1 \input texinfo @c -*- texinfo -*-
2 @setfilename gdbint.info
4 @settitle @value{GDBN} Internals
6 @dircategory Software development
8 * Gdb-Internals: (gdbint). The GNU debugger's internals.
12 Copyright @copyright{} 1990, 1991, 1992, 1993, 1994, 1996, 1998, 1999,
13 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2008, 2009, 2010
14 Free Software Foundation, Inc.
15 Contributed by Cygnus Solutions. Written by John Gilmore.
16 Second Edition by Stan Shebs.
18 Permission is granted to copy, distribute and/or modify this document
19 under the terms of the GNU Free Documentation License, Version 1.1 or
20 any later version published by the Free Software Foundation; with no
21 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
22 Texts. A copy of the license is included in the section entitled ``GNU
23 Free Documentation License''.
27 This file documents the internals of the GNU debugger @value{GDBN}.
37 @title @value{GDBN} Internals
38 @subtitle{A guide to the internals of the GNU debugger}
40 @author Cygnus Solutions
41 @author Second Edition:
43 @author Cygnus Solutions
46 \def\$#1${{#1}} % Kluge: collect RCS revision info without $...$
47 \xdef\manvers{\$Revision$} % For use in headers, footers too
49 \hfill Cygnus Solutions\par
51 \hfill \TeX{}info \texinfoversion\par
55 @vskip 0pt plus 1filll
62 @c Perhaps this should be the title of the document (but only for info,
63 @c not for TeX). Existing GNU manuals seem inconsistent on this point.
64 @top Scope of this Document
66 This document documents the internals of the GNU debugger, @value{GDBN}. It
67 includes description of @value{GDBN}'s key algorithms and operations, as well
68 as the mechanisms that adapt @value{GDBN} to specific hosts and targets.
81 * Target Architecture Definition::
82 * Target Descriptions::
83 * Target Vector Definition::
88 * Versions and Branches::
89 * Start of New Year Procedure::
94 * GDB Observers:: @value{GDBN} Currently available observers
95 * GNU Free Documentation License:: The license for this documentation
108 @section Requirements
109 @cindex requirements for @value{GDBN}
111 Before diving into the internals, you should understand the formal
112 requirements and other expectations for @value{GDBN}. Although some
113 of these may seem obvious, there have been proposals for @value{GDBN}
114 that have run counter to these requirements.
116 First of all, @value{GDBN} is a debugger. It's not designed to be a
117 front panel for embedded systems. It's not a text editor. It's not a
118 shell. It's not a programming environment.
120 @value{GDBN} is an interactive tool. Although a batch mode is
121 available, @value{GDBN}'s primary role is to interact with a human
124 @value{GDBN} should be responsive to the user. A programmer hot on
125 the trail of a nasty bug, and operating under a looming deadline, is
126 going to be very impatient of everything, including the response time
127 to debugger commands.
129 @value{GDBN} should be relatively permissive, such as for expressions.
130 While the compiler should be picky (or have the option to be made
131 picky), since source code lives for a long time usually, the
132 programmer doing debugging shouldn't be spending time figuring out to
133 mollify the debugger.
135 @value{GDBN} will be called upon to deal with really large programs.
136 Executable sizes of 50 to 100 megabytes occur regularly, and we've
137 heard reports of programs approaching 1 gigabyte in size.
139 @value{GDBN} should be able to run everywhere. No other debugger is
140 available for even half as many configurations as @value{GDBN}
144 @section Contributors
146 The first edition of this document was written by John Gilmore of
147 Cygnus Solutions. The current second edition was written by Stan Shebs
148 of Cygnus Solutions, who continues to update the manual.
150 Over the years, many others have made additions and changes to this
151 document. This section attempts to record the significant contributors
152 to that effort. One of the virtues of free software is that everyone
153 is free to contribute to it; with regret, we cannot actually
154 acknowledge everyone here.
157 @emph{Plea:} This section has only been added relatively recently (four
158 years after publication of the second edition). Additions to this
159 section are particularly welcome. If you or your friends (or enemies,
160 to be evenhanded) have been unfairly omitted from this list, we would
161 like to add your names!
164 A document such as this relies on being kept up to date by numerous
165 small updates by contributing engineers as they make changes to the
166 code base. The file @file{ChangeLog} in the @value{GDBN} distribution
167 approximates a blow-by-blow account. The most prolific contributors to
168 this important, but low profile task are Andrew Cagney (responsible
169 for over half the entries), Daniel Jacobowitz, Mark Kettenis, Jim
170 Blandy and Eli Zaretskii.
172 Eli Zaretskii and Daniel Jacobowitz wrote the sections documenting
175 Jeremy Bennett updated the sections on initializing a new architecture
176 and register representation, and added the section on Frame Interpretation.
179 @node Overall Structure
181 @chapter Overall Structure
183 @value{GDBN} consists of three major subsystems: user interface,
184 symbol handling (the @dfn{symbol side}), and target system handling (the
187 The user interface consists of several actual interfaces, plus
190 The symbol side consists of object file readers, debugging info
191 interpreters, symbol table management, source language expression
192 parsing, type and value printing.
194 The target side consists of execution control, stack frame analysis, and
195 physical target manipulation.
197 The target side/symbol side division is not formal, and there are a
198 number of exceptions. For instance, core file support involves symbolic
199 elements (the basic core file reader is in BFD) and target elements (it
200 supplies the contents of memory and the values of registers). Instead,
201 this division is useful for understanding how the minor subsystems
204 @section The Symbol Side
206 The symbolic side of @value{GDBN} can be thought of as ``everything
207 you can do in @value{GDBN} without having a live program running''.
208 For instance, you can look at the types of variables, and evaluate
209 many kinds of expressions.
211 @section The Target Side
213 The target side of @value{GDBN} is the ``bits and bytes manipulator''.
214 Although it may make reference to symbolic info here and there, most
215 of the target side will run with only a stripped executable
216 available---or even no executable at all, in remote debugging cases.
218 Operations such as disassembly, stack frame crawls, and register
219 display, are able to work with no symbolic info at all. In some cases,
220 such as disassembly, @value{GDBN} will use symbolic info to present addresses
221 relative to symbols rather than as raw numbers, but it will work either
224 @section Configurations
228 @dfn{Host} refers to attributes of the system where @value{GDBN} runs.
229 @dfn{Target} refers to the system where the program being debugged
230 executes. In most cases they are the same machine, in which case a
231 third type of @dfn{Native} attributes come into play.
233 Defines and include files needed to build on the host are host
234 support. Examples are tty support, system defined types, host byte
235 order, host float format. These are all calculated by @code{autoconf}
236 when the debugger is built.
238 Defines and information needed to handle the target format are target
239 dependent. Examples are the stack frame format, instruction set,
240 breakpoint instruction, registers, and how to set up and tear down the stack
243 Information that is only needed when the host and target are the same,
244 is native dependent. One example is Unix child process support; if the
245 host and target are not the same, calling @code{fork} to start the target
246 process is a bad idea. The various macros needed for finding the
247 registers in the @code{upage}, running @code{ptrace}, and such are all
248 in the native-dependent files.
250 Another example of native-dependent code is support for features that
251 are really part of the target environment, but which require
252 @code{#include} files that are only available on the host system. Core
253 file handling and @code{setjmp} handling are two common cases.
255 When you want to make @value{GDBN} work as the traditional native debugger
256 on a system, you will need to supply both target and native information.
258 @section Source Tree Structure
259 @cindex @value{GDBN} source tree structure
261 The @value{GDBN} source directory has a mostly flat structure---there
262 are only a few subdirectories. A file's name usually gives a hint as
263 to what it does; for example, @file{stabsread.c} reads stabs,
264 @file{dwarf2read.c} reads @sc{DWARF 2}, etc.
266 Files that are related to some common task have names that share
267 common substrings. For example, @file{*-thread.c} files deal with
268 debugging threads on various platforms; @file{*read.c} files deal with
269 reading various kinds of symbol and object files; @file{inf*.c} files
270 deal with direct control of the @dfn{inferior program} (@value{GDBN}
271 parlance for the program being debugged).
273 There are several dozens of files in the @file{*-tdep.c} family.
274 @samp{tdep} stands for @dfn{target-dependent code}---each of these
275 files implements debug support for a specific target architecture
276 (sparc, mips, etc). Usually, only one of these will be used in a
277 specific @value{GDBN} configuration (sometimes two, closely related).
279 Similarly, there are many @file{*-nat.c} files, each one for native
280 debugging on a specific system (e.g., @file{sparc-linux-nat.c} is for
281 native debugging of Sparc machines running the Linux kernel).
283 The few subdirectories of the source tree are:
287 Code that implements @dfn{CLI}, the @value{GDBN} Command-Line
288 Interpreter. @xref{User Interface, Command Interpreter}.
291 Code for the @value{GDBN} remote server.
294 Code for Insight, the @value{GDBN} TK-based GUI front-end.
297 The @dfn{GDB/MI}, the @value{GDBN} Machine Interface interpreter.
300 Target signal translation code.
303 Code for @dfn{TUI}, the @value{GDBN} Text-mode full-screen User
304 Interface. @xref{User Interface, TUI}.
312 @value{GDBN} uses a number of debugging-specific algorithms. They are
313 often not very complicated, but get lost in the thicket of special
314 cases and real-world issues. This chapter describes the basic
315 algorithms and mentions some of the specific target definitions that
318 @section Prologue Analysis
320 @cindex prologue analysis
321 @cindex call frame information
322 @cindex CFI (call frame information)
323 To produce a backtrace and allow the user to manipulate older frames'
324 variables and arguments, @value{GDBN} needs to find the base addresses
325 of older frames, and discover where those frames' registers have been
326 saved. Since a frame's ``callee-saves'' registers get saved by
327 younger frames if and when they're reused, a frame's registers may be
328 scattered unpredictably across younger frames. This means that
329 changing the value of a register-allocated variable in an older frame
330 may actually entail writing to a save slot in some younger frame.
332 Modern versions of GCC emit Dwarf call frame information (``CFI''),
333 which describes how to find frame base addresses and saved registers.
334 But CFI is not always available, so as a fallback @value{GDBN} uses a
335 technique called @dfn{prologue analysis} to find frame sizes and saved
336 registers. A prologue analyzer disassembles the function's machine
337 code starting from its entry point, and looks for instructions that
338 allocate frame space, save the stack pointer in a frame pointer
339 register, save registers, and so on. Obviously, this can't be done
340 accurately in general, but it's tractable to do well enough to be very
341 helpful. Prologue analysis predates the GNU toolchain's support for
342 CFI; at one time, prologue analysis was the only mechanism
343 @value{GDBN} used for stack unwinding at all, when the function
344 calling conventions didn't specify a fixed frame layout.
346 In the olden days, function prologues were generated by hand-written,
347 target-specific code in GCC, and treated as opaque and untouchable by
348 optimizers. Looking at this code, it was usually straightforward to
349 write a prologue analyzer for @value{GDBN} that would accurately
350 understand all the prologues GCC would generate. However, over time
351 GCC became more aggressive about instruction scheduling, and began to
352 understand more about the semantics of the prologue instructions
353 themselves; in response, @value{GDBN}'s analyzers became more complex
354 and fragile. Keeping the prologue analyzers working as GCC (and the
355 instruction sets themselves) evolved became a substantial task.
357 @cindex @file{prologue-value.c}
358 @cindex abstract interpretation of function prologues
359 @cindex pseudo-evaluation of function prologues
360 To try to address this problem, the code in @file{prologue-value.h}
361 and @file{prologue-value.c} provides a general framework for writing
362 prologue analyzers that are simpler and more robust than ad-hoc
363 analyzers. When we analyze a prologue using the prologue-value
364 framework, we're really doing ``abstract interpretation'' or
365 ``pseudo-evaluation'': running the function's code in simulation, but
366 using conservative approximations of the values registers and memory
367 would hold when the code actually runs. For example, if our function
368 starts with the instruction:
371 addi r1, 42 # add 42 to r1
374 we don't know exactly what value will be in @code{r1} after executing
375 this instruction, but we do know it'll be 42 greater than its original
378 If we then see an instruction like:
381 addi r1, 22 # add 22 to r1
384 we still don't know what @code{r1's} value is, but again, we can say
385 it is now 64 greater than its original value.
387 If the next instruction were:
390 mov r2, r1 # set r2 to r1's value
393 then we can say that @code{r2's} value is now the original value of
396 It's common for prologues to save registers on the stack, so we'll
397 need to track the values of stack frame slots, as well as the
398 registers. So after an instruction like this:
404 then we'd know that the stack slot four bytes above the frame pointer
405 holds the original value of @code{r1} plus 64.
409 Of course, this can only go so far before it gets unreasonable. If we
410 wanted to be able to say anything about the value of @code{r1} after
414 xor r1, r3 # exclusive-or r1 and r3, place result in r1
417 then things would get pretty complex. But remember, we're just doing
418 a conservative approximation; if exclusive-or instructions aren't
419 relevant to prologues, we can just say @code{r1}'s value is now
420 ``unknown''. We can ignore things that are too complex, if that loss of
421 information is acceptable for our application.
423 So when we say ``conservative approximation'' here, what we mean is an
424 approximation that is either accurate, or marked ``unknown'', but
427 Using this framework, a prologue analyzer is simply an interpreter for
428 machine code, but one that uses conservative approximations for the
429 contents of registers and memory instead of actual values. Starting
430 from the function's entry point, you simulate instructions up to the
431 current PC, or an instruction that you don't know how to simulate.
432 Now you can examine the state of the registers and stack slots you've
438 To see how large your stack frame is, just check the value of the
439 stack pointer register; if it's the original value of the SP
440 minus a constant, then that constant is the stack frame's size.
441 If the SP's value has been marked as ``unknown'', then that means
442 the prologue has done something too complex for us to track, and
443 we don't know the frame size.
446 To see where we've saved the previous frame's registers, we just
447 search the values we've tracked --- stack slots, usually, but
448 registers, too, if you want --- for something equal to the register's
449 original value. If the calling conventions suggest a standard place
450 to save a given register, then we can check there first, but really,
451 anything that will get us back the original value will probably work.
454 This does take some work. But prologue analyzers aren't
455 quick-and-simple pattern patching to recognize a few fixed prologue
456 forms any more; they're big, hairy functions. Along with inferior
457 function calls, prologue analysis accounts for a substantial portion
458 of the time needed to stabilize a @value{GDBN} port. So it's
459 worthwhile to look for an approach that will be easier to understand
460 and maintain. In the approach described above:
465 It's easier to see that the analyzer is correct: you just see
466 whether the analyzer properly (albeit conservatively) simulates
467 the effect of each instruction.
470 It's easier to extend the analyzer: you can add support for new
471 instructions, and know that you haven't broken anything that
472 wasn't already broken before.
475 It's orthogonal: to gather new information, you don't need to
476 complicate the code for each instruction. As long as your domain
477 of conservative values is already detailed enough to tell you
478 what you need, then all the existing instruction simulations are
479 already gathering the right data for you.
483 The file @file{prologue-value.h} contains detailed comments explaining
484 the framework and how to use it.
487 @section Breakpoint Handling
490 In general, a breakpoint is a user-designated location in the program
491 where the user wants to regain control if program execution ever reaches
494 There are two main ways to implement breakpoints; either as ``hardware''
495 breakpoints or as ``software'' breakpoints.
497 @cindex hardware breakpoints
498 @cindex program counter
499 Hardware breakpoints are sometimes available as a builtin debugging
500 features with some chips. Typically these work by having dedicated
501 register into which the breakpoint address may be stored. If the PC
502 (shorthand for @dfn{program counter})
503 ever matches a value in a breakpoint registers, the CPU raises an
504 exception and reports it to @value{GDBN}.
506 Another possibility is when an emulator is in use; many emulators
507 include circuitry that watches the address lines coming out from the
508 processor, and force it to stop if the address matches a breakpoint's
511 A third possibility is that the target already has the ability to do
512 breakpoints somehow; for instance, a ROM monitor may do its own
513 software breakpoints. So although these are not literally ``hardware
514 breakpoints'', from @value{GDBN}'s point of view they work the same;
515 @value{GDBN} need not do anything more than set the breakpoint and wait
516 for something to happen.
518 Since they depend on hardware resources, hardware breakpoints may be
519 limited in number; when the user asks for more, @value{GDBN} will
520 start trying to set software breakpoints. (On some architectures,
521 notably the 32-bit x86 platforms, @value{GDBN} cannot always know
522 whether there's enough hardware resources to insert all the hardware
523 breakpoints and watchpoints. On those platforms, @value{GDBN} prints
524 an error message only when the program being debugged is continued.)
526 @cindex software breakpoints
527 Software breakpoints require @value{GDBN} to do somewhat more work.
528 The basic theory is that @value{GDBN} will replace a program
529 instruction with a trap, illegal divide, or some other instruction
530 that will cause an exception, and then when it's encountered,
531 @value{GDBN} will take the exception and stop the program. When the
532 user says to continue, @value{GDBN} will restore the original
533 instruction, single-step, re-insert the trap, and continue on.
535 Since it literally overwrites the program being tested, the program area
536 must be writable, so this technique won't work on programs in ROM. It
537 can also distort the behavior of programs that examine themselves,
538 although such a situation would be highly unusual.
540 Also, the software breakpoint instruction should be the smallest size of
541 instruction, so it doesn't overwrite an instruction that might be a jump
542 target, and cause disaster when the program jumps into the middle of the
543 breakpoint instruction. (Strictly speaking, the breakpoint must be no
544 larger than the smallest interval between instructions that may be jump
545 targets; perhaps there is an architecture where only even-numbered
546 instructions may jumped to.) Note that it's possible for an instruction
547 set not to have any instructions usable for a software breakpoint,
548 although in practice only the ARC has failed to define such an
551 Basic breakpoint object handling is in @file{breakpoint.c}. However,
552 much of the interesting breakpoint action is in @file{infrun.c}.
555 @cindex insert or remove software breakpoint
556 @findex target_remove_breakpoint
557 @findex target_insert_breakpoint
558 @item target_remove_breakpoint (@var{bp_tgt})
559 @itemx target_insert_breakpoint (@var{bp_tgt})
560 Insert or remove a software breakpoint at address
561 @code{@var{bp_tgt}->placed_address}. Returns zero for success,
562 non-zero for failure. On input, @var{bp_tgt} contains the address of the
563 breakpoint, and is otherwise initialized to zero. The fields of the
564 @code{struct bp_target_info} pointed to by @var{bp_tgt} are updated
565 to contain other information about the breakpoint on output. The field
566 @code{placed_address} may be updated if the breakpoint was placed at a
567 related address; the field @code{shadow_contents} contains the real
568 contents of the bytes where the breakpoint has been inserted,
569 if reading memory would return the breakpoint instead of the
570 underlying memory; the field @code{shadow_len} is the length of
571 memory cached in @code{shadow_contents}, if any; and the field
572 @code{placed_size} is optionally set and used by the target, if
573 it could differ from @code{shadow_len}.
575 For example, the remote target @samp{Z0} packet does not require
576 shadowing memory, so @code{shadow_len} is left at zero. However,
577 the length reported by @code{gdbarch_breakpoint_from_pc} is cached in
578 @code{placed_size}, so that a matching @samp{z0} packet can be
579 used to remove the breakpoint.
581 @cindex insert or remove hardware breakpoint
582 @findex target_remove_hw_breakpoint
583 @findex target_insert_hw_breakpoint
584 @item target_remove_hw_breakpoint (@var{bp_tgt})
585 @itemx target_insert_hw_breakpoint (@var{bp_tgt})
586 Insert or remove a hardware-assisted breakpoint at address
587 @code{@var{bp_tgt}->placed_address}. Returns zero for success,
588 non-zero for failure. See @code{target_insert_breakpoint} for
589 a description of the @code{struct bp_target_info} pointed to by
590 @var{bp_tgt}; the @code{shadow_contents} and
591 @code{shadow_len} members are not used for hardware breakpoints,
592 but @code{placed_size} may be.
595 @section Single Stepping
597 @section Signal Handling
599 @section Thread Handling
601 @section Inferior Function Calls
603 @section Longjmp Support
605 @cindex @code{longjmp} debugging
606 @value{GDBN} has support for figuring out that the target is doing a
607 @code{longjmp} and for stopping at the target of the jump, if we are
608 stepping. This is done with a few specialized internal breakpoints,
609 which are visible in the output of the @samp{maint info breakpoint}
612 @findex gdbarch_get_longjmp_target
613 To make this work, you need to define a function called
614 @code{gdbarch_get_longjmp_target}, which will examine the
615 @code{jmp_buf} structure and extract the @code{longjmp} target address.
616 Since @code{jmp_buf} is target specific and typically defined in a
617 target header not available to @value{GDBN}, you will need to
618 determine the offset of the PC manually and return that; many targets
619 define a @code{jb_pc_offset} field in the tdep structure to save the
620 value once calculated.
625 Watchpoints are a special kind of breakpoints (@pxref{Algorithms,
626 breakpoints}) which break when data is accessed rather than when some
627 instruction is executed. When you have data which changes without
628 your knowing what code does that, watchpoints are the silver bullet to
629 hunt down and kill such bugs.
631 @cindex hardware watchpoints
632 @cindex software watchpoints
633 Watchpoints can be either hardware-assisted or not; the latter type is
634 known as ``software watchpoints.'' @value{GDBN} always uses
635 hardware-assisted watchpoints if they are available, and falls back on
636 software watchpoints otherwise. Typical situations where @value{GDBN}
637 will use software watchpoints are:
641 The watched memory region is too large for the underlying hardware
642 watchpoint support. For example, each x86 debug register can watch up
643 to 4 bytes of memory, so trying to watch data structures whose size is
644 more than 16 bytes will cause @value{GDBN} to use software
648 The value of the expression to be watched depends on data held in
649 registers (as opposed to memory).
652 Too many different watchpoints requested. (On some architectures,
653 this situation is impossible to detect until the debugged program is
654 resumed.) Note that x86 debug registers are used both for hardware
655 breakpoints and for watchpoints, so setting too many hardware
656 breakpoints might cause watchpoint insertion to fail.
659 No hardware-assisted watchpoints provided by the target
663 Software watchpoints are very slow, since @value{GDBN} needs to
664 single-step the program being debugged and test the value of the
665 watched expression(s) after each instruction. The rest of this
666 section is mostly irrelevant for software watchpoints.
668 When the inferior stops, @value{GDBN} tries to establish, among other
669 possible reasons, whether it stopped due to a watchpoint being hit.
670 It first uses @code{STOPPED_BY_WATCHPOINT} to see if any watchpoint
671 was hit. If not, all watchpoint checking is skipped.
673 Then @value{GDBN} calls @code{target_stopped_data_address} exactly
674 once. This method returns the address of the watchpoint which
675 triggered, if the target can determine it. If the triggered address
676 is available, @value{GDBN} compares the address returned by this
677 method with each watched memory address in each active watchpoint.
678 For data-read and data-access watchpoints, @value{GDBN} announces
679 every watchpoint that watches the triggered address as being hit.
680 For this reason, data-read and data-access watchpoints
681 @emph{require} that the triggered address be available; if not, read
682 and access watchpoints will never be considered hit. For data-write
683 watchpoints, if the triggered address is available, @value{GDBN}
684 considers only those watchpoints which match that address;
685 otherwise, @value{GDBN} considers all data-write watchpoints. For
686 each data-write watchpoint that @value{GDBN} considers, it evaluates
687 the expression whose value is being watched, and tests whether the
688 watched value has changed. Watchpoints whose watched values have
689 changed are announced as hit.
691 @c FIXME move these to the main lists of target/native defns
693 @value{GDBN} uses several macros and primitives to support hardware
697 @findex TARGET_CAN_USE_HARDWARE_WATCHPOINT
698 @item TARGET_CAN_USE_HARDWARE_WATCHPOINT (@var{type}, @var{count}, @var{other})
699 Return the number of hardware watchpoints of type @var{type} that are
700 possible to be set. The value is positive if @var{count} watchpoints
701 of this type can be set, zero if setting watchpoints of this type is
702 not supported, and negative if @var{count} is more than the maximum
703 number of watchpoints of type @var{type} that can be set. @var{other}
704 is non-zero if other types of watchpoints are currently enabled (there
705 are architectures which cannot set watchpoints of different types at
708 @findex TARGET_REGION_OK_FOR_HW_WATCHPOINT
709 @item TARGET_REGION_OK_FOR_HW_WATCHPOINT (@var{addr}, @var{len})
710 Return non-zero if hardware watchpoints can be used to watch a region
711 whose address is @var{addr} and whose length in bytes is @var{len}.
713 @cindex insert or remove hardware watchpoint
714 @findex target_insert_watchpoint
715 @findex target_remove_watchpoint
716 @item target_insert_watchpoint (@var{addr}, @var{len}, @var{type})
717 @itemx target_remove_watchpoint (@var{addr}, @var{len}, @var{type})
718 Insert or remove a hardware watchpoint starting at @var{addr}, for
719 @var{len} bytes. @var{type} is the watchpoint type, one of the
720 possible values of the enumerated data type @code{target_hw_bp_type},
721 defined by @file{breakpoint.h} as follows:
724 enum target_hw_bp_type
726 hw_write = 0, /* Common (write) HW watchpoint */
727 hw_read = 1, /* Read HW watchpoint */
728 hw_access = 2, /* Access (read or write) HW watchpoint */
729 hw_execute = 3 /* Execute HW breakpoint */
734 These two macros should return 0 for success, non-zero for failure.
736 @findex target_stopped_data_address
737 @item target_stopped_data_address (@var{addr_p})
738 If the inferior has some watchpoint that triggered, place the address
739 associated with the watchpoint at the location pointed to by
740 @var{addr_p} and return non-zero. Otherwise, return zero. This
741 is required for data-read and data-access watchpoints. It is
742 not required for data-write watchpoints, but @value{GDBN} uses
743 it to improve handling of those also.
745 @value{GDBN} will only call this method once per watchpoint stop,
746 immediately after calling @code{STOPPED_BY_WATCHPOINT}. If the
747 target's watchpoint indication is sticky, i.e., stays set after
748 resuming, this method should clear it. For instance, the x86 debug
749 control register has sticky triggered flags.
751 @findex target_watchpoint_addr_within_range
752 @item target_watchpoint_addr_within_range (@var{target}, @var{addr}, @var{start}, @var{length})
753 Check whether @var{addr} (as returned by @code{target_stopped_data_address})
754 lies within the hardware-defined watchpoint region described by
755 @var{start} and @var{length}. This only needs to be provided if the
756 granularity of a watchpoint is greater than one byte, i.e., if the
757 watchpoint can also trigger on nearby addresses outside of the watched
760 @findex HAVE_STEPPABLE_WATCHPOINT
761 @item HAVE_STEPPABLE_WATCHPOINT
762 If defined to a non-zero value, it is not necessary to disable a
763 watchpoint to step over it. Like @code{gdbarch_have_nonsteppable_watchpoint},
764 this is usually set when watchpoints trigger at the instruction
765 which will perform an interesting read or write. It should be
766 set if there is a temporary disable bit which allows the processor
767 to step over the interesting instruction without raising the
768 watchpoint exception again.
770 @findex gdbarch_have_nonsteppable_watchpoint
771 @item int gdbarch_have_nonsteppable_watchpoint (@var{gdbarch})
772 If it returns a non-zero value, @value{GDBN} should disable a
773 watchpoint to step the inferior over it. This is usually set when
774 watchpoints trigger at the instruction which will perform an
775 interesting read or write.
777 @findex HAVE_CONTINUABLE_WATCHPOINT
778 @item HAVE_CONTINUABLE_WATCHPOINT
779 If defined to a non-zero value, it is possible to continue the
780 inferior after a watchpoint has been hit. This is usually set
781 when watchpoints trigger at the instruction following an interesting
784 @findex STOPPED_BY_WATCHPOINT
785 @item STOPPED_BY_WATCHPOINT (@var{wait_status})
786 Return non-zero if stopped by a watchpoint. @var{wait_status} is of
787 the type @code{struct target_waitstatus}, defined by @file{target.h}.
788 Normally, this macro is defined to invoke the function pointed to by
789 the @code{to_stopped_by_watchpoint} member of the structure (of the
790 type @code{target_ops}, defined on @file{target.h}) that describes the
791 target-specific operations; @code{to_stopped_by_watchpoint} ignores
792 the @var{wait_status} argument.
794 @value{GDBN} does not require the non-zero value returned by
795 @code{STOPPED_BY_WATCHPOINT} to be 100% correct, so if a target cannot
796 determine for sure whether the inferior stopped due to a watchpoint,
797 it could return non-zero ``just in case''.
800 @subsection Watchpoints and Threads
801 @cindex watchpoints, with threads
803 @value{GDBN} only supports process-wide watchpoints, which trigger
804 in all threads. @value{GDBN} uses the thread ID to make watchpoints
805 act as if they were thread-specific, but it cannot set hardware
806 watchpoints that only trigger in a specific thread. Therefore, even
807 if the target supports threads, per-thread debug registers, and
808 watchpoints which only affect a single thread, it should set the
809 per-thread debug registers for all threads to the same value. On
810 @sc{gnu}/Linux native targets, this is accomplished by using
811 @code{ALL_LWPS} in @code{target_insert_watchpoint} and
812 @code{target_remove_watchpoint} and by using
813 @code{linux_set_new_thread} to register a handler for newly created
816 @value{GDBN}'s @sc{gnu}/Linux support only reports a single event
817 at a time, although multiple events can trigger simultaneously for
818 multi-threaded programs. When multiple events occur, @file{linux-nat.c}
819 queues subsequent events and returns them the next time the program
820 is resumed. This means that @code{STOPPED_BY_WATCHPOINT} and
821 @code{target_stopped_data_address} only need to consult the current
822 thread's state---the thread indicated by @code{inferior_ptid}. If
823 two threads have hit watchpoints simultaneously, those routines
824 will be called a second time for the second thread.
826 @subsection x86 Watchpoints
827 @cindex x86 debug registers
828 @cindex watchpoints, on x86
830 The 32-bit Intel x86 (a.k.a.@: ia32) processors feature special debug
831 registers designed to facilitate debugging. @value{GDBN} provides a
832 generic library of functions that x86-based ports can use to implement
833 support for watchpoints and hardware-assisted breakpoints. This
834 subsection documents the x86 watchpoint facilities in @value{GDBN}.
836 (At present, the library functions read and write debug registers directly, and are
837 thus only available for native configurations.)
839 To use the generic x86 watchpoint support, a port should do the
843 @findex I386_USE_GENERIC_WATCHPOINTS
845 Define the macro @code{I386_USE_GENERIC_WATCHPOINTS} somewhere in the
846 target-dependent headers.
849 Include the @file{config/i386/nm-i386.h} header file @emph{after}
850 defining @code{I386_USE_GENERIC_WATCHPOINTS}.
853 Add @file{i386-nat.o} to the value of the Make variable
854 @code{NATDEPFILES} (@pxref{Native Debugging, NATDEPFILES}).
857 Provide implementations for the @code{I386_DR_LOW_*} macros described
858 below. Typically, each macro should call a target-specific function
859 which does the real work.
862 The x86 watchpoint support works by maintaining mirror images of the
863 debug registers. Values are copied between the mirror images and the
864 real debug registers via a set of macros which each target needs to
868 @findex I386_DR_LOW_SET_CONTROL
869 @item I386_DR_LOW_SET_CONTROL (@var{val})
870 Set the Debug Control (DR7) register to the value @var{val}.
872 @findex I386_DR_LOW_SET_ADDR
873 @item I386_DR_LOW_SET_ADDR (@var{idx}, @var{addr})
874 Put the address @var{addr} into the debug register number @var{idx}.
876 @findex I386_DR_LOW_RESET_ADDR
877 @item I386_DR_LOW_RESET_ADDR (@var{idx})
878 Reset (i.e.@: zero out) the address stored in the debug register
881 @findex I386_DR_LOW_GET_STATUS
882 @item I386_DR_LOW_GET_STATUS
883 Return the value of the Debug Status (DR6) register. This value is
884 used immediately after it is returned by
885 @code{I386_DR_LOW_GET_STATUS}, so as to support per-thread status
889 For each one of the 4 debug registers (whose indices are from 0 to 3)
890 that store addresses, a reference count is maintained by @value{GDBN},
891 to allow sharing of debug registers by several watchpoints. This
892 allows users to define several watchpoints that watch the same
893 expression, but with different conditions and/or commands, without
894 wasting debug registers which are in short supply. @value{GDBN}
895 maintains the reference counts internally, targets don't have to do
896 anything to use this feature.
898 The x86 debug registers can each watch a region that is 1, 2, or 4
899 bytes long. The ia32 architecture requires that each watched region
900 be appropriately aligned: 2-byte region on 2-byte boundary, 4-byte
901 region on 4-byte boundary. However, the x86 watchpoint support in
902 @value{GDBN} can watch unaligned regions and regions larger than 4
903 bytes (up to 16 bytes) by allocating several debug registers to watch
904 a single region. This allocation of several registers per a watched
905 region is also done automatically without target code intervention.
907 The generic x86 watchpoint support provides the following API for the
908 @value{GDBN}'s application code:
911 @findex i386_region_ok_for_watchpoint
912 @item i386_region_ok_for_watchpoint (@var{addr}, @var{len})
913 The macro @code{TARGET_REGION_OK_FOR_HW_WATCHPOINT} is set to call
914 this function. It counts the number of debug registers required to
915 watch a given region, and returns a non-zero value if that number is
916 less than 4, the number of debug registers available to x86
919 @findex i386_stopped_data_address
920 @item i386_stopped_data_address (@var{addr_p})
922 @code{target_stopped_data_address} is set to call this function.
924 function examines the breakpoint condition bits in the DR6 Debug
925 Status register, as returned by the @code{I386_DR_LOW_GET_STATUS}
926 macro, and returns the address associated with the first bit that is
929 @findex i386_stopped_by_watchpoint
930 @item i386_stopped_by_watchpoint (void)
931 The macro @code{STOPPED_BY_WATCHPOINT}
932 is set to call this function. The
933 argument passed to @code{STOPPED_BY_WATCHPOINT} is ignored. This
934 function examines the breakpoint condition bits in the DR6 Debug
935 Status register, as returned by the @code{I386_DR_LOW_GET_STATUS}
936 macro, and returns true if any bit is set. Otherwise, false is
939 @findex i386_insert_watchpoint
940 @findex i386_remove_watchpoint
941 @item i386_insert_watchpoint (@var{addr}, @var{len}, @var{type})
942 @itemx i386_remove_watchpoint (@var{addr}, @var{len}, @var{type})
943 Insert or remove a watchpoint. The macros
944 @code{target_insert_watchpoint} and @code{target_remove_watchpoint}
945 are set to call these functions. @code{i386_insert_watchpoint} first
946 looks for a debug register which is already set to watch the same
947 region for the same access types; if found, it just increments the
948 reference count of that debug register, thus implementing debug
949 register sharing between watchpoints. If no such register is found,
950 the function looks for a vacant debug register, sets its mirrored
951 value to @var{addr}, sets the mirrored value of DR7 Debug Control
952 register as appropriate for the @var{len} and @var{type} parameters,
953 and then passes the new values of the debug register and DR7 to the
954 inferior by calling @code{I386_DR_LOW_SET_ADDR} and
955 @code{I386_DR_LOW_SET_CONTROL}. If more than one debug register is
956 required to cover the given region, the above process is repeated for
959 @code{i386_remove_watchpoint} does the opposite: it resets the address
960 in the mirrored value of the debug register and its read/write and
961 length bits in the mirrored value of DR7, then passes these new
962 values to the inferior via @code{I386_DR_LOW_RESET_ADDR} and
963 @code{I386_DR_LOW_SET_CONTROL}. If a register is shared by several
964 watchpoints, each time a @code{i386_remove_watchpoint} is called, it
965 decrements the reference count, and only calls
966 @code{I386_DR_LOW_RESET_ADDR} and @code{I386_DR_LOW_SET_CONTROL} when
967 the count goes to zero.
969 @findex i386_insert_hw_breakpoint
970 @findex i386_remove_hw_breakpoint
971 @item i386_insert_hw_breakpoint (@var{bp_tgt})
972 @itemx i386_remove_hw_breakpoint (@var{bp_tgt})
973 These functions insert and remove hardware-assisted breakpoints. The
974 macros @code{target_insert_hw_breakpoint} and
975 @code{target_remove_hw_breakpoint} are set to call these functions.
976 The argument is a @code{struct bp_target_info *}, as described in
977 the documentation for @code{target_insert_breakpoint}.
978 These functions work like @code{i386_insert_watchpoint} and
979 @code{i386_remove_watchpoint}, respectively, except that they set up
980 the debug registers to watch instruction execution, and each
981 hardware-assisted breakpoint always requires exactly one debug
984 @findex i386_cleanup_dregs
985 @item i386_cleanup_dregs (void)
986 This function clears all the reference counts, addresses, and control
987 bits in the mirror images of the debug registers. It doesn't affect
988 the actual debug registers in the inferior process.
995 x86 processors support setting watchpoints on I/O reads or writes.
996 However, since no target supports this (as of March 2001), and since
997 @code{enum target_hw_bp_type} doesn't even have an enumeration for I/O
998 watchpoints, this feature is not yet available to @value{GDBN} running
1002 x86 processors can enable watchpoints locally, for the current task
1003 only, or globally, for all the tasks. For each debug register,
1004 there's a bit in the DR7 Debug Control register that determines
1005 whether the associated address is watched locally or globally. The
1006 current implementation of x86 watchpoint support in @value{GDBN}
1007 always sets watchpoints to be locally enabled, since global
1008 watchpoints might interfere with the underlying OS and are probably
1009 unavailable in many platforms.
1012 @section Checkpoints
1015 In the abstract, a checkpoint is a point in the execution history of
1016 the program, which the user may wish to return to at some later time.
1018 Internally, a checkpoint is a saved copy of the program state, including
1019 whatever information is required in order to restore the program to that
1020 state at a later time. This can be expected to include the state of
1021 registers and memory, and may include external state such as the state
1022 of open files and devices.
1024 There are a number of ways in which checkpoints may be implemented
1025 in gdb, e.g.@: as corefiles, as forked processes, and as some opaque
1026 method implemented on the target side.
1028 A corefile can be used to save an image of target memory and register
1029 state, which can in principle be restored later --- but corefiles do
1030 not typically include information about external entities such as
1031 open files. Currently this method is not implemented in gdb.
1033 A forked process can save the state of user memory and registers,
1034 as well as some subset of external (kernel) state. This method
1035 is used to implement checkpoints on Linux, and in principle might
1036 be used on other systems.
1038 Some targets, e.g.@: simulators, might have their own built-in
1039 method for saving checkpoints, and gdb might be able to take
1040 advantage of that capability without necessarily knowing any
1041 details of how it is done.
1044 @section Observing changes in @value{GDBN} internals
1045 @cindex observer pattern interface
1046 @cindex notifications about changes in internals
1048 In order to function properly, several modules need to be notified when
1049 some changes occur in the @value{GDBN} internals. Traditionally, these
1050 modules have relied on several paradigms, the most common ones being
1051 hooks and gdb-events. Unfortunately, none of these paradigms was
1052 versatile enough to become the standard notification mechanism in
1053 @value{GDBN}. The fact that they only supported one ``client'' was also
1054 a strong limitation.
1056 A new paradigm, based on the Observer pattern of the @cite{Design
1057 Patterns} book, has therefore been implemented. The goal was to provide
1058 a new interface overcoming the issues with the notification mechanisms
1059 previously available. This new interface needed to be strongly typed,
1060 easy to extend, and versatile enough to be used as the standard
1061 interface when adding new notifications.
1063 See @ref{GDB Observers} for a brief description of the observers
1064 currently implemented in GDB. The rationale for the current
1065 implementation is also briefly discussed.
1067 @node User Interface
1069 @chapter User Interface
1071 @value{GDBN} has several user interfaces, of which the traditional
1072 command-line interface is perhaps the most familiar.
1074 @section Command Interpreter
1076 @cindex command interpreter
1078 The command interpreter in @value{GDBN} is fairly simple. It is designed to
1079 allow for the set of commands to be augmented dynamically, and also
1080 has a recursive subcommand capability, where the first argument to
1081 a command may itself direct a lookup on a different command list.
1083 For instance, the @samp{set} command just starts a lookup on the
1084 @code{setlist} command list, while @samp{set thread} recurses
1085 to the @code{set_thread_cmd_list}.
1089 To add commands in general, use @code{add_cmd}. @code{add_com} adds to
1090 the main command list, and should be used for those commands. The usual
1091 place to add commands is in the @code{_initialize_@var{xyz}} routines at
1092 the ends of most source files.
1094 @findex add_setshow_cmd
1095 @findex add_setshow_cmd_full
1096 To add paired @samp{set} and @samp{show} commands, use
1097 @code{add_setshow_cmd} or @code{add_setshow_cmd_full}. The former is
1098 a slightly simpler interface which is useful when you don't need to
1099 further modify the new command structures, while the latter returns
1100 the new command structures for manipulation.
1102 @cindex deprecating commands
1103 @findex deprecate_cmd
1104 Before removing commands from the command set it is a good idea to
1105 deprecate them for some time. Use @code{deprecate_cmd} on commands or
1106 aliases to set the deprecated flag. @code{deprecate_cmd} takes a
1107 @code{struct cmd_list_element} as it's first argument. You can use the
1108 return value from @code{add_com} or @code{add_cmd} to deprecate the
1109 command immediately after it is created.
1111 The first time a command is used the user will be warned and offered a
1112 replacement (if one exists). Note that the replacement string passed to
1113 @code{deprecate_cmd} should be the full name of the command, i.e., the
1114 entire string the user should type at the command line.
1116 @anchor{UI-Independent Output}
1117 @section UI-Independent Output---the @code{ui_out} Functions
1118 @c This section is based on the documentation written by Fernando
1119 @c Nasser <fnasser@redhat.com>.
1121 @cindex @code{ui_out} functions
1122 The @code{ui_out} functions present an abstraction level for the
1123 @value{GDBN} output code. They hide the specifics of different user
1124 interfaces supported by @value{GDBN}, and thus free the programmer
1125 from the need to write several versions of the same code, one each for
1126 every UI, to produce output.
1128 @subsection Overview and Terminology
1130 In general, execution of each @value{GDBN} command produces some sort
1131 of output, and can even generate an input request.
1133 Output can be generated for the following purposes:
1137 to display a @emph{result} of an operation;
1140 to convey @emph{info} or produce side-effects of a requested
1144 to provide a @emph{notification} of an asynchronous event (including
1145 progress indication of a prolonged asynchronous operation);
1148 to display @emph{error messages} (including warnings);
1151 to show @emph{debug data};
1154 to @emph{query} or prompt a user for input (a special case).
1158 This section mainly concentrates on how to build result output,
1159 although some of it also applies to other kinds of output.
1161 Generation of output that displays the results of an operation
1162 involves one or more of the following:
1166 output of the actual data
1169 formatting the output as appropriate for console output, to make it
1170 easily readable by humans
1173 machine oriented formatting--a more terse formatting to allow for easy
1174 parsing by programs which read @value{GDBN}'s output
1177 annotation, whose purpose is to help legacy GUIs to identify interesting
1181 The @code{ui_out} routines take care of the first three aspects.
1182 Annotations are provided by separate annotation routines. Note that use
1183 of annotations for an interface between a GUI and @value{GDBN} is
1186 Output can be in the form of a single item, which we call a @dfn{field};
1187 a @dfn{list} consisting of identical fields; a @dfn{tuple} consisting of
1188 non-identical fields; or a @dfn{table}, which is a tuple consisting of a
1189 header and a body. In a BNF-like form:
1192 @item <table> @expansion{}
1193 @code{<header> <body>}
1194 @item <header> @expansion{}
1195 @code{@{ <column> @}}
1196 @item <column> @expansion{}
1197 @code{<width> <alignment> <title>}
1198 @item <body> @expansion{}
1203 @subsection General Conventions
1205 Most @code{ui_out} routines are of type @code{void}, the exceptions are
1206 @code{ui_out_stream_new} (which returns a pointer to the newly created
1207 object) and the @code{make_cleanup} routines.
1209 The first parameter is always the @code{ui_out} vector object, a pointer
1210 to a @code{struct ui_out}.
1212 The @var{format} parameter is like in @code{printf} family of functions.
1213 When it is present, there must also be a variable list of arguments
1214 sufficient used to satisfy the @code{%} specifiers in the supplied
1217 When a character string argument is not used in a @code{ui_out} function
1218 call, a @code{NULL} pointer has to be supplied instead.
1221 @subsection Table, Tuple and List Functions
1223 @cindex list output functions
1224 @cindex table output functions
1225 @cindex tuple output functions
1226 This section introduces @code{ui_out} routines for building lists,
1227 tuples and tables. The routines to output the actual data items
1228 (fields) are presented in the next section.
1230 To recap: A @dfn{tuple} is a sequence of @dfn{fields}, each field
1231 containing information about an object; a @dfn{list} is a sequence of
1232 fields where each field describes an identical object.
1234 Use the @dfn{table} functions when your output consists of a list of
1235 rows (tuples) and the console output should include a heading. Use this
1236 even when you are listing just one object but you still want the header.
1238 @cindex nesting level in @code{ui_out} functions
1239 Tables can not be nested. Tuples and lists can be nested up to a
1240 maximum of five levels.
1242 The overall structure of the table output code is something like this:
1257 Here is the description of table-, tuple- and list-related @code{ui_out}
1260 @deftypefun void ui_out_table_begin (struct ui_out *@var{uiout}, int @var{nbrofcols}, int @var{nr_rows}, const char *@var{tblid})
1261 The function @code{ui_out_table_begin} marks the beginning of the output
1262 of a table. It should always be called before any other @code{ui_out}
1263 function for a given table. @var{nbrofcols} is the number of columns in
1264 the table. @var{nr_rows} is the number of rows in the table.
1265 @var{tblid} is an optional string identifying the table. The string
1266 pointed to by @var{tblid} is copied by the implementation of
1267 @code{ui_out_table_begin}, so the application can free the string if it
1268 was @code{malloc}ed.
1270 The companion function @code{ui_out_table_end}, described below, marks
1271 the end of the table's output.
1274 @deftypefun void ui_out_table_header (struct ui_out *@var{uiout}, int @var{width}, enum ui_align @var{alignment}, const char *@var{colhdr})
1275 @code{ui_out_table_header} provides the header information for a single
1276 table column. You call this function several times, one each for every
1277 column of the table, after @code{ui_out_table_begin}, but before
1278 @code{ui_out_table_body}.
1280 The value of @var{width} gives the column width in characters. The
1281 value of @var{alignment} is one of @code{left}, @code{center}, and
1282 @code{right}, and it specifies how to align the header: left-justify,
1283 center, or right-justify it. @var{colhdr} points to a string that
1284 specifies the column header; the implementation copies that string, so
1285 column header strings in @code{malloc}ed storage can be freed after the
1289 @deftypefun void ui_out_table_body (struct ui_out *@var{uiout})
1290 This function delimits the table header from the table body.
1293 @deftypefun void ui_out_table_end (struct ui_out *@var{uiout})
1294 This function signals the end of a table's output. It should be called
1295 after the table body has been produced by the list and field output
1298 There should be exactly one call to @code{ui_out_table_end} for each
1299 call to @code{ui_out_table_begin}, otherwise the @code{ui_out} functions
1300 will signal an internal error.
1303 The output of the tuples that represent the table rows must follow the
1304 call to @code{ui_out_table_body} and precede the call to
1305 @code{ui_out_table_end}. You build a tuple by calling
1306 @code{ui_out_tuple_begin} and @code{ui_out_tuple_end}, with suitable
1307 calls to functions which actually output fields between them.
1309 @deftypefun void ui_out_tuple_begin (struct ui_out *@var{uiout}, const char *@var{id})
1310 This function marks the beginning of a tuple output. @var{id} points
1311 to an optional string that identifies the tuple; it is copied by the
1312 implementation, and so strings in @code{malloc}ed storage can be freed
1316 @deftypefun void ui_out_tuple_end (struct ui_out *@var{uiout})
1317 This function signals an end of a tuple output. There should be exactly
1318 one call to @code{ui_out_tuple_end} for each call to
1319 @code{ui_out_tuple_begin}, otherwise an internal @value{GDBN} error will
1323 @deftypefun {struct cleanup *} make_cleanup_ui_out_tuple_begin_end (struct ui_out *@var{uiout}, const char *@var{id})
1324 This function first opens the tuple and then establishes a cleanup
1325 (@pxref{Coding, Cleanups}) to close the tuple. It provides a convenient
1326 and correct implementation of the non-portable@footnote{The function
1327 cast is not portable ISO C.} code sequence:
1329 struct cleanup *old_cleanup;
1330 ui_out_tuple_begin (uiout, "...");
1331 old_cleanup = make_cleanup ((void(*)(void *)) ui_out_tuple_end,
1336 @deftypefun void ui_out_list_begin (struct ui_out *@var{uiout}, const char *@var{id})
1337 This function marks the beginning of a list output. @var{id} points to
1338 an optional string that identifies the list; it is copied by the
1339 implementation, and so strings in @code{malloc}ed storage can be freed
1343 @deftypefun void ui_out_list_end (struct ui_out *@var{uiout})
1344 This function signals an end of a list output. There should be exactly
1345 one call to @code{ui_out_list_end} for each call to
1346 @code{ui_out_list_begin}, otherwise an internal @value{GDBN} error will
1350 @deftypefun {struct cleanup *} make_cleanup_ui_out_list_begin_end (struct ui_out *@var{uiout}, const char *@var{id})
1351 Similar to @code{make_cleanup_ui_out_tuple_begin_end}, this function
1352 opens a list and then establishes cleanup (@pxref{Coding, Cleanups})
1353 that will close the list.
1356 @subsection Item Output Functions
1358 @cindex item output functions
1359 @cindex field output functions
1361 The functions described below produce output for the actual data
1362 items, or fields, which contain information about the object.
1364 Choose the appropriate function accordingly to your particular needs.
1366 @deftypefun void ui_out_field_fmt (struct ui_out *@var{uiout}, char *@var{fldname}, char *@var{format}, ...)
1367 This is the most general output function. It produces the
1368 representation of the data in the variable-length argument list
1369 according to formatting specifications in @var{format}, a
1370 @code{printf}-like format string. The optional argument @var{fldname}
1371 supplies the name of the field. The data items themselves are
1372 supplied as additional arguments after @var{format}.
1374 This generic function should be used only when it is not possible to
1375 use one of the specialized versions (see below).
1378 @deftypefun void ui_out_field_int (struct ui_out *@var{uiout}, const char *@var{fldname}, int @var{value})
1379 This function outputs a value of an @code{int} variable. It uses the
1380 @code{"%d"} output conversion specification. @var{fldname} specifies
1381 the name of the field.
1384 @deftypefun void ui_out_field_fmt_int (struct ui_out *@var{uiout}, int @var{width}, enum ui_align @var{alignment}, const char *@var{fldname}, int @var{value})
1385 This function outputs a value of an @code{int} variable. It differs from
1386 @code{ui_out_field_int} in that the caller specifies the desired @var{width} and @var{alignment} of the output.
1387 @var{fldname} specifies
1388 the name of the field.
1391 @deftypefun void ui_out_field_core_addr (struct ui_out *@var{uiout}, const char *@var{fldname}, struct gdbarch *@var{gdbarch}, CORE_ADDR @var{address})
1392 This function outputs an address as appropriate for @var{gdbarch}.
1395 @deftypefun void ui_out_field_string (struct ui_out *@var{uiout}, const char *@var{fldname}, const char *@var{string})
1396 This function outputs a string using the @code{"%s"} conversion
1400 Sometimes, there's a need to compose your output piece by piece using
1401 functions that operate on a stream, such as @code{value_print} or
1402 @code{fprintf_symbol_filtered}. These functions accept an argument of
1403 the type @code{struct ui_file *}, a pointer to a @code{ui_file} object
1404 used to store the data stream used for the output. When you use one
1405 of these functions, you need a way to pass their results stored in a
1406 @code{ui_file} object to the @code{ui_out} functions. To this end,
1407 you first create a @code{ui_stream} object by calling
1408 @code{ui_out_stream_new}, pass the @code{stream} member of that
1409 @code{ui_stream} object to @code{value_print} and similar functions,
1410 and finally call @code{ui_out_field_stream} to output the field you
1411 constructed. When the @code{ui_stream} object is no longer needed,
1412 you should destroy it and free its memory by calling
1413 @code{ui_out_stream_delete}.
1415 @deftypefun {struct ui_stream *} ui_out_stream_new (struct ui_out *@var{uiout})
1416 This function creates a new @code{ui_stream} object which uses the
1417 same output methods as the @code{ui_out} object whose pointer is
1418 passed in @var{uiout}. It returns a pointer to the newly created
1419 @code{ui_stream} object.
1422 @deftypefun void ui_out_stream_delete (struct ui_stream *@var{streambuf})
1423 This functions destroys a @code{ui_stream} object specified by
1427 @deftypefun void ui_out_field_stream (struct ui_out *@var{uiout}, const char *@var{fieldname}, struct ui_stream *@var{streambuf})
1428 This function consumes all the data accumulated in
1429 @code{streambuf->stream} and outputs it like
1430 @code{ui_out_field_string} does. After a call to
1431 @code{ui_out_field_stream}, the accumulated data no longer exists, but
1432 the stream is still valid and may be used for producing more fields.
1435 @strong{Important:} If there is any chance that your code could bail
1436 out before completing output generation and reaching the point where
1437 @code{ui_out_stream_delete} is called, it is necessary to set up a
1438 cleanup, to avoid leaking memory and other resources. Here's a
1439 skeleton code to do that:
1442 struct ui_stream *mybuf = ui_out_stream_new (uiout);
1443 struct cleanup *old = make_cleanup (ui_out_stream_delete, mybuf);
1448 If the function already has the old cleanup chain set (for other kinds
1449 of cleanups), you just have to add your cleanup to it:
1452 mybuf = ui_out_stream_new (uiout);
1453 make_cleanup (ui_out_stream_delete, mybuf);
1456 Note that with cleanups in place, you should not call
1457 @code{ui_out_stream_delete} directly, or you would attempt to free the
1460 @subsection Utility Output Functions
1462 @deftypefun void ui_out_field_skip (struct ui_out *@var{uiout}, const char *@var{fldname})
1463 This function skips a field in a table. Use it if you have to leave
1464 an empty field without disrupting the table alignment. The argument
1465 @var{fldname} specifies a name for the (missing) filed.
1468 @deftypefun void ui_out_text (struct ui_out *@var{uiout}, const char *@var{string})
1469 This function outputs the text in @var{string} in a way that makes it
1470 easy to be read by humans. For example, the console implementation of
1471 this method filters the text through a built-in pager, to prevent it
1472 from scrolling off the visible portion of the screen.
1474 Use this function for printing relatively long chunks of text around
1475 the actual field data: the text it produces is not aligned according
1476 to the table's format. Use @code{ui_out_field_string} to output a
1477 string field, and use @code{ui_out_message}, described below, to
1478 output short messages.
1481 @deftypefun void ui_out_spaces (struct ui_out *@var{uiout}, int @var{nspaces})
1482 This function outputs @var{nspaces} spaces. It is handy to align the
1483 text produced by @code{ui_out_text} with the rest of the table or
1487 @deftypefun void ui_out_message (struct ui_out *@var{uiout}, int @var{verbosity}, const char *@var{format}, ...)
1488 This function produces a formatted message, provided that the current
1489 verbosity level is at least as large as given by @var{verbosity}. The
1490 current verbosity level is specified by the user with the @samp{set
1491 verbositylevel} command.@footnote{As of this writing (April 2001),
1492 setting verbosity level is not yet implemented, and is always returned
1493 as zero. So calling @code{ui_out_message} with a @var{verbosity}
1494 argument more than zero will cause the message to never be printed.}
1497 @deftypefun void ui_out_wrap_hint (struct ui_out *@var{uiout}, char *@var{indent})
1498 This function gives the console output filter (a paging filter) a hint
1499 of where to break lines which are too long. Ignored for all other
1500 output consumers. @var{indent}, if non-@code{NULL}, is the string to
1501 be printed to indent the wrapped text on the next line; it must remain
1502 accessible until the next call to @code{ui_out_wrap_hint}, or until an
1503 explicit newline is produced by one of the other functions. If
1504 @var{indent} is @code{NULL}, the wrapped text will not be indented.
1507 @deftypefun void ui_out_flush (struct ui_out *@var{uiout})
1508 This function flushes whatever output has been accumulated so far, if
1509 the UI buffers output.
1513 @subsection Examples of Use of @code{ui_out} functions
1515 @cindex using @code{ui_out} functions
1516 @cindex @code{ui_out} functions, usage examples
1517 This section gives some practical examples of using the @code{ui_out}
1518 functions to generalize the old console-oriented code in
1519 @value{GDBN}. The examples all come from functions defined on the
1520 @file{breakpoints.c} file.
1522 This example, from the @code{breakpoint_1} function, shows how to
1525 The original code was:
1528 if (!found_a_breakpoint++)
1530 annotate_breakpoints_headers ();
1533 printf_filtered ("Num ");
1535 printf_filtered ("Type ");
1537 printf_filtered ("Disp ");
1539 printf_filtered ("Enb ");
1543 printf_filtered ("Address ");
1546 printf_filtered ("What\n");
1548 annotate_breakpoints_table ();
1552 Here's the new version:
1555 nr_printable_breakpoints = @dots{};
1558 ui_out_table_begin (ui, 6, nr_printable_breakpoints, "BreakpointTable");
1560 ui_out_table_begin (ui, 5, nr_printable_breakpoints, "BreakpointTable");
1562 if (nr_printable_breakpoints > 0)
1563 annotate_breakpoints_headers ();
1564 if (nr_printable_breakpoints > 0)
1566 ui_out_table_header (uiout, 3, ui_left, "number", "Num"); /* 1 */
1567 if (nr_printable_breakpoints > 0)
1569 ui_out_table_header (uiout, 14, ui_left, "type", "Type"); /* 2 */
1570 if (nr_printable_breakpoints > 0)
1572 ui_out_table_header (uiout, 4, ui_left, "disp", "Disp"); /* 3 */
1573 if (nr_printable_breakpoints > 0)
1575 ui_out_table_header (uiout, 3, ui_left, "enabled", "Enb"); /* 4 */
1578 if (nr_printable_breakpoints > 0)
1580 if (print_address_bits <= 32)
1581 ui_out_table_header (uiout, 10, ui_left, "addr", "Address");/* 5 */
1583 ui_out_table_header (uiout, 18, ui_left, "addr", "Address");/* 5 */
1585 if (nr_printable_breakpoints > 0)
1587 ui_out_table_header (uiout, 40, ui_noalign, "what", "What"); /* 6 */
1588 ui_out_table_body (uiout);
1589 if (nr_printable_breakpoints > 0)
1590 annotate_breakpoints_table ();
1593 This example, from the @code{print_one_breakpoint} function, shows how
1594 to produce the actual data for the table whose structure was defined
1595 in the above example. The original code was:
1600 printf_filtered ("%-3d ", b->number);
1602 if ((int)b->type > (sizeof(bptypes)/sizeof(bptypes[0]))
1603 || ((int) b->type != bptypes[(int) b->type].type))
1604 internal_error ("bptypes table does not describe type #%d.",
1606 printf_filtered ("%-14s ", bptypes[(int)b->type].description);
1608 printf_filtered ("%-4s ", bpdisps[(int)b->disposition]);
1610 printf_filtered ("%-3c ", bpenables[(int)b->enable]);
1614 This is the new version:
1618 ui_out_tuple_begin (uiout, "bkpt");
1620 ui_out_field_int (uiout, "number", b->number);
1622 if (((int) b->type > (sizeof (bptypes) / sizeof (bptypes[0])))
1623 || ((int) b->type != bptypes[(int) b->type].type))
1624 internal_error ("bptypes table does not describe type #%d.",
1626 ui_out_field_string (uiout, "type", bptypes[(int)b->type].description);
1628 ui_out_field_string (uiout, "disp", bpdisps[(int)b->disposition]);
1630 ui_out_field_fmt (uiout, "enabled", "%c", bpenables[(int)b->enable]);
1634 This example, also from @code{print_one_breakpoint}, shows how to
1635 produce a complicated output field using the @code{print_expression}
1636 functions which requires a stream to be passed. It also shows how to
1637 automate stream destruction with cleanups. The original code was:
1641 print_expression (b->exp, gdb_stdout);
1647 struct ui_stream *stb = ui_out_stream_new (uiout);
1648 struct cleanup *old_chain = make_cleanup_ui_out_stream_delete (stb);
1651 print_expression (b->exp, stb->stream);
1652 ui_out_field_stream (uiout, "what", local_stream);
1655 This example, also from @code{print_one_breakpoint}, shows how to use
1656 @code{ui_out_text} and @code{ui_out_field_string}. The original code
1661 if (b->dll_pathname == NULL)
1662 printf_filtered ("<any library> ");
1664 printf_filtered ("library \"%s\" ", b->dll_pathname);
1671 if (b->dll_pathname == NULL)
1673 ui_out_field_string (uiout, "what", "<any library>");
1674 ui_out_spaces (uiout, 1);
1678 ui_out_text (uiout, "library \"");
1679 ui_out_field_string (uiout, "what", b->dll_pathname);
1680 ui_out_text (uiout, "\" ");
1684 The following example from @code{print_one_breakpoint} shows how to
1685 use @code{ui_out_field_int} and @code{ui_out_spaces}. The original
1690 if (b->forked_inferior_pid != 0)
1691 printf_filtered ("process %d ", b->forked_inferior_pid);
1698 if (b->forked_inferior_pid != 0)
1700 ui_out_text (uiout, "process ");
1701 ui_out_field_int (uiout, "what", b->forked_inferior_pid);
1702 ui_out_spaces (uiout, 1);
1706 Here's an example of using @code{ui_out_field_string}. The original
1711 if (b->exec_pathname != NULL)
1712 printf_filtered ("program \"%s\" ", b->exec_pathname);
1719 if (b->exec_pathname != NULL)
1721 ui_out_text (uiout, "program \"");
1722 ui_out_field_string (uiout, "what", b->exec_pathname);
1723 ui_out_text (uiout, "\" ");
1727 Finally, here's an example of printing an address. The original code:
1731 printf_filtered ("%s ",
1732 hex_string_custom ((unsigned long) b->address, 8));
1739 ui_out_field_core_addr (uiout, "Address", b->address);
1743 @section Console Printing
1752 @cindex @code{libgdb}
1753 @code{libgdb} 1.0 was an abortive project of years ago. The theory was
1754 to provide an API to @value{GDBN}'s functionality.
1757 @cindex @code{libgdb}
1758 @code{libgdb} 2.0 is an ongoing effort to update @value{GDBN} so that is
1759 better able to support graphical and other environments.
1761 Since @code{libgdb} development is on-going, its architecture is still
1762 evolving. The following components have so far been identified:
1766 Observer - @file{gdb-events.h}.
1768 Builder - @file{ui-out.h}
1770 Event Loop - @file{event-loop.h}
1772 Library - @file{gdb.h}
1775 The model that ties these components together is described below.
1777 @section The @code{libgdb} Model
1779 A client of @code{libgdb} interacts with the library in two ways.
1783 As an observer (using @file{gdb-events}) receiving notifications from
1784 @code{libgdb} of any internal state changes (break point changes, run
1787 As a client querying @code{libgdb} (using the @file{ui-out} builder) to
1788 obtain various status values from @value{GDBN}.
1791 Since @code{libgdb} could have multiple clients (e.g., a GUI supporting
1792 the existing @value{GDBN} CLI), those clients must co-operate when
1793 controlling @code{libgdb}. In particular, a client must ensure that
1794 @code{libgdb} is idle (i.e.@: no other client is using @code{libgdb})
1795 before responding to a @file{gdb-event} by making a query.
1797 @section CLI support
1799 At present @value{GDBN}'s CLI is very much entangled in with the core of
1800 @code{libgdb}. Consequently, a client wishing to include the CLI in
1801 their interface needs to carefully co-ordinate its own and the CLI's
1804 It is suggested that the client set @code{libgdb} up to be bi-modal
1805 (alternate between CLI and client query modes). The notes below sketch
1810 The client registers itself as an observer of @code{libgdb}.
1812 The client create and install @code{cli-out} builder using its own
1813 versions of the @code{ui-file} @code{gdb_stderr}, @code{gdb_stdtarg} and
1814 @code{gdb_stdout} streams.
1816 The client creates a separate custom @code{ui-out} builder that is only
1817 used while making direct queries to @code{libgdb}.
1820 When the client receives input intended for the CLI, it simply passes it
1821 along. Since the @code{cli-out} builder is installed by default, all
1822 the CLI output in response to that command is routed (pronounced rooted)
1823 through to the client controlled @code{gdb_stdout} et.@: al.@: streams.
1824 At the same time, the client is kept abreast of internal changes by
1825 virtue of being a @code{libgdb} observer.
1827 The only restriction on the client is that it must wait until
1828 @code{libgdb} becomes idle before initiating any queries (using the
1829 client's custom builder).
1831 @section @code{libgdb} components
1833 @subheading Observer - @file{gdb-events.h}
1834 @file{gdb-events} provides the client with a very raw mechanism that can
1835 be used to implement an observer. At present it only allows for one
1836 observer and that observer must, internally, handle the need to delay
1837 the processing of any event notifications until after @code{libgdb} has
1838 finished the current command.
1840 @subheading Builder - @file{ui-out.h}
1841 @file{ui-out} provides the infrastructure necessary for a client to
1842 create a builder. That builder is then passed down to @code{libgdb}
1843 when doing any queries.
1845 @subheading Event Loop - @file{event-loop.h}
1846 @c There could be an entire section on the event-loop
1847 @file{event-loop}, currently non-re-entrant, provides a simple event
1848 loop. A client would need to either plug its self into this loop or,
1849 implement a new event-loop that @value{GDBN} would use.
1851 The event-loop will eventually be made re-entrant. This is so that
1852 @value{GDBN} can better handle the problem of some commands blocking
1853 instead of returning.
1855 @subheading Library - @file{gdb.h}
1856 @file{libgdb} is the most obvious component of this system. It provides
1857 the query interface. Each function is parameterized by a @code{ui-out}
1858 builder. The result of the query is constructed using that builder
1859 before the query function returns.
1866 @cindex @code{value} structure
1867 @value{GDBN} uses @code{struct value}, or @dfn{values}, as an internal
1868 abstraction for the representation of a variety of inferior objects
1869 and @value{GDBN} convenience objects.
1871 Values have an associated @code{struct type}, that describes a virtual
1872 view of the raw data or object stored in or accessed through the
1875 A value is in addition discriminated by its lvalue-ness, given its
1876 @code{enum lval_type} enumeration type:
1878 @cindex @code{lval_type} enumeration, for values.
1880 @item @code{not_lval}
1881 This value is not an lval. It can't be assigned to.
1883 @item @code{lval_memory}
1884 This value represents an object in memory.
1886 @item @code{lval_register}
1887 This value represents an object that lives in a register.
1889 @item @code{lval_internalvar}
1890 Represents the value of an internal variable.
1892 @item @code{lval_internalvar_component}
1893 Represents part of a @value{GDBN} internal variable. E.g., a
1896 @cindex computed values
1897 @item @code{lval_computed}
1898 These are ``computed'' values. They allow creating specialized value
1899 objects for specific purposes, all abstracted away from the core value
1900 support code. The creator of such a value writes specialized
1901 functions to handle the reading and writing to/from the value's
1902 backend data, and optionally, a ``copy operator'' and a
1905 Pointers to these functions are stored in a @code{struct lval_funcs}
1906 instance (declared in @file{value.h}), and passed to the
1907 @code{allocate_computed_value} function, as in the example below.
1911 nil_value_read (struct value *v)
1913 /* This callback reads data from some backend, and stores it in V.
1914 In this case, we always read null data. You'll want to fill in
1915 something more interesting. */
1917 memset (value_contents_all_raw (v),
1919 TYPE_LENGTH (value_type (v)));
1923 nil_value_write (struct value *v, struct value *fromval)
1925 /* Takes the data from FROMVAL and stores it in the backend of V. */
1927 to_oblivion (value_contents_all_raw (fromval),
1929 TYPE_LENGTH (value_type (fromval)));
1932 static struct lval_funcs nil_value_funcs =
1939 make_nil_value (void)
1944 type = make_nils_type ();
1945 v = allocate_computed_value (type, &nil_value_funcs, NULL);
1951 See the implementation of the @code{$_siginfo} convenience variable in
1952 @file{infrun.c} as a real example use of lval_computed.
1957 @chapter Stack Frames
1960 @cindex call stack frame
1961 A frame is a construct that @value{GDBN} uses to keep track of calling
1962 and called functions.
1964 @cindex unwind frame
1965 @value{GDBN}'s frame model, a fresh design, was implemented with the
1966 need to support @sc{dwarf}'s Call Frame Information in mind. In fact,
1967 the term ``unwind'' is taken directly from that specification.
1968 Developers wishing to learn more about unwinders, are encouraged to
1969 read the @sc{dwarf} specification, available from
1970 @url{http://www.dwarfstd.org}.
1972 @findex frame_register_unwind
1973 @findex get_frame_register
1974 @value{GDBN}'s model is that you find a frame's registers by
1975 ``unwinding'' them from the next younger frame. That is,
1976 @samp{get_frame_register} which returns the value of a register in
1977 frame #1 (the next-to-youngest frame), is implemented by calling frame
1978 #0's @code{frame_register_unwind} (the youngest frame). But then the
1979 obvious question is: how do you access the registers of the youngest
1982 @cindex sentinel frame
1983 @findex get_frame_type
1984 @vindex SENTINEL_FRAME
1985 To answer this question, @value{GDBN} has the @dfn{sentinel} frame, the
1986 ``-1st'' frame. Unwinding registers from the sentinel frame gives you
1987 the current values of the youngest real frame's registers. If @var{f}
1988 is a sentinel frame, then @code{get_frame_type (@var{f}) @equiv{}
1991 @section Selecting an Unwinder
1993 @findex frame_unwind_prepend_unwinder
1994 @findex frame_unwind_append_unwinder
1995 The architecture registers a list of frame unwinders (@code{struct
1996 frame_unwind}), using the functions
1997 @code{frame_unwind_prepend_unwinder} and
1998 @code{frame_unwind_append_unwinder}. Each unwinder includes a
1999 sniffer. Whenever @value{GDBN} needs to unwind a frame (to fetch the
2000 previous frame's registers or the current frame's ID), it calls
2001 registered sniffers in order to find one which recognizes the frame.
2002 The first time a sniffer returns non-zero, the corresponding unwinder
2003 is assigned to the frame.
2005 @section Unwinding the Frame ID
2008 Every frame has an associated ID, of type @code{struct frame_id}.
2009 The ID includes the stack base and function start address for
2010 the frame. The ID persists through the entire life of the frame,
2011 including while other called frames are running; it is used to
2012 locate an appropriate @code{struct frame_info} from the cache.
2014 Every time the inferior stops, and at various other times, the frame
2015 cache is flushed. Because of this, parts of @value{GDBN} which need
2016 to keep track of individual frames cannot use pointers to @code{struct
2017 frame_info}. A frame ID provides a stable reference to a frame, even
2018 when the unwinder must be run again to generate a new @code{struct
2019 frame_info} for the same frame.
2021 The frame's unwinder's @code{this_id} method is called to find the ID.
2022 Note that this is different from register unwinding, where the next
2023 frame's @code{prev_register} is called to unwind this frame's
2026 Both stack base and function address are required to identify the
2027 frame, because a recursive function has the same function address for
2028 two consecutive frames and a leaf function may have the same stack
2029 address as its caller. On some platforms, a third address is part of
2030 the ID to further disambiguate frames---for instance, on IA-64
2031 the separate register stack address is included in the ID.
2033 An invalid frame ID (@code{outer_frame_id}) returned from the
2034 @code{this_id} method means to stop unwinding after this frame.
2036 @code{null_frame_id} is another invalid frame ID which should be used
2037 when there is no frame. For instance, certain breakpoints are attached
2038 to a specific frame, and that frame is identified through its frame ID
2039 (we use this to implement the "finish" command). Using
2040 @code{null_frame_id} as the frame ID for a given breakpoint means
2041 that the breakpoint is not specific to any frame. The @code{this_id}
2042 method should never return @code{null_frame_id}.
2044 @section Unwinding Registers
2046 Each unwinder includes a @code{prev_register} method. This method
2047 takes a frame, an associated cache pointer, and a register number.
2048 It returns a @code{struct value *} describing the requested register,
2049 as saved by this frame. This is the value of the register that is
2050 current in this frame's caller.
2052 The returned value must have the same type as the register. It may
2053 have any lvalue type. In most circumstances one of these routines
2054 will generate the appropriate value:
2057 @item frame_unwind_got_optimized
2058 @findex frame_unwind_got_optimized
2059 This register was not saved.
2061 @item frame_unwind_got_register
2062 @findex frame_unwind_got_register
2063 This register was copied into another register in this frame. This
2064 is also used for unchanged registers; they are ``copied'' into the
2067 @item frame_unwind_got_memory
2068 @findex frame_unwind_got_memory
2069 This register was saved in memory.
2071 @item frame_unwind_got_constant
2072 @findex frame_unwind_got_constant
2073 This register was not saved, but the unwinder can compute the previous
2074 value some other way.
2076 @item frame_unwind_got_address
2077 @findex frame_unwind_got_address
2078 Same as @code{frame_unwind_got_constant}, except that the value is a target
2079 address. This is frequently used for the stack pointer, which is not
2080 explicitly saved but has a known offset from this frame's stack
2081 pointer. For architectures with a flat unified address space, this is
2082 generally the same as @code{frame_unwind_got_constant}.
2085 @node Symbol Handling
2087 @chapter Symbol Handling
2089 Symbols are a key part of @value{GDBN}'s operation. Symbols include
2090 variables, functions, and types.
2092 Symbol information for a large program can be truly massive, and
2093 reading of symbol information is one of the major performance
2094 bottlenecks in @value{GDBN}; it can take many minutes to process it
2095 all. Studies have shown that nearly all the time spent is
2096 computational, rather than file reading.
2098 One of the ways for @value{GDBN} to provide a good user experience is
2099 to start up quickly, taking no more than a few seconds. It is simply
2100 not possible to process all of a program's debugging info in that
2101 time, and so we attempt to handle symbols incrementally. For instance,
2102 we create @dfn{partial symbol tables} consisting of only selected
2103 symbols, and only expand them to full symbol tables when necessary.
2105 @section Symbol Reading
2107 @cindex symbol reading
2108 @cindex reading of symbols
2109 @cindex symbol files
2110 @value{GDBN} reads symbols from @dfn{symbol files}. The usual symbol
2111 file is the file containing the program which @value{GDBN} is
2112 debugging. @value{GDBN} can be directed to use a different file for
2113 symbols (with the @samp{symbol-file} command), and it can also read
2114 more symbols via the @samp{add-file} and @samp{load} commands. In
2115 addition, it may bring in more symbols while loading shared
2118 @findex find_sym_fns
2119 Symbol files are initially opened by code in @file{symfile.c} using
2120 the BFD library (@pxref{Support Libraries}). BFD identifies the type
2121 of the file by examining its header. @code{find_sym_fns} then uses
2122 this identification to locate a set of symbol-reading functions.
2124 @findex add_symtab_fns
2125 @cindex @code{sym_fns} structure
2126 @cindex adding a symbol-reading module
2127 Symbol-reading modules identify themselves to @value{GDBN} by calling
2128 @code{add_symtab_fns} during their module initialization. The argument
2129 to @code{add_symtab_fns} is a @code{struct sym_fns} which contains the
2130 name (or name prefix) of the symbol format, the length of the prefix,
2131 and pointers to four functions. These functions are called at various
2132 times to process symbol files whose identification matches the specified
2135 The functions supplied by each module are:
2138 @item @var{xyz}_symfile_init(struct sym_fns *sf)
2140 @cindex secondary symbol file
2141 Called from @code{symbol_file_add} when we are about to read a new
2142 symbol file. This function should clean up any internal state (possibly
2143 resulting from half-read previous files, for example) and prepare to
2144 read a new symbol file. Note that the symbol file which we are reading
2145 might be a new ``main'' symbol file, or might be a secondary symbol file
2146 whose symbols are being added to the existing symbol table.
2148 The argument to @code{@var{xyz}_symfile_init} is a newly allocated
2149 @code{struct sym_fns} whose @code{bfd} field contains the BFD for the
2150 new symbol file being read. Its @code{private} field has been zeroed,
2151 and can be modified as desired. Typically, a struct of private
2152 information will be @code{malloc}'d, and a pointer to it will be placed
2153 in the @code{private} field.
2155 There is no result from @code{@var{xyz}_symfile_init}, but it can call
2156 @code{error} if it detects an unavoidable problem.
2158 @item @var{xyz}_new_init()
2160 Called from @code{symbol_file_add} when discarding existing symbols.
2161 This function needs only handle the symbol-reading module's internal
2162 state; the symbol table data structures visible to the rest of
2163 @value{GDBN} will be discarded by @code{symbol_file_add}. It has no
2164 arguments and no result. It may be called after
2165 @code{@var{xyz}_symfile_init}, if a new symbol table is being read, or
2166 may be called alone if all symbols are simply being discarded.
2168 @item @var{xyz}_symfile_read(struct sym_fns *sf, CORE_ADDR addr, int mainline)
2170 Called from @code{symbol_file_add} to actually read the symbols from a
2171 symbol-file into a set of psymtabs or symtabs.
2173 @code{sf} points to the @code{struct sym_fns} originally passed to
2174 @code{@var{xyz}_sym_init} for possible initialization. @code{addr} is
2175 the offset between the file's specified start address and its true
2176 address in memory. @code{mainline} is 1 if this is the main symbol
2177 table being read, and 0 if a secondary symbol file (e.g., shared library
2178 or dynamically loaded file) is being read.@refill
2181 In addition, if a symbol-reading module creates psymtabs when
2182 @var{xyz}_symfile_read is called, these psymtabs will contain a pointer
2183 to a function @code{@var{xyz}_psymtab_to_symtab}, which can be called
2184 from any point in the @value{GDBN} symbol-handling code.
2187 @item @var{xyz}_psymtab_to_symtab (struct partial_symtab *pst)
2189 Called from @code{psymtab_to_symtab} (or the @code{PSYMTAB_TO_SYMTAB} macro) if
2190 the psymtab has not already been read in and had its @code{pst->symtab}
2191 pointer set. The argument is the psymtab to be fleshed-out into a
2192 symtab. Upon return, @code{pst->readin} should have been set to 1, and
2193 @code{pst->symtab} should contain a pointer to the new corresponding symtab, or
2194 zero if there were no symbols in that part of the symbol file.
2197 @section Partial Symbol Tables
2199 @value{GDBN} has three types of symbol tables:
2202 @cindex full symbol table
2205 Full symbol tables (@dfn{symtabs}). These contain the main
2206 information about symbols and addresses.
2210 Partial symbol tables (@dfn{psymtabs}). These contain enough
2211 information to know when to read the corresponding part of the full
2214 @cindex minimal symbol table
2217 Minimal symbol tables (@dfn{msymtabs}). These contain information
2218 gleaned from non-debugging symbols.
2221 @cindex partial symbol table
2222 This section describes partial symbol tables.
2224 A psymtab is constructed by doing a very quick pass over an executable
2225 file's debugging information. Small amounts of information are
2226 extracted---enough to identify which parts of the symbol table will
2227 need to be re-read and fully digested later, when the user needs the
2228 information. The speed of this pass causes @value{GDBN} to start up very
2229 quickly. Later, as the detailed rereading occurs, it occurs in small
2230 pieces, at various times, and the delay therefrom is mostly invisible to
2232 @c (@xref{Symbol Reading}.)
2234 The symbols that show up in a file's psymtab should be, roughly, those
2235 visible to the debugger's user when the program is not running code from
2236 that file. These include external symbols and types, static symbols and
2237 types, and @code{enum} values declared at file scope.
2239 The psymtab also contains the range of instruction addresses that the
2240 full symbol table would represent.
2242 @cindex finding a symbol
2243 @cindex symbol lookup
2244 The idea is that there are only two ways for the user (or much of the
2245 code in the debugger) to reference a symbol:
2248 @findex find_pc_function
2249 @findex find_pc_line
2251 By its address (e.g., execution stops at some address which is inside a
2252 function in this file). The address will be noticed to be in the
2253 range of this psymtab, and the full symtab will be read in.
2254 @code{find_pc_function}, @code{find_pc_line}, and other
2255 @code{find_pc_@dots{}} functions handle this.
2257 @cindex lookup_symbol
2260 (e.g., the user asks to print a variable, or set a breakpoint on a
2261 function). Global names and file-scope names will be found in the
2262 psymtab, which will cause the symtab to be pulled in. Local names will
2263 have to be qualified by a global name, or a file-scope name, in which
2264 case we will have already read in the symtab as we evaluated the
2265 qualifier. Or, a local symbol can be referenced when we are ``in'' a
2266 local scope, in which case the first case applies. @code{lookup_symbol}
2267 does most of the work here.
2270 The only reason that psymtabs exist is to cause a symtab to be read in
2271 at the right moment. Any symbol that can be elided from a psymtab,
2272 while still causing that to happen, should not appear in it. Since
2273 psymtabs don't have the idea of scope, you can't put local symbols in
2274 them anyway. Psymtabs don't have the idea of the type of a symbol,
2275 either, so types need not appear, unless they will be referenced by
2278 It is a bug for @value{GDBN} to behave one way when only a psymtab has
2279 been read, and another way if the corresponding symtab has been read
2280 in. Such bugs are typically caused by a psymtab that does not contain
2281 all the visible symbols, or which has the wrong instruction address
2284 The psymtab for a particular section of a symbol file (objfile) could be
2285 thrown away after the symtab has been read in. The symtab should always
2286 be searched before the psymtab, so the psymtab will never be used (in a
2287 bug-free environment). Currently, psymtabs are allocated on an obstack,
2288 and all the psymbols themselves are allocated in a pair of large arrays
2289 on an obstack, so there is little to be gained by trying to free them
2290 unless you want to do a lot more work.
2292 Whether or not psymtabs are created depends on the objfile's symbol
2293 reader. The core of @value{GDBN} hides the details of partial symbols
2294 and partial symbol tables behind a set of function pointers known as
2295 the @dfn{quick symbol functions}. These are documented in
2300 @unnumberedsubsec Fundamental Types (e.g., @code{FT_VOID}, @code{FT_BOOLEAN}).
2302 @cindex fundamental types
2303 These are the fundamental types that @value{GDBN} uses internally. Fundamental
2304 types from the various debugging formats (stabs, ELF, etc) are mapped
2305 into one of these. They are basically a union of all fundamental types
2306 that @value{GDBN} knows about for all the languages that @value{GDBN}
2309 @unnumberedsubsec Type Codes (e.g., @code{TYPE_CODE_PTR}, @code{TYPE_CODE_ARRAY}).
2312 Each time @value{GDBN} builds an internal type, it marks it with one
2313 of these types. The type may be a fundamental type, such as
2314 @code{TYPE_CODE_INT}, or a derived type, such as @code{TYPE_CODE_PTR}
2315 which is a pointer to another type. Typically, several @code{FT_*}
2316 types map to one @code{TYPE_CODE_*} type, and are distinguished by
2317 other members of the type struct, such as whether the type is signed
2318 or unsigned, and how many bits it uses.
2320 @unnumberedsubsec Builtin Types (e.g., @code{builtin_type_void}, @code{builtin_type_char}).
2322 These are instances of type structs that roughly correspond to
2323 fundamental types and are created as global types for @value{GDBN} to
2324 use for various ugly historical reasons. We eventually want to
2325 eliminate these. Note for example that @code{builtin_type_int}
2326 initialized in @file{gdbtypes.c} is basically the same as a
2327 @code{TYPE_CODE_INT} type that is initialized in @file{c-lang.c} for
2328 an @code{FT_INTEGER} fundamental type. The difference is that the
2329 @code{builtin_type} is not associated with any particular objfile, and
2330 only one instance exists, while @file{c-lang.c} builds as many
2331 @code{TYPE_CODE_INT} types as needed, with each one associated with
2332 some particular objfile.
2334 @section Object File Formats
2335 @cindex object file formats
2339 @cindex @code{a.out} format
2340 The @code{a.out} format is the original file format for Unix. It
2341 consists of three sections: @code{text}, @code{data}, and @code{bss},
2342 which are for program code, initialized data, and uninitialized data,
2345 The @code{a.out} format is so simple that it doesn't have any reserved
2346 place for debugging information. (Hey, the original Unix hackers used
2347 @samp{adb}, which is a machine-language debugger!) The only debugging
2348 format for @code{a.out} is stabs, which is encoded as a set of normal
2349 symbols with distinctive attributes.
2351 The basic @code{a.out} reader is in @file{dbxread.c}.
2356 The COFF format was introduced with System V Release 3 (SVR3) Unix.
2357 COFF files may have multiple sections, each prefixed by a header. The
2358 number of sections is limited.
2360 The COFF specification includes support for debugging. Although this
2361 was a step forward, the debugging information was woefully limited.
2362 For instance, it was not possible to represent code that came from an
2363 included file. GNU's COFF-using configs often use stabs-type info,
2364 encapsulated in special sections.
2366 The COFF reader is in @file{coffread.c}.
2370 @cindex ECOFF format
2371 ECOFF is an extended COFF originally introduced for Mips and Alpha
2374 The basic ECOFF reader is in @file{mipsread.c}.
2378 @cindex XCOFF format
2379 The IBM RS/6000 running AIX uses an object file format called XCOFF.
2380 The COFF sections, symbols, and line numbers are used, but debugging
2381 symbols are @code{dbx}-style stabs whose strings are located in the
2382 @code{.debug} section (rather than the string table). For more
2383 information, see @ref{Top,,,stabs,The Stabs Debugging Format}.
2385 The shared library scheme has a clean interface for figuring out what
2386 shared libraries are in use, but the catch is that everything which
2387 refers to addresses (symbol tables and breakpoints at least) needs to be
2388 relocated for both shared libraries and the main executable. At least
2389 using the standard mechanism this can only be done once the program has
2390 been run (or the core file has been read).
2394 @cindex PE-COFF format
2395 Windows 95 and NT use the PE (@dfn{Portable Executable}) format for their
2396 executables. PE is basically COFF with additional headers.
2398 While BFD includes special PE support, @value{GDBN} needs only the basic
2404 The ELF format came with System V Release 4 (SVR4) Unix. ELF is
2405 similar to COFF in being organized into a number of sections, but it
2406 removes many of COFF's limitations. Debugging info may be either stabs
2407 encapsulated in ELF sections, or more commonly these days, DWARF.
2409 The basic ELF reader is in @file{elfread.c}.
2414 SOM is HP's object file and debug format (not to be confused with IBM's
2415 SOM, which is a cross-language ABI).
2417 The SOM reader is in @file{somread.c}.
2419 @section Debugging File Formats
2421 This section describes characteristics of debugging information that
2422 are independent of the object file format.
2426 @cindex stabs debugging info
2427 @code{stabs} started out as special symbols within the @code{a.out}
2428 format. Since then, it has been encapsulated into other file
2429 formats, such as COFF and ELF.
2431 While @file{dbxread.c} does some of the basic stab processing,
2432 including for encapsulated versions, @file{stabsread.c} does
2437 @cindex COFF debugging info
2438 The basic COFF definition includes debugging information. The level
2439 of support is minimal and non-extensible, and is not often used.
2441 @subsection Mips debug (Third Eye)
2443 @cindex ECOFF debugging info
2444 ECOFF includes a definition of a special debug format.
2446 The file @file{mdebugread.c} implements reading for this format.
2448 @c mention DWARF 1 as a formerly-supported format
2452 @cindex DWARF 2 debugging info
2453 DWARF 2 is an improved but incompatible version of DWARF 1.
2455 The DWARF 2 reader is in @file{dwarf2read.c}.
2457 @subsection Compressed DWARF 2
2459 @cindex Compressed DWARF 2 debugging info
2460 Compressed DWARF 2 is not technically a separate debugging format, but
2461 merely DWARF 2 debug information that has been compressed. In this
2462 format, every object-file section holding DWARF 2 debugging
2463 information is compressed and prepended with a header. (The section
2464 is also typically renamed, so a section called @code{.debug_info} in a
2465 DWARF 2 binary would be called @code{.zdebug_info} in a compressed
2466 DWARF 2 binary.) The header is 12 bytes long:
2470 4 bytes: the literal string ``ZLIB''
2472 8 bytes: the uncompressed size of the section, in big-endian byte
2476 The same reader is used for both compressed an normal DWARF 2 info.
2477 Section decompression is done in @code{zlib_decompress_section} in
2478 @file{dwarf2read.c}.
2482 @cindex DWARF 3 debugging info
2483 DWARF 3 is an improved version of DWARF 2.
2487 @cindex SOM debugging info
2488 Like COFF, the SOM definition includes debugging information.
2490 @section Adding a New Symbol Reader to @value{GDBN}
2492 @cindex adding debugging info reader
2493 If you are using an existing object file format (@code{a.out}, COFF, ELF, etc),
2494 there is probably little to be done.
2496 If you need to add a new object file format, you must first add it to
2497 BFD. This is beyond the scope of this document.
2499 You must then arrange for the BFD code to provide access to the
2500 debugging symbols. Generally @value{GDBN} will have to call swapping
2501 routines from BFD and a few other BFD internal routines to locate the
2502 debugging information. As much as possible, @value{GDBN} should not
2503 depend on the BFD internal data structures.
2505 For some targets (e.g., COFF), there is a special transfer vector used
2506 to call swapping routines, since the external data structures on various
2507 platforms have different sizes and layouts. Specialized routines that
2508 will only ever be implemented by one object file format may be called
2509 directly. This interface should be described in a file
2510 @file{bfd/lib@var{xyz}.h}, which is included by @value{GDBN}.
2512 @section Memory Management for Symbol Files
2514 Most memory associated with a loaded symbol file is stored on
2515 its @code{objfile_obstack}. This includes symbols, types,
2516 namespace data, and other information produced by the symbol readers.
2518 Because this data lives on the objfile's obstack, it is automatically
2519 released when the objfile is unloaded or reloaded. Therefore one
2520 objfile must not reference symbol or type data from another objfile;
2521 they could be unloaded at different times.
2523 User convenience variables, et cetera, have associated types. Normally
2524 these types live in the associated objfile. However, when the objfile
2525 is unloaded, those types are deep copied to global memory, so that
2526 the values of the user variables and history items are not lost.
2529 @node Language Support
2531 @chapter Language Support
2533 @cindex language support
2534 @value{GDBN}'s language support is mainly driven by the symbol reader,
2535 although it is possible for the user to set the source language
2538 @value{GDBN} chooses the source language by looking at the extension
2539 of the file recorded in the debug info; @file{.c} means C, @file{.f}
2540 means Fortran, etc. It may also use a special-purpose language
2541 identifier if the debug format supports it, like with DWARF.
2543 @section Adding a Source Language to @value{GDBN}
2545 @cindex adding source language
2546 To add other languages to @value{GDBN}'s expression parser, follow the
2550 @item Create the expression parser.
2552 @cindex expression parser
2553 This should reside in a file @file{@var{lang}-exp.y}. Routines for
2554 building parsed expressions into a @code{union exp_element} list are in
2557 @cindex language parser
2558 Since we can't depend upon everyone having Bison, and YACC produces
2559 parsers that define a bunch of global names, the following lines
2560 @strong{must} be included at the top of the YACC parser, to prevent the
2561 various parsers from defining the same global names:
2564 #define yyparse @var{lang}_parse
2565 #define yylex @var{lang}_lex
2566 #define yyerror @var{lang}_error
2567 #define yylval @var{lang}_lval
2568 #define yychar @var{lang}_char
2569 #define yydebug @var{lang}_debug
2570 #define yypact @var{lang}_pact
2571 #define yyr1 @var{lang}_r1
2572 #define yyr2 @var{lang}_r2
2573 #define yydef @var{lang}_def
2574 #define yychk @var{lang}_chk
2575 #define yypgo @var{lang}_pgo
2576 #define yyact @var{lang}_act
2577 #define yyexca @var{lang}_exca
2578 #define yyerrflag @var{lang}_errflag
2579 #define yynerrs @var{lang}_nerrs
2582 At the bottom of your parser, define a @code{struct language_defn} and
2583 initialize it with the right values for your language. Define an
2584 @code{initialize_@var{lang}} routine and have it call
2585 @samp{add_language(@var{lang}_language_defn)} to tell the rest of @value{GDBN}
2586 that your language exists. You'll need some other supporting variables
2587 and functions, which will be used via pointers from your
2588 @code{@var{lang}_language_defn}. See the declaration of @code{struct
2589 language_defn} in @file{language.h}, and the other @file{*-exp.y} files,
2590 for more information.
2592 @item Add any evaluation routines, if necessary
2594 @cindex expression evaluation routines
2595 @findex evaluate_subexp
2596 @findex prefixify_subexp
2597 @findex length_of_subexp
2598 If you need new opcodes (that represent the operations of the language),
2599 add them to the enumerated type in @file{expression.h}. Add support
2600 code for these operations in the @code{evaluate_subexp} function
2601 defined in the file @file{eval.c}. Add cases
2602 for new opcodes in two functions from @file{parse.c}:
2603 @code{prefixify_subexp} and @code{length_of_subexp}. These compute
2604 the number of @code{exp_element}s that a given operation takes up.
2606 @item Update some existing code
2608 Add an enumerated identifier for your language to the enumerated type
2609 @code{enum language} in @file{defs.h}.
2611 Update the routines in @file{language.c} so your language is included.
2612 These routines include type predicates and such, which (in some cases)
2613 are language dependent. If your language does not appear in the switch
2614 statement, an error is reported.
2616 @vindex current_language
2617 Also included in @file{language.c} is the code that updates the variable
2618 @code{current_language}, and the routines that translate the
2619 @code{language_@var{lang}} enumerated identifier into a printable
2622 @findex _initialize_language
2623 Update the function @code{_initialize_language} to include your
2624 language. This function picks the default language upon startup, so is
2625 dependent upon which languages that @value{GDBN} is built for.
2627 @findex allocate_symtab
2628 Update @code{allocate_symtab} in @file{symfile.c} and/or symbol-reading
2629 code so that the language of each symtab (source file) is set properly.
2630 This is used to determine the language to use at each stack frame level.
2631 Currently, the language is set based upon the extension of the source
2632 file. If the language can be better inferred from the symbol
2633 information, please set the language of the symtab in the symbol-reading
2636 @findex print_subexp
2637 @findex op_print_tab
2638 Add helper code to @code{print_subexp} (in @file{expprint.c}) to handle any new
2639 expression opcodes you have added to @file{expression.h}. Also, add the
2640 printed representations of your operators to @code{op_print_tab}.
2642 @item Add a place of call
2645 Add a call to @code{@var{lang}_parse()} and @code{@var{lang}_error} in
2646 @code{parse_exp_1} (defined in @file{parse.c}).
2648 @item Edit @file{Makefile.in}
2650 Add dependencies in @file{Makefile.in}. Make sure you update the macro
2651 variables such as @code{HFILES} and @code{OBJS}, otherwise your code may
2652 not get linked in, or, worse yet, it may not get @code{tar}red into the
2657 @node Host Definition
2659 @chapter Host Definition
2661 With the advent of Autoconf, it's rarely necessary to have host
2662 definition machinery anymore. The following information is provided,
2663 mainly, as an historical reference.
2665 @section Adding a New Host
2667 @cindex adding a new host
2668 @cindex host, adding
2669 @value{GDBN}'s host configuration support normally happens via Autoconf.
2670 New host-specific definitions should not be needed. Older hosts
2671 @value{GDBN} still use the host-specific definitions and files listed
2672 below, but these mostly exist for historical reasons, and will
2673 eventually disappear.
2676 @item gdb/config/@var{arch}/@var{xyz}.mh
2677 This file is a Makefile fragment that once contained both host and
2678 native configuration information (@pxref{Native Debugging}) for the
2679 machine @var{xyz}. The host configuration information is now handled
2682 Host configuration information included definitions for @code{CC},
2683 @code{SYSV_DEFINE}, @code{XM_CFLAGS}, @code{XM_ADD_FILES},
2684 @code{XM_CLIBS}, @code{XM_CDEPS}, etc.; see @file{Makefile.in}.
2686 New host-only configurations do not need this file.
2690 (Files named @file{gdb/config/@var{arch}/xm-@var{xyz}.h} were once
2691 used to define host-specific macros, but were no longer needed and
2692 have all been removed.)
2694 @subheading Generic Host Support Files
2696 @cindex generic host support
2697 There are some ``generic'' versions of routines that can be used by
2701 @cindex remote debugging support
2702 @cindex serial line support
2704 This contains serial line support for Unix systems. It is included by
2705 default on all Unix-like hosts.
2708 This contains serial pipe support for Unix systems. It is included by
2709 default on all Unix-like hosts.
2712 This contains serial line support for 32-bit programs running under
2713 Windows using MinGW.
2716 This contains serial line support for 32-bit programs running under DOS,
2717 using the DJGPP (a.k.a.@: GO32) execution environment.
2719 @cindex TCP remote support
2721 This contains generic TCP support using sockets. It is included by
2722 default on all Unix-like hosts and with MinGW.
2725 @section Host Conditionals
2727 When @value{GDBN} is configured and compiled, various macros are
2728 defined or left undefined, to control compilation based on the
2729 attributes of the host system. While formerly they could be set in
2730 host-specific header files, at present they can be changed only by
2731 setting @code{CFLAGS} when building, or by editing the source code.
2733 These macros and their meanings (or if the meaning is not documented
2734 here, then one of the source files where they are used is indicated)
2738 @item @value{GDBN}INIT_FILENAME
2739 The default name of @value{GDBN}'s initialization file (normally
2742 @item SIGWINCH_HANDLER
2743 If your host defines @code{SIGWINCH}, you can define this to be the name
2744 of a function to be called if @code{SIGWINCH} is received.
2746 @item SIGWINCH_HANDLER_BODY
2747 Define this to expand into code that will define the function named by
2748 the expansion of @code{SIGWINCH_HANDLER}.
2750 @item CRLF_SOURCE_FILES
2751 @cindex DOS text files
2752 Define this if host files use @code{\r\n} rather than @code{\n} as a
2753 line terminator. This will cause source file listings to omit @code{\r}
2754 characters when printing and it will allow @code{\r\n} line endings of files
2755 which are ``sourced'' by gdb. It must be possible to open files in binary
2756 mode using @code{O_BINARY} or, for fopen, @code{"rb"}.
2758 @item DEFAULT_PROMPT
2760 The default value of the prompt string (normally @code{"(gdb) "}).
2763 @cindex terminal device
2764 The name of the generic TTY device, defaults to @code{"/dev/tty"}.
2767 Substitute for isatty, if not available.
2770 Define this if binary files are opened the same way as text files.
2772 @item CC_HAS_LONG_LONG
2773 @cindex @code{long long} data type
2774 Define this if the host C compiler supports @code{long long}. This is set
2775 by the @code{configure} script.
2777 @item PRINTF_HAS_LONG_LONG
2778 Define this if the host can handle printing of long long integers via
2779 the printf format conversion specifier @code{ll}. This is set by the
2780 @code{configure} script.
2782 @item LSEEK_NOT_LINEAR
2783 Define this if @code{lseek (n)} does not necessarily move to byte number
2784 @code{n} in the file. This is only used when reading source files. It
2785 is normally faster to define @code{CRLF_SOURCE_FILES} when possible.
2788 Define this to help placate @code{lint} in some situations.
2791 Define this to override the defaults of @code{__volatile__} or
2796 @node Target Architecture Definition
2798 @chapter Target Architecture Definition
2800 @cindex target architecture definition
2801 @value{GDBN}'s target architecture defines what sort of
2802 machine-language programs @value{GDBN} can work with, and how it works
2805 The target architecture object is implemented as the C structure
2806 @code{struct gdbarch *}. The structure, and its methods, are generated
2807 using the Bourne shell script @file{gdbarch.sh}.
2810 * OS ABI Variant Handling::
2811 * Initialize New Architecture::
2812 * Registers and Memory::
2813 * Pointers and Addresses::
2815 * Register Representation::
2816 * Frame Interpretation::
2817 * Inferior Call Setup::
2818 * Adding support for debugging core files::
2819 * Defining Other Architecture Features::
2820 * Adding a New Target::
2823 @node OS ABI Variant Handling
2824 @section Operating System ABI Variant Handling
2825 @cindex OS ABI variants
2827 @value{GDBN} provides a mechanism for handling variations in OS
2828 ABIs. An OS ABI variant may have influence over any number of
2829 variables in the target architecture definition. There are two major
2830 components in the OS ABI mechanism: sniffers and handlers.
2832 A @dfn{sniffer} examines a file matching a BFD architecture/flavour pair
2833 (the architecture may be wildcarded) in an attempt to determine the
2834 OS ABI of that file. Sniffers with a wildcarded architecture are considered
2835 to be @dfn{generic}, while sniffers for a specific architecture are
2836 considered to be @dfn{specific}. A match from a specific sniffer
2837 overrides a match from a generic sniffer. Multiple sniffers for an
2838 architecture/flavour may exist, in order to differentiate between two
2839 different operating systems which use the same basic file format. The
2840 OS ABI framework provides a generic sniffer for ELF-format files which
2841 examines the @code{EI_OSABI} field of the ELF header, as well as note
2842 sections known to be used by several operating systems.
2844 @cindex fine-tuning @code{gdbarch} structure
2845 A @dfn{handler} is used to fine-tune the @code{gdbarch} structure for the
2846 selected OS ABI. There may be only one handler for a given OS ABI
2847 for each BFD architecture.
2849 The following OS ABI variants are defined in @file{defs.h}:
2853 @findex GDB_OSABI_UNINITIALIZED
2854 @item GDB_OSABI_UNINITIALIZED
2855 Used for struct gdbarch_info if ABI is still uninitialized.
2857 @findex GDB_OSABI_UNKNOWN
2858 @item GDB_OSABI_UNKNOWN
2859 The ABI of the inferior is unknown. The default @code{gdbarch}
2860 settings for the architecture will be used.
2862 @findex GDB_OSABI_SVR4
2863 @item GDB_OSABI_SVR4
2864 UNIX System V Release 4.
2866 @findex GDB_OSABI_HURD
2867 @item GDB_OSABI_HURD
2868 GNU using the Hurd kernel.
2870 @findex GDB_OSABI_SOLARIS
2871 @item GDB_OSABI_SOLARIS
2874 @findex GDB_OSABI_OSF1
2875 @item GDB_OSABI_OSF1
2876 OSF/1, including Digital UNIX and Compaq Tru64 UNIX.
2878 @findex GDB_OSABI_LINUX
2879 @item GDB_OSABI_LINUX
2880 GNU using the Linux kernel.
2882 @findex GDB_OSABI_FREEBSD_AOUT
2883 @item GDB_OSABI_FREEBSD_AOUT
2884 FreeBSD using the @code{a.out} executable format.
2886 @findex GDB_OSABI_FREEBSD_ELF
2887 @item GDB_OSABI_FREEBSD_ELF
2888 FreeBSD using the ELF executable format.
2890 @findex GDB_OSABI_NETBSD_AOUT
2891 @item GDB_OSABI_NETBSD_AOUT
2892 NetBSD using the @code{a.out} executable format.
2894 @findex GDB_OSABI_NETBSD_ELF
2895 @item GDB_OSABI_NETBSD_ELF
2896 NetBSD using the ELF executable format.
2898 @findex GDB_OSABI_OPENBSD_ELF
2899 @item GDB_OSABI_OPENBSD_ELF
2900 OpenBSD using the ELF executable format.
2902 @findex GDB_OSABI_WINCE
2903 @item GDB_OSABI_WINCE
2906 @findex GDB_OSABI_GO32
2907 @item GDB_OSABI_GO32
2910 @findex GDB_OSABI_IRIX
2911 @item GDB_OSABI_IRIX
2914 @findex GDB_OSABI_INTERIX
2915 @item GDB_OSABI_INTERIX
2916 Interix (Posix layer for MS-Windows systems).
2918 @findex GDB_OSABI_HPUX_ELF
2919 @item GDB_OSABI_HPUX_ELF
2920 HP/UX using the ELF executable format.
2922 @findex GDB_OSABI_HPUX_SOM
2923 @item GDB_OSABI_HPUX_SOM
2924 HP/UX using the SOM executable format.
2926 @findex GDB_OSABI_QNXNTO
2927 @item GDB_OSABI_QNXNTO
2930 @findex GDB_OSABI_CYGWIN
2931 @item GDB_OSABI_CYGWIN
2934 @findex GDB_OSABI_AIX
2940 Here are the functions that make up the OS ABI framework:
2942 @deftypefun {const char *} gdbarch_osabi_name (enum gdb_osabi @var{osabi})
2943 Return the name of the OS ABI corresponding to @var{osabi}.
2946 @deftypefun void gdbarch_register_osabi (enum bfd_architecture @var{arch}, unsigned long @var{machine}, enum gdb_osabi @var{osabi}, void (*@var{init_osabi})(struct gdbarch_info @var{info}, struct gdbarch *@var{gdbarch}))
2947 Register the OS ABI handler specified by @var{init_osabi} for the
2948 architecture, machine type and OS ABI specified by @var{arch},
2949 @var{machine} and @var{osabi}. In most cases, a value of zero for the
2950 machine type, which implies the architecture's default machine type,
2954 @deftypefun void gdbarch_register_osabi_sniffer (enum bfd_architecture @var{arch}, enum bfd_flavour @var{flavour}, enum gdb_osabi (*@var{sniffer})(bfd *@var{abfd}))
2955 Register the OS ABI file sniffer specified by @var{sniffer} for the
2956 BFD architecture/flavour pair specified by @var{arch} and @var{flavour}.
2957 If @var{arch} is @code{bfd_arch_unknown}, the sniffer is considered to
2958 be generic, and is allowed to examine @var{flavour}-flavoured files for
2962 @deftypefun {enum gdb_osabi} gdbarch_lookup_osabi (bfd *@var{abfd})
2963 Examine the file described by @var{abfd} to determine its OS ABI.
2964 The value @code{GDB_OSABI_UNKNOWN} is returned if the OS ABI cannot
2968 @deftypefun void gdbarch_init_osabi (struct gdbarch info @var{info}, struct gdbarch *@var{gdbarch}, enum gdb_osabi @var{osabi})
2969 Invoke the OS ABI handler corresponding to @var{osabi} to fine-tune the
2970 @code{gdbarch} structure specified by @var{gdbarch}. If a handler
2971 corresponding to @var{osabi} has not been registered for @var{gdbarch}'s
2972 architecture, a warning will be issued and the debugging session will continue
2973 with the defaults already established for @var{gdbarch}.
2976 @deftypefun void generic_elf_osabi_sniff_abi_tag_sections (bfd *@var{abfd}, asection *@var{sect}, void *@var{obj})
2977 Helper routine for ELF file sniffers. Examine the file described by
2978 @var{abfd} and look at ABI tag note sections to determine the OS ABI
2979 from the note. This function should be called via
2980 @code{bfd_map_over_sections}.
2983 @node Initialize New Architecture
2984 @section Initializing a New Architecture
2987 * How an Architecture is Represented::
2988 * Looking Up an Existing Architecture::
2989 * Creating a New Architecture::
2992 @node How an Architecture is Represented
2993 @subsection How an Architecture is Represented
2994 @cindex architecture representation
2995 @cindex representation of architecture
2997 Each @code{gdbarch} is associated with a single @sc{bfd} architecture,
2998 via a @code{bfd_arch_@var{arch}} in the @code{bfd_architecture}
2999 enumeration. The @code{gdbarch} is registered by a call to
3000 @code{register_gdbarch_init}, usually from the file's
3001 @code{_initialize_@var{filename}} routine, which will be automatically
3002 called during @value{GDBN} startup. The arguments are a @sc{bfd}
3003 architecture constant and an initialization function.
3005 @findex _initialize_@var{arch}_tdep
3006 @cindex @file{@var{arch}-tdep.c}
3007 A @value{GDBN} description for a new architecture, @var{arch} is created by
3008 defining a global function @code{_initialize_@var{arch}_tdep}, by
3009 convention in the source file @file{@var{arch}-tdep.c}. For example,
3010 in the case of the OpenRISC 1000, this function is called
3011 @code{_initialize_or1k_tdep} and is found in the file
3014 @cindex @file{configure.tgt}
3015 @cindex @code{gdbarch}
3016 @findex gdbarch_register
3017 The resulting object files containing the implementation of the
3018 @code{_initialize_@var{arch}_tdep} function are specified in the @value{GDBN}
3019 @file{configure.tgt} file, which includes a large case statement
3020 pattern matching against the @code{--target} option of the
3021 @code{configure} script. The new @code{struct gdbarch} is created
3022 within the @code{_initialize_@var{arch}_tdep} function by calling
3023 @code{gdbarch_register}:
3026 void gdbarch_register (enum bfd_architecture @var{architecture},
3027 gdbarch_init_ftype *@var{init_func},
3028 gdbarch_dump_tdep_ftype *@var{tdep_dump_func});
3031 The @var{architecture} will identify the unique @sc{bfd} to be
3032 associated with this @code{gdbarch}. The @var{init_func} funciton is
3033 called to create and return the new @code{struct gdbarch}. The
3034 @var{tdep_dump_func} function will dump the target specific details
3035 associated with this architecture.
3037 For example the function @code{_initialize_or1k_tdep} creates its
3038 architecture for 32-bit OpenRISC 1000 architectures by calling:
3041 gdbarch_register (bfd_arch_or32, or1k_gdbarch_init, or1k_dump_tdep);
3044 @node Looking Up an Existing Architecture
3045 @subsection Looking Up an Existing Architecture
3046 @cindex @code{gdbarch} lookup
3048 The initialization function has this prototype:
3051 static struct gdbarch *
3052 @var{arch}_gdbarch_init (struct gdbarch_info @var{info},
3053 struct gdbarch_list *@var{arches})
3056 The @var{info} argument contains parameters used to select the correct
3057 architecture, and @var{arches} is a list of architectures which
3058 have already been created with the same @code{bfd_arch_@var{arch}}
3061 The initialization function should first make sure that @var{info}
3062 is acceptable, and return @code{NULL} if it is not. Then, it should
3063 search through @var{arches} for an exact match to @var{info}, and
3064 return one if found. Lastly, if no exact match was found, it should
3065 create a new architecture based on @var{info} and return it.
3067 @findex gdbarch_list_lookup_by_info
3068 @cindex @code{gdbarch_info}
3069 The lookup is done using @code{gdbarch_list_lookup_by_info}. It is
3070 passed the list of existing architectures, @var{arches}, and the
3071 @code{struct gdbarch_info}, @var{info}, and returns the first matching
3072 architecture it finds, or @code{NULL} if none are found. If an
3073 architecture is found it can be returned as the result from the
3074 initialization function, otherwise a new @code{struct gdbach} will need
3077 The struct gdbarch_info has the following components:
3082 const struct bfd_arch_info *bfd_arch_info;
3085 struct gdbarch_tdep_info *tdep_info;
3086 enum gdb_osabi osabi;
3087 const struct target_desc *target_desc;
3091 @vindex bfd_arch_info
3092 The @code{bfd_arch_info} member holds the key details about the
3093 architecture. The @code{byte_order} member is a value in an
3094 enumeration indicating the endianism. The @code{abfd} member is a
3095 pointer to the full @sc{bfd}, the @code{tdep_info} member is
3096 additional custom target specific information, @code{osabi} identifies
3097 which (if any) of a number of operating specific ABIs are used by this
3098 architecture and the @code{target_desc} member is a set of name-value
3099 pairs with information about register usage in this target.
3101 When the @code{struct gdbarch} initialization function is called, not
3102 all the fields are provided---only those which can be deduced from the
3103 @sc{bfd}. The @code{struct gdbarch_info}, @var{info} is used as a
3104 look-up key with the list of existing architectures, @var{arches} to
3105 see if a suitable architecture already exists. The @var{tdep_info},
3106 @var{osabi} and @var{target_desc} fields may be added before this
3107 lookup to refine the search.
3109 Only information in @var{info} should be used to choose the new
3110 architecture. Historically, @var{info} could be sparse, and
3111 defaults would be collected from the first element on @var{arches}.
3112 However, @value{GDBN} now fills in @var{info} more thoroughly,
3113 so new @code{gdbarch} initialization functions should not take
3114 defaults from @var{arches}.
3116 @node Creating a New Architecture
3117 @subsection Creating a New Architecture
3118 @cindex @code{struct gdbarch} creation
3120 @findex gdbarch_alloc
3121 @cindex @code{gdbarch_tdep} when allocating new @code{gdbarch}
3122 If no architecture is found, then a new architecture must be created,
3123 by calling @code{gdbarch_alloc} using the supplied @code{@w{struct
3124 gdbarch_info}} and any additional custom target specific
3125 information in a @code{struct gdbarch_tdep}. The prototype for
3126 @code{gdbarch_alloc} is:
3129 struct gdbarch *gdbarch_alloc (const struct gdbarch_info *@var{info},
3130 struct gdbarch_tdep *@var{tdep});
3133 @cindex @code{set_gdbarch} functions
3134 @cindex @code{gdbarch} accessor functions
3135 The newly created struct gdbarch must then be populated. Although
3136 there are default values, in most cases they are not what is
3139 For each element, @var{X}, there is are a pair of corresponding accessor
3140 functions, one to set the value of that element,
3141 @code{set_gdbarch_@var{X}}, the second to either get the value of an
3142 element (if it is a variable) or to apply the element (if it is a
3143 function), @code{gdbarch_@var{X}}. Note that both accessor functions
3144 take a pointer to the @code{@w{struct gdbarch}} as first
3145 argument. Populating the new @code{gdbarch} should use the
3146 @code{set_gdbarch} functions.
3148 The following sections identify the main elements that should be set
3149 in this way. This is not the complete list, but represents the
3150 functions and elements that must commonly be specified for a new
3151 architecture. Many of the functions and variables are described in the
3152 header file @file{gdbarch.h}.
3154 This is the main work in defining a new architecture. Implementing the
3155 set of functions to populate the @code{struct gdbarch}.
3157 @cindex @code{gdbarch_tdep} definition
3158 @code{struct gdbarch_tdep} is not defined within @value{GDBN}---it is up
3159 to the user to define this struct if it is needed to hold custom target
3160 information that is not covered by the standard @code{@w{struct
3161 gdbarch}}. For example with the OpenRISC 1000 architecture it is used to
3162 hold the number of matchpoints available in the target (along with other
3165 If there is no additional target specific information, it can be set to
3168 @node Registers and Memory
3169 @section Registers and Memory
3171 @value{GDBN}'s model of the target machine is rather simple.
3172 @value{GDBN} assumes the machine includes a bank of registers and a
3173 block of memory. Each register may have a different size.
3175 @value{GDBN} does not have a magical way to match up with the
3176 compiler's idea of which registers are which; however, it is critical
3177 that they do match up accurately. The only way to make this work is
3178 to get accurate information about the order that the compiler uses,
3179 and to reflect that in the @code{gdbarch_register_name} and related functions.
3181 @value{GDBN} can handle big-endian, little-endian, and bi-endian architectures.
3183 @node Pointers and Addresses
3184 @section Pointers Are Not Always Addresses
3185 @cindex pointer representation
3186 @cindex address representation
3187 @cindex word-addressed machines
3188 @cindex separate data and code address spaces
3189 @cindex spaces, separate data and code address
3190 @cindex address spaces, separate data and code
3191 @cindex code pointers, word-addressed
3192 @cindex converting between pointers and addresses
3193 @cindex D10V addresses
3195 On almost all 32-bit architectures, the representation of a pointer is
3196 indistinguishable from the representation of some fixed-length number
3197 whose value is the byte address of the object pointed to. On such
3198 machines, the words ``pointer'' and ``address'' can be used interchangeably.
3199 However, architectures with smaller word sizes are often cramped for
3200 address space, so they may choose a pointer representation that breaks this
3201 identity, and allows a larger code address space.
3203 @c D10V is gone from sources - more current example?
3205 For example, the Renesas D10V is a 16-bit VLIW processor whose
3206 instructions are 32 bits long@footnote{Some D10V instructions are
3207 actually pairs of 16-bit sub-instructions. However, since you can't
3208 jump into the middle of such a pair, code addresses can only refer to
3209 full 32 bit instructions, which is what matters in this explanation.}.
3210 If the D10V used ordinary byte addresses to refer to code locations,
3211 then the processor would only be able to address 64kb of instructions.
3212 However, since instructions must be aligned on four-byte boundaries, the
3213 low two bits of any valid instruction's byte address are always
3214 zero---byte addresses waste two bits. So instead of byte addresses,
3215 the D10V uses word addresses---byte addresses shifted right two bits---to
3216 refer to code. Thus, the D10V can use 16-bit words to address 256kb of
3219 However, this means that code pointers and data pointers have different
3220 forms on the D10V. The 16-bit word @code{0xC020} refers to byte address
3221 @code{0xC020} when used as a data address, but refers to byte address
3222 @code{0x30080} when used as a code address.
3224 (The D10V also uses separate code and data address spaces, which also
3225 affects the correspondence between pointers and addresses, but we're
3226 going to ignore that here; this example is already too long.)
3228 To cope with architectures like this---the D10V is not the only
3229 one!---@value{GDBN} tries to distinguish between @dfn{addresses}, which are
3230 byte numbers, and @dfn{pointers}, which are the target's representation
3231 of an address of a particular type of data. In the example above,
3232 @code{0xC020} is the pointer, which refers to one of the addresses
3233 @code{0xC020} or @code{0x30080}, depending on the type imposed upon it.
3234 @value{GDBN} provides functions for turning a pointer into an address
3235 and vice versa, in the appropriate way for the current architecture.
3237 Unfortunately, since addresses and pointers are identical on almost all
3238 processors, this distinction tends to bit-rot pretty quickly. Thus,
3239 each time you port @value{GDBN} to an architecture which does
3240 distinguish between pointers and addresses, you'll probably need to
3241 clean up some architecture-independent code.
3243 Here are functions which convert between pointers and addresses:
3245 @deftypefun CORE_ADDR extract_typed_address (void *@var{buf}, struct type *@var{type})
3246 Treat the bytes at @var{buf} as a pointer or reference of type
3247 @var{type}, and return the address it represents, in a manner
3248 appropriate for the current architecture. This yields an address
3249 @value{GDBN} can use to read target memory, disassemble, etc. Note that
3250 @var{buf} refers to a buffer in @value{GDBN}'s memory, not the
3253 For example, if the current architecture is the Intel x86, this function
3254 extracts a little-endian integer of the appropriate length from
3255 @var{buf} and returns it. However, if the current architecture is the
3256 D10V, this function will return a 16-bit integer extracted from
3257 @var{buf}, multiplied by four if @var{type} is a pointer to a function.
3259 If @var{type} is not a pointer or reference type, then this function
3260 will signal an internal error.
3263 @deftypefun CORE_ADDR store_typed_address (void *@var{buf}, struct type *@var{type}, CORE_ADDR @var{addr})
3264 Store the address @var{addr} in @var{buf}, in the proper format for a
3265 pointer of type @var{type} in the current architecture. Note that
3266 @var{buf} refers to a buffer in @value{GDBN}'s memory, not the
3269 For example, if the current architecture is the Intel x86, this function
3270 stores @var{addr} unmodified as a little-endian integer of the
3271 appropriate length in @var{buf}. However, if the current architecture
3272 is the D10V, this function divides @var{addr} by four if @var{type} is
3273 a pointer to a function, and then stores it in @var{buf}.
3275 If @var{type} is not a pointer or reference type, then this function
3276 will signal an internal error.
3279 @deftypefun CORE_ADDR value_as_address (struct value *@var{val})
3280 Assuming that @var{val} is a pointer, return the address it represents,
3281 as appropriate for the current architecture.
3283 This function actually works on integral values, as well as pointers.
3284 For pointers, it performs architecture-specific conversions as
3285 described above for @code{extract_typed_address}.
3288 @deftypefun CORE_ADDR value_from_pointer (struct type *@var{type}, CORE_ADDR @var{addr})
3289 Create and return a value representing a pointer of type @var{type} to
3290 the address @var{addr}, as appropriate for the current architecture.
3291 This function performs architecture-specific conversions as described
3292 above for @code{store_typed_address}.
3295 Here are two functions which architectures can define to indicate the
3296 relationship between pointers and addresses. These have default
3297 definitions, appropriate for architectures on which all pointers are
3298 simple unsigned byte addresses.
3300 @deftypefun CORE_ADDR gdbarch_pointer_to_address (struct gdbarch *@var{gdbarch}, struct type *@var{type}, char *@var{buf})
3301 Assume that @var{buf} holds a pointer of type @var{type}, in the
3302 appropriate format for the current architecture. Return the byte
3303 address the pointer refers to.
3305 This function may safely assume that @var{type} is either a pointer or a
3306 C@t{++} reference type.
3309 @deftypefun void gdbarch_address_to_pointer (struct gdbarch *@var{gdbarch}, struct type *@var{type}, char *@var{buf}, CORE_ADDR @var{addr})
3310 Store in @var{buf} a pointer of type @var{type} representing the address
3311 @var{addr}, in the appropriate format for the current architecture.
3313 This function may safely assume that @var{type} is either a pointer or a
3314 C@t{++} reference type.
3317 @node Address Classes
3318 @section Address Classes
3319 @cindex address classes
3320 @cindex DW_AT_byte_size
3321 @cindex DW_AT_address_class
3323 Sometimes information about different kinds of addresses is available
3324 via the debug information. For example, some programming environments
3325 define addresses of several different sizes. If the debug information
3326 distinguishes these kinds of address classes through either the size
3327 info (e.g, @code{DW_AT_byte_size} in @w{DWARF 2}) or through an explicit
3328 address class attribute (e.g, @code{DW_AT_address_class} in @w{DWARF 2}), the
3329 following macros should be defined in order to disambiguate these
3330 types within @value{GDBN} as well as provide the added information to
3331 a @value{GDBN} user when printing type expressions.
3333 @deftypefun int gdbarch_address_class_type_flags (struct gdbarch *@var{gdbarch}, int @var{byte_size}, int @var{dwarf2_addr_class})
3334 Returns the type flags needed to construct a pointer type whose size
3335 is @var{byte_size} and whose address class is @var{dwarf2_addr_class}.
3336 This function is normally called from within a symbol reader. See
3337 @file{dwarf2read.c}.
3340 @deftypefun {char *} gdbarch_address_class_type_flags_to_name (struct gdbarch *@var{gdbarch}, int @var{type_flags})
3341 Given the type flags representing an address class qualifier, return
3344 @deftypefun int gdbarch_address_class_name_to_type_flags (struct gdbarch *@var{gdbarch}, int @var{name}, int *@var{type_flags_ptr})
3345 Given an address qualifier name, set the @code{int} referenced by @var{type_flags_ptr} to the type flags
3346 for that address class qualifier.
3349 Since the need for address classes is rather rare, none of
3350 the address class functions are defined by default. Predicate
3351 functions are provided to detect when they are defined.
3353 Consider a hypothetical architecture in which addresses are normally
3354 32-bits wide, but 16-bit addresses are also supported. Furthermore,
3355 suppose that the @w{DWARF 2} information for this architecture simply
3356 uses a @code{DW_AT_byte_size} value of 2 to indicate the use of one
3357 of these "short" pointers. The following functions could be defined
3358 to implement the address class functions:
3361 somearch_address_class_type_flags (int byte_size,
3362 int dwarf2_addr_class)
3365 return TYPE_FLAG_ADDRESS_CLASS_1;
3371 somearch_address_class_type_flags_to_name (int type_flags)
3373 if (type_flags & TYPE_FLAG_ADDRESS_CLASS_1)
3380 somearch_address_class_name_to_type_flags (char *name,
3381 int *type_flags_ptr)
3383 if (strcmp (name, "short") == 0)
3385 *type_flags_ptr = TYPE_FLAG_ADDRESS_CLASS_1;
3393 The qualifier @code{@@short} is used in @value{GDBN}'s type expressions
3394 to indicate the presence of one of these ``short'' pointers. For
3395 example if the debug information indicates that @code{short_ptr_var} is
3396 one of these short pointers, @value{GDBN} might show the following
3400 (gdb) ptype short_ptr_var
3401 type = int * @@short
3405 @node Register Representation
3406 @section Register Representation
3409 * Raw and Cooked Registers::
3410 * Register Architecture Functions & Variables::
3411 * Register Information Functions::
3412 * Register and Memory Data::
3413 * Register Caching::
3416 @node Raw and Cooked Registers
3417 @subsection Raw and Cooked Registers
3418 @cindex raw register representation
3419 @cindex cooked register representation
3420 @cindex representations, raw and cooked registers
3422 @value{GDBN} considers registers to be a set with members numbered
3423 linearly from 0 upwards. The first part of that set corresponds to real
3424 physical registers, the second part to any @dfn{pseudo-registers}.
3425 Pseudo-registers have no independent physical existence, but are useful
3426 representations of information within the architecture. For example the
3427 OpenRISC 1000 architecture has up to 32 general purpose registers, which
3428 are typically represented as 32-bit (or 64-bit) integers. However the
3429 GPRs are also used as operands to the floating point operations, and it
3430 could be convenient to define a set of pseudo-registers, to show the
3431 GPRs represented as floating point values.
3433 For any architecture, the implementer will decide on a mapping from
3434 hardware to @value{GDBN} register numbers. The registers corresponding to real
3435 hardware are referred to as @dfn{raw} registers, the remaining registers are
3436 @dfn{pseudo-registers}. The total register set (raw and pseudo) is called
3437 the @dfn{cooked} register set.
3440 @node Register Architecture Functions & Variables
3441 @subsection Functions and Variables Specifying the Register Architecture
3442 @cindex @code{gdbarch} register architecture functions
3444 These @code{struct gdbarch} functions and variables specify the number
3445 and type of registers in the architecture.
3447 @deftypefn {Architecture Function} CORE_ADDR read_pc (struct regcache *@var{regcache})
3449 @deftypefn {Architecture Function} void write_pc (struct regcache *@var{regcache}, CORE_ADDR @var{val})
3451 Read or write the program counter. The default value of both
3452 functions is @code{NULL} (no function available). If the program
3453 counter is just an ordinary register, it can be specified in
3454 @code{struct gdbarch} instead (see @code{pc_regnum} below) and it will
3455 be read or written using the standard routines to access registers. This
3456 function need only be specified if the program counter is not an
3459 Any register information can be obtained using the supplied register
3460 cache, @var{regcache}. @xref{Register Caching, , Register Caching}.
3464 @deftypefn {Architecture Function} void pseudo_register_read (struct gdbarch *@var{gdbarch}, struct regcache *@var{regcache}, int @var{regnum}, const gdb_byte *@var{buf})
3466 @deftypefn {Architecture Function} void pseudo_register_write (struct gdbarch *@var{gdbarch}, struct regcache *@var{regcache}, int @var{regnum}, const gdb_byte *@var{buf})
3468 These functions should be defined if there are any pseudo-registers.
3469 The default value is @code{NULL}. @var{regnum} is the number of the
3470 register to read or write (which will be a @dfn{cooked} register
3471 number) and @var{buf} is the buffer where the value read will be
3472 placed, or from which the value to be written will be taken. The
3473 value in the buffer may be converted to or from a signed or unsigned
3474 integral value using one of the utility functions (@pxref{Register and
3475 Memory Data, , Using Different Register and Memory Data
3478 The access should be for the specified architecture,
3479 @var{gdbarch}. Any register information can be obtained using the
3480 supplied register cache, @var{regcache}. @xref{Register Caching, ,
3485 @deftypevr {Architecture Variable} int sp_regnum
3487 @cindex stack pointer
3490 This specifies the register holding the stack pointer, which may be a
3491 raw or pseudo-register. It defaults to -1 (not defined), but it is an
3492 error for it not to be defined.
3494 The value of the stack pointer register can be accessed withing
3495 @value{GDBN} as the variable @kbd{$sp}.
3499 @deftypevr {Architecture Variable} int pc_regnum
3501 @cindex program counter
3504 This specifies the register holding the program counter, which may be a
3505 raw or pseudo-register. It defaults to -1 (not defined). If
3506 @code{pc_regnum} is not defined, then the functions @code{read_pc} and
3507 @code{write_pc} (see above) must be defined.
3509 The value of the program counter (whether defined as a register, or
3510 through @code{read_pc} and @code{write_pc}) can be accessed withing
3511 @value{GDBN} as the variable @kbd{$pc}.
3515 @deftypevr {Architecture Variable} int ps_regnum
3517 @cindex processor status register
3518 @cindex status register
3521 This specifies the register holding the processor status (often called
3522 the status register), which may be a raw or pseudo-register. It
3523 defaults to -1 (not defined).
3525 If defined, the value of this register can be accessed withing
3526 @value{GDBN} as the variable @kbd{$ps}.
3530 @deftypevr {Architecture Variable} int fp0_regnum
3532 @cindex first floating point register
3534 This specifies the first floating point register. It defaults to
3535 0. @code{fp0_regnum} is not needed unless the target offers support
3540 @node Register Information Functions
3541 @subsection Functions Giving Register Information
3542 @cindex @code{gdbarch} register information functions
3544 These functions return information about registers.
3546 @deftypefn {Architecture Function} {const char *} register_name (struct gdbarch *@var{gdbarch}, int @var{regnum})
3548 This function should convert a register number (raw or pseudo) to a
3549 register name (as a C @code{const char *}). This is used both to
3550 determine the name of a register for output and to work out the meaning
3551 of any register names used as input. The function may also return
3552 @code{NULL}, to indicate that @var{regnum} is not a valid register.
3554 For example with the OpenRISC 1000, @value{GDBN} registers 0-31 are the
3555 General Purpose Registers, register 32 is the program counter and
3556 register 33 is the supervision register (i.e.@: the processor status
3557 register), which map to the strings @code{"gpr00"} through
3558 @code{"gpr31"}, @code{"pc"} and @code{"sr"} respectively. This means
3559 that the @value{GDBN} command @kbd{print $gpr5} should print the value of
3560 the OR1K general purpose register 5@footnote{
3561 @cindex frame pointer
3563 Historically, @value{GDBN} always had a concept of a frame pointer
3564 register, which could be accessed via the @value{GDBN} variable,
3565 @kbd{$fp}. That concept is now deprecated, recognizing that not all
3566 architectures have a frame pointer. However if an architecture does
3567 have a frame pointer register, and defines a register or
3568 pseudo-register with the name @code{"fp"}, then that register will be
3569 used as the value of the @kbd{$fp} variable.}.
3571 The default value for this function is @code{NULL}, meaning
3572 undefined. It should always be defined.
3574 The access should be for the specified architecture, @var{gdbarch}.
3578 @deftypefn {Architecture Function} {struct type *} register_type (struct gdbarch *@var{gdbarch}, int @var{regnum})
3580 Given a register number, this function identifies the type of data it
3581 may be holding, specified as a @code{struct type}. @value{GDBN} allows
3582 creation of arbitrary types, but a number of built in types are
3583 provided (@code{builtin_type_void}, @code{builtin_type_int32} etc),
3584 together with functions to derive types from these.
3586 Typically the program counter will have a type of ``pointer to
3587 function'' (it points to code), the frame pointer and stack pointer
3588 will have types of ``pointer to void'' (they point to data on the stack)
3589 and all other integer registers will have a type of 32-bit integer or
3592 This information guides the formatting when displaying register
3593 information. The default value is @code{NULL} meaning no information is
3594 available to guide formatting when displaying registers.
3598 @deftypefn {Architecture Function} void print_registers_info (struct gdbarch *@var{gdbarch}, struct ui_file *@var{file}, struct frame_info *@var{frame}, int @var{regnum}, int @var{all})
3600 Define this function to print out one or all of the registers for the
3601 @value{GDBN} @kbd{info registers} command. The default value is the
3602 function @code{default_print_registers_info}, which uses the register
3603 type information (see @code{register_type} above) to determine how each
3604 register should be printed. Define a custom version of this function
3605 for fuller control over how the registers are displayed.
3607 The access should be for the specified architecture, @var{gdbarch},
3608 with output to the the file specified by the User Interface
3609 Independent Output file handle, @var{file} (@pxref{UI-Independent
3610 Output, , UI-Independent Output---the @code{ui_out}
3613 The registers should show their values in the frame specified by
3614 @var{frame}. If @var{regnum} is -1 and @var{all} is zero, then all
3615 the ``significant'' registers should be shown (the implementer should
3616 decide which registers are ``significant''). Otherwise only the value of
3617 the register specified by @var{regnum} should be output. If
3618 @var{regnum} is -1 and @var{all} is non-zero (true), then the value of
3619 all registers should be shown.
3621 By default @code{default_print_registers_info} prints one register per
3622 line, and if @var{all} is zero omits floating-point registers.
3626 @deftypefn {Architecture Function} void print_float_info (struct gdbarch *@var{gdbarch}, struct ui_file *@var{file}, struct frame_info *@var{frame}, const char *@var{args})
3628 Define this function to provide output about the floating point unit and
3629 registers for the @value{GDBN} @kbd{info float} command respectively.
3630 The default value is @code{NULL} (not defined), meaning no information
3633 The @var{gdbarch} and @var{file} and @var{frame} arguments have the same
3634 meaning as in the @code{print_registers_info} function above. The string
3635 @var{args} contains any supplementary arguments to the @kbd{info float}
3638 Define this function if the target supports floating point operations.
3642 @deftypefn {Architecture Function} void print_vector_info (struct gdbarch *@var{gdbarch}, struct ui_file *@var{file}, struct frame_info *@var{frame}, const char *@var{args})
3644 Define this function to provide output about the vector unit and
3645 registers for the @value{GDBN} @kbd{info vector} command respectively.
3646 The default value is @code{NULL} (not defined), meaning no information
3649 The @var{gdbarch}, @var{file} and @var{frame} arguments have the
3650 same meaning as in the @code{print_registers_info} function above. The
3651 string @var{args} contains any supplementary arguments to the @kbd{info
3654 Define this function if the target supports vector operations.
3658 @deftypefn {Architecture Function} int register_reggroup_p (struct gdbarch *@var{gdbarch}, int @var{regnum}, struct reggroup *@var{group})
3660 @value{GDBN} groups registers into different categories (general,
3661 vector, floating point etc). This function, given a register,
3662 @var{regnum}, and group, @var{group}, returns 1 (true) if the register
3663 is in the group and 0 (false) otherwise.
3665 The information should be for the specified architecture,
3668 The default value is the function @code{default_register_reggroup_p}
3669 which will do a reasonable job based on the type of the register (see
3670 the function @code{register_type} above), with groups for general
3671 purpose registers, floating point registers, vector registers and raw
3672 (i.e not pseudo) registers.
3676 @node Register and Memory Data
3677 @subsection Using Different Register and Memory Data Representations
3678 @cindex register representation
3679 @cindex memory representation
3680 @cindex representations, register and memory
3681 @cindex register data formats, converting
3682 @cindex @code{struct value}, converting register contents to
3684 Some architectures have different representations of data objects,
3685 depending whether the object is held in a register or memory. For
3691 The Alpha architecture can represent 32 bit integer values in
3692 floating-point registers.
3695 The x86 architecture supports 80-bit floating-point registers. The
3696 @code{long double} data type occupies 96 bits in memory but only 80
3697 bits when stored in a register.
3701 In general, the register representation of a data type is determined by
3702 the architecture, or @value{GDBN}'s interface to the architecture, while
3703 the memory representation is determined by the Application Binary
3706 For almost all data types on almost all architectures, the two
3707 representations are identical, and no special handling is needed.
3708 However, they do occasionally differ. An architecture may define the
3709 following @code{struct gdbarch} functions to request conversions
3710 between the register and memory representations of a data type:
3712 @deftypefn {Architecture Function} int gdbarch_convert_register_p (struct gdbarch *@var{gdbarch}, int @var{reg})
3714 Return non-zero (true) if the representation of a data value stored in
3715 this register may be different to the representation of that same data
3716 value when stored in memory. The default value is @code{NULL}
3719 If this function is defined and returns non-zero, the @code{struct
3720 gdbarch} functions @code{gdbarch_register_to_value} and
3721 @code{gdbarch_value_to_register} (see below) should be used to perform
3722 any necessary conversion.
3724 If defined, this function should return zero for the register's native
3725 type, when no conversion is necessary.
3728 @deftypefn {Architecture Function} void gdbarch_register_to_value (struct gdbarch *@var{gdbarch}, int @var{reg}, struct type *@var{type}, char *@var{from}, char *@var{to})
3730 Convert the value of register number @var{reg} to a data object of
3731 type @var{type}. The buffer at @var{from} holds the register's value
3732 in raw format; the converted value should be placed in the buffer at
3736 @emph{Note:} @code{gdbarch_register_to_value} and
3737 @code{gdbarch_value_to_register} take their @var{reg} and @var{type}
3738 arguments in different orders.
3741 @code{gdbarch_register_to_value} should only be used with registers
3742 for which the @code{gdbarch_convert_register_p} function returns a
3747 @deftypefn {Architecture Function} void gdbarch_value_to_register (struct gdbarch *@var{gdbarch}, struct type *@var{type}, int @var{reg}, char *@var{from}, char *@var{to})
3749 Convert a data value of type @var{type} to register number @var{reg}'
3753 @emph{Note:} @code{gdbarch_register_to_value} and
3754 @code{gdbarch_value_to_register} take their @var{reg} and @var{type}
3755 arguments in different orders.
3758 @code{gdbarch_value_to_register} should only be used with registers
3759 for which the @code{gdbarch_convert_register_p} function returns a
3764 @node Register Caching
3765 @subsection Register Caching
3766 @cindex register caching
3768 Caching of registers is used, so that the target does not need to be
3769 accessed and reanalyzed multiple times for each register in
3770 circumstances where the register value cannot have changed.
3772 @cindex @code{struct regcache}
3773 @value{GDBN} provides @code{struct regcache}, associated with a
3774 particular @code{struct gdbarch} to hold the cached values of the raw
3775 registers. A set of functions is provided to access both the raw
3776 registers (with @code{raw} in their name) and the full set of cooked
3777 registers (with @code{cooked} in their name). Functions are provided
3778 to ensure the register cache is kept synchronized with the values of
3779 the actual registers in the target.
3781 Accessing registers through the @code{struct regcache} routines will
3782 ensure that the appropriate @code{struct gdbarch} functions are called
3783 when necessary to access the underlying target architecture. In general
3784 users should use the @dfn{cooked} functions, since these will map to the
3785 @dfn{raw} functions automatically as appropriate.
3787 @findex regcache_cooked_read
3788 @findex regcache_cooked_write
3789 @cindex @code{gdb_byte}
3790 @findex regcache_cooked_read_signed
3791 @findex regcache_cooked_read_unsigned
3792 @findex regcache_cooked_write_signed
3793 @findex regcache_cooked_write_unsigned
3794 The two key functions are @code{regcache_cooked_read} and
3795 @code{regcache_cooked_write} which read or write a register from or to
3796 a byte buffer (type @code{gdb_byte *}). For convenience the wrapper
3797 functions @code{regcache_cooked_read_signed},
3798 @code{regcache_cooked_read_unsigned},
3799 @code{regcache_cooked_write_signed} and
3800 @code{regcache_cooked_write_unsigned} are provided, which read or
3801 write the value using the buffer and convert to or from an integral
3802 value as appropriate.
3804 @node Frame Interpretation
3805 @section Frame Interpretation
3808 * All About Stack Frames::
3809 * Frame Handling Terminology::
3811 * Functions and Variable to Analyze Frames::
3812 * Functions to Access Frame Data::
3813 * Analyzing Stacks---Frame Sniffers::
3816 @node All About Stack Frames
3817 @subsection All About Stack Frames
3819 @value{GDBN} needs to understand the stack on which local (automatic)
3820 variables are stored. The area of the stack containing all the local
3821 variables for a function invocation is known as the @dfn{stack frame}
3822 for that function (or colloquially just as the @dfn{frame}). In turn the
3823 function that called the function will have its stack frame, and so on
3824 back through the chain of functions that have been called.
3826 Almost all architectures have one register dedicated to point to the
3827 end of the stack (the @dfn{stack pointer}). Many have a second register
3828 which points to the start of the currently active stack frame (the
3829 @dfn{frame pointer}). The specific arrangements for an architecture are
3830 a key part of the ABI.
3832 A diagram helps to explain this. Here is a simple program to compute
3845 return n * fact (n - 1);
3853 for (i = 0; i < 10; i++)
3856 printf ("%d! = %d\n", i, f);
3861 Consider the state of the stack when the code reaches line 6 after the
3862 main program has called @code{fact@w{ }(3)}. The chain of function
3863 calls will be @code{main ()}, @code{fact@w{ }(3)}, @code{fact@w{
3864 }(2)}, @code{@w{fact (1)}} and @code{fact@w{ }(0)}.
3866 In this illustration the stack is falling (as used for example by the
3867 OpenRISC 1000 ABI). The stack pointer (SP) is at the end of the stack
3868 (lowest address) and the frame pointer (FP) is at the highest address
3869 in the current stack frame. The following diagram shows how the stack
3872 @center @image{stack_frame,14cm}
3874 In each stack frame, offset 0 from the stack pointer is the frame
3875 pointer of the previous frame and offset 4 (this is illustrating a
3876 32-bit architecture) from the stack pointer is the return address.
3877 Local variables are indexed from the frame pointer, with negative
3878 indexes. In the function @code{fact}, offset -4 from the frame
3879 pointer is the argument @var{n}. In the @code{main} function, offset
3880 -4 from the frame pointer is the local variable @var{i} and offset -8
3881 from the frame pointer is the local variable @var{f}@footnote{This is
3882 a simplified example for illustrative purposes only. Good optimizing
3883 compilers would not put anything on the stack for such simple
3884 functions. Indeed they might eliminate the recursion and use of the
3887 It is very easy to get confused when examining stacks. @value{GDBN}
3888 has terminology it uses rigorously throughout. The stack frame of the
3889 function currently executing, or where execution stopped is numbered
3890 zero. In this example frame #0 is the stack frame of the call to
3891 @code{fact@w{ }(0)}. The stack frame of its calling function
3892 (@code{fact@w{ }(1)} in this case) is numbered #1 and so on back
3893 through the chain of calls.
3895 The main @value{GDBN} data structure describing frames is
3896 @code{@w{struct frame_info}}. It is not used directly, but only via
3897 its accessor functions. @code{frame_info} includes information about
3898 the registers in the frame and a pointer to the code of the function
3899 with which the frame is associated. The entire stack is represented as
3900 a linked list of @code{frame_info} structs.
3902 @node Frame Handling Terminology
3903 @subsection Frame Handling Terminology
3905 It is easy to get confused when referencing stack frames. @value{GDBN}
3906 uses some precise terminology.
3912 @cindex stack frame, definition of THIS frame
3913 @cindex frame, definition of THIS frame
3914 @dfn{THIS} frame is the frame currently under consideration.
3918 @cindex stack frame, definition of NEXT frame
3919 @cindex frame, definition of NEXT frame
3920 The @dfn{NEXT} frame, also sometimes called the inner or newer frame is the
3921 frame of the function called by the function of THIS frame.
3924 @cindex PREVIOUS frame
3925 @cindex stack frame, definition of PREVIOUS frame
3926 @cindex frame, definition of PREVIOUS frame
3927 The @dfn{PREVIOUS} frame, also sometimes called the outer or older frame is
3928 the frame of the function which called the function of THIS frame.
3932 So in the example in the previous section (@pxref{All About Stack
3933 Frames, , All About Stack Frames}), if THIS frame is #3 (the call to
3934 @code{fact@w{ }(3)}), the NEXT frame is frame #2 (the call to
3935 @code{fact@w{ }(2)}) and the PREVIOUS frame is frame #4 (the call to
3936 @code{main@w{ }()}).
3938 @cindex innermost frame
3939 @cindex stack frame, definition of innermost frame
3940 @cindex frame, definition of innermost frame
3941 The @dfn{innermost} frame is the frame of the current executing
3942 function, or where the program stopped, in this example, in the middle
3943 of the call to @code{@w{fact (0))}}. It is always numbered frame #0.
3945 @cindex base of a frame
3946 @cindex stack frame, definition of base of a frame
3947 @cindex frame, definition of base of a frame
3948 The @dfn{base} of a frame is the address immediately before the start
3949 of the NEXT frame. For a stack which grows down in memory (a
3950 @dfn{falling} stack) this will be the lowest address and for a stack
3951 which grows up in memory (a @dfn{rising} stack) this will be the
3952 highest address in the frame.
3954 @value{GDBN} functions to analyze the stack are typically given a
3955 pointer to the NEXT frame to determine information about THIS
3956 frame. Information about THIS frame includes data on where the
3957 registers of the PREVIOUS frame are stored in this stack frame. In
3958 this example the frame pointer of the PREVIOUS frame is stored at
3959 offset 0 from the stack pointer of THIS frame.
3962 @cindex stack frame, definition of unwinding
3963 @cindex frame, definition of unwinding
3964 The process whereby a function is given a pointer to the NEXT
3965 frame to work out information about THIS frame is referred to as
3966 @dfn{unwinding}. The @value{GDBN} functions involved in this typically
3967 include unwind in their name.
3970 @cindex stack frame, definition of sniffing
3971 @cindex frame, definition of sniffing
3972 The process of analyzing a target to determine the information that
3973 should go in struct frame_info is called @dfn{sniffing}. The functions
3974 that carry this out are called sniffers and typically include sniffer
3975 in their name. More than one sniffer may be required to extract all
3976 the information for a particular frame.
3978 @cindex sentinel frame
3979 @cindex stack frame, definition of sentinel frame
3980 @cindex frame, definition of sentinel frame
3981 Because so many functions work using the NEXT frame, there is an issue
3982 about addressing the innermost frame---it has no NEXT frame. To solve
3983 this @value{GDBN} creates a dummy frame #-1, known as the
3984 @dfn{sentinel} frame.
3986 @node Prologue Caches
3987 @subsection Prologue Caches
3989 @cindex function prologue
3990 @cindex prologue of a function
3991 All the frame sniffing functions typically examine the code at the
3992 start of the corresponding function, to determine the state of
3993 registers. The ABI will save old values and set new values of key
3994 registers at the start of each function in what is known as the
3995 function @dfn{prologue}.
3997 @cindex prologue cache
3998 For any particular stack frame this data does not change, so all the
3999 standard unwinding functions, in addition to receiving a pointer to
4000 the NEXT frame as their first argument, receive a pointer to a
4001 @dfn{prologue cache} as their second argument. This can be used to store
4002 values associated with a particular frame, for reuse on subsequent
4003 calls involving the same frame.
4005 It is up to the user to define the structure used (it is a
4006 @code{void@w{ }*} pointer) and arrange allocation and deallocation of
4007 storage. However for general use, @value{GDBN} provides
4008 @code{@w{struct trad_frame_cache}}, with a set of accessor
4009 routines. This structure holds the stack and code address of
4010 THIS frame, the base address of the frame, a pointer to the
4011 struct @code{frame_info} for the NEXT frame and details of
4012 where the registers of the PREVIOUS frame may be found in THIS
4015 Typically the first time any sniffer function is called with NEXT
4016 frame, the prologue sniffer for THIS frame will be @code{NULL}. The
4017 sniffer will analyze the frame, allocate a prologue cache structure
4018 and populate it. Subsequent calls using the same NEXT frame will
4019 pass in this prologue cache, so the data can be returned with no
4020 additional analysis.
4022 @node Functions and Variable to Analyze Frames
4023 @subsection Functions and Variable to Analyze Frames
4025 These struct @code{gdbarch} functions and variable should be defined
4026 to provide analysis of the stack frame and allow it to be adjusted as
4029 @deftypefn {Architecture Function} CORE_ADDR skip_prologue (struct gdbarch *@var{gdbarch}, CORE_ADDR @var{pc})
4031 The prologue of a function is the code at the beginning of the
4032 function which sets up the stack frame, saves the return address
4033 etc. The code representing the behavior of the function starts after
4036 This function skips past the prologue of a function if the program
4037 counter, @var{pc}, is within the prologue of a function. The result is
4038 the program counter immediately after the prologue. With modern
4039 optimizing compilers, this may be a far from trivial exercise. However
4040 the required information may be within the binary as DWARF2 debugging
4041 information, making the job much easier.
4043 The default value is @code{NULL} (not defined). This function should always
4044 be provided, but can take advantage of DWARF2 debugging information,
4045 if that is available.
4049 @deftypefn {Architecture Function} int inner_than (CORE_ADDR @var{lhs}, CORE_ADDR @var{rhs})
4050 @findex core_addr_lessthan
4051 @findex core_addr_greaterthan
4053 Given two frame or stack pointers, return non-zero (true) if the first
4054 represents the @dfn{inner} stack frame and 0 (false) otherwise. This
4055 is used to determine whether the target has a stack which grows up in
4056 memory (rising stack) or grows down in memory (falling stack).
4057 @xref{All About Stack Frames, , All About Stack Frames}, for an
4058 explanation of @dfn{inner} frames.
4060 The default value of this function is @code{NULL} and it should always
4061 be defined. However for almost all architectures one of the built-in
4062 functions can be used: @code{core_addr_lessthan} (for stacks growing
4063 down in memory) or @code{core_addr_greaterthan} (for stacks growing up
4068 @anchor{frame_align}
4069 @deftypefn {Architecture Function} CORE_ADDR frame_align (struct gdbarch *@var{gdbarch}, CORE_ADDR @var{address})
4073 The architecture may have constraints on how its frames are
4074 aligned. For example the OpenRISC 1000 ABI requires stack frames to be
4075 double-word aligned, but 32-bit versions of the architecture allocate
4076 single-word values to the stack. Thus extra padding may be needed at
4077 the end of a stack frame.
4079 Given a proposed address for the stack pointer, this function
4080 returns a suitably aligned address (by expanding the stack frame).
4082 The default value is @code{NULL} (undefined). This function should be defined
4083 for any architecture where it is possible the stack could become
4084 misaligned. The utility functions @code{align_down} (for falling
4085 stacks) and @code{align_up} (for rising stacks) will facilitate the
4086 implementation of this function.
4090 @deftypevr {Architecture Variable} int frame_red_zone_size
4092 Some ABIs reserve space beyond the end of the stack for use by leaf
4093 functions without prologue or epilogue or by exception handlers (for
4094 example the OpenRISC 1000).
4096 This is known as a @dfn{red zone} (AMD terminology). The @sc{amd64}
4097 (nee x86-64) ABI documentation refers to the @dfn{red zone} when
4098 describing this scratch area.
4100 The default value is 0. Set this field if the architecture has such a
4101 red zone. The value must be aligned as required by the ABI (see
4102 @code{frame_align} above for an explanation of stack frame alignment).
4106 @node Functions to Access Frame Data
4107 @subsection Functions to Access Frame Data
4109 These functions provide access to key registers and arguments in the
4112 @deftypefn {Architecture Function} CORE_ADDR unwind_pc (struct gdbarch *@var{gdbarch}, struct frame_info *@var{next_frame})
4114 This function is given a pointer to the NEXT stack frame (@pxref{All
4115 About Stack Frames, , All About Stack Frames}, for how frames are
4116 represented) and returns the value of the program counter in the
4117 PREVIOUS frame (i.e.@: the frame of the function that called THIS
4118 one). This is commonly referred to as the @dfn{return address}.
4120 The implementation, which must be frame agnostic (work with any frame),
4121 is typically no more than:
4125 pc = frame_unwind_register_unsigned (next_frame, @var{ARCH}_PC_REGNUM);
4126 return gdbarch_addr_bits_remove (gdbarch, pc);
4131 @deftypefn {Architecture Function} CORE_ADDR unwind_sp (struct gdbarch *@var{gdbarch}, struct frame_info *@var{next_frame})
4133 This function is given a pointer to the NEXT stack frame
4134 (@pxref{All About Stack Frames, , All About Stack Frames} for how
4135 frames are represented) and returns the value of the stack pointer in
4136 the PREVIOUS frame (i.e.@: the frame of the function that called
4139 The implementation, which must be frame agnostic (work with any frame),
4140 is typically no more than:
4144 sp = frame_unwind_register_unsigned (next_frame, @var{ARCH}_SP_REGNUM);
4145 return gdbarch_addr_bits_remove (gdbarch, sp);
4150 @deftypefn {Architecture Function} int frame_num_args (struct gdbarch *@var{gdbarch}, struct frame_info *@var{this_frame})
4152 This function is given a pointer to THIS stack frame (@pxref{All
4153 About Stack Frames, , All About Stack Frames} for how frames are
4154 represented), and returns the number of arguments that are being
4155 passed, or -1 if not known.
4157 The default value is @code{NULL} (undefined), in which case the number of
4158 arguments passed on any stack frame is always unknown. For many
4159 architectures this will be a suitable default.
4163 @node Analyzing Stacks---Frame Sniffers
4164 @subsection Analyzing Stacks---Frame Sniffers
4166 When a program stops, @value{GDBN} needs to construct the chain of
4167 struct @code{frame_info} representing the state of the stack using
4168 appropriate @dfn{sniffers}.
4170 Each architecture requires appropriate sniffers, but they do not form
4171 entries in @code{@w{struct gdbarch}}, since more than one sniffer may
4172 be required and a sniffer may be suitable for more than one
4173 @code{@w{struct gdbarch}}. Instead sniffers are associated with
4174 architectures using the following functions.
4179 @findex frame_unwind_append_sniffer
4180 @code{frame_unwind_append_sniffer} is used to add a new sniffer to
4181 analyze THIS frame when given a pointer to the NEXT frame.
4184 @findex frame_base_append_sniffer
4185 @code{frame_base_append_sniffer} is used to add a new sniffer
4186 which can determine information about the base of a stack frame.
4189 @findex frame_base_set_default
4190 @code{frame_base_set_default} is used to specify the default base
4195 These functions all take a reference to @code{@w{struct gdbarch}}, so
4196 they are associated with a specific architecture. They are usually
4197 called in the @code{gdbarch} initialization function, after the
4198 @code{gdbarch} struct has been set up. Unless a default has been set, the
4199 most recently appended sniffer will be tried first.
4201 The main frame unwinding sniffer (as set by
4202 @code{frame_unwind_append_sniffer)} returns a structure specifying
4203 a set of sniffing functions:
4205 @cindex @code{frame_unwind}
4209 enum frame_type type;
4210 frame_this_id_ftype *this_id;
4211 frame_prev_register_ftype *prev_register;
4212 const struct frame_data *unwind_data;
4213 frame_sniffer_ftype *sniffer;
4214 frame_prev_pc_ftype *prev_pc;
4215 frame_dealloc_cache_ftype *dealloc_cache;
4219 The @code{type} field indicates the type of frame this sniffer can
4220 handle: normal, dummy (@pxref{Functions Creating Dummy Frames, ,
4221 Functions Creating Dummy Frames}), signal handler or sentinel. Signal
4222 handlers sometimes have their own simplified stack structure for
4223 efficiency, so may need their own handlers.
4225 The @code{unwind_data} field holds additional information which may be
4226 relevant to particular types of frame. For example it may hold
4227 additional information for signal handler frames.
4229 The remaining fields define functions that yield different types of
4230 information when given a pointer to the NEXT stack frame. Not all
4231 functions need be provided. If an entry is @code{NULL}, the next sniffer will
4237 @code{this_id} determines the stack pointer and function (code
4238 entry point) for THIS stack frame.
4241 @code{prev_register} determines where the values of registers for
4242 the PREVIOUS stack frame are stored in THIS stack frame.
4245 @code{sniffer} takes a look at THIS frame's registers to
4246 determine if this is the appropriate unwinder.
4249 @code{prev_pc} determines the program counter for THIS
4250 frame. Only needed if the program counter is not an ordinary register
4251 (@pxref{Register Architecture Functions & Variables,
4252 , Functions and Variables Specifying the Register Architecture}).
4255 @code{dealloc_cache} frees any additional memory associated with
4256 the prologue cache for this frame (@pxref{Prologue Caches, , Prologue
4261 In general it is only the @code{this_id} and @code{prev_register}
4262 fields that need be defined for custom sniffers.
4264 The frame base sniffer is much simpler. It is a @code{@w{struct
4265 frame_base}}, which refers to the corresponding @code{frame_unwind}
4266 struct and whose fields refer to functions yielding various addresses
4269 @cindex @code{frame_base}
4273 const struct frame_unwind *unwind;
4274 frame_this_base_ftype *this_base;
4275 frame_this_locals_ftype *this_locals;
4276 frame_this_args_ftype *this_args;
4280 All the functions referred to take a pointer to the NEXT frame as
4281 argument. The function referred to by @code{this_base} returns the
4282 base address of THIS frame, the function referred to by
4283 @code{this_locals} returns the base address of local variables in THIS
4284 frame and the function referred to by @code{this_args} returns the
4285 base address of the function arguments in this frame.
4287 As described above, the base address of a frame is the address
4288 immediately before the start of the NEXT frame. For a falling
4289 stack, this is the lowest address in the frame and for a rising stack
4290 it is the highest address in the frame. For most architectures the
4291 same address is also the base address for local variables and
4292 arguments, in which case the same function can be used for all three
4293 entries@footnote{It is worth noting that if it cannot be determined in any
4294 other way (for example by there being a register with the name
4295 @code{"fp"}), then the result of the @code{this_base} function will be
4296 used as the value of the frame pointer variable @kbd{$fp} in
4297 @value{GDBN}. This is very often not correct (for example with the
4298 OpenRISC 1000, this value is the stack pointer, @kbd{$sp}). In this
4299 case a register (raw or pseudo) with the name @code{"fp"} should be
4300 defined. It will be used in preference as the value of @kbd{$fp}.}.
4302 @node Inferior Call Setup
4303 @section Inferior Call Setup
4304 @cindex calls to the inferior
4307 * About Dummy Frames::
4308 * Functions Creating Dummy Frames::
4311 @node About Dummy Frames
4312 @subsection About Dummy Frames
4313 @cindex dummy frames
4315 @value{GDBN} can call functions in the target code (for example by
4316 using the @kbd{call} or @kbd{print} commands). These functions may be
4317 breakpointed, and it is essential that if a function does hit a
4318 breakpoint, commands like @kbd{backtrace} work correctly.
4320 This is achieved by making the stack look as though the function had
4321 been called from the point where @value{GDBN} had previously stopped.
4322 This requires that @value{GDBN} can set up stack frames appropriate for
4323 such function calls.
4325 @node Functions Creating Dummy Frames
4326 @subsection Functions Creating Dummy Frames
4328 The following functions provide the functionality to set up such
4329 @dfn{dummy} stack frames.
4331 @deftypefn {Architecture Function} CORE_ADDR push_dummy_call (struct gdbarch *@var{gdbarch}, struct value *@var{function}, struct regcache *@var{regcache}, CORE_ADDR @var{bp_addr}, int @var{nargs}, struct value **@var{args}, CORE_ADDR @var{sp}, int @var{struct_return}, CORE_ADDR @var{struct_addr})
4333 This function sets up a dummy stack frame for the function about to be
4334 called. @code{push_dummy_call} is given the arguments to be passed
4335 and must copy them into registers or push them on to the stack as
4336 appropriate for the ABI.
4338 @var{function} is a pointer to the function
4339 that will be called and @var{regcache} the register cache from which
4340 values should be obtained. @var{bp_addr} is the address to which the
4341 function should return (which is breakpointed, so @value{GDBN} can
4342 regain control, hence the name). @var{nargs} is the number of
4343 arguments to pass and @var{args} an array containing the argument
4344 values. @var{struct_return} is non-zero (true) if the function returns
4345 a structure, and if so @var{struct_addr} is the address in which the
4346 structure should be returned.
4348 After calling this function, @value{GDBN} will pass control to the
4349 target at the address of the function, which will find the stack and
4350 registers set up just as expected.
4352 The default value of this function is @code{NULL} (undefined). If the
4353 function is not defined, then @value{GDBN} will not allow the user to
4354 call functions within the target being debugged.
4358 @deftypefn {Architecture Function} {struct frame_id} unwind_dummy_id (struct gdbarch *@var{gdbarch}, struct frame_info *@var{next_frame})
4360 This is the inverse of @code{push_dummy_call} which restores the stack
4361 pointer and program counter after a call to evaluate a function using
4362 a dummy stack frame. The result is a @code{@w{struct frame_id}}, which
4363 contains the value of the stack pointer and program counter to be
4366 The NEXT frame pointer is provided as argument,
4367 @var{next_frame}. THIS frame is the frame of the dummy function,
4368 which can be unwound, to yield the required stack pointer and program
4369 counter from the PREVIOUS frame.
4371 The default value is @code{NULL} (undefined). If @code{push_dummy_call} is
4372 defined, then this function should also be defined.
4376 @deftypefn {Architecture Function} CORE_ADDR push_dummy_code (struct gdbarch *@var{gdbarch}, CORE_ADDR @var{sp}, CORE_ADDR @var{funaddr}, struct value **@var{args}, int @var{nargs}, struct type *@var{value_type}, CORE_ADDR *@var{real_pc}, CORE_ADDR *@var{bp_addr}, struct regcache *@var{regcache})
4378 If this function is not defined (its default value is @code{NULL}), a dummy
4379 call will use the entry point of the currently loaded code on the
4380 target as its return address. A temporary breakpoint will be set
4381 there, so the location must be writable and have room for a
4384 It is possible that this default is not suitable. It might not be
4385 writable (in ROM possibly), or the ABI might require code to be
4386 executed on return from a call to unwind the stack before the
4387 breakpoint is encountered.
4389 If either of these is the case, then push_dummy_code should be defined
4390 to push an instruction sequence onto the end of the stack to which the
4391 dummy call should return.
4393 The arguments are essentially the same as those to
4394 @code{push_dummy_call}. However the function is provided with the
4395 type of the function result, @var{value_type}, @var{bp_addr} is used
4396 to return a value (the address at which the breakpoint instruction
4397 should be inserted) and @var{real pc} is used to specify the resume
4398 address when starting the call sequence. The function should return
4399 the updated innermost stack address.
4402 @emph{Note:} This does require that code in the stack can be executed.
4403 Some Harvard architectures may not allow this.
4408 @node Adding support for debugging core files
4409 @section Adding support for debugging core files
4412 The prerequisite for adding core file support in @value{GDBN} is to have
4413 core file support in BFD.
4415 Once BFD support is available, writing the apropriate
4416 @code{regset_from_core_section} architecture function should be all
4417 that is needed in order to add support for core files in @value{GDBN}.
4419 @node Defining Other Architecture Features
4420 @section Defining Other Architecture Features
4422 This section describes other functions and values in @code{gdbarch},
4423 together with some useful macros, that you can use to define the
4424 target architecture.
4428 @item CORE_ADDR gdbarch_addr_bits_remove (@var{gdbarch}, @var{addr})
4429 @findex gdbarch_addr_bits_remove
4430 If a raw machine instruction address includes any bits that are not
4431 really part of the address, then this function is used to zero those bits in
4432 @var{addr}. This is only used for addresses of instructions, and even then not
4435 For example, the two low-order bits of the PC on the Hewlett-Packard PA
4436 2.0 architecture contain the privilege level of the corresponding
4437 instruction. Since instructions must always be aligned on four-byte
4438 boundaries, the processor masks out these bits to generate the actual
4439 address of the instruction. @code{gdbarch_addr_bits_remove} would then for
4440 example look like that:
4442 arch_addr_bits_remove (CORE_ADDR addr)
4444 return (addr &= ~0x3);
4448 @item int address_class_name_to_type_flags (@var{gdbarch}, @var{name}, @var{type_flags_ptr})
4449 @findex address_class_name_to_type_flags
4450 If @var{name} is a valid address class qualifier name, set the @code{int}
4451 referenced by @var{type_flags_ptr} to the mask representing the qualifier
4452 and return 1. If @var{name} is not a valid address class qualifier name,
4455 The value for @var{type_flags_ptr} should be one of
4456 @code{TYPE_FLAG_ADDRESS_CLASS_1}, @code{TYPE_FLAG_ADDRESS_CLASS_2}, or
4457 possibly some combination of these values or'd together.
4458 @xref{Target Architecture Definition, , Address Classes}.
4460 @item int address_class_name_to_type_flags_p (@var{gdbarch})
4461 @findex address_class_name_to_type_flags_p
4462 Predicate which indicates whether @code{address_class_name_to_type_flags}
4465 @item int gdbarch_address_class_type_flags (@var{gdbarch}, @var{byte_size}, @var{dwarf2_addr_class})
4466 @findex gdbarch_address_class_type_flags
4467 Given a pointers byte size (as described by the debug information) and
4468 the possible @code{DW_AT_address_class} value, return the type flags
4469 used by @value{GDBN} to represent this address class. The value
4470 returned should be one of @code{TYPE_FLAG_ADDRESS_CLASS_1},
4471 @code{TYPE_FLAG_ADDRESS_CLASS_2}, or possibly some combination of these
4472 values or'd together.
4473 @xref{Target Architecture Definition, , Address Classes}.
4475 @item int gdbarch_address_class_type_flags_p (@var{gdbarch})
4476 @findex gdbarch_address_class_type_flags_p
4477 Predicate which indicates whether @code{gdbarch_address_class_type_flags_p} has
4480 @item const char *gdbarch_address_class_type_flags_to_name (@var{gdbarch}, @var{type_flags})
4481 @findex gdbarch_address_class_type_flags_to_name
4482 Return the name of the address class qualifier associated with the type
4483 flags given by @var{type_flags}.
4485 @item int gdbarch_address_class_type_flags_to_name_p (@var{gdbarch})
4486 @findex gdbarch_address_class_type_flags_to_name_p
4487 Predicate which indicates whether @code{gdbarch_address_class_type_flags_to_name} has been defined.
4488 @xref{Target Architecture Definition, , Address Classes}.
4490 @item void gdbarch_address_to_pointer (@var{gdbarch}, @var{type}, @var{buf}, @var{addr})
4491 @findex gdbarch_address_to_pointer
4492 Store in @var{buf} a pointer of type @var{type} representing the address
4493 @var{addr}, in the appropriate format for the current architecture.
4494 This function may safely assume that @var{type} is either a pointer or a
4495 C@t{++} reference type.
4496 @xref{Target Architecture Definition, , Pointers Are Not Always Addresses}.
4498 @item int gdbarch_believe_pcc_promotion (@var{gdbarch})
4499 @findex gdbarch_believe_pcc_promotion
4500 Used to notify if the compiler promotes a @code{short} or @code{char}
4501 parameter to an @code{int}, but still reports the parameter as its
4502 original type, rather than the promoted type.
4504 @item gdbarch_bits_big_endian (@var{gdbarch})
4505 @findex gdbarch_bits_big_endian
4506 This is used if the numbering of bits in the targets does @strong{not} match
4507 the endianism of the target byte order. A value of 1 means that the bits
4508 are numbered in a big-endian bit order, 0 means little-endian.
4510 @item set_gdbarch_bits_big_endian (@var{gdbarch}, @var{bits_big_endian})
4511 @findex set_gdbarch_bits_big_endian
4512 Calling set_gdbarch_bits_big_endian with a value of 1 indicates that the
4513 bits in the target are numbered in a big-endian bit order, 0 indicates
4518 This is the character array initializer for the bit pattern to put into
4519 memory where a breakpoint is set. Although it's common to use a trap
4520 instruction for a breakpoint, it's not required; for instance, the bit
4521 pattern could be an invalid instruction. The breakpoint must be no
4522 longer than the shortest instruction of the architecture.
4524 @code{BREAKPOINT} has been deprecated in favor of
4525 @code{gdbarch_breakpoint_from_pc}.
4527 @item BIG_BREAKPOINT
4528 @itemx LITTLE_BREAKPOINT
4529 @findex LITTLE_BREAKPOINT
4530 @findex BIG_BREAKPOINT
4531 Similar to BREAKPOINT, but used for bi-endian targets.
4533 @code{BIG_BREAKPOINT} and @code{LITTLE_BREAKPOINT} have been deprecated in
4534 favor of @code{gdbarch_breakpoint_from_pc}.
4536 @item const gdb_byte *gdbarch_breakpoint_from_pc (@var{gdbarch}, @var{pcptr}, @var{lenptr})
4537 @findex gdbarch_breakpoint_from_pc
4538 @anchor{gdbarch_breakpoint_from_pc} Use the program counter to determine the
4539 contents and size of a breakpoint instruction. It returns a pointer to
4540 a static string of bytes that encode a breakpoint instruction, stores the
4541 length of the string to @code{*@var{lenptr}}, and adjusts the program
4542 counter (if necessary) to point to the actual memory location where the
4543 breakpoint should be inserted. May return @code{NULL} to indicate that
4544 software breakpoints are not supported.
4546 Although it is common to use a trap instruction for a breakpoint, it's
4547 not required; for instance, the bit pattern could be an invalid
4548 instruction. The breakpoint must be no longer than the shortest
4549 instruction of the architecture.
4551 Provided breakpoint bytes can be also used by @code{bp_loc_is_permanent} to
4552 detect permanent breakpoints. @code{gdbarch_breakpoint_from_pc} should return
4553 an unchanged memory copy if it was called for a location with permanent
4554 breakpoint as some architectures use breakpoint instructions containing
4555 arbitrary parameter value.
4557 Replaces all the other @var{BREAKPOINT} macros.
4559 @item int gdbarch_memory_insert_breakpoint (@var{gdbarch}, @var{bp_tgt})
4560 @itemx gdbarch_memory_remove_breakpoint (@var{gdbarch}, @var{bp_tgt})
4561 @findex gdbarch_memory_remove_breakpoint
4562 @findex gdbarch_memory_insert_breakpoint
4563 Insert or remove memory based breakpoints. Reasonable defaults
4564 (@code{default_memory_insert_breakpoint} and
4565 @code{default_memory_remove_breakpoint} respectively) have been
4566 provided so that it is not necessary to set these for most
4567 architectures. Architectures which may want to set
4568 @code{gdbarch_memory_insert_breakpoint} and @code{gdbarch_memory_remove_breakpoint} will likely have instructions that are oddly sized or are not stored in a
4569 conventional manner.
4571 It may also be desirable (from an efficiency standpoint) to define
4572 custom breakpoint insertion and removal routines if
4573 @code{gdbarch_breakpoint_from_pc} needs to read the target's memory for some
4576 @item CORE_ADDR gdbarch_adjust_breakpoint_address (@var{gdbarch}, @var{bpaddr})
4577 @findex gdbarch_adjust_breakpoint_address
4578 @cindex breakpoint address adjusted
4579 Given an address at which a breakpoint is desired, return a breakpoint
4580 address adjusted to account for architectural constraints on
4581 breakpoint placement. This method is not needed by most targets.
4583 The FR-V target (see @file{frv-tdep.c}) requires this method.
4584 The FR-V is a VLIW architecture in which a number of RISC-like
4585 instructions are grouped (packed) together into an aggregate
4586 instruction or instruction bundle. When the processor executes
4587 one of these bundles, the component instructions are executed
4590 In the course of optimization, the compiler may group instructions
4591 from distinct source statements into the same bundle. The line number
4592 information associated with one of the latter statements will likely
4593 refer to some instruction other than the first one in the bundle. So,
4594 if the user attempts to place a breakpoint on one of these latter
4595 statements, @value{GDBN} must be careful to @emph{not} place the break
4596 instruction on any instruction other than the first one in the bundle.
4597 (Remember though that the instructions within a bundle execute
4598 in parallel, so the @emph{first} instruction is the instruction
4599 at the lowest address and has nothing to do with execution order.)
4601 The FR-V's @code{gdbarch_adjust_breakpoint_address} method will adjust a
4602 breakpoint's address by scanning backwards for the beginning of
4603 the bundle, returning the address of the bundle.
4605 Since the adjustment of a breakpoint may significantly alter a user's
4606 expectation, @value{GDBN} prints a warning when an adjusted breakpoint
4607 is initially set and each time that that breakpoint is hit.
4609 @item int gdbarch_call_dummy_location (@var{gdbarch})
4610 @findex gdbarch_call_dummy_location
4611 See the file @file{inferior.h}.
4613 This method has been replaced by @code{gdbarch_push_dummy_code}
4614 (@pxref{gdbarch_push_dummy_code}).
4616 @item int gdbarch_cannot_fetch_register (@var{gdbarch}, @var{regum})
4617 @findex gdbarch_cannot_fetch_register
4618 This function should return nonzero if @var{regno} cannot be fetched
4619 from an inferior process.
4621 @item int gdbarch_cannot_store_register (@var{gdbarch}, @var{regnum})
4622 @findex gdbarch_cannot_store_register
4623 This function should return nonzero if @var{regno} should not be
4624 written to the target. This is often the case for program counters,
4625 status words, and other special registers. This function returns 0 as
4626 default so that @value{GDBN} will assume that all registers may be written.
4628 @item int gdbarch_convert_register_p (@var{gdbarch}, @var{regnum}, struct type *@var{type})
4629 @findex gdbarch_convert_register_p
4630 Return non-zero if register @var{regnum} represents data values of type
4631 @var{type} in a non-standard form.
4632 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
4634 @item int gdbarch_fp0_regnum (@var{gdbarch})
4635 @findex gdbarch_fp0_regnum
4636 This function returns the number of the first floating point register,
4637 if the machine has such registers. Otherwise, it returns -1.
4639 @item CORE_ADDR gdbarch_decr_pc_after_break (@var{gdbarch})
4640 @findex gdbarch_decr_pc_after_break
4641 This function shall return the amount by which to decrement the PC after the
4642 program encounters a breakpoint. This is often the number of bytes in
4643 @code{BREAKPOINT}, though not always. For most targets this value will be 0.
4645 @item DISABLE_UNSETTABLE_BREAK (@var{addr})
4646 @findex DISABLE_UNSETTABLE_BREAK
4647 If defined, this should evaluate to 1 if @var{addr} is in a shared
4648 library in which breakpoints cannot be set and so should be disabled.
4650 @item int gdbarch_dwarf2_reg_to_regnum (@var{gdbarch}, @var{dwarf2_regnr})
4651 @findex gdbarch_dwarf2_reg_to_regnum
4652 Convert DWARF2 register number @var{dwarf2_regnr} into @value{GDBN} regnum.
4653 If not defined, no conversion will be performed.
4655 @item int gdbarch_ecoff_reg_to_regnum (@var{gdbarch}, @var{ecoff_regnr})
4656 @findex gdbarch_ecoff_reg_to_regnum
4657 Convert ECOFF register number @var{ecoff_regnr} into @value{GDBN} regnum. If
4658 not defined, no conversion will be performed.
4660 @item GCC_COMPILED_FLAG_SYMBOL
4661 @itemx GCC2_COMPILED_FLAG_SYMBOL
4662 @findex GCC2_COMPILED_FLAG_SYMBOL
4663 @findex GCC_COMPILED_FLAG_SYMBOL
4664 If defined, these are the names of the symbols that @value{GDBN} will
4665 look for to detect that GCC compiled the file. The default symbols
4666 are @code{gcc_compiled.} and @code{gcc2_compiled.},
4667 respectively. (Currently only defined for the Delta 68.)
4669 @item gdbarch_get_longjmp_target
4670 @findex gdbarch_get_longjmp_target
4671 This function determines the target PC address that @code{longjmp}
4672 will jump to, assuming that we have just stopped at a @code{longjmp}
4673 breakpoint. It takes a @code{CORE_ADDR *} as argument, and stores the
4674 target PC value through this pointer. It examines the current state
4675 of the machine as needed, typically by using a manually-determined
4676 offset into the @code{jmp_buf}. (While we might like to get the offset
4677 from the target's @file{jmpbuf.h}, that header file cannot be assumed
4678 to be available when building a cross-debugger.)
4680 @item DEPRECATED_IBM6000_TARGET
4681 @findex DEPRECATED_IBM6000_TARGET
4682 Shows that we are configured for an IBM RS/6000 system. This
4683 conditional should be eliminated (FIXME) and replaced by
4684 feature-specific macros. It was introduced in haste and we are
4685 repenting at leisure.
4687 @item I386_USE_GENERIC_WATCHPOINTS
4688 An x86-based target can define this to use the generic x86 watchpoint
4689 support; see @ref{Algorithms, I386_USE_GENERIC_WATCHPOINTS}.
4691 @item gdbarch_in_function_epilogue_p (@var{gdbarch}, @var{addr})
4692 @findex gdbarch_in_function_epilogue_p
4693 Returns non-zero if the given @var{addr} is in the epilogue of a function.
4694 The epilogue of a function is defined as the part of a function where
4695 the stack frame of the function already has been destroyed up to the
4696 final `return from function call' instruction.
4698 @item int gdbarch_in_solib_return_trampoline (@var{gdbarch}, @var{pc}, @var{name})
4699 @findex gdbarch_in_solib_return_trampoline
4700 Define this function to return nonzero if the program is stopped in the
4701 trampoline that returns from a shared library.
4703 @item target_so_ops.in_dynsym_resolve_code (@var{pc})
4704 @findex in_dynsym_resolve_code
4705 Define this to return nonzero if the program is stopped in the
4708 @item SKIP_SOLIB_RESOLVER (@var{pc})
4709 @findex SKIP_SOLIB_RESOLVER
4710 Define this to evaluate to the (nonzero) address at which execution
4711 should continue to get past the dynamic linker's symbol resolution
4712 function. A zero value indicates that it is not important or necessary
4713 to set a breakpoint to get through the dynamic linker and that single
4714 stepping will suffice.
4716 @item CORE_ADDR gdbarch_integer_to_address (@var{gdbarch}, @var{type}, @var{buf})
4717 @findex gdbarch_integer_to_address
4718 @cindex converting integers to addresses
4719 Define this when the architecture needs to handle non-pointer to address
4720 conversions specially. Converts that value to an address according to
4721 the current architectures conventions.
4723 @emph{Pragmatics: When the user copies a well defined expression from
4724 their source code and passes it, as a parameter, to @value{GDBN}'s
4725 @code{print} command, they should get the same value as would have been
4726 computed by the target program. Any deviation from this rule can cause
4727 major confusion and annoyance, and needs to be justified carefully. In
4728 other words, @value{GDBN} doesn't really have the freedom to do these
4729 conversions in clever and useful ways. It has, however, been pointed
4730 out that users aren't complaining about how @value{GDBN} casts integers
4731 to pointers; they are complaining that they can't take an address from a
4732 disassembly listing and give it to @code{x/i}. Adding an architecture
4733 method like @code{gdbarch_integer_to_address} certainly makes it possible for
4734 @value{GDBN} to ``get it right'' in all circumstances.}
4736 @xref{Target Architecture Definition, , Pointers Are Not Always
4739 @item CORE_ADDR gdbarch_pointer_to_address (@var{gdbarch}, @var{type}, @var{buf})
4740 @findex gdbarch_pointer_to_address
4741 Assume that @var{buf} holds a pointer of type @var{type}, in the
4742 appropriate format for the current architecture. Return the byte
4743 address the pointer refers to.
4744 @xref{Target Architecture Definition, , Pointers Are Not Always Addresses}.
4746 @item void gdbarch_register_to_value(@var{gdbarch}, @var{frame}, @var{regnum}, @var{type}, @var{fur})
4747 @findex gdbarch_register_to_value
4748 Convert the raw contents of register @var{regnum} into a value of type
4750 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
4752 @item REGISTER_CONVERT_TO_VIRTUAL(@var{reg}, @var{type}, @var{from}, @var{to})
4753 @findex REGISTER_CONVERT_TO_VIRTUAL
4754 Convert the value of register @var{reg} from its raw form to its virtual
4756 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
4758 @item REGISTER_CONVERT_TO_RAW(@var{type}, @var{reg}, @var{from}, @var{to})
4759 @findex REGISTER_CONVERT_TO_RAW
4760 Convert the value of register @var{reg} from its virtual form to its raw
4762 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
4764 @item const struct regset *regset_from_core_section (struct gdbarch * @var{gdbarch}, const char * @var{sect_name}, size_t @var{sect_size})
4765 @findex regset_from_core_section
4766 Return the appropriate register set for a core file section with name
4767 @var{sect_name} and size @var{sect_size}.
4769 @item SOFTWARE_SINGLE_STEP_P()
4770 @findex SOFTWARE_SINGLE_STEP_P
4771 Define this as 1 if the target does not have a hardware single-step
4772 mechanism. The macro @code{SOFTWARE_SINGLE_STEP} must also be defined.
4774 @item SOFTWARE_SINGLE_STEP(@var{signal}, @var{insert_breakpoints_p})
4775 @findex SOFTWARE_SINGLE_STEP
4776 A function that inserts or removes (depending on
4777 @var{insert_breakpoints_p}) breakpoints at each possible destinations of
4778 the next instruction. See @file{sparc-tdep.c} and @file{rs6000-tdep.c}
4781 @item set_gdbarch_sofun_address_maybe_missing (@var{gdbarch}, @var{set})
4782 @findex set_gdbarch_sofun_address_maybe_missing
4783 Somebody clever observed that, the more actual addresses you have in the
4784 debug information, the more time the linker has to spend relocating
4785 them. So whenever there's some other way the debugger could find the
4786 address it needs, you should omit it from the debug info, to make
4789 Calling @code{set_gdbarch_sofun_address_maybe_missing} with a non-zero
4790 argument @var{set} indicates that a particular set of hacks of this sort
4791 are in use, affecting @code{N_SO} and @code{N_FUN} entries in stabs-format
4792 debugging information. @code{N_SO} stabs mark the beginning and ending
4793 addresses of compilation units in the text segment. @code{N_FUN} stabs
4794 mark the starts and ends of functions.
4796 In this case, @value{GDBN} assumes two things:
4800 @code{N_FUN} stabs have an address of zero. Instead of using those
4801 addresses, you should find the address where the function starts by
4802 taking the function name from the stab, and then looking that up in the
4803 minsyms (the linker/assembler symbol table). In other words, the stab
4804 has the name, and the linker/assembler symbol table is the only place
4805 that carries the address.
4808 @code{N_SO} stabs have an address of zero, too. You just look at the
4809 @code{N_FUN} stabs that appear before and after the @code{N_SO} stab, and
4810 guess the starting and ending addresses of the compilation unit from them.
4813 @item int gdbarch_stabs_argument_has_addr (@var{gdbarch}, @var{type})
4814 @findex gdbarch_stabs_argument_has_addr
4815 @anchor{gdbarch_stabs_argument_has_addr} Define this function to return
4816 nonzero if a function argument of type @var{type} is passed by reference
4819 @item CORE_ADDR gdbarch_push_dummy_call (@var{gdbarch}, @var{function}, @var{regcache}, @var{bp_addr}, @var{nargs}, @var{args}, @var{sp}, @var{struct_return}, @var{struct_addr})
4820 @findex gdbarch_push_dummy_call
4821 @anchor{gdbarch_push_dummy_call} Define this to push the dummy frame's call to
4822 the inferior function onto the stack. In addition to pushing @var{nargs}, the
4823 code should push @var{struct_addr} (when @var{struct_return} is non-zero), and
4824 the return address (@var{bp_addr}).
4826 @var{function} is a pointer to a @code{struct value}; on architectures that use
4827 function descriptors, this contains the function descriptor value.
4829 Returns the updated top-of-stack pointer.
4831 @item CORE_ADDR gdbarch_push_dummy_code (@var{gdbarch}, @var{sp}, @var{funaddr}, @var{using_gcc}, @var{args}, @var{nargs}, @var{value_type}, @var{real_pc}, @var{bp_addr}, @var{regcache})
4832 @findex gdbarch_push_dummy_code
4833 @anchor{gdbarch_push_dummy_code} Given a stack based call dummy, push the
4834 instruction sequence (including space for a breakpoint) to which the
4835 called function should return.
4837 Set @var{bp_addr} to the address at which the breakpoint instruction
4838 should be inserted, @var{real_pc} to the resume address when starting
4839 the call sequence, and return the updated inner-most stack address.
4841 By default, the stack is grown sufficient to hold a frame-aligned
4842 (@pxref{frame_align}) breakpoint, @var{bp_addr} is set to the address
4843 reserved for that breakpoint, and @var{real_pc} set to @var{funaddr}.
4845 This method replaces @w{@code{gdbarch_call_dummy_location (@var{gdbarch})}}.
4847 @item int gdbarch_sdb_reg_to_regnum (@var{gdbarch}, @var{sdb_regnr})
4848 @findex gdbarch_sdb_reg_to_regnum
4849 Use this function to convert sdb register @var{sdb_regnr} into @value{GDBN}
4850 regnum. If not defined, no conversion will be done.
4852 @item enum return_value_convention gdbarch_return_value (struct gdbarch *@var{gdbarch}, struct type *@var{valtype}, struct regcache *@var{regcache}, void *@var{readbuf}, const void *@var{writebuf})
4853 @findex gdbarch_return_value
4854 @anchor{gdbarch_return_value} Given a function with a return-value of
4855 type @var{rettype}, return which return-value convention that function
4858 @value{GDBN} currently recognizes two function return-value conventions:
4859 @code{RETURN_VALUE_REGISTER_CONVENTION} where the return value is found
4860 in registers; and @code{RETURN_VALUE_STRUCT_CONVENTION} where the return
4861 value is found in memory and the address of that memory location is
4862 passed in as the function's first parameter.
4864 If the register convention is being used, and @var{writebuf} is
4865 non-@code{NULL}, also copy the return-value in @var{writebuf} into
4868 If the register convention is being used, and @var{readbuf} is
4869 non-@code{NULL}, also copy the return value from @var{regcache} into
4870 @var{readbuf} (@var{regcache} contains a copy of the registers from the
4871 just returned function).
4873 @emph{Maintainer note: This method replaces separate predicate, extract,
4874 store methods. By having only one method, the logic needed to determine
4875 the return-value convention need only be implemented in one place. If
4876 @value{GDBN} were written in an @sc{oo} language, this method would
4877 instead return an object that knew how to perform the register
4878 return-value extract and store.}
4880 @emph{Maintainer note: This method does not take a @var{gcc_p}
4881 parameter, and such a parameter should not be added. If an architecture
4882 that requires per-compiler or per-function information be identified,
4883 then the replacement of @var{rettype} with @code{struct value}
4884 @var{function} should be pursued.}
4886 @emph{Maintainer note: The @var{regcache} parameter limits this methods
4887 to the inner most frame. While replacing @var{regcache} with a
4888 @code{struct frame_info} @var{frame} parameter would remove that
4889 limitation there has yet to be a demonstrated need for such a change.}
4891 @item void gdbarch_skip_permanent_breakpoint (@var{gdbarch}, @var{regcache})
4892 @findex gdbarch_skip_permanent_breakpoint
4893 Advance the inferior's PC past a permanent breakpoint. @value{GDBN} normally
4894 steps over a breakpoint by removing it, stepping one instruction, and
4895 re-inserting the breakpoint. However, permanent breakpoints are
4896 hardwired into the inferior, and can't be removed, so this strategy
4897 doesn't work. Calling @code{gdbarch_skip_permanent_breakpoint} adjusts the
4898 processor's state so that execution will resume just after the breakpoint.
4899 This function does the right thing even when the breakpoint is in the delay slot
4900 of a branch or jump.
4902 @item CORE_ADDR gdbarch_skip_trampoline_code (@var{gdbarch}, @var{frame}, @var{pc})
4903 @findex gdbarch_skip_trampoline_code
4904 If the target machine has trampoline code that sits between callers and
4905 the functions being called, then define this function to return a new PC
4906 that is at the start of the real function.
4908 @item int gdbarch_deprecated_fp_regnum (@var{gdbarch})
4909 @findex gdbarch_deprecated_fp_regnum
4910 If the frame pointer is in a register, use this function to return the
4911 number of that register.
4913 @item int gdbarch_stab_reg_to_regnum (@var{gdbarch}, @var{stab_regnr})
4914 @findex gdbarch_stab_reg_to_regnum
4915 Use this function to convert stab register @var{stab_regnr} into @value{GDBN}
4916 regnum. If not defined, no conversion will be done.
4918 @item SYMBOL_RELOADING_DEFAULT
4919 @findex SYMBOL_RELOADING_DEFAULT
4920 The default value of the ``symbol-reloading'' variable. (Never defined in
4923 @item TARGET_CHAR_BIT
4924 @findex TARGET_CHAR_BIT
4925 Number of bits in a char; defaults to 8.
4927 @item int gdbarch_char_signed (@var{gdbarch})
4928 @findex gdbarch_char_signed
4929 Non-zero if @code{char} is normally signed on this architecture; zero if
4930 it should be unsigned.
4932 The ISO C standard requires the compiler to treat @code{char} as
4933 equivalent to either @code{signed char} or @code{unsigned char}; any
4934 character in the standard execution set is supposed to be positive.
4935 Most compilers treat @code{char} as signed, but @code{char} is unsigned
4936 on the IBM S/390, RS6000, and PowerPC targets.
4938 @item int gdbarch_double_bit (@var{gdbarch})
4939 @findex gdbarch_double_bit
4940 Number of bits in a double float; defaults to @w{@code{8 * TARGET_CHAR_BIT}}.
4942 @item int gdbarch_float_bit (@var{gdbarch})
4943 @findex gdbarch_float_bit
4944 Number of bits in a float; defaults to @w{@code{4 * TARGET_CHAR_BIT}}.
4946 @item int gdbarch_int_bit (@var{gdbarch})
4947 @findex gdbarch_int_bit
4948 Number of bits in an integer; defaults to @w{@code{4 * TARGET_CHAR_BIT}}.
4950 @item int gdbarch_long_bit (@var{gdbarch})
4951 @findex gdbarch_long_bit
4952 Number of bits in a long integer; defaults to @w{@code{4 * TARGET_CHAR_BIT}}.
4954 @item int gdbarch_long_double_bit (@var{gdbarch})
4955 @findex gdbarch_long_double_bit
4956 Number of bits in a long double float;
4957 defaults to @w{@code{2 * gdbarch_double_bit (@var{gdbarch})}}.
4959 @item int gdbarch_long_long_bit (@var{gdbarch})
4960 @findex gdbarch_long_long_bit
4961 Number of bits in a long long integer; defaults to
4962 @w{@code{2 * gdbarch_long_bit (@var{gdbarch})}}.
4964 @item int gdbarch_ptr_bit (@var{gdbarch})
4965 @findex gdbarch_ptr_bit
4966 Number of bits in a pointer; defaults to
4967 @w{@code{gdbarch_int_bit (@var{gdbarch})}}.
4969 @item int gdbarch_short_bit (@var{gdbarch})
4970 @findex gdbarch_short_bit
4971 Number of bits in a short integer; defaults to @w{@code{2 * TARGET_CHAR_BIT}}.
4973 @item void gdbarch_virtual_frame_pointer (@var{gdbarch}, @var{pc}, @var{frame_regnum}, @var{frame_offset})
4974 @findex gdbarch_virtual_frame_pointer
4975 Returns a @code{(@var{register}, @var{offset})} pair representing the virtual
4976 frame pointer in use at the code address @var{pc}. If virtual frame
4977 pointers are not used, a default definition simply returns
4978 @code{gdbarch_deprecated_fp_regnum} (or @code{gdbarch_sp_regnum}, if
4979 no frame pointer is defined), with an offset of zero.
4981 @c need to explain virtual frame pointers, they are recorded in agent
4982 @c expressions for tracepoints
4984 @item TARGET_HAS_HARDWARE_WATCHPOINTS
4985 If non-zero, the target has support for hardware-assisted
4986 watchpoints. @xref{Algorithms, watchpoints}, for more details and
4987 other related macros.
4989 @item int gdbarch_print_insn (@var{gdbarch}, @var{vma}, @var{info})
4990 @findex gdbarch_print_insn
4991 This is the function used by @value{GDBN} to print an assembly
4992 instruction. It prints the instruction at address @var{vma} in
4993 debugged memory and returns the length of the instruction, in bytes.
4994 This usually points to a function in the @code{opcodes} library
4995 (@pxref{Support Libraries, ,Opcodes}). @var{info} is a structure (of
4996 type @code{disassemble_info}) defined in the header file
4997 @file{include/dis-asm.h}, and used to pass information to the
4998 instruction decoding routine.
5000 @item frame_id gdbarch_dummy_id (@var{gdbarch}, @var{frame})
5001 @findex gdbarch_dummy_id
5002 @anchor{gdbarch_dummy_id} Given @var{frame} return a @w{@code{struct
5003 frame_id}} that uniquely identifies an inferior function call's dummy
5004 frame. The value returned must match the dummy frame stack value
5005 previously saved by @code{call_function_by_hand}.
5007 @item void gdbarch_value_to_register (@var{gdbarch}, @var{frame}, @var{type}, @var{buf})
5008 @findex gdbarch_value_to_register
5009 Convert a value of type @var{type} into the raw contents of a register.
5010 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
5014 Motorola M68K target conditionals.
5018 Define this to be the 4-bit location of the breakpoint trap vector. If
5019 not defined, it will default to @code{0xf}.
5021 @item REMOTE_BPT_VECTOR
5022 Defaults to @code{1}.
5026 @node Adding a New Target
5027 @section Adding a New Target
5029 @cindex adding a target
5030 The following files add a target to @value{GDBN}:
5033 @cindex target dependent files
5035 @item gdb/@var{ttt}-tdep.c
5036 Contains any miscellaneous code required for this target machine. On
5037 some machines it doesn't exist at all.
5039 @item gdb/@var{arch}-tdep.c
5040 @itemx gdb/@var{arch}-tdep.h
5041 This is required to describe the basic layout of the target machine's
5042 processor chip (registers, stack, etc.). It can be shared among many
5043 targets that use the same processor architecture.
5047 (Target header files such as
5048 @file{gdb/config/@var{arch}/tm-@var{ttt}.h},
5049 @file{gdb/config/@var{arch}/tm-@var{arch}.h}, and
5050 @file{config/tm-@var{os}.h} are no longer used.)
5052 @findex _initialize_@var{arch}_tdep
5053 A @value{GDBN} description for a new architecture, arch is created by
5054 defining a global function @code{_initialize_@var{arch}_tdep}, by
5055 convention in the source file @file{@var{arch}-tdep.c}. For
5056 example, in the case of the OpenRISC 1000, this function is called
5057 @code{_initialize_or1k_tdep} and is found in the file
5060 The object file resulting from compiling this source file, which will
5061 contain the implementation of the
5062 @code{_initialize_@var{arch}_tdep} function is specified in the
5063 @value{GDBN} @file{configure.tgt} file, which includes a large case
5064 statement pattern matching against the @code{--target} option of the
5065 @kbd{configure} script.
5068 @emph{Note:} If the architecture requires multiple source files, the
5069 corresponding binaries should be included in
5070 @file{configure.tgt}. However if there are header files, the
5071 dependencies on these will not be picked up from the entries in
5072 @file{configure.tgt}. The @file{Makefile.in} file will need extending to
5073 show these dependencies.
5076 @findex gdbarch_register
5077 A new struct gdbarch, defining the new architecture, is created within
5078 the @code{_initialize_@var{arch}_tdep} function by calling
5079 @code{gdbarch_register}:
5082 void gdbarch_register (enum bfd_architecture architecture,
5083 gdbarch_init_ftype *init_func,
5084 gdbarch_dump_tdep_ftype *tdep_dump_func);
5087 This function has been described fully in an earlier
5088 section. @xref{How an Architecture is Represented, , How an
5089 Architecture is Represented}.
5091 The new @code{@w{struct gdbarch}} should contain implementations of
5092 the necessary functions (described in the previous sections) to
5093 describe the basic layout of the target machine's processor chip
5094 (registers, stack, etc.). It can be shared among many targets that use
5095 the same processor architecture.
5097 @node Target Descriptions
5098 @chapter Target Descriptions
5099 @cindex target descriptions
5101 The target architecture definition (@pxref{Target Architecture Definition})
5102 contains @value{GDBN}'s hard-coded knowledge about an architecture. For
5103 some platforms, it is handy to have more flexible knowledge about a specific
5104 instance of the architecture---for instance, a processor or development board.
5105 @dfn{Target descriptions} provide a mechanism for the user to tell @value{GDBN}
5106 more about what their target supports, or for the target to tell @value{GDBN}
5109 For details on writing, automatically supplying, and manually selecting
5110 target descriptions, see @ref{Target Descriptions, , , gdb,
5111 Debugging with @value{GDBN}}. This section will cover some related
5112 topics about the @value{GDBN} internals.
5115 * Target Descriptions Implementation::
5116 * Adding Target Described Register Support::
5119 @node Target Descriptions Implementation
5120 @section Target Descriptions Implementation
5121 @cindex target descriptions, implementation
5123 Before @value{GDBN} connects to a new target, or runs a new program on
5124 an existing target, it discards any existing target description and
5125 reverts to a default gdbarch. Then, after connecting, it looks for a
5126 new target description by calling @code{target_find_description}.
5128 A description may come from a user specified file (XML), the remote
5129 @samp{qXfer:features:read} packet (also XML), or from any custom
5130 @code{to_read_description} routine in the target vector. For instance,
5131 the remote target supports guessing whether a MIPS target is 32-bit or
5132 64-bit based on the size of the @samp{g} packet.
5134 If any target description is found, @value{GDBN} creates a new gdbarch
5135 incorporating the description by calling @code{gdbarch_update_p}. Any
5136 @samp{<architecture>} element is handled first, to determine which
5137 architecture's gdbarch initialization routine is called to create the
5138 new architecture. Then the initialization routine is called, and has
5139 a chance to adjust the constructed architecture based on the contents
5140 of the target description. For instance, it can recognize any
5141 properties set by a @code{to_read_description} routine. Also
5142 see @ref{Adding Target Described Register Support}.
5144 @node Adding Target Described Register Support
5145 @section Adding Target Described Register Support
5146 @cindex target descriptions, adding register support
5148 Target descriptions can report additional registers specific to an
5149 instance of the target. But it takes a little work in the architecture
5150 specific routines to support this.
5152 A target description must either have no registers or a complete
5153 set---this avoids complexity in trying to merge standard registers
5154 with the target defined registers. It is the architecture's
5155 responsibility to validate that a description with registers has
5156 everything it needs. To keep architecture code simple, the same
5157 mechanism is used to assign fixed internal register numbers to
5160 If @code{tdesc_has_registers} returns 1, the description contains
5161 registers. The architecture's @code{gdbarch_init} routine should:
5166 Call @code{tdesc_data_alloc} to allocate storage, early, before
5167 searching for a matching gdbarch or allocating a new one.
5170 Use @code{tdesc_find_feature} to locate standard features by name.
5173 Use @code{tdesc_numbered_register} and @code{tdesc_numbered_register_choices}
5174 to locate the expected registers in the standard features.
5177 Return @code{NULL} if a required feature is missing, or if any standard
5178 feature is missing expected registers. This will produce a warning that
5179 the description was incomplete.
5182 Free the allocated data before returning, unless @code{tdesc_use_registers}
5186 Call @code{set_gdbarch_num_regs} as usual, with a number higher than any
5187 fixed number passed to @code{tdesc_numbered_register}.
5190 Call @code{tdesc_use_registers} after creating a new gdbarch, before
5195 After @code{tdesc_use_registers} has been called, the architecture's
5196 @code{register_name}, @code{register_type}, and @code{register_reggroup_p}
5197 routines will not be called; that information will be taken from
5198 the target description. @code{num_regs} may be increased to account
5199 for any additional registers in the description.
5201 Pseudo-registers require some extra care:
5206 Using @code{tdesc_numbered_register} allows the architecture to give
5207 constant register numbers to standard architectural registers, e.g.@:
5208 as an @code{enum} in @file{@var{arch}-tdep.h}. But because
5209 pseudo-registers are always numbered above @code{num_regs},
5210 which may be increased by the description, constant numbers
5211 can not be used for pseudos. They must be numbered relative to
5212 @code{num_regs} instead.
5215 The description will not describe pseudo-registers, so the
5216 architecture must call @code{set_tdesc_pseudo_register_name},
5217 @code{set_tdesc_pseudo_register_type}, and
5218 @code{set_tdesc_pseudo_register_reggroup_p} to supply routines
5219 describing pseudo registers. These routines will be passed
5220 internal register numbers, so the same routines used for the
5221 gdbarch equivalents are usually suitable.
5226 @node Target Vector Definition
5228 @chapter Target Vector Definition
5229 @cindex target vector
5231 The target vector defines the interface between @value{GDBN}'s
5232 abstract handling of target systems, and the nitty-gritty code that
5233 actually exercises control over a process or a serial port.
5234 @value{GDBN} includes some 30-40 different target vectors; however,
5235 each configuration of @value{GDBN} includes only a few of them.
5238 * Managing Execution State::
5239 * Existing Targets::
5242 @node Managing Execution State
5243 @section Managing Execution State
5244 @cindex execution state
5246 A target vector can be completely inactive (not pushed on the target
5247 stack), active but not running (pushed, but not connected to a fully
5248 manifested inferior), or completely active (pushed, with an accessible
5249 inferior). Most targets are only completely inactive or completely
5250 active, but some support persistent connections to a target even
5251 when the target has exited or not yet started.
5253 For example, connecting to the simulator using @code{target sim} does
5254 not create a running program. Neither registers nor memory are
5255 accessible until @code{run}. Similarly, after @code{kill}, the
5256 program can not continue executing. But in both cases @value{GDBN}
5257 remains connected to the simulator, and target-specific commands
5258 are directed to the simulator.
5260 A target which only supports complete activation should push itself
5261 onto the stack in its @code{to_open} routine (by calling
5262 @code{push_target}), and unpush itself from the stack in its
5263 @code{to_mourn_inferior} routine (by calling @code{unpush_target}).
5265 A target which supports both partial and complete activation should
5266 still call @code{push_target} in @code{to_open}, but not call
5267 @code{unpush_target} in @code{to_mourn_inferior}. Instead, it should
5268 call either @code{target_mark_running} or @code{target_mark_exited}
5269 in its @code{to_open}, depending on whether the target is fully active
5270 after connection. It should also call @code{target_mark_running} any
5271 time the inferior becomes fully active (e.g.@: in
5272 @code{to_create_inferior} and @code{to_attach}), and
5273 @code{target_mark_exited} when the inferior becomes inactive (in
5274 @code{to_mourn_inferior}). The target should also make sure to call
5275 @code{target_mourn_inferior} from its @code{to_kill}, to return the
5276 target to inactive state.
5278 @node Existing Targets
5279 @section Existing Targets
5282 @subsection File Targets
5284 Both executables and core files have target vectors.
5286 @subsection Standard Protocol and Remote Stubs
5288 @value{GDBN}'s file @file{remote.c} talks a serial protocol to code that
5289 runs in the target system. @value{GDBN} provides several sample
5290 @dfn{stubs} that can be integrated into target programs or operating
5291 systems for this purpose; they are named @file{@var{cpu}-stub.c}. Many
5292 operating systems, embedded targets, emulators, and simulators already
5293 have a @value{GDBN} stub built into them, and maintenance of the remote
5294 protocol must be careful to preserve compatibility.
5296 The @value{GDBN} user's manual describes how to put such a stub into
5297 your target code. What follows is a discussion of integrating the
5298 SPARC stub into a complicated operating system (rather than a simple
5299 program), by Stu Grossman, the author of this stub.
5301 The trap handling code in the stub assumes the following upon entry to
5306 %l1 and %l2 contain pc and npc respectively at the time of the trap;
5312 you are in the correct trap window.
5315 As long as your trap handler can guarantee those conditions, then there
5316 is no reason why you shouldn't be able to ``share'' traps with the stub.
5317 The stub has no requirement that it be jumped to directly from the
5318 hardware trap vector. That is why it calls @code{exceptionHandler()},
5319 which is provided by the external environment. For instance, this could
5320 set up the hardware traps to actually execute code which calls the stub
5321 first, and then transfers to its own trap handler.
5323 For the most point, there probably won't be much of an issue with
5324 ``sharing'' traps, as the traps we use are usually not used by the kernel,
5325 and often indicate unrecoverable error conditions. Anyway, this is all
5326 controlled by a table, and is trivial to modify. The most important
5327 trap for us is for @code{ta 1}. Without that, we can't single step or
5328 do breakpoints. Everything else is unnecessary for the proper operation
5329 of the debugger/stub.
5331 From reading the stub, it's probably not obvious how breakpoints work.
5332 They are simply done by deposit/examine operations from @value{GDBN}.
5334 @subsection ROM Monitor Interface
5336 @subsection Custom Protocols
5338 @subsection Transport Layer
5340 @subsection Builtin Simulator
5343 @node Native Debugging
5345 @chapter Native Debugging
5346 @cindex native debugging
5348 Several files control @value{GDBN}'s configuration for native support:
5352 @item gdb/config/@var{arch}/@var{xyz}.mh
5353 Specifies Makefile fragments needed by a @emph{native} configuration on
5354 machine @var{xyz}. In particular, this lists the required
5355 native-dependent object files, by defining @samp{NATDEPFILES=@dots{}}.
5356 Also specifies the header file which describes native support on
5357 @var{xyz}, by defining @samp{NAT_FILE= nm-@var{xyz}.h}. You can also
5358 define @samp{NAT_CFLAGS}, @samp{NAT_ADD_FILES}, @samp{NAT_CLIBS},
5359 @samp{NAT_CDEPS}, @samp{NAT_GENERATED_FILES}, etc.; see @file{Makefile.in}.
5361 @emph{Maintainer's note: The @file{.mh} suffix is because this file
5362 originally contained @file{Makefile} fragments for hosting @value{GDBN}
5363 on machine @var{xyz}. While the file is no longer used for this
5364 purpose, the @file{.mh} suffix remains. Perhaps someone will
5365 eventually rename these fragments so that they have a @file{.mn}
5368 @item gdb/config/@var{arch}/nm-@var{xyz}.h
5369 (@file{nm.h} is a link to this file, created by @code{configure}). Contains C
5370 macro definitions describing the native system environment, such as
5371 child process control and core file support.
5373 @item gdb/@var{xyz}-nat.c
5374 Contains any miscellaneous C code required for this native support of
5375 this machine. On some machines it doesn't exist at all.
5378 There are some ``generic'' versions of routines that can be used by
5379 various systems. These can be customized in various ways by macros
5380 defined in your @file{nm-@var{xyz}.h} file. If these routines work for
5381 the @var{xyz} host, you can just include the generic file's name (with
5382 @samp{.o}, not @samp{.c}) in @code{NATDEPFILES}.
5384 Otherwise, if your machine needs custom support routines, you will need
5385 to write routines that perform the same functions as the generic file.
5386 Put them into @file{@var{xyz}-nat.c}, and put @file{@var{xyz}-nat.o}
5387 into @code{NATDEPFILES}.
5391 This contains the @emph{target_ops vector} that supports Unix child
5392 processes on systems which use ptrace and wait to control the child.
5395 This contains the @emph{target_ops vector} that supports Unix child
5396 processes on systems which use /proc to control the child.
5399 This does the low-level grunge that uses Unix system calls to do a ``fork
5400 and exec'' to start up a child process.
5403 This is the low level interface to inferior processes for systems using
5404 the Unix @code{ptrace} call in a vanilla way.
5413 @section shared libraries
5415 @section Native Conditionals
5416 @cindex native conditionals
5418 When @value{GDBN} is configured and compiled, various macros are
5419 defined or left undefined, to control compilation when the host and
5420 target systems are the same. These macros should be defined (or left
5421 undefined) in @file{nm-@var{system}.h}.
5425 @item I386_USE_GENERIC_WATCHPOINTS
5426 An x86-based machine can define this to use the generic x86 watchpoint
5427 support; see @ref{Algorithms, I386_USE_GENERIC_WATCHPOINTS}.
5429 @item SOLIB_ADD (@var{filename}, @var{from_tty}, @var{targ}, @var{readsyms})
5431 Define this to expand into an expression that will cause the symbols in
5432 @var{filename} to be added to @value{GDBN}'s symbol table. If
5433 @var{readsyms} is zero symbols are not read but any necessary low level
5434 processing for @var{filename} is still done.
5436 @item SOLIB_CREATE_INFERIOR_HOOK
5437 @findex SOLIB_CREATE_INFERIOR_HOOK
5438 Define this to expand into any shared-library-relocation code that you
5439 want to be run just after the child process has been forked.
5441 @item START_INFERIOR_TRAPS_EXPECTED
5442 @findex START_INFERIOR_TRAPS_EXPECTED
5443 When starting an inferior, @value{GDBN} normally expects to trap
5445 the shell execs, and once when the program itself execs. If the actual
5446 number of traps is something other than 2, then define this macro to
5447 expand into the number expected.
5451 @node Support Libraries
5453 @chapter Support Libraries
5458 BFD provides support for @value{GDBN} in several ways:
5461 @item identifying executable and core files
5462 BFD will identify a variety of file types, including a.out, coff, and
5463 several variants thereof, as well as several kinds of core files.
5465 @item access to sections of files
5466 BFD parses the file headers to determine the names, virtual addresses,
5467 sizes, and file locations of all the various named sections in files
5468 (such as the text section or the data section). @value{GDBN} simply
5469 calls BFD to read or write section @var{x} at byte offset @var{y} for
5472 @item specialized core file support
5473 BFD provides routines to determine the failing command name stored in a
5474 core file, the signal with which the program failed, and whether a core
5475 file matches (i.e.@: could be a core dump of) a particular executable
5478 @item locating the symbol information
5479 @value{GDBN} uses an internal interface of BFD to determine where to find the
5480 symbol information in an executable file or symbol-file. @value{GDBN} itself
5481 handles the reading of symbols, since BFD does not ``understand'' debug
5482 symbols, but @value{GDBN} uses BFD's cached information to find the symbols,
5487 @cindex opcodes library
5489 The opcodes library provides @value{GDBN}'s disassembler. (It's a separate
5490 library because it's also used in binutils, for @file{objdump}).
5493 @cindex readline library
5494 The @code{readline} library provides a set of functions for use by applications
5495 that allow users to edit command lines as they are typed in.
5498 @cindex @code{libiberty} library
5500 The @code{libiberty} library provides a set of functions and features
5501 that integrate and improve on functionality found in modern operating
5502 systems. Broadly speaking, such features can be divided into three
5503 groups: supplemental functions (functions that may be missing in some
5504 environments and operating systems), replacement functions (providing
5505 a uniform and easier to use interface for commonly used standard
5506 functions), and extensions (which provide additional functionality
5507 beyond standard functions).
5509 @value{GDBN} uses various features provided by the @code{libiberty}
5510 library, for instance the C@t{++} demangler, the @acronym{IEEE}
5511 floating format support functions, the input options parser
5512 @samp{getopt}, the @samp{obstack} extension, and other functions.
5514 @subsection @code{obstacks} in @value{GDBN}
5515 @cindex @code{obstacks}
5517 The obstack mechanism provides a convenient way to allocate and free
5518 chunks of memory. Each obstack is a pool of memory that is managed
5519 like a stack. Objects (of any nature, size and alignment) are
5520 allocated and freed in a @acronym{LIFO} fashion on an obstack (see
5521 @code{libiberty}'s documentation for a more detailed explanation of
5524 The most noticeable use of the @code{obstacks} in @value{GDBN} is in
5525 object files. There is an obstack associated with each internal
5526 representation of an object file. Lots of things get allocated on
5527 these @code{obstacks}: dictionary entries, blocks, blockvectors,
5528 symbols, minimal symbols, types, vectors of fundamental types, class
5529 fields of types, object files section lists, object files section
5530 offset lists, line tables, symbol tables, partial symbol tables,
5531 string tables, symbol table private data, macros tables, debug
5532 information sections and entries, import and export lists (som),
5533 unwind information (hppa), dwarf2 location expressions data. Plus
5534 various strings such as directory names strings, debug format strings,
5537 An essential and convenient property of all data on @code{obstacks} is
5538 that memory for it gets allocated (with @code{obstack_alloc}) at
5539 various times during a debugging session, but it is released all at
5540 once using the @code{obstack_free} function. The @code{obstack_free}
5541 function takes a pointer to where in the stack it must start the
5542 deletion from (much like the cleanup chains have a pointer to where to
5543 start the cleanups). Because of the stack like structure of the
5544 @code{obstacks}, this allows to free only a top portion of the
5545 obstack. There are a few instances in @value{GDBN} where such thing
5546 happens. Calls to @code{obstack_free} are done after some local data
5547 is allocated to the obstack. Only the local data is deleted from the
5548 obstack. Of course this assumes that nothing between the
5549 @code{obstack_alloc} and the @code{obstack_free} allocates anything
5550 else on the same obstack. For this reason it is best and safest to
5551 use temporary @code{obstacks}.
5553 Releasing the whole obstack is also not safe per se. It is safe only
5554 under the condition that we know the @code{obstacks} memory is no
5555 longer needed. In @value{GDBN} we get rid of the @code{obstacks} only
5556 when we get rid of the whole objfile(s), for instance upon reading a
5560 @cindex regular expressions library
5571 @item SIGN_EXTEND_CHAR
5573 @item SWITCH_ENUM_BUG
5582 @section Array Containers
5583 @cindex Array Containers
5586 Often it is necessary to manipulate a dynamic array of a set of
5587 objects. C forces some bookkeeping on this, which can get cumbersome
5588 and repetitive. The @file{vec.h} file contains macros for defining
5589 and using a typesafe vector type. The functions defined will be
5590 inlined when compiling, and so the abstraction cost should be zero.
5591 Domain checks are added to detect programming errors.
5593 An example use would be an array of symbols or section information.
5594 The array can be grown as symbols are read in (or preallocated), and
5595 the accessor macros provided keep care of all the necessary
5596 bookkeeping. Because the arrays are type safe, there is no danger of
5597 accidentally mixing up the contents. Think of these as C++ templates,
5598 but implemented in C.
5600 Because of the different behavior of structure objects, scalar objects
5601 and of pointers, there are three flavors of vector, one for each of
5602 these variants. Both the structure object and pointer variants pass
5603 pointers to objects around --- in the former case the pointers are
5604 stored into the vector and in the latter case the pointers are
5605 dereferenced and the objects copied into the vector. The scalar
5606 object variant is suitable for @code{int}-like objects, and the vector
5607 elements are returned by value.
5609 There are both @code{index} and @code{iterate} accessors. The iterator
5610 returns a boolean iteration condition and updates the iteration
5611 variable passed by reference. Because the iterator will be inlined,
5612 the address-of can be optimized away.
5614 The vectors are implemented using the trailing array idiom, thus they
5615 are not resizeable without changing the address of the vector object
5616 itself. This means you cannot have variables or fields of vector type
5617 --- always use a pointer to a vector. The one exception is the final
5618 field of a structure, which could be a vector type. You will have to
5619 use the @code{embedded_size} & @code{embedded_init} calls to create
5620 such objects, and they will probably not be resizeable (so don't use
5621 the @dfn{safe} allocation variants). The trailing array idiom is used
5622 (rather than a pointer to an array of data), because, if we allow
5623 @code{NULL} to also represent an empty vector, empty vectors occupy
5624 minimal space in the structure containing them.
5626 Each operation that increases the number of active elements is
5627 available in @dfn{quick} and @dfn{safe} variants. The former presumes
5628 that there is sufficient allocated space for the operation to succeed
5629 (it dies if there is not). The latter will reallocate the vector, if
5630 needed. Reallocation causes an exponential increase in vector size.
5631 If you know you will be adding N elements, it would be more efficient
5632 to use the reserve operation before adding the elements with the
5633 @dfn{quick} operation. This will ensure there are at least as many
5634 elements as you ask for, it will exponentially increase if there are
5635 too few spare slots. If you want reserve a specific number of slots,
5636 but do not want the exponential increase (for instance, you know this
5637 is the last allocation), use a negative number for reservation. You
5638 can also create a vector of a specific size from the get go.
5640 You should prefer the push and pop operations, as they append and
5641 remove from the end of the vector. If you need to remove several items
5642 in one go, use the truncate operation. The insert and remove
5643 operations allow you to change elements in the middle of the vector.
5644 There are two remove operations, one which preserves the element
5645 ordering @code{ordered_remove}, and one which does not
5646 @code{unordered_remove}. The latter function copies the end element
5647 into the removed slot, rather than invoke a memmove operation. The
5648 @code{lower_bound} function will determine where to place an item in
5649 the array using insert that will maintain sorted order.
5651 If you need to directly manipulate a vector, then the @code{address}
5652 accessor will return the address of the start of the vector. Also the
5653 @code{space} predicate will tell you whether there is spare capacity in the
5654 vector. You will not normally need to use these two functions.
5656 Vector types are defined using a
5657 @code{DEF_VEC_@{O,P,I@}(@var{typename})} macro. Variables of vector
5658 type are declared using a @code{VEC(@var{typename})} macro. The
5659 characters @code{O}, @code{P} and @code{I} indicate whether
5660 @var{typename} is an object (@code{O}), pointer (@code{P}) or integral
5661 (@code{I}) type. Be careful to pick the correct one, as you'll get an
5662 awkward and inefficient API if you use the wrong one. There is a
5663 check, which results in a compile-time warning, for the @code{P} and
5664 @code{I} versions, but there is no check for the @code{O} versions, as
5665 that is not possible in plain C.
5667 An example of their use would be,
5670 DEF_VEC_P(tree); // non-managed tree vector.
5673 VEC(tree) *v; // A (pointer to) a vector of tree pointers.
5676 struct my_struct *s;
5678 if (VEC_length(tree, s->v)) @{ we have some contents @}
5679 VEC_safe_push(tree, s->v, decl); // append some decl onto the end
5680 for (ix = 0; VEC_iterate(tree, s->v, ix, elt); ix++)
5681 @{ do something with elt @}
5685 The @file{vec.h} file provides details on how to invoke the various
5686 accessors provided. They are enumerated here:
5690 Return the number of items in the array,
5693 Return true if the array has no elements.
5697 Return the last or arbitrary item in the array.
5700 Access an array element and indicate whether the array has been
5705 Create and destroy an array.
5707 @item VEC_embedded_size
5708 @itemx VEC_embedded_init
5709 Helpers for embedding an array as the final element of another struct.
5715 Return the amount of free space in an array.
5718 Ensure a certain amount of free space.
5720 @item VEC_quick_push
5721 @itemx VEC_safe_push
5722 Append to an array, either assuming the space is available, or making
5726 Remove the last item from an array.
5729 Remove several items from the end of an array.
5732 Add several items to the end of an array.
5735 Overwrite an item in the array.
5737 @item VEC_quick_insert
5738 @itemx VEC_safe_insert
5739 Insert an item into the middle of the array. Either the space must
5740 already exist, or the space is created.
5742 @item VEC_ordered_remove
5743 @itemx VEC_unordered_remove
5744 Remove an item from the array, preserving order or not.
5746 @item VEC_block_remove
5747 Remove a set of items from the array.
5750 Provide the address of the first element.
5752 @item VEC_lower_bound
5753 Binary search the array.
5763 This chapter covers topics that are lower-level than the major
5764 algorithms of @value{GDBN}.
5769 Cleanups are a structured way to deal with things that need to be done
5772 When your code does something (e.g., @code{xmalloc} some memory, or
5773 @code{open} a file) that needs to be undone later (e.g., @code{xfree}
5774 the memory or @code{close} the file), it can make a cleanup. The
5775 cleanup will be done at some future point: when the command is finished
5776 and control returns to the top level; when an error occurs and the stack
5777 is unwound; or when your code decides it's time to explicitly perform
5778 cleanups. Alternatively you can elect to discard the cleanups you
5784 @item struct cleanup *@var{old_chain};
5785 Declare a variable which will hold a cleanup chain handle.
5787 @findex make_cleanup
5788 @item @var{old_chain} = make_cleanup (@var{function}, @var{arg});
5789 Make a cleanup which will cause @var{function} to be called with
5790 @var{arg} (a @code{char *}) later. The result, @var{old_chain}, is a
5791 handle that can later be passed to @code{do_cleanups} or
5792 @code{discard_cleanups}. Unless you are going to call
5793 @code{do_cleanups} or @code{discard_cleanups}, you can ignore the result
5794 from @code{make_cleanup}.
5797 @item do_cleanups (@var{old_chain});
5798 Do all cleanups added to the chain since the corresponding
5799 @code{make_cleanup} call was made.
5801 @findex discard_cleanups
5802 @item discard_cleanups (@var{old_chain});
5803 Same as @code{do_cleanups} except that it just removes the cleanups from
5804 the chain and does not call the specified functions.
5807 Cleanups are implemented as a chain. The handle returned by
5808 @code{make_cleanups} includes the cleanup passed to the call and any
5809 later cleanups appended to the chain (but not yet discarded or
5813 make_cleanup (a, 0);
5815 struct cleanup *old = make_cleanup (b, 0);
5823 will call @code{c()} and @code{b()} but will not call @code{a()}. The
5824 cleanup that calls @code{a()} will remain in the cleanup chain, and will
5825 be done later unless otherwise discarded.@refill
5827 Your function should explicitly do or discard the cleanups it creates.
5828 Failing to do this leads to non-deterministic behavior since the caller
5829 will arbitrarily do or discard your functions cleanups. This need leads
5830 to two common cleanup styles.
5832 The first style is try/finally. Before it exits, your code-block calls
5833 @code{do_cleanups} with the old cleanup chain and thus ensures that your
5834 code-block's cleanups are always performed. For instance, the following
5835 code-segment avoids a memory leak problem (even when @code{error} is
5836 called and a forced stack unwind occurs) by ensuring that the
5837 @code{xfree} will always be called:
5840 struct cleanup *old = make_cleanup (null_cleanup, 0);
5841 data = xmalloc (sizeof blah);
5842 make_cleanup (xfree, data);
5847 The second style is try/except. Before it exits, your code-block calls
5848 @code{discard_cleanups} with the old cleanup chain and thus ensures that
5849 any created cleanups are not performed. For instance, the following
5850 code segment, ensures that the file will be closed but only if there is
5854 FILE *file = fopen ("afile", "r");
5855 struct cleanup *old = make_cleanup (close_file, file);
5857 discard_cleanups (old);
5861 Some functions, e.g., @code{fputs_filtered()} or @code{error()}, specify
5862 that they ``should not be called when cleanups are not in place''. This
5863 means that any actions you need to reverse in the case of an error or
5864 interruption must be on the cleanup chain before you call these
5865 functions, since they might never return to your code (they
5866 @samp{longjmp} instead).
5868 @section Per-architecture module data
5869 @cindex per-architecture module data
5870 @cindex multi-arch data
5871 @cindex data-pointer, per-architecture/per-module
5873 The multi-arch framework includes a mechanism for adding module
5874 specific per-architecture data-pointers to the @code{struct gdbarch}
5875 architecture object.
5877 A module registers one or more per-architecture data-pointers using:
5879 @deftypefn {Architecture Function} {struct gdbarch_data *} gdbarch_data_register_pre_init (gdbarch_data_pre_init_ftype *@var{pre_init})
5880 @var{pre_init} is used to, on-demand, allocate an initial value for a
5881 per-architecture data-pointer using the architecture's obstack (passed
5882 in as a parameter). Since @var{pre_init} can be called during
5883 architecture creation, it is not parameterized with the architecture.
5884 and must not call modules that use per-architecture data.
5887 @deftypefn {Architecture Function} {struct gdbarch_data *} gdbarch_data_register_post_init (gdbarch_data_post_init_ftype *@var{post_init})
5888 @var{post_init} is used to obtain an initial value for a
5889 per-architecture data-pointer @emph{after}. Since @var{post_init} is
5890 always called after architecture creation, it both receives the fully
5891 initialized architecture and is free to call modules that use
5892 per-architecture data (care needs to be taken to ensure that those
5893 other modules do not try to call back to this module as that will
5894 create in cycles in the initialization call graph).
5897 These functions return a @code{struct gdbarch_data} that is used to
5898 identify the per-architecture data-pointer added for that module.
5900 The per-architecture data-pointer is accessed using the function:
5902 @deftypefn {Architecture Function} {void *} gdbarch_data (struct gdbarch *@var{gdbarch}, struct gdbarch_data *@var{data_handle})
5903 Given the architecture @var{arch} and module data handle
5904 @var{data_handle} (returned by @code{gdbarch_data_register_pre_init}
5905 or @code{gdbarch_data_register_post_init}), this function returns the
5906 current value of the per-architecture data-pointer. If the data
5907 pointer is @code{NULL}, it is first initialized by calling the
5908 corresponding @var{pre_init} or @var{post_init} method.
5911 The examples below assume the following definitions:
5914 struct nozel @{ int total; @};
5915 static struct gdbarch_data *nozel_handle;
5918 A module can extend the architecture vector, adding additional
5919 per-architecture data, using the @var{pre_init} method. The module's
5920 per-architecture data is then initialized during architecture
5923 In the below, the module's per-architecture @emph{nozel} is added. An
5924 architecture can specify its nozel by calling @code{set_gdbarch_nozel}
5925 from @code{gdbarch_init}.
5929 nozel_pre_init (struct obstack *obstack)
5931 struct nozel *data = OBSTACK_ZALLOC (obstack, struct nozel);
5938 set_gdbarch_nozel (struct gdbarch *gdbarch, int total)
5940 struct nozel *data = gdbarch_data (gdbarch, nozel_handle);
5941 data->total = nozel;
5945 A module can on-demand create architecture dependent data structures
5946 using @code{post_init}.
5948 In the below, the nozel's total is computed on-demand by
5949 @code{nozel_post_init} using information obtained from the
5954 nozel_post_init (struct gdbarch *gdbarch)
5956 struct nozel *data = GDBARCH_OBSTACK_ZALLOC (gdbarch, struct nozel);
5957 nozel->total = gdbarch@dots{} (gdbarch);
5964 nozel_total (struct gdbarch *gdbarch)
5966 struct nozel *data = gdbarch_data (gdbarch, nozel_handle);
5971 @section Wrapping Output Lines
5972 @cindex line wrap in output
5975 Output that goes through @code{printf_filtered} or @code{fputs_filtered}
5976 or @code{fputs_demangled} needs only to have calls to @code{wrap_here}
5977 added in places that would be good breaking points. The utility
5978 routines will take care of actually wrapping if the line width is
5981 The argument to @code{wrap_here} is an indentation string which is
5982 printed @emph{only} if the line breaks there. This argument is saved
5983 away and used later. It must remain valid until the next call to
5984 @code{wrap_here} or until a newline has been printed through the
5985 @code{*_filtered} functions. Don't pass in a local variable and then
5988 It is usually best to call @code{wrap_here} after printing a comma or
5989 space. If you call it before printing a space, make sure that your
5990 indentation properly accounts for the leading space that will print if
5991 the line wraps there.
5993 Any function or set of functions that produce filtered output must
5994 finish by printing a newline, to flush the wrap buffer, before switching
5995 to unfiltered (@code{printf}) output. Symbol reading routines that
5996 print warnings are a good example.
5998 @section @value{GDBN} Coding Standards
5999 @cindex coding standards
6001 @value{GDBN} follows the GNU coding standards, as described in
6002 @file{etc/standards.texi}. This file is also available for anonymous
6003 FTP from GNU archive sites. @value{GDBN} takes a strict interpretation
6004 of the standard; in general, when the GNU standard recommends a practice
6005 but does not require it, @value{GDBN} requires it.
6007 @value{GDBN} follows an additional set of coding standards specific to
6008 @value{GDBN}, as described in the following sections.
6013 @value{GDBN} assumes an ISO/IEC 9899:1990 (a.k.a.@: ISO C90) compliant
6016 @value{GDBN} does not assume an ISO C or POSIX compliant C library.
6019 @subsection Memory Management
6021 @value{GDBN} does not use the functions @code{malloc}, @code{realloc},
6022 @code{calloc}, @code{free} and @code{asprintf}.
6024 @value{GDBN} uses the functions @code{xmalloc}, @code{xrealloc} and
6025 @code{xcalloc} when allocating memory. Unlike @code{malloc} et.al.@:
6026 these functions do not return when the memory pool is empty. Instead,
6027 they unwind the stack using cleanups. These functions return
6028 @code{NULL} when requested to allocate a chunk of memory of size zero.
6030 @emph{Pragmatics: By using these functions, the need to check every
6031 memory allocation is removed. These functions provide portable
6034 @value{GDBN} does not use the function @code{free}.
6036 @value{GDBN} uses the function @code{xfree} to return memory to the
6037 memory pool. Consistent with ISO-C, this function ignores a request to
6038 free a @code{NULL} pointer.
6040 @emph{Pragmatics: On some systems @code{free} fails when passed a
6041 @code{NULL} pointer.}
6043 @value{GDBN} can use the non-portable function @code{alloca} for the
6044 allocation of small temporary values (such as strings).
6046 @emph{Pragmatics: This function is very non-portable. Some systems
6047 restrict the memory being allocated to no more than a few kilobytes.}
6049 @value{GDBN} uses the string function @code{xstrdup} and the print
6050 function @code{xstrprintf}.
6052 @emph{Pragmatics: @code{asprintf} and @code{strdup} can fail. Print
6053 functions such as @code{sprintf} are very prone to buffer overflow
6057 @subsection Compiler Warnings
6058 @cindex compiler warnings
6060 With few exceptions, developers should avoid the configuration option
6061 @samp{--disable-werror} when building @value{GDBN}. The exceptions
6062 are listed in the file @file{gdb/MAINTAINERS}. The default, when
6063 building with @sc{gcc}, is @samp{--enable-werror}.
6065 This option causes @value{GDBN} (when built using GCC) to be compiled
6066 with a carefully selected list of compiler warning flags. Any warnings
6067 from those flags are treated as errors.
6069 The current list of warning flags includes:
6073 Recommended @sc{gcc} warnings.
6075 @item -Wdeclaration-after-statement
6077 @sc{gcc} 3.x (and later) and @sc{c99} allow declarations mixed with
6078 code, but @sc{gcc} 2.x and @sc{c89} do not.
6080 @item -Wpointer-arith
6082 @item -Wformat-nonliteral
6083 Non-literal format strings, with a few exceptions, are bugs - they
6084 might contain unintended user-supplied format specifiers.
6085 Since @value{GDBN} uses the @code{format printf} attribute on all
6086 @code{printf} like functions this checks not just @code{printf} calls
6087 but also calls to functions such as @code{fprintf_unfiltered}.
6089 @item -Wno-pointer-sign
6090 In version 4.0, GCC began warning about pointer argument passing or
6091 assignment even when the source and destination differed only in
6092 signedness. However, most @value{GDBN} code doesn't distinguish
6093 carefully between @code{char} and @code{unsigned char}. In early 2006
6094 the @value{GDBN} developers decided correcting these warnings wasn't
6095 worth the time it would take.
6097 @item -Wno-unused-parameter
6098 Due to the way that @value{GDBN} is implemented many functions have
6099 unused parameters. Consequently this warning is avoided. The macro
6100 @code{ATTRIBUTE_UNUSED} is not used as it leads to false negatives ---
6101 it is not an error to have @code{ATTRIBUTE_UNUSED} on a parameter that
6106 @itemx -Wno-char-subscripts
6107 These are warnings which might be useful for @value{GDBN}, but are
6108 currently too noisy to enable with @samp{-Werror}.
6112 @subsection Formatting
6114 @cindex source code formatting
6115 The standard GNU recommendations for formatting must be followed
6118 A function declaration should not have its name in column zero. A
6119 function definition should have its name in column zero.
6123 static void foo (void);
6131 @emph{Pragmatics: This simplifies scripting. Function definitions can
6132 be found using @samp{^function-name}.}
6134 There must be a space between a function or macro name and the opening
6135 parenthesis of its argument list (except for macro definitions, as
6136 required by C). There must not be a space after an open paren/bracket
6137 or before a close paren/bracket.
6139 While additional whitespace is generally helpful for reading, do not use
6140 more than one blank line to separate blocks, and avoid adding whitespace
6141 after the end of a program line (as of 1/99, some 600 lines had
6142 whitespace after the semicolon). Excess whitespace causes difficulties
6143 for @code{diff} and @code{patch} utilities.
6145 Pointers are declared using the traditional K&R C style:
6159 @subsection Comments
6161 @cindex comment formatting
6162 The standard GNU requirements on comments must be followed strictly.
6164 Block comments must appear in the following form, with no @code{/*}- or
6165 @code{*/}-only lines, and no leading @code{*}:
6168 /* Wait for control to return from inferior to debugger. If inferior
6169 gets a signal, we may decide to start it up again instead of
6170 returning. That is why there is a loop in this function. When
6171 this function actually returns it means the inferior should be left
6172 stopped and @value{GDBN} should read more commands. */
6175 (Note that this format is encouraged by Emacs; tabbing for a multi-line
6176 comment works correctly, and @kbd{M-q} fills the block consistently.)
6178 Put a blank line between the block comments preceding function or
6179 variable definitions, and the definition itself.
6181 In general, put function-body comments on lines by themselves, rather
6182 than trying to fit them into the 20 characters left at the end of a
6183 line, since either the comment or the code will inevitably get longer
6184 than will fit, and then somebody will have to move it anyhow.
6188 @cindex C data types
6189 Code must not depend on the sizes of C data types, the format of the
6190 host's floating point numbers, the alignment of anything, or the order
6191 of evaluation of expressions.
6193 @cindex function usage
6194 Use functions freely. There are only a handful of compute-bound areas
6195 in @value{GDBN} that might be affected by the overhead of a function
6196 call, mainly in symbol reading. Most of @value{GDBN}'s performance is
6197 limited by the target interface (whether serial line or system call).
6199 However, use functions with moderation. A thousand one-line functions
6200 are just as hard to understand as a single thousand-line function.
6202 @emph{Macros are bad, M'kay.}
6203 (But if you have to use a macro, make sure that the macro arguments are
6204 protected with parentheses.)
6208 Declarations like @samp{struct foo *} should be used in preference to
6209 declarations like @samp{typedef struct foo @{ @dots{} @} *foo_ptr}.
6212 @subsection Function Prototypes
6213 @cindex function prototypes
6215 Prototypes must be used when both @emph{declaring} and @emph{defining}
6216 a function. Prototypes for @value{GDBN} functions must include both the
6217 argument type and name, with the name matching that used in the actual
6218 function definition.
6220 All external functions should have a declaration in a header file that
6221 callers include, except for @code{_initialize_*} functions, which must
6222 be external so that @file{init.c} construction works, but shouldn't be
6223 visible to random source files.
6225 Where a source file needs a forward declaration of a static function,
6226 that declaration must appear in a block near the top of the source file.
6229 @subsection Internal Error Recovery
6231 During its execution, @value{GDBN} can encounter two types of errors.
6232 User errors and internal errors. User errors include not only a user
6233 entering an incorrect command but also problems arising from corrupt
6234 object files and system errors when interacting with the target.
6235 Internal errors include situations where @value{GDBN} has detected, at
6236 run time, a corrupt or erroneous situation.
6238 When reporting an internal error, @value{GDBN} uses
6239 @code{internal_error} and @code{gdb_assert}.
6241 @value{GDBN} must not call @code{abort} or @code{assert}.
6243 @emph{Pragmatics: There is no @code{internal_warning} function. Either
6244 the code detected a user error, recovered from it and issued a
6245 @code{warning} or the code failed to correctly recover from the user
6246 error and issued an @code{internal_error}.}
6248 @subsection Command Names
6250 GDB U/I commands are written @samp{foo-bar}, not @samp{foo_bar}.
6252 @subsection File Names
6254 Any file used when building the core of @value{GDBN} must be in lower
6255 case. Any file used when building the core of @value{GDBN} must be 8.3
6256 unique. These requirements apply to both source and generated files.
6258 @emph{Pragmatics: The core of @value{GDBN} must be buildable on many
6259 platforms including DJGPP and MacOS/HFS. Every time an unfriendly file
6260 is introduced to the build process both @file{Makefile.in} and
6261 @file{configure.in} need to be modified accordingly. Compare the
6262 convoluted conversion process needed to transform @file{COPYING} into
6263 @file{copying.c} with the conversion needed to transform
6264 @file{version.in} into @file{version.c}.}
6266 Any file non 8.3 compliant file (that is not used when building the core
6267 of @value{GDBN}) must be added to @file{gdb/config/djgpp/fnchange.lst}.
6269 @emph{Pragmatics: This is clearly a compromise.}
6271 When @value{GDBN} has a local version of a system header file (ex
6272 @file{string.h}) the file name based on the POSIX header prefixed with
6273 @file{gdb_} (@file{gdb_string.h}). These headers should be relatively
6274 independent: they should use only macros defined by @file{configure},
6275 the compiler, or the host; they should include only system headers; they
6276 should refer only to system types. They may be shared between multiple
6277 programs, e.g.@: @value{GDBN} and @sc{gdbserver}.
6279 For other files @samp{-} is used as the separator.
6282 @subsection Include Files
6284 A @file{.c} file should include @file{defs.h} first.
6286 A @file{.c} file should directly include the @code{.h} file of every
6287 declaration and/or definition it directly refers to. It cannot rely on
6290 A @file{.h} file should directly include the @code{.h} file of every
6291 declaration and/or definition it directly refers to. It cannot rely on
6292 indirect inclusion. Exception: The file @file{defs.h} does not need to
6293 be directly included.
6295 An external declaration should only appear in one include file.
6297 An external declaration should never appear in a @code{.c} file.
6298 Exception: a declaration for the @code{_initialize} function that
6299 pacifies @option{-Wmissing-declaration}.
6301 A @code{typedef} definition should only appear in one include file.
6303 An opaque @code{struct} declaration can appear in multiple @file{.h}
6304 files. Where possible, a @file{.h} file should use an opaque
6305 @code{struct} declaration instead of an include.
6307 All @file{.h} files should be wrapped in:
6310 #ifndef INCLUDE_FILE_NAME_H
6311 #define INCLUDE_FILE_NAME_H
6317 @subsection Clean Design and Portable Implementation
6320 In addition to getting the syntax right, there's the little question of
6321 semantics. Some things are done in certain ways in @value{GDBN} because long
6322 experience has shown that the more obvious ways caused various kinds of
6325 @cindex assumptions about targets
6326 You can't assume the byte order of anything that comes from a target
6327 (including @var{value}s, object files, and instructions). Such things
6328 must be byte-swapped using @code{SWAP_TARGET_AND_HOST} in
6329 @value{GDBN}, or one of the swap routines defined in @file{bfd.h},
6330 such as @code{bfd_get_32}.
6332 You can't assume that you know what interface is being used to talk to
6333 the target system. All references to the target must go through the
6334 current @code{target_ops} vector.
6336 You can't assume that the host and target machines are the same machine
6337 (except in the ``native'' support modules). In particular, you can't
6338 assume that the target machine's header files will be available on the
6339 host machine. Target code must bring along its own header files --
6340 written from scratch or explicitly donated by their owner, to avoid
6344 Insertion of new @code{#ifdef}'s will be frowned upon. It's much better
6345 to write the code portably than to conditionalize it for various
6348 @cindex system dependencies
6349 New @code{#ifdef}'s which test for specific compilers or manufacturers
6350 or operating systems are unacceptable. All @code{#ifdef}'s should test
6351 for features. The information about which configurations contain which
6352 features should be segregated into the configuration files. Experience
6353 has proven far too often that a feature unique to one particular system
6354 often creeps into other systems; and that a conditional based on some
6355 predefined macro for your current system will become worthless over
6356 time, as new versions of your system come out that behave differently
6357 with regard to this feature.
6359 Adding code that handles specific architectures, operating systems,
6360 target interfaces, or hosts, is not acceptable in generic code.
6362 @cindex portable file name handling
6363 @cindex file names, portability
6364 One particularly notorious area where system dependencies tend to
6365 creep in is handling of file names. The mainline @value{GDBN} code
6366 assumes Posix semantics of file names: absolute file names begin with
6367 a forward slash @file{/}, slashes are used to separate leading
6368 directories, case-sensitive file names. These assumptions are not
6369 necessarily true on non-Posix systems such as MS-Windows. To avoid
6370 system-dependent code where you need to take apart or construct a file
6371 name, use the following portable macros:
6374 @findex HAVE_DOS_BASED_FILE_SYSTEM
6375 @item HAVE_DOS_BASED_FILE_SYSTEM
6376 This preprocessing symbol is defined to a non-zero value on hosts
6377 whose filesystems belong to the MS-DOS/MS-Windows family. Use this
6378 symbol to write conditional code which should only be compiled for
6381 @findex IS_DIR_SEPARATOR
6382 @item IS_DIR_SEPARATOR (@var{c})
6383 Evaluates to a non-zero value if @var{c} is a directory separator
6384 character. On Unix and GNU/Linux systems, only a slash @file{/} is
6385 such a character, but on Windows, both @file{/} and @file{\} will
6388 @findex IS_ABSOLUTE_PATH
6389 @item IS_ABSOLUTE_PATH (@var{file})
6390 Evaluates to a non-zero value if @var{file} is an absolute file name.
6391 For Unix and GNU/Linux hosts, a name which begins with a slash
6392 @file{/} is absolute. On DOS and Windows, @file{d:/foo} and
6393 @file{x:\bar} are also absolute file names.
6395 @findex FILENAME_CMP
6396 @item FILENAME_CMP (@var{f1}, @var{f2})
6397 Calls a function which compares file names @var{f1} and @var{f2} as
6398 appropriate for the underlying host filesystem. For Posix systems,
6399 this simply calls @code{strcmp}; on case-insensitive filesystems it
6400 will call @code{strcasecmp} instead.
6402 @findex DIRNAME_SEPARATOR
6403 @item DIRNAME_SEPARATOR
6404 Evaluates to a character which separates directories in
6405 @code{PATH}-style lists, typically held in environment variables.
6406 This character is @samp{:} on Unix, @samp{;} on DOS and Windows.
6408 @findex SLASH_STRING
6410 This evaluates to a constant string you should use to produce an
6411 absolute filename from leading directories and the file's basename.
6412 @code{SLASH_STRING} is @code{"/"} on most systems, but might be
6413 @code{"\\"} for some Windows-based ports.
6416 In addition to using these macros, be sure to use portable library
6417 functions whenever possible. For example, to extract a directory or a
6418 basename part from a file name, use the @code{dirname} and
6419 @code{basename} library functions (available in @code{libiberty} for
6420 platforms which don't provide them), instead of searching for a slash
6421 with @code{strrchr}.
6423 Another way to generalize @value{GDBN} along a particular interface is with an
6424 attribute struct. For example, @value{GDBN} has been generalized to handle
6425 multiple kinds of remote interfaces---not by @code{#ifdef}s everywhere, but
6426 by defining the @code{target_ops} structure and having a current target (as
6427 well as a stack of targets below it, for memory references). Whenever
6428 something needs to be done that depends on which remote interface we are
6429 using, a flag in the current target_ops structure is tested (e.g.,
6430 @code{target_has_stack}), or a function is called through a pointer in the
6431 current target_ops structure. In this way, when a new remote interface
6432 is added, only one module needs to be touched---the one that actually
6433 implements the new remote interface. Other examples of
6434 attribute-structs are BFD access to multiple kinds of object file
6435 formats, or @value{GDBN}'s access to multiple source languages.
6437 Please avoid duplicating code. For example, in @value{GDBN} 3.x all
6438 the code interfacing between @code{ptrace} and the rest of
6439 @value{GDBN} was duplicated in @file{*-dep.c}, and so changing
6440 something was very painful. In @value{GDBN} 4.x, these have all been
6441 consolidated into @file{infptrace.c}. @file{infptrace.c} can deal
6442 with variations between systems the same way any system-independent
6443 file would (hooks, @code{#if defined}, etc.), and machines which are
6444 radically different don't need to use @file{infptrace.c} at all.
6446 All debugging code must be controllable using the @samp{set debug
6447 @var{module}} command. Do not use @code{printf} to print trace
6448 messages. Use @code{fprintf_unfiltered(gdb_stdlog, ...}. Do not use
6449 @code{#ifdef DEBUG}.
6454 @chapter Porting @value{GDBN}
6455 @cindex porting to new machines
6457 Most of the work in making @value{GDBN} compile on a new machine is in
6458 specifying the configuration of the machine. Porting a new
6459 architecture to @value{GDBN} can be broken into a number of steps.
6464 Ensure a @sc{bfd} exists for executables of the target architecture in
6465 the @file{bfd} directory. If one does not exist, create one by
6466 modifying an existing similar one.
6469 Implement a disassembler for the target architecture in the @file{opcodes}
6473 Define the target architecture in the @file{gdb} directory
6474 (@pxref{Adding a New Target, , Adding a New Target}). Add the pattern
6475 for the new target to @file{configure.tgt} with the names of the files
6476 that contain the code. By convention the target architecture
6477 definition for an architecture @var{arch} is placed in
6478 @file{@var{arch}-tdep.c}.
6480 Within @file{@var{arch}-tdep.c} define the function
6481 @code{_initialize_@var{arch}_tdep} which calls
6482 @code{gdbarch_register} to create the new @code{@w{struct
6483 gdbarch}} for the architecture.
6486 If a new remote target is needed, consider adding a new remote target
6487 by defining a function
6488 @code{_initialize_remote_@var{arch}}. However if at all possible
6489 use the @value{GDBN} @emph{Remote Serial Protocol} for this and implement
6490 the server side protocol independently with the target.
6493 If desired implement a simulator in the @file{sim} directory. This
6494 should create the library @file{libsim.a} implementing the interface
6495 in @file{remote-sim.h} (found in the @file{include} directory).
6498 Build and test. If desired, lobby the @sc{gdb} steering group to
6499 have the new port included in the main distribution!
6502 Add a description of the new architecture to the main @value{GDBN} user
6503 guide (@pxref{Configuration Specific Information, , Configuration
6504 Specific Information, gdb, Debugging with @value{GDBN}}).
6508 @node Versions and Branches
6509 @chapter Versions and Branches
6513 @value{GDBN}'s version is determined by the file
6514 @file{gdb/version.in} and takes one of the following forms:
6517 @item @var{major}.@var{minor}
6518 @itemx @var{major}.@var{minor}.@var{patchlevel}
6519 an official release (e.g., 6.2 or 6.2.1)
6520 @item @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD}
6521 a snapshot taken at @var{YYYY}-@var{MM}-@var{DD}-gmt (e.g.,
6522 6.1.50.20020302, 6.1.90.20020304, or 6.1.0.20020308)
6523 @item @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD}-cvs
6524 a @sc{cvs} check out drawn on @var{YYYY}-@var{MM}-@var{DD} (e.g.,
6525 6.1.50.20020302-cvs, 6.1.90.20020304-cvs, or 6.1.0.20020308-cvs)
6526 @item @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD} (@var{vendor})
6527 a vendor specific release of @value{GDBN}, that while based on@*
6528 @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD},
6529 may include additional changes
6532 @value{GDBN}'s mainline uses the @var{major} and @var{minor} version
6533 numbers from the most recent release branch, with a @var{patchlevel}
6534 of 50. At the time each new release branch is created, the mainline's
6535 @var{major} and @var{minor} version numbers are updated.
6537 @value{GDBN}'s release branch is similar. When the branch is cut, the
6538 @var{patchlevel} is changed from 50 to 90. As draft releases are
6539 drawn from the branch, the @var{patchlevel} is incremented. Once the
6540 first release (@var{major}.@var{minor}) has been made, the
6541 @var{patchlevel} is set to 0 and updates have an incremented
6544 For snapshots, and @sc{cvs} check outs, it is also possible to
6545 identify the @sc{cvs} origin:
6548 @item @var{major}.@var{minor}.50.@var{YYYY}@var{MM}@var{DD}
6549 drawn from the @sc{head} of mainline @sc{cvs} (e.g., 6.1.50.20020302)
6550 @item @var{major}.@var{minor}.90.@var{YYYY}@var{MM}@var{DD}
6551 @itemx @var{major}.@var{minor}.91.@var{YYYY}@var{MM}@var{DD} @dots{}
6552 drawn from a release branch prior to the release (e.g.,
6554 @item @var{major}.@var{minor}.0.@var{YYYY}@var{MM}@var{DD}
6555 @itemx @var{major}.@var{minor}.1.@var{YYYY}@var{MM}@var{DD} @dots{}
6556 drawn from a release branch after the release (e.g., 6.2.0.20020308)
6559 If the previous @value{GDBN} version is 6.1 and the current version is
6560 6.2, then, substituting 6 for @var{major} and 1 or 2 for @var{minor},
6561 here's an illustration of a typical sequence:
6568 +--------------------------.
6571 6.2.50.20020303-cvs 6.1.90 (draft #1)
6573 6.2.50.20020304-cvs 6.1.90.20020304-cvs
6575 6.2.50.20020305-cvs 6.1.91 (draft #2)
6577 6.2.50.20020306-cvs 6.1.91.20020306-cvs
6579 6.2.50.20020307-cvs 6.2 (release)
6581 6.2.50.20020308-cvs 6.2.0.20020308-cvs
6583 6.2.50.20020309-cvs 6.2.1 (update)
6585 6.2.50.20020310-cvs <branch closed>
6589 +--------------------------.
6592 6.3.50.20020312-cvs 6.2.90 (draft #1)
6596 @section Release Branches
6597 @cindex Release Branches
6599 @value{GDBN} draws a release series (6.2, 6.2.1, @dots{}) from a
6600 single release branch, and identifies that branch using the @sc{cvs}
6604 gdb_@var{major}_@var{minor}-@var{YYYY}@var{MM}@var{DD}-branchpoint
6605 gdb_@var{major}_@var{minor}-branch
6606 gdb_@var{major}_@var{minor}-@var{YYYY}@var{MM}@var{DD}-release
6609 @emph{Pragmatics: To help identify the date at which a branch or
6610 release is made, both the branchpoint and release tags include the
6611 date that they are cut (@var{YYYY}@var{MM}@var{DD}) in the tag. The
6612 branch tag, denoting the head of the branch, does not need this.}
6614 @section Vendor Branches
6615 @cindex vendor branches
6617 To avoid version conflicts, vendors are expected to modify the file
6618 @file{gdb/version.in} to include a vendor unique alphabetic identifier
6619 (an official @value{GDBN} release never uses alphabetic characters in
6620 its version identifier). E.g., @samp{6.2widgit2}, or @samp{6.2 (Widgit
6623 @section Experimental Branches
6624 @cindex experimental branches
6626 @subsection Guidelines
6628 @value{GDBN} permits the creation of branches, cut from the @sc{cvs}
6629 repository, for experimental development. Branches make it possible
6630 for developers to share preliminary work, and maintainers to examine
6631 significant new developments.
6633 The following are a set of guidelines for creating such branches:
6637 @item a branch has an owner
6638 The owner can set further policy for a branch, but may not change the
6639 ground rules. In particular, they can set a policy for commits (be it
6640 adding more reviewers or deciding who can commit).
6642 @item all commits are posted
6643 All changes committed to a branch shall also be posted to
6644 @email{gdb-patches@@sourceware.org, the @value{GDBN} patches
6645 mailing list}. While commentary on such changes are encouraged, people
6646 should remember that the changes only apply to a branch.
6648 @item all commits are covered by an assignment
6649 This ensures that all changes belong to the Free Software Foundation,
6650 and avoids the possibility that the branch may become contaminated.
6652 @item a branch is focused
6653 A focused branch has a single objective or goal, and does not contain
6654 unnecessary or irrelevant changes. Cleanups, where identified, being
6655 be pushed into the mainline as soon as possible.
6657 @item a branch tracks mainline
6658 This keeps the level of divergence under control. It also keeps the
6659 pressure on developers to push cleanups and other stuff into the
6662 @item a branch shall contain the entire @value{GDBN} module
6663 The @value{GDBN} module @code{gdb} should be specified when creating a
6664 branch (branches of individual files should be avoided). @xref{Tags}.
6666 @item a branch shall be branded using @file{version.in}
6667 The file @file{gdb/version.in} shall be modified so that it identifies
6668 the branch @var{owner} and branch @var{name}, e.g.,
6669 @samp{6.2.50.20030303_owner_name} or @samp{6.2 (Owner Name)}.
6676 To simplify the identification of @value{GDBN} branches, the following
6677 branch tagging convention is strongly recommended:
6681 @item @var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint
6682 @itemx @var{owner}_@var{name}-@var{YYYYMMDD}-branch
6683 The branch point and corresponding branch tag. @var{YYYYMMDD} is the
6684 date that the branch was created. A branch is created using the
6685 sequence: @anchor{experimental branch tags}
6687 cvs rtag @var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint gdb
6688 cvs rtag -b -r @var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint \
6689 @var{owner}_@var{name}-@var{YYYYMMDD}-branch gdb
6692 @item @var{owner}_@var{name}-@var{yyyymmdd}-mergepoint
6693 The tagged point, on the mainline, that was used when merging the branch
6694 on @var{yyyymmdd}. To merge in all changes since the branch was cut,
6695 use a command sequence like:
6697 cvs rtag @var{owner}_@var{name}-@var{yyyymmdd}-mergepoint gdb
6699 -j@var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint
6700 -j@var{owner}_@var{name}-@var{yyyymmdd}-mergepoint
6703 Similar sequences can be used to just merge in changes since the last
6709 For further information on @sc{cvs}, see
6710 @uref{http://www.gnu.org/software/cvs/, Concurrent Versions System}.
6712 @node Start of New Year Procedure
6713 @chapter Start of New Year Procedure
6714 @cindex new year procedure
6716 At the start of each new year, the following actions should be performed:
6720 Rotate the ChangeLog file
6722 The current @file{ChangeLog} file should be renamed into
6723 @file{ChangeLog-YYYY} where YYYY is the year that has just passed.
6724 A new @file{ChangeLog} file should be created, and its contents should
6725 contain a reference to the previous ChangeLog. The following should
6726 also be preserved at the end of the new ChangeLog, in order to provide
6727 the appropriate settings when editing this file with Emacs:
6733 version-control: never
6739 Add an entry for the newly created ChangeLog file (@file{ChangeLog-YYYY})
6740 in @file{gdb/config/djgpp/fnchange.lst}.
6743 Update the copyright year in the startup message
6745 Update the copyright year in:
6748 file @file{top.c}, function @code{print_gdb_version}
6750 file @file{gdbserver/server.c}, function @code{gdbserver_version}
6752 file @file{gdbserver/gdbreplay.c}, function @code{gdbreplay_version}
6756 Run the @file{copyright.sh} script to add the new year in the copyright
6757 notices of most source files. This script requires Emacs 22 or later to
6761 The new year also needs to be added manually in all other files that
6762 are not already taken care of by the @file{copyright.sh} script:
6790 @chapter Releasing @value{GDBN}
6791 @cindex making a new release of gdb
6793 @section Branch Commit Policy
6795 The branch commit policy is pretty slack. @value{GDBN} releases 5.0,
6796 5.1 and 5.2 all used the below:
6800 The @file{gdb/MAINTAINERS} file still holds.
6802 Don't fix something on the branch unless/until it is also fixed in the
6803 trunk. If this isn't possible, mentioning it in the @file{gdb/PROBLEMS}
6804 file is better than committing a hack.
6806 When considering a patch for the branch, suggested criteria include:
6807 Does it fix a build? Does it fix the sequence @kbd{break main; run}
6808 when debugging a static binary?
6810 The further a change is from the core of @value{GDBN}, the less likely
6811 the change will worry anyone (e.g., target specific code).
6813 Only post a proposal to change the core of @value{GDBN} after you've
6814 sent individual bribes to all the people listed in the
6815 @file{MAINTAINERS} file @t{;-)}
6818 @emph{Pragmatics: Provided updates are restricted to non-core
6819 functionality there is little chance that a broken change will be fatal.
6820 This means that changes such as adding a new architectures or (within
6821 reason) support for a new host are considered acceptable.}
6824 @section Obsoleting code
6826 Before anything else, poke the other developers (and around the source
6827 code) to see if there is anything that can be removed from @value{GDBN}
6828 (an old target, an unused file).
6830 Obsolete code is identified by adding an @code{OBSOLETE} prefix to every
6831 line. Doing this means that it is easy to identify something that has
6832 been obsoleted when greping through the sources.
6834 The process is done in stages --- this is mainly to ensure that the
6835 wider @value{GDBN} community has a reasonable opportunity to respond.
6836 Remember, everything on the Internet takes a week.
6840 Post the proposal on @email{gdb@@sourceware.org, the GDB mailing
6841 list} Creating a bug report to track the task's state, is also highly
6846 Post the proposal on @email{gdb-announce@@sourceware.org, the GDB
6847 Announcement mailing list}.
6851 Go through and edit all relevant files and lines so that they are
6852 prefixed with the word @code{OBSOLETE}.
6854 Wait until the next GDB version, containing this obsolete code, has been
6857 Remove the obsolete code.
6861 @emph{Maintainer note: While removing old code is regrettable it is
6862 hopefully better for @value{GDBN}'s long term development. Firstly it
6863 helps the developers by removing code that is either no longer relevant
6864 or simply wrong. Secondly since it removes any history associated with
6865 the file (effectively clearing the slate) the developer has a much freer
6866 hand when it comes to fixing broken files.}
6870 @section Before the Branch
6872 The most important objective at this stage is to find and fix simple
6873 changes that become a pain to track once the branch is created. For
6874 instance, configuration problems that stop @value{GDBN} from even
6875 building. If you can't get the problem fixed, document it in the
6876 @file{gdb/PROBLEMS} file.
6878 @subheading Prompt for @file{gdb/NEWS}
6880 People always forget. Send a post reminding them but also if you know
6881 something interesting happened add it yourself. The @code{schedule}
6882 script will mention this in its e-mail.
6884 @subheading Review @file{gdb/README}
6886 Grab one of the nightly snapshots and then walk through the
6887 @file{gdb/README} looking for anything that can be improved. The
6888 @code{schedule} script will mention this in its e-mail.
6890 @subheading Refresh any imported files.
6892 A number of files are taken from external repositories. They include:
6896 @file{texinfo/texinfo.tex}
6898 @file{config.guess} et.@: al.@: (see the top-level @file{MAINTAINERS}
6901 @file{etc/standards.texi}, @file{etc/make-stds.texi}
6904 @subheading Check the ARI
6906 @uref{http://sourceware.org/gdb/ari,,A.R.I.} is an @code{awk} script
6907 (Awk Regression Index ;-) that checks for a number of errors and coding
6908 conventions. The checks include things like using @code{malloc} instead
6909 of @code{xmalloc} and file naming problems. There shouldn't be any
6912 @subsection Review the bug data base
6914 Close anything obviously fixed.
6916 @subsection Check all cross targets build
6918 The targets are listed in @file{gdb/MAINTAINERS}.
6921 @section Cut the Branch
6923 @subheading Create the branch
6928 $ V=`echo $v | sed 's/\./_/g'`
6929 $ D=`date -u +%Y-%m-%d`
6932 $ echo cvs -f -d :ext:sourceware.org:/cvs/src rtag \
6933 -D $D-gmt gdb_$V-$D-branchpoint insight
6934 cvs -f -d :ext:sourceware.org:/cvs/src rtag
6935 -D 2002-03-03-gmt gdb_5_2-2002-03-03-branchpoint insight
6938 $ echo cvs -f -d :ext:sourceware.org:/cvs/src rtag \
6939 -b -r gdb_$V-$D-branchpoint gdb_$V-branch insight
6940 cvs -f -d :ext:sourceware.org:/cvs/src rtag \
6941 -b -r gdb_5_2-2002-03-03-branchpoint gdb_5_2-branch insight
6949 By using @kbd{-D YYYY-MM-DD-gmt}, the branch is forced to an exact
6952 The trunk is first tagged so that the branch point can easily be found.
6954 Insight, which includes @value{GDBN}, is tagged at the same time.
6956 @file{version.in} gets bumped to avoid version number conflicts.
6958 The reading of @file{.cvsrc} is disabled using @file{-f}.
6961 @subheading Update @file{version.in}
6966 $ V=`echo $v | sed 's/\./_/g'`
6970 $ echo cvs -f -d :ext:sourceware.org:/cvs/src co \
6971 -r gdb_$V-branch src/gdb/version.in
6972 cvs -f -d :ext:sourceware.org:/cvs/src co
6973 -r gdb_5_2-branch src/gdb/version.in
6975 U src/gdb/version.in
6977 $ echo $u.90-0000-00-00-cvs > version.in
6979 5.1.90-0000-00-00-cvs
6980 $ cvs -f commit version.in
6985 @file{0000-00-00} is used as a date to pump prime the version.in update
6988 @file{.90} and the previous branch version are used as fairly arbitrary
6989 initial branch version number.
6993 @subheading Update the web and news pages
6997 @subheading Tweak cron to track the new branch
6999 The file @file{gdbadmin/cron/crontab} contains gdbadmin's cron table.
7000 This file needs to be updated so that:
7004 A daily timestamp is added to the file @file{version.in}.
7006 The new branch is included in the snapshot process.
7010 See the file @file{gdbadmin/cron/README} for how to install the updated
7013 The file @file{gdbadmin/ss/README} should also be reviewed to reflect
7014 any changes. That file is copied to both the branch/ and current/
7015 snapshot directories.
7018 @subheading Update the NEWS and README files
7020 The @file{NEWS} file needs to be updated so that on the branch it refers
7021 to @emph{changes in the current release} while on the trunk it also
7022 refers to @emph{changes since the current release}.
7024 The @file{README} file needs to be updated so that it refers to the
7027 @subheading Post the branch info
7029 Send an announcement to the mailing lists:
7033 @email{gdb-announce@@sourceware.org, GDB Announcement mailing list}
7035 @email{gdb@@sourceware.org, GDB Discussion mailing list} and
7036 @email{gdb-testers@@sourceware.org, GDB Testers mailing list}
7039 @emph{Pragmatics: The branch creation is sent to the announce list to
7040 ensure that people people not subscribed to the higher volume discussion
7043 The announcement should include:
7049 How to check out the branch using CVS.
7051 The date/number of weeks until the release.
7053 The branch commit policy still holds.
7056 @section Stabilize the branch
7058 Something goes here.
7060 @section Create a Release
7062 The process of creating and then making available a release is broken
7063 down into a number of stages. The first part addresses the technical
7064 process of creating a releasable tar ball. The later stages address the
7065 process of releasing that tar ball.
7067 When making a release candidate just the first section is needed.
7069 @subsection Create a release candidate
7071 The objective at this stage is to create a set of tar balls that can be
7072 made available as a formal release (or as a less formal release
7075 @subsubheading Freeze the branch
7077 Send out an e-mail notifying everyone that the branch is frozen to
7078 @email{gdb-patches@@sourceware.org}.
7080 @subsubheading Establish a few defaults.
7085 $ t=/sourceware/snapshot-tmp/gdbadmin-tmp
7087 /sourceware/snapshot-tmp/gdbadmin-tmp/gdb_5_2-branch/5.2
7091 /sourceware/snapshot-tmp/gdbadmin-tmp/gdb_5_2-branch/5.2
7093 /home/gdbadmin/bin/autoconf
7102 Check the @code{autoconf} version carefully. You want to be using the
7103 version documented in the toplevel @file{README-maintainer-mode} file.
7104 It is very unlikely that the version of @code{autoconf} installed in
7105 system directories (e.g., @file{/usr/bin/autoconf}) is correct.
7108 @subsubheading Check out the relevant modules:
7111 $ for m in gdb insight
7113 ( mkdir -p $m && cd $m && cvs -q -f -d /cvs/src co -P -r $b $m )
7123 The reading of @file{.cvsrc} is disabled (@file{-f}) so that there isn't
7124 any confusion between what is written here and what your local
7125 @code{cvs} really does.
7128 @subsubheading Update relevant files.
7134 Major releases get their comments added as part of the mainline. Minor
7135 releases should probably mention any significant bugs that were fixed.
7137 Don't forget to include the @file{ChangeLog} entry.
7140 $ emacs gdb/src/gdb/NEWS
7145 $ cp gdb/src/gdb/NEWS insight/src/gdb/NEWS
7146 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
7151 You'll need to update:
7163 $ emacs gdb/src/gdb/README
7168 $ cp gdb/src/gdb/README insight/src/gdb/README
7169 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
7172 @emph{Maintainer note: Hopefully the @file{README} file was reviewed
7173 before the initial branch was cut so just a simple substitute is needed
7176 @emph{Maintainer note: Other projects generate @file{README} and
7177 @file{INSTALL} from the core documentation. This might be worth
7180 @item gdb/version.in
7183 $ echo $v > gdb/src/gdb/version.in
7184 $ cat gdb/src/gdb/version.in
7186 $ emacs gdb/src/gdb/version.in
7189 ... Bump to version ...
7191 $ cp gdb/src/gdb/version.in insight/src/gdb/version.in
7192 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
7197 @subsubheading Do the dirty work
7199 This is identical to the process used to create the daily snapshot.
7202 $ for m in gdb insight
7204 ( cd $m/src && gmake -f src-release $m.tar )
7208 If the top level source directory does not have @file{src-release}
7209 (@value{GDBN} version 5.3.1 or earlier), try these commands instead:
7212 $ for m in gdb insight
7214 ( cd $m/src && gmake -f Makefile.in $m.tar )
7218 @subsubheading Check the source files
7220 You're looking for files that have mysteriously disappeared.
7221 @kbd{distclean} has the habit of deleting files it shouldn't. Watch out
7222 for the @file{version.in} update @kbd{cronjob}.
7225 $ ( cd gdb/src && cvs -f -q -n update )
7229 @dots{} lots of generated files @dots{}
7234 @dots{} lots of generated files @dots{}
7239 @emph{Don't worry about the @file{gdb.info-??} or
7240 @file{gdb/p-exp.tab.c}. They were generated (and yes @file{gdb.info-1}
7241 was also generated only something strange with CVS means that they
7242 didn't get suppressed). Fixing it would be nice though.}
7244 @subsubheading Create compressed versions of the release
7250 gdb/ gdb-5.2.tar insight/ insight-5.2.tar
7251 $ for m in gdb insight
7253 bzip2 -v -9 -c $m-$v.tar > $m-$v.tar.bz2
7254 gzip -v -9 -c $m-$v.tar > $m-$v.tar.gz
7264 A pipe such as @kbd{bunzip2 < xxx.bz2 | gzip -9 > xxx.gz} is not since,
7265 in that mode, @code{gzip} does not know the name of the file and, hence,
7266 can not include it in the compressed file. This is also why the release
7267 process runs @code{tar} and @code{bzip2} as separate passes.
7270 @subsection Sanity check the tar ball
7272 Pick a popular machine (Solaris/PPC?) and try the build on that.
7275 $ bunzip2 < gdb-5.2.tar.bz2 | tar xpf -
7280 $ ./gdb/gdb ./gdb/gdb
7284 Breakpoint 1 at 0x80732bc: file main.c, line 734.
7286 Starting program: /tmp/gdb-5.2/gdb/gdb
7288 Breakpoint 1, main (argc=1, argv=0xbffff8b4) at main.c:734
7289 734 catch_errors (captured_main, &args, "", RETURN_MASK_ALL);
7291 $1 = @{argc = 136426532, argv = 0x821b7f0@}
7295 @subsection Make a release candidate available
7297 If this is a release candidate then the only remaining steps are:
7301 Commit @file{version.in} and @file{ChangeLog}
7303 Tweak @file{version.in} (and @file{ChangeLog} to read
7304 @var{L}.@var{M}.@var{N}-0000-00-00-cvs so that the version update
7305 process can restart.
7307 Make the release candidate available in
7308 @uref{ftp://sourceware.org/pub/gdb/snapshots/branch}
7310 Notify the relevant mailing lists ( @email{gdb@@sourceware.org} and
7311 @email{gdb-testers@@sourceware.org} that the candidate is available.
7314 @subsection Make a formal release available
7316 (And you thought all that was required was to post an e-mail.)
7318 @subsubheading Install on sware
7320 Copy the new files to both the release and the old release directory:
7323 $ cp *.bz2 *.gz ~ftp/pub/gdb/old-releases/
7324 $ cp *.bz2 *.gz ~ftp/pub/gdb/releases
7328 Clean up the releases directory so that only the most recent releases
7329 are available (e.g.@: keep 5.2 and 5.2.1 but remove 5.1):
7332 $ cd ~ftp/pub/gdb/releases
7337 Update the file @file{README} and @file{.message} in the releases
7344 $ ln README .message
7347 @subsubheading Update the web pages.
7351 @item htdocs/download/ANNOUNCEMENT
7352 This file, which is posted as the official announcement, includes:
7355 General announcement.
7357 News. If making an @var{M}.@var{N}.1 release, retain the news from
7358 earlier @var{M}.@var{N} release.
7363 @item htdocs/index.html
7364 @itemx htdocs/news/index.html
7365 @itemx htdocs/download/index.html
7366 These files include:
7369 Announcement of the most recent release.
7371 News entry (remember to update both the top level and the news directory).
7373 These pages also need to be regenerate using @code{index.sh}.
7375 @item download/onlinedocs/
7376 You need to find the magic command that is used to generate the online
7377 docs from the @file{.tar.bz2}. The best way is to look in the output
7378 from one of the nightly @code{cron} jobs and then just edit accordingly.
7382 $ ~/ss/update-web-docs \
7383 ~ftp/pub/gdb/releases/gdb-5.2.tar.bz2 \
7385 /www/sourceware/htdocs/gdb/download/onlinedocs \
7390 Just like the online documentation. Something like:
7393 $ /bin/sh ~/ss/update-web-ari \
7394 ~ftp/pub/gdb/releases/gdb-5.2.tar.bz2 \
7396 /www/sourceware/htdocs/gdb/download/ari \
7402 @subsubheading Shadow the pages onto gnu
7404 Something goes here.
7407 @subsubheading Install the @value{GDBN} tar ball on GNU
7409 At the time of writing, the GNU machine was @kbd{gnudist.gnu.org} in
7410 @file{~ftp/gnu/gdb}.
7412 @subsubheading Make the @file{ANNOUNCEMENT}
7414 Post the @file{ANNOUNCEMENT} file you created above to:
7418 @email{gdb-announce@@sourceware.org, GDB Announcement mailing list}
7420 @email{info-gnu@@gnu.org, General GNU Announcement list} (but delay it a
7421 day or so to let things get out)
7423 @email{bug-gdb@@gnu.org, GDB Bug Report mailing list}
7428 The release is out but you're still not finished.
7430 @subsubheading Commit outstanding changes
7432 In particular you'll need to commit any changes to:
7436 @file{gdb/ChangeLog}
7438 @file{gdb/version.in}
7445 @subsubheading Tag the release
7450 $ d=`date -u +%Y-%m-%d`
7453 $ ( cd insight/src/gdb && cvs -f -q update )
7454 $ ( cd insight/src && cvs -f -q tag gdb_5_2-$d-release )
7457 Insight is used since that contains more of the release than
7460 @subsubheading Mention the release on the trunk
7462 Just put something in the @file{ChangeLog} so that the trunk also
7463 indicates when the release was made.
7465 @subsubheading Restart @file{gdb/version.in}
7467 If @file{gdb/version.in} does not contain an ISO date such as
7468 @kbd{2002-01-24} then the daily @code{cronjob} won't update it. Having
7469 committed all the release changes it can be set to
7470 @file{5.2.0_0000-00-00-cvs} which will restart things (yes the @kbd{_}
7471 is important - it affects the snapshot process).
7473 Don't forget the @file{ChangeLog}.
7475 @subsubheading Merge into trunk
7477 The files committed to the branch may also need changes merged into the
7480 @subsubheading Revise the release schedule
7482 Post a revised release schedule to @email{gdb@@sourceware.org, GDB
7483 Discussion List} with an updated announcement. The schedule can be
7484 generated by running:
7487 $ ~/ss/schedule `date +%s` schedule
7491 The first parameter is approximate date/time in seconds (from the epoch)
7492 of the most recent release.
7494 Also update the schedule @code{cronjob}.
7496 @section Post release
7498 Remove any @code{OBSOLETE} code.
7505 The testsuite is an important component of the @value{GDBN} package.
7506 While it is always worthwhile to encourage user testing, in practice
7507 this is rarely sufficient; users typically use only a small subset of
7508 the available commands, and it has proven all too common for a change
7509 to cause a significant regression that went unnoticed for some time.
7511 The @value{GDBN} testsuite uses the DejaGNU testing framework. The
7512 tests themselves are calls to various @code{Tcl} procs; the framework
7513 runs all the procs and summarizes the passes and fails.
7515 @section Using the Testsuite
7517 @cindex running the test suite
7518 To run the testsuite, simply go to the @value{GDBN} object directory (or to the
7519 testsuite's objdir) and type @code{make check}. This just sets up some
7520 environment variables and invokes DejaGNU's @code{runtest} script. While
7521 the testsuite is running, you'll get mentions of which test file is in use,
7522 and a mention of any unexpected passes or fails. When the testsuite is
7523 finished, you'll get a summary that looks like this:
7528 # of expected passes 6016
7529 # of unexpected failures 58
7530 # of unexpected successes 5
7531 # of expected failures 183
7532 # of unresolved testcases 3
7533 # of untested testcases 5
7536 To run a specific test script, type:
7538 make check RUNTESTFLAGS='@var{tests}'
7540 where @var{tests} is a list of test script file names, separated by
7543 If you use GNU make, you can use its @option{-j} option to run the
7544 testsuite in parallel. This can greatly reduce the amount of time it
7545 takes for the testsuite to run. In this case, if you set
7546 @code{RUNTESTFLAGS} then, by default, the tests will be run serially
7547 even under @option{-j}. You can override this and force a parallel run
7548 by setting the @code{make} variable @code{FORCE_PARALLEL} to any
7549 non-empty value. Note that the parallel @kbd{make check} assumes
7550 that you want to run the entire testsuite, so it is not compatible
7551 with some dejagnu options, like @option{--directory}.
7553 The ideal test run consists of expected passes only; however, reality
7554 conspires to keep us from this ideal. Unexpected failures indicate
7555 real problems, whether in @value{GDBN} or in the testsuite. Expected
7556 failures are still failures, but ones which have been decided are too
7557 hard to deal with at the time; for instance, a test case might work
7558 everywhere except on AIX, and there is no prospect of the AIX case
7559 being fixed in the near future. Expected failures should not be added
7560 lightly, since you may be masking serious bugs in @value{GDBN}.
7561 Unexpected successes are expected fails that are passing for some
7562 reason, while unresolved and untested cases often indicate some minor
7563 catastrophe, such as the compiler being unable to deal with a test
7566 When making any significant change to @value{GDBN}, you should run the
7567 testsuite before and after the change, to confirm that there are no
7568 regressions. Note that truly complete testing would require that you
7569 run the testsuite with all supported configurations and a variety of
7570 compilers; however this is more than really necessary. In many cases
7571 testing with a single configuration is sufficient. Other useful
7572 options are to test one big-endian (Sparc) and one little-endian (x86)
7573 host, a cross config with a builtin simulator (powerpc-eabi,
7574 mips-elf), or a 64-bit host (Alpha).
7576 If you add new functionality to @value{GDBN}, please consider adding
7577 tests for it as well; this way future @value{GDBN} hackers can detect
7578 and fix their changes that break the functionality you added.
7579 Similarly, if you fix a bug that was not previously reported as a test
7580 failure, please add a test case for it. Some cases are extremely
7581 difficult to test, such as code that handles host OS failures or bugs
7582 in particular versions of compilers, and it's OK not to try to write
7583 tests for all of those.
7585 DejaGNU supports separate build, host, and target machines. However,
7586 some @value{GDBN} test scripts do not work if the build machine and
7587 the host machine are not the same. In such an environment, these scripts
7588 will give a result of ``UNRESOLVED'', like this:
7591 UNRESOLVED: gdb.base/example.exp: This test script does not work on a remote host.
7594 @section Testsuite Parameters
7596 Several variables exist to modify the behavior of the testsuite.
7600 @item @code{TRANSCRIPT}
7602 Sometimes it is convenient to get a transcript of the commands which
7603 the testsuite sends to @value{GDBN}. For example, if @value{GDBN}
7604 crashes during testing, a transcript can be used to more easily
7605 reconstruct the failure when running @value{GDBN} under @value{GDBN}.
7607 You can instruct the @value{GDBN} testsuite to write transcripts by
7608 setting the DejaGNU variable @code{TRANSCRIPT} (to any value)
7609 before invoking @code{runtest} or @kbd{make check}. The transcripts
7610 will be written into DejaGNU's output directory. One transcript will
7611 be made for each invocation of @value{GDBN}; they will be named
7612 @file{transcript.@var{n}}, where @var{n} is an integer. The first
7613 line of the transcript file will show how @value{GDBN} was invoked;
7614 each subsequent line is a command sent as input to @value{GDBN}.
7617 make check RUNTESTFLAGS=TRANSCRIPT=y
7620 Note that the transcript is not always complete. In particular, tests
7621 of completion can yield partial command lines.
7625 Sometimes one wishes to test a different @value{GDBN} than the one in the build
7626 directory. For example, one may wish to run the testsuite on
7627 @file{/usr/bin/gdb}.
7630 make check RUNTESTFLAGS=GDB=/usr/bin/gdb
7633 @item @code{GDBSERVER}
7635 When testing a different @value{GDBN}, it is often useful to also test a
7636 different gdbserver.
7639 make check RUNTESTFLAGS="GDB=/usr/bin/gdb GDBSERVER=/usr/bin/gdbserver"
7642 @item @code{INTERNAL_GDBFLAGS}
7644 When running the testsuite normally one doesn't want whatever is in
7645 @file{~/.gdbinit} to interfere with the tests, therefore the test harness
7646 passes @option{-nx} to @value{GDBN}. One also doesn't want any windowed
7647 version of @value{GDBN}, e.g., @command{gdbtui}, to run.
7648 This is achieved via @code{INTERNAL_GDBFLAGS}.
7651 set INTERNAL_GDBFLAGS "-nw -nx"
7654 This is all well and good, except when testing an installed @value{GDBN}
7655 that has been configured with @option{--with-system-gdbinit}. Here one
7656 does not want @file{~/.gdbinit} loaded but one may want the system
7657 @file{.gdbinit} file loaded. This can be achieved by pointing @code{$HOME}
7658 at a directory without a @file{.gdbinit} and by overriding
7659 @code{INTERNAL_GDBFLAGS} and removing @option{-nx}.
7663 HOME=`pwd` runtest \
7665 GDBSERVER=/usr/bin/gdbserver \
7666 INTERNAL_GDBFLAGS=-nw
7671 There are two ways to run the testsuite and pass additional parameters
7672 to DejaGnu. The first is with @kbd{make check} and specifying the
7673 makefile variable @samp{RUNTESTFLAGS}.
7676 make check RUNTESTFLAGS=TRANSCRIPT=y
7679 The second is to cd to the @file{testsuite} directory and invoke the DejaGnu
7680 @command{runtest} command directly.
7685 runtest TRANSCRIPT=y
7688 @section Testsuite Configuration
7689 @cindex Testsuite Configuration
7691 It is possible to adjust the behavior of the testsuite by defining
7692 the global variables listed below, either in a @file{site.exp} file,
7697 @item @code{gdb_test_timeout}
7699 Defining this variable changes the default timeout duration used during
7700 communication with @value{GDBN}. More specifically, the global variable
7701 used during testing is @code{timeout}, but this variable gets reset to
7702 @code{gdb_test_timeout} at the beginning of each testcase, making sure
7703 that any local change to @code{timeout} in a testcase does not affect
7704 subsequent testcases.
7706 This global variable comes in handy when the debugger is slower than
7707 normal due to the testing environment, triggering unexpected @code{TIMEOUT}
7708 test failures. Examples include when testing on a remote machine, or
7709 against a system where communications are slow.
7711 If not specifically defined, this variable gets automatically defined
7712 to the same value as @code{timeout} during the testsuite initialization.
7713 The default value of the timeout is defined in the file
7714 @file{gdb/testsuite/config/unix.exp} that is part of the @value{GDBN}
7715 test suite@footnote{If you are using a board file, it could override
7716 the test-suite default; search the board file for "timeout".}.
7720 @section Testsuite Organization
7722 @cindex test suite organization
7723 The testsuite is entirely contained in @file{gdb/testsuite}. While the
7724 testsuite includes some makefiles and configury, these are very minimal,
7725 and used for little besides cleaning up, since the tests themselves
7726 handle the compilation of the programs that @value{GDBN} will run. The file
7727 @file{testsuite/lib/gdb.exp} contains common utility procs useful for
7728 all @value{GDBN} tests, while the directory @file{testsuite/config} contains
7729 configuration-specific files, typically used for special-purpose
7730 definitions of procs like @code{gdb_load} and @code{gdb_start}.
7732 The tests themselves are to be found in @file{testsuite/gdb.*} and
7733 subdirectories of those. The names of the test files must always end
7734 with @file{.exp}. DejaGNU collects the test files by wildcarding
7735 in the test directories, so both subdirectories and individual files
7736 get chosen and run in alphabetical order.
7738 The following table lists the main types of subdirectories and what they
7739 are for. Since DejaGNU finds test files no matter where they are
7740 located, and since each test file sets up its own compilation and
7741 execution environment, this organization is simply for convenience and
7746 This is the base testsuite. The tests in it should apply to all
7747 configurations of @value{GDBN} (but generic native-only tests may live here).
7748 The test programs should be in the subset of C that is valid K&R,
7749 ANSI/ISO, and C@t{++} (@code{#ifdef}s are allowed if necessary, for instance
7752 @item gdb.@var{lang}
7753 Language-specific tests for any language @var{lang} besides C. Examples are
7754 @file{gdb.cp} and @file{gdb.java}.
7756 @item gdb.@var{platform}
7757 Non-portable tests. The tests are specific to a specific configuration
7758 (host or target), such as HP-UX or eCos. Example is @file{gdb.hp}, for
7761 @item gdb.@var{compiler}
7762 Tests specific to a particular compiler. As of this writing (June
7763 1999), there aren't currently any groups of tests in this category that
7764 couldn't just as sensibly be made platform-specific, but one could
7765 imagine a @file{gdb.gcc}, for tests of @value{GDBN}'s handling of GCC
7768 @item gdb.@var{subsystem}
7769 Tests that exercise a specific @value{GDBN} subsystem in more depth. For
7770 instance, @file{gdb.disasm} exercises various disassemblers, while
7771 @file{gdb.stabs} tests pathways through the stabs symbol reader.
7774 @section Writing Tests
7775 @cindex writing tests
7777 In many areas, the @value{GDBN} tests are already quite comprehensive; you
7778 should be able to copy existing tests to handle new cases.
7780 You should try to use @code{gdb_test} whenever possible, since it
7781 includes cases to handle all the unexpected errors that might happen.
7782 However, it doesn't cost anything to add new test procedures; for
7783 instance, @file{gdb.base/exprs.exp} defines a @code{test_expr} that
7784 calls @code{gdb_test} multiple times.
7786 Only use @code{send_gdb} and @code{gdb_expect} when absolutely
7787 necessary. Even if @value{GDBN} has several valid responses to
7788 a command, you can use @code{gdb_test_multiple}. Like @code{gdb_test},
7789 @code{gdb_test_multiple} recognizes internal errors and unexpected
7792 Do not write tests which expect a literal tab character from @value{GDBN}.
7793 On some operating systems (e.g.@: OpenBSD) the TTY layer expands tabs to
7794 spaces, so by the time @value{GDBN}'s output reaches expect the tab is gone.
7796 The source language programs do @emph{not} need to be in a consistent
7797 style. Since @value{GDBN} is used to debug programs written in many different
7798 styles, it's worth having a mix of styles in the testsuite; for
7799 instance, some @value{GDBN} bugs involving the display of source lines would
7800 never manifest themselves if the programs used GNU coding style
7807 Check the @file{README} file, it often has useful information that does not
7808 appear anywhere else in the directory.
7811 * Getting Started:: Getting started working on @value{GDBN}
7812 * Debugging GDB:: Debugging @value{GDBN} with itself
7815 @node Getting Started
7817 @section Getting Started
7819 @value{GDBN} is a large and complicated program, and if you first starting to
7820 work on it, it can be hard to know where to start. Fortunately, if you
7821 know how to go about it, there are ways to figure out what is going on.
7823 This manual, the @value{GDBN} Internals manual, has information which applies
7824 generally to many parts of @value{GDBN}.
7826 Information about particular functions or data structures are located in
7827 comments with those functions or data structures. If you run across a
7828 function or a global variable which does not have a comment correctly
7829 explaining what is does, this can be thought of as a bug in @value{GDBN}; feel
7830 free to submit a bug report, with a suggested comment if you can figure
7831 out what the comment should say. If you find a comment which is
7832 actually wrong, be especially sure to report that.
7834 Comments explaining the function of macros defined in host, target, or
7835 native dependent files can be in several places. Sometimes they are
7836 repeated every place the macro is defined. Sometimes they are where the
7837 macro is used. Sometimes there is a header file which supplies a
7838 default definition of the macro, and the comment is there. This manual
7839 also documents all the available macros.
7840 @c (@pxref{Host Conditionals}, @pxref{Target
7841 @c Conditionals}, @pxref{Native Conditionals}, and @pxref{Obsolete
7844 Start with the header files. Once you have some idea of how
7845 @value{GDBN}'s internal symbol tables are stored (see @file{symtab.h},
7846 @file{gdbtypes.h}), you will find it much easier to understand the
7847 code which uses and creates those symbol tables.
7849 You may wish to process the information you are getting somehow, to
7850 enhance your understanding of it. Summarize it, translate it to another
7851 language, add some (perhaps trivial or non-useful) feature to @value{GDBN}, use
7852 the code to predict what a test case would do and write the test case
7853 and verify your prediction, etc. If you are reading code and your eyes
7854 are starting to glaze over, this is a sign you need to use a more active
7857 Once you have a part of @value{GDBN} to start with, you can find more
7858 specifically the part you are looking for by stepping through each
7859 function with the @code{next} command. Do not use @code{step} or you
7860 will quickly get distracted; when the function you are stepping through
7861 calls another function try only to get a big-picture understanding
7862 (perhaps using the comment at the beginning of the function being
7863 called) of what it does. This way you can identify which of the
7864 functions being called by the function you are stepping through is the
7865 one which you are interested in. You may need to examine the data
7866 structures generated at each stage, with reference to the comments in
7867 the header files explaining what the data structures are supposed to
7870 Of course, this same technique can be used if you are just reading the
7871 code, rather than actually stepping through it. The same general
7872 principle applies---when the code you are looking at calls something
7873 else, just try to understand generally what the code being called does,
7874 rather than worrying about all its details.
7876 @cindex command implementation
7877 A good place to start when tracking down some particular area is with
7878 a command which invokes that feature. Suppose you want to know how
7879 single-stepping works. As a @value{GDBN} user, you know that the
7880 @code{step} command invokes single-stepping. The command is invoked
7881 via command tables (see @file{command.h}); by convention the function
7882 which actually performs the command is formed by taking the name of
7883 the command and adding @samp{_command}, or in the case of an
7884 @code{info} subcommand, @samp{_info}. For example, the @code{step}
7885 command invokes the @code{step_command} function and the @code{info
7886 display} command invokes @code{display_info}. When this convention is
7887 not followed, you might have to use @code{grep} or @kbd{M-x
7888 tags-search} in emacs, or run @value{GDBN} on itself and set a
7889 breakpoint in @code{execute_command}.
7891 @cindex @code{bug-gdb} mailing list
7892 If all of the above fail, it may be appropriate to ask for information
7893 on @code{bug-gdb}. But @emph{never} post a generic question like ``I was
7894 wondering if anyone could give me some tips about understanding
7895 @value{GDBN}''---if we had some magic secret we would put it in this manual.
7896 Suggestions for improving the manual are always welcome, of course.
7900 @section Debugging @value{GDBN} with itself
7901 @cindex debugging @value{GDBN}
7903 If @value{GDBN} is limping on your machine, this is the preferred way to get it
7904 fully functional. Be warned that in some ancient Unix systems, like
7905 Ultrix 4.2, a program can't be running in one process while it is being
7906 debugged in another. Rather than typing the command @kbd{@w{./gdb
7907 ./gdb}}, which works on Suns and such, you can copy @file{gdb} to
7908 @file{gdb2} and then type @kbd{@w{./gdb ./gdb2}}.
7910 When you run @value{GDBN} in the @value{GDBN} source directory, it will read a
7911 @file{.gdbinit} file that sets up some simple things to make debugging
7912 gdb easier. The @code{info} command, when executed without a subcommand
7913 in a @value{GDBN} being debugged by gdb, will pop you back up to the top level
7914 gdb. See @file{.gdbinit} for details.
7916 If you use emacs, you will probably want to do a @code{make TAGS} after
7917 you configure your distribution; this will put the machine dependent
7918 routines for your local machine where they will be accessed first by
7921 Also, make sure that you've either compiled @value{GDBN} with your local cc, or
7922 have run @code{fixincludes} if you are compiling with gcc.
7924 @section Submitting Patches
7926 @cindex submitting patches
7927 Thanks for thinking of offering your changes back to the community of
7928 @value{GDBN} users. In general we like to get well designed enhancements.
7929 Thanks also for checking in advance about the best way to transfer the
7932 The @value{GDBN} maintainers will only install ``cleanly designed'' patches.
7933 This manual summarizes what we believe to be clean design for @value{GDBN}.
7935 If the maintainers don't have time to put the patch in when it arrives,
7936 or if there is any question about a patch, it goes into a large queue
7937 with everyone else's patches and bug reports.
7939 @cindex legal papers for code contributions
7940 The legal issue is that to incorporate substantial changes requires a
7941 copyright assignment from you and/or your employer, granting ownership
7942 of the changes to the Free Software Foundation. You can get the
7943 standard documents for doing this by sending mail to @code{gnu@@gnu.org}
7944 and asking for it. We recommend that people write in "All programs
7945 owned by the Free Software Foundation" as "NAME OF PROGRAM", so that
7946 changes in many programs (not just @value{GDBN}, but GAS, Emacs, GCC,
7948 contributed with only one piece of legalese pushed through the
7949 bureaucracy and filed with the FSF. We can't start merging changes until
7950 this paperwork is received by the FSF (their rules, which we follow
7951 since we maintain it for them).
7953 Technically, the easiest way to receive changes is to receive each
7954 feature as a small context diff or unidiff, suitable for @code{patch}.
7955 Each message sent to me should include the changes to C code and
7956 header files for a single feature, plus @file{ChangeLog} entries for
7957 each directory where files were modified, and diffs for any changes
7958 needed to the manuals (@file{gdb/doc/gdb.texinfo} or
7959 @file{gdb/doc/gdbint.texinfo}). If there are a lot of changes for a
7960 single feature, they can be split down into multiple messages.
7962 In this way, if we read and like the feature, we can add it to the
7963 sources with a single patch command, do some testing, and check it in.
7964 If you leave out the @file{ChangeLog}, we have to write one. If you leave
7965 out the doc, we have to puzzle out what needs documenting. Etc., etc.
7967 The reason to send each change in a separate message is that we will not
7968 install some of the changes. They'll be returned to you with questions
7969 or comments. If we're doing our job correctly, the message back to you
7970 will say what you have to fix in order to make the change acceptable.
7971 The reason to have separate messages for separate features is so that
7972 the acceptable changes can be installed while one or more changes are
7973 being reworked. If multiple features are sent in a single message, we
7974 tend to not put in the effort to sort out the acceptable changes from
7975 the unacceptable, so none of the features get installed until all are
7978 If this sounds painful or authoritarian, well, it is. But we get a lot
7979 of bug reports and a lot of patches, and many of them don't get
7980 installed because we don't have the time to finish the job that the bug
7981 reporter or the contributor could have done. Patches that arrive
7982 complete, working, and well designed, tend to get installed on the day
7983 they arrive. The others go into a queue and get installed as time
7984 permits, which, since the maintainers have many demands to meet, may not
7985 be for quite some time.
7987 Please send patches directly to
7988 @email{gdb-patches@@sourceware.org, the @value{GDBN} maintainers}.
7990 @section Build Script
7992 @cindex build script
7994 The script @file{gdb_buildall.sh} builds @value{GDBN} with flag
7995 @option{--enable-targets=all} set. This builds @value{GDBN} with all supported
7996 targets activated. This helps testing @value{GDBN} when doing changes that
7997 affect more than one architecture and is much faster than using
7998 @file{gdb_mbuild.sh}.
8000 After building @value{GDBN} the script checks which architectures are
8001 supported and then switches the current architecture to each of those to get
8002 information about the architecture. The test results are stored in log files
8003 in the directory the script was called from.
8005 @include observer.texi