1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
3 @c 2001, 2002, 2003, 2004
4 @c Free Software Foundation, Inc.
5 @c UPDATE!! On future updates--
6 @c (1) check for new machine-dep cmdline options in
7 @c md_parse_option definitions in config/tc-*.c
8 @c (2) for platform-specific directives, examine md_pseudo_op
10 @c (3) for object-format specific directives, examine obj_pseudo_op
12 @c (4) portable directives in potable[] in read.c
16 @macro gcctabopt{body}
19 @c defaults, config file may override:
24 @include asconfig.texi
29 @c common OR combinations of conditions
55 @set abnormal-separator
59 @settitle Using @value{AS}
62 @settitle Using @value{AS} (@value{TARGET})
64 @setchapternewpage odd
69 @c WARE! Some of the machine-dependent sections contain tables of machine
70 @c instructions. Except in multi-column format, these tables look silly.
71 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
72 @c the multi-col format is faked within @example sections.
74 @c Again unfortunately, the natural size that fits on a page, for these tables,
75 @c is different depending on whether or not smallbook is turned on.
76 @c This matters, because of order: text flow switches columns at each page
79 @c The format faked in this source works reasonably well for smallbook,
80 @c not well for the default large-page format. This manual expects that if you
81 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
82 @c tables in question. You can turn on one without the other at your
83 @c discretion, of course.
86 @c the insn tables look just as silly in info files regardless of smallbook,
87 @c might as well show 'em anyways.
93 * As: (as). The GNU assembler.
94 * Gas: (as). The GNU assembler.
103 This file documents the GNU Assembler "@value{AS}".
105 @c man begin COPYRIGHT
106 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002 Free Software Foundation, Inc.
108 Permission is granted to copy, distribute and/or modify this document
109 under the terms of the GNU Free Documentation License, Version 1.1
110 or any later version published by the Free Software Foundation;
111 with no Invariant Sections, with no Front-Cover Texts, and with no
112 Back-Cover Texts. A copy of the license is included in the
113 section entitled ``GNU Free Documentation License''.
118 Permission is granted to process this file through Tex and print the
119 results, provided the printed document carries copying permission
120 notice identical to this one except for the removal of this paragraph
121 (this paragraph not being relevant to the printed manual).
127 @title Using @value{AS}
128 @subtitle The @sc{gnu} Assembler
130 @subtitle for the @value{TARGET} family
133 @subtitle Version @value{VERSION}
136 The Free Software Foundation Inc. thanks The Nice Computer
137 Company of Australia for loaning Dean Elsner to write the
138 first (Vax) version of @command{as} for Project @sc{gnu}.
139 The proprietors, management and staff of TNCCA thank FSF for
140 distracting the boss while they got some work
143 @author Dean Elsner, Jay Fenlason & friends
147 \hfill {\it Using {\tt @value{AS}}}\par
148 \hfill Edited by Cygnus Support\par
150 %"boxit" macro for figures:
151 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
152 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
153 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
154 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
155 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
158 @vskip 0pt plus 1filll
159 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002 Free Software Foundation, Inc.
161 Permission is granted to copy, distribute and/or modify this document
162 under the terms of the GNU Free Documentation License, Version 1.1
163 or any later version published by the Free Software Foundation;
164 with no Invariant Sections, with no Front-Cover Texts, and with no
165 Back-Cover Texts. A copy of the license is included in the
166 section entitled ``GNU Free Documentation License''.
172 @top Using @value{AS}
174 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}} version
177 This version of the file describes @command{@value{AS}} configured to generate
178 code for @value{TARGET} architectures.
181 This document is distributed under the terms of the GNU Free
182 Documentation License. A copy of the license is included in the
183 section entitled ``GNU Free Documentation License''.
186 * Overview:: Overview
187 * Invoking:: Command-Line Options
189 * Sections:: Sections and Relocation
191 * Expressions:: Expressions
192 * Pseudo Ops:: Assembler Directives
193 * Machine Dependencies:: Machine Dependent Features
194 * Reporting Bugs:: Reporting Bugs
195 * Acknowledgements:: Who Did What
196 * GNU Free Documentation License:: GNU Free Documentation License
204 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
206 This version of the manual describes @command{@value{AS}} configured to generate
207 code for @value{TARGET} architectures.
211 @cindex invocation summary
212 @cindex option summary
213 @cindex summary of options
214 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
215 @pxref{Invoking,,Command-Line Options}.
217 @c man title AS the portable GNU assembler.
221 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
225 @c We don't use deffn and friends for the following because they seem
226 @c to be limited to one line for the header.
228 @c man begin SYNOPSIS
229 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
230 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{--gstabs}] [@b{--gstabs+}]
231 [@b{--gdwarf2}] [@b{--help}] [@b{-I} @var{dir}] [@b{-J}] [@b{-K}] [@b{-L}]
232 [@b{--listing-lhs-width}=@var{NUM}] [@b{--listing-lhs-width2}=@var{NUM}]
233 [@b{--listing-rhs-width}=@var{NUM}] [@b{--listing-cont-lines}=@var{NUM}]
234 [@b{--keep-locals}] [@b{-o} @var{objfile}] [@b{-R}] [@b{--statistics}] [@b{-v}]
235 [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}] [@b{--fatal-warnings}]
236 [@b{-w}] [@b{-x}] [@b{-Z}] [@b{--target-help}] [@var{target-options}]
237 [@b{--}|@var{files} @dots{}]
239 @c Target dependent options are listed below. Keep the list sorted.
240 @c Add an empty line for separation.
242 @c am29k has no machine-dependent assembler options
246 @emph{Target Alpha options:}
248 [@b{-mdebug} | @b{-no-mdebug}]
249 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
250 [@b{-F}] [@b{-32addr}]
254 @emph{Target ARC options:}
260 @emph{Target ARM options:}
261 @c Don't document the deprecated options
262 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
263 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
264 [@b{-mfpu}=@var{floating-point-format}]
265 [@b{-mfloat-abi}=@var{abi}]
266 [@b{-meabi}=@var{ver}]
269 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
270 @b{-mapcs-reentrant}]
271 [@b{-mthumb-interwork}] [@b{-moabi}] [@b{-k}]
275 @emph{Target CRIS options:}
276 [@b{--underscore} | @b{--no-underscore}]
278 [@b{--emulation=criself} | @b{--emulation=crisaout}]
279 @c Deprecated -- deliberately not documented.
284 @emph{Target D10V options:}
289 @emph{Target D30V options:}
290 [@b{-O}|@b{-n}|@b{-N}]
293 @c Renesas family chips have no machine-dependent assembler options
296 @c HPPA has no machine-dependent assembler options (yet).
300 @emph{Target i386 options:}
301 [@b{--32}|@b{--64}] [@b{-n}]
305 @emph{Target i960 options:}
306 @c see md_parse_option in tc-i960.c
307 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
309 [@b{-b}] [@b{-no-relax}]
313 @emph{Target IA-64 options:}
314 [@b{-mconstant-gp}|@b{-mauto-pic}]
315 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
317 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
321 @emph{Target IP2K options:}
322 [@b{-mip2022}|@b{-mip2022ext}]
326 @emph{Target M32R options:}
327 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
332 @emph{Target M680X0 options:}
333 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
337 @emph{Target M68HC11 options:}
338 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
339 [@b{-mshort}|@b{-mlong}]
340 [@b{-mshort-double}|@b{-mlong-double}]
341 [@b{--force-long-branchs}] [@b{--short-branchs}]
342 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
343 [@b{--print-opcodes}] [@b{--generate-example}]
347 @emph{Target MCORE options:}
348 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
349 [@b{-mcpu=[210|340]}]
353 @emph{Target MIPS options:}
354 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
355 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
356 [@b{-non_shared}] [@b{-xgot}]
357 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
358 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
359 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
360 [@b{-mips64}] [@b{-mips64r2}]
361 [@b{-construct-floats}] [@b{-no-construct-floats}]
362 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
363 [@b{-mfix7000}] [@b{-mno-fix7000}]
364 [@b{-mips16}] [@b{-no-mips16}]
365 [@b{-mips3d}] [@b{-no-mips3d}]
366 [@b{-mdmx}] [@b{-no-mdmx}]
367 [@b{-mdebug}] [@b{-no-mdebug}]
368 [@b{-mpdr}] [@b{-mno-pdr}]
372 @emph{Target MMIX options:}
373 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
374 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
375 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
376 [@b{--linker-allocated-gregs}]
380 @emph{Target PDP11 options:}
381 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
382 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
383 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
387 @emph{Target picoJava options:}
392 @emph{Target PowerPC options:}
393 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
394 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
395 @b{-mbooke32}|@b{-mbooke64}]
396 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
397 [@b{-mregnames}|@b{-mno-regnames}]
398 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
399 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
400 [@b{-msolaris}|@b{-mno-solaris}]
404 @emph{Target SPARC options:}
405 @c The order here is important. See c-sparc.texi.
406 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
407 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
408 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
413 @emph{Target TIC54X options:}
414 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
415 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
418 @c Z8000 has no machine-dependent assembler options
422 @emph{Target Xtensa options:}
423 [@b{--[no-]density}] [@b{--[no-]relax}] [@b{--[no-]generics}]
424 [@b{--[no-]text-section-literals}]
425 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
434 Turn on listings, in any of a variety of ways:
438 omit false conditionals
441 omit debugging directives
444 include high-level source
450 include macro expansions
453 omit forms processing
459 set the name of the listing file
462 You may combine these options; for example, use @samp{-aln} for assembly
463 listing without forms processing. The @samp{=file} option, if used, must be
464 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
467 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
470 Ignored. This option is accepted for script compatibility with calls to
473 @item --defsym @var{sym}=@var{value}
474 Define the symbol @var{sym} to be @var{value} before assembling the input file.
475 @var{value} must be an integer constant. As in C, a leading @samp{0x}
476 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
479 ``fast''---skip whitespace and comment preprocessing (assume source is
483 Generate stabs debugging information for each assembler line. This
484 may help debugging assembler code, if the debugger can handle it.
487 Generate stabs debugging information for each assembler line, with GNU
488 extensions that probably only gdb can handle, and that could make other
489 debuggers crash or refuse to read your program. This
490 may help debugging assembler code. Currently the only GNU extension is
491 the location of the current working directory at assembling time.
494 Generate DWARF2 debugging information for each assembler line. This
495 may help debugging assembler code, if the debugger can handle it. Note---this
496 option is only supported by some targets, not all of them.
499 Print a summary of the command line options and exit.
502 Print a summary of all target specific options and exit.
505 Add directory @var{dir} to the search list for @code{.include} directives.
508 Don't warn about signed overflow.
511 @ifclear DIFF-TBL-KLUGE
512 This option is accepted but has no effect on the @value{TARGET} family.
514 @ifset DIFF-TBL-KLUGE
515 Issue warnings when difference tables altered for long displacements.
520 Keep (in the symbol table) local symbols. On traditional a.out systems
521 these start with @samp{L}, but different systems have different local
524 @item --listing-lhs-width=@var{number}
525 Set the maximum width, in words, of the output data column for an assembler
526 listing to @var{number}.
528 @item --listing-lhs-width2=@var{number}
529 Set the maximum width, in words, of the output data column for continuation
530 lines in an assembler listing to @var{number}.
532 @item --listing-rhs-width=@var{number}
533 Set the maximum width of an input source line, as displayed in a listing, to
536 @item --listing-cont-lines=@var{number}
537 Set the maximum number of lines printed in a listing for a single line of input
540 @item -o @var{objfile}
541 Name the object-file output from @command{@value{AS}} @var{objfile}.
544 Fold the data section into the text section.
547 Print the maximum space (in bytes) and total time (in seconds) used by
550 @item --strip-local-absolute
551 Remove local absolute symbols from the outgoing symbol table.
555 Print the @command{as} version.
558 Print the @command{as} version and exit.
562 Suppress warning messages.
564 @item --fatal-warnings
565 Treat warnings as errors.
568 Don't suppress warning messages or treat them as errors.
577 Generate an object file even after errors.
579 @item -- | @var{files} @dots{}
580 Standard input, or source files to assemble.
585 The following options are available when @value{AS} is configured for
590 This option selects the core processor variant.
592 Select either big-endian (-EB) or little-endian (-EL) output.
597 The following options are available when @value{AS} is configured for the ARM
601 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
602 Specify which ARM processor variant is the target.
603 @item -march=@var{architecture}[+@var{extension}@dots{}]
604 Specify which ARM architecture variant is used by the target.
605 @item -mfpu=@var{floating-point-format}
606 Select which Floating Point architecture is the target.
607 @item -mfloat-abi=@var{abi}
608 Select which floating point ABI is in use.
610 Enable Thumb only instruction decoding.
611 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant | -moabi
612 Select which procedure calling convention is in use.
614 Select either big-endian (-EB) or little-endian (-EL) output.
615 @item -mthumb-interwork
616 Specify that the code has been generated with interworking between Thumb and
619 Specify that PIC code has been generated.
624 See the info pages for documentation of the CRIS-specific options.
628 The following options are available when @value{AS} is configured for
631 @cindex D10V optimization
632 @cindex optimization, D10V
634 Optimize output by parallelizing instructions.
639 The following options are available when @value{AS} is configured for a D30V
642 @cindex D30V optimization
643 @cindex optimization, D30V
645 Optimize output by parallelizing instructions.
649 Warn when nops are generated.
651 @cindex D30V nops after 32-bit multiply
653 Warn when a nop after a 32-bit multiply instruction is generated.
658 The following options are available when @value{AS} is configured for the
659 Intel 80960 processor.
662 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
663 Specify which variant of the 960 architecture is the target.
666 Add code to collect statistics about branches taken.
669 Do not alter compare-and-branch instructions for long displacements;
676 The following options are available when @value{AS} is configured for the
682 Specifies that the extended IP2022 instructions are allowed.
685 Restores the default behaviour, which restricts the permitted instructions to
686 just the basic IP2022 ones.
692 The following options are available when @value{AS} is configured for the
693 Renesas M32R (formerly Mitsubishi M32R) series.
698 Specify which processor in the M32R family is the target. The default
699 is normally the M32R, but this option changes it to the M32RX.
701 @item --warn-explicit-parallel-conflicts or --Wp
702 Produce warning messages when questionable parallel constructs are
705 @item --no-warn-explicit-parallel-conflicts or --Wnp
706 Do not produce warning messages when questionable parallel constructs are
713 The following options are available when @value{AS} is configured for the
714 Motorola 68000 series.
719 Shorten references to undefined symbols, to one word instead of two.
721 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
722 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
723 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
724 Specify what processor in the 68000 family is the target. The default
725 is normally the 68020, but this can be changed at configuration time.
727 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
728 The target machine does (or does not) have a floating-point coprocessor.
729 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
730 the basic 68000 is not compatible with the 68881, a combination of the
731 two can be specified, since it's possible to do emulation of the
732 coprocessor instructions with the main processor.
734 @item -m68851 | -mno-68851
735 The target machine does (or does not) have a memory-management
736 unit coprocessor. The default is to assume an MMU for 68020 and up.
743 For details about the PDP-11 machine dependent features options,
744 see @ref{PDP-11-Options}.
747 @item -mpic | -mno-pic
748 Generate position-independent (or position-dependent) code. The
749 default is @option{-mpic}.
752 @itemx -mall-extensions
753 Enable all instruction set extensions. This is the default.
755 @item -mno-extensions
756 Disable all instruction set extensions.
758 @item -m@var{extension} | -mno-@var{extension}
759 Enable (or disable) a particular instruction set extension.
762 Enable the instruction set extensions supported by a particular CPU, and
763 disable all other extensions.
765 @item -m@var{machine}
766 Enable the instruction set extensions supported by a particular machine
767 model, and disable all other extensions.
773 The following options are available when @value{AS} is configured for
774 a picoJava processor.
778 @cindex PJ endianness
779 @cindex endianness, PJ
780 @cindex big endian output, PJ
782 Generate ``big endian'' format output.
784 @cindex little endian output, PJ
786 Generate ``little endian'' format output.
792 The following options are available when @value{AS} is configured for the
793 Motorola 68HC11 or 68HC12 series.
797 @item -m68hc11 | -m68hc12 | -m68hcs12
798 Specify what processor is the target. The default is
799 defined by the configuration option when building the assembler.
802 Specify to use the 16-bit integer ABI.
805 Specify to use the 32-bit integer ABI.
808 Specify to use the 32-bit double ABI.
811 Specify to use the 64-bit double ABI.
813 @item --force-long-branchs
814 Relative branches are turned into absolute ones. This concerns
815 conditional branches, unconditional branches and branches to a
818 @item -S | --short-branchs
819 Do not turn relative branchs into absolute ones
820 when the offset is out of range.
822 @item --strict-direct-mode
823 Do not turn the direct addressing mode into extended addressing mode
824 when the instruction does not support direct addressing mode.
826 @item --print-insn-syntax
827 Print the syntax of instruction in case of error.
829 @item --print-opcodes
830 print the list of instructions with syntax and then exit.
832 @item --generate-example
833 print an example of instruction for each possible instruction and then exit.
834 This option is only useful for testing @command{@value{AS}}.
840 The following options are available when @command{@value{AS}} is configured
841 for the SPARC architecture:
844 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
845 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
846 Explicitly select a variant of the SPARC architecture.
848 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
849 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
851 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
852 UltraSPARC extensions.
854 @item -xarch=v8plus | -xarch=v8plusa
855 For compatibility with the Solaris v9 assembler. These options are
856 equivalent to -Av8plus and -Av8plusa, respectively.
859 Warn when the assembler switches to another architecture.
864 The following options are available when @value{AS} is configured for the 'c54x
869 Enable extended addressing mode. All addresses and relocations will assume
870 extended addressing (usually 23 bits).
871 @item -mcpu=@var{CPU_VERSION}
872 Sets the CPU version being compiled for.
873 @item -merrors-to-file @var{FILENAME}
874 Redirect error output to a file, for broken systems which don't support such
875 behaviour in the shell.
880 The following options are available when @value{AS} is configured for
881 a @sc{mips} processor.
885 This option sets the largest size of an object that can be referenced
886 implicitly with the @code{gp} register. It is only accepted for targets that
887 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
889 @cindex MIPS endianness
890 @cindex endianness, MIPS
891 @cindex big endian output, MIPS
893 Generate ``big endian'' format output.
895 @cindex little endian output, MIPS
897 Generate ``little endian'' format output.
909 Generate code for a particular @sc{mips} Instruction Set Architecture level.
910 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
911 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
912 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
913 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
915 correspond to generic
916 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
917 and @samp{MIPS64 Release 2}
918 ISA processors, respectively.
920 @item -march=@var{CPU}
921 Generate code for a particular @sc{mips} cpu.
923 @item -mtune=@var{cpu}
924 Schedule and tune for a particular @sc{mips} cpu.
928 Cause nops to be inserted if the read of the destination register
929 of an mfhi or mflo instruction occurs in the following two instructions.
933 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
934 section instead of the standard ELF .stabs sections.
938 Control generation of @code{.pdr} sections.
942 The register sizes are normally inferred from the ISA and ABI, but these
943 flags force a certain group of registers to be treated as 32 bits wide at
944 all times. @samp{-mgp32} controls the size of general-purpose registers
945 and @samp{-mfp32} controls the size of floating-point registers.
949 Generate code for the MIPS 16 processor. This is equivalent to putting
950 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
951 turns off this option.
955 Generate code for the MIPS-3D Application Specific Extension.
956 This tells the assembler to accept MIPS-3D instructions.
957 @samp{-no-mips3d} turns off this option.
961 Generate code for the MDMX Application Specific Extension.
962 This tells the assembler to accept MDMX instructions.
963 @samp{-no-mdmx} turns off this option.
965 @item --construct-floats
966 @itemx --no-construct-floats
967 The @samp{--no-construct-floats} option disables the construction of
968 double width floating point constants by loading the two halves of the
969 value into the two single width floating point registers that make up
970 the double width register. By default @samp{--construct-floats} is
971 selected, allowing construction of these floating point constants.
974 @item --emulation=@var{name}
975 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
976 for some other target, in all respects, including output format (choosing
977 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
978 debugging information or store symbol table information, and default
979 endianness. The available configuration names are: @samp{mipsecoff},
980 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
981 @samp{mipsbelf}. The first two do not alter the default endianness from that
982 of the primary target for which the assembler was configured; the others change
983 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
984 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
985 selection in any case.
987 This option is currently supported only when the primary target
988 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
989 Furthermore, the primary target or others specified with
990 @samp{--enable-targets=@dots{}} at configuration time must include support for
991 the other format, if both are to be available. For example, the Irix 5
992 configuration includes support for both.
994 Eventually, this option will support more configurations, with more
995 fine-grained control over the assembler's behavior, and will be supported for
999 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1006 Control how to deal with multiplication overflow and division by zero.
1007 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1008 (and only work for Instruction Set Architecture level 2 and higher);
1009 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1013 When this option is used, @command{@value{AS}} will issue a warning every
1014 time it generates a nop instruction from a macro.
1019 The following options are available when @value{AS} is configured for
1025 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1026 The command line option @samp{-nojsri2bsr} can be used to disable it.
1030 Enable or disable the silicon filter behaviour. By default this is disabled.
1031 The default can be overridden by the @samp{-sifilter} command line option.
1034 Alter jump instructions for long displacements.
1036 @item -mcpu=[210|340]
1037 Select the cpu type on the target hardware. This controls which instructions
1041 Assemble for a big endian target.
1044 Assemble for a little endian target.
1050 See the info pages for documentation of the MMIX-specific options.
1054 The following options are available when @value{AS} is configured for
1055 an Xtensa processor.
1058 @item --density | --no-density
1059 Enable or disable use of instructions from the Xtensa code density
1060 option. This is enabled by default when the Xtensa processor supports
1061 the code density option.
1063 @item --relax | --no-relax
1064 Enable or disable instruction relaxation. This is enabled by default.
1065 Note: In the current implementation, these options also control whether
1066 assembler optimizations are performed, making these options equivalent
1067 to @option{--generics} and @option{--no-generics}.
1069 @item --generics | --no-generics
1070 Enable or disable all assembler transformations of Xtensa instructions.
1071 The default is @option{--generics};
1072 @option{--no-generics} should be used only in the rare cases when the
1073 instructions must be exactly as specified in the assembly source.
1075 @item --text-section-literals | --no-text-section-literals
1076 With @option{--text-@-section-@-literals}, literal pools are interspersed
1077 in the text section. The default is
1078 @option{--no-@-text-@-section-@-literals}, which places literals in a
1079 separate section in the output file.
1081 @item --target-align | --no-target-align
1082 Enable or disable automatic alignment to reduce branch penalties at the
1083 expense of some code density. The default is @option{--target-@-align}.
1085 @item --longcalls | --no-longcalls
1086 Enable or disable transformation of call instructions to allow calls
1087 across a greater range of addresses. The default is
1088 @option{--no-@-longcalls}.
1095 * Manual:: Structure of this Manual
1096 * GNU Assembler:: The GNU Assembler
1097 * Object Formats:: Object File Formats
1098 * Command Line:: Command Line
1099 * Input Files:: Input Files
1100 * Object:: Output (Object) File
1101 * Errors:: Error and Warning Messages
1105 @section Structure of this Manual
1107 @cindex manual, structure and purpose
1108 This manual is intended to describe what you need to know to use
1109 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1110 notation for symbols, constants, and expressions; the directives that
1111 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1114 We also cover special features in the @value{TARGET}
1115 configuration of @command{@value{AS}}, including assembler directives.
1118 This manual also describes some of the machine-dependent features of
1119 various flavors of the assembler.
1122 @cindex machine instructions (not covered)
1123 On the other hand, this manual is @emph{not} intended as an introduction
1124 to programming in assembly language---let alone programming in general!
1125 In a similar vein, we make no attempt to introduce the machine
1126 architecture; we do @emph{not} describe the instruction set, standard
1127 mnemonics, registers or addressing modes that are standard to a
1128 particular architecture.
1130 You may want to consult the manufacturer's
1131 machine architecture manual for this information.
1135 For information on the H8/300 machine instruction set, see @cite{H8/300
1136 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1137 Programming Manual} (Renesas).
1140 For information on the H8/500 machine instruction set, see @cite{H8/500
1141 Series Programming Manual} (Renesas M21T001).
1144 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1145 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1146 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1147 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1150 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1154 @c I think this is premature---doc@cygnus.com, 17jan1991
1156 Throughout this manual, we assume that you are running @dfn{GNU},
1157 the portable operating system from the @dfn{Free Software
1158 Foundation, Inc.}. This restricts our attention to certain kinds of
1159 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1160 once this assumption is granted examples and definitions need less
1163 @command{@value{AS}} is part of a team of programs that turn a high-level
1164 human-readable series of instructions into a low-level
1165 computer-readable series of instructions. Different versions of
1166 @command{@value{AS}} are used for different kinds of computer.
1169 @c There used to be a section "Terminology" here, which defined
1170 @c "contents", "byte", "word", and "long". Defining "word" to any
1171 @c particular size is confusing when the .word directive may generate 16
1172 @c bits on one machine and 32 bits on another; in general, for the user
1173 @c version of this manual, none of these terms seem essential to define.
1174 @c They were used very little even in the former draft of the manual;
1175 @c this draft makes an effort to avoid them (except in names of
1179 @section The GNU Assembler
1181 @c man begin DESCRIPTION
1183 @sc{gnu} @command{as} is really a family of assemblers.
1185 This manual describes @command{@value{AS}}, a member of that family which is
1186 configured for the @value{TARGET} architectures.
1188 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1189 should find a fairly similar environment when you use it on another
1190 architecture. Each version has much in common with the others,
1191 including object file formats, most assembler directives (often called
1192 @dfn{pseudo-ops}) and assembler syntax.@refill
1194 @cindex purpose of @sc{gnu} assembler
1195 @command{@value{AS}} is primarily intended to assemble the output of the
1196 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1197 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1198 assemble correctly everything that other assemblers for the same
1199 machine would assemble.
1201 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1204 @c This remark should appear in generic version of manual; assumption
1205 @c here is that generic version sets M680x0.
1206 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1207 assembler for the same architecture; for example, we know of several
1208 incompatible versions of 680x0 assembly language syntax.
1213 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1214 program in one pass of the source file. This has a subtle impact on the
1215 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1217 @node Object Formats
1218 @section Object File Formats
1220 @cindex object file format
1221 The @sc{gnu} assembler can be configured to produce several alternative
1222 object file formats. For the most part, this does not affect how you
1223 write assembly language programs; but directives for debugging symbols
1224 are typically different in different file formats. @xref{Symbol
1225 Attributes,,Symbol Attributes}.
1228 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1229 @value{OBJ-NAME} format object files.
1231 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1233 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1234 @code{a.out} or COFF format object files.
1237 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1238 @code{b.out} or COFF format object files.
1241 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1242 SOM or ELF format object files.
1247 @section Command Line
1249 @cindex command line conventions
1251 After the program name @command{@value{AS}}, the command line may contain
1252 options and file names. Options may appear in any order, and may be
1253 before, after, or between file names. The order of file names is
1256 @cindex standard input, as input file
1258 @file{--} (two hyphens) by itself names the standard input file
1259 explicitly, as one of the files for @command{@value{AS}} to assemble.
1261 @cindex options, command line
1262 Except for @samp{--} any command line argument that begins with a
1263 hyphen (@samp{-}) is an option. Each option changes the behavior of
1264 @command{@value{AS}}. No option changes the way another option works. An
1265 option is a @samp{-} followed by one or more letters; the case of
1266 the letter is important. All options are optional.
1268 Some options expect exactly one file name to follow them. The file
1269 name may either immediately follow the option's letter (compatible
1270 with older assemblers) or it may be the next command argument (@sc{gnu}
1271 standard). These two command lines are equivalent:
1274 @value{AS} -o my-object-file.o mumble.s
1275 @value{AS} -omy-object-file.o mumble.s
1279 @section Input Files
1282 @cindex source program
1283 @cindex files, input
1284 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1285 describe the program input to one run of @command{@value{AS}}. The program may
1286 be in one or more files; how the source is partitioned into files
1287 doesn't change the meaning of the source.
1289 @c I added "con" prefix to "catenation" just to prove I can overcome my
1290 @c APL training... doc@cygnus.com
1291 The source program is a concatenation of the text in all the files, in the
1294 @c man begin DESCRIPTION
1295 Each time you run @command{@value{AS}} it assembles exactly one source
1296 program. The source program is made up of one or more files.
1297 (The standard input is also a file.)
1299 You give @command{@value{AS}} a command line that has zero or more input file
1300 names. The input files are read (from left file name to right). A
1301 command line argument (in any position) that has no special meaning
1302 is taken to be an input file name.
1304 If you give @command{@value{AS}} no file names it attempts to read one input file
1305 from the @command{@value{AS}} standard input, which is normally your terminal. You
1306 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1309 Use @samp{--} if you need to explicitly name the standard input file
1310 in your command line.
1312 If the source is empty, @command{@value{AS}} produces a small, empty object
1317 @subheading Filenames and Line-numbers
1319 @cindex input file linenumbers
1320 @cindex line numbers, in input files
1321 There are two ways of locating a line in the input file (or files) and
1322 either may be used in reporting error messages. One way refers to a line
1323 number in a physical file; the other refers to a line number in a
1324 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1326 @dfn{Physical files} are those files named in the command line given
1327 to @command{@value{AS}}.
1329 @dfn{Logical files} are simply names declared explicitly by assembler
1330 directives; they bear no relation to physical files. Logical file names help
1331 error messages reflect the original source file, when @command{@value{AS}} source
1332 is itself synthesized from other files. @command{@value{AS}} understands the
1333 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1334 @ref{File,,@code{.file}}.
1337 @section Output (Object) File
1343 Every time you run @command{@value{AS}} it produces an output file, which is
1344 your assembly language program translated into numbers. This file
1345 is the object file. Its default name is
1353 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1355 You can give it another name by using the @option{-o} option. Conventionally,
1356 object file names end with @file{.o}. The default name is used for historical
1357 reasons: older assemblers were capable of assembling self-contained programs
1358 directly into a runnable program. (For some formats, this isn't currently
1359 possible, but it can be done for the @code{a.out} format.)
1363 The object file is meant for input to the linker @code{@value{LD}}. It contains
1364 assembled program code, information to help @code{@value{LD}} integrate
1365 the assembled program into a runnable file, and (optionally) symbolic
1366 information for the debugger.
1368 @c link above to some info file(s) like the description of a.out.
1369 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1372 @section Error and Warning Messages
1374 @c man begin DESCRIPTION
1376 @cindex error messages
1377 @cindex warning messages
1378 @cindex messages from assembler
1379 @command{@value{AS}} may write warnings and error messages to the standard error
1380 file (usually your terminal). This should not happen when a compiler
1381 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1382 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1383 grave problem that stops the assembly.
1387 @cindex format of warning messages
1388 Warning messages have the format
1391 file_name:@b{NNN}:Warning Message Text
1395 @cindex line numbers, in warnings/errors
1396 (where @b{NNN} is a line number). If a logical file name has been given
1397 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1398 the current input file is used. If a logical line number was given
1400 (@pxref{Line,,@code{.line}})
1404 (@pxref{Line,,@code{.line}})
1407 (@pxref{Ln,,@code{.ln}})
1410 then it is used to calculate the number printed,
1411 otherwise the actual line in the current source file is printed. The
1412 message text is intended to be self explanatory (in the grand Unix
1415 @cindex format of error messages
1416 Error messages have the format
1418 file_name:@b{NNN}:FATAL:Error Message Text
1420 The file name and line number are derived as for warning
1421 messages. The actual message text may be rather less explanatory
1422 because many of them aren't supposed to happen.
1425 @chapter Command-Line Options
1427 @cindex options, all versions of assembler
1428 This chapter describes command-line options available in @emph{all}
1429 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1431 to the @value{TARGET} target.
1434 to particular machine architectures.
1437 @c man begin DESCRIPTION
1439 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1440 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1441 The assembler arguments must be separated from each other (and the @samp{-Wa})
1442 by commas. For example:
1445 gcc -c -g -O -Wa,-alh,-L file.c
1449 This passes two options to the assembler: @samp{-alh} (emit a listing to
1450 standard output with high-level and assembly source) and @samp{-L} (retain
1451 local symbols in the symbol table).
1453 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1454 command-line options are automatically passed to the assembler by the compiler.
1455 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1456 precisely what options it passes to each compilation pass, including the
1462 * a:: -a[cdhlns] enable listings
1463 * alternate:: --alternate enable alternate macro syntax
1464 * D:: -D for compatibility
1465 * f:: -f to work faster
1466 * I:: -I for .include search path
1467 @ifclear DIFF-TBL-KLUGE
1468 * K:: -K for compatibility
1470 @ifset DIFF-TBL-KLUGE
1471 * K:: -K for difference tables
1474 * L:: -L to retain local labels
1475 * listing:: --listing-XXX to configure listing output
1476 * M:: -M or --mri to assemble in MRI compatibility mode
1477 * MD:: --MD for dependency tracking
1478 * o:: -o to name the object file
1479 * R:: -R to join data and text sections
1480 * statistics:: --statistics to see statistics about assembly
1481 * traditional-format:: --traditional-format for compatible output
1482 * v:: -v to announce version
1483 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1484 * Z:: -Z to make object file even after errors
1488 @section Enable Listings: @option{-a[cdhlns]}
1497 @cindex listings, enabling
1498 @cindex assembly listings, enabling
1500 These options enable listing output from the assembler. By itself,
1501 @samp{-a} requests high-level, assembly, and symbols listing.
1502 You can use other letters to select specific options for the list:
1503 @samp{-ah} requests a high-level language listing,
1504 @samp{-al} requests an output-program assembly listing, and
1505 @samp{-as} requests a symbol table listing.
1506 High-level listings require that a compiler debugging option like
1507 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1510 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1511 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1512 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1513 omitted from the listing.
1515 Use the @samp{-ad} option to omit debugging directives from the
1518 Once you have specified one of these options, you can further control
1519 listing output and its appearance using the directives @code{.list},
1520 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1522 The @samp{-an} option turns off all forms processing.
1523 If you do not request listing output with one of the @samp{-a} options, the
1524 listing-control directives have no effect.
1526 The letters after @samp{-a} may be combined into one option,
1527 @emph{e.g.}, @samp{-aln}.
1529 Note if the assembler source is coming from the standard input (eg because it
1530 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1531 is being used) then the listing will not contain any comments or preprocessor
1532 directives. This is because the listing code buffers input source lines from
1533 stdin only after they have been preprocessed by the assembler. This reduces
1534 memory usage and makes the code more efficient.
1537 @section @option{--alternate}
1540 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1543 @section @option{-D}
1546 This option has no effect whatsoever, but it is accepted to make it more
1547 likely that scripts written for other assemblers also work with
1548 @command{@value{AS}}.
1551 @section Work Faster: @option{-f}
1554 @cindex trusted compiler
1555 @cindex faster processing (@option{-f})
1556 @samp{-f} should only be used when assembling programs written by a
1557 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1558 and comment preprocessing on
1559 the input file(s) before assembling them. @xref{Preprocessing,
1563 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1564 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1569 @section @code{.include} Search Path: @option{-I} @var{path}
1571 @kindex -I @var{path}
1572 @cindex paths for @code{.include}
1573 @cindex search path for @code{.include}
1574 @cindex @code{include} directive search path
1575 Use this option to add a @var{path} to the list of directories
1576 @command{@value{AS}} searches for files specified in @code{.include}
1577 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1578 many times as necessary to include a variety of paths. The current
1579 working directory is always searched first; after that, @command{@value{AS}}
1580 searches any @samp{-I} directories in the same order as they were
1581 specified (left to right) on the command line.
1584 @section Difference Tables: @option{-K}
1587 @ifclear DIFF-TBL-KLUGE
1588 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1589 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1590 where it can be used to warn when the assembler alters the machine code
1591 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1592 family does not have the addressing limitations that sometimes lead to this
1593 alteration on other platforms.
1596 @ifset DIFF-TBL-KLUGE
1597 @cindex difference tables, warning
1598 @cindex warning for altered difference tables
1599 @command{@value{AS}} sometimes alters the code emitted for directives of the form
1600 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1601 You can use the @samp{-K} option if you want a warning issued when this
1606 @section Include Local Labels: @option{-L}
1609 @cindex local labels, retaining in output
1610 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1611 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1612 debugging, because they are intended for the use of programs (like
1613 compilers) that compose assembler programs, not for your notice.
1614 Normally both @command{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1615 normally debug with them.
1617 This option tells @command{@value{AS}} to retain those @samp{L@dots{}} symbols
1618 in the object file. Usually if you do this you also tell the linker
1619 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1621 By default, a local label is any label beginning with @samp{L}, but each
1622 target is allowed to redefine the local label prefix.
1624 On the HPPA local labels begin with @samp{L$}.
1628 @section Configuring listing output: @option{--listing}
1630 The listing feature of the assembler can be enabled via the command line switch
1631 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1632 hex dump of the corresponding locations in the output object file, and displays
1633 them as a listing file. The format of this listing can be controlled by pseudo
1634 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1635 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1638 @item --listing-lhs-width=@samp{number}
1639 @kindex --listing-lhs-width
1640 @cindex Width of first line disassembly output
1641 Sets the maximum width, in words, of the first line of the hex byte dump. This
1642 dump appears on the left hand side of the listing output.
1644 @item --listing-lhs-width2=@samp{number}
1645 @kindex --listing-lhs-width2
1646 @cindex Width of continuation lines of disassembly output
1647 Sets the maximum width, in words, of any further lines of the hex byte dump for
1648 a given input source line. If this value is not specified, it defaults to being
1649 the same as the value specified for @samp{--listing-lhs-width}. If neither
1650 switch is used the default is to one.
1652 @item --listing-rhs-width=@samp{number}
1653 @kindex --listing-rhs-width
1654 @cindex Width of source line output
1655 Sets the maximum width, in characters, of the source line that is displayed
1656 alongside the hex dump. The default value for this parameter is 100. The
1657 source line is displayed on the right hand side of the listing output.
1659 @item --listing-cont-lines=@samp{number}
1660 @kindex --listing-cont-lines
1661 @cindex Maximum number of continuation lines
1662 Sets the maximum number of continuation lines of hex dump that will be
1663 displayed for a given single line of source input. The default value is 4.
1667 @section Assemble in MRI Compatibility Mode: @option{-M}
1670 @cindex MRI compatibility mode
1671 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1672 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1673 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1674 configured target) assembler from Microtec Research. The exact nature of the
1675 MRI syntax will not be documented here; see the MRI manuals for more
1676 information. Note in particular that the handling of macros and macro
1677 arguments is somewhat different. The purpose of this option is to permit
1678 assembling existing MRI assembler code using @command{@value{AS}}.
1680 The MRI compatibility is not complete. Certain operations of the MRI assembler
1681 depend upon its object file format, and can not be supported using other object
1682 file formats. Supporting these would require enhancing each object file format
1683 individually. These are:
1686 @item global symbols in common section
1688 The m68k MRI assembler supports common sections which are merged by the linker.
1689 Other object file formats do not support this. @command{@value{AS}} handles
1690 common sections by treating them as a single common symbol. It permits local
1691 symbols to be defined within a common section, but it can not support global
1692 symbols, since it has no way to describe them.
1694 @item complex relocations
1696 The MRI assemblers support relocations against a negated section address, and
1697 relocations which combine the start addresses of two or more sections. These
1698 are not support by other object file formats.
1700 @item @code{END} pseudo-op specifying start address
1702 The MRI @code{END} pseudo-op permits the specification of a start address.
1703 This is not supported by other object file formats. The start address may
1704 instead be specified using the @option{-e} option to the linker, or in a linker
1707 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1709 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1710 name to the output file. This is not supported by other object file formats.
1712 @item @code{ORG} pseudo-op
1714 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1715 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1716 which changes the location within the current section. Absolute sections are
1717 not supported by other object file formats. The address of a section may be
1718 assigned within a linker script.
1721 There are some other features of the MRI assembler which are not supported by
1722 @command{@value{AS}}, typically either because they are difficult or because they
1723 seem of little consequence. Some of these may be supported in future releases.
1727 @item EBCDIC strings
1729 EBCDIC strings are not supported.
1731 @item packed binary coded decimal
1733 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1734 and @code{DCB.P} pseudo-ops are not supported.
1736 @item @code{FEQU} pseudo-op
1738 The m68k @code{FEQU} pseudo-op is not supported.
1740 @item @code{NOOBJ} pseudo-op
1742 The m68k @code{NOOBJ} pseudo-op is not supported.
1744 @item @code{OPT} branch control options
1746 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1747 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1748 relaxes all branches, whether forward or backward, to an appropriate size, so
1749 these options serve no purpose.
1751 @item @code{OPT} list control options
1753 The following m68k @code{OPT} list control options are ignored: @code{C},
1754 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1755 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1757 @item other @code{OPT} options
1759 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1760 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1762 @item @code{OPT} @code{D} option is default
1764 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1765 @code{OPT NOD} may be used to turn it off.
1767 @item @code{XREF} pseudo-op.
1769 The m68k @code{XREF} pseudo-op is ignored.
1771 @item @code{.debug} pseudo-op
1773 The i960 @code{.debug} pseudo-op is not supported.
1775 @item @code{.extended} pseudo-op
1777 The i960 @code{.extended} pseudo-op is not supported.
1779 @item @code{.list} pseudo-op.
1781 The various options of the i960 @code{.list} pseudo-op are not supported.
1783 @item @code{.optimize} pseudo-op
1785 The i960 @code{.optimize} pseudo-op is not supported.
1787 @item @code{.output} pseudo-op
1789 The i960 @code{.output} pseudo-op is not supported.
1791 @item @code{.setreal} pseudo-op
1793 The i960 @code{.setreal} pseudo-op is not supported.
1798 @section Dependency Tracking: @option{--MD}
1801 @cindex dependency tracking
1804 @command{@value{AS}} can generate a dependency file for the file it creates. This
1805 file consists of a single rule suitable for @code{make} describing the
1806 dependencies of the main source file.
1808 The rule is written to the file named in its argument.
1810 This feature is used in the automatic updating of makefiles.
1813 @section Name the Object File: @option{-o}
1816 @cindex naming object file
1817 @cindex object file name
1818 There is always one object file output when you run @command{@value{AS}}. By
1819 default it has the name
1822 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1836 You use this option (which takes exactly one filename) to give the
1837 object file a different name.
1839 Whatever the object file is called, @command{@value{AS}} overwrites any
1840 existing file of the same name.
1843 @section Join Data and Text Sections: @option{-R}
1846 @cindex data and text sections, joining
1847 @cindex text and data sections, joining
1848 @cindex joining text and data sections
1849 @cindex merging text and data sections
1850 @option{-R} tells @command{@value{AS}} to write the object file as if all
1851 data-section data lives in the text section. This is only done at
1852 the very last moment: your binary data are the same, but data
1853 section parts are relocated differently. The data section part of
1854 your object file is zero bytes long because all its bytes are
1855 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1857 When you specify @option{-R} it would be possible to generate shorter
1858 address displacements (because we do not have to cross between text and
1859 data section). We refrain from doing this simply for compatibility with
1860 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1863 When @command{@value{AS}} is configured for COFF or ELF output,
1864 this option is only useful if you use sections named @samp{.text} and
1869 @option{-R} is not supported for any of the HPPA targets. Using
1870 @option{-R} generates a warning from @command{@value{AS}}.
1874 @section Display Assembly Statistics: @option{--statistics}
1876 @kindex --statistics
1877 @cindex statistics, about assembly
1878 @cindex time, total for assembly
1879 @cindex space used, maximum for assembly
1880 Use @samp{--statistics} to display two statistics about the resources used by
1881 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1882 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1885 @node traditional-format
1886 @section Compatible Output: @option{--traditional-format}
1888 @kindex --traditional-format
1889 For some targets, the output of @command{@value{AS}} is different in some ways
1890 from the output of some existing assembler. This switch requests
1891 @command{@value{AS}} to use the traditional format instead.
1893 For example, it disables the exception frame optimizations which
1894 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
1897 @section Announce Version: @option{-v}
1901 @cindex assembler version
1902 @cindex version of assembler
1903 You can find out what version of as is running by including the
1904 option @samp{-v} (which you can also spell as @samp{-version}) on the
1908 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
1910 @command{@value{AS}} should never give a warning or error message when
1911 assembling compiler output. But programs written by people often
1912 cause @command{@value{AS}} to give a warning that a particular assumption was
1913 made. All such warnings are directed to the standard error file.
1917 @cindex suppressing warnings
1918 @cindex warnings, suppressing
1919 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
1920 This only affects the warning messages: it does not change any particular of
1921 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
1924 @kindex --fatal-warnings
1925 @cindex errors, caused by warnings
1926 @cindex warnings, causing error
1927 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
1928 files that generate warnings to be in error.
1931 @cindex warnings, switching on
1932 You can switch these options off again by specifying @option{--warn}, which
1933 causes warnings to be output as usual.
1936 @section Generate Object File in Spite of Errors: @option{-Z}
1937 @cindex object file, after errors
1938 @cindex errors, continuing after
1939 After an error message, @command{@value{AS}} normally produces no output. If for
1940 some reason you are interested in object file output even after
1941 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
1942 option. If there are any errors, @command{@value{AS}} continues anyways, and
1943 writes an object file after a final warning message of the form @samp{@var{n}
1944 errors, @var{m} warnings, generating bad object file.}
1949 @cindex machine-independent syntax
1950 @cindex syntax, machine-independent
1951 This chapter describes the machine-independent syntax allowed in a
1952 source file. @command{@value{AS}} syntax is similar to what many other
1953 assemblers use; it is inspired by the BSD 4.2
1958 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
1962 * Preprocessing:: Preprocessing
1963 * Whitespace:: Whitespace
1964 * Comments:: Comments
1965 * Symbol Intro:: Symbols
1966 * Statements:: Statements
1967 * Constants:: Constants
1971 @section Preprocessing
1973 @cindex preprocessing
1974 The @command{@value{AS}} internal preprocessor:
1976 @cindex whitespace, removed by preprocessor
1978 adjusts and removes extra whitespace. It leaves one space or tab before
1979 the keywords on a line, and turns any other whitespace on the line into
1982 @cindex comments, removed by preprocessor
1984 removes all comments, replacing them with a single space, or an
1985 appropriate number of newlines.
1987 @cindex constants, converted by preprocessor
1989 converts character constants into the appropriate numeric values.
1992 It does not do macro processing, include file handling, or
1993 anything else you may get from your C compiler's preprocessor. You can
1994 do include file processing with the @code{.include} directive
1995 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
1996 to get other ``CPP'' style preprocessing by giving the input file a
1997 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
1998 Output, gcc.info, Using GNU CC}.
2000 Excess whitespace, comments, and character constants
2001 cannot be used in the portions of the input text that are not
2004 @cindex turning preprocessing on and off
2005 @cindex preprocessing, turning on and off
2008 If the first line of an input file is @code{#NO_APP} or if you use the
2009 @samp{-f} option, whitespace and comments are not removed from the input file.
2010 Within an input file, you can ask for whitespace and comment removal in
2011 specific portions of the by putting a line that says @code{#APP} before the
2012 text that may contain whitespace or comments, and putting a line that says
2013 @code{#NO_APP} after this text. This feature is mainly intend to support
2014 @code{asm} statements in compilers whose output is otherwise free of comments
2021 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2022 Whitespace is used to separate symbols, and to make programs neater for
2023 people to read. Unless within character constants
2024 (@pxref{Characters,,Character Constants}), any whitespace means the same
2025 as exactly one space.
2031 There are two ways of rendering comments to @command{@value{AS}}. In both
2032 cases the comment is equivalent to one space.
2034 Anything from @samp{/*} through the next @samp{*/} is a comment.
2035 This means you may not nest these comments.
2039 The only way to include a newline ('\n') in a comment
2040 is to use this sort of comment.
2043 /* This sort of comment does not nest. */
2046 @cindex line comment character
2047 Anything from the @dfn{line comment} character to the next newline
2048 is considered a comment and is ignored. The line comment character is
2050 @samp{;} for the AMD 29K family;
2053 @samp{;} on the ARC;
2056 @samp{@@} on the ARM;
2059 @samp{;} for the H8/300 family;
2062 @samp{!} for the H8/500 family;
2065 @samp{;} for the HPPA;
2068 @samp{#} on the i386 and x86-64;
2071 @samp{#} on the i960;
2074 @samp{;} for the PDP-11;
2077 @samp{;} for picoJava;
2080 @samp{#} for Motorola PowerPC;
2083 @samp{!} for the Renesas / SuperH SH;
2086 @samp{!} on the SPARC;
2089 @samp{#} on the ip2k;
2092 @samp{#} on the m32r;
2095 @samp{|} on the 680x0;
2098 @samp{#} on the 68HC11 and 68HC12;
2101 @samp{;} on the M880x0;
2104 @samp{#} on the Vax;
2107 @samp{!} for the Z8000;
2110 @samp{#} on the V850;
2113 @samp{#} for Xtensa systems;
2115 see @ref{Machine Dependencies}. @refill
2116 @c FIXME What about i860?
2119 On some machines there are two different line comment characters. One
2120 character only begins a comment if it is the first non-whitespace character on
2121 a line, while the other always begins a comment.
2125 The V850 assembler also supports a double dash as starting a comment that
2126 extends to the end of the line.
2132 @cindex lines starting with @code{#}
2133 @cindex logical line numbers
2134 To be compatible with past assemblers, lines that begin with @samp{#} have a
2135 special interpretation. Following the @samp{#} should be an absolute
2136 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2137 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
2138 new logical file name. The rest of the line, if any, should be whitespace.
2140 If the first non-whitespace characters on the line are not numeric,
2141 the line is ignored. (Just like a comment.)
2144 # This is an ordinary comment.
2145 # 42-6 "new_file_name" # New logical file name
2146 # This is logical line # 36.
2148 This feature is deprecated, and may disappear from future versions
2149 of @command{@value{AS}}.
2154 @cindex characters used in symbols
2155 @ifclear SPECIAL-SYMS
2156 A @dfn{symbol} is one or more characters chosen from the set of all
2157 letters (both upper and lower case), digits and the three characters
2163 A @dfn{symbol} is one or more characters chosen from the set of all
2164 letters (both upper and lower case), digits and the three characters
2165 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2171 On most machines, you can also use @code{$} in symbol names; exceptions
2172 are noted in @ref{Machine Dependencies}.
2174 No symbol may begin with a digit. Case is significant.
2175 There is no length limit: all characters are significant. Symbols are
2176 delimited by characters not in that set, or by the beginning of a file
2177 (since the source program must end with a newline, the end of a file is
2178 not a possible symbol delimiter). @xref{Symbols}.
2179 @cindex length of symbols
2184 @cindex statements, structure of
2185 @cindex line separator character
2186 @cindex statement separator character
2188 @ifclear abnormal-separator
2189 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2190 semicolon (@samp{;}). The newline or semicolon is considered part of
2191 the preceding statement. Newlines and semicolons within character
2192 constants are an exception: they do not end statements.
2194 @ifset abnormal-separator
2196 A @dfn{statement} ends at a newline character (@samp{\n}) or an ``at''
2197 sign (@samp{@@}). The newline or at sign is considered part of the
2198 preceding statement. Newlines and at signs within character constants
2199 are an exception: they do not end statements.
2202 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2203 point (@samp{!}). The newline or exclamation point is considered part of the
2204 preceding statement. Newlines and exclamation points within character
2205 constants are an exception: they do not end statements.
2208 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2209 H8/300) a dollar sign (@samp{$}); or (for the
2212 (@samp{;}). The newline or separator character is considered part of
2213 the preceding statement. Newlines and separators within character
2214 constants are an exception: they do not end statements.
2219 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2220 separator character. (The line separator is usually @samp{;}, unless
2221 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
2222 newline or separator character is considered part of the preceding
2223 statement. Newlines and separators within character constants are an
2224 exception: they do not end statements.
2227 @cindex newline, required at file end
2228 @cindex EOF, newline must precede
2229 It is an error to end any statement with end-of-file: the last
2230 character of any input file should be a newline.@refill
2232 An empty statement is allowed, and may include whitespace. It is ignored.
2234 @cindex instructions and directives
2235 @cindex directives and instructions
2236 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2237 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2239 A statement begins with zero or more labels, optionally followed by a
2240 key symbol which determines what kind of statement it is. The key
2241 symbol determines the syntax of the rest of the statement. If the
2242 symbol begins with a dot @samp{.} then the statement is an assembler
2243 directive: typically valid for any computer. If the symbol begins with
2244 a letter the statement is an assembly language @dfn{instruction}: it
2245 assembles into a machine language instruction.
2247 Different versions of @command{@value{AS}} for different computers
2248 recognize different instructions. In fact, the same symbol may
2249 represent a different instruction in a different computer's assembly
2253 @cindex @code{:} (label)
2254 @cindex label (@code{:})
2255 A label is a symbol immediately followed by a colon (@code{:}).
2256 Whitespace before a label or after a colon is permitted, but you may not
2257 have whitespace between a label's symbol and its colon. @xref{Labels}.
2260 For HPPA targets, labels need not be immediately followed by a colon, but
2261 the definition of a label must begin in column zero. This also implies that
2262 only one label may be defined on each line.
2266 label: .directive followed by something
2267 another_label: # This is an empty statement.
2268 instruction operand_1, operand_2, @dots{}
2275 A constant is a number, written so that its value is known by
2276 inspection, without knowing any context. Like this:
2279 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2280 .ascii "Ring the bell\7" # A string constant.
2281 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2282 .float 0f-314159265358979323846264338327\
2283 95028841971.693993751E-40 # - pi, a flonum.
2288 * Characters:: Character Constants
2289 * Numbers:: Number Constants
2293 @subsection Character Constants
2295 @cindex character constants
2296 @cindex constants, character
2297 There are two kinds of character constants. A @dfn{character} stands
2298 for one character in one byte and its value may be used in
2299 numeric expressions. String constants (properly called string
2300 @emph{literals}) are potentially many bytes and their values may not be
2301 used in arithmetic expressions.
2305 * Chars:: Characters
2309 @subsubsection Strings
2311 @cindex string constants
2312 @cindex constants, string
2313 A @dfn{string} is written between double-quotes. It may contain
2314 double-quotes or null characters. The way to get special characters
2315 into a string is to @dfn{escape} these characters: precede them with
2316 a backslash @samp{\} character. For example @samp{\\} represents
2317 one backslash: the first @code{\} is an escape which tells
2318 @command{@value{AS}} to interpret the second character literally as a backslash
2319 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2320 escape character). The complete list of escapes follows.
2322 @cindex escape codes, character
2323 @cindex character escape codes
2326 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2328 @cindex @code{\b} (backspace character)
2329 @cindex backspace (@code{\b})
2331 Mnemonic for backspace; for ASCII this is octal code 010.
2334 @c Mnemonic for EOText; for ASCII this is octal code 004.
2336 @cindex @code{\f} (formfeed character)
2337 @cindex formfeed (@code{\f})
2339 Mnemonic for FormFeed; for ASCII this is octal code 014.
2341 @cindex @code{\n} (newline character)
2342 @cindex newline (@code{\n})
2344 Mnemonic for newline; for ASCII this is octal code 012.
2347 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2349 @cindex @code{\r} (carriage return character)
2350 @cindex carriage return (@code{\r})
2352 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2355 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2356 @c other assemblers.
2358 @cindex @code{\t} (tab)
2359 @cindex tab (@code{\t})
2361 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2364 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2365 @c @item \x @var{digit} @var{digit} @var{digit}
2366 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2368 @cindex @code{\@var{ddd}} (octal character code)
2369 @cindex octal character code (@code{\@var{ddd}})
2370 @item \ @var{digit} @var{digit} @var{digit}
2371 An octal character code. The numeric code is 3 octal digits.
2372 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2373 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2375 @cindex @code{\@var{xd...}} (hex character code)
2376 @cindex hex character code (@code{\@var{xd...}})
2377 @item \@code{x} @var{hex-digits...}
2378 A hex character code. All trailing hex digits are combined. Either upper or
2379 lower case @code{x} works.
2381 @cindex @code{\\} (@samp{\} character)
2382 @cindex backslash (@code{\\})
2384 Represents one @samp{\} character.
2387 @c Represents one @samp{'} (accent acute) character.
2388 @c This is needed in single character literals
2389 @c (@xref{Characters,,Character Constants}.) to represent
2392 @cindex @code{\"} (doublequote character)
2393 @cindex doublequote (@code{\"})
2395 Represents one @samp{"} character. Needed in strings to represent
2396 this character, because an unescaped @samp{"} would end the string.
2398 @item \ @var{anything-else}
2399 Any other character when escaped by @kbd{\} gives a warning, but
2400 assembles as if the @samp{\} was not present. The idea is that if
2401 you used an escape sequence you clearly didn't want the literal
2402 interpretation of the following character. However @command{@value{AS}} has no
2403 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2404 code and warns you of the fact.
2407 Which characters are escapable, and what those escapes represent,
2408 varies widely among assemblers. The current set is what we think
2409 the BSD 4.2 assembler recognizes, and is a subset of what most C
2410 compilers recognize. If you are in doubt, do not use an escape
2414 @subsubsection Characters
2416 @cindex single character constant
2417 @cindex character, single
2418 @cindex constant, single character
2419 A single character may be written as a single quote immediately
2420 followed by that character. The same escapes apply to characters as
2421 to strings. So if you want to write the character backslash, you
2422 must write @kbd{'\\} where the first @code{\} escapes the second
2423 @code{\}. As you can see, the quote is an acute accent, not a
2424 grave accent. A newline
2426 @ifclear abnormal-separator
2427 (or semicolon @samp{;})
2429 @ifset abnormal-separator
2431 (or at sign @samp{@@})
2434 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2435 Renesas SH or H8/500)
2439 immediately following an acute accent is taken as a literal character
2440 and does not count as the end of a statement. The value of a character
2441 constant in a numeric expression is the machine's byte-wide code for
2442 that character. @command{@value{AS}} assumes your character code is ASCII:
2443 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2446 @subsection Number Constants
2448 @cindex constants, number
2449 @cindex number constants
2450 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2451 are stored in the target machine. @emph{Integers} are numbers that
2452 would fit into an @code{int} in the C language. @emph{Bignums} are
2453 integers, but they are stored in more than 32 bits. @emph{Flonums}
2454 are floating point numbers, described below.
2457 * Integers:: Integers
2462 * Bit Fields:: Bit Fields
2468 @subsubsection Integers
2470 @cindex constants, integer
2472 @cindex binary integers
2473 @cindex integers, binary
2474 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2475 the binary digits @samp{01}.
2477 @cindex octal integers
2478 @cindex integers, octal
2479 An octal integer is @samp{0} followed by zero or more of the octal
2480 digits (@samp{01234567}).
2482 @cindex decimal integers
2483 @cindex integers, decimal
2484 A decimal integer starts with a non-zero digit followed by zero or
2485 more digits (@samp{0123456789}).
2487 @cindex hexadecimal integers
2488 @cindex integers, hexadecimal
2489 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2490 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2492 Integers have the usual values. To denote a negative integer, use
2493 the prefix operator @samp{-} discussed under expressions
2494 (@pxref{Prefix Ops,,Prefix Operators}).
2497 @subsubsection Bignums
2500 @cindex constants, bignum
2501 A @dfn{bignum} has the same syntax and semantics as an integer
2502 except that the number (or its negative) takes more than 32 bits to
2503 represent in binary. The distinction is made because in some places
2504 integers are permitted while bignums are not.
2507 @subsubsection Flonums
2509 @cindex floating point numbers
2510 @cindex constants, floating point
2512 @cindex precision, floating point
2513 A @dfn{flonum} represents a floating point number. The translation is
2514 indirect: a decimal floating point number from the text is converted by
2515 @command{@value{AS}} to a generic binary floating point number of more than
2516 sufficient precision. This generic floating point number is converted
2517 to a particular computer's floating point format (or formats) by a
2518 portion of @command{@value{AS}} specialized to that computer.
2520 A flonum is written by writing (in order)
2525 (@samp{0} is optional on the HPPA.)
2529 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2531 @kbd{e} is recommended. Case is not important.
2533 @c FIXME: verify if flonum syntax really this vague for most cases
2534 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2535 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2538 On the H8/300, H8/500,
2539 Renesas / SuperH SH,
2540 and AMD 29K architectures, the letter must be
2541 one of the letters @samp{DFPRSX} (in upper or lower case).
2543 On the ARC, the letter must be one of the letters @samp{DFRS}
2544 (in upper or lower case).
2546 On the Intel 960 architecture, the letter must be
2547 one of the letters @samp{DFT} (in upper or lower case).
2549 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2553 One of the letters @samp{DFPRSX} (in upper or lower case).
2556 One of the letters @samp{DFRS} (in upper or lower case).
2559 One of the letters @samp{DFPRSX} (in upper or lower case).
2562 The letter @samp{E} (upper case only).
2565 One of the letters @samp{DFT} (in upper or lower case).
2570 An optional sign: either @samp{+} or @samp{-}.
2573 An optional @dfn{integer part}: zero or more decimal digits.
2576 An optional @dfn{fractional part}: @samp{.} followed by zero
2577 or more decimal digits.
2580 An optional exponent, consisting of:
2584 An @samp{E} or @samp{e}.
2585 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2586 @c principle this can perfectly well be different on different targets.
2588 Optional sign: either @samp{+} or @samp{-}.
2590 One or more decimal digits.
2595 At least one of the integer part or the fractional part must be
2596 present. The floating point number has the usual base-10 value.
2598 @command{@value{AS}} does all processing using integers. Flonums are computed
2599 independently of any floating point hardware in the computer running
2600 @command{@value{AS}}.
2604 @c Bit fields are written as a general facility but are also controlled
2605 @c by a conditional-compilation flag---which is as of now (21mar91)
2606 @c turned on only by the i960 config of GAS.
2608 @subsubsection Bit Fields
2611 @cindex constants, bit field
2612 You can also define numeric constants as @dfn{bit fields}.
2613 specify two numbers separated by a colon---
2615 @var{mask}:@var{value}
2618 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2621 The resulting number is then packed
2623 @c this conditional paren in case bit fields turned on elsewhere than 960
2624 (in host-dependent byte order)
2626 into a field whose width depends on which assembler directive has the
2627 bit-field as its argument. Overflow (a result from the bitwise and
2628 requiring more binary digits to represent) is not an error; instead,
2629 more constants are generated, of the specified width, beginning with the
2630 least significant digits.@refill
2632 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2633 @code{.short}, and @code{.word} accept bit-field arguments.
2638 @chapter Sections and Relocation
2643 * Secs Background:: Background
2644 * Ld Sections:: Linker Sections
2645 * As Sections:: Assembler Internal Sections
2646 * Sub-Sections:: Sub-Sections
2650 @node Secs Background
2653 Roughly, a section is a range of addresses, with no gaps; all data
2654 ``in'' those addresses is treated the same for some particular purpose.
2655 For example there may be a ``read only'' section.
2657 @cindex linker, and assembler
2658 @cindex assembler, and linker
2659 The linker @code{@value{LD}} reads many object files (partial programs) and
2660 combines their contents to form a runnable program. When @command{@value{AS}}
2661 emits an object file, the partial program is assumed to start at address 0.
2662 @code{@value{LD}} assigns the final addresses for the partial program, so that
2663 different partial programs do not overlap. This is actually an
2664 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2667 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2668 addresses. These blocks slide to their run-time addresses as rigid
2669 units; their length does not change and neither does the order of bytes
2670 within them. Such a rigid unit is called a @emph{section}. Assigning
2671 run-time addresses to sections is called @dfn{relocation}. It includes
2672 the task of adjusting mentions of object-file addresses so they refer to
2673 the proper run-time addresses.
2675 For the H8/300 and H8/500,
2676 and for the Renesas / SuperH SH,
2677 @command{@value{AS}} pads sections if needed to
2678 ensure they end on a word (sixteen bit) boundary.
2681 @cindex standard assembler sections
2682 An object file written by @command{@value{AS}} has at least three sections, any
2683 of which may be empty. These are named @dfn{text}, @dfn{data} and
2688 When it generates COFF or ELF output,
2690 @command{@value{AS}} can also generate whatever other named sections you specify
2691 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2692 If you do not use any directives that place output in the @samp{.text}
2693 or @samp{.data} sections, these sections still exist, but are empty.
2698 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2700 @command{@value{AS}} can also generate whatever other named sections you
2701 specify using the @samp{.space} and @samp{.subspace} directives. See
2702 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2703 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2704 assembler directives.
2707 Additionally, @command{@value{AS}} uses different names for the standard
2708 text, data, and bss sections when generating SOM output. Program text
2709 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2710 BSS into @samp{$BSS$}.
2714 Within the object file, the text section starts at address @code{0}, the
2715 data section follows, and the bss section follows the data section.
2718 When generating either SOM or ELF output files on the HPPA, the text
2719 section starts at address @code{0}, the data section at address
2720 @code{0x4000000}, and the bss section follows the data section.
2723 To let @code{@value{LD}} know which data changes when the sections are
2724 relocated, and how to change that data, @command{@value{AS}} also writes to the
2725 object file details of the relocation needed. To perform relocation
2726 @code{@value{LD}} must know, each time an address in the object
2730 Where in the object file is the beginning of this reference to
2733 How long (in bytes) is this reference?
2735 Which section does the address refer to? What is the numeric value of
2737 (@var{address}) @minus{} (@var{start-address of section})?
2740 Is the reference to an address ``Program-Counter relative''?
2743 @cindex addresses, format of
2744 @cindex section-relative addressing
2745 In fact, every address @command{@value{AS}} ever uses is expressed as
2747 (@var{section}) + (@var{offset into section})
2750 Further, most expressions @command{@value{AS}} computes have this section-relative
2753 (For some object formats, such as SOM for the HPPA, some expressions are
2754 symbol-relative instead.)
2757 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2758 @var{N} into section @var{secname}.''
2760 Apart from text, data and bss sections you need to know about the
2761 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2762 addresses in the absolute section remain unchanged. For example, address
2763 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2764 @code{@value{LD}}. Although the linker never arranges two partial programs'
2765 data sections with overlapping addresses after linking, @emph{by definition}
2766 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2767 part of a program is always the same address when the program is running as
2768 address @code{@{absolute@ 239@}} in any other part of the program.
2770 The idea of sections is extended to the @dfn{undefined} section. Any
2771 address whose section is unknown at assembly time is by definition
2772 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2773 Since numbers are always defined, the only way to generate an undefined
2774 address is to mention an undefined symbol. A reference to a named
2775 common block would be such a symbol: its value is unknown at assembly
2776 time so it has section @emph{undefined}.
2778 By analogy the word @emph{section} is used to describe groups of sections in
2779 the linked program. @code{@value{LD}} puts all partial programs' text
2780 sections in contiguous addresses in the linked program. It is
2781 customary to refer to the @emph{text section} of a program, meaning all
2782 the addresses of all partial programs' text sections. Likewise for
2783 data and bss sections.
2785 Some sections are manipulated by @code{@value{LD}}; others are invented for
2786 use of @command{@value{AS}} and have no meaning except during assembly.
2789 @section Linker Sections
2790 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2795 @cindex named sections
2796 @cindex sections, named
2797 @item named sections
2800 @cindex text section
2801 @cindex data section
2805 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2806 separate but equal sections. Anything you can say of one section is
2809 When the program is running, however, it is
2810 customary for the text section to be unalterable. The
2811 text section is often shared among processes: it contains
2812 instructions, constants and the like. The data section of a running
2813 program is usually alterable: for example, C variables would be stored
2814 in the data section.
2819 This section contains zeroed bytes when your program begins running. It
2820 is used to hold uninitialized variables or common storage. The length of
2821 each partial program's bss section is important, but because it starts
2822 out containing zeroed bytes there is no need to store explicit zero
2823 bytes in the object file. The bss section was invented to eliminate
2824 those explicit zeros from object files.
2826 @cindex absolute section
2827 @item absolute section
2828 Address 0 of this section is always ``relocated'' to runtime address 0.
2829 This is useful if you want to refer to an address that @code{@value{LD}} must
2830 not change when relocating. In this sense we speak of absolute
2831 addresses being ``unrelocatable'': they do not change during relocation.
2833 @cindex undefined section
2834 @item undefined section
2835 This ``section'' is a catch-all for address references to objects not in
2836 the preceding sections.
2837 @c FIXME: ref to some other doc on obj-file formats could go here.
2840 @cindex relocation example
2841 An idealized example of three relocatable sections follows.
2843 The example uses the traditional section names @samp{.text} and @samp{.data}.
2845 Memory addresses are on the horizontal axis.
2849 @c END TEXI2ROFF-KILL
2852 partial program # 1: |ttttt|dddd|00|
2859 partial program # 2: |TTT|DDD|000|
2862 +--+---+-----+--+----+---+-----+~~
2863 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2864 +--+---+-----+--+----+---+-----+~~
2866 addresses: 0 @dots{}
2873 \line{\it Partial program \#1: \hfil}
2874 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2875 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2877 \line{\it Partial program \#2: \hfil}
2878 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2879 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2881 \line{\it linked program: \hfil}
2882 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2883 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2884 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2885 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2887 \line{\it addresses: \hfil}
2891 @c END TEXI2ROFF-KILL
2894 @section Assembler Internal Sections
2896 @cindex internal assembler sections
2897 @cindex sections in messages, internal
2898 These sections are meant only for the internal use of @command{@value{AS}}. They
2899 have no meaning at run-time. You do not really need to know about these
2900 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2901 warning messages, so it might be helpful to have an idea of their
2902 meanings to @command{@value{AS}}. These sections are used to permit the
2903 value of every expression in your assembly language program to be a
2904 section-relative address.
2907 @cindex assembler internal logic error
2908 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2909 An internal assembler logic error has been found. This means there is a
2910 bug in the assembler.
2912 @cindex expr (internal section)
2914 The assembler stores complex expression internally as combinations of
2915 symbols. When it needs to represent an expression as a symbol, it puts
2916 it in the expr section.
2918 @c FIXME item transfer[t] vector preload
2919 @c FIXME item transfer[t] vector postload
2920 @c FIXME item register
2924 @section Sub-Sections
2926 @cindex numbered subsections
2927 @cindex grouping data
2933 fall into two sections: text and data.
2935 You may have separate groups of
2937 data in named sections
2941 data in named sections
2947 that you want to end up near to each other in the object file, even though they
2948 are not contiguous in the assembler source. @command{@value{AS}} allows you to
2949 use @dfn{subsections} for this purpose. Within each section, there can be
2950 numbered subsections with values from 0 to 8192. Objects assembled into the
2951 same subsection go into the object file together with other objects in the same
2952 subsection. For example, a compiler might want to store constants in the text
2953 section, but might not want to have them interspersed with the program being
2954 assembled. In this case, the compiler could issue a @samp{.text 0} before each
2955 section of code being output, and a @samp{.text 1} before each group of
2956 constants being output.
2958 Subsections are optional. If you do not use subsections, everything
2959 goes in subsection number zero.
2962 Each subsection is zero-padded up to a multiple of four bytes.
2963 (Subsections may be padded a different amount on different flavors
2964 of @command{@value{AS}}.)
2968 On the H8/300 and H8/500 platforms, each subsection is zero-padded to a word
2969 boundary (two bytes).
2970 The same is true on the Renesas SH.
2973 @c FIXME section padding (alignment)?
2974 @c Rich Pixley says padding here depends on target obj code format; that
2975 @c doesn't seem particularly useful to say without further elaboration,
2976 @c so for now I say nothing about it. If this is a generic BFD issue,
2977 @c these paragraphs might need to vanish from this manual, and be
2978 @c discussed in BFD chapter of binutils (or some such).
2981 On the AMD 29K family, no particular padding is added to section or
2982 subsection sizes; @value{AS} forces no alignment on this platform.
2986 Subsections appear in your object file in numeric order, lowest numbered
2987 to highest. (All this to be compatible with other people's assemblers.)
2988 The object file contains no representation of subsections; @code{@value{LD}} and
2989 other programs that manipulate object files see no trace of them.
2990 They just see all your text subsections as a text section, and all your
2991 data subsections as a data section.
2993 To specify which subsection you want subsequent statements assembled
2994 into, use a numeric argument to specify it, in a @samp{.text
2995 @var{expression}} or a @samp{.data @var{expression}} statement.
2998 When generating COFF output, you
3003 can also use an extra subsection
3004 argument with arbitrary named sections: @samp{.section @var{name},
3009 When generating ELF output, you
3014 can also use the @code{.subsection} directive (@pxref{SubSection})
3015 to specify a subsection: @samp{.subsection @var{expression}}.
3017 @var{Expression} should be an absolute expression.
3018 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
3019 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3020 begins in @code{text 0}. For instance:
3022 .text 0 # The default subsection is text 0 anyway.
3023 .ascii "This lives in the first text subsection. *"
3025 .ascii "But this lives in the second text subsection."
3027 .ascii "This lives in the data section,"
3028 .ascii "in the first data subsection."
3030 .ascii "This lives in the first text section,"
3031 .ascii "immediately following the asterisk (*)."
3034 Each section has a @dfn{location counter} incremented by one for every byte
3035 assembled into that section. Because subsections are merely a convenience
3036 restricted to @command{@value{AS}} there is no concept of a subsection location
3037 counter. There is no way to directly manipulate a location counter---but the
3038 @code{.align} directive changes it, and any label definition captures its
3039 current value. The location counter of the section where statements are being
3040 assembled is said to be the @dfn{active} location counter.
3043 @section bss Section
3046 @cindex common variable storage
3047 The bss section is used for local common variable storage.
3048 You may allocate address space in the bss section, but you may
3049 not dictate data to load into it before your program executes. When
3050 your program starts running, all the contents of the bss
3051 section are zeroed bytes.
3053 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3054 @ref{Lcomm,,@code{.lcomm}}.
3056 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3057 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
3060 When assembling for a target which supports multiple sections, such as ELF or
3061 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3062 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3063 section. Typically the section will only contain symbol definitions and
3064 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3071 Symbols are a central concept: the programmer uses symbols to name
3072 things, the linker uses symbols to link, and the debugger uses symbols
3076 @cindex debuggers, and symbol order
3077 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3078 the same order they were declared. This may break some debuggers.
3083 * Setting Symbols:: Giving Symbols Other Values
3084 * Symbol Names:: Symbol Names
3085 * Dot:: The Special Dot Symbol
3086 * Symbol Attributes:: Symbol Attributes
3093 A @dfn{label} is written as a symbol immediately followed by a colon
3094 @samp{:}. The symbol then represents the current value of the
3095 active location counter, and is, for example, a suitable instruction
3096 operand. You are warned if you use the same symbol to represent two
3097 different locations: the first definition overrides any other
3101 On the HPPA, the usual form for a label need not be immediately followed by a
3102 colon, but instead must start in column zero. Only one label may be defined on
3103 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3104 provides a special directive @code{.label} for defining labels more flexibly.
3107 @node Setting Symbols
3108 @section Giving Symbols Other Values
3110 @cindex assigning values to symbols
3111 @cindex symbol values, assigning
3112 A symbol can be given an arbitrary value by writing a symbol, followed
3113 by an equals sign @samp{=}, followed by an expression
3114 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3115 directive. @xref{Set,,@code{.set}}.
3118 @section Symbol Names
3120 @cindex symbol names
3121 @cindex names, symbol
3122 @ifclear SPECIAL-SYMS
3123 Symbol names begin with a letter or with one of @samp{._}. On most
3124 machines, you can also use @code{$} in symbol names; exceptions are
3125 noted in @ref{Machine Dependencies}. That character may be followed by any
3126 string of digits, letters, dollar signs (unless otherwise noted in
3127 @ref{Machine Dependencies}), and underscores.
3130 For the AMD 29K family, @samp{?} is also allowed in the
3131 body of a symbol name, though not at its beginning.
3136 Symbol names begin with a letter or with one of @samp{._}. On the
3137 Renesas SH or the H8/500, you can also use @code{$} in symbol names. That
3138 character may be followed by any string of digits, letters, dollar signs (save
3139 on the H8/300), and underscores.
3143 Case of letters is significant: @code{foo} is a different symbol name
3146 Each symbol has exactly one name. Each name in an assembly language program
3147 refers to exactly one symbol. You may use that symbol name any number of times
3150 @subheading Local Symbol Names
3152 @cindex local symbol names
3153 @cindex symbol names, local
3154 @cindex temporary symbol names
3155 @cindex symbol names, temporary
3156 Local symbols help compilers and programmers use names temporarily.
3157 They create symbols which are guaranteed to be unique over the entire scope of
3158 the input source code and which can be referred to by a simple notation.
3159 To define a local symbol, write a label of the form @samp{@b{N}:} (where @b{N}
3160 represents any positive integer). To refer to the most recent previous
3161 definition of that symbol write @samp{@b{N}b}, using the same number as when
3162 you defined the label. To refer to the next definition of a local label, write
3163 @samp{@b{N}f}--- The @samp{b} stands for``backwards'' and the @samp{f} stands
3166 There is no restriction on how you can use these labels, and you can reuse them
3167 too. So that it is possible to repeatedly define the same local label (using
3168 the same number @samp{@b{N}}), although you can only refer to the most recently
3169 defined local label of that number (for a backwards reference) or the next
3170 definition of a specific local label for a forward reference. It is also worth
3171 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3172 implemented in a slightly more efficient manner than the others.
3183 Which is the equivalent of:
3186 label_1: branch label_3
3187 label_2: branch label_1
3188 label_3: branch label_4
3189 label_4: branch label_3
3192 Local symbol names are only a notational device. They are immediately
3193 transformed into more conventional symbol names before the assembler uses them.
3194 The symbol names stored in the symbol table, appearing in error messages and
3195 optionally emitted to the object file. The names are constructed using these
3200 All local labels begin with @samp{L}. Normally both @command{@value{AS}} and
3201 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
3202 used for symbols you are never intended to see. If you use the
3203 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3204 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3205 you may use them in debugging.
3208 This is the number that was used in the local label definition. So if the
3209 label is written @samp{55:} then the number is @samp{55}.
3212 This unusual character is included so you do not accidentally invent a symbol
3213 of the same name. The character has ASCII value of @samp{\002} (control-B).
3215 @item @emph{ordinal number}
3216 This is a serial number to keep the labels distinct. The first definition of
3217 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3218 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3219 the number @samp{1} and its 15th defintion gets @samp{15} as well.
3222 So for example, the first @code{1:} is named @code{L1@kbd{C-B}1}, the 44th
3223 @code{3:} is named @code{L3@kbd{C-B}44}.
3225 @subheading Dollar Local Labels
3226 @cindex dollar local symbols
3228 @code{@value{AS}} also supports an even more local form of local labels called
3229 dollar labels. These labels go out of scope (ie they become undefined) as soon
3230 as a non-local label is defined. Thus they remain valid for only a small
3231 region of the input source code. Normal local labels, by contrast, remain in
3232 scope for the entire file, or until they are redefined by another occurrence of
3233 the same local label.
3235 Dollar labels are defined in exactly the same way as ordinary local labels,
3236 except that instead of being terminated by a colon, they are terminated by a
3237 dollar sign. eg @samp{@b{55$}}.
3239 They can also be distinguished from ordinary local labels by their transformed
3240 name which uses ASCII character @samp{\001} (control-A) as the magic character
3241 to distinguish them from ordinary labels. Thus the 5th defintion of @samp{6$}
3242 is named @samp{L6@kbd{C-A}5}.
3245 @section The Special Dot Symbol
3247 @cindex dot (symbol)
3248 @cindex @code{.} (symbol)
3249 @cindex current address
3250 @cindex location counter
3251 The special symbol @samp{.} refers to the current address that
3252 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3253 .long .} defines @code{melvin} to contain its own address.
3254 Assigning a value to @code{.} is treated the same as a @code{.org}
3255 directive. Thus, the expression @samp{.=.+4} is the same as saying
3256 @ifclear no-space-dir
3265 @node Symbol Attributes
3266 @section Symbol Attributes
3268 @cindex symbol attributes
3269 @cindex attributes, symbol
3270 Every symbol has, as well as its name, the attributes ``Value'' and
3271 ``Type''. Depending on output format, symbols can also have auxiliary
3274 The detailed definitions are in @file{a.out.h}.
3277 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3278 all these attributes, and probably won't warn you. This makes the
3279 symbol an externally defined symbol, which is generally what you
3283 * Symbol Value:: Value
3284 * Symbol Type:: Type
3287 * a.out Symbols:: Symbol Attributes: @code{a.out}
3291 * a.out Symbols:: Symbol Attributes: @code{a.out}
3294 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3299 * COFF Symbols:: Symbol Attributes for COFF
3302 * SOM Symbols:: Symbol Attributes for SOM
3309 @cindex value of a symbol
3310 @cindex symbol value
3311 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3312 location in the text, data, bss or absolute sections the value is the
3313 number of addresses from the start of that section to the label.
3314 Naturally for text, data and bss sections the value of a symbol changes
3315 as @code{@value{LD}} changes section base addresses during linking. Absolute
3316 symbols' values do not change during linking: that is why they are
3319 The value of an undefined symbol is treated in a special way. If it is
3320 0 then the symbol is not defined in this assembler source file, and
3321 @code{@value{LD}} tries to determine its value from other files linked into the
3322 same program. You make this kind of symbol simply by mentioning a symbol
3323 name without defining it. A non-zero value represents a @code{.comm}
3324 common declaration. The value is how much common storage to reserve, in
3325 bytes (addresses). The symbol refers to the first address of the
3331 @cindex type of a symbol
3333 The type attribute of a symbol contains relocation (section)
3334 information, any flag settings indicating that a symbol is external, and
3335 (optionally), other information for linkers and debuggers. The exact
3336 format depends on the object-code output format in use.
3341 @c The following avoids a "widow" subsection title. @group would be
3342 @c better if it were available outside examples.
3345 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3347 @cindex @code{b.out} symbol attributes
3348 @cindex symbol attributes, @code{b.out}
3349 These symbol attributes appear only when @command{@value{AS}} is configured for
3350 one of the Berkeley-descended object output formats---@code{a.out} or
3356 @subsection Symbol Attributes: @code{a.out}
3358 @cindex @code{a.out} symbol attributes
3359 @cindex symbol attributes, @code{a.out}
3365 @subsection Symbol Attributes: @code{a.out}
3367 @cindex @code{a.out} symbol attributes
3368 @cindex symbol attributes, @code{a.out}
3372 * Symbol Desc:: Descriptor
3373 * Symbol Other:: Other
3377 @subsubsection Descriptor
3379 @cindex descriptor, of @code{a.out} symbol
3380 This is an arbitrary 16-bit value. You may establish a symbol's
3381 descriptor value by using a @code{.desc} statement
3382 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3383 @command{@value{AS}}.
3386 @subsubsection Other
3388 @cindex other attribute, of @code{a.out} symbol
3389 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3394 @subsection Symbol Attributes for COFF
3396 @cindex COFF symbol attributes
3397 @cindex symbol attributes, COFF
3399 The COFF format supports a multitude of auxiliary symbol attributes;
3400 like the primary symbol attributes, they are set between @code{.def} and
3401 @code{.endef} directives.
3403 @subsubsection Primary Attributes
3405 @cindex primary attributes, COFF symbols
3406 The symbol name is set with @code{.def}; the value and type,
3407 respectively, with @code{.val} and @code{.type}.
3409 @subsubsection Auxiliary Attributes
3411 @cindex auxiliary attributes, COFF symbols
3412 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3413 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3414 table information for COFF.
3419 @subsection Symbol Attributes for SOM
3421 @cindex SOM symbol attributes
3422 @cindex symbol attributes, SOM
3424 The SOM format for the HPPA supports a multitude of symbol attributes set with
3425 the @code{.EXPORT} and @code{.IMPORT} directives.
3427 The attributes are described in @cite{HP9000 Series 800 Assembly
3428 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3429 @code{EXPORT} assembler directive documentation.
3433 @chapter Expressions
3437 @cindex numeric values
3438 An @dfn{expression} specifies an address or numeric value.
3439 Whitespace may precede and/or follow an expression.
3441 The result of an expression must be an absolute number, or else an offset into
3442 a particular section. If an expression is not absolute, and there is not
3443 enough information when @command{@value{AS}} sees the expression to know its
3444 section, a second pass over the source program might be necessary to interpret
3445 the expression---but the second pass is currently not implemented.
3446 @command{@value{AS}} aborts with an error message in this situation.
3449 * Empty Exprs:: Empty Expressions
3450 * Integer Exprs:: Integer Expressions
3454 @section Empty Expressions
3456 @cindex empty expressions
3457 @cindex expressions, empty
3458 An empty expression has no value: it is just whitespace or null.
3459 Wherever an absolute expression is required, you may omit the
3460 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3461 is compatible with other assemblers.
3464 @section Integer Expressions
3466 @cindex integer expressions
3467 @cindex expressions, integer
3468 An @dfn{integer expression} is one or more @emph{arguments} delimited
3469 by @emph{operators}.
3472 * Arguments:: Arguments
3473 * Operators:: Operators
3474 * Prefix Ops:: Prefix Operators
3475 * Infix Ops:: Infix Operators
3479 @subsection Arguments
3481 @cindex expression arguments
3482 @cindex arguments in expressions
3483 @cindex operands in expressions
3484 @cindex arithmetic operands
3485 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3486 contexts arguments are sometimes called ``arithmetic operands''. In
3487 this manual, to avoid confusing them with the ``instruction operands'' of
3488 the machine language, we use the term ``argument'' to refer to parts of
3489 expressions only, reserving the word ``operand'' to refer only to machine
3490 instruction operands.
3492 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3493 @var{section} is one of text, data, bss, absolute,
3494 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3497 Numbers are usually integers.
3499 A number can be a flonum or bignum. In this case, you are warned
3500 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3501 these 32 bits are an integer. You may write integer-manipulating
3502 instructions that act on exotic constants, compatible with other
3505 @cindex subexpressions
3506 Subexpressions are a left parenthesis @samp{(} followed by an integer
3507 expression, followed by a right parenthesis @samp{)}; or a prefix
3508 operator followed by an argument.
3511 @subsection Operators
3513 @cindex operators, in expressions
3514 @cindex arithmetic functions
3515 @cindex functions, in expressions
3516 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3517 operators are followed by an argument. Infix operators appear
3518 between their arguments. Operators may be preceded and/or followed by
3522 @subsection Prefix Operator
3524 @cindex prefix operators
3525 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3526 one argument, which must be absolute.
3528 @c the tex/end tex stuff surrounding this small table is meant to make
3529 @c it align, on the printed page, with the similar table in the next
3530 @c section (which is inside an enumerate).
3532 \global\advance\leftskip by \itemindent
3537 @dfn{Negation}. Two's complement negation.
3539 @dfn{Complementation}. Bitwise not.
3543 \global\advance\leftskip by -\itemindent
3547 @subsection Infix Operators
3549 @cindex infix operators
3550 @cindex operators, permitted arguments
3551 @dfn{Infix operators} take two arguments, one on either side. Operators
3552 have precedence, but operations with equal precedence are performed left
3553 to right. Apart from @code{+} or @option{-}, both arguments must be
3554 absolute, and the result is absolute.
3557 @cindex operator precedence
3558 @cindex precedence of operators
3565 @dfn{Multiplication}.
3568 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3575 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3579 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3583 Intermediate precedence
3588 @dfn{Bitwise Inclusive Or}.
3594 @dfn{Bitwise Exclusive Or}.
3597 @dfn{Bitwise Or Not}.
3604 @cindex addition, permitted arguments
3605 @cindex plus, permitted arguments
3606 @cindex arguments for addition
3608 @dfn{Addition}. If either argument is absolute, the result has the section of
3609 the other argument. You may not add together arguments from different
3612 @cindex subtraction, permitted arguments
3613 @cindex minus, permitted arguments
3614 @cindex arguments for subtraction
3616 @dfn{Subtraction}. If the right argument is absolute, the
3617 result has the section of the left argument.
3618 If both arguments are in the same section, the result is absolute.
3619 You may not subtract arguments from different sections.
3620 @c FIXME is there still something useful to say about undefined - undefined ?
3622 @cindex comparison expressions
3623 @cindex expressions, comparison
3627 @dfn{Is Not Equal To}
3631 @dfn{Is Greater Than}
3633 @dfn{Is Greater Than Or Equal To}
3635 @dfn{Is Less Than Or Equal To}
3637 The comparison operators can be used as infix operators. A true results has a
3638 value of -1 whereas a false result has a value of 0. Note, these operators
3639 perform signed comparisons.
3642 @item Lowest Precedence
3651 These two logical operations can be used to combine the results of sub
3652 expressions. Note, unlike the comparison operators a true result returns a
3653 value of 1 but a false results does still return 0. Also note that the logical
3654 or operator has a slightly lower precedence than logical and.
3659 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3660 address; you can only have a defined section in one of the two arguments.
3663 @chapter Assembler Directives
3665 @cindex directives, machine independent
3666 @cindex pseudo-ops, machine independent
3667 @cindex machine independent directives
3668 All assembler directives have names that begin with a period (@samp{.}).
3669 The rest of the name is letters, usually in lower case.
3671 This chapter discusses directives that are available regardless of the
3672 target machine configuration for the @sc{gnu} assembler.
3674 Some machine configurations provide additional directives.
3675 @xref{Machine Dependencies}.
3678 @ifset machine-directives
3679 @xref{Machine Dependencies} for additional directives.
3684 * Abort:: @code{.abort}
3686 * ABORT:: @code{.ABORT}
3689 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3690 * Altmacro:: @code{.altmacro}
3691 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3692 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3693 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3694 * Byte:: @code{.byte @var{expressions}}
3695 * Comm:: @code{.comm @var{symbol} , @var{length} }
3697 * CFI directives:: @code{.cfi_startproc}, @code{.cfi_endproc}, etc.
3699 * Data:: @code{.data @var{subsection}}
3701 * Def:: @code{.def @var{name}}
3704 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3710 * Double:: @code{.double @var{flonums}}
3711 * Eject:: @code{.eject}
3712 * Else:: @code{.else}
3713 * Elseif:: @code{.elseif}
3716 * Endef:: @code{.endef}
3719 * Endfunc:: @code{.endfunc}
3720 * Endif:: @code{.endif}
3721 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3722 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3724 * Exitm:: @code{.exitm}
3725 * Extern:: @code{.extern}
3726 * Fail:: @code{.fail}
3727 @ifclear no-file-dir
3728 * File:: @code{.file @var{string}}
3731 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3732 * Float:: @code{.float @var{flonums}}
3733 * Func:: @code{.func}
3734 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3736 * Hidden:: @code{.hidden @var{names}}
3739 * hword:: @code{.hword @var{expressions}}
3740 * Ident:: @code{.ident}
3741 * If:: @code{.if @var{absolute expression}}
3742 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3743 * Include:: @code{.include "@var{file}"}
3744 * Int:: @code{.int @var{expressions}}
3746 * Internal:: @code{.internal @var{names}}
3749 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3750 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3751 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3752 * Lflags:: @code{.lflags}
3753 @ifclear no-line-dir
3754 * Line:: @code{.line @var{line-number}}
3757 * Ln:: @code{.ln @var{line-number}}
3758 * Linkonce:: @code{.linkonce [@var{type}]}
3759 * List:: @code{.list}
3760 * Long:: @code{.long @var{expressions}}
3762 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3765 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3766 * MRI:: @code{.mri @var{val}}
3767 * Noaltmacro:: @code{.noaltmacro}
3768 * Nolist:: @code{.nolist}
3769 * Octa:: @code{.octa @var{bignums}}
3770 * Org:: @code{.org @var{new-lc} , @var{fill}}
3771 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3773 * PopSection:: @code{.popsection}
3774 * Previous:: @code{.previous}
3777 * Print:: @code{.print @var{string}}
3779 * Protected:: @code{.protected @var{names}}
3782 * Psize:: @code{.psize @var{lines}, @var{columns}}
3783 * Purgem:: @code{.purgem @var{name}}
3785 * PushSection:: @code{.pushsection @var{name}}
3788 * Quad:: @code{.quad @var{bignums}}
3789 * Rept:: @code{.rept @var{count}}
3790 * Sbttl:: @code{.sbttl "@var{subheading}"}
3792 * Scl:: @code{.scl @var{class}}
3795 * Section:: @code{.section @var{name}}
3798 * Set:: @code{.set @var{symbol}, @var{expression}}
3799 * Short:: @code{.short @var{expressions}}
3800 * Single:: @code{.single @var{flonums}}
3802 * Size:: @code{.size [@var{name} , @var{expression}]}
3805 * Skip:: @code{.skip @var{size} , @var{fill}}
3806 * Sleb128:: @code{.sleb128 @var{expressions}}
3807 * Space:: @code{.space @var{size} , @var{fill}}
3809 * Stab:: @code{.stabd, .stabn, .stabs}
3812 * String:: @code{.string "@var{str}"}
3813 * Struct:: @code{.struct @var{expression}}
3815 * SubSection:: @code{.subsection}
3816 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3820 * Tag:: @code{.tag @var{structname}}
3823 * Text:: @code{.text @var{subsection}}
3824 * Title:: @code{.title "@var{heading}"}
3826 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3829 * Uleb128:: @code{.uleb128 @var{expressions}}
3831 * Val:: @code{.val @var{addr}}
3835 * Version:: @code{.version "@var{string}"}
3836 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3837 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3840 * Weak:: @code{.weak @var{names}}
3841 * Word:: @code{.word @var{expressions}}
3842 * Deprecated:: Deprecated Directives
3846 @section @code{.abort}
3848 @cindex @code{abort} directive
3849 @cindex stopping the assembly
3850 This directive stops the assembly immediately. It is for
3851 compatibility with other assemblers. The original idea was that the
3852 assembly language source would be piped into the assembler. If the sender
3853 of the source quit, it could use this directive tells @command{@value{AS}} to
3854 quit also. One day @code{.abort} will not be supported.
3858 @section @code{.ABORT}
3860 @cindex @code{ABORT} directive
3861 When producing COFF output, @command{@value{AS}} accepts this directive as a
3862 synonym for @samp{.abort}.
3865 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3871 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3873 @cindex padding the location counter
3874 @cindex @code{align} directive
3875 Pad the location counter (in the current subsection) to a particular storage
3876 boundary. The first expression (which must be absolute) is the alignment
3877 required, as described below.
3879 The second expression (also absolute) gives the fill value to be stored in the
3880 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3881 padding bytes are normally zero. However, on some systems, if the section is
3882 marked as containing code and the fill value is omitted, the space is filled
3883 with no-op instructions.
3885 The third expression is also absolute, and is also optional. If it is present,
3886 it is the maximum number of bytes that should be skipped by this alignment
3887 directive. If doing the alignment would require skipping more bytes than the
3888 specified maximum, then the alignment is not done at all. You can omit the
3889 fill value (the second argument) entirely by simply using two commas after the
3890 required alignment; this can be useful if you want the alignment to be filled
3891 with no-op instructions when appropriate.
3893 The way the required alignment is specified varies from system to system.
3894 For the a29k, arc, hppa, i386 using ELF, i860, iq2000, m68k, m88k, or32,
3895 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
3896 alignment request in bytes. For example @samp{.align 8} advances
3897 the location counter until it is a multiple of 8. If the location counter
3898 is already a multiple of 8, no change is needed. For the tic54x, the
3899 first expression is the alignment request in words.
3901 For other systems, including the i386 using a.out format, and the arm and
3902 strongarm, it is the
3903 number of low-order zero bits the location counter must have after
3904 advancement. For example @samp{.align 3} advances the location
3905 counter until it a multiple of 8. If the location counter is already a
3906 multiple of 8, no change is needed.
3908 This inconsistency is due to the different behaviors of the various
3909 native assemblers for these systems which GAS must emulate.
3910 GAS also provides @code{.balign} and @code{.p2align} directives,
3911 described later, which have a consistent behavior across all
3912 architectures (but are specific to GAS).
3915 @section @code{.ascii "@var{string}"}@dots{}
3917 @cindex @code{ascii} directive
3918 @cindex string literals
3919 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3920 separated by commas. It assembles each string (with no automatic
3921 trailing zero byte) into consecutive addresses.
3924 @section @code{.asciz "@var{string}"}@dots{}
3926 @cindex @code{asciz} directive
3927 @cindex zero-terminated strings
3928 @cindex null-terminated strings
3929 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3930 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3933 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3935 @cindex padding the location counter given number of bytes
3936 @cindex @code{balign} directive
3937 Pad the location counter (in the current subsection) to a particular
3938 storage boundary. The first expression (which must be absolute) is the
3939 alignment request in bytes. For example @samp{.balign 8} advances
3940 the location counter until it is a multiple of 8. If the location counter
3941 is already a multiple of 8, no change is needed.
3943 The second expression (also absolute) gives the fill value to be stored in the
3944 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3945 padding bytes are normally zero. However, on some systems, if the section is
3946 marked as containing code and the fill value is omitted, the space is filled
3947 with no-op instructions.
3949 The third expression is also absolute, and is also optional. If it is present,
3950 it is the maximum number of bytes that should be skipped by this alignment
3951 directive. If doing the alignment would require skipping more bytes than the
3952 specified maximum, then the alignment is not done at all. You can omit the
3953 fill value (the second argument) entirely by simply using two commas after the
3954 required alignment; this can be useful if you want the alignment to be filled
3955 with no-op instructions when appropriate.
3957 @cindex @code{balignw} directive
3958 @cindex @code{balignl} directive
3959 The @code{.balignw} and @code{.balignl} directives are variants of the
3960 @code{.balign} directive. The @code{.balignw} directive treats the fill
3961 pattern as a two byte word value. The @code{.balignl} directives treats the
3962 fill pattern as a four byte longword value. For example, @code{.balignw
3963 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
3964 filled in with the value 0x368d (the exact placement of the bytes depends upon
3965 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
3969 @section @code{.byte @var{expressions}}
3971 @cindex @code{byte} directive
3972 @cindex integers, one byte
3973 @code{.byte} expects zero or more expressions, separated by commas.
3974 Each expression is assembled into the next byte.
3977 @section @code{.comm @var{symbol} , @var{length} }
3979 @cindex @code{comm} directive
3980 @cindex symbol, common
3981 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
3982 common symbol in one object file may be merged with a defined or common symbol
3983 of the same name in another object file. If @code{@value{LD}} does not see a
3984 definition for the symbol--just one or more common symbols--then it will
3985 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
3986 absolute expression. If @code{@value{LD}} sees multiple common symbols with
3987 the same name, and they do not all have the same size, it will allocate space
3988 using the largest size.
3991 When using ELF, the @code{.comm} directive takes an optional third argument.
3992 This is the desired alignment of the symbol, specified as a byte boundary (for
3993 example, an alignment of 16 means that the least significant 4 bits of the
3994 address should be zero). The alignment must be an absolute expression, and it
3995 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
3996 for the common symbol, it will use the alignment when placing the symbol. If
3997 no alignment is specified, @command{@value{AS}} will set the alignment to the
3998 largest power of two less than or equal to the size of the symbol, up to a
4003 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4004 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4007 @node CFI directives
4008 @section @code{.cfi_startproc}
4009 @cindex @code{cfi_startproc} directive
4010 @code{.cfi_startproc} is used at the beginning of each function that
4011 should have an entry in @code{.eh_frame}. It initializes some internal
4012 data structures and emits architecture dependent initial CFI instructions.
4013 Don't forget to close the function by
4014 @code{.cfi_endproc}.
4016 @section @code{.cfi_endproc}
4017 @cindex @code{cfi_endproc} directive
4018 @code{.cfi_endproc} is used at the end of a function where it closes its
4019 unwind entry previously opened by
4020 @code{.cfi_startproc}. and emits it to @code{.eh_frame}.
4022 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4023 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4024 address from @var{register} and add @var{offset} to it}.
4026 @section @code{.cfi_def_cfa_register @var{register}}
4027 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4028 now on @var{register} will be used instead of the old one. Offset
4031 @section @code{.cfi_def_cfa_offset @var{offset}}
4032 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4033 remains the same, but @var{offset} is new. Note that it is the
4034 absolute offset that will be added to a defined register to compute
4037 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4038 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4039 value that is added/substracted from the previous offset.
4041 @section @code{.cfi_offset @var{register}, @var{offset}}
4042 Previous value of @var{register} is saved at offset @var{offset} from
4045 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4046 Previous value of @var{register} is saved at offset @var{offset} from
4047 the current CFA register. This is transformed to @code{.cfi_offset}
4048 using the known displacement of the CFA register from the CFA.
4049 This is often easier to use, because the number will match the
4050 code it's annotating.
4052 @section @code{.cfi_window_save}
4053 SPARC register window has been saved.
4055 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4056 Allows the user to add arbitrary bytes to the unwind info. One
4057 might use this to add OS-specific CFI opcodes, or generic CFI
4058 opcodes that GAS does not yet support.
4061 @section @code{.data @var{subsection}}
4063 @cindex @code{data} directive
4064 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4065 end of the data subsection numbered @var{subsection} (which is an
4066 absolute expression). If @var{subsection} is omitted, it defaults
4071 @section @code{.def @var{name}}
4073 @cindex @code{def} directive
4074 @cindex COFF symbols, debugging
4075 @cindex debugging COFF symbols
4076 Begin defining debugging information for a symbol @var{name}; the
4077 definition extends until the @code{.endef} directive is encountered.
4080 This directive is only observed when @command{@value{AS}} is configured for COFF
4081 format output; when producing @code{b.out}, @samp{.def} is recognized,
4088 @section @code{.desc @var{symbol}, @var{abs-expression}}
4090 @cindex @code{desc} directive
4091 @cindex COFF symbol descriptor
4092 @cindex symbol descriptor, COFF
4093 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4094 to the low 16 bits of an absolute expression.
4097 The @samp{.desc} directive is not available when @command{@value{AS}} is
4098 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4099 object format. For the sake of compatibility, @command{@value{AS}} accepts
4100 it, but produces no output, when configured for COFF.
4106 @section @code{.dim}
4108 @cindex @code{dim} directive
4109 @cindex COFF auxiliary symbol information
4110 @cindex auxiliary symbol information, COFF
4111 This directive is generated by compilers to include auxiliary debugging
4112 information in the symbol table. It is only permitted inside
4113 @code{.def}/@code{.endef} pairs.
4116 @samp{.dim} is only meaningful when generating COFF format output; when
4117 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4123 @section @code{.double @var{flonums}}
4125 @cindex @code{double} directive
4126 @cindex floating point numbers (double)
4127 @code{.double} expects zero or more flonums, separated by commas. It
4128 assembles floating point numbers.
4130 The exact kind of floating point numbers emitted depends on how
4131 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4135 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4136 in @sc{ieee} format.
4141 @section @code{.eject}
4143 @cindex @code{eject} directive
4144 @cindex new page, in listings
4145 @cindex page, in listings
4146 @cindex listing control: new page
4147 Force a page break at this point, when generating assembly listings.
4150 @section @code{.else}
4152 @cindex @code{else} directive
4153 @code{.else} is part of the @command{@value{AS}} support for conditional
4154 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
4155 of code to be assembled if the condition for the preceding @code{.if}
4159 @section @code{.elseif}
4161 @cindex @code{elseif} directive
4162 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4163 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
4164 @code{.if} block that would otherwise fill the entire @code{.else} section.
4167 @section @code{.end}
4169 @cindex @code{end} directive
4170 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4171 process anything in the file past the @code{.end} directive.
4175 @section @code{.endef}
4177 @cindex @code{endef} directive
4178 This directive flags the end of a symbol definition begun with
4182 @samp{.endef} is only meaningful when generating COFF format output; if
4183 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4184 directive but ignores it.
4189 @section @code{.endfunc}
4190 @cindex @code{endfunc} directive
4191 @code{.endfunc} marks the end of a function specified with @code{.func}.
4194 @section @code{.endif}
4196 @cindex @code{endif} directive
4197 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4198 it marks the end of a block of code that is only assembled
4199 conditionally. @xref{If,,@code{.if}}.
4202 @section @code{.equ @var{symbol}, @var{expression}}
4204 @cindex @code{equ} directive
4205 @cindex assigning values to symbols
4206 @cindex symbols, assigning values to
4207 This directive sets the value of @var{symbol} to @var{expression}.
4208 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
4211 The syntax for @code{equ} on the HPPA is
4212 @samp{@var{symbol} .equ @var{expression}}.
4216 @section @code{.equiv @var{symbol}, @var{expression}}
4217 @cindex @code{equiv} directive
4218 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4219 the assembler will signal an error if @var{symbol} is already defined. Note a
4220 symbol which has been referenced but not actually defined is considered to be
4223 Except for the contents of the error message, this is roughly equivalent to
4232 @section @code{.err}
4233 @cindex @code{err} directive
4234 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4235 message and, unless the @option{-Z} option was used, it will not generate an
4236 object file. This can be used to signal error an conditionally compiled code.
4239 @section @code{.exitm}
4240 Exit early from the current macro definition. @xref{Macro}.
4243 @section @code{.extern}
4245 @cindex @code{extern} directive
4246 @code{.extern} is accepted in the source program---for compatibility
4247 with other assemblers---but it is ignored. @command{@value{AS}} treats
4248 all undefined symbols as external.
4251 @section @code{.fail @var{expression}}
4253 @cindex @code{fail} directive
4254 Generates an error or a warning. If the value of the @var{expression} is 500
4255 or more, @command{@value{AS}} will print a warning message. If the value is less
4256 than 500, @command{@value{AS}} will print an error message. The message will
4257 include the value of @var{expression}. This can occasionally be useful inside
4258 complex nested macros or conditional assembly.
4260 @ifclear no-file-dir
4262 @section @code{.file @var{string}}
4264 @cindex @code{file} directive
4265 @cindex logical file name
4266 @cindex file name, logical
4267 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4268 file. @var{string} is the new file name. In general, the filename is
4269 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4270 to specify an empty file name, you must give the quotes--@code{""}. This
4271 statement may go away in future: it is only recognized to be compatible with
4272 old @command{@value{AS}} programs.
4274 In some configurations of @command{@value{AS}}, @code{.file} has already been
4275 removed to avoid conflicts with other assemblers. @xref{Machine Dependencies}.
4280 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4282 @cindex @code{fill} directive
4283 @cindex writing patterns in memory
4284 @cindex patterns, writing in memory
4285 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4286 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4287 may be zero or more. @var{Size} may be zero or more, but if it is
4288 more than 8, then it is deemed to have the value 8, compatible with
4289 other people's assemblers. The contents of each @var{repeat} bytes
4290 is taken from an 8-byte number. The highest order 4 bytes are
4291 zero. The lowest order 4 bytes are @var{value} rendered in the
4292 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4293 Each @var{size} bytes in a repetition is taken from the lowest order
4294 @var{size} bytes of this number. Again, this bizarre behavior is
4295 compatible with other people's assemblers.
4297 @var{size} and @var{value} are optional.
4298 If the second comma and @var{value} are absent, @var{value} is
4299 assumed zero. If the first comma and following tokens are absent,
4300 @var{size} is assumed to be 1.
4303 @section @code{.float @var{flonums}}
4305 @cindex floating point numbers (single)
4306 @cindex @code{float} directive
4307 This directive assembles zero or more flonums, separated by commas. It
4308 has the same effect as @code{.single}.
4310 The exact kind of floating point numbers emitted depends on how
4311 @command{@value{AS}} is configured.
4312 @xref{Machine Dependencies}.
4316 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4317 in @sc{ieee} format.
4322 @section @code{.func @var{name}[,@var{label}]}
4323 @cindex @code{func} directive
4324 @code{.func} emits debugging information to denote function @var{name}, and
4325 is ignored unless the file is assembled with debugging enabled.
4326 Only @samp{--gstabs[+]} is currently supported.
4327 @var{label} is the entry point of the function and if omitted @var{name}
4328 prepended with the @samp{leading char} is used.
4329 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4330 All functions are currently defined to have @code{void} return type.
4331 The function must be terminated with @code{.endfunc}.
4334 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4336 @cindex @code{global} directive
4337 @cindex symbol, making visible to linker
4338 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4339 @var{symbol} in your partial program, its value is made available to
4340 other partial programs that are linked with it. Otherwise,
4341 @var{symbol} takes its attributes from a symbol of the same name
4342 from another file linked into the same program.
4344 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4345 compatibility with other assemblers.
4348 On the HPPA, @code{.global} is not always enough to make it accessible to other
4349 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4350 @xref{HPPA Directives,, HPPA Assembler Directives}.
4355 @section @code{.hidden @var{names}}
4357 @cindex @code{hidden} directive
4359 This is one of the ELF visibility directives. The other two are
4360 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4361 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4363 This directive overrides the named symbols default visibility (which is set by
4364 their binding: local, global or weak). The directive sets the visibility to
4365 @code{hidden} which means that the symbols are not visible to other components.
4366 Such symbols are always considered to be @code{protected} as well.
4370 @section @code{.hword @var{expressions}}
4372 @cindex @code{hword} directive
4373 @cindex integers, 16-bit
4374 @cindex numbers, 16-bit
4375 @cindex sixteen bit integers
4376 This expects zero or more @var{expressions}, and emits
4377 a 16 bit number for each.
4380 This directive is a synonym for @samp{.short}; depending on the target
4381 architecture, it may also be a synonym for @samp{.word}.
4385 This directive is a synonym for @samp{.short}.
4388 This directive is a synonym for both @samp{.short} and @samp{.word}.
4393 @section @code{.ident}
4395 @cindex @code{ident} directive
4396 This directive is used by some assemblers to place tags in object files.
4397 @command{@value{AS}} simply accepts the directive for source-file
4398 compatibility with such assemblers, but does not actually emit anything
4402 @section @code{.if @var{absolute expression}}
4404 @cindex conditional assembly
4405 @cindex @code{if} directive
4406 @code{.if} marks the beginning of a section of code which is only
4407 considered part of the source program being assembled if the argument
4408 (which must be an @var{absolute expression}) is non-zero. The end of
4409 the conditional section of code must be marked by @code{.endif}
4410 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4411 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4412 If you have several conditions to check, @code{.elseif} may be used to avoid
4413 nesting blocks if/else within each subsequent @code{.else} block.
4415 The following variants of @code{.if} are also supported:
4417 @cindex @code{ifdef} directive
4418 @item .ifdef @var{symbol}
4419 Assembles the following section of code if the specified @var{symbol}
4420 has been defined. Note a symbol which has been referenced but not yet defined
4421 is considered to be undefined.
4423 @cindex @code{ifc} directive
4424 @item .ifc @var{string1},@var{string2}
4425 Assembles the following section of code if the two strings are the same. The
4426 strings may be optionally quoted with single quotes. If they are not quoted,
4427 the first string stops at the first comma, and the second string stops at the
4428 end of the line. Strings which contain whitespace should be quoted. The
4429 string comparison is case sensitive.
4431 @cindex @code{ifeq} directive
4432 @item .ifeq @var{absolute expression}
4433 Assembles the following section of code if the argument is zero.
4435 @cindex @code{ifeqs} directive
4436 @item .ifeqs @var{string1},@var{string2}
4437 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4439 @cindex @code{ifge} directive
4440 @item .ifge @var{absolute expression}
4441 Assembles the following section of code if the argument is greater than or
4444 @cindex @code{ifgt} directive
4445 @item .ifgt @var{absolute expression}
4446 Assembles the following section of code if the argument is greater than zero.
4448 @cindex @code{ifle} directive
4449 @item .ifle @var{absolute expression}
4450 Assembles the following section of code if the argument is less than or equal
4453 @cindex @code{iflt} directive
4454 @item .iflt @var{absolute expression}
4455 Assembles the following section of code if the argument is less than zero.
4457 @cindex @code{ifnc} directive
4458 @item .ifnc @var{string1},@var{string2}.
4459 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4460 following section of code if the two strings are not the same.
4462 @cindex @code{ifndef} directive
4463 @cindex @code{ifnotdef} directive
4464 @item .ifndef @var{symbol}
4465 @itemx .ifnotdef @var{symbol}
4466 Assembles the following section of code if the specified @var{symbol}
4467 has not been defined. Both spelling variants are equivalent. Note a symbol
4468 which has been referenced but not yet defined is considered to be undefined.
4470 @cindex @code{ifne} directive
4471 @item .ifne @var{absolute expression}
4472 Assembles the following section of code if the argument is not equal to zero
4473 (in other words, this is equivalent to @code{.if}).
4475 @cindex @code{ifnes} directive
4476 @item .ifnes @var{string1},@var{string2}
4477 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4478 following section of code if the two strings are not the same.
4482 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4484 @cindex @code{incbin} directive
4485 @cindex binary files, including
4486 The @code{incbin} directive includes @var{file} verbatim at the current
4487 location. You can control the search paths used with the @samp{-I} command-line
4488 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4491 The @var{skip} argument skips a number of bytes from the start of the
4492 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4493 read. Note that the data is not aligned in any way, so it is the user's
4494 responsibility to make sure that proper alignment is provided both before and
4495 after the @code{incbin} directive.
4498 @section @code{.include "@var{file}"}
4500 @cindex @code{include} directive
4501 @cindex supporting files, including
4502 @cindex files, including
4503 This directive provides a way to include supporting files at specified
4504 points in your source program. The code from @var{file} is assembled as
4505 if it followed the point of the @code{.include}; when the end of the
4506 included file is reached, assembly of the original file continues. You
4507 can control the search paths used with the @samp{-I} command-line option
4508 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4512 @section @code{.int @var{expressions}}
4514 @cindex @code{int} directive
4515 @cindex integers, 32-bit
4516 Expect zero or more @var{expressions}, of any section, separated by commas.
4517 For each expression, emit a number that, at run time, is the value of that
4518 expression. The byte order and bit size of the number depends on what kind
4519 of target the assembly is for.
4523 On the H8/500 and most forms of the H8/300, @code{.int} emits 16-bit
4524 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4531 @section @code{.internal @var{names}}
4533 @cindex @code{internal} directive
4535 This is one of the ELF visibility directives. The other two are
4536 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4537 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4539 This directive overrides the named symbols default visibility (which is set by
4540 their binding: local, global or weak). The directive sets the visibility to
4541 @code{internal} which means that the symbols are considered to be @code{hidden}
4542 (i.e., not visible to other components), and that some extra, processor specific
4543 processing must also be performed upon the symbols as well.
4547 @section @code{.irp @var{symbol},@var{values}}@dots{}
4549 @cindex @code{irp} directive
4550 Evaluate a sequence of statements assigning different values to @var{symbol}.
4551 The sequence of statements starts at the @code{.irp} directive, and is
4552 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4553 set to @var{value}, and the sequence of statements is assembled. If no
4554 @var{value} is listed, the sequence of statements is assembled once, with
4555 @var{symbol} set to the null string. To refer to @var{symbol} within the
4556 sequence of statements, use @var{\symbol}.
4558 For example, assembling
4566 is equivalent to assembling
4575 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4577 @cindex @code{irpc} directive
4578 Evaluate a sequence of statements assigning different values to @var{symbol}.
4579 The sequence of statements starts at the @code{.irpc} directive, and is
4580 terminated by an @code{.endr} directive. For each character in @var{value},
4581 @var{symbol} is set to the character, and the sequence of statements is
4582 assembled. If no @var{value} is listed, the sequence of statements is
4583 assembled once, with @var{symbol} set to the null string. To refer to
4584 @var{symbol} within the sequence of statements, use @var{\symbol}.
4586 For example, assembling
4594 is equivalent to assembling
4603 @section @code{.lcomm @var{symbol} , @var{length}}
4605 @cindex @code{lcomm} directive
4606 @cindex local common symbols
4607 @cindex symbols, local common
4608 Reserve @var{length} (an absolute expression) bytes for a local common
4609 denoted by @var{symbol}. The section and value of @var{symbol} are
4610 those of the new local common. The addresses are allocated in the bss
4611 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4612 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4613 not visible to @code{@value{LD}}.
4616 Some targets permit a third argument to be used with @code{.lcomm}. This
4617 argument specifies the desired alignment of the symbol in the bss section.
4621 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4622 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4626 @section @code{.lflags}
4628 @cindex @code{lflags} directive (ignored)
4629 @command{@value{AS}} accepts this directive, for compatibility with other
4630 assemblers, but ignores it.
4632 @ifclear no-line-dir
4634 @section @code{.line @var{line-number}}
4636 @cindex @code{line} directive
4640 @section @code{.ln @var{line-number}}
4642 @cindex @code{ln} directive
4644 @cindex logical line number
4646 Change the logical line number. @var{line-number} must be an absolute
4647 expression. The next line has that logical line number. Therefore any other
4648 statements on the current line (after a statement separator character) are
4649 reported as on logical line number @var{line-number} @minus{} 1. One day
4650 @command{@value{AS}} will no longer support this directive: it is recognized only
4651 for compatibility with existing assembler programs.
4655 @emph{Warning:} In the AMD29K configuration of @value{AS}, this command is
4656 not available; use the synonym @code{.ln} in that context.
4661 @ifclear no-line-dir
4662 Even though this is a directive associated with the @code{a.out} or
4663 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4664 when producing COFF output, and treats @samp{.line} as though it
4665 were the COFF @samp{.ln} @emph{if} it is found outside a
4666 @code{.def}/@code{.endef} pair.
4668 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4669 used by compilers to generate auxiliary symbol information for
4674 @section @code{.linkonce [@var{type}]}
4676 @cindex @code{linkonce} directive
4677 @cindex common sections
4678 Mark the current section so that the linker only includes a single copy of it.
4679 This may be used to include the same section in several different object files,
4680 but ensure that the linker will only include it once in the final output file.
4681 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4682 Duplicate sections are detected based on the section name, so it should be
4685 This directive is only supported by a few object file formats; as of this
4686 writing, the only object file format which supports it is the Portable
4687 Executable format used on Windows NT.
4689 The @var{type} argument is optional. If specified, it must be one of the
4690 following strings. For example:
4694 Not all types may be supported on all object file formats.
4698 Silently discard duplicate sections. This is the default.
4701 Warn if there are duplicate sections, but still keep only one copy.
4704 Warn if any of the duplicates have different sizes.
4707 Warn if any of the duplicates do not have exactly the same contents.
4711 @section @code{.ln @var{line-number}}
4713 @cindex @code{ln} directive
4714 @ifclear no-line-dir
4715 @samp{.ln} is a synonym for @samp{.line}.
4718 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4719 must be an absolute expression. The next line has that logical
4720 line number, so any other statements on the current line (after a
4721 statement separator character @code{;}) are reported as on logical
4722 line number @var{line-number} @minus{} 1.
4725 This directive is accepted, but ignored, when @command{@value{AS}} is
4726 configured for @code{b.out}; its effect is only associated with COFF
4732 @section @code{.mri @var{val}}
4734 @cindex @code{mri} directive
4735 @cindex MRI mode, temporarily
4736 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4737 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4738 affects code assembled until the next @code{.mri} directive, or until the end
4739 of the file. @xref{M, MRI mode, MRI mode}.
4742 @section @code{.list}
4744 @cindex @code{list} directive
4745 @cindex listing control, turning on
4746 Control (in conjunction with the @code{.nolist} directive) whether or
4747 not assembly listings are generated. These two directives maintain an
4748 internal counter (which is zero initially). @code{.list} increments the
4749 counter, and @code{.nolist} decrements it. Assembly listings are
4750 generated whenever the counter is greater than zero.
4752 By default, listings are disabled. When you enable them (with the
4753 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4754 the initial value of the listing counter is one.
4757 @section @code{.long @var{expressions}}
4759 @cindex @code{long} directive
4760 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4763 @c no one seems to know what this is for or whether this description is
4764 @c what it really ought to do
4766 @section @code{.lsym @var{symbol}, @var{expression}}
4768 @cindex @code{lsym} directive
4769 @cindex symbol, not referenced in assembly
4770 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4771 the hash table, ensuring it cannot be referenced by name during the
4772 rest of the assembly. This sets the attributes of the symbol to be
4773 the same as the expression value:
4775 @var{other} = @var{descriptor} = 0
4776 @var{type} = @r{(section of @var{expression})}
4777 @var{value} = @var{expression}
4780 The new symbol is not flagged as external.
4784 @section @code{.macro}
4787 The commands @code{.macro} and @code{.endm} allow you to define macros that
4788 generate assembly output. For example, this definition specifies a macro
4789 @code{sum} that puts a sequence of numbers into memory:
4792 .macro sum from=0, to=5
4801 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4813 @item .macro @var{macname}
4814 @itemx .macro @var{macname} @var{macargs} @dots{}
4815 @cindex @code{macro} directive
4816 Begin the definition of a macro called @var{macname}. If your macro
4817 definition requires arguments, specify their names after the macro name,
4818 separated by commas or spaces. You can supply a default value for any
4819 macro argument by following the name with @samp{=@var{deflt}}. For
4820 example, these are all valid @code{.macro} statements:
4824 Begin the definition of a macro called @code{comm}, which takes no
4827 @item .macro plus1 p, p1
4828 @itemx .macro plus1 p p1
4829 Either statement begins the definition of a macro called @code{plus1},
4830 which takes two arguments; within the macro definition, write
4831 @samp{\p} or @samp{\p1} to evaluate the arguments.
4833 @item .macro reserve_str p1=0 p2
4834 Begin the definition of a macro called @code{reserve_str}, with two
4835 arguments. The first argument has a default value, but not the second.
4836 After the definition is complete, you can call the macro either as
4837 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
4838 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
4839 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
4840 @samp{0}, and @samp{\p2} evaluating to @var{b}).
4843 When you call a macro, you can specify the argument values either by
4844 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
4845 @samp{sum to=17, from=9}.
4848 @cindex @code{endm} directive
4849 Mark the end of a macro definition.
4852 @cindex @code{exitm} directive
4853 Exit early from the current macro definition.
4855 @cindex number of macros executed
4856 @cindex macros, count executed
4858 @command{@value{AS}} maintains a counter of how many macros it has
4859 executed in this pseudo-variable; you can copy that number to your
4860 output with @samp{\@@}, but @emph{only within a macro definition}.
4862 @item LOCAL @var{name} [ , @dots{} ]
4863 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
4864 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
4865 @xref{Altmacro,,@code{.altmacro}}.
4869 @section @code{.altmacro}
4870 Enable alternate macro mode, enabling:
4873 @item LOCAL @var{name} [ , @dots{} ]
4874 One additional directive, @code{LOCAL}, is available. It is used to
4875 generate a string replacement for each of the @var{name} arguments, and
4876 replace any instances of @var{name} in each macro expansion. The
4877 replacement string is unique in the assembly, and different for each
4878 separate macro expansion. @code{LOCAL} allows you to write macros that
4879 define symbols, without fear of conflict between separate macro expansions.
4881 @item String delimiters
4882 You can write strings delimited in these other ways besides
4883 @code{"@var{string}"}:
4886 @item '@var{string}'
4887 You can delimit strings with single-quote charaters.
4889 @item <@var{string}>
4890 You can delimit strings with matching angle brackets.
4893 @item single-character string escape
4894 To include any single character literally in a string (even if the
4895 character would otherwise have some special meaning), you can prefix the
4896 character with @samp{!} (an exclamation mark). For example, you can
4897 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4899 @item Expression results as strings
4900 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4901 and use the result as a string.
4905 @section @code{.noaltmacro}
4906 Disable alternate macro mode. @ref{Altmacro}
4909 @section @code{.nolist}
4911 @cindex @code{nolist} directive
4912 @cindex listing control, turning off
4913 Control (in conjunction with the @code{.list} directive) whether or
4914 not assembly listings are generated. These two directives maintain an
4915 internal counter (which is zero initially). @code{.list} increments the
4916 counter, and @code{.nolist} decrements it. Assembly listings are
4917 generated whenever the counter is greater than zero.
4920 @section @code{.octa @var{bignums}}
4922 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
4923 @cindex @code{octa} directive
4924 @cindex integer, 16-byte
4925 @cindex sixteen byte integer
4926 This directive expects zero or more bignums, separated by commas. For each
4927 bignum, it emits a 16-byte integer.
4929 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
4930 hence @emph{octa}-word for 16 bytes.
4933 @section @code{.org @var{new-lc} , @var{fill}}
4935 @cindex @code{org} directive
4936 @cindex location counter, advancing
4937 @cindex advancing location counter
4938 @cindex current address, advancing
4939 Advance the location counter of the current section to
4940 @var{new-lc}. @var{new-lc} is either an absolute expression or an
4941 expression with the same section as the current subsection. That is,
4942 you can't use @code{.org} to cross sections: if @var{new-lc} has the
4943 wrong section, the @code{.org} directive is ignored. To be compatible
4944 with former assemblers, if the section of @var{new-lc} is absolute,
4945 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
4946 is the same as the current subsection.
4948 @code{.org} may only increase the location counter, or leave it
4949 unchanged; you cannot use @code{.org} to move the location counter
4952 @c double negative used below "not undefined" because this is a specific
4953 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
4954 @c section. doc@cygnus.com 18feb91
4955 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
4956 may not be undefined. If you really detest this restriction we eagerly await
4957 a chance to share your improved assembler.
4959 Beware that the origin is relative to the start of the section, not
4960 to the start of the subsection. This is compatible with other
4961 people's assemblers.
4963 When the location counter (of the current subsection) is advanced, the
4964 intervening bytes are filled with @var{fill} which should be an
4965 absolute expression. If the comma and @var{fill} are omitted,
4966 @var{fill} defaults to zero.
4969 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4971 @cindex padding the location counter given a power of two
4972 @cindex @code{p2align} directive
4973 Pad the location counter (in the current subsection) to a particular
4974 storage boundary. The first expression (which must be absolute) is the
4975 number of low-order zero bits the location counter must have after
4976 advancement. For example @samp{.p2align 3} advances the location
4977 counter until it a multiple of 8. If the location counter is already a
4978 multiple of 8, no change is needed.
4980 The second expression (also absolute) gives the fill value to be stored in the
4981 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4982 padding bytes are normally zero. However, on some systems, if the section is
4983 marked as containing code and the fill value is omitted, the space is filled
4984 with no-op instructions.
4986 The third expression is also absolute, and is also optional. If it is present,
4987 it is the maximum number of bytes that should be skipped by this alignment
4988 directive. If doing the alignment would require skipping more bytes than the
4989 specified maximum, then the alignment is not done at all. You can omit the
4990 fill value (the second argument) entirely by simply using two commas after the
4991 required alignment; this can be useful if you want the alignment to be filled
4992 with no-op instructions when appropriate.
4994 @cindex @code{p2alignw} directive
4995 @cindex @code{p2alignl} directive
4996 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
4997 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
4998 pattern as a two byte word value. The @code{.p2alignl} directives treats the
4999 fill pattern as a four byte longword value. For example, @code{.p2alignw
5000 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5001 filled in with the value 0x368d (the exact placement of the bytes depends upon
5002 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5007 @section @code{.previous}
5009 @cindex @code{previous} directive
5010 @cindex Section Stack
5011 This is one of the ELF section stack manipulation directives. The others are
5012 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5013 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5014 (@pxref{PopSection}).
5016 This directive swaps the current section (and subsection) with most recently
5017 referenced section (and subsection) prior to this one. Multiple
5018 @code{.previous} directives in a row will flip between two sections (and their
5021 In terms of the section stack, this directive swaps the current section with
5022 the top section on the section stack.
5027 @section @code{.popsection}
5029 @cindex @code{popsection} directive
5030 @cindex Section Stack
5031 This is one of the ELF section stack manipulation directives. The others are
5032 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5033 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5036 This directive replaces the current section (and subsection) with the top
5037 section (and subsection) on the section stack. This section is popped off the
5042 @section @code{.print @var{string}}
5044 @cindex @code{print} directive
5045 @command{@value{AS}} will print @var{string} on the standard output during
5046 assembly. You must put @var{string} in double quotes.
5050 @section @code{.protected @var{names}}
5052 @cindex @code{protected} directive
5054 This is one of the ELF visibility directives. The other two are
5055 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5057 This directive overrides the named symbols default visibility (which is set by
5058 their binding: local, global or weak). The directive sets the visibility to
5059 @code{protected} which means that any references to the symbols from within the
5060 components that defines them must be resolved to the definition in that
5061 component, even if a definition in another component would normally preempt
5066 @section @code{.psize @var{lines} , @var{columns}}
5068 @cindex @code{psize} directive
5069 @cindex listing control: paper size
5070 @cindex paper size, for listings
5071 Use this directive to declare the number of lines---and, optionally, the
5072 number of columns---to use for each page, when generating listings.
5074 If you do not use @code{.psize}, listings use a default line-count
5075 of 60. You may omit the comma and @var{columns} specification; the
5076 default width is 200 columns.
5078 @command{@value{AS}} generates formfeeds whenever the specified number of
5079 lines is exceeded (or whenever you explicitly request one, using
5082 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5083 those explicitly specified with @code{.eject}.
5086 @section @code{.purgem @var{name}}
5088 @cindex @code{purgem} directive
5089 Undefine the macro @var{name}, so that later uses of the string will not be
5090 expanded. @xref{Macro}.
5094 @section @code{.pushsection @var{name} , @var{subsection}}
5096 @cindex @code{pushsection} directive
5097 @cindex Section Stack
5098 This is one of the ELF section stack manipulation directives. The others are
5099 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5100 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5103 This directive pushes the current section (and subsection) onto the
5104 top of the section stack, and then replaces the current section and
5105 subsection with @code{name} and @code{subsection}.
5109 @section @code{.quad @var{bignums}}
5111 @cindex @code{quad} directive
5112 @code{.quad} expects zero or more bignums, separated by commas. For
5113 each bignum, it emits
5115 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5116 warning message; and just takes the lowest order 8 bytes of the bignum.
5117 @cindex eight-byte integer
5118 @cindex integer, 8-byte
5120 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5121 hence @emph{quad}-word for 8 bytes.
5124 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5125 warning message; and just takes the lowest order 16 bytes of the bignum.
5126 @cindex sixteen-byte integer
5127 @cindex integer, 16-byte
5131 @section @code{.rept @var{count}}
5133 @cindex @code{rept} directive
5134 Repeat the sequence of lines between the @code{.rept} directive and the next
5135 @code{.endr} directive @var{count} times.
5137 For example, assembling
5145 is equivalent to assembling
5154 @section @code{.sbttl "@var{subheading}"}
5156 @cindex @code{sbttl} directive
5157 @cindex subtitles for listings
5158 @cindex listing control: subtitle
5159 Use @var{subheading} as the title (third line, immediately after the
5160 title line) when generating assembly listings.
5162 This directive affects subsequent pages, as well as the current page if
5163 it appears within ten lines of the top of a page.
5167 @section @code{.scl @var{class}}
5169 @cindex @code{scl} directive
5170 @cindex symbol storage class (COFF)
5171 @cindex COFF symbol storage class
5172 Set the storage-class value for a symbol. This directive may only be
5173 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5174 whether a symbol is static or external, or it may record further
5175 symbolic debugging information.
5178 The @samp{.scl} directive is primarily associated with COFF output; when
5179 configured to generate @code{b.out} output format, @command{@value{AS}}
5180 accepts this directive but ignores it.
5186 @section @code{.section @var{name}}
5188 @cindex named section
5189 Use the @code{.section} directive to assemble the following code into a section
5192 This directive is only supported for targets that actually support arbitrarily
5193 named sections; on @code{a.out} targets, for example, it is not accepted, even
5194 with a standard @code{a.out} section name.
5198 @c only print the extra heading if both COFF and ELF are set
5199 @subheading COFF Version
5202 @cindex @code{section} directive (COFF version)
5203 For COFF targets, the @code{.section} directive is used in one of the following
5207 .section @var{name}[, "@var{flags}"]
5208 .section @var{name}[, @var{subsegment}]
5211 If the optional argument is quoted, it is taken as flags to use for the
5212 section. Each flag is a single character. The following flags are recognized:
5215 bss section (uninitialized data)
5217 section is not loaded
5227 shared section (meaningful for PE targets)
5229 ignored. (For compatibility with the ELF version)
5232 If no flags are specified, the default flags depend upon the section name. If
5233 the section name is not recognized, the default will be for the section to be
5234 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5235 from the section, rather than adding them, so if they are used on their own it
5236 will be as if no flags had been specified at all.
5238 If the optional argument to the @code{.section} directive is not quoted, it is
5239 taken as a subsegment number (@pxref{Sub-Sections}).
5244 @c only print the extra heading if both COFF and ELF are set
5245 @subheading ELF Version
5248 @cindex Section Stack
5249 This is one of the ELF section stack manipulation directives. The others are
5250 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5251 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5252 @code{.previous} (@pxref{Previous}).
5254 @cindex @code{section} directive (ELF version)
5255 For ELF targets, the @code{.section} directive is used like this:
5258 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]
5261 The optional @var{flags} argument is a quoted string which may contain any
5262 combination of the following characters:
5265 section is allocatable
5269 section is executable
5271 section is mergeable
5273 section contains zero terminated strings
5275 section is a member of a section group
5277 section is used for thread-local-storage
5280 The optional @var{type} argument may contain one of the following constants:
5283 section contains data
5285 section does not contain data (i.e., section only occupies space)
5287 section contains data which is used by things other than the program
5290 Note on targets where the @code{@@} character is the start of a comment (eg
5291 ARM) then another character is used instead. For example the ARM port uses the
5294 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
5295 be specified as well as an extra argument - @var{entsize} - like this:
5298 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
5301 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
5302 constants, each @var{entsize} octets long. Sections with both @code{M} and
5303 @code{S} must contain zero terminated strings where each character is
5304 @var{entsize} bytes long. The linker may remove duplicates within sections with
5305 the same name, same entity size and same flags. @var{entsize} must be an
5306 absolute expression.
5308 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
5309 be present along with an additional field like this:
5312 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
5315 The @var{GroupName} field specifies the name of the section group to which this
5316 particular section belongs. The optional linkage field can contain:
5319 indicates that only one copy of this section should be retained
5324 Note - if both the @var{M} and @var{G} flags are present then the fields for
5325 the Merge flag should come first, like this:
5328 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
5331 If no flags are specified, the default flags depend upon the section name. If
5332 the section name is not recognized, the default will be for the section to have
5333 none of the above flags: it will not be allocated in memory, nor writable, nor
5334 executable. The section will contain data.
5336 For ELF targets, the assembler supports another type of @code{.section}
5337 directive for compatibility with the Solaris assembler:
5340 .section "@var{name}"[, @var{flags}...]
5343 Note that the section name is quoted. There may be a sequence of comma
5347 section is allocatable
5351 section is executable
5353 section is used for thread local storage
5356 This directive replaces the current section and subsection. See the
5357 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
5358 some examples of how this directive and the other section stack directives
5364 @section @code{.set @var{symbol}, @var{expression}}
5366 @cindex @code{set} directive
5367 @cindex symbol value, setting
5368 Set the value of @var{symbol} to @var{expression}. This
5369 changes @var{symbol}'s value and type to conform to
5370 @var{expression}. If @var{symbol} was flagged as external, it remains
5371 flagged (@pxref{Symbol Attributes}).
5373 You may @code{.set} a symbol many times in the same assembly.
5375 If you @code{.set} a global symbol, the value stored in the object
5376 file is the last value stored into it.
5379 The syntax for @code{set} on the HPPA is
5380 @samp{@var{symbol} .set @var{expression}}.
5384 @section @code{.short @var{expressions}}
5386 @cindex @code{short} directive
5388 @code{.short} is normally the same as @samp{.word}.
5389 @xref{Word,,@code{.word}}.
5391 In some configurations, however, @code{.short} and @code{.word} generate
5392 numbers of different lengths; @pxref{Machine Dependencies}.
5396 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5399 This expects zero or more @var{expressions}, and emits
5400 a 16 bit number for each.
5405 @section @code{.single @var{flonums}}
5407 @cindex @code{single} directive
5408 @cindex floating point numbers (single)
5409 This directive assembles zero or more flonums, separated by commas. It
5410 has the same effect as @code{.float}.
5412 The exact kind of floating point numbers emitted depends on how
5413 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5417 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5418 numbers in @sc{ieee} format.
5424 @section @code{.size}
5426 This directive is used to set the size associated with a symbol.
5430 @c only print the extra heading if both COFF and ELF are set
5431 @subheading COFF Version
5434 @cindex @code{size} directive (COFF version)
5435 For COFF targets, the @code{.size} directive is only permitted inside
5436 @code{.def}/@code{.endef} pairs. It is used like this:
5439 .size @var{expression}
5443 @samp{.size} is only meaningful when generating COFF format output; when
5444 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5451 @c only print the extra heading if both COFF and ELF are set
5452 @subheading ELF Version
5455 @cindex @code{size} directive (ELF version)
5456 For ELF targets, the @code{.size} directive is used like this:
5459 .size @var{name} , @var{expression}
5462 This directive sets the size associated with a symbol @var{name}.
5463 The size in bytes is computed from @var{expression} which can make use of label
5464 arithmetic. This directive is typically used to set the size of function
5470 @section @code{.sleb128 @var{expressions}}
5472 @cindex @code{sleb128} directive
5473 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5474 compact, variable length representation of numbers used by the DWARF
5475 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
5477 @ifclear no-space-dir
5479 @section @code{.skip @var{size} , @var{fill}}
5481 @cindex @code{skip} directive
5482 @cindex filling memory
5483 This directive emits @var{size} bytes, each of value @var{fill}. Both
5484 @var{size} and @var{fill} are absolute expressions. If the comma and
5485 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5489 @section @code{.space @var{size} , @var{fill}}
5491 @cindex @code{space} directive
5492 @cindex filling memory
5493 This directive emits @var{size} bytes, each of value @var{fill}. Both
5494 @var{size} and @var{fill} are absolute expressions. If the comma
5495 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5500 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5501 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5502 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5503 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5512 @section @code{.space}
5513 @cindex @code{space} directive
5515 On the AMD 29K, this directive is ignored; it is accepted for
5516 compatibility with other AMD 29K assemblers.
5519 @emph{Warning:} In most versions of the @sc{gnu} assembler, the directive
5520 @code{.space} has the effect of @code{.block} @xref{Machine Dependencies}.
5526 @section @code{.stabd, .stabn, .stabs}
5528 @cindex symbolic debuggers, information for
5529 @cindex @code{stab@var{x}} directives
5530 There are three directives that begin @samp{.stab}.
5531 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5532 The symbols are not entered in the @command{@value{AS}} hash table: they
5533 cannot be referenced elsewhere in the source file.
5534 Up to five fields are required:
5538 This is the symbol's name. It may contain any character except
5539 @samp{\000}, so is more general than ordinary symbol names. Some
5540 debuggers used to code arbitrarily complex structures into symbol names
5544 An absolute expression. The symbol's type is set to the low 8 bits of
5545 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5546 and debuggers choke on silly bit patterns.
5549 An absolute expression. The symbol's ``other'' attribute is set to the
5550 low 8 bits of this expression.
5553 An absolute expression. The symbol's descriptor is set to the low 16
5554 bits of this expression.
5557 An absolute expression which becomes the symbol's value.
5560 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5561 or @code{.stabs} statement, the symbol has probably already been created;
5562 you get a half-formed symbol in your object file. This is
5563 compatible with earlier assemblers!
5566 @cindex @code{stabd} directive
5567 @item .stabd @var{type} , @var{other} , @var{desc}
5569 The ``name'' of the symbol generated is not even an empty string.
5570 It is a null pointer, for compatibility. Older assemblers used a
5571 null pointer so they didn't waste space in object files with empty
5574 The symbol's value is set to the location counter,
5575 relocatably. When your program is linked, the value of this symbol
5576 is the address of the location counter when the @code{.stabd} was
5579 @cindex @code{stabn} directive
5580 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5581 The name of the symbol is set to the empty string @code{""}.
5583 @cindex @code{stabs} directive
5584 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5585 All five fields are specified.
5591 @section @code{.string} "@var{str}"
5593 @cindex string, copying to object file
5594 @cindex @code{string} directive
5596 Copy the characters in @var{str} to the object file. You may specify more than
5597 one string to copy, separated by commas. Unless otherwise specified for a
5598 particular machine, the assembler marks the end of each string with a 0 byte.
5599 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5602 @section @code{.struct @var{expression}}
5604 @cindex @code{struct} directive
5605 Switch to the absolute section, and set the section offset to @var{expression},
5606 which must be an absolute expression. You might use this as follows:
5615 This would define the symbol @code{field1} to have the value 0, the symbol
5616 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5617 value 8. Assembly would be left in the absolute section, and you would need to
5618 use a @code{.section} directive of some sort to change to some other section
5619 before further assembly.
5623 @section @code{.subsection @var{name}}
5625 @cindex @code{subsection} directive
5626 @cindex Section Stack
5627 This is one of the ELF section stack manipulation directives. The others are
5628 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5629 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5632 This directive replaces the current subsection with @code{name}. The current
5633 section is not changed. The replaced subsection is put onto the section stack
5634 in place of the then current top of stack subsection.
5639 @section @code{.symver}
5640 @cindex @code{symver} directive
5641 @cindex symbol versioning
5642 @cindex versions of symbols
5643 Use the @code{.symver} directive to bind symbols to specific version nodes
5644 within a source file. This is only supported on ELF platforms, and is
5645 typically used when assembling files to be linked into a shared library.
5646 There are cases where it may make sense to use this in objects to be bound
5647 into an application itself so as to override a versioned symbol from a
5650 For ELF targets, the @code{.symver} directive can be used like this:
5652 .symver @var{name}, @var{name2@@nodename}
5654 If the symbol @var{name} is defined within the file
5655 being assembled, the @code{.symver} directive effectively creates a symbol
5656 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5657 just don't try and create a regular alias is that the @var{@@} character isn't
5658 permitted in symbol names. The @var{name2} part of the name is the actual name
5659 of the symbol by which it will be externally referenced. The name @var{name}
5660 itself is merely a name of convenience that is used so that it is possible to
5661 have definitions for multiple versions of a function within a single source
5662 file, and so that the compiler can unambiguously know which version of a
5663 function is being mentioned. The @var{nodename} portion of the alias should be
5664 the name of a node specified in the version script supplied to the linker when
5665 building a shared library. If you are attempting to override a versioned
5666 symbol from a shared library, then @var{nodename} should correspond to the
5667 nodename of the symbol you are trying to override.
5669 If the symbol @var{name} is not defined within the file being assembled, all
5670 references to @var{name} will be changed to @var{name2@@nodename}. If no
5671 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5674 Another usage of the @code{.symver} directive is:
5676 .symver @var{name}, @var{name2@@@@nodename}
5678 In this case, the symbol @var{name} must exist and be defined within
5679 the file being assembled. It is similar to @var{name2@@nodename}. The
5680 difference is @var{name2@@@@nodename} will also be used to resolve
5681 references to @var{name2} by the linker.
5683 The third usage of the @code{.symver} directive is:
5685 .symver @var{name}, @var{name2@@@@@@nodename}
5687 When @var{name} is not defined within the
5688 file being assembled, it is treated as @var{name2@@nodename}. When
5689 @var{name} is defined within the file being assembled, the symbol
5690 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5695 @section @code{.tag @var{structname}}
5697 @cindex COFF structure debugging
5698 @cindex structure debugging, COFF
5699 @cindex @code{tag} directive
5700 This directive is generated by compilers to include auxiliary debugging
5701 information in the symbol table. It is only permitted inside
5702 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5703 definitions in the symbol table with instances of those structures.
5706 @samp{.tag} is only used when generating COFF format output; when
5707 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5713 @section @code{.text @var{subsection}}
5715 @cindex @code{text} directive
5716 Tells @command{@value{AS}} to assemble the following statements onto the end of
5717 the text subsection numbered @var{subsection}, which is an absolute
5718 expression. If @var{subsection} is omitted, subsection number zero
5722 @section @code{.title "@var{heading}"}
5724 @cindex @code{title} directive
5725 @cindex listing control: title line
5726 Use @var{heading} as the title (second line, immediately after the
5727 source file name and pagenumber) when generating assembly listings.
5729 This directive affects subsequent pages, as well as the current page if
5730 it appears within ten lines of the top of a page.
5734 @section @code{.type}
5736 This directive is used to set the type of a symbol.
5740 @c only print the extra heading if both COFF and ELF are set
5741 @subheading COFF Version
5744 @cindex COFF symbol type
5745 @cindex symbol type, COFF
5746 @cindex @code{type} directive (COFF version)
5747 For COFF targets, this directive is permitted only within
5748 @code{.def}/@code{.endef} pairs. It is used like this:
5754 This records the integer @var{int} as the type attribute of a symbol table
5758 @samp{.type} is associated only with COFF format output; when
5759 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
5760 directive but ignores it.
5766 @c only print the extra heading if both COFF and ELF are set
5767 @subheading ELF Version
5770 @cindex ELF symbol type
5771 @cindex symbol type, ELF
5772 @cindex @code{type} directive (ELF version)
5773 For ELF targets, the @code{.type} directive is used like this:
5776 .type @var{name} , @var{type description}
5779 This sets the type of symbol @var{name} to be either a
5780 function symbol or an object symbol. There are five different syntaxes
5781 supported for the @var{type description} field, in order to provide
5782 compatibility with various other assemblers. The syntaxes supported are:
5785 .type <name>,#function
5786 .type <name>,#object
5788 .type <name>,@@function
5789 .type <name>,@@object
5791 .type <name>,%function
5792 .type <name>,%object
5794 .type <name>,"function"
5795 .type <name>,"object"
5797 .type <name> STT_FUNCTION
5798 .type <name> STT_OBJECT
5804 @section @code{.uleb128 @var{expressions}}
5806 @cindex @code{uleb128} directive
5807 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
5808 compact, variable length representation of numbers used by the DWARF
5809 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
5813 @section @code{.val @var{addr}}
5815 @cindex @code{val} directive
5816 @cindex COFF value attribute
5817 @cindex value attribute, COFF
5818 This directive, permitted only within @code{.def}/@code{.endef} pairs,
5819 records the address @var{addr} as the value attribute of a symbol table
5823 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
5824 configured for @code{b.out}, it accepts this directive but ignores it.
5830 @section @code{.version "@var{string}"}
5832 @cindex @code{version} directive
5833 This directive creates a @code{.note} section and places into it an ELF
5834 formatted note of type NT_VERSION. The note's name is set to @code{string}.
5839 @section @code{.vtable_entry @var{table}, @var{offset}}
5841 @cindex @code{vtable_entry}
5842 This directive finds or creates a symbol @code{table} and creates a
5843 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
5846 @section @code{.vtable_inherit @var{child}, @var{parent}}
5848 @cindex @code{vtable_inherit}
5849 This directive finds the symbol @code{child} and finds or creates the symbol
5850 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
5851 parent whose addend is the value of the child symbol. As a special case the
5852 parent name of @code{0} is treated as refering the @code{*ABS*} section.
5856 @section @code{.weak @var{names}}
5858 @cindex @code{weak} directive
5859 This directive sets the weak attribute on the comma separated list of symbol
5860 @code{names}. If the symbols do not already exist, they will be created.
5862 Weak symbols are supported in COFF as a GNU extension. This directive
5863 sets the weak attribute on the comma separated list of symbol
5864 @code{names}. If the symbols do not already exist, they will be created.
5867 @code{.weak @var{name} [ < = | == > @var{alternate}] [, ...]}
5870 On the PE target, weak aliases are supported natively. Weak aliases
5871 (usually called "weak externals" in PE) are created when an alternate
5872 name is specified. When a weak symbol is linked and the symbol is not
5873 defined, the weak symbol becomes an alias for the alternate symbol. If
5874 one equal sign is used, the linker searches for defined symbols within
5875 other objects and libraries. This is the usual mode, historically
5876 called "lazy externals." Otherwise, when two equal signs are used,
5877 the linker searches for defined symbols only within other objects.
5879 Non-alias weak symbols are supported on PE as a GNU extension.
5882 @section @code{.word @var{expressions}}
5884 @cindex @code{word} directive
5885 This directive expects zero or more @var{expressions}, of any section,
5886 separated by commas.
5889 For each expression, @command{@value{AS}} emits a 32-bit number.
5892 For each expression, @command{@value{AS}} emits a 16-bit number.
5897 The size of the number emitted, and its byte order,
5898 depend on what target computer the assembly is for.
5901 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
5902 @c happen---32-bit addressability, period; no long/short jumps.
5903 @ifset DIFF-TBL-KLUGE
5904 @cindex difference tables altered
5905 @cindex altered difference tables
5907 @emph{Warning: Special Treatment to support Compilers}
5911 Machines with a 32-bit address space, but that do less than 32-bit
5912 addressing, require the following special treatment. If the machine of
5913 interest to you does 32-bit addressing (or doesn't require it;
5914 @pxref{Machine Dependencies}), you can ignore this issue.
5917 In order to assemble compiler output into something that works,
5918 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
5919 Directives of the form @samp{.word sym1-sym2} are often emitted by
5920 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
5921 directive of the form @samp{.word sym1-sym2}, and the difference between
5922 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
5923 creates a @dfn{secondary jump table}, immediately before the next label.
5924 This secondary jump table is preceded by a short-jump to the
5925 first byte after the secondary table. This short-jump prevents the flow
5926 of control from accidentally falling into the new table. Inside the
5927 table is a long-jump to @code{sym2}. The original @samp{.word}
5928 contains @code{sym1} minus the address of the long-jump to
5931 If there were several occurrences of @samp{.word sym1-sym2} before the
5932 secondary jump table, all of them are adjusted. If there was a
5933 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
5934 long-jump to @code{sym4} is included in the secondary jump table,
5935 and the @code{.word} directives are adjusted to contain @code{sym3}
5936 minus the address of the long-jump to @code{sym4}; and so on, for as many
5937 entries in the original jump table as necessary.
5940 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
5941 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
5942 assembly language programmers.
5945 @c end DIFF-TBL-KLUGE
5948 @section Deprecated Directives
5950 @cindex deprecated directives
5951 @cindex obsolescent directives
5952 One day these directives won't work.
5953 They are included for compatibility with older assemblers.
5960 @node Machine Dependencies
5961 @chapter Machine Dependent Features
5963 @cindex machine dependencies
5964 The machine instruction sets are (almost by definition) different on
5965 each machine where @command{@value{AS}} runs. Floating point representations
5966 vary as well, and @command{@value{AS}} often supports a few additional
5967 directives or command-line options for compatibility with other
5968 assemblers on a particular platform. Finally, some versions of
5969 @command{@value{AS}} support special pseudo-instructions for branch
5972 This chapter discusses most of these differences, though it does not
5973 include details on any machine's instruction set. For details on that
5974 subject, see the hardware manufacturer's manual.
5978 * AMD29K-Dependent:: AMD 29K Dependent Features
5981 * Alpha-Dependent:: Alpha Dependent Features
5984 * ARC-Dependent:: ARC Dependent Features
5987 * ARM-Dependent:: ARM Dependent Features
5990 * CRIS-Dependent:: CRIS Dependent Features
5993 * D10V-Dependent:: D10V Dependent Features
5996 * D30V-Dependent:: D30V Dependent Features
5999 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6002 * H8/500-Dependent:: Renesas H8/500 Dependent Features
6005 * HPPA-Dependent:: HPPA Dependent Features
6008 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6011 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6014 * i860-Dependent:: Intel 80860 Dependent Features
6017 * i960-Dependent:: Intel 80960 Dependent Features
6020 * IP2K-Dependent:: IP2K Dependent Features
6023 * M32R-Dependent:: M32R Dependent Features
6026 * M68K-Dependent:: M680x0 Dependent Features
6029 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6032 * M88K-Dependent:: M880x0 Dependent Features
6035 * MIPS-Dependent:: MIPS Dependent Features
6038 * MMIX-Dependent:: MMIX Dependent Features
6041 * MSP430-Dependent:: MSP430 Dependent Features
6044 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6045 * SH64-Dependent:: SuperH SH64 Dependent Features
6048 * PDP-11-Dependent:: PDP-11 Dependent Features
6051 * PJ-Dependent:: picoJava Dependent Features
6054 * PPC-Dependent:: PowerPC Dependent Features
6057 * Sparc-Dependent:: SPARC Dependent Features
6060 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
6063 * V850-Dependent:: V850 Dependent Features
6066 * Xtensa-Dependent:: Xtensa Dependent Features
6069 * Z8000-Dependent:: Z8000 Dependent Features
6072 * Vax-Dependent:: VAX Dependent Features
6079 @c The following major nodes are *sections* in the GENERIC version, *chapters*
6080 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
6081 @c peculiarity: to preserve cross-references, there must be a node called
6082 @c "Machine Dependencies". Hence the conditional nodenames in each
6083 @c major node below. Node defaulting in makeinfo requires adjacency of
6084 @c node and sectioning commands; hence the repetition of @chapter BLAH
6085 @c in both conditional blocks.
6088 @include c-a29k.texi
6092 @include c-alpha.texi
6104 @include c-cris.texi
6109 @node Machine Dependencies
6110 @chapter Machine Dependent Features
6112 The machine instruction sets are different on each Renesas chip family,
6113 and there are also some syntax differences among the families. This
6114 chapter describes the specific @command{@value{AS}} features for each
6118 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6119 * H8/500-Dependent:: Renesas H8/500 Dependent Features
6120 * SH-Dependent:: Renesas SH Dependent Features
6127 @include c-d10v.texi
6131 @include c-d30v.texi
6135 @include c-h8300.texi
6139 @include c-h8500.texi
6143 @include c-hppa.texi
6147 @include c-i370.texi
6151 @include c-i386.texi
6155 @include c-i860.texi
6159 @include c-i960.texi
6163 @include c-ia64.texi
6167 @include c-ip2k.texi
6171 @include c-m32r.texi
6175 @include c-m68k.texi
6179 @include c-m68hc11.texi
6183 @include c-m88k.texi
6187 @include c-mips.texi
6191 @include c-mmix.texi
6195 @include c-msp430.texi
6199 @include c-ns32k.texi
6203 @include c-pdp11.texi
6216 @include c-sh64.texi
6220 @include c-sparc.texi
6224 @include c-tic54x.texi
6236 @include c-v850.texi
6240 @include c-xtensa.texi
6244 @c reverse effect of @down at top of generic Machine-Dep chapter
6248 @node Reporting Bugs
6249 @chapter Reporting Bugs
6250 @cindex bugs in assembler
6251 @cindex reporting bugs in assembler
6253 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6255 Reporting a bug may help you by bringing a solution to your problem, or it may
6256 not. But in any case the principal function of a bug report is to help the
6257 entire community by making the next version of @command{@value{AS}} work better.
6258 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6260 In order for a bug report to serve its purpose, you must include the
6261 information that enables us to fix the bug.
6264 * Bug Criteria:: Have you found a bug?
6265 * Bug Reporting:: How to report bugs
6269 @section Have You Found a Bug?
6270 @cindex bug criteria
6272 If you are not sure whether you have found a bug, here are some guidelines:
6275 @cindex fatal signal
6276 @cindex assembler crash
6277 @cindex crash of assembler
6279 If the assembler gets a fatal signal, for any input whatever, that is a
6280 @command{@value{AS}} bug. Reliable assemblers never crash.
6282 @cindex error on valid input
6284 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6286 @cindex invalid input
6288 If @command{@value{AS}} does not produce an error message for invalid input, that
6289 is a bug. However, you should note that your idea of ``invalid input'' might
6290 be our idea of ``an extension'' or ``support for traditional practice''.
6293 If you are an experienced user of assemblers, your suggestions for improvement
6294 of @command{@value{AS}} are welcome in any case.
6298 @section How to Report Bugs
6300 @cindex assembler bugs, reporting
6302 A number of companies and individuals offer support for @sc{gnu} products. If
6303 you obtained @command{@value{AS}} from a support organization, we recommend you
6304 contact that organization first.
6306 You can find contact information for many support companies and
6307 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6310 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6311 to @samp{bug-binutils@@gnu.org}.
6313 The fundamental principle of reporting bugs usefully is this:
6314 @strong{report all the facts}. If you are not sure whether to state a
6315 fact or leave it out, state it!
6317 Often people omit facts because they think they know what causes the problem
6318 and assume that some details do not matter. Thus, you might assume that the
6319 name of a symbol you use in an example does not matter. Well, probably it does
6320 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6321 happens to fetch from the location where that name is stored in memory;
6322 perhaps, if the name were different, the contents of that location would fool
6323 the assembler into doing the right thing despite the bug. Play it safe and
6324 give a specific, complete example. That is the easiest thing for you to do,
6325 and the most helpful.
6327 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6328 it is new to us. Therefore, always write your bug reports on the assumption
6329 that the bug has not been reported previously.
6331 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6332 bell?'' This cannot help us fix a bug, so it is basically useless. We
6333 respond by asking for enough details to enable us to investigate.
6334 You might as well expedite matters by sending them to begin with.
6336 To enable us to fix the bug, you should include all these things:
6340 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6341 it with the @samp{--version} argument.
6343 Without this, we will not know whether there is any point in looking for
6344 the bug in the current version of @command{@value{AS}}.
6347 Any patches you may have applied to the @command{@value{AS}} source.
6350 The type of machine you are using, and the operating system name and
6354 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6358 The command arguments you gave the assembler to assemble your example and
6359 observe the bug. To guarantee you will not omit something important, list them
6360 all. A copy of the Makefile (or the output from make) is sufficient.
6362 If we were to try to guess the arguments, we would probably guess wrong
6363 and then we might not encounter the bug.
6366 A complete input file that will reproduce the bug. If the bug is observed when
6367 the assembler is invoked via a compiler, send the assembler source, not the
6368 high level language source. Most compilers will produce the assembler source
6369 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6370 the options @samp{-v --save-temps}; this will save the assembler source in a
6371 file with an extension of @file{.s}, and also show you exactly how
6372 @command{@value{AS}} is being run.
6375 A description of what behavior you observe that you believe is
6376 incorrect. For example, ``It gets a fatal signal.''
6378 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6379 will certainly notice it. But if the bug is incorrect output, we might not
6380 notice unless it is glaringly wrong. You might as well not give us a chance to
6383 Even if the problem you experience is a fatal signal, you should still say so
6384 explicitly. Suppose something strange is going on, such as, your copy of
6385 @command{@value{AS}} is out of synch, or you have encountered a bug in the C
6386 library on your system. (This has happened!) Your copy might crash and ours
6387 would not. If you told us to expect a crash, then when ours fails to crash, we
6388 would know that the bug was not happening for us. If you had not told us to
6389 expect a crash, then we would not be able to draw any conclusion from our
6393 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6394 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6395 option. Always send diffs from the old file to the new file. If you even
6396 discuss something in the @command{@value{AS}} source, refer to it by context, not
6399 The line numbers in our development sources will not match those in your
6400 sources. Your line numbers would convey no useful information to us.
6403 Here are some things that are not necessary:
6407 A description of the envelope of the bug.
6409 Often people who encounter a bug spend a lot of time investigating
6410 which changes to the input file will make the bug go away and which
6411 changes will not affect it.
6413 This is often time consuming and not very useful, because the way we
6414 will find the bug is by running a single example under the debugger
6415 with breakpoints, not by pure deduction from a series of examples.
6416 We recommend that you save your time for something else.
6418 Of course, if you can find a simpler example to report @emph{instead}
6419 of the original one, that is a convenience for us. Errors in the
6420 output will be easier to spot, running under the debugger will take
6421 less time, and so on.
6423 However, simplification is not vital; if you do not want to do this,
6424 report the bug anyway and send us the entire test case you used.
6427 A patch for the bug.
6429 A patch for the bug does help us if it is a good one. But do not omit
6430 the necessary information, such as the test case, on the assumption that
6431 a patch is all we need. We might see problems with your patch and decide
6432 to fix the problem another way, or we might not understand it at all.
6434 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6435 construct an example that will make the program follow a certain path through
6436 the code. If you do not send us the example, we will not be able to construct
6437 one, so we will not be able to verify that the bug is fixed.
6439 And if we cannot understand what bug you are trying to fix, or why your
6440 patch should be an improvement, we will not install it. A test case will
6441 help us to understand.
6444 A guess about what the bug is or what it depends on.
6446 Such guesses are usually wrong. Even we cannot guess right about such
6447 things without first using the debugger to find the facts.
6450 @node Acknowledgements
6451 @chapter Acknowledgements
6453 If you have contributed to @command{@value{AS}} and your name isn't listed here,
6454 it is not meant as a slight. We just don't know about it. Send mail to the
6455 maintainer, and we'll correct the situation. Currently
6457 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6459 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6462 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6463 information and the 68k series machines, most of the preprocessing pass, and
6464 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6466 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6467 many bug fixes, including merging support for several processors, breaking GAS
6468 up to handle multiple object file format back ends (including heavy rewrite,
6469 testing, an integration of the coff and b.out back ends), adding configuration
6470 including heavy testing and verification of cross assemblers and file splits
6471 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6472 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6473 port (including considerable amounts of reverse engineering), a SPARC opcode
6474 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6475 assertions and made them work, much other reorganization, cleanup, and lint.
6477 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6478 in format-specific I/O modules.
6480 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6481 has done much work with it since.
6483 The Intel 80386 machine description was written by Eliot Dresselhaus.
6485 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6487 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6488 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6490 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6491 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6492 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6493 support a.out format.
6495 Support for the Zilog Z8k and Renesas H8/300 and H8/500 processors (tc-z8k,
6496 tc-h8300, tc-h8500), and IEEE 695 object file format (obj-ieee), was written by
6497 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6498 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6501 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6502 simplified the configuration of which versions accept which directives. He
6503 updated the 68k machine description so that Motorola's opcodes always produced
6504 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6505 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6506 cross-compilation support, and one bug in relaxation that took a week and
6507 required the proverbial one-bit fix.
6509 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6510 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6511 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6512 PowerPC assembler, and made a few other minor patches.
6514 Steve Chamberlain made @command{@value{AS}} able to generate listings.
6516 Hewlett-Packard contributed support for the HP9000/300.
6518 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6519 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6520 formats). This work was supported by both the Center for Software Science at
6521 the University of Utah and Cygnus Support.
6523 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6524 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6525 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6526 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6527 and some initial 64-bit support).
6529 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6531 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6532 support for openVMS/Alpha.
6534 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6537 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6538 Inc. added support for Xtensa processors.
6540 Several engineers at Cygnus Support have also provided many small bug fixes and
6541 configuration enhancements.
6543 Many others have contributed large or small bugfixes and enhancements. If
6544 you have contributed significant work and are not mentioned on this list, and
6545 want to be, let us know. Some of the history has been lost; we are not
6546 intentionally leaving anyone out.