1 \input texinfo @c -*- texinfo -*-
2 @setfilename gdbint.info
4 @dircategory Software development
6 * Gdb-Internals: (gdbint). The GNU debugger's internals.
10 This file documents the internals of the GNU debugger @value{GDBN}.
11 Copyright 1990,1991,1992,1993,1994,1996,1998,1999,2000,2001,2002,2003,2004
12 Free Software Foundation, Inc.
13 Contributed by Cygnus Solutions. Written by John Gilmore.
14 Second Edition by Stan Shebs.
16 Permission is granted to copy, distribute and/or modify this document
17 under the terms of the GNU Free Documentation License, Version 1.1 or
18 any later version published by the Free Software Foundation; with no
19 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
20 Texts. A copy of the license is included in the section entitled ``GNU
21 Free Documentation License''.
24 @setchapternewpage off
25 @settitle @value{GDBN} Internals
31 @title @value{GDBN} Internals
32 @subtitle{A guide to the internals of the GNU debugger}
34 @author Cygnus Solutions
35 @author Second Edition:
37 @author Cygnus Solutions
40 \def\$#1${{#1}} % Kluge: collect RCS revision info without $...$
41 \xdef\manvers{\$Revision$} % For use in headers, footers too
43 \hfill Cygnus Solutions\par
45 \hfill \TeX{}info \texinfoversion\par
49 @vskip 0pt plus 1filll
50 Copyright @copyright{} 1990,1991,1992,1993,1994,1996,1998,1999,2000,2001,
51 2002, 2003, 2004 Free Software Foundation, Inc.
53 Permission is granted to copy, distribute and/or modify this document
54 under the terms of the GNU Free Documentation License, Version 1.1 or
55 any later version published by the Free Software Foundation; with no
56 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
57 Texts. A copy of the license is included in the section entitled ``GNU
58 Free Documentation License''.
64 @c Perhaps this should be the title of the document (but only for info,
65 @c not for TeX). Existing GNU manuals seem inconsistent on this point.
66 @top Scope of this Document
68 This document documents the internals of the GNU debugger, @value{GDBN}. It
69 includes description of @value{GDBN}'s key algorithms and operations, as well
70 as the mechanisms that adapt @value{GDBN} to specific hosts and targets.
81 * Target Architecture Definition::
82 * Target Vector Definition::
87 * Versions and Branches::
92 * GDB Observers:: @value{GDBN} Currently available observers
93 * GNU Free Documentation License:: The license for this documentation
100 @cindex requirements for @value{GDBN}
102 Before diving into the internals, you should understand the formal
103 requirements and other expectations for @value{GDBN}. Although some
104 of these may seem obvious, there have been proposals for @value{GDBN}
105 that have run counter to these requirements.
107 First of all, @value{GDBN} is a debugger. It's not designed to be a
108 front panel for embedded systems. It's not a text editor. It's not a
109 shell. It's not a programming environment.
111 @value{GDBN} is an interactive tool. Although a batch mode is
112 available, @value{GDBN}'s primary role is to interact with a human
115 @value{GDBN} should be responsive to the user. A programmer hot on
116 the trail of a nasty bug, and operating under a looming deadline, is
117 going to be very impatient of everything, including the response time
118 to debugger commands.
120 @value{GDBN} should be relatively permissive, such as for expressions.
121 While the compiler should be picky (or have the option to be made
122 picky), since source code lives for a long time usually, the
123 programmer doing debugging shouldn't be spending time figuring out to
124 mollify the debugger.
126 @value{GDBN} will be called upon to deal with really large programs.
127 Executable sizes of 50 to 100 megabytes occur regularly, and we've
128 heard reports of programs approaching 1 gigabyte in size.
130 @value{GDBN} should be able to run everywhere. No other debugger is
131 available for even half as many configurations as @value{GDBN}
135 @node Overall Structure
137 @chapter Overall Structure
139 @value{GDBN} consists of three major subsystems: user interface,
140 symbol handling (the @dfn{symbol side}), and target system handling (the
143 The user interface consists of several actual interfaces, plus
146 The symbol side consists of object file readers, debugging info
147 interpreters, symbol table management, source language expression
148 parsing, type and value printing.
150 The target side consists of execution control, stack frame analysis, and
151 physical target manipulation.
153 The target side/symbol side division is not formal, and there are a
154 number of exceptions. For instance, core file support involves symbolic
155 elements (the basic core file reader is in BFD) and target elements (it
156 supplies the contents of memory and the values of registers). Instead,
157 this division is useful for understanding how the minor subsystems
160 @section The Symbol Side
162 The symbolic side of @value{GDBN} can be thought of as ``everything
163 you can do in @value{GDBN} without having a live program running''.
164 For instance, you can look at the types of variables, and evaluate
165 many kinds of expressions.
167 @section The Target Side
169 The target side of @value{GDBN} is the ``bits and bytes manipulator''.
170 Although it may make reference to symbolic info here and there, most
171 of the target side will run with only a stripped executable
172 available---or even no executable at all, in remote debugging cases.
174 Operations such as disassembly, stack frame crawls, and register
175 display, are able to work with no symbolic info at all. In some cases,
176 such as disassembly, @value{GDBN} will use symbolic info to present addresses
177 relative to symbols rather than as raw numbers, but it will work either
180 @section Configurations
184 @dfn{Host} refers to attributes of the system where @value{GDBN} runs.
185 @dfn{Target} refers to the system where the program being debugged
186 executes. In most cases they are the same machine, in which case a
187 third type of @dfn{Native} attributes come into play.
189 Defines and include files needed to build on the host are host support.
190 Examples are tty support, system defined types, host byte order, host
193 Defines and information needed to handle the target format are target
194 dependent. Examples are the stack frame format, instruction set,
195 breakpoint instruction, registers, and how to set up and tear down the stack
198 Information that is only needed when the host and target are the same,
199 is native dependent. One example is Unix child process support; if the
200 host and target are not the same, doing a fork to start the target
201 process is a bad idea. The various macros needed for finding the
202 registers in the @code{upage}, running @code{ptrace}, and such are all
203 in the native-dependent files.
205 Another example of native-dependent code is support for features that
206 are really part of the target environment, but which require
207 @code{#include} files that are only available on the host system. Core
208 file handling and @code{setjmp} handling are two common cases.
210 When you want to make @value{GDBN} work ``native'' on a particular machine, you
211 have to include all three kinds of information.
219 @value{GDBN} uses a number of debugging-specific algorithms. They are
220 often not very complicated, but get lost in the thicket of special
221 cases and real-world issues. This chapter describes the basic
222 algorithms and mentions some of the specific target definitions that
228 @cindex call stack frame
229 A frame is a construct that @value{GDBN} uses to keep track of calling
230 and called functions.
232 @findex create_new_frame
234 @code{FRAME_FP} in the machine description has no meaning to the
235 machine-independent part of @value{GDBN}, except that it is used when
236 setting up a new frame from scratch, as follows:
239 create_new_frame (read_register (DEPRECATED_FP_REGNUM), read_pc ()));
242 @cindex frame pointer register
243 Other than that, all the meaning imparted to @code{DEPRECATED_FP_REGNUM}
244 is imparted by the machine-dependent code. So,
245 @code{DEPRECATED_FP_REGNUM} can have any value that is convenient for
246 the code that creates new frames. (@code{create_new_frame} calls
247 @code{DEPRECATED_INIT_EXTRA_FRAME_INFO} if it is defined; that is where
248 you should use the @code{DEPRECATED_FP_REGNUM} value, if your frames are
252 Given a @value{GDBN} frame, define @code{DEPRECATED_FRAME_CHAIN} to
253 determine the address of the calling function's frame. This will be
254 used to create a new @value{GDBN} frame struct, and then
255 @code{DEPRECATED_INIT_EXTRA_FRAME_INFO} and
256 @code{DEPRECATED_INIT_FRAME_PC} will be called for the new frame.
258 @section Breakpoint Handling
261 In general, a breakpoint is a user-designated location in the program
262 where the user wants to regain control if program execution ever reaches
265 There are two main ways to implement breakpoints; either as ``hardware''
266 breakpoints or as ``software'' breakpoints.
268 @cindex hardware breakpoints
269 @cindex program counter
270 Hardware breakpoints are sometimes available as a builtin debugging
271 features with some chips. Typically these work by having dedicated
272 register into which the breakpoint address may be stored. If the PC
273 (shorthand for @dfn{program counter})
274 ever matches a value in a breakpoint registers, the CPU raises an
275 exception and reports it to @value{GDBN}.
277 Another possibility is when an emulator is in use; many emulators
278 include circuitry that watches the address lines coming out from the
279 processor, and force it to stop if the address matches a breakpoint's
282 A third possibility is that the target already has the ability to do
283 breakpoints somehow; for instance, a ROM monitor may do its own
284 software breakpoints. So although these are not literally ``hardware
285 breakpoints'', from @value{GDBN}'s point of view they work the same;
286 @value{GDBN} need not do anything more than set the breakpoint and wait
287 for something to happen.
289 Since they depend on hardware resources, hardware breakpoints may be
290 limited in number; when the user asks for more, @value{GDBN} will
291 start trying to set software breakpoints. (On some architectures,
292 notably the 32-bit x86 platforms, @value{GDBN} cannot always know
293 whether there's enough hardware resources to insert all the hardware
294 breakpoints and watchpoints. On those platforms, @value{GDBN} prints
295 an error message only when the program being debugged is continued.)
297 @cindex software breakpoints
298 Software breakpoints require @value{GDBN} to do somewhat more work.
299 The basic theory is that @value{GDBN} will replace a program
300 instruction with a trap, illegal divide, or some other instruction
301 that will cause an exception, and then when it's encountered,
302 @value{GDBN} will take the exception and stop the program. When the
303 user says to continue, @value{GDBN} will restore the original
304 instruction, single-step, re-insert the trap, and continue on.
306 Since it literally overwrites the program being tested, the program area
307 must be writable, so this technique won't work on programs in ROM. It
308 can also distort the behavior of programs that examine themselves,
309 although such a situation would be highly unusual.
311 Also, the software breakpoint instruction should be the smallest size of
312 instruction, so it doesn't overwrite an instruction that might be a jump
313 target, and cause disaster when the program jumps into the middle of the
314 breakpoint instruction. (Strictly speaking, the breakpoint must be no
315 larger than the smallest interval between instructions that may be jump
316 targets; perhaps there is an architecture where only even-numbered
317 instructions may jumped to.) Note that it's possible for an instruction
318 set not to have any instructions usable for a software breakpoint,
319 although in practice only the ARC has failed to define such an
323 The basic definition of the software breakpoint is the macro
326 Basic breakpoint object handling is in @file{breakpoint.c}. However,
327 much of the interesting breakpoint action is in @file{infrun.c}.
329 @section Single Stepping
331 @section Signal Handling
333 @section Thread Handling
335 @section Inferior Function Calls
337 @section Longjmp Support
339 @cindex @code{longjmp} debugging
340 @value{GDBN} has support for figuring out that the target is doing a
341 @code{longjmp} and for stopping at the target of the jump, if we are
342 stepping. This is done with a few specialized internal breakpoints,
343 which are visible in the output of the @samp{maint info breakpoint}
346 @findex GET_LONGJMP_TARGET
347 To make this work, you need to define a macro called
348 @code{GET_LONGJMP_TARGET}, which will examine the @code{jmp_buf}
349 structure and extract the longjmp target address. Since @code{jmp_buf}
350 is target specific, you will need to define it in the appropriate
351 @file{tm-@var{target}.h} file. Look in @file{tm-sun4os4.h} and
352 @file{sparc-tdep.c} for examples of how to do this.
357 Watchpoints are a special kind of breakpoints (@pxref{Algorithms,
358 breakpoints}) which break when data is accessed rather than when some
359 instruction is executed. When you have data which changes without
360 your knowing what code does that, watchpoints are the silver bullet to
361 hunt down and kill such bugs.
363 @cindex hardware watchpoints
364 @cindex software watchpoints
365 Watchpoints can be either hardware-assisted or not; the latter type is
366 known as ``software watchpoints.'' @value{GDBN} always uses
367 hardware-assisted watchpoints if they are available, and falls back on
368 software watchpoints otherwise. Typical situations where @value{GDBN}
369 will use software watchpoints are:
373 The watched memory region is too large for the underlying hardware
374 watchpoint support. For example, each x86 debug register can watch up
375 to 4 bytes of memory, so trying to watch data structures whose size is
376 more than 16 bytes will cause @value{GDBN} to use software
380 The value of the expression to be watched depends on data held in
381 registers (as opposed to memory).
384 Too many different watchpoints requested. (On some architectures,
385 this situation is impossible to detect until the debugged program is
386 resumed.) Note that x86 debug registers are used both for hardware
387 breakpoints and for watchpoints, so setting too many hardware
388 breakpoints might cause watchpoint insertion to fail.
391 No hardware-assisted watchpoints provided by the target
395 Software watchpoints are very slow, since @value{GDBN} needs to
396 single-step the program being debugged and test the value of the
397 watched expression(s) after each instruction. The rest of this
398 section is mostly irrelevant for software watchpoints.
400 When the inferior stops, @value{GDBN} tries to establish, among other
401 possible reasons, whether it stopped due to a watchpoint being hit.
402 For a data-write watchpoint, it does so by evaluating, for each
403 watchpoint, the expression whose value is being watched, and testing
404 whether the watched value has changed. For data-read and data-access
405 watchpoints, @value{GDBN} needs the target to supply a primitive that
406 returns the address of the data that was accessed or read (see the
407 description of @code{target_stopped_data_address} below): if this
408 primitive returns a valid address, @value{GDBN} infers that a
409 watchpoint triggered if it watches an expression whose evaluation uses
412 @value{GDBN} uses several macros and primitives to support hardware
416 @findex TARGET_HAS_HARDWARE_WATCHPOINTS
417 @item TARGET_HAS_HARDWARE_WATCHPOINTS
418 If defined, the target supports hardware watchpoints.
420 @findex TARGET_CAN_USE_HARDWARE_WATCHPOINT
421 @item TARGET_CAN_USE_HARDWARE_WATCHPOINT (@var{type}, @var{count}, @var{other})
422 Return the number of hardware watchpoints of type @var{type} that are
423 possible to be set. The value is positive if @var{count} watchpoints
424 of this type can be set, zero if setting watchpoints of this type is
425 not supported, and negative if @var{count} is more than the maximum
426 number of watchpoints of type @var{type} that can be set. @var{other}
427 is non-zero if other types of watchpoints are currently enabled (there
428 are architectures which cannot set watchpoints of different types at
431 @findex TARGET_REGION_OK_FOR_HW_WATCHPOINT
432 @item TARGET_REGION_OK_FOR_HW_WATCHPOINT (@var{addr}, @var{len})
433 Return non-zero if hardware watchpoints can be used to watch a region
434 whose address is @var{addr} and whose length in bytes is @var{len}.
436 @findex TARGET_REGION_SIZE_OK_FOR_HW_WATCHPOINT
437 @item TARGET_REGION_SIZE_OK_FOR_HW_WATCHPOINT (@var{size})
438 Return non-zero if hardware watchpoints can be used to watch a region
439 whose size is @var{size}. @value{GDBN} only uses this macro as a
440 fall-back, in case @code{TARGET_REGION_OK_FOR_HW_WATCHPOINT} is not
443 @cindex insert or remove hardware watchpoint
444 @findex target_insert_watchpoint
445 @findex target_remove_watchpoint
446 @item target_insert_watchpoint (@var{addr}, @var{len}, @var{type})
447 @itemx target_remove_watchpoint (@var{addr}, @var{len}, @var{type})
448 Insert or remove a hardware watchpoint starting at @var{addr}, for
449 @var{len} bytes. @var{type} is the watchpoint type, one of the
450 possible values of the enumerated data type @code{target_hw_bp_type},
451 defined by @file{breakpoint.h} as follows:
454 enum target_hw_bp_type
456 hw_write = 0, /* Common (write) HW watchpoint */
457 hw_read = 1, /* Read HW watchpoint */
458 hw_access = 2, /* Access (read or write) HW watchpoint */
459 hw_execute = 3 /* Execute HW breakpoint */
464 These two macros should return 0 for success, non-zero for failure.
466 @cindex insert or remove hardware breakpoint
467 @findex target_remove_hw_breakpoint
468 @findex target_insert_hw_breakpoint
469 @item target_remove_hw_breakpoint (@var{addr}, @var{shadow})
470 @itemx target_insert_hw_breakpoint (@var{addr}, @var{shadow})
471 Insert or remove a hardware-assisted breakpoint at address @var{addr}.
472 Returns zero for success, non-zero for failure. @var{shadow} is the
473 real contents of the byte where the breakpoint has been inserted; it
474 is generally not valid when hardware breakpoints are used, but since
475 no other code touches these values, the implementations of the above
476 two macros can use them for their internal purposes.
478 @findex target_stopped_data_address
479 @item target_stopped_data_address (@var{addr_p})
480 If the inferior has some watchpoint that triggered, place the address
481 associated with the watchpoint at the location pointed to by
482 @var{addr_p} and return non-zero. Otherwise, return zero. Note that
483 this primitive is used by @value{GDBN} only on targets that support
484 data-read or data-access type watchpoints, so targets that have
485 support only for data-write watchpoints need not implement these
488 @findex HAVE_STEPPABLE_WATCHPOINT
489 @item HAVE_STEPPABLE_WATCHPOINT
490 If defined to a non-zero value, it is not necessary to disable a
491 watchpoint to step over it.
493 @findex HAVE_NONSTEPPABLE_WATCHPOINT
494 @item HAVE_NONSTEPPABLE_WATCHPOINT
495 If defined to a non-zero value, @value{GDBN} should disable a
496 watchpoint to step the inferior over it.
498 @findex HAVE_CONTINUABLE_WATCHPOINT
499 @item HAVE_CONTINUABLE_WATCHPOINT
500 If defined to a non-zero value, it is possible to continue the
501 inferior after a watchpoint has been hit.
503 @findex CANNOT_STEP_HW_WATCHPOINTS
504 @item CANNOT_STEP_HW_WATCHPOINTS
505 If this is defined to a non-zero value, @value{GDBN} will remove all
506 watchpoints before stepping the inferior.
508 @findex STOPPED_BY_WATCHPOINT
509 @item STOPPED_BY_WATCHPOINT (@var{wait_status})
510 Return non-zero if stopped by a watchpoint. @var{wait_status} is of
511 the type @code{struct target_waitstatus}, defined by @file{target.h}.
512 Normally, this macro is defined to invoke the function pointed to by
513 the @code{to_stopped_by_watchpoint} member of the structure (of the
514 type @code{target_ops}, defined on @file{target.h}) that describes the
515 target-specific operations; @code{to_stopped_by_watchpoint} ignores
516 the @var{wait_status} argument.
518 @value{GDBN} does not require the non-zero value returned by
519 @code{STOPPED_BY_WATCHPOINT} to be 100% correct, so if a target cannot
520 determine for sure whether the inferior stopped due to a watchpoint,
521 it could return non-zero ``just in case''.
524 @subsection x86 Watchpoints
525 @cindex x86 debug registers
526 @cindex watchpoints, on x86
528 The 32-bit Intel x86 (a.k.a.@: ia32) processors feature special debug
529 registers designed to facilitate debugging. @value{GDBN} provides a
530 generic library of functions that x86-based ports can use to implement
531 support for watchpoints and hardware-assisted breakpoints. This
532 subsection documents the x86 watchpoint facilities in @value{GDBN}.
534 To use the generic x86 watchpoint support, a port should do the
538 @findex I386_USE_GENERIC_WATCHPOINTS
540 Define the macro @code{I386_USE_GENERIC_WATCHPOINTS} somewhere in the
541 target-dependent headers.
544 Include the @file{config/i386/nm-i386.h} header file @emph{after}
545 defining @code{I386_USE_GENERIC_WATCHPOINTS}.
548 Add @file{i386-nat.o} to the value of the Make variable
549 @code{NATDEPFILES} (@pxref{Native Debugging, NATDEPFILES}) or
550 @code{TDEPFILES} (@pxref{Target Architecture Definition, TDEPFILES}).
553 Provide implementations for the @code{I386_DR_LOW_*} macros described
554 below. Typically, each macro should call a target-specific function
555 which does the real work.
558 The x86 watchpoint support works by maintaining mirror images of the
559 debug registers. Values are copied between the mirror images and the
560 real debug registers via a set of macros which each target needs to
564 @findex I386_DR_LOW_SET_CONTROL
565 @item I386_DR_LOW_SET_CONTROL (@var{val})
566 Set the Debug Control (DR7) register to the value @var{val}.
568 @findex I386_DR_LOW_SET_ADDR
569 @item I386_DR_LOW_SET_ADDR (@var{idx}, @var{addr})
570 Put the address @var{addr} into the debug register number @var{idx}.
572 @findex I386_DR_LOW_RESET_ADDR
573 @item I386_DR_LOW_RESET_ADDR (@var{idx})
574 Reset (i.e.@: zero out) the address stored in the debug register
577 @findex I386_DR_LOW_GET_STATUS
578 @item I386_DR_LOW_GET_STATUS
579 Return the value of the Debug Status (DR6) register. This value is
580 used immediately after it is returned by
581 @code{I386_DR_LOW_GET_STATUS}, so as to support per-thread status
585 For each one of the 4 debug registers (whose indices are from 0 to 3)
586 that store addresses, a reference count is maintained by @value{GDBN},
587 to allow sharing of debug registers by several watchpoints. This
588 allows users to define several watchpoints that watch the same
589 expression, but with different conditions and/or commands, without
590 wasting debug registers which are in short supply. @value{GDBN}
591 maintains the reference counts internally, targets don't have to do
592 anything to use this feature.
594 The x86 debug registers can each watch a region that is 1, 2, or 4
595 bytes long. The ia32 architecture requires that each watched region
596 be appropriately aligned: 2-byte region on 2-byte boundary, 4-byte
597 region on 4-byte boundary. However, the x86 watchpoint support in
598 @value{GDBN} can watch unaligned regions and regions larger than 4
599 bytes (up to 16 bytes) by allocating several debug registers to watch
600 a single region. This allocation of several registers per a watched
601 region is also done automatically without target code intervention.
603 The generic x86 watchpoint support provides the following API for the
604 @value{GDBN}'s application code:
607 @findex i386_region_ok_for_watchpoint
608 @item i386_region_ok_for_watchpoint (@var{addr}, @var{len})
609 The macro @code{TARGET_REGION_OK_FOR_HW_WATCHPOINT} is set to call
610 this function. It counts the number of debug registers required to
611 watch a given region, and returns a non-zero value if that number is
612 less than 4, the number of debug registers available to x86
615 @findex i386_stopped_data_address
616 @item i386_stopped_data_address (@var{addr_p})
618 @code{target_stopped_data_address} is set to call this function.
620 function examines the breakpoint condition bits in the DR6 Debug
621 Status register, as returned by the @code{I386_DR_LOW_GET_STATUS}
622 macro, and returns the address associated with the first bit that is
625 @findex i386_stopped_by_watchpoint
626 @item i386_stopped_by_watchpoint (void)
627 The macro @code{STOPPED_BY_WATCHPOINT}
628 is set to call this function. The
629 argument passed to @code{STOPPED_BY_WATCHPOINT} is ignored. This
630 function examines the breakpoint condition bits in the DR6 Debug
631 Status register, as returned by the @code{I386_DR_LOW_GET_STATUS}
632 macro, and returns true if any bit is set. Otherwise, false is
635 @findex i386_insert_watchpoint
636 @findex i386_remove_watchpoint
637 @item i386_insert_watchpoint (@var{addr}, @var{len}, @var{type})
638 @itemx i386_remove_watchpoint (@var{addr}, @var{len}, @var{type})
639 Insert or remove a watchpoint. The macros
640 @code{target_insert_watchpoint} and @code{target_remove_watchpoint}
641 are set to call these functions. @code{i386_insert_watchpoint} first
642 looks for a debug register which is already set to watch the same
643 region for the same access types; if found, it just increments the
644 reference count of that debug register, thus implementing debug
645 register sharing between watchpoints. If no such register is found,
646 the function looks for a vacant debug register, sets its mirrored
647 value to @var{addr}, sets the mirrored value of DR7 Debug Control
648 register as appropriate for the @var{len} and @var{type} parameters,
649 and then passes the new values of the debug register and DR7 to the
650 inferior by calling @code{I386_DR_LOW_SET_ADDR} and
651 @code{I386_DR_LOW_SET_CONTROL}. If more than one debug register is
652 required to cover the given region, the above process is repeated for
655 @code{i386_remove_watchpoint} does the opposite: it resets the address
656 in the mirrored value of the debug register and its read/write and
657 length bits in the mirrored value of DR7, then passes these new
658 values to the inferior via @code{I386_DR_LOW_RESET_ADDR} and
659 @code{I386_DR_LOW_SET_CONTROL}. If a register is shared by several
660 watchpoints, each time a @code{i386_remove_watchpoint} is called, it
661 decrements the reference count, and only calls
662 @code{I386_DR_LOW_RESET_ADDR} and @code{I386_DR_LOW_SET_CONTROL} when
663 the count goes to zero.
665 @findex i386_insert_hw_breakpoint
666 @findex i386_remove_hw_breakpoint
667 @item i386_insert_hw_breakpoint (@var{addr}, @var{shadow}
668 @itemx i386_remove_hw_breakpoint (@var{addr}, @var{shadow})
669 These functions insert and remove hardware-assisted breakpoints. The
670 macros @code{target_insert_hw_breakpoint} and
671 @code{target_remove_hw_breakpoint} are set to call these functions.
672 These functions work like @code{i386_insert_watchpoint} and
673 @code{i386_remove_watchpoint}, respectively, except that they set up
674 the debug registers to watch instruction execution, and each
675 hardware-assisted breakpoint always requires exactly one debug
678 @findex i386_stopped_by_hwbp
679 @item i386_stopped_by_hwbp (void)
680 This function returns non-zero if the inferior has some watchpoint or
681 hardware breakpoint that triggered. It works like
682 @code{i386_stopped_data_address}, except that it doesn't record the
683 address whose watchpoint triggered.
685 @findex i386_cleanup_dregs
686 @item i386_cleanup_dregs (void)
687 This function clears all the reference counts, addresses, and control
688 bits in the mirror images of the debug registers. It doesn't affect
689 the actual debug registers in the inferior process.
696 x86 processors support setting watchpoints on I/O reads or writes.
697 However, since no target supports this (as of March 2001), and since
698 @code{enum target_hw_bp_type} doesn't even have an enumeration for I/O
699 watchpoints, this feature is not yet available to @value{GDBN} running
703 x86 processors can enable watchpoints locally, for the current task
704 only, or globally, for all the tasks. For each debug register,
705 there's a bit in the DR7 Debug Control register that determines
706 whether the associated address is watched locally or globally. The
707 current implementation of x86 watchpoint support in @value{GDBN}
708 always sets watchpoints to be locally enabled, since global
709 watchpoints might interfere with the underlying OS and are probably
710 unavailable in many platforms.
713 @section Observing changes in @value{GDBN} internals
714 @cindex observer pattern interface
715 @cindex notifications about changes in internals
717 In order to function properly, several modules need to be notified when
718 some changes occur in the @value{GDBN} internals. Traditionally, these
719 modules have relied on several paradigms, the most common ones being
720 hooks and gdb-events. Unfortunately, none of these paradigms was
721 versatile enough to become the standard notification mechanism in
722 @value{GDBN}. The fact that they only supported one ``client'' was also
725 A new paradigm, based on the Observer pattern of the @cite{Design
726 Patterns} book, has therefore been implemented. The goal was to provide
727 a new interface overcoming the issues with the notification mechanisms
728 previously available. This new interface needed to be strongly typed,
729 easy to extend, and versatile enough to be used as the standard
730 interface when adding new notifications.
732 See @ref{GDB Observers} for a brief description of the observers
733 currently implemented in GDB. The rationale for the current
734 implementation is also briefly discussed.
738 @chapter User Interface
740 @value{GDBN} has several user interfaces. Although the command-line interface
741 is the most common and most familiar, there are others.
743 @section Command Interpreter
745 @cindex command interpreter
747 The command interpreter in @value{GDBN} is fairly simple. It is designed to
748 allow for the set of commands to be augmented dynamically, and also
749 has a recursive subcommand capability, where the first argument to
750 a command may itself direct a lookup on a different command list.
752 For instance, the @samp{set} command just starts a lookup on the
753 @code{setlist} command list, while @samp{set thread} recurses
754 to the @code{set_thread_cmd_list}.
758 To add commands in general, use @code{add_cmd}. @code{add_com} adds to
759 the main command list, and should be used for those commands. The usual
760 place to add commands is in the @code{_initialize_@var{xyz}} routines at
761 the ends of most source files.
763 @findex add_setshow_cmd
764 @findex add_setshow_cmd_full
765 To add paired @samp{set} and @samp{show} commands, use
766 @code{add_setshow_cmd} or @code{add_setshow_cmd_full}. The former is
767 a slightly simpler interface which is useful when you don't need to
768 further modify the new command structures, while the latter returns
769 the new command structures for manipulation.
771 @cindex deprecating commands
772 @findex deprecate_cmd
773 Before removing commands from the command set it is a good idea to
774 deprecate them for some time. Use @code{deprecate_cmd} on commands or
775 aliases to set the deprecated flag. @code{deprecate_cmd} takes a
776 @code{struct cmd_list_element} as it's first argument. You can use the
777 return value from @code{add_com} or @code{add_cmd} to deprecate the
778 command immediately after it is created.
780 The first time a command is used the user will be warned and offered a
781 replacement (if one exists). Note that the replacement string passed to
782 @code{deprecate_cmd} should be the full name of the command, i.e. the
783 entire string the user should type at the command line.
785 @section UI-Independent Output---the @code{ui_out} Functions
786 @c This section is based on the documentation written by Fernando
787 @c Nasser <fnasser@redhat.com>.
789 @cindex @code{ui_out} functions
790 The @code{ui_out} functions present an abstraction level for the
791 @value{GDBN} output code. They hide the specifics of different user
792 interfaces supported by @value{GDBN}, and thus free the programmer
793 from the need to write several versions of the same code, one each for
794 every UI, to produce output.
796 @subsection Overview and Terminology
798 In general, execution of each @value{GDBN} command produces some sort
799 of output, and can even generate an input request.
801 Output can be generated for the following purposes:
805 to display a @emph{result} of an operation;
808 to convey @emph{info} or produce side-effects of a requested
812 to provide a @emph{notification} of an asynchronous event (including
813 progress indication of a prolonged asynchronous operation);
816 to display @emph{error messages} (including warnings);
819 to show @emph{debug data};
822 to @emph{query} or prompt a user for input (a special case).
826 This section mainly concentrates on how to build result output,
827 although some of it also applies to other kinds of output.
829 Generation of output that displays the results of an operation
830 involves one or more of the following:
834 output of the actual data
837 formatting the output as appropriate for console output, to make it
838 easily readable by humans
841 machine oriented formatting--a more terse formatting to allow for easy
842 parsing by programs which read @value{GDBN}'s output
845 annotation, whose purpose is to help legacy GUIs to identify interesting
849 The @code{ui_out} routines take care of the first three aspects.
850 Annotations are provided by separate annotation routines. Note that use
851 of annotations for an interface between a GUI and @value{GDBN} is
854 Output can be in the form of a single item, which we call a @dfn{field};
855 a @dfn{list} consisting of identical fields; a @dfn{tuple} consisting of
856 non-identical fields; or a @dfn{table}, which is a tuple consisting of a
857 header and a body. In a BNF-like form:
860 @item <table> @expansion{}
861 @code{<header> <body>}
862 @item <header> @expansion{}
863 @code{@{ <column> @}}
864 @item <column> @expansion{}
865 @code{<width> <alignment> <title>}
866 @item <body> @expansion{}
871 @subsection General Conventions
873 Most @code{ui_out} routines are of type @code{void}, the exceptions are
874 @code{ui_out_stream_new} (which returns a pointer to the newly created
875 object) and the @code{make_cleanup} routines.
877 The first parameter is always the @code{ui_out} vector object, a pointer
878 to a @code{struct ui_out}.
880 The @var{format} parameter is like in @code{printf} family of functions.
881 When it is present, there must also be a variable list of arguments
882 sufficient used to satisfy the @code{%} specifiers in the supplied
885 When a character string argument is not used in a @code{ui_out} function
886 call, a @code{NULL} pointer has to be supplied instead.
889 @subsection Table, Tuple and List Functions
891 @cindex list output functions
892 @cindex table output functions
893 @cindex tuple output functions
894 This section introduces @code{ui_out} routines for building lists,
895 tuples and tables. The routines to output the actual data items
896 (fields) are presented in the next section.
898 To recap: A @dfn{tuple} is a sequence of @dfn{fields}, each field
899 containing information about an object; a @dfn{list} is a sequence of
900 fields where each field describes an identical object.
902 Use the @dfn{table} functions when your output consists of a list of
903 rows (tuples) and the console output should include a heading. Use this
904 even when you are listing just one object but you still want the header.
906 @cindex nesting level in @code{ui_out} functions
907 Tables can not be nested. Tuples and lists can be nested up to a
908 maximum of five levels.
910 The overall structure of the table output code is something like this:
925 Here is the description of table-, tuple- and list-related @code{ui_out}
928 @deftypefun void ui_out_table_begin (struct ui_out *@var{uiout}, int @var{nbrofcols}, int @var{nr_rows}, const char *@var{tblid})
929 The function @code{ui_out_table_begin} marks the beginning of the output
930 of a table. It should always be called before any other @code{ui_out}
931 function for a given table. @var{nbrofcols} is the number of columns in
932 the table. @var{nr_rows} is the number of rows in the table.
933 @var{tblid} is an optional string identifying the table. The string
934 pointed to by @var{tblid} is copied by the implementation of
935 @code{ui_out_table_begin}, so the application can free the string if it
938 The companion function @code{ui_out_table_end}, described below, marks
939 the end of the table's output.
942 @deftypefun void ui_out_table_header (struct ui_out *@var{uiout}, int @var{width}, enum ui_align @var{alignment}, const char *@var{colhdr})
943 @code{ui_out_table_header} provides the header information for a single
944 table column. You call this function several times, one each for every
945 column of the table, after @code{ui_out_table_begin}, but before
946 @code{ui_out_table_body}.
948 The value of @var{width} gives the column width in characters. The
949 value of @var{alignment} is one of @code{left}, @code{center}, and
950 @code{right}, and it specifies how to align the header: left-justify,
951 center, or right-justify it. @var{colhdr} points to a string that
952 specifies the column header; the implementation copies that string, so
953 column header strings in @code{malloc}ed storage can be freed after the
957 @deftypefun void ui_out_table_body (struct ui_out *@var{uiout})
958 This function delimits the table header from the table body.
961 @deftypefun void ui_out_table_end (struct ui_out *@var{uiout})
962 This function signals the end of a table's output. It should be called
963 after the table body has been produced by the list and field output
966 There should be exactly one call to @code{ui_out_table_end} for each
967 call to @code{ui_out_table_begin}, otherwise the @code{ui_out} functions
968 will signal an internal error.
971 The output of the tuples that represent the table rows must follow the
972 call to @code{ui_out_table_body} and precede the call to
973 @code{ui_out_table_end}. You build a tuple by calling
974 @code{ui_out_tuple_begin} and @code{ui_out_tuple_end}, with suitable
975 calls to functions which actually output fields between them.
977 @deftypefun void ui_out_tuple_begin (struct ui_out *@var{uiout}, const char *@var{id})
978 This function marks the beginning of a tuple output. @var{id} points
979 to an optional string that identifies the tuple; it is copied by the
980 implementation, and so strings in @code{malloc}ed storage can be freed
984 @deftypefun void ui_out_tuple_end (struct ui_out *@var{uiout})
985 This function signals an end of a tuple output. There should be exactly
986 one call to @code{ui_out_tuple_end} for each call to
987 @code{ui_out_tuple_begin}, otherwise an internal @value{GDBN} error will
991 @deftypefun struct cleanup *make_cleanup_ui_out_tuple_begin_end (struct ui_out *@var{uiout}, const char *@var{id})
992 This function first opens the tuple and then establishes a cleanup
993 (@pxref{Coding, Cleanups}) to close the tuple. It provides a convenient
994 and correct implementation of the non-portable@footnote{The function
995 cast is not portable ISO C.} code sequence:
997 struct cleanup *old_cleanup;
998 ui_out_tuple_begin (uiout, "...");
999 old_cleanup = make_cleanup ((void(*)(void *)) ui_out_tuple_end,
1004 @deftypefun void ui_out_list_begin (struct ui_out *@var{uiout}, const char *@var{id})
1005 This function marks the beginning of a list output. @var{id} points to
1006 an optional string that identifies the list; it is copied by the
1007 implementation, and so strings in @code{malloc}ed storage can be freed
1011 @deftypefun void ui_out_list_end (struct ui_out *@var{uiout})
1012 This function signals an end of a list output. There should be exactly
1013 one call to @code{ui_out_list_end} for each call to
1014 @code{ui_out_list_begin}, otherwise an internal @value{GDBN} error will
1018 @deftypefun struct cleanup *make_cleanup_ui_out_list_begin_end (struct ui_out *@var{uiout}, const char *@var{id})
1019 Similar to @code{make_cleanup_ui_out_tuple_begin_end}, this function
1020 opens a list and then establishes cleanup (@pxref{Coding, Cleanups})
1021 that will close the list.list.
1024 @subsection Item Output Functions
1026 @cindex item output functions
1027 @cindex field output functions
1029 The functions described below produce output for the actual data
1030 items, or fields, which contain information about the object.
1032 Choose the appropriate function accordingly to your particular needs.
1034 @deftypefun void ui_out_field_fmt (struct ui_out *@var{uiout}, char *@var{fldname}, char *@var{format}, ...)
1035 This is the most general output function. It produces the
1036 representation of the data in the variable-length argument list
1037 according to formatting specifications in @var{format}, a
1038 @code{printf}-like format string. The optional argument @var{fldname}
1039 supplies the name of the field. The data items themselves are
1040 supplied as additional arguments after @var{format}.
1042 This generic function should be used only when it is not possible to
1043 use one of the specialized versions (see below).
1046 @deftypefun void ui_out_field_int (struct ui_out *@var{uiout}, const char *@var{fldname}, int @var{value})
1047 This function outputs a value of an @code{int} variable. It uses the
1048 @code{"%d"} output conversion specification. @var{fldname} specifies
1049 the name of the field.
1052 @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})
1053 This function outputs a value of an @code{int} variable. It differs from
1054 @code{ui_out_field_int} in that the caller specifies the desired @var{width} and @var{alignment} of the output.
1055 @var{fldname} specifies
1056 the name of the field.
1059 @deftypefun void ui_out_field_core_addr (struct ui_out *@var{uiout}, const char *@var{fldname}, CORE_ADDR @var{address})
1060 This function outputs an address.
1063 @deftypefun void ui_out_field_string (struct ui_out *@var{uiout}, const char *@var{fldname}, const char *@var{string})
1064 This function outputs a string using the @code{"%s"} conversion
1068 Sometimes, there's a need to compose your output piece by piece using
1069 functions that operate on a stream, such as @code{value_print} or
1070 @code{fprintf_symbol_filtered}. These functions accept an argument of
1071 the type @code{struct ui_file *}, a pointer to a @code{ui_file} object
1072 used to store the data stream used for the output. When you use one
1073 of these functions, you need a way to pass their results stored in a
1074 @code{ui_file} object to the @code{ui_out} functions. To this end,
1075 you first create a @code{ui_stream} object by calling
1076 @code{ui_out_stream_new}, pass the @code{stream} member of that
1077 @code{ui_stream} object to @code{value_print} and similar functions,
1078 and finally call @code{ui_out_field_stream} to output the field you
1079 constructed. When the @code{ui_stream} object is no longer needed,
1080 you should destroy it and free its memory by calling
1081 @code{ui_out_stream_delete}.
1083 @deftypefun struct ui_stream *ui_out_stream_new (struct ui_out *@var{uiout})
1084 This function creates a new @code{ui_stream} object which uses the
1085 same output methods as the @code{ui_out} object whose pointer is
1086 passed in @var{uiout}. It returns a pointer to the newly created
1087 @code{ui_stream} object.
1090 @deftypefun void ui_out_stream_delete (struct ui_stream *@var{streambuf})
1091 This functions destroys a @code{ui_stream} object specified by
1095 @deftypefun void ui_out_field_stream (struct ui_out *@var{uiout}, const char *@var{fieldname}, struct ui_stream *@var{streambuf})
1096 This function consumes all the data accumulated in
1097 @code{streambuf->stream} and outputs it like
1098 @code{ui_out_field_string} does. After a call to
1099 @code{ui_out_field_stream}, the accumulated data no longer exists, but
1100 the stream is still valid and may be used for producing more fields.
1103 @strong{Important:} If there is any chance that your code could bail
1104 out before completing output generation and reaching the point where
1105 @code{ui_out_stream_delete} is called, it is necessary to set up a
1106 cleanup, to avoid leaking memory and other resources. Here's a
1107 skeleton code to do that:
1110 struct ui_stream *mybuf = ui_out_stream_new (uiout);
1111 struct cleanup *old = make_cleanup (ui_out_stream_delete, mybuf);
1116 If the function already has the old cleanup chain set (for other kinds
1117 of cleanups), you just have to add your cleanup to it:
1120 mybuf = ui_out_stream_new (uiout);
1121 make_cleanup (ui_out_stream_delete, mybuf);
1124 Note that with cleanups in place, you should not call
1125 @code{ui_out_stream_delete} directly, or you would attempt to free the
1128 @subsection Utility Output Functions
1130 @deftypefun void ui_out_field_skip (struct ui_out *@var{uiout}, const char *@var{fldname})
1131 This function skips a field in a table. Use it if you have to leave
1132 an empty field without disrupting the table alignment. The argument
1133 @var{fldname} specifies a name for the (missing) filed.
1136 @deftypefun void ui_out_text (struct ui_out *@var{uiout}, const char *@var{string})
1137 This function outputs the text in @var{string} in a way that makes it
1138 easy to be read by humans. For example, the console implementation of
1139 this method filters the text through a built-in pager, to prevent it
1140 from scrolling off the visible portion of the screen.
1142 Use this function for printing relatively long chunks of text around
1143 the actual field data: the text it produces is not aligned according
1144 to the table's format. Use @code{ui_out_field_string} to output a
1145 string field, and use @code{ui_out_message}, described below, to
1146 output short messages.
1149 @deftypefun void ui_out_spaces (struct ui_out *@var{uiout}, int @var{nspaces})
1150 This function outputs @var{nspaces} spaces. It is handy to align the
1151 text produced by @code{ui_out_text} with the rest of the table or
1155 @deftypefun void ui_out_message (struct ui_out *@var{uiout}, int @var{verbosity}, const char *@var{format}, ...)
1156 This function produces a formatted message, provided that the current
1157 verbosity level is at least as large as given by @var{verbosity}. The
1158 current verbosity level is specified by the user with the @samp{set
1159 verbositylevel} command.@footnote{As of this writing (April 2001),
1160 setting verbosity level is not yet implemented, and is always returned
1161 as zero. So calling @code{ui_out_message} with a @var{verbosity}
1162 argument more than zero will cause the message to never be printed.}
1165 @deftypefun void ui_out_wrap_hint (struct ui_out *@var{uiout}, char *@var{indent})
1166 This function gives the console output filter (a paging filter) a hint
1167 of where to break lines which are too long. Ignored for all other
1168 output consumers. @var{indent}, if non-@code{NULL}, is the string to
1169 be printed to indent the wrapped text on the next line; it must remain
1170 accessible until the next call to @code{ui_out_wrap_hint}, or until an
1171 explicit newline is produced by one of the other functions. If
1172 @var{indent} is @code{NULL}, the wrapped text will not be indented.
1175 @deftypefun void ui_out_flush (struct ui_out *@var{uiout})
1176 This function flushes whatever output has been accumulated so far, if
1177 the UI buffers output.
1181 @subsection Examples of Use of @code{ui_out} functions
1183 @cindex using @code{ui_out} functions
1184 @cindex @code{ui_out} functions, usage examples
1185 This section gives some practical examples of using the @code{ui_out}
1186 functions to generalize the old console-oriented code in
1187 @value{GDBN}. The examples all come from functions defined on the
1188 @file{breakpoints.c} file.
1190 This example, from the @code{breakpoint_1} function, shows how to
1193 The original code was:
1196 if (!found_a_breakpoint++)
1198 annotate_breakpoints_headers ();
1201 printf_filtered ("Num ");
1203 printf_filtered ("Type ");
1205 printf_filtered ("Disp ");
1207 printf_filtered ("Enb ");
1211 printf_filtered ("Address ");
1214 printf_filtered ("What\n");
1216 annotate_breakpoints_table ();
1220 Here's the new version:
1223 nr_printable_breakpoints = @dots{};
1226 ui_out_table_begin (ui, 6, nr_printable_breakpoints, "BreakpointTable");
1228 ui_out_table_begin (ui, 5, nr_printable_breakpoints, "BreakpointTable");
1230 if (nr_printable_breakpoints > 0)
1231 annotate_breakpoints_headers ();
1232 if (nr_printable_breakpoints > 0)
1234 ui_out_table_header (uiout, 3, ui_left, "number", "Num"); /* 1 */
1235 if (nr_printable_breakpoints > 0)
1237 ui_out_table_header (uiout, 14, ui_left, "type", "Type"); /* 2 */
1238 if (nr_printable_breakpoints > 0)
1240 ui_out_table_header (uiout, 4, ui_left, "disp", "Disp"); /* 3 */
1241 if (nr_printable_breakpoints > 0)
1243 ui_out_table_header (uiout, 3, ui_left, "enabled", "Enb"); /* 4 */
1246 if (nr_printable_breakpoints > 0)
1248 if (TARGET_ADDR_BIT <= 32)
1249 ui_out_table_header (uiout, 10, ui_left, "addr", "Address");/* 5 */
1251 ui_out_table_header (uiout, 18, ui_left, "addr", "Address");/* 5 */
1253 if (nr_printable_breakpoints > 0)
1255 ui_out_table_header (uiout, 40, ui_noalign, "what", "What"); /* 6 */
1256 ui_out_table_body (uiout);
1257 if (nr_printable_breakpoints > 0)
1258 annotate_breakpoints_table ();
1261 This example, from the @code{print_one_breakpoint} function, shows how
1262 to produce the actual data for the table whose structure was defined
1263 in the above example. The original code was:
1268 printf_filtered ("%-3d ", b->number);
1270 if ((int)b->type > (sizeof(bptypes)/sizeof(bptypes[0]))
1271 || ((int) b->type != bptypes[(int) b->type].type))
1272 internal_error ("bptypes table does not describe type #%d.",
1274 printf_filtered ("%-14s ", bptypes[(int)b->type].description);
1276 printf_filtered ("%-4s ", bpdisps[(int)b->disposition]);
1278 printf_filtered ("%-3c ", bpenables[(int)b->enable]);
1282 This is the new version:
1286 ui_out_tuple_begin (uiout, "bkpt");
1288 ui_out_field_int (uiout, "number", b->number);
1290 if (((int) b->type > (sizeof (bptypes) / sizeof (bptypes[0])))
1291 || ((int) b->type != bptypes[(int) b->type].type))
1292 internal_error ("bptypes table does not describe type #%d.",
1294 ui_out_field_string (uiout, "type", bptypes[(int)b->type].description);
1296 ui_out_field_string (uiout, "disp", bpdisps[(int)b->disposition]);
1298 ui_out_field_fmt (uiout, "enabled", "%c", bpenables[(int)b->enable]);
1302 This example, also from @code{print_one_breakpoint}, shows how to
1303 produce a complicated output field using the @code{print_expression}
1304 functions which requires a stream to be passed. It also shows how to
1305 automate stream destruction with cleanups. The original code was:
1309 print_expression (b->exp, gdb_stdout);
1315 struct ui_stream *stb = ui_out_stream_new (uiout);
1316 struct cleanup *old_chain = make_cleanup_ui_out_stream_delete (stb);
1319 print_expression (b->exp, stb->stream);
1320 ui_out_field_stream (uiout, "what", local_stream);
1323 This example, also from @code{print_one_breakpoint}, shows how to use
1324 @code{ui_out_text} and @code{ui_out_field_string}. The original code
1329 if (b->dll_pathname == NULL)
1330 printf_filtered ("<any library> ");
1332 printf_filtered ("library \"%s\" ", b->dll_pathname);
1339 if (b->dll_pathname == NULL)
1341 ui_out_field_string (uiout, "what", "<any library>");
1342 ui_out_spaces (uiout, 1);
1346 ui_out_text (uiout, "library \"");
1347 ui_out_field_string (uiout, "what", b->dll_pathname);
1348 ui_out_text (uiout, "\" ");
1352 The following example from @code{print_one_breakpoint} shows how to
1353 use @code{ui_out_field_int} and @code{ui_out_spaces}. The original
1358 if (b->forked_inferior_pid != 0)
1359 printf_filtered ("process %d ", b->forked_inferior_pid);
1366 if (b->forked_inferior_pid != 0)
1368 ui_out_text (uiout, "process ");
1369 ui_out_field_int (uiout, "what", b->forked_inferior_pid);
1370 ui_out_spaces (uiout, 1);
1374 Here's an example of using @code{ui_out_field_string}. The original
1379 if (b->exec_pathname != NULL)
1380 printf_filtered ("program \"%s\" ", b->exec_pathname);
1387 if (b->exec_pathname != NULL)
1389 ui_out_text (uiout, "program \"");
1390 ui_out_field_string (uiout, "what", b->exec_pathname);
1391 ui_out_text (uiout, "\" ");
1395 Finally, here's an example of printing an address. The original code:
1399 printf_filtered ("%s ",
1400 hex_string_custom ((unsigned long) b->address, 8));
1407 ui_out_field_core_addr (uiout, "Address", b->address);
1411 @section Console Printing
1420 @cindex @code{libgdb}
1421 @code{libgdb} 1.0 was an abortive project of years ago. The theory was
1422 to provide an API to @value{GDBN}'s functionality.
1425 @cindex @code{libgdb}
1426 @code{libgdb} 2.0 is an ongoing effort to update @value{GDBN} so that is
1427 better able to support graphical and other environments.
1429 Since @code{libgdb} development is on-going, its architecture is still
1430 evolving. The following components have so far been identified:
1434 Observer - @file{gdb-events.h}.
1436 Builder - @file{ui-out.h}
1438 Event Loop - @file{event-loop.h}
1440 Library - @file{gdb.h}
1443 The model that ties these components together is described below.
1445 @section The @code{libgdb} Model
1447 A client of @code{libgdb} interacts with the library in two ways.
1451 As an observer (using @file{gdb-events}) receiving notifications from
1452 @code{libgdb} of any internal state changes (break point changes, run
1455 As a client querying @code{libgdb} (using the @file{ui-out} builder) to
1456 obtain various status values from @value{GDBN}.
1459 Since @code{libgdb} could have multiple clients (e.g. a GUI supporting
1460 the existing @value{GDBN} CLI), those clients must co-operate when
1461 controlling @code{libgdb}. In particular, a client must ensure that
1462 @code{libgdb} is idle (i.e. no other client is using @code{libgdb})
1463 before responding to a @file{gdb-event} by making a query.
1465 @section CLI support
1467 At present @value{GDBN}'s CLI is very much entangled in with the core of
1468 @code{libgdb}. Consequently, a client wishing to include the CLI in
1469 their interface needs to carefully co-ordinate its own and the CLI's
1472 It is suggested that the client set @code{libgdb} up to be bi-modal
1473 (alternate between CLI and client query modes). The notes below sketch
1478 The client registers itself as an observer of @code{libgdb}.
1480 The client create and install @code{cli-out} builder using its own
1481 versions of the @code{ui-file} @code{gdb_stderr}, @code{gdb_stdtarg} and
1482 @code{gdb_stdout} streams.
1484 The client creates a separate custom @code{ui-out} builder that is only
1485 used while making direct queries to @code{libgdb}.
1488 When the client receives input intended for the CLI, it simply passes it
1489 along. Since the @code{cli-out} builder is installed by default, all
1490 the CLI output in response to that command is routed (pronounced rooted)
1491 through to the client controlled @code{gdb_stdout} et.@: al.@: streams.
1492 At the same time, the client is kept abreast of internal changes by
1493 virtue of being a @code{libgdb} observer.
1495 The only restriction on the client is that it must wait until
1496 @code{libgdb} becomes idle before initiating any queries (using the
1497 client's custom builder).
1499 @section @code{libgdb} components
1501 @subheading Observer - @file{gdb-events.h}
1502 @file{gdb-events} provides the client with a very raw mechanism that can
1503 be used to implement an observer. At present it only allows for one
1504 observer and that observer must, internally, handle the need to delay
1505 the processing of any event notifications until after @code{libgdb} has
1506 finished the current command.
1508 @subheading Builder - @file{ui-out.h}
1509 @file{ui-out} provides the infrastructure necessary for a client to
1510 create a builder. That builder is then passed down to @code{libgdb}
1511 when doing any queries.
1513 @subheading Event Loop - @file{event-loop.h}
1514 @c There could be an entire section on the event-loop
1515 @file{event-loop}, currently non-re-entrant, provides a simple event
1516 loop. A client would need to either plug its self into this loop or,
1517 implement a new event-loop that GDB would use.
1519 The event-loop will eventually be made re-entrant. This is so that
1520 @value{GDBN} can better handle the problem of some commands blocking
1521 instead of returning.
1523 @subheading Library - @file{gdb.h}
1524 @file{libgdb} is the most obvious component of this system. It provides
1525 the query interface. Each function is parameterized by a @code{ui-out}
1526 builder. The result of the query is constructed using that builder
1527 before the query function returns.
1529 @node Symbol Handling
1531 @chapter Symbol Handling
1533 Symbols are a key part of @value{GDBN}'s operation. Symbols include variables,
1534 functions, and types.
1536 @section Symbol Reading
1538 @cindex symbol reading
1539 @cindex reading of symbols
1540 @cindex symbol files
1541 @value{GDBN} reads symbols from @dfn{symbol files}. The usual symbol
1542 file is the file containing the program which @value{GDBN} is
1543 debugging. @value{GDBN} can be directed to use a different file for
1544 symbols (with the @samp{symbol-file} command), and it can also read
1545 more symbols via the @samp{add-file} and @samp{load} commands, or while
1546 reading symbols from shared libraries.
1548 @findex find_sym_fns
1549 Symbol files are initially opened by code in @file{symfile.c} using
1550 the BFD library (@pxref{Support Libraries}). BFD identifies the type
1551 of the file by examining its header. @code{find_sym_fns} then uses
1552 this identification to locate a set of symbol-reading functions.
1554 @findex add_symtab_fns
1555 @cindex @code{sym_fns} structure
1556 @cindex adding a symbol-reading module
1557 Symbol-reading modules identify themselves to @value{GDBN} by calling
1558 @code{add_symtab_fns} during their module initialization. The argument
1559 to @code{add_symtab_fns} is a @code{struct sym_fns} which contains the
1560 name (or name prefix) of the symbol format, the length of the prefix,
1561 and pointers to four functions. These functions are called at various
1562 times to process symbol files whose identification matches the specified
1565 The functions supplied by each module are:
1568 @item @var{xyz}_symfile_init(struct sym_fns *sf)
1570 @cindex secondary symbol file
1571 Called from @code{symbol_file_add} when we are about to read a new
1572 symbol file. This function should clean up any internal state (possibly
1573 resulting from half-read previous files, for example) and prepare to
1574 read a new symbol file. Note that the symbol file which we are reading
1575 might be a new ``main'' symbol file, or might be a secondary symbol file
1576 whose symbols are being added to the existing symbol table.
1578 The argument to @code{@var{xyz}_symfile_init} is a newly allocated
1579 @code{struct sym_fns} whose @code{bfd} field contains the BFD for the
1580 new symbol file being read. Its @code{private} field has been zeroed,
1581 and can be modified as desired. Typically, a struct of private
1582 information will be @code{malloc}'d, and a pointer to it will be placed
1583 in the @code{private} field.
1585 There is no result from @code{@var{xyz}_symfile_init}, but it can call
1586 @code{error} if it detects an unavoidable problem.
1588 @item @var{xyz}_new_init()
1590 Called from @code{symbol_file_add} when discarding existing symbols.
1591 This function needs only handle the symbol-reading module's internal
1592 state; the symbol table data structures visible to the rest of
1593 @value{GDBN} will be discarded by @code{symbol_file_add}. It has no
1594 arguments and no result. It may be called after
1595 @code{@var{xyz}_symfile_init}, if a new symbol table is being read, or
1596 may be called alone if all symbols are simply being discarded.
1598 @item @var{xyz}_symfile_read(struct sym_fns *sf, CORE_ADDR addr, int mainline)
1600 Called from @code{symbol_file_add} to actually read the symbols from a
1601 symbol-file into a set of psymtabs or symtabs.
1603 @code{sf} points to the @code{struct sym_fns} originally passed to
1604 @code{@var{xyz}_sym_init} for possible initialization. @code{addr} is
1605 the offset between the file's specified start address and its true
1606 address in memory. @code{mainline} is 1 if this is the main symbol
1607 table being read, and 0 if a secondary symbol file (e.g. shared library
1608 or dynamically loaded file) is being read.@refill
1611 In addition, if a symbol-reading module creates psymtabs when
1612 @var{xyz}_symfile_read is called, these psymtabs will contain a pointer
1613 to a function @code{@var{xyz}_psymtab_to_symtab}, which can be called
1614 from any point in the @value{GDBN} symbol-handling code.
1617 @item @var{xyz}_psymtab_to_symtab (struct partial_symtab *pst)
1619 Called from @code{psymtab_to_symtab} (or the @code{PSYMTAB_TO_SYMTAB} macro) if
1620 the psymtab has not already been read in and had its @code{pst->symtab}
1621 pointer set. The argument is the psymtab to be fleshed-out into a
1622 symtab. Upon return, @code{pst->readin} should have been set to 1, and
1623 @code{pst->symtab} should contain a pointer to the new corresponding symtab, or
1624 zero if there were no symbols in that part of the symbol file.
1627 @section Partial Symbol Tables
1629 @value{GDBN} has three types of symbol tables:
1632 @cindex full symbol table
1635 Full symbol tables (@dfn{symtabs}). These contain the main
1636 information about symbols and addresses.
1640 Partial symbol tables (@dfn{psymtabs}). These contain enough
1641 information to know when to read the corresponding part of the full
1644 @cindex minimal symbol table
1647 Minimal symbol tables (@dfn{msymtabs}). These contain information
1648 gleaned from non-debugging symbols.
1651 @cindex partial symbol table
1652 This section describes partial symbol tables.
1654 A psymtab is constructed by doing a very quick pass over an executable
1655 file's debugging information. Small amounts of information are
1656 extracted---enough to identify which parts of the symbol table will
1657 need to be re-read and fully digested later, when the user needs the
1658 information. The speed of this pass causes @value{GDBN} to start up very
1659 quickly. Later, as the detailed rereading occurs, it occurs in small
1660 pieces, at various times, and the delay therefrom is mostly invisible to
1662 @c (@xref{Symbol Reading}.)
1664 The symbols that show up in a file's psymtab should be, roughly, those
1665 visible to the debugger's user when the program is not running code from
1666 that file. These include external symbols and types, static symbols and
1667 types, and @code{enum} values declared at file scope.
1669 The psymtab also contains the range of instruction addresses that the
1670 full symbol table would represent.
1672 @cindex finding a symbol
1673 @cindex symbol lookup
1674 The idea is that there are only two ways for the user (or much of the
1675 code in the debugger) to reference a symbol:
1678 @findex find_pc_function
1679 @findex find_pc_line
1681 By its address (e.g. execution stops at some address which is inside a
1682 function in this file). The address will be noticed to be in the
1683 range of this psymtab, and the full symtab will be read in.
1684 @code{find_pc_function}, @code{find_pc_line}, and other
1685 @code{find_pc_@dots{}} functions handle this.
1687 @cindex lookup_symbol
1690 (e.g. the user asks to print a variable, or set a breakpoint on a
1691 function). Global names and file-scope names will be found in the
1692 psymtab, which will cause the symtab to be pulled in. Local names will
1693 have to be qualified by a global name, or a file-scope name, in which
1694 case we will have already read in the symtab as we evaluated the
1695 qualifier. Or, a local symbol can be referenced when we are ``in'' a
1696 local scope, in which case the first case applies. @code{lookup_symbol}
1697 does most of the work here.
1700 The only reason that psymtabs exist is to cause a symtab to be read in
1701 at the right moment. Any symbol that can be elided from a psymtab,
1702 while still causing that to happen, should not appear in it. Since
1703 psymtabs don't have the idea of scope, you can't put local symbols in
1704 them anyway. Psymtabs don't have the idea of the type of a symbol,
1705 either, so types need not appear, unless they will be referenced by
1708 It is a bug for @value{GDBN} to behave one way when only a psymtab has
1709 been read, and another way if the corresponding symtab has been read
1710 in. Such bugs are typically caused by a psymtab that does not contain
1711 all the visible symbols, or which has the wrong instruction address
1714 The psymtab for a particular section of a symbol file (objfile) could be
1715 thrown away after the symtab has been read in. The symtab should always
1716 be searched before the psymtab, so the psymtab will never be used (in a
1717 bug-free environment). Currently, psymtabs are allocated on an obstack,
1718 and all the psymbols themselves are allocated in a pair of large arrays
1719 on an obstack, so there is little to be gained by trying to free them
1720 unless you want to do a lot more work.
1724 @unnumberedsubsec Fundamental Types (e.g., @code{FT_VOID}, @code{FT_BOOLEAN}).
1726 @cindex fundamental types
1727 These are the fundamental types that @value{GDBN} uses internally. Fundamental
1728 types from the various debugging formats (stabs, ELF, etc) are mapped
1729 into one of these. They are basically a union of all fundamental types
1730 that @value{GDBN} knows about for all the languages that @value{GDBN}
1733 @unnumberedsubsec Type Codes (e.g., @code{TYPE_CODE_PTR}, @code{TYPE_CODE_ARRAY}).
1736 Each time @value{GDBN} builds an internal type, it marks it with one
1737 of these types. The type may be a fundamental type, such as
1738 @code{TYPE_CODE_INT}, or a derived type, such as @code{TYPE_CODE_PTR}
1739 which is a pointer to another type. Typically, several @code{FT_*}
1740 types map to one @code{TYPE_CODE_*} type, and are distinguished by
1741 other members of the type struct, such as whether the type is signed
1742 or unsigned, and how many bits it uses.
1744 @unnumberedsubsec Builtin Types (e.g., @code{builtin_type_void}, @code{builtin_type_char}).
1746 These are instances of type structs that roughly correspond to
1747 fundamental types and are created as global types for @value{GDBN} to
1748 use for various ugly historical reasons. We eventually want to
1749 eliminate these. Note for example that @code{builtin_type_int}
1750 initialized in @file{gdbtypes.c} is basically the same as a
1751 @code{TYPE_CODE_INT} type that is initialized in @file{c-lang.c} for
1752 an @code{FT_INTEGER} fundamental type. The difference is that the
1753 @code{builtin_type} is not associated with any particular objfile, and
1754 only one instance exists, while @file{c-lang.c} builds as many
1755 @code{TYPE_CODE_INT} types as needed, with each one associated with
1756 some particular objfile.
1758 @section Object File Formats
1759 @cindex object file formats
1763 @cindex @code{a.out} format
1764 The @code{a.out} format is the original file format for Unix. It
1765 consists of three sections: @code{text}, @code{data}, and @code{bss},
1766 which are for program code, initialized data, and uninitialized data,
1769 The @code{a.out} format is so simple that it doesn't have any reserved
1770 place for debugging information. (Hey, the original Unix hackers used
1771 @samp{adb}, which is a machine-language debugger!) The only debugging
1772 format for @code{a.out} is stabs, which is encoded as a set of normal
1773 symbols with distinctive attributes.
1775 The basic @code{a.out} reader is in @file{dbxread.c}.
1780 The COFF format was introduced with System V Release 3 (SVR3) Unix.
1781 COFF files may have multiple sections, each prefixed by a header. The
1782 number of sections is limited.
1784 The COFF specification includes support for debugging. Although this
1785 was a step forward, the debugging information was woefully limited. For
1786 instance, it was not possible to represent code that came from an
1789 The COFF reader is in @file{coffread.c}.
1793 @cindex ECOFF format
1794 ECOFF is an extended COFF originally introduced for Mips and Alpha
1797 The basic ECOFF reader is in @file{mipsread.c}.
1801 @cindex XCOFF format
1802 The IBM RS/6000 running AIX uses an object file format called XCOFF.
1803 The COFF sections, symbols, and line numbers are used, but debugging
1804 symbols are @code{dbx}-style stabs whose strings are located in the
1805 @code{.debug} section (rather than the string table). For more
1806 information, see @ref{Top,,,stabs,The Stabs Debugging Format}.
1808 The shared library scheme has a clean interface for figuring out what
1809 shared libraries are in use, but the catch is that everything which
1810 refers to addresses (symbol tables and breakpoints at least) needs to be
1811 relocated for both shared libraries and the main executable. At least
1812 using the standard mechanism this can only be done once the program has
1813 been run (or the core file has been read).
1817 @cindex PE-COFF format
1818 Windows 95 and NT use the PE (@dfn{Portable Executable}) format for their
1819 executables. PE is basically COFF with additional headers.
1821 While BFD includes special PE support, @value{GDBN} needs only the basic
1827 The ELF format came with System V Release 4 (SVR4) Unix. ELF is similar
1828 to COFF in being organized into a number of sections, but it removes
1829 many of COFF's limitations.
1831 The basic ELF reader is in @file{elfread.c}.
1836 SOM is HP's object file and debug format (not to be confused with IBM's
1837 SOM, which is a cross-language ABI).
1839 The SOM reader is in @file{hpread.c}.
1841 @subsection Other File Formats
1843 @cindex Netware Loadable Module format
1844 Other file formats that have been supported by @value{GDBN} include Netware
1845 Loadable Modules (@file{nlmread.c}).
1847 @section Debugging File Formats
1849 This section describes characteristics of debugging information that
1850 are independent of the object file format.
1854 @cindex stabs debugging info
1855 @code{stabs} started out as special symbols within the @code{a.out}
1856 format. Since then, it has been encapsulated into other file
1857 formats, such as COFF and ELF.
1859 While @file{dbxread.c} does some of the basic stab processing,
1860 including for encapsulated versions, @file{stabsread.c} does
1865 @cindex COFF debugging info
1866 The basic COFF definition includes debugging information. The level
1867 of support is minimal and non-extensible, and is not often used.
1869 @subsection Mips debug (Third Eye)
1871 @cindex ECOFF debugging info
1872 ECOFF includes a definition of a special debug format.
1874 The file @file{mdebugread.c} implements reading for this format.
1878 @cindex DWARF 1 debugging info
1879 DWARF 1 is a debugging format that was originally designed to be
1880 used with ELF in SVR4 systems.
1885 @c If defined, these are the producer strings in a DWARF 1 file. All of
1886 @c these have reasonable defaults already.
1888 The DWARF 1 reader is in @file{dwarfread.c}.
1892 @cindex DWARF 2 debugging info
1893 DWARF 2 is an improved but incompatible version of DWARF 1.
1895 The DWARF 2 reader is in @file{dwarf2read.c}.
1899 @cindex SOM debugging info
1900 Like COFF, the SOM definition includes debugging information.
1902 @section Adding a New Symbol Reader to @value{GDBN}
1904 @cindex adding debugging info reader
1905 If you are using an existing object file format (@code{a.out}, COFF, ELF, etc),
1906 there is probably little to be done.
1908 If you need to add a new object file format, you must first add it to
1909 BFD. This is beyond the scope of this document.
1911 You must then arrange for the BFD code to provide access to the
1912 debugging symbols. Generally @value{GDBN} will have to call swapping routines
1913 from BFD and a few other BFD internal routines to locate the debugging
1914 information. As much as possible, @value{GDBN} should not depend on the BFD
1915 internal data structures.
1917 For some targets (e.g., COFF), there is a special transfer vector used
1918 to call swapping routines, since the external data structures on various
1919 platforms have different sizes and layouts. Specialized routines that
1920 will only ever be implemented by one object file format may be called
1921 directly. This interface should be described in a file
1922 @file{bfd/lib@var{xyz}.h}, which is included by @value{GDBN}.
1925 @node Language Support
1927 @chapter Language Support
1929 @cindex language support
1930 @value{GDBN}'s language support is mainly driven by the symbol reader,
1931 although it is possible for the user to set the source language
1934 @value{GDBN} chooses the source language by looking at the extension
1935 of the file recorded in the debug info; @file{.c} means C, @file{.f}
1936 means Fortran, etc. It may also use a special-purpose language
1937 identifier if the debug format supports it, like with DWARF.
1939 @section Adding a Source Language to @value{GDBN}
1941 @cindex adding source language
1942 To add other languages to @value{GDBN}'s expression parser, follow the
1946 @item Create the expression parser.
1948 @cindex expression parser
1949 This should reside in a file @file{@var{lang}-exp.y}. Routines for
1950 building parsed expressions into a @code{union exp_element} list are in
1953 @cindex language parser
1954 Since we can't depend upon everyone having Bison, and YACC produces
1955 parsers that define a bunch of global names, the following lines
1956 @strong{must} be included at the top of the YACC parser, to prevent the
1957 various parsers from defining the same global names:
1960 #define yyparse @var{lang}_parse
1961 #define yylex @var{lang}_lex
1962 #define yyerror @var{lang}_error
1963 #define yylval @var{lang}_lval
1964 #define yychar @var{lang}_char
1965 #define yydebug @var{lang}_debug
1966 #define yypact @var{lang}_pact
1967 #define yyr1 @var{lang}_r1
1968 #define yyr2 @var{lang}_r2
1969 #define yydef @var{lang}_def
1970 #define yychk @var{lang}_chk
1971 #define yypgo @var{lang}_pgo
1972 #define yyact @var{lang}_act
1973 #define yyexca @var{lang}_exca
1974 #define yyerrflag @var{lang}_errflag
1975 #define yynerrs @var{lang}_nerrs
1978 At the bottom of your parser, define a @code{struct language_defn} and
1979 initialize it with the right values for your language. Define an
1980 @code{initialize_@var{lang}} routine and have it call
1981 @samp{add_language(@var{lang}_language_defn)} to tell the rest of @value{GDBN}
1982 that your language exists. You'll need some other supporting variables
1983 and functions, which will be used via pointers from your
1984 @code{@var{lang}_language_defn}. See the declaration of @code{struct
1985 language_defn} in @file{language.h}, and the other @file{*-exp.y} files,
1986 for more information.
1988 @item Add any evaluation routines, if necessary
1990 @cindex expression evaluation routines
1991 @findex evaluate_subexp
1992 @findex prefixify_subexp
1993 @findex length_of_subexp
1994 If you need new opcodes (that represent the operations of the language),
1995 add them to the enumerated type in @file{expression.h}. Add support
1996 code for these operations in the @code{evaluate_subexp} function
1997 defined in the file @file{eval.c}. Add cases
1998 for new opcodes in two functions from @file{parse.c}:
1999 @code{prefixify_subexp} and @code{length_of_subexp}. These compute
2000 the number of @code{exp_element}s that a given operation takes up.
2002 @item Update some existing code
2004 Add an enumerated identifier for your language to the enumerated type
2005 @code{enum language} in @file{defs.h}.
2007 Update the routines in @file{language.c} so your language is included.
2008 These routines include type predicates and such, which (in some cases)
2009 are language dependent. If your language does not appear in the switch
2010 statement, an error is reported.
2012 @vindex current_language
2013 Also included in @file{language.c} is the code that updates the variable
2014 @code{current_language}, and the routines that translate the
2015 @code{language_@var{lang}} enumerated identifier into a printable
2018 @findex _initialize_language
2019 Update the function @code{_initialize_language} to include your
2020 language. This function picks the default language upon startup, so is
2021 dependent upon which languages that @value{GDBN} is built for.
2023 @findex allocate_symtab
2024 Update @code{allocate_symtab} in @file{symfile.c} and/or symbol-reading
2025 code so that the language of each symtab (source file) is set properly.
2026 This is used to determine the language to use at each stack frame level.
2027 Currently, the language is set based upon the extension of the source
2028 file. If the language can be better inferred from the symbol
2029 information, please set the language of the symtab in the symbol-reading
2032 @findex print_subexp
2033 @findex op_print_tab
2034 Add helper code to @code{print_subexp} (in @file{expprint.c}) to handle any new
2035 expression opcodes you have added to @file{expression.h}. Also, add the
2036 printed representations of your operators to @code{op_print_tab}.
2038 @item Add a place of call
2041 Add a call to @code{@var{lang}_parse()} and @code{@var{lang}_error} in
2042 @code{parse_exp_1} (defined in @file{parse.c}).
2044 @item Use macros to trim code
2046 @cindex trimming language-dependent code
2047 The user has the option of building @value{GDBN} for some or all of the
2048 languages. If the user decides to build @value{GDBN} for the language
2049 @var{lang}, then every file dependent on @file{language.h} will have the
2050 macro @code{_LANG_@var{lang}} defined in it. Use @code{#ifdef}s to
2051 leave out large routines that the user won't need if he or she is not
2052 using your language.
2054 Note that you do not need to do this in your YACC parser, since if @value{GDBN}
2055 is not build for @var{lang}, then @file{@var{lang}-exp.tab.o} (the
2056 compiled form of your parser) is not linked into @value{GDBN} at all.
2058 See the file @file{configure.in} for how @value{GDBN} is configured
2059 for different languages.
2061 @item Edit @file{Makefile.in}
2063 Add dependencies in @file{Makefile.in}. Make sure you update the macro
2064 variables such as @code{HFILES} and @code{OBJS}, otherwise your code may
2065 not get linked in, or, worse yet, it may not get @code{tar}red into the
2070 @node Host Definition
2072 @chapter Host Definition
2074 With the advent of Autoconf, it's rarely necessary to have host
2075 definition machinery anymore. The following information is provided,
2076 mainly, as an historical reference.
2078 @section Adding a New Host
2080 @cindex adding a new host
2081 @cindex host, adding
2082 @value{GDBN}'s host configuration support normally happens via Autoconf.
2083 New host-specific definitions should not be needed. Older hosts
2084 @value{GDBN} still use the host-specific definitions and files listed
2085 below, but these mostly exist for historical reasons, and will
2086 eventually disappear.
2089 @item gdb/config/@var{arch}/@var{xyz}.mh
2090 This file once contained both host and native configuration information
2091 (@pxref{Native Debugging}) for the machine @var{xyz}. The host
2092 configuration information is now handed by Autoconf.
2094 Host configuration information included a definition of
2095 @code{XM_FILE=xm-@var{xyz}.h} and possibly definitions for @code{CC},
2096 @code{SYSV_DEFINE}, @code{XM_CFLAGS}, @code{XM_ADD_FILES},
2097 @code{XM_CLIBS}, @code{XM_CDEPS}, etc.; see @file{Makefile.in}.
2099 New host only configurations do not need this file.
2101 @item gdb/config/@var{arch}/xm-@var{xyz}.h
2102 This file once contained definitions and includes required when hosting
2103 gdb on machine @var{xyz}. Those definitions and includes are now
2104 handled by Autoconf.
2106 New host and native configurations do not need this file.
2108 @emph{Maintainer's note: Some hosts continue to use the @file{xm-xyz.h}
2109 file to define the macros @var{HOST_FLOAT_FORMAT},
2110 @var{HOST_DOUBLE_FORMAT} and @var{HOST_LONG_DOUBLE_FORMAT}. That code
2111 also needs to be replaced with either an Autoconf or run-time test.}
2115 @subheading Generic Host Support Files
2117 @cindex generic host support
2118 There are some ``generic'' versions of routines that can be used by
2119 various systems. These can be customized in various ways by macros
2120 defined in your @file{xm-@var{xyz}.h} file. If these routines work for
2121 the @var{xyz} host, you can just include the generic file's name (with
2122 @samp{.o}, not @samp{.c}) in @code{XDEPFILES}.
2124 Otherwise, if your machine needs custom support routines, you will need
2125 to write routines that perform the same functions as the generic file.
2126 Put them into @code{@var{xyz}-xdep.c}, and put @code{@var{xyz}-xdep.o}
2127 into @code{XDEPFILES}.
2130 @cindex remote debugging support
2131 @cindex serial line support
2133 This contains serial line support for Unix systems. This is always
2134 included, via the makefile variable @code{SER_HARDWIRE}; override this
2135 variable in the @file{.mh} file to avoid it.
2138 This contains serial line support for 32-bit programs running under DOS,
2139 using the DJGPP (a.k.a.@: GO32) execution environment.
2141 @cindex TCP remote support
2143 This contains generic TCP support using sockets.
2146 @section Host Conditionals
2148 When @value{GDBN} is configured and compiled, various macros are
2149 defined or left undefined, to control compilation based on the
2150 attributes of the host system. These macros and their meanings (or if
2151 the meaning is not documented here, then one of the source files where
2152 they are used is indicated) are:
2155 @item @value{GDBN}INIT_FILENAME
2156 The default name of @value{GDBN}'s initialization file (normally
2160 This macro is deprecated.
2162 @item SIGWINCH_HANDLER
2163 If your host defines @code{SIGWINCH}, you can define this to be the name
2164 of a function to be called if @code{SIGWINCH} is received.
2166 @item SIGWINCH_HANDLER_BODY
2167 Define this to expand into code that will define the function named by
2168 the expansion of @code{SIGWINCH_HANDLER}.
2170 @item ALIGN_STACK_ON_STARTUP
2171 @cindex stack alignment
2172 Define this if your system is of a sort that will crash in
2173 @code{tgetent} if the stack happens not to be longword-aligned when
2174 @code{main} is called. This is a rare situation, but is known to occur
2175 on several different types of systems.
2177 @item CRLF_SOURCE_FILES
2178 @cindex DOS text files
2179 Define this if host files use @code{\r\n} rather than @code{\n} as a
2180 line terminator. This will cause source file listings to omit @code{\r}
2181 characters when printing and it will allow @code{\r\n} line endings of files
2182 which are ``sourced'' by gdb. It must be possible to open files in binary
2183 mode using @code{O_BINARY} or, for fopen, @code{"rb"}.
2185 @item DEFAULT_PROMPT
2187 The default value of the prompt string (normally @code{"(gdb) "}).
2190 @cindex terminal device
2191 The name of the generic TTY device, defaults to @code{"/dev/tty"}.
2194 Define this if binary files are opened the same way as text files.
2198 In some cases, use the system call @code{mmap} for reading symbol
2199 tables. For some machines this allows for sharing and quick updates.
2202 Define this if the host system has @code{termio.h}.
2209 Values for host-side constants.
2212 Substitute for isatty, if not available.
2215 This is the longest integer type available on the host. If not defined,
2216 it will default to @code{long long} or @code{long}, depending on
2217 @code{CC_HAS_LONG_LONG}.
2219 @item CC_HAS_LONG_LONG
2220 @cindex @code{long long} data type
2221 Define this if the host C compiler supports @code{long long}. This is set
2222 by the @code{configure} script.
2224 @item PRINTF_HAS_LONG_LONG
2225 Define this if the host can handle printing of long long integers via
2226 the printf format conversion specifier @code{ll}. This is set by the
2227 @code{configure} script.
2229 @item HAVE_LONG_DOUBLE
2230 Define this if the host C compiler supports @code{long double}. This is
2231 set by the @code{configure} script.
2233 @item PRINTF_HAS_LONG_DOUBLE
2234 Define this if the host can handle printing of long double float-point
2235 numbers via the printf format conversion specifier @code{Lg}. This is
2236 set by the @code{configure} script.
2238 @item SCANF_HAS_LONG_DOUBLE
2239 Define this if the host can handle the parsing of long double
2240 float-point numbers via the scanf format conversion specifier
2241 @code{Lg}. This is set by the @code{configure} script.
2243 @item LSEEK_NOT_LINEAR
2244 Define this if @code{lseek (n)} does not necessarily move to byte number
2245 @code{n} in the file. This is only used when reading source files. It
2246 is normally faster to define @code{CRLF_SOURCE_FILES} when possible.
2249 This macro is used as the argument to @code{lseek} (or, most commonly,
2250 @code{bfd_seek}). FIXME, should be replaced by SEEK_SET instead,
2251 which is the POSIX equivalent.
2254 If defined, this should be one or more tokens, such as @code{volatile},
2255 that can be used in both the declaration and definition of functions to
2256 indicate that they never return. The default is already set correctly
2257 if compiling with GCC. This will almost never need to be defined.
2260 If defined, this should be one or more tokens, such as
2261 @code{__attribute__ ((noreturn))}, that can be used in the declarations
2262 of functions to indicate that they never return. The default is already
2263 set correctly if compiling with GCC. This will almost never need to be
2268 Define these to appropriate value for the system @code{lseek}, if not already
2272 This is the signal for stopping @value{GDBN}. Defaults to
2273 @code{SIGTSTP}. (Only redefined for the Convex.)
2276 Means that System V (prior to SVR4) include files are in use. (FIXME:
2277 This symbol is abused in @file{infrun.c}, @file{regex.c}, and
2278 @file{utils.c} for other things, at the moment.)
2281 Define this to help placate @code{lint} in some situations.
2284 Define this to override the defaults of @code{__volatile__} or
2289 @node Target Architecture Definition
2291 @chapter Target Architecture Definition
2293 @cindex target architecture definition
2294 @value{GDBN}'s target architecture defines what sort of
2295 machine-language programs @value{GDBN} can work with, and how it works
2298 The target architecture object is implemented as the C structure
2299 @code{struct gdbarch *}. The structure, and its methods, are generated
2300 using the Bourne shell script @file{gdbarch.sh}.
2302 @section Operating System ABI Variant Handling
2303 @cindex OS ABI variants
2305 @value{GDBN} provides a mechanism for handling variations in OS
2306 ABIs. An OS ABI variant may have influence over any number of
2307 variables in the target architecture definition. There are two major
2308 components in the OS ABI mechanism: sniffers and handlers.
2310 A @dfn{sniffer} examines a file matching a BFD architecture/flavour pair
2311 (the architecture may be wildcarded) in an attempt to determine the
2312 OS ABI of that file. Sniffers with a wildcarded architecture are considered
2313 to be @dfn{generic}, while sniffers for a specific architecture are
2314 considered to be @dfn{specific}. A match from a specific sniffer
2315 overrides a match from a generic sniffer. Multiple sniffers for an
2316 architecture/flavour may exist, in order to differentiate between two
2317 different operating systems which use the same basic file format. The
2318 OS ABI framework provides a generic sniffer for ELF-format files which
2319 examines the @code{EI_OSABI} field of the ELF header, as well as note
2320 sections known to be used by several operating systems.
2322 @cindex fine-tuning @code{gdbarch} structure
2323 A @dfn{handler} is used to fine-tune the @code{gdbarch} structure for the
2324 selected OS ABI. There may be only one handler for a given OS ABI
2325 for each BFD architecture.
2327 The following OS ABI variants are defined in @file{osabi.h}:
2331 @findex GDB_OSABI_UNKNOWN
2332 @item GDB_OSABI_UNKNOWN
2333 The ABI of the inferior is unknown. The default @code{gdbarch}
2334 settings for the architecture will be used.
2336 @findex GDB_OSABI_SVR4
2337 @item GDB_OSABI_SVR4
2338 UNIX System V Release 4
2340 @findex GDB_OSABI_HURD
2341 @item GDB_OSABI_HURD
2342 GNU using the Hurd kernel
2344 @findex GDB_OSABI_SOLARIS
2345 @item GDB_OSABI_SOLARIS
2348 @findex GDB_OSABI_OSF1
2349 @item GDB_OSABI_OSF1
2350 OSF/1, including Digital UNIX and Compaq Tru64 UNIX
2352 @findex GDB_OSABI_LINUX
2353 @item GDB_OSABI_LINUX
2354 GNU using the Linux kernel
2356 @findex GDB_OSABI_FREEBSD_AOUT
2357 @item GDB_OSABI_FREEBSD_AOUT
2358 FreeBSD using the a.out executable format
2360 @findex GDB_OSABI_FREEBSD_ELF
2361 @item GDB_OSABI_FREEBSD_ELF
2362 FreeBSD using the ELF executable format
2364 @findex GDB_OSABI_NETBSD_AOUT
2365 @item GDB_OSABI_NETBSD_AOUT
2366 NetBSD using the a.out executable format
2368 @findex GDB_OSABI_NETBSD_ELF
2369 @item GDB_OSABI_NETBSD_ELF
2370 NetBSD using the ELF executable format
2372 @findex GDB_OSABI_WINCE
2373 @item GDB_OSABI_WINCE
2376 @findex GDB_OSABI_GO32
2377 @item GDB_OSABI_GO32
2380 @findex GDB_OSABI_NETWARE
2381 @item GDB_OSABI_NETWARE
2384 @findex GDB_OSABI_ARM_EABI_V1
2385 @item GDB_OSABI_ARM_EABI_V1
2386 ARM Embedded ABI version 1
2388 @findex GDB_OSABI_ARM_EABI_V2
2389 @item GDB_OSABI_ARM_EABI_V2
2390 ARM Embedded ABI version 2
2392 @findex GDB_OSABI_ARM_APCS
2393 @item GDB_OSABI_ARM_APCS
2394 Generic ARM Procedure Call Standard
2398 Here are the functions that make up the OS ABI framework:
2400 @deftypefun const char *gdbarch_osabi_name (enum gdb_osabi @var{osabi})
2401 Return the name of the OS ABI corresponding to @var{osabi}.
2404 @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}))
2405 Register the OS ABI handler specified by @var{init_osabi} for the
2406 architecture, machine type and OS ABI specified by @var{arch},
2407 @var{machine} and @var{osabi}. In most cases, a value of zero for the
2408 machine type, which implies the architecture's default machine type,
2412 @deftypefun void gdbarch_register_osabi_sniffer (enum bfd_architecture @var{arch}, enum bfd_flavour @var{flavour}, enum gdb_osabi (*@var{sniffer})(bfd *@var{abfd}))
2413 Register the OS ABI file sniffer specified by @var{sniffer} for the
2414 BFD architecture/flavour pair specified by @var{arch} and @var{flavour}.
2415 If @var{arch} is @code{bfd_arch_unknown}, the sniffer is considered to
2416 be generic, and is allowed to examine @var{flavour}-flavoured files for
2420 @deftypefun enum gdb_osabi gdbarch_lookup_osabi (bfd *@var{abfd})
2421 Examine the file described by @var{abfd} to determine its OS ABI.
2422 The value @code{GDB_OSABI_UNKNOWN} is returned if the OS ABI cannot
2426 @deftypefun void gdbarch_init_osabi (struct gdbarch info @var{info}, struct gdbarch *@var{gdbarch}, enum gdb_osabi @var{osabi})
2427 Invoke the OS ABI handler corresponding to @var{osabi} to fine-tune the
2428 @code{gdbarch} structure specified by @var{gdbarch}. If a handler
2429 corresponding to @var{osabi} has not been registered for @var{gdbarch}'s
2430 architecture, a warning will be issued and the debugging session will continue
2431 with the defaults already established for @var{gdbarch}.
2434 @section Registers and Memory
2436 @value{GDBN}'s model of the target machine is rather simple.
2437 @value{GDBN} assumes the machine includes a bank of registers and a
2438 block of memory. Each register may have a different size.
2440 @value{GDBN} does not have a magical way to match up with the
2441 compiler's idea of which registers are which; however, it is critical
2442 that they do match up accurately. The only way to make this work is
2443 to get accurate information about the order that the compiler uses,
2444 and to reflect that in the @code{REGISTER_NAME} and related macros.
2446 @value{GDBN} can handle big-endian, little-endian, and bi-endian architectures.
2448 @section Pointers Are Not Always Addresses
2449 @cindex pointer representation
2450 @cindex address representation
2451 @cindex word-addressed machines
2452 @cindex separate data and code address spaces
2453 @cindex spaces, separate data and code address
2454 @cindex address spaces, separate data and code
2455 @cindex code pointers, word-addressed
2456 @cindex converting between pointers and addresses
2457 @cindex D10V addresses
2459 On almost all 32-bit architectures, the representation of a pointer is
2460 indistinguishable from the representation of some fixed-length number
2461 whose value is the byte address of the object pointed to. On such
2462 machines, the words ``pointer'' and ``address'' can be used interchangeably.
2463 However, architectures with smaller word sizes are often cramped for
2464 address space, so they may choose a pointer representation that breaks this
2465 identity, and allows a larger code address space.
2467 For example, the Renesas D10V is a 16-bit VLIW processor whose
2468 instructions are 32 bits long@footnote{Some D10V instructions are
2469 actually pairs of 16-bit sub-instructions. However, since you can't
2470 jump into the middle of such a pair, code addresses can only refer to
2471 full 32 bit instructions, which is what matters in this explanation.}.
2472 If the D10V used ordinary byte addresses to refer to code locations,
2473 then the processor would only be able to address 64kb of instructions.
2474 However, since instructions must be aligned on four-byte boundaries, the
2475 low two bits of any valid instruction's byte address are always
2476 zero---byte addresses waste two bits. So instead of byte addresses,
2477 the D10V uses word addresses---byte addresses shifted right two bits---to
2478 refer to code. Thus, the D10V can use 16-bit words to address 256kb of
2481 However, this means that code pointers and data pointers have different
2482 forms on the D10V. The 16-bit word @code{0xC020} refers to byte address
2483 @code{0xC020} when used as a data address, but refers to byte address
2484 @code{0x30080} when used as a code address.
2486 (The D10V also uses separate code and data address spaces, which also
2487 affects the correspondence between pointers and addresses, but we're
2488 going to ignore that here; this example is already too long.)
2490 To cope with architectures like this---the D10V is not the only
2491 one!---@value{GDBN} tries to distinguish between @dfn{addresses}, which are
2492 byte numbers, and @dfn{pointers}, which are the target's representation
2493 of an address of a particular type of data. In the example above,
2494 @code{0xC020} is the pointer, which refers to one of the addresses
2495 @code{0xC020} or @code{0x30080}, depending on the type imposed upon it.
2496 @value{GDBN} provides functions for turning a pointer into an address
2497 and vice versa, in the appropriate way for the current architecture.
2499 Unfortunately, since addresses and pointers are identical on almost all
2500 processors, this distinction tends to bit-rot pretty quickly. Thus,
2501 each time you port @value{GDBN} to an architecture which does
2502 distinguish between pointers and addresses, you'll probably need to
2503 clean up some architecture-independent code.
2505 Here are functions which convert between pointers and addresses:
2507 @deftypefun CORE_ADDR extract_typed_address (void *@var{buf}, struct type *@var{type})
2508 Treat the bytes at @var{buf} as a pointer or reference of type
2509 @var{type}, and return the address it represents, in a manner
2510 appropriate for the current architecture. This yields an address
2511 @value{GDBN} can use to read target memory, disassemble, etc. Note that
2512 @var{buf} refers to a buffer in @value{GDBN}'s memory, not the
2515 For example, if the current architecture is the Intel x86, this function
2516 extracts a little-endian integer of the appropriate length from
2517 @var{buf} and returns it. However, if the current architecture is the
2518 D10V, this function will return a 16-bit integer extracted from
2519 @var{buf}, multiplied by four if @var{type} is a pointer to a function.
2521 If @var{type} is not a pointer or reference type, then this function
2522 will signal an internal error.
2525 @deftypefun CORE_ADDR store_typed_address (void *@var{buf}, struct type *@var{type}, CORE_ADDR @var{addr})
2526 Store the address @var{addr} in @var{buf}, in the proper format for a
2527 pointer of type @var{type} in the current architecture. Note that
2528 @var{buf} refers to a buffer in @value{GDBN}'s memory, not the
2531 For example, if the current architecture is the Intel x86, this function
2532 stores @var{addr} unmodified as a little-endian integer of the
2533 appropriate length in @var{buf}. However, if the current architecture
2534 is the D10V, this function divides @var{addr} by four if @var{type} is
2535 a pointer to a function, and then stores it in @var{buf}.
2537 If @var{type} is not a pointer or reference type, then this function
2538 will signal an internal error.
2541 @deftypefun CORE_ADDR value_as_address (struct value *@var{val})
2542 Assuming that @var{val} is a pointer, return the address it represents,
2543 as appropriate for the current architecture.
2545 This function actually works on integral values, as well as pointers.
2546 For pointers, it performs architecture-specific conversions as
2547 described above for @code{extract_typed_address}.
2550 @deftypefun CORE_ADDR value_from_pointer (struct type *@var{type}, CORE_ADDR @var{addr})
2551 Create and return a value representing a pointer of type @var{type} to
2552 the address @var{addr}, as appropriate for the current architecture.
2553 This function performs architecture-specific conversions as described
2554 above for @code{store_typed_address}.
2557 Here are some macros which architectures can define to indicate the
2558 relationship between pointers and addresses. These have default
2559 definitions, appropriate for architectures on which all pointers are
2560 simple unsigned byte addresses.
2562 @deftypefn {Target Macro} CORE_ADDR POINTER_TO_ADDRESS (struct type *@var{type}, char *@var{buf})
2563 Assume that @var{buf} holds a pointer of type @var{type}, in the
2564 appropriate format for the current architecture. Return the byte
2565 address the pointer refers to.
2567 This function may safely assume that @var{type} is either a pointer or a
2568 C@t{++} reference type.
2571 @deftypefn {Target Macro} void ADDRESS_TO_POINTER (struct type *@var{type}, char *@var{buf}, CORE_ADDR @var{addr})
2572 Store in @var{buf} a pointer of type @var{type} representing the address
2573 @var{addr}, in the appropriate format for the current architecture.
2575 This function may safely assume that @var{type} is either a pointer or a
2576 C@t{++} reference type.
2579 @section Address Classes
2580 @cindex address classes
2581 @cindex DW_AT_byte_size
2582 @cindex DW_AT_address_class
2584 Sometimes information about different kinds of addresses is available
2585 via the debug information. For example, some programming environments
2586 define addresses of several different sizes. If the debug information
2587 distinguishes these kinds of address classes through either the size
2588 info (e.g, @code{DW_AT_byte_size} in @w{DWARF 2}) or through an explicit
2589 address class attribute (e.g, @code{DW_AT_address_class} in @w{DWARF 2}), the
2590 following macros should be defined in order to disambiguate these
2591 types within @value{GDBN} as well as provide the added information to
2592 a @value{GDBN} user when printing type expressions.
2594 @deftypefn {Target Macro} int ADDRESS_CLASS_TYPE_FLAGS (int @var{byte_size}, int @var{dwarf2_addr_class})
2595 Returns the type flags needed to construct a pointer type whose size
2596 is @var{byte_size} and whose address class is @var{dwarf2_addr_class}.
2597 This function is normally called from within a symbol reader. See
2598 @file{dwarf2read.c}.
2601 @deftypefn {Target Macro} char *ADDRESS_CLASS_TYPE_FLAGS_TO_NAME (int @var{type_flags})
2602 Given the type flags representing an address class qualifier, return
2605 @deftypefn {Target Macro} int ADDRESS_CLASS_NAME_to_TYPE_FLAGS (int @var{name}, int *var{type_flags_ptr})
2606 Given an address qualifier name, set the @code{int} refererenced by @var{type_flags_ptr} to the type flags
2607 for that address class qualifier.
2610 Since the need for address classes is rather rare, none of
2611 the address class macros defined by default. Predicate
2612 macros are provided to detect when they are defined.
2614 Consider a hypothetical architecture in which addresses are normally
2615 32-bits wide, but 16-bit addresses are also supported. Furthermore,
2616 suppose that the @w{DWARF 2} information for this architecture simply
2617 uses a @code{DW_AT_byte_size} value of 2 to indicate the use of one
2618 of these "short" pointers. The following functions could be defined
2619 to implement the address class macros:
2622 somearch_address_class_type_flags (int byte_size,
2623 int dwarf2_addr_class)
2626 return TYPE_FLAG_ADDRESS_CLASS_1;
2632 somearch_address_class_type_flags_to_name (int type_flags)
2634 if (type_flags & TYPE_FLAG_ADDRESS_CLASS_1)
2641 somearch_address_class_name_to_type_flags (char *name,
2642 int *type_flags_ptr)
2644 if (strcmp (name, "short") == 0)
2646 *type_flags_ptr = TYPE_FLAG_ADDRESS_CLASS_1;
2654 The qualifier @code{@@short} is used in @value{GDBN}'s type expressions
2655 to indicate the presence of one of these "short" pointers. E.g, if
2656 the debug information indicates that @code{short_ptr_var} is one of these
2657 short pointers, @value{GDBN} might show the following behavior:
2660 (gdb) ptype short_ptr_var
2661 type = int * @@short
2665 @section Raw and Virtual Register Representations
2666 @cindex raw register representation
2667 @cindex virtual register representation
2668 @cindex representations, raw and virtual registers
2670 @emph{Maintainer note: This section is pretty much obsolete. The
2671 functionality described here has largely been replaced by
2672 pseudo-registers and the mechanisms described in @ref{Target
2673 Architecture Definition, , Using Different Register and Memory Data
2674 Representations}. See also @uref{http://www.gnu.org/software/gdb/bugs/,
2675 Bug Tracking Database} and
2676 @uref{http://sources.redhat.com/gdb/current/ari/, ARI Index} for more
2677 up-to-date information.}
2679 Some architectures use one representation for a value when it lives in a
2680 register, but use a different representation when it lives in memory.
2681 In @value{GDBN}'s terminology, the @dfn{raw} representation is the one used in
2682 the target registers, and the @dfn{virtual} representation is the one
2683 used in memory, and within @value{GDBN} @code{struct value} objects.
2685 @emph{Maintainer note: Notice that the same mechanism is being used to
2686 both convert a register to a @code{struct value} and alternative
2689 For almost all data types on almost all architectures, the virtual and
2690 raw representations are identical, and no special handling is needed.
2691 However, they do occasionally differ. For example:
2695 The x86 architecture supports an 80-bit @code{long double} type. However, when
2696 we store those values in memory, they occupy twelve bytes: the
2697 floating-point number occupies the first ten, and the final two bytes
2698 are unused. This keeps the values aligned on four-byte boundaries,
2699 allowing more efficient access. Thus, the x86 80-bit floating-point
2700 type is the raw representation, and the twelve-byte loosely-packed
2701 arrangement is the virtual representation.
2704 Some 64-bit MIPS targets present 32-bit registers to @value{GDBN} as 64-bit
2705 registers, with garbage in their upper bits. @value{GDBN} ignores the top 32
2706 bits. Thus, the 64-bit form, with garbage in the upper 32 bits, is the
2707 raw representation, and the trimmed 32-bit representation is the
2708 virtual representation.
2711 In general, the raw representation is determined by the architecture, or
2712 @value{GDBN}'s interface to the architecture, while the virtual representation
2713 can be chosen for @value{GDBN}'s convenience. @value{GDBN}'s register file,
2714 @code{registers}, holds the register contents in raw format, and the
2715 @value{GDBN} remote protocol transmits register values in raw format.
2717 Your architecture may define the following macros to request
2718 conversions between the raw and virtual format:
2720 @deftypefn {Target Macro} int REGISTER_CONVERTIBLE (int @var{reg})
2721 Return non-zero if register number @var{reg}'s value needs different raw
2722 and virtual formats.
2724 You should not use @code{REGISTER_CONVERT_TO_VIRTUAL} for a register
2725 unless this macro returns a non-zero value for that register.
2728 @deftypefn {Target Macro} int DEPRECATED_REGISTER_RAW_SIZE (int @var{reg})
2729 The size of register number @var{reg}'s raw value. This is the number
2730 of bytes the register will occupy in @code{registers}, or in a @value{GDBN}
2731 remote protocol packet.
2734 @deftypefn {Target Macro} int DEPRECATED_REGISTER_VIRTUAL_SIZE (int @var{reg})
2735 The size of register number @var{reg}'s value, in its virtual format.
2736 This is the size a @code{struct value}'s buffer will have, holding that
2740 @deftypefn {Target Macro} struct type *DEPRECATED_REGISTER_VIRTUAL_TYPE (int @var{reg})
2741 This is the type of the virtual representation of register number
2742 @var{reg}. Note that there is no need for a macro giving a type for the
2743 register's raw form; once the register's value has been obtained, @value{GDBN}
2744 always uses the virtual form.
2747 @deftypefn {Target Macro} void REGISTER_CONVERT_TO_VIRTUAL (int @var{reg}, struct type *@var{type}, char *@var{from}, char *@var{to})
2748 Convert the value of register number @var{reg} to @var{type}, which
2749 should always be @code{DEPRECATED_REGISTER_VIRTUAL_TYPE (@var{reg})}. The buffer
2750 at @var{from} holds the register's value in raw format; the macro should
2751 convert the value to virtual format, and place it at @var{to}.
2753 Note that @code{REGISTER_CONVERT_TO_VIRTUAL} and
2754 @code{REGISTER_CONVERT_TO_RAW} take their @var{reg} and @var{type}
2755 arguments in different orders.
2757 You should only use @code{REGISTER_CONVERT_TO_VIRTUAL} with registers
2758 for which the @code{REGISTER_CONVERTIBLE} macro returns a non-zero
2762 @deftypefn {Target Macro} void REGISTER_CONVERT_TO_RAW (struct type *@var{type}, int @var{reg}, char *@var{from}, char *@var{to})
2763 Convert the value of register number @var{reg} to @var{type}, which
2764 should always be @code{DEPRECATED_REGISTER_VIRTUAL_TYPE (@var{reg})}. The buffer
2765 at @var{from} holds the register's value in raw format; the macro should
2766 convert the value to virtual format, and place it at @var{to}.
2768 Note that REGISTER_CONVERT_TO_VIRTUAL and REGISTER_CONVERT_TO_RAW take
2769 their @var{reg} and @var{type} arguments in different orders.
2773 @section Using Different Register and Memory Data Representations
2774 @cindex register representation
2775 @cindex memory representation
2776 @cindex representations, register and memory
2777 @cindex register data formats, converting
2778 @cindex @code{struct value}, converting register contents to
2780 @emph{Maintainer's note: The way GDB manipulates registers is undergoing
2781 significant change. Many of the macros and functions refered to in this
2782 section are likely to be subject to further revision. See
2783 @uref{http://sources.redhat.com/gdb/current/ari/, A.R. Index} and
2784 @uref{http://www.gnu.org/software/gdb/bugs, Bug Tracking Database} for
2785 further information. cagney/2002-05-06.}
2787 Some architectures can represent a data object in a register using a
2788 form that is different to the objects more normal memory representation.
2794 The Alpha architecture can represent 32 bit integer values in
2795 floating-point registers.
2798 The x86 architecture supports 80-bit floating-point registers. The
2799 @code{long double} data type occupies 96 bits in memory but only 80 bits
2800 when stored in a register.
2804 In general, the register representation of a data type is determined by
2805 the architecture, or @value{GDBN}'s interface to the architecture, while
2806 the memory representation is determined by the Application Binary
2809 For almost all data types on almost all architectures, the two
2810 representations are identical, and no special handling is needed.
2811 However, they do occasionally differ. Your architecture may define the
2812 following macros to request conversions between the register and memory
2813 representations of a data type:
2815 @deftypefn {Target Macro} int CONVERT_REGISTER_P (int @var{reg})
2816 Return non-zero if the representation of a data value stored in this
2817 register may be different to the representation of that same data value
2818 when stored in memory.
2820 When non-zero, the macros @code{REGISTER_TO_VALUE} and
2821 @code{VALUE_TO_REGISTER} are used to perform any necessary conversion.
2824 @deftypefn {Target Macro} void REGISTER_TO_VALUE (int @var{reg}, struct type *@var{type}, char *@var{from}, char *@var{to})
2825 Convert the value of register number @var{reg} to a data object of type
2826 @var{type}. The buffer at @var{from} holds the register's value in raw
2827 format; the converted value should be placed in the buffer at @var{to}.
2829 Note that @code{REGISTER_TO_VALUE} and @code{VALUE_TO_REGISTER} take
2830 their @var{reg} and @var{type} arguments in different orders.
2832 You should only use @code{REGISTER_TO_VALUE} with registers for which
2833 the @code{CONVERT_REGISTER_P} macro returns a non-zero value.
2836 @deftypefn {Target Macro} void VALUE_TO_REGISTER (struct type *@var{type}, int @var{reg}, char *@var{from}, char *@var{to})
2837 Convert a data value of type @var{type} to register number @var{reg}'
2840 Note that @code{REGISTER_TO_VALUE} and @code{VALUE_TO_REGISTER} take
2841 their @var{reg} and @var{type} arguments in different orders.
2843 You should only use @code{VALUE_TO_REGISTER} with registers for which
2844 the @code{CONVERT_REGISTER_P} macro returns a non-zero value.
2847 @deftypefn {Target Macro} void REGISTER_CONVERT_TO_TYPE (int @var{regnum}, struct type *@var{type}, char *@var{buf})
2848 See @file{mips-tdep.c}. It does not do what you want.
2852 @section Frame Interpretation
2854 @section Inferior Call Setup
2856 @section Compiler Characteristics
2858 @section Target Conditionals
2860 This section describes the macros that you can use to define the target
2865 @item ADDR_BITS_REMOVE (addr)
2866 @findex ADDR_BITS_REMOVE
2867 If a raw machine instruction address includes any bits that are not
2868 really part of the address, then define this macro to expand into an
2869 expression that zeroes those bits in @var{addr}. This is only used for
2870 addresses of instructions, and even then not in all contexts.
2872 For example, the two low-order bits of the PC on the Hewlett-Packard PA
2873 2.0 architecture contain the privilege level of the corresponding
2874 instruction. Since instructions must always be aligned on four-byte
2875 boundaries, the processor masks out these bits to generate the actual
2876 address of the instruction. ADDR_BITS_REMOVE should filter out these
2877 bits with an expression such as @code{((addr) & ~3)}.
2879 @item ADDRESS_CLASS_NAME_TO_TYPE_FLAGS (@var{name}, @var{type_flags_ptr})
2880 @findex ADDRESS_CLASS_NAME_TO_TYPE_FLAGS
2881 If @var{name} is a valid address class qualifier name, set the @code{int}
2882 referenced by @var{type_flags_ptr} to the mask representing the qualifier
2883 and return 1. If @var{name} is not a valid address class qualifier name,
2886 The value for @var{type_flags_ptr} should be one of
2887 @code{TYPE_FLAG_ADDRESS_CLASS_1}, @code{TYPE_FLAG_ADDRESS_CLASS_2}, or
2888 possibly some combination of these values or'd together.
2889 @xref{Target Architecture Definition, , Address Classes}.
2891 @item ADDRESS_CLASS_NAME_TO_TYPE_FLAGS_P ()
2892 @findex ADDRESS_CLASS_NAME_TO_TYPE_FLAGS_P
2893 Predicate which indicates whether @code{ADDRESS_CLASS_NAME_TO_TYPE_FLAGS}
2896 @item ADDRESS_CLASS_TYPE_FLAGS (@var{byte_size}, @var{dwarf2_addr_class})
2897 @findex ADDRESS_CLASS_TYPE_FLAGS (@var{byte_size}, @var{dwarf2_addr_class})
2898 Given a pointers byte size (as described by the debug information) and
2899 the possible @code{DW_AT_address_class} value, return the type flags
2900 used by @value{GDBN} to represent this address class. The value
2901 returned should be one of @code{TYPE_FLAG_ADDRESS_CLASS_1},
2902 @code{TYPE_FLAG_ADDRESS_CLASS_2}, or possibly some combination of these
2903 values or'd together.
2904 @xref{Target Architecture Definition, , Address Classes}.
2906 @item ADDRESS_CLASS_TYPE_FLAGS_P ()
2907 @findex ADDRESS_CLASS_TYPE_FLAGS_P
2908 Predicate which indicates whether @code{ADDRESS_CLASS_TYPE_FLAGS} has
2911 @item ADDRESS_CLASS_TYPE_FLAGS_TO_NAME (@var{type_flags})
2912 @findex ADDRESS_CLASS_TYPE_FLAGS_TO_NAME
2913 Return the name of the address class qualifier associated with the type
2914 flags given by @var{type_flags}.
2916 @item ADDRESS_CLASS_TYPE_FLAGS_TO_NAME_P ()
2917 @findex ADDRESS_CLASS_TYPE_FLAGS_TO_NAME_P
2918 Predicate which indicates whether @code{ADDRESS_CLASS_TYPE_FLAGS_TO_NAME} has
2920 @xref{Target Architecture Definition, , Address Classes}.
2922 @item ADDRESS_TO_POINTER (@var{type}, @var{buf}, @var{addr})
2923 @findex ADDRESS_TO_POINTER
2924 Store in @var{buf} a pointer of type @var{type} representing the address
2925 @var{addr}, in the appropriate format for the current architecture.
2926 This macro may safely assume that @var{type} is either a pointer or a
2927 C@t{++} reference type.
2928 @xref{Target Architecture Definition, , Pointers Are Not Always Addresses}.
2930 @item BELIEVE_PCC_PROMOTION
2931 @findex BELIEVE_PCC_PROMOTION
2932 Define if the compiler promotes a @code{short} or @code{char}
2933 parameter to an @code{int}, but still reports the parameter as its
2934 original type, rather than the promoted type.
2936 @item BITS_BIG_ENDIAN
2937 @findex BITS_BIG_ENDIAN
2938 Define this if the numbering of bits in the targets does @strong{not} match the
2939 endianness of the target byte order. A value of 1 means that the bits
2940 are numbered in a big-endian bit order, 0 means little-endian.
2944 This is the character array initializer for the bit pattern to put into
2945 memory where a breakpoint is set. Although it's common to use a trap
2946 instruction for a breakpoint, it's not required; for instance, the bit
2947 pattern could be an invalid instruction. The breakpoint must be no
2948 longer than the shortest instruction of the architecture.
2950 @code{BREAKPOINT} has been deprecated in favor of
2951 @code{BREAKPOINT_FROM_PC}.
2953 @item BIG_BREAKPOINT
2954 @itemx LITTLE_BREAKPOINT
2955 @findex LITTLE_BREAKPOINT
2956 @findex BIG_BREAKPOINT
2957 Similar to BREAKPOINT, but used for bi-endian targets.
2959 @code{BIG_BREAKPOINT} and @code{LITTLE_BREAKPOINT} have been deprecated in
2960 favor of @code{BREAKPOINT_FROM_PC}.
2962 @item DEPRECATED_REMOTE_BREAKPOINT
2963 @itemx DEPRECATED_LITTLE_REMOTE_BREAKPOINT
2964 @itemx DEPRECATED_BIG_REMOTE_BREAKPOINT
2965 @findex DEPRECATED_BIG_REMOTE_BREAKPOINT
2966 @findex DEPRECATED_LITTLE_REMOTE_BREAKPOINT
2967 @findex DEPRECATED_REMOTE_BREAKPOINT
2968 Specify the breakpoint instruction sequence for a remote target.
2969 @code{DEPRECATED_REMOTE_BREAKPOINT},
2970 @code{DEPRECATED_BIG_REMOTE_BREAKPOINT} and
2971 @code{DEPRECATED_LITTLE_REMOTE_BREAKPOINT} have been deprecated in
2972 favor of @code{BREAKPOINT_FROM_PC} (@pxref{BREAKPOINT_FROM_PC}).
2974 @item BREAKPOINT_FROM_PC (@var{pcptr}, @var{lenptr})
2975 @findex BREAKPOINT_FROM_PC
2976 @anchor{BREAKPOINT_FROM_PC} Use the program counter to determine the
2977 contents and size of a breakpoint instruction. It returns a pointer to
2978 a string of bytes that encode a breakpoint instruction, stores the
2979 length of the string to @code{*@var{lenptr}}, and adjusts the program
2980 counter (if necessary) to point to the actual memory location where the
2981 breakpoint should be inserted.
2983 Although it is common to use a trap instruction for a breakpoint, it's
2984 not required; for instance, the bit pattern could be an invalid
2985 instruction. The breakpoint must be no longer than the shortest
2986 instruction of the architecture.
2988 Replaces all the other @var{BREAKPOINT} macros.
2990 @item MEMORY_INSERT_BREAKPOINT (@var{addr}, @var{contents_cache})
2991 @itemx MEMORY_REMOVE_BREAKPOINT (@var{addr}, @var{contents_cache})
2992 @findex MEMORY_REMOVE_BREAKPOINT
2993 @findex MEMORY_INSERT_BREAKPOINT
2994 Insert or remove memory based breakpoints. Reasonable defaults
2995 (@code{default_memory_insert_breakpoint} and
2996 @code{default_memory_remove_breakpoint} respectively) have been
2997 provided so that it is not necessary to define these for most
2998 architectures. Architectures which may want to define
2999 @code{MEMORY_INSERT_BREAKPOINT} and @code{MEMORY_REMOVE_BREAKPOINT} will
3000 likely have instructions that are oddly sized or are not stored in a
3001 conventional manner.
3003 It may also be desirable (from an efficiency standpoint) to define
3004 custom breakpoint insertion and removal routines if
3005 @code{BREAKPOINT_FROM_PC} needs to read the target's memory for some
3008 @item ADJUST_BREAKPOINT_ADDRESS (@var{address})
3009 @findex ADJUST_BREAKPOINT_ADDRESS
3010 @cindex breakpoint address adjusted
3011 Given an address at which a breakpoint is desired, return a breakpoint
3012 address adjusted to account for architectural constraints on
3013 breakpoint placement. This method is not needed by most targets.
3015 The FR-V target (see @file{frv-tdep.c}) requires this method.
3016 The FR-V is a VLIW architecture in which a number of RISC-like
3017 instructions are grouped (packed) together into an aggregate
3018 instruction or instruction bundle. When the processor executes
3019 one of these bundles, the component instructions are executed
3022 In the course of optimization, the compiler may group instructions
3023 from distinct source statements into the same bundle. The line number
3024 information associated with one of the latter statements will likely
3025 refer to some instruction other than the first one in the bundle. So,
3026 if the user attempts to place a breakpoint on one of these latter
3027 statements, @value{GDBN} must be careful to @emph{not} place the break
3028 instruction on any instruction other than the first one in the bundle.
3029 (Remember though that the instructions within a bundle execute
3030 in parallel, so the @emph{first} instruction is the instruction
3031 at the lowest address and has nothing to do with execution order.)
3033 The FR-V's @code{ADJUST_BREAKPOINT_ADDRESS} method will adjust a
3034 breakpoint's address by scanning backwards for the beginning of
3035 the bundle, returning the address of the bundle.
3037 Since the adjustment of a breakpoint may significantly alter a user's
3038 expectation, @value{GDBN} prints a warning when an adjusted breakpoint
3039 is initially set and each time that that breakpoint is hit.
3041 @item CALL_DUMMY_LOCATION
3042 @findex CALL_DUMMY_LOCATION
3043 See the file @file{inferior.h}.
3045 This method has been replaced by @code{push_dummy_code}
3046 (@pxref{push_dummy_code}).
3048 @item CANNOT_FETCH_REGISTER (@var{regno})
3049 @findex CANNOT_FETCH_REGISTER
3050 A C expression that should be nonzero if @var{regno} cannot be fetched
3051 from an inferior process. This is only relevant if
3052 @code{FETCH_INFERIOR_REGISTERS} is not defined.
3054 @item CANNOT_STORE_REGISTER (@var{regno})
3055 @findex CANNOT_STORE_REGISTER
3056 A C expression that should be nonzero if @var{regno} should not be
3057 written to the target. This is often the case for program counters,
3058 status words, and other special registers. If this is not defined,
3059 @value{GDBN} will assume that all registers may be written.
3061 @item int CONVERT_REGISTER_P(@var{regnum})
3062 @findex CONVERT_REGISTER_P
3063 Return non-zero if register @var{regnum} can represent data values in a
3065 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
3067 @item DECR_PC_AFTER_BREAK
3068 @findex DECR_PC_AFTER_BREAK
3069 Define this to be the amount by which to decrement the PC after the
3070 program encounters a breakpoint. This is often the number of bytes in
3071 @code{BREAKPOINT}, though not always. For most targets this value will be 0.
3073 @item DISABLE_UNSETTABLE_BREAK (@var{addr})
3074 @findex DISABLE_UNSETTABLE_BREAK
3075 If defined, this should evaluate to 1 if @var{addr} is in a shared
3076 library in which breakpoints cannot be set and so should be disabled.
3078 @item PRINT_FLOAT_INFO()
3079 @findex PRINT_FLOAT_INFO
3080 If defined, then the @samp{info float} command will print information about
3081 the processor's floating point unit.
3083 @item print_registers_info (@var{gdbarch}, @var{frame}, @var{regnum}, @var{all})
3084 @findex print_registers_info
3085 If defined, pretty print the value of the register @var{regnum} for the
3086 specified @var{frame}. If the value of @var{regnum} is -1, pretty print
3087 either all registers (@var{all} is non zero) or a select subset of
3088 registers (@var{all} is zero).
3090 The default method prints one register per line, and if @var{all} is
3091 zero omits floating-point registers.
3093 @item PRINT_VECTOR_INFO()
3094 @findex PRINT_VECTOR_INFO
3095 If defined, then the @samp{info vector} command will call this function
3096 to print information about the processor's vector unit.
3098 By default, the @samp{info vector} command will print all vector
3099 registers (the register's type having the vector attribute).
3101 @item DWARF_REG_TO_REGNUM
3102 @findex DWARF_REG_TO_REGNUM
3103 Convert DWARF register number into @value{GDBN} regnum. If not defined,
3104 no conversion will be performed.
3106 @item DWARF2_REG_TO_REGNUM
3107 @findex DWARF2_REG_TO_REGNUM
3108 Convert DWARF2 register number into @value{GDBN} regnum. If not
3109 defined, no conversion will be performed.
3111 @item ECOFF_REG_TO_REGNUM
3112 @findex ECOFF_REG_TO_REGNUM
3113 Convert ECOFF register number into @value{GDBN} regnum. If not defined,
3114 no conversion will be performed.
3116 @item END_OF_TEXT_DEFAULT
3117 @findex END_OF_TEXT_DEFAULT
3118 This is an expression that should designate the end of the text section.
3121 @item EXTRACT_RETURN_VALUE(@var{type}, @var{regbuf}, @var{valbuf})
3122 @findex EXTRACT_RETURN_VALUE
3123 Define this to extract a function's return value of type @var{type} from
3124 the raw register state @var{regbuf} and copy that, in virtual format,
3127 This method has been deprecated in favour of @code{gdbarch_return_value}
3128 (@pxref{gdbarch_return_value}).
3130 @item DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS(@var{regbuf})
3131 @findex DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS
3132 @anchor{DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS}
3133 When defined, extract from the array @var{regbuf} (containing the raw
3134 register state) the @code{CORE_ADDR} at which a function should return
3135 its structure value.
3137 @xref{gdbarch_return_value}.
3139 @item DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS_P()
3140 @findex DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS_P
3141 Predicate for @code{DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS}.
3143 @item DEPRECATED_FP_REGNUM
3144 @findex DEPRECATED_FP_REGNUM
3145 If the virtual frame pointer is kept in a register, then define this
3146 macro to be the number (greater than or equal to zero) of that register.
3148 This should only need to be defined if @code{DEPRECATED_TARGET_READ_FP}
3151 @item DEPRECATED_FRAMELESS_FUNCTION_INVOCATION(@var{fi})
3152 @findex DEPRECATED_FRAMELESS_FUNCTION_INVOCATION
3153 Define this to an expression that returns 1 if the function invocation
3154 represented by @var{fi} does not have a stack frame associated with it.
3157 @item frame_align (@var{address})
3158 @anchor{frame_align}
3160 Define this to adjust @var{address} so that it meets the alignment
3161 requirements for the start of a new stack frame. A stack frame's
3162 alignment requirements are typically stronger than a target processors
3163 stack alignment requirements (@pxref{DEPRECATED_STACK_ALIGN}).
3165 This function is used to ensure that, when creating a dummy frame, both
3166 the initial stack pointer and (if needed) the address of the return
3167 value are correctly aligned.
3169 Unlike @code{DEPRECATED_STACK_ALIGN}, this function always adjusts the
3170 address in the direction of stack growth.
3172 By default, no frame based stack alignment is performed.
3174 @item int frame_red_zone_size
3176 The number of bytes, beyond the innermost-stack-address, reserved by the
3177 @sc{abi}. A function is permitted to use this scratch area (instead of
3178 allocating extra stack space).
3180 When performing an inferior function call, to ensure that it does not
3181 modify this area, @value{GDBN} adjusts the innermost-stack-address by
3182 @var{frame_red_zone_size} bytes before pushing parameters onto the
3185 By default, zero bytes are allocated. The value must be aligned
3186 (@pxref{frame_align}).
3188 The @sc{amd64} (nee x86-64) @sc{abi} documentation refers to the
3189 @emph{red zone} when describing this scratch area.
3192 @item DEPRECATED_FRAME_CHAIN(@var{frame})
3193 @findex DEPRECATED_FRAME_CHAIN
3194 Given @var{frame}, return a pointer to the calling frame.
3196 @item DEPRECATED_FRAME_CHAIN_VALID(@var{chain}, @var{thisframe})
3197 @findex DEPRECATED_FRAME_CHAIN_VALID
3198 Define this to be an expression that returns zero if the given frame is an
3199 outermost frame, with no caller, and nonzero otherwise. Most normal
3200 situations can be handled without defining this macro, including @code{NULL}
3201 chain pointers, dummy frames, and frames whose PC values are inside the
3202 startup file (e.g.@: @file{crt0.o}), inside @code{main}, or inside
3205 @item DEPRECATED_FRAME_INIT_SAVED_REGS(@var{frame})
3206 @findex DEPRECATED_FRAME_INIT_SAVED_REGS
3207 See @file{frame.h}. Determines the address of all registers in the
3208 current stack frame storing each in @code{frame->saved_regs}. Space for
3209 @code{frame->saved_regs} shall be allocated by
3210 @code{DEPRECATED_FRAME_INIT_SAVED_REGS} using
3211 @code{frame_saved_regs_zalloc}.
3213 @code{FRAME_FIND_SAVED_REGS} is deprecated.
3215 @item FRAME_NUM_ARGS (@var{fi})
3216 @findex FRAME_NUM_ARGS
3217 For the frame described by @var{fi} return the number of arguments that
3218 are being passed. If the number of arguments is not known, return
3221 @item DEPRECATED_FRAME_SAVED_PC(@var{frame})
3222 @findex DEPRECATED_FRAME_SAVED_PC
3223 @anchor{DEPRECATED_FRAME_SAVED_PC} Given @var{frame}, return the pc
3224 saved there. This is the return address.
3226 This method is deprecated. @xref{unwind_pc}.
3228 @item CORE_ADDR unwind_pc (struct frame_info *@var{this_frame})
3230 @anchor{unwind_pc} Return the instruction address, in @var{this_frame}'s
3231 caller, at which execution will resume after @var{this_frame} returns.
3232 This is commonly refered to as the return address.
3234 The implementation, which must be frame agnostic (work with any frame),
3235 is typically no more than:
3239 frame_unwind_unsigned_register (this_frame, D10V_PC_REGNUM, &pc);
3240 return d10v_make_iaddr (pc);
3244 @xref{DEPRECATED_FRAME_SAVED_PC}, which this method replaces.
3246 @item CORE_ADDR unwind_sp (struct frame_info *@var{this_frame})
3248 @anchor{unwind_sp} Return the frame's inner most stack address. This is
3249 commonly refered to as the frame's @dfn{stack pointer}.
3251 The implementation, which must be frame agnostic (work with any frame),
3252 is typically no more than:
3256 frame_unwind_unsigned_register (this_frame, D10V_SP_REGNUM, &sp);
3257 return d10v_make_daddr (sp);
3261 @xref{TARGET_READ_SP}, which this method replaces.
3263 @item FUNCTION_EPILOGUE_SIZE
3264 @findex FUNCTION_EPILOGUE_SIZE
3265 For some COFF targets, the @code{x_sym.x_misc.x_fsize} field of the
3266 function end symbol is 0. For such targets, you must define
3267 @code{FUNCTION_EPILOGUE_SIZE} to expand into the standard size of a
3268 function's epilogue.
3270 @item DEPRECATED_FUNCTION_START_OFFSET
3271 @findex DEPRECATED_FUNCTION_START_OFFSET
3272 An integer, giving the offset in bytes from a function's address (as
3273 used in the values of symbols, function pointers, etc.), and the
3274 function's first genuine instruction.
3276 This is zero on almost all machines: the function's address is usually
3277 the address of its first instruction. However, on the VAX, for
3278 example, each function starts with two bytes containing a bitmask
3279 indicating which registers to save upon entry to the function. The
3280 VAX @code{call} instructions check this value, and save the
3281 appropriate registers automatically. Thus, since the offset from the
3282 function's address to its first instruction is two bytes,
3283 @code{DEPRECATED_FUNCTION_START_OFFSET} would be 2 on the VAX.
3285 @item GCC_COMPILED_FLAG_SYMBOL
3286 @itemx GCC2_COMPILED_FLAG_SYMBOL
3287 @findex GCC2_COMPILED_FLAG_SYMBOL
3288 @findex GCC_COMPILED_FLAG_SYMBOL
3289 If defined, these are the names of the symbols that @value{GDBN} will
3290 look for to detect that GCC compiled the file. The default symbols
3291 are @code{gcc_compiled.} and @code{gcc2_compiled.},
3292 respectively. (Currently only defined for the Delta 68.)
3294 @item @value{GDBN}_MULTI_ARCH
3295 @findex @value{GDBN}_MULTI_ARCH
3296 If defined and non-zero, enables support for multiple architectures
3297 within @value{GDBN}.
3299 This support can be enabled at two levels. At level one, only
3300 definitions for previously undefined macros are provided; at level two,
3301 a multi-arch definition of all architecture dependent macros will be
3304 @item @value{GDBN}_TARGET_IS_HPPA
3305 @findex @value{GDBN}_TARGET_IS_HPPA
3306 This determines whether horrible kludge code in @file{dbxread.c} and
3307 @file{partial-stab.h} is used to mangle multiple-symbol-table files from
3308 HPPA's. This should all be ripped out, and a scheme like @file{elfread.c}
3311 @item GET_LONGJMP_TARGET
3312 @findex GET_LONGJMP_TARGET
3313 For most machines, this is a target-dependent parameter. On the
3314 DECstation and the Iris, this is a native-dependent parameter, since
3315 the header file @file{setjmp.h} is needed to define it.
3317 This macro determines the target PC address that @code{longjmp} will jump to,
3318 assuming that we have just stopped at a @code{longjmp} breakpoint. It takes a
3319 @code{CORE_ADDR *} as argument, and stores the target PC value through this
3320 pointer. It examines the current state of the machine as needed.
3322 @item DEPRECATED_GET_SAVED_REGISTER
3323 @findex DEPRECATED_GET_SAVED_REGISTER
3324 Define this if you need to supply your own definition for the function
3325 @code{DEPRECATED_GET_SAVED_REGISTER}.
3327 @item DEPRECATED_IBM6000_TARGET
3328 @findex DEPRECATED_IBM6000_TARGET
3329 Shows that we are configured for an IBM RS/6000 system. This
3330 conditional should be eliminated (FIXME) and replaced by
3331 feature-specific macros. It was introduced in a haste and we are
3332 repenting at leisure.
3334 @item I386_USE_GENERIC_WATCHPOINTS
3335 An x86-based target can define this to use the generic x86 watchpoint
3336 support; see @ref{Algorithms, I386_USE_GENERIC_WATCHPOINTS}.
3338 @item SYMBOLS_CAN_START_WITH_DOLLAR
3339 @findex SYMBOLS_CAN_START_WITH_DOLLAR
3340 Some systems have routines whose names start with @samp{$}. Giving this
3341 macro a non-zero value tells @value{GDBN}'s expression parser to check for such
3342 routines when parsing tokens that begin with @samp{$}.
3344 On HP-UX, certain system routines (millicode) have names beginning with
3345 @samp{$} or @samp{$$}. For example, @code{$$dyncall} is a millicode
3346 routine that handles inter-space procedure calls on PA-RISC.
3348 @item DEPRECATED_INIT_EXTRA_FRAME_INFO (@var{fromleaf}, @var{frame})
3349 @findex DEPRECATED_INIT_EXTRA_FRAME_INFO
3350 If additional information about the frame is required this should be
3351 stored in @code{frame->extra_info}. Space for @code{frame->extra_info}
3352 is allocated using @code{frame_extra_info_zalloc}.
3354 @item DEPRECATED_INIT_FRAME_PC (@var{fromleaf}, @var{prev})
3355 @findex DEPRECATED_INIT_FRAME_PC
3356 This is a C statement that sets the pc of the frame pointed to by
3357 @var{prev}. [By default...]
3359 @item INNER_THAN (@var{lhs}, @var{rhs})
3361 Returns non-zero if stack address @var{lhs} is inner than (nearer to the
3362 stack top) stack address @var{rhs}. Define this as @code{lhs < rhs} if
3363 the target's stack grows downward in memory, or @code{lhs > rsh} if the
3366 @item gdbarch_in_function_epilogue_p (@var{gdbarch}, @var{pc})
3367 @findex gdbarch_in_function_epilogue_p
3368 Returns non-zero if the given @var{pc} is in the epilogue of a function.
3369 The epilogue of a function is defined as the part of a function where
3370 the stack frame of the function already has been destroyed up to the
3371 final `return from function call' instruction.
3373 @item DEPRECATED_SIGTRAMP_START (@var{pc})
3374 @findex DEPRECATED_SIGTRAMP_START
3375 @itemx DEPRECATED_SIGTRAMP_END (@var{pc})
3376 @findex DEPRECATED_SIGTRAMP_END
3377 Define these to be the start and end address of the @code{sigtramp} for the
3378 given @var{pc}. On machines where the address is just a compile time
3379 constant, the macro expansion will typically just ignore the supplied
3382 @item IN_SOLIB_CALL_TRAMPOLINE (@var{pc}, @var{name})
3383 @findex IN_SOLIB_CALL_TRAMPOLINE
3384 Define this to evaluate to nonzero if the program is stopped in the
3385 trampoline that connects to a shared library.
3387 @item IN_SOLIB_RETURN_TRAMPOLINE (@var{pc}, @var{name})
3388 @findex IN_SOLIB_RETURN_TRAMPOLINE
3389 Define this to evaluate to nonzero if the program is stopped in the
3390 trampoline that returns from a shared library.
3392 @item IN_SOLIB_DYNSYM_RESOLVE_CODE (@var{pc})
3393 @findex IN_SOLIB_DYNSYM_RESOLVE_CODE
3394 Define this to evaluate to nonzero if the program is stopped in the
3397 @item SKIP_SOLIB_RESOLVER (@var{pc})
3398 @findex SKIP_SOLIB_RESOLVER
3399 Define this to evaluate to the (nonzero) address at which execution
3400 should continue to get past the dynamic linker's symbol resolution
3401 function. A zero value indicates that it is not important or necessary
3402 to set a breakpoint to get through the dynamic linker and that single
3403 stepping will suffice.
3405 @item INTEGER_TO_ADDRESS (@var{type}, @var{buf})
3406 @findex INTEGER_TO_ADDRESS
3407 @cindex converting integers to addresses
3408 Define this when the architecture needs to handle non-pointer to address
3409 conversions specially. Converts that value to an address according to
3410 the current architectures conventions.
3412 @emph{Pragmatics: When the user copies a well defined expression from
3413 their source code and passes it, as a parameter, to @value{GDBN}'s
3414 @code{print} command, they should get the same value as would have been
3415 computed by the target program. Any deviation from this rule can cause
3416 major confusion and annoyance, and needs to be justified carefully. In
3417 other words, @value{GDBN} doesn't really have the freedom to do these
3418 conversions in clever and useful ways. It has, however, been pointed
3419 out that users aren't complaining about how @value{GDBN} casts integers
3420 to pointers; they are complaining that they can't take an address from a
3421 disassembly listing and give it to @code{x/i}. Adding an architecture
3422 method like @code{INTEGER_TO_ADDRESS} certainly makes it possible for
3423 @value{GDBN} to ``get it right'' in all circumstances.}
3425 @xref{Target Architecture Definition, , Pointers Are Not Always
3428 @item NO_HIF_SUPPORT
3429 @findex NO_HIF_SUPPORT
3430 (Specific to the a29k.)
3432 @item POINTER_TO_ADDRESS (@var{type}, @var{buf})
3433 @findex POINTER_TO_ADDRESS
3434 Assume that @var{buf} holds a pointer of type @var{type}, in the
3435 appropriate format for the current architecture. Return the byte
3436 address the pointer refers to.
3437 @xref{Target Architecture Definition, , Pointers Are Not Always Addresses}.
3439 @item REGISTER_CONVERTIBLE (@var{reg})
3440 @findex REGISTER_CONVERTIBLE
3441 Return non-zero if @var{reg} uses different raw and virtual formats.
3442 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3444 @item REGISTER_TO_VALUE(@var{regnum}, @var{type}, @var{from}, @var{to})
3445 @findex REGISTER_TO_VALUE
3446 Convert the raw contents of register @var{regnum} into a value of type
3448 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
3450 @item DEPRECATED_REGISTER_RAW_SIZE (@var{reg})
3451 @findex DEPRECATED_REGISTER_RAW_SIZE
3452 Return the raw size of @var{reg}; defaults to the size of the register's
3454 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3456 @item register_reggroup_p (@var{gdbarch}, @var{regnum}, @var{reggroup})
3457 @findex register_reggroup_p
3458 @cindex register groups
3459 Return non-zero if register @var{regnum} is a member of the register
3460 group @var{reggroup}.
3462 By default, registers are grouped as follows:
3465 @item float_reggroup
3466 Any register with a valid name and a floating-point type.
3467 @item vector_reggroup
3468 Any register with a valid name and a vector type.
3469 @item general_reggroup
3470 Any register with a valid name and a type other than vector or
3471 floating-point. @samp{float_reggroup}.
3473 @itemx restore_reggroup
3475 Any register with a valid name.
3478 @item DEPRECATED_REGISTER_VIRTUAL_SIZE (@var{reg})
3479 @findex DEPRECATED_REGISTER_VIRTUAL_SIZE
3480 Return the virtual size of @var{reg}; defaults to the size of the
3481 register's virtual type.
3482 Return the virtual size of @var{reg}.
3483 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3485 @item DEPRECATED_REGISTER_VIRTUAL_TYPE (@var{reg})
3486 @findex REGISTER_VIRTUAL_TYPE
3487 Return the virtual type of @var{reg}.
3488 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3490 @item struct type *register_type (@var{gdbarch}, @var{reg})
3491 @findex register_type
3492 If defined, return the type of register @var{reg}. This function
3493 superseeds @code{DEPRECATED_REGISTER_VIRTUAL_TYPE}. @xref{Target Architecture
3494 Definition, , Raw and Virtual Register Representations}.
3496 @item REGISTER_CONVERT_TO_VIRTUAL(@var{reg}, @var{type}, @var{from}, @var{to})
3497 @findex REGISTER_CONVERT_TO_VIRTUAL
3498 Convert the value of register @var{reg} from its raw form to its virtual
3500 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3502 @item REGISTER_CONVERT_TO_RAW(@var{type}, @var{reg}, @var{from}, @var{to})
3503 @findex REGISTER_CONVERT_TO_RAW
3504 Convert the value of register @var{reg} from its virtual form to its raw
3506 @xref{Target Architecture Definition, , Raw and Virtual Register Representations}.
3508 @item const struct regset *regset_from_core_section (struct gdbarch * @var{gdbarch}, const char * @var{sect_name}, size_t @var{sect_size})
3509 @findex regset_from_core_section
3510 Return the appropriate register set for a core file section with name
3511 @var{sect_name} and size @var{sect_size}.
3513 @item SOFTWARE_SINGLE_STEP_P()
3514 @findex SOFTWARE_SINGLE_STEP_P
3515 Define this as 1 if the target does not have a hardware single-step
3516 mechanism. The macro @code{SOFTWARE_SINGLE_STEP} must also be defined.
3518 @item SOFTWARE_SINGLE_STEP(@var{signal}, @var{insert_breapoints_p})
3519 @findex SOFTWARE_SINGLE_STEP
3520 A function that inserts or removes (depending on
3521 @var{insert_breapoints_p}) breakpoints at each possible destinations of
3522 the next instruction. See @file{sparc-tdep.c} and @file{rs6000-tdep.c}
3525 @item SOFUN_ADDRESS_MAYBE_MISSING
3526 @findex SOFUN_ADDRESS_MAYBE_MISSING
3527 Somebody clever observed that, the more actual addresses you have in the
3528 debug information, the more time the linker has to spend relocating
3529 them. So whenever there's some other way the debugger could find the
3530 address it needs, you should omit it from the debug info, to make
3533 @code{SOFUN_ADDRESS_MAYBE_MISSING} indicates that a particular set of
3534 hacks of this sort are in use, affecting @code{N_SO} and @code{N_FUN}
3535 entries in stabs-format debugging information. @code{N_SO} stabs mark
3536 the beginning and ending addresses of compilation units in the text
3537 segment. @code{N_FUN} stabs mark the starts and ends of functions.
3539 @code{SOFUN_ADDRESS_MAYBE_MISSING} means two things:
3543 @code{N_FUN} stabs have an address of zero. Instead, you should find the
3544 addresses where the function starts by taking the function name from
3545 the stab, and then looking that up in the minsyms (the
3546 linker/assembler symbol table). In other words, the stab has the
3547 name, and the linker/assembler symbol table is the only place that carries
3551 @code{N_SO} stabs have an address of zero, too. You just look at the
3552 @code{N_FUN} stabs that appear before and after the @code{N_SO} stab,
3553 and guess the starting and ending addresses of the compilation unit from
3557 @item PC_LOAD_SEGMENT
3558 @findex PC_LOAD_SEGMENT
3559 If defined, print information about the load segment for the program
3560 counter. (Defined only for the RS/6000.)
3564 If the program counter is kept in a register, then define this macro to
3565 be the number (greater than or equal to zero) of that register.
3567 This should only need to be defined if @code{TARGET_READ_PC} and
3568 @code{TARGET_WRITE_PC} are not defined.
3571 @findex PARM_BOUNDARY
3572 If non-zero, round arguments to a boundary of this many bits before
3573 pushing them on the stack.
3575 @item stabs_argument_has_addr (@var{gdbarch}, @var{type})
3576 @findex stabs_argument_has_addr
3577 @findex DEPRECATED_REG_STRUCT_HAS_ADDR
3578 @anchor{stabs_argument_has_addr} Define this to return nonzero if a
3579 function argument of type @var{type} is passed by reference instead of
3582 This method replaces @code{DEPRECATED_REG_STRUCT_HAS_ADDR}
3583 (@pxref{DEPRECATED_REG_STRUCT_HAS_ADDR}).
3585 @item PROCESS_LINENUMBER_HOOK
3586 @findex PROCESS_LINENUMBER_HOOK
3587 A hook defined for XCOFF reading.
3589 @item PROLOGUE_FIRSTLINE_OVERLAP
3590 @findex PROLOGUE_FIRSTLINE_OVERLAP
3591 (Only used in unsupported Convex configuration.)
3595 If defined, this is the number of the processor status register. (This
3596 definition is only used in generic code when parsing "$ps".)
3598 @item DEPRECATED_POP_FRAME
3599 @findex DEPRECATED_POP_FRAME
3601 If defined, used by @code{frame_pop} to remove a stack frame. This
3602 method has been superseeded by generic code.
3604 @item push_dummy_call (@var{gdbarch}, @var{function}, @var{regcache}, @var{pc_addr}, @var{nargs}, @var{args}, @var{sp}, @var{struct_return}, @var{struct_addr})
3605 @findex push_dummy_call
3606 @findex DEPRECATED_PUSH_ARGUMENTS.
3607 @anchor{push_dummy_call} Define this to push the dummy frame's call to
3608 the inferior function onto the stack. In addition to pushing
3609 @var{nargs}, the code should push @var{struct_addr} (when
3610 @var{struct_return}), and the return address (@var{bp_addr}).
3612 @var{function} is a pointer to a @code{struct value}; on architectures that use
3613 function descriptors, this contains the function descriptor value.
3615 Returns the updated top-of-stack pointer.
3617 This method replaces @code{DEPRECATED_PUSH_ARGUMENTS}.
3619 @item CORE_ADDR push_dummy_code (@var{gdbarch}, @var{sp}, @var{funaddr}, @var{using_gcc}, @var{args}, @var{nargs}, @var{value_type}, @var{real_pc}, @var{bp_addr})
3620 @findex push_dummy_code
3621 @anchor{push_dummy_code} Given a stack based call dummy, push the
3622 instruction sequence (including space for a breakpoint) to which the
3623 called function should return.
3625 Set @var{bp_addr} to the address at which the breakpoint instruction
3626 should be inserted, @var{real_pc} to the resume address when starting
3627 the call sequence, and return the updated inner-most stack address.
3629 By default, the stack is grown sufficient to hold a frame-aligned
3630 (@pxref{frame_align}) breakpoint, @var{bp_addr} is set to the address
3631 reserved for that breakpoint, and @var{real_pc} set to @var{funaddr}.
3633 This method replaces @code{CALL_DUMMY_LOCATION},
3634 @code{DEPRECATED_REGISTER_SIZE}.
3636 @item REGISTER_NAME(@var{i})
3637 @findex REGISTER_NAME
3638 Return the name of register @var{i} as a string. May return @code{NULL}
3639 or @code{NUL} to indicate that register @var{i} is not valid.
3641 @item DEPRECATED_REG_STRUCT_HAS_ADDR (@var{gcc_p}, @var{type})
3642 @findex DEPRECATED_REG_STRUCT_HAS_ADDR
3643 @anchor{DEPRECATED_REG_STRUCT_HAS_ADDR}Define this to return 1 if the
3644 given type will be passed by pointer rather than directly.
3646 This method has been replaced by @code{stabs_argument_has_addr}
3647 (@pxref{stabs_argument_has_addr}).
3649 @item SAVE_DUMMY_FRAME_TOS (@var{sp})
3650 @findex SAVE_DUMMY_FRAME_TOS
3651 @anchor{SAVE_DUMMY_FRAME_TOS} Used in @samp{call_function_by_hand} to
3652 notify the target dependent code of the top-of-stack value that will be
3653 passed to the the inferior code. This is the value of the @code{SP}
3654 after both the dummy frame and space for parameters/results have been
3655 allocated on the stack. @xref{unwind_dummy_id}.
3657 @item SDB_REG_TO_REGNUM
3658 @findex SDB_REG_TO_REGNUM
3659 Define this to convert sdb register numbers into @value{GDBN} regnums. If not
3660 defined, no conversion will be done.
3662 @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})
3663 @findex gdbarch_return_value
3664 @anchor{gdbarch_return_value} Given a function with a return-value of
3665 type @var{rettype}, return which return-value convention that function
3668 @value{GDBN} currently recognizes two function return-value conventions:
3669 @code{RETURN_VALUE_REGISTER_CONVENTION} where the return value is found
3670 in registers; and @code{RETURN_VALUE_STRUCT_CONVENTION} where the return
3671 value is found in memory and the address of that memory location is
3672 passed in as the function's first parameter.
3674 If the register convention is being used, and @var{writebuf} is
3675 non-@code{NULL}, also copy the return-value in @var{writebuf} into
3678 If the register convention is being used, and @var{readbuf} is
3679 non-@code{NULL}, also copy the return value from @var{regcache} into
3680 @var{readbuf} (@var{regcache} contains a copy of the registers from the
3681 just returned function).
3683 @xref{DEPRECATED_EXTRACT_STRUCT_VALUE_ADDRESS}, for a description of how
3684 return-values that use the struct convention are handled.
3686 @emph{Maintainer note: This method replaces separate predicate, extract,
3687 store methods. By having only one method, the logic needed to determine
3688 the return-value convention need only be implemented in one place. If
3689 @value{GDBN} were written in an @sc{oo} language, this method would
3690 instead return an object that knew how to perform the register
3691 return-value extract and store.}
3693 @emph{Maintainer note: This method does not take a @var{gcc_p}
3694 parameter, and such a parameter should not be added. If an architecture
3695 that requires per-compiler or per-function information be identified,
3696 then the replacement of @var{rettype} with @code{struct value}
3697 @var{function} should be persued.}
3699 @emph{Maintainer note: The @var{regcache} parameter limits this methods
3700 to the inner most frame. While replacing @var{regcache} with a
3701 @code{struct frame_info} @var{frame} parameter would remove that
3702 limitation there has yet to be a demonstrated need for such a change.}
3704 @item SKIP_PERMANENT_BREAKPOINT
3705 @findex SKIP_PERMANENT_BREAKPOINT
3706 Advance the inferior's PC past a permanent breakpoint. @value{GDBN} normally
3707 steps over a breakpoint by removing it, stepping one instruction, and
3708 re-inserting the breakpoint. However, permanent breakpoints are
3709 hardwired into the inferior, and can't be removed, so this strategy
3710 doesn't work. Calling @code{SKIP_PERMANENT_BREAKPOINT} adjusts the processor's
3711 state so that execution will resume just after the breakpoint. This
3712 macro does the right thing even when the breakpoint is in the delay slot
3713 of a branch or jump.
3715 @item SKIP_PROLOGUE (@var{pc})
3716 @findex SKIP_PROLOGUE
3717 A C expression that returns the address of the ``real'' code beyond the
3718 function entry prologue found at @var{pc}.
3720 @item SKIP_TRAMPOLINE_CODE (@var{pc})
3721 @findex SKIP_TRAMPOLINE_CODE
3722 If the target machine has trampoline code that sits between callers and
3723 the functions being called, then define this macro to return a new PC
3724 that is at the start of the real function.
3728 If the stack-pointer is kept in a register, then define this macro to be
3729 the number (greater than or equal to zero) of that register, or -1 if
3730 there is no such register.
3732 @item STAB_REG_TO_REGNUM
3733 @findex STAB_REG_TO_REGNUM
3734 Define this to convert stab register numbers (as gotten from `r'
3735 declarations) into @value{GDBN} regnums. If not defined, no conversion will be
3738 @item DEPRECATED_STACK_ALIGN (@var{addr})
3739 @anchor{DEPRECATED_STACK_ALIGN}
3740 @findex DEPRECATED_STACK_ALIGN
3741 Define this to increase @var{addr} so that it meets the alignment
3742 requirements for the processor's stack.
3744 Unlike @ref{frame_align}, this function always adjusts @var{addr}
3747 By default, no stack alignment is performed.
3749 @item STEP_SKIPS_DELAY (@var{addr})
3750 @findex STEP_SKIPS_DELAY
3751 Define this to return true if the address is of an instruction with a
3752 delay slot. If a breakpoint has been placed in the instruction's delay
3753 slot, @value{GDBN} will single-step over that instruction before resuming
3754 normally. Currently only defined for the Mips.
3756 @item STORE_RETURN_VALUE (@var{type}, @var{regcache}, @var{valbuf})
3757 @findex STORE_RETURN_VALUE
3758 A C expression that writes the function return value, found in
3759 @var{valbuf}, into the @var{regcache}. @var{type} is the type of the
3760 value that is to be returned.
3762 This method has been deprecated in favour of @code{gdbarch_return_value}
3763 (@pxref{gdbarch_return_value}).
3765 @item SYMBOL_RELOADING_DEFAULT
3766 @findex SYMBOL_RELOADING_DEFAULT
3767 The default value of the ``symbol-reloading'' variable. (Never defined in
3770 @item TARGET_CHAR_BIT
3771 @findex TARGET_CHAR_BIT
3772 Number of bits in a char; defaults to 8.
3774 @item TARGET_CHAR_SIGNED
3775 @findex TARGET_CHAR_SIGNED
3776 Non-zero if @code{char} is normally signed on this architecture; zero if
3777 it should be unsigned.
3779 The ISO C standard requires the compiler to treat @code{char} as
3780 equivalent to either @code{signed char} or @code{unsigned char}; any
3781 character in the standard execution set is supposed to be positive.
3782 Most compilers treat @code{char} as signed, but @code{char} is unsigned
3783 on the IBM S/390, RS6000, and PowerPC targets.
3785 @item TARGET_COMPLEX_BIT
3786 @findex TARGET_COMPLEX_BIT
3787 Number of bits in a complex number; defaults to @code{2 * TARGET_FLOAT_BIT}.
3789 At present this macro is not used.
3791 @item TARGET_DOUBLE_BIT
3792 @findex TARGET_DOUBLE_BIT
3793 Number of bits in a double float; defaults to @code{8 * TARGET_CHAR_BIT}.
3795 @item TARGET_DOUBLE_COMPLEX_BIT
3796 @findex TARGET_DOUBLE_COMPLEX_BIT
3797 Number of bits in a double complex; defaults to @code{2 * TARGET_DOUBLE_BIT}.
3799 At present this macro is not used.
3801 @item TARGET_FLOAT_BIT
3802 @findex TARGET_FLOAT_BIT
3803 Number of bits in a float; defaults to @code{4 * TARGET_CHAR_BIT}.
3805 @item TARGET_INT_BIT
3806 @findex TARGET_INT_BIT
3807 Number of bits in an integer; defaults to @code{4 * TARGET_CHAR_BIT}.
3809 @item TARGET_LONG_BIT
3810 @findex TARGET_LONG_BIT
3811 Number of bits in a long integer; defaults to @code{4 * TARGET_CHAR_BIT}.
3813 @item TARGET_LONG_DOUBLE_BIT
3814 @findex TARGET_LONG_DOUBLE_BIT
3815 Number of bits in a long double float;
3816 defaults to @code{2 * TARGET_DOUBLE_BIT}.
3818 @item TARGET_LONG_LONG_BIT
3819 @findex TARGET_LONG_LONG_BIT
3820 Number of bits in a long long integer; defaults to @code{2 * TARGET_LONG_BIT}.
3822 @item TARGET_PTR_BIT
3823 @findex TARGET_PTR_BIT
3824 Number of bits in a pointer; defaults to @code{TARGET_INT_BIT}.
3826 @item TARGET_SHORT_BIT
3827 @findex TARGET_SHORT_BIT
3828 Number of bits in a short integer; defaults to @code{2 * TARGET_CHAR_BIT}.
3830 @item TARGET_READ_PC
3831 @findex TARGET_READ_PC
3832 @itemx TARGET_WRITE_PC (@var{val}, @var{pid})
3833 @findex TARGET_WRITE_PC
3834 @anchor{TARGET_WRITE_PC}
3835 @itemx TARGET_READ_SP
3836 @findex TARGET_READ_SP
3837 @itemx TARGET_READ_FP
3838 @findex TARGET_READ_FP
3843 @anchor{TARGET_READ_SP} These change the behavior of @code{read_pc},
3844 @code{write_pc}, and @code{read_sp}. For most targets, these may be
3845 left undefined. @value{GDBN} will call the read and write register
3846 functions with the relevant @code{_REGNUM} argument.
3848 These macros are useful when a target keeps one of these registers in a
3849 hard to get at place; for example, part in a segment register and part
3850 in an ordinary register.
3852 @xref{unwind_sp}, which replaces @code{TARGET_READ_SP}.
3854 @item TARGET_VIRTUAL_FRAME_POINTER(@var{pc}, @var{regp}, @var{offsetp})
3855 @findex TARGET_VIRTUAL_FRAME_POINTER
3856 Returns a @code{(register, offset)} pair representing the virtual frame
3857 pointer in use at the code address @var{pc}. If virtual frame pointers
3858 are not used, a default definition simply returns
3859 @code{DEPRECATED_FP_REGNUM}, with an offset of zero.
3861 @item TARGET_HAS_HARDWARE_WATCHPOINTS
3862 If non-zero, the target has support for hardware-assisted
3863 watchpoints. @xref{Algorithms, watchpoints}, for more details and
3864 other related macros.
3866 @item TARGET_PRINT_INSN (@var{addr}, @var{info})
3867 @findex TARGET_PRINT_INSN
3868 This is the function used by @value{GDBN} to print an assembly
3869 instruction. It prints the instruction at address @var{addr} in
3870 debugged memory and returns the length of the instruction, in bytes. If
3871 a target doesn't define its own printing routine, it defaults to an
3872 accessor function for the global pointer
3873 @code{deprecated_tm_print_insn}. This usually points to a function in
3874 the @code{opcodes} library (@pxref{Support Libraries, ,Opcodes}).
3875 @var{info} is a structure (of type @code{disassemble_info}) defined in
3876 @file{include/dis-asm.h} used to pass information to the instruction
3879 @item struct frame_id unwind_dummy_id (struct frame_info *@var{frame})
3880 @findex unwind_dummy_id
3881 @anchor{unwind_dummy_id} Given @var{frame} return a @code{struct
3882 frame_id} that uniquely identifies an inferior function call's dummy
3883 frame. The value returned must match the dummy frame stack value
3884 previously saved using @code{SAVE_DUMMY_FRAME_TOS}.
3885 @xref{SAVE_DUMMY_FRAME_TOS}.
3887 @item DEPRECATED_USE_STRUCT_CONVENTION (@var{gcc_p}, @var{type})
3888 @findex DEPRECATED_USE_STRUCT_CONVENTION
3889 If defined, this must be an expression that is nonzero if a value of the
3890 given @var{type} being returned from a function must have space
3891 allocated for it on the stack. @var{gcc_p} is true if the function
3892 being considered is known to have been compiled by GCC; this is helpful
3893 for systems where GCC is known to use different calling convention than
3896 This method has been deprecated in favour of @code{gdbarch_return_value}
3897 (@pxref{gdbarch_return_value}).
3899 @item VALUE_TO_REGISTER(@var{type}, @var{regnum}, @var{from}, @var{to})
3900 @findex VALUE_TO_REGISTER
3901 Convert a value of type @var{type} into the raw contents of register
3903 @xref{Target Architecture Definition, , Using Different Register and Memory Data Representations}.
3905 @item VARIABLES_INSIDE_BLOCK (@var{desc}, @var{gcc_p})
3906 @findex VARIABLES_INSIDE_BLOCK
3907 For dbx-style debugging information, if the compiler puts variable
3908 declarations inside LBRAC/RBRAC blocks, this should be defined to be
3909 nonzero. @var{desc} is the value of @code{n_desc} from the
3910 @code{N_RBRAC} symbol, and @var{gcc_p} is true if @value{GDBN} has noticed the
3911 presence of either the @code{GCC_COMPILED_SYMBOL} or the
3912 @code{GCC2_COMPILED_SYMBOL}. By default, this is 0.
3914 @item OS9K_VARIABLES_INSIDE_BLOCK (@var{desc}, @var{gcc_p})
3915 @findex OS9K_VARIABLES_INSIDE_BLOCK
3916 Similarly, for OS/9000. Defaults to 1.
3919 Motorola M68K target conditionals.
3923 Define this to be the 4-bit location of the breakpoint trap vector. If
3924 not defined, it will default to @code{0xf}.
3926 @item REMOTE_BPT_VECTOR
3927 Defaults to @code{1}.
3929 @item NAME_OF_MALLOC
3930 @findex NAME_OF_MALLOC
3931 A string containing the name of the function to call in order to
3932 allocate some memory in the inferior. The default value is "malloc".
3936 @section Adding a New Target
3938 @cindex adding a target
3939 The following files add a target to @value{GDBN}:
3943 @item gdb/config/@var{arch}/@var{ttt}.mt
3944 Contains a Makefile fragment specific to this target. Specifies what
3945 object files are needed for target @var{ttt}, by defining
3946 @samp{TDEPFILES=@dots{}} and @samp{TDEPLIBS=@dots{}}. Also specifies
3947 the header file which describes @var{ttt}, by defining @samp{TM_FILE=
3950 You can also define @samp{TM_CFLAGS}, @samp{TM_CLIBS}, @samp{TM_CDEPS},
3951 but these are now deprecated, replaced by autoconf, and may go away in
3952 future versions of @value{GDBN}.
3954 @item gdb/@var{ttt}-tdep.c
3955 Contains any miscellaneous code required for this target machine. On
3956 some machines it doesn't exist at all. Sometimes the macros in
3957 @file{tm-@var{ttt}.h} become very complicated, so they are implemented
3958 as functions here instead, and the macro is simply defined to call the
3959 function. This is vastly preferable, since it is easier to understand
3962 @item gdb/@var{arch}-tdep.c
3963 @itemx gdb/@var{arch}-tdep.h
3964 This often exists to describe the basic layout of the target machine's
3965 processor chip (registers, stack, etc.). If used, it is included by
3966 @file{@var{ttt}-tdep.h}. It can be shared among many targets that use
3969 @item gdb/config/@var{arch}/tm-@var{ttt}.h
3970 (@file{tm.h} is a link to this file, created by @code{configure}). Contains
3971 macro definitions about the target machine's registers, stack frame
3972 format and instructions.
3974 New targets do not need this file and should not create it.
3976 @item gdb/config/@var{arch}/tm-@var{arch}.h
3977 This often exists to describe the basic layout of the target machine's
3978 processor chip (registers, stack, etc.). If used, it is included by
3979 @file{tm-@var{ttt}.h}. It can be shared among many targets that use the
3982 New targets do not need this file and should not create it.
3986 If you are adding a new operating system for an existing CPU chip, add a
3987 @file{config/tm-@var{os}.h} file that describes the operating system
3988 facilities that are unusual (extra symbol table info; the breakpoint
3989 instruction needed; etc.). Then write a @file{@var{arch}/tm-@var{os}.h}
3990 that just @code{#include}s @file{tm-@var{arch}.h} and
3991 @file{config/tm-@var{os}.h}.
3994 @section Converting an existing Target Architecture to Multi-arch
3995 @cindex converting targets to multi-arch
3997 This section describes the current accepted best practice for converting
3998 an existing target architecture to the multi-arch framework.
4000 The process consists of generating, testing, posting and committing a
4001 sequence of patches. Each patch must contain a single change, for
4007 Directly convert a group of functions into macros (the conversion does
4008 not change the behavior of any of the functions).
4011 Replace a non-multi-arch with a multi-arch mechanism (e.g.,
4015 Enable multi-arch level one.
4018 Delete one or more files.
4023 There isn't a size limit on a patch, however, a developer is strongly
4024 encouraged to keep the patch size down.
4026 Since each patch is well defined, and since each change has been tested
4027 and shows no regressions, the patches are considered @emph{fairly}
4028 obvious. Such patches, when submitted by developers listed in the
4029 @file{MAINTAINERS} file, do not need approval. Occasional steps in the
4030 process may be more complicated and less clear. The developer is
4031 expected to use their judgment and is encouraged to seek advice as
4034 @subsection Preparation
4036 The first step is to establish control. Build (with @option{-Werror}
4037 enabled) and test the target so that there is a baseline against which
4038 the debugger can be compared.
4040 At no stage can the test results regress or @value{GDBN} stop compiling
4041 with @option{-Werror}.
4043 @subsection Add the multi-arch initialization code
4045 The objective of this step is to establish the basic multi-arch
4046 framework. It involves
4051 The addition of a @code{@var{arch}_gdbarch_init} function@footnote{The
4052 above is from the original example and uses K&R C. @value{GDBN}
4053 has since converted to ISO C but lets ignore that.} that creates
4056 static struct gdbarch *
4057 d10v_gdbarch_init (info, arches)
4058 struct gdbarch_info info;
4059 struct gdbarch_list *arches;
4061 struct gdbarch *gdbarch;
4062 /* there is only one d10v architecture */
4064 return arches->gdbarch;
4065 gdbarch = gdbarch_alloc (&info, NULL);
4073 A per-architecture dump function to print any architecture specific
4077 mips_dump_tdep (struct gdbarch *current_gdbarch,
4078 struct ui_file *file)
4080 @dots{} code to print architecture specific info @dots{}
4085 A change to @code{_initialize_@var{arch}_tdep} to register this new
4089 _initialize_mips_tdep (void)
4091 gdbarch_register (bfd_arch_mips, mips_gdbarch_init,
4096 Add the macro @code{GDB_MULTI_ARCH}, defined as 0 (zero), to the file@*
4097 @file{config/@var{arch}/tm-@var{arch}.h}.
4101 @subsection Update multi-arch incompatible mechanisms
4103 Some mechanisms do not work with multi-arch. They include:
4106 @item FRAME_FIND_SAVED_REGS
4107 Replaced with @code{DEPRECATED_FRAME_INIT_SAVED_REGS}
4111 At this stage you could also consider converting the macros into
4114 @subsection Prepare for multi-arch level to one
4116 Temporally set @code{GDB_MULTI_ARCH} to @code{GDB_MULTI_ARCH_PARTIAL}
4117 and then build and start @value{GDBN} (the change should not be
4118 committed). @value{GDBN} may not build, and once built, it may die with
4119 an internal error listing the architecture methods that must be
4122 Fix any build problems (patch(es)).
4124 Convert all the architecture methods listed, which are only macros, into
4125 functions (patch(es)).
4127 Update @code{@var{arch}_gdbarch_init} to set all the missing
4128 architecture methods and wrap the corresponding macros in @code{#if
4129 !GDB_MULTI_ARCH} (patch(es)).
4131 @subsection Set multi-arch level one
4133 Change the value of @code{GDB_MULTI_ARCH} to GDB_MULTI_ARCH_PARTIAL (a
4136 Any problems with throwing ``the switch'' should have been fixed
4139 @subsection Convert remaining macros
4141 Suggest converting macros into functions (and setting the corresponding
4142 architecture method) in small batches.
4144 @subsection Set multi-arch level to two
4146 This should go smoothly.
4148 @subsection Delete the TM file
4150 The @file{tm-@var{arch}.h} can be deleted. @file{@var{arch}.mt} and
4151 @file{configure.in} updated.
4154 @node Target Vector Definition
4156 @chapter Target Vector Definition
4157 @cindex target vector
4159 The target vector defines the interface between @value{GDBN}'s
4160 abstract handling of target systems, and the nitty-gritty code that
4161 actually exercises control over a process or a serial port.
4162 @value{GDBN} includes some 30-40 different target vectors; however,
4163 each configuration of @value{GDBN} includes only a few of them.
4165 @section File Targets
4167 Both executables and core files have target vectors.
4169 @section Standard Protocol and Remote Stubs
4171 @value{GDBN}'s file @file{remote.c} talks a serial protocol to code
4172 that runs in the target system. @value{GDBN} provides several sample
4173 @dfn{stubs} that can be integrated into target programs or operating
4174 systems for this purpose; they are named @file{*-stub.c}.
4176 The @value{GDBN} user's manual describes how to put such a stub into
4177 your target code. What follows is a discussion of integrating the
4178 SPARC stub into a complicated operating system (rather than a simple
4179 program), by Stu Grossman, the author of this stub.
4181 The trap handling code in the stub assumes the following upon entry to
4186 %l1 and %l2 contain pc and npc respectively at the time of the trap;
4192 you are in the correct trap window.
4195 As long as your trap handler can guarantee those conditions, then there
4196 is no reason why you shouldn't be able to ``share'' traps with the stub.
4197 The stub has no requirement that it be jumped to directly from the
4198 hardware trap vector. That is why it calls @code{exceptionHandler()},
4199 which is provided by the external environment. For instance, this could
4200 set up the hardware traps to actually execute code which calls the stub
4201 first, and then transfers to its own trap handler.
4203 For the most point, there probably won't be much of an issue with
4204 ``sharing'' traps, as the traps we use are usually not used by the kernel,
4205 and often indicate unrecoverable error conditions. Anyway, this is all
4206 controlled by a table, and is trivial to modify. The most important
4207 trap for us is for @code{ta 1}. Without that, we can't single step or
4208 do breakpoints. Everything else is unnecessary for the proper operation
4209 of the debugger/stub.
4211 From reading the stub, it's probably not obvious how breakpoints work.
4212 They are simply done by deposit/examine operations from @value{GDBN}.
4214 @section ROM Monitor Interface
4216 @section Custom Protocols
4218 @section Transport Layer
4220 @section Builtin Simulator
4223 @node Native Debugging
4225 @chapter Native Debugging
4226 @cindex native debugging
4228 Several files control @value{GDBN}'s configuration for native support:
4232 @item gdb/config/@var{arch}/@var{xyz}.mh
4233 Specifies Makefile fragments needed by a @emph{native} configuration on
4234 machine @var{xyz}. In particular, this lists the required
4235 native-dependent object files, by defining @samp{NATDEPFILES=@dots{}}.
4236 Also specifies the header file which describes native support on
4237 @var{xyz}, by defining @samp{NAT_FILE= nm-@var{xyz}.h}. You can also
4238 define @samp{NAT_CFLAGS}, @samp{NAT_ADD_FILES}, @samp{NAT_CLIBS},
4239 @samp{NAT_CDEPS}, etc.; see @file{Makefile.in}.
4241 @emph{Maintainer's note: The @file{.mh} suffix is because this file
4242 originally contained @file{Makefile} fragments for hosting @value{GDBN}
4243 on machine @var{xyz}. While the file is no longer used for this
4244 purpose, the @file{.mh} suffix remains. Perhaps someone will
4245 eventually rename these fragments so that they have a @file{.mn}
4248 @item gdb/config/@var{arch}/nm-@var{xyz}.h
4249 (@file{nm.h} is a link to this file, created by @code{configure}). Contains C
4250 macro definitions describing the native system environment, such as
4251 child process control and core file support.
4253 @item gdb/@var{xyz}-nat.c
4254 Contains any miscellaneous C code required for this native support of
4255 this machine. On some machines it doesn't exist at all.
4258 There are some ``generic'' versions of routines that can be used by
4259 various systems. These can be customized in various ways by macros
4260 defined in your @file{nm-@var{xyz}.h} file. If these routines work for
4261 the @var{xyz} host, you can just include the generic file's name (with
4262 @samp{.o}, not @samp{.c}) in @code{NATDEPFILES}.
4264 Otherwise, if your machine needs custom support routines, you will need
4265 to write routines that perform the same functions as the generic file.
4266 Put them into @file{@var{xyz}-nat.c}, and put @file{@var{xyz}-nat.o}
4267 into @code{NATDEPFILES}.
4271 This contains the @emph{target_ops vector} that supports Unix child
4272 processes on systems which use ptrace and wait to control the child.
4275 This contains the @emph{target_ops vector} that supports Unix child
4276 processes on systems which use /proc to control the child.
4279 This does the low-level grunge that uses Unix system calls to do a ``fork
4280 and exec'' to start up a child process.
4283 This is the low level interface to inferior processes for systems using
4284 the Unix @code{ptrace} call in a vanilla way.
4287 @section Native core file Support
4288 @cindex native core files
4291 @findex fetch_core_registers
4292 @item core-aout.c::fetch_core_registers()
4293 Support for reading registers out of a core file. This routine calls
4294 @code{register_addr()}, see below. Now that BFD is used to read core
4295 files, virtually all machines should use @code{core-aout.c}, and should
4296 just provide @code{fetch_core_registers} in @code{@var{xyz}-nat.c} (or
4297 @code{REGISTER_U_ADDR} in @code{nm-@var{xyz}.h}).
4299 @item core-aout.c::register_addr()
4300 If your @code{nm-@var{xyz}.h} file defines the macro
4301 @code{REGISTER_U_ADDR(addr, blockend, regno)}, it should be defined to
4302 set @code{addr} to the offset within the @samp{user} struct of @value{GDBN}
4303 register number @code{regno}. @code{blockend} is the offset within the
4304 ``upage'' of @code{u.u_ar0}. If @code{REGISTER_U_ADDR} is defined,
4305 @file{core-aout.c} will define the @code{register_addr()} function and
4306 use the macro in it. If you do not define @code{REGISTER_U_ADDR}, but
4307 you are using the standard @code{fetch_core_registers()}, you will need
4308 to define your own version of @code{register_addr()}, put it into your
4309 @code{@var{xyz}-nat.c} file, and be sure @code{@var{xyz}-nat.o} is in
4310 the @code{NATDEPFILES} list. If you have your own
4311 @code{fetch_core_registers()}, you may not need a separate
4312 @code{register_addr()}. Many custom @code{fetch_core_registers()}
4313 implementations simply locate the registers themselves.@refill
4316 When making @value{GDBN} run native on a new operating system, to make it
4317 possible to debug core files, you will need to either write specific
4318 code for parsing your OS's core files, or customize
4319 @file{bfd/trad-core.c}. First, use whatever @code{#include} files your
4320 machine uses to define the struct of registers that is accessible
4321 (possibly in the u-area) in a core file (rather than
4322 @file{machine/reg.h}), and an include file that defines whatever header
4323 exists on a core file (e.g. the u-area or a @code{struct core}). Then
4324 modify @code{trad_unix_core_file_p} to use these values to set up the
4325 section information for the data segment, stack segment, any other
4326 segments in the core file (perhaps shared library contents or control
4327 information), ``registers'' segment, and if there are two discontiguous
4328 sets of registers (e.g. integer and float), the ``reg2'' segment. This
4329 section information basically delimits areas in the core file in a
4330 standard way, which the section-reading routines in BFD know how to seek
4333 Then back in @value{GDBN}, you need a matching routine called
4334 @code{fetch_core_registers}. If you can use the generic one, it's in
4335 @file{core-aout.c}; if not, it's in your @file{@var{xyz}-nat.c} file.
4336 It will be passed a char pointer to the entire ``registers'' segment,
4337 its length, and a zero; or a char pointer to the entire ``regs2''
4338 segment, its length, and a 2. The routine should suck out the supplied
4339 register values and install them into @value{GDBN}'s ``registers'' array.
4341 If your system uses @file{/proc} to control processes, and uses ELF
4342 format core files, then you may be able to use the same routines for
4343 reading the registers out of processes and out of core files.
4351 @section shared libraries
4353 @section Native Conditionals
4354 @cindex native conditionals
4356 When @value{GDBN} is configured and compiled, various macros are
4357 defined or left undefined, to control compilation when the host and
4358 target systems are the same. These macros should be defined (or left
4359 undefined) in @file{nm-@var{system}.h}.
4363 @item CHILD_PREPARE_TO_STORE
4364 @findex CHILD_PREPARE_TO_STORE
4365 If the machine stores all registers at once in the child process, then
4366 define this to ensure that all values are correct. This usually entails
4367 a read from the child.
4369 [Note that this is incorrectly defined in @file{xm-@var{system}.h} files
4372 @item FETCH_INFERIOR_REGISTERS
4373 @findex FETCH_INFERIOR_REGISTERS
4374 Define this if the native-dependent code will provide its own routines
4375 @code{fetch_inferior_registers} and @code{store_inferior_registers} in
4376 @file{@var{host}-nat.c}. If this symbol is @emph{not} defined, and
4377 @file{infptrace.c} is included in this configuration, the default
4378 routines in @file{infptrace.c} are used for these functions.
4382 This macro is normally defined to be the number of the first floating
4383 point register, if the machine has such registers. As such, it would
4384 appear only in target-specific code. However, @file{/proc} support uses this
4385 to decide whether floats are in use on this target.
4387 @item GET_LONGJMP_TARGET
4388 @findex GET_LONGJMP_TARGET
4389 For most machines, this is a target-dependent parameter. On the
4390 DECstation and the Iris, this is a native-dependent parameter, since
4391 @file{setjmp.h} is needed to define it.
4393 This macro determines the target PC address that @code{longjmp} will jump to,
4394 assuming that we have just stopped at a longjmp breakpoint. It takes a
4395 @code{CORE_ADDR *} as argument, and stores the target PC value through this
4396 pointer. It examines the current state of the machine as needed.
4398 @item I386_USE_GENERIC_WATCHPOINTS
4399 An x86-based machine can define this to use the generic x86 watchpoint
4400 support; see @ref{Algorithms, I386_USE_GENERIC_WATCHPOINTS}.
4403 @findex KERNEL_U_ADDR
4404 Define this to the address of the @code{u} structure (the ``user
4405 struct'', also known as the ``u-page'') in kernel virtual memory. @value{GDBN}
4406 needs to know this so that it can subtract this address from absolute
4407 addresses in the upage, that are obtained via ptrace or from core files.
4408 On systems that don't need this value, set it to zero.
4410 @item KERNEL_U_ADDR_HPUX
4411 @findex KERNEL_U_ADDR_HPUX
4412 Define this to cause @value{GDBN} to determine the address of @code{u} at
4413 runtime, by using HP-style @code{nlist} on the kernel's image in the
4416 @item ONE_PROCESS_WRITETEXT
4417 @findex ONE_PROCESS_WRITETEXT
4418 Define this to be able to, when a breakpoint insertion fails, warn the
4419 user that another process may be running with the same executable.
4422 @findex PROC_NAME_FMT
4423 Defines the format for the name of a @file{/proc} device. Should be
4424 defined in @file{nm.h} @emph{only} in order to override the default
4425 definition in @file{procfs.c}.
4427 @item PTRACE_ARG3_TYPE
4428 @findex PTRACE_ARG3_TYPE
4429 The type of the third argument to the @code{ptrace} system call, if it
4430 exists and is different from @code{int}.
4432 @item REGISTER_U_ADDR
4433 @findex REGISTER_U_ADDR
4434 Defines the offset of the registers in the ``u area''.
4436 @item SHELL_COMMAND_CONCAT
4437 @findex SHELL_COMMAND_CONCAT
4438 If defined, is a string to prefix on the shell command used to start the
4443 If defined, this is the name of the shell to use to run the inferior.
4444 Defaults to @code{"/bin/sh"}.
4446 @item SOLIB_ADD (@var{filename}, @var{from_tty}, @var{targ}, @var{readsyms})
4448 Define this to expand into an expression that will cause the symbols in
4449 @var{filename} to be added to @value{GDBN}'s symbol table. If
4450 @var{readsyms} is zero symbols are not read but any necessary low level
4451 processing for @var{filename} is still done.
4453 @item SOLIB_CREATE_INFERIOR_HOOK
4454 @findex SOLIB_CREATE_INFERIOR_HOOK
4455 Define this to expand into any shared-library-relocation code that you
4456 want to be run just after the child process has been forked.
4458 @item START_INFERIOR_TRAPS_EXPECTED
4459 @findex START_INFERIOR_TRAPS_EXPECTED
4460 When starting an inferior, @value{GDBN} normally expects to trap
4462 the shell execs, and once when the program itself execs. If the actual
4463 number of traps is something other than 2, then define this macro to
4464 expand into the number expected.
4468 This determines whether small routines in @file{*-tdep.c}, which
4469 translate register values between @value{GDBN}'s internal
4470 representation and the @file{/proc} representation, are compiled.
4473 @findex U_REGS_OFFSET
4474 This is the offset of the registers in the upage. It need only be
4475 defined if the generic ptrace register access routines in
4476 @file{infptrace.c} are being used (that is, @file{infptrace.c} is
4477 configured in, and @code{FETCH_INFERIOR_REGISTERS} is not defined). If
4478 the default value from @file{infptrace.c} is good enough, leave it
4481 The default value means that u.u_ar0 @emph{points to} the location of
4482 the registers. I'm guessing that @code{#define U_REGS_OFFSET 0} means
4483 that @code{u.u_ar0} @emph{is} the location of the registers.
4487 See @file{objfiles.c}.
4490 @findex DEBUG_PTRACE
4491 Define this to debug @code{ptrace} calls.
4495 @node Support Libraries
4497 @chapter Support Libraries
4502 BFD provides support for @value{GDBN} in several ways:
4505 @item identifying executable and core files
4506 BFD will identify a variety of file types, including a.out, coff, and
4507 several variants thereof, as well as several kinds of core files.
4509 @item access to sections of files
4510 BFD parses the file headers to determine the names, virtual addresses,
4511 sizes, and file locations of all the various named sections in files
4512 (such as the text section or the data section). @value{GDBN} simply
4513 calls BFD to read or write section @var{x} at byte offset @var{y} for
4516 @item specialized core file support
4517 BFD provides routines to determine the failing command name stored in a
4518 core file, the signal with which the program failed, and whether a core
4519 file matches (i.e.@: could be a core dump of) a particular executable
4522 @item locating the symbol information
4523 @value{GDBN} uses an internal interface of BFD to determine where to find the
4524 symbol information in an executable file or symbol-file. @value{GDBN} itself
4525 handles the reading of symbols, since BFD does not ``understand'' debug
4526 symbols, but @value{GDBN} uses BFD's cached information to find the symbols,
4531 @cindex opcodes library
4533 The opcodes library provides @value{GDBN}'s disassembler. (It's a separate
4534 library because it's also used in binutils, for @file{objdump}).
4541 @cindex @code{libiberty} library
4543 The @code{libiberty} library provides a set of functions and features
4544 that integrate and improve on functionality found in modern operating
4545 systems. Broadly speaking, such features can be divided into three
4546 groups: supplemental functions (functions that may be missing in some
4547 environments and operating systems), replacement functions (providing
4548 a uniform and easier to use interface for commonly used standard
4549 functions), and extensions (which provide additional functionality
4550 beyond standard functions).
4552 @value{GDBN} uses various features provided by the @code{libiberty}
4553 library, for instance the C@t{++} demangler, the @acronym{IEEE}
4554 floating format support functions, the input options parser
4555 @samp{getopt}, the @samp{obstack} extension, and other functions.
4557 @subsection @code{obstacks} in @value{GDBN}
4558 @cindex @code{obstacks}
4560 The obstack mechanism provides a convenient way to allocate and free
4561 chunks of memory. Each obstack is a pool of memory that is managed
4562 like a stack. Objects (of any nature, size and alignment) are
4563 allocated and freed in a @acronym{LIFO} fashion on an obstack (see
4564 @code{libiberty}'s documenatation for a more detailed explanation of
4567 The most noticeable use of the @code{obstacks} in @value{GDBN} is in
4568 object files. There is an obstack associated with each internal
4569 representation of an object file. Lots of things get allocated on
4570 these @code{obstacks}: dictionary entries, blocks, blockvectors,
4571 symbols, minimal symbols, types, vectors of fundamental types, class
4572 fields of types, object files section lists, object files section
4573 offets lists, line tables, symbol tables, partial symbol tables,
4574 string tables, symbol table private data, macros tables, debug
4575 information sections and entries, import and export lists (som),
4576 unwind information (hppa), dwarf2 location expressions data. Plus
4577 various strings such as directory names strings, debug format strings,
4580 An essential and convenient property of all data on @code{obstacks} is
4581 that memory for it gets allocated (with @code{obstack_alloc}) at
4582 various times during a debugging sesssion, but it is released all at
4583 once using the @code{obstack_free} function. The @code{obstack_free}
4584 function takes a pointer to where in the stack it must start the
4585 deletion from (much like the cleanup chains have a pointer to where to
4586 start the cleanups). Because of the stack like structure of the
4587 @code{obstacks}, this allows to free only a top portion of the
4588 obstack. There are a few instances in @value{GDBN} where such thing
4589 happens. Calls to @code{obstack_free} are done after some local data
4590 is allocated to the obstack. Only the local data is deleted from the
4591 obstack. Of course this assumes that nothing between the
4592 @code{obstack_alloc} and the @code{obstack_free} allocates anything
4593 else on the same obstack. For this reason it is best and safest to
4594 use temporary @code{obstacks}.
4596 Releasing the whole obstack is also not safe per se. It is safe only
4597 under the condition that we know the @code{obstacks} memory is no
4598 longer needed. In @value{GDBN} we get rid of the @code{obstacks} only
4599 when we get rid of the whole objfile(s), for instance upon reading a
4603 @cindex regular expressions library
4614 @item SIGN_EXTEND_CHAR
4616 @item SWITCH_ENUM_BUG
4631 This chapter covers topics that are lower-level than the major
4632 algorithms of @value{GDBN}.
4637 Cleanups are a structured way to deal with things that need to be done
4640 When your code does something (e.g., @code{xmalloc} some memory, or
4641 @code{open} a file) that needs to be undone later (e.g., @code{xfree}
4642 the memory or @code{close} the file), it can make a cleanup. The
4643 cleanup will be done at some future point: when the command is finished
4644 and control returns to the top level; when an error occurs and the stack
4645 is unwound; or when your code decides it's time to explicitly perform
4646 cleanups. Alternatively you can elect to discard the cleanups you
4652 @item struct cleanup *@var{old_chain};
4653 Declare a variable which will hold a cleanup chain handle.
4655 @findex make_cleanup
4656 @item @var{old_chain} = make_cleanup (@var{function}, @var{arg});
4657 Make a cleanup which will cause @var{function} to be called with
4658 @var{arg} (a @code{char *}) later. The result, @var{old_chain}, is a
4659 handle that can later be passed to @code{do_cleanups} or
4660 @code{discard_cleanups}. Unless you are going to call
4661 @code{do_cleanups} or @code{discard_cleanups}, you can ignore the result
4662 from @code{make_cleanup}.
4665 @item do_cleanups (@var{old_chain});
4666 Do all cleanups added to the chain since the corresponding
4667 @code{make_cleanup} call was made.
4669 @findex discard_cleanups
4670 @item discard_cleanups (@var{old_chain});
4671 Same as @code{do_cleanups} except that it just removes the cleanups from
4672 the chain and does not call the specified functions.
4675 Cleanups are implemented as a chain. The handle returned by
4676 @code{make_cleanups} includes the cleanup passed to the call and any
4677 later cleanups appended to the chain (but not yet discarded or
4681 make_cleanup (a, 0);
4683 struct cleanup *old = make_cleanup (b, 0);
4691 will call @code{c()} and @code{b()} but will not call @code{a()}. The
4692 cleanup that calls @code{a()} will remain in the cleanup chain, and will
4693 be done later unless otherwise discarded.@refill
4695 Your function should explicitly do or discard the cleanups it creates.
4696 Failing to do this leads to non-deterministic behavior since the caller
4697 will arbitrarily do or discard your functions cleanups. This need leads
4698 to two common cleanup styles.
4700 The first style is try/finally. Before it exits, your code-block calls
4701 @code{do_cleanups} with the old cleanup chain and thus ensures that your
4702 code-block's cleanups are always performed. For instance, the following
4703 code-segment avoids a memory leak problem (even when @code{error} is
4704 called and a forced stack unwind occurs) by ensuring that the
4705 @code{xfree} will always be called:
4708 struct cleanup *old = make_cleanup (null_cleanup, 0);
4709 data = xmalloc (sizeof blah);
4710 make_cleanup (xfree, data);
4715 The second style is try/except. Before it exits, your code-block calls
4716 @code{discard_cleanups} with the old cleanup chain and thus ensures that
4717 any created cleanups are not performed. For instance, the following
4718 code segment, ensures that the file will be closed but only if there is
4722 FILE *file = fopen ("afile", "r");
4723 struct cleanup *old = make_cleanup (close_file, file);
4725 discard_cleanups (old);
4729 Some functions, e.g. @code{fputs_filtered()} or @code{error()}, specify
4730 that they ``should not be called when cleanups are not in place''. This
4731 means that any actions you need to reverse in the case of an error or
4732 interruption must be on the cleanup chain before you call these
4733 functions, since they might never return to your code (they
4734 @samp{longjmp} instead).
4736 @section Per-architecture module data
4737 @cindex per-architecture module data
4738 @cindex multi-arch data
4739 @cindex data-pointer, per-architecture/per-module
4741 The multi-arch framework includes a mechanism for adding module
4742 specific per-architecture data-pointers to the @code{struct gdbarch}
4743 architecture object.
4745 A module registers one or more per-architecture data-pointers using:
4747 @deftypefun struct gdbarch_data *gdbarch_data_register_pre_init (gdbarch_data_pre_init_ftype *@var{pre_init})
4748 @var{pre_init} is used to, on-demand, allocate an initial value for a
4749 per-architecture data-pointer using the architecture's obstack (passed
4750 in as a parameter). Since @var{pre_init} can be called during
4751 architecture creation, it is not parameterized with the architecture.
4752 and must not call modules that use per-architecture data.
4755 @deftypefun struct gdbarch_data *gdbarch_data_register_post_init (gdbarch_data_post_init_ftype *@var{post_init})
4756 @var{post_init} is used to obtain an initial value for a
4757 per-architecture data-pointer @emph{after}. Since @var{post_init} is
4758 always called after architecture creation, it both receives the fully
4759 initialized architecture and is free to call modules that use
4760 per-architecture data (care needs to be taken to ensure that those
4761 other modules do not try to call back to this module as that will
4762 create in cycles in the initialization call graph).
4765 These functions return a @code{struct gdbarch_data} that is used to
4766 identify the per-architecture data-pointer added for that module.
4768 The per-architecture data-pointer is accessed using the function:
4770 @deftypefun void *gdbarch_data (struct gdbarch *@var{gdbarch}, struct gdbarch_data *@var{data_handle})
4771 Given the architecture @var{arch} and module data handle
4772 @var{data_handle} (returned by @code{gdbarch_data_register_pre_init}
4773 or @code{gdbarch_data_register_post_init}), this function returns the
4774 current value of the per-architecture data-pointer. If the data
4775 pointer is @code{NULL}, it is first initialized by calling the
4776 corresponding @var{pre_init} or @var{post_init} method.
4779 The examples below assume the following definitions:
4782 struct nozel @{ int total; @};
4783 static struct gdbarch_data *nozel_handle;
4786 A module can extend the architecture vector, adding additional
4787 per-architecture data, using the @var{pre_init} method. The module's
4788 per-architecture data is then initialized during architecture
4791 In the below, the module's per-architecture @emph{nozel} is added. An
4792 architecture can specify its nozel by calling @code{set_gdbarch_nozel}
4793 from @code{gdbarch_init}.
4797 nozel_pre_init (struct obstack *obstack)
4799 struct nozel *data = OBSTACK_ZALLOC (obstack, struct nozel);
4806 set_gdbarch_nozel (struct gdbarch *gdbarch, int total)
4808 struct nozel *data = gdbarch_data (gdbarch, nozel_handle);
4809 data->total = nozel;
4813 A module can on-demand create architecture dependant data structures
4814 using @code{post_init}.
4816 In the below, the nozel's total is computed on-demand by
4817 @code{nozel_post_init} using information obtained from the
4822 nozel_post_init (struct gdbarch *gdbarch)
4824 struct nozel *data = GDBARCH_OBSTACK_ZALLOC (gdbarch, struct nozel);
4825 nozel->total = gdbarch@dots{} (gdbarch);
4832 nozel_total (struct gdbarch *gdbarch)
4834 struct nozel *data = gdbarch_data (gdbarch, nozel_handle);
4839 @section Wrapping Output Lines
4840 @cindex line wrap in output
4843 Output that goes through @code{printf_filtered} or @code{fputs_filtered}
4844 or @code{fputs_demangled} needs only to have calls to @code{wrap_here}
4845 added in places that would be good breaking points. The utility
4846 routines will take care of actually wrapping if the line width is
4849 The argument to @code{wrap_here} is an indentation string which is
4850 printed @emph{only} if the line breaks there. This argument is saved
4851 away and used later. It must remain valid until the next call to
4852 @code{wrap_here} or until a newline has been printed through the
4853 @code{*_filtered} functions. Don't pass in a local variable and then
4856 It is usually best to call @code{wrap_here} after printing a comma or
4857 space. If you call it before printing a space, make sure that your
4858 indentation properly accounts for the leading space that will print if
4859 the line wraps there.
4861 Any function or set of functions that produce filtered output must
4862 finish by printing a newline, to flush the wrap buffer, before switching
4863 to unfiltered (@code{printf}) output. Symbol reading routines that
4864 print warnings are a good example.
4866 @section @value{GDBN} Coding Standards
4867 @cindex coding standards
4869 @value{GDBN} follows the GNU coding standards, as described in
4870 @file{etc/standards.texi}. This file is also available for anonymous
4871 FTP from GNU archive sites. @value{GDBN} takes a strict interpretation
4872 of the standard; in general, when the GNU standard recommends a practice
4873 but does not require it, @value{GDBN} requires it.
4875 @value{GDBN} follows an additional set of coding standards specific to
4876 @value{GDBN}, as described in the following sections.
4881 @value{GDBN} assumes an ISO/IEC 9899:1990 (a.k.a.@: ISO C90) compliant
4884 @value{GDBN} does not assume an ISO C or POSIX compliant C library.
4887 @subsection Memory Management
4889 @value{GDBN} does not use the functions @code{malloc}, @code{realloc},
4890 @code{calloc}, @code{free} and @code{asprintf}.
4892 @value{GDBN} uses the functions @code{xmalloc}, @code{xrealloc} and
4893 @code{xcalloc} when allocating memory. Unlike @code{malloc} et.al.@:
4894 these functions do not return when the memory pool is empty. Instead,
4895 they unwind the stack using cleanups. These functions return
4896 @code{NULL} when requested to allocate a chunk of memory of size zero.
4898 @emph{Pragmatics: By using these functions, the need to check every
4899 memory allocation is removed. These functions provide portable
4902 @value{GDBN} does not use the function @code{free}.
4904 @value{GDBN} uses the function @code{xfree} to return memory to the
4905 memory pool. Consistent with ISO-C, this function ignores a request to
4906 free a @code{NULL} pointer.
4908 @emph{Pragmatics: On some systems @code{free} fails when passed a
4909 @code{NULL} pointer.}
4911 @value{GDBN} can use the non-portable function @code{alloca} for the
4912 allocation of small temporary values (such as strings).
4914 @emph{Pragmatics: This function is very non-portable. Some systems
4915 restrict the memory being allocated to no more than a few kilobytes.}
4917 @value{GDBN} uses the string function @code{xstrdup} and the print
4918 function @code{xstrprintf}.
4920 @emph{Pragmatics: @code{asprintf} and @code{strdup} can fail. Print
4921 functions such as @code{sprintf} are very prone to buffer overflow
4925 @subsection Compiler Warnings
4926 @cindex compiler warnings
4928 With few exceptions, developers should include the configuration option
4929 @samp{--enable-gdb-build-warnings=,-Werror} when building @value{GDBN}.
4930 The exceptions are listed in the file @file{gdb/MAINTAINERS}.
4932 This option causes @value{GDBN} (when built using GCC) to be compiled
4933 with a carefully selected list of compiler warning flags. Any warnings
4934 from those flags being treated as errors.
4936 The current list of warning flags includes:
4940 Since @value{GDBN} coding standard requires all functions to be declared
4941 using a prototype, the flag has the side effect of ensuring that
4942 prototyped functions are always visible with out resorting to
4943 @samp{-Wstrict-prototypes}.
4946 Such code often appears to work except on instruction set architectures
4947 that use register windows.
4954 @itemx -Wformat-nonliteral
4955 Since @value{GDBN} uses the @code{format printf} attribute on all
4956 @code{printf} like functions these check not just @code{printf} calls
4957 but also calls to functions such as @code{fprintf_unfiltered}.
4960 This warning includes uses of the assignment operator within an
4961 @code{if} statement.
4963 @item -Wpointer-arith
4965 @item -Wuninitialized
4967 @item -Wunused-label
4968 This warning has the additional benefit of detecting the absence of the
4969 @code{case} reserved word in a switch statement:
4971 enum @{ FD_SCHEDULED, NOTHING_SCHEDULED @} sched;
4984 @item -Wunused-function
4987 @emph{Pragmatics: Due to the way that @value{GDBN} is implemented most
4988 functions have unused parameters. Consequently the warning
4989 @samp{-Wunused-parameter} is precluded from the list. The macro
4990 @code{ATTRIBUTE_UNUSED} is not used as it leads to false negatives ---
4991 it is not an error to have @code{ATTRIBUTE_UNUSED} on a parameter that
4992 is being used. The options @samp{-Wall} and @samp{-Wunused} are also
4993 precluded because they both include @samp{-Wunused-parameter}.}
4995 @emph{Pragmatics: @value{GDBN} has not simply accepted the warnings
4996 enabled by @samp{-Wall -Werror -W...}. Instead it is selecting warnings
4997 when and where their benefits can be demonstrated.}
4999 @subsection Formatting
5001 @cindex source code formatting
5002 The standard GNU recommendations for formatting must be followed
5005 A function declaration should not have its name in column zero. A
5006 function definition should have its name in column zero.
5010 static void foo (void);
5018 @emph{Pragmatics: This simplifies scripting. Function definitions can
5019 be found using @samp{^function-name}.}
5021 There must be a space between a function or macro name and the opening
5022 parenthesis of its argument list (except for macro definitions, as
5023 required by C). There must not be a space after an open paren/bracket
5024 or before a close paren/bracket.
5026 While additional whitespace is generally helpful for reading, do not use
5027 more than one blank line to separate blocks, and avoid adding whitespace
5028 after the end of a program line (as of 1/99, some 600 lines had
5029 whitespace after the semicolon). Excess whitespace causes difficulties
5030 for @code{diff} and @code{patch} utilities.
5032 Pointers are declared using the traditional K&R C style:
5046 @subsection Comments
5048 @cindex comment formatting
5049 The standard GNU requirements on comments must be followed strictly.
5051 Block comments must appear in the following form, with no @code{/*}- or
5052 @code{*/}-only lines, and no leading @code{*}:
5055 /* Wait for control to return from inferior to debugger. If inferior
5056 gets a signal, we may decide to start it up again instead of
5057 returning. That is why there is a loop in this function. When
5058 this function actually returns it means the inferior should be left
5059 stopped and @value{GDBN} should read more commands. */
5062 (Note that this format is encouraged by Emacs; tabbing for a multi-line
5063 comment works correctly, and @kbd{M-q} fills the block consistently.)
5065 Put a blank line between the block comments preceding function or
5066 variable definitions, and the definition itself.
5068 In general, put function-body comments on lines by themselves, rather
5069 than trying to fit them into the 20 characters left at the end of a
5070 line, since either the comment or the code will inevitably get longer
5071 than will fit, and then somebody will have to move it anyhow.
5075 @cindex C data types
5076 Code must not depend on the sizes of C data types, the format of the
5077 host's floating point numbers, the alignment of anything, or the order
5078 of evaluation of expressions.
5080 @cindex function usage
5081 Use functions freely. There are only a handful of compute-bound areas
5082 in @value{GDBN} that might be affected by the overhead of a function
5083 call, mainly in symbol reading. Most of @value{GDBN}'s performance is
5084 limited by the target interface (whether serial line or system call).
5086 However, use functions with moderation. A thousand one-line functions
5087 are just as hard to understand as a single thousand-line function.
5089 @emph{Macros are bad, M'kay.}
5090 (But if you have to use a macro, make sure that the macro arguments are
5091 protected with parentheses.)
5095 Declarations like @samp{struct foo *} should be used in preference to
5096 declarations like @samp{typedef struct foo @{ @dots{} @} *foo_ptr}.
5099 @subsection Function Prototypes
5100 @cindex function prototypes
5102 Prototypes must be used when both @emph{declaring} and @emph{defining}
5103 a function. Prototypes for @value{GDBN} functions must include both the
5104 argument type and name, with the name matching that used in the actual
5105 function definition.
5107 All external functions should have a declaration in a header file that
5108 callers include, except for @code{_initialize_*} functions, which must
5109 be external so that @file{init.c} construction works, but shouldn't be
5110 visible to random source files.
5112 Where a source file needs a forward declaration of a static function,
5113 that declaration must appear in a block near the top of the source file.
5116 @subsection Internal Error Recovery
5118 During its execution, @value{GDBN} can encounter two types of errors.
5119 User errors and internal errors. User errors include not only a user
5120 entering an incorrect command but also problems arising from corrupt
5121 object files and system errors when interacting with the target.
5122 Internal errors include situations where @value{GDBN} has detected, at
5123 run time, a corrupt or erroneous situation.
5125 When reporting an internal error, @value{GDBN} uses
5126 @code{internal_error} and @code{gdb_assert}.
5128 @value{GDBN} must not call @code{abort} or @code{assert}.
5130 @emph{Pragmatics: There is no @code{internal_warning} function. Either
5131 the code detected a user error, recovered from it and issued a
5132 @code{warning} or the code failed to correctly recover from the user
5133 error and issued an @code{internal_error}.}
5135 @subsection File Names
5137 Any file used when building the core of @value{GDBN} must be in lower
5138 case. Any file used when building the core of @value{GDBN} must be 8.3
5139 unique. These requirements apply to both source and generated files.
5141 @emph{Pragmatics: The core of @value{GDBN} must be buildable on many
5142 platforms including DJGPP and MacOS/HFS. Every time an unfriendly file
5143 is introduced to the build process both @file{Makefile.in} and
5144 @file{configure.in} need to be modified accordingly. Compare the
5145 convoluted conversion process needed to transform @file{COPYING} into
5146 @file{copying.c} with the conversion needed to transform
5147 @file{version.in} into @file{version.c}.}
5149 Any file non 8.3 compliant file (that is not used when building the core
5150 of @value{GDBN}) must be added to @file{gdb/config/djgpp/fnchange.lst}.
5152 @emph{Pragmatics: This is clearly a compromise.}
5154 When @value{GDBN} has a local version of a system header file (ex
5155 @file{string.h}) the file name based on the POSIX header prefixed with
5156 @file{gdb_} (@file{gdb_string.h}). These headers should be relatively
5157 independent: they should use only macros defined by @file{configure},
5158 the compiler, or the host; they should include only system headers; they
5159 should refer only to system types. They may be shared between multiple
5160 programs, e.g.@: @value{GDBN} and @sc{gdbserver}.
5162 For other files @samp{-} is used as the separator.
5165 @subsection Include Files
5167 A @file{.c} file should include @file{defs.h} first.
5169 A @file{.c} file should directly include the @code{.h} file of every
5170 declaration and/or definition it directly refers to. It cannot rely on
5173 A @file{.h} file should directly include the @code{.h} file of every
5174 declaration and/or definition it directly refers to. It cannot rely on
5175 indirect inclusion. Exception: The file @file{defs.h} does not need to
5176 be directly included.
5178 An external declaration should only appear in one include file.
5180 An external declaration should never appear in a @code{.c} file.
5181 Exception: a declaration for the @code{_initialize} function that
5182 pacifies @option{-Wmissing-declaration}.
5184 A @code{typedef} definition should only appear in one include file.
5186 An opaque @code{struct} declaration can appear in multiple @file{.h}
5187 files. Where possible, a @file{.h} file should use an opaque
5188 @code{struct} declaration instead of an include.
5190 All @file{.h} files should be wrapped in:
5193 #ifndef INCLUDE_FILE_NAME_H
5194 #define INCLUDE_FILE_NAME_H
5200 @subsection Clean Design and Portable Implementation
5203 In addition to getting the syntax right, there's the little question of
5204 semantics. Some things are done in certain ways in @value{GDBN} because long
5205 experience has shown that the more obvious ways caused various kinds of
5208 @cindex assumptions about targets
5209 You can't assume the byte order of anything that comes from a target
5210 (including @var{value}s, object files, and instructions). Such things
5211 must be byte-swapped using @code{SWAP_TARGET_AND_HOST} in
5212 @value{GDBN}, or one of the swap routines defined in @file{bfd.h},
5213 such as @code{bfd_get_32}.
5215 You can't assume that you know what interface is being used to talk to
5216 the target system. All references to the target must go through the
5217 current @code{target_ops} vector.
5219 You can't assume that the host and target machines are the same machine
5220 (except in the ``native'' support modules). In particular, you can't
5221 assume that the target machine's header files will be available on the
5222 host machine. Target code must bring along its own header files --
5223 written from scratch or explicitly donated by their owner, to avoid
5227 Insertion of new @code{#ifdef}'s will be frowned upon. It's much better
5228 to write the code portably than to conditionalize it for various
5231 @cindex system dependencies
5232 New @code{#ifdef}'s which test for specific compilers or manufacturers
5233 or operating systems are unacceptable. All @code{#ifdef}'s should test
5234 for features. The information about which configurations contain which
5235 features should be segregated into the configuration files. Experience
5236 has proven far too often that a feature unique to one particular system
5237 often creeps into other systems; and that a conditional based on some
5238 predefined macro for your current system will become worthless over
5239 time, as new versions of your system come out that behave differently
5240 with regard to this feature.
5242 Adding code that handles specific architectures, operating systems,
5243 target interfaces, or hosts, is not acceptable in generic code.
5245 @cindex portable file name handling
5246 @cindex file names, portability
5247 One particularly notorious area where system dependencies tend to
5248 creep in is handling of file names. The mainline @value{GDBN} code
5249 assumes Posix semantics of file names: absolute file names begin with
5250 a forward slash @file{/}, slashes are used to separate leading
5251 directories, case-sensitive file names. These assumptions are not
5252 necessarily true on non-Posix systems such as MS-Windows. To avoid
5253 system-dependent code where you need to take apart or construct a file
5254 name, use the following portable macros:
5257 @findex HAVE_DOS_BASED_FILE_SYSTEM
5258 @item HAVE_DOS_BASED_FILE_SYSTEM
5259 This preprocessing symbol is defined to a non-zero value on hosts
5260 whose filesystems belong to the MS-DOS/MS-Windows family. Use this
5261 symbol to write conditional code which should only be compiled for
5264 @findex IS_DIR_SEPARATOR
5265 @item IS_DIR_SEPARATOR (@var{c})
5266 Evaluates to a non-zero value if @var{c} is a directory separator
5267 character. On Unix and GNU/Linux systems, only a slash @file{/} is
5268 such a character, but on Windows, both @file{/} and @file{\} will
5271 @findex IS_ABSOLUTE_PATH
5272 @item IS_ABSOLUTE_PATH (@var{file})
5273 Evaluates to a non-zero value if @var{file} is an absolute file name.
5274 For Unix and GNU/Linux hosts, a name which begins with a slash
5275 @file{/} is absolute. On DOS and Windows, @file{d:/foo} and
5276 @file{x:\bar} are also absolute file names.
5278 @findex FILENAME_CMP
5279 @item FILENAME_CMP (@var{f1}, @var{f2})
5280 Calls a function which compares file names @var{f1} and @var{f2} as
5281 appropriate for the underlying host filesystem. For Posix systems,
5282 this simply calls @code{strcmp}; on case-insensitive filesystems it
5283 will call @code{strcasecmp} instead.
5285 @findex DIRNAME_SEPARATOR
5286 @item DIRNAME_SEPARATOR
5287 Evaluates to a character which separates directories in
5288 @code{PATH}-style lists, typically held in environment variables.
5289 This character is @samp{:} on Unix, @samp{;} on DOS and Windows.
5291 @findex SLASH_STRING
5293 This evaluates to a constant string you should use to produce an
5294 absolute filename from leading directories and the file's basename.
5295 @code{SLASH_STRING} is @code{"/"} on most systems, but might be
5296 @code{"\\"} for some Windows-based ports.
5299 In addition to using these macros, be sure to use portable library
5300 functions whenever possible. For example, to extract a directory or a
5301 basename part from a file name, use the @code{dirname} and
5302 @code{basename} library functions (available in @code{libiberty} for
5303 platforms which don't provide them), instead of searching for a slash
5304 with @code{strrchr}.
5306 Another way to generalize @value{GDBN} along a particular interface is with an
5307 attribute struct. For example, @value{GDBN} has been generalized to handle
5308 multiple kinds of remote interfaces---not by @code{#ifdef}s everywhere, but
5309 by defining the @code{target_ops} structure and having a current target (as
5310 well as a stack of targets below it, for memory references). Whenever
5311 something needs to be done that depends on which remote interface we are
5312 using, a flag in the current target_ops structure is tested (e.g.,
5313 @code{target_has_stack}), or a function is called through a pointer in the
5314 current target_ops structure. In this way, when a new remote interface
5315 is added, only one module needs to be touched---the one that actually
5316 implements the new remote interface. Other examples of
5317 attribute-structs are BFD access to multiple kinds of object file
5318 formats, or @value{GDBN}'s access to multiple source languages.
5320 Please avoid duplicating code. For example, in @value{GDBN} 3.x all
5321 the code interfacing between @code{ptrace} and the rest of
5322 @value{GDBN} was duplicated in @file{*-dep.c}, and so changing
5323 something was very painful. In @value{GDBN} 4.x, these have all been
5324 consolidated into @file{infptrace.c}. @file{infptrace.c} can deal
5325 with variations between systems the same way any system-independent
5326 file would (hooks, @code{#if defined}, etc.), and machines which are
5327 radically different don't need to use @file{infptrace.c} at all.
5329 All debugging code must be controllable using the @samp{set debug
5330 @var{module}} command. Do not use @code{printf} to print trace
5331 messages. Use @code{fprintf_unfiltered(gdb_stdlog, ...}. Do not use
5332 @code{#ifdef DEBUG}.
5337 @chapter Porting @value{GDBN}
5338 @cindex porting to new machines
5340 Most of the work in making @value{GDBN} compile on a new machine is in
5341 specifying the configuration of the machine. This is done in a
5342 dizzying variety of header files and configuration scripts, which we
5343 hope to make more sensible soon. Let's say your new host is called an
5344 @var{xyz} (e.g., @samp{sun4}), and its full three-part configuration
5345 name is @code{@var{arch}-@var{xvend}-@var{xos}} (e.g.,
5346 @samp{sparc-sun-sunos4}). In particular:
5350 In the top level directory, edit @file{config.sub} and add @var{arch},
5351 @var{xvend}, and @var{xos} to the lists of supported architectures,
5352 vendors, and operating systems near the bottom of the file. Also, add
5353 @var{xyz} as an alias that maps to
5354 @code{@var{arch}-@var{xvend}-@var{xos}}. You can test your changes by
5358 ./config.sub @var{xyz}
5365 ./config.sub @code{@var{arch}-@var{xvend}-@var{xos}}
5369 which should both respond with @code{@var{arch}-@var{xvend}-@var{xos}}
5370 and no error messages.
5373 You need to port BFD, if that hasn't been done already. Porting BFD is
5374 beyond the scope of this manual.
5377 To configure @value{GDBN} itself, edit @file{gdb/configure.host} to recognize
5378 your system and set @code{gdb_host} to @var{xyz}, and (unless your
5379 desired target is already available) also edit @file{gdb/configure.tgt},
5380 setting @code{gdb_target} to something appropriate (for instance,
5383 @emph{Maintainer's note: Work in progress. The file
5384 @file{gdb/configure.host} originally needed to be modified when either a
5385 new native target or a new host machine was being added to @value{GDBN}.
5386 Recent changes have removed this requirement. The file now only needs
5387 to be modified when adding a new native configuration. This will likely
5388 changed again in the future.}
5391 Finally, you'll need to specify and define @value{GDBN}'s host-, native-, and
5392 target-dependent @file{.h} and @file{.c} files used for your
5396 @node Versions and Branches
5397 @chapter Versions and Branches
5401 @value{GDBN}'s version is determined by the file
5402 @file{gdb/version.in} and takes one of the following forms:
5405 @item @var{major}.@var{minor}
5406 @itemx @var{major}.@var{minor}.@var{patchlevel}
5407 an official release (e.g., 6.2 or 6.2.1)
5408 @item @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD}
5409 a snapshot taken at @var{YYYY}-@var{MM}-@var{DD}-gmt (e.g.,
5410 6.1.50.20020302, 6.1.90.20020304, or 6.1.0.20020308)
5411 @item @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD}-cvs
5412 a @sc{cvs} check out drawn on @var{YYYY}-@var{MM}-@var{DD} (e.g.,
5413 6.1.50.20020302-cvs, 6.1.90.20020304-cvs, or 6.1.0.20020308-cvs)
5414 @item @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD} (@var{vendor})
5415 a vendor specific release of @value{GDBN}, that while based on@*
5416 @var{major}.@var{minor}.@var{patchlevel}.@var{YYYY}@var{MM}@var{DD},
5417 may include additional changes
5420 @value{GDBN}'s mainline uses the @var{major} and @var{minor} version
5421 numbers from the most recent release branch, with a @var{patchlevel}
5422 of 50. At the time each new release branch is created, the mainline's
5423 @var{major} and @var{minor} version numbers are updated.
5425 @value{GDBN}'s release branch is similar. When the branch is cut, the
5426 @var{patchlevel} is changed from 50 to 90. As draft releases are
5427 drawn from the branch, the @var{patchlevel} is incremented. Once the
5428 first release (@var{major}.@var{minor}) has been made, the
5429 @var{patchlevel} is set to 0 and updates have an incremented
5432 For snapshots, and @sc{cvs} check outs, it is also possible to
5433 identify the @sc{cvs} origin:
5436 @item @var{major}.@var{minor}.50.@var{YYYY}@var{MM}@var{DD}
5437 drawn from the @sc{head} of mainline @sc{cvs} (e.g., 6.1.50.20020302)
5438 @item @var{major}.@var{minor}.90.@var{YYYY}@var{MM}@var{DD}
5439 @itemx @var{major}.@var{minor}.91.@var{YYYY}@var{MM}@var{DD} @dots{}
5440 drawn from a release branch prior to the release (e.g.,
5442 @item @var{major}.@var{minor}.0.@var{YYYY}@var{MM}@var{DD}
5443 @itemx @var{major}.@var{minor}.1.@var{YYYY}@var{MM}@var{DD} @dots{}
5444 drawn from a release branch after the release (e.g., 6.2.0.20020308)
5447 If the previous @value{GDBN} version is 6.1 and the current version is
5448 6.2, then, substituting 6 for @var{major} and 1 or 2 for @var{minor},
5449 here's an illustration of a typical sequence:
5456 +--------------------------.
5459 6.2.50.20020303-cvs 6.1.90 (draft #1)
5461 6.2.50.20020304-cvs 6.1.90.20020304-cvs
5463 6.2.50.20020305-cvs 6.1.91 (draft #2)
5465 6.2.50.20020306-cvs 6.1.91.20020306-cvs
5467 6.2.50.20020307-cvs 6.2 (release)
5469 6.2.50.20020308-cvs 6.2.0.20020308-cvs
5471 6.2.50.20020309-cvs 6.2.1 (update)
5473 6.2.50.20020310-cvs <branch closed>
5477 +--------------------------.
5480 6.3.50.20020312-cvs 6.2.90 (draft #1)
5484 @section Release Branches
5485 @cindex Release Branches
5487 @value{GDBN} draws a release series (6.2, 6.2.1, @dots{}) from a
5488 single release branch, and identifies that branch using the @sc{cvs}
5492 gdb_@var{major}_@var{minor}-@var{YYYY}@var{MM}@var{DD}-branchpoint
5493 gdb_@var{major}_@var{minor}-branch
5494 gdb_@var{major}_@var{minor}-@var{YYYY}@var{MM}@var{DD}-release
5497 @emph{Pragmatics: To help identify the date at which a branch or
5498 release is made, both the branchpoint and release tags include the
5499 date that they are cut (@var{YYYY}@var{MM}@var{DD}) in the tag. The
5500 branch tag, denoting the head of the branch, does not need this.}
5502 @section Vendor Branches
5503 @cindex vendor branches
5505 To avoid version conflicts, vendors are expected to modify the file
5506 @file{gdb/version.in} to include a vendor unique alphabetic identifier
5507 (an official @value{GDBN} release never uses alphabetic characters in
5508 its version identifer). E.g., @samp{6.2widgit2}, or @samp{6.2 (Widgit
5511 @section Experimental Branches
5512 @cindex experimental branches
5514 @subsection Guidelines
5516 @value{GDBN} permits the creation of branches, cut from the @sc{cvs}
5517 repository, for experimental development. Branches make it possible
5518 for developers to share preliminary work, and maintainers to examine
5519 significant new developments.
5521 The following are a set of guidelines for creating such branches:
5525 @item a branch has an owner
5526 The owner can set further policy for a branch, but may not change the
5527 ground rules. In particular, they can set a policy for commits (be it
5528 adding more reviewers or deciding who can commit).
5530 @item all commits are posted
5531 All changes committed to a branch shall also be posted to
5532 @email{gdb-patches@@sources.redhat.com, the @value{GDBN} patches
5533 mailing list}. While commentary on such changes are encouraged, people
5534 should remember that the changes only apply to a branch.
5536 @item all commits are covered by an assignment
5537 This ensures that all changes belong to the Free Software Foundation,
5538 and avoids the possibility that the branch may become contaminated.
5540 @item a branch is focused
5541 A focused branch has a single objective or goal, and does not contain
5542 unnecessary or irrelevant changes. Cleanups, where identified, being
5543 be pushed into the mainline as soon as possible.
5545 @item a branch tracks mainline
5546 This keeps the level of divergence under control. It also keeps the
5547 pressure on developers to push cleanups and other stuff into the
5550 @item a branch shall contain the entire @value{GDBN} module
5551 The @value{GDBN} module @code{gdb} should be specified when creating a
5552 branch (branches of individual files should be avoided). @xref{Tags}.
5554 @item a branch shall be branded using @file{version.in}
5555 The file @file{gdb/version.in} shall be modified so that it identifies
5556 the branch @var{owner} and branch @var{name}, e.g.,
5557 @samp{6.2.50.20030303_owner_name} or @samp{6.2 (Owner Name)}.
5564 To simplify the identification of @value{GDBN} branches, the following
5565 branch tagging convention is strongly recommended:
5569 @item @var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint
5570 @itemx @var{owner}_@var{name}-@var{YYYYMMDD}-branch
5571 The branch point and corresponding branch tag. @var{YYYYMMDD} is the
5572 date that the branch was created. A branch is created using the
5573 sequence: @anchor{experimental branch tags}
5575 cvs rtag @var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint gdb
5576 cvs rtag -b -r @var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint \
5577 @var{owner}_@var{name}-@var{YYYYMMDD}-branch gdb
5580 @item @var{owner}_@var{name}-@var{yyyymmdd}-mergepoint
5581 The tagged point, on the mainline, that was used when merging the branch
5582 on @var{yyyymmdd}. To merge in all changes since the branch was cut,
5583 use a command sequence like:
5585 cvs rtag @var{owner}_@var{name}-@var{yyyymmdd}-mergepoint gdb
5587 -j@var{owner}_@var{name}-@var{YYYYMMDD}-branchpoint
5588 -j@var{owner}_@var{name}-@var{yyyymmdd}-mergepoint
5591 Similar sequences can be used to just merge in changes since the last
5597 For further information on @sc{cvs}, see
5598 @uref{http://www.gnu.org/software/cvs/, Concurrent Versions System}.
5602 @chapter Releasing @value{GDBN}
5603 @cindex making a new release of gdb
5605 @section Branch Commit Policy
5607 The branch commit policy is pretty slack. @value{GDBN} releases 5.0,
5608 5.1 and 5.2 all used the below:
5612 The @file{gdb/MAINTAINERS} file still holds.
5614 Don't fix something on the branch unless/until it is also fixed in the
5615 trunk. If this isn't possible, mentioning it in the @file{gdb/PROBLEMS}
5616 file is better than committing a hack.
5618 When considering a patch for the branch, suggested criteria include:
5619 Does it fix a build? Does it fix the sequence @kbd{break main; run}
5620 when debugging a static binary?
5622 The further a change is from the core of @value{GDBN}, the less likely
5623 the change will worry anyone (e.g., target specific code).
5625 Only post a proposal to change the core of @value{GDBN} after you've
5626 sent individual bribes to all the people listed in the
5627 @file{MAINTAINERS} file @t{;-)}
5630 @emph{Pragmatics: Provided updates are restricted to non-core
5631 functionality there is little chance that a broken change will be fatal.
5632 This means that changes such as adding a new architectures or (within
5633 reason) support for a new host are considered acceptable.}
5636 @section Obsoleting code
5638 Before anything else, poke the other developers (and around the source
5639 code) to see if there is anything that can be removed from @value{GDBN}
5640 (an old target, an unused file).
5642 Obsolete code is identified by adding an @code{OBSOLETE} prefix to every
5643 line. Doing this means that it is easy to identify something that has
5644 been obsoleted when greping through the sources.
5646 The process is done in stages --- this is mainly to ensure that the
5647 wider @value{GDBN} community has a reasonable opportunity to respond.
5648 Remember, everything on the Internet takes a week.
5652 Post the proposal on @email{gdb@@sources.redhat.com, the GDB mailing
5653 list} Creating a bug report to track the task's state, is also highly
5658 Post the proposal on @email{gdb-announce@@sources.redhat.com, the GDB
5659 Announcement mailing list}.
5663 Go through and edit all relevant files and lines so that they are
5664 prefixed with the word @code{OBSOLETE}.
5666 Wait until the next GDB version, containing this obsolete code, has been
5669 Remove the obsolete code.
5673 @emph{Maintainer note: While removing old code is regrettable it is
5674 hopefully better for @value{GDBN}'s long term development. Firstly it
5675 helps the developers by removing code that is either no longer relevant
5676 or simply wrong. Secondly since it removes any history associated with
5677 the file (effectively clearing the slate) the developer has a much freer
5678 hand when it comes to fixing broken files.}
5682 @section Before the Branch
5684 The most important objective at this stage is to find and fix simple
5685 changes that become a pain to track once the branch is created. For
5686 instance, configuration problems that stop @value{GDBN} from even
5687 building. If you can't get the problem fixed, document it in the
5688 @file{gdb/PROBLEMS} file.
5690 @subheading Prompt for @file{gdb/NEWS}
5692 People always forget. Send a post reminding them but also if you know
5693 something interesting happened add it yourself. The @code{schedule}
5694 script will mention this in its e-mail.
5696 @subheading Review @file{gdb/README}
5698 Grab one of the nightly snapshots and then walk through the
5699 @file{gdb/README} looking for anything that can be improved. The
5700 @code{schedule} script will mention this in its e-mail.
5702 @subheading Refresh any imported files.
5704 A number of files are taken from external repositories. They include:
5708 @file{texinfo/texinfo.tex}
5710 @file{config.guess} et.@: al.@: (see the top-level @file{MAINTAINERS}
5713 @file{etc/standards.texi}, @file{etc/make-stds.texi}
5716 @subheading Check the ARI
5718 @uref{http://sources.redhat.com/gdb/ari,,A.R.I.} is an @code{awk} script
5719 (Awk Regression Index ;-) that checks for a number of errors and coding
5720 conventions. The checks include things like using @code{malloc} instead
5721 of @code{xmalloc} and file naming problems. There shouldn't be any
5724 @subsection Review the bug data base
5726 Close anything obviously fixed.
5728 @subsection Check all cross targets build
5730 The targets are listed in @file{gdb/MAINTAINERS}.
5733 @section Cut the Branch
5735 @subheading Create the branch
5740 $ V=`echo $v | sed 's/\./_/g'`
5741 $ D=`date -u +%Y-%m-%d`
5744 $ echo cvs -f -d :ext:sources.redhat.com:/cvs/src rtag \
5745 -D $D-gmt gdb_$V-$D-branchpoint insight+dejagnu
5746 cvs -f -d :ext:sources.redhat.com:/cvs/src rtag
5747 -D 2002-03-03-gmt gdb_5_2-2002-03-03-branchpoint insight+dejagnu
5750 $ echo cvs -f -d :ext:sources.redhat.com:/cvs/src rtag \
5751 -b -r gdb_$V-$D-branchpoint gdb_$V-branch insight+dejagnu
5752 cvs -f -d :ext:sources.redhat.com:/cvs/src rtag \
5753 -b -r gdb_5_2-2002-03-03-branchpoint gdb_5_2-branch insight+dejagnu
5761 by using @kbd{-D YYYY-MM-DD-gmt} the branch is forced to an exact
5764 the trunk is first taged so that the branch point can easily be found
5766 Insight (which includes GDB) and dejagnu are all tagged at the same time
5768 @file{version.in} gets bumped to avoid version number conflicts
5770 the reading of @file{.cvsrc} is disabled using @file{-f}
5773 @subheading Update @file{version.in}
5778 $ V=`echo $v | sed 's/\./_/g'`
5782 $ echo cvs -f -d :ext:sources.redhat.com:/cvs/src co \
5783 -r gdb_$V-branch src/gdb/version.in
5784 cvs -f -d :ext:sources.redhat.com:/cvs/src co
5785 -r gdb_5_2-branch src/gdb/version.in
5787 U src/gdb/version.in
5789 $ echo $u.90-0000-00-00-cvs > version.in
5791 5.1.90-0000-00-00-cvs
5792 $ cvs -f commit version.in
5797 @file{0000-00-00} is used as a date to pump prime the version.in update
5800 @file{.90} and the previous branch version are used as fairly arbitrary
5801 initial branch version number
5805 @subheading Update the web and news pages
5809 @subheading Tweak cron to track the new branch
5811 The file @file{gdbadmin/cron/crontab} contains gdbadmin's cron table.
5812 This file needs to be updated so that:
5816 a daily timestamp is added to the file @file{version.in}
5818 the new branch is included in the snapshot process
5822 See the file @file{gdbadmin/cron/README} for how to install the updated
5825 The file @file{gdbadmin/ss/README} should also be reviewed to reflect
5826 any changes. That file is copied to both the branch/ and current/
5827 snapshot directories.
5830 @subheading Update the NEWS and README files
5832 The @file{NEWS} file needs to be updated so that on the branch it refers
5833 to @emph{changes in the current release} while on the trunk it also
5834 refers to @emph{changes since the current release}.
5836 The @file{README} file needs to be updated so that it refers to the
5839 @subheading Post the branch info
5841 Send an announcement to the mailing lists:
5845 @email{gdb-announce@@sources.redhat.com, GDB Announcement mailing list}
5847 @email{gdb@@sources.redhat.com, GDB Discsussion mailing list} and
5848 @email{gdb-testers@@sources.redhat.com, GDB Discsussion mailing list}
5851 @emph{Pragmatics: The branch creation is sent to the announce list to
5852 ensure that people people not subscribed to the higher volume discussion
5855 The announcement should include:
5861 how to check out the branch using CVS
5863 the date/number of weeks until the release
5865 the branch commit policy
5869 @section Stabilize the branch
5871 Something goes here.
5873 @section Create a Release
5875 The process of creating and then making available a release is broken
5876 down into a number of stages. The first part addresses the technical
5877 process of creating a releasable tar ball. The later stages address the
5878 process of releasing that tar ball.
5880 When making a release candidate just the first section is needed.
5882 @subsection Create a release candidate
5884 The objective at this stage is to create a set of tar balls that can be
5885 made available as a formal release (or as a less formal release
5888 @subsubheading Freeze the branch
5890 Send out an e-mail notifying everyone that the branch is frozen to
5891 @email{gdb-patches@@sources.redhat.com}.
5893 @subsubheading Establish a few defaults.
5898 $ t=/sourceware/snapshot-tmp/gdbadmin-tmp
5900 /sourceware/snapshot-tmp/gdbadmin-tmp/gdb_5_2-branch/5.2
5904 /sourceware/snapshot-tmp/gdbadmin-tmp/gdb_5_2-branch/5.2
5906 /home/gdbadmin/bin/autoconf
5915 Check the @code{autoconf} version carefully. You want to be using the
5916 version taken from the @file{binutils} snapshot directory, which can be
5917 found at @uref{ftp://sources.redhat.com/pub/binutils/}. It is very
5918 unlikely that a system installed version of @code{autoconf} (e.g.,
5919 @file{/usr/bin/autoconf}) is correct.
5922 @subsubheading Check out the relevant modules:
5925 $ for m in gdb insight dejagnu
5927 ( mkdir -p $m && cd $m && cvs -q -f -d /cvs/src co -P -r $b $m )
5937 The reading of @file{.cvsrc} is disabled (@file{-f}) so that there isn't
5938 any confusion between what is written here and what your local
5939 @code{cvs} really does.
5942 @subsubheading Update relevant files.
5948 Major releases get their comments added as part of the mainline. Minor
5949 releases should probably mention any significant bugs that were fixed.
5951 Don't forget to include the @file{ChangeLog} entry.
5954 $ emacs gdb/src/gdb/NEWS
5959 $ cp gdb/src/gdb/NEWS insight/src/gdb/NEWS
5960 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
5965 You'll need to update:
5977 $ emacs gdb/src/gdb/README
5982 $ cp gdb/src/gdb/README insight/src/gdb/README
5983 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
5986 @emph{Maintainer note: Hopefully the @file{README} file was reviewed
5987 before the initial branch was cut so just a simple substitute is needed
5990 @emph{Maintainer note: Other projects generate @file{README} and
5991 @file{INSTALL} from the core documentation. This might be worth
5994 @item gdb/version.in
5997 $ echo $v > gdb/src/gdb/version.in
5998 $ cat gdb/src/gdb/version.in
6000 $ emacs gdb/src/gdb/version.in
6003 ... Bump to version ...
6005 $ cp gdb/src/gdb/version.in insight/src/gdb/version.in
6006 $ cp gdb/src/gdb/ChangeLog insight/src/gdb/ChangeLog
6009 @item dejagnu/src/dejagnu/configure.in
6011 Dejagnu is more complicated. The version number is a parameter to
6012 @code{AM_INIT_AUTOMAKE}. Tweak it to read something like gdb-5.1.91.
6014 Don't forget to re-generate @file{configure}.
6016 Don't forget to include a @file{ChangeLog} entry.
6019 $ emacs dejagnu/src/dejagnu/configure.in
6024 $ ( cd dejagnu/src/dejagnu && autoconf )
6029 @subsubheading Do the dirty work
6031 This is identical to the process used to create the daily snapshot.
6034 $ for m in gdb insight
6036 ( cd $m/src && gmake -f src-release $m.tar )
6038 $ ( m=dejagnu; cd $m/src && gmake -f src-release $m.tar.bz2 )
6041 If the top level source directory does not have @file{src-release}
6042 (@value{GDBN} version 5.3.1 or earlier), try these commands instead:
6045 $ for m in gdb insight
6047 ( cd $m/src && gmake -f Makefile.in $m.tar )
6049 $ ( m=dejagnu; cd $m/src && gmake -f Makefile.in $m.tar.bz2 )
6052 @subsubheading Check the source files
6054 You're looking for files that have mysteriously disappeared.
6055 @kbd{distclean} has the habit of deleting files it shouldn't. Watch out
6056 for the @file{version.in} update @kbd{cronjob}.
6059 $ ( cd gdb/src && cvs -f -q -n update )
6063 @dots{} lots of generated files @dots{}
6068 @dots{} lots of generated files @dots{}
6073 @emph{Don't worry about the @file{gdb.info-??} or
6074 @file{gdb/p-exp.tab.c}. They were generated (and yes @file{gdb.info-1}
6075 was also generated only something strange with CVS means that they
6076 didn't get supressed). Fixing it would be nice though.}
6078 @subsubheading Create compressed versions of the release
6084 dejagnu/ dejagnu-gdb-5.2.tar.bz2 gdb/ gdb-5.2.tar insight/ insight-5.2.tar
6085 $ for m in gdb insight
6087 bzip2 -v -9 -c $m-$v.tar > $m-$v.tar.bz2
6088 gzip -v -9 -c $m-$v.tar > $m-$v.tar.gz
6098 A pipe such as @kbd{bunzip2 < xxx.bz2 | gzip -9 > xxx.gz} is not since,
6099 in that mode, @code{gzip} does not know the name of the file and, hence,
6100 can not include it in the compressed file. This is also why the release
6101 process runs @code{tar} and @code{bzip2} as separate passes.
6104 @subsection Sanity check the tar ball
6106 Pick a popular machine (Solaris/PPC?) and try the build on that.
6109 $ bunzip2 < gdb-5.2.tar.bz2 | tar xpf -
6114 $ ./gdb/gdb ./gdb/gdb
6118 Breakpoint 1 at 0x80732bc: file main.c, line 734.
6120 Starting program: /tmp/gdb-5.2/gdb/gdb
6122 Breakpoint 1, main (argc=1, argv=0xbffff8b4) at main.c:734
6123 734 catch_errors (captured_main, &args, "", RETURN_MASK_ALL);
6125 $1 = @{argc = 136426532, argv = 0x821b7f0@}
6129 @subsection Make a release candidate available
6131 If this is a release candidate then the only remaining steps are:
6135 Commit @file{version.in} and @file{ChangeLog}
6137 Tweak @file{version.in} (and @file{ChangeLog} to read
6138 @var{L}.@var{M}.@var{N}-0000-00-00-cvs so that the version update
6139 process can restart.
6141 Make the release candidate available in
6142 @uref{ftp://sources.redhat.com/pub/gdb/snapshots/branch}
6144 Notify the relevant mailing lists ( @email{gdb@@sources.redhat.com} and
6145 @email{gdb-testers@@sources.redhat.com} that the candidate is available.
6148 @subsection Make a formal release available
6150 (And you thought all that was required was to post an e-mail.)
6152 @subsubheading Install on sware
6154 Copy the new files to both the release and the old release directory:
6157 $ cp *.bz2 *.gz ~ftp/pub/gdb/old-releases/
6158 $ cp *.bz2 *.gz ~ftp/pub/gdb/releases
6162 Clean up the releases directory so that only the most recent releases
6163 are available (e.g. keep 5.2 and 5.2.1 but remove 5.1):
6166 $ cd ~ftp/pub/gdb/releases
6171 Update the file @file{README} and @file{.message} in the releases
6178 $ ln README .message
6181 @subsubheading Update the web pages.
6185 @item htdocs/download/ANNOUNCEMENT
6186 This file, which is posted as the official announcement, includes:
6189 General announcement
6191 News. If making an @var{M}.@var{N}.1 release, retain the news from
6192 earlier @var{M}.@var{N} release.
6197 @item htdocs/index.html
6198 @itemx htdocs/news/index.html
6199 @itemx htdocs/download/index.html
6200 These files include:
6203 announcement of the most recent release
6205 news entry (remember to update both the top level and the news directory).
6207 These pages also need to be regenerate using @code{index.sh}.
6209 @item download/onlinedocs/
6210 You need to find the magic command that is used to generate the online
6211 docs from the @file{.tar.bz2}. The best way is to look in the output
6212 from one of the nightly @code{cron} jobs and then just edit accordingly.
6216 $ ~/ss/update-web-docs \
6217 ~ftp/pub/gdb/releases/gdb-5.2.tar.bz2 \
6219 /www/sourceware/htdocs/gdb/download/onlinedocs \
6224 Just like the online documentation. Something like:
6227 $ /bin/sh ~/ss/update-web-ari \
6228 ~ftp/pub/gdb/releases/gdb-5.2.tar.bz2 \
6230 /www/sourceware/htdocs/gdb/download/ari \
6236 @subsubheading Shadow the pages onto gnu
6238 Something goes here.
6241 @subsubheading Install the @value{GDBN} tar ball on GNU
6243 At the time of writing, the GNU machine was @kbd{gnudist.gnu.org} in
6244 @file{~ftp/gnu/gdb}.
6246 @subsubheading Make the @file{ANNOUNCEMENT}
6248 Post the @file{ANNOUNCEMENT} file you created above to:
6252 @email{gdb-announce@@sources.redhat.com, GDB Announcement mailing list}
6254 @email{info-gnu@@gnu.org, General GNU Announcement list} (but delay it a
6255 day or so to let things get out)
6257 @email{bug-gdb@@gnu.org, GDB Bug Report mailing list}
6262 The release is out but you're still not finished.
6264 @subsubheading Commit outstanding changes
6266 In particular you'll need to commit any changes to:
6270 @file{gdb/ChangeLog}
6272 @file{gdb/version.in}
6279 @subsubheading Tag the release
6284 $ d=`date -u +%Y-%m-%d`
6287 $ ( cd insight/src/gdb && cvs -f -q update )
6288 $ ( cd insight/src && cvs -f -q tag gdb_5_2-$d-release )
6291 Insight is used since that contains more of the release than
6292 @value{GDBN} (@code{dejagnu} doesn't get tagged but I think we can live
6295 @subsubheading Mention the release on the trunk
6297 Just put something in the @file{ChangeLog} so that the trunk also
6298 indicates when the release was made.
6300 @subsubheading Restart @file{gdb/version.in}
6302 If @file{gdb/version.in} does not contain an ISO date such as
6303 @kbd{2002-01-24} then the daily @code{cronjob} won't update it. Having
6304 committed all the release changes it can be set to
6305 @file{5.2.0_0000-00-00-cvs} which will restart things (yes the @kbd{_}
6306 is important - it affects the snapshot process).
6308 Don't forget the @file{ChangeLog}.
6310 @subsubheading Merge into trunk
6312 The files committed to the branch may also need changes merged into the
6315 @subsubheading Revise the release schedule
6317 Post a revised release schedule to @email{gdb@@sources.redhat.com, GDB
6318 Discussion List} with an updated announcement. The schedule can be
6319 generated by running:
6322 $ ~/ss/schedule `date +%s` schedule
6326 The first parameter is approximate date/time in seconds (from the epoch)
6327 of the most recent release.
6329 Also update the schedule @code{cronjob}.
6331 @section Post release
6333 Remove any @code{OBSOLETE} code.
6340 The testsuite is an important component of the @value{GDBN} package.
6341 While it is always worthwhile to encourage user testing, in practice
6342 this is rarely sufficient; users typically use only a small subset of
6343 the available commands, and it has proven all too common for a change
6344 to cause a significant regression that went unnoticed for some time.
6346 The @value{GDBN} testsuite uses the DejaGNU testing framework.
6347 DejaGNU is built using @code{Tcl} and @code{expect}. The tests
6348 themselves are calls to various @code{Tcl} procs; the framework runs all the
6349 procs and summarizes the passes and fails.
6351 @section Using the Testsuite
6353 @cindex running the test suite
6354 To run the testsuite, simply go to the @value{GDBN} object directory (or to the
6355 testsuite's objdir) and type @code{make check}. This just sets up some
6356 environment variables and invokes DejaGNU's @code{runtest} script. While
6357 the testsuite is running, you'll get mentions of which test file is in use,
6358 and a mention of any unexpected passes or fails. When the testsuite is
6359 finished, you'll get a summary that looks like this:
6364 # of expected passes 6016
6365 # of unexpected failures 58
6366 # of unexpected successes 5
6367 # of expected failures 183
6368 # of unresolved testcases 3
6369 # of untested testcases 5
6372 The ideal test run consists of expected passes only; however, reality
6373 conspires to keep us from this ideal. Unexpected failures indicate
6374 real problems, whether in @value{GDBN} or in the testsuite. Expected
6375 failures are still failures, but ones which have been decided are too
6376 hard to deal with at the time; for instance, a test case might work
6377 everywhere except on AIX, and there is no prospect of the AIX case
6378 being fixed in the near future. Expected failures should not be added
6379 lightly, since you may be masking serious bugs in @value{GDBN}.
6380 Unexpected successes are expected fails that are passing for some
6381 reason, while unresolved and untested cases often indicate some minor
6382 catastrophe, such as the compiler being unable to deal with a test
6385 When making any significant change to @value{GDBN}, you should run the
6386 testsuite before and after the change, to confirm that there are no
6387 regressions. Note that truly complete testing would require that you
6388 run the testsuite with all supported configurations and a variety of
6389 compilers; however this is more than really necessary. In many cases
6390 testing with a single configuration is sufficient. Other useful
6391 options are to test one big-endian (Sparc) and one little-endian (x86)
6392 host, a cross config with a builtin simulator (powerpc-eabi,
6393 mips-elf), or a 64-bit host (Alpha).
6395 If you add new functionality to @value{GDBN}, please consider adding
6396 tests for it as well; this way future @value{GDBN} hackers can detect
6397 and fix their changes that break the functionality you added.
6398 Similarly, if you fix a bug that was not previously reported as a test
6399 failure, please add a test case for it. Some cases are extremely
6400 difficult to test, such as code that handles host OS failures or bugs
6401 in particular versions of compilers, and it's OK not to try to write
6402 tests for all of those.
6404 DejaGNU supports separate build, host, and target machines. However,
6405 some @value{GDBN} test scripts do not work if the build machine and
6406 the host machine are not the same. In such an environment, these scripts
6407 will give a result of ``UNRESOLVED'', like this:
6410 UNRESOLVED: gdb.base/example.exp: This test script does not work on a remote host.
6413 @section Testsuite Organization
6415 @cindex test suite organization
6416 The testsuite is entirely contained in @file{gdb/testsuite}. While the
6417 testsuite includes some makefiles and configury, these are very minimal,
6418 and used for little besides cleaning up, since the tests themselves
6419 handle the compilation of the programs that @value{GDBN} will run. The file
6420 @file{testsuite/lib/gdb.exp} contains common utility procs useful for
6421 all @value{GDBN} tests, while the directory @file{testsuite/config} contains
6422 configuration-specific files, typically used for special-purpose
6423 definitions of procs like @code{gdb_load} and @code{gdb_start}.
6425 The tests themselves are to be found in @file{testsuite/gdb.*} and
6426 subdirectories of those. The names of the test files must always end
6427 with @file{.exp}. DejaGNU collects the test files by wildcarding
6428 in the test directories, so both subdirectories and individual files
6429 get chosen and run in alphabetical order.
6431 The following table lists the main types of subdirectories and what they
6432 are for. Since DejaGNU finds test files no matter where they are
6433 located, and since each test file sets up its own compilation and
6434 execution environment, this organization is simply for convenience and
6439 This is the base testsuite. The tests in it should apply to all
6440 configurations of @value{GDBN} (but generic native-only tests may live here).
6441 The test programs should be in the subset of C that is valid K&R,
6442 ANSI/ISO, and C@t{++} (@code{#ifdef}s are allowed if necessary, for instance
6445 @item gdb.@var{lang}
6446 Language-specific tests for any language @var{lang} besides C. Examples are
6447 @file{gdb.cp} and @file{gdb.java}.
6449 @item gdb.@var{platform}
6450 Non-portable tests. The tests are specific to a specific configuration
6451 (host or target), such as HP-UX or eCos. Example is @file{gdb.hp}, for
6454 @item gdb.@var{compiler}
6455 Tests specific to a particular compiler. As of this writing (June
6456 1999), there aren't currently any groups of tests in this category that
6457 couldn't just as sensibly be made platform-specific, but one could
6458 imagine a @file{gdb.gcc}, for tests of @value{GDBN}'s handling of GCC
6461 @item gdb.@var{subsystem}
6462 Tests that exercise a specific @value{GDBN} subsystem in more depth. For
6463 instance, @file{gdb.disasm} exercises various disassemblers, while
6464 @file{gdb.stabs} tests pathways through the stabs symbol reader.
6467 @section Writing Tests
6468 @cindex writing tests
6470 In many areas, the @value{GDBN} tests are already quite comprehensive; you
6471 should be able to copy existing tests to handle new cases.
6473 You should try to use @code{gdb_test} whenever possible, since it
6474 includes cases to handle all the unexpected errors that might happen.
6475 However, it doesn't cost anything to add new test procedures; for
6476 instance, @file{gdb.base/exprs.exp} defines a @code{test_expr} that
6477 calls @code{gdb_test} multiple times.
6479 Only use @code{send_gdb} and @code{gdb_expect} when absolutely
6480 necessary, such as when @value{GDBN} has several valid responses to a command.
6482 The source language programs do @emph{not} need to be in a consistent
6483 style. Since @value{GDBN} is used to debug programs written in many different
6484 styles, it's worth having a mix of styles in the testsuite; for
6485 instance, some @value{GDBN} bugs involving the display of source lines would
6486 never manifest themselves if the programs used GNU coding style
6493 Check the @file{README} file, it often has useful information that does not
6494 appear anywhere else in the directory.
6497 * Getting Started:: Getting started working on @value{GDBN}
6498 * Debugging GDB:: Debugging @value{GDBN} with itself
6501 @node Getting Started,,, Hints
6503 @section Getting Started
6505 @value{GDBN} is a large and complicated program, and if you first starting to
6506 work on it, it can be hard to know where to start. Fortunately, if you
6507 know how to go about it, there are ways to figure out what is going on.
6509 This manual, the @value{GDBN} Internals manual, has information which applies
6510 generally to many parts of @value{GDBN}.
6512 Information about particular functions or data structures are located in
6513 comments with those functions or data structures. If you run across a
6514 function or a global variable which does not have a comment correctly
6515 explaining what is does, this can be thought of as a bug in @value{GDBN}; feel
6516 free to submit a bug report, with a suggested comment if you can figure
6517 out what the comment should say. If you find a comment which is
6518 actually wrong, be especially sure to report that.
6520 Comments explaining the function of macros defined in host, target, or
6521 native dependent files can be in several places. Sometimes they are
6522 repeated every place the macro is defined. Sometimes they are where the
6523 macro is used. Sometimes there is a header file which supplies a
6524 default definition of the macro, and the comment is there. This manual
6525 also documents all the available macros.
6526 @c (@pxref{Host Conditionals}, @pxref{Target
6527 @c Conditionals}, @pxref{Native Conditionals}, and @pxref{Obsolete
6530 Start with the header files. Once you have some idea of how
6531 @value{GDBN}'s internal symbol tables are stored (see @file{symtab.h},
6532 @file{gdbtypes.h}), you will find it much easier to understand the
6533 code which uses and creates those symbol tables.
6535 You may wish to process the information you are getting somehow, to
6536 enhance your understanding of it. Summarize it, translate it to another
6537 language, add some (perhaps trivial or non-useful) feature to @value{GDBN}, use
6538 the code to predict what a test case would do and write the test case
6539 and verify your prediction, etc. If you are reading code and your eyes
6540 are starting to glaze over, this is a sign you need to use a more active
6543 Once you have a part of @value{GDBN} to start with, you can find more
6544 specifically the part you are looking for by stepping through each
6545 function with the @code{next} command. Do not use @code{step} or you
6546 will quickly get distracted; when the function you are stepping through
6547 calls another function try only to get a big-picture understanding
6548 (perhaps using the comment at the beginning of the function being
6549 called) of what it does. This way you can identify which of the
6550 functions being called by the function you are stepping through is the
6551 one which you are interested in. You may need to examine the data
6552 structures generated at each stage, with reference to the comments in
6553 the header files explaining what the data structures are supposed to
6556 Of course, this same technique can be used if you are just reading the
6557 code, rather than actually stepping through it. The same general
6558 principle applies---when the code you are looking at calls something
6559 else, just try to understand generally what the code being called does,
6560 rather than worrying about all its details.
6562 @cindex command implementation
6563 A good place to start when tracking down some particular area is with
6564 a command which invokes that feature. Suppose you want to know how
6565 single-stepping works. As a @value{GDBN} user, you know that the
6566 @code{step} command invokes single-stepping. The command is invoked
6567 via command tables (see @file{command.h}); by convention the function
6568 which actually performs the command is formed by taking the name of
6569 the command and adding @samp{_command}, or in the case of an
6570 @code{info} subcommand, @samp{_info}. For example, the @code{step}
6571 command invokes the @code{step_command} function and the @code{info
6572 display} command invokes @code{display_info}. When this convention is
6573 not followed, you might have to use @code{grep} or @kbd{M-x
6574 tags-search} in emacs, or run @value{GDBN} on itself and set a
6575 breakpoint in @code{execute_command}.
6577 @cindex @code{bug-gdb} mailing list
6578 If all of the above fail, it may be appropriate to ask for information
6579 on @code{bug-gdb}. But @emph{never} post a generic question like ``I was
6580 wondering if anyone could give me some tips about understanding
6581 @value{GDBN}''---if we had some magic secret we would put it in this manual.
6582 Suggestions for improving the manual are always welcome, of course.
6584 @node Debugging GDB,,,Hints
6586 @section Debugging @value{GDBN} with itself
6587 @cindex debugging @value{GDBN}
6589 If @value{GDBN} is limping on your machine, this is the preferred way to get it
6590 fully functional. Be warned that in some ancient Unix systems, like
6591 Ultrix 4.2, a program can't be running in one process while it is being
6592 debugged in another. Rather than typing the command @kbd{@w{./gdb
6593 ./gdb}}, which works on Suns and such, you can copy @file{gdb} to
6594 @file{gdb2} and then type @kbd{@w{./gdb ./gdb2}}.
6596 When you run @value{GDBN} in the @value{GDBN} source directory, it will read a
6597 @file{.gdbinit} file that sets up some simple things to make debugging
6598 gdb easier. The @code{info} command, when executed without a subcommand
6599 in a @value{GDBN} being debugged by gdb, will pop you back up to the top level
6600 gdb. See @file{.gdbinit} for details.
6602 If you use emacs, you will probably want to do a @code{make TAGS} after
6603 you configure your distribution; this will put the machine dependent
6604 routines for your local machine where they will be accessed first by
6607 Also, make sure that you've either compiled @value{GDBN} with your local cc, or
6608 have run @code{fixincludes} if you are compiling with gcc.
6610 @section Submitting Patches
6612 @cindex submitting patches
6613 Thanks for thinking of offering your changes back to the community of
6614 @value{GDBN} users. In general we like to get well designed enhancements.
6615 Thanks also for checking in advance about the best way to transfer the
6618 The @value{GDBN} maintainers will only install ``cleanly designed'' patches.
6619 This manual summarizes what we believe to be clean design for @value{GDBN}.
6621 If the maintainers don't have time to put the patch in when it arrives,
6622 or if there is any question about a patch, it goes into a large queue
6623 with everyone else's patches and bug reports.
6625 @cindex legal papers for code contributions
6626 The legal issue is that to incorporate substantial changes requires a
6627 copyright assignment from you and/or your employer, granting ownership
6628 of the changes to the Free Software Foundation. You can get the
6629 standard documents for doing this by sending mail to @code{gnu@@gnu.org}
6630 and asking for it. We recommend that people write in "All programs
6631 owned by the Free Software Foundation" as "NAME OF PROGRAM", so that
6632 changes in many programs (not just @value{GDBN}, but GAS, Emacs, GCC,
6634 contributed with only one piece of legalese pushed through the
6635 bureaucracy and filed with the FSF. We can't start merging changes until
6636 this paperwork is received by the FSF (their rules, which we follow
6637 since we maintain it for them).
6639 Technically, the easiest way to receive changes is to receive each
6640 feature as a small context diff or unidiff, suitable for @code{patch}.
6641 Each message sent to me should include the changes to C code and
6642 header files for a single feature, plus @file{ChangeLog} entries for
6643 each directory where files were modified, and diffs for any changes
6644 needed to the manuals (@file{gdb/doc/gdb.texinfo} or
6645 @file{gdb/doc/gdbint.texinfo}). If there are a lot of changes for a
6646 single feature, they can be split down into multiple messages.
6648 In this way, if we read and like the feature, we can add it to the
6649 sources with a single patch command, do some testing, and check it in.
6650 If you leave out the @file{ChangeLog}, we have to write one. If you leave
6651 out the doc, we have to puzzle out what needs documenting. Etc., etc.
6653 The reason to send each change in a separate message is that we will not
6654 install some of the changes. They'll be returned to you with questions
6655 or comments. If we're doing our job correctly, the message back to you
6656 will say what you have to fix in order to make the change acceptable.
6657 The reason to have separate messages for separate features is so that
6658 the acceptable changes can be installed while one or more changes are
6659 being reworked. If multiple features are sent in a single message, we
6660 tend to not put in the effort to sort out the acceptable changes from
6661 the unacceptable, so none of the features get installed until all are
6664 If this sounds painful or authoritarian, well, it is. But we get a lot
6665 of bug reports and a lot of patches, and many of them don't get
6666 installed because we don't have the time to finish the job that the bug
6667 reporter or the contributor could have done. Patches that arrive
6668 complete, working, and well designed, tend to get installed on the day
6669 they arrive. The others go into a queue and get installed as time
6670 permits, which, since the maintainers have many demands to meet, may not
6671 be for quite some time.
6673 Please send patches directly to
6674 @email{gdb-patches@@sources.redhat.com, the @value{GDBN} maintainers}.
6676 @section Obsolete Conditionals
6677 @cindex obsolete code
6679 Fragments of old code in @value{GDBN} sometimes reference or set the following
6680 configuration macros. They should not be used by new code, and old uses
6681 should be removed as those parts of the debugger are otherwise touched.
6684 @item STACK_END_ADDR
6685 This macro used to define where the end of the stack appeared, for use
6686 in interpreting core file formats that don't record this address in the
6687 core file itself. This information is now configured in BFD, and @value{GDBN}
6688 gets the info portably from there. The values in @value{GDBN}'s configuration
6689 files should be moved into BFD configuration files (if needed there),
6690 and deleted from all of @value{GDBN}'s config files.
6692 Any @file{@var{foo}-xdep.c} file that references STACK_END_ADDR
6693 is so old that it has never been converted to use BFD. Now that's old!
6697 @include observer.texi