1 This is stabs.info, produced by makeinfo version 4.8 from
2 ../.././gdb/doc/stabs.texinfo.
4 INFO-DIR-SECTION Software development
6 * Stabs: (stabs). The "stabs" debugging information format.
9 This document describes the stabs debugging symbol tables.
11 Copyright (C) 1992,1993,1994,1995,1997,1998,2000,2001 Free
12 Software Foundation, Inc. Contributed by Cygnus Support. Written by
13 Julia Menapace, Jim Kingdon, and David MacKenzie.
15 Permission is granted to copy, distribute and/or modify this document
16 under the terms of the GNU Free Documentation License, Version 1.1 or
17 any later version published by the Free Software Foundation; with no
18 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
19 Texts. A copy of the license is included in the section entitled "GNU
20 Free Documentation License".
23 File: stabs.info, Node: Top, Next: Overview, Up: (dir)
25 The "stabs" representation of debugging information
26 ***************************************************
28 This document describes the stabs debugging format.
32 * Overview:: Overview of stabs
33 * Program Structure:: Encoding of the structure of the program
34 * Constants:: Constants
36 * Types:: Type definitions
37 * Macro define and undefine:: Representation of #define and #undef
38 * Symbol Tables:: Symbol information in symbol tables
39 * Cplusplus:: Stabs specific to C++
40 * Stab Types:: Symbol types in a.out files
41 * Symbol Descriptors:: Table of symbol descriptors
42 * Type Descriptors:: Table of type descriptors
43 * Expanded Reference:: Reference information by stab type
44 * Questions:: Questions and anomalies
45 * Stab Sections:: In some object file formats, stabs are
47 * Symbol Types Index:: Index of symbolic stab symbol type names.
48 * GNU Free Documentation License:: The license for this documentation
51 File: stabs.info, Node: Overview, Next: Program Structure, Prev: Top, Up: Top
56 "Stabs" refers to a format for information that describes a program to
57 a debugger. This format was apparently invented by Peter Kessler at
58 the University of California at Berkeley, for the `pdx' Pascal
59 debugger; the format has spread widely since then.
61 This document is one of the few published sources of documentation on
62 stabs. It is believed to be comprehensive for stabs used by C. The
63 lists of symbol descriptors (*note Symbol Descriptors::) and type
64 descriptors (*note Type Descriptors::) are believed to be completely
65 comprehensive. Stabs for COBOL-specific features and for variant
66 records (used by Pascal and Modula-2) are poorly documented here.
68 Other sources of information on stabs are `Dbx and Dbxtool
69 Interfaces', 2nd edition, by Sun, 1988, and `AIX Version 3.2 Files
70 Reference', Fourth Edition, September 1992, "dbx Stabstring Grammar" in
71 the a.out section, page 2-31. This document is believed to incorporate
72 the information from those two sources except where it explicitly
73 directs you to them for more information.
77 * Flow:: Overview of debugging information flow
78 * Stabs Format:: Overview of stab format
79 * String Field:: The string field
80 * C Example:: A simple example in C source
81 * Assembly Code:: The simple example at the assembly level
84 File: stabs.info, Node: Flow, Next: Stabs Format, Up: Overview
86 1.1 Overview of Debugging Information Flow
87 ==========================================
89 The GNU C compiler compiles C source in a `.c' file into assembly
90 language in a `.s' file, which the assembler translates into a `.o'
91 file, which the linker combines with other `.o' files and libraries to
92 produce an executable file.
94 With the `-g' option, GCC puts in the `.s' file additional debugging
95 information, which is slightly transformed by the assembler and linker,
96 and carried through into the final executable. This debugging
97 information describes features of the source file like line numbers,
98 the types and scopes of variables, and function names, parameters, and
101 For some object file formats, the debugging information is
102 encapsulated in assembler directives known collectively as "stab"
103 (symbol table) directives, which are interspersed with the generated
104 code. Stabs are the native format for debugging information in the
105 a.out and XCOFF object file formats. The GNU tools can also emit stabs
106 in the COFF and ECOFF object file formats.
108 The assembler adds the information from stabs to the symbol
109 information it places by default in the symbol table and the string
110 table of the `.o' file it is building. The linker consolidates the `.o'
111 files into one executable file, with one symbol table and one string
112 table. Debuggers use the symbol and string tables in the executable as
113 a source of debugging information about the program.
116 File: stabs.info, Node: Stabs Format, Next: String Field, Prev: Flow, Up: Overview
118 1.2 Overview of Stab Format
119 ===========================
121 There are three overall formats for stab assembler directives,
122 differentiated by the first word of the stab. The name of the directive
123 describes which combination of four possible data fields follows. It is
124 either `.stabs' (string), `.stabn' (number), or `.stabd' (dot). IBM's
125 XCOFF assembler uses `.stabx' (and some other directives such as
126 `.file' and `.bi') instead of `.stabs', `.stabn' or `.stabd'.
128 The overall format of each class of stab is:
130 .stabs "STRING",TYPE,OTHER,DESC,VALUE
131 .stabn TYPE,OTHER,DESC,VALUE
132 .stabd TYPE,OTHER,DESC
133 .stabx "STRING",VALUE,TYPE,SDB-TYPE
135 For `.stabn' and `.stabd', there is no STRING (the `n_strx' field is
136 zero; see *Note Symbol Tables::). For `.stabd', the VALUE field is
137 implicit and has the value of the current file location. For `.stabx',
138 the SDB-TYPE field is unused for stabs and can always be set to zero.
139 The OTHER field is almost always unused and can be set to zero.
141 The number in the TYPE field gives some basic information about
142 which type of stab this is (or whether it _is_ a stab, as opposed to an
143 ordinary symbol). Each valid type number defines a different stab
144 type; further, the stab type defines the exact interpretation of, and
145 possible values for, any remaining STRING, DESC, or VALUE fields
146 present in the stab. *Note Stab Types::, for a list in numeric order
147 of the valid TYPE field values for stab directives.
150 File: stabs.info, Node: String Field, Next: C Example, Prev: Stabs Format, Up: Overview
155 For most stabs the string field holds the meat of the debugging
156 information. The flexible nature of this field is what makes stabs
157 extensible. For some stab types the string field contains only a name.
158 For other stab types the contents can be a great deal more complex.
160 The overall format of the string field for most stab types is:
162 "NAME:SYMBOL-DESCRIPTOR TYPE-INFORMATION"
164 NAME is the name of the symbol represented by the stab; it can
165 contain a pair of colons (*note Nested Symbols::). NAME can be
166 omitted, which means the stab represents an unnamed object. For
167 example, `:t10=*2' defines type 10 as a pointer to type 2, but does not
168 give the type a name. Omitting the NAME field is supported by AIX dbx
169 and GDB after about version 4.8, but not other debuggers. GCC
170 sometimes uses a single space as the name instead of omitting the name
171 altogether; apparently that is supported by most debuggers.
173 The SYMBOL-DESCRIPTOR following the `:' is an alphabetic character
174 that tells more specifically what kind of symbol the stab represents.
175 If the SYMBOL-DESCRIPTOR is omitted, but type information follows, then
176 the stab represents a local variable. For a list of symbol
177 descriptors, see *Note Symbol Descriptors::. The `c' symbol descriptor
178 is an exception in that it is not followed by type information. *Note
181 TYPE-INFORMATION is either a TYPE-NUMBER, or `TYPE-NUMBER='. A
182 TYPE-NUMBER alone is a type reference, referring directly to a type
183 that has already been defined.
185 The `TYPE-NUMBER=' form is a type definition, where the number
186 represents a new type which is about to be defined. The type
187 definition may refer to other types by number, and those type numbers
188 may be followed by `=' and nested definitions. Also, the Lucid
189 compiler will repeat `TYPE-NUMBER=' more than once if it wants to
190 define several type numbers at once.
192 In a type definition, if the character that follows the equals sign
193 is non-numeric then it is a TYPE-DESCRIPTOR, and tells what kind of
194 type is about to be defined. Any other values following the
195 TYPE-DESCRIPTOR vary, depending on the TYPE-DESCRIPTOR. *Note Type
196 Descriptors::, for a list of TYPE-DESCRIPTOR values. If a number
197 follows the `=' then the number is a TYPE-REFERENCE. For a full
198 description of types, *Note Types::.
200 A TYPE-NUMBER is often a single number. The GNU and Sun tools
201 additionally permit a TYPE-NUMBER to be a pair
202 (FILE-NUMBER,FILETYPE-NUMBER) (the parentheses appear in the string,
203 and serve to distinguish the two cases). The FILE-NUMBER is 0 for the
204 base source file, 1 for the first included file, 2 for the next, and so
205 on. The FILETYPE-NUMBER is a number starting with 1 which is
206 incremented for each new type defined in the file. (Separating the
207 file number and the type number permits the `N_BINCL' optimization to
208 succeed more often; see *Note Include Files::).
210 There is an AIX extension for type attributes. Following the `='
211 are any number of type attributes. Each one starts with `@' and ends
212 with `;'. Debuggers, including AIX's dbx and GDB 4.10, skip any type
213 attributes they do not recognize. GDB 4.9 and other versions of dbx
214 may not do this. Because of a conflict with C++ (*note Cplusplus::),
215 new attributes should not be defined which begin with a digit, `(', or
216 `-'; GDB may be unable to distinguish those from the C++ type
217 descriptor `@'. The attributes are:
220 BOUNDARY is an integer specifying the alignment. I assume it
221 applies to all variables of this type.
224 Pointer class (for checking). Not sure what this means, or how
225 INTEGER is interpreted.
228 Indicate this is a packed type, meaning that structure fields or
229 array elements are placed more closely in memory, to save memory
230 at the expense of speed.
233 Size in bits of a variable of this type. This is fully supported
234 by GDB 4.11 and later.
237 Indicate that this type is a string instead of an array of
238 characters, or a bitstring instead of a set. It doesn't change
239 the layout of the data being represented, but does enable the
240 debugger to know which type it is.
243 Indicate that this type is a vector instead of an array. The only
244 major difference between vectors and arrays is that vectors are
245 passed by value instead of by reference (vector coprocessor
249 All of this can make the string field quite long. All versions of
250 GDB, and some versions of dbx, can handle arbitrarily long strings.
251 But many versions of dbx (or assemblers or linkers, I'm not sure which)
252 cretinously limit the strings to about 80 characters, so compilers which
253 must work with such systems need to split the `.stabs' directive into
254 several `.stabs' directives. Each stab duplicates every field except
255 the string field. The string field of every stab except the last is
256 marked as continued with a backslash at the end (in the assembly code
257 this may be written as a double backslash, depending on the assembler).
258 Removing the backslashes and concatenating the string fields of each
259 stab produces the original, long string. Just to be incompatible (or so
260 they don't have to worry about what the assembler does with
261 backslashes), AIX can use `?' instead of backslash.
264 File: stabs.info, Node: C Example, Next: Assembly Code, Prev: String Field, Up: Overview
266 1.4 A Simple Example in C Source
267 ================================
269 To get the flavor of how stabs describe source information for a C
270 program, let's look at the simple program:
274 printf("Hello world");
277 When compiled with `-g', the program above yields the following `.s'
278 file. Line numbers have been added to make it easier to refer to parts
279 of the `.s' file in the description of the stabs that follows.
282 File: stabs.info, Node: Assembly Code, Prev: C Example, Up: Overview
284 1.5 The Simple Example at the Assembly Level
285 ============================================
287 This simple "hello world" example demonstrates several of the stab
288 types used to describe C language source files.
291 2 .stabs "/cygint/s1/users/jcm/play/",100,0,0,Ltext0
292 3 .stabs "hello.c",100,0,0,Ltext0
295 6 .stabs "int:t1=r1;-2147483648;2147483647;",128,0,0,0
296 7 .stabs "char:t2=r2;0;127;",128,0,0,0
297 8 .stabs "long int:t3=r1;-2147483648;2147483647;",128,0,0,0
298 9 .stabs "unsigned int:t4=r1;0;-1;",128,0,0,0
299 10 .stabs "long unsigned int:t5=r1;0;-1;",128,0,0,0
300 11 .stabs "short int:t6=r1;-32768;32767;",128,0,0,0
301 12 .stabs "long long int:t7=r1;0;-1;",128,0,0,0
302 13 .stabs "short unsigned int:t8=r1;0;65535;",128,0,0,0
303 14 .stabs "long long unsigned int:t9=r1;0;-1;",128,0,0,0
304 15 .stabs "signed char:t10=r1;-128;127;",128,0,0,0
305 16 .stabs "unsigned char:t11=r1;0;255;",128,0,0,0
306 17 .stabs "float:t12=r1;4;0;",128,0,0,0
307 18 .stabs "double:t13=r1;8;0;",128,0,0,0
308 19 .stabs "long double:t14=r1;8;0;",128,0,0,0
309 20 .stabs "void:t15=15",128,0,0,0
312 23 .ascii "Hello, world!\12\0"
327 38 sethi %hi(LC0),%o1
328 39 or %o1,%lo(LC0),%o0
339 50 .stabs "main:F1",36,0,0,_main
340 51 .stabn 192,0,0,LBB2
341 52 .stabn 224,0,0,LBE2
344 File: stabs.info, Node: Program Structure, Next: Constants, Prev: Overview, Up: Top
346 2 Encoding the Structure of the Program
347 ***************************************
349 The elements of the program structure that stabs encode include the name
350 of the main function, the names of the source and include files, the
351 line numbers, procedure names and types, and the beginnings and ends of
356 * Main Program:: Indicate what the main program is
357 * Source Files:: The path and name of the source file
358 * Include Files:: Names of include files
361 * Nested Procedures::
363 * Alternate Entry Points:: Entering procedures except at the beginning.
366 File: stabs.info, Node: Main Program, Next: Source Files, Up: Program Structure
371 Most languages allow the main program to have any name. The `N_MAIN'
372 stab type tells the debugger the name that is used in this program.
373 Only the string field is significant; it is the name of a function
374 which is the main program. Most C compilers do not use this stab (they
375 expect the debugger to assume that the name is `main'), but some C
376 compilers emit an `N_MAIN' stab for the `main' function. I'm not sure
377 how XCOFF handles this.
380 File: stabs.info, Node: Source Files, Next: Include Files, Prev: Main Program, Up: Program Structure
382 2.2 Paths and Names of the Source Files
383 =======================================
385 Before any other stabs occur, there must be a stab specifying the source
386 file. This information is contained in a symbol of stab type `N_SO';
387 the string field contains the name of the file. The value of the
388 symbol is the start address of the portion of the text section
389 corresponding to that file.
391 Some compilers use the desc field to indicate the language of the
392 source file. Sun's compilers started this usage, and the first
393 constants are derived from their documentation. Languages added by
394 gcc/gdb start at 0x32 to avoid conflict with languages Sun may add in
395 the future. A desc field with a value 0 indicates that no language has
396 been specified via this mechanism.
416 `N_SO_FORTRAN90' (0x7)
422 `N_SO_OBJCPLUS' (0x33)
425 Some compilers (for example, GCC2 and SunOS4 `/bin/cc') also include
426 the directory in which the source was compiled, in a second `N_SO'
427 symbol preceding the one containing the file name. This symbol can be
428 distinguished by the fact that it ends in a slash. Code from the
429 `cfront' C++ compiler can have additional `N_SO' symbols for
430 nonexistent source files after the `N_SO' for the real source file;
431 these are believed to contain no useful information.
435 .stabs "/cygint/s1/users/jcm/play/",100,0,0,Ltext0 # 100 is N_SO
436 .stabs "hello.c",100,0,0,Ltext0
440 Instead of `N_SO' symbols, XCOFF uses a `.file' assembler directive
441 which assembles to a `C_FILE' symbol; explaining this in detail is
442 outside the scope of this document.
444 If it is useful to indicate the end of a source file, this is done
445 with an `N_SO' symbol with an empty string for the name. The value is
446 the address of the end of the text section for the file. For some
447 systems, there is no indication of the end of a source file, and you
448 just need to figure it ended when you see an `N_SO' for a different
449 source file, or a symbol ending in `.o' (which at least some linkers
450 insert to mark the start of a new `.o' file).
453 File: stabs.info, Node: Include Files, Next: Line Numbers, Prev: Source Files, Up: Program Structure
455 2.3 Names of Include Files
456 ==========================
458 There are several schemes for dealing with include files: the
459 traditional `N_SOL' approach, Sun's `N_BINCL' approach, and the XCOFF
460 `C_BINCL' approach (which despite the similar name has little in common
463 An `N_SOL' symbol specifies which include file subsequent symbols
464 refer to. The string field is the name of the file and the value is the
465 text address corresponding to the end of the previous include file and
466 the start of this one. To specify the main source file again, use an
467 `N_SOL' symbol with the name of the main source file.
469 The `N_BINCL' approach works as follows. An `N_BINCL' symbol
470 specifies the start of an include file. In an object file, only the
471 string is significant; the linker puts data into some of the other
472 fields. The end of the include file is marked by an `N_EINCL' symbol
473 (which has no string field). In an object file, there is no
474 significant data in the `N_EINCL' symbol. `N_BINCL' and `N_EINCL' can
477 If the linker detects that two source files have identical stabs
478 between an `N_BINCL' and `N_EINCL' pair (as will generally be the case
479 for a header file), then it only puts out the stabs once. Each
480 additional occurrence is replaced by an `N_EXCL' symbol. I believe the
481 GNU linker and the Sun (both SunOS4 and Solaris) linker are the only
482 ones which supports this feature.
484 A linker which supports this feature will set the value of a
485 `N_BINCL' symbol to the total of all the characters in the stabs
486 strings included in the header file, omitting any file numbers. The
487 value of an `N_EXCL' symbol is the same as the value of the `N_BINCL'
488 symbol it replaces. This information can be used to match up `N_EXCL'
489 and `N_BINCL' symbols which have the same filename. The `N_EINCL'
490 value, and the values of the other and description fields for all
491 three, appear to always be zero.
493 For the start of an include file in XCOFF, use the `.bi' assembler
494 directive, which generates a `C_BINCL' symbol. A `.ei' directive,
495 which generates a `C_EINCL' symbol, denotes the end of the include
496 file. Both directives are followed by the name of the source file in
497 quotes, which becomes the string for the symbol. The value of each
498 symbol, produced automatically by the assembler and linker, is the
499 offset into the executable of the beginning (inclusive, as you'd
500 expect) or end (inclusive, as you would not expect) of the portion of
501 the COFF line table that corresponds to this include file. `C_BINCL'
502 and `C_EINCL' do not nest.
505 File: stabs.info, Node: Line Numbers, Next: Procedures, Prev: Include Files, Up: Program Structure
510 An `N_SLINE' symbol represents the start of a source line. The desc
511 field contains the line number and the value contains the code address
512 for the start of that source line. On most machines the address is
513 absolute; for stabs in sections (*note Stab Sections::), it is relative
514 to the function in which the `N_SLINE' symbol occurs.
516 GNU documents `N_DSLINE' and `N_BSLINE' symbols for line numbers in
517 the data or bss segments, respectively. They are identical to
518 `N_SLINE' but are relocated differently by the linker. They were
519 intended to be used to describe the source location of a variable
520 declaration, but I believe that GCC2 actually puts the line number in
521 the desc field of the stab for the variable itself. GDB has been
522 ignoring these symbols (unless they contain a string field) since at
525 For single source lines that generate discontiguous code, such as
526 flow of control statements, there may be more than one line number
527 entry for the same source line. In this case there is a line number
528 entry at the start of each code range, each with the same line number.
530 XCOFF does not use stabs for line numbers. Instead, it uses COFF
531 line numbers (which are outside the scope of this document). Standard
532 COFF line numbers cannot deal with include files, but in XCOFF this is
533 fixed with the `C_BINCL' method of marking include files (*note Include
537 File: stabs.info, Node: Procedures, Next: Nested Procedures, Prev: Line Numbers, Up: Program Structure
542 All of the following stabs normally use the `N_FUN' symbol type.
543 However, Sun's `acc' compiler on SunOS4 uses `N_GSYM' and `N_STSYM',
544 which means that the value of the stab for the function is useless and
545 the debugger must get the address of the function from the non-stab
546 symbols instead. On systems where non-stab symbols have leading
547 underscores, the stabs will lack underscores and the debugger needs to
548 know about the leading underscore to match up the stab and the non-stab
549 symbol. BSD Fortran is said to use `N_FNAME' with the same
550 restriction; the value of the symbol is not useful (I'm not sure it
551 really does use this, because GDB doesn't handle this and no one has
554 A function is represented by an `F' symbol descriptor for a global
555 (extern) function, and `f' for a static (local) function. For a.out,
556 the value of the symbol is the address of the start of the function; it
557 is already relocated. For stabs in ELF, the SunPRO compiler version
558 2.0.1 and GCC put out an address which gets relocated by the linker.
559 In a future release SunPRO is planning to put out zero, in which case
560 the address can be found from the ELF (non-stab) symbol. Because
561 looking things up in the ELF symbols would probably be slow, I'm not
562 sure how to find which symbol of that name is the right one, and this
563 doesn't provide any way to deal with nested functions, it would
564 probably be better to make the value of the stab an address relative to
565 the start of the file, or just absolute. See *Note ELF Linker
566 Relocation:: for more information on linker relocation of stabs in ELF
567 files. For XCOFF, the stab uses the `C_FUN' storage class and the
568 value of the stab is meaningless; the address of the function can be
569 found from the csect symbol (XTY_LD/XMC_PR).
571 The type information of the stab represents the return type of the
572 function; thus `foo:f5' means that foo is a function returning type 5.
573 There is no need to try to get the line number of the start of the
574 function from the stab for the function; it is in the next `N_SLINE'
577 Some compilers (such as Sun's Solaris compiler) support an extension
578 for specifying the types of the arguments. I suspect this extension is
579 not used for old (non-prototyped) function definitions in C. If the
580 extension is in use, the type information of the stab for the function
581 is followed by type information for each argument, with each argument
582 preceded by `;'. An argument type of 0 means that additional arguments
583 are being passed, whose types and number may vary (`...' in ANSI C).
584 GDB has tolerated this extension (parsed the syntax, if not necessarily
585 used the information) since at least version 4.8; I don't know whether
586 all versions of dbx tolerate it. The argument types given here are not
587 redundant with the symbols for the formal parameters (*note
588 Parameters::); they are the types of the arguments as they are passed,
589 before any conversions might take place. For example, if a C function
590 which is declared without a prototype takes a `float' argument, the
591 value is passed as a `double' but then converted to a `float'.
592 Debuggers need to use the types given in the arguments when printing
593 values, but when calling the function they need to use the types given
594 in the symbol defining the function.
596 If the return type and types of arguments of a function which is
597 defined in another source file are specified (i.e., a function
598 prototype in ANSI C), traditionally compilers emit no stab; the only
599 way for the debugger to find the information is if the source file
600 where the function is defined was also compiled with debugging symbols.
601 As an extension the Solaris compiler uses symbol descriptor `P'
602 followed by the return type of the function, followed by the arguments,
603 each preceded by `;', as in a stab with symbol descriptor `f' or `F'.
604 This use of symbol descriptor `P' can be distinguished from its use for
605 register parameters (*note Register Parameters::) by the fact that it
606 has symbol type `N_FUN'.
608 The AIX documentation also defines symbol descriptor `J' as an
609 internal function. I assume this means a function nested within another
610 function. It also says symbol descriptor `m' is a module in Modula-2
613 Procedures (functions which do not return values) are represented as
614 functions returning the `void' type in C. I don't see why this couldn't
615 be used for all languages (inventing a `void' type for this purpose if
616 necessary), but the AIX documentation defines `I', `P', and `Q' for
617 internal, global, and static procedures, respectively. These symbol
618 descriptors are unusual in that they are not followed by type
621 The following example shows a stab for a function `main' which
622 returns type number `1'. The `_main' specified for the value is a
623 reference to an assembler label which is used to fill in the start
624 address of the function.
626 .stabs "main:F1",36,0,0,_main # 36 is N_FUN
628 The stab representing a procedure is located immediately following
629 the code of the procedure. This stab is in turn directly followed by a
630 group of other stabs describing elements of the procedure. These other
631 stabs describe the procedure's parameters, its block local variables,
632 and its block structure.
634 If functions can appear in different sections, then the debugger may
635 not be able to find the end of a function. Recent versions of GCC will
636 mark the end of a function with an `N_FUN' symbol with an empty string
637 for the name. The value is the address of the end of the current
638 function. Without such a symbol, there is no indication of the address
639 of the end of a function, and you must assume that it ended at the
640 starting address of the next function or at the end of the text section
644 File: stabs.info, Node: Nested Procedures, Next: Block Structure, Prev: Procedures, Up: Program Structure
646 2.6 Nested Procedures
647 =====================
649 For any of the symbol descriptors representing procedures, after the
650 symbol descriptor and the type information is optionally a scope
651 specifier. This consists of a comma, the name of the procedure, another
652 comma, and the name of the enclosing procedure. The first name is local
653 to the scope specified, and seems to be redundant with the name of the
654 symbol (before the `:'). This feature is used by GCC, and presumably
655 Pascal, Modula-2, etc., compilers, for nested functions.
657 If procedures are nested more than one level deep, only the
658 immediately containing scope is specified. For example, this code:
669 return baz (x + 2 * y);
671 return x + bar (3 * x);
676 .stabs "baz:f1,baz,bar",36,0,0,_baz.15 # 36 is N_FUN
677 .stabs "bar:f1,bar,foo",36,0,0,_bar.12
678 .stabs "foo:F1",36,0,0,_foo
681 File: stabs.info, Node: Block Structure, Next: Alternate Entry Points, Prev: Nested Procedures, Up: Program Structure
686 The program's block structure is represented by the `N_LBRAC' (left
687 brace) and the `N_RBRAC' (right brace) stab types. The variables
688 defined inside a block precede the `N_LBRAC' symbol for most compilers,
689 including GCC. Other compilers, such as the Convex, Acorn RISC
690 machine, and Sun `acc' compilers, put the variables after the `N_LBRAC'
691 symbol. The values of the `N_LBRAC' and `N_RBRAC' symbols are the
692 start and end addresses of the code of the block, respectively. For
693 most machines, they are relative to the starting address of this source
694 file. For the Gould NP1, they are absolute. For stabs in sections
695 (*note Stab Sections::), they are relative to the function in which
698 The `N_LBRAC' and `N_RBRAC' stabs that describe the block scope of a
699 procedure are located after the `N_FUN' stab that represents the
702 Sun documents the desc field of `N_LBRAC' and `N_RBRAC' symbols as
703 containing the nesting level of the block. However, dbx seems to not
704 care, and GCC always sets desc to zero.
706 For XCOFF, block scope is indicated with `C_BLOCK' symbols. If the
707 name of the symbol is `.bb', then it is the beginning of the block; if
708 the name of the symbol is `.be'; it is the end of the block.
711 File: stabs.info, Node: Alternate Entry Points, Prev: Block Structure, Up: Program Structure
713 2.8 Alternate Entry Points
714 ==========================
716 Some languages, like Fortran, have the ability to enter procedures at
717 some place other than the beginning. One can declare an alternate entry
718 point. The `N_ENTRY' stab is for this; however, the Sun FORTRAN
719 compiler doesn't use it. According to AIX documentation, only the name
720 of a `C_ENTRY' stab is significant; the address of the alternate entry
721 point comes from the corresponding external symbol. A previous
722 revision of this document said that the value of an `N_ENTRY' stab was
723 the address of the alternate entry point, but I don't know the source
724 for that information.
727 File: stabs.info, Node: Constants, Next: Variables, Prev: Program Structure, Up: Top
732 The `c' symbol descriptor indicates that this stab represents a
733 constant. This symbol descriptor is an exception to the general rule
734 that symbol descriptors are followed by type information. Instead, it
735 is followed by `=' and one of the following:
738 Boolean constant. VALUE is a numeric value; I assume it is 0 for
742 Character constant. VALUE is the numeric value of the constant.
744 `e TYPE-INFORMATION , VALUE'
745 Constant whose value can be represented as integral.
746 TYPE-INFORMATION is the type of the constant, as it would appear
747 after a symbol descriptor (*note String Field::). VALUE is the
748 numeric value of the constant. GDB 4.9 does not actually get the
749 right value if VALUE does not fit in a host `int', but it does not
750 do anything violent, and future debuggers could be extended to
751 accept integers of any size (whether unsigned or not). This
752 constant type is usually documented as being only for enumeration
753 constants, but GDB has never imposed that restriction; I don't
754 know about other debuggers.
757 Integer constant. VALUE is the numeric value. The type is some
758 sort of generic integer type (for GDB, a host `int'); to specify
759 the type explicitly, use `e' instead.
762 Real constant. VALUE is the real value, which can be `INF'
763 (optionally preceded by a sign) for infinity, `QNAN' for a quiet
764 NaN (not-a-number), or `SNAN' for a signalling NaN. If it is a
765 normal number the format is that accepted by the C library function
769 String constant. STRING is a string enclosed in either `'' (in
770 which case `'' characters within the string are represented as
771 `\'' or `"' (in which case `"' characters within the string are
772 represented as `\"').
774 `S TYPE-INFORMATION , ELEMENTS , BITS , PATTERN'
775 Set constant. TYPE-INFORMATION is the type of the constant, as it
776 would appear after a symbol descriptor (*note String Field::).
777 ELEMENTS is the number of elements in the set (does this means how
778 many bits of PATTERN are actually used, which would be redundant
779 with the type, or perhaps the number of bits set in PATTERN? I
780 don't get it), BITS is the number of bits in the constant (meaning
781 it specifies the length of PATTERN, I think), and PATTERN is a
782 hexadecimal representation of the set. AIX documentation refers
783 to a limit of 32 bytes, but I see no reason why this limit should
784 exist. This form could probably be used for arbitrary constants,
785 not just sets; the only catch is that PATTERN should be understood
786 to be target, not host, byte order and format.
788 The boolean, character, string, and set constants are not supported
789 by GDB 4.9, but it ignores them. GDB 4.8 and earlier gave an error
790 message and refused to read symbols from the file containing the
793 The above information is followed by `;'.
796 File: stabs.info, Node: Variables, Next: Types, Prev: Constants, Up: Top
801 Different types of stabs describe the various ways that variables can be
802 allocated: on the stack, globally, in registers, in common blocks,
803 statically, or as arguments to a function.
807 * Stack Variables:: Variables allocated on the stack.
808 * Global Variables:: Variables used by more than one source file.
809 * Register Variables:: Variables in registers.
810 * Common Blocks:: Variables statically allocated together.
811 * Statics:: Variables local to one source file.
812 * Based Variables:: Fortran pointer based variables.
813 * Parameters:: Variables for arguments to functions.
816 File: stabs.info, Node: Stack Variables, Next: Global Variables, Up: Variables
818 4.1 Automatic Variables Allocated on the Stack
819 ==============================================
821 If a variable's scope is local to a function and its lifetime is only as
822 long as that function executes (C calls such variables "automatic"), it
823 can be allocated in a register (*note Register Variables::) or on the
826 Each variable allocated on the stack has a stab with the symbol
827 descriptor omitted. Since type information should begin with a digit,
828 `-', or `(', only those characters precluded from being used for symbol
829 descriptors. However, the Acorn RISC machine (ARM) is said to get this
830 wrong: it puts out a mere type definition here, without the preceding
831 `TYPE-NUMBER='. This is a bad idea; there is no guarantee that type
832 descriptors are distinct from symbol descriptors. Stabs for stack
833 variables use the `N_LSYM' stab type, or `C_LSYM' for XCOFF.
835 The value of the stab is the offset of the variable within the local
836 variables. On most machines this is an offset from the frame pointer
837 and is negative. The location of the stab specifies which block it is
838 defined in; see *Note Block Structure::.
840 For example, the following C code:
848 produces the following stabs:
850 .stabs "main:F1",36,0,0,_main # 36 is N_FUN
851 .stabs "x:1",128,0,0,-12 # 128 is N_LSYM
852 .stabn 192,0,0,LBB2 # 192 is N_LBRAC
853 .stabn 224,0,0,LBE2 # 224 is N_RBRAC
855 See *Note Procedures:: for more information on the `N_FUN' stab, and
856 *Note Block Structure:: for more information on the `N_LBRAC' and
860 File: stabs.info, Node: Global Variables, Next: Register Variables, Prev: Stack Variables, Up: Variables
865 A variable whose scope is not specific to just one source file is
866 represented by the `G' symbol descriptor. These stabs use the `N_GSYM'
867 stab type (C_GSYM for XCOFF). The type information for the stab (*note
868 String Field::) gives the type of the variable.
870 For example, the following source code:
874 yields the following assembly code:
876 .stabs "g_foo:G2",32,0,0,0 # 32 is N_GSYM
882 The address of the variable represented by the `N_GSYM' is not
883 contained in the `N_GSYM' stab. The debugger gets this information
884 from the external symbol for the global variable. In the example above,
885 the `.global _g_foo' and `_g_foo:' lines tell the assembler to produce
888 Some compilers, like GCC, output `N_GSYM' stabs only once, where the
889 variable is defined. Other compilers, like SunOS4 /bin/cc, output a
890 `N_GSYM' stab for each compilation unit which references the variable.
893 File: stabs.info, Node: Register Variables, Next: Common Blocks, Prev: Global Variables, Up: Variables
895 4.3 Register Variables
896 ======================
898 Register variables have their own stab type, `N_RSYM' (`C_RSYM' for
899 XCOFF), and their own symbol descriptor, `r'. The stab's value is the
900 number of the register where the variable data will be stored.
902 AIX defines a separate symbol descriptor `d' for floating point
903 registers. This seems unnecessary; why not just just give floating
904 point registers different register numbers? I have not verified whether
905 the compiler actually uses `d'.
907 If the register is explicitly allocated to a global variable, but not
910 register int g_bar asm ("%g5");
912 then the stab may be emitted at the end of the object file, with the
916 File: stabs.info, Node: Common Blocks, Next: Statics, Prev: Register Variables, Up: Variables
921 A common block is a statically allocated section of memory which can be
922 referred to by several source files. It may contain several variables.
923 I believe Fortran is the only language with this feature.
925 A `N_BCOMM' stab begins a common block and an `N_ECOMM' stab ends
926 it. The only field that is significant in these two stabs is the
927 string, which names a normal (non-debugging) symbol that gives the
928 address of the common block. According to IBM documentation, only the
929 `N_BCOMM' has the name of the common block (even though their compiler
930 actually puts it both places).
932 The stabs for the members of the common block are between the
933 `N_BCOMM' and the `N_ECOMM'; the value of each stab is the offset
934 within the common block of that variable. IBM uses the `C_ECOML' stab
935 type, and there is a corresponding `N_ECOML' stab type, but Sun's
936 Fortran compiler uses `N_GSYM' instead. The variables within a common
937 block use the `V' symbol descriptor (I believe this is true of all
938 Fortran variables). Other stabs (at least type declarations using
939 `C_DECL') can also be between the `N_BCOMM' and the `N_ECOMM'.
942 File: stabs.info, Node: Statics, Next: Based Variables, Prev: Common Blocks, Up: Variables
947 Initialized static variables are represented by the `S' and `V' symbol
948 descriptors. `S' means file scope static, and `V' means procedure
949 scope static. One exception: in XCOFF, IBM's xlc compiler always uses
950 `V', and whether it is file scope or not is distinguished by whether
951 the stab is located within a function.
953 In a.out files, `N_STSYM' means the data section, `N_FUN' means the
954 text section, and `N_LCSYM' means the bss section. For those systems
955 with a read-only data section separate from the text section (Solaris),
956 `N_ROSYM' means the read-only data section.
958 For example, the source lines:
960 static const int var_const = 5;
961 static int var_init = 2;
962 static int var_noinit;
964 yield the following stabs:
966 .stabs "var_const:S1",36,0,0,_var_const # 36 is N_FUN
968 .stabs "var_init:S1",38,0,0,_var_init # 38 is N_STSYM
970 .stabs "var_noinit:S1",40,0,0,_var_noinit # 40 is N_LCSYM
972 In XCOFF files, the stab type need not indicate the section;
973 `C_STSYM' can be used for all statics. Also, each static variable is
974 enclosed in a static block. A `C_BSTAT' (emitted with a `.bs'
975 assembler directive) symbol begins the static block; its value is the
976 symbol number of the csect symbol whose value is the address of the
977 static block, its section is the section of the variables in that
978 static block, and its name is `.bs'. A `C_ESTAT' (emitted with a `.es'
979 assembler directive) symbol ends the static block; its name is `.es'
980 and its value and section are ignored.
982 In ECOFF files, the storage class is used to specify the section, so
983 the stab type need not indicate the section.
985 In ELF files, for the SunPRO compiler version 2.0.1, symbol
986 descriptor `S' means that the address is absolute (the linker relocates
987 it) and symbol descriptor `V' means that the address is relative to the
988 start of the relevant section for that compilation unit. SunPRO has
989 plans to have the linker stop relocating stabs; I suspect that their the
990 debugger gets the address from the corresponding ELF (not stab) symbol.
991 I'm not sure how to find which symbol of that name is the right one.
992 The clean way to do all this would be to have the value of a symbol
993 descriptor `S' symbol be an offset relative to the start of the file,
994 just like everything else, but that introduces obvious compatibility
995 problems. For more information on linker stab relocation, *Note ELF
999 File: stabs.info, Node: Based Variables, Next: Parameters, Prev: Statics, Up: Variables
1001 4.6 Fortran Based Variables
1002 ===========================
1004 Fortran (at least, the Sun and SGI dialects of FORTRAN-77) has a feature
1005 which allows allocating arrays with `malloc', but which avoids blurring
1006 the line between arrays and pointers the way that C does. In stabs
1007 such a variable uses the `b' symbol descriptor.
1009 For example, the Fortran declarations
1011 real foo, foo10(10), foo10_5(10,5)
1013 pointer (foo10p, foo10)
1014 pointer (foo105p, foo10_5)
1020 foo10_5:bar3;1;5;ar3;1;10;6
1022 In this example, `real' is type 6 and type 3 is an integral type
1023 which is the type of the subscripts of the array (probably `integer').
1025 The `b' symbol descriptor is like `V' in that it denotes a
1026 statically allocated symbol whose scope is local to a function; see
1027 *Note Statics::. The value of the symbol, instead of being the address
1028 of the variable itself, is the address of a pointer to that variable.
1029 So in the above example, the value of the `foo' stab is the address of
1030 a pointer to a real, the value of the `foo10' stab is the address of a
1031 pointer to a 10-element array of reals, and the value of the `foo10_5'
1032 stab is the address of a pointer to a 5-element array of 10-element
1036 File: stabs.info, Node: Parameters, Prev: Based Variables, Up: Variables
1041 Formal parameters to a function are represented by a stab (or sometimes
1042 two; see below) for each parameter. The stabs are in the order in which
1043 the debugger should print the parameters (i.e., the order in which the
1044 parameters are declared in the source file). The exact form of the stab
1045 depends on how the parameter is being passed.
1047 Parameters passed on the stack use the symbol descriptor `p' and the
1048 `N_PSYM' symbol type (or `C_PSYM' for XCOFF). The value of the symbol
1049 is an offset used to locate the parameter on the stack; its exact
1050 meaning is machine-dependent, but on most machines it is an offset from
1053 As a simple example, the code:
1061 .stabs "main:F1",36,0,0,_main # 36 is N_FUN
1062 .stabs "argc:p1",160,0,0,68 # 160 is N_PSYM
1063 .stabs "argv:p20=*21=*2",160,0,0,72
1065 The type definition of `argv' is interesting because it contains
1066 several type definitions. Type 21 is pointer to type 2 (char) and
1067 `argv' (type 20) is pointer to type 21.
1069 The following symbol descriptors are also said to go with `N_PSYM'.
1070 The value of the symbol is said to be an offset from the argument
1071 pointer (I'm not sure whether this is true or not).
1074 pF Fortran function parameter
1075 X (function result variable)
1079 * Register Parameters::
1080 * Local Variable Parameters::
1081 * Reference Parameters::
1082 * Conformant Arrays::
1085 File: stabs.info, Node: Register Parameters, Next: Local Variable Parameters, Up: Parameters
1087 4.7.1 Passing Parameters in Registers
1088 -------------------------------------
1090 If the parameter is passed in a register, then traditionally there are
1091 two symbols for each argument:
1093 .stabs "arg:p1" . . . ; N_PSYM
1094 .stabs "arg:r1" . . . ; N_RSYM
1096 Debuggers use the second one to find the value, and the first one to
1097 know that it is an argument.
1099 Because that approach is kind of ugly, some compilers use symbol
1100 descriptor `P' or `R' to indicate an argument which is in a register.
1101 Symbol type `C_RPSYM' is used in XCOFF and `N_RSYM' is used otherwise.
1102 The symbol's value is the register number. `P' and `R' mean the same
1103 thing; the difference is that `P' is a GNU invention and `R' is an IBM
1104 (XCOFF) invention. As of version 4.9, GDB should handle either one.
1106 There is at least one case where GCC uses a `p' and `r' pair rather
1107 than `P'; this is where the argument is passed in the argument list and
1108 then loaded into a register.
1110 According to the AIX documentation, symbol descriptor `D' is for a
1111 parameter passed in a floating point register. This seems
1112 unnecessary--why not just use `R' with a register number which
1113 indicates that it's a floating point register? I haven't verified
1114 whether the system actually does what the documentation indicates.
1116 On the sparc and hppa, for a `P' symbol whose type is a structure or
1117 union, the register contains the address of the structure. On the
1118 sparc, this is also true of a `p' and `r' pair (using Sun `cc') or a
1119 `p' symbol. However, if a (small) structure is really in a register,
1120 `r' is used. And, to top it all off, on the hppa it might be a
1121 structure which was passed on the stack and loaded into a register and
1122 for which there is a `p' and `r' pair! I believe that symbol
1123 descriptor `i' is supposed to deal with this case (it is said to mean
1124 "value parameter by reference, indirect access"; I don't know the
1125 source for this information), but I don't know details or what
1126 compilers or debuggers use it, if any (not GDB or GCC). It is not
1127 clear to me whether this case needs to be dealt with differently than
1128 parameters passed by reference (*note Reference Parameters::).
1131 File: stabs.info, Node: Local Variable Parameters, Next: Reference Parameters, Prev: Register Parameters, Up: Parameters
1133 4.7.2 Storing Parameters as Local Variables
1134 -------------------------------------------
1136 There is a case similar to an argument in a register, which is an
1137 argument that is actually stored as a local variable. Sometimes this
1138 happens when the argument was passed in a register and then the compiler
1139 stores it as a local variable. If possible, the compiler should claim
1140 that it's in a register, but this isn't always done.
1142 If a parameter is passed as one type and converted to a smaller type
1143 by the prologue (for example, the parameter is declared as a `float',
1144 but the calling conventions specify that it is passed as a `double'),
1145 then GCC2 (sometimes) uses a pair of symbols. The first symbol uses
1146 symbol descriptor `p' and the type which is passed. The second symbol
1147 has the type and location which the parameter actually has after the
1148 prologue. For example, suppose the following C code appears with no
1149 prototypes involved:
1156 if `f' is passed as a double at stack offset 8, and the prologue
1157 converts it to a float in register number 0, then the stabs look like:
1159 .stabs "f:p13",160,0,3,8 # 160 is `N_PSYM', here 13 is `double'
1160 .stabs "f:r12",64,0,3,0 # 64 is `N_RSYM', here 12 is `float'
1162 In both stabs 3 is the line number where `f' is declared (*note Line
1165 GCC, at least on the 960, has another solution to the same problem.
1166 It uses a single `p' symbol descriptor for an argument which is stored
1167 as a local variable but uses `N_LSYM' instead of `N_PSYM'. In this
1168 case, the value of the symbol is an offset relative to the local
1169 variables for that function, not relative to the arguments; on some
1170 machines those are the same thing, but not on all.
1172 On the VAX or on other machines in which the calling convention
1173 includes the number of words of arguments actually passed, the debugger
1174 (GDB at least) uses the parameter symbols to keep track of whether it
1175 needs to print nameless arguments in addition to the formal parameters
1176 which it has printed because each one has a stab. For example, in
1178 extern int fprintf (FILE *stream, char *format, ...);
1180 fprintf (stdout, "%d\n", x);
1182 there are stabs for `stream' and `format'. On most machines, the
1183 debugger can only print those two arguments (because it has no way of
1184 knowing that additional arguments were passed), but on the VAX or other
1185 machines with a calling convention which indicates the number of words
1186 of arguments, the debugger can print all three arguments. To do so,
1187 the parameter symbol (symbol descriptor `p') (not necessarily `r' or
1188 symbol descriptor omitted symbols) needs to contain the actual type as
1189 passed (for example, `double' not `float' if it is passed as a double
1190 and converted to a float).
1193 File: stabs.info, Node: Reference Parameters, Next: Conformant Arrays, Prev: Local Variable Parameters, Up: Parameters
1195 4.7.3 Passing Parameters by Reference
1196 -------------------------------------
1198 If the parameter is passed by reference (e.g., Pascal `VAR'
1199 parameters), then the symbol descriptor is `v' if it is in the argument
1200 list, or `a' if it in a register. Other than the fact that these
1201 contain the address of the parameter rather than the parameter itself,
1202 they are identical to `p' and `R', respectively. I believe `a' is an
1203 AIX invention; `v' is supported by all stabs-using systems as far as I
1207 File: stabs.info, Node: Conformant Arrays, Prev: Reference Parameters, Up: Parameters
1209 4.7.4 Passing Conformant Array Parameters
1210 -----------------------------------------
1212 Conformant arrays are a feature of Modula-2, and perhaps other
1213 languages, in which the size of an array parameter is not known to the
1214 called function until run-time. Such parameters have two stabs: a `x'
1215 for the array itself, and a `C', which represents the size of the
1216 array. The value of the `x' stab is the offset in the argument list
1217 where the address of the array is stored (it this right? it is a
1218 guess); the value of the `C' stab is the offset in the argument list
1219 where the size of the array (in elements? in bytes?) is stored.
1222 File: stabs.info, Node: Types, Next: Macro define and undefine, Prev: Variables, Up: Top
1227 The examples so far have described types as references to previously
1228 defined types, or defined in terms of subranges of or pointers to
1229 previously defined types. This chapter describes the other type
1230 descriptors that may follow the `=' in a type definition.
1234 * Builtin Types:: Integers, floating point, void, etc.
1235 * Miscellaneous Types:: Pointers, sets, files, etc.
1236 * Cross-References:: Referring to a type not yet defined.
1237 * Subranges:: A type with a specific range.
1238 * Arrays:: An aggregate type of same-typed elements.
1239 * Strings:: Like an array but also has a length.
1240 * Enumerations:: Like an integer but the values have names.
1241 * Structures:: An aggregate type of different-typed elements.
1242 * Typedefs:: Giving a type a name.
1243 * Unions:: Different types sharing storage.
1247 File: stabs.info, Node: Builtin Types, Next: Miscellaneous Types, Up: Types
1252 Certain types are built in (`int', `short', `void', `float', etc.); the
1253 debugger recognizes these types and knows how to handle them. Thus,
1254 don't be surprised if some of the following ways of specifying builtin
1255 types do not specify everything that a debugger would need to know
1256 about the type--in some cases they merely specify enough information to
1257 distinguish the type from other types.
1259 The traditional way to define builtin types is convoluted, so new
1260 ways have been invented to describe them. Sun's `acc' uses special
1261 builtin type descriptors (`b' and `R'), and IBM uses negative type
1262 numbers. GDB accepts all three ways, as of version 4.8; dbx just
1263 accepts the traditional builtin types and perhaps one of the other two
1264 formats. The following sections describe each of these formats.
1268 * Traditional Builtin Types:: Put on your seat belts and prepare for kludgery
1269 * Builtin Type Descriptors:: Builtin types with special type descriptors
1270 * Negative Type Numbers:: Builtin types using negative type numbers
1273 File: stabs.info, Node: Traditional Builtin Types, Next: Builtin Type Descriptors, Up: Builtin Types
1275 5.1.1 Traditional Builtin Types
1276 -------------------------------
1278 This is the traditional, convoluted method for defining builtin types.
1279 There are several classes of such type definitions: integer, floating
1284 * Traditional Integer Types::
1285 * Traditional Other Types::
1288 File: stabs.info, Node: Traditional Integer Types, Next: Traditional Other Types, Up: Traditional Builtin Types
1290 5.1.1.1 Traditional Integer Types
1291 .................................
1293 Often types are defined as subranges of themselves. If the bounding
1294 values fit within an `int', then they are given normally. For example:
1296 .stabs "int:t1=r1;-2147483648;2147483647;",128,0,0,0 # 128 is N_LSYM
1297 .stabs "char:t2=r2;0;127;",128,0,0,0
1299 Builtin types can also be described as subranges of `int':
1301 .stabs "unsigned short:t6=r1;0;65535;",128,0,0,0
1303 If the lower bound of a subrange is 0 and the upper bound is -1, the
1304 type is an unsigned integral type whose bounds are too big to describe
1305 in an `int'. Traditionally this is only used for `unsigned int' and
1308 .stabs "unsigned int:t4=r1;0;-1;",128,0,0,0
1310 For larger types, GCC 2.4.5 puts out bounds in octal, with one or
1311 more leading zeroes. In this case a negative bound consists of a number
1312 which is a 1 bit (for the sign bit) followed by a 0 bit for each bit in
1313 the number (except the sign bit), and a positive bound is one which is a
1314 1 bit for each bit in the number (except possibly the sign bit). All
1315 known versions of dbx and GDB version 4 accept this (at least in the
1316 sense of not refusing to process the file), but GDB 3.5 refuses to read
1317 the whole file containing such symbols. So GCC 2.3.3 did not output the
1318 proper size for these types. As an example of octal bounds, the string
1319 fields of the stabs for 64 bit integer types look like:
1321 long int:t3=r1;001000000000000000000000;000777777777777777777777;
1322 long unsigned int:t5=r1;000000000000000000000000;001777777777777777777777;
1324 If the lower bound of a subrange is 0 and the upper bound is
1325 negative, the type is an unsigned integral type whose size in bytes is
1326 the absolute value of the upper bound. I believe this is a Convex
1327 convention for `unsigned long long'.
1329 If the lower bound of a subrange is negative and the upper bound is
1330 0, the type is a signed integral type whose size in bytes is the
1331 absolute value of the lower bound. I believe this is a Convex
1332 convention for `long long'. To distinguish this from a legitimate
1333 subrange, the type should be a subrange of itself. I'm not sure whether
1334 this is the case for Convex.
1337 File: stabs.info, Node: Traditional Other Types, Prev: Traditional Integer Types, Up: Traditional Builtin Types
1339 5.1.1.2 Traditional Other Types
1340 ...............................
1342 If the upper bound of a subrange is 0 and the lower bound is positive,
1343 the type is a floating point type, and the lower bound of the subrange
1344 indicates the number of bytes in the type:
1346 .stabs "float:t12=r1;4;0;",128,0,0,0
1347 .stabs "double:t13=r1;8;0;",128,0,0,0
1349 However, GCC writes `long double' the same way it writes `double',
1350 so there is no way to distinguish.
1352 .stabs "long double:t14=r1;8;0;",128,0,0,0
1354 Complex types are defined the same way as floating-point types;
1355 there is no way to distinguish a single-precision complex from a
1356 double-precision floating-point type.
1358 The C `void' type is defined as itself:
1360 .stabs "void:t15=15",128,0,0,0
1362 I'm not sure how a boolean type is represented.
1365 File: stabs.info, Node: Builtin Type Descriptors, Next: Negative Type Numbers, Prev: Traditional Builtin Types, Up: Builtin Types
1367 5.1.2 Defining Builtin Types Using Builtin Type Descriptors
1368 -----------------------------------------------------------
1370 This is the method used by Sun's `acc' for defining builtin types.
1371 These are the type descriptors to define builtin types:
1373 `b SIGNED CHAR-FLAG WIDTH ; OFFSET ; NBITS ;'
1374 Define an integral type. SIGNED is `u' for unsigned or `s' for
1375 signed. CHAR-FLAG is `c' which indicates this is a character
1376 type, or is omitted. I assume this is to distinguish an integral
1377 type from a character type of the same size, for example it might
1378 make sense to set it for the C type `wchar_t' so the debugger can
1379 print such variables differently (Solaris does not do this). Sun
1380 sets it on the C types `signed char' and `unsigned char' which
1381 arguably is wrong. WIDTH and OFFSET appear to be for small
1382 objects stored in larger ones, for example a `short' in an `int'
1383 register. WIDTH is normally the number of bytes in the type.
1384 OFFSET seems to always be zero. NBITS is the number of bits in
1387 Note that type descriptor `b' used for builtin types conflicts with
1388 its use for Pascal space types (*note Miscellaneous Types::); they
1389 can be distinguished because the character following the type
1390 descriptor will be a digit, `(', or `-' for a Pascal space type, or
1391 `u' or `s' for a builtin type.
1394 Documented by AIX to define a wide character type, but their
1395 compiler actually uses negative type numbers (*note Negative Type
1398 `R FP-TYPE ; BYTES ;'
1399 Define a floating point type. FP-TYPE has one of the following
1403 IEEE 32-bit (single precision) floating point format.
1406 IEEE 64-bit (double precision) floating point format.
1413 These are for complex numbers. A comment in the GDB source
1414 describes them as Fortran `complex', `double complex', and
1415 `complex*16', respectively, but what does that mean? (i.e.,
1416 Single precision? Double precision?).
1419 Long double. This should probably only be used for Sun format
1420 `long double', and new codes should be used for other floating
1421 point formats (`NF_DOUBLE' can be used if a `long double' is
1422 really just an IEEE double, of course).
1424 BYTES is the number of bytes occupied by the type. This allows a
1425 debugger to perform some operations with the type even if it
1426 doesn't understand FP-TYPE.
1428 `g TYPE-INFORMATION ; NBITS'
1429 Documented by AIX to define a floating type, but their compiler
1430 actually uses negative type numbers (*note Negative Type
1433 `c TYPE-INFORMATION ; NBITS'
1434 Documented by AIX to define a complex type, but their compiler
1435 actually uses negative type numbers (*note Negative Type
1438 The C `void' type is defined as a signed integral type 0 bits long:
1439 .stabs "void:t19=bs0;0;0",128,0,0,0
1440 The Solaris compiler seems to omit the trailing semicolon in this
1441 case. Getting sloppy in this way is not a swift move because if a type
1442 is embedded in a more complex expression it is necessary to be able to
1445 I'm not sure how a boolean type is represented.
1448 File: stabs.info, Node: Negative Type Numbers, Prev: Builtin Type Descriptors, Up: Builtin Types
1450 5.1.3 Negative Type Numbers
1451 ---------------------------
1453 This is the method used in XCOFF for defining builtin types. Since the
1454 debugger knows about the builtin types anyway, the idea of negative
1455 type numbers is simply to give a special type number which indicates
1456 the builtin type. There is no stab defining these types.
1458 There are several subtle issues with negative type numbers.
1460 One is the size of the type. A builtin type (for example the C types
1461 `int' or `long') might have different sizes depending on compiler
1462 options, the target architecture, the ABI, etc. This issue doesn't
1463 come up for IBM tools since (so far) they just target the RS/6000; the
1464 sizes indicated below for each size are what the IBM RS/6000 tools use.
1465 To deal with differing sizes, either define separate negative type
1466 numbers for each size (which works but requires changing the debugger,
1467 and, unless you get both AIX dbx and GDB to accept the change,
1468 introduces an incompatibility), or use a type attribute (*note String
1469 Field::) to define a new type with the appropriate size (which merely
1470 requires a debugger which understands type attributes, like AIX dbx or
1473 .stabs "boolean:t10=@s8;-16",128,0,0,0
1475 defines an 8-bit boolean type, and
1477 .stabs "boolean:t10=@s64;-16",128,0,0,0
1479 defines a 64-bit boolean type.
1481 A similar issue is the format of the type. This comes up most often
1482 for floating-point types, which could have various formats (particularly
1483 extended doubles, which vary quite a bit even among IEEE systems).
1484 Again, it is best to define a new negative type number for each
1485 different format; changing the format based on the target system has
1486 various problems. One such problem is that the Alpha has both VAX and
1487 IEEE floating types. One can easily imagine one library using the VAX
1488 types and another library in the same executable using the IEEE types.
1489 Another example is that the interpretation of whether a boolean is true
1490 or false can be based on the least significant bit, most significant
1491 bit, whether it is zero, etc., and different compilers (or different
1492 options to the same compiler) might provide different kinds of boolean.
1494 The last major issue is the names of the types. The name of a given
1495 type depends _only_ on the negative type number given; these do not
1496 vary depending on the language, the target system, or anything else.
1497 One can always define separate type numbers--in the following list you
1498 will see for example separate `int' and `integer*4' types which are
1499 identical except for the name. But compatibility can be maintained by
1500 not inventing new negative type numbers and instead just defining a new
1501 type with a new name. For example:
1503 .stabs "CARDINAL:t10=-8",128,0,0,0
1505 Here is the list of negative type numbers. The phrase "integral
1506 type" is used to mean twos-complement (I strongly suspect that all
1507 machines which use stabs use twos-complement; most machines use
1508 twos-complement these days).
1511 `int', 32 bit signed integral type.
1514 `char', 8 bit type holding a character. Both GDB and dbx on AIX
1515 treat this as signed. GCC uses this type whether `char' is signed
1516 or not, which seems like a bad idea. The AIX compiler (`xlc')
1517 seems to avoid this type; it uses -5 instead for `char'.
1520 `short', 16 bit signed integral type.
1523 `long', 32 bit signed integral type.
1526 `unsigned char', 8 bit unsigned integral type.
1529 `signed char', 8 bit signed integral type.
1532 `unsigned short', 16 bit unsigned integral type.
1535 `unsigned int', 32 bit unsigned integral type.
1538 `unsigned', 32 bit unsigned integral type.
1541 `unsigned long', 32 bit unsigned integral type.
1544 `void', type indicating the lack of a value.
1547 `float', IEEE single precision.
1550 `double', IEEE double precision.
1553 `long double', IEEE double precision. The compiler claims the size
1554 will increase in a future release, and for binary compatibility
1555 you have to avoid using `long double'. I hope when they increase
1556 it they use a new negative type number.
1559 `integer'. 32 bit signed integral type.
1562 `boolean'. 32 bit type. GDB and GCC assume that zero is false,
1563 one is true, and other values have unspecified meaning. I hope
1564 this agrees with how the IBM tools use the type.
1567 `short real'. IEEE single precision.
1570 `real'. IEEE double precision.
1573 `stringptr'. *Note Strings::.
1576 `character', 8 bit unsigned character type.
1579 `logical*1', 8 bit type. This Fortran type has a split
1580 personality in that it is used for boolean variables, but can also
1581 be used for unsigned integers. 0 is false, 1 is true, and other
1582 values are non-boolean.
1585 `logical*2', 16 bit type. This Fortran type has a split
1586 personality in that it is used for boolean variables, but can also
1587 be used for unsigned integers. 0 is false, 1 is true, and other
1588 values are non-boolean.
1591 `logical*4', 32 bit type. This Fortran type has a split
1592 personality in that it is used for boolean variables, but can also
1593 be used for unsigned integers. 0 is false, 1 is true, and other
1594 values are non-boolean.
1597 `logical', 32 bit type. This Fortran type has a split personality
1598 in that it is used for boolean variables, but can also be used for
1599 unsigned integers. 0 is false, 1 is true, and other values are
1603 `complex'. A complex type consisting of two IEEE single-precision
1604 floating point values.
1607 `complex'. A complex type consisting of two IEEE double-precision
1608 floating point values.
1611 `integer*1', 8 bit signed integral type.
1614 `integer*2', 16 bit signed integral type.
1617 `integer*4', 32 bit signed integral type.
1620 `wchar'. Wide character, 16 bits wide, unsigned (what format?
1624 `long long', 64 bit signed integral type.
1627 `unsigned long long', 64 bit unsigned integral type.
1630 `logical*8', 64 bit unsigned integral type.
1633 `integer*8', 64 bit signed integral type.
1636 File: stabs.info, Node: Miscellaneous Types, Next: Cross-References, Prev: Builtin Types, Up: Types
1638 5.2 Miscellaneous Types
1639 =======================
1641 `b TYPE-INFORMATION ; BYTES'
1642 Pascal space type. This is documented by IBM; what does it mean?
1644 This use of the `b' type descriptor can be distinguished from its
1645 use for builtin integral types (*note Builtin Type Descriptors::)
1646 because the character following the type descriptor is always a
1649 `B TYPE-INFORMATION'
1650 A volatile-qualified version of TYPE-INFORMATION. This is a Sun
1651 extension. References and stores to a variable with a
1652 volatile-qualified type must not be optimized or cached; they must
1653 occur as the user specifies them.
1655 `d TYPE-INFORMATION'
1656 File of type TYPE-INFORMATION. As far as I know this is only used
1659 `k TYPE-INFORMATION'
1660 A const-qualified version of TYPE-INFORMATION. This is a Sun
1661 extension. A variable with a const-qualified type cannot be
1664 `M TYPE-INFORMATION ; LENGTH'
1665 Multiple instance type. The type seems to composed of LENGTH
1666 repetitions of TYPE-INFORMATION, for example `character*3' is
1667 represented by `M-2;3', where `-2' is a reference to a character
1668 type (*note Negative Type Numbers::). I'm not sure how this
1669 differs from an array. This appears to be a Fortran feature.
1670 LENGTH is a bound, like those in range types; see *Note
1673 `S TYPE-INFORMATION'
1674 Pascal set type. TYPE-INFORMATION must be a small type such as an
1675 enumeration or a subrange, and the type is a bitmask whose length
1676 is specified by the number of elements in TYPE-INFORMATION.
1678 In CHILL, if it is a bitstring instead of a set, also use the `S'
1679 type attribute (*note String Field::).
1681 `* TYPE-INFORMATION'
1682 Pointer to TYPE-INFORMATION.
1685 File: stabs.info, Node: Cross-References, Next: Subranges, Prev: Miscellaneous Types, Up: Types
1687 5.3 Cross-References to Other Types
1688 ===================================
1690 A type can be used before it is defined; one common way to deal with
1691 that situation is just to use a type reference to a type which has not
1694 Another way is with the `x' type descriptor, which is followed by
1695 `s' for a structure tag, `u' for a union tag, or `e' for a enumerator
1696 tag, followed by the name of the tag, followed by `:'. If the name
1697 contains `::' between a `<' and `>' pair (for C++ templates), such a
1698 `::' does not end the name--only a single `:' ends the name; see *Note
1701 For example, the following C declarations:
1708 .stabs "bar:G16=*17=xsfoo:",32,0,0,0
1710 Not all debuggers support the `x' type descriptor, so on some
1711 machines GCC does not use it. I believe that for the above example it
1712 would just emit a reference to type 17 and never define it, but I
1713 haven't verified that.
1715 Modula-2 imported types, at least on AIX, use the `i' type
1716 descriptor, which is followed by the name of the module from which the
1717 type is imported, followed by `:', followed by the name of the type.
1718 There is then optionally a comma followed by type information for the
1719 type. This differs from merely naming the type (*note Typedefs::) in
1720 that it identifies the module; I don't understand whether the name of
1721 the type given here is always just the same as the name we are giving
1722 it, or whether this type descriptor is used with a nameless stab (*note
1723 String Field::), or what. The symbol ends with `;'.
1726 File: stabs.info, Node: Subranges, Next: Arrays, Prev: Cross-References, Up: Types
1731 The `r' type descriptor defines a type as a subrange of another type.
1732 It is followed by type information for the type of which it is a
1733 subrange, a semicolon, an integral lower bound, a semicolon, an
1734 integral upper bound, and a semicolon. The AIX documentation does not
1735 specify the trailing semicolon, in an effort to specify array indexes
1736 more cleanly, but a subrange which is not an array index has always
1737 included a trailing semicolon (*note Arrays::).
1739 Instead of an integer, either bound can be one of the following:
1742 The bound is passed by reference on the stack at offset OFFSET
1743 from the argument list. *Note Parameters::, for more information
1747 The bound is passed by value on the stack at offset OFFSET from
1751 The bound is passed by reference in register number
1755 The bound is passed by value in register number REGISTER-NUMBER.
1760 Subranges are also used for builtin types; see *Note Traditional
1764 File: stabs.info, Node: Arrays, Next: Strings, Prev: Subranges, Up: Types
1769 Arrays use the `a' type descriptor. Following the type descriptor is
1770 the type of the index and the type of the array elements. If the index
1771 type is a range type, it ends in a semicolon; otherwise (for example,
1772 if it is a type reference), there does not appear to be any way to tell
1773 where the types are separated. In an effort to clean up this mess, IBM
1774 documents the two types as being separated by a semicolon, and a range
1775 type as not ending in a semicolon (but this is not right for range
1776 types which are not array indexes, *note Subranges::). I think
1777 probably the best solution is to specify that a semicolon ends a range
1778 type, and that the index type and element type of an array are
1779 separated by a semicolon, but that if the index type is a range type,
1780 the extra semicolon can be omitted. GDB (at least through version 4.9)
1781 doesn't support any kind of index type other than a range anyway; I'm
1784 It is well established, and widely used, that the type of the index,
1785 unlike most types found in the stabs, is merely a type definition, not
1786 type information (*note String Field::) (that is, it need not start with
1787 `TYPE-NUMBER=' if it is defining a new type). According to a comment
1788 in GDB, this is also true of the type of the array elements; it gives
1789 `ar1;1;10;ar1;1;10;4' as a legitimate way to express a two dimensional
1790 array. According to AIX documentation, the element type must be type
1791 information. GDB accepts either.
1793 The type of the index is often a range type, expressed as the type
1794 descriptor `r' and some parameters. It defines the size of the array.
1795 In the example below, the range `r1;0;2;' defines an index type which
1796 is a subrange of type 1 (integer), with a lower bound of 0 and an upper
1797 bound of 2. This defines the valid range of subscripts of a
1798 three-element C array.
1800 For example, the definition:
1802 char char_vec[3] = {'a','b','c'};
1804 produces the output:
1806 .stabs "char_vec:G19=ar1;0;2;2",32,0,0,0
1814 If an array is "packed", the elements are spaced more closely than
1815 normal, saving memory at the expense of speed. For example, an array
1816 of 3-byte objects might, if unpacked, have each element aligned on a
1817 4-byte boundary, but if packed, have no padding. One way to specify
1818 that something is packed is with type attributes (*note String
1819 Field::). In the case of arrays, another is to use the `P' type
1820 descriptor instead of `a'. Other than specifying a packed array, `P'
1821 is identical to `a'.
1823 An open array is represented by the `A' type descriptor followed by
1824 type information specifying the type of the array elements.
1826 An N-dimensional dynamic array is represented by
1828 D DIMENSIONS ; TYPE-INFORMATION
1830 DIMENSIONS is the number of dimensions; TYPE-INFORMATION specifies
1831 the type of the array elements.
1833 A subarray of an N-dimensional array is represented by
1835 E DIMENSIONS ; TYPE-INFORMATION
1837 DIMENSIONS is the number of dimensions; TYPE-INFORMATION specifies
1838 the type of the array elements.
1841 File: stabs.info, Node: Strings, Next: Enumerations, Prev: Arrays, Up: Types
1846 Some languages, like C or the original Pascal, do not have string types,
1847 they just have related things like arrays of characters. But most
1848 Pascals and various other languages have string types, which are
1849 indicated as follows:
1851 `n TYPE-INFORMATION ; BYTES'
1852 BYTES is the maximum length. I'm not sure what TYPE-INFORMATION
1853 is; I suspect that it means that this is a string of
1854 TYPE-INFORMATION (thus allowing a string of integers, a string of
1855 wide characters, etc., as well as a string of characters). Not
1856 sure what the format of this type is. This is an AIX feature.
1858 `z TYPE-INFORMATION ; BYTES'
1859 Just like `n' except that this is a gstring, not an ordinary
1860 string. I don't know the difference.
1863 Pascal Stringptr. What is this? This is an AIX feature.
1865 Languages, such as CHILL which have a string type which is basically
1866 just an array of characters use the `S' type attribute (*note String
1870 File: stabs.info, Node: Enumerations, Next: Structures, Prev: Strings, Up: Types
1875 Enumerations are defined with the `e' type descriptor.
1877 The source line below declares an enumeration type at file scope.
1878 The type definition is located after the `N_RBRAC' that marks the end of
1879 the previous procedure's block scope, and before the `N_FUN' that marks
1880 the beginning of the next procedure's block scope. Therefore it does
1881 not describe a block local symbol, but a file local one.
1885 enum e_places {first,second=3,last};
1887 generates the following stab:
1889 .stabs "e_places:T22=efirst:0,second:3,last:4,;",128,0,0,0
1891 The symbol descriptor (`T') says that the stab describes a
1892 structure, enumeration, or union tag. The type descriptor `e',
1893 following the `22=' of the type definition narrows it down to an
1894 enumeration type. Following the `e' is a list of the elements of the
1895 enumeration. The format is `NAME:VALUE,'. The list of elements ends
1896 with `;'. The fact that VALUE is specified as an integer can cause
1897 problems if the value is large. GCC 2.5.2 tries to output it in octal
1898 in that case with a leading zero, which is probably a good thing,
1899 although GDB 4.11 supports octal only in cases where decimal is
1900 perfectly good. Negative decimal values are supported by both GDB and
1903 There is no standard way to specify the size of an enumeration type;
1904 it is determined by the architecture (normally all enumerations types
1905 are 32 bits). Type attributes can be used to specify an enumeration
1906 type of another size for debuggers which support them; see *Note String
1909 Enumeration types are unusual in that they define symbols for the
1910 enumeration values (`first', `second', and `third' in the above
1911 example), and even though these symbols are visible in the file as a
1912 whole (rather than being in a more local namespace like structure
1913 member names), they are defined in the type definition for the
1914 enumeration type rather than each having their own symbol. In order to
1915 be fast, GDB will only get symbols from such types (in its initial scan
1916 of the stabs) if the type is the first thing defined after a `T' or `t'
1917 symbol descriptor (the above example fulfills this requirement). If
1918 the type does not have a name, the compiler should emit it in a
1919 nameless stab (*note String Field::); GCC does this.
1922 File: stabs.info, Node: Structures, Next: Typedefs, Prev: Enumerations, Up: Types
1927 The encoding of structures in stabs can be shown with an example.
1929 The following source code declares a structure tag and defines an
1930 instance of the structure in global scope. Then a `typedef' equates the
1931 structure tag with a new type. Separate stabs are generated for the
1932 structure tag, the structure `typedef', and the structure instance. The
1933 stabs for the tag and the `typedef' are emitted when the definitions are
1934 encountered. Since the structure elements are not initialized, the
1935 stab and code for the structure variable itself is located at the end
1936 of the program in the bss section.
1942 struct s_tag* s_next;
1945 typedef struct s_tag s_typedef;
1947 The structure tag has an `N_LSYM' stab type because, like the
1948 enumeration, the symbol has file scope. Like the enumeration, the
1949 symbol descriptor is `T', for enumeration, structure, or tag type. The
1950 type descriptor `s' following the `16=' of the type definition narrows
1951 the symbol type to structure.
1953 Following the `s' type descriptor is the number of bytes the
1954 structure occupies, followed by a description of each structure element.
1955 The structure element descriptions are of the form `NAME:TYPE, BIT
1956 OFFSET FROM THE START OF THE STRUCT, NUMBER OF BITS IN THE ELEMENT'.
1959 .stabs "s_tag:T16=s20s_int:1,0,32;s_float:12,32,32;
1960 s_char_vec:17=ar1;0;7;2,64,64;s_next:18=*16,128,32;;",128,0,0,0
1962 In this example, the first two structure elements are previously
1963 defined types. For these, the type following the `NAME:' part of the
1964 element description is a simple type reference. The other two structure
1965 elements are new types. In this case there is a type definition
1966 embedded after the `NAME:'. The type definition for the array element
1967 looks just like a type definition for a stand-alone array. The
1968 `s_next' field is a pointer to the same kind of structure that the
1969 field is an element of. So the definition of structure type 16
1970 contains a type definition for an element which is a pointer to type 16.
1972 If a field is a static member (this is a C++ feature in which a
1973 single variable appears to be a field of every structure of a given
1974 type) it still starts out with the field name, a colon, and the type,
1975 but then instead of a comma, bit position, comma, and bit size, there
1976 is a colon followed by the name of the variable which each such field
1979 If the structure has methods (a C++ feature), they follow the
1980 non-method fields; see *Note Cplusplus::.
1983 File: stabs.info, Node: Typedefs, Next: Unions, Prev: Structures, Up: Types
1985 5.9 Giving a Type a Name
1986 ========================
1988 To give a type a name, use the `t' symbol descriptor. The type is
1989 specified by the type information (*note String Field::) for the stab.
1992 .stabs "s_typedef:t16",128,0,0,0 # 128 is N_LSYM
1994 specifies that `s_typedef' refers to type number 16. Such stabs
1995 have symbol type `N_LSYM' (or `C_DECL' for XCOFF). (The Sun
1996 documentation mentions using `N_GSYM' in some cases).
1998 If you are specifying the tag name for a structure, union, or
1999 enumeration, use the `T' symbol descriptor instead. I believe C is the
2000 only language with this feature.
2002 If the type is an opaque type (I believe this is a Modula-2 feature),
2003 AIX provides a type descriptor to specify it. The type descriptor is
2004 `o' and is followed by a name. I don't know what the name means--is it
2005 always the same as the name of the type, or is this type descriptor
2006 used with a nameless stab (*note String Field::)? There optionally
2007 follows a comma followed by type information which defines the type of
2008 this type. If omitted, a semicolon is used in place of the comma and
2009 the type information, and the type is much like a generic pointer
2010 type--it has a known size but little else about it is specified.
2013 File: stabs.info, Node: Unions, Next: Function Types, Prev: Typedefs, Up: Types
2024 This code generates a stab for a union tag and a stab for a union
2025 variable. Both use the `N_LSYM' stab type. If a union variable is
2026 scoped locally to the procedure in which it is defined, its stab is
2027 located immediately preceding the `N_LBRAC' for the procedure's block
2030 The stab for the union tag, however, is located preceding the code
2031 for the procedure in which it is defined. The stab type is `N_LSYM'.
2032 This would seem to imply that the union type is file scope, like the
2033 struct type `s_tag'. This is not true. The contents and position of
2034 the stab for `u_type' do not convey any information about its procedure
2038 .stabs "u_tag:T23=u4u_int:1,0,32;u_float:12,0,32;u_char:21,0,32;;",
2041 The symbol descriptor `T', following the `name:' means that the stab
2042 describes an enumeration, structure, or union tag. The type descriptor
2043 `u', following the `23=' of the type definition, narrows it down to a
2044 union type definition. Following the `u' is the number of bytes in the
2045 union. After that is a list of union element descriptions. Their
2046 format is `NAME:TYPE, BIT OFFSET INTO THE UNION, NUMBER OF BYTES FOR
2049 The stab for the union variable is:
2051 .stabs "an_u:23",128,0,0,-20 # 128 is N_LSYM
2053 `-20' specifies where the variable is stored (*note Stack
2057 File: stabs.info, Node: Function Types, Prev: Unions, Up: Types
2062 Various types can be defined for function variables. These types are
2063 not used in defining functions (*note Procedures::); they are used for
2064 things like pointers to functions.
2066 The simple, traditional, type is type descriptor `f' is followed by
2067 type information for the return type of the function, followed by a
2070 This does not deal with functions for which the number and types of
2071 the parameters are part of the type, as in Modula-2 or ANSI C. AIX
2072 provides extensions to specify these, using the `f', `F', `p', and `R'
2075 First comes the type descriptor. If it is `f' or `F', this type
2076 involves a function rather than a procedure, and the type information
2077 for the return type of the function follows, followed by a comma. Then
2078 comes the number of parameters to the function and a semicolon. Then,
2079 for each parameter, there is the name of the parameter followed by a
2080 colon (this is only present for type descriptors `R' and `F' which
2081 represent Pascal function or procedure parameters), type information
2082 for the parameter, a comma, 0 if passed by reference or 1 if passed by
2083 value, and a semicolon. The type definition ends with a semicolon.
2085 For example, this variable definition:
2089 generates the following code:
2091 .stabs "g_pf:G24=*25=f1",32,0,0,0
2092 .common _g_pf,4,"bss"
2094 The variable defines a new type, 24, which is a pointer to another
2095 new type, 25, which is a function returning `int'.
2098 File: stabs.info, Node: Macro define and undefine, Next: Symbol Tables, Prev: Types, Up: Top
2100 6 Representation of #define and #undef
2101 **************************************
2103 This section describes the stabs support for macro define and undefine
2104 information, supported on some systems. (e.g., with `-g3' `-gstabs'
2107 A `#define MACRO-NAME MACRO-BODY' is represented with an
2108 `N_MAC_DEFINE' stab with a string field of `MACRO-NAME MACRO-BODY'.
2110 An `#undef MACRO-NAME' is represented with an `N_MAC_UNDEF' stabs
2111 with a string field of simply `MACRO-NAME'.
2113 For both `N_MAC_DEFINE' and `N_MAC_UNDEF', the desc field is the
2114 line number within the file where the corresponding `#define' or
2117 For example, the following C code:
2120 #define TWO(a, b) (a + (a) + 2 * b)
2121 #define ONE(c) (c + 19)
2123 main(int argc, char *argv[])
2125 func(NONE, TWO(10, 11));
2126 func(NONE, ONE(23));
2129 #define ONE(c) (c + 23)
2131 func(NONE, ONE(-23));
2138 func(int arg1, int arg2)
2140 global = arg1 + arg2;
2143 produces the following stabs (as well as many others):
2145 .stabs "NONE 42",54,0,1,0
2146 .stabs "TWO(a,b) (a + (a) + 2 * b)",54,0,2,0
2147 .stabs "ONE(c) (c + 19)",54,0,3,0
2148 .stabs "ONE",58,0,10,0
2149 .stabs "ONE(c) (c + 23)",54,0,11,0
2151 NOTE: In the above example, `54' is `N_MAC_DEFINE' and `58' is
2155 File: stabs.info, Node: Symbol Tables, Next: Cplusplus, Prev: Macro define and undefine, Up: Top
2157 7 Symbol Information in Symbol Tables
2158 *************************************
2160 This chapter describes the format of symbol table entries and how stab
2161 assembler directives map to them. It also describes the
2162 transformations that the assembler and linker make on data from stabs.
2166 * Symbol Table Format::
2167 * Transformations On Symbol Tables::
2170 File: stabs.info, Node: Symbol Table Format, Next: Transformations On Symbol Tables, Up: Symbol Tables
2172 7.1 Symbol Table Format
2173 =======================
2175 Each time the assembler encounters a stab directive, it puts each field
2176 of the stab into a corresponding field in a symbol table entry of its
2177 output file. If the stab contains a string field, the symbol table
2178 entry for that stab points to a string table entry containing the
2179 string data from the stab. Assembler labels become relocatable
2180 addresses. Symbol table entries in a.out have the format:
2182 struct internal_nlist {
2183 unsigned long n_strx; /* index into string table of name */
2184 unsigned char n_type; /* type of symbol */
2185 unsigned char n_other; /* misc info (usually empty) */
2186 unsigned short n_desc; /* description field */
2187 bfd_vma n_value; /* value of symbol */
2190 If the stab has a string, the `n_strx' field holds the offset in
2191 bytes of the string within the string table. The string is terminated
2192 by a NUL character. If the stab lacks a string (for example, it was
2193 produced by a `.stabn' or `.stabd' directive), the `n_strx' field is
2196 Symbol table entries with `n_type' field values greater than 0x1f
2197 originated as stabs generated by the compiler (with one random
2198 exception). The other entries were placed in the symbol table of the
2199 executable by the assembler or the linker.
2202 File: stabs.info, Node: Transformations On Symbol Tables, Prev: Symbol Table Format, Up: Symbol Tables
2204 7.2 Transformations on Symbol Tables
2205 ====================================
2207 The linker concatenates object files and does fixups of externally
2210 You can see the transformations made on stab data by the assembler
2211 and linker by examining the symbol table after each pass of the build.
2212 To do this, use `nm -ap', which dumps the symbol table, including
2213 debugging information, unsorted. For stab entries the columns are:
2214 VALUE, OTHER, DESC, TYPE, STRING. For assembler and linker symbols,
2215 the columns are: VALUE, TYPE, STRING.
2217 The low 5 bits of the stab type tell the linker how to relocate the
2218 value of the stab. Thus for stab types like `N_RSYM' and `N_LSYM',
2219 where the value is an offset or a register number, the low 5 bits are
2220 `N_ABS', which tells the linker not to relocate the value.
2222 Where the value of a stab contains an assembly language label, it is
2223 transformed by each build step. The assembler turns it into a
2224 relocatable address and the linker turns it into an absolute address.
2228 * Transformations On Static Variables::
2229 * Transformations On Global Variables::
2230 * Stab Section Transformations:: For some object file formats,
2231 things are a bit different.
2234 File: stabs.info, Node: Transformations On Static Variables, Next: Transformations On Global Variables, Up: Transformations On Symbol Tables
2236 7.2.1 Transformations on Static Variables
2237 -----------------------------------------
2239 This source line defines a static variable at file scope:
2241 static int s_g_repeat
2243 The following stab describes the symbol:
2245 .stabs "s_g_repeat:S1",38,0,0,_s_g_repeat
2247 The assembler transforms the stab into this symbol table entry in the
2248 `.o' file. The location is expressed as a data segment offset.
2250 00000084 - 00 0000 STSYM s_g_repeat:S1
2252 In the symbol table entry from the executable, the linker has made the
2253 relocatable address absolute.
2255 0000e00c - 00 0000 STSYM s_g_repeat:S1
2258 File: stabs.info, Node: Transformations On Global Variables, Next: Stab Section Transformations, Prev: Transformations On Static Variables, Up: Transformations On Symbol Tables
2260 7.2.2 Transformations on Global Variables
2261 -----------------------------------------
2263 Stabs for global variables do not contain location information. In this
2264 case, the debugger finds location information in the assembler or
2265 linker symbol table entry describing the variable. The source line:
2271 .stabs "g_foo:G2",32,0,0,0
2273 The variable is represented by two symbol table entries in the object
2274 file (see below). The first one originated as a stab. The second one
2275 is an external symbol. The upper case `D' signifies that the `n_type'
2276 field of the symbol table contains 7, `N_DATA' with local linkage. The
2277 stab's value is zero since the value is not used for `N_GSYM' stabs.
2278 The value of the linker symbol is the relocatable address corresponding
2281 00000000 - 00 0000 GSYM g_foo:G2
2284 These entries as transformed by the linker. The linker symbol table
2285 entry now holds an absolute address:
2287 00000000 - 00 0000 GSYM g_foo:G2
2292 File: stabs.info, Node: Stab Section Transformations, Prev: Transformations On Global Variables, Up: Transformations On Symbol Tables
2294 7.2.3 Transformations of Stabs in separate sections
2295 ---------------------------------------------------
2297 For object file formats using stabs in separate sections (*note Stab
2298 Sections::), use `objdump --stabs' instead of `nm' to show the stabs in
2299 an object or executable file. `objdump' is a GNU utility; Sun does not
2300 provide any equivalent.
2302 The following example is for a stab whose value is an address is
2303 relative to the compilation unit (*note ELF Linker Relocation::). For
2304 example, if the source line
2308 appears within a function, then the assembly language output from the
2313 .stabs "ld:V(0,3)",0x26,0,4,.L18-Ddata.data # 0x26 is N_STSYM
2319 Because the value is formed by subtracting one symbol from another,
2320 the value is absolute, not relocatable, and so the object file contains
2322 Symnum n_type n_othr n_desc n_value n_strx String
2323 31 STSYM 0 4 00000004 680 ld:V(0,3)
2325 without any relocations, and the executable file also contains
2327 Symnum n_type n_othr n_desc n_value n_strx String
2328 31 STSYM 0 4 00000004 680 ld:V(0,3)
2331 File: stabs.info, Node: Cplusplus, Next: Stab Types, Prev: Symbol Tables, Up: Top
2338 * Class Names:: C++ class names are both tags and typedefs.
2339 * Nested Symbols:: C++ symbol names can be within other types.
2340 * Basic Cplusplus Types::
2343 * Methods:: Method definition
2344 * Method Type Descriptor:: The `#' type descriptor
2345 * Member Type Descriptor:: The `@' type descriptor
2347 * Method Modifiers::
2350 * Virtual Base Classes::
2354 File: stabs.info, Node: Class Names, Next: Nested Symbols, Up: Cplusplus
2359 In C++, a class name which is declared with `class', `struct', or
2360 `union', is not only a tag, as in C, but also a type name. Thus there
2361 should be stabs with both `t' and `T' symbol descriptors (*note
2364 To save space, there is a special abbreviation for this case. If the
2365 `T' symbol descriptor is followed by `t', then the stab defines both a
2366 type name and a tag.
2368 For example, the C++ code
2370 struct foo {int x;};
2372 can be represented as either
2374 .stabs "foo:T19=s4x:1,0,32;;",128,0,0,0 # 128 is N_LSYM
2375 .stabs "foo:t19",128,0,0,0
2379 .stabs "foo:Tt19=s4x:1,0,32;;",128,0,0,0
2382 File: stabs.info, Node: Nested Symbols, Next: Basic Cplusplus Types, Prev: Class Names, Up: Cplusplus
2384 8.2 Defining a Symbol Within Another Type
2385 =========================================
2387 In C++, a symbol (such as a type name) can be defined within another
2390 In stabs, this is sometimes represented by making the name of a
2391 symbol which contains `::'. Such a pair of colons does not end the name
2392 of the symbol, the way a single colon would (*note String Field::). I'm
2393 not sure how consistently used or well thought out this mechanism is.
2394 So that a pair of colons in this position always has this meaning, `:'
2395 cannot be used as a symbol descriptor.
2397 For example, if the string for a stab is `foo::bar::baz:t5=*6', then
2398 `foo::bar::baz' is the name of the symbol, `t' is the symbol
2399 descriptor, and `5=*6' is the type information.
2402 File: stabs.info, Node: Basic Cplusplus Types, Next: Simple Classes, Prev: Nested Symbols, Up: Cplusplus
2404 8.3 Basic Types For C++
2405 =======================
2407 << the examples that follow are based on a01.C >>
2409 C++ adds two more builtin types to the set defined for C. These are
2410 the unknown type and the vtable record type. The unknown type, type
2411 16, is defined in terms of itself like the void type.
2413 The vtable record type, type 17, is defined as a structure type and
2414 then as a structure tag. The structure has four fields: delta, index,
2415 pfn, and delta2. pfn is the function pointer.
2417 << In boilerplate $vtbl_ptr_type, what are the fields delta, index,
2418 and delta2 used for? >>
2420 This basic type is present in all C++ programs even if there are no
2421 virtual methods defined.
2423 .stabs "struct_name:sym_desc(type)type_def(17)=type_desc(struct)struct_bytes(8)
2424 elem_name(delta):type_ref(short int),bit_offset(0),field_bits(16);
2425 elem_name(index):type_ref(short int),bit_offset(16),field_bits(16);
2426 elem_name(pfn):type_def(18)=type_desc(ptr to)type_ref(void),
2427 bit_offset(32),field_bits(32);
2428 elem_name(delta2):type_def(short int);bit_offset(32),field_bits(16);;"
2431 .stabs "$vtbl_ptr_type:t17=s8
2432 delta:6,0,16;index:6,16,16;pfn:18=*15,32,32;delta2:6,32,16;;"
2435 .stabs "name:sym_dec(struct tag)type_ref($vtbl_ptr_type)",N_LSYM,NIL,NIL,NIL
2437 .stabs "$vtbl_ptr_type:T17",128,0,0,0
2440 File: stabs.info, Node: Simple Classes, Next: Class Instance, Prev: Basic Cplusplus Types, Up: Cplusplus
2442 8.4 Simple Class Definition
2443 ===========================
2445 The stabs describing C++ language features are an extension of the
2446 stabs describing C. Stabs representing C++ class types elaborate
2447 extensively on the stab format used to describe structure types in C.
2448 Stabs representing class type variables look just like stabs
2449 representing C language variables.
2451 Consider the following very simple class definition.
2456 int Ameth(int in, char other);
2459 The class `baseA' is represented by two stabs. The first stab
2460 describes the class as a structure type. The second stab describes a
2461 structure tag of the class type. Both stabs are of stab type `N_LSYM'.
2462 Since the stab is not located between an `N_FUN' and an `N_LBRAC' stab
2463 this indicates that the class is defined at file scope. If it were,
2464 then the `N_LSYM' would signify a local variable.
2466 A stab describing a C++ class type is similar in format to a stab
2467 describing a C struct, with each class member shown as a field in the
2468 structure. The part of the struct format describing fields is expanded
2469 to include extra information relevant to C++ class members. In
2470 addition, if the class has multiple base classes or virtual functions
2471 the struct format outside of the field parts is also augmented.
2473 In this simple example the field part of the C++ class stab
2474 representing member data looks just like the field part of a C struct
2475 stab. The section on protections describes how its format is sometimes
2476 extended for member data.
2478 The field part of a C++ class stab representing a member function
2479 differs substantially from the field part of a C struct stab. It still
2480 begins with `name:' but then goes on to define a new type number for
2481 the member function, describe its return type, its argument types, its
2482 protection level, any qualifiers applied to the method definition, and
2483 whether the method is virtual or not. If the method is virtual then
2484 the method description goes on to give the vtable index of the method,
2485 and the type number of the first base class defining the method.
2487 When the field name is a method name it is followed by two colons
2488 rather than one. This is followed by a new type definition for the
2489 method. This is a number followed by an equal sign and the type of the
2490 method. Normally this will be a type declared using the `#' type
2491 descriptor; see *Note Method Type Descriptor::; static member functions
2492 are declared using the `f' type descriptor instead; see *Note Function
2495 The format of an overloaded operator method name differs from that of
2496 other methods. It is `op$::OPERATOR-NAME.' where OPERATOR-NAME is the
2497 operator name such as `+' or `+='. The name ends with a period, and
2498 any characters except the period can occur in the OPERATOR-NAME string.
2500 The next part of the method description represents the arguments to
2501 the method, preceded by a colon and ending with a semi-colon. The
2502 types of the arguments are expressed in the same way argument types are
2503 expressed in C++ name mangling. In this example an `int' and a `char'
2506 This is followed by a number, a letter, and an asterisk or period,
2507 followed by another semicolon. The number indicates the protections
2508 that apply to the member function. Here the 2 means public. The
2509 letter encodes any qualifier applied to the method definition. In this
2510 case, `A' means that it is a normal function definition. The dot shows
2511 that the method is not virtual. The sections that follow elaborate
2512 further on these fields and describe the additional information present
2513 for virtual methods.
2515 .stabs "class_name:sym_desc(type)type_def(20)=type_desc(struct)struct_bytes(4)
2516 field_name(Adat):type(int),bit_offset(0),field_bits(32);
2518 method_name(Ameth)::type_def(21)=type_desc(method)return_type(int);
2519 :arg_types(int char);
2520 protection(public)qualifier(normal)virtual(no);;"
2523 .stabs "baseA:t20=s4Adat:1,0,32;Ameth::21=##1;:ic;2A.;;",128,0,0,0
2525 .stabs "class_name:sym_desc(struct tag)",N_LSYM,NIL,NIL,NIL
2527 .stabs "baseA:T20",128,0,0,0
2530 File: stabs.info, Node: Class Instance, Next: Methods, Prev: Simple Classes, Up: Cplusplus
2535 As shown above, describing even a simple C++ class definition is
2536 accomplished by massively extending the stab format used in C to
2537 describe structure types. However, once the class is defined, C stabs
2538 with no modifications can be used to describe class instances. The
2545 yields the following stab describing the class instance. It looks no
2546 different from a standard C stab describing a local variable.
2548 .stabs "name:type_ref(baseA)", N_LSYM, NIL, NIL, frame_ptr_offset
2550 .stabs "AbaseA:20",128,0,0,-20
2553 File: stabs.info, Node: Methods, Next: Method Type Descriptor, Prev: Class Instance, Up: Cplusplus
2555 8.6 Method Definition
2556 =====================
2558 The class definition shown above declares Ameth. The C++ source below
2562 baseA::Ameth(int in, char other)
2567 This method definition yields three stabs following the code of the
2568 method. One stab describes the method itself and following two describe
2569 its parameters. Although there is only one formal argument all methods
2570 have an implicit argument which is the `this' pointer. The `this'
2571 pointer is a pointer to the object on which the method was called. Note
2572 that the method name is mangled to encode the class name and argument
2573 types. Name mangling is described in the ARM (`The Annotated C++
2574 Reference Manual', by Ellis and Stroustrup, ISBN 0-201-51459-1);
2575 `gpcompare.texi' in Cygnus GCC distributions describes the differences
2576 between GNU mangling and ARM mangling.
2578 .stabs "name:symbol_descriptor(global function)return_type(int)",
2579 N_FUN, NIL, NIL, code_addr_of_method_start
2581 .stabs "Ameth__5baseAic:F1",36,0,0,_Ameth__5baseAic
2583 Here is the stab for the `this' pointer implicit argument. The name
2584 of the `this' pointer is always `this'. Type 19, the `this' pointer is
2585 defined as a pointer to type 20, `baseA', but a stab defining `baseA'
2586 has not yet been emitted. Since the compiler knows it will be emitted
2587 shortly, here it just outputs a cross reference to the undefined
2588 symbol, by prefixing the symbol name with `xs'.
2590 .stabs "name:sym_desc(register param)type_def(19)=
2591 type_desc(ptr to)type_ref(baseA)=
2592 type_desc(cross-reference to)baseA:",N_RSYM,NIL,NIL,register_number
2594 .stabs "this:P19=*20=xsbaseA:",64,0,0,8
2596 The stab for the explicit integer argument looks just like a
2597 parameter to a C function. The last field of the stab is the offset
2598 from the argument pointer, which in most systems is the same as the
2601 .stabs "name:sym_desc(value parameter)type_ref(int)",
2602 N_PSYM,NIL,NIL,offset_from_arg_ptr
2604 .stabs "in:p1",160,0,0,72
2606 << The examples that follow are based on A1.C >>
2609 File: stabs.info, Node: Method Type Descriptor, Next: Member Type Descriptor, Prev: Methods, Up: Cplusplus
2611 8.7 The `#' Type Descriptor
2612 ===========================
2614 This is used to describe a class method. This is a function which takes
2615 an extra argument as its first argument, for the `this' pointer.
2617 If the `#' is immediately followed by another `#', the second one
2618 will be followed by the return type and a semicolon. The class and
2619 argument types are not specified, and must be determined by demangling
2620 the name of the method if it is available.
2622 Otherwise, the single `#' is followed by the class type, a comma,
2623 the return type, a comma, and zero or more parameter types separated by
2624 commas. The list of arguments is terminated by a semicolon. In the
2625 debugging output generated by gcc, a final argument type of `void'
2626 indicates a method which does not take a variable number of arguments.
2627 If the final argument type of `void' does not appear, the method was
2628 declared with an ellipsis.
2630 Note that although such a type will normally be used to describe
2631 fields in structures, unions, or classes, for at least some versions of
2632 the compiler it can also be used in other contexts.
2635 File: stabs.info, Node: Member Type Descriptor, Next: Protections, Prev: Method Type Descriptor, Up: Cplusplus
2637 8.8 The `@' Type Descriptor
2638 ===========================
2640 The `@' type descriptor is used for a pointer-to-non-static-member-data
2641 type. It is followed by type information for the class (or union), a
2642 comma, and type information for the member data.
2644 The following C++ source:
2646 typedef int A::*int_in_a;
2648 generates the following stab:
2650 .stabs "int_in_a:t20=21=@19,1",128,0,0,0
2652 Note that there is a conflict between this and type attributes
2653 (*note String Field::); both use type descriptor `@'. Fortunately, the
2654 `@' type descriptor used in this C++ sense always will be followed by a
2655 digit, `(', or `-', and type attributes never start with those things.
2658 File: stabs.info, Node: Protections, Next: Method Modifiers, Prev: Member Type Descriptor, Up: Cplusplus
2663 In the simple class definition shown above all member data and
2664 functions were publicly accessible. The example that follows contrasts
2665 public, protected and privately accessible fields and shows how these
2666 protections are encoded in C++ stabs.
2668 If the character following the `FIELD-NAME:' part of the string is
2669 `/', then the next character is the visibility. `0' means private, `1'
2670 means protected, and `2' means public. Debuggers should ignore
2671 visibility characters they do not recognize, and assume a reasonable
2672 default (such as public) (GDB 4.11 does not, but this should be fixed
2673 in the next GDB release). If no visibility is specified the field is
2674 public. The visibility `9' means that the field has been optimized out
2675 and is public (there is no way to specify an optimized out field with a
2676 private or protected visibility). Visibility `9' is not supported by
2677 GDB 4.11; this should be fixed in the next GDB release.
2679 The following C++ source:
2690 generates the following stab:
2693 .stabs "vis:T19=s12priv:/01,0,32;prot:/12,32,8;pub:12,64,32;;",128,0,0,0
2695 `vis:T19=s12' indicates that type number 19 is a 12 byte structure
2696 named `vis' The `priv' field has public visibility (`/0'), type int
2697 (`1'), and offset and size `,0,32;'. The `prot' field has protected
2698 visibility (`/1'), type char (`2') and offset and size `,32,8;'. The
2699 `pub' field has type float (`12'), and offset and size `,64,32;'.
2701 Protections for member functions are signified by one digit embedded
2702 in the field part of the stab describing the method. The digit is 0 if
2703 private, 1 if protected and 2 if public. Consider the C++ class
2708 int priv_meth(int in){return in;};
2710 char protMeth(char in){return in;};
2712 float pubMeth(float in){return in;};
2715 It generates the following stab. The digit in question is to the
2716 left of an `A' in each case. Notice also that in this case two symbol
2717 descriptors apply to the class name struct tag and struct type.
2719 .stabs "class_name:sym_desc(struct tag&type)type_def(21)=
2720 sym_desc(struct)struct_bytes(1)
2721 meth_name::type_def(22)=sym_desc(method)returning(int);
2722 :args(int);protection(private)modifier(normal)virtual(no);
2723 meth_name::type_def(23)=sym_desc(method)returning(char);
2724 :args(char);protection(protected)modifier(normal)virtual(no);
2725 meth_name::type_def(24)=sym_desc(method)returning(float);
2726 :args(float);protection(public)modifier(normal)virtual(no);;",
2729 .stabs "all_methods:Tt21=s1priv_meth::22=##1;:i;0A.;protMeth::23=##2;:c;1A.;
2730 pubMeth::24=##12;:f;2A.;;",128,0,0,0
2733 File: stabs.info, Node: Method Modifiers, Next: Virtual Methods, Prev: Protections, Up: Cplusplus
2735 8.10 Method Modifiers (`const', `volatile', `const volatile')
2736 =============================================================
2740 In the class example described above all the methods have the normal
2741 modifier. This method modifier information is located just after the
2742 protection information for the method. This field has four possible
2743 character values. Normal methods use `A', const methods use `B',
2744 volatile methods use `C', and const volatile methods use `D'. Consider
2745 the class definition below:
2749 int ConstMeth (int arg) const { return arg; };
2750 char VolatileMeth (char arg) volatile { return arg; };
2751 float ConstVolMeth (float arg) const volatile {return arg; };
2754 This class is described by the following stab:
2756 .stabs "class(A):sym_desc(struct)type_def(20)=type_desc(struct)struct_bytes(1)
2757 meth_name(ConstMeth)::type_def(21)sym_desc(method)
2758 returning(int);:arg(int);protection(public)modifier(const)virtual(no);
2759 meth_name(VolatileMeth)::type_def(22)=sym_desc(method)
2760 returning(char);:arg(char);protection(public)modifier(volatile)virt(no)
2761 meth_name(ConstVolMeth)::type_def(23)=sym_desc(method)
2762 returning(float);:arg(float);protection(public)modifier(const volatile)
2765 .stabs "A:T20=s1ConstMeth::21=##1;:i;2B.;VolatileMeth::22=##2;:c;2C.;
2766 ConstVolMeth::23=##12;:f;2D.;;",128,0,0,0
2769 File: stabs.info, Node: Virtual Methods, Next: Inheritance, Prev: Method Modifiers, Up: Cplusplus
2771 8.11 Virtual Methods
2772 ====================
2774 << The following examples are based on a4.C >>
2776 The presence of virtual methods in a class definition adds additional
2777 data to the class description. The extra data is appended to the
2778 description of the virtual method and to the end of the class
2779 description. Consider the class definition below:
2784 virtual int A_virt (int arg) { return arg; };
2787 This results in the stab below describing class A. It defines a new
2788 type (20) which is an 8 byte structure. The first field of the class
2789 struct is `Adat', an integer, starting at structure offset 0 and
2792 The second field in the class struct is not explicitly defined by the
2793 C++ class definition but is implied by the fact that the class contains
2794 a virtual method. This field is the vtable pointer. The name of the
2795 vtable pointer field starts with `$vf' and continues with a type
2796 reference to the class it is part of. In this example the type
2797 reference for class A is 20 so the name of its vtable pointer field is
2798 `$vf20', followed by the usual colon.
2800 Next there is a type definition for the vtable pointer type (21).
2801 This is in turn defined as a pointer to another new type (22).
2803 Type 22 is the vtable itself, which is defined as an array, indexed
2804 by a range of integers between 0 and 1, and whose elements are of type
2805 17. Type 17 was the vtable record type defined by the boilerplate C++
2806 type definitions, as shown earlier.
2808 The bit offset of the vtable pointer field is 32. The number of bits
2809 in the field are not specified when the field is a vtable pointer.
2811 Next is the method definition for the virtual member function
2812 `A_virt'. Its description starts out using the same format as the
2813 non-virtual member functions described above, except instead of a dot
2814 after the `A' there is an asterisk, indicating that the function is
2815 virtual. Since is is virtual some addition information is appended to
2816 the end of the method description.
2818 The first number represents the vtable index of the method. This is
2819 a 32 bit unsigned number with the high bit set, followed by a
2822 The second number is a type reference to the first base class in the
2823 inheritance hierarchy defining the virtual member function. In this
2824 case the class stab describes a base class so the virtual function is
2825 not overriding any other definition of the method. Therefore the
2826 reference is to the type number of the class that the stab is
2829 This is followed by three semi-colons. One marks the end of the
2830 current sub-section, one marks the end of the method field, and the
2831 third marks the end of the struct definition.
2833 For classes containing virtual functions the very last section of the
2834 string part of the stab holds a type reference to the first base class.
2835 This is preceded by `~%' and followed by a final semi-colon.
2837 .stabs "class_name(A):type_def(20)=sym_desc(struct)struct_bytes(8)
2838 field_name(Adat):type_ref(int),bit_offset(0),field_bits(32);
2839 field_name(A virt func ptr):type_def(21)=type_desc(ptr to)type_def(22)=
2840 sym_desc(array)index_type_ref(range of int from 0 to 1);
2841 elem_type_ref(vtbl elem type),
2843 meth_name(A_virt)::typedef(23)=sym_desc(method)returning(int);
2844 :arg_type(int),protection(public)normal(yes)virtual(yes)
2845 vtable_index(1);class_first_defining(A);;;~%first_base(A);",
2848 .stabs "A:t20=s8Adat:1,0,32;$vf20:21=*22=ar1;0;1;17,32;
2849 A_virt::23=##1;:i;2A*-2147483647;20;;;~%20;",128,0,0,0
2852 File: stabs.info, Node: Inheritance, Next: Virtual Base Classes, Prev: Virtual Methods, Up: Cplusplus
2857 Stabs describing C++ derived classes include additional sections that
2858 describe the inheritance hierarchy of the class. A derived class stab
2859 also encodes the number of base classes. For each base class it tells
2860 if the base class is virtual or not, and if the inheritance is private
2861 or public. It also gives the offset into the object of the portion of
2862 the object corresponding to each base class.
2864 This additional information is embedded in the class stab following
2865 the number of bytes in the struct. First the number of base classes
2866 appears bracketed by an exclamation point and a comma.
2868 Then for each base type there repeats a series: a virtual character,
2869 a visibility character, a number, a comma, another number, and a
2872 The virtual character is `1' if the base class is virtual and `0' if
2873 not. The visibility character is `2' if the derivation is public, `1'
2874 if it is protected, and `0' if it is private. Debuggers should ignore
2875 virtual or visibility characters they do not recognize, and assume a
2876 reasonable default (such as public and non-virtual) (GDB 4.11 does not,
2877 but this should be fixed in the next GDB release).
2879 The number following the virtual and visibility characters is the
2880 offset from the start of the object to the part of the object
2881 pertaining to the base class.
2883 After the comma, the second number is a type_descriptor for the base
2884 type. Finally a semi-colon ends the series, which repeats for each
2887 The source below defines three base classes `A', `B', and `C' and
2888 the derived class `D'.
2893 virtual int A_virt (int arg) { return arg; };
2899 virtual int B_virt (int arg) {return arg; };
2905 virtual int C_virt (int arg) {return arg; };
2908 class D : A, virtual B, public C {
2911 virtual int A_virt (int arg ) { return arg+1; };
2912 virtual int B_virt (int arg) { return arg+2; };
2913 virtual int C_virt (int arg) { return arg+3; };
2914 virtual int D_virt (int arg) { return arg; };
2917 Class stabs similar to the ones described earlier are generated for
2920 .stabs "A:T20=s8Adat:1,0,32;$vf20:21=*22=ar1;0;1;17,32;
2921 A_virt::23=##1;:i;2A*-2147483647;20;;;~%20;",128,0,0,0
2923 .stabs "B:Tt25=s8Bdat:1,0,32;$vf25:21,32;B_virt::26=##1;
2924 :i;2A*-2147483647;25;;;~%25;",128,0,0,0
2926 .stabs "C:Tt28=s8Cdat:1,0,32;$vf28:21,32;C_virt::29=##1;
2927 :i;2A*-2147483647;28;;;~%28;",128,0,0,0
2929 In the stab describing derived class `D' below, the information about
2930 the derivation of this class is encoded as follows.
2932 .stabs "derived_class_name:symbol_descriptors(struct tag&type)=
2933 type_descriptor(struct)struct_bytes(32)!num_bases(3),
2934 base_virtual(no)inheritance_public(no)base_offset(0),
2935 base_class_type_ref(A);
2936 base_virtual(yes)inheritance_public(no)base_offset(NIL),
2937 base_class_type_ref(B);
2938 base_virtual(no)inheritance_public(yes)base_offset(64),
2939 base_class_type_ref(C); ...
2941 .stabs "D:Tt31=s32!3,000,20;100,25;0264,28;$vb25:24,128;Ddat:
2942 1,160,32;A_virt::32=##1;:i;2A*-2147483647;20;;B_virt:
2943 :32:i;2A*-2147483647;25;;C_virt::32:i;2A*-2147483647;
2944 28;;D_virt::32:i;2A*-2147483646;31;;;~%20;",128,0,0,0
2947 File: stabs.info, Node: Virtual Base Classes, Next: Static Members, Prev: Inheritance, Up: Cplusplus
2949 8.13 Virtual Base Classes
2950 =========================
2952 A derived class object consists of a concatenation in memory of the data
2953 areas defined by each base class, starting with the leftmost and ending
2954 with the rightmost in the list of base classes. The exception to this
2955 rule is for virtual inheritance. In the example above, class `D'
2956 inherits virtually from base class `B'. This means that an instance of
2957 a `D' object will not contain its own `B' part but merely a pointer to
2958 a `B' part, known as a virtual base pointer.
2960 In a derived class stab, the base offset part of the derivation
2961 information, described above, shows how the base class parts are
2962 ordered. The base offset for a virtual base class is always given as 0.
2963 Notice that the base offset for `B' is given as 0 even though `B' is
2964 not the first base class. The first base class `A' starts at offset 0.
2966 The field information part of the stab for class `D' describes the
2967 field which is the pointer to the virtual base class `B'. The vbase
2968 pointer name is `$vb' followed by a type reference to the virtual base
2969 class. Since the type id for `B' in this example is 25, the vbase
2970 pointer name is `$vb25'.
2972 .stabs "D:Tt31=s32!3,000,20;100,25;0264,28;$vb25:24,128;Ddat:1,
2973 160,32;A_virt::32=##1;:i;2A*-2147483647;20;;B_virt::32:i;
2974 2A*-2147483647;25;;C_virt::32:i;2A*-2147483647;28;;D_virt:
2975 :32:i;2A*-2147483646;31;;;~%20;",128,0,0,0
2977 Following the name and a semicolon is a type reference describing the
2978 type of the virtual base class pointer, in this case 24. Type 24 was
2979 defined earlier as the type of the `B' class `this' pointer. The
2980 `this' pointer for a class is a pointer to the class type.
2982 .stabs "this:P24=*25=xsB:",64,0,0,8
2984 Finally the field offset part of the vbase pointer field description
2985 shows that the vbase pointer is the first field in the `D' object,
2986 before any data fields defined by the class. The layout of a `D' class
2987 object is a follows, `Adat' at 0, the vtable pointer for `A' at 32,
2988 `Cdat' at 64, the vtable pointer for C at 96, the virtual base pointer
2989 for `B' at 128, and `Ddat' at 160.
2992 File: stabs.info, Node: Static Members, Prev: Virtual Base Classes, Up: Cplusplus
2997 The data area for a class is a concatenation of the space used by the
2998 data members of the class. If the class has virtual methods, a vtable
2999 pointer follows the class data. The field offset part of each field
3000 description in the class stab shows this ordering.
3002 << How is this reflected in stabs? See Cygnus bug #677 for some
3006 File: stabs.info, Node: Stab Types, Next: Symbol Descriptors, Prev: Cplusplus, Up: Top
3008 Appendix A Table of Stab Types
3009 ******************************
3011 The following are all the possible values for the stab type field, for
3012 a.out files, in numeric order. This does not apply to XCOFF, but it
3013 does apply to stabs in sections (*note Stab Sections::). Stabs in
3014 ECOFF use these values but add 0x8f300 to distinguish them from non-stab
3017 The symbolic names are defined in the file `include/aout/stabs.def'.
3021 * Non-Stab Symbol Types:: Types from 0 to 0x1f
3022 * Stab Symbol Types:: Types from 0x20 to 0xff
3025 File: stabs.info, Node: Non-Stab Symbol Types, Next: Stab Symbol Types, Up: Stab Types
3027 A.1 Non-Stab Symbol Types
3028 =========================
3030 The following types are used by the linker and assembler, not by stab
3031 directives. Since this document does not attempt to describe aspects of
3032 object file format other than the debugging format, no details are
3039 File scope absolute symbol
3042 External absolute symbol
3045 File scope text symbol
3047 `0x5 N_TEXT | N_EXT'
3048 External text symbol
3051 File scope data symbol
3053 `0x7 N_DATA | N_EXT'
3054 External data symbol
3057 File scope BSS symbol
3063 Same as `N_FN', for Sequent compilers
3066 Symbol is indirected to another symbol
3069 Common--visible after shared library dynamic link
3072 `0x15 N_SETA | N_EXT'
3073 Absolute set element
3076 `0x17 N_SETT | N_EXT'
3077 Text segment set element
3080 `0x19 N_SETD | N_EXT'
3081 Data segment set element
3084 `0x1b N_SETB | N_EXT'
3085 BSS segment set element
3088 `0x1d N_SETV | N_EXT'
3089 Pointer to set vector
3092 Print a warning message during linking
3095 File name of a `.o' file
3098 File: stabs.info, Node: Stab Symbol Types, Prev: Non-Stab Symbol Types, Up: Stab Types
3100 A.2 Stab Symbol Types
3101 =====================
3103 The following symbol types indicate that this is a stab. This is the
3104 full list of stab numbers, including stab types that are used in
3105 languages other than C.
3108 Global symbol; see *Note Global Variables::.
3111 Function name (for BSD Fortran); see *Note Procedures::.
3114 Function name (*note Procedures::) or text segment variable (*note
3118 Data segment file-scope variable; see *Note Statics::.
3121 BSS segment file-scope variable; see *Note Statics::.
3124 Name of main routine; see *Note Main Program::.
3127 Variable in `.rodata' section; see *Note Statics::.
3130 Global symbol (for Pascal); see *Note N_PC::.
3133 Number of symbols (according to Ultrix V4.0); see *Note N_NSYMS::.
3136 No DST map; see *Note N_NOMAP::.
3139 Name and body of a `#define'd macro; see *Note Macro define and
3143 Object file (Solaris2).
3146 Name of an `#undef'ed macro; see *Note Macro define and undefine::.
3149 Debugger options (Solaris2).
3152 Register variable; see *Note Register Variables::.
3155 Modula-2 compilation unit; see *Note N_M2C::.
3158 Line number in text segment; see *Note Line Numbers::.
3161 Line number in data segment; see *Note Line Numbers::.
3164 Line number in bss segment; see *Note Line Numbers::.
3167 Sun source code browser, path to `.cb' file; see *Note N_BROWS::.
3170 GNU Modula2 definition module dependency; see *Note N_DEFD::.
3173 Function start/body/end line numbers (Solaris2).
3176 GNU C++ exception variable; see *Note N_EHDECL::.
3179 Modula2 info "for imc" (according to Ultrix V4.0); see *Note
3183 GNU C++ `catch' clause; see *Note N_CATCH::.
3186 Structure of union element; see *Note N_SSYM::.
3189 Last stab for module (Solaris2).
3192 Path and name of source file; see *Note Source Files::.
3195 Stack variable (*note Stack Variables::) or type (*note
3199 Beginning of an include file (Sun only); see *Note Include Files::.
3202 Name of include file; see *Note Include Files::.
3205 Parameter variable; see *Note Parameters::.
3208 End of an include file; see *Note Include Files::.
3211 Alternate entry point; see *Note Alternate Entry Points::.
3214 Beginning of a lexical block; see *Note Block Structure::.
3217 Place holder for a deleted include file; see *Note Include Files::.
3220 Modula2 scope information (Sun linker); see *Note N_SCOPE::.
3223 End of a lexical block; see *Note Block Structure::.
3226 Begin named common block; see *Note Common Blocks::.
3229 End named common block; see *Note Common Blocks::.
3232 Member of a common block; see *Note Common Blocks::.
3235 Pascal `with' statement: type,,0,0,offset (Solaris2).
3238 Gould non-base registers; see *Note Gould::.
3241 Gould non-base registers; see *Note Gould::.
3244 Gould non-base registers; see *Note Gould::.
3247 Gould non-base registers; see *Note Gould::.
3250 Gould non-base registers; see *Note Gould::.
3253 File: stabs.info, Node: Symbol Descriptors, Next: Type Descriptors, Prev: Stab Types, Up: Top
3255 Appendix B Table of Symbol Descriptors
3256 **************************************
3258 The symbol descriptor is the character which follows the colon in many
3259 stabs, and which tells what kind of stab it is. *Note String Field::,
3260 for more information about their use.
3265 Variable on the stack; see *Note Stack Variables::.
3268 C++ nested symbol; see *Note Nested Symbols::.
3271 Parameter passed by reference in register; see *Note Reference
3275 Based variable; see *Note Based Variables::.
3278 Constant; see *Note Constants::.
3281 Conformant array bound (Pascal, maybe other languages); *Note
3282 Conformant Arrays::. Name of a caught exception (GNU C++). These
3283 can be distinguished because the latter uses `N_CATCH' and the
3284 former uses another symbol type.
3287 Floating point register variable; see *Note Register Variables::.
3290 Parameter in floating point register; see *Note Register
3294 File scope function; see *Note Procedures::.
3297 Global function; see *Note Procedures::.
3300 Global variable; see *Note Global Variables::.
3303 *Note Register Parameters::.
3306 Internal (nested) procedure; see *Note Nested Procedures::.
3309 Internal (nested) function; see *Note Nested Procedures::.
3312 Label name (documented by AIX, no further information known).
3315 Module; see *Note Procedures::.
3318 Argument list parameter; see *Note Parameters::.
3324 Fortran Function parameter; see *Note Parameters::.
3327 Unfortunately, three separate meanings have been independently
3328 invented for this symbol descriptor. At least the GNU and Sun
3329 uses can be distinguished by the symbol type. Global Procedure
3330 (AIX) (symbol type used unknown); see *Note Procedures::.
3331 Register parameter (GNU) (symbol type `N_PSYM'); see *Note
3332 Parameters::. Prototype of function referenced by this file (Sun
3333 `acc') (symbol type `N_FUN').
3336 Static Procedure; see *Note Procedures::.
3339 Register parameter; see *Note Register Parameters::.
3342 Register variable; see *Note Register Variables::.
3345 File scope variable; see *Note Statics::.
3348 Local variable (OS9000).
3351 Type name; see *Note Typedefs::.
3354 Enumeration, structure, or union tag; see *Note Typedefs::.
3357 Parameter passed by reference; see *Note Reference Parameters::.
3360 Procedure scope static variable; see *Note Statics::.
3363 Conformant array; see *Note Conformant Arrays::.
3366 Function return variable; see *Note Parameters::.
3369 File: stabs.info, Node: Type Descriptors, Next: Expanded Reference, Prev: Symbol Descriptors, Up: Top
3371 Appendix C Table of Type Descriptors
3372 ************************************
3374 The type descriptor is the character which follows the type number and
3375 an equals sign. It specifies what kind of type is being defined.
3376 *Note String Field::, for more information about their use.
3380 Type reference; see *Note String Field::.
3383 Reference to builtin type; see *Note Negative Type Numbers::.
3386 Method (C++); see *Note Method Type Descriptor::.
3389 Pointer; see *Note Miscellaneous Types::.
3395 Type Attributes (AIX); see *Note String Field::. Member (class
3396 and variable) type (GNU C++); see *Note Member Type Descriptor::.
3399 Array; see *Note Arrays::.
3402 Open array; see *Note Arrays::.
3405 Pascal space type (AIX); see *Note Miscellaneous Types::. Builtin
3406 integer type (Sun); see *Note Builtin Type Descriptors::. Const
3407 and volatile qualified type (OS9000).
3410 Volatile-qualified type; see *Note Miscellaneous Types::.
3413 Complex builtin type (AIX); see *Note Builtin Type Descriptors::.
3414 Const-qualified type (OS9000).
3417 COBOL Picture type. See AIX documentation for details.
3420 File type; see *Note Miscellaneous Types::.
3423 N-dimensional dynamic array; see *Note Arrays::.
3426 Enumeration type; see *Note Enumerations::.
3429 N-dimensional subarray; see *Note Arrays::.
3432 Function type; see *Note Function Types::.
3435 Pascal function parameter; see *Note Function Types::
3438 Builtin floating point type; see *Note Builtin Type Descriptors::.
3441 COBOL Group. See AIX documentation for details.
3444 Imported type (AIX); see *Note Cross-References::.
3445 Volatile-qualified type (OS9000).
3448 Const-qualified type; see *Note Miscellaneous Types::.
3451 COBOL File Descriptor. See AIX documentation for details.
3454 Multiple instance type; see *Note Miscellaneous Types::.
3457 String type; see *Note Strings::.
3460 Stringptr; see *Note Strings::.
3463 Opaque type; see *Note Typedefs::.
3466 Procedure; see *Note Function Types::.
3469 Packed array; see *Note Arrays::.
3472 Range type; see *Note Subranges::.
3475 Builtin floating type; see *Note Builtin Type Descriptors:: (Sun).
3476 Pascal subroutine parameter; see *Note Function Types:: (AIX).
3477 Detecting this conflict is possible with careful parsing (hint: a
3478 Pascal subroutine parameter type will always contain a comma, and
3479 a builtin type descriptor never will).
3482 Structure type; see *Note Structures::.
3485 Set type; see *Note Miscellaneous Types::.
3488 Union; see *Note Unions::.
3491 Variant record. This is a Pascal and Modula-2 feature which is
3492 like a union within a struct in C. See AIX documentation for
3496 Wide character; see *Note Builtin Type Descriptors::.
3499 Cross-reference; see *Note Cross-References::.
3502 Used by IBM's xlC C++ compiler (for structures, I think).
3505 gstring; see *Note Strings::.
3508 File: stabs.info, Node: Expanded Reference, Next: Questions, Prev: Type Descriptors, Up: Top
3510 Appendix D Expanded Reference by Stab Type
3511 ******************************************
3513 For a full list of stab types, and cross-references to where they are
3514 described, see *Note Stab Types::. This appendix just covers certain
3515 stabs which are not yet described in the main body of this document;
3516 eventually the information will all be in one place.
3520 The first line is the symbol type (see `include/aout/stab.def').
3522 The second line describes the language constructs the symbol type
3525 The third line is the stab format with the significant stab fields
3526 named and the rest NIL.
3528 Subsequent lines expand upon the meaning and possible values for each
3529 significant stab field.
3531 Finally, any further information.
3535 * N_PC:: Pascal global symbol
3536 * N_NSYMS:: Number of symbols
3537 * N_NOMAP:: No DST map
3538 * N_M2C:: Modula-2 compilation unit
3539 * N_BROWS:: Path to .cb file for Sun source code browser
3540 * N_DEFD:: GNU Modula2 definition module dependency
3541 * N_EHDECL:: GNU C++ exception variable
3542 * N_MOD2:: Modula2 information "for imc"
3543 * N_CATCH:: GNU C++ "catch" clause
3544 * N_SSYM:: Structure or union element
3545 * N_SCOPE:: Modula2 scope information (Sun only)
3546 * Gould:: non-base register symbols used on Gould systems
3547 * N_LENG:: Length of preceding entry
3550 File: stabs.info, Node: N_PC, Next: N_NSYMS, Up: Expanded Reference
3556 Global symbol (for Pascal).
3558 "name" -> "symbol_name" <<?>>
3559 value -> supposedly the line number (stab.def is skeptical)
3563 global pascal symbol: name,,0,subtype,line
3567 File: stabs.info, Node: N_NSYMS, Next: N_NOMAP, Prev: N_PC, Up: Expanded Reference
3572 -- `.stabn': N_NSYMS
3573 Number of symbols (according to Ultrix V4.0).
3575 0, files,,funcs,lines (stab.def)
3578 File: stabs.info, Node: N_NOMAP, Next: N_M2C, Prev: N_NSYMS, Up: Expanded Reference
3583 -- `.stabs': N_NOMAP
3584 No DST map for symbol (according to Ultrix V4.0). I think this
3585 means a variable has been optimized out.
3587 name, ,0,type,ignored (stab.def)
3590 File: stabs.info, Node: N_M2C, Next: N_BROWS, Prev: N_NOMAP, Up: Expanded Reference
3596 Modula-2 compilation unit.
3598 "string" -> "unit_name,unit_time_stamp[,code_time_stamp]"
3600 value -> 0 (main unit)
3603 See `Dbx and Dbxtool Interfaces', 2nd edition, by Sun, 1988, for
3608 File: stabs.info, Node: N_BROWS, Next: N_DEFD, Prev: N_M2C, Up: Expanded Reference
3613 -- `.stabs': N_BROWS
3614 Sun source code browser, path to `.cb' file
3616 <<?>> "path to associated `.cb' file"
3618 Note: N_BROWS has the same value as N_BSLINE.
3621 File: stabs.info, Node: N_DEFD, Next: N_EHDECL, Prev: N_BROWS, Up: Expanded Reference
3627 GNU Modula2 definition module dependency.
3629 GNU Modula-2 definition module dependency. The value is the
3630 modification time of the definition file. The other field is
3631 non-zero if it is imported with the GNU M2 keyword `%INITIALIZE'.
3632 Perhaps `N_M2C' can be used if there are enough empty fields?
3635 File: stabs.info, Node: N_EHDECL, Next: N_MOD2, Prev: N_DEFD, Up: Expanded Reference
3640 -- `.stabs': N_EHDECL
3641 GNU C++ exception variable <<?>>.
3643 "STRING is variable name"
3645 Note: conflicts with `N_MOD2'.
3648 File: stabs.info, Node: N_MOD2, Next: N_CATCH, Prev: N_EHDECL, Up: Expanded Reference
3654 Modula2 info "for imc" (according to Ultrix V4.0)
3656 Note: conflicts with `N_EHDECL' <<?>>
3659 File: stabs.info, Node: N_CATCH, Next: N_SSYM, Prev: N_MOD2, Up: Expanded Reference
3664 -- `.stabn': N_CATCH
3665 GNU C++ `catch' clause
3667 GNU C++ `catch' clause. The value is its address. The desc field
3668 is nonzero if this entry is immediately followed by a `CAUGHT' stab
3669 saying what exception was caught. Multiple `CAUGHT' stabs means
3670 that multiple exceptions can be caught here. If desc is 0, it
3671 means all exceptions are caught here.
3674 File: stabs.info, Node: N_SSYM, Next: N_SCOPE, Prev: N_CATCH, Up: Expanded Reference
3680 Structure or union element.
3682 The value is the offset in the structure.
3684 <<?looking at structs and unions in C I didn't see these>>
3687 File: stabs.info, Node: N_SCOPE, Next: Gould, Prev: N_SSYM, Up: Expanded Reference
3692 -- `.stab?': N_SCOPE
3693 Modula2 scope information (Sun linker) <<?>>
3696 File: stabs.info, Node: Gould, Next: N_LENG, Prev: N_SCOPE, Up: Expanded Reference
3698 D.12 Non-base registers on Gould systems
3699 ========================================
3701 -- `.stab?': N_NBTEXT
3702 -- `.stab?': N_NBDATA
3703 -- `.stab?': N_NBBSS
3704 -- `.stab?': N_NBSTS
3705 -- `.stab?': N_NBLCS
3706 These are used on Gould systems for non-base registers syms.
3708 However, the following values are not the values used by Gould;
3709 they are the values which GNU has been documenting for these
3710 values for a long time, without actually checking what Gould uses.
3711 I include these values only because perhaps some someone actually
3712 did something with the GNU information (I hope not, why GNU
3713 knowingly assigned wrong values to these in the header file is a
3714 complete mystery to me).
3716 240 0xf0 N_NBTEXT ??
3717 242 0xf2 N_NBDATA ??
3723 File: stabs.info, Node: N_LENG, Prev: Gould, Up: Expanded Reference
3729 Second symbol entry containing a length-value for the preceding
3730 entry. The value is the length.
3733 File: stabs.info, Node: Questions, Next: Stab Sections, Prev: Expanded Reference, Up: Top
3735 Appendix E Questions and Anomalies
3736 **********************************
3738 * For GNU C stabs defining local and global variables (`N_LSYM' and
3739 `N_GSYM'), the desc field is supposed to contain the source line
3740 number on which the variable is defined. In reality the desc
3741 field is always 0. (This behavior is defined in `dbxout.c' and
3742 putting a line number in desc is controlled by `#ifdef
3743 WINNING_GDB', which defaults to false). GDB supposedly uses this
3744 information if you say `list VAR'. In reality, VAR can be a
3745 variable defined in the program and GDB says `function VAR not
3748 * In GNU C stabs, there seems to be no way to differentiate tag
3749 types: structures, unions, and enums (symbol descriptor `T') and
3750 typedefs (symbol descriptor `t') defined at file scope from types
3751 defined locally to a procedure or other more local scope. They
3752 all use the `N_LSYM' stab type. Types defined at procedure scope
3753 are emitted after the `N_RBRAC' of the preceding function and
3754 before the code of the procedure in which they are defined. This
3755 is exactly the same as types defined in the source file between
3756 the two procedure bodies. GDB over-compensates by placing all
3757 types in block #1, the block for symbols of file scope. This is
3758 true for default, `-ansi' and `-traditional' compiler options.
3759 (Bugs gcc/1063, gdb/1066.)
3761 * What ends the procedure scope? Is it the proc block's `N_RBRAC'
3762 or the next `N_FUN'? (I believe its the first.)
3765 File: stabs.info, Node: Stab Sections, Next: Symbol Types Index, Prev: Questions, Up: Top
3767 Appendix F Using Stabs in Their Own Sections
3768 ********************************************
3770 Many object file formats allow tools to create object files with custom
3771 sections containing any arbitrary data. For any such object file
3772 format, stabs can be embedded in special sections. This is how stabs
3773 are used with ELF and SOM, and aside from ECOFF and XCOFF, is how stabs
3778 * Stab Section Basics:: How to embed stabs in sections
3779 * ELF Linker Relocation:: Sun ELF hacks
3782 File: stabs.info, Node: Stab Section Basics, Next: ELF Linker Relocation, Up: Stab Sections
3784 F.1 How to Embed Stabs in Sections
3785 ==================================
3787 The assembler creates two custom sections, a section named `.stab'
3788 which contains an array of fixed length structures, one struct per stab,
3789 and a section named `.stabstr' containing all the variable length
3790 strings that are referenced by stabs in the `.stab' section. The byte
3791 order of the stabs binary data depends on the object file format. For
3792 ELF, it matches the byte order of the ELF file itself, as determined
3793 from the `EI_DATA' field in the `e_ident' member of the ELF header.
3794 For SOM, it is always big-endian (is this true??? FIXME). For COFF, it
3795 matches the byte order of the COFF headers. The meaning of the fields
3796 is the same as for a.out (*note Symbol Table Format::), except that the
3797 `n_strx' field is relative to the strings for the current compilation
3798 unit (which can be found using the synthetic N_UNDF stab described
3799 below), rather than the entire string table.
3801 The first stab in the `.stab' section for each compilation unit is
3802 synthetic, generated entirely by the assembler, with no corresponding
3803 `.stab' directive as input to the assembler. This stab contains the
3807 Offset in the `.stabstr' section to the source filename.
3813 Unused field, always zero. This may eventually be used to hold
3814 overflows from the count in the `n_desc' field.
3817 Count of upcoming symbols, i.e., the number of remaining stabs for
3821 Size of the string table fragment associated with this source
3824 The `.stabstr' section always starts with a null byte (so that string
3825 offsets of zero reference a null string), followed by random length
3826 strings, each of which is null byte terminated.
3828 The ELF section header for the `.stab' section has its `sh_link'
3829 member set to the section number of the `.stabstr' section, and the
3830 `.stabstr' section has its ELF section header `sh_type' member set to
3831 `SHT_STRTAB' to mark it as a string table. SOM and COFF have no way of
3832 linking the sections together or marking them as string tables.
3834 For COFF, the `.stab' and `.stabstr' sections may be simply
3835 concatenated by the linker. GDB then uses the `n_desc' fields to
3836 figure out the extent of the original sections. Similarly, the
3837 `n_value' fields of the header symbols are added together in order to
3838 get the actual position of the strings in a desired `.stabstr' section.
3839 Although this design obviates any need for the linker to relocate or
3840 otherwise manipulate `.stab' and `.stabstr' sections, it also requires
3841 some care to ensure that the offsets are calculated correctly. For
3842 instance, if the linker were to pad in between the `.stabstr' sections
3843 before concatenating, then the offsets to strings in the middle of the
3844 executable's `.stabstr' section would be wrong.
3846 The GNU linker is able to optimize stabs information by merging
3847 duplicate strings and removing duplicate header file information (*note
3848 Include Files::). When some versions of the GNU linker optimize stabs
3849 in sections, they remove the leading `N_UNDF' symbol and arranges for
3850 all the `n_strx' fields to be relative to the start of the `.stabstr'
3854 File: stabs.info, Node: ELF Linker Relocation, Prev: Stab Section Basics, Up: Stab Sections
3856 F.2 Having the Linker Relocate Stabs in ELF
3857 ===========================================
3859 This section describes some Sun hacks for Stabs in ELF; it does not
3860 apply to COFF or SOM.
3862 To keep linking fast, you don't want the linker to have to relocate
3863 very many stabs. Making sure this is done for `N_SLINE', `N_RBRAC',
3864 and `N_LBRAC' stabs is the most important thing (see the descriptions
3865 of those stabs for more information). But Sun's stabs in ELF has taken
3866 this further, to make all addresses in the `n_value' field (functions
3867 and static variables) relative to the source file. For the `N_SO'
3868 symbol itself, Sun simply omits the address. To find the address of
3869 each section corresponding to a given source file, the compiler puts
3870 out symbols giving the address of each section for a given source file.
3871 Since these are ELF (not stab) symbols, the linker relocates them
3872 correctly without having to touch the stabs section. They are named
3873 `Bbss.bss' for the bss section, `Ddata.data' for the data section, and
3874 `Drodata.rodata' for the rodata section. For the text section, there
3875 is no such symbol (but there should be, see below). For an example of
3876 how these symbols work, *Note Stab Section Transformations::. GCC does
3877 not provide these symbols; it instead relies on the stabs getting
3878 relocated. Thus addresses which would normally be relative to
3879 `Bbss.bss', etc., are already relocated. The Sun linker provided with
3880 Solaris 2.2 and earlier relocates stabs using normal ELF relocation
3881 information, as it would do for any section. Sun has been threatening
3882 to kludge their linker to not do this (to speed up linking), even
3883 though the correct way to avoid having the linker do these relocations
3884 is to have the compiler no longer output relocatable values. Last I
3885 heard they had been talked out of the linker kludge. See Sun point
3886 patch 101052-01 and Sun bug 1142109. With the Sun compiler this
3887 affects `S' symbol descriptor stabs (*note Statics::) and functions
3888 (*note Procedures::). In the latter case, to adopt the clean solution
3889 (making the value of the stab relative to the start of the compilation
3890 unit), it would be necessary to invent a `Ttext.text' symbol, analogous
3891 to the `Bbss.bss', etc., symbols. I recommend this rather than using a
3892 zero value and getting the address from the ELF symbols.
3894 Finding the correct `Bbss.bss', etc., symbol is difficult, because
3895 the linker simply concatenates the `.stab' sections from each `.o' file
3896 without including any information about which part of a `.stab' section
3897 comes from which `.o' file. The way GDB does this is to look for an
3898 ELF `STT_FILE' symbol which has the same name as the last component of
3899 the file name from the `N_SO' symbol in the stabs (for example, if the
3900 file name is `../../gdb/main.c', it looks for an ELF `STT_FILE' symbol
3901 named `main.c'). This loses if different files have the same name
3902 (they could be in different directories, a library could have been
3903 copied from one system to another, etc.). It would be much cleaner to
3904 have the `Bbss.bss' symbols in the stabs themselves. Having the linker
3905 relocate them there is no more work than having the linker relocate ELF
3906 symbols, and it solves the problem of having to associate the ELF and
3907 stab symbols. However, no one has yet designed or implemented such a
3911 File: stabs.info, Node: GNU Free Documentation License, Prev: Symbol Types Index, Up: Top
3913 Appendix G GNU Free Documentation License
3914 *****************************************
3916 Version 1.2, November 2002
3918 Copyright (C) 2000,2001,2002 Free Software Foundation, Inc.
3919 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
3921 Everyone is permitted to copy and distribute verbatim copies
3922 of this license document, but changing it is not allowed.
3926 The purpose of this License is to make a manual, textbook, or other
3927 functional and useful document "free" in the sense of freedom: to
3928 assure everyone the effective freedom to copy and redistribute it,
3929 with or without modifying it, either commercially or
3930 noncommercially. Secondarily, this License preserves for the
3931 author and publisher a way to get credit for their work, while not
3932 being considered responsible for modifications made by others.
3934 This License is a kind of "copyleft", which means that derivative
3935 works of the document must themselves be free in the same sense.
3936 It complements the GNU General Public License, which is a copyleft
3937 license designed for free software.
3939 We have designed this License in order to use it for manuals for
3940 free software, because free software needs free documentation: a
3941 free program should come with manuals providing the same freedoms
3942 that the software does. But this License is not limited to
3943 software manuals; it can be used for any textual work, regardless
3944 of subject matter or whether it is published as a printed book.
3945 We recommend this License principally for works whose purpose is
3946 instruction or reference.
3948 1. APPLICABILITY AND DEFINITIONS
3950 This License applies to any manual or other work, in any medium,
3951 that contains a notice placed by the copyright holder saying it
3952 can be distributed under the terms of this License. Such a notice
3953 grants a world-wide, royalty-free license, unlimited in duration,
3954 to use that work under the conditions stated herein. The
3955 "Document", below, refers to any such manual or work. Any member
3956 of the public is a licensee, and is addressed as "you". You
3957 accept the license if you copy, modify or distribute the work in a
3958 way requiring permission under copyright law.
3960 A "Modified Version" of the Document means any work containing the
3961 Document or a portion of it, either copied verbatim, or with
3962 modifications and/or translated into another language.
3964 A "Secondary Section" is a named appendix or a front-matter section
3965 of the Document that deals exclusively with the relationship of the
3966 publishers or authors of the Document to the Document's overall
3967 subject (or to related matters) and contains nothing that could
3968 fall directly within that overall subject. (Thus, if the Document
3969 is in part a textbook of mathematics, a Secondary Section may not
3970 explain any mathematics.) The relationship could be a matter of
3971 historical connection with the subject or with related matters, or
3972 of legal, commercial, philosophical, ethical or political position
3975 The "Invariant Sections" are certain Secondary Sections whose
3976 titles are designated, as being those of Invariant Sections, in
3977 the notice that says that the Document is released under this
3978 License. If a section does not fit the above definition of
3979 Secondary then it is not allowed to be designated as Invariant.
3980 The Document may contain zero Invariant Sections. If the Document
3981 does not identify any Invariant Sections then there are none.
3983 The "Cover Texts" are certain short passages of text that are
3984 listed, as Front-Cover Texts or Back-Cover Texts, in the notice
3985 that says that the Document is released under this License. A
3986 Front-Cover Text may be at most 5 words, and a Back-Cover Text may
3987 be at most 25 words.
3989 A "Transparent" copy of the Document means a machine-readable copy,
3990 represented in a format whose specification is available to the
3991 general public, that is suitable for revising the document
3992 straightforwardly with generic text editors or (for images
3993 composed of pixels) generic paint programs or (for drawings) some
3994 widely available drawing editor, and that is suitable for input to
3995 text formatters or for automatic translation to a variety of
3996 formats suitable for input to text formatters. A copy made in an
3997 otherwise Transparent file format whose markup, or absence of
3998 markup, has been arranged to thwart or discourage subsequent
3999 modification by readers is not Transparent. An image format is
4000 not Transparent if used for any substantial amount of text. A
4001 copy that is not "Transparent" is called "Opaque".
4003 Examples of suitable formats for Transparent copies include plain
4004 ASCII without markup, Texinfo input format, LaTeX input format,
4005 SGML or XML using a publicly available DTD, and
4006 standard-conforming simple HTML, PostScript or PDF designed for
4007 human modification. Examples of transparent image formats include
4008 PNG, XCF and JPG. Opaque formats include proprietary formats that
4009 can be read and edited only by proprietary word processors, SGML or
4010 XML for which the DTD and/or processing tools are not generally
4011 available, and the machine-generated HTML, PostScript or PDF
4012 produced by some word processors for output purposes only.
4014 The "Title Page" means, for a printed book, the title page itself,
4015 plus such following pages as are needed to hold, legibly, the
4016 material this License requires to appear in the title page. For
4017 works in formats which do not have any title page as such, "Title
4018 Page" means the text near the most prominent appearance of the
4019 work's title, preceding the beginning of the body of the text.
4021 A section "Entitled XYZ" means a named subunit of the Document
4022 whose title either is precisely XYZ or contains XYZ in parentheses
4023 following text that translates XYZ in another language. (Here XYZ
4024 stands for a specific section name mentioned below, such as
4025 "Acknowledgements", "Dedications", "Endorsements", or "History".)
4026 To "Preserve the Title" of such a section when you modify the
4027 Document means that it remains a section "Entitled XYZ" according
4030 The Document may include Warranty Disclaimers next to the notice
4031 which states that this License applies to the Document. These
4032 Warranty Disclaimers are considered to be included by reference in
4033 this License, but only as regards disclaiming warranties: any other
4034 implication that these Warranty Disclaimers may have is void and
4035 has no effect on the meaning of this License.
4039 You may copy and distribute the Document in any medium, either
4040 commercially or noncommercially, provided that this License, the
4041 copyright notices, and the license notice saying this License
4042 applies to the Document are reproduced in all copies, and that you
4043 add no other conditions whatsoever to those of this License. You
4044 may not use technical measures to obstruct or control the reading
4045 or further copying of the copies you make or distribute. However,
4046 you may accept compensation in exchange for copies. If you
4047 distribute a large enough number of copies you must also follow
4048 the conditions in section 3.
4050 You may also lend copies, under the same conditions stated above,
4051 and you may publicly display copies.
4053 3. COPYING IN QUANTITY
4055 If you publish printed copies (or copies in media that commonly
4056 have printed covers) of the Document, numbering more than 100, and
4057 the Document's license notice requires Cover Texts, you must
4058 enclose the copies in covers that carry, clearly and legibly, all
4059 these Cover Texts: Front-Cover Texts on the front cover, and
4060 Back-Cover Texts on the back cover. Both covers must also clearly
4061 and legibly identify you as the publisher of these copies. The
4062 front cover must present the full title with all words of the
4063 title equally prominent and visible. You may add other material
4064 on the covers in addition. Copying with changes limited to the
4065 covers, as long as they preserve the title of the Document and
4066 satisfy these conditions, can be treated as verbatim copying in
4069 If the required texts for either cover are too voluminous to fit
4070 legibly, you should put the first ones listed (as many as fit
4071 reasonably) on the actual cover, and continue the rest onto
4074 If you publish or distribute Opaque copies of the Document
4075 numbering more than 100, you must either include a
4076 machine-readable Transparent copy along with each Opaque copy, or
4077 state in or with each Opaque copy a computer-network location from
4078 which the general network-using public has access to download
4079 using public-standard network protocols a complete Transparent
4080 copy of the Document, free of added material. If you use the
4081 latter option, you must take reasonably prudent steps, when you
4082 begin distribution of Opaque copies in quantity, to ensure that
4083 this Transparent copy will remain thus accessible at the stated
4084 location until at least one year after the last time you
4085 distribute an Opaque copy (directly or through your agents or
4086 retailers) of that edition to the public.
4088 It is requested, but not required, that you contact the authors of
4089 the Document well before redistributing any large number of
4090 copies, to give them a chance to provide you with an updated
4091 version of the Document.
4095 You may copy and distribute a Modified Version of the Document
4096 under the conditions of sections 2 and 3 above, provided that you
4097 release the Modified Version under precisely this License, with
4098 the Modified Version filling the role of the Document, thus
4099 licensing distribution and modification of the Modified Version to
4100 whoever possesses a copy of it. In addition, you must do these
4101 things in the Modified Version:
4103 A. Use in the Title Page (and on the covers, if any) a title
4104 distinct from that of the Document, and from those of
4105 previous versions (which should, if there were any, be listed
4106 in the History section of the Document). You may use the
4107 same title as a previous version if the original publisher of
4108 that version gives permission.
4110 B. List on the Title Page, as authors, one or more persons or
4111 entities responsible for authorship of the modifications in
4112 the Modified Version, together with at least five of the
4113 principal authors of the Document (all of its principal
4114 authors, if it has fewer than five), unless they release you
4115 from this requirement.
4117 C. State on the Title page the name of the publisher of the
4118 Modified Version, as the publisher.
4120 D. Preserve all the copyright notices of the Document.
4122 E. Add an appropriate copyright notice for your modifications
4123 adjacent to the other copyright notices.
4125 F. Include, immediately after the copyright notices, a license
4126 notice giving the public permission to use the Modified
4127 Version under the terms of this License, in the form shown in
4130 G. Preserve in that license notice the full lists of Invariant
4131 Sections and required Cover Texts given in the Document's
4134 H. Include an unaltered copy of this License.
4136 I. Preserve the section Entitled "History", Preserve its Title,
4137 and add to it an item stating at least the title, year, new
4138 authors, and publisher of the Modified Version as given on
4139 the Title Page. If there is no section Entitled "History" in
4140 the Document, create one stating the title, year, authors,
4141 and publisher of the Document as given on its Title Page,
4142 then add an item describing the Modified Version as stated in
4143 the previous sentence.
4145 J. Preserve the network location, if any, given in the Document
4146 for public access to a Transparent copy of the Document, and
4147 likewise the network locations given in the Document for
4148 previous versions it was based on. These may be placed in
4149 the "History" section. You may omit a network location for a
4150 work that was published at least four years before the
4151 Document itself, or if the original publisher of the version
4152 it refers to gives permission.
4154 K. For any section Entitled "Acknowledgements" or "Dedications",
4155 Preserve the Title of the section, and preserve in the
4156 section all the substance and tone of each of the contributor
4157 acknowledgements and/or dedications given therein.
4159 L. Preserve all the Invariant Sections of the Document,
4160 unaltered in their text and in their titles. Section numbers
4161 or the equivalent are not considered part of the section
4164 M. Delete any section Entitled "Endorsements". Such a section
4165 may not be included in the Modified Version.
4167 N. Do not retitle any existing section to be Entitled
4168 "Endorsements" or to conflict in title with any Invariant
4171 O. Preserve any Warranty Disclaimers.
4173 If the Modified Version includes new front-matter sections or
4174 appendices that qualify as Secondary Sections and contain no
4175 material copied from the Document, you may at your option
4176 designate some or all of these sections as invariant. To do this,
4177 add their titles to the list of Invariant Sections in the Modified
4178 Version's license notice. These titles must be distinct from any
4179 other section titles.
4181 You may add a section Entitled "Endorsements", provided it contains
4182 nothing but endorsements of your Modified Version by various
4183 parties--for example, statements of peer review or that the text
4184 has been approved by an organization as the authoritative
4185 definition of a standard.
4187 You may add a passage of up to five words as a Front-Cover Text,
4188 and a passage of up to 25 words as a Back-Cover Text, to the end
4189 of the list of Cover Texts in the Modified Version. Only one
4190 passage of Front-Cover Text and one of Back-Cover Text may be
4191 added by (or through arrangements made by) any one entity. If the
4192 Document already includes a cover text for the same cover,
4193 previously added by you or by arrangement made by the same entity
4194 you are acting on behalf of, you may not add another; but you may
4195 replace the old one, on explicit permission from the previous
4196 publisher that added the old one.
4198 The author(s) and publisher(s) of the Document do not by this
4199 License give permission to use their names for publicity for or to
4200 assert or imply endorsement of any Modified Version.
4202 5. COMBINING DOCUMENTS
4204 You may combine the Document with other documents released under
4205 this License, under the terms defined in section 4 above for
4206 modified versions, provided that you include in the combination
4207 all of the Invariant Sections of all of the original documents,
4208 unmodified, and list them all as Invariant Sections of your
4209 combined work in its license notice, and that you preserve all
4210 their Warranty Disclaimers.
4212 The combined work need only contain one copy of this License, and
4213 multiple identical Invariant Sections may be replaced with a single
4214 copy. If there are multiple Invariant Sections with the same name
4215 but different contents, make the title of each such section unique
4216 by adding at the end of it, in parentheses, the name of the
4217 original author or publisher of that section if known, or else a
4218 unique number. Make the same adjustment to the section titles in
4219 the list of Invariant Sections in the license notice of the
4222 In the combination, you must combine any sections Entitled
4223 "History" in the various original documents, forming one section
4224 Entitled "History"; likewise combine any sections Entitled
4225 "Acknowledgements", and any sections Entitled "Dedications". You
4226 must delete all sections Entitled "Endorsements."
4228 6. COLLECTIONS OF DOCUMENTS
4230 You may make a collection consisting of the Document and other
4231 documents released under this License, and replace the individual
4232 copies of this License in the various documents with a single copy
4233 that is included in the collection, provided that you follow the
4234 rules of this License for verbatim copying of each of the
4235 documents in all other respects.
4237 You may extract a single document from such a collection, and
4238 distribute it individually under this License, provided you insert
4239 a copy of this License into the extracted document, and follow
4240 this License in all other respects regarding verbatim copying of
4243 7. AGGREGATION WITH INDEPENDENT WORKS
4245 A compilation of the Document or its derivatives with other
4246 separate and independent documents or works, in or on a volume of
4247 a storage or distribution medium, is called an "aggregate" if the
4248 copyright resulting from the compilation is not used to limit the
4249 legal rights of the compilation's users beyond what the individual
4250 works permit. When the Document is included in an aggregate, this
4251 License does not apply to the other works in the aggregate which
4252 are not themselves derivative works of the Document.
4254 If the Cover Text requirement of section 3 is applicable to these
4255 copies of the Document, then if the Document is less than one half
4256 of the entire aggregate, the Document's Cover Texts may be placed
4257 on covers that bracket the Document within the aggregate, or the
4258 electronic equivalent of covers if the Document is in electronic
4259 form. Otherwise they must appear on printed covers that bracket
4260 the whole aggregate.
4264 Translation is considered a kind of modification, so you may
4265 distribute translations of the Document under the terms of section
4266 4. Replacing Invariant Sections with translations requires special
4267 permission from their copyright holders, but you may include
4268 translations of some or all Invariant Sections in addition to the
4269 original versions of these Invariant Sections. You may include a
4270 translation of this License, and all the license notices in the
4271 Document, and any Warranty Disclaimers, provided that you also
4272 include the original English version of this License and the
4273 original versions of those notices and disclaimers. In case of a
4274 disagreement between the translation and the original version of
4275 this License or a notice or disclaimer, the original version will
4278 If a section in the Document is Entitled "Acknowledgements",
4279 "Dedications", or "History", the requirement (section 4) to
4280 Preserve its Title (section 1) will typically require changing the
4285 You may not copy, modify, sublicense, or distribute the Document
4286 except as expressly provided for under this License. Any other
4287 attempt to copy, modify, sublicense or distribute the Document is
4288 void, and will automatically terminate your rights under this
4289 License. However, parties who have received copies, or rights,
4290 from you under this License will not have their licenses
4291 terminated so long as such parties remain in full compliance.
4293 10. FUTURE REVISIONS OF THIS LICENSE
4295 The Free Software Foundation may publish new, revised versions of
4296 the GNU Free Documentation License from time to time. Such new
4297 versions will be similar in spirit to the present version, but may
4298 differ in detail to address new problems or concerns. See
4299 `http://www.gnu.org/copyleft/'.
4301 Each version of the License is given a distinguishing version
4302 number. If the Document specifies that a particular numbered
4303 version of this License "or any later version" applies to it, you
4304 have the option of following the terms and conditions either of
4305 that specified version or of any later version that has been
4306 published (not as a draft) by the Free Software Foundation. If
4307 the Document does not specify a version number of this License,
4308 you may choose any version ever published (not as a draft) by the
4309 Free Software Foundation.
4311 G.1 ADDENDUM: How to use this License for your documents
4312 ========================================================
4314 To use this License in a document you have written, include a copy of
4315 the License in the document and put the following copyright and license
4316 notices just after the title page:
4318 Copyright (C) YEAR YOUR NAME.
4319 Permission is granted to copy, distribute and/or modify this document
4320 under the terms of the GNU Free Documentation License, Version 1.2
4321 or any later version published by the Free Software Foundation;
4322 with no Invariant Sections, no Front-Cover Texts, and no Back-Cover
4323 Texts. A copy of the license is included in the section entitled ``GNU
4324 Free Documentation License''.
4326 If you have Invariant Sections, Front-Cover Texts and Back-Cover
4327 Texts, replace the "with...Texts." line with this:
4329 with the Invariant Sections being LIST THEIR TITLES, with
4330 the Front-Cover Texts being LIST, and with the Back-Cover Texts
4333 If you have Invariant Sections without Cover Texts, or some other
4334 combination of the three, merge those two alternatives to suit the
4337 If your document contains nontrivial examples of program code, we
4338 recommend releasing these examples in parallel under your choice of
4339 free software license, such as the GNU General Public License, to
4340 permit their use in free software.
4343 File: stabs.info, Node: Symbol Types Index, Next: GNU Free Documentation License, Prev: Stab Sections, Up: Top
4351 * .bb: Block Structure. (line 26)
4352 * .be: Block Structure. (line 26)
4353 * C_BCOMM: Common Blocks. (line 10)
4354 * C_BINCL: Include Files. (line 41)
4355 * C_BLOCK: Block Structure. (line 26)
4356 * C_BSTAT: Statics. (line 31)
4357 * C_DECL, for types: Typedefs. (line 6)
4358 * C_ECOML: Common Blocks. (line 17)
4359 * C_ECOMM: Common Blocks. (line 10)
4360 * C_EINCL: Include Files. (line 41)
4361 * C_ENTRY: Alternate Entry Points.
4363 * C_ESTAT: Statics. (line 31)
4364 * C_FILE: Source Files. (line 61)
4365 * C_FUN: Procedures. (line 18)
4366 * C_GSYM: Global Variables. (line 6)
4367 * C_LSYM: Stack Variables. (line 11)
4368 * C_PSYM: Parameters. (line 12)
4369 * C_RPSYM: Register Parameters. (line 15)
4370 * C_RSYM: Register Variables. (line 6)
4371 * C_STSYM: Statics. (line 31)
4372 * N_BCOMM: Common Blocks. (line 10)
4373 * N_BINCL: Include Files. (line 17)
4374 * N_BROWS: N_BROWS. (line 7)
4375 * N_BSLINE: Line Numbers. (line 12)
4376 * N_CATCH: N_CATCH. (line 7)
4377 * N_DEFD: N_DEFD. (line 7)
4378 * N_DSLINE: Line Numbers. (line 12)
4379 * N_ECOML: Common Blocks. (line 17)
4380 * N_ECOMM: Common Blocks. (line 10)
4381 * N_EHDECL: N_EHDECL. (line 7)
4382 * N_EINCL: Include Files. (line 17)
4383 * N_ENTRY: Alternate Entry Points.
4385 * N_EXCL: Include Files. (line 17)
4386 * N_FNAME: Procedures. (line 6)
4387 * N_FUN, for functions: Procedures. (line 6)
4388 * N_FUN, for variables: Statics. (line 12)
4389 * N_GSYM: Global Variables. (line 6)
4390 * N_GSYM, for functions (Sun acc): Procedures. (line 6)
4391 * N_LBRAC: Block Structure. (line 6)
4392 * N_LCSYM: Statics. (line 12)
4393 * N_LENG: N_LENG. (line 7)
4394 * N_LSYM, for parameter: Local Variable Parameters.
4396 * N_LSYM, for stack variables: Stack Variables. (line 11)
4397 * N_LSYM, for types: Typedefs. (line 6)
4398 * N_M2C: N_M2C. (line 7)
4399 * N_MAC_DEFINE: Macro define and undefine.
4401 * N_MAC_UNDEF: Macro define and undefine.
4403 * N_MAIN: Main Program. (line 6)
4404 * N_MOD2: N_MOD2. (line 7)
4405 * N_NBBSS: Gould. (line 9)
4406 * N_NBDATA: Gould. (line 8)
4407 * N_NBLCS: Gould. (line 11)
4408 * N_NBSTS: Gould. (line 10)
4409 * N_NBTEXT: Gould. (line 7)
4410 * N_NOMAP: N_NOMAP. (line 7)
4411 * N_NSYMS: N_NSYMS. (line 7)
4412 * N_PC: N_PC. (line 7)
4413 * N_PSYM: Parameters. (line 12)
4414 * N_RBRAC: Block Structure. (line 6)
4415 * N_ROSYM: Statics. (line 12)
4416 * N_RSYM: Register Variables. (line 6)
4417 * N_RSYM, for parameters: Register Parameters. (line 15)
4418 * N_SCOPE: N_SCOPE. (line 7)
4419 * N_SLINE: Line Numbers. (line 6)
4420 * N_SO: Source Files. (line 6)
4421 * N_SOL: Include Files. (line 11)
4422 * N_SSYM: N_SSYM. (line 7)
4423 * N_STSYM: Statics. (line 12)
4424 * N_STSYM, for functions (Sun acc): Procedures. (line 6)
4430 Node: Overview
\x7f1924
4432 Node: Stabs Format
\x7f4865
4433 Node: String Field
\x7f6427
4434 Node: C Example
\x7f11858
4435 Node: Assembly Code
\x7f12403
4436 Node: Program Structure
\x7f14374
4437 Node: Main Program
\x7f15100
4438 Node: Source Files
\x7f15661
4439 Node: Include Files
\x7f18113
4440 Node: Line Numbers
\x7f20778
4441 Node: Procedures
\x7f22312
4442 Node: Nested Procedures
\x7f28202
4443 Node: Block Structure
\x7f29378
4444 Node: Alternate Entry Points
\x7f30784
4445 Node: Constants
\x7f31517
4446 Node: Variables
\x7f34629
4447 Node: Stack Variables
\x7f35317
4448 Node: Global Variables
\x7f37018
4449 Node: Register Variables
\x7f38174
4450 Node: Common Blocks
\x7f38996
4451 Node: Statics
\x7f40250
4452 Node: Based Variables
\x7f42829
4453 Node: Parameters
\x7f44214
4454 Node: Register Parameters
\x7f45826
4455 Node: Local Variable Parameters
\x7f48087
4456 Node: Reference Parameters
\x7f51002
4457 Node: Conformant Arrays
\x7f51622
4458 Node: Types
\x7f52339
4459 Node: Builtin Types
\x7f53286
4460 Node: Traditional Builtin Types
\x7f54432
4461 Node: Traditional Integer Types
\x7f54833
4462 Node: Traditional Other Types
\x7f57141
4463 Node: Builtin Type Descriptors
\x7f58055
4464 Node: Negative Type Numbers
\x7f61555
4465 Node: Miscellaneous Types
\x7f67910
4466 Node: Cross-References
\x7f69796
4467 Node: Subranges
\x7f71471
4468 Node: Arrays
\x7f72710
4469 Node: Strings
\x7f75935
4470 Node: Enumerations
\x7f76997
4471 Node: Structures
\x7f79382
4472 Node: Typedefs
\x7f82089
4473 Node: Unions
\x7f83413
4474 Node: Function Types
\x7f84994
4475 Node: Macro define and undefine
\x7f86576
4476 Node: Symbol Tables
\x7f88153
4477 Node: Symbol Table Format
\x7f88605
4478 Node: Transformations On Symbol Tables
\x7f90053
4479 Node: Transformations On Static Variables
\x7f91407
4480 Node: Transformations On Global Variables
\x7f92143
4481 Node: Stab Section Transformations
\x7f93386
4482 Node: Cplusplus
\x7f94769
4483 Node: Class Names
\x7f95352
4484 Node: Nested Symbols
\x7f96097
4485 Node: Basic Cplusplus Types
\x7f96943
4486 Node: Simple Classes
\x7f98503
4487 Node: Class Instance
\x7f102797
4488 Node: Methods
\x7f103514
4489 Node: Method Type Descriptor
\x7f105733
4490 Node: Member Type Descriptor
\x7f106933
4491 Node: Protections
\x7f107725
4492 Node: Method Modifiers
\x7f110815
4493 Node: Virtual Methods
\x7f112443
4494 Node: Inheritance
\x7f116244
4495 Node: Virtual Base Classes
\x7f119940
4496 Node: Static Members
\x7f122184
4497 Node: Stab Types
\x7f122654
4498 Node: Non-Stab Symbol Types
\x7f123278
4499 Node: Stab Symbol Types
\x7f124709
4500 Node: Symbol Descriptors
\x7f128640
4501 Node: Type Descriptors
\x7f131419
4502 Node: Expanded Reference
\x7f134631
4503 Node: N_PC
\x7f136049
4504 Node: N_NSYMS
\x7f136417
4505 Node: N_NOMAP
\x7f136658
4506 Node: N_M2C
\x7f136964
4507 Node: N_BROWS
\x7f137398
4508 Node: N_DEFD
\x7f137681
4509 Node: N_EHDECL
\x7f138138
4510 Node: N_MOD2
\x7f138389
4511 Node: N_CATCH
\x7f138627
4512 Node: N_SSYM
\x7f139121
4513 Node: N_SCOPE
\x7f139406
4514 Node: Gould
\x7f139596
4515 Node: N_LENG
\x7f140588
4516 Node: Questions
\x7f140816
4517 Node: Stab Sections
\x7f142460
4518 Node: Stab Section Basics
\x7f143058
4519 Node: ELF Linker Relocation
\x7f146399
4520 Node: GNU Free Documentation License
\x7f149809
4521 Node: Symbol Types Index
\x7f172243