands help you find out which language is the
6949 working language, and also what language source files were written in.
6951 @kindex show language
6952 @kindex info frame@r{, show the source language}
6953 @kindex info source@r{, show the source language}
6956 Display the current working language. This is the
6957 language you can use with commands such as @code{print} to
6958 build and compute expressions that may involve variables in your program.
6961 Display the source language for this frame. This language becomes the
6962 working language if you use an identifier from this frame.
6963 @xref{Frame Info, ,Information about a frame}, to identify the other
6964 information listed here.
6967 Display the source language of this source file.
6968 @xref{Symbols, ,Examining the Symbol Table}, to identify the other
6969 information listed here.
6972 In unusual circumstances, you may have source files with extensions
6973 not in the standard list. You can then set the extension associated
6974 with a language explicitly:
6976 @kindex set extension-language
6977 @kindex info extensions
6979 @item set extension-language @var{.ext} @var{language}
6980 Set source files with extension @var{.ext} to be assumed to be in
6981 the source language @var{language}.
6983 @item info extensions
6984 List all the filename extensions and the associated languages.
6988 @section Type and range checking
6991 @emph{Warning:} In this release, the @value{GDBN} commands for type and range
6992 checking are included, but they do not yet have any effect. This
6993 section documents the intended facilities.
6995 @c FIXME remove warning when type/range code added
6997 Some languages are designed to guard you against making seemingly common
6998 errors through a series of compile- and run-time checks. These include
6999 checking the type of arguments to functions and operators, and making
7000 sure mathematical overflows are caught at run time. Checks such as
7001 these help to ensure a program's correctness once it has been compiled
7002 by eliminating type mismatches, and providing active checks for range
7003 errors when your program is running.
7005 @value{GDBN} can check for conditions like the above if you wish.
7006 Although @value{GDBN} does not check the statements in your program, it
7007 can check expressions entered directly into @value{GDBN} for evaluation via
7008 the @code{print} command, for example. As with the working language,
7009 @value{GDBN} can also decide whether or not to check automatically based on
7010 your program's source language. @xref{Support, ,Supported languages},
7011 for the default settings of supported languages.
7014 * Type Checking:: An overview of type checking
7015 * Range Checking:: An overview of range checking
7018 @cindex type checking
7019 @cindex checks, type
7021 @subsection An overview of type checking
7023 Some languages, such as Modula-2, are strongly typed, meaning that the
7024 arguments to operators and functions have to be of the correct type,
7025 otherwise an error occurs. These checks prevent type mismatch
7026 errors from ever causing any run-time problems. For example,
7034 The second example fails because the @code{CARDINAL} 1 is not
7035 type-compatible with the @code{REAL} 2.3.
7037 For the expressions you use in @value{GDBN} commands, you can tell the
7038 @value{GDBN} type checker to skip checking;
7039 to treat any mismatches as errors and abandon the expression;
7040 or to only issue warnings when type mismatches occur,
7041 but evaluate the expression anyway. When you choose the last of
7042 these, @value{GDBN} evaluates expressions like the second example above, but
7043 also issues a warning.
7045 Even if you turn type checking off, there may be other reasons
7046 related to type that prevent @value{GDBN} from evaluating an expression.
7047 For instance, @value{GDBN} does not know how to add an @code{int} and
7048 a @code{struct foo}. These particular type errors have nothing to do
7049 with the language in use, and usually arise from expressions, such as
7050 the one described above, which make little sense to evaluate anyway.
7052 Each language defines to what degree it is strict about type. For
7053 instance, both Modula-2 and C require the arguments to arithmetical
7054 operators to be numbers. In C, enumerated types and pointers can be
7055 represented as numbers, so that they are valid arguments to mathematical
7056 operators. @xref{Support, ,Supported languages}, for further
7057 details on specific languages.
7059 @value{GDBN} provides some additional commands for controlling the type checker:
7061 @kindex set check@r{, type}
7062 @kindex set check type
7063 @kindex show check type
7065 @item set check type auto
7066 Set type checking on or off based on the current working language.
7067 @xref{Support, ,Supported languages}, for the default settings for
7070 @item set check type on
7071 @itemx set check type off
7072 Set type checking on or off, overriding the default setting for the
7073 current working language. Issue a warning if the setting does not
7074 match the language default. If any type mismatches occur in
7075 evaluating an expression while type checking is on, @value{GDBN} prints a
7076 message and aborts evaluation of the expression.
7078 @item set check type warn
7079 Cause the type checker to issue warnings, but to always attempt to
7080 evaluate the expression. Evaluating the expression may still
7081 be impossible for other reasons. For example, @value{GDBN} cannot add
7082 numbers and structures.
7085 Show the current setting of the type checker, and whether or not @value{GDBN}
7086 is setting it automatically.
7089 @cindex range checking
7090 @cindex checks, range
7091 @node Range Checking
7092 @subsection An overview of range checking
7094 In some languages (such as Modula-2), it is an error to exceed the
7095 bounds of a type; this is enforced with run-time checks. Such range
7096 checking is meant to ensure program correctness by making sure
7097 computations do not overflow, or indices on an array element access do
7098 not exceed the bounds of the array.
7100 For expressions you use in @value{GDBN} commands, you can tell
7101 @value{GDBN} to treat range errors in one of three ways: ignore them,
7102 always treat them as errors and abandon the expression, or issue
7103 warnings but evaluate the expression anyway.
7105 A range error can result from numerical overflow, from exceeding an
7106 array index bound, or when you type a constant that is not a member
7107 of any type. Some languages, however, do not treat overflows as an
7108 error. In many implementations of C, mathematical overflow causes the
7109 result to ``wrap around'' to lower values---for example, if @var{m} is
7110 the largest integer value, and @var{s} is the smallest, then
7113 @var{m} + 1 @result{} @var{s}
7116 This, too, is specific to individual languages, and in some cases
7117 specific to individual compilers or machines. @xref{Support, ,
7118 Supported languages}, for further details on specific languages.
7120 @value{GDBN} provides some additional commands for controlling the range checker:
7122 @kindex set check@r{, range}
7123 @kindex set check range
7124 @kindex show check range
7126 @item set check range auto
7127 Set range checking on or off based on the current working language.
7128 @xref{Support, ,Supported languages}, for the default settings for
7131 @item set check range on
7132 @itemx set check range off
7133 Set range checking on or off, overriding the default setting for the
7134 current working language. A warning is issued if the setting does not
7135 match the language default. If a range error occurs and range checking is on,
7136 then a message is printed and evaluation of the expression is aborted.
7138 @item set check range warn
7139 Output messages when the @value{GDBN} range checker detects a range error,
7140 but attempt to evaluate the expression anyway. Evaluating the
7141 expression may still be impossible for other reasons, such as accessing
7142 memory that the process does not own (a typical example from many Unix
7146 Show the current setting of the range checker, and whether or not it is
7147 being set automatically by @value{GDBN}.
7151 @section Supported languages
7153 @value{GDBN} supports C, C@t{++}, Fortran, Java, Chill, assembly, and Modula-2.
7154 @c This is false ...
7155 Some @value{GDBN} features may be used in expressions regardless of the
7156 language you use: the @value{GDBN} @code{@@} and @code{::} operators,
7157 and the @samp{@{type@}addr} construct (@pxref{Expressions,
7158 ,Expressions}) can be used with the constructs of any supported
7161 The following sections detail to what degree each source language is
7162 supported by @value{GDBN}. These sections are not meant to be language
7163 tutorials or references, but serve only as a reference guide to what the
7164 @value{GDBN} expression parser accepts, and what input and output
7165 formats should look like for different languages. There are many good
7166 books written on each of these languages; please look to these for a
7167 language reference or tutorial.
7171 * Modula-2:: Modula-2
7176 @subsection C and C@t{++}
7178 @cindex C and C@t{++}
7179 @cindex expressions in C or C@t{++}
7181 Since C and C@t{++} are so closely related, many features of @value{GDBN} apply
7182 to both languages. Whenever this is the case, we discuss those languages
7186 @cindex @code{g++}, @sc{gnu} C@t{++} compiler
7187 @cindex @sc{gnu} C@t{++}
7188 The C@t{++} debugging facilities are jointly implemented by the C@t{++}
7189 compiler and @value{GDBN}. Therefore, to debug your C@t{++} code
7190 effectively, you must compile your C@t{++} programs with a supported
7191 C@t{++} compiler, such as @sc{gnu} @code{g++}, or the HP ANSI C@t{++}
7192 compiler (@code{aCC}).
7194 For best results when using @sc{gnu} C@t{++}, use the stabs debugging
7195 format. You can select that format explicitly with the @code{g++}
7196 command-line options @samp{-gstabs} or @samp{-gstabs+}. See
7197 @ref{Debugging Options,,Options for Debugging Your Program or @sc{gnu}
7198 CC, gcc.info, Using @sc{gnu} CC}, for more information.
7201 * C Operators:: C and C@t{++} operators
7202 * C Constants:: C and C@t{++} constants
7203 * C plus plus expressions:: C@t{++} expressions
7204 * C Defaults:: Default settings for C and C@t{++}
7205 * C Checks:: C and C@t{++} type and range checks
7206 * Debugging C:: @value{GDBN} and C
7207 * Debugging C plus plus:: @value{GDBN} features for C@t{++}
7211 @subsubsection C and C@t{++} operators
7213 @cindex C and C@t{++} operators
7215 Operators must be defined on values of specific types. For instance,
7216 @code{+} is defined on numbers, but not on structures. Operators are
7217 often defined on groups of types.
7219 For the purposes of C and C@t{++}, the following definitions hold:
7224 @emph{Integral types} include @code{int} with any of its storage-class
7225 specifiers; @code{char}; @code{enum}; and, for C@t{++}, @code{bool}.
7228 @emph{Floating-point types} include @code{float}, @code{double}, and
7229 @code{long double} (if supported by the target platform).
7232 @emph{Pointer types} include all types defined as @code{(@var{type} *)}.
7235 @emph{Scalar types} include all of the above.
7240 The following operators are supported. They are listed here
7241 in order of increasing precedence:
7245 The comma or sequencing operator. Expressions in a comma-separated list
7246 are evaluated from left to right, with the result of the entire
7247 expression being the last expression evaluated.
7250 Assignment. The value of an assignment expression is the value
7251 assigned. Defined on scalar types.
7254 Used in an expression of the form @w{@code{@var{a} @var{op}= @var{b}}},
7255 and translated to @w{@code{@var{a} = @var{a op b}}}.
7256 @w{@code{@var{op}=}} and @code{=} have the same precedence.
7257 @var{op} is any one of the operators @code{|}, @code{^}, @code{&},
7258 @code{<<}, @code{>>}, @code{+}, @code{-}, @code{*}, @code{/}, @code{%}.
7261 The ternary operator. @code{@var{a} ? @var{b} : @var{c}} can be thought
7262 of as: if @var{a} then @var{b} else @var{c}. @var{a} should be of an
7266 Logical @sc{or}. Defined on integral types.
7269 Logical @sc{and}. Defined on integral types.
7272 Bitwise @sc{or}. Defined on integral types.
7275 Bitwise exclusive-@sc{or}. Defined on integral types.
7278 Bitwise @sc{and}. Defined on integral types.
7281 Equality and inequality. Defined on scalar types. The value of these
7282 expressions is 0 for false and non-zero for true.
7284 @item <@r{, }>@r{, }<=@r{, }>=
7285 Less than, greater than, less than or equal, greater than or equal.
7286 Defined on scalar types. The value of these expressions is 0 for false
7287 and non-zero for true.
7290 left shift, and right shift. Defined on integral types.
7293 The @value{GDBN} ``artificial array'' operator (@pxref{Expressions, ,Expressions}).
7296 Addition and subtraction. Defined on integral types, floating-point types and
7299 @item *@r{, }/@r{, }%
7300 Multiplication, division, and modulus. Multiplication and division are
7301 defined on integral and floating-point types. Modulus is defined on
7305 Increment and decrement. When appearing before a variable, the
7306 operation is performed before the variable is used in an expression;
7307 when appearing after it, the variable's value is used before the
7308 operation takes place.
7311 Pointer dereferencing. Defined on pointer types. Same precedence as
7315 Address operator. Defined on variables. Same precedence as @code{++}.
7317 For debugging C@t{++}, @value{GDBN} implements a use of @samp{&} beyond what is
7318 allowed in the C@t{++} language itself: you can use @samp{&(&@var{ref})}
7319 (or, if you prefer, simply @samp{&&@var{ref}}) to examine the address
7320 where a C@t{++} reference variable (declared with @samp{&@var{ref}}) is
7324 Negative. Defined on integral and floating-point types. Same
7325 precedence as @code{++}.
7328 Logical negation. Defined on integral types. Same precedence as
7332 Bitwise complement operator. Defined on integral types. Same precedence as
7337 Structure member, and pointer-to-structure member. For convenience,
7338 @value{GDBN} regards the two as equivalent, choosing whether to dereference a
7339 pointer based on the stored type information.
7340 Defined on @code{struct} and @code{union} data.
7343 Dereferences of pointers to members.
7346 Array indexing. @code{@var{a}[@var{i}]} is defined as
7347 @code{*(@var{a}+@var{i})}. Same precedence as @code{->}.
7350 Function parameter list. Same precedence as @code{->}.
7353 C@t{++} scope resolution operator. Defined on @code{struct}, @code{union},
7354 and @code{class} types.
7357 Doubled colons also represent the @value{GDBN} scope operator
7358 (@pxref{Expressions, ,Expressions}). Same precedence as @code{::},
7362 If an operator is redefined in the user code, @value{GDBN} usually
7363 attempts to invoke the redefined version instead of using the operator's
7371 @subsubsection C and C@t{++} constants
7373 @cindex C and C@t{++} constants
7375 @value{GDBN} allows you to express the constants of C and C@t{++} in the
7380 Integer constants are a sequence of digits. Octal constants are
7381 specified by a leading @samp{0} (i.e. zero), and hexadecimal constants by
7382 a leading @samp{0x} or @samp{0X}. Constants may also end with a letter
7383 @samp{l}, specifying that the constant should be treated as a
7387 Floating point constants are a sequence of digits, followed by a decimal
7388 point, followed by a sequence of digits, and optionally followed by an
7389 exponent. An exponent is of the form:
7390 @samp{@w{e@r{[[}+@r{]|}-@r{]}@var{nnn}}}, where @var{nnn} is another
7391 sequence of digits. The @samp{+} is optional for positive exponents.
7392 A floating-point constant may also end with a letter @samp{f} or
7393 @samp{F}, specifying that the constant should be treated as being of
7394 the @code{float} (as opposed to the default @code{double}) type; or with
7395 a letter @samp{l} or @samp{L}, which specifies a @code{long double}
7399 Enumerated constants consist of enumerated identifiers, or their
7400 integral equivalents.
7403 Character constants are a single character surrounded by single quotes
7404 (@code{'}), or a number---the ordinal value of the corresponding character
7405 (usually its @sc{ascii} value). Within quotes, the single character may
7406 be represented by a letter or by @dfn{escape sequences}, which are of
7407 the form @samp{\@var{nnn}}, where @var{nnn} is the octal representation
7408 of the character's ordinal value; or of the form @samp{\@var{x}}, where
7409 @samp{@var{x}} is a predefined special character---for example,
7410 @samp{\n} for newline.
7413 String constants are a sequence of character constants surrounded by
7414 double quotes (@code{"}). Any valid character constant (as described
7415 above) may appear. Double quotes within the string must be preceded by
7416 a backslash, so for instance @samp{"a\"b'c"} is a string of five
7420 Pointer constants are an integral value. You can also write pointers
7421 to constants using the C operator @samp{&}.
7424 Array constants are comma-separated lists surrounded by braces @samp{@{}
7425 and @samp{@}}; for example, @samp{@{1,2,3@}} is a three-element array of
7426 integers, @samp{@{@{1,2@}, @{3,4@}, @{5,6@}@}} is a three-by-two array,
7427 and @samp{@{&"hi", &"there", &"fred"@}} is a three-element array of pointers.
7431 * C plus plus expressions::
7438 @node C plus plus expressions
7439 @subsubsection C@t{++} expressions
7441 @cindex expressions in C@t{++}
7442 @value{GDBN} expression handling can interpret most C@t{++} expressions.
7444 @cindex C@t{++} support, not in @sc{coff}
7445 @cindex @sc{coff} versus C@t{++}
7446 @cindex C@t{++} and object formats
7447 @cindex object formats and C@t{++}
7448 @cindex a.out and C@t{++}
7449 @cindex @sc{ecoff} and C@t{++}
7450 @cindex @sc{xcoff} and C@t{++}
7451 @cindex @sc{elf}/stabs and C@t{++}
7452 @cindex @sc{elf}/@sc{dwarf} and C@t{++}
7453 @c FIXME!! GDB may eventually be able to debug C++ using DWARF; check
7454 @c periodically whether this has happened...
7456 @emph{Warning:} @value{GDBN} can only debug C@t{++} code if you use the
7457 proper compiler. Typically, C@t{++} debugging depends on the use of
7458 additional debugging information in the symbol table, and thus requires
7459 special support. In particular, if your compiler generates a.out, MIPS
7460 @sc{ecoff}, RS/6000 @sc{xcoff}, or @sc{elf} with stabs extensions to the
7461 symbol table, these facilities are all available. (With @sc{gnu} CC,
7462 you can use the @samp{-gstabs} option to request stabs debugging
7463 extensions explicitly.) Where the object code format is standard
7464 @sc{coff} or @sc{dwarf} in @sc{elf}, on the other hand, most of the C@t{++}
7465 support in @value{GDBN} does @emph{not} work.
7470 @cindex member functions
7472 Member function calls are allowed; you can use expressions like
7475 count = aml->GetOriginal(x, y)
7478 @vindex this@r{, inside C@t{++} member functions}
7479 @cindex namespace in C@t{++}
7481 While a member function is active (in the selected stack frame), your
7482 expressions have the same namespace available as the member function;
7483 that is, @value{GDBN} allows implicit references to the class instance
7484 pointer @code{this} following the same rules as C@t{++}.
7486 @cindex call overloaded functions
7487 @cindex overloaded functions, calling
7488 @cindex type conversions in C@t{++}
7490 You can call overloaded functions; @value{GDBN} resolves the function
7491 call to the right definition, with some restrictions. @value{GDBN} does not
7492 perform overload resolution involving user-defined type conversions,
7493 calls to constructors, or instantiations of templates that do not exist
7494 in the program. It also cannot handle ellipsis argument lists or
7497 It does perform integral conversions and promotions, floating-point
7498 promotions, arithmetic conversions, pointer conversions, conversions of
7499 class objects to base classes, and standard conversions such as those of
7500 functions or arrays to pointers; it requires an exact match on the
7501 number of function arguments.
7503 Overload resolution is always performed, unless you have specified
7504 @code{set overload-resolution off}. @xref{Debugging C plus plus,
7505 ,@value{GDBN} features for C@t{++}}.
7507 You must specify @code{set overload-resolution off} in order to use an
7508 explicit function signature to call an overloaded function, as in
7510 p 'foo(char,int)'('x', 13)
7513 The @value{GDBN} command-completion facility can simplify this;
7514 see @ref{Completion, ,Command completion}.
7516 @cindex reference declarations
7518 @value{GDBN} understands variables declared as C@t{++} references; you can use
7519 them in expressions just as you do in C@t{++} source---they are automatically
7522 In the parameter list shown when @value{GDBN} displays a frame, the values of
7523 reference variables are not displayed (unlike other variables); this
7524 avoids clutter, since references are often used for large structures.
7525 The @emph{address} of a reference variable is always shown, unless
7526 you have specified @samp{set print address off}.
7529 @value{GDBN} supports the C@t{++} name resolution operator @code{::}---your
7530 expressions can use it just as expressions in your program do. Since
7531 one scope may be defined in another, you can use @code{::} repeatedly if
7532 necessary, for example in an expression like
7533 @samp{@var{scope1}::@var{scope2}::@var{name}}. @value{GDBN} also allows
7534 resolving name scope by reference to source files, in both C and C@t{++}
7535 debugging (@pxref{Variables, ,Program variables}).
7538 In addition, when used with HP's C@t{++} compiler, @value{GDBN} supports
7539 calling virtual functions correctly, printing out virtual bases of
7540 objects, calling functions in a base subobject, casting objects, and
7541 invoking user-defined operators.
7544 @subsubsection C and C@t{++} defaults
7546 @cindex C and C@t{++} defaults
7548 If you allow @value{GDBN} to set type and range checking automatically, they
7549 both default to @code{off} whenever the working language changes to
7550 C or C@t{++}. This happens regardless of whether you or @value{GDBN}
7551 selects the working language.
7553 If you allow @value{GDBN} to set the language automatically, it
7554 recognizes source files whose names end with @file{.c}, @file{.C}, or
7555 @file{.cc}, etc, and when @value{GDBN} enters code compiled from one of
7556 these files, it sets the working language to C or C@t{++}.
7557 @xref{Automatically, ,Having @value{GDBN} infer the source language},
7558 for further details.
7560 @c Type checking is (a) primarily motivated by Modula-2, and (b)
7561 @c unimplemented. If (b) changes, it might make sense to let this node
7562 @c appear even if Mod-2 does not, but meanwhile ignore it. roland 16jul93.
7565 @subsubsection C and C@t{++} type and range checks
7567 @cindex C and C@t{++} checks
7569 By default, when @value{GDBN} parses C or C@t{++} expressions, type checking
7570 is not used. However, if you turn type checking on, @value{GDBN}
7571 considers two variables type equivalent if:
7575 The two variables are structured and have the same structure, union, or
7579 The two variables have the same type name, or types that have been
7580 declared equivalent through @code{typedef}.
7583 @c leaving this out because neither J Gilmore nor R Pesch understand it.
7586 The two @code{struct}, @code{union}, or @code{enum} variables are
7587 declared in the same declaration. (Note: this may not be true for all C
7592 Range checking, if turned on, is done on mathematical operations. Array
7593 indices are not checked, since they are often used to index a pointer
7594 that is not itself an array.
7597 @subsubsection @value{GDBN} and C
7599 The @code{set print union} and @code{show print union} commands apply to
7600 the @code{union} type. When set to @samp{on}, any @code{union} that is
7601 inside a @code{struct} or @code{class} is also printed. Otherwise, it
7602 appears as @samp{@{...@}}.
7604 The @code{@@} operator aids in the debugging of dynamic arrays, formed
7605 with pointers and a memory allocation function. @xref{Expressions,
7609 * Debugging C plus plus::
7612 @node Debugging C plus plus
7613 @subsubsection @value{GDBN} features for C@t{++}
7615 @cindex commands for C@t{++}
7617 Some @value{GDBN} commands are particularly useful with C@t{++}, and some are
7618 designed specifically for use with C@t{++}. Here is a summary:
7621 @cindex break in overloaded functions
7622 @item @r{breakpoint menus}
7623 When you want a breakpoint in a function whose name is overloaded,
7624 @value{GDBN} breakpoint menus help you specify which function definition
7625 you want. @xref{Breakpoint Menus,,Breakpoint menus}.
7627 @cindex overloading in C@t{++}
7628 @item rbreak @var{regex}
7629 Setting breakpoints using regular expressions is helpful for setting
7630 breakpoints on overloaded functions that are not members of any special
7632 @xref{Set Breaks, ,Setting breakpoints}.
7634 @cindex C@t{++} exception handling
7637 Debug C@t{++} exception handling using these commands. @xref{Set
7638 Catchpoints, , Setting catchpoints}.
7641 @item ptype @var{typename}
7642 Print inheritance relationships as well as other information for type
7644 @xref{Symbols, ,Examining the Symbol Table}.
7646 @cindex C@t{++} symbol display
7647 @item set print demangle
7648 @itemx show print demangle
7649 @itemx set print asm-demangle
7650 @itemx show print asm-demangle
7651 Control whether C@t{++} symbols display in their source form, both when
7652 displaying code as C@t{++} source and when displaying disassemblies.
7653 @xref{Print Settings, ,Print settings}.
7655 @item set print object
7656 @itemx show print object
7657 Choose whether to print derived (actual) or declared types of objects.
7658 @xref{Print Settings, ,Print settings}.
7660 @item set print vtbl
7661 @itemx show print vtbl
7662 Control the format for printing virtual function tables.
7663 @xref{Print Settings, ,Print settings}.
7664 (The @code{vtbl} commands do not work on programs compiled with the HP
7665 ANSI C@t{++} compiler (@code{aCC}).)
7667 @kindex set overload-resolution
7668 @cindex overloaded functions, overload resolution
7669 @item set overload-resolution on
7670 Enable overload resolution for C@t{++} expression evaluation. The default
7671 is on. For overloaded functions, @value{GDBN} evaluates the arguments
7672 and searches for a function whose signature matches the argument types,
7673 using the standard C@t{++} conversion rules (see @ref{C plus plus expressions, ,C@t{++}
7674 expressions}, for details). If it cannot find a match, it emits a
7677 @item set overload-resolution off
7678 Disable overload resolution for C@t{++} expression evaluation. For
7679 overloaded functions that are not class member functions, @value{GDBN}
7680 chooses the first function of the specified name that it finds in the
7681 symbol table, whether or not its arguments are of the correct type. For
7682 overloaded functions that are class member functions, @value{GDBN}
7683 searches for a function whose signature @emph{exactly} matches the
7686 @item @r{Overloaded symbol names}
7687 You can specify a particular definition of an overloaded symbol, using
7688 the same notation that is used to declare such symbols in C@t{++}: type
7689 @code{@var{symbol}(@var{types})} rather than just @var{symbol}. You can
7690 also use the @value{GDBN} command-line word completion facilities to list the
7691 available choices, or to finish the type list for you.
7692 @xref{Completion,, Command completion}, for details on how to do this.
7696 @subsection Modula-2
7698 @cindex Modula-2, @value{GDBN} support
7700 The extensions made to @value{GDBN} to support Modula-2 only support
7701 output from the @sc{gnu} Modula-2 compiler (which is currently being
7702 developed). Other Modula-2 compilers are not currently supported, and
7703 attempting to debug executables produced by them is most likely
7704 to give an error as @value{GDBN} reads in the executable's symbol
7707 @cindex expressions in Modula-2
7709 * M2 Operators:: Built-in operators
7710 * Built-In Func/Proc:: Built-in functions and procedures
7711 * M2 Constants:: Modula-2 constants
7712 * M2 Defaults:: Default settings for Modula-2
7713 * Deviations:: Deviations from standard Modula-2
7714 * M2 Checks:: Modula-2 type and range checks
7715 * M2 Scope:: The scope operators @code{::} and @code{.}
7716 * GDB/M2:: @value{GDBN} and Modula-2
7720 @subsubsection Operators
7721 @cindex Modula-2 operators
7723 Operators must be defined on values of specific types. For instance,
7724 @code{+} is defined on numbers, but not on structures. Operators are
7725 often defined on groups of types. For the purposes of Modula-2, the
7726 following definitions hold:
7731 @emph{Integral types} consist of @code{INTEGER}, @code{CARDINAL}, and
7735 @emph{Character types} consist of @code{CHAR} and its subranges.
7738 @emph{Floating-point types} consist of @code{REAL}.
7741 @emph{Pointer types} consist of anything declared as @code{POINTER TO
7745 @emph{Scalar types} consist of all of the above.
7748 @emph{Set types} consist of @code{SET} and @code{BITSET} types.
7751 @emph{Boolean types} consist of @code{BOOLEAN}.
7755 The following operators are supported, and appear in order of
7756 increasing precedence:
7760 Function argument or array index separator.
7763 Assignment. The value of @var{var} @code{:=} @var{value} is
7767 Less than, greater than on integral, floating-point, or enumerated
7771 Less than or equal to, greater than or equal to
7772 on integral, floating-point and enumerated types, or set inclusion on
7773 set types. Same precedence as @code{<}.
7775 @item =@r{, }<>@r{, }#
7776 Equality and two ways of expressing inequality, valid on scalar types.
7777 Same precedence as @code{<}. In @value{GDBN} scripts, only @code{<>} is
7778 available for inequality, since @code{#} conflicts with the script
7782 Set membership. Defined on set types and the types of their members.
7783 Same precedence as @code{<}.
7786 Boolean disjunction. Defined on boolean types.
7789 Boolean conjunction. Defined on boolean types.
7792 The @value{GDBN} ``artificial array'' operator (@pxref{Expressions, ,Expressions}).
7795 Addition and subtraction on integral and floating-point types, or union
7796 and difference on set types.
7799 Multiplication on integral and floating-point types, or set intersection
7803 Division on floating-point types, or symmetric set difference on set
7804 types. Same precedence as @code{*}.
7807 Integer division and remainder. Defined on integral types. Same
7808 precedence as @code{*}.
7811 Negative. Defined on @code{INTEGER} and @code{REAL} data.
7814 Pointer dereferencing. Defined on pointer types.
7817 Boolean negation. Defined on boolean types. Same precedence as
7821 @code{RECORD} field selector. Defined on @code{RECORD} data. Same
7822 precedence as @code{^}.
7825 Array indexing. Defined on @code{ARRAY} data. Same precedence as @code{^}.
7828 Procedure argument list. Defined on @code{PROCEDURE} objects. Same precedence
7832 @value{GDBN} and Modula-2 scope operators.
7836 @emph{Warning:} Sets and their operations are not yet supported, so @value{GDBN}
7837 treats the use of the operator @code{IN}, or the use of operators
7838 @code{+}, @code{-}, @code{*}, @code{/}, @code{=}, , @code{<>}, @code{#},
7839 @code{<=}, and @code{>=} on sets as an error.
7843 @node Built-In Func/Proc
7844 @subsubsection Built-in functions and procedures
7845 @cindex Modula-2 built-ins
7847 Modula-2 also makes available several built-in procedures and functions.
7848 In describing these, the following metavariables are used:
7853 represents an @code{ARRAY} variable.
7856 represents a @code{CHAR} constant or variable.
7859 represents a variable or constant of integral type.
7862 represents an identifier that belongs to a set. Generally used in the
7863 same function with the metavariable @var{s}. The type of @var{s} should
7864 be @code{SET OF @var{mtype}} (where @var{mtype} is the type of @var{m}).
7867 represents a variable or constant of integral or floating-point type.
7870 represents a variable or constant of floating-point type.
7876 represents a variable.
7879 represents a variable or constant of one of many types. See the
7880 explanation of the function for details.
7883 All Modula-2 built-in procedures also return a result, described below.
7887 Returns the absolute value of @var{n}.
7890 If @var{c} is a lower case letter, it returns its upper case
7891 equivalent, otherwise it returns its argument.
7894 Returns the character whose ordinal value is @var{i}.
7897 Decrements the value in the variable @var{v} by one. Returns the new value.
7899 @item DEC(@var{v},@var{i})
7900 Decrements the value in the variable @var{v} by @var{i}. Returns the
7903 @item EXCL(@var{m},@var{s})
7904 Removes the element @var{m} from the set @var{s}. Returns the new
7907 @item FLOAT(@var{i})
7908 Returns the floating point equivalent of the integer @var{i}.
7911 Returns the index of the last member of @var{a}.
7914 Increments the value in the variable @var{v} by one. Returns the new value.
7916 @item INC(@var{v},@var{i})
7917 Increments the value in the variable @var{v} by @var{i}. Returns the
7920 @item INCL(@var{m},@var{s})
7921 Adds the element @var{m} to the set @var{s} if it is not already
7922 there. Returns the new set.
7925 Returns the maximum value of the type @var{t}.
7928 Returns the minimum value of the type @var{t}.
7931 Returns boolean TRUE if @var{i} is an odd number.
7934 Returns the ordinal value of its argument. For example, the ordinal
7935 value of a character is its @sc{ascii} value (on machines supporting the
7936 @sc{ascii} character set). @var{x} must be of an ordered type, which include
7937 integral, character and enumerated types.
7940 Returns the size of its argument. @var{x} can be a variable or a type.
7942 @item TRUNC(@var{r})
7943 Returns the integral part of @var{r}.
7945 @item VAL(@var{t},@var{i})
7946 Returns the member of the type @var{t} whose ordinal value is @var{i}.
7950 @emph{Warning:} Sets and their operations are not yet supported, so
7951 @value{GDBN} treats the use of procedures @code{INCL} and @code{EXCL} as
7955 @cindex Modula-2 constants
7957 @subsubsection Constants
7959 @value{GDBN} allows you to express the constants of Modula-2 in the following
7965 Integer constants are simply a sequence of digits. When used in an
7966 expression, a constant is interpreted to be type-compatible with the
7967 rest of the expression. Hexadecimal integers are specified by a
7968 trailing @samp{H}, and octal integers by a trailing @samp{B}.
7971 Floating point constants appear as a sequence of digits, followed by a
7972 decimal point and another sequence of digits. An optional exponent can
7973 then be specified, in the form @samp{E@r{[}+@r{|}-@r{]}@var{nnn}}, where
7974 @samp{@r{[}+@r{|}-@r{]}@var{nnn}} is the desired exponent. All of the
7975 digits of the floating point constant must be valid decimal (base 10)
7979 Character constants consist of a single character enclosed by a pair of
7980 like quotes, either single (@code{'}) or double (@code{"}). They may
7981 also be expressed by their ordinal value (their @sc{ascii} value, usually)
7982 followed by a @samp{C}.
7985 String constants consist of a sequence of characters enclosed by a
7986 pair of like quotes, either single (@code{'}) or double (@code{"}).
7987 Escape sequences in the style of C are also allowed. @xref{C
7988 Constants, ,C and C@t{++} constants}, for a brief explanation of escape
7992 Enumerated constants consist of an enumerated identifier.
7995 Boolean constants consist of the identifiers @code{TRUE} and
7999 Pointer constants consist of integral values only.
8002 Set constants are not yet supported.
8006 @subsubsection Modula-2 defaults
8007 @cindex Modula-2 defaults
8009 If type and range checking are set automatically by @value{GDBN}, they
8010 both default to @code{on} whenever the working language changes to
8011 Modula-2. This happens regardless of whether you or @value{GDBN}
8012 selected the working language.
8014 If you allow @value{GDBN} to set the language automatically, then entering
8015 code compiled from a file whose name ends with @file{.mod} sets the
8016 working language to Modula-2. @xref{Automatically, ,Having @value{GDBN} set
8017 the language automatically}, for further details.
8020 @subsubsection Deviations from standard Modula-2
8021 @cindex Modula-2, deviations from
8023 A few changes have been made to make Modula-2 programs easier to debug.
8024 This is done primarily via loosening its type strictness:
8028 Unlike in standard Modula-2, pointer constants can be formed by
8029 integers. This allows you to modify pointer variables during
8030 debugging. (In standard Modula-2, the actual address contained in a
8031 pointer variable is hidden from you; it can only be modified
8032 through direct assignment to another pointer variable or expression that
8033 returned a pointer.)
8036 C escape sequences can be used in strings and characters to represent
8037 non-printable characters. @value{GDBN} prints out strings with these
8038 escape sequences embedded. Single non-printable characters are
8039 printed using the @samp{CHR(@var{nnn})} format.
8042 The assignment operator (@code{:=}) returns the value of its right-hand
8046 All built-in procedures both modify @emph{and} return their argument.
8050 @subsubsection Modula-2 type and range checks
8051 @cindex Modula-2 checks
8054 @emph{Warning:} in this release, @value{GDBN} does not yet perform type or
8057 @c FIXME remove warning when type/range checks added
8059 @value{GDBN} considers two Modula-2 variables type equivalent if:
8063 They are of types that have been declared equivalent via a @code{TYPE
8064 @var{t1} = @var{t2}} statement
8067 They have been declared on the same line. (Note: This is true of the
8068 @sc{gnu} Modula-2 compiler, but it may not be true of other compilers.)
8071 As long as type checking is enabled, any attempt to combine variables
8072 whose types are not equivalent is an error.
8074 Range checking is done on all mathematical operations, assignment, array
8075 index bounds, and all built-in functions and procedures.
8078 @subsubsection The scope operators @code{::} and @code{.}
8080 @cindex @code{.}, Modula-2 scope operator
8081 @cindex colon, doubled as scope operator
8083 @vindex colon-colon@r{, in Modula-2}
8084 @c Info cannot handle :: but TeX can.
8087 @vindex ::@r{, in Modula-2}
8090 There are a few subtle differences between the Modula-2 scope operator
8091 (@code{.}) and the @value{GDBN} scope operator (@code{::}). The two have
8096 @var{module} . @var{id}
8097 @var{scope} :: @var{id}
8101 where @var{scope} is the name of a module or a procedure,
8102 @var{module} the name of a module, and @var{id} is any declared
8103 identifier within your program, except another module.
8105 Using the @code{::} operator makes @value{GDBN} search the scope
8106 specified by @var{scope} for the identifier @var{id}. If it is not
8107 found in the specified scope, then @value{GDBN} searches all scopes
8108 enclosing the one specified by @var{scope}.
8110 Using the @code{.} operator makes @value{GDBN} search the current scope for
8111 the identifier specified by @var{id} that was imported from the
8112 definition module specified by @var{module}. With this operator, it is
8113 an error if the identifier @var{id} was not imported from definition
8114 module @var{module}, or if @var{id} is not an identifier in
8118 @subsubsection @value{GDBN} and Modula-2
8120 Some @value{GDBN} commands have little use when debugging Modula-2 programs.
8121 Five subcommands of @code{set print} and @code{show print} apply
8122 specifically to C and C@t{++}: @samp{vtbl}, @samp{demangle},
8123 @samp{asm-demangle}, @samp{object}, and @samp{union}. The first four
8124 apply to C@t{++}, and the last to the C @code{union} type, which has no direct
8125 analogue in Modula-2.
8127 The @code{@@} operator (@pxref{Expressions, ,Expressions}), while available
8128 with any language, is not useful with Modula-2. Its
8129 intent is to aid the debugging of @dfn{dynamic arrays}, which cannot be
8130 created in Modula-2 as they can in C or C@t{++}. However, because an
8131 address can be specified by an integral constant, the construct
8132 @samp{@{@var{type}@}@var{adrexp}} is still useful.
8134 @cindex @code{#} in Modula-2
8135 In @value{GDBN} scripts, the Modula-2 inequality operator @code{#} is
8136 interpreted as the beginning of a comment. Use @code{<>} instead.
8141 The extensions made to @value{GDBN} to support Chill only support output
8142 from the @sc{gnu} Chill compiler. Other Chill compilers are not currently
8143 supported, and attempting to debug executables produced by them is most
8144 likely to give an error as @value{GDBN} reads in the executable's symbol
8147 @c This used to say "... following Chill related topics ...", but since
8148 @c menus are not shown in the printed manual, it would look awkward.
8149 This section covers the Chill related topics and the features
8150 of @value{GDBN} which support these topics.
8153 * How modes are displayed:: How modes are displayed
8154 * Locations:: Locations and their accesses
8155 * Values and their Operations:: Values and their Operations
8156 * Chill type and range checks::
8160 @node How modes are displayed
8161 @subsubsection How modes are displayed
8163 The Chill Datatype- (Mode) support of @value{GDBN} is directly related
8164 with the functionality of the @sc{gnu} Chill compiler, and therefore deviates
8165 slightly from the standard specification of the Chill language. The
8168 @c FIXME: this @table's contents effectively disable @code by using @r
8169 @c on every @item. So why does it need @code?
8171 @item @r{@emph{Discrete modes:}}
8174 @emph{Integer Modes} which are predefined by @code{BYTE, UBYTE, INT,
8177 @emph{Boolean Mode} which is predefined by @code{BOOL},
8179 @emph{Character Mode} which is predefined by @code{CHAR},
8181 @emph{Set Mode} which is displayed by the keyword @code{SET}.
8183 (@value{GDBP}) ptype x
8184 type = SET (karli = 10, susi = 20, fritzi = 100)
8186 If the type is an unnumbered set the set element values are omitted.
8188 @emph{Range Mode} which is displayed by
8190 @code{type = <basemode>(<lower bound> : <upper bound>)}
8192 where @code{<lower bound>, <upper bound>} can be of any discrete literal
8193 expression (e.g. set element names).
8196 @item @r{@emph{Powerset Mode:}}
8197 A Powerset Mode is displayed by the keyword @code{POWERSET} followed by
8198 the member mode of the powerset. The member mode can be any discrete mode.
8200 (@value{GDBP}) ptype x
8201 type = POWERSET SET (egon, hugo, otto)
8204 @item @r{@emph{Reference Modes:}}
8207 @emph{Bound Reference Mode} which is displayed by the keyword @code{REF}
8208 followed by the mode name to which the reference is bound.
8210 @emph{Free Reference Mode} which is displayed by the keyword @code{PTR}.
8213 @item @r{@emph{Procedure mode}}
8214 The procedure mode is displayed by @code{type = PROC(<parameter list>)
8215 <return mode> EXCEPTIONS (<exception list>)}. The @code{<parameter
8216 list>} is a list of the parameter modes. @code{<return mode>} indicates
8217 the mode of the result of the procedure if any. The exceptionlist lists
8218 all possible exceptions which can be raised by the procedure.
8221 @item @r{@emph{Instance mode}}
8222 The instance mode is represented by a structure, which has a static
8223 type, and is therefore not really of interest.
8226 @item @r{@emph{Synchronization Modes:}}
8229 @emph{Event Mode} which is displayed by
8231 @code{EVENT (<event length>)}
8233 where @code{(<event length>)} is optional.
8235 @emph{Buffer Mode} which is displayed by
8237 @code{BUFFER (<buffer length>)<buffer element mode>}
8239 where @code{(<buffer length>)} is optional.
8242 @item @r{@emph{Timing Modes:}}
8245 @emph{Duration Mode} which is predefined by @code{DURATION}
8247 @emph{Absolute Time Mode} which is predefined by @code{TIME}
8250 @item @r{@emph{Real Modes:}}
8251 Real Modes are predefined with @code{REAL} and @code{LONG_REAL}.
8253 @item @r{@emph{String Modes:}}
8256 @emph{Character String Mode} which is displayed by
8258 @code{CHARS(<string length>)}
8260 followed by the keyword @code{VARYING} if the String Mode is a varying
8263 @emph{Bit String Mode} which is displayed by
8270 @item @r{@emph{Array Mode:}}
8271 The Array Mode is displayed by the keyword @code{ARRAY(<range>)}
8272 followed by the element mode (which may in turn be an array mode).
8274 (@value{GDBP}) ptype x
8277 SET (karli = 10, susi = 20, fritzi = 100)
8280 @item @r{@emph{Structure Mode}}
8281 The Structure mode is displayed by the keyword @code{STRUCT(<field
8282 list>)}. The @code{<field list>} consists of names and modes of fields
8283 of the structure. Variant structures have the keyword @code{CASE <field>
8284 OF <variant fields> ESAC} in their field list. Since the current version
8285 of the GNU Chill compiler doesn't implement tag processing (no runtime
8286 checks of variant fields, and therefore no debugging info), the output
8287 always displays all variant fields.
8289 (@value{GDBP}) ptype str
8304 @subsubsection Locations and their accesses
8306 A location in Chill is an object which can contain values.
8308 A value of a location is generally accessed by the (declared) name of
8309 the location. The output conforms to the specification of values in
8310 Chill programs. How values are specified
8311 is the topic of the next section, @ref{Values and their Operations}.
8313 The pseudo-location @code{RESULT} (or @code{result}) can be used to
8314 display or change the result of a currently-active procedure:
8321 This does the same as the Chill action @code{RESULT EXPR} (which
8322 is not available in @value{GDBN}).
8324 Values of reference mode locations are printed by @code{PTR(<hex
8325 value>)} in case of a free reference mode, and by @code{(REF <reference
8326 mode>) (<hex-value>)} in case of a bound reference. @code{<hex value>}
8327 represents the address where the reference points to. To access the
8328 value of the location referenced by the pointer, use the dereference
8331 Values of procedure mode locations are displayed by
8334 (<argument modes> ) <return mode> @} <address> <name of procedure
8337 @code{<argument modes>} is a list of modes according to the parameter
8338 specification of the procedure and @code{<address>} shows the address of
8342 Locations of instance modes are displayed just like a structure with two
8343 fields specifying the @emph{process type} and the @emph{copy number} of
8344 the investigated instance location@footnote{This comes from the current
8345 implementation of instances. They are implemented as a structure (no
8346 na). The output should be something like @code{[<name of the process>;
8347 <instance number>]}.}. The field names are @code{__proc_type} and
8350 Locations of synchronization modes are displayed like a structure with
8351 the field name @code{__event_data} in case of a event mode location, and
8352 like a structure with the field @code{__buffer_data} in case of a buffer
8353 mode location (refer to previous paragraph).
8355 Structure Mode locations are printed by @code{[.<field name>: <value>,
8356 ...]}. The @code{<field name>} corresponds to the structure mode
8357 definition and the layout of @code{<value>} varies depending of the mode
8358 of the field. If the investigated structure mode location is of variant
8359 structure mode, the variant parts of the structure are enclosed in curled
8360 braces (@samp{@{@}}). Fields enclosed by @samp{@{,@}} are residing
8361 on the same memory location and represent the current values of the
8362 memory location in their specific modes. Since no tag processing is done
8363 all variants are displayed. A variant field is printed by
8364 @code{(<variant name>) = .<field name>: <value>}. (who implements the
8367 (@value{GDBP}) print str1 $4 = [.as: 0, .bs: karli, .<TAG>: { (karli) =
8368 [.cs: []], (susi) = [.ds: susi]}]
8372 Substructures of string mode-, array mode- or structure mode-values
8373 (e.g. array slices, fields of structure locations) are accessed using
8374 certain operations which are described in the next section, @ref{Values
8375 and their Operations}.
8377 A location value may be interpreted as having a different mode using the
8378 location conversion. This mode conversion is written as @code{<mode
8379 name>(<location>)}. The user has to consider that the sizes of the modes
8380 have to be equal otherwise an error occurs. Furthermore, no range
8381 checking of the location against the destination mode is performed, and
8382 therefore the result can be quite confusing.
8385 (@value{GDBP}) print int (s(3 up 4)) XXX TO be filled in !! XXX
8388 @node Values and their Operations
8389 @subsubsection Values and their Operations
8391 Values are used to alter locations, to investigate complex structures in
8392 more detail or to filter relevant information out of a large amount of
8393 data. There are several (mode dependent) operations defined which enable
8394 such investigations. These operations are not only applicable to
8395 constant values but also to locations, which can become quite useful
8396 when debugging complex structures. During parsing the command line
8397 (e.g. evaluating an expression) @value{GDBN} treats location names as
8398 the values behind these locations.
8400 This section describes how values have to be specified and which
8401 operations are legal to be used with such values.
8404 @item Literal Values
8405 Literal values are specified in the same manner as in @sc{gnu} Chill programs.
8406 For detailed specification refer to the @sc{gnu} Chill implementation Manual
8408 @c FIXME: if the Chill Manual is a Texinfo documents, the above should
8409 @c be converted to a @ref.
8414 @emph{Integer Literals} are specified in the same manner as in Chill
8415 programs (refer to the Chill Standard z200/88 chpt 5.2.4.2)
8417 @emph{Boolean Literals} are defined by @code{TRUE} and @code{FALSE}.
8419 @emph{Character Literals} are defined by @code{'<character>'}. (e.g.
8422 @emph{Set Literals} are defined by a name which was specified in a set
8423 mode. The value delivered by a Set Literal is the set value. This is
8424 comparable to an enumeration in C/C@t{++} language.
8426 @emph{Emptiness Literal} is predefined by @code{NULL}. The value of the
8427 emptiness literal delivers either the empty reference value, the empty
8428 procedure value or the empty instance value.
8431 @emph{Character String Literals} are defined by a sequence of characters
8432 enclosed in single- or double quotes. If a single- or double quote has
8433 to be part of the string literal it has to be stuffed (specified twice).
8435 @emph{Bitstring Literals} are specified in the same manner as in Chill
8436 programs (refer z200/88 chpt 5.2.4.8).
8438 @emph{Floating point literals} are specified in the same manner as in
8439 (gnu-)Chill programs (refer @sc{gnu} Chill implementation Manual chapter 1.5).
8444 A tuple is specified by @code{<mode name>[<tuple>]}, where @code{<mode
8445 name>} can be omitted if the mode of the tuple is unambiguous. This
8446 unambiguity is derived from the context of a evaluated expression.
8447 @code{<tuple>} can be one of the following:
8450 @item @emph{Powerset Tuple}
8451 @item @emph{Array Tuple}
8452 @item @emph{Structure Tuple}
8453 Powerset tuples, array tuples and structure tuples are specified in the
8454 same manner as in Chill programs refer to z200/88 chpt 5.2.5.
8457 @item String Element Value
8458 A string element value is specified by
8460 @code{<string value>(<index>)}
8462 where @code{<index>} is a integer expression. It delivers a character
8463 value which is equivalent to the character indexed by @code{<index>} in
8466 @item String Slice Value
8467 A string slice value is specified by @code{<string value>(<slice
8468 spec>)}, where @code{<slice spec>} can be either a range of integer
8469 expressions or specified by @code{<start expr> up <size>}.
8470 @code{<size>} denotes the number of elements which the slice contains.
8471 The delivered value is a string value, which is part of the specified
8474 @item Array Element Values
8475 An array element value is specified by @code{<array value>(<expr>)} and
8476 delivers a array element value of the mode of the specified array.
8478 @item Array Slice Values
8479 An array slice is specified by @code{<array value>(<slice spec>)}, where
8480 @code{<slice spec>} can be either a range specified by expressions or by
8481 @code{<start expr> up <size>}. @code{<size>} denotes the number of
8482 arrayelements the slice contains. The delivered value is an array value
8483 which is part of the specified array.
8485 @item Structure Field Values
8486 A structure field value is derived by @code{<structure value>.<field
8487 name>}, where @code{<field name>} indicates the name of a field specified
8488 in the mode definition of the structure. The mode of the delivered value
8489 corresponds to this mode definition in the structure definition.
8491 @item Procedure Call Value
8492 The procedure call value is derived from the return value of the
8493 procedure@footnote{If a procedure call is used for instance in an
8494 expression, then this procedure is called with all its side
8495 effects. This can lead to confusing results if used carelessly.}.
8497 Values of duration mode locations are represented by @code{ULONG} literals.
8499 Values of time mode locations appear as
8501 @code{TIME(<secs>:<nsecs>)}
8506 This is not implemented yet:
8507 @item Built-in Value
8509 The following built in functions are provided:
8521 @item @code{UPPER()}
8522 @item @code{LOWER()}
8523 @item @code{LENGTH()}
8527 @item @code{ARCSIN()}
8528 @item @code{ARCCOS()}
8529 @item @code{ARCTAN()}
8536 For a detailed description refer to the GNU Chill implementation manual
8540 @item Zero-adic Operator Value
8541 The zero-adic operator value is derived from the instance value for the
8542 current active process.
8544 @item Expression Values
8545 The value delivered by an expression is the result of the evaluation of
8546 the specified expression. If there are error conditions (mode
8547 incompatibility, etc.) the evaluation of expressions is aborted with a
8548 corresponding error message. Expressions may be parenthesised which
8549 causes the evaluation of this expression before any other expression
8550 which uses the result of the parenthesised expression. The following
8551 operators are supported by @value{GDBN}:
8554 @item @code{OR, ORIF, XOR}
8555 @itemx @code{AND, ANDIF}
8557 Logical operators defined over operands of boolean mode.
8560 Equality and inequality operators defined over all modes.
8564 Relational operators defined over predefined modes.
8567 @itemx @code{*, /, MOD, REM}
8568 Arithmetic operators defined over predefined modes.
8571 Change sign operator.
8574 String concatenation operator.
8577 String repetition operator.
8580 Referenced location operator which can be used either to take the
8581 address of a location (@code{->loc}), or to dereference a reference
8582 location (@code{loc->}).
8584 @item @code{OR, XOR}
8587 Powerset and bitstring operators.
8591 Powerset inclusion operators.
8594 Membership operator.
8598 @node Chill type and range checks
8599 @subsubsection Chill type and range checks
8601 @value{GDBN} considers two Chill variables mode equivalent if the sizes
8602 of the two modes are equal. This rule applies recursively to more
8603 complex datatypes which means that complex modes are treated
8604 equivalent if all element modes (which also can be complex modes like
8605 structures, arrays, etc.) have the same size.
8607 Range checking is done on all mathematical operations, assignment, array
8608 index bounds and all built in procedures.
8610 Strong type checks are forced using the @value{GDBN} command @code{set
8611 check strong}. This enforces strong type and range checks on all
8612 operations where Chill constructs are used (expressions, built in
8613 functions, etc.) in respect to the semantics as defined in the z.200
8614 language specification.
8616 All checks can be disabled by the @value{GDBN} command @code{set check
8620 @c Deviations from the Chill Standard Z200/88
8621 see last paragraph ?
8624 @node Chill defaults
8625 @subsubsection Chill defaults
8627 If type and range checking are set automatically by @value{GDBN}, they
8628 both default to @code{on} whenever the working language changes to
8629 Chill. This happens regardless of whether you or @value{GDBN}
8630 selected the working language.
8632 If you allow @value{GDBN} to set the language automatically, then entering
8633 code compiled from a file whose name ends with @file{.ch} sets the
8634 working language to Chill. @xref{Automatically, ,Having @value{GDBN} set
8635 the language automatically}, for further details.
8638 @chapter Examining the Symbol Table
8640 The commands described in this chapter allow you to inquire about the
8641 symbols (names of variables, functions and types) defined in your
8642 program. This information is inherent in the text of your program and
8643 does not change as your program executes. @value{GDBN} finds it in your
8644 program's symbol table, in the file indicated when you started @value{GDBN}
8645 (@pxref{File Options, ,Choosing files}), or by one of the
8646 file-management commands (@pxref{Files, ,Commands to specify files}).
8648 @cindex symbol names
8649 @cindex names of symbols
8650 @cindex quoting names
8651 Occasionally, you may need to refer to symbols that contain unusual
8652 characters, which @value{GDBN} ordinarily treats as word delimiters. The
8653 most frequent case is in referring to static variables in other
8654 source files (@pxref{Variables,,Program variables}). File names
8655 are recorded in object files as debugging symbols, but @value{GDBN} would
8656 ordinarily parse a typical file name, like @file{foo.c}, as the three words
8657 @samp{foo} @samp{.} @samp{c}. To allow @value{GDBN} to recognize
8658 @samp{foo.c} as a single symbol, enclose it in single quotes; for example,
8665 looks up the value of @code{x} in the scope of the file @file{foo.c}.
8668 @kindex info address
8669 @cindex address of a symbol
8670 @item info address @var{symbol}
8671 Describe where the data for @var{symbol} is stored. For a register
8672 variable, this says which register it is kept in. For a non-register
8673 local variable, this prints the stack-frame offset at which the variable
8676 Note the contrast with @samp{print &@var{symbol}}, which does not work
8677 at all for a register variable, and for a stack local variable prints
8678 the exact address of the current instantiation of the variable.
8681 @cindex symbol from address
8682 @item info symbol @var{addr}
8683 Print the name of a symbol which is stored at the address @var{addr}.
8684 If no symbol is stored exactly at @var{addr}, @value{GDBN} prints the
8685 nearest symbol and an offset from it:
8688 (@value{GDBP}) info symbol 0x54320
8689 _initialize_vx + 396 in section .text
8693 This is the opposite of the @code{info address} command. You can use
8694 it to find out the name of a variable or a function given its address.
8697 @item whatis @var{expr}
8698 Print the data type of expression @var{expr}. @var{expr} is not
8699 actually evaluated, and any side-effecting operations (such as
8700 assignments or function calls) inside it do not take place.
8701 @xref{Expressions, ,Expressions}.
8704 Print the data type of @code{$}, the last value in the value history.
8707 @item ptype @var{typename}
8708 Print a description of data type @var{typename}. @var{typename} may be
8709 the name of a type, or for C code it may have the form @samp{class
8710 @var{class-name}}, @samp{struct @var{struct-tag}}, @samp{union
8711 @var{union-tag}} or @samp{enum @var{enum-tag}}.
8713 @item ptype @var{expr}
8715 Print a description of the type of expression @var{expr}. @code{ptype}
8716 differs from @code{whatis} by printing a detailed description, instead
8717 of just the name of the type.
8719 For example, for this variable declaration:
8722 struct complex @{double real; double imag;@} v;
8726 the two commands give this output:
8730 (@value{GDBP}) whatis v
8731 type = struct complex
8732 (@value{GDBP}) ptype v
8733 type = struct complex @{
8741 As with @code{whatis}, using @code{ptype} without an argument refers to
8742 the type of @code{$}, the last value in the value history.
8745 @item info types @var{regexp}
8747 Print a brief description of all types whose names match @var{regexp}
8748 (or all types in your program, if you supply no argument). Each
8749 complete typename is matched as though it were a complete line; thus,
8750 @samp{i type value} gives information on all types in your program whose
8751 names include the string @code{value}, but @samp{i type ^value$} gives
8752 information only on types whose complete name is @code{value}.
8754 This command differs from @code{ptype} in two ways: first, like
8755 @code{whatis}, it does not print a detailed description; second, it
8756 lists all source files where a type is defined.
8759 @cindex local variables
8760 @item info scope @var{addr}
8761 List all the variables local to a particular scope. This command
8762 accepts a location---a function name, a source line, or an address
8763 preceded by a @samp{*}, and prints all the variables local to the
8764 scope defined by that location. For example:
8767 (@value{GDBP}) @b{info scope command_line_handler}
8768 Scope for command_line_handler:
8769 Symbol rl is an argument at stack/frame offset 8, length 4.
8770 Symbol linebuffer is in static storage at address 0x150a18, length 4.
8771 Symbol linelength is in static storage at address 0x150a1c, length 4.
8772 Symbol p is a local variable in register $esi, length 4.
8773 Symbol p1 is a local variable in register $ebx, length 4.
8774 Symbol nline is a local variable in register $edx, length 4.
8775 Symbol repeat is a local variable at frame offset -8, length 4.
8779 This command is especially useful for determining what data to collect
8780 during a @dfn{trace experiment}, see @ref{Tracepoint Actions,
8785 Show the name of the current source file---that is, the source file for
8786 the function containing the current point of execution---and the language
8789 @kindex info sources
8791 Print the names of all source files in your program for which there is
8792 debugging information, organized into two lists: files whose symbols
8793 have already been read, and files whose symbols will be read when needed.
8795 @kindex info functions
8796 @item info functions
8797 Print the names and data types of all defined functions.
8799 @item info functions @var{regexp}
8800 Print the names and data types of all defined functions
8801 whose names contain a match for regular expression @var{regexp}.
8802 Thus, @samp{info fun step} finds all functions whose names
8803 include @code{step}; @samp{info fun ^step} finds those whose names
8804 start with @code{step}. If a function name contains characters
8805 that conflict with the regular expression language (eg.
8806 @samp{operator*()}), they may be quoted with a backslash.
8808 @kindex info variables
8809 @item info variables
8810 Print the names and data types of all variables that are declared
8811 outside of functions (i.e., excluding local variables).
8813 @item info variables @var{regexp}
8814 Print the names and data types of all variables (except for local
8815 variables) whose names contain a match for regular expression
8819 This was never implemented.
8820 @kindex info methods
8822 @itemx info methods @var{regexp}
8823 The @code{info methods} command permits the user to examine all defined
8824 methods within C@t{++} program, or (with the @var{regexp} argument) a
8825 specific set of methods found in the various C@t{++} classes. Many
8826 C@t{++} classes provide a large number of methods. Thus, the output
8827 from the @code{ptype} command can be overwhelming and hard to use. The
8828 @code{info-methods} command filters the methods, printing only those
8829 which match the regular-expression @var{regexp}.
8832 @cindex reloading symbols
8833 Some systems allow individual object files that make up your program to
8834 be replaced without stopping and restarting your program. For example,
8835 in VxWorks you can simply recompile a defective object file and keep on
8836 running. If you are running on one of these systems, you can allow
8837 @value{GDBN} to reload the symbols for automatically relinked modules:
8840 @kindex set symbol-reloading
8841 @item set symbol-reloading on
8842 Replace symbol definitions for the corresponding source file when an
8843 object file with a particular name is seen again.
8845 @item set symbol-reloading off
8846 Do not replace symbol definitions when encountering object files of the
8847 same name more than once. This is the default state; if you are not
8848 running on a system that permits automatic relinking of modules, you
8849 should leave @code{symbol-reloading} off, since otherwise @value{GDBN}
8850 may discard symbols when linking large programs, that may contain
8851 several modules (from different directories or libraries) with the same
8854 @kindex show symbol-reloading
8855 @item show symbol-reloading
8856 Show the current @code{on} or @code{off} setting.
8859 @kindex set opaque-type-resolution
8860 @item set opaque-type-resolution on
8861 Tell @value{GDBN} to resolve opaque types. An opaque type is a type
8862 declared as a pointer to a @code{struct}, @code{class}, or
8863 @code{union}---for example, @code{struct MyType *}---that is used in one
8864 source file although the full declaration of @code{struct MyType} is in
8865 another source file. The default is on.
8867 A change in the setting of this subcommand will not take effect until
8868 the next time symbols for a file are loaded.
8870 @item set opaque-type-resolution off
8871 Tell @value{GDBN} not to resolve opaque types. In this case, the type
8872 is printed as follows:
8874 @{<no data fields>@}
8877 @kindex show opaque-type-resolution
8878 @item show opaque-type-resolution
8879 Show whether opaque types are resolved or not.
8881 @kindex maint print symbols
8883 @kindex maint print psymbols
8884 @cindex partial symbol dump
8885 @item maint print symbols @var{filename}
8886 @itemx maint print psymbols @var{filename}
8887 @itemx maint print msymbols @var{filename}
8888 Write a dump of debugging symbol data into the file @var{filename}.
8889 These commands are used to debug the @value{GDBN} symbol-reading code. Only
8890 symbols with debugging data are included. If you use @samp{maint print
8891 symbols}, @value{GDBN} includes all the symbols for which it has already
8892 collected full details: that is, @var{filename} reflects symbols for
8893 only those files whose symbols @value{GDBN} has read. You can use the
8894 command @code{info sources} to find out which files these are. If you
8895 use @samp{maint print psymbols} instead, the dump shows information about
8896 symbols that @value{GDBN} only knows partially---that is, symbols defined in
8897 files that @value{GDBN} has skimmed, but not yet read completely. Finally,
8898 @samp{maint print msymbols} dumps just the minimal symbol information
8899 required for each object file from which @value{GDBN} has read some symbols.
8900 @xref{Files, ,Commands to specify files}, for a discussion of how
8901 @value{GDBN} reads symbols (in the description of @code{symbol-file}).
8905 @chapter Altering Execution
8907 Once you think you have found an error in your program, you might want to
8908 find out for certain whether correcting the apparent error would lead to
8909 correct results in the rest of the run. You can find the answer by
8910 experiment, using the @value{GDBN} features for altering execution of the
8913 For example, you can store new values into variables or memory
8914 locations, give your program a signal, restart it at a different
8915 address, or even return prematurely from a function.
8918 * Assignment:: Assignment to variables
8919 * Jumping:: Continuing at a different address
8920 * Signaling:: Giving your program a signal
8921 * Returning:: Returning from a function
8922 * Calling:: Calling your program's functions
8923 * Patching:: Patching your program
8927 @section Assignment to variables
8930 @cindex setting variables
8931 To alter the value of a variable, evaluate an assignment expression.
8932 @xref{Expressions, ,Expressions}. For example,
8939 stores the value 4 into the variable @code{x}, and then prints the
8940 value of the assignment expression (which is 4).
8941 @xref{Languages, ,Using @value{GDBN} with Different Languages}, for more
8942 information on operators in supported languages.
8944 @kindex set variable
8945 @cindex variables, setting
8946 If you are not interested in seeing the value of the assignment, use the
8947 @code{set} command instead of the @code{print} command. @code{set} is
8948 really the same as @code{print} except that the expression's value is
8949 not printed and is not put in the value history (@pxref{Value History,
8950 ,Value history}). The expression is evaluated only for its effects.
8952 If the beginning of the argument string of the @code{set} command
8953 appears identical to a @code{set} subcommand, use the @code{set
8954 variable} command instead of just @code{set}. This command is identical
8955 to @code{set} except for its lack of subcommands. For example, if your
8956 program has a variable @code{width}, you get an error if you try to set
8957 a new value with just @samp{set width=13}, because @value{GDBN} has the
8958 command @code{set width}:
8961 (@value{GDBP}) whatis width
8963 (@value{GDBP}) p width
8965 (@value{GDBP}) set width=47
8966 Invalid syntax in expression.
8970 The invalid expression, of course, is @samp{=47}. In
8971 order to actually set the program's variable @code{width}, use
8974 (@value{GDBP}) set var width=47
8977 Because the @code{set} command has many subcommands that can conflict
8978 with the names of program variables, it is a good idea to use the
8979 @code{set variable} command instead of just @code{set}. For example, if
8980 your program has a variable @code{g}, you run into problems if you try
8981 to set a new value with just @samp{set g=4}, because @value{GDBN} has
8982 the command @code{set gnutarget}, abbreviated @code{set g}:
8986 (@value{GDBP}) whatis g
8990 (@value{GDBP}) set g=4
8994 The program being debugged has been started already.
8995 Start it from the beginning? (y or n) y
8996 Starting program: /home/smith/cc_progs/a.out
8997 "/home/smith/cc_progs/a.out": can't open to read symbols:
8999 (@value{GDBP}) show g
9000 The current BFD target is "=4".
9005 The program variable @code{g} did not change, and you silently set the
9006 @code{gnutarget} to an invalid value. In order to set the variable
9010 (@value{GDBP}) set var g=4
9013 @value{GDBN} allows more implicit conversions in assignments than C; you can
9014 freely store an integer value into a pointer variable or vice versa,
9015 and you can convert any structure to any other structure that is the
9016 same length or shorter.
9017 @comment FIXME: how do structs align/pad in these conversions?
9018 @comment /doc@cygnus.com 18dec1990
9020 To store values into arbitrary places in memory, use the @samp{@{@dots{}@}}
9021 construct to generate a value of specified type at a specified address
9022 (@pxref{Expressions, ,Expressions}). For example, @code{@{int@}0x83040} refers
9023 to memory location @code{0x83040} as an integer (which implies a certain size
9024 and representation in memory), and
9027 set @{int@}0x83040 = 4
9031 stores the value 4 into that memory location.
9034 @section Continuing at a different address
9036 Ordinarily, when you continue your program, you do so at the place where
9037 it stopped, with the @code{continue} command. You can instead continue at
9038 an address of your own choosing, with the following commands:
9042 @item jump @var{linespec}
9043 Resume execution at line @var{linespec}. Execution stops again
9044 immediately if there is a breakpoint there. @xref{List, ,Printing
9045 source lines}, for a description of the different forms of
9046 @var{linespec}. It is common practice to use the @code{tbreak} command
9047 in conjunction with @code{jump}. @xref{Set Breaks, ,Setting
9050 The @code{jump} command does not change the current stack frame, or
9051 the stack pointer, or the contents of any memory location or any
9052 register other than the program counter. If line @var{linespec} is in
9053 a different function from the one currently executing, the results may
9054 be bizarre if the two functions expect different patterns of arguments or
9055 of local variables. For this reason, the @code{jump} command requests
9056 confirmation if the specified line is not in the function currently
9057 executing. However, even bizarre results are predictable if you are
9058 well acquainted with the machine-language code of your program.
9060 @item jump *@var{address}
9061 Resume execution at the instruction at address @var{address}.
9064 @c Doesn't work on HP-UX; have to set $pcoqh and $pcoqt.
9065 On many systems, you can get much the same effect as the @code{jump}
9066 command by storing a new value into the register @code{$pc}. The
9067 difference is that this does not start your program running; it only
9068 changes the address of where it @emph{will} run when you continue. For
9076 makes the next @code{continue} command or stepping command execute at
9077 address @code{0x485}, rather than at the address where your program stopped.
9078 @xref{Continuing and Stepping, ,Continuing and stepping}.
9080 The most common occasion to use the @code{jump} command is to back
9081 up---perhaps with more breakpoints set---over a portion of a program
9082 that has already executed, in order to examine its execution in more
9087 @section Giving your program a signal
9091 @item signal @var{signal}
9092 Resume execution where your program stopped, but immediately give it the
9093 signal @var{signal}. @var{signal} can be the name or the number of a
9094 signal. For example, on many systems @code{signal 2} and @code{signal
9095 SIGINT} are both ways of sending an interrupt signal.
9097 Alternatively, if @var{signal} is zero, continue execution without
9098 giving a signal. This is useful when your program stopped on account of
9099 a signal and would ordinary see the signal when resumed with the
9100 @code{continue} command; @samp{signal 0} causes it to resume without a
9103 @code{signal} does not repeat when you press @key{RET} a second time
9104 after executing the command.
9108 Invoking the @code{signal} command is not the same as invoking the
9109 @code{kill} utility from the shell. Sending a signal with @code{kill}
9110 causes @value{GDBN} to decide what to do with the signal depending on
9111 the signal handling tables (@pxref{Signals}). The @code{signal} command
9112 passes the signal directly to your program.
9116 @section Returning from a function
9119 @cindex returning from a function
9122 @itemx return @var{expression}
9123 You can cancel execution of a function call with the @code{return}
9124 command. If you give an
9125 @var{expression} argument, its value is used as the function's return
9129 When you use @code{return}, @value{GDBN} discards the selected stack frame
9130 (and all frames within it). You can think of this as making the
9131 discarded frame return prematurely. If you wish to specify a value to
9132 be returned, give that value as the argument to @code{return}.
9134 This pops the selected stack frame (@pxref{Selection, ,Selecting a
9135 frame}), and any other frames inside of it, leaving its caller as the
9136 innermost remaining frame. That frame becomes selected. The
9137 specified value is stored in the registers used for returning values
9140 The @code{return} command does not resume execution; it leaves the
9141 program stopped in the state that would exist if the function had just
9142 returned. In contrast, the @code{finish} command (@pxref{Continuing
9143 and Stepping, ,Continuing and stepping}) resumes execution until the
9144 selected stack frame returns naturally.
9147 @section Calling program functions
9149 @cindex calling functions
9152 @item call @var{expr}
9153 Evaluate the expression @var{expr} without displaying @code{void}
9157 You can use this variant of the @code{print} command if you want to
9158 execute a function from your program, but without cluttering the output
9159 with @code{void} returned values. If the result is not void, it
9160 is printed and saved in the value history.
9162 @c OBSOLETE For the A29K, a user-controlled variable @code{call_scratch_address},
9163 @c OBSOLETE specifies the location of a scratch area to be used when @value{GDBN}
9164 @c OBSOLETE calls a function in the target. This is necessary because the usual
9165 @c OBSOLETE method of putting the scratch area on the stack does not work in systems
9166 @c OBSOLETE that have separate instruction and data spaces.
9169 @section Patching programs
9171 @cindex patching binaries
9172 @cindex writing into executables
9173 @cindex writing into corefiles
9175 By default, @value{GDBN} opens the file containing your program's
9176 executable code (or the corefile) read-only. This prevents accidental
9177 alterations to machine code; but it also prevents you from intentionally
9178 patching your program's binary.
9180 If you'd like to be able to patch the binary, you can specify that
9181 explicitly with the @code{set write} command. For example, you might
9182 want to turn on internal debugging flags, or even to make emergency
9188 @itemx set write off
9189 If you specify @samp{set write on}, @value{GDBN} opens executable and
9190 core files for both reading and writing; if you specify @samp{set write
9191 off} (the default), @value{GDBN} opens them read-only.
9193 If you have already loaded a file, you must load it again (using the
9194 @code{exec-file} or @code{core-file} command) after changing @code{set
9195 write}, for your new setting to take effect.
9199 Display whether executable files and core files are opened for writing
9204 @chapter @value{GDBN} Files
9206 @value{GDBN} needs to know the file name of the program to be debugged,
9207 both in order to read its symbol table and in order to start your
9208 program. To debug a core dump of a previous run, you must also tell
9209 @value{GDBN} the name of the core dump file.
9212 * Files:: Commands to specify files
9213 * Symbol Errors:: Errors reading symbol files
9217 @section Commands to specify files
9219 @cindex symbol table
9220 @cindex core dump file
9222 You may want to specify executable and core dump file names. The usual
9223 way to do this is at start-up time, using the arguments to
9224 @value{GDBN}'s start-up commands (@pxref{Invocation, , Getting In and
9225 Out of @value{GDBN}}).
9227 Occasionally it is necessary to change to a different file during a
9228 @value{GDBN} session. Or you may run @value{GDBN} and forget to specify
9229 a file you want to use. In these situations the @value{GDBN} commands
9230 to specify new files are useful.
9233 @cindex executable file
9235 @item file @var{filename}
9236 Use @var{filename} as the program to be debugged. It is read for its
9237 symbols and for the contents of pure memory. It is also the program
9238 executed when you use the @code{run} command. If you do not specify a
9239 directory and the file is not found in the @value{GDBN} working directory,
9240 @value{GDBN} uses the environment variable @code{PATH} as a list of
9241 directories to search, just as the shell does when looking for a program
9242 to run. You can change the value of this variable, for both @value{GDBN}
9243 and your program, using the @code{path} command.
9245 On systems with memory-mapped files, an auxiliary file named
9246 @file{@var{filename}.syms} may hold symbol table information for
9247 @var{filename}. If so, @value{GDBN} maps in the symbol table from
9248 @file{@var{filename}.syms}, starting up more quickly. See the
9249 descriptions of the file options @samp{-mapped} and @samp{-readnow}
9250 (available on the command line, and with the commands @code{file},
9251 @code{symbol-file}, or @code{add-symbol-file}, described below),
9252 for more information.
9255 @code{file} with no argument makes @value{GDBN} discard any information it
9256 has on both executable file and the symbol table.
9259 @item exec-file @r{[} @var{filename} @r{]}
9260 Specify that the program to be run (but not the symbol table) is found
9261 in @var{filename}. @value{GDBN} searches the environment variable @code{PATH}
9262 if necessary to locate your program. Omitting @var{filename} means to
9263 discard information on the executable file.
9266 @item symbol-file @r{[} @var{filename} @r{]}
9267 Read symbol table information from file @var{filename}. @code{PATH} is
9268 searched when necessary. Use the @code{file} command to get both symbol
9269 table and program to run from the same file.
9271 @code{symbol-file} with no argument clears out @value{GDBN} information on your
9272 program's symbol table.
9274 The @code{symbol-file} command causes @value{GDBN} to forget the contents
9275 of its convenience variables, the value history, and all breakpoints and
9276 auto-display expressions. This is because they may contain pointers to
9277 the internal data recording symbols and data types, which are part of
9278 the old symbol table data being discarded inside @value{GDBN}.
9280 @code{symbol-file} does not repeat if you press @key{RET} again after
9283 When @value{GDBN} is configured for a particular environment, it
9284 understands debugging information in whatever format is the standard
9285 generated for that environment; you may use either a @sc{gnu} compiler, or
9286 other compilers that adhere to the local conventions.
9287 Best results are usually obtained from @sc{gnu} compilers; for example,
9288 using @code{@value{GCC}} you can generate debugging information for
9291 For most kinds of object files, with the exception of old SVR3 systems
9292 using COFF, the @code{symbol-file} command does not normally read the
9293 symbol table in full right away. Instead, it scans the symbol table
9294 quickly to find which source files and which symbols are present. The
9295 details are read later, one source file at a time, as they are needed.
9297 The purpose of this two-stage reading strategy is to make @value{GDBN}
9298 start up faster. For the most part, it is invisible except for
9299 occasional pauses while the symbol table details for a particular source
9300 file are being read. (The @code{set verbose} command can turn these
9301 pauses into messages if desired. @xref{Messages/Warnings, ,Optional
9302 warnings and messages}.)
9304 We have not implemented the two-stage strategy for COFF yet. When the
9305 symbol table is stored in COFF format, @code{symbol-file} reads the
9306 symbol table data in full right away. Note that ``stabs-in-COFF''
9307 still does the two-stage strategy, since the debug info is actually
9311 @cindex reading symbols immediately
9312 @cindex symbols, reading immediately
9314 @cindex memory-mapped symbol file
9315 @cindex saving symbol table
9316 @item symbol-file @var{filename} @r{[} -readnow @r{]} @r{[} -mapped @r{]}
9317 @itemx file @var{filename} @r{[} -readnow @r{]} @r{[} -mapped @r{]}
9318 You can override the @value{GDBN} two-stage strategy for reading symbol
9319 tables by using the @samp{-readnow} option with any of the commands that
9320 load symbol table information, if you want to be sure @value{GDBN} has the
9321 entire symbol table available.
9323 If memory-mapped files are available on your system through the
9324 @code{mmap} system call, you can use another option, @samp{-mapped}, to
9325 cause @value{GDBN} to write the symbols for your program into a reusable
9326 file. Future @value{GDBN} debugging sessions map in symbol information
9327 from this auxiliary symbol file (if the program has not changed), rather
9328 than spending time reading the symbol table from the executable
9329 program. Using the @samp{-mapped} option has the same effect as
9330 starting @value{GDBN} with the @samp{-mapped} command-line option.
9332 You can use both options together, to make sure the auxiliary symbol
9333 file has all the symbol information for your program.
9335 The auxiliary symbol file for a program called @var{myprog} is called
9336 @samp{@var{myprog}.syms}. Once this file exists (so long as it is newer
9337 than the corresponding executable), @value{GDBN} always attempts to use
9338 it when you debug @var{myprog}; no special options or commands are
9341 The @file{.syms} file is specific to the host machine where you run
9342 @value{GDBN}. It holds an exact image of the internal @value{GDBN}
9343 symbol table. It cannot be shared across multiple host platforms.
9345 @c FIXME: for now no mention of directories, since this seems to be in
9346 @c flux. 13mar1992 status is that in theory GDB would look either in
9347 @c current dir or in same dir as myprog; but issues like competing
9348 @c GDB's, or clutter in system dirs, mean that in practice right now
9349 @c only current dir is used. FFish says maybe a special GDB hierarchy
9350 @c (eg rooted in val of env var GDBSYMS) could exist for mappable symbol
9355 @item core-file @r{[} @var{filename} @r{]}
9356 Specify the whereabouts of a core dump file to be used as the ``contents
9357 of memory''. Traditionally, core files contain only some parts of the
9358 address space of the process that generated them; @value{GDBN} can access the
9359 executable file itself for other parts.
9361 @code{core-file} with no argument specifies that no core file is
9364 Note that the core file is ignored when your program is actually running
9365 under @value{GDBN}. So, if you have been running your program and you
9366 wish to debug a core file instead, you must kill the subprocess in which
9367 the program is running. To do this, use the @code{kill} command
9368 (@pxref{Kill Process, ,Killing the child process}).
9370 @kindex add-symbol-file
9371 @cindex dynamic linking
9372 @item add-symbol-file @var{filename} @var{address}
9373 @itemx add-symbol-file @var{filename} @var{address} @r{[} -readnow @r{]} @r{[} -mapped @r{]}
9374 @itemx add-symbol-file @var{filename} @r{-s}@var{section} @var{address} @dots{}
9375 The @code{add-symbol-file} command reads additional symbol table
9376 information from the file @var{filename}. You would use this command
9377 when @var{filename} has been dynamically loaded (by some other means)
9378 into the program that is running. @var{address} should be the memory
9379 address at which the file has been loaded; @value{GDBN} cannot figure
9380 this out for itself. You can additionally specify an arbitrary number
9381 of @samp{@r{-s}@var{section} @var{address}} pairs, to give an explicit
9382 section name and base address for that section. You can specify any
9383 @var{address} as an expression.
9385 The symbol table of the file @var{filename} is added to the symbol table
9386 originally read with the @code{symbol-file} command. You can use the
9387 @code{add-symbol-file} command any number of times; the new symbol data
9388 thus read keeps adding to the old. To discard all old symbol data
9389 instead, use the @code{symbol-file} command without any arguments.
9391 @cindex relocatable object files, reading symbols from
9392 @cindex object files, relocatable, reading symbols from
9393 @cindex reading symbols from relocatable object files
9394 @cindex symbols, reading from relocatable object files
9395 @cindex @file{.o} files, reading symbols from
9396 Although @var{filename} is typically a shared library file, an
9397 executable file, or some other object file which has been fully
9398 relocated for loading into a process, you can also load symbolic
9399 information from relocatable @file{.o} files, as long as:
9403 the file's symbolic information refers only to linker symbols defined in
9404 that file, not to symbols defined by other object files,
9406 every section the file's symbolic information refers to has actually
9407 been loaded into the inferior, as it appears in the file, and
9409 you can determine the address at which every section was loaded, and
9410 provide these to the @code{add-symbol-file} command.
9414 Some embedded operating systems, like Sun Chorus and VxWorks, can load
9415 relocatable files into an already running program; such systems
9416 typically make the requirements above easy to meet. However, it's
9417 important to recognize that many native systems use complex link
9418 procedures (@code{.linkonce} section factoring and C++ constructor table
9419 assembly, for example) that make the requirements difficult to meet. In
9420 general, one cannot assume that using @code{add-symbol-file} to read a
9421 relocatable object file's symbolic information will have the same effect
9422 as linking the relocatable object file into the program in the normal
9425 @code{add-symbol-file} does not repeat if you press @key{RET} after using it.
9427 You can use the @samp{-mapped} and @samp{-readnow} options just as with
9428 the @code{symbol-file} command, to change how @value{GDBN} manages the symbol
9429 table information for @var{filename}.
9431 @kindex add-shared-symbol-file
9432 @item add-shared-symbol-file
9433 The @code{add-shared-symbol-file} command can be used only under Harris' CXUX
9434 operating system for the Motorola 88k. @value{GDBN} automatically looks for
9435 shared libraries, however if @value{GDBN} does not find yours, you can run
9436 @code{add-shared-symbol-file}. It takes no arguments.
9440 The @code{section} command changes the base address of section SECTION of
9441 the exec file to ADDR. This can be used if the exec file does not contain
9442 section addresses, (such as in the a.out format), or when the addresses
9443 specified in the file itself are wrong. Each section must be changed
9444 separately. The @code{info files} command, described below, lists all
9445 the sections and their addresses.
9451 @code{info files} and @code{info target} are synonymous; both print the
9452 current target (@pxref{Targets, ,Specifying a Debugging Target}),
9453 including the names of the executable and core dump files currently in
9454 use by @value{GDBN}, and the files from which symbols were loaded. The
9455 command @code{help target} lists all possible targets rather than
9458 @kindex maint info sections
9459 @item maint info sections
9460 Another command that can give you extra information about program sections
9461 is @code{maint info sections}. In addition to the section information
9462 displayed by @code{info files}, this command displays the flags and file
9463 offset of each section in the executable and core dump files. In addition,
9464 @code{maint info sections} provides the following command options (which
9465 may be arbitrarily combined):
9469 Display sections for all loaded object files, including shared libraries.
9470 @item @var{sections}
9471 Display info only for named @var{sections}.
9472 @item @var{section-flags}
9473 Display info only for sections for which @var{section-flags} are true.
9474 The section flags that @value{GDBN} currently knows about are:
9477 Section will have space allocated in the process when loaded.
9478 Set for all sections except those containing debug information.
9480 Section will be loaded from the file into the child process memory.
9481 Set for pre-initialized code and data, clear for @code{.bss} sections.
9483 Section needs to be relocated before loading.
9485 Section cannot be modified by the child process.
9487 Section contains executable code only.
9489 Section contains data only (no executable code).
9491 Section will reside in ROM.
9493 Section contains data for constructor/destructor lists.
9495 Section is not empty.
9497 An instruction to the linker to not output the section.
9498 @item COFF_SHARED_LIBRARY
9499 A notification to the linker that the section contains
9500 COFF shared library information.
9502 Section contains common symbols.
9507 All file-specifying commands allow both absolute and relative file names
9508 as arguments. @value{GDBN} always converts the file name to an absolute file
9509 name and remembers it that way.
9511 @cindex shared libraries
9512 @value{GDBN} supports HP-UX, SunOS, SVr4, Irix 5, and IBM RS/6000 shared
9515 @value{GDBN} automatically loads symbol definitions from shared libraries
9516 when you use the @code{run} command, or when you examine a core file.
9517 (Before you issue the @code{run} command, @value{GDBN} does not understand
9518 references to a function in a shared library, however---unless you are
9519 debugging a core file).
9521 On HP-UX, if the program loads a library explicitly, @value{GDBN}
9522 automatically loads the symbols at the time of the @code{shl_load} call.
9524 @c FIXME: some @value{GDBN} release may permit some refs to undef
9525 @c FIXME...symbols---eg in a break cmd---assuming they are from a shared
9526 @c FIXME...lib; check this from time to time when updating manual
9528 There are times, however, when you may wish to not automatically load
9529 symbol definitions from shared libraries, such as when they are
9530 particularly large or there are many of them.
9532 To control the automatic loading of shared library symbols, use the
9536 @kindex set auto-solib-add
9537 @item set auto-solib-add @var{mode}
9538 If @var{mode} is @code{on}, symbols from all shared object libraries
9539 will be loaded automatically when the inferior begins execution, you
9540 attach to an independently started inferior, or when the dynamic linker
9541 informs @value{GDBN} that a new library has been loaded. If @var{mode}
9542 is @code{off}, symbols must be loaded manually, using the
9543 @code{sharedlibrary} command. The default value is @code{on}.
9545 @kindex show auto-solib-add
9546 @item show auto-solib-add
9547 Display the current autoloading mode.
9550 To explicitly load shared library symbols, use the @code{sharedlibrary}
9554 @kindex info sharedlibrary
9557 @itemx info sharedlibrary
9558 Print the names of the shared libraries which are currently loaded.
9560 @kindex sharedlibrary
9562 @item sharedlibrary @var{regex}
9563 @itemx share @var{regex}
9564 Load shared object library symbols for files matching a
9565 Unix regular expression.
9566 As with files loaded automatically, it only loads shared libraries
9567 required by your program for a core file or after typing @code{run}. If
9568 @var{regex} is omitted all shared libraries required by your program are
9572 On some systems, such as HP-UX systems, @value{GDBN} supports
9573 autoloading shared library symbols until a limiting threshold size is
9574 reached. This provides the benefit of allowing autoloading to remain on
9575 by default, but avoids autoloading excessively large shared libraries,
9576 up to a threshold that is initially set, but which you can modify if you
9579 Beyond that threshold, symbols from shared libraries must be explicitly
9580 loaded. To load these symbols, use the command @code{sharedlibrary
9581 @var{filename}}. The base address of the shared library is determined
9582 automatically by @value{GDBN} and need not be specified.
9584 To display or set the threshold, use the commands:
9587 @kindex set auto-solib-limit
9588 @item set auto-solib-limit @var{threshold}
9589 Set the autoloading size threshold, in an integral number of megabytes.
9590 If @var{threshold} is nonzero and shared library autoloading is enabled,
9591 symbols from all shared object libraries will be loaded until the total
9592 size of the loaded shared library symbols exceeds this threshold.
9593 Otherwise, symbols must be loaded manually, using the
9594 @code{sharedlibrary} command. The default threshold is 100 (i.e. 100
9597 @kindex show auto-solib-limit
9598 @item show auto-solib-limit
9599 Display the current autoloading size threshold, in megabytes.
9603 @section Errors reading symbol files
9605 While reading a symbol file, @value{GDBN} occasionally encounters problems,
9606 such as symbol types it does not recognize, or known bugs in compiler
9607 output. By default, @value{GDBN} does not notify you of such problems, since
9608 they are relatively common and primarily of interest to people
9609 debugging compilers. If you are interested in seeing information
9610 about ill-constructed symbol tables, you can either ask @value{GDBN} to print
9611 only one message about each such type of problem, no matter how many
9612 times the problem occurs; or you can ask @value{GDBN} to print more messages,
9613 to see how many times the problems occur, with the @code{set
9614 complaints} command (@pxref{Messages/Warnings, ,Optional warnings and
9617 The messages currently printed, and their meanings, include:
9620 @item inner block not inside outer block in @var{symbol}
9622 The symbol information shows where symbol scopes begin and end
9623 (such as at the start of a function or a block of statements). This
9624 error indicates that an inner scope block is not fully contained
9625 in its outer scope blocks.
9627 @value{GDBN} circumvents the problem by treating the inner block as if it had
9628 the same scope as the outer block. In the error message, @var{symbol}
9629 may be shown as ``@code{(don't know)}'' if the outer block is not a
9632 @item block at @var{address} out of order
9634 The symbol information for symbol scope blocks should occur in
9635 order of increasing addresses. This error indicates that it does not
9638 @value{GDBN} does not circumvent this problem, and has trouble
9639 locating symbols in the source file whose symbols it is reading. (You
9640 can often determine what source file is affected by specifying
9641 @code{set verbose on}. @xref{Messages/Warnings, ,Optional warnings and
9644 @item bad block start address patched
9646 The symbol information for a symbol scope block has a start address
9647 smaller than the address of the preceding source line. This is known
9648 to occur in the SunOS 4.1.1 (and earlier) C compiler.
9650 @value{GDBN} circumvents the problem by treating the symbol scope block as
9651 starting on the previous source line.
9653 @item bad string table offset in symbol @var{n}
9656 Symbol number @var{n} contains a pointer into the string table which is
9657 larger than the size of the string table.
9659 @value{GDBN} circumvents the problem by considering the symbol to have the
9660 name @code{foo}, which may cause other problems if many symbols end up
9663 @item unknown symbol type @code{0x@var{nn}}
9665 The symbol information contains new data types that @value{GDBN} does
9666 not yet know how to read. @code{0x@var{nn}} is the symbol type of the
9667 uncomprehended information, in hexadecimal.
9669 @value{GDBN} circumvents the error by ignoring this symbol information.
9670 This usually allows you to debug your program, though certain symbols
9671 are not accessible. If you encounter such a problem and feel like
9672 debugging it, you can debug @code{@value{GDBP}} with itself, breakpoint
9673 on @code{complain}, then go up to the function @code{read_dbx_symtab}
9674 and examine @code{*bufp} to see the symbol.
9676 @item stub type has NULL name
9678 @value{GDBN} could not find the full definition for a struct or class.
9680 @item const/volatile indicator missing (ok if using g++ v1.x), got@dots{}
9681 The symbol information for a C@t{++} member function is missing some
9682 information that recent versions of the compiler should have output for
9685 @item info mismatch between compiler and debugger
9687 @value{GDBN} could not parse a type specification output by the compiler.
9692 @chapter Specifying a Debugging Target
9694 @cindex debugging target
9697 A @dfn{target} is the execution environment occupied by your program.
9699 Often, @value{GDBN} runs in the same host environment as your program;
9700 in that case, the debugging target is specified as a side effect when
9701 you use the @code{file} or @code{core} commands. When you need more
9702 flexibility---for example, running @value{GDBN} on a physically separate
9703 host, or controlling a standalone system over a serial port or a
9704 realtime system over a TCP/IP connection---you can use the @code{target}
9705 command to specify one of the target types configured for @value{GDBN}
9706 (@pxref{Target Commands, ,Commands for managing targets}).
9709 * Active Targets:: Active targets
9710 * Target Commands:: Commands for managing targets
9711 * Byte Order:: Choosing target byte order
9712 * Remote:: Remote debugging
9713 * KOD:: Kernel Object Display
9717 @node Active Targets
9718 @section Active targets
9720 @cindex stacking targets
9721 @cindex active targets
9722 @cindex multiple targets
9724 There are three classes of targets: processes, core files, and
9725 executable files. @value{GDBN} can work concurrently on up to three
9726 active targets, one in each class. This allows you to (for example)
9727 start a process and inspect its activity without abandoning your work on
9730 For example, if you execute @samp{gdb a.out}, then the executable file
9731 @code{a.out} is the only active target. If you designate a core file as
9732 well---presumably from a prior run that crashed and coredumped---then
9733 @value{GDBN} has two active targets and uses them in tandem, looking
9734 first in the corefile target, then in the executable file, to satisfy
9735 requests for memory addresses. (Typically, these two classes of target
9736 are complementary, since core files contain only a program's
9737 read-write memory---variables and so on---plus machine status, while
9738 executable files contain only the program text and initialized data.)
9740 When you type @code{run}, your executable file becomes an active process
9741 target as well. When a process target is active, all @value{GDBN}
9742 commands requesting memory addresses refer to that target; addresses in
9743 an active core file or executable file target are obscured while the
9744 process target is active.
9746 Use the @code{core-file} and @code{exec-file} commands to select a new
9747 core file or executable target (@pxref{Files, ,Commands to specify
9748 files}). To specify as a target a process that is already running, use
9749 the @code{attach} command (@pxref{Attach, ,Debugging an already-running
9752 @node Target Commands
9753 @section Commands for managing targets
9756 @item target @var{type} @var{parameters}
9757 Connects the @value{GDBN} host environment to a target machine or
9758 process. A target is typically a protocol for talking to debugging
9759 facilities. You use the argument @var{type} to specify the type or
9760 protocol of the target machine.
9762 Further @var{parameters} are interpreted by the target protocol, but
9763 typically include things like device names or host names to connect
9764 with, process numbers, and baud rates.
9766 The @code{target} command does not repeat if you press @key{RET} again
9767 after executing the command.
9771 Displays the names of all targets available. To display targets
9772 currently selected, use either @code{info target} or @code{info files}
9773 (@pxref{Files, ,Commands to specify files}).
9775 @item help target @var{name}
9776 Describe a particular target, including any parameters necessary to
9779 @kindex set gnutarget
9780 @item set gnutarget @var{args}
9781 @value{GDBN} uses its own library BFD to read your files. @value{GDBN}
9782 knows whether it is reading an @dfn{executable},
9783 a @dfn{core}, or a @dfn{.o} file; however, you can specify the file format
9784 with the @code{set gnutarget} command. Unlike most @code{target} commands,
9785 with @code{gnutarget} the @code{target} refers to a program, not a machine.
9788 @emph{Warning:} To specify a file format with @code{set gnutarget},
9789 you must know the actual BFD name.
9793 @xref{Files, , Commands to specify files}.
9795 @kindex show gnutarget
9796 @item show gnutarget
9797 Use the @code{show gnutarget} command to display what file format
9798 @code{gnutarget} is set to read. If you have not set @code{gnutarget},
9799 @value{GDBN} will determine the file format for each file automatically,
9800 and @code{show gnutarget} displays @samp{The current BDF target is "auto"}.
9803 Here are some common targets (available, or not, depending on the GDB
9808 @item target exec @var{program}
9809 An executable file. @samp{target exec @var{program}} is the same as
9810 @samp{exec-file @var{program}}.
9813 @item target core @var{filename}
9814 A core dump file. @samp{target core @var{filename}} is the same as
9815 @samp{core-file @var{filename}}.
9817 @kindex target remote
9818 @item target remote @var{dev}
9819 Remote serial target in GDB-specific protocol. The argument @var{dev}
9820 specifies what serial device to use for the connection (e.g.
9821 @file{/dev/ttya}). @xref{Remote, ,Remote debugging}. @code{target remote}
9822 supports the @code{load} command. This is only useful if you have
9823 some other way of getting the stub to the target system, and you can put
9824 it somewhere in memory where it won't get clobbered by the download.
9828 Builtin CPU simulator. @value{GDBN} includes simulators for most architectures.
9836 works; however, you cannot assume that a specific memory map, device
9837 drivers, or even basic I/O is available, although some simulators do
9838 provide these. For info about any processor-specific simulator details,
9839 see the appropriate section in @ref{Embedded Processors, ,Embedded
9844 Some configurations may include these targets as well:
9849 @item target nrom @var{dev}
9850 NetROM ROM emulator. This target only supports downloading.
9854 Different targets are available on different configurations of @value{GDBN};
9855 your configuration may have more or fewer targets.
9857 Many remote targets require you to download the executable's code
9858 once you've successfully established a connection.
9862 @kindex load @var{filename}
9863 @item load @var{filename}
9864 Depending on what remote debugging facilities are configured into
9865 @value{GDBN}, the @code{load} command may be available. Where it exists, it
9866 is meant to make @var{filename} (an executable) available for debugging
9867 on the remote system---by downloading, or dynamic linking, for example.
9868 @code{load} also records the @var{filename} symbol table in @value{GDBN}, like
9869 the @code{add-symbol-file} command.
9871 If your @value{GDBN} does not have a @code{load} command, attempting to
9872 execute it gets the error message ``@code{You can't do that when your
9873 target is @dots{}}''
9875 The file is loaded at whatever address is specified in the executable.
9876 For some object file formats, you can specify the load address when you
9877 link the program; for other formats, like a.out, the object file format
9878 specifies a fixed address.
9879 @c FIXME! This would be a good place for an xref to the GNU linker doc.
9881 @code{load} does not repeat if you press @key{RET} again after using it.
9885 @section Choosing target byte order
9887 @cindex choosing target byte order
9888 @cindex target byte order
9890 Some types of processors, such as the MIPS, PowerPC, and Hitachi SH,
9891 offer the ability to run either big-endian or little-endian byte
9892 orders. Usually the executable or symbol will include a bit to
9893 designate the endian-ness, and you will not need to worry about
9894 which to use. However, you may still find it useful to adjust
9895 @value{GDBN}'s idea of processor endian-ness manually.
9898 @kindex set endian big
9899 @item set endian big
9900 Instruct @value{GDBN} to assume the target is big-endian.
9902 @kindex set endian little
9903 @item set endian little
9904 Instruct @value{GDBN} to assume the target is little-endian.
9906 @kindex set endian auto
9907 @item set endian auto
9908 Instruct @value{GDBN} to use the byte order associated with the
9912 Display @value{GDBN}'s current idea of the target byte order.
9916 Note that these commands merely adjust interpretation of symbolic
9917 data on the host, and that they have absolutely no effect on the
9921 @section Remote debugging
9922 @cindex remote debugging
9924 If you are trying to debug a program running on a machine that cannot run
9925 @value{GDBN} in the usual way, it is often useful to use remote debugging.
9926 For example, you might use remote debugging on an operating system kernel,
9927 or on a small system which does not have a general purpose operating system
9928 powerful enough to run a full-featured debugger.
9930 Some configurations of @value{GDBN} have special serial or TCP/IP interfaces
9931 to make this work with particular debugging targets. In addition,
9932 @value{GDBN} comes with a generic serial protocol (specific to @value{GDBN},
9933 but not specific to any particular target system) which you can use if you
9934 write the remote stubs---the code that runs on the remote system to
9935 communicate with @value{GDBN}.
9937 Other remote targets may be available in your
9938 configuration of @value{GDBN}; use @code{help target} to list them.
9941 * Remote Serial:: @value{GDBN} remote serial protocol
9945 @subsection The @value{GDBN} remote serial protocol
9947 @cindex remote serial debugging, overview
9948 To debug a program running on another machine (the debugging
9949 @dfn{target} machine), you must first arrange for all the usual
9950 prerequisites for the program to run by itself. For example, for a C
9955 A startup routine to set up the C runtime environment; these usually
9956 have a name like @file{crt0}. The startup routine may be supplied by
9957 your hardware supplier, or you may have to write your own.
9960 A C subroutine library to support your program's
9961 subroutine calls, notably managing input and output.
9964 A way of getting your program to the other machine---for example, a
9965 download program. These are often supplied by the hardware
9966 manufacturer, but you may have to write your own from hardware
9970 The next step is to arrange for your program to use a serial port to
9971 communicate with the machine where @value{GDBN} is running (the @dfn{host}
9972 machine). In general terms, the scheme looks like this:
9976 @value{GDBN} already understands how to use this protocol; when everything
9977 else is set up, you can simply use the @samp{target remote} command
9978 (@pxref{Targets,,Specifying a Debugging Target}).
9980 @item On the target,
9981 you must link with your program a few special-purpose subroutines that
9982 implement the @value{GDBN} remote serial protocol. The file containing these
9983 subroutines is called a @dfn{debugging stub}.
9985 On certain remote targets, you can use an auxiliary program
9986 @code{gdbserver} instead of linking a stub into your program.
9987 @xref{Server,,Using the @code{gdbserver} program}, for details.
9990 The debugging stub is specific to the architecture of the remote
9991 machine; for example, use @file{sparc-stub.c} to debug programs on
9994 @cindex remote serial stub list
9995 These working remote stubs are distributed with @value{GDBN}:
10000 @cindex @file{i386-stub.c}
10003 For Intel 386 and compatible architectures.
10006 @cindex @file{m68k-stub.c}
10007 @cindex Motorola 680x0
10009 For Motorola 680x0 architectures.
10012 @cindex @file{sh-stub.c}
10015 For Hitachi SH architectures.
10018 @cindex @file{sparc-stub.c}
10020 For @sc{sparc} architectures.
10022 @item sparcl-stub.c
10023 @cindex @file{sparcl-stub.c}
10026 For Fujitsu @sc{sparclite} architectures.
10030 The @file{README} file in the @value{GDBN} distribution may list other
10031 recently added stubs.
10034 * Stub Contents:: What the stub can do for you
10035 * Bootstrapping:: What you must do for the stub
10036 * Debug Session:: Putting it all together
10037 * Protocol:: Definition of the communication protocol
10038 * Server:: Using the `gdbserver' program
10039 * NetWare:: Using the `gdbserve.nlm' program
10042 @node Stub Contents
10043 @subsubsection What the stub can do for you
10045 @cindex remote serial stub
10046 The debugging stub for your architecture supplies these three
10050 @item set_debug_traps
10051 @kindex set_debug_traps
10052 @cindex remote serial stub, initialization
10053 This routine arranges for @code{handle_exception} to run when your
10054 program stops. You must call this subroutine explicitly near the
10055 beginning of your program.
10057 @item handle_exception
10058 @kindex handle_exception
10059 @cindex remote serial stub, main routine
10060 This is the central workhorse, but your program never calls it
10061 explicitly---the setup code arranges for @code{handle_exception} to
10062 run when a trap is triggered.
10064 @code{handle_exception} takes control when your program stops during
10065 execution (for example, on a breakpoint), and mediates communications
10066 with @value{GDBN} on the host machine. This is where the communications
10067 protocol is implemented; @code{handle_exception} acts as the @value{GDBN}
10068 representative on the target machine. It begins by sending summary
10069 information on the state of your program, then continues to execute,
10070 retrieving and transmitting any information @value{GDBN} needs, until you
10071 execute a @value{GDBN} command that makes your program resume; at that point,
10072 @code{handle_exception} returns control to your own code on the target
10076 @cindex @code{breakpoint} subroutine, remote
10077 Use this auxiliary subroutine to make your program contain a
10078 breakpoint. Depending on the particular situation, this may be the only
10079 way for @value{GDBN} to get control. For instance, if your target
10080 machine has some sort of interrupt button, you won't need to call this;
10081 pressing the interrupt button transfers control to
10082 @code{handle_exception}---in effect, to @value{GDBN}. On some machines,
10083 simply receiving characters on the serial port may also trigger a trap;
10084 again, in that situation, you don't need to call @code{breakpoint} from
10085 your own program---simply running @samp{target remote} from the host
10086 @value{GDBN} session gets control.
10088 Call @code{breakpoint} if none of these is true, or if you simply want
10089 to make certain your program stops at a predetermined point for the
10090 start of your debugging session.
10093 @node Bootstrapping
10094 @subsubsection What you must do for the stub
10096 @cindex remote stub, support routines
10097 The debugging stubs that come with @value{GDBN} are set up for a particular
10098 chip architecture, but they have no information about the rest of your
10099 debugging target machine.
10101 First of all you need to tell the stub how to communicate with the
10105 @item int getDebugChar()
10106 @kindex getDebugChar
10107 Write this subroutine to read a single character from the serial port.
10108 It may be identical to @code{getchar} for your target system; a
10109 different name is used to allow you to distinguish the two if you wish.
10111 @item void putDebugChar(int)
10112 @kindex putDebugChar
10113 Write this subroutine to write a single character to the serial port.
10114 It may be identical to @code{putchar} for your target system; a
10115 different name is used to allow you to distinguish the two if you wish.
10118 @cindex control C, and remote debugging
10119 @cindex interrupting remote targets
10120 If you want @value{GDBN} to be able to stop your program while it is
10121 running, you need to use an interrupt-driven serial driver, and arrange
10122 for it to stop when it receives a @code{^C} (@samp{\003}, the control-C
10123 character). That is the character which @value{GDBN} uses to tell the
10124 remote system to stop.
10126 Getting the debugging target to return the proper status to @value{GDBN}
10127 probably requires changes to the standard stub; one quick and dirty way
10128 is to just execute a breakpoint instruction (the ``dirty'' part is that
10129 @value{GDBN} reports a @code{SIGTRAP} instead of a @code{SIGINT}).
10131 Other routines you need to supply are:
10134 @item void exceptionHandler (int @var{exception_number}, void *@var{exception_address})
10135 @kindex exceptionHandler
10136 Write this function to install @var{exception_address} in the exception
10137 handling tables. You need to do this because the stub does not have any
10138 way of knowing what the exception handling tables on your target system
10139 are like (for example, the processor's table might be in @sc{rom},
10140 containing entries which point to a table in @sc{ram}).
10141 @var{exception_number} is the exception number which should be changed;
10142 its meaning is architecture-dependent (for example, different numbers
10143 might represent divide by zero, misaligned access, etc). When this
10144 exception occurs, control should be transferred directly to
10145 @var{exception_address}, and the processor state (stack, registers,
10146 and so on) should be just as it is when a processor exception occurs. So if
10147 you want to use a jump instruction to reach @var{exception_address}, it
10148 should be a simple jump, not a jump to subroutine.
10150 For the 386, @var{exception_address} should be installed as an interrupt
10151 gate so that interrupts are masked while the handler runs. The gate
10152 should be at privilege level 0 (the most privileged level). The
10153 @sc{sparc} and 68k stubs are able to mask interrupts themselves without
10154 help from @code{exceptionHandler}.
10156 @item void flush_i_cache()
10157 @kindex flush_i_cache
10158 On @sc{sparc} and @sc{sparclite} only, write this subroutine to flush the
10159 instruction cache, if any, on your target machine. If there is no
10160 instruction cache, this subroutine may be a no-op.
10162 On target machines that have instruction caches, @value{GDBN} requires this
10163 function to make certain that the state of your program is stable.
10167 You must also make sure this library routine is available:
10170 @item void *memset(void *, int, int)
10172 This is the standard library function @code{memset} that sets an area of
10173 memory to a known value. If you have one of the free versions of
10174 @code{libc.a}, @code{memset} can be found there; otherwise, you must
10175 either obtain it from your hardware manufacturer, or write your own.
10178 If you do not use the GNU C compiler, you may need other standard
10179 library subroutines as well; this varies from one stub to another,
10180 but in general the stubs are likely to use any of the common library
10181 subroutines which @code{@value{GCC}} generates as inline code.
10184 @node Debug Session
10185 @subsubsection Putting it all together
10187 @cindex remote serial debugging summary
10188 In summary, when your program is ready to debug, you must follow these
10193 Make sure you have defined the supporting low-level routines
10194 (@pxref{Bootstrapping,,What you must do for the stub}):
10196 @code{getDebugChar}, @code{putDebugChar},
10197 @code{flush_i_cache}, @code{memset}, @code{exceptionHandler}.
10201 Insert these lines near the top of your program:
10209 For the 680x0 stub only, you need to provide a variable called
10210 @code{exceptionHook}. Normally you just use:
10213 void (*exceptionHook)() = 0;
10217 but if before calling @code{set_debug_traps}, you set it to point to a
10218 function in your program, that function is called when
10219 @code{@value{GDBN}} continues after stopping on a trap (for example, bus
10220 error). The function indicated by @code{exceptionHook} is called with
10221 one parameter: an @code{int} which is the exception number.
10224 Compile and link together: your program, the @value{GDBN} debugging stub for
10225 your target architecture, and the supporting subroutines.
10228 Make sure you have a serial connection between your target machine and
10229 the @value{GDBN} host, and identify the serial port on the host.
10232 @c The "remote" target now provides a `load' command, so we should
10233 @c document that. FIXME.
10234 Download your program to your target machine (or get it there by
10235 whatever means the manufacturer provides), and start it.
10238 To start remote debugging, run @value{GDBN} on the host machine, and specify
10239 as an executable file the program that is running in the remote machine.
10240 This tells @value{GDBN} how to find your program's symbols and the contents
10244 @cindex serial line, @code{target remote}
10245 Establish communication using the @code{target remote} command.
10246 Its argument specifies how to communicate with the target
10247 machine---either via a devicename attached to a direct serial line, or a
10248 TCP port (usually to a terminal server which in turn has a serial line
10249 to the target). For example, to use a serial line connected to the
10250 device named @file{/dev/ttyb}:
10253 target remote /dev/ttyb
10256 @cindex TCP port, @code{target remote}
10257 To use a TCP connection, use an argument of the form
10258 @code{@var{host}:port}. For example, to connect to port 2828 on a
10259 terminal server named @code{manyfarms}:
10262 target remote manyfarms:2828
10265 If your remote target is actually running on the same machine as
10266 your debugger session (e.g.@: a simulator of your target running on
10267 the same host), you can omit the hostname. For example, to connect
10268 to port 1234 on your local machine:
10271 target remote :1234
10275 Note that the colon is still required here.
10278 Now you can use all the usual commands to examine and change data and to
10279 step and continue the remote program.
10281 To resume the remote program and stop debugging it, use the @code{detach}
10284 @cindex interrupting remote programs
10285 @cindex remote programs, interrupting
10286 Whenever @value{GDBN} is waiting for the remote program, if you type the
10287 interrupt character (often @key{C-C}), @value{GDBN} attempts to stop the
10288 program. This may or may not succeed, depending in part on the hardware
10289 and the serial drivers the remote system uses. If you type the
10290 interrupt character once again, @value{GDBN} displays this prompt:
10293 Interrupted while waiting for the program.
10294 Give up (and stop debugging it)? (y or n)
10297 If you type @kbd{y}, @value{GDBN} abandons the remote debugging session.
10298 (If you decide you want to try again later, you can use @samp{target
10299 remote} again to connect once more.) If you type @kbd{n}, @value{GDBN}
10300 goes back to waiting.
10303 @subsubsection Communication protocol
10305 @cindex debugging stub, example
10306 @cindex remote stub, example
10307 @cindex stub example, remote debugging
10308 The stub files provided with @value{GDBN} implement the target side of the
10309 communication protocol, and the @value{GDBN} side is implemented in the
10310 @value{GDBN} source file @file{remote.c}. Normally, you can simply allow
10311 these subroutines to communicate, and ignore the details. (If you're
10312 implementing your own stub file, you can still ignore the details: start
10313 with one of the existing stub files. @file{sparc-stub.c} is the best
10314 organized, and therefore the easiest to read.)
10316 However, there may be occasions when you need to know something about
10317 the protocol---for example, if there is only one serial port to your
10318 target machine, you might want your program to do something special if
10319 it recognizes a packet meant for @value{GDBN}.
10321 In the examples below, @samp{<-} and @samp{->} are used to indicate
10322 transmitted and received data respectfully.
10324 @cindex protocol, @value{GDBN} remote serial
10325 @cindex serial protocol, @value{GDBN} remote
10326 @cindex remote serial protocol
10327 All @value{GDBN} commands and responses (other than acknowledgments) are
10328 sent as a @var{packet}. A @var{packet} is introduced with the character
10329 @samp{$}, the actual @var{packet-data}, and the terminating character
10330 @samp{#} followed by a two-digit @var{checksum}:
10333 @code{$}@var{packet-data}@code{#}@var{checksum}
10337 @cindex checksum, for @value{GDBN} remote
10339 The two-digit @var{checksum} is computed as the modulo 256 sum of all
10340 characters between the leading @samp{$} and the trailing @samp{#} (an
10341 eight bit unsigned checksum).
10343 Implementors should note that prior to @value{GDBN} 5.0 the protocol
10344 specification also included an optional two-digit @var{sequence-id}:
10347 @code{$}@var{sequence-id}@code{:}@var{packet-data}@code{#}@var{checksum}
10350 @cindex sequence-id, for @value{GDBN} remote
10352 That @var{sequence-id} was appended to the acknowledgment. @value{GDBN}
10353 has never output @var{sequence-id}s. Stubs that handle packets added
10354 since @value{GDBN} 5.0 must not accept @var{sequence-id}.
10356 @cindex acknowledgment, for @value{GDBN} remote
10357 When either the host or the target machine receives a packet, the first
10358 response expected is an acknowledgment: either @samp{+} (to indicate
10359 the package was received correctly) or @samp{-} (to request
10363 <- @code{$}@var{packet-data}@code{#}@var{checksum}
10368 The host (@value{GDBN}) sends @var{command}s, and the target (the
10369 debugging stub incorporated in your program) sends a @var{response}. In
10370 the case of step and continue @var{command}s, the response is only sent
10371 when the operation has completed (the target has again stopped).
10373 @var{packet-data} consists of a sequence of characters with the
10374 exception of @samp{#} and @samp{$} (see @samp{X} packet for additional
10377 Fields within the packet should be separated using @samp{,} @samp{;} or
10378 @samp{:}. Except where otherwise noted all numbers are represented in
10379 HEX with leading zeros suppressed.
10381 Implementors should note that prior to @value{GDBN} 5.0, the character
10382 @samp{:} could not appear as the third character in a packet (as it
10383 would potentially conflict with the @var{sequence-id}).
10385 Response @var{data} can be run-length encoded to save space. A @samp{*}
10386 means that the next character is an @sc{ascii} encoding giving a repeat count
10387 which stands for that many repetitions of the character preceding the
10388 @samp{*}. The encoding is @code{n+29}, yielding a printable character
10389 where @code{n >=3} (which is where rle starts to win). The printable
10390 characters @samp{$}, @samp{#}, @samp{+} and @samp{-} or with a numeric
10391 value greater than 126 should not be used.
10393 Some remote systems have used a different run-length encoding mechanism
10394 loosely refered to as the cisco encoding. Following the @samp{*}
10395 character are two hex digits that indicate the size of the packet.
10402 means the same as "0000".
10404 The error response returned for some packets includes a two character
10405 error number. That number is not well defined.
10407 For any @var{command} not supported by the stub, an empty response
10408 (@samp{$#00}) should be returned. That way it is possible to extend the
10409 protocol. A newer @value{GDBN} can tell if a packet is supported based
10412 A stub is required to support the @samp{g}, @samp{G}, @samp{m}, @samp{M},
10413 @samp{c}, and @samp{s} @var{command}s. All other @var{command}s are
10416 Below is a complete list of all currently defined @var{command}s and
10417 their corresponding response @var{data}:
10419 @multitable @columnfractions .30 .30 .40
10424 @item extended mode
10427 Enable extended mode. In extended mode, the remote server is made
10428 persistent. The @samp{R} packet is used to restart the program being
10431 @tab reply @samp{OK}
10433 The remote target both supports and has enabled extended mode.
10438 Indicate the reason the target halted. The reply is the same as for step
10447 @tab Reserved for future use
10449 @item set program arguments @strong{(reserved)}
10450 @tab @code{A}@var{arglen}@code{,}@var{argnum}@code{,}@var{arg}@code{,...}
10455 Initialized @samp{argv[]} array passed into program. @var{arglen}
10456 specifies the number of bytes in the hex encoded byte stream @var{arg}.
10457 See @file{gdbserver} for more details.
10459 @tab reply @code{OK}
10461 @tab reply @code{E}@var{NN}
10463 @item set baud @strong{(deprecated)}
10464 @tab @code{b}@var{baud}
10466 Change the serial line speed to @var{baud}. JTC: @emph{When does the
10467 transport layer state change? When it's received, or after the ACK is
10468 transmitted. In either case, there are problems if the command or the
10469 acknowledgment packet is dropped.} Stan: @emph{If people really wanted
10470 to add something like this, and get it working for the first time, they
10471 ought to modify ser-unix.c to send some kind of out-of-band message to a
10472 specially-setup stub and have the switch happen "in between" packets, so
10473 that from remote protocol's point of view, nothing actually
10476 @item set breakpoint @strong{(deprecated)}
10477 @tab @code{B}@var{addr},@var{mode}
10479 Set (@var{mode} is @samp{S}) or clear (@var{mode} is @samp{C}) a
10480 breakpoint at @var{addr}. @emph{This has been replaced by the @samp{Z} and
10484 @tab @code{c}@var{addr}
10486 @var{addr} is address to resume. If @var{addr} is omitted, resume at
10492 @item continue with signal
10493 @tab @code{C}@var{sig}@code{;}@var{addr}
10495 Continue with signal @var{sig} (hex signal number). If
10496 @code{;}@var{addr} is omitted, resume at same address.
10501 @item toggle debug @strong{(deprecated)}
10509 Detach @value{GDBN} from the remote system. Sent to the remote target before
10510 @value{GDBN} disconnects.
10512 @tab reply @emph{no response}
10514 @value{GDBN} does not check for any response after sending this packet.
10518 @tab Reserved for future use
10522 @tab Reserved for future use
10526 @tab Reserved for future use
10530 @tab Reserved for future use
10532 @item read registers
10534 @tab Read general registers.
10536 @tab reply @var{XX...}
10538 Each byte of register data is described by two hex digits. The bytes
10539 with the register are transmitted in target byte order. The size of
10540 each register and their position within the @samp{g} @var{packet} are
10541 determined by the @value{GDBN} internal macros @var{REGISTER_RAW_SIZE} and
10542 @var{REGISTER_NAME} macros. The specification of several standard
10543 @code{g} packets is specified below.
10545 @tab @code{E}@var{NN}
10549 @tab @code{G}@var{XX...}
10551 See @samp{g} for a description of the @var{XX...} data.
10553 @tab reply @code{OK}
10556 @tab reply @code{E}@var{NN}
10561 @tab Reserved for future use
10564 @tab @code{H}@var{c}@var{t...}
10566 Set thread for subsequent operations (@samp{m}, @samp{M}, @samp{g},
10567 @samp{G}, et.al.). @var{c} = @samp{c} for thread used in step and
10568 continue; @var{t...} can be -1 for all threads. @var{c} = @samp{g} for
10569 thread used in other operations. If zero, pick a thread, any thread.
10571 @tab reply @code{OK}
10574 @tab reply @code{E}@var{NN}
10578 @c 'H': How restrictive (or permissive) is the thread model. If a
10579 @c thread is selected and stopped, are other threads allowed
10580 @c to continue to execute? As I mentioned above, I think the
10581 @c semantics of each command when a thread is selected must be
10582 @c described. For example:
10584 @c 'g': If the stub supports threads and a specific thread is
10585 @c selected, returns the register block from that thread;
10586 @c otherwise returns current registers.
10588 @c 'G' If the stub supports threads and a specific thread is
10589 @c selected, sets the registers of the register block of
10590 @c that thread; otherwise sets current registers.
10592 @item cycle step @strong{(draft)}
10593 @tab @code{i}@var{addr}@code{,}@var{nnn}
10595 Step the remote target by a single clock cycle. If @code{,}@var{nnn} is
10596 present, cycle step @var{nnn} cycles. If @var{addr} is present, cycle
10597 step starting at that address.
10599 @item signal then cycle step @strong{(reserved)}
10602 See @samp{i} and @samp{S} for likely syntax and semantics.
10606 @tab Reserved for future use
10610 @tab Reserved for future use
10615 FIXME: @emph{There is no description of how operate when a specific
10616 thread context has been selected (ie. does 'k' kill only that thread?)}.
10620 @tab Reserved for future use
10624 @tab Reserved for future use
10627 @tab @code{m}@var{addr}@code{,}@var{length}
10629 Read @var{length} bytes of memory starting at address @var{addr}.
10630 Neither @value{GDBN} nor the stub assume that sized memory transfers are assumed
10631 using word alligned accesses. FIXME: @emph{A word aligned memory
10632 transfer mechanism is needed.}
10634 @tab reply @var{XX...}
10636 @var{XX...} is mem contents. Can be fewer bytes than requested if able
10637 to read only part of the data. Neither @value{GDBN} nor the stub assume that
10638 sized memory transfers are assumed using word alligned accesses. FIXME:
10639 @emph{A word aligned memory transfer mechanism is needed.}
10641 @tab reply @code{E}@var{NN}
10642 @tab @var{NN} is errno
10645 @tab @code{M}@var{addr},@var{length}@code{:}@var{XX...}
10647 Write @var{length} bytes of memory starting at address @var{addr}.
10648 @var{XX...} is the data.
10650 @tab reply @code{OK}
10653 @tab reply @code{E}@var{NN}
10655 for an error (this includes the case where only part of the data was
10660 @tab Reserved for future use
10664 @tab Reserved for future use
10668 @tab Reserved for future use
10672 @tab Reserved for future use
10674 @item read reg @strong{(reserved)}
10675 @tab @code{p}@var{n...}
10677 See write register.
10679 @tab return @var{r....}
10680 @tab The hex encoded value of the register in target byte order.
10683 @tab @code{P}@var{n...}@code{=}@var{r...}
10685 Write register @var{n...} with value @var{r...}, which contains two hex
10686 digits for each byte in the register (target byte order).
10688 @tab reply @code{OK}
10691 @tab reply @code{E}@var{NN}
10694 @item general query
10695 @tab @code{q}@var{query}
10697 Request info about @var{query}. In general @value{GDBN} queries
10698 have a leading upper case letter. Custom vendor queries should use a
10699 company prefix (in lower case) ex: @samp{qfsf.var}. @var{query} may
10700 optionally be followed by a @samp{,} or @samp{;} separated list. Stubs
10701 must ensure that they match the full @var{query} name.
10703 @tab reply @code{XX...}
10704 @tab Hex encoded data from query. The reply can not be empty.
10706 @tab reply @code{E}@var{NN}
10710 @tab Indicating an unrecognized @var{query}.
10713 @tab @code{Q}@var{var}@code{=}@var{val}
10715 Set value of @var{var} to @var{val}. See @samp{q} for a discussing of
10716 naming conventions.
10718 @item reset @strong{(deprecated)}
10721 Reset the entire system.
10723 @item remote restart
10724 @tab @code{R}@var{XX}
10726 Restart the program being debugged. @var{XX}, while needed, is ignored.
10727 This packet is only available in extended mode.
10732 The @samp{R} packet has no reply.
10735 @tab @code{s}@var{addr}
10737 @var{addr} is address to resume. If @var{addr} is omitted, resume at
10743 @item step with signal
10744 @tab @code{S}@var{sig}@code{;}@var{addr}
10746 Like @samp{C} but step not continue.
10752 @tab @code{t}@var{addr}@code{:}@var{PP}@code{,}@var{MM}
10754 Search backwards starting at address @var{addr} for a match with pattern
10755 @var{PP} and mask @var{MM}. @var{PP} and @var{MM} are 4
10756 bytes. @var{addr} must be at least 3 digits.
10759 @tab @code{T}@var{XX}
10760 @tab Find out if the thread XX is alive.
10762 @tab reply @code{OK}
10763 @tab thread is still alive
10765 @tab reply @code{E}@var{NN}
10766 @tab thread is dead
10770 @tab Reserved for future use
10774 @tab Reserved for future use
10778 @tab Reserved for future use
10782 @tab Reserved for future use
10786 @tab Reserved for future use
10790 @tab Reserved for future use
10794 @tab Reserved for future use
10796 @item write mem (binary)
10797 @tab @code{X}@var{addr}@code{,}@var{length}@var{:}@var{XX...}
10799 @var{addr} is address, @var{length} is number of bytes, @var{XX...} is
10800 binary data. The characters @code{$}, @code{#}, and @code{0x7d} are
10801 escaped using @code{0x7d}.
10803 @tab reply @code{OK}
10806 @tab reply @code{E}@var{NN}
10811 @tab Reserved for future use
10815 @tab Reserved for future use
10817 @item remove break or watchpoint @strong{(draft)}
10818 @tab @code{z}@var{t}@code{,}@var{addr}@code{,}@var{length}
10822 @item insert break or watchpoint @strong{(draft)}
10823 @tab @code{Z}@var{t}@code{,}@var{addr}@code{,}@var{length}
10825 @var{t} is type: @samp{0} - software breakpoint, @samp{1} - hardware
10826 breakpoint, @samp{2} - write watchpoint, @samp{3} - read watchpoint,
10827 @samp{4} - access watchpoint; @var{addr} is address; @var{length} is in
10828 bytes. For a software breakpoint, @var{length} specifies the size of
10829 the instruction to be patched. For hardware breakpoints and watchpoints
10830 @var{length} specifies the memory region to be monitored. To avoid
10831 potential problems with duplicate packets, the operations should be
10832 implemented in an idempotent way.
10834 @tab reply @code{E}@var{NN}
10837 @tab reply @code{OK}
10841 @tab If not supported.
10845 @tab Reserved for future use
10849 The @samp{C}, @samp{c}, @samp{S}, @samp{s} and @samp{?} packets can
10850 receive any of the below as a reply. In the case of the @samp{C},
10851 @samp{c}, @samp{S} and @samp{s} packets, that reply is only returned
10852 when the target halts. In the below the exact meaning of @samp{signal
10853 number} is poorly defined. In general one of the UNIX signal numbering
10854 conventions is used.
10856 @multitable @columnfractions .4 .6
10858 @item @code{S}@var{AA}
10859 @tab @var{AA} is the signal number
10861 @item @code{T}@var{AA}@var{n...}@code{:}@var{r...}@code{;}@var{n...}@code{:}@var{r...}@code{;}@var{n...}@code{:}@var{r...}@code{;}
10863 @var{AA} = two hex digit signal number; @var{n...} = register number
10864 (hex), @var{r...} = target byte ordered register contents, size defined
10865 by @code{REGISTER_RAW_SIZE}; @var{n...} = @samp{thread}, @var{r...} =
10866 thread process ID, this is a hex integer; @var{n...} = other string not
10867 starting with valid hex digit. @value{GDBN} should ignore this
10868 @var{n...}, @var{r...} pair and go on to the next. This way we can
10869 extend the protocol.
10871 @item @code{W}@var{AA}
10873 The process exited, and @var{AA} is the exit status. This is only
10874 applicable for certains sorts of targets.
10876 @item @code{X}@var{AA}
10878 The process terminated with signal @var{AA}.
10880 @item @code{N}@var{AA}@code{;}@var{t...}@code{;}@var{d...}@code{;}@var{b...} @strong{(obsolete)}
10882 @var{AA} = signal number; @var{t...} = address of symbol "_start";
10883 @var{d...} = base of data section; @var{b...} = base of bss section.
10884 @emph{Note: only used by Cisco Systems targets. The difference between
10885 this reply and the "qOffsets" query is that the 'N' packet may arrive
10886 spontaneously whereas the 'qOffsets' is a query initiated by the host
10889 @item @code{O}@var{XX...}
10891 @var{XX...} is hex encoding of @sc{ascii} data. This can happen at any time
10892 while the program is running and the debugger should continue to wait
10897 The following set and query packets have already been defined.
10899 @multitable @columnfractions .2 .2 .6
10901 @item current thread
10902 @tab @code{q}@code{C}
10903 @tab Return the current thread id.
10905 @tab reply @code{QC}@var{pid}
10907 Where @var{pid} is a HEX encoded 16 bit process id.
10910 @tab Any other reply implies the old pid.
10912 @item all thread ids
10913 @tab @code{q}@code{fThreadInfo}
10915 @tab @code{q}@code{sThreadInfo}
10917 Obtain a list of active thread ids from the target (OS). Since there
10918 may be too many active threads to fit into one reply packet, this query
10919 works iteratively: it may require more than one query/reply sequence to
10920 obtain the entire list of threads. The first query of the sequence will
10921 be the @code{qf}@code{ThreadInfo} query; subsequent queries in the
10922 sequence will be the @code{qs}@code{ThreadInfo} query.
10925 @tab NOTE: replaces the @code{qL} query (see below).
10927 @tab reply @code{m}@var{<id>}
10928 @tab A single thread id
10930 @tab reply @code{m}@var{<id>},@var{<id>...}
10931 @tab a comma-separated list of thread ids
10933 @tab reply @code{l}
10934 @tab (lower case 'el') denotes end of list.
10938 In response to each query, the target will reply with a list of one
10939 or more thread ids, in big-endian hex, separated by commas. GDB will
10940 respond to each reply with a request for more thread ids (using the
10941 @code{qs} form of the query), until the target responds with @code{l}
10942 (lower-case el, for @code{'last'}).
10944 @item extra thread info
10945 @tab @code{q}@code{ThreadExtraInfo}@code{,}@var{id}
10950 Where @var{<id>} is a thread-id in big-endian hex.
10951 Obtain a printable string description of a thread's attributes from
10952 the target OS. This string may contain anything that the target OS
10953 thinks is interesting for @value{GDBN} to tell the user about the thread.
10954 The string is displayed in @value{GDBN}'s @samp{info threads} display.
10955 Some examples of possible thread extra info strings are "Runnable", or
10956 "Blocked on Mutex".
10958 @tab reply @var{XX...}
10960 Where @var{XX...} is a hex encoding of @sc{ascii} data, comprising the
10961 printable string containing the extra information about the thread's
10964 @item query @var{LIST} or @var{threadLIST} @strong{(deprecated)}
10965 @tab @code{q}@code{L}@var{startflag}@var{threadcount}@var{nextthread}
10970 Obtain thread information from RTOS. Where: @var{startflag} (one hex
10971 digit) is one to indicate the first query and zero to indicate a
10972 subsequent query; @var{threadcount} (two hex digits) is the maximum
10973 number of threads the response packet can contain; and @var{nextthread}
10974 (eight hex digits), for subsequent queries (@var{startflag} is zero), is
10975 returned in the response as @var{argthread}.
10978 @tab NOTE: this query is replaced by the @code{q}@code{fThreadInfo}
10981 @tab reply @code{q}@code{M}@var{count}@var{done}@var{argthread}@var{thread...}
10986 Where: @var{count} (two hex digits) is the number of threads being
10987 returned; @var{done} (one hex digit) is zero to indicate more threads
10988 and one indicates no further threads; @var{argthreadid} (eight hex
10989 digits) is @var{nextthread} from the request packet; @var{thread...} is
10990 a sequence of thread IDs from the target. @var{threadid} (eight hex
10991 digits). See @code{remote.c:parse_threadlist_response()}.
10993 @item compute CRC of memory block
10994 @tab @code{q}@code{CRC:}@var{addr}@code{,}@var{length}
10997 @tab reply @code{E}@var{NN}
10998 @tab An error (such as memory fault)
11000 @tab reply @code{C}@var{CRC32}
11001 @tab A 32 bit cyclic redundancy check of the specified memory region.
11003 @item query sect offs
11004 @tab @code{q}@code{Offsets}
11006 Get section offsets that the target used when re-locating the downloaded
11007 image. @emph{Note: while a @code{Bss} offset is included in the
11008 response, @value{GDBN} ignores this and instead applies the @code{Data}
11009 offset to the @code{Bss} section.}
11011 @tab reply @code{Text=}@var{xxx}@code{;Data=}@var{yyy}@code{;Bss=}@var{zzz}
11013 @item thread info request
11014 @tab @code{q}@code{P}@var{mode}@var{threadid}
11019 Returns information on @var{threadid}. Where: @var{mode} is a hex
11020 encoded 32 bit mode; @var{threadid} is a hex encoded 64 bit thread ID.
11024 See @code{remote.c:remote_unpack_thread_info_response()}.
11026 @item remote command
11027 @tab @code{q}@code{Rcmd,}@var{COMMAND}
11032 @var{COMMAND} (hex encoded) is passed to the local interpreter for
11033 execution. Invalid commands should be reported using the output string.
11034 Before the final result packet, the target may also respond with a
11035 number of intermediate @code{O}@var{OUTPUT} console output
11036 packets. @emph{Implementors should note that providing access to a
11037 stubs's interpreter may have security implications}.
11039 @tab reply @code{OK}
11041 A command response with no output.
11043 @tab reply @var{OUTPUT}
11045 A command response with the hex encoded output string @var{OUTPUT}.
11047 @tab reply @code{E}@var{NN}
11049 Indicate a badly formed request.
11054 When @samp{q}@samp{Rcmd} is not recognized.
11056 @item symbol lookup
11057 @tab @code{qSymbol::}
11059 Notify the target that @value{GDBN} is prepared to serve symbol lookup
11060 requests. Accept requests from the target for the values of symbols.
11065 @tab reply @code{OK}
11067 The target does not need to look up any (more) symbols.
11069 @tab reply @code{qSymbol:}@var{sym_name}
11073 The target requests the value of symbol @var{sym_name} (hex encoded).
11074 @value{GDBN} may provide the value by using the
11075 @code{qSymbol:}@var{sym_value}:@var{sym_name}
11076 message, described below.
11079 @tab @code{qSymbol:}@var{sym_value}:@var{sym_name}
11083 Set the value of SYM_NAME to SYM_VALUE.
11087 @var{sym_name} (hex encoded) is the name of a symbol whose value
11088 the target has previously requested.
11092 @var{sym_value} (hex) is the value for symbol @var{sym_name}.
11093 If @value{GDBN} cannot supply a value for @var{sym_name}, then this
11094 field will be empty.
11096 @tab reply @code{OK}
11098 The target does not need to look up any (more) symbols.
11100 @tab reply @code{qSymbol:}@var{sym_name}
11104 The target requests the value of a new symbol @var{sym_name} (hex encoded).
11105 @value{GDBN} will continue to supply the values of symbols (if available),
11106 until the target ceases to request them.
11110 The following @samp{g}/@samp{G} packets have previously been defined.
11111 In the below, some thirty-two bit registers are transferred as sixty-four
11112 bits. Those registers should be zero/sign extended (which?) to fill the
11113 space allocated. Register bytes are transfered in target byte order.
11114 The two nibbles within a register byte are transfered most-significant -
11117 @multitable @columnfractions .5 .5
11121 All registers are transfered as thirty-two bit quantities in the order:
11122 32 general-purpose; sr; lo; hi; bad; cause; pc; 32 floating-point
11123 registers; fsr; fir; fp.
11127 All registers are transfered as sixty-four bit quantities (including
11128 thirty-two bit registers such as @code{sr}). The ordering is the same
11133 Example sequence of a target being re-started. Notice how the restart
11134 does not get any direct output:
11139 @emph{target restarts}
11142 -> @code{T001:1234123412341234}
11146 Example sequence of a target being stepped by a single instruction:
11154 -> @code{T001:1234123412341234}
11163 @subsubsection Using the @code{gdbserver} program
11166 @cindex remote connection without stubs
11167 @code{gdbserver} is a control program for Unix-like systems, which
11168 allows you to connect your program with a remote @value{GDBN} via
11169 @code{target remote}---but without linking in the usual debugging stub.
11171 @code{gdbserver} is not a complete replacement for the debugging stubs,
11172 because it requires essentially the same operating-system facilities
11173 that @value{GDBN} itself does. In fact, a system that can run
11174 @code{gdbserver} to connect to a remote @value{GDBN} could also run
11175 @value{GDBN} locally! @code{gdbserver} is sometimes useful nevertheless,
11176 because it is a much smaller program than @value{GDBN} itself. It is
11177 also easier to port than all of @value{GDBN}, so you may be able to get
11178 started more quickly on a new system by using @code{gdbserver}.
11179 Finally, if you develop code for real-time systems, you may find that
11180 the tradeoffs involved in real-time operation make it more convenient to
11181 do as much development work as possible on another system, for example
11182 by cross-compiling. You can use @code{gdbserver} to make a similar
11183 choice for debugging.
11185 @value{GDBN} and @code{gdbserver} communicate via either a serial line
11186 or a TCP connection, using the standard @value{GDBN} remote serial
11190 @item On the target machine,
11191 you need to have a copy of the program you want to debug.
11192 @code{gdbserver} does not need your program's symbol table, so you can
11193 strip the program if necessary to save space. @value{GDBN} on the host
11194 system does all the symbol handling.
11196 To use the server, you must tell it how to communicate with @value{GDBN};
11197 the name of your program; and the arguments for your program. The
11201 target> gdbserver @var{comm} @var{program} [ @var{args} @dots{} ]
11204 @var{comm} is either a device name (to use a serial line) or a TCP
11205 hostname and portnumber. For example, to debug Emacs with the argument
11206 @samp{foo.txt} and communicate with @value{GDBN} over the serial port
11210 target> gdbserver /dev/com1 emacs foo.txt
11213 @code{gdbserver} waits passively for the host @value{GDBN} to communicate
11216 To use a TCP connection instead of a serial line:
11219 target> gdbserver host:2345 emacs foo.txt
11222 The only difference from the previous example is the first argument,
11223 specifying that you are communicating with the host @value{GDBN} via
11224 TCP. The @samp{host:2345} argument means that @code{gdbserver} is to
11225 expect a TCP connection from machine @samp{host} to local TCP port 2345.
11226 (Currently, the @samp{host} part is ignored.) You can choose any number
11227 you want for the port number as long as it does not conflict with any
11228 TCP ports already in use on the target system (for example, @code{23} is
11229 reserved for @code{telnet}).@footnote{If you choose a port number that
11230 conflicts with another service, @code{gdbserver} prints an error message
11231 and exits.} You must use the same port number with the host @value{GDBN}
11232 @code{target remote} command.
11234 @item On the @value{GDBN} host machine,
11235 you need an unstripped copy of your program, since @value{GDBN} needs
11236 symbols and debugging information. Start up @value{GDBN} as usual,
11237 using the name of the local copy of your program as the first argument.
11238 (You may also need the @w{@samp{--baud}} option if the serial line is
11239 running at anything other than 9600@dmn{bps}.) After that, use @code{target
11240 remote} to establish communications with @code{gdbserver}. Its argument
11241 is either a device name (usually a serial device, like
11242 @file{/dev/ttyb}), or a TCP port descriptor in the form
11243 @code{@var{host}:@var{PORT}}. For example:
11246 (@value{GDBP}) target remote /dev/ttyb
11250 communicates with the server via serial line @file{/dev/ttyb}, and
11253 (@value{GDBP}) target remote the-target:2345
11257 communicates via a TCP connection to port 2345 on host @w{@file{the-target}}.
11258 For TCP connections, you must start up @code{gdbserver} prior to using
11259 the @code{target remote} command. Otherwise you may get an error whose
11260 text depends on the host system, but which usually looks something like
11261 @samp{Connection refused}.
11265 @subsubsection Using the @code{gdbserve.nlm} program
11267 @kindex gdbserve.nlm
11268 @code{gdbserve.nlm} is a control program for NetWare systems, which
11269 allows you to connect your program with a remote @value{GDBN} via
11270 @code{target remote}.
11272 @value{GDBN} and @code{gdbserve.nlm} communicate via a serial line,
11273 using the standard @value{GDBN} remote serial protocol.
11276 @item On the target machine,
11277 you need to have a copy of the program you want to debug.
11278 @code{gdbserve.nlm} does not need your program's symbol table, so you
11279 can strip the program if necessary to save space. @value{GDBN} on the
11280 host system does all the symbol handling.
11282 To use the server, you must tell it how to communicate with
11283 @value{GDBN}; the name of your program; and the arguments for your
11284 program. The syntax is:
11287 load gdbserve [ BOARD=@var{board} ] [ PORT=@var{port} ]
11288 [ BAUD=@var{baud} ] @var{program} [ @var{args} @dots{} ]
11291 @var{board} and @var{port} specify the serial line; @var{baud} specifies
11292 the baud rate used by the connection. @var{port} and @var{node} default
11293 to 0, @var{baud} defaults to 9600@dmn{bps}.
11295 For example, to debug Emacs with the argument @samp{foo.txt}and
11296 communicate with @value{GDBN} over serial port number 2 or board 1
11297 using a 19200@dmn{bps} connection:
11300 load gdbserve BOARD=1 PORT=2 BAUD=19200 emacs foo.txt
11303 @item On the @value{GDBN} host machine,
11304 you need an unstripped copy of your program, since @value{GDBN} needs
11305 symbols and debugging information. Start up @value{GDBN} as usual,
11306 using the name of the local copy of your program as the first argument.
11307 (You may also need the @w{@samp{--baud}} option if the serial line is
11308 running at anything other than 9600@dmn{bps}. After that, use @code{target
11309 remote} to establish communications with @code{gdbserve.nlm}. Its
11310 argument is a device name (usually a serial device, like
11311 @file{/dev/ttyb}). For example:
11314 (@value{GDBP}) target remote /dev/ttyb
11318 communications with the server via serial line @file{/dev/ttyb}.
11322 @section Kernel Object Display
11324 @cindex kernel object display
11325 @cindex kernel object
11328 Some targets support kernel object display. Using this facility,
11329 @value{GDBN} communicates specially with the underlying operating system
11330 and can display information about operating system-level objects such as
11331 mutexes and other synchronization objects. Exactly which objects can be
11332 displayed is determined on a per-OS basis.
11334 Use the @code{set os} command to set the operating system. This tells
11335 @value{GDBN} which kernel object display module to initialize:
11338 (@value{GDBP}) set os cisco
11341 If @code{set os} succeeds, @value{GDBN} will display some information
11342 about the operating system, and will create a new @code{info} command
11343 which can be used to query the target. The @code{info} command is named
11344 after the operating system:
11347 (@value{GDBP}) info cisco
11348 List of Cisco Kernel Objects
11350 any Any and all objects
11353 Further subcommands can be used to query about particular objects known
11356 There is currently no way to determine whether a given operating system
11357 is supported other than to try it.
11360 @node Configurations
11361 @chapter Configuration-Specific Information
11363 While nearly all @value{GDBN} commands are available for all native and
11364 cross versions of the debugger, there are some exceptions. This chapter
11365 describes things that are only available in certain configurations.
11367 There are three major categories of configurations: native
11368 configurations, where the host and target are the same, embedded
11369 operating system configurations, which are usually the same for several
11370 different processor architectures, and bare embedded processors, which
11371 are quite different from each other.
11376 * Embedded Processors::
11383 This section describes details specific to particular native
11388 * SVR4 Process Information:: SVR4 process information
11389 * DJGPP Native:: Features specific to the DJGPP port
11395 On HP-UX systems, if you refer to a function or variable name that
11396 begins with a dollar sign, @value{GDBN} searches for a user or system
11397 name first, before it searches for a convenience variable.
11399 @node SVR4 Process Information
11400 @subsection SVR4 process information
11403 @cindex process image
11405 Many versions of SVR4 provide a facility called @samp{/proc} that can be
11406 used to examine the image of a running process using file-system
11407 subroutines. If @value{GDBN} is configured for an operating system with
11408 this facility, the command @code{info proc} is available to report on
11409 several kinds of information about the process running your program.
11410 @code{info proc} works only on SVR4 systems that include the
11411 @code{procfs} code. This includes OSF/1 (Digital Unix), Solaris, Irix,
11412 and Unixware, but not HP-UX or Linux, for example.
11417 Summarize available information about the process.
11419 @kindex info proc mappings
11420 @item info proc mappings
11421 Report on the address ranges accessible in the program, with information
11422 on whether your program may read, write, or execute each range.
11424 @comment These sub-options of 'info proc' were not included when
11425 @comment procfs.c was re-written. Keep their descriptions around
11426 @comment against the day when someone finds the time to put them back in.
11427 @kindex info proc times
11428 @item info proc times
11429 Starting time, user CPU time, and system CPU time for your program and
11432 @kindex info proc id
11434 Report on the process IDs related to your program: its own process ID,
11435 the ID of its parent, the process group ID, and the session ID.
11437 @kindex info proc status
11438 @item info proc status
11439 General information on the state of the process. If the process is
11440 stopped, this report includes the reason for stopping, and any signal
11443 @item info proc all
11444 Show all the above information about the process.
11449 @subsection Features for Debugging @sc{djgpp} Programs
11450 @cindex @sc{djgpp} debugging
11451 @cindex native @sc{djgpp} debugging
11452 @cindex MS-DOS-specific commands
11454 @sc{djgpp} is the port of @sc{gnu} development tools to MS-DOS and
11455 MS-Windows. @sc{djgpp} programs are 32-bit protected-mode programs
11456 that use the @dfn{DPMI} (DOS Protected-Mode Interface) API to run on
11457 top of real-mode DOS systems and their emulations.
11459 @value{GDBN} supports native debugging of @sc{djgpp} programs, and
11460 defines a few commands specific to the @sc{djgpp} port. This
11461 subsection describes those commands.
11466 This is a prefix of @sc{djgpp}-specific commands which print
11467 information about the target system and important OS structures.
11470 @cindex MS-DOS system info
11471 @cindex free memory information (MS-DOS)
11472 @item info dos sysinfo
11473 This command displays assorted information about the underlying
11474 platform: the CPU type and features, the OS version and flavor, the
11475 DPMI version, and the available conventional and DPMI memory.
11480 @cindex segment descriptor tables
11481 @cindex descriptor tables display
11483 @itemx info dos ldt
11484 @itemx info dos idt
11485 These 3 commands display entries from, respectively, Global, Local,
11486 and Interrupt Descriptor Tables (GDT, LDT, and IDT). The descriptor
11487 tables are data structures which store a descriptor for each segment
11488 that is currently in use. The segment's selector is an index into a
11489 descriptor table; the table entry for that index holds the
11490 descriptor's base address and limit, and its attributes and access
11493 A typical @sc{djgpp} program uses 3 segments: a code segment, a data
11494 segment (used for both data and the stack), and a DOS segment (which
11495 allows access to DOS/BIOS data structures and absolute addresses in
11496 conventional memory). However, the DPMI host will usually define
11497 additional segments in order to support the DPMI environment.
11499 @cindex garbled pointers
11500 These commands allow to display entries from the descriptor tables.
11501 Without an argument, all entries from the specified table are
11502 displayed. An argument, which should be an integer expression, means
11503 display a single entry whose index is given by the argument. For
11504 example, here's a convenient way to display information about the
11505 debugged program's data segment:
11508 (@value{GDBP}) info dos ldt $ds
11509 0x13f: base=0x11970000 limit=0x0009ffff 32-Bit Data (Read/Write, Exp-up)
11513 This comes in handy when you want to see whether a pointer is outside
11514 the data segment's limit (i.e.@: @dfn{garbled}).
11516 @cindex page tables display (MS-DOS)
11518 @itemx info dos pte
11519 These two commands display entries from, respectively, the Page
11520 Directory and the Page Tables. Page Directories and Page Tables are
11521 data structures which control how virtual memory addresses are mapped
11522 into physical addresses. A Page Table includes an entry for every
11523 page of memory that is mapped into the program's address space; there
11524 may be several Page Tables, each one holding up to 4096 entries. A
11525 Page Directory has up to 4096 entries, one each for every Page Table
11526 that is currently in use.
11528 Without an argument, @kbd{info dos pde} displays the entire Page
11529 Directory, and @kbd{info dos pte} displays all the entries in all of
11530 the Page Tables. An argument, an integer expression, given to the
11531 @kbd{info dos pde} command means display only that entry from the Page
11532 Directory table. An argument given to the @kbd{info dos pte} command
11533 means display entries from a single Page Table, the one pointed to by
11534 the specified entry in the Page Directory.
11536 These commands are useful when your program uses @dfn{DMA} (Direct
11537 Memory Access), which needs physical addresses to program the DMA
11540 These commands are supported only with some DPMI servers.
11542 @cindex physical address from linear address
11543 @item info dos address-pte
11544 This command displays the Page Table entry for a specified linear
11545 address. The argument linear address should already have the
11546 appropriate segment's base address added to it, because this command
11547 accepts addresses which may belong to @emph{any} segment. For
11548 example, here's how to display the Page Table entry for the page where
11549 the variable @code{i} is stored:
11552 (@value{GDBP}) info dos address-pte __djgpp_base_address + (char *)&i
11553 Page Table entry for address 0x11a00d30:
11554 Base=0x02698000 Dirty Acc. Not-Cached Write-Back Usr Read-Write +0xd30
11558 This says that @code{i} is stored at offset @code{0xd30} from the page
11559 whose physical base address is @code{0x02698000}, and prints all the
11560 attributes of that page.
11562 Note that you must cast the addresses of variables to a @code{char *},
11563 since otherwise the value of @code{__djgpp_base_address}, the base
11564 address of all variables and functions in a @sc{djgpp} program, will
11565 be added using the rules of C pointer arithmetics: if @code{i} is
11566 declared an @code{int}, @value{GDBN} will add 4 times the value of
11567 @code{__djgpp_base_address} to the address of @code{i}.
11569 Here's another example, it displays the Page Table entry for the
11573 (@value{GDBP}) info dos address-pte *((unsigned *)&_go32_info_block + 3)
11574 Page Table entry for address 0x29110:
11575 Base=0x00029000 Dirty Acc. Not-Cached Write-Back Usr Read-Write +0x110
11579 (The @code{+ 3} offset is because the transfer buffer's address is the
11580 3rd member of the @code{_go32_info_block} structure.) The output of
11581 this command clearly shows that addresses in conventional memory are
11582 mapped 1:1, i.e.@: the physical and linear addresses are identical.
11584 This command is supported only with some DPMI servers.
11588 @section Embedded Operating Systems
11590 This section describes configurations involving the debugging of
11591 embedded operating systems that are available for several different
11595 * VxWorks:: Using @value{GDBN} with VxWorks
11598 @value{GDBN} includes the ability to debug programs running on
11599 various real-time operating systems.
11602 @subsection Using @value{GDBN} with VxWorks
11608 @kindex target vxworks
11609 @item target vxworks @var{machinename}
11610 A VxWorks system, attached via TCP/IP. The argument @var{machinename}
11611 is the target system's machine name or IP address.
11615 On VxWorks, @code{load} links @var{filename} dynamically on the
11616 current target system as well as adding its symbols in @value{GDBN}.
11618 @value{GDBN} enables developers to spawn and debug tasks running on networked
11619 VxWorks targets from a Unix host. Already-running tasks spawned from
11620 the VxWorks shell can also be debugged. @value{GDBN} uses code that runs on
11621 both the Unix host and on the VxWorks target. The program
11622 @code{@value{GDBP}} is installed and executed on the Unix host. (It may be
11623 installed with the name @code{vxgdb}, to distinguish it from a
11624 @value{GDBN} for debugging programs on the host itself.)
11627 @item VxWorks-timeout @var{args}
11628 @kindex vxworks-timeout
11629 All VxWorks-based targets now support the option @code{vxworks-timeout}.
11630 This option is set by the user, and @var{args} represents the number of
11631 seconds @value{GDBN} waits for responses to rpc's. You might use this if
11632 your VxWorks target is a slow software simulator or is on the far side
11633 of a thin network line.
11636 The following information on connecting to VxWorks was current when
11637 this manual was produced; newer releases of VxWorks may use revised
11640 @kindex INCLUDE_RDB
11641 To use @value{GDBN} with VxWorks, you must rebuild your VxWorks kernel
11642 to include the remote debugging interface routines in the VxWorks
11643 library @file{rdb.a}. To do this, define @code{INCLUDE_RDB} in the
11644 VxWorks configuration file @file{configAll.h} and rebuild your VxWorks
11645 kernel. The resulting kernel contains @file{rdb.a}, and spawns the
11646 source debugging task @code{tRdbTask} when VxWorks is booted. For more
11647 information on configuring and remaking VxWorks, see the manufacturer's
11649 @c VxWorks, see the @cite{VxWorks Programmer's Guide}.
11651 Once you have included @file{rdb.a} in your VxWorks system image and set
11652 your Unix execution search path to find @value{GDBN}, you are ready to
11653 run @value{GDBN}. From your Unix host, run @code{@value{GDBP}} (or
11654 @code{vxgdb}, depending on your installation).
11656 @value{GDBN} comes up showing the prompt:
11663 * VxWorks Connection:: Connecting to VxWorks
11664 * VxWorks Download:: VxWorks download
11665 * VxWorks Attach:: Running tasks
11668 @node VxWorks Connection
11669 @subsubsection Connecting to VxWorks
11671 The @value{GDBN} command @code{target} lets you connect to a VxWorks target on the
11672 network. To connect to a target whose host name is ``@code{tt}'', type:
11675 (vxgdb) target vxworks tt
11679 @value{GDBN} displays messages like these:
11682 Attaching remote machine across net...
11687 @value{GDBN} then attempts to read the symbol tables of any object modules
11688 loaded into the VxWorks target since it was last booted. @value{GDBN} locates
11689 these files by searching the directories listed in the command search
11690 path (@pxref{Environment, ,Your program's environment}); if it fails
11691 to find an object file, it displays a message such as:
11694 prog.o: No such file or directory.
11697 When this happens, add the appropriate directory to the search path with
11698 the @value{GDBN} command @code{path}, and execute the @code{target}
11701 @node VxWorks Download
11702 @subsubsection VxWorks download
11704 @cindex download to VxWorks
11705 If you have connected to the VxWorks target and you want to debug an
11706 object that has not yet been loaded, you can use the @value{GDBN}
11707 @code{load} command to download a file from Unix to VxWorks
11708 incrementally. The object file given as an argument to the @code{load}
11709 command is actually opened twice: first by the VxWorks target in order
11710 to download the code, then by @value{GDBN} in order to read the symbol
11711 table. This can lead to problems if the current working directories on
11712 the two systems differ. If both systems have NFS mounted the same
11713 filesystems, you can avoid these problems by using absolute paths.
11714 Otherwise, it is simplest to set the working directory on both systems
11715 to the directory in which the object file resides, and then to reference
11716 the file by its name, without any path. For instance, a program
11717 @file{prog.o} may reside in @file{@var{vxpath}/vw/demo/rdb} in VxWorks
11718 and in @file{@var{hostpath}/vw/demo/rdb} on the host. To load this
11719 program, type this on VxWorks:
11722 -> cd "@var{vxpath}/vw/demo/rdb"
11726 Then, in @value{GDBN}, type:
11729 (vxgdb) cd @var{hostpath}/vw/demo/rdb
11730 (vxgdb) load prog.o
11733 @value{GDBN} displays a response similar to this:
11736 Reading symbol data from wherever/vw/demo/rdb/prog.o... done.
11739 You can also use the @code{load} command to reload an object module
11740 after editing and recompiling the corresponding source file. Note that
11741 this makes @value{GDBN} delete all currently-defined breakpoints,
11742 auto-displays, and convenience variables, and to clear the value
11743 history. (This is necessary in order to preserve the integrity of
11744 debugger's data structures that reference the target system's symbol
11747 @node VxWorks Attach
11748 @subsubsection Running tasks
11750 @cindex running VxWorks tasks
11751 You can also attach to an existing task using the @code{attach} command as
11755 (vxgdb) attach @var{task}
11759 where @var{task} is the VxWorks hexadecimal task ID. The task can be running
11760 or suspended when you attach to it. Running tasks are suspended at
11761 the time of attachment.
11763 @node Embedded Processors
11764 @section Embedded Processors
11766 This section goes into details specific to particular embedded
11770 @c OBSOLETE * A29K Embedded:: AMD A29K Embedded
11773 * H8/300:: Hitachi H8/300
11774 * H8/500:: Hitachi H8/500
11775 * i960:: Intel i960
11776 * M32R/D:: Mitsubishi M32R/D
11777 * M68K:: Motorola M68K
11778 * M88K:: Motorola M88K
11779 * MIPS Embedded:: MIPS Embedded
11780 * PA:: HP PA Embedded
11783 * Sparclet:: Tsqware Sparclet
11784 * Sparclite:: Fujitsu Sparclite
11785 * ST2000:: Tandem ST2000
11786 * Z8000:: Zilog Z8000
11789 @c OBSOLETE @node A29K Embedded
11790 @c OBSOLETE @subsection AMD A29K Embedded
11793 @c OBSOLETE * A29K UDI::
11794 @c OBSOLETE * A29K EB29K::
11795 @c OBSOLETE * Comms (EB29K):: Communications setup
11796 @c OBSOLETE * gdb-EB29K:: EB29K cross-debugging
11797 @c OBSOLETE * Remote Log:: Remote log
11798 @c OBSOLETE @end menu
11800 @c OBSOLETE @table @code
11802 @c OBSOLETE @kindex target adapt
11803 @c OBSOLETE @item target adapt @var{dev}
11804 @c OBSOLETE Adapt monitor for A29K.
11806 @c OBSOLETE @kindex target amd-eb
11807 @c OBSOLETE @item target amd-eb @var{dev} @var{speed} @var{PROG}
11808 @c OBSOLETE @cindex AMD EB29K
11809 @c OBSOLETE Remote PC-resident AMD EB29K board, attached over serial lines.
11810 @c OBSOLETE @var{dev} is the serial device, as for @code{target remote};
11811 @c OBSOLETE @var{speed} allows you to specify the linespeed; and @var{PROG} is the
11812 @c OBSOLETE name of the program to be debugged, as it appears to DOS on the PC.
11813 @c OBSOLETE @xref{A29K EB29K, ,EBMON protocol for AMD29K}.
11815 @c OBSOLETE @end table
11817 @c OBSOLETE @node A29K UDI
11818 @c OBSOLETE @subsubsection A29K UDI
11820 @c OBSOLETE @cindex UDI
11821 @c OBSOLETE @cindex AMD29K via UDI
11823 @c OBSOLETE @value{GDBN} supports AMD's UDI (``Universal Debugger Interface'')
11824 @c OBSOLETE protocol for debugging the a29k processor family. To use this
11825 @c OBSOLETE configuration with AMD targets running the MiniMON monitor, you need the
11826 @c OBSOLETE program @code{MONTIP}, available from AMD at no charge. You can also
11827 @c OBSOLETE use @value{GDBN} with the UDI-conformant a29k simulator program
11828 @c OBSOLETE @code{ISSTIP}, also available from AMD.
11830 @c OBSOLETE @table @code
11831 @c OBSOLETE @item target udi @var{keyword}
11832 @c OBSOLETE @kindex udi
11833 @c OBSOLETE Select the UDI interface to a remote a29k board or simulator, where
11834 @c OBSOLETE @var{keyword} is an entry in the AMD configuration file @file{udi_soc}.
11835 @c OBSOLETE This file contains keyword entries which specify parameters used to
11836 @c OBSOLETE connect to a29k targets. If the @file{udi_soc} file is not in your
11837 @c OBSOLETE working directory, you must set the environment variable @samp{UDICONF}
11838 @c OBSOLETE to its pathname.
11839 @c OBSOLETE @end table
11841 @c OBSOLETE @node A29K EB29K
11842 @c OBSOLETE @subsubsection EBMON protocol for AMD29K
11844 @c OBSOLETE @cindex EB29K board
11845 @c OBSOLETE @cindex running 29K programs
11847 @c OBSOLETE AMD distributes a 29K development board meant to fit in a PC, together
11848 @c OBSOLETE with a DOS-hosted monitor program called @code{EBMON}. As a shorthand
11849 @c OBSOLETE term, this development system is called the ``EB29K''. To use
11850 @c OBSOLETE @value{GDBN} from a Unix system to run programs on the EB29K board, you
11851 @c OBSOLETE must first connect a serial cable between the PC (which hosts the EB29K
11852 @c OBSOLETE board) and a serial port on the Unix system. In the following, we
11853 @c OBSOLETE assume you've hooked the cable between the PC's @file{COM1} port and
11854 @c OBSOLETE @file{/dev/ttya} on the Unix system.
11856 @c OBSOLETE @node Comms (EB29K)
11857 @c OBSOLETE @subsubsection Communications setup
11859 @c OBSOLETE The next step is to set up the PC's port, by doing something like this
11860 @c OBSOLETE in DOS on the PC:
11862 @c OBSOLETE @example
11863 @c OBSOLETE C:\> MODE com1:9600,n,8,1,none
11864 @c OBSOLETE @end example
11866 @c OBSOLETE @noindent
11867 @c OBSOLETE This example---run on an MS DOS 4.0 system---sets the PC port to 9600
11868 @c OBSOLETE bps, no parity, eight data bits, one stop bit, and no ``retry'' action;
11869 @c OBSOLETE you must match the communications parameters when establishing the Unix
11870 @c OBSOLETE end of the connection as well.
11871 @c OBSOLETE @c FIXME: Who knows what this "no retry action" crud from the DOS manual may
11872 @c OBSOLETE @c mean? It's optional; leave it out? ---doc@cygnus.com, 25feb91
11874 @c OBSOLETE @c It's optional, but it's unwise to omit it: who knows what is the
11875 @c OBSOLETE @c default value set when the DOS machines boots? "No retry" means that
11876 @c OBSOLETE @c the DOS serial device driver won't retry the operation if it fails;
11877 @c OBSOLETE @c I understand that this is needed because the GDB serial protocol
11878 @c OBSOLETE @c handles any errors and retransmissions itself. ---Eli Zaretskii, 3sep99
11880 @c OBSOLETE To give control of the PC to the Unix side of the serial line, type
11881 @c OBSOLETE the following at the DOS console:
11883 @c OBSOLETE @example
11884 @c OBSOLETE C:\> CTTY com1
11885 @c OBSOLETE @end example
11887 @c OBSOLETE @noindent
11888 @c OBSOLETE (Later, if you wish to return control to the DOS console, you can use
11889 @c OBSOLETE the command @code{CTTY con}---but you must send it over the device that
11890 @c OBSOLETE had control, in our example over the @file{COM1} serial line.)
11892 @c OBSOLETE From the Unix host, use a communications program such as @code{tip} or
11893 @c OBSOLETE @code{cu} to communicate with the PC; for example,
11895 @c OBSOLETE @example
11896 @c OBSOLETE cu -s 9600 -l /dev/ttya
11897 @c OBSOLETE @end example
11899 @c OBSOLETE @noindent
11900 @c OBSOLETE The @code{cu} options shown specify, respectively, the linespeed and the
11901 @c OBSOLETE serial port to use. If you use @code{tip} instead, your command line
11902 @c OBSOLETE may look something like the following:
11904 @c OBSOLETE @example
11905 @c OBSOLETE tip -9600 /dev/ttya
11906 @c OBSOLETE @end example
11908 @c OBSOLETE @noindent
11909 @c OBSOLETE Your system may require a different name where we show
11910 @c OBSOLETE @file{/dev/ttya} as the argument to @code{tip}. The communications
11911 @c OBSOLETE parameters, including which port to use, are associated with the
11912 @c OBSOLETE @code{tip} argument in the ``remote'' descriptions file---normally the
11913 @c OBSOLETE system table @file{/etc/remote}.
11914 @c OBSOLETE @c FIXME: What if anything needs doing to match the "n,8,1,none" part of
11915 @c OBSOLETE @c the DOS side's comms setup? cu can support -o (odd
11916 @c OBSOLETE @c parity), -e (even parity)---apparently no settings for no parity or
11917 @c OBSOLETE @c for character size. Taken from stty maybe...? John points out tip
11918 @c OBSOLETE @c can set these as internal variables, eg ~s parity=none; man stty
11919 @c OBSOLETE @c suggests that it *might* work to stty these options with stdin or
11920 @c OBSOLETE @c stdout redirected... ---doc@cygnus.com, 25feb91
11922 @c OBSOLETE @c There's nothing to be done for the "none" part of the DOS MODE
11923 @c OBSOLETE @c command. The rest of the parameters should be matched by the
11924 @c OBSOLETE @c baudrate, bits, and parity used by the Unix side. ---Eli Zaretskii, 3Sep99
11926 @c OBSOLETE @kindex EBMON
11927 @c OBSOLETE Using the @code{tip} or @code{cu} connection, change the DOS working
11928 @c OBSOLETE directory to the directory containing a copy of your 29K program, then
11929 @c OBSOLETE start the PC program @code{EBMON} (an EB29K control program supplied
11930 @c OBSOLETE with your board by AMD). You should see an initial display from
11931 @c OBSOLETE @code{EBMON} similar to the one that follows, ending with the
11932 @c OBSOLETE @code{EBMON} prompt @samp{#}---
11934 @c OBSOLETE @example
11935 @c OBSOLETE C:\> G:
11937 @c OBSOLETE G:\> CD \usr\joe\work29k
11939 @c OBSOLETE G:\USR\JOE\WORK29K> EBMON
11940 @c OBSOLETE Am29000 PC Coprocessor Board Monitor, version 3.0-18
11941 @c OBSOLETE Copyright 1990 Advanced Micro Devices, Inc.
11942 @c OBSOLETE Written by Gibbons and Associates, Inc.
11944 @c OBSOLETE Enter '?' or 'H' for help
11946 @c OBSOLETE PC Coprocessor Type = EB29K
11947 @c OBSOLETE I/O Base = 0x208
11948 @c OBSOLETE Memory Base = 0xd0000
11950 @c OBSOLETE Data Memory Size = 2048KB
11951 @c OBSOLETE Available I-RAM Range = 0x8000 to 0x1fffff
11952 @c OBSOLETE Available D-RAM Range = 0x80002000 to 0x801fffff
11954 @c OBSOLETE PageSize = 0x400
11955 @c OBSOLETE Register Stack Size = 0x800
11956 @c OBSOLETE Memory Stack Size = 0x1800
11958 @c OBSOLETE CPU PRL = 0x3
11959 @c OBSOLETE Am29027 Available = No
11960 @c OBSOLETE Byte Write Available = Yes
11963 @c OBSOLETE @end example
11965 @c OBSOLETE Then exit the @code{cu} or @code{tip} program (done in the example by
11966 @c OBSOLETE typing @code{~.} at the @code{EBMON} prompt). @code{EBMON} keeps
11967 @c OBSOLETE running, ready for @value{GDBN} to take over.
11969 @c OBSOLETE For this example, we've assumed what is probably the most convenient
11970 @c OBSOLETE way to make sure the same 29K program is on both the PC and the Unix
11971 @c OBSOLETE system: a PC/NFS connection that establishes ``drive @file{G:}'' on the
11972 @c OBSOLETE PC as a file system on the Unix host. If you do not have PC/NFS or
11973 @c OBSOLETE something similar connecting the two systems, you must arrange some
11974 @c OBSOLETE other way---perhaps floppy-disk transfer---of getting the 29K program
11975 @c OBSOLETE from the Unix system to the PC; @value{GDBN} does @emph{not} download it over the
11976 @c OBSOLETE serial line.
11978 @c OBSOLETE @node gdb-EB29K
11979 @c OBSOLETE @subsubsection EB29K cross-debugging
11981 @c OBSOLETE Finally, @code{cd} to the directory containing an image of your 29K
11982 @c OBSOLETE program on the Unix system, and start @value{GDBN}---specifying as argument the
11983 @c OBSOLETE name of your 29K program:
11985 @c OBSOLETE @example
11986 @c OBSOLETE cd /usr/joe/work29k
11987 @c OBSOLETE @value{GDBP} myfoo
11988 @c OBSOLETE @end example
11990 @c OBSOLETE @need 500
11991 @c OBSOLETE Now you can use the @code{target} command:
11993 @c OBSOLETE @example
11994 @c OBSOLETE target amd-eb /dev/ttya 9600 MYFOO
11995 @c OBSOLETE @c FIXME: test above 'target amd-eb' as spelled, with caps! caps are meant to
11996 @c OBSOLETE @c emphasize that this is the name as seen by DOS (since I think DOS is
11997 @c OBSOLETE @c single-minded about case of letters). ---doc@cygnus.com, 25feb91
11998 @c OBSOLETE @end example
12000 @c OBSOLETE @noindent
12001 @c OBSOLETE In this example, we've assumed your program is in a file called
12002 @c OBSOLETE @file{myfoo}. Note that the filename given as the last argument to
12003 @c OBSOLETE @code{target amd-eb} should be the name of the program as it appears to DOS.
12004 @c OBSOLETE In our example this is simply @code{MYFOO}, but in general it can include
12005 @c OBSOLETE a DOS path, and depending on your transfer mechanism may not resemble
12006 @c OBSOLETE the name on the Unix side.
12008 @c OBSOLETE At this point, you can set any breakpoints you wish; when you are ready
12009 @c OBSOLETE to see your program run on the 29K board, use the @value{GDBN} command
12010 @c OBSOLETE @code{run}.
12012 @c OBSOLETE To stop debugging the remote program, use the @value{GDBN} @code{detach}
12013 @c OBSOLETE command.
12015 @c OBSOLETE To return control of the PC to its console, use @code{tip} or @code{cu}
12016 @c OBSOLETE once again, after your @value{GDBN} session has concluded, to attach to
12017 @c OBSOLETE @code{EBMON}. You can then type the command @code{q} to shut down
12018 @c OBSOLETE @code{EBMON}, returning control to the DOS command-line interpreter.
12019 @c OBSOLETE Type @kbd{CTTY con} to return command input to the main DOS console,
12020 @c OBSOLETE and type @kbd{~.} to leave @code{tip} or @code{cu}.
12022 @c OBSOLETE @node Remote Log
12023 @c OBSOLETE @subsubsection Remote log
12024 @c OBSOLETE @cindex @file{eb.log}, a log file for EB29K
12025 @c OBSOLETE @cindex log file for EB29K
12027 @c OBSOLETE The @code{target amd-eb} command creates a file @file{eb.log} in the
12028 @c OBSOLETE current working directory, to help debug problems with the connection.
12029 @c OBSOLETE @file{eb.log} records all the output from @code{EBMON}, including echoes
12030 @c OBSOLETE of the commands sent to it. Running @samp{tail -f} on this file in
12031 @c OBSOLETE another window often helps to understand trouble with @code{EBMON}, or
12032 @c OBSOLETE unexpected events on the PC side of the connection.
12040 @item target rdi @var{dev}
12041 ARM Angel monitor, via RDI library interface to ADP protocol. You may
12042 use this target to communicate with both boards running the Angel
12043 monitor, or with the EmbeddedICE JTAG debug device.
12046 @item target rdp @var{dev}
12052 @subsection Hitachi H8/300
12056 @kindex target hms@r{, with H8/300}
12057 @item target hms @var{dev}
12058 A Hitachi SH, H8/300, or H8/500 board, attached via serial line to your host.
12059 Use special commands @code{device} and @code{speed} to control the serial
12060 line and the communications speed used.
12062 @kindex target e7000@r{, with H8/300}
12063 @item target e7000 @var{dev}
12064 E7000 emulator for Hitachi H8 and SH.
12066 @kindex target sh3@r{, with H8/300}
12067 @kindex target sh3e@r{, with H8/300}
12068 @item target sh3 @var{dev}
12069 @itemx target sh3e @var{dev}
12070 Hitachi SH-3 and SH-3E target systems.
12074 @cindex download to H8/300 or H8/500
12075 @cindex H8/300 or H8/500 download
12076 @cindex download to Hitachi SH
12077 @cindex Hitachi SH download
12078 When you select remote debugging to a Hitachi SH, H8/300, or H8/500
12079 board, the @code{load} command downloads your program to the Hitachi
12080 board and also opens it as the current executable target for
12081 @value{GDBN} on your host (like the @code{file} command).
12083 @value{GDBN} needs to know these things to talk to your
12084 Hitachi SH, H8/300, or H8/500:
12088 that you want to use @samp{target hms}, the remote debugging interface
12089 for Hitachi microprocessors, or @samp{target e7000}, the in-circuit
12090 emulator for the Hitachi SH and the Hitachi 300H. (@samp{target hms} is
12091 the default when @value{GDBN} is configured specifically for the Hitachi SH,
12092 H8/300, or H8/500.)
12095 what serial device connects your host to your Hitachi board (the first
12096 serial device available on your host is the default).
12099 what speed to use over the serial device.
12103 * Hitachi Boards:: Connecting to Hitachi boards.
12104 * Hitachi ICE:: Using the E7000 In-Circuit Emulator.
12105 * Hitachi Special:: Special @value{GDBN} commands for Hitachi micros.
12108 @node Hitachi Boards
12109 @subsubsection Connecting to Hitachi boards
12111 @c only for Unix hosts
12113 @cindex serial device, Hitachi micros
12114 Use the special @code{@value{GDBN}} command @samp{device @var{port}} if you
12115 need to explicitly set the serial device. The default @var{port} is the
12116 first available port on your host. This is only necessary on Unix
12117 hosts, where it is typically something like @file{/dev/ttya}.
12120 @cindex serial line speed, Hitachi micros
12121 @code{@value{GDBN}} has another special command to set the communications
12122 speed: @samp{speed @var{bps}}. This command also is only used from Unix
12123 hosts; on DOS hosts, set the line speed as usual from outside @value{GDBN} with
12124 the DOS @code{mode} command (for instance,
12125 @w{@kbd{mode com2:9600,n,8,1,p}} for a 9600@dmn{bps} connection).
12127 The @samp{device} and @samp{speed} commands are available only when you
12128 use a Unix host to debug your Hitachi microprocessor programs. If you
12130 @value{GDBN} depends on an auxiliary terminate-and-stay-resident program
12131 called @code{asynctsr} to communicate with the development board
12132 through a PC serial port. You must also use the DOS @code{mode} command
12133 to set up the serial port on the DOS side.
12135 The following sample session illustrates the steps needed to start a
12136 program under @value{GDBN} control on an H8/300. The example uses a
12137 sample H8/300 program called @file{t.x}. The procedure is the same for
12138 the Hitachi SH and the H8/500.
12140 First hook up your development board. In this example, we use a
12141 board attached to serial port @code{COM2}; if you use a different serial
12142 port, substitute its name in the argument of the @code{mode} command.
12143 When you call @code{asynctsr}, the auxiliary comms program used by the
12144 debugger, you give it just the numeric part of the serial port's name;
12145 for example, @samp{asyncstr 2} below runs @code{asyncstr} on
12149 C:\H8300\TEST> asynctsr 2
12150 C:\H8300\TEST> mode com2:9600,n,8,1,p
12152 Resident portion of MODE loaded
12154 COM2: 9600, n, 8, 1, p
12159 @emph{Warning:} We have noticed a bug in PC-NFS that conflicts with
12160 @code{asynctsr}. If you also run PC-NFS on your DOS host, you may need to
12161 disable it, or even boot without it, to use @code{asynctsr} to control
12162 your development board.
12165 @kindex target hms@r{, and serial protocol}
12166 Now that serial communications are set up, and the development board is
12167 connected, you can start up @value{GDBN}. Call @code{@value{GDBP}} with
12168 the name of your program as the argument. @code{@value{GDBN}} prompts
12169 you, as usual, with the prompt @samp{(@value{GDBP})}. Use two special
12170 commands to begin your debugging session: @samp{target hms} to specify
12171 cross-debugging to the Hitachi board, and the @code{load} command to
12172 download your program to the board. @code{load} displays the names of
12173 the program's sections, and a @samp{*} for each 2K of data downloaded.
12174 (If you want to refresh @value{GDBN} data on symbols or on the
12175 executable file without downloading, use the @value{GDBN} commands
12176 @code{file} or @code{symbol-file}. These commands, and @code{load}
12177 itself, are described in @ref{Files,,Commands to specify files}.)
12180 (eg-C:\H8300\TEST) @value{GDBP} t.x
12181 @value{GDBN} is free software and you are welcome to distribute copies
12182 of it under certain conditions; type "show copying" to see
12184 There is absolutely no warranty for @value{GDBN}; type "show warranty"
12186 @value{GDBN} @value{GDBVN}, Copyright 1992 Free Software Foundation, Inc...
12187 (@value{GDBP}) target hms
12188 Connected to remote H8/300 HMS system.
12189 (@value{GDBP}) load t.x
12190 .text : 0x8000 .. 0xabde ***********
12191 .data : 0xabde .. 0xad30 *
12192 .stack : 0xf000 .. 0xf014 *
12195 At this point, you're ready to run or debug your program. From here on,
12196 you can use all the usual @value{GDBN} commands. The @code{break} command
12197 sets breakpoints; the @code{run} command starts your program;
12198 @code{print} or @code{x} display data; the @code{continue} command
12199 resumes execution after stopping at a breakpoint. You can use the
12200 @code{help} command at any time to find out more about @value{GDBN} commands.
12202 Remember, however, that @emph{operating system} facilities aren't
12203 available on your development board; for example, if your program hangs,
12204 you can't send an interrupt---but you can press the @sc{reset} switch!
12206 Use the @sc{reset} button on the development board
12209 to interrupt your program (don't use @kbd{ctl-C} on the DOS host---it has
12210 no way to pass an interrupt signal to the development board); and
12213 to return to the @value{GDBN} command prompt after your program finishes
12214 normally. The communications protocol provides no other way for @value{GDBN}
12215 to detect program completion.
12218 In either case, @value{GDBN} sees the effect of a @sc{reset} on the
12219 development board as a ``normal exit'' of your program.
12222 @subsubsection Using the E7000 in-circuit emulator
12224 @kindex target e7000@r{, with Hitachi ICE}
12225 You can use the E7000 in-circuit emulator to develop code for either the
12226 Hitachi SH or the H8/300H. Use one of these forms of the @samp{target
12227 e7000} command to connect @value{GDBN} to your E7000:
12230 @item target e7000 @var{port} @var{speed}
12231 Use this form if your E7000 is connected to a serial port. The
12232 @var{port} argument identifies what serial port to use (for example,
12233 @samp{com2}). The third argument is the line speed in bits per second
12234 (for example, @samp{9600}).
12236 @item target e7000 @var{hostname}
12237 If your E7000 is installed as a host on a TCP/IP network, you can just
12238 specify its hostname; @value{GDBN} uses @code{telnet} to connect.
12241 @node Hitachi Special
12242 @subsubsection Special @value{GDBN} commands for Hitachi micros
12244 Some @value{GDBN} commands are available only for the H8/300:
12248 @kindex set machine
12249 @kindex show machine
12250 @item set machine h8300
12251 @itemx set machine h8300h
12252 Condition @value{GDBN} for one of the two variants of the H8/300
12253 architecture with @samp{set machine}. You can use @samp{show machine}
12254 to check which variant is currently in effect.
12263 @kindex set memory @var{mod}
12264 @cindex memory models, H8/500
12265 @item set memory @var{mod}
12267 Specify which H8/500 memory model (@var{mod}) you are using with
12268 @samp{set memory}; check which memory model is in effect with @samp{show
12269 memory}. The accepted values for @var{mod} are @code{small},
12270 @code{big}, @code{medium}, and @code{compact}.
12275 @subsection Intel i960
12279 @kindex target mon960
12280 @item target mon960 @var{dev}
12281 MON960 monitor for Intel i960.
12283 @kindex target nindy
12284 @item target nindy @var{devicename}
12285 An Intel 960 board controlled by a Nindy Monitor. @var{devicename} is
12286 the name of the serial device to use for the connection, e.g.
12293 @dfn{Nindy} is a ROM Monitor program for Intel 960 target systems. When
12294 @value{GDBN} is configured to control a remote Intel 960 using Nindy, you can
12295 tell @value{GDBN} how to connect to the 960 in several ways:
12299 Through command line options specifying serial port, version of the
12300 Nindy protocol, and communications speed;
12303 By responding to a prompt on startup;
12306 By using the @code{target} command at any point during your @value{GDBN}
12307 session. @xref{Target Commands, ,Commands for managing targets}.
12311 @cindex download to Nindy-960
12312 With the Nindy interface to an Intel 960 board, @code{load}
12313 downloads @var{filename} to the 960 as well as adding its symbols in
12317 * Nindy Startup:: Startup with Nindy
12318 * Nindy Options:: Options for Nindy
12319 * Nindy Reset:: Nindy reset command
12322 @node Nindy Startup
12323 @subsubsection Startup with Nindy
12325 If you simply start @code{@value{GDBP}} without using any command-line
12326 options, you are prompted for what serial port to use, @emph{before} you
12327 reach the ordinary @value{GDBN} prompt:
12330 Attach /dev/ttyNN -- specify NN, or "quit" to quit:
12334 Respond to the prompt with whatever suffix (after @samp{/dev/tty})
12335 identifies the serial port you want to use. You can, if you choose,
12336 simply start up with no Nindy connection by responding to the prompt
12337 with an empty line. If you do this and later wish to attach to Nindy,
12338 use @code{target} (@pxref{Target Commands, ,Commands for managing targets}).
12340 @node Nindy Options
12341 @subsubsection Options for Nindy
12343 These are the startup options for beginning your @value{GDBN} session with a
12344 Nindy-960 board attached:
12347 @item -r @var{port}
12348 Specify the serial port name of a serial interface to be used to connect
12349 to the target system. This option is only available when @value{GDBN} is
12350 configured for the Intel 960 target architecture. You may specify
12351 @var{port} as any of: a full pathname (e.g. @samp{-r /dev/ttya}), a
12352 device name in @file{/dev} (e.g. @samp{-r ttya}), or simply the unique
12353 suffix for a specific @code{tty} (e.g. @samp{-r a}).
12356 (An uppercase letter ``O'', not a zero.) Specify that @value{GDBN} should use
12357 the ``old'' Nindy monitor protocol to connect to the target system.
12358 This option is only available when @value{GDBN} is configured for the Intel 960
12359 target architecture.
12362 @emph{Warning:} if you specify @samp{-O}, but are actually trying to
12363 connect to a target system that expects the newer protocol, the connection
12364 fails, appearing to be a speed mismatch. @value{GDBN} repeatedly
12365 attempts to reconnect at several different line speeds. You can abort
12366 this process with an interrupt.
12370 Specify that @value{GDBN} should first send a @code{BREAK} signal to the target
12371 system, in an attempt to reset it, before connecting to a Nindy target.
12374 @emph{Warning:} Many target systems do not have the hardware that this
12375 requires; it only works with a few boards.
12379 The standard @samp{-b} option controls the line speed used on the serial
12384 @subsubsection Nindy reset command
12389 For a Nindy target, this command sends a ``break'' to the remote target
12390 system; this is only useful if the target has been equipped with a
12391 circuit to perform a hard reset (or some other interesting action) when
12392 a break is detected.
12397 @subsection Mitsubishi M32R/D
12401 @kindex target m32r
12402 @item target m32r @var{dev}
12403 Mitsubishi M32R/D ROM monitor.
12410 The Motorola m68k configuration includes ColdFire support, and
12411 target command for the following ROM monitors.
12415 @kindex target abug
12416 @item target abug @var{dev}
12417 ABug ROM monitor for M68K.
12419 @kindex target cpu32bug
12420 @item target cpu32bug @var{dev}
12421 CPU32BUG monitor, running on a CPU32 (M68K) board.
12423 @kindex target dbug
12424 @item target dbug @var{dev}
12425 dBUG ROM monitor for Motorola ColdFire.
12428 @item target est @var{dev}
12429 EST-300 ICE monitor, running on a CPU32 (M68K) board.
12431 @kindex target rom68k
12432 @item target rom68k @var{dev}
12433 ROM 68K monitor, running on an M68K IDP board.
12437 If @value{GDBN} is configured with @code{m68*-ericsson-*}, it will
12438 instead have only a single special target command:
12442 @kindex target es1800
12443 @item target es1800 @var{dev}
12444 ES-1800 emulator for M68K.
12452 @kindex target rombug
12453 @item target rombug @var{dev}
12454 ROMBUG ROM monitor for OS/9000.
12464 @item target bug @var{dev}
12465 BUG monitor, running on a MVME187 (m88k) board.
12469 @node MIPS Embedded
12470 @subsection MIPS Embedded
12472 @cindex MIPS boards
12473 @value{GDBN} can use the MIPS remote debugging protocol to talk to a
12474 MIPS board attached to a serial line. This is available when
12475 you configure @value{GDBN} with @samp{--target=mips-idt-ecoff}.
12478 Use these @value{GDBN} commands to specify the connection to your target board:
12481 @item target mips @var{port}
12482 @kindex target mips @var{port}
12483 To run a program on the board, start up @code{@value{GDBP}} with the
12484 name of your program as the argument. To connect to the board, use the
12485 command @samp{target mips @var{port}}, where @var{port} is the name of
12486 the serial port connected to the board. If the program has not already
12487 been downloaded to the board, you may use the @code{load} command to
12488 download it. You can then use all the usual @value{GDBN} commands.
12490 For example, this sequence connects to the target board through a serial
12491 port, and loads and runs a program called @var{prog} through the
12495 host$ @value{GDBP} @var{prog}
12496 @value{GDBN} is free software and @dots{}
12497 (@value{GDBP}) target mips /dev/ttyb
12498 (@value{GDBP}) load @var{prog}
12502 @item target mips @var{hostname}:@var{portnumber}
12503 On some @value{GDBN} host configurations, you can specify a TCP
12504 connection (for instance, to a serial line managed by a terminal
12505 concentrator) instead of a serial port, using the syntax
12506 @samp{@var{hostname}:@var{portnumber}}.
12508 @item target pmon @var{port}
12509 @kindex target pmon @var{port}
12512 @item target ddb @var{port}
12513 @kindex target ddb @var{port}
12514 NEC's DDB variant of PMON for Vr4300.
12516 @item target lsi @var{port}
12517 @kindex target lsi @var{port}
12518 LSI variant of PMON.
12520 @kindex target r3900
12521 @item target r3900 @var{dev}
12522 Densan DVE-R3900 ROM monitor for Toshiba R3900 Mips.
12524 @kindex target array
12525 @item target array @var{dev}
12526 Array Tech LSI33K RAID controller board.
12532 @value{GDBN} also supports these special commands for MIPS targets:
12535 @item set processor @var{args}
12536 @itemx show processor
12537 @kindex set processor @var{args}
12538 @kindex show processor
12539 Use the @code{set processor} command to set the type of MIPS
12540 processor when you want to access processor-type-specific registers.
12541 For example, @code{set processor @var{r3041}} tells @value{GDBN}
12542 to use the CPU registers appropriate for the 3041 chip.
12543 Use the @code{show processor} command to see what MIPS processor @value{GDBN}
12544 is using. Use the @code{info reg} command to see what registers
12545 @value{GDBN} is using.
12547 @item set mipsfpu double
12548 @itemx set mipsfpu single
12549 @itemx set mipsfpu none
12550 @itemx show mipsfpu
12551 @kindex set mipsfpu
12552 @kindex show mipsfpu
12553 @cindex MIPS remote floating point
12554 @cindex floating point, MIPS remote
12555 If your target board does not support the MIPS floating point
12556 coprocessor, you should use the command @samp{set mipsfpu none} (if you
12557 need this, you may wish to put the command in your @value{GDBN} init
12558 file). This tells @value{GDBN} how to find the return value of
12559 functions which return floating point values. It also allows
12560 @value{GDBN} to avoid saving the floating point registers when calling
12561 functions on the board. If you are using a floating point coprocessor
12562 with only single precision floating point support, as on the @sc{r4650}
12563 processor, use the command @samp{set mipsfpu single}. The default
12564 double precision floating point coprocessor may be selected using
12565 @samp{set mipsfpu double}.
12567 In previous versions the only choices were double precision or no
12568 floating point, so @samp{set mipsfpu on} will select double precision
12569 and @samp{set mipsfpu off} will select no floating point.
12571 As usual, you can inquire about the @code{mipsfpu} variable with
12572 @samp{show mipsfpu}.
12574 @item set remotedebug @var{n}
12575 @itemx show remotedebug
12576 @kindex set remotedebug@r{, MIPS protocol}
12577 @kindex show remotedebug@r{, MIPS protocol}
12578 @cindex @code{remotedebug}, MIPS protocol
12579 @cindex MIPS @code{remotedebug} protocol
12580 @c FIXME! For this to be useful, you must know something about the MIPS
12581 @c FIXME...protocol. Where is it described?
12582 You can see some debugging information about communications with the board
12583 by setting the @code{remotedebug} variable. If you set it to @code{1} using
12584 @samp{set remotedebug 1}, every packet is displayed. If you set it
12585 to @code{2}, every character is displayed. You can check the current value
12586 at any time with the command @samp{show remotedebug}.
12588 @item set timeout @var{seconds}
12589 @itemx set retransmit-timeout @var{seconds}
12590 @itemx show timeout
12591 @itemx show retransmit-timeout
12592 @cindex @code{timeout}, MIPS protocol
12593 @cindex @code{retransmit-timeout}, MIPS protocol
12594 @kindex set timeout
12595 @kindex show timeout
12596 @kindex set retransmit-timeout
12597 @kindex show retransmit-timeout
12598 You can control the timeout used while waiting for a packet, in the MIPS
12599 remote protocol, with the @code{set timeout @var{seconds}} command. The
12600 default is 5 seconds. Similarly, you can control the timeout used while
12601 waiting for an acknowledgement of a packet with the @code{set
12602 retransmit-timeout @var{seconds}} command. The default is 3 seconds.
12603 You can inspect both values with @code{show timeout} and @code{show
12604 retransmit-timeout}. (These commands are @emph{only} available when
12605 @value{GDBN} is configured for @samp{--target=mips-idt-ecoff}.)
12607 The timeout set by @code{set timeout} does not apply when @value{GDBN}
12608 is waiting for your program to stop. In that case, @value{GDBN} waits
12609 forever because it has no way of knowing how long the program is going
12610 to run before stopping.
12614 @subsection PowerPC
12618 @kindex target dink32
12619 @item target dink32 @var{dev}
12620 DINK32 ROM monitor.
12622 @kindex target ppcbug
12623 @item target ppcbug @var{dev}
12624 @kindex target ppcbug1
12625 @item target ppcbug1 @var{dev}
12626 PPCBUG ROM monitor for PowerPC.
12629 @item target sds @var{dev}
12630 SDS monitor, running on a PowerPC board (such as Motorola's ADS).
12635 @subsection HP PA Embedded
12639 @kindex target op50n
12640 @item target op50n @var{dev}
12641 OP50N monitor, running on an OKI HPPA board.
12643 @kindex target w89k
12644 @item target w89k @var{dev}
12645 W89K monitor, running on a Winbond HPPA board.
12650 @subsection Hitachi SH
12654 @kindex target hms@r{, with Hitachi SH}
12655 @item target hms @var{dev}
12656 A Hitachi SH board attached via serial line to your host. Use special
12657 commands @code{device} and @code{speed} to control the serial line and
12658 the communications speed used.
12660 @kindex target e7000@r{, with Hitachi SH}
12661 @item target e7000 @var{dev}
12662 E7000 emulator for Hitachi SH.
12664 @kindex target sh3@r{, with SH}
12665 @kindex target sh3e@r{, with SH}
12666 @item target sh3 @var{dev}
12667 @item target sh3e @var{dev}
12668 Hitachi SH-3 and SH-3E target systems.
12673 @subsection Tsqware Sparclet
12677 @value{GDBN} enables developers to debug tasks running on
12678 Sparclet targets from a Unix host.
12679 @value{GDBN} uses code that runs on
12680 both the Unix host and on the Sparclet target. The program
12681 @code{@value{GDBP}} is installed and executed on the Unix host.
12684 @item remotetimeout @var{args}
12685 @kindex remotetimeout
12686 @value{GDBN} supports the option @code{remotetimeout}.
12687 This option is set by the user, and @var{args} represents the number of
12688 seconds @value{GDBN} waits for responses.
12691 @cindex compiling, on Sparclet
12692 When compiling for debugging, include the options @samp{-g} to get debug
12693 information and @samp{-Ttext} to relocate the program to where you wish to
12694 load it on the target. You may also want to add the options @samp{-n} or
12695 @samp{-N} in order to reduce the size of the sections. Example:
12698 sparclet-aout-gcc prog.c -Ttext 0x12010000 -g -o prog -N
12701 You can use @code{objdump} to verify that the addresses are what you intended:
12704 sparclet-aout-objdump --headers --syms prog
12707 @cindex running, on Sparclet
12709 your Unix execution search path to find @value{GDBN}, you are ready to
12710 run @value{GDBN}. From your Unix host, run @code{@value{GDBP}}
12711 (or @code{sparclet-aout-gdb}, depending on your installation).
12713 @value{GDBN} comes up showing the prompt:
12720 * Sparclet File:: Setting the file to debug
12721 * Sparclet Connection:: Connecting to Sparclet
12722 * Sparclet Download:: Sparclet download
12723 * Sparclet Execution:: Running and debugging
12726 @node Sparclet File
12727 @subsubsection Setting file to debug
12729 The @value{GDBN} command @code{file} lets you choose with program to debug.
12732 (gdbslet) file prog
12736 @value{GDBN} then attempts to read the symbol table of @file{prog}.
12737 @value{GDBN} locates
12738 the file by searching the directories listed in the command search
12740 If the file was compiled with debug information (option "-g"), source
12741 files will be searched as well.
12742 @value{GDBN} locates
12743 the source files by searching the directories listed in the directory search
12744 path (@pxref{Environment, ,Your program's environment}).
12746 to find a file, it displays a message such as:
12749 prog: No such file or directory.
12752 When this happens, add the appropriate directories to the search paths with
12753 the @value{GDBN} commands @code{path} and @code{dir}, and execute the
12754 @code{target} command again.
12756 @node Sparclet Connection
12757 @subsubsection Connecting to Sparclet
12759 The @value{GDBN} command @code{target} lets you connect to a Sparclet target.
12760 To connect to a target on serial port ``@code{ttya}'', type:
12763 (gdbslet) target sparclet /dev/ttya
12764 Remote target sparclet connected to /dev/ttya
12765 main () at ../prog.c:3
12769 @value{GDBN} displays messages like these:
12775 @node Sparclet Download
12776 @subsubsection Sparclet download
12778 @cindex download to Sparclet
12779 Once connected to the Sparclet target,
12780 you can use the @value{GDBN}
12781 @code{load} command to download the file from the host to the target.
12782 The file name and load offset should be given as arguments to the @code{load}
12784 Since the file format is aout, the program must be loaded to the starting
12785 address. You can use @code{objdump} to find out what this value is. The load
12786 offset is an offset which is added to the VMA (virtual memory address)
12787 of each of the file's sections.
12788 For instance, if the program
12789 @file{prog} was linked to text address 0x1201000, with data at 0x12010160
12790 and bss at 0x12010170, in @value{GDBN}, type:
12793 (gdbslet) load prog 0x12010000
12794 Loading section .text, size 0xdb0 vma 0x12010000
12797 If the code is loaded at a different address then what the program was linked
12798 to, you may need to use the @code{section} and @code{add-symbol-file} commands
12799 to tell @value{GDBN} where to map the symbol table.
12801 @node Sparclet Execution
12802 @subsubsection Running and debugging
12804 @cindex running and debugging Sparclet programs
12805 You can now begin debugging the task using @value{GDBN}'s execution control
12806 commands, @code{b}, @code{step}, @code{run}, etc. See the @value{GDBN}
12807 manual for the list of commands.
12811 Breakpoint 1 at 0x12010000: file prog.c, line 3.
12813 Starting program: prog
12814 Breakpoint 1, main (argc=1, argv=0xeffff21c) at prog.c:3
12815 3 char *symarg = 0;
12817 4 char *execarg = "hello!";
12822 @subsection Fujitsu Sparclite
12826 @kindex target sparclite
12827 @item target sparclite @var{dev}
12828 Fujitsu sparclite boards, used only for the purpose of loading.
12829 You must use an additional command to debug the program.
12830 For example: target remote @var{dev} using @value{GDBN} standard
12836 @subsection Tandem ST2000
12838 @value{GDBN} may be used with a Tandem ST2000 phone switch, running Tandem's
12841 To connect your ST2000 to the host system, see the manufacturer's
12842 manual. Once the ST2000 is physically attached, you can run:
12845 target st2000 @var{dev} @var{speed}
12849 to establish it as your debugging environment. @var{dev} is normally
12850 the name of a serial device, such as @file{/dev/ttya}, connected to the
12851 ST2000 via a serial line. You can instead specify @var{dev} as a TCP
12852 connection (for example, to a serial line attached via a terminal
12853 concentrator) using the syntax @code{@var{hostname}:@var{portnumber}}.
12855 The @code{load} and @code{attach} commands are @emph{not} defined for
12856 this target; you must load your program into the ST2000 as you normally
12857 would for standalone operation. @value{GDBN} reads debugging information
12858 (such as symbols) from a separate, debugging version of the program
12859 available on your host computer.
12860 @c FIXME!! This is terribly vague; what little content is here is
12861 @c basically hearsay.
12863 @cindex ST2000 auxiliary commands
12864 These auxiliary @value{GDBN} commands are available to help you with the ST2000
12868 @item st2000 @var{command}
12869 @kindex st2000 @var{cmd}
12870 @cindex STDBUG commands (ST2000)
12871 @cindex commands to STDBUG (ST2000)
12872 Send a @var{command} to the STDBUG monitor. See the manufacturer's
12873 manual for available commands.
12876 @cindex connect (to STDBUG)
12877 Connect the controlling terminal to the STDBUG command monitor. When
12878 you are done interacting with STDBUG, typing either of two character
12879 sequences gets you back to the @value{GDBN} command prompt:
12880 @kbd{@key{RET}~.} (Return, followed by tilde and period) or
12881 @kbd{@key{RET}~@key{C-d}} (Return, followed by tilde and control-D).
12885 @subsection Zilog Z8000
12888 @cindex simulator, Z8000
12889 @cindex Zilog Z8000 simulator
12891 When configured for debugging Zilog Z8000 targets, @value{GDBN} includes
12894 For the Z8000 family, @samp{target sim} simulates either the Z8002 (the
12895 unsegmented variant of the Z8000 architecture) or the Z8001 (the
12896 segmented variant). The simulator recognizes which architecture is
12897 appropriate by inspecting the object code.
12900 @item target sim @var{args}
12902 @kindex target sim@r{, with Z8000}
12903 Debug programs on a simulated CPU. If the simulator supports setup
12904 options, specify them via @var{args}.
12908 After specifying this target, you can debug programs for the simulated
12909 CPU in the same style as programs for your host computer; use the
12910 @code{file} command to load a new program image, the @code{run} command
12911 to run your program, and so on.
12913 As well as making available all the usual machine registers
12914 (@pxref{Registers, ,Registers}), the Z8000 simulator provides three
12915 additional items of information as specially named registers:
12920 Counts clock-ticks in the simulator.
12923 Counts instructions run in the simulator.
12926 Execution time in 60ths of a second.
12930 You can refer to these values in @value{GDBN} expressions with the usual
12931 conventions; for example, @w{@samp{b fputc if $cycles>5000}} sets a
12932 conditional breakpoint that suspends only after at least 5000
12933 simulated clock ticks.
12935 @node Architectures
12936 @section Architectures
12938 This section describes characteristics of architectures that affect
12939 all uses of @value{GDBN} with the architecture, both native and cross.
12952 @kindex set rstack_high_address
12953 @cindex AMD 29K register stack
12954 @cindex register stack, AMD29K
12955 @item set rstack_high_address @var{address}
12956 On AMD 29000 family processors, registers are saved in a separate
12957 @dfn{register stack}. There is no way for @value{GDBN} to determine the
12958 extent of this stack. Normally, @value{GDBN} just assumes that the
12959 stack is ``large enough''. This may result in @value{GDBN} referencing
12960 memory locations that do not exist. If necessary, you can get around
12961 this problem by specifying the ending address of the register stack with
12962 the @code{set rstack_high_address} command. The argument should be an
12963 address, which you probably want to precede with @samp{0x} to specify in
12966 @kindex show rstack_high_address
12967 @item show rstack_high_address
12968 Display the current limit of the register stack, on AMD 29000 family
12976 See the following section.
12981 @cindex stack on Alpha
12982 @cindex stack on MIPS
12983 @cindex Alpha stack
12985 Alpha- and MIPS-based computers use an unusual stack frame, which
12986 sometimes requires @value{GDBN} to search backward in the object code to
12987 find the beginning of a function.
12989 @cindex response time, MIPS debugging
12990 To improve response time (especially for embedded applications, where
12991 @value{GDBN} may be restricted to a slow serial line for this search)
12992 you may want to limit the size of this search, using one of these
12996 @cindex @code{heuristic-fence-post} (Alpha, MIPS)
12997 @item set heuristic-fence-post @var{limit}
12998 Restrict @value{GDBN} to examining at most @var{limit} bytes in its
12999 search for the beginning of a function. A value of @var{0} (the
13000 default) means there is no limit. However, except for @var{0}, the
13001 larger the limit the more bytes @code{heuristic-fence-post} must search
13002 and therefore the longer it takes to run.
13004 @item show heuristic-fence-post
13005 Display the current limit.
13009 These commands are available @emph{only} when @value{GDBN} is configured
13010 for debugging programs on Alpha or MIPS processors.
13013 @node Controlling GDB
13014 @chapter Controlling @value{GDBN}
13016 You can alter the way @value{GDBN} interacts with you by using the
13017 @code{set} command. For commands controlling how @value{GDBN} displays
13018 data, see @ref{Print Settings, ,Print settings}. Other settings are
13023 * Editing:: Command editing
13024 * History:: Command history
13025 * Screen Size:: Screen size
13026 * Numbers:: Numbers
13027 * Messages/Warnings:: Optional warnings and messages
13028 * Debugging Output:: Optional messages about internal happenings
13036 @value{GDBN} indicates its readiness to read a command by printing a string
13037 called the @dfn{prompt}. This string is normally @samp{(@value{GDBP})}. You
13038 can change the prompt string with the @code{set prompt} command. For
13039 instance, when debugging @value{GDBN} with @value{GDBN}, it is useful to change
13040 the prompt in one of the @value{GDBN} sessions so that you can always tell
13041 which one you are talking to.
13043 @emph{Note:} @code{set prompt} does not add a space for you after the
13044 prompt you set. This allows you to set a prompt which ends in a space
13045 or a prompt that does not.
13049 @item set prompt @var{newprompt}
13050 Directs @value{GDBN} to use @var{newprompt} as its prompt string henceforth.
13052 @kindex show prompt
13054 Prints a line of the form: @samp{Gdb's prompt is: @var{your-prompt}}
13058 @section Command editing
13060 @cindex command line editing
13062 @value{GDBN} reads its input commands via the @dfn{readline} interface. This
13063 @sc{gnu} library provides consistent behavior for programs which provide a
13064 command line interface to the user. Advantages are @sc{gnu} Emacs-style
13065 or @dfn{vi}-style inline editing of commands, @code{csh}-like history
13066 substitution, and a storage and recall of command history across
13067 debugging sessions.
13069 You may control the behavior of command line editing in @value{GDBN} with the
13070 command @code{set}.
13073 @kindex set editing
13076 @itemx set editing on
13077 Enable command line editing (enabled by default).
13079 @item set editing off
13080 Disable command line editing.
13082 @kindex show editing
13084 Show whether command line editing is enabled.
13088 @section Command history
13090 @value{GDBN} can keep track of the commands you type during your
13091 debugging sessions, so that you can be certain of precisely what
13092 happened. Use these commands to manage the @value{GDBN} command
13096 @cindex history substitution
13097 @cindex history file
13098 @kindex set history filename
13099 @kindex GDBHISTFILE
13100 @item set history filename @var{fname}
13101 Set the name of the @value{GDBN} command history file to @var{fname}.
13102 This is the file where @value{GDBN} reads an initial command history
13103 list, and where it writes the command history from this session when it
13104 exits. You can access this list through history expansion or through
13105 the history command editing characters listed below. This file defaults
13106 to the value of the environment variable @code{GDBHISTFILE}, or to
13107 @file{./.gdb_history} (@file{./_gdb_history} on MS-DOS) if this variable
13110 @cindex history save
13111 @kindex set history save
13112 @item set history save
13113 @itemx set history save on
13114 Record command history in a file, whose name may be specified with the
13115 @code{set history filename} command. By default, this option is disabled.
13117 @item set history save off
13118 Stop recording command history in a file.
13120 @cindex history size
13121 @kindex set history size
13122 @item set history size @var{size}
13123 Set the number of commands which @value{GDBN} keeps in its history list.
13124 This defaults to the value of the environment variable
13125 @code{HISTSIZE}, or to 256 if this variable is not set.
13128 @cindex history expansion
13129 History expansion assigns special meaning to the character @kbd{!}.
13130 @ifset have-readline-appendices
13131 @xref{Event Designators}.
13134 Since @kbd{!} is also the logical not operator in C, history expansion
13135 is off by default. If you decide to enable history expansion with the
13136 @code{set history expansion on} command, you may sometimes need to
13137 follow @kbd{!} (when it is used as logical not, in an expression) with
13138 a space or a tab to prevent it from being expanded. The readline
13139 history facilities do not attempt substitution on the strings
13140 @kbd{!=} and @kbd{!(}, even when history expansion is enabled.
13142 The commands to control history expansion are:
13145 @kindex set history expansion
13146 @item set history expansion on
13147 @itemx set history expansion
13148 Enable history expansion. History expansion is off by default.
13150 @item set history expansion off
13151 Disable history expansion.
13153 The readline code comes with more complete documentation of
13154 editing and history expansion features. Users unfamiliar with @sc{gnu} Emacs
13155 or @code{vi} may wish to read it.
13156 @ifset have-readline-appendices
13157 @xref{Command Line Editing}.
13161 @kindex show history
13163 @itemx show history filename
13164 @itemx show history save
13165 @itemx show history size
13166 @itemx show history expansion
13167 These commands display the state of the @value{GDBN} history parameters.
13168 @code{show history} by itself displays all four states.
13174 @item show commands
13175 Display the last ten commands in the command history.
13177 @item show commands @var{n}
13178 Print ten commands centered on command number @var{n}.
13180 @item show commands +
13181 Print ten commands just after the commands last printed.
13185 @section Screen size
13186 @cindex size of screen
13187 @cindex pauses in output
13189 Certain commands to @value{GDBN} may produce large amounts of
13190 information output to the screen. To help you read all of it,
13191 @value{GDBN} pauses and asks you for input at the end of each page of
13192 output. Type @key{RET} when you want to continue the output, or @kbd{q}
13193 to discard the remaining output. Also, the screen width setting
13194 determines when to wrap lines of output. Depending on what is being
13195 printed, @value{GDBN} tries to break the line at a readable place,
13196 rather than simply letting it overflow onto the following line.
13198 Normally @value{GDBN} knows the size of the screen from the terminal
13199 driver software. For example, on Unix @value{GDBN} uses the termcap data base
13200 together with the value of the @code{TERM} environment variable and the
13201 @code{stty rows} and @code{stty cols} settings. If this is not correct,
13202 you can override it with the @code{set height} and @code{set
13209 @kindex show height
13210 @item set height @var{lpp}
13212 @itemx set width @var{cpl}
13214 These @code{set} commands specify a screen height of @var{lpp} lines and
13215 a screen width of @var{cpl} characters. The associated @code{show}
13216 commands display the current settings.
13218 If you specify a height of zero lines, @value{GDBN} does not pause during
13219 output no matter how long the output is. This is useful if output is to a
13220 file or to an editor buffer.
13222 Likewise, you can specify @samp{set width 0} to prevent @value{GDBN}
13223 from wrapping its output.
13228 @cindex number representation
13229 @cindex entering numbers
13231 You can always enter numbers in octal, decimal, or hexadecimal in
13232 @value{GDBN} by the usual conventions: octal numbers begin with
13233 @samp{0}, decimal numbers end with @samp{.}, and hexadecimal numbers
13234 begin with @samp{0x}. Numbers that begin with none of these are, by
13235 default, entered in base 10; likewise, the default display for
13236 numbers---when no particular format is specified---is base 10. You can
13237 change the default base for both input and output with the @code{set
13241 @kindex set input-radix
13242 @item set input-radix @var{base}
13243 Set the default base for numeric input. Supported choices
13244 for @var{base} are decimal 8, 10, or 16. @var{base} must itself be
13245 specified either unambiguously or using the current default radix; for
13255 sets the base to decimal. On the other hand, @samp{set radix 10}
13256 leaves the radix unchanged no matter what it was.
13258 @kindex set output-radix
13259 @item set output-radix @var{base}
13260 Set the default base for numeric display. Supported choices
13261 for @var{base} are decimal 8, 10, or 16. @var{base} must itself be
13262 specified either unambiguously or using the current default radix.
13264 @kindex show input-radix
13265 @item show input-radix
13266 Display the current default base for numeric input.
13268 @kindex show output-radix
13269 @item show output-radix
13270 Display the current default base for numeric display.
13273 @node Messages/Warnings
13274 @section Optional warnings and messages
13276 By default, @value{GDBN} is silent about its inner workings. If you are
13277 running on a slow machine, you may want to use the @code{set verbose}
13278 command. This makes @value{GDBN} tell you when it does a lengthy
13279 internal operation, so you will not think it has crashed.
13281 Currently, the messages controlled by @code{set verbose} are those
13282 which announce that the symbol table for a source file is being read;
13283 see @code{symbol-file} in @ref{Files, ,Commands to specify files}.
13286 @kindex set verbose
13287 @item set verbose on
13288 Enables @value{GDBN} output of certain informational messages.
13290 @item set verbose off
13291 Disables @value{GDBN} output of certain informational messages.
13293 @kindex show verbose
13295 Displays whether @code{set verbose} is on or off.
13298 By default, if @value{GDBN} encounters bugs in the symbol table of an
13299 object file, it is silent; but if you are debugging a compiler, you may
13300 find this information useful (@pxref{Symbol Errors, ,Errors reading
13305 @kindex set complaints
13306 @item set complaints @var{limit}
13307 Permits @value{GDBN} to output @var{limit} complaints about each type of
13308 unusual symbols before becoming silent about the problem. Set
13309 @var{limit} to zero to suppress all complaints; set it to a large number
13310 to prevent complaints from being suppressed.
13312 @kindex show complaints
13313 @item show complaints
13314 Displays how many symbol complaints @value{GDBN} is permitted to produce.
13318 By default, @value{GDBN} is cautious, and asks what sometimes seems to be a
13319 lot of stupid questions to confirm certain commands. For example, if
13320 you try to run a program which is already running:
13324 The program being debugged has been started already.
13325 Start it from the beginning? (y or n)
13328 If you are willing to unflinchingly face the consequences of your own
13329 commands, you can disable this ``feature'':
13333 @kindex set confirm
13335 @cindex confirmation
13336 @cindex stupid questions
13337 @item set confirm off
13338 Disables confirmation requests.
13340 @item set confirm on
13341 Enables confirmation requests (the default).
13343 @kindex show confirm
13345 Displays state of confirmation requests.
13349 @node Debugging Output
13350 @section Optional messages about internal happenings
13352 @kindex set debug arch
13353 @item set debug arch
13354 Turns on or off display of gdbarch debugging info. The default is off
13355 @kindex show debug arch
13356 @item show debug arch
13357 Displays the current state of displaying gdbarch debugging info.
13358 @kindex set debug event
13359 @item set debug event
13360 Turns on or off display of @value{GDBN} event debugging info. The
13362 @kindex show debug event
13363 @item show debug event
13364 Displays the current state of displaying @value{GDBN} event debugging
13366 @kindex set debug expression
13367 @item set debug expression
13368 Turns on or off display of @value{GDBN} expression debugging info. The
13370 @kindex show debug expression
13371 @item show debug expression
13372 Displays the current state of displaying @value{GDBN} expression
13374 @kindex set debug overload
13375 @item set debug overload
13376 Turns on or off display of @value{GDBN} C@t{++} overload debugging
13377 info. This includes info such as ranking of functions, etc. The default
13379 @kindex show debug overload
13380 @item show debug overload
13381 Displays the current state of displaying @value{GDBN} C@t{++} overload
13383 @kindex set debug remote
13384 @cindex packets, reporting on stdout
13385 @cindex serial connections, debugging
13386 @item set debug remote
13387 Turns on or off display of reports on all packets sent back and forth across
13388 the serial line to the remote machine. The info is printed on the
13389 @value{GDBN} standard output stream. The default is off.
13390 @kindex show debug remote
13391 @item show debug remote
13392 Displays the state of display of remote packets.
13393 @kindex set debug serial
13394 @item set debug serial
13395 Turns on or off display of @value{GDBN} serial debugging info. The
13397 @kindex show debug serial
13398 @item show debug serial
13399 Displays the current state of displaying @value{GDBN} serial debugging
13401 @kindex set debug target
13402 @item set debug target
13403 Turns on or off display of @value{GDBN} target debugging info. This info
13404 includes what is going on at the target level of GDB, as it happens. The
13406 @kindex show debug target
13407 @item show debug target
13408 Displays the current state of displaying @value{GDBN} target debugging
13410 @kindex set debug varobj
13411 @item set debug varobj
13412 Turns on or off display of @value{GDBN} variable object debugging
13413 info. The default is off.
13414 @kindex show debug varobj
13415 @item show debug varobj
13416 Displays the current state of displaying @value{GDBN} variable object
13421 @chapter Canned Sequences of Commands
13423 Aside from breakpoint commands (@pxref{Break Commands, ,Breakpoint
13424 command lists}), @value{GDBN} provides two ways to store sequences of
13425 commands for execution as a unit: user-defined commands and command
13429 * Define:: User-defined commands
13430 * Hooks:: User-defined command hooks
13431 * Command Files:: Command files
13432 * Output:: Commands for controlled output
13436 @section User-defined commands
13438 @cindex user-defined command
13439 A @dfn{user-defined command} is a sequence of @value{GDBN} commands to
13440 which you assign a new name as a command. This is done with the
13441 @code{define} command. User commands may accept up to 10 arguments
13442 separated by whitespace. Arguments are accessed within the user command
13443 via @var{$arg0@dots{}$arg9}. A trivial example:
13447 print $arg0 + $arg1 + $arg2
13451 To execute the command use:
13458 This defines the command @code{adder}, which prints the sum of
13459 its three arguments. Note the arguments are text substitutions, so they may
13460 reference variables, use complex expressions, or even perform inferior
13466 @item define @var{commandname}
13467 Define a command named @var{commandname}. If there is already a command
13468 by that name, you are asked to confirm that you want to redefine it.
13470 The definition of the command is made up of other @value{GDBN} command lines,
13471 which are given following the @code{define} command. The end of these
13472 commands is marked by a line containing @code{end}.
13477 Takes a single argument, which is an expression to evaluate.
13478 It is followed by a series of commands that are executed
13479 only if the expression is true (nonzero).
13480 There can then optionally be a line @code{else}, followed
13481 by a series of commands that are only executed if the expression
13482 was false. The end of the list is marked by a line containing @code{end}.
13486 The syntax is similar to @code{if}: the command takes a single argument,
13487 which is an expression to evaluate, and must be followed by the commands to
13488 execute, one per line, terminated by an @code{end}.
13489 The commands are executed repeatedly as long as the expression
13493 @item document @var{commandname}
13494 Document the user-defined command @var{commandname}, so that it can be
13495 accessed by @code{help}. The command @var{commandname} must already be
13496 defined. This command reads lines of documentation just as @code{define}
13497 reads the lines of the command definition, ending with @code{end}.
13498 After the @code{document} command is finished, @code{help} on command
13499 @var{commandname} displays the documentation you have written.
13501 You may use the @code{document} command again to change the
13502 documentation of a command. Redefining the command with @code{define}
13503 does not change the documentation.
13505 @kindex help user-defined
13506 @item help user-defined
13507 List all user-defined commands, with the first line of the documentation
13512 @itemx show user @var{commandname}
13513 Display the @value{GDBN} commands used to define @var{commandname} (but
13514 not its documentation). If no @var{commandname} is given, display the
13515 definitions for all user-defined commands.
13519 When user-defined commands are executed, the
13520 commands of the definition are not printed. An error in any command
13521 stops execution of the user-defined command.
13523 If used interactively, commands that would ask for confirmation proceed
13524 without asking when used inside a user-defined command. Many @value{GDBN}
13525 commands that normally print messages to say what they are doing omit the
13526 messages when used in a user-defined command.
13529 @section User-defined command hooks
13530 @cindex command hooks
13531 @cindex hooks, for commands
13532 @cindex hooks, pre-command
13536 You may define @dfn{hooks}, which are a special kind of user-defined
13537 command. Whenever you run the command @samp{foo}, if the user-defined
13538 command @samp{hook-foo} exists, it is executed (with no arguments)
13539 before that command.
13541 @cindex hooks, post-command
13544 A hook may also be defined which is run after the command you executed.
13545 Whenever you run the command @samp{foo}, if the user-defined command
13546 @samp{hookpost-foo} exists, it is executed (with no arguments) after
13547 that command. Post-execution hooks may exist simultaneously with
13548 pre-execution hooks, for the same command.
13550 It is valid for a hook to call the command which it hooks. If this
13551 occurs, the hook is not re-executed, thereby avoiding infinte recursion.
13553 @c It would be nice if hookpost could be passed a parameter indicating
13554 @c if the command it hooks executed properly or not. FIXME!
13556 @kindex stop@r{, a pseudo-command}
13557 In addition, a pseudo-command, @samp{stop} exists. Defining
13558 (@samp{hook-stop}) makes the associated commands execute every time
13559 execution stops in your program: before breakpoint commands are run,
13560 displays are printed, or the stack frame is printed.
13562 For example, to ignore @code{SIGALRM} signals while
13563 single-stepping, but treat them normally during normal execution,
13568 handle SIGALRM nopass
13572 handle SIGALRM pass
13575 define hook-continue
13576 handle SIGLARM pass
13580 As a further example, to hook at the begining and end of the @code{echo}
13581 command, and to add extra text to the beginning and end of the message,
13589 define hookpost-echo
13593 (@value{GDBP}) echo Hello World
13594 <<<---Hello World--->>>
13599 You can define a hook for any single-word command in @value{GDBN}, but
13600 not for command aliases; you should define a hook for the basic command
13601 name, e.g. @code{backtrace} rather than @code{bt}.
13602 @c FIXME! So how does Joe User discover whether a command is an alias
13604 If an error occurs during the execution of your hook, execution of
13605 @value{GDBN} commands stops and @value{GDBN} issues a prompt
13606 (before the command that you actually typed had a chance to run).
13608 If you try to define a hook which does not match any known command, you
13609 get a warning from the @code{define} command.
13611 @node Command Files
13612 @section Command files
13614 @cindex command files
13615 A command file for @value{GDBN} is a file of lines that are @value{GDBN}
13616 commands. Comments (lines starting with @kbd{#}) may also be included.
13617 An empty line in a command file does nothing; it does not mean to repeat
13618 the last command, as it would from the terminal.
13621 @cindex @file{.gdbinit}
13622 @cindex @file{gdb.ini}
13623 When you start @value{GDBN}, it automatically executes commands from its
13624 @dfn{init files}, normally called @file{.gdbinit}@footnote{The DJGPP
13625 port of @value{GDBN} uses the name @file{gdb.ini} instead, due to the
13626 limitations of file names imposed by DOS filesystems.}.
13627 During startup, @value{GDBN} does the following:
13631 Reads the init file (if any) in your home directory@footnote{On
13632 DOS/Windows systems, the home directory is the one pointed to by the
13633 @code{HOME} environment variable.}.
13636 Processes command line options and operands.
13639 Reads the init file (if any) in the current working directory.
13642 Reads command files specified by the @samp{-x} option.
13645 The init file in your home directory can set options (such as @samp{set
13646 complaints}) that affect subsequent processing of command line options
13647 and operands. Init files are not executed if you use the @samp{-nx}
13648 option (@pxref{Mode Options, ,Choosing modes}).
13650 @cindex init file name
13651 On some configurations of @value{GDBN}, the init file is known by a
13652 different name (these are typically environments where a specialized
13653 form of @value{GDBN} may need to coexist with other forms, hence a
13654 different name for the specialized version's init file). These are the
13655 environments with special init file names:
13657 @cindex @file{.vxgdbinit}
13660 VxWorks (Wind River Systems real-time OS): @file{.vxgdbinit}
13662 @cindex @file{.os68gdbinit}
13664 OS68K (Enea Data Systems real-time OS): @file{.os68gdbinit}
13666 @cindex @file{.esgdbinit}
13668 ES-1800 (Ericsson Telecom AB M68000 emulator): @file{.esgdbinit}
13671 You can also request the execution of a command file with the
13672 @code{source} command:
13676 @item source @var{filename}
13677 Execute the command file @var{filename}.
13680 The lines in a command file are executed sequentially. They are not
13681 printed as they are executed. An error in any command terminates execution
13682 of the command file.
13684 Commands that would ask for confirmation if used interactively proceed
13685 without asking when used in a command file. Many @value{GDBN} commands that
13686 normally print messages to say what they are doing omit the messages
13687 when called from command files.
13689 @value{GDBN} also accepts command input from standard input. In this
13690 mode, normal output goes to standard output and error output goes to
13691 standard error. Errors in a command file supplied on standard input do
13692 not terminate execution of the command file --- execution continues with
13696 gdb < cmds > log 2>&1
13699 (The syntax above will vary depending on the shell used.) This example
13700 will execute commands from the file @file{cmds}. All output and errors
13701 would be directed to @file{log}.
13704 @section Commands for controlled output
13706 During the execution of a command file or a user-defined command, normal
13707 @value{GDBN} output is suppressed; the only output that appears is what is
13708 explicitly printed by the commands in the definition. This section
13709 describes three commands useful for generating exactly the output you
13714 @item echo @var{text}
13715 @c I do not consider backslash-space a standard C escape sequence
13716 @c because it is not in ANSI.
13717 Print @var{text}. Nonprinting characters can be included in
13718 @var{text} using C escape sequences, such as @samp{\n} to print a
13719 newline. @strong{No newline is printed unless you specify one.}
13720 In addition to the standard C escape sequences, a backslash followed
13721 by a space stands for a space. This is useful for displaying a
13722 string with spaces at the beginning or the end, since leading and
13723 trailing spaces are otherwise trimmed from all arguments.
13724 To print @samp{@w{ }and foo =@w{ }}, use the command
13725 @samp{echo \@w{ }and foo = \@w{ }}.
13727 A backslash at the end of @var{text} can be used, as in C, to continue
13728 the command onto subsequent lines. For example,
13731 echo This is some text\n\
13732 which is continued\n\
13733 onto several lines.\n
13736 produces the same output as
13739 echo This is some text\n
13740 echo which is continued\n
13741 echo onto several lines.\n
13745 @item output @var{expression}
13746 Print the value of @var{expression} and nothing but that value: no
13747 newlines, no @samp{$@var{nn} = }. The value is not entered in the
13748 value history either. @xref{Expressions, ,Expressions}, for more information
13751 @item output/@var{fmt} @var{expression}
13752 Print the value of @var{expression} in format @var{fmt}. You can use
13753 the same formats as for @code{print}. @xref{Output Formats,,Output
13754 formats}, for more information.
13757 @item printf @var{string}, @var{expressions}@dots{}
13758 Print the values of the @var{expressions} under the control of
13759 @var{string}. The @var{expressions} are separated by commas and may be
13760 either numbers or pointers. Their values are printed as specified by
13761 @var{string}, exactly as if your program were to execute the C
13763 @c FIXME: the above implies that at least all ANSI C formats are
13764 @c supported, but it isn't true: %E and %G don't work (or so it seems).
13765 @c Either this is a bug, or the manual should document what formats are
13769 printf (@var{string}, @var{expressions}@dots{});
13772 For example, you can print two values in hex like this:
13775 printf "foo, bar-foo = 0x%x, 0x%x\n", foo, bar-foo
13778 The only backslash-escape sequences that you can use in the format
13779 string are the simple ones that consist of backslash followed by a
13784 @chapter @value{GDBN} Text User Interface
13788 * TUI Overview:: TUI overview
13789 * TUI Keys:: TUI key bindings
13790 * TUI Commands:: TUI specific commands
13791 * TUI Configuration:: TUI configuration variables
13794 The @value{GDBN} Text User Interface, TUI in short,
13795 is a terminal interface which uses the @code{curses} library
13796 to show the source file, the assembly output, the program registers
13797 and @value{GDBN} commands in separate text windows.
13798 The TUI is available only when @value{GDBN} is configured
13799 with the @code{--enable-tui} configure option (@pxref{Configure Options}).
13802 @section TUI overview
13804 The TUI has two display modes that can be switched while
13809 A curses (or TUI) mode in which it displays several text
13810 windows on the terminal.
13813 A standard mode which corresponds to the @value{GDBN} configured without
13817 In the TUI mode, @value{GDBN} can display several text window
13822 This window is the @value{GDBN} command window with the @value{GDBN}
13823 prompt and the @value{GDBN} outputs. The @value{GDBN} input is still
13824 managed using readline but through the TUI. The @emph{command}
13825 window is always visible.
13828 The source window shows the source file of the program. The current
13829 line as well as active breakpoints are displayed in this window.
13830 The current program position is shown with the @samp{>} marker and
13831 active breakpoints are shown with @samp{*} markers.
13834 The assembly window shows the disassembly output of the program.
13837 This window shows the processor registers. It detects when
13838 a register is changed and when this is the case, registers that have
13839 changed are highlighted.
13843 The source, assembly and register windows are attached to the thread
13844 and the frame position. They are updated when the current thread
13845 changes, when the frame changes or when the program counter changes.
13846 These three windows are arranged by the TUI according to several
13847 layouts. The layout defines which of these three windows are visible.
13848 The following layouts are available:
13858 source and assembly
13861 source and registers
13864 assembly and registers
13869 @section TUI Key Bindings
13870 @cindex TUI key bindings
13872 The TUI installs several key bindings in the readline keymaps
13873 (@pxref{Command Line Editing}).
13874 They allow to leave or enter in the TUI mode or they operate
13875 directly on the TUI layout and windows. The following key bindings
13876 are installed for both TUI mode and the @value{GDBN} standard mode.
13885 Enter or leave the TUI mode. When the TUI mode is left,
13886 the curses window management is left and @value{GDBN} operates using
13887 its standard mode writing on the terminal directly. When the TUI
13888 mode is entered, the control is given back to the curses windows.
13889 The screen is then refreshed.
13893 Use a TUI layout with only one window. The layout will
13894 either be @samp{source} or @samp{assembly}. When the TUI mode
13895 is not active, it will switch to the TUI mode.
13897 Think of this key binding as the Emacs @kbd{C-x 1} binding.
13901 Use a TUI layout with at least two windows. When the current
13902 layout shows already two windows, a next layout with two windows is used.
13903 When a new layout is chosen, one window will always be common to the
13904 previous layout and the new one.
13906 Think of it as the Emacs @kbd{C-x 2} binding.
13910 The following key bindings are handled only by the TUI mode:
13915 Scroll the active window one page up.
13919 Scroll the active window one page down.
13923 Scroll the active window one line up.
13927 Scroll the active window one line down.
13931 Scroll the active window one column left.
13935 Scroll the active window one column right.
13939 Refresh the screen.
13943 In the TUI mode, the arrow keys are used by the active window
13944 for scrolling. This means they are not available for readline. It is
13945 necessary to use other readline key bindings such as @key{C-p}, @key{C-n},
13946 @key{C-b} and @key{C-f}.
13949 @section TUI specific commands
13950 @cindex TUI commands
13952 The TUI has specific commands to control the text windows.
13953 These commands are always available, that is they do not depend on
13954 the current terminal mode in which @value{GDBN} runs. When @value{GDBN}
13955 is in the standard mode, using these commands will automatically switch
13960 @kindex layout next
13961 Display the next layout.
13964 @kindex layout prev
13965 Display the previous layout.
13969 Display the source window only.
13973 Display the assembly window only.
13976 @kindex layout split
13977 Display the source and assembly window.
13980 @kindex layout regs
13981 Display the register window together with the source or assembly window.
13983 @item focus next | prev | src | asm | regs | split
13985 Set the focus to the named window.
13986 This command allows to change the active window so that scrolling keys
13987 can be affected to another window.
13991 Refresh the screen. This is similar to using @key{C-L} key.
13995 Update the source window and the current execution point.
13997 @item winheight @var{name} +@var{count}
13998 @itemx winheight @var{name} -@var{count}
14000 Change the height of the window @var{name} by @var{count}
14001 lines. Positive counts increase the height, while negative counts
14006 @node TUI Configuration
14007 @section TUI configuration variables
14008 @cindex TUI configuration variables
14010 The TUI has several configuration variables that control the
14011 appearance of windows on the terminal.
14014 @item set tui border-kind @var{kind}
14015 @kindex set tui border-kind
14016 Select the border appearance for the source, assembly and register windows.
14017 The possible values are the following:
14020 Use a space character to draw the border.
14023 Use ascii characters + - and | to draw the border.
14026 Use the Alternate Character Set to draw the border. The border is
14027 drawn using character line graphics if the terminal supports them.
14031 @item set tui active-border-mode @var{mode}
14032 @kindex set tui active-border-mode
14033 Select the attributes to display the border of the active window.
14034 The possible values are @code{normal}, @code{standout}, @code{reverse},
14035 @code{half}, @code{half-standout}, @code{bold} and @code{bold-standout}.
14037 @item set tui border-mode @var{mode}
14038 @kindex set tui border-mode
14039 Select the attributes to display the border of other windows.
14040 The @var{mode} can be one of the following:
14043 Use normal attributes to display the border.
14049 Use reverse video mode.
14052 Use half bright mode.
14054 @item half-standout
14055 Use half bright and standout mode.
14058 Use extra bright or bold mode.
14060 @item bold-standout
14061 Use extra bright or bold and standout mode.
14068 @chapter Using @value{GDBN} under @sc{gnu} Emacs
14071 @cindex @sc{gnu} Emacs
14072 A special interface allows you to use @sc{gnu} Emacs to view (and
14073 edit) the source files for the program you are debugging with
14076 To use this interface, use the command @kbd{M-x gdb} in Emacs. Give the
14077 executable file you want to debug as an argument. This command starts
14078 @value{GDBN} as a subprocess of Emacs, with input and output through a newly
14079 created Emacs buffer.
14080 @c (Do not use the @code{-tui} option to run @value{GDBN} from Emacs.)
14082 Using @value{GDBN} under Emacs is just like using @value{GDBN} normally except for two
14087 All ``terminal'' input and output goes through the Emacs buffer.
14090 This applies both to @value{GDBN} commands and their output, and to the input
14091 and output done by the program you are debugging.
14093 This is useful because it means that you can copy the text of previous
14094 commands and input them again; you can even use parts of the output
14097 All the facilities of Emacs' Shell mode are available for interacting
14098 with your program. In particular, you can send signals the usual
14099 way---for example, @kbd{C-c C-c} for an interrupt, @kbd{C-c C-z} for a
14104 @value{GDBN} displays source code through Emacs.
14107 Each time @value{GDBN} displays a stack frame, Emacs automatically finds the
14108 source file for that frame and puts an arrow (@samp{=>}) at the
14109 left margin of the current line. Emacs uses a separate buffer for
14110 source display, and splits the screen to show both your @value{GDBN} session
14113 Explicit @value{GDBN} @code{list} or search commands still produce output as
14114 usual, but you probably have no reason to use them from Emacs.
14117 @emph{Warning:} If the directory where your program resides is not your
14118 current directory, it can be easy to confuse Emacs about the location of
14119 the source files, in which case the auxiliary display buffer does not
14120 appear to show your source. @value{GDBN} can find programs by searching your
14121 environment's @code{PATH} variable, so the @value{GDBN} input and output
14122 session proceeds normally; but Emacs does not get enough information
14123 back from @value{GDBN} to locate the source files in this situation. To
14124 avoid this problem, either start @value{GDBN} mode from the directory where
14125 your program resides, or specify an absolute file name when prompted for the
14126 @kbd{M-x gdb} argument.
14128 A similar confusion can result if you use the @value{GDBN} @code{file} command to
14129 switch to debugging a program in some other location, from an existing
14130 @value{GDBN} buffer in Emacs.
14133 By default, @kbd{M-x gdb} calls the program called @file{gdb}. If
14134 you need to call @value{GDBN} by a different name (for example, if you keep
14135 several configurations around, with different names) you can set the
14136 Emacs variable @code{gdb-command-name}; for example,
14139 (setq gdb-command-name "mygdb")
14143 (preceded by @kbd{M-:} or @kbd{ESC :}, or typed in the @code{*scratch*} buffer, or
14144 in your @file{.emacs} file) makes Emacs call the program named
14145 ``@code{mygdb}'' instead.
14147 In the @value{GDBN} I/O buffer, you can use these special Emacs commands in
14148 addition to the standard Shell mode commands:
14152 Describe the features of Emacs' @value{GDBN} Mode.
14155 Execute to another source line, like the @value{GDBN} @code{step} command; also
14156 update the display window to show the current file and location.
14159 Execute to next source line in this function, skipping all function
14160 calls, like the @value{GDBN} @code{next} command. Then update the display window
14161 to show the current file and location.
14164 Execute one instruction, like the @value{GDBN} @code{stepi} command; update
14165 display window accordingly.
14167 @item M-x gdb-nexti
14168 Execute to next instruction, using the @value{GDBN} @code{nexti} command; update
14169 display window accordingly.
14172 Execute until exit from the selected stack frame, like the @value{GDBN}
14173 @code{finish} command.
14176 Continue execution of your program, like the @value{GDBN} @code{continue}
14179 @emph{Warning:} In Emacs v19, this command is @kbd{C-c C-p}.
14182 Go up the number of frames indicated by the numeric argument
14183 (@pxref{Arguments, , Numeric Arguments, Emacs, The @sc{gnu} Emacs Manual}),
14184 like the @value{GDBN} @code{up} command.
14186 @emph{Warning:} In Emacs v19, this command is @kbd{C-c C-u}.
14189 Go down the number of frames indicated by the numeric argument, like the
14190 @value{GDBN} @code{down} command.
14192 @emph{Warning:} In Emacs v19, this command is @kbd{C-c C-d}.
14195 Read the number where the cursor is positioned, and insert it at the end
14196 of the @value{GDBN} I/O buffer. For example, if you wish to disassemble code
14197 around an address that was displayed earlier, type @kbd{disassemble};
14198 then move the cursor to the address display, and pick up the
14199 argument for @code{disassemble} by typing @kbd{C-x &}.
14201 You can customize this further by defining elements of the list
14202 @code{gdb-print-command}; once it is defined, you can format or
14203 otherwise process numbers picked up by @kbd{C-x &} before they are
14204 inserted. A numeric argument to @kbd{C-x &} indicates that you
14205 wish special formatting, and also acts as an index to pick an element of the
14206 list. If the list element is a string, the number to be inserted is
14207 formatted using the Emacs function @code{format}; otherwise the number
14208 is passed as an argument to the corresponding list element.
14211 In any source file, the Emacs command @kbd{C-x SPC} (@code{gdb-break})
14212 tells @value{GDBN} to set a breakpoint on the source line point is on.
14214 If you accidentally delete the source-display buffer, an easy way to get
14215 it back is to type the command @code{f} in the @value{GDBN} buffer, to
14216 request a frame display; when you run under Emacs, this recreates
14217 the source buffer if necessary to show you the context of the current
14220 The source files displayed in Emacs are in ordinary Emacs buffers
14221 which are visiting the source files in the usual way. You can edit
14222 the files with these buffers if you wish; but keep in mind that @value{GDBN}
14223 communicates with Emacs in terms of line numbers. If you add or
14224 delete lines from the text, the line numbers that @value{GDBN} knows cease
14225 to correspond properly with the code.
14227 @c The following dropped because Epoch is nonstandard. Reactivate
14228 @c if/when v19 does something similar. ---doc@cygnus.com 19dec1990
14230 @kindex Emacs Epoch environment
14234 Version 18 of @sc{gnu} Emacs has a built-in window system
14235 called the @code{epoch}
14236 environment. Users of this environment can use a new command,
14237 @code{inspect} which performs identically to @code{print} except that
14238 each value is printed in its own window.
14241 @include annotate.texi
14242 @include gdbmi.texinfo
14245 @chapter Reporting Bugs in @value{GDBN}
14246 @cindex bugs in @value{GDBN}
14247 @cindex reporting bugs in @value{GDBN}
14249 Your bug reports play an essential role in making @value{GDBN} reliable.
14251 Reporting a bug may help you by bringing a solution to your problem, or it
14252 may not. But in any case the principal function of a bug report is to help
14253 the entire community by making the next version of @value{GDBN} work better. Bug
14254 reports are your contribution to the maintenance of @value{GDBN}.
14256 In order for a bug report to serve its purpose, you must include the
14257 information that enables us to fix the bug.
14260 * Bug Criteria:: Have you found a bug?
14261 * Bug Reporting:: How to report bugs
14265 @section Have you found a bug?
14266 @cindex bug criteria
14268 If you are not sure whether you have found a bug, here are some guidelines:
14271 @cindex fatal signal
14272 @cindex debugger crash
14273 @cindex crash of debugger
14275 If the debugger gets a fatal signal, for any input whatever, that is a
14276 @value{GDBN} bug. Reliable debuggers never crash.
14278 @cindex error on valid input
14280 If @value{GDBN} produces an error message for valid input, that is a
14281 bug. (Note that if you're cross debugging, the problem may also be
14282 somewhere in the connection to the target.)
14284 @cindex invalid input
14286 If @value{GDBN} does not produce an error message for invalid input,
14287 that is a bug. However, you should note that your idea of
14288 ``invalid input'' might be our idea of ``an extension'' or ``support
14289 for traditional practice''.
14292 If you are an experienced user of debugging tools, your suggestions
14293 for improvement of @value{GDBN} are welcome in any case.
14296 @node Bug Reporting
14297 @section How to report bugs
14298 @cindex bug reports
14299 @cindex @value{GDBN} bugs, reporting
14301 A number of companies and individuals offer support for @sc{gnu} products.
14302 If you obtained @value{GDBN} from a support organization, we recommend you
14303 contact that organization first.
14305 You can find contact information for many support companies and
14306 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
14308 @c should add a web page ref...
14310 In any event, we also recommend that you send bug reports for
14311 @value{GDBN} to this addresses:
14317 @strong{Do not send bug reports to @samp{info-gdb}, or to
14318 @samp{help-gdb}, or to any newsgroups.} Most users of @value{GDBN} do
14319 not want to receive bug reports. Those that do have arranged to receive
14322 The mailing list @samp{bug-gdb} has a newsgroup @samp{gnu.gdb.bug} which
14323 serves as a repeater. The mailing list and the newsgroup carry exactly
14324 the same messages. Often people think of posting bug reports to the
14325 newsgroup instead of mailing them. This appears to work, but it has one
14326 problem which can be crucial: a newsgroup posting often lacks a mail
14327 path back to the sender. Thus, if we need to ask for more information,
14328 we may be unable to reach you. For this reason, it is better to send
14329 bug reports to the mailing list.
14331 As a last resort, send bug reports on paper to:
14334 @sc{gnu} Debugger Bugs
14335 Free Software Foundation Inc.
14336 59 Temple Place - Suite 330
14337 Boston, MA 02111-1307
14341 The fundamental principle of reporting bugs usefully is this:
14342 @strong{report all the facts}. If you are not sure whether to state a
14343 fact or leave it out, state it!
14345 Often people omit facts because they think they know what causes the
14346 problem and assume that some details do not matter. Thus, you might
14347 assume that the name of the variable you use in an example does not matter.
14348 Well, probably it does not, but one cannot be sure. Perhaps the bug is a
14349 stray memory reference which happens to fetch from the location where that
14350 name is stored in memory; perhaps, if the name were different, the contents
14351 of that location would fool the debugger into doing the right thing despite
14352 the bug. Play it safe and give a specific, complete example. That is the
14353 easiest thing for you to do, and the most helpful.
14355 Keep in mind that the purpose of a bug report is to enable us to fix the
14356 bug. It may be that the bug has been reported previously, but neither
14357 you nor we can know that unless your bug report is complete and
14360 Sometimes people give a few sketchy facts and ask, ``Does this ring a
14361 bell?'' Those bug reports are useless, and we urge everyone to
14362 @emph{refuse to respond to them} except to chide the sender to report
14365 To enable us to fix the bug, you should include all these things:
14369 The version of @value{GDBN}. @value{GDBN} announces it if you start
14370 with no arguments; you can also print it at any time using @code{show
14373 Without this, we will not know whether there is any point in looking for
14374 the bug in the current version of @value{GDBN}.
14377 The type of machine you are using, and the operating system name and
14381 What compiler (and its version) was used to compile @value{GDBN}---e.g.
14382 ``@value{GCC}--2.8.1''.
14385 What compiler (and its version) was used to compile the program you are
14386 debugging---e.g. ``@value{GCC}--2.8.1'', or ``HP92453-01 A.10.32.03 HP
14387 C Compiler''. For GCC, you can say @code{gcc --version} to get this
14388 information; for other compilers, see the documentation for those
14392 The command arguments you gave the compiler to compile your example and
14393 observe the bug. For example, did you use @samp{-O}? To guarantee
14394 you will not omit something important, list them all. A copy of the
14395 Makefile (or the output from make) is sufficient.
14397 If we were to try to guess the arguments, we would probably guess wrong
14398 and then we might not encounter the bug.
14401 A complete input script, and all necessary source files, that will
14405 A description of what behavior you observe that you believe is
14406 incorrect. For example, ``It gets a fatal signal.''
14408 Of course, if the bug is that @value{GDBN} gets a fatal signal, then we
14409 will certainly notice it. But if the bug is incorrect output, we might
14410 not notice unless it is glaringly wrong. You might as well not give us
14411 a chance to make a mistake.
14413 Even if the problem you experience is a fatal signal, you should still
14414 say so explicitly. Suppose something strange is going on, such as, your
14415 copy of @value{GDBN} is out of synch, or you have encountered a bug in
14416 the C library on your system. (This has happened!) Your copy might
14417 crash and ours would not. If you told us to expect a crash, then when
14418 ours fails to crash, we would know that the bug was not happening for
14419 us. If you had not told us to expect a crash, then we would not be able
14420 to draw any conclusion from our observations.
14423 If you wish to suggest changes to the @value{GDBN} source, send us context
14424 diffs. If you even discuss something in the @value{GDBN} source, refer to
14425 it by context, not by line number.
14427 The line numbers in our development sources will not match those in your
14428 sources. Your line numbers would convey no useful information to us.
14432 Here are some things that are not necessary:
14436 A description of the envelope of the bug.
14438 Often people who encounter a bug spend a lot of time investigating
14439 which changes to the input file will make the bug go away and which
14440 changes will not affect it.
14442 This is often time consuming and not very useful, because the way we
14443 will find the bug is by running a single example under the debugger
14444 with breakpoints, not by pure deduction from a series of examples.
14445 We recommend that you save your time for something else.
14447 Of course, if you can find a simpler example to report @emph{instead}
14448 of the original one, that is a convenience for us. Errors in the
14449 output will be easier to spot, running under the debugger will take
14450 less time, and so on.
14452 However, simplification is not vital; if you do not want to do this,
14453 report the bug anyway and send us the entire test case you used.
14456 A patch for the bug.
14458 A patch for the bug does help us if it is a good one. But do not omit
14459 the necessary information, such as the test case, on the assumption that
14460 a patch is all we need. We might see problems with your patch and decide
14461 to fix the problem another way, or we might not understand it at all.
14463 Sometimes with a program as complicated as @value{GDBN} it is very hard to
14464 construct an example that will make the program follow a certain path
14465 through the code. If you do not send us the example, we will not be able
14466 to construct one, so we will not be able to verify that the bug is fixed.
14468 And if we cannot understand what bug you are trying to fix, or why your
14469 patch should be an improvement, we will not install it. A test case will
14470 help us to understand.
14473 A guess about what the bug is or what it depends on.
14475 Such guesses are usually wrong. Even we cannot guess right about such
14476 things without first using the debugger to find the facts.
14479 @c The readline documentation is distributed with the readline code
14480 @c and consists of the two following files:
14482 @c inc-hist.texinfo
14483 @c Use -I with makeinfo to point to the appropriate directory,
14484 @c environment var TEXINPUTS with TeX.
14485 @include rluser.texinfo
14486 @include inc-hist.texinfo
14489 @node Formatting Documentation
14490 @appendix Formatting Documentation
14492 @cindex @value{GDBN} reference card
14493 @cindex reference card
14494 The @value{GDBN} 4 release includes an already-formatted reference card, ready
14495 for printing with PostScript or Ghostscript, in the @file{gdb}
14496 subdirectory of the main source directory@footnote{In
14497 @file{gdb-@value{GDBVN}/gdb/refcard.ps} of the version @value{GDBVN}
14498 release.}. If you can use PostScript or Ghostscript with your printer,
14499 you can print the reference card immediately with @file{refcard.ps}.
14501 The release also includes the source for the reference card. You
14502 can format it, using @TeX{}, by typing:
14508 The @value{GDBN} reference card is designed to print in @dfn{landscape}
14509 mode on US ``letter'' size paper;
14510 that is, on a sheet 11 inches wide by 8.5 inches
14511 high. You will need to specify this form of printing as an option to
14512 your @sc{dvi} output program.
14514 @cindex documentation
14516 All the documentation for @value{GDBN} comes as part of the machine-readable
14517 distribution. The documentation is written in Texinfo format, which is
14518 a documentation system that uses a single source file to produce both
14519 on-line information and a printed manual. You can use one of the Info
14520 formatting commands to create the on-line version of the documentation
14521 and @TeX{} (or @code{texi2roff}) to typeset the printed version.
14523 @value{GDBN} includes an already formatted copy of the on-line Info
14524 version of this manual in the @file{gdb} subdirectory. The main Info
14525 file is @file{gdb-@value{GDBVN}/gdb/gdb.info}, and it refers to
14526 subordinate files matching @samp{gdb.info*} in the same directory. If
14527 necessary, you can print out these files, or read them with any editor;
14528 but they are easier to read using the @code{info} subsystem in @sc{gnu}
14529 Emacs or the standalone @code{info} program, available as part of the
14530 @sc{gnu} Texinfo distribution.
14532 If you want to format these Info files yourself, you need one of the
14533 Info formatting programs, such as @code{texinfo-format-buffer} or
14536 If you have @code{makeinfo} installed, and are in the top level
14537 @value{GDBN} source directory (@file{gdb-@value{GDBVN}}, in the case of
14538 version @value{GDBVN}), you can make the Info file by typing:
14545 If you want to typeset and print copies of this manual, you need @TeX{},
14546 a program to print its @sc{dvi} output files, and @file{texinfo.tex}, the
14547 Texinfo definitions file.
14549 @TeX{} is a typesetting program; it does not print files directly, but
14550 produces output files called @sc{dvi} files. To print a typeset
14551 document, you need a program to print @sc{dvi} files. If your system
14552 has @TeX{} installed, chances are it has such a program. The precise
14553 command to use depends on your system; @kbd{lpr -d} is common; another
14554 (for PostScript devices) is @kbd{dvips}. The @sc{dvi} print command may
14555 require a file name without any extension or a @samp{.dvi} extension.
14557 @TeX{} also requires a macro definitions file called
14558 @file{texinfo.tex}. This file tells @TeX{} how to typeset a document
14559 written in Texinfo format. On its own, @TeX{} cannot either read or
14560 typeset a Texinfo file. @file{texinfo.tex} is distributed with GDB
14561 and is located in the @file{gdb-@var{version-number}/texinfo}
14564 If you have @TeX{} and a @sc{dvi} printer program installed, you can
14565 typeset and print this manual. First switch to the the @file{gdb}
14566 subdirectory of the main source directory (for example, to
14567 @file{gdb-@value{GDBVN}/gdb}) and type:
14573 Then give @file{gdb.dvi} to your @sc{dvi} printing program.
14575 @node Installing GDB
14576 @appendix Installing @value{GDBN}
14577 @cindex configuring @value{GDBN}
14578 @cindex installation
14580 @value{GDBN} comes with a @code{configure} script that automates the process
14581 of preparing @value{GDBN} for installation; you can then use @code{make} to
14582 build the @code{gdb} program.
14584 @c irrelevant in info file; it's as current as the code it lives with.
14585 @footnote{If you have a more recent version of @value{GDBN} than @value{GDBVN},
14586 look at the @file{README} file in the sources; we may have improved the
14587 installation procedures since publishing this manual.}
14590 The @value{GDBN} distribution includes all the source code you need for
14591 @value{GDBN} in a single directory, whose name is usually composed by
14592 appending the version number to @samp{gdb}.
14594 For example, the @value{GDBN} version @value{GDBVN} distribution is in the
14595 @file{gdb-@value{GDBVN}} directory. That directory contains:
14598 @item gdb-@value{GDBVN}/configure @r{(and supporting files)}
14599 script for configuring @value{GDBN} and all its supporting libraries
14601 @item gdb-@value{GDBVN}/gdb
14602 the source specific to @value{GDBN} itself
14604 @item gdb-@value{GDBVN}/bfd
14605 source for the Binary File Descriptor library
14607 @item gdb-@value{GDBVN}/include
14608 @sc{gnu} include files
14610 @item gdb-@value{GDBVN}/libiberty
14611 source for the @samp{-liberty} free software library
14613 @item gdb-@value{GDBVN}/opcodes
14614 source for the library of opcode tables and disassemblers
14616 @item gdb-@value{GDBVN}/readline
14617 source for the @sc{gnu} command-line interface
14619 @item gdb-@value{GDBVN}/glob
14620 source for the @sc{gnu} filename pattern-matching subroutine
14622 @item gdb-@value{GDBVN}/mmalloc
14623 source for the @sc{gnu} memory-mapped malloc package
14626 The simplest way to configure and build @value{GDBN} is to run @code{configure}
14627 from the @file{gdb-@var{version-number}} source directory, which in
14628 this example is the @file{gdb-@value{GDBVN}} directory.
14630 First switch to the @file{gdb-@var{version-number}} source directory
14631 if you are not already in it; then run @code{configure}. Pass the
14632 identifier for the platform on which @value{GDBN} will run as an
14638 cd gdb-@value{GDBVN}
14639 ./configure @var{host}
14644 where @var{host} is an identifier such as @samp{sun4} or
14645 @samp{decstation}, that identifies the platform where @value{GDBN} will run.
14646 (You can often leave off @var{host}; @code{configure} tries to guess the
14647 correct value by examining your system.)
14649 Running @samp{configure @var{host}} and then running @code{make} builds the
14650 @file{bfd}, @file{readline}, @file{mmalloc}, and @file{libiberty}
14651 libraries, then @code{gdb} itself. The configured source files, and the
14652 binaries, are left in the corresponding source directories.
14655 @code{configure} is a Bourne-shell (@code{/bin/sh}) script; if your
14656 system does not recognize this automatically when you run a different
14657 shell, you may need to run @code{sh} on it explicitly:
14660 sh configure @var{host}
14663 If you run @code{configure} from a directory that contains source
14664 directories for multiple libraries or programs, such as the
14665 @file{gdb-@value{GDBVN}} source directory for version @value{GDBVN}, @code{configure}
14666 creates configuration files for every directory level underneath (unless
14667 you tell it not to, with the @samp{--norecursion} option).
14669 You can run the @code{configure} script from any of the
14670 subordinate directories in the @value{GDBN} distribution if you only want to
14671 configure that subdirectory, but be sure to specify a path to it.
14673 For example, with version @value{GDBVN}, type the following to configure only
14674 the @code{bfd} subdirectory:
14678 cd gdb-@value{GDBVN}/bfd
14679 ../configure @var{host}
14683 You can install @code{@value{GDBP}} anywhere; it has no hardwired paths.
14684 However, you should make sure that the shell on your path (named by
14685 the @samp{SHELL} environment variable) is publicly readable. Remember
14686 that @value{GDBN} uses the shell to start your program---some systems refuse to
14687 let @value{GDBN} debug child processes whose programs are not readable.
14690 * Separate Objdir:: Compiling @value{GDBN} in another directory
14691 * Config Names:: Specifying names for hosts and targets
14692 * Configure Options:: Summary of options for configure
14695 @node Separate Objdir
14696 @section Compiling @value{GDBN} in another directory
14698 If you want to run @value{GDBN} versions for several host or target machines,
14699 you need a different @code{gdb} compiled for each combination of
14700 host and target. @code{configure} is designed to make this easy by
14701 allowing you to generate each configuration in a separate subdirectory,
14702 rather than in the source directory. If your @code{make} program
14703 handles the @samp{VPATH} feature (@sc{gnu} @code{make} does), running
14704 @code{make} in each of these directories builds the @code{gdb}
14705 program specified there.
14707 To build @code{gdb} in a separate directory, run @code{configure}
14708 with the @samp{--srcdir} option to specify where to find the source.
14709 (You also need to specify a path to find @code{configure}
14710 itself from your working directory. If the path to @code{configure}
14711 would be the same as the argument to @samp{--srcdir}, you can leave out
14712 the @samp{--srcdir} option; it is assumed.)
14714 For example, with version @value{GDBVN}, you can build @value{GDBN} in a
14715 separate directory for a Sun 4 like this:
14719 cd gdb-@value{GDBVN}
14722 ../gdb-@value{GDBVN}/configure sun4
14727 When @code{configure} builds a configuration using a remote source
14728 directory, it creates a tree for the binaries with the same structure
14729 (and using the same names) as the tree under the source directory. In
14730 the example, you'd find the Sun 4 library @file{libiberty.a} in the
14731 directory @file{gdb-sun4/libiberty}, and @value{GDBN} itself in
14732 @file{gdb-sun4/gdb}.
14734 One popular reason to build several @value{GDBN} configurations in separate
14735 directories is to configure @value{GDBN} for cross-compiling (where
14736 @value{GDBN} runs on one machine---the @dfn{host}---while debugging
14737 programs that run on another machine---the @dfn{target}).
14738 You specify a cross-debugging target by
14739 giving the @samp{--target=@var{target}} option to @code{configure}.
14741 When you run @code{make} to build a program or library, you must run
14742 it in a configured directory---whatever directory you were in when you
14743 called @code{configure} (or one of its subdirectories).
14745 The @code{Makefile} that @code{configure} generates in each source
14746 directory also runs recursively. If you type @code{make} in a source
14747 directory such as @file{gdb-@value{GDBVN}} (or in a separate configured
14748 directory configured with @samp{--srcdir=@var{dirname}/gdb-@value{GDBVN}}), you
14749 will build all the required libraries, and then build GDB.
14751 When you have multiple hosts or targets configured in separate
14752 directories, you can run @code{make} on them in parallel (for example,
14753 if they are NFS-mounted on each of the hosts); they will not interfere
14757 @section Specifying names for hosts and targets
14759 The specifications used for hosts and targets in the @code{configure}
14760 script are based on a three-part naming scheme, but some short predefined
14761 aliases are also supported. The full naming scheme encodes three pieces
14762 of information in the following pattern:
14765 @var{architecture}-@var{vendor}-@var{os}
14768 For example, you can use the alias @code{sun4} as a @var{host} argument,
14769 or as the value for @var{target} in a @code{--target=@var{target}}
14770 option. The equivalent full name is @samp{sparc-sun-sunos4}.
14772 The @code{configure} script accompanying @value{GDBN} does not provide
14773 any query facility to list all supported host and target names or
14774 aliases. @code{configure} calls the Bourne shell script
14775 @code{config.sub} to map abbreviations to full names; you can read the
14776 script, if you wish, or you can use it to test your guesses on
14777 abbreviations---for example:
14780 % sh config.sub i386-linux
14782 % sh config.sub alpha-linux
14783 alpha-unknown-linux-gnu
14784 % sh config.sub hp9k700
14786 % sh config.sub sun4
14787 sparc-sun-sunos4.1.1
14788 % sh config.sub sun3
14789 m68k-sun-sunos4.1.1
14790 % sh config.sub i986v
14791 Invalid configuration `i986v': machine `i986v' not recognized
14795 @code{config.sub} is also distributed in the @value{GDBN} source
14796 directory (@file{gdb-@value{GDBVN}}, for version @value{GDBVN}).
14798 @node Configure Options
14799 @section @code{configure} options
14801 Here is a summary of the @code{configure} options and arguments that
14802 are most often useful for building @value{GDBN}. @code{configure} also has
14803 several other options not listed here. @inforef{What Configure
14804 Does,,configure.info}, for a full explanation of @code{configure}.
14807 configure @r{[}--help@r{]}
14808 @r{[}--prefix=@var{dir}@r{]}
14809 @r{[}--exec-prefix=@var{dir}@r{]}
14810 @r{[}--srcdir=@var{dirname}@r{]}
14811 @r{[}--norecursion@r{]} @r{[}--rm@r{]}
14812 @r{[}--target=@var{target}@r{]}
14817 You may introduce options with a single @samp{-} rather than
14818 @samp{--} if you prefer; but you may abbreviate option names if you use
14823 Display a quick summary of how to invoke @code{configure}.
14825 @item --prefix=@var{dir}
14826 Configure the source to install programs and files under directory
14829 @item --exec-prefix=@var{dir}
14830 Configure the source to install programs under directory
14833 @c avoid splitting the warning from the explanation:
14835 @item --srcdir=@var{dirname}
14836 @strong{Warning: using this option requires @sc{gnu} @code{make}, or another
14837 @code{make} that implements the @code{VPATH} feature.}@*
14838 Use this option to make configurations in directories separate from the
14839 @value{GDBN} source directories. Among other things, you can use this to
14840 build (or maintain) several configurations simultaneously, in separate
14841 directories. @code{configure} writes configuration specific files in
14842 the current directory, but arranges for them to use the source in the
14843 directory @var{dirname}. @code{configure} creates directories under
14844 the working directory in parallel to the source directories below
14847 @item --norecursion
14848 Configure only the directory level where @code{configure} is executed; do not
14849 propagate configuration to subdirectories.
14851 @item --target=@var{target}
14852 Configure @value{GDBN} for cross-debugging programs running on the specified
14853 @var{target}. Without this option, @value{GDBN} is configured to debug
14854 programs that run on the same machine (@var{host}) as @value{GDBN} itself.
14856 There is no convenient way to generate a list of all available targets.
14858 @item @var{host} @dots{}
14859 Configure @value{GDBN} to run on the specified @var{host}.
14861 There is no convenient way to generate a list of all available hosts.
14864 There are many other options available as well, but they are generally
14865 needed for special purposes only.
14875 % I think something like @colophon should be in texinfo. In the
14877 \long\def\colophon{\hbox to0pt{}\vfill
14878 \centerline{The body of this manual is set in}
14879 \centerline{\fontname\tenrm,}
14880 \centerline{with headings in {\bf\fontname\tenbf}}
14881 \centerline{and examples in {\tt\fontname\tentt}.}
14882 \centerline{{\it\fontname\tenit\/},}
14883 \centerline{{\bf\fontname\tenbf}, and}
14884 \centerline{{\sl\fontname\tensl\/}}
14885 \centerline{are used for emphasis.}\vfill}
14887 % Blame: doc@cygnus.com, 1991.
14890 @c TeX can handle the contents at the start but makeinfo 3.12 can not