2 @setfilename gdbint.info
7 * Gdb-Internals: (gdbint). The GNU debugger's internals.
13 This file documents the internals of the GNU debugger GDB.
15 Copyright 1990-1999 Free Software Foundation, Inc.
16 Contributed by Cygnus Solutions. Written by John Gilmore.
17 Second Edition by Stan Shebs.
19 Permission is granted to make and distribute verbatim copies of this
20 manual provided the copyright notice and this permission notice are
21 preserved on all copies.
24 Permission is granted to process this file through Tex and print the
25 results, provided the printed document carries copying permission notice
26 identical to this one except for the removal of this paragraph (this
27 paragraph not being relevant to the printed manual).
30 Permission is granted to copy or distribute modified versions of this
31 manual under the terms of the GPL (for which purpose this text may be
32 regarded as a program in the language TeX).
35 @setchapternewpage off
36 @settitle GDB Internals
40 @subtitle{A guide to the internals of the GNU debugger}
42 @author Cygnus Solutions
43 @author Second Edition:
45 @author Cygnus Solutions
48 \def\$#1${{#1}} % Kluge: collect RCS revision info without $...$
49 \xdef\manvers{\$Revision$} % For use in headers, footers too
51 \hfill Cygnus Solutions\par
53 \hfill \TeX{}info \texinfoversion\par
57 @vskip 0pt plus 1filll
58 Copyright @copyright{} 1990-1999 Free Software Foundation, Inc.
60 Permission is granted to make and distribute verbatim copies of
61 this manual provided the copyright notice and this permission notice
62 are preserved on all copies.
67 @c Perhaps this should be the title of the document (but only for info,
68 @c not for TeX). Existing GNU manuals seem inconsistent on this point.
69 @top Scope of this Document
71 This document documents the internals of the GNU debugger, GDB. It
72 includes description of GDB's key algorithms and operations, as well
73 as the mechanisms that adapt GDB to specific hosts and targets.
83 * Target Architecture Definition::
84 * Target Vector Definition::
96 Before diving into the internals, you should understand the formal
97 requirements and other expectations for GDB. Although some of these may
98 seem obvious, there have been proposals for GDB that have run counter to
101 First of all, GDB is a debugger. It's not designed to be a front panel
102 for embedded systems. It's not a text editor. It's not a shell. It's
103 not a programming environment.
105 GDB is an interactive tool. Although a batch mode is available, GDB's
106 primary role is to interact with a human programmer.
108 GDB should be responsive to the user. A programmer hot on the trail of
109 a nasty bug, and operating under a looming deadline, is going to be very
110 impatient of everything, including the response time to debugger
113 GDB should be relatively permissive, such as for expressions. While the
114 compiler should be picky (or have the option to be made picky), since
115 source code lives for a long time usually, the programmer doing
116 debugging shouldn't be spending time figuring out to mollify the
119 GDB will be called upon to deal with really large programs. Executable
120 sizes of 50 to 100 megabytes occur regularly, and we've heard reports of
121 programs approaching 1 gigabyte in size.
123 GDB should be able to run everywhere. No other debugger is available
124 for even half as many configurations as GDB supports.
127 @node Overall Structure
129 @chapter Overall Structure
131 GDB consists of three major subsystems: user interface, symbol handling
132 (the ``symbol side''), and target system handling (the ``target side'').
134 Ther user interface consists of several actual interfaces, plus
137 The symbol side consists of object file readers, debugging info
138 interpreters, symbol table management, source language expression
139 parsing, type and value printing.
141 The target side consists of execution control, stack frame analysis, and
142 physical target manipulation.
144 The target side/symbol side division is not formal, and there are a
145 number of exceptions. For instance, core file support involves symbolic
146 elements (the basic core file reader is in BFD) and target elements (it
147 supplies the contents of memory and the values of registers). Instead,
148 this division is useful for understanding how the minor subsystems
151 @section The Symbol Side
153 The symbolic side of GDB can be thought of as ``everything you can do in
154 GDB without having a live program running''. For instance, you can look
155 at the types of variables, and evaluate many kinds of expressions.
157 @section The Target Side
159 The target side of GDB is the ``bits and bytes manipulator''. Although
160 it may make reference to symbolic info here and there, most of the
161 target side will run with only a stripped executable available -- or
162 even no executable at all, in remote debugging cases.
164 Operations such as disassembly, stack frame crawls, and register
165 display, are able to work with no symbolic info at all. In some cases,
166 such as disassembly, GDB will use symbolic info to present addresses
167 relative to symbols rather than as raw numbers, but it will work either
170 @section Configurations
172 @dfn{Host} refers to attributes of the system where GDB runs.
173 @dfn{Target} refers to the system where the program being debugged
174 executes. In most cases they are the same machine, in which case a
175 third type of @dfn{Native} attributes come into play.
177 Defines and include files needed to build on the host are host support.
178 Examples are tty support, system defined types, host byte order, host
181 Defines and information needed to handle the target format are target
182 dependent. Examples are the stack frame format, instruction set,
183 breakpoint instruction, registers, and how to set up and tear down the stack
186 Information that is only needed when the host and target are the same,
187 is native dependent. One example is Unix child process support; if the
188 host and target are not the same, doing a fork to start the target
189 process is a bad idea. The various macros needed for finding the
190 registers in the @code{upage}, running @code{ptrace}, and such are all
191 in the native-dependent files.
193 Another example of native-dependent code is support for features that
194 are really part of the target environment, but which require
195 @code{#include} files that are only available on the host system. Core
196 file handling and @code{setjmp} handling are two common cases.
198 When you want to make GDB work ``native'' on a particular machine, you
199 have to include all three kinds of information.
206 GDB uses a number of debugging-specific algorithms. They are often not
207 very complicated, but get lost in the thicket of special cases and
208 real-world issues. This chapter describes the basic algorithms and
209 mentions some of the specific target definitions that they use.
213 A frame is a construct that GDB uses to keep track of calling and called
216 @code{FRAME_FP} in the machine description has no meaning to the
217 machine-independent part of GDB, except that it is used when setting up
218 a new frame from scratch, as follows:
221 create_new_frame (read_register (FP_REGNUM), read_pc ()));
224 Other than that, all the meaning imparted to @code{FP_REGNUM} is
225 imparted by the machine-dependent code. So, @code{FP_REGNUM} can have
226 any value that is convenient for the code that creates new frames.
227 (@code{create_new_frame} calls @code{INIT_EXTRA_FRAME_INFO} if it is
228 defined; that is where you should use the @code{FP_REGNUM} value, if
229 your frames are nonstandard.)
231 Given a GDB frame, define @code{FRAME_CHAIN} to determine the address of
232 the calling function's frame. This will be used to create a new GDB
233 frame struct, and then @code{INIT_EXTRA_FRAME_INFO} and
234 @code{INIT_FRAME_PC} will be called for the new frame.
236 @section Breakpoint Handling
238 In general, a breakpoint is a user-designated location in the program
239 where the user wants to regain control if program execution ever reaches
242 There are two main ways to implement breakpoints; either as ``hardware''
243 breakpoints or as ``software'' breakpoints.
245 Hardware breakpoints are sometimes available as a builtin debugging
246 features with some chips. Typically these work by having dedicated
247 register into which the breakpoint address may be stored. If the PC
248 ever matches a value in a breakpoint registers, the CPU raises an
249 exception and reports it to GDB. Another possibility is when an
250 emulator is in use; many emulators include circuitry that watches the
251 address lines coming out from the processor, and force it to stop if the
252 address matches a breakpoint's address. A third possibility is that the
253 target already has the ability to do breakpoints somehow; for instance,
254 a ROM monitor may do its own software breakpoints. So although these
255 are not literally ``hardware breakpoints'', from GDB's point of view
256 they work the same; GDB need not do nothing more than set the breakpoint
257 and wait for something to happen.
259 Since they depend on hardware resources, hardware breakpoints may be
260 limited in number; when the user asks for more, GDB will start trying to
261 set software breakpoints.
263 Software breakpoints require GDB to do somewhat more work. The basic
264 theory is that GDB will replace a program instruction a trap, illegal
265 divide, or some other instruction that will cause an exception, and then
266 when it's encountered, GDB will take the exception and stop the program.
267 When the user says to continue, GDB will restore the original
268 instruction, single-step, re-insert the trap, and continue on.
270 Since it literally overwrites the program being tested, the program area
271 must be writeable, so this technique won't work on programs in ROM. It
272 can also distort the behavior of programs that examine themselves,
273 although the situation would be highly unusual.
275 Also, the software breakpoint instruction should be the smallest size of
276 instruction, so it doesn't overwrite an instruction that might be a jump
277 target, and cause disaster when the program jumps into the middle of the
278 breakpoint instruction. (Strictly speaking, the breakpoint must be no
279 larger than the smallest interval between instructions that may be jump
280 targets; perhaps there is an architecture where only even-numbered
281 instructions may jumped to.) Note that it's possible for an instruction
282 set not to have any instructions usable for a software breakpoint,
283 although in practice only the ARC has failed to define such an
286 The basic definition of the software breakpoint is the macro
289 Basic breakpoint object handling is in @file{breakpoint.c}. However,
290 much of the interesting breakpoint action is in @file{infrun.c}.
292 @section Single Stepping
294 @section Signal Handling
296 @section Thread Handling
298 @section Inferior Function Calls
300 @section Longjmp Support
302 GDB has support for figuring out that the target is doing a
303 @code{longjmp} and for stopping at the target of the jump, if we are
304 stepping. This is done with a few specialized internal breakpoints,
305 which are visible in the @code{maint info breakpoint} command.
307 To make this work, you need to define a macro called
308 @code{GET_LONGJMP_TARGET}, which will examine the @code{jmp_buf}
309 structure and extract the longjmp target address. Since @code{jmp_buf}
310 is target specific, you will need to define it in the appropriate
311 @file{tm-@var{xyz}.h} file. Look in @file{tm-sun4os4.h} and
312 @file{sparc-tdep.c} for examples of how to do this.
316 @chapter User Interface
318 GDB has several user interfaces. Although the command-line interface
319 is the most common and most familiar, there are others.
321 @section Command Interpreter
323 The command interpreter in GDB is fairly simple. It is designed to
324 allow for the set of commands to be augmented dynamically, and also
325 has a recursive subcommand capability, where the first argument to
326 a command may itself direct a lookup on a different command list.
328 For instance, the @code{set} command just starts a lookup on the
329 @code{setlist} command list, while @code{set thread} recurses
330 to the @code{set_thread_cmd_list}.
332 To add commands in general, use @code{add_cmd}. @code{add_com} adds to
333 the main command list, and should be used for those commands. The usual
334 place to add commands is in the @code{_initialize_@var{xyz}} routines at the
335 ends of most source files.
337 @section Console Printing
343 @code{libgdb} was an abortive project of years ago. The theory was to
344 provide an API to GDB's functionality.
346 @node Symbol Handling
348 @chapter Symbol Handling
350 Symbols are a key part of GDB's operation. Symbols include variables,
351 functions, and types.
353 @section Symbol Reading
355 GDB reads symbols from ``symbol files''. The usual symbol file is the
356 file containing the program which GDB is debugging. GDB can be directed
357 to use a different file for symbols (with the @code{symbol-file}
358 command), and it can also read more symbols via the ``add-file'' and
359 ``load'' commands, or while reading symbols from shared libraries.
361 Symbol files are initially opened by code in @file{symfile.c} using the
362 BFD library. BFD identifies the type of the file by examining its
363 header. @code{symfile_init} then uses this identification to locate a
364 set of symbol-reading functions.
366 Symbol reading modules identify themselves to GDB by calling
367 @code{add_symtab_fns} during their module initialization. The argument
368 to @code{add_symtab_fns} is a @code{struct sym_fns} which contains the
369 name (or name prefix) of the symbol format, the length of the prefix,
370 and pointers to four functions. These functions are called at various
371 times to process symbol-files whose identification matches the specified
374 The functions supplied by each module are:
377 @item @var{xyz}_symfile_init(struct sym_fns *sf)
379 Called from @code{symbol_file_add} when we are about to read a new
380 symbol file. This function should clean up any internal state (possibly
381 resulting from half-read previous files, for example) and prepare to
382 read a new symbol file. Note that the symbol file which we are reading
383 might be a new "main" symbol file, or might be a secondary symbol file
384 whose symbols are being added to the existing symbol table.
386 The argument to @code{@var{xyz}_symfile_init} is a newly allocated
387 @code{struct sym_fns} whose @code{bfd} field contains the BFD for the
388 new symbol file being read. Its @code{private} field has been zeroed,
389 and can be modified as desired. Typically, a struct of private
390 information will be @code{malloc}'d, and a pointer to it will be placed
391 in the @code{private} field.
393 There is no result from @code{@var{xyz}_symfile_init}, but it can call
394 @code{error} if it detects an unavoidable problem.
396 @item @var{xyz}_new_init()
398 Called from @code{symbol_file_add} when discarding existing symbols.
399 This function need only handle the symbol-reading module's internal
400 state; the symbol table data structures visible to the rest of GDB will
401 be discarded by @code{symbol_file_add}. It has no arguments and no
402 result. It may be called after @code{@var{xyz}_symfile_init}, if a new
403 symbol table is being read, or may be called alone if all symbols are
404 simply being discarded.
406 @item @var{xyz}_symfile_read(struct sym_fns *sf, CORE_ADDR addr, int mainline)
408 Called from @code{symbol_file_add} to actually read the symbols from a
409 symbol-file into a set of psymtabs or symtabs.
411 @code{sf} points to the struct sym_fns originally passed to
412 @code{@var{xyz}_sym_init} for possible initialization. @code{addr} is
413 the offset between the file's specified start address and its true
414 address in memory. @code{mainline} is 1 if this is the main symbol
415 table being read, and 0 if a secondary symbol file (e.g. shared library
416 or dynamically loaded file) is being read.@refill
419 In addition, if a symbol-reading module creates psymtabs when
420 @var{xyz}_symfile_read is called, these psymtabs will contain a pointer
421 to a function @code{@var{xyz}_psymtab_to_symtab}, which can be called
422 from any point in the GDB symbol-handling code.
425 @item @var{xyz}_psymtab_to_symtab (struct partial_symtab *pst)
427 Called from @code{psymtab_to_symtab} (or the PSYMTAB_TO_SYMTAB macro) if
428 the psymtab has not already been read in and had its @code{pst->symtab}
429 pointer set. The argument is the psymtab to be fleshed-out into a
430 symtab. Upon return, pst->readin should have been set to 1, and
431 pst->symtab should contain a pointer to the new corresponding symtab, or
432 zero if there were no symbols in that part of the symbol file.
435 @section Partial Symbol Tables
437 GDB has three types of symbol tables.
441 @item full symbol tables (symtabs). These contain the main information
442 about symbols and addresses.
444 @item partial symbol tables (psymtabs). These contain enough
445 information to know when to read the corresponding part of the full
448 @item minimal symbol tables (msymtabs). These contain information
449 gleaned from non-debugging symbols.
453 This section describes partial symbol tables.
455 A psymtab is constructed by doing a very quick pass over an executable
456 file's debugging information. Small amounts of information are
457 extracted -- enough to identify which parts of the symbol table will
458 need to be re-read and fully digested later, when the user needs the
459 information. The speed of this pass causes GDB to start up very
460 quickly. Later, as the detailed rereading occurs, it occurs in small
461 pieces, at various times, and the delay therefrom is mostly invisible to
463 @c (@xref{Symbol Reading}.)
465 The symbols that show up in a file's psymtab should be, roughly, those
466 visible to the debugger's user when the program is not running code from
467 that file. These include external symbols and types, static symbols and
468 types, and enum values declared at file scope.
470 The psymtab also contains the range of instruction addresses that the
471 full symbol table would represent.
473 The idea is that there are only two ways for the user (or much of the
474 code in the debugger) to reference a symbol:
479 (e.g. execution stops at some address which is inside a function in this
480 file). The address will be noticed to be in the range of this psymtab,
481 and the full symtab will be read in. @code{find_pc_function},
482 @code{find_pc_line}, and other @code{find_pc_@dots{}} functions handle
486 (e.g. the user asks to print a variable, or set a breakpoint on a
487 function). Global names and file-scope names will be found in the
488 psymtab, which will cause the symtab to be pulled in. Local names will
489 have to be qualified by a global name, or a file-scope name, in which
490 case we will have already read in the symtab as we evaluated the
491 qualifier. Or, a local symbol can be referenced when we are "in" a
492 local scope, in which case the first case applies. @code{lookup_symbol}
493 does most of the work here.
497 The only reason that psymtabs exist is to cause a symtab to be read in
498 at the right moment. Any symbol that can be elided from a psymtab,
499 while still causing that to happen, should not appear in it. Since
500 psymtabs don't have the idea of scope, you can't put local symbols in
501 them anyway. Psymtabs don't have the idea of the type of a symbol,
502 either, so types need not appear, unless they will be referenced by
505 It is a bug for GDB to behave one way when only a psymtab has been read,
506 and another way if the corresponding symtab has been read in. Such bugs
507 are typically caused by a psymtab that does not contain all the visible
508 symbols, or which has the wrong instruction address ranges.
510 The psymtab for a particular section of a symbol-file (objfile) could be
511 thrown away after the symtab has been read in. The symtab should always
512 be searched before the psymtab, so the psymtab will never be used (in a
513 bug-free environment). Currently, psymtabs are allocated on an obstack,
514 and all the psymbols themselves are allocated in a pair of large arrays
515 on an obstack, so there is little to be gained by trying to free them
516 unless you want to do a lot more work.
520 Fundamental Types (e.g., FT_VOID, FT_BOOLEAN).
522 These are the fundamental types that GDB uses internally. Fundamental
523 types from the various debugging formats (stabs, ELF, etc) are mapped
524 into one of these. They are basically a union of all fundamental types
525 that gdb knows about for all the languages that GDB knows about.
527 Type Codes (e.g., TYPE_CODE_PTR, TYPE_CODE_ARRAY).
529 Each time GDB builds an internal type, it marks it with one of these
530 types. The type may be a fundamental type, such as TYPE_CODE_INT, or a
531 derived type, such as TYPE_CODE_PTR which is a pointer to another type.
532 Typically, several FT_* types map to one TYPE_CODE_* type, and are
533 distinguished by other members of the type struct, such as whether the
534 type is signed or unsigned, and how many bits it uses.
536 Builtin Types (e.g., builtin_type_void, builtin_type_char).
538 These are instances of type structs that roughly correspond to
539 fundamental types and are created as global types for GDB to use for
540 various ugly historical reasons. We eventually want to eliminate these.
541 Note for example that builtin_type_int initialized in gdbtypes.c is
542 basically the same as a TYPE_CODE_INT type that is initialized in
543 c-lang.c for an FT_INTEGER fundamental type. The difference is that the
544 builtin_type is not associated with any particular objfile, and only one
545 instance exists, while c-lang.c builds as many TYPE_CODE_INT types as
546 needed, with each one associated with some particular objfile.
548 @section Object File Formats
552 The @file{a.out} format is the original file format for Unix. It
553 consists of three sections: text, data, and bss, which are for program
554 code, initialized data, and uninitialized data, respectively.
556 The @file{a.out} format is so simple that it doesn't have any reserved
557 place for debugging information. (Hey, the original Unix hackers used
558 @file{adb}, which is a machine-language debugger.) The only debugging
559 format for @file{a.out} is stabs, which is encoded as a set of normal
560 symbols with distinctive attributes.
562 The basic @file{a.out} reader is in @file{dbxread.c}.
566 The COFF format was introduced with System V Release 3 (SVR3) Unix.
567 COFF files may have multiple sections, each prefixed by a header. The
568 number of sections is limited.
570 The COFF specification includes support for debugging. Although this
571 was a step forward, the debugging information was woefully limited. For
572 instance, it was not possible to represent code that came from an
575 The COFF reader is in @file{coffread.c}.
579 ECOFF is an extended COFF originally introduced for Mips and Alpha
582 The basic ECOFF reader is in @file{mipsread.c}.
586 The IBM RS/6000 running AIX uses an object file format called XCOFF.
587 The COFF sections, symbols, and line numbers are used, but debugging
588 symbols are dbx-style stabs whose strings are located in the
589 @samp{.debug} section (rather than the string table). For more
590 information, see @xref{Top,,,stabs,The Stabs Debugging Format}.
592 The shared library scheme has a clean interface for figuring out what
593 shared libraries are in use, but the catch is that everything which
594 refers to addresses (symbol tables and breakpoints at least) needs to be
595 relocated for both shared libraries and the main executable. At least
596 using the standard mechanism this can only be done once the program has
597 been run (or the core file has been read).
601 Windows 95 and NT use the PE (Portable Executable) format for their
602 executables. PE is basically COFF with additional headers.
604 While BFD includes special PE support, GDB needs only the basic
609 The ELF format came with System V Release 4 (SVR4) Unix. ELF is similar
610 to COFF in being organized into a number of sections, but it removes
611 many of COFF's limitations.
613 The basic ELF reader is in @file{elfread.c}.
617 SOM is HP's object file and debug format (not to be confused with IBM's
618 SOM, which is a cross-language ABI).
620 The SOM reader is in @file{hpread.c}.
622 @subsection Other File Formats
624 Other file formats that have been supported by GDB include Netware
625 Loadable Modules (@file{nlmread.c}.
627 @section Debugging File Formats
629 This section describes characteristics of debugging information that
630 are independent of the object file format.
634 @code{stabs} started out as special symbols within the @code{a.out}
635 format. Since then, it has been encapsulated into other file
636 formats, such as COFF and ELF.
638 While @file{dbxread.c} does some of the basic stab processing,
639 including for encapsulated versions, @file{stabsread.c} does
644 The basic COFF definition includes debugging information. The level
645 of support is minimal and non-extensible, and is not often used.
647 @subsection Mips debug (Third Eye)
649 ECOFF includes a definition of a special debug format.
651 The file @file{mdebugread.c} implements reading for this format.
655 DWARF 1 is a debugging format that was originally designed to be
656 used with ELF in SVR4 systems.
662 @c If defined, these are the producer strings in a DWARF 1 file. All of
663 @c these have reasonable defaults already.
665 The DWARF 1 reader is in @file{dwarfread.c}.
669 DWARF 2 is an improved but incompatible version of DWARF 1.
671 The DWARF 2 reader is in @file{dwarf2read.c}.
675 Like COFF, the SOM definition includes debugging information.
677 @section Adding a New Symbol Reader to GDB
679 If you are using an existing object file format (a.out, COFF, ELF, etc),
680 there is probably little to be done.
682 If you need to add a new object file format, you must first add it to
683 BFD. This is beyond the scope of this document.
685 You must then arrange for the BFD code to provide access to the
686 debugging symbols. Generally GDB will have to call swapping routines
687 from BFD and a few other BFD internal routines to locate the debugging
688 information. As much as possible, GDB should not depend on the BFD
689 internal data structures.
691 For some targets (e.g., COFF), there is a special transfer vector used
692 to call swapping routines, since the external data structures on various
693 platforms have different sizes and layouts. Specialized routines that
694 will only ever be implemented by one object file format may be called
695 directly. This interface should be described in a file
696 @file{bfd/libxyz.h}, which is included by GDB.
699 @node Language Support
701 @chapter Language Support
703 GDB's language support is mainly driven by the symbol reader, although
704 it is possible for the user to set the source language manually.
706 GDB chooses the source language by looking at the extension of the file
707 recorded in the debug info; @code{.c} means C, @code{.f} means Fortran,
708 etc. It may also use a special-purpose language identifier if the debug
709 format supports it, such as DWARF.
711 @section Adding a Source Language to GDB
713 To add other languages to GDB's expression parser, follow the following
717 @item Create the expression parser.
719 This should reside in a file @file{@var{lang}-exp.y}. Routines for
720 building parsed expressions into a @samp{union exp_element} list are in
723 Since we can't depend upon everyone having Bison, and YACC produces
724 parsers that define a bunch of global names, the following lines
725 @emph{must} be included at the top of the YACC parser, to prevent the
726 various parsers from defining the same global names:
729 #define yyparse @var{lang}_parse
730 #define yylex @var{lang}_lex
731 #define yyerror @var{lang}_error
732 #define yylval @var{lang}_lval
733 #define yychar @var{lang}_char
734 #define yydebug @var{lang}_debug
735 #define yypact @var{lang}_pact
736 #define yyr1 @var{lang}_r1
737 #define yyr2 @var{lang}_r2
738 #define yydef @var{lang}_def
739 #define yychk @var{lang}_chk
740 #define yypgo @var{lang}_pgo
741 #define yyact @var{lang}_act
742 #define yyexca @var{lang}_exca
743 #define yyerrflag @var{lang}_errflag
744 #define yynerrs @var{lang}_nerrs
747 At the bottom of your parser, define a @code{struct language_defn} and
748 initialize it with the right values for your language. Define an
749 @code{initialize_@var{lang}} routine and have it call
750 @samp{add_language(@var{lang}_language_defn)} to tell the rest of GDB
751 that your language exists. You'll need some other supporting variables
752 and functions, which will be used via pointers from your
753 @code{@var{lang}_language_defn}. See the declaration of @code{struct
754 language_defn} in @file{language.h}, and the other @file{*-exp.y} files,
755 for more information.
757 @item Add any evaluation routines, if necessary
759 If you need new opcodes (that represent the operations of the language),
760 add them to the enumerated type in @file{expression.h}. Add support
761 code for these operations in @code{eval.c:evaluate_subexp()}. Add cases
762 for new opcodes in two functions from @file{parse.c}:
763 @code{prefixify_subexp()} and @code{length_of_subexp()}. These compute
764 the number of @code{exp_element}s that a given operation takes up.
766 @item Update some existing code
768 Add an enumerated identifier for your language to the enumerated type
769 @code{enum language} in @file{defs.h}.
771 Update the routines in @file{language.c} so your language is included.
772 These routines include type predicates and such, which (in some cases)
773 are language dependent. If your language does not appear in the switch
774 statement, an error is reported.
776 Also included in @file{language.c} is the code that updates the variable
777 @code{current_language}, and the routines that translate the
778 @code{language_@var{lang}} enumerated identifier into a printable
781 Update the function @code{_initialize_language} to include your
782 language. This function picks the default language upon startup, so is
783 dependent upon which languages that GDB is built for.
785 Update @code{allocate_symtab} in @file{symfile.c} and/or symbol-reading
786 code so that the language of each symtab (source file) is set properly.
787 This is used to determine the language to use at each stack frame level.
788 Currently, the language is set based upon the extension of the source
789 file. If the language can be better inferred from the symbol
790 information, please set the language of the symtab in the symbol-reading
793 Add helper code to @code{expprint.c:print_subexp()} to handle any new
794 expression opcodes you have added to @file{expression.h}. Also, add the
795 printed representations of your operators to @code{op_print_tab}.
797 @item Add a place of call
799 Add a call to @code{@var{lang}_parse()} and @code{@var{lang}_error} in
800 @code{parse.c:parse_exp_1()}.
802 @item Use macros to trim code
804 The user has the option of building GDB for some or all of the
805 languages. If the user decides to build GDB for the language
806 @var{lang}, then every file dependent on @file{language.h} will have the
807 macro @code{_LANG_@var{lang}} defined in it. Use @code{#ifdef}s to
808 leave out large routines that the user won't need if he or she is not
811 Note that you do not need to do this in your YACC parser, since if GDB
812 is not build for @var{lang}, then @file{@var{lang}-exp.tab.o} (the
813 compiled form of your parser) is not linked into GDB at all.
815 See the file @file{configure.in} for how GDB is configured for different
818 @item Edit @file{Makefile.in}
820 Add dependencies in @file{Makefile.in}. Make sure you update the macro
821 variables such as @code{HFILES} and @code{OBJS}, otherwise your code may
822 not get linked in, or, worse yet, it may not get @code{tar}red into the
828 @node Host Definition
830 @chapter Host Definition
832 With the advent of autoconf, it's rarely necessary to have host
833 definition machinery anymore.
835 @section Adding a New Host
837 Most of GDB's host configuration support happens via autoconf. It
838 should be rare to need new host-specific definitions. GDB still uses
839 the host-specific definitions and files listed below, but these mostly
840 exist for historical reasons, and should eventually disappear.
842 Several files control GDB's configuration for host systems:
846 @item gdb/config/@var{arch}/@var{xyz}.mh
847 Specifies Makefile fragments needed when hosting on machine @var{xyz}.
848 In particular, this lists the required machine-dependent object files,
849 by defining @samp{XDEPFILES=@dots{}}. Also specifies the header file
850 which describes host @var{xyz}, by defining @code{XM_FILE=
851 xm-@var{xyz}.h}. You can also define @code{CC}, @code{SYSV_DEFINE},
852 @code{XM_CFLAGS}, @code{XM_ADD_FILES}, @code{XM_CLIBS}, @code{XM_CDEPS},
853 etc.; see @file{Makefile.in}.
855 @item gdb/config/@var{arch}/xm-@var{xyz}.h
856 (@file{xm.h} is a link to this file, created by configure). Contains C
857 macro definitions describing the host system environment, such as byte
858 order, host C compiler and library.
860 @item gdb/@var{xyz}-xdep.c
861 Contains any miscellaneous C code required for this machine as a host.
862 On most machines it doesn't exist at all. If it does exist, put
863 @file{@var{xyz}-xdep.o} into the @code{XDEPFILES} line in
864 @file{gdb/config/@var{arch}/@var{xyz}.mh}.
868 @subheading Generic Host Support Files
870 There are some ``generic'' versions of routines that can be used by
871 various systems. These can be customized in various ways by macros
872 defined in your @file{xm-@var{xyz}.h} file. If these routines work for
873 the @var{xyz} host, you can just include the generic file's name (with
874 @samp{.o}, not @samp{.c}) in @code{XDEPFILES}.
876 Otherwise, if your machine needs custom support routines, you will need
877 to write routines that perform the same functions as the generic file.
878 Put them into @code{@var{xyz}-xdep.c}, and put @code{@var{xyz}-xdep.o}
879 into @code{XDEPFILES}.
884 This contains serial line support for Unix systems. This is always
885 included, via the makefile variable @code{SER_HARDWIRE}; override this
886 variable in the @file{.mh} file to avoid it.
889 This contains serial line support for 32-bit programs running under DOS,
890 using the GO32 execution environment.
893 This contains generic TCP support using sockets.
897 @section Host Conditionals
899 When GDB is configured and compiled, various macros are defined or left
900 undefined, to control compilation based on the attributes of the host
901 system. These macros and their meanings (or if the meaning is not
902 documented here, then one of the source files where they are used is
907 @item GDBINIT_FILENAME
908 The default name of GDB's initialization file (normally @file{.gdbinit}).
910 @item MEM_FNS_DECLARED
911 Your host config file defines this if it includes declarations of
912 @code{memcpy} and @code{memset}. Define this to avoid conflicts between
913 the native include files and the declarations in @file{defs.h}.
916 Define this if your system does not have a @code{<sys/file.h>}.
918 @item SIGWINCH_HANDLER
919 If your host defines @code{SIGWINCH}, you can define this to be the name
920 of a function to be called if @code{SIGWINCH} is received.
922 @item SIGWINCH_HANDLER_BODY
923 Define this to expand into code that will define the function named by
924 the expansion of @code{SIGWINCH_HANDLER}.
926 @item ALIGN_STACK_ON_STARTUP
927 Define this if your system is of a sort that will crash in
928 @code{tgetent} if the stack happens not to be longword-aligned when
929 @code{main} is called. This is a rare situation, but is known to occur
930 on several different types of systems.
932 @item CRLF_SOURCE_FILES
933 Define this if host files use @code{\r\n} rather than @code{\n} as a
934 line terminator. This will cause source file listings to omit @code{\r}
935 characters when printing and it will allow \r\n line endings of files
936 which are "sourced" by gdb. It must be possible to open files in binary
937 mode using @code{O_BINARY} or, for fopen, @code{"rb"}.
940 The default value of the prompt string (normally @code{"(gdb) "}).
943 The name of the generic TTY device, defaults to @code{"/dev/tty"}.
945 @item FCLOSE_PROVIDED
946 Define this if the system declares @code{fclose} in the headers included
947 in @code{defs.h}. This isn't needed unless your compiler is unusually
951 Define this if binary files are opened the same way as text files.
953 @item GETENV_PROVIDED
954 Define this if the system declares @code{getenv} in its headers included
955 in @code{defs.h}. This isn't needed unless your compiler is unusually
959 In some cases, use the system call @code{mmap} for reading symbol
960 tables. For some machines this allows for sharing and quick updates.
962 @item HAVE_SIGSETMASK
963 Define this if the host system has job control, but does not define
964 @code{sigsetmask()}. Currently, this is only true of the RS/6000.
967 Define this if the host system has @code{termio.h}.
969 @item HOST_BYTE_ORDER
970 The ordering of bytes in the host. This must be defined to be either
971 @code{BIG_ENDIAN} or @code{LITTLE_ENDIAN}.
978 Values for host-side constants.
981 Substitute for isatty, if not available.
984 This is the longest integer type available on the host. If not defined,
985 it will default to @code{long long} or @code{long}, depending on
986 @code{CC_HAS_LONG_LONG}.
988 @item CC_HAS_LONG_LONG
989 Define this if the host C compiler supports ``long long''. This is set
990 by the configure script.
992 @item PRINTF_HAS_LONG_LONG
993 Define this if the host can handle printing of long long integers via
994 the printf format directive ``ll''. This is set by the configure script.
996 @item HAVE_LONG_DOUBLE
997 Define this if the host C compiler supports ``long double''. This is
998 set by the configure script.
1000 @item PRINTF_HAS_LONG_DOUBLE
1001 Define this if the host can handle printing of long double float-point
1002 numbers via the printf format directive ``Lg''. This is set by the
1005 @item SCANF_HAS_LONG_DOUBLE
1006 Define this if the host can handle the parsing of long double
1007 float-point numbers via the scanf format directive directive
1008 ``Lg''. This is set by the configure script.
1010 @item LSEEK_NOT_LINEAR
1011 Define this if @code{lseek (n)} does not necessarily move to byte number
1012 @code{n} in the file. This is only used when reading source files. It
1013 is normally faster to define @code{CRLF_SOURCE_FILES} when possible.
1016 This macro is used as the argument to lseek (or, most commonly,
1017 bfd_seek). FIXME, should be replaced by SEEK_SET instead, which is the
1020 @item MALLOC_INCOMPATIBLE
1021 Define this if the system's prototype for @code{malloc} differs from the
1022 @sc{ANSI} definition.
1024 @item MMAP_BASE_ADDRESS
1025 When using HAVE_MMAP, the first mapping should go at this address.
1027 @item MMAP_INCREMENT
1028 when using HAVE_MMAP, this is the increment between mappings.
1030 @item NEED_POSIX_SETPGID
1031 Define this to use the POSIX version of @code{setpgid} to determine
1032 whether job control is available.
1035 If defined, this should be one or more tokens, such as @code{volatile},
1036 that can be used in both the declaration and definition of functions to
1037 indicate that they never return. The default is already set correctly
1038 if compiling with GCC. This will almost never need to be defined.
1041 If defined, this should be one or more tokens, such as
1042 @code{__attribute__ ((noreturn))}, that can be used in the declarations
1043 of functions to indicate that they never return. The default is already
1044 set correctly if compiling with GCC. This will almost never need to be
1047 @item USE_GENERIC_DUMMY_FRAMES
1048 Define this to 1 if the target is using the generic inferior function
1049 call code. See @code{blockframe.c} for more information.
1052 GDB will use the @code{mmalloc} library for memory allocation for symbol
1053 reading if this symbol is defined. Be careful defining it since there
1054 are systems on which @code{mmalloc} does not work for some reason. One
1055 example is the DECstation, where its RPC library can't cope with our
1056 redefinition of @code{malloc} to call @code{mmalloc}. When defining
1057 @code{USE_MMALLOC}, you will also have to set @code{MMALLOC} in the
1058 Makefile, to point to the mmalloc library. This define is set when you
1059 configure with --with-mmalloc.
1062 Define this if you are using @code{mmalloc}, but don't want the overhead
1063 of checking the heap with @code{mmcheck}. Note that on some systems,
1064 the C runtime makes calls to malloc prior to calling @code{main}, and if
1065 @code{free} is ever called with these pointers after calling
1066 @code{mmcheck} to enable checking, a memory corruption abort is certain
1067 to occur. These systems can still use mmalloc, but must define
1071 Define this to 1 if the C runtime allocates memory prior to
1072 @code{mmcheck} being called, but that memory is never freed so we don't
1073 have to worry about it triggering a memory corruption abort. The
1074 default is 0, which means that @code{mmcheck} will only install the heap
1075 checking functions if there has not yet been any memory allocation
1076 calls, and if it fails to install the functions, gdb will issue a
1077 warning. This is currently defined if you configure using
1080 @item NO_SIGINTERRUPT
1081 Define this to indicate that siginterrupt() is not available.
1084 Define if this is not in a system .h file.
1088 Define these to appropriate value for the system lseek(), if not already
1092 This is the signal for stopping GDB. Defaults to SIGTSTP. (Only
1093 redefined for the Convex.)
1096 Define this if the interior's tty should be opened with the O_NOCTTY
1097 flag. (FIXME: This should be a native-only flag, but @file{inflow.c} is
1101 Means that System V (prior to SVR4) include files are in use. (FIXME:
1102 This symbol is abused in @file{infrun.c}, @file{regex.c},
1103 @file{remote-nindy.c}, and @file{utils.c} for other things, at the
1107 Define this to help placate lint in some situations.
1110 Define this to override the defaults of @code{__volatile__} or
1116 @node Target Architecture Definition
1118 @chapter Target Architecture Definition
1120 GDB's target architecture defines what sort of machine-language programs
1121 GDB can work with, and how it works with them.
1123 At present, the target architecture definition consists of a number of C
1126 @section Registers and Memory
1128 GDB's model of the target machine is rather simple. GDB assumes the
1129 machine includes a bank of registers and a block of memory. Each
1130 register may have a different size.
1132 GDB does not have a magical way to match up with the compiler's idea of
1133 which registers are which; however, it is critical that they do match up
1134 accurately. The only way to make this work is to get accurate
1135 information about the order that the compiler uses, and to reflect that
1136 in the @code{REGISTER_NAME} and related macros.
1138 GDB can handle big-endian, little-endian, and bi-endian architectures.
1140 @section Frame Interpretation
1142 @section Inferior Call Setup
1144 @section Compiler Characteristics
1146 @section Target Conditionals
1148 This section describes the macros that you can use to define the target
1153 @item ADDITIONAL_OPTIONS
1154 @item ADDITIONAL_OPTION_CASES
1155 @item ADDITIONAL_OPTION_HANDLER
1156 @item ADDITIONAL_OPTION_HELP
1157 These are a set of macros that allow the addition of additional command
1158 line options to GDB. They are currently used only for the unsupported
1159 i960 Nindy target, and should not be used in any other configuration.
1161 @item ADDR_BITS_REMOVE (addr)
1162 If a raw machine address includes any bits that are not really part of
1163 the address, then define this macro to expand into an expression that
1164 zeros those bits in @var{addr}. For example, the two low-order bits of
1165 a Motorola 88K address may be used by some kernels for their own
1166 purposes, since addresses must always be 4-byte aligned, and so are of
1167 no use for addressing. Those bits should be filtered out with an
1168 expression such as @code{((addr) & ~3)}.
1170 @item BEFORE_MAIN_LOOP_HOOK
1171 Define this to expand into any code that you want to execute before the
1172 main loop starts. Although this is not, strictly speaking, a target
1173 conditional, that is how it is currently being used. Note that if a
1174 configuration were to define it one way for a host and a different way
1175 for the target, GDB will probably not compile, let alone run correctly.
1176 This is currently used only for the unsupported i960 Nindy target, and
1177 should not be used in any other configuration.
1179 @item BELIEVE_PCC_PROMOTION
1180 Define if the compiler promotes a short or char parameter to an int, but
1181 still reports the parameter as its original type, rather than the
1184 @item BELIEVE_PCC_PROMOTION_TYPE
1185 Define this if GDB should believe the type of a short argument when
1186 compiled by pcc, but look within a full int space to get its value.
1187 Only defined for Sun-3 at present.
1189 @item BITS_BIG_ENDIAN
1190 Define this if the numbering of bits in the targets does *not* match the
1191 endianness of the target byte order. A value of 1 means that the bits
1192 are numbered in a big-endian order, 0 means little-endian.
1195 This is the character array initializer for the bit pattern to put into
1196 memory where a breakpoint is set. Although it's common to use a trap
1197 instruction for a breakpoint, it's not required; for instance, the bit
1198 pattern could be an invalid instruction. The breakpoint must be no
1199 longer than the shortest instruction of the architecture.
1201 @var{BREAKPOINT} has been deprecated in favour of
1202 @var{BREAKPOINT_FROM_PC}.
1204 @item BIG_BREAKPOINT
1205 @item LITTLE_BREAKPOINT
1206 Similar to BREAKPOINT, but used for bi-endian targets.
1208 @var{BIG_BREAKPOINT} and @var{LITTLE_BREAKPOINT} have been deprecated in
1209 favour of @var{BREAKPOINT_FROM_PC}.
1211 @item REMOTE_BREAKPOINT
1212 @item LITTLE_REMOTE_BREAKPOINT
1213 @item BIG_REMOTE_BREAKPOINT
1214 Similar to BREAKPOINT, but used for remote targets.
1216 @var{BIG_REMOTE_BREAKPOINT} and @var{LITTLE_REMOTE_BREAKPOINT} have been
1217 deprecated in favour of @var{BREAKPOINT_FROM_PC}.
1219 @item BREAKPOINT_FROM_PC (pcptr, lenptr)
1221 Use the program counter to determine the contents and size of a
1222 breakpoint instruction. It returns a pointer to a string of bytes that
1223 encode a breakpoint instruction, stores the length of the string to
1224 *lenptr, and adjusts pc (if necessary) to point to the actual memory
1225 location where the breakpoint should be inserted.
1227 Although it is common to use a trap instruction for a breakpoint, it's
1228 not required; for instance, the bit pattern could be an invalid
1229 instruction. The breakpoint must be no longer than the shortest
1230 instruction of the architecture.
1232 Replaces all the other @var{BREAKPOINT} macros.
1235 A C expresson that is non-zero when the target suports inferior function
1238 @item CALL_DUMMY_WORDS
1239 Pointer to an array of @var{LONGEST} words of data containing
1240 host-byte-ordered @var{REGISTER_BYTES} sized values that partially
1241 specify the sequence of instructions needed for an inferior function
1244 Should be deprecated in favour of a macro that uses target-byte-ordered
1247 @item SIZEOF_CALL_DUMMY_WORDS
1248 The size of @var{CALL_DUMMY_WORDS}. When @var{CALL_DUMMY_P} this must
1249 return a positive value. See also @var{CALL_DUMMY_LENGTH}.
1252 A static initializer for @var{CALL_DUMMY_WORDS}. Deprecated.
1254 @item CALL_DUMMY_LOCATION
1257 @item CALL_DUMMY_STACK_ADJUST
1258 Stack adjustment needed when performing an inferior function call.
1260 Should be deprecated in favor of something like @var{STACK_ALIGN}.
1262 @item CALL_DUMMY_STACK_ADJUST_P
1263 Predicate for use of @var{CALL_DUMMY_STACK_ADJUST}.
1265 Should be deprecated in favor of something like @var{STACK_ALIGN}.
1267 @item CANNOT_FETCH_REGISTER (regno)
1268 A C expression that should be nonzero if @var{regno} cannot be fetched
1269 from an inferior process. This is only relevant if
1270 @code{FETCH_INFERIOR_REGISTERS} is not defined.
1272 @item CANNOT_STORE_REGISTER (regno)
1273 A C expression that should be nonzero if @var{regno} should not be
1274 written to the target. This is often the case for program counters,
1275 status words, and other special registers. If this is not defined, GDB
1276 will assume that all registers may be written.
1278 @item DO_DEFERRED_STORES
1279 @item CLEAR_DEFERRED_STORES
1280 Define this to execute any deferred stores of registers into the inferior,
1281 and to cancel any deferred stores.
1283 Currently only implemented correctly for native Sparc configurations?
1286 Define this to expand into the character that G++ uses to distinguish
1287 compiler-generated identifiers from programmer-specified identifiers.
1288 By default, this expands into @code{'$'}. Most System V targets should
1289 define this to @code{'.'}.
1291 @item DBX_PARM_SYMBOL_CLASS
1292 Hook for the @code{SYMBOL_CLASS} of a parameter when decoding DBX symbol
1293 information. In the i960, parameters can be stored as locals or as
1294 args, depending on the type of the debug record.
1296 @item DECR_PC_AFTER_BREAK
1297 Define this to be the amount by which to decrement the PC after the
1298 program encounters a breakpoint. This is often the number of bytes in
1299 BREAKPOINT, though not always. For most targets this value will be 0.
1301 @item DECR_PC_AFTER_HW_BREAK
1302 Similarly, for hardware breakpoints.
1304 @item DISABLE_UNSETTABLE_BREAK addr
1305 If defined, this should evaluate to 1 if @var{addr} is in a shared
1306 library in which breakpoints cannot be set and so should be disabled.
1308 @item DO_REGISTERS_INFO
1309 If defined, use this to print the value of a register or all registers.
1311 @item END_OF_TEXT_DEFAULT
1312 This is an expression that should designate the end of the text section
1315 @item EXTRACT_RETURN_VALUE(type,regbuf,valbuf)
1316 Define this to extract a function's return value of type @var{type} from
1317 the raw register state @var{regbuf} and copy that, in virtual format,
1320 @item EXTRACT_STRUCT_VALUE_ADDRESS(regbuf)
1321 When @var{EXTRACT_STRUCT_VALUE_ADDRESS_P} this is used to to extract
1322 from an array @var{regbuf} (containing the raw register state) the
1323 address in which a function should return its structure value, as a
1324 CORE_ADDR (or an expression that can be used as one).
1326 @item EXTRACT_STRUCT_VALUE_ADDRESS_P
1327 Predicate for @var{EXTRACT_STRUCT_VALUE_ADDRESS}.
1330 If defined, then the `info float' command will print information about
1331 the processor's floating point unit.
1334 The number of the frame pointer register.
1336 @item FRAMELESS_FUNCTION_INVOCATION(fi)
1337 Define this to an expression that returns 1 if the function invocation
1338 represented by @var{fi} does not have a stack frame associated with it.
1341 @item FRAME_ARGS_ADDRESS_CORRECT
1344 @item FRAME_CHAIN(frame)
1345 Given @var{frame}, return a pointer to the calling frame.
1347 @item FRAME_CHAIN_COMBINE(chain,frame)
1348 Define this to take the frame chain pointer and the frame's nominal
1349 address and produce the nominal address of the caller's frame.
1350 Presently only defined for HP PA.
1352 @item FRAME_CHAIN_VALID(chain,thisframe)
1354 Define this to be an expression that returns zero if the given frame is
1355 an outermost frame, with no caller, and nonzero otherwise. Three common
1356 definitions are available. @code{default_frame_chain_valid} (the
1357 default) is nonzero if the chain pointer is nonzero and given frame's PC
1358 is not inside the startup file (such as @file{crt0.o}).
1359 @code{alternate_frame_chain_valid} is nonzero if the chain pointer is
1360 nonzero and the given frame's PC is not in @code{main()} or a known
1361 entry point function (such as @code{_start()}).
1363 @item FRAME_INIT_SAVED_REGS(frame)
1364 See @file{frame.h}. Determines the address of all registers in the
1365 current stack frame storing each in @code{frame->saved_regs}. Space for
1366 @code{frame->saved_regs} shall be allocated by
1367 @code{FRAME_INIT_SAVED_REGS} using either
1368 @code{frame_saved_regs_zalloc} or @code{frame_obstack_alloc}.
1370 @var{FRAME_FIND_SAVED_REGS} and @var{EXTRA_FRAME_INFO} are deprecated.
1372 @item FRAME_NUM_ARGS (fi)
1373 For the frame described by @var{fi} return the number of arguments that
1374 are being passed. If the number of arguments is not known, return
1377 @item FRAME_SAVED_PC(frame)
1378 Given @var{frame}, return the pc saved there. That is, the return
1381 @item FUNCTION_EPILOGUE_SIZE
1382 For some COFF targets, the @code{x_sym.x_misc.x_fsize} field of the
1383 function end symbol is 0. For such targets, you must define
1384 @code{FUNCTION_EPILOGUE_SIZE} to expand into the standard size of a
1385 function's epilogue.
1387 @item GCC_COMPILED_FLAG_SYMBOL
1388 @item GCC2_COMPILED_FLAG_SYMBOL
1389 If defined, these are the names of the symbols that GDB will look for to
1390 detect that GCC compiled the file. The default symbols are
1391 @code{gcc_compiled.} and @code{gcc2_compiled.}, respectively. (Currently
1392 only defined for the Delta 68.)
1394 @item GDB_TARGET_IS_HPPA
1395 This determines whether horrible kludge code in dbxread.c and
1396 partial-stab.h is used to mangle multiple-symbol-table files from
1397 HPPA's. This should all be ripped out, and a scheme like elfread.c
1400 @item GDB_TARGET_IS_MACH386
1401 @item GDB_TARGET_IS_SUN3
1402 @item GDB_TARGET_IS_SUN386
1403 Kludges that should go away.
1405 @item GET_LONGJMP_TARGET
1406 For most machines, this is a target-dependent parameter. On the
1407 DECstation and the Iris, this is a native-dependent parameter, since
1408 <setjmp.h> is needed to define it.
1410 This macro determines the target PC address that longjmp() will jump to,
1411 assuming that we have just stopped at a longjmp breakpoint. It takes a
1412 CORE_ADDR * as argument, and stores the target PC value through this
1413 pointer. It examines the current state of the machine as needed.
1415 @item GET_SAVED_REGISTER
1416 Define this if you need to supply your own definition for the function
1417 @code{get_saved_register}.
1419 @item HAVE_REGISTER_WINDOWS
1420 Define this if the target has register windows.
1421 @item REGISTER_IN_WINDOW_P (regnum)
1422 Define this to be an expression that is 1 if the given register is in
1425 @item IBM6000_TARGET
1426 Shows that we are configured for an IBM RS/6000 target. This
1427 conditional should be eliminated (FIXME) and replaced by
1428 feature-specific macros. It was introduced in haste and we are
1429 repenting at leisure.
1432 Define this if the target system uses IEEE-format floating point numbers.
1434 @item INIT_EXTRA_FRAME_INFO (fromleaf, frame)
1435 If additional information about the frame is required this should be
1436 stored in @code{frame->extra_info}. Space for @code{frame->extra_info}
1437 is allocated using @code{frame_obstack_alloc}.
1439 @item INIT_FRAME_PC (fromleaf, prev)
1440 This is a C statement that sets the pc of the frame pointed to by
1441 @var{prev}. [By default...]
1443 @item INNER_THAN (lhs,rhs)
1444 Returns non-zero if stack address @var{lhs} is inner than (nearer to the
1445 stack top) stack address @var{rhs}. Define this as @code{lhs < rhs} if
1446 the target's stack grows downward in memory, or @code{lhs > rsh} if the
1449 @item IN_SIGTRAMP (pc, name)
1450 Define this to return true if the given @var{pc} and/or @var{name}
1451 indicates that the current function is a sigtramp.
1453 @item SIGTRAMP_START (pc)
1454 @item SIGTRAMP_END (pc)
1455 Define these to be the start and end address of the sigtramp for the
1456 given @var{pc}. On machines where the address is just a compile time
1457 constant, the macro expansion will typically just ignore the supplied
1460 @item IN_SOLIB_CALL_TRAMPOLINE pc name
1461 Define this to evaluate to nonzero if the program is stopped in the
1462 trampoline that connects to a shared library.
1464 @item IN_SOLIB_RETURN_TRAMPOLINE pc name
1465 Define this to evaluate to nonzero if the program is stopped in the
1466 trampoline that returns from a shared library.
1468 @item IS_TRAPPED_INTERNALVAR (name)
1469 This is an ugly hook to allow the specification of special actions that
1470 should occur as a side-effect of setting the value of a variable
1471 internal to GDB. Currently only used by the h8500. Note that this
1472 could be either a host or target conditional.
1474 @item NEED_TEXT_START_END
1475 Define this if GDB should determine the start and end addresses of the
1476 text section. (Seems dubious.)
1478 @item NO_HIF_SUPPORT
1479 (Specific to the a29k.)
1481 @item SOFTWARE_SINGLE_STEP_P
1482 Define this as 1 if the target does not have a hardware single-step
1483 mechanism. The macro @code{SOFTWARE_SINGLE_STEP} must also be defined.
1485 @item SOFTWARE_SINGLE_STEP(signal,insert_breapoints_p)
1486 A function that inserts or removes (dependant on
1487 @var{insert_breapoints_p}) breakpoints at each possible destinations of
1488 the next instruction. See @code{sparc-tdep.c} and @code{rs6000-tdep.c}
1491 @item PCC_SOL_BROKEN
1492 (Used only in the Convex target.)
1494 @item PC_IN_CALL_DUMMY
1497 @item PC_LOAD_SEGMENT
1498 If defined, print information about the load segment for the program
1499 counter. (Defined only for the RS/6000.)
1502 If the program counter is kept in a register, then define this macro to
1503 be the number of that register. This need be defined only if
1504 @code{TARGET_WRITE_PC} is not defined.
1507 The number of the ``next program counter'' register, if defined.
1510 The number of the ``next next program counter'' register, if defined.
1511 Currently, this is only defined for the Motorola 88K.
1513 @item PRINT_REGISTER_HOOK (regno)
1514 If defined, this must be a function that prints the contents of the
1515 given register to standard output.
1517 @item PRINT_TYPELESS_INTEGER
1518 This is an obscure substitute for @code{print_longest} that seems to
1519 have been defined for the Convex target.
1521 @item PROCESS_LINENUMBER_HOOK
1522 A hook defined for XCOFF reading.
1524 @item PROLOGUE_FIRSTLINE_OVERLAP
1525 (Only used in unsupported Convex configuration.)
1528 If defined, this is the number of the processor status register. (This
1529 definition is only used in generic code when parsing "$ps".)
1532 Used in @samp{call_function_by_hand} to remove an artificial stack
1535 @item PUSH_ARGUMENTS (nargs, args, sp, struct_return, struct_addr)
1536 Define this to push arguments onto the stack for inferior function
1537 call. Return the updated stack pointer value.
1539 @item PUSH_DUMMY_FRAME
1540 Used in @samp{call_function_by_hand} to create an artificial stack frame.
1542 @item REGISTER_BYTES
1543 The total amount of space needed to store GDB's copy of the machine's
1546 @item REGISTER_NAME(i)
1547 Return the name of register @var{i} as a string. May return @var{NULL}
1548 or @var{NUL} to indicate that register @var{i} is not valid.
1550 @item REGISTER_NAMES
1551 Deprecated in favor of @var{REGISTER_NAME}.
1553 @item REG_STRUCT_HAS_ADDR (gcc_p, type)
1554 Define this to return 1 if the given type will be passed by pointer
1555 rather than directly.
1557 @item SDB_REG_TO_REGNUM
1558 Define this to convert sdb register numbers into GDB regnums. If not
1559 defined, no conversion will be done.
1561 @item SHIFT_INST_REGS
1562 (Only used for m88k targets.)
1564 @item SKIP_PROLOGUE (pc)
1565 A C expression that returns the address of the ``real'' code beyond the
1566 function entry prologue found at @var{pc}.
1568 @item SKIP_PROLOGUE_FRAMELESS_P
1569 A C expression that should behave similarly, but that can stop as soon
1570 as the function is known to have a frame. If not defined,
1571 @code{SKIP_PROLOGUE} will be used instead.
1573 @item SKIP_TRAMPOLINE_CODE (pc)
1574 If the target machine has trampoline code that sits between callers and
1575 the functions being called, then define this macro to return a new PC
1576 that is at the start of the real function.
1579 Define this to be the number of the register that serves as the stack
1582 @item STAB_REG_TO_REGNUM
1583 Define this to convert stab register numbers (as gotten from `r'
1584 declarations) into GDB regnums. If not defined, no conversion will be
1587 @item STACK_ALIGN (addr)
1588 Define this to adjust the address to the alignment required for the
1591 @item STEP_SKIPS_DELAY (addr)
1592 Define this to return true if the address is of an instruction with a
1593 delay slot. If a breakpoint has been placed in the instruction's delay
1594 slot, GDB will single-step over that instruction before resuming
1595 normally. Currently only defined for the Mips.
1597 @item STORE_RETURN_VALUE (type, valbuf)
1598 A C expression that stores a function return value of type @var{type},
1599 where @var{valbuf} is the address of the value to be stored.
1601 @item SUN_FIXED_LBRAC_BUG
1602 (Used only for Sun-3 and Sun-4 targets.)
1604 @item SYMBOL_RELOADING_DEFAULT
1605 The default value of the `symbol-reloading' variable. (Never defined in
1608 @item TARGET_BYTE_ORDER_DEFAULT
1609 The ordering of bytes in the target. This must be either
1610 @code{BIG_ENDIAN} or @code{LITTLE_ENDIAN}. This macro replaces
1611 @var{TARGET_BYTE_ORDER} which is deprecated.
1613 @item TARGET_BYTE_ORDER_SELECTABLE_P
1614 Non-zero if the target has both @code{BIG_ENDIAN} and
1615 @code{LITTLE_ENDIAN} variants. This macro replaces
1616 @var{TARGET_BYTE_ORDER_SELECTABLE} which is deprecated.
1618 @item TARGET_CHAR_BIT
1619 Number of bits in a char; defaults to 8.
1621 @item TARGET_COMPLEX_BIT
1622 Number of bits in a complex number; defaults to @code{2 * TARGET_FLOAT_BIT}.
1624 At present this macro is not used.
1626 @item TARGET_DOUBLE_BIT
1627 Number of bits in a double float; defaults to @code{8 * TARGET_CHAR_BIT}.
1629 @item TARGET_DOUBLE_COMPLEX_BIT
1630 Number of bits in a double complex; defaults to @code{2 * TARGET_DOUBLE_BIT}.
1632 At present this macro is not used.
1634 @item TARGET_FLOAT_BIT
1635 Number of bits in a float; defaults to @code{4 * TARGET_CHAR_BIT}.
1637 @item TARGET_INT_BIT
1638 Number of bits in an integer; defaults to @code{4 * TARGET_CHAR_BIT}.
1640 @item TARGET_LONG_BIT
1641 Number of bits in a long integer; defaults to @code{4 * TARGET_CHAR_BIT}.
1643 @item TARGET_LONG_DOUBLE_BIT
1644 Number of bits in a long double float;
1645 defaults to @code{2 * TARGET_DOUBLE_BIT}.
1647 @item TARGET_LONG_LONG_BIT
1648 Number of bits in a long long integer; defaults to @code{2 * TARGET_LONG_BIT}.
1650 @item TARGET_PTR_BIT
1651 Number of bits in a pointer; defaults to @code{TARGET_INT_BIT}.
1653 @item TARGET_SHORT_BIT
1654 Number of bits in a short integer; defaults to @code{2 * TARGET_CHAR_BIT}.
1656 @item TARGET_READ_PC
1657 @item TARGET_WRITE_PC (val, pid)
1658 @item TARGET_READ_SP
1659 @item TARGET_WRITE_SP
1660 @item TARGET_READ_FP
1661 @item TARGET_WRITE_FP
1662 These change the behavior of @code{read_pc}, @code{write_pc},
1663 @code{read_sp}, @code{write_sp}, @code{read_fp} and @code{write_fp}.
1664 For most targets, these may be left undefined. GDB will call the read
1665 and write register functions with the relevant @code{_REGNUM} argument.
1667 These macros are useful when a target keeps one of these registers in a
1668 hard to get at place; for example, part in a segment register and part
1669 in an ordinary register.
1671 @item TARGET_VIRTUAL_FRAME_POINTER(pc,regp,offsetp)
1672 Returns a @code{(register, offset)} pair representing the virtual
1673 frame pointer in use at the code address @code{"pc"}. If virtual
1674 frame pointers are not used, a default definition simply returns
1675 @code{FP_REGNUM}, with an offset of zero.
1677 @item USE_STRUCT_CONVENTION (gcc_p, type)
1678 If defined, this must be an expression that is nonzero if a value of the
1679 given @var{type} being returned from a function must have space
1680 allocated for it on the stack. @var{gcc_p} is true if the function
1681 being considered is known to have been compiled by GCC; this is helpful
1682 for systems where GCC is known to use different calling convention than
1685 @item VARIABLES_INSIDE_BLOCK (desc, gcc_p)
1686 For dbx-style debugging information, if the compiler puts variable
1687 declarations inside LBRAC/RBRAC blocks, this should be defined to be
1688 nonzero. @var{desc} is the value of @code{n_desc} from the
1689 @code{N_RBRAC} symbol, and @var{gcc_p} is true if GDB has noticed the
1690 presence of either the @code{GCC_COMPILED_SYMBOL} or the
1691 @code{GCC2_COMPILED_SYMBOL}. By default, this is 0.
1693 @item OS9K_VARIABLES_INSIDE_BLOCK (desc, gcc_p)
1694 Similarly, for OS/9000. Defaults to 1.
1698 Motorola M68K target conditionals.
1703 Define this to be the 4-bit location of the breakpoint trap vector. If
1704 not defined, it will default to @code{0xf}.
1706 @item REMOTE_BPT_VECTOR
1707 Defaults to @code{1}.
1711 @section Adding a New Target
1713 The following files define a target to GDB:
1717 @item gdb/config/@var{arch}/@var{ttt}.mt
1718 Contains a Makefile fragment specific to this target. Specifies what
1719 object files are needed for target @var{ttt}, by defining
1720 @samp{TDEPFILES=@dots{}}. Also specifies the header file which
1721 describes @var{ttt}, by defining @samp{TM_FILE= tm-@var{ttt}.h}. You
1722 can also define @samp{TM_CFLAGS}, @samp{TM_CLIBS}, @samp{TM_CDEPS}, but
1723 these are now deprecated and may go away in future versions of GDB.
1725 @item gdb/config/@var{arch}/tm-@var{ttt}.h
1726 (@file{tm.h} is a link to this file, created by configure). Contains
1727 macro definitions about the target machine's registers, stack frame
1728 format and instructions.
1730 @item gdb/@var{ttt}-tdep.c
1731 Contains any miscellaneous code required for this target machine. On
1732 some machines it doesn't exist at all. Sometimes the macros in
1733 @file{tm-@var{ttt}.h} become very complicated, so they are implemented
1734 as functions here instead, and the macro is simply defined to call the
1735 function. This is vastly preferable, since it is easier to understand
1738 @item gdb/config/@var{arch}/tm-@var{arch}.h
1739 This often exists to describe the basic layout of the target machine's
1740 processor chip (registers, stack, etc). If used, it is included by
1741 @file{tm-@var{ttt}.h}. It can be shared among many targets that use the
1744 @item gdb/@var{arch}-tdep.c
1745 Similarly, there are often common subroutines that are shared by all
1746 target machines that use this particular architecture.
1750 If you are adding a new operating system for an existing CPU chip, add a
1751 @file{config/tm-@var{os}.h} file that describes the operating system
1752 facilities that are unusual (extra symbol table info; the breakpoint
1753 instruction needed; etc). Then write a @file{@var{arch}/tm-@var{os}.h}
1754 that just @code{#include}s @file{tm-@var{arch}.h} and
1755 @file{config/tm-@var{os}.h}.
1758 @node Target Vector Definition
1760 @chapter Target Vector Definition
1762 The target vector defines the interface between GDB's abstract handling
1763 of target systems, and the nitty-gritty code that actually exercises
1764 control over a process or a serial port. GDB includes some 30-40
1765 different target vectors; however, each configuration of GDB includes
1768 @section File Targets
1770 Both executables and core files have target vectors.
1772 @section Standard Protocol and Remote Stubs
1774 GDB's file @file{remote.c} talks a serial protocol to code that runs in
1775 the target system. GDB provides several sample ``stubs'' that can be
1776 integrated into target programs or operating systems for this purpose;
1777 they are named @file{*-stub.c}.
1779 The GDB user's manual describes how to put such a stub into your target
1780 code. What follows is a discussion of integrating the SPARC stub into a
1781 complicated operating system (rather than a simple program), by Stu
1782 Grossman, the author of this stub.
1784 The trap handling code in the stub assumes the following upon entry to
1789 @item %l1 and %l2 contain pc and npc respectively at the time of the trap
1791 @item traps are disabled
1793 @item you are in the correct trap window
1797 As long as your trap handler can guarantee those conditions, then there
1798 is no reason why you shouldn't be able to `share' traps with the stub.
1799 The stub has no requirement that it be jumped to directly from the
1800 hardware trap vector. That is why it calls @code{exceptionHandler()},
1801 which is provided by the external environment. For instance, this could
1802 setup the hardware traps to actually execute code which calls the stub
1803 first, and then transfers to its own trap handler.
1805 For the most point, there probably won't be much of an issue with
1806 `sharing' traps, as the traps we use are usually not used by the kernel,
1807 and often indicate unrecoverable error conditions. Anyway, this is all
1808 controlled by a table, and is trivial to modify. The most important
1809 trap for us is for @code{ta 1}. Without that, we can't single step or
1810 do breakpoints. Everything else is unnecessary for the proper operation
1811 of the debugger/stub.
1813 From reading the stub, it's probably not obvious how breakpoints work.
1814 They are simply done by deposit/examine operations from GDB.
1816 @section ROM Monitor Interface
1818 @section Custom Protocols
1820 @section Transport Layer
1822 @section Builtin Simulator
1825 @node Native Debugging
1827 @chapter Native Debugging
1829 Several files control GDB's configuration for native support:
1833 @item gdb/config/@var{arch}/@var{xyz}.mh
1834 Specifies Makefile fragments needed when hosting @emph{or native} on
1835 machine @var{xyz}. In particular, this lists the required
1836 native-dependent object files, by defining @samp{NATDEPFILES=@dots{}}.
1837 Also specifies the header file which describes native support on
1838 @var{xyz}, by defining @samp{NAT_FILE= nm-@var{xyz}.h}. You can also
1839 define @samp{NAT_CFLAGS}, @samp{NAT_ADD_FILES}, @samp{NAT_CLIBS},
1840 @samp{NAT_CDEPS}, etc.; see @file{Makefile.in}.
1842 @item gdb/config/@var{arch}/nm-@var{xyz}.h
1843 (@file{nm.h} is a link to this file, created by configure). Contains C
1844 macro definitions describing the native system environment, such as
1845 child process control and core file support.
1847 @item gdb/@var{xyz}-nat.c
1848 Contains any miscellaneous C code required for this native support of
1849 this machine. On some machines it doesn't exist at all.
1853 There are some ``generic'' versions of routines that can be used by
1854 various systems. These can be customized in various ways by macros
1855 defined in your @file{nm-@var{xyz}.h} file. If these routines work for
1856 the @var{xyz} host, you can just include the generic file's name (with
1857 @samp{.o}, not @samp{.c}) in @code{NATDEPFILES}.
1859 Otherwise, if your machine needs custom support routines, you will need
1860 to write routines that perform the same functions as the generic file.
1861 Put them into @code{@var{xyz}-nat.c}, and put @code{@var{xyz}-nat.o}
1862 into @code{NATDEPFILES}.
1867 This contains the @emph{target_ops vector} that supports Unix child
1868 processes on systems which use ptrace and wait to control the child.
1871 This contains the @emph{target_ops vector} that supports Unix child
1872 processes on systems which use /proc to control the child.
1875 This does the low-level grunge that uses Unix system calls to do a "fork
1876 and exec" to start up a child process.
1879 This is the low level interface to inferior processes for systems using
1880 the Unix @code{ptrace} call in a vanilla way.
1884 @section Native core file Support
1888 @item core-aout.c::fetch_core_registers()
1889 Support for reading registers out of a core file. This routine calls
1890 @code{register_addr()}, see below. Now that BFD is used to read core
1891 files, virtually all machines should use @code{core-aout.c}, and should
1892 just provide @code{fetch_core_registers} in @code{@var{xyz}-nat.c} (or
1893 @code{REGISTER_U_ADDR} in @code{nm-@var{xyz}.h}).
1895 @item core-aout.c::register_addr()
1896 If your @code{nm-@var{xyz}.h} file defines the macro
1897 @code{REGISTER_U_ADDR(addr, blockend, regno)}, it should be defined to
1898 set @code{addr} to the offset within the @samp{user} struct of GDB
1899 register number @code{regno}. @code{blockend} is the offset within the
1900 ``upage'' of @code{u.u_ar0}. If @code{REGISTER_U_ADDR} is defined,
1901 @file{core-aout.c} will define the @code{register_addr()} function and
1902 use the macro in it. If you do not define @code{REGISTER_U_ADDR}, but
1903 you are using the standard @code{fetch_core_registers()}, you will need
1904 to define your own version of @code{register_addr()}, put it into your
1905 @code{@var{xyz}-nat.c} file, and be sure @code{@var{xyz}-nat.o} is in
1906 the @code{NATDEPFILES} list. If you have your own
1907 @code{fetch_core_registers()}, you may not need a separate
1908 @code{register_addr()}. Many custom @code{fetch_core_registers()}
1909 implementations simply locate the registers themselves.@refill
1913 When making GDB run native on a new operating system, to make it
1914 possible to debug core files, you will need to either write specific
1915 code for parsing your OS's core files, or customize
1916 @file{bfd/trad-core.c}. First, use whatever @code{#include} files your
1917 machine uses to define the struct of registers that is accessible
1918 (possibly in the u-area) in a core file (rather than
1919 @file{machine/reg.h}), and an include file that defines whatever header
1920 exists on a core file (e.g. the u-area or a @samp{struct core}). Then
1921 modify @code{trad_unix_core_file_p()} to use these values to set up the
1922 section information for the data segment, stack segment, any other
1923 segments in the core file (perhaps shared library contents or control
1924 information), ``registers'' segment, and if there are two discontiguous
1925 sets of registers (e.g. integer and float), the ``reg2'' segment. This
1926 section information basically delimits areas in the core file in a
1927 standard way, which the section-reading routines in BFD know how to seek
1930 Then back in GDB, you need a matching routine called
1931 @code{fetch_core_registers()}. If you can use the generic one, it's in
1932 @file{core-aout.c}; if not, it's in your @file{@var{xyz}-nat.c} file.
1933 It will be passed a char pointer to the entire ``registers'' segment,
1934 its length, and a zero; or a char pointer to the entire ``regs2''
1935 segment, its length, and a 2. The routine should suck out the supplied
1936 register values and install them into GDB's ``registers'' array.
1938 If your system uses @file{/proc} to control processes, and uses ELF
1939 format core files, then you may be able to use the same routines for
1940 reading the registers out of processes and out of core files.
1948 @section shared libraries
1950 @section Native Conditionals
1952 When GDB is configured and compiled, various macros are defined or left
1953 undefined, to control compilation when the host and target systems are
1954 the same. These macros should be defined (or left undefined) in
1955 @file{nm-@var{system}.h}.
1960 If defined, then GDB will include support for the @code{attach} and
1961 @code{detach} commands.
1963 @item CHILD_PREPARE_TO_STORE
1964 If the machine stores all registers at once in the child process, then
1965 define this to ensure that all values are correct. This usually entails
1966 a read from the child.
1968 [Note that this is incorrectly defined in @file{xm-@var{system}.h} files
1971 @item FETCH_INFERIOR_REGISTERS
1972 Define this if the native-dependent code will provide its own routines
1973 @code{fetch_inferior_registers} and @code{store_inferior_registers} in
1974 @file{@var{HOST}-nat.c}. If this symbol is @emph{not} defined, and
1975 @file{infptrace.c} is included in this configuration, the default
1976 routines in @file{infptrace.c} are used for these functions.
1978 @item FILES_INFO_HOOK
1979 (Only defined for Convex.)
1982 This macro is normally defined to be the number of the first floating
1983 point register, if the machine has such registers. As such, it would
1984 appear only in target-specific code. However, /proc support uses this
1985 to decide whether floats are in use on this target.
1987 @item GET_LONGJMP_TARGET
1988 For most machines, this is a target-dependent parameter. On the
1989 DECstation and the Iris, this is a native-dependent parameter, since
1990 <setjmp.h> is needed to define it.
1992 This macro determines the target PC address that longjmp() will jump to,
1993 assuming that we have just stopped at a longjmp breakpoint. It takes a
1994 CORE_ADDR * as argument, and stores the target PC value through this
1995 pointer. It examines the current state of the machine as needed.
1998 Define this to the address of the @code{u} structure (the ``user
1999 struct'', also known as the ``u-page'') in kernel virtual memory. GDB
2000 needs to know this so that it can subtract this address from absolute
2001 addresses in the upage, that are obtained via ptrace or from core files.
2002 On systems that don't need this value, set it to zero.
2004 @item KERNEL_U_ADDR_BSD
2005 Define this to cause GDB to determine the address of @code{u} at
2006 runtime, by using Berkeley-style @code{nlist} on the kernel's image in
2009 @item KERNEL_U_ADDR_HPUX
2010 Define this to cause GDB to determine the address of @code{u} at
2011 runtime, by using HP-style @code{nlist} on the kernel's image in the
2014 @item ONE_PROCESS_WRITETEXT
2015 Define this to be able to, when a breakpoint insertion fails, warn the
2016 user that another process may be running with the same executable.
2019 Defines the format for the name of a @file{/proc} device. Should be
2020 defined in @file{nm.h} @emph{only} in order to override the default
2021 definition in @file{procfs.c}.
2026 @item PTRACE_ARG3_TYPE
2027 The type of the third argument to the @code{ptrace} system call, if it
2028 exists and is different from @code{int}.
2030 @item REGISTER_U_ADDR
2031 Defines the offset of the registers in the ``u area''.
2033 @item SHELL_COMMAND_CONCAT
2034 If defined, is a string to prefix on the shell command used to start the
2038 If defined, this is the name of the shell to use to run the inferior.
2039 Defaults to @code{"/bin/sh"}.
2041 @item SOLIB_ADD (filename, from_tty, targ)
2042 Define this to expand into an expression that will cause the symbols in
2043 @var{filename} to be added to GDB's symbol table.
2045 @item SOLIB_CREATE_INFERIOR_HOOK
2046 Define this to expand into any shared-library-relocation code that you
2047 want to be run just after the child process has been forked.
2049 @item START_INFERIOR_TRAPS_EXPECTED
2050 When starting an inferior, GDB normally expects to trap twice; once when
2051 the shell execs, and once when the program itself execs. If the actual
2052 number of traps is something other than 2, then define this macro to
2053 expand into the number expected.
2055 @item SVR4_SHARED_LIBS
2056 Define this to indicate that SVR4-style shared libraries are in use.
2059 This determines whether small routines in @file{*-tdep.c}, which
2060 translate register values between GDB's internal representation and the
2061 /proc representation, are compiled.
2064 This is the offset of the registers in the upage. It need only be
2065 defined if the generic ptrace register access routines in
2066 @file{infptrace.c} are being used (that is, @file{infptrace.c} is
2067 configured in, and @code{FETCH_INFERIOR_REGISTERS} is not defined). If
2068 the default value from @file{infptrace.c} is good enough, leave it
2071 The default value means that u.u_ar0 @emph{points to} the location of
2072 the registers. I'm guessing that @code{#define U_REGS_OFFSET 0} means
2073 that u.u_ar0 @emph{is} the location of the registers.
2079 Define this to debug ptrace calls.
2084 @node Support Libraries
2086 @chapter Support Libraries
2090 BFD provides support for GDB in several ways:
2094 @item identifying executable and core files
2095 BFD will identify a variety of file types, including a.out, coff, and
2096 several variants thereof, as well as several kinds of core files.
2098 @item access to sections of files
2099 BFD parses the file headers to determine the names, virtual addresses,
2100 sizes, and file locations of all the various named sections in files
2101 (such as the text section or the data section). GDB simply calls BFD to
2102 read or write section X at byte offset Y for length Z.
2104 @item specialized core file support
2105 BFD provides routines to determine the failing command name stored in a
2106 core file, the signal with which the program failed, and whether a core
2107 file matches (i.e. could be a core dump of) a particular executable
2110 @item locating the symbol information
2111 GDB uses an internal interface of BFD to determine where to find the
2112 symbol information in an executable file or symbol-file. GDB itself
2113 handles the reading of symbols, since BFD does not ``understand'' debug
2114 symbols, but GDB uses BFD's cached information to find the symbols,
2121 The opcodes library provides GDB's disassembler. (It's a separate
2122 library because it's also used in binutils, for @file{objdump}).
2142 @item SIGN_EXTEND_CHAR
2144 @item SWITCH_ENUM_BUG
2160 This chapter covers topics that are lower-level than the major
2165 Cleanups are a structured way to deal with things that need to be done
2166 later. When your code does something (like @code{malloc} some memory,
2167 or open a file) that needs to be undone later (e.g. free the memory or
2168 close the file), it can make a cleanup. The cleanup will be done at
2169 some future point: when the command is finished, when an error occurs,
2170 or when your code decides it's time to do cleanups.
2172 You can also discard cleanups, that is, throw them away without doing
2173 what they say. This is only done if you ask that it be done.
2179 @item struct cleanup *@var{old_chain};
2180 Declare a variable which will hold a cleanup chain handle.
2182 @item @var{old_chain} = make_cleanup (@var{function}, @var{arg});
2183 Make a cleanup which will cause @var{function} to be called with
2184 @var{arg} (a @code{char *}) later. The result, @var{old_chain}, is a
2185 handle that can be passed to @code{do_cleanups} or
2186 @code{discard_cleanups} later. Unless you are going to call
2187 @code{do_cleanups} or @code{discard_cleanups} yourself, you can ignore
2188 the result from @code{make_cleanup}.
2190 @item do_cleanups (@var{old_chain});
2191 Perform all cleanups done since @code{make_cleanup} returned
2192 @var{old_chain}. E.g.:
2194 make_cleanup (a, 0);
2195 old = make_cleanup (b, 0);
2199 will call @code{b()} but will not call @code{a()}. The cleanup that
2200 calls @code{a()} will remain in the cleanup chain, and will be done
2201 later unless otherwise discarded.@refill
2203 @item discard_cleanups (@var{old_chain});
2204 Same as @code{do_cleanups} except that it just removes the cleanups from
2205 the chain and does not call the specified functions.
2209 Some functions, e.g. @code{fputs_filtered()} or @code{error()}, specify
2210 that they ``should not be called when cleanups are not in place''. This
2211 means that any actions you need to reverse in the case of an error or
2212 interruption must be on the cleanup chain before you call these
2213 functions, since they might never return to your code (they
2214 @samp{longjmp} instead).
2216 @section Wrapping Output Lines
2218 Output that goes through @code{printf_filtered} or @code{fputs_filtered}
2219 or @code{fputs_demangled} needs only to have calls to @code{wrap_here}
2220 added in places that would be good breaking points. The utility
2221 routines will take care of actually wrapping if the line width is
2224 The argument to @code{wrap_here} is an indentation string which is
2225 printed @emph{only} if the line breaks there. This argument is saved
2226 away and used later. It must remain valid until the next call to
2227 @code{wrap_here} or until a newline has been printed through the
2228 @code{*_filtered} functions. Don't pass in a local variable and then
2231 It is usually best to call @code{wrap_here()} after printing a comma or
2232 space. If you call it before printing a space, make sure that your
2233 indentation properly accounts for the leading space that will print if
2234 the line wraps there.
2236 Any function or set of functions that produce filtered output must
2237 finish by printing a newline, to flush the wrap buffer, before switching
2238 to unfiltered (``@code{printf}'') output. Symbol reading routines that
2239 print warnings are a good example.
2241 @section GDB Coding Standards
2243 GDB follows the GNU coding standards, as described in
2244 @file{etc/standards.texi}. This file is also available for anonymous
2245 FTP from GNU archive sites. GDB takes a strict interpretation of the
2246 standard; in general, when the GNU standard recommends a practice but
2247 does not require it, GDB requires it.
2249 GDB follows an additional set of coding standards specific to GDB,
2250 as described in the following sections.
2252 You can configure with @samp{--enable-build-warnings} to get GCC to
2253 check on a number of these rules. GDB sources ought not to engender any
2254 complaints, unless they are caused by bogus host systems. (The exact
2255 set of enabled warnings is currently @samp{-Wall -Wpointer-arith
2256 -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations}.
2258 @subsection Formatting
2260 The standard GNU recommendations for formatting must be followed
2263 Note that while in a definition, the function's name must be in column
2264 zero; in a function declaration, the name must be on the same line as
2267 In addition, there must be a space between a function or macro name and
2268 the opening parenthesis of its argument list (except for macro
2269 definitions, as required by C). There must not be a space after an open
2270 paren/bracket or before a close paren/bracket.
2272 While additional whitespace is generally helpful for reading, do not use
2273 more than one blank line to separate blocks, and avoid adding whitespace
2274 after the end of a program line (as of 1/99, some 600 lines had whitespace
2275 after the semicolon). Excess whitespace causes difficulties for diff and
2278 @subsection Comments
2280 The standard GNU requirements on comments must be followed strictly.
2282 Block comments must appear in the following form, with no `/*'- or
2283 '*/'-only lines, and no leading `*':
2286 /* Wait for control to return from inferior to debugger. If inferior
2287 gets a signal, we may decide to start it up again instead of
2288 returning. That is why there is a loop in this function. When
2289 this function actually returns it means the inferior should be left
2290 stopped and GDB should read more commands. */
2293 (Note that this format is encouraged by Emacs; tabbing for a multi-line
2294 comment works correctly, and M-Q fills the block consistently.)
2296 Put a blank line between the block comments preceding function or
2297 variable definitions, and the definition itself.
2299 In general, put function-body comments on lines by themselves, rather
2300 than trying to fit them into the 20 characters left at the end of a
2301 line, since either the comment or the code will inevitably get longer
2302 than will fit, and then somebody will have to move it anyhow.
2306 Code must not depend on the sizes of C data types, the format of the
2307 host's floating point numbers, the alignment of anything, or the order
2308 of evaluation of expressions.
2310 Use functions freely. There are only a handful of compute-bound areas
2311 in GDB that might be affected by the overhead of a function call, mainly
2312 in symbol reading. Most of GDB's performance is limited by the target
2313 interface (whether serial line or system call).
2315 However, use functions with moderation. A thousand one-line functions
2316 are just as hard to understand as a single thousand-line function.
2318 @subsection Function Prototypes
2320 Prototypes must be used to @emph{declare} functions but never to
2321 @emph{define} them. Prototypes for GDB functions must include both the
2322 argument type and name, with the name matching that used in the actual
2323 function definition.
2325 For the sake of compatibility with pre-ANSI compilers, define prototypes
2326 with the @code{PARAMS} macro:
2329 extern int memory_remove_breakpoint PARAMS ((CORE_ADDR addr,
2330 char *contents_cache));
2333 Note the double parentheses around the parameter types. This allows an
2334 arbitrary number of parameters to be described, without freaking out the
2335 C preprocessor. When the function has no parameters, it should be
2339 extern void noprocess PARAMS ((void));
2342 The @code{PARAMS} macro expands to its argument in ANSI C, or to a
2343 simple @code{()} in traditional C.
2345 All external functions should have a @code{PARAMS} declaration in a
2346 header file that callers include, except for @code{_initialize_*}
2347 functions, which must be external so that @file{init.c} construction
2348 works, but shouldn't be visible to random source files.
2350 All static functions must be declared in a block near the top of the
2353 @subsection Clean Design
2355 In addition to getting the syntax right, there's the little question of
2356 semantics. Some things are done in certain ways in GDB because long
2357 experience has shown that the more obvious ways caused various kinds of
2360 You can't assume the byte order of anything that comes from a target
2361 (including @var{value}s, object files, and instructions). Such things
2362 must be byte-swapped using @code{SWAP_TARGET_AND_HOST} in GDB, or one of
2363 the swap routines defined in @file{bfd.h}, such as @code{bfd_get_32}.
2365 You can't assume that you know what interface is being used to talk to
2366 the target system. All references to the target must go through the
2367 current @code{target_ops} vector.
2369 You can't assume that the host and target machines are the same machine
2370 (except in the ``native'' support modules). In particular, you can't
2371 assume that the target machine's header files will be available on the
2372 host machine. Target code must bring along its own header files --
2373 written from scratch or explicitly donated by their owner, to avoid
2376 Insertion of new @code{#ifdef}'s will be frowned upon. It's much better
2377 to write the code portably than to conditionalize it for various
2380 New @code{#ifdef}'s which test for specific compilers or manufacturers
2381 or operating systems are unacceptable. All @code{#ifdef}'s should test
2382 for features. The information about which configurations contain which
2383 features should be segregated into the configuration files. Experience
2384 has proven far too often that a feature unique to one particular system
2385 often creeps into other systems; and that a conditional based on some
2386 predefined macro for your current system will become worthless over
2387 time, as new versions of your system come out that behave differently
2388 with regard to this feature.
2390 Adding code that handles specific architectures, operating systems,
2391 target interfaces, or hosts, is not acceptable in generic code. If a
2392 hook is needed at that point, invent a generic hook and define it for
2393 your configuration, with something like:
2396 #ifdef WRANGLE_SIGNALS
2397 WRANGLE_SIGNALS (signo);
2401 In your host, target, or native configuration file, as appropriate,
2402 define @code{WRANGLE_SIGNALS} to do the machine-dependent thing. Take a
2403 bit of care in defining the hook, so that it can be used by other ports
2404 in the future, if they need a hook in the same place.
2406 If the hook is not defined, the code should do whatever "most" machines
2407 want. Using @code{#ifdef}, as above, is the preferred way to do this,
2408 but sometimes that gets convoluted, in which case use
2411 #ifndef SPECIAL_FOO_HANDLING
2412 #define SPECIAL_FOO_HANDLING(pc, sp) (0)
2416 where the macro is used or in an appropriate header file.
2418 Whether to include a @dfn{small} hook, a hook around the exact pieces of
2419 code which are system-dependent, or whether to replace a whole function
2420 with a hook depends on the case. A good example of this dilemma can be
2421 found in @code{get_saved_register}. All machines that GDB 2.8 ran on
2422 just needed the @code{FRAME_FIND_SAVED_REGS} hook to find the saved
2423 registers. Then the SPARC and Pyramid came along, and
2424 @code{HAVE_REGISTER_WINDOWS} and @code{REGISTER_IN_WINDOW_P} were
2425 introduced. Then the 29k and 88k required the @code{GET_SAVED_REGISTER}
2426 hook. The first three are examples of small hooks; the latter replaces
2427 a whole function. In this specific case, it is useful to have both
2428 kinds; it would be a bad idea to replace all the uses of the small hooks
2429 with @code{GET_SAVED_REGISTER}, since that would result in much
2430 duplicated code. Other times, duplicating a few lines of code here or
2431 there is much cleaner than introducing a large number of small hooks.
2433 Another way to generalize GDB along a particular interface is with an
2434 attribute struct. For example, GDB has been generalized to handle
2435 multiple kinds of remote interfaces -- not by #ifdef's everywhere, but
2436 by defining the "target_ops" structure and having a current target (as
2437 well as a stack of targets below it, for memory references). Whenever
2438 something needs to be done that depends on which remote interface we are
2439 using, a flag in the current target_ops structure is tested (e.g.
2440 `target_has_stack'), or a function is called through a pointer in the
2441 current target_ops structure. In this way, when a new remote interface
2442 is added, only one module needs to be touched -- the one that actually
2443 implements the new remote interface. Other examples of
2444 attribute-structs are BFD access to multiple kinds of object file
2445 formats, or GDB's access to multiple source languages.
2447 Please avoid duplicating code. For example, in GDB 3.x all the code
2448 interfacing between @code{ptrace} and the rest of GDB was duplicated in
2449 @file{*-dep.c}, and so changing something was very painful. In GDB 4.x,
2450 these have all been consolidated into @file{infptrace.c}.
2451 @file{infptrace.c} can deal with variations between systems the same way
2452 any system-independent file would (hooks, #if defined, etc.), and
2453 machines which are radically different don't need to use infptrace.c at
2459 @chapter Porting GDB
2461 Most of the work in making GDB compile on a new machine is in specifying
2462 the configuration of the machine. This is done in a dizzying variety of
2463 header files and configuration scripts, which we hope to make more
2464 sensible soon. Let's say your new host is called an @var{xyz} (e.g.
2465 @samp{sun4}), and its full three-part configuration name is
2466 @code{@var{arch}-@var{xvend}-@var{xos}} (e.g. @samp{sparc-sun-sunos4}).
2469 In the top level directory, edit @file{config.sub} and add @var{arch},
2470 @var{xvend}, and @var{xos} to the lists of supported architectures,
2471 vendors, and operating systems near the bottom of the file. Also, add
2472 @var{xyz} as an alias that maps to
2473 @code{@var{arch}-@var{xvend}-@var{xos}}. You can test your changes by
2477 ./config.sub @var{xyz}
2482 ./config.sub @code{@var{arch}-@var{xvend}-@var{xos}}
2485 which should both respond with @code{@var{arch}-@var{xvend}-@var{xos}}
2486 and no error messages.
2488 You need to port BFD, if that hasn't been done already. Porting BFD is
2489 beyond the scope of this manual.
2491 To configure GDB itself, edit @file{gdb/configure.host} to recognize
2492 your system and set @code{gdb_host} to @var{xyz}, and (unless your
2493 desired target is already available) also edit @file{gdb/configure.tgt},
2494 setting @code{gdb_target} to something appropriate (for instance,
2497 Finally, you'll need to specify and define GDB's host-, native-, and
2498 target-dependent @file{.h} and @file{.c} files used for your
2501 @section Configuring GDB for Release
2503 From the top level directory (containing @file{gdb}, @file{bfd},
2504 @file{libiberty}, and so on):
2506 make -f Makefile.in gdb.tar.gz
2509 This will properly configure, clean, rebuild any files that are
2510 distributed pre-built (e.g. @file{c-exp.tab.c} or @file{refcard.ps}),
2511 and will then make a tarfile. (If the top level directory has already
2512 been configured, you can just do @code{make gdb.tar.gz} instead.)
2514 This procedure requires:
2516 @item symbolic links
2517 @item @code{makeinfo} (texinfo2 level)
2520 @item @code{yacc} or @code{bison}
2523 @dots{} and the usual slew of utilities (@code{sed}, @code{tar}, etc.).
2525 @subheading TEMPORARY RELEASE PROCEDURE FOR DOCUMENTATION
2527 @file{gdb.texinfo} is currently marked up using the texinfo-2 macros,
2528 which are not yet a default for anything (but we have to start using
2531 For making paper, the only thing this implies is the right generation of
2532 @file{texinfo.tex} needs to be included in the distribution.
2534 For making info files, however, rather than duplicating the texinfo2
2535 distribution, generate @file{gdb-all.texinfo} locally, and include the
2536 files @file{gdb.info*} in the distribution. Note the plural;
2537 @code{makeinfo} will split the document into one overall file and five
2538 or so included files.
2544 Check the @file{README} file, it often has useful information that does not
2545 appear anywhere else in the directory.
2548 * Getting Started:: Getting started working on GDB
2549 * Debugging GDB:: Debugging GDB with itself
2552 @node Getting Started,,, Hints
2554 @section Getting Started
2556 GDB is a large and complicated program, and if you first starting to
2557 work on it, it can be hard to know where to start. Fortunately, if you
2558 know how to go about it, there are ways to figure out what is going on.
2560 This manual, the GDB Internals manual, has information which applies
2561 generally to many parts of GDB.
2563 Information about particular functions or data structures are located in
2564 comments with those functions or data structures. If you run across a
2565 function or a global variable which does not have a comment correctly
2566 explaining what is does, this can be thought of as a bug in GDB; feel
2567 free to submit a bug report, with a suggested comment if you can figure
2568 out what the comment should say. If you find a comment which is
2569 actually wrong, be especially sure to report that.
2571 Comments explaining the function of macros defined in host, target, or
2572 native dependent files can be in several places. Sometimes they are
2573 repeated every place the macro is defined. Sometimes they are where the
2574 macro is used. Sometimes there is a header file which supplies a
2575 default definition of the macro, and the comment is there. This manual
2576 also documents all the available macros.
2577 @c (@pxref{Host Conditionals}, @pxref{Target
2578 @c Conditionals}, @pxref{Native Conditionals}, and @pxref{Obsolete
2581 Start with the header files. Once you some idea of how GDB's internal
2582 symbol tables are stored (see @file{symtab.h}, @file{gdbtypes.h}), you
2583 will find it much easier to understand the code which uses and creates
2584 those symbol tables.
2586 You may wish to process the information you are getting somehow, to
2587 enhance your understanding of it. Summarize it, translate it to another
2588 language, add some (perhaps trivial or non-useful) feature to GDB, use
2589 the code to predict what a test case would do and write the test case
2590 and verify your prediction, etc. If you are reading code and your eyes
2591 are starting to glaze over, this is a sign you need to use a more active
2594 Once you have a part of GDB to start with, you can find more
2595 specifically the part you are looking for by stepping through each
2596 function with the @code{next} command. Do not use @code{step} or you
2597 will quickly get distracted; when the function you are stepping through
2598 calls another function try only to get a big-picture understanding
2599 (perhaps using the comment at the beginning of the function being
2600 called) of what it does. This way you can identify which of the
2601 functions being called by the function you are stepping through is the
2602 one which you are interested in. You may need to examine the data
2603 structures generated at each stage, with reference to the comments in
2604 the header files explaining what the data structures are supposed to
2607 Of course, this same technique can be used if you are just reading the
2608 code, rather than actually stepping through it. The same general
2609 principle applies---when the code you are looking at calls something
2610 else, just try to understand generally what the code being called does,
2611 rather than worrying about all its details.
2613 A good place to start when tracking down some particular area is with a
2614 command which invokes that feature. Suppose you want to know how
2615 single-stepping works. As a GDB user, you know that the @code{step}
2616 command invokes single-stepping. The command is invoked via command
2617 tables (see @file{command.h}); by convention the function which actually
2618 performs the command is formed by taking the name of the command and
2619 adding @samp{_command}, or in the case of an @code{info} subcommand,
2620 @samp{_info}. For example, the @code{step} command invokes the
2621 @code{step_command} function and the @code{info display} command invokes
2622 @code{display_info}. When this convention is not followed, you might
2623 have to use @code{grep} or @kbd{M-x tags-search} in emacs, or run GDB on
2624 itself and set a breakpoint in @code{execute_command}.
2626 If all of the above fail, it may be appropriate to ask for information
2627 on @code{bug-gdb}. But @emph{never} post a generic question like ``I was
2628 wondering if anyone could give me some tips about understanding
2629 GDB''---if we had some magic secret we would put it in this manual.
2630 Suggestions for improving the manual are always welcome, of course.
2632 @node Debugging GDB,,,Hints
2634 @section Debugging GDB with itself
2636 If GDB is limping on your machine, this is the preferred way to get it
2637 fully functional. Be warned that in some ancient Unix systems, like
2638 Ultrix 4.2, a program can't be running in one process while it is being
2639 debugged in another. Rather than typing the command @code{@w{./gdb
2640 ./gdb}}, which works on Suns and such, you can copy @file{gdb} to
2641 @file{gdb2} and then type @code{@w{./gdb ./gdb2}}.
2643 When you run GDB in the GDB source directory, it will read a
2644 @file{.gdbinit} file that sets up some simple things to make debugging
2645 gdb easier. The @code{info} command, when executed without a subcommand
2646 in a GDB being debugged by gdb, will pop you back up to the top level
2647 gdb. See @file{.gdbinit} for details.
2649 If you use emacs, you will probably want to do a @code{make TAGS} after
2650 you configure your distribution; this will put the machine dependent
2651 routines for your local machine where they will be accessed first by
2654 Also, make sure that you've either compiled GDB with your local cc, or
2655 have run @code{fixincludes} if you are compiling with gcc.
2657 @section Submitting Patches
2659 Thanks for thinking of offering your changes back to the community of
2660 GDB users. In general we like to get well designed enhancements.
2661 Thanks also for checking in advance about the best way to transfer the
2664 The GDB maintainers will only install ``cleanly designed'' patches. You
2665 may not always agree on what is clean design.
2666 @c @pxref{Coding Style}, @pxref{Clean Design}.
2668 If the maintainers don't have time to put the patch in when it arrives,
2669 or if there is any question about a patch, it goes into a large queue
2670 with everyone else's patches and bug reports.
2672 The legal issue is that to incorporate substantial changes requires a
2673 copyright assignment from you and/or your employer, granting ownership
2674 of the changes to the Free Software Foundation. You can get the
2675 standard document for doing this by sending mail to
2676 @code{gnu@@prep.ai.mit.edu} and asking for it. I recommend that people
2677 write in "All programs owned by the Free Software Foundation" as "NAME
2678 OF PROGRAM", so that changes in many programs (not just GDB, but GAS,
2679 Emacs, GCC, etc) can be contributed with only one piece of legalese
2680 pushed through the bureacracy and filed with the FSF. I can't start
2681 merging changes until this paperwork is received by the FSF (their
2682 rules, which I follow since I maintain it for them).
2684 Technically, the easiest way to receive changes is to receive each
2685 feature as a small context diff or unidiff, suitable for "patch".
2686 Each message sent to me should include the changes to C code and
2687 header files for a single feature, plus ChangeLog entries for each
2688 directory where files were modified, and diffs for any changes needed
2689 to the manuals (gdb/doc/gdb.texi or gdb/doc/gdbint.texi). If there
2690 are a lot of changes for a single feature, they can be split down
2691 into multiple messages.
2693 In this way, if I read and like the feature, I can add it to the
2694 sources with a single patch command, do some testing, and check it in.
2695 If you leave out the ChangeLog, I have to write one. If you leave
2696 out the doc, I have to puzzle out what needs documenting. Etc.
2698 The reason to send each change in a separate message is that I will
2699 not install some of the changes. They'll be returned to you with
2700 questions or comments. If I'm doing my job, my message back to you
2701 will say what you have to fix in order to make the change acceptable.
2702 The reason to have separate messages for separate features is so
2703 that other changes (which I @emph{am} willing to accept) can be installed
2704 while one or more changes are being reworked. If multiple features
2705 are sent in a single message, I tend to not put in the effort to sort
2706 out the acceptable changes from the unacceptable, so none of the
2707 features get installed until all are acceptable.
2709 If this sounds painful or authoritarian, well, it is. But I get a lot
2710 of bug reports and a lot of patches, and most of them don't get
2711 installed because I don't have the time to finish the job that the bug
2712 reporter or the contributor could have done. Patches that arrive
2713 complete, working, and well designed, tend to get installed on the day
2714 they arrive. The others go into a queue and get installed if and when
2715 I scan back over the queue -- which can literally take months
2716 sometimes. It's in both our interests to make patch installation easy
2717 -- you get your changes installed, and I make some forward progress on
2718 GDB in a normal 12-hour day (instead of them having to wait until I
2719 have a 14-hour or 16-hour day to spend cleaning up patches before I
2722 Please send patches directly to the GDB maintainers at
2723 @code{gdb-patches@@cygnus.com}.
2725 @section Obsolete Conditionals
2727 Fragments of old code in GDB sometimes reference or set the following
2728 configuration macros. They should not be used by new code, and old uses
2729 should be removed as those parts of the debugger are otherwise touched.
2733 @item STACK_END_ADDR
2734 This macro used to define where the end of the stack appeared, for use
2735 in interpreting core file formats that don't record this address in the
2736 core file itself. This information is now configured in BFD, and GDB
2737 gets the info portably from there. The values in GDB's configuration
2738 files should be moved into BFD configuration files (if needed there),
2739 and deleted from all of GDB's config files.
2741 Any @file{@var{foo}-xdep.c} file that references STACK_END_ADDR
2742 is so old that it has never been converted to use BFD. Now that's old!
2744 @item PYRAMID_CONTROL_FRAME_DEBUGGING
2748 @item PYRAMID_PTRACE
2751 @item REG_STACK_SEGMENT