1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
3 @c 2001, 2002, 2003, 2004, 2005, 2006
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
52 @set abnormal-separator
56 @settitle Using @value{AS}
59 @settitle Using @value{AS} (@value{TARGET})
61 @setchapternewpage odd
66 @c WARE! Some of the machine-dependent sections contain tables of machine
67 @c instructions. Except in multi-column format, these tables look silly.
68 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
69 @c the multi-col format is faked within @example sections.
71 @c Again unfortunately, the natural size that fits on a page, for these tables,
72 @c is different depending on whether or not smallbook is turned on.
73 @c This matters, because of order: text flow switches columns at each page
76 @c The format faked in this source works reasonably well for smallbook,
77 @c not well for the default large-page format. This manual expects that if you
78 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
79 @c tables in question. You can turn on one without the other at your
80 @c discretion, of course.
83 @c the insn tables look just as silly in info files regardless of smallbook,
84 @c might as well show 'em anyways.
90 * As: (as). The GNU assembler.
91 * Gas: (as). The GNU assembler.
100 This file documents the GNU Assembler "@value{AS}".
102 @c man begin COPYRIGHT
103 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002, 2006 Free Software Foundation, Inc.
105 Permission is granted to copy, distribute and/or modify this document
106 under the terms of the GNU Free Documentation License, Version 1.1
107 or any later version published by the Free Software Foundation;
108 with no Invariant Sections, with no Front-Cover Texts, and with no
109 Back-Cover Texts. A copy of the license is included in the
110 section entitled ``GNU Free Documentation License''.
115 Permission is granted to process this file through Tex and print the
116 results, provided the printed document carries copying permission
117 notice identical to this one except for the removal of this paragraph
118 (this paragraph not being relevant to the printed manual).
124 @title Using @value{AS}
125 @subtitle The @sc{gnu} Assembler
127 @subtitle for the @value{TARGET} family
130 @subtitle Version @value{VERSION}
133 The Free Software Foundation Inc. thanks The Nice Computer
134 Company of Australia for loaning Dean Elsner to write the
135 first (Vax) version of @command{as} for Project @sc{gnu}.
136 The proprietors, management and staff of TNCCA thank FSF for
137 distracting the boss while they got some work
140 @author Dean Elsner, Jay Fenlason & friends
144 \hfill {\it Using {\tt @value{AS}}}\par
145 \hfill Edited by Cygnus Support\par
147 %"boxit" macro for figures:
148 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
149 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
150 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
151 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
152 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
155 @vskip 0pt plus 1filll
156 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002, 2006 Free Software Foundation, Inc.
158 Permission is granted to copy, distribute and/or modify this document
159 under the terms of the GNU Free Documentation License, Version 1.1
160 or any later version published by the Free Software Foundation;
161 with no Invariant Sections, with no Front-Cover Texts, and with no
162 Back-Cover Texts. A copy of the license is included in the
163 section entitled ``GNU Free Documentation License''.
169 @top Using @value{AS}
171 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}} version
174 This version of the file describes @command{@value{AS}} configured to generate
175 code for @value{TARGET} architectures.
178 This document is distributed under the terms of the GNU Free
179 Documentation License. A copy of the license is included in the
180 section entitled ``GNU Free Documentation License''.
183 * Overview:: Overview
184 * Invoking:: Command-Line Options
186 * Sections:: Sections and Relocation
188 * Expressions:: Expressions
189 * Pseudo Ops:: Assembler Directives
190 * Machine Dependencies:: Machine Dependent Features
191 * Reporting Bugs:: Reporting Bugs
192 * Acknowledgements:: Who Did What
193 * GNU Free Documentation License:: GNU Free Documentation License
194 * AS Index:: AS Index
201 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
203 This version of the manual describes @command{@value{AS}} configured to generate
204 code for @value{TARGET} architectures.
208 @cindex invocation summary
209 @cindex option summary
210 @cindex summary of options
211 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
212 @pxref{Invoking,,Command-Line Options}.
214 @c man title AS the portable GNU assembler.
218 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
222 @c We don't use deffn and friends for the following because they seem
223 @c to be limited to one line for the header.
225 @c man begin SYNOPSIS
226 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
227 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
228 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
229 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
230 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
231 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
232 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
233 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
234 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
235 [@b{--target-help}] [@var{target-options}]
236 [@b{--}|@var{files} @dots{}]
238 @c Target dependent options are listed below. Keep the list sorted.
239 @c Add an empty line for separation.
242 @emph{Target Alpha options:}
244 [@b{-mdebug} | @b{-no-mdebug}]
245 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
246 [@b{-F}] [@b{-32addr}]
250 @emph{Target ARC options:}
256 @emph{Target ARM options:}
257 @c Don't document the deprecated options
258 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
259 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
260 [@b{-mfpu}=@var{floating-point-format}]
261 [@b{-mfloat-abi}=@var{abi}]
262 [@b{-meabi}=@var{ver}]
265 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
266 @b{-mapcs-reentrant}]
267 [@b{-mthumb-interwork}] [@b{-k}]
271 @emph{Target CRIS options:}
272 [@b{--underscore} | @b{--no-underscore}]
274 [@b{--emulation=criself} | @b{--emulation=crisaout}]
275 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
276 @c Deprecated -- deliberately not documented.
281 @emph{Target D10V options:}
286 @emph{Target D30V options:}
287 [@b{-O}|@b{-n}|@b{-N}]
290 @c Renesas family chips have no machine-dependent assembler options
293 @c HPPA has no machine-dependent assembler options (yet).
297 @emph{Target i386 options:}
298 [@b{--32}|@b{--64}] [@b{-n}]
302 @emph{Target i960 options:}
303 @c see md_parse_option in tc-i960.c
304 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
306 [@b{-b}] [@b{-no-relax}]
310 @emph{Target IA-64 options:}
311 [@b{-mconstant-gp}|@b{-mauto-pic}]
312 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
314 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
315 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
316 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
317 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
321 @emph{Target IP2K options:}
322 [@b{-mip2022}|@b{-mip2022ext}]
326 @emph{Target M32C options:}
327 [@b{-m32c}|@b{-m16c}]
331 @emph{Target M32R options:}
332 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
337 @emph{Target M680X0 options:}
338 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
342 @emph{Target M68HC11 options:}
343 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
344 [@b{-mshort}|@b{-mlong}]
345 [@b{-mshort-double}|@b{-mlong-double}]
346 [@b{--force-long-branchs}] [@b{--short-branchs}]
347 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
348 [@b{--print-opcodes}] [@b{--generate-example}]
352 @emph{Target MCORE options:}
353 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
354 [@b{-mcpu=[210|340]}]
358 @emph{Target MIPS options:}
359 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
360 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
361 [@b{-non_shared}] [@b{-xgot}]
362 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
363 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
364 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
365 [@b{-mips64}] [@b{-mips64r2}]
366 [@b{-construct-floats}] [@b{-no-construct-floats}]
367 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
368 [@b{-mfix7000}] [@b{-mno-fix7000}]
369 [@b{-mips16}] [@b{-no-mips16}]
370 [@b{-msmartmips}] [@b{-mno-smartmips}]
371 [@b{-mips3d}] [@b{-no-mips3d}]
372 [@b{-mdmx}] [@b{-no-mdmx}]
373 [@b{-mdsp}] [@b{-mno-dsp}]
374 [@b{-mmt}] [@b{-mno-mt}]
375 [@b{-mdebug}] [@b{-no-mdebug}]
376 [@b{-mpdr}] [@b{-mno-pdr}]
380 @emph{Target MMIX options:}
381 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
382 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
383 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
384 [@b{--linker-allocated-gregs}]
388 @emph{Target PDP11 options:}
389 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
390 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
391 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
395 @emph{Target picoJava options:}
400 @emph{Target PowerPC options:}
401 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
402 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
403 @b{-mbooke32}|@b{-mbooke64}]
404 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
405 [@b{-mregnames}|@b{-mno-regnames}]
406 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
407 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
408 [@b{-msolaris}|@b{-mno-solaris}]
412 @emph{Target SPARC options:}
413 @c The order here is important. See c-sparc.texi.
414 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
415 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
416 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
421 @emph{Target TIC54X options:}
422 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
423 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
428 @emph{Target Z80 options:}
429 [@b{-z80}] [@b{-r800}]
430 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
431 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
432 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
433 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
434 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
435 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
439 @c Z8000 has no machine-dependent assembler options
443 @emph{Target Xtensa options:}
444 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
445 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
446 [@b{--[no-]transform}]
447 [@b{--rename-section} @var{oldname}=@var{newname}]
455 @include at-file.texi
458 Turn on listings, in any of a variety of ways:
462 omit false conditionals
465 omit debugging directives
468 include high-level source
474 include macro expansions
477 omit forms processing
483 set the name of the listing file
486 You may combine these options; for example, use @samp{-aln} for assembly
487 listing without forms processing. The @samp{=file} option, if used, must be
488 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
491 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
494 Ignored. This option is accepted for script compatibility with calls to
497 @item --defsym @var{sym}=@var{value}
498 Define the symbol @var{sym} to be @var{value} before assembling the input file.
499 @var{value} must be an integer constant. As in C, a leading @samp{0x}
500 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
503 ``fast''---skip whitespace and comment preprocessing (assume source is
508 Generate debugging information for each assembler source line using whichever
509 debug format is preferred by the target. This currently means either STABS,
513 Generate stabs debugging information for each assembler line. This
514 may help debugging assembler code, if the debugger can handle it.
517 Generate stabs debugging information for each assembler line, with GNU
518 extensions that probably only gdb can handle, and that could make other
519 debuggers crash or refuse to read your program. This
520 may help debugging assembler code. Currently the only GNU extension is
521 the location of the current working directory at assembling time.
524 Generate DWARF2 debugging information for each assembler line. This
525 may help debugging assembler code, if the debugger can handle it. Note---this
526 option is only supported by some targets, not all of them.
529 Print a summary of the command line options and exit.
532 Print a summary of all target specific options and exit.
535 Add directory @var{dir} to the search list for @code{.include} directives.
538 Don't warn about signed overflow.
541 @ifclear DIFF-TBL-KLUGE
542 This option is accepted but has no effect on the @value{TARGET} family.
544 @ifset DIFF-TBL-KLUGE
545 Issue warnings when difference tables altered for long displacements.
550 Keep (in the symbol table) local symbols. On traditional a.out systems
551 these start with @samp{L}, but different systems have different local
554 @item --listing-lhs-width=@var{number}
555 Set the maximum width, in words, of the output data column for an assembler
556 listing to @var{number}.
558 @item --listing-lhs-width2=@var{number}
559 Set the maximum width, in words, of the output data column for continuation
560 lines in an assembler listing to @var{number}.
562 @item --listing-rhs-width=@var{number}
563 Set the maximum width of an input source line, as displayed in a listing, to
566 @item --listing-cont-lines=@var{number}
567 Set the maximum number of lines printed in a listing for a single line of input
570 @item -o @var{objfile}
571 Name the object-file output from @command{@value{AS}} @var{objfile}.
574 Fold the data section into the text section.
576 @kindex --hash-size=@var{number}
577 Set the default size of GAS's hash tables to a prime number close to
578 @var{number}. Increasing this value can reduce the length of time it takes the
579 assembler to perform its tasks, at the expense of increasing the assembler's
580 memory requirements. Similarly reducing this value can reduce the memory
581 requirements at the expense of speed.
583 @item --reduce-memory-overheads
584 This option reduces GAS's memory requirements, at the expense of making the
585 assembly processes slower. Currently this switch is a synonym for
586 @samp{--hash-size=4051}, but in the future it may have other effects as well.
589 Print the maximum space (in bytes) and total time (in seconds) used by
592 @item --strip-local-absolute
593 Remove local absolute symbols from the outgoing symbol table.
597 Print the @command{as} version.
600 Print the @command{as} version and exit.
604 Suppress warning messages.
606 @item --fatal-warnings
607 Treat warnings as errors.
610 Don't suppress warning messages or treat them as errors.
619 Generate an object file even after errors.
621 @item -- | @var{files} @dots{}
622 Standard input, or source files to assemble.
627 The following options are available when @value{AS} is configured for
632 This option selects the core processor variant.
634 Select either big-endian (-EB) or little-endian (-EL) output.
639 The following options are available when @value{AS} is configured for the ARM
643 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
644 Specify which ARM processor variant is the target.
645 @item -march=@var{architecture}[+@var{extension}@dots{}]
646 Specify which ARM architecture variant is used by the target.
647 @item -mfpu=@var{floating-point-format}
648 Select which Floating Point architecture is the target.
649 @item -mfloat-abi=@var{abi}
650 Select which floating point ABI is in use.
652 Enable Thumb only instruction decoding.
653 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
654 Select which procedure calling convention is in use.
656 Select either big-endian (-EB) or little-endian (-EL) output.
657 @item -mthumb-interwork
658 Specify that the code has been generated with interworking between Thumb and
661 Specify that PIC code has been generated.
666 See the info pages for documentation of the CRIS-specific options.
670 The following options are available when @value{AS} is configured for
673 @cindex D10V optimization
674 @cindex optimization, D10V
676 Optimize output by parallelizing instructions.
681 The following options are available when @value{AS} is configured for a D30V
684 @cindex D30V optimization
685 @cindex optimization, D30V
687 Optimize output by parallelizing instructions.
691 Warn when nops are generated.
693 @cindex D30V nops after 32-bit multiply
695 Warn when a nop after a 32-bit multiply instruction is generated.
700 The following options are available when @value{AS} is configured for the
701 Intel 80960 processor.
704 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
705 Specify which variant of the 960 architecture is the target.
708 Add code to collect statistics about branches taken.
711 Do not alter compare-and-branch instructions for long displacements;
718 The following options are available when @value{AS} is configured for the
724 Specifies that the extended IP2022 instructions are allowed.
727 Restores the default behaviour, which restricts the permitted instructions to
728 just the basic IP2022 ones.
734 The following options are available when @value{AS} is configured for the
735 Renesas M32C and M16C processors.
740 Assemble M32C instructions.
743 Assemble M16C instructions (the default).
749 The following options are available when @value{AS} is configured for the
750 Renesas M32R (formerly Mitsubishi M32R) series.
755 Specify which processor in the M32R family is the target. The default
756 is normally the M32R, but this option changes it to the M32RX.
758 @item --warn-explicit-parallel-conflicts or --Wp
759 Produce warning messages when questionable parallel constructs are
762 @item --no-warn-explicit-parallel-conflicts or --Wnp
763 Do not produce warning messages when questionable parallel constructs are
770 The following options are available when @value{AS} is configured for the
771 Motorola 68000 series.
776 Shorten references to undefined symbols, to one word instead of two.
778 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
779 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
780 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
781 Specify what processor in the 68000 family is the target. The default
782 is normally the 68020, but this can be changed at configuration time.
784 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
785 The target machine does (or does not) have a floating-point coprocessor.
786 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
787 the basic 68000 is not compatible with the 68881, a combination of the
788 two can be specified, since it's possible to do emulation of the
789 coprocessor instructions with the main processor.
791 @item -m68851 | -mno-68851
792 The target machine does (or does not) have a memory-management
793 unit coprocessor. The default is to assume an MMU for 68020 and up.
800 For details about the PDP-11 machine dependent features options,
801 see @ref{PDP-11-Options}.
804 @item -mpic | -mno-pic
805 Generate position-independent (or position-dependent) code. The
806 default is @option{-mpic}.
809 @itemx -mall-extensions
810 Enable all instruction set extensions. This is the default.
812 @item -mno-extensions
813 Disable all instruction set extensions.
815 @item -m@var{extension} | -mno-@var{extension}
816 Enable (or disable) a particular instruction set extension.
819 Enable the instruction set extensions supported by a particular CPU, and
820 disable all other extensions.
822 @item -m@var{machine}
823 Enable the instruction set extensions supported by a particular machine
824 model, and disable all other extensions.
830 The following options are available when @value{AS} is configured for
831 a picoJava processor.
835 @cindex PJ endianness
836 @cindex endianness, PJ
837 @cindex big endian output, PJ
839 Generate ``big endian'' format output.
841 @cindex little endian output, PJ
843 Generate ``little endian'' format output.
849 The following options are available when @value{AS} is configured for the
850 Motorola 68HC11 or 68HC12 series.
854 @item -m68hc11 | -m68hc12 | -m68hcs12
855 Specify what processor is the target. The default is
856 defined by the configuration option when building the assembler.
859 Specify to use the 16-bit integer ABI.
862 Specify to use the 32-bit integer ABI.
865 Specify to use the 32-bit double ABI.
868 Specify to use the 64-bit double ABI.
870 @item --force-long-branchs
871 Relative branches are turned into absolute ones. This concerns
872 conditional branches, unconditional branches and branches to a
875 @item -S | --short-branchs
876 Do not turn relative branchs into absolute ones
877 when the offset is out of range.
879 @item --strict-direct-mode
880 Do not turn the direct addressing mode into extended addressing mode
881 when the instruction does not support direct addressing mode.
883 @item --print-insn-syntax
884 Print the syntax of instruction in case of error.
886 @item --print-opcodes
887 print the list of instructions with syntax and then exit.
889 @item --generate-example
890 print an example of instruction for each possible instruction and then exit.
891 This option is only useful for testing @command{@value{AS}}.
897 The following options are available when @command{@value{AS}} is configured
898 for the SPARC architecture:
901 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
902 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
903 Explicitly select a variant of the SPARC architecture.
905 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
906 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
908 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
909 UltraSPARC extensions.
911 @item -xarch=v8plus | -xarch=v8plusa
912 For compatibility with the Solaris v9 assembler. These options are
913 equivalent to -Av8plus and -Av8plusa, respectively.
916 Warn when the assembler switches to another architecture.
921 The following options are available when @value{AS} is configured for the 'c54x
926 Enable extended addressing mode. All addresses and relocations will assume
927 extended addressing (usually 23 bits).
928 @item -mcpu=@var{CPU_VERSION}
929 Sets the CPU version being compiled for.
930 @item -merrors-to-file @var{FILENAME}
931 Redirect error output to a file, for broken systems which don't support such
932 behaviour in the shell.
937 The following options are available when @value{AS} is configured for
938 a @sc{mips} processor.
942 This option sets the largest size of an object that can be referenced
943 implicitly with the @code{gp} register. It is only accepted for targets that
944 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
946 @cindex MIPS endianness
947 @cindex endianness, MIPS
948 @cindex big endian output, MIPS
950 Generate ``big endian'' format output.
952 @cindex little endian output, MIPS
954 Generate ``little endian'' format output.
966 Generate code for a particular @sc{mips} Instruction Set Architecture level.
967 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
968 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
969 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
970 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
972 correspond to generic
973 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
974 and @samp{MIPS64 Release 2}
975 ISA processors, respectively.
977 @item -march=@var{CPU}
978 Generate code for a particular @sc{mips} cpu.
980 @item -mtune=@var{cpu}
981 Schedule and tune for a particular @sc{mips} cpu.
985 Cause nops to be inserted if the read of the destination register
986 of an mfhi or mflo instruction occurs in the following two instructions.
990 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
991 section instead of the standard ELF .stabs sections.
995 Control generation of @code{.pdr} sections.
999 The register sizes are normally inferred from the ISA and ABI, but these
1000 flags force a certain group of registers to be treated as 32 bits wide at
1001 all times. @samp{-mgp32} controls the size of general-purpose registers
1002 and @samp{-mfp32} controls the size of floating-point registers.
1006 Generate code for the MIPS 16 processor. This is equivalent to putting
1007 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1008 turns off this option.
1011 @itemx -mno-smartmips
1012 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1013 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1014 @samp{-mno-smartmips} turns off this option.
1018 Generate code for the MIPS-3D Application Specific Extension.
1019 This tells the assembler to accept MIPS-3D instructions.
1020 @samp{-no-mips3d} turns off this option.
1024 Generate code for the MDMX Application Specific Extension.
1025 This tells the assembler to accept MDMX instructions.
1026 @samp{-no-mdmx} turns off this option.
1030 Generate code for the DSP Application Specific Extension.
1031 This tells the assembler to accept DSP instructions.
1032 @samp{-mno-dsp} turns off this option.
1036 Generate code for the MT Application Specific Extension.
1037 This tells the assembler to accept MT instructions.
1038 @samp{-mno-mt} turns off this option.
1040 @item --construct-floats
1041 @itemx --no-construct-floats
1042 The @samp{--no-construct-floats} option disables the construction of
1043 double width floating point constants by loading the two halves of the
1044 value into the two single width floating point registers that make up
1045 the double width register. By default @samp{--construct-floats} is
1046 selected, allowing construction of these floating point constants.
1049 @item --emulation=@var{name}
1050 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1051 for some other target, in all respects, including output format (choosing
1052 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1053 debugging information or store symbol table information, and default
1054 endianness. The available configuration names are: @samp{mipsecoff},
1055 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1056 @samp{mipsbelf}. The first two do not alter the default endianness from that
1057 of the primary target for which the assembler was configured; the others change
1058 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1059 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1060 selection in any case.
1062 This option is currently supported only when the primary target
1063 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1064 Furthermore, the primary target or others specified with
1065 @samp{--enable-targets=@dots{}} at configuration time must include support for
1066 the other format, if both are to be available. For example, the Irix 5
1067 configuration includes support for both.
1069 Eventually, this option will support more configurations, with more
1070 fine-grained control over the assembler's behavior, and will be supported for
1074 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1081 Control how to deal with multiplication overflow and division by zero.
1082 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1083 (and only work for Instruction Set Architecture level 2 and higher);
1084 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1088 When this option is used, @command{@value{AS}} will issue a warning every
1089 time it generates a nop instruction from a macro.
1094 The following options are available when @value{AS} is configured for
1100 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1101 The command line option @samp{-nojsri2bsr} can be used to disable it.
1105 Enable or disable the silicon filter behaviour. By default this is disabled.
1106 The default can be overridden by the @samp{-sifilter} command line option.
1109 Alter jump instructions for long displacements.
1111 @item -mcpu=[210|340]
1112 Select the cpu type on the target hardware. This controls which instructions
1116 Assemble for a big endian target.
1119 Assemble for a little endian target.
1125 See the info pages for documentation of the MMIX-specific options.
1129 The following options are available when @value{AS} is configured for
1130 an Xtensa processor.
1133 @item --text-section-literals | --no-text-section-literals
1134 With @option{--text-@-section-@-literals}, literal pools are interspersed
1135 in the text section. The default is
1136 @option{--no-@-text-@-section-@-literals}, which places literals in a
1137 separate section in the output file. These options only affect literals
1138 referenced via PC-relative @code{L32R} instructions; literals for
1139 absolute mode @code{L32R} instructions are handled separately.
1141 @item --absolute-literals | --no-absolute-literals
1142 Indicate to the assembler whether @code{L32R} instructions use absolute
1143 or PC-relative addressing. The default is to assume absolute addressing
1144 if the Xtensa processor includes the absolute @code{L32R} addressing
1145 option. Otherwise, only the PC-relative @code{L32R} mode can be used.
1147 @item --target-align | --no-target-align
1148 Enable or disable automatic alignment to reduce branch penalties at the
1149 expense of some code density. The default is @option{--target-@-align}.
1151 @item --longcalls | --no-longcalls
1152 Enable or disable transformation of call instructions to allow calls
1153 across a greater range of addresses. The default is
1154 @option{--no-@-longcalls}.
1156 @item --transform | --no-transform
1157 Enable or disable all assembler transformations of Xtensa instructions.
1158 The default is @option{--transform};
1159 @option{--no-transform} should be used only in the rare cases when the
1160 instructions must be exactly as specified in the assembly source.
1165 The following options are available when @value{AS} is configured for
1166 a Z80 family processor.
1169 Assemble for Z80 processor.
1171 Assemble for R800 processor.
1172 @item -ignore-undocumented-instructions
1174 Assemble undocumented Z80 instructions that also work on R800 without warning.
1175 @item -ignore-unportable-instructions
1177 Assemble all undocumented Z80 instructions without warning.
1178 @item -warn-undocumented-instructions
1180 Issue a warning for undocumented Z80 instructions that also work on R800.
1181 @item -warn-unportable-instructions
1183 Issue a warning for undocumented Z80 instructions that do notwork on R800.
1184 @item -forbid-undocumented-instructions
1186 Treat all undocumented instructions as errors.
1187 @item -forbid-unportable-instructions
1189 Treat undocumented Z80 intructions that do notwork on R800 as errors.
1196 * Manual:: Structure of this Manual
1197 * GNU Assembler:: The GNU Assembler
1198 * Object Formats:: Object File Formats
1199 * Command Line:: Command Line
1200 * Input Files:: Input Files
1201 * Object:: Output (Object) File
1202 * Errors:: Error and Warning Messages
1206 @section Structure of this Manual
1208 @cindex manual, structure and purpose
1209 This manual is intended to describe what you need to know to use
1210 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1211 notation for symbols, constants, and expressions; the directives that
1212 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1215 We also cover special features in the @value{TARGET}
1216 configuration of @command{@value{AS}}, including assembler directives.
1219 This manual also describes some of the machine-dependent features of
1220 various flavors of the assembler.
1223 @cindex machine instructions (not covered)
1224 On the other hand, this manual is @emph{not} intended as an introduction
1225 to programming in assembly language---let alone programming in general!
1226 In a similar vein, we make no attempt to introduce the machine
1227 architecture; we do @emph{not} describe the instruction set, standard
1228 mnemonics, registers or addressing modes that are standard to a
1229 particular architecture.
1231 You may want to consult the manufacturer's
1232 machine architecture manual for this information.
1236 For information on the H8/300 machine instruction set, see @cite{H8/300
1237 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1238 Programming Manual} (Renesas).
1241 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1242 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1243 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1244 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1247 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1251 @c I think this is premature---doc@cygnus.com, 17jan1991
1253 Throughout this manual, we assume that you are running @dfn{GNU},
1254 the portable operating system from the @dfn{Free Software
1255 Foundation, Inc.}. This restricts our attention to certain kinds of
1256 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1257 once this assumption is granted examples and definitions need less
1260 @command{@value{AS}} is part of a team of programs that turn a high-level
1261 human-readable series of instructions into a low-level
1262 computer-readable series of instructions. Different versions of
1263 @command{@value{AS}} are used for different kinds of computer.
1266 @c There used to be a section "Terminology" here, which defined
1267 @c "contents", "byte", "word", and "long". Defining "word" to any
1268 @c particular size is confusing when the .word directive may generate 16
1269 @c bits on one machine and 32 bits on another; in general, for the user
1270 @c version of this manual, none of these terms seem essential to define.
1271 @c They were used very little even in the former draft of the manual;
1272 @c this draft makes an effort to avoid them (except in names of
1276 @section The GNU Assembler
1278 @c man begin DESCRIPTION
1280 @sc{gnu} @command{as} is really a family of assemblers.
1282 This manual describes @command{@value{AS}}, a member of that family which is
1283 configured for the @value{TARGET} architectures.
1285 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1286 should find a fairly similar environment when you use it on another
1287 architecture. Each version has much in common with the others,
1288 including object file formats, most assembler directives (often called
1289 @dfn{pseudo-ops}) and assembler syntax.@refill
1291 @cindex purpose of @sc{gnu} assembler
1292 @command{@value{AS}} is primarily intended to assemble the output of the
1293 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1294 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1295 assemble correctly everything that other assemblers for the same
1296 machine would assemble.
1298 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1301 @c This remark should appear in generic version of manual; assumption
1302 @c here is that generic version sets M680x0.
1303 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1304 assembler for the same architecture; for example, we know of several
1305 incompatible versions of 680x0 assembly language syntax.
1310 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1311 program in one pass of the source file. This has a subtle impact on the
1312 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1314 @node Object Formats
1315 @section Object File Formats
1317 @cindex object file format
1318 The @sc{gnu} assembler can be configured to produce several alternative
1319 object file formats. For the most part, this does not affect how you
1320 write assembly language programs; but directives for debugging symbols
1321 are typically different in different file formats. @xref{Symbol
1322 Attributes,,Symbol Attributes}.
1325 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1326 @value{OBJ-NAME} format object files.
1328 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1330 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1331 @code{b.out} or COFF format object files.
1334 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1335 SOM or ELF format object files.
1340 @section Command Line
1342 @cindex command line conventions
1344 After the program name @command{@value{AS}}, the command line may contain
1345 options and file names. Options may appear in any order, and may be
1346 before, after, or between file names. The order of file names is
1349 @cindex standard input, as input file
1351 @file{--} (two hyphens) by itself names the standard input file
1352 explicitly, as one of the files for @command{@value{AS}} to assemble.
1354 @cindex options, command line
1355 Except for @samp{--} any command line argument that begins with a
1356 hyphen (@samp{-}) is an option. Each option changes the behavior of
1357 @command{@value{AS}}. No option changes the way another option works. An
1358 option is a @samp{-} followed by one or more letters; the case of
1359 the letter is important. All options are optional.
1361 Some options expect exactly one file name to follow them. The file
1362 name may either immediately follow the option's letter (compatible
1363 with older assemblers) or it may be the next command argument (@sc{gnu}
1364 standard). These two command lines are equivalent:
1367 @value{AS} -o my-object-file.o mumble.s
1368 @value{AS} -omy-object-file.o mumble.s
1372 @section Input Files
1375 @cindex source program
1376 @cindex files, input
1377 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1378 describe the program input to one run of @command{@value{AS}}. The program may
1379 be in one or more files; how the source is partitioned into files
1380 doesn't change the meaning of the source.
1382 @c I added "con" prefix to "catenation" just to prove I can overcome my
1383 @c APL training... doc@cygnus.com
1384 The source program is a concatenation of the text in all the files, in the
1387 @c man begin DESCRIPTION
1388 Each time you run @command{@value{AS}} it assembles exactly one source
1389 program. The source program is made up of one or more files.
1390 (The standard input is also a file.)
1392 You give @command{@value{AS}} a command line that has zero or more input file
1393 names. The input files are read (from left file name to right). A
1394 command line argument (in any position) that has no special meaning
1395 is taken to be an input file name.
1397 If you give @command{@value{AS}} no file names it attempts to read one input file
1398 from the @command{@value{AS}} standard input, which is normally your terminal. You
1399 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1402 Use @samp{--} if you need to explicitly name the standard input file
1403 in your command line.
1405 If the source is empty, @command{@value{AS}} produces a small, empty object
1410 @subheading Filenames and Line-numbers
1412 @cindex input file linenumbers
1413 @cindex line numbers, in input files
1414 There are two ways of locating a line in the input file (or files) and
1415 either may be used in reporting error messages. One way refers to a line
1416 number in a physical file; the other refers to a line number in a
1417 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1419 @dfn{Physical files} are those files named in the command line given
1420 to @command{@value{AS}}.
1422 @dfn{Logical files} are simply names declared explicitly by assembler
1423 directives; they bear no relation to physical files. Logical file names help
1424 error messages reflect the original source file, when @command{@value{AS}} source
1425 is itself synthesized from other files. @command{@value{AS}} understands the
1426 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1427 @ref{File,,@code{.file}}.
1430 @section Output (Object) File
1436 Every time you run @command{@value{AS}} it produces an output file, which is
1437 your assembly language program translated into numbers. This file
1438 is the object file. Its default name is
1446 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1448 You can give it another name by using the @option{-o} option. Conventionally,
1449 object file names end with @file{.o}. The default name is used for historical
1450 reasons: older assemblers were capable of assembling self-contained programs
1451 directly into a runnable program. (For some formats, this isn't currently
1452 possible, but it can be done for the @code{a.out} format.)
1456 The object file is meant for input to the linker @code{@value{LD}}. It contains
1457 assembled program code, information to help @code{@value{LD}} integrate
1458 the assembled program into a runnable file, and (optionally) symbolic
1459 information for the debugger.
1461 @c link above to some info file(s) like the description of a.out.
1462 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1465 @section Error and Warning Messages
1467 @c man begin DESCRIPTION
1469 @cindex error messages
1470 @cindex warning messages
1471 @cindex messages from assembler
1472 @command{@value{AS}} may write warnings and error messages to the standard error
1473 file (usually your terminal). This should not happen when a compiler
1474 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1475 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1476 grave problem that stops the assembly.
1480 @cindex format of warning messages
1481 Warning messages have the format
1484 file_name:@b{NNN}:Warning Message Text
1488 @cindex line numbers, in warnings/errors
1489 (where @b{NNN} is a line number). If a logical file name has been given
1490 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1491 the current input file is used. If a logical line number was given
1493 (@pxref{Line,,@code{.line}})
1495 then it is used to calculate the number printed,
1496 otherwise the actual line in the current source file is printed. The
1497 message text is intended to be self explanatory (in the grand Unix
1500 @cindex format of error messages
1501 Error messages have the format
1503 file_name:@b{NNN}:FATAL:Error Message Text
1505 The file name and line number are derived as for warning
1506 messages. The actual message text may be rather less explanatory
1507 because many of them aren't supposed to happen.
1510 @chapter Command-Line Options
1512 @cindex options, all versions of assembler
1513 This chapter describes command-line options available in @emph{all}
1514 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1516 to the @value{TARGET} target.
1519 to particular machine architectures.
1522 @c man begin DESCRIPTION
1524 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1525 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1526 The assembler arguments must be separated from each other (and the @samp{-Wa})
1527 by commas. For example:
1530 gcc -c -g -O -Wa,-alh,-L file.c
1534 This passes two options to the assembler: @samp{-alh} (emit a listing to
1535 standard output with high-level and assembly source) and @samp{-L} (retain
1536 local symbols in the symbol table).
1538 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1539 command-line options are automatically passed to the assembler by the compiler.
1540 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1541 precisely what options it passes to each compilation pass, including the
1547 * a:: -a[cdhlns] enable listings
1548 * alternate:: --alternate enable alternate macro syntax
1549 * D:: -D for compatibility
1550 * f:: -f to work faster
1551 * I:: -I for .include search path
1552 @ifclear DIFF-TBL-KLUGE
1553 * K:: -K for compatibility
1555 @ifset DIFF-TBL-KLUGE
1556 * K:: -K for difference tables
1559 * L:: -L to retain local labels
1560 * listing:: --listing-XXX to configure listing output
1561 * M:: -M or --mri to assemble in MRI compatibility mode
1562 * MD:: --MD for dependency tracking
1563 * o:: -o to name the object file
1564 * R:: -R to join data and text sections
1565 * statistics:: --statistics to see statistics about assembly
1566 * traditional-format:: --traditional-format for compatible output
1567 * v:: -v to announce version
1568 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1569 * Z:: -Z to make object file even after errors
1573 @section Enable Listings: @option{-a[cdhlns]}
1582 @cindex listings, enabling
1583 @cindex assembly listings, enabling
1585 These options enable listing output from the assembler. By itself,
1586 @samp{-a} requests high-level, assembly, and symbols listing.
1587 You can use other letters to select specific options for the list:
1588 @samp{-ah} requests a high-level language listing,
1589 @samp{-al} requests an output-program assembly listing, and
1590 @samp{-as} requests a symbol table listing.
1591 High-level listings require that a compiler debugging option like
1592 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1595 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1596 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1597 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1598 omitted from the listing.
1600 Use the @samp{-ad} option to omit debugging directives from the
1603 Once you have specified one of these options, you can further control
1604 listing output and its appearance using the directives @code{.list},
1605 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1607 The @samp{-an} option turns off all forms processing.
1608 If you do not request listing output with one of the @samp{-a} options, the
1609 listing-control directives have no effect.
1611 The letters after @samp{-a} may be combined into one option,
1612 @emph{e.g.}, @samp{-aln}.
1614 Note if the assembler source is coming from the standard input (eg because it
1615 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1616 is being used) then the listing will not contain any comments or preprocessor
1617 directives. This is because the listing code buffers input source lines from
1618 stdin only after they have been preprocessed by the assembler. This reduces
1619 memory usage and makes the code more efficient.
1622 @section @option{--alternate}
1625 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1628 @section @option{-D}
1631 This option has no effect whatsoever, but it is accepted to make it more
1632 likely that scripts written for other assemblers also work with
1633 @command{@value{AS}}.
1636 @section Work Faster: @option{-f}
1639 @cindex trusted compiler
1640 @cindex faster processing (@option{-f})
1641 @samp{-f} should only be used when assembling programs written by a
1642 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1643 and comment preprocessing on
1644 the input file(s) before assembling them. @xref{Preprocessing,
1648 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1649 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1654 @section @code{.include} Search Path: @option{-I} @var{path}
1656 @kindex -I @var{path}
1657 @cindex paths for @code{.include}
1658 @cindex search path for @code{.include}
1659 @cindex @code{include} directive search path
1660 Use this option to add a @var{path} to the list of directories
1661 @command{@value{AS}} searches for files specified in @code{.include}
1662 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1663 many times as necessary to include a variety of paths. The current
1664 working directory is always searched first; after that, @command{@value{AS}}
1665 searches any @samp{-I} directories in the same order as they were
1666 specified (left to right) on the command line.
1669 @section Difference Tables: @option{-K}
1672 @ifclear DIFF-TBL-KLUGE
1673 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1674 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1675 where it can be used to warn when the assembler alters the machine code
1676 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1677 family does not have the addressing limitations that sometimes lead to this
1678 alteration on other platforms.
1681 @ifset DIFF-TBL-KLUGE
1682 @cindex difference tables, warning
1683 @cindex warning for altered difference tables
1684 @command{@value{AS}} sometimes alters the code emitted for directives of the form
1685 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1686 You can use the @samp{-K} option if you want a warning issued when this
1691 @section Include Local Labels: @option{-L}
1694 @cindex local labels, retaining in output
1695 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1696 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1697 debugging, because they are intended for the use of programs (like
1698 compilers) that compose assembler programs, not for your notice.
1699 Normally both @command{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1700 normally debug with them.
1702 This option tells @command{@value{AS}} to retain those @samp{L@dots{}} symbols
1703 in the object file. Usually if you do this you also tell the linker
1704 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1706 By default, a local label is any label beginning with @samp{L}, but each
1707 target is allowed to redefine the local label prefix.
1709 On the HPPA local labels begin with @samp{L$}.
1713 @section Configuring listing output: @option{--listing}
1715 The listing feature of the assembler can be enabled via the command line switch
1716 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1717 hex dump of the corresponding locations in the output object file, and displays
1718 them as a listing file. The format of this listing can be controlled by pseudo
1719 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1720 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1723 @item --listing-lhs-width=@samp{number}
1724 @kindex --listing-lhs-width
1725 @cindex Width of first line disassembly output
1726 Sets the maximum width, in words, of the first line of the hex byte dump. This
1727 dump appears on the left hand side of the listing output.
1729 @item --listing-lhs-width2=@samp{number}
1730 @kindex --listing-lhs-width2
1731 @cindex Width of continuation lines of disassembly output
1732 Sets the maximum width, in words, of any further lines of the hex byte dump for
1733 a given input source line. If this value is not specified, it defaults to being
1734 the same as the value specified for @samp{--listing-lhs-width}. If neither
1735 switch is used the default is to one.
1737 @item --listing-rhs-width=@samp{number}
1738 @kindex --listing-rhs-width
1739 @cindex Width of source line output
1740 Sets the maximum width, in characters, of the source line that is displayed
1741 alongside the hex dump. The default value for this parameter is 100. The
1742 source line is displayed on the right hand side of the listing output.
1744 @item --listing-cont-lines=@samp{number}
1745 @kindex --listing-cont-lines
1746 @cindex Maximum number of continuation lines
1747 Sets the maximum number of continuation lines of hex dump that will be
1748 displayed for a given single line of source input. The default value is 4.
1752 @section Assemble in MRI Compatibility Mode: @option{-M}
1755 @cindex MRI compatibility mode
1756 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1757 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1758 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1759 configured target) assembler from Microtec Research. The exact nature of the
1760 MRI syntax will not be documented here; see the MRI manuals for more
1761 information. Note in particular that the handling of macros and macro
1762 arguments is somewhat different. The purpose of this option is to permit
1763 assembling existing MRI assembler code using @command{@value{AS}}.
1765 The MRI compatibility is not complete. Certain operations of the MRI assembler
1766 depend upon its object file format, and can not be supported using other object
1767 file formats. Supporting these would require enhancing each object file format
1768 individually. These are:
1771 @item global symbols in common section
1773 The m68k MRI assembler supports common sections which are merged by the linker.
1774 Other object file formats do not support this. @command{@value{AS}} handles
1775 common sections by treating them as a single common symbol. It permits local
1776 symbols to be defined within a common section, but it can not support global
1777 symbols, since it has no way to describe them.
1779 @item complex relocations
1781 The MRI assemblers support relocations against a negated section address, and
1782 relocations which combine the start addresses of two or more sections. These
1783 are not support by other object file formats.
1785 @item @code{END} pseudo-op specifying start address
1787 The MRI @code{END} pseudo-op permits the specification of a start address.
1788 This is not supported by other object file formats. The start address may
1789 instead be specified using the @option{-e} option to the linker, or in a linker
1792 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1794 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1795 name to the output file. This is not supported by other object file formats.
1797 @item @code{ORG} pseudo-op
1799 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1800 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1801 which changes the location within the current section. Absolute sections are
1802 not supported by other object file formats. The address of a section may be
1803 assigned within a linker script.
1806 There are some other features of the MRI assembler which are not supported by
1807 @command{@value{AS}}, typically either because they are difficult or because they
1808 seem of little consequence. Some of these may be supported in future releases.
1812 @item EBCDIC strings
1814 EBCDIC strings are not supported.
1816 @item packed binary coded decimal
1818 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1819 and @code{DCB.P} pseudo-ops are not supported.
1821 @item @code{FEQU} pseudo-op
1823 The m68k @code{FEQU} pseudo-op is not supported.
1825 @item @code{NOOBJ} pseudo-op
1827 The m68k @code{NOOBJ} pseudo-op is not supported.
1829 @item @code{OPT} branch control options
1831 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1832 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1833 relaxes all branches, whether forward or backward, to an appropriate size, so
1834 these options serve no purpose.
1836 @item @code{OPT} list control options
1838 The following m68k @code{OPT} list control options are ignored: @code{C},
1839 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1840 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1842 @item other @code{OPT} options
1844 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1845 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1847 @item @code{OPT} @code{D} option is default
1849 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1850 @code{OPT NOD} may be used to turn it off.
1852 @item @code{XREF} pseudo-op.
1854 The m68k @code{XREF} pseudo-op is ignored.
1856 @item @code{.debug} pseudo-op
1858 The i960 @code{.debug} pseudo-op is not supported.
1860 @item @code{.extended} pseudo-op
1862 The i960 @code{.extended} pseudo-op is not supported.
1864 @item @code{.list} pseudo-op.
1866 The various options of the i960 @code{.list} pseudo-op are not supported.
1868 @item @code{.optimize} pseudo-op
1870 The i960 @code{.optimize} pseudo-op is not supported.
1872 @item @code{.output} pseudo-op
1874 The i960 @code{.output} pseudo-op is not supported.
1876 @item @code{.setreal} pseudo-op
1878 The i960 @code{.setreal} pseudo-op is not supported.
1883 @section Dependency Tracking: @option{--MD}
1886 @cindex dependency tracking
1889 @command{@value{AS}} can generate a dependency file for the file it creates. This
1890 file consists of a single rule suitable for @code{make} describing the
1891 dependencies of the main source file.
1893 The rule is written to the file named in its argument.
1895 This feature is used in the automatic updating of makefiles.
1898 @section Name the Object File: @option{-o}
1901 @cindex naming object file
1902 @cindex object file name
1903 There is always one object file output when you run @command{@value{AS}}. By
1904 default it has the name
1907 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1921 You use this option (which takes exactly one filename) to give the
1922 object file a different name.
1924 Whatever the object file is called, @command{@value{AS}} overwrites any
1925 existing file of the same name.
1928 @section Join Data and Text Sections: @option{-R}
1931 @cindex data and text sections, joining
1932 @cindex text and data sections, joining
1933 @cindex joining text and data sections
1934 @cindex merging text and data sections
1935 @option{-R} tells @command{@value{AS}} to write the object file as if all
1936 data-section data lives in the text section. This is only done at
1937 the very last moment: your binary data are the same, but data
1938 section parts are relocated differently. The data section part of
1939 your object file is zero bytes long because all its bytes are
1940 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1942 When you specify @option{-R} it would be possible to generate shorter
1943 address displacements (because we do not have to cross between text and
1944 data section). We refrain from doing this simply for compatibility with
1945 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1948 When @command{@value{AS}} is configured for COFF or ELF output,
1949 this option is only useful if you use sections named @samp{.text} and
1954 @option{-R} is not supported for any of the HPPA targets. Using
1955 @option{-R} generates a warning from @command{@value{AS}}.
1959 @section Display Assembly Statistics: @option{--statistics}
1961 @kindex --statistics
1962 @cindex statistics, about assembly
1963 @cindex time, total for assembly
1964 @cindex space used, maximum for assembly
1965 Use @samp{--statistics} to display two statistics about the resources used by
1966 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1967 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1970 @node traditional-format
1971 @section Compatible Output: @option{--traditional-format}
1973 @kindex --traditional-format
1974 For some targets, the output of @command{@value{AS}} is different in some ways
1975 from the output of some existing assembler. This switch requests
1976 @command{@value{AS}} to use the traditional format instead.
1978 For example, it disables the exception frame optimizations which
1979 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
1982 @section Announce Version: @option{-v}
1986 @cindex assembler version
1987 @cindex version of assembler
1988 You can find out what version of as is running by including the
1989 option @samp{-v} (which you can also spell as @samp{-version}) on the
1993 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
1995 @command{@value{AS}} should never give a warning or error message when
1996 assembling compiler output. But programs written by people often
1997 cause @command{@value{AS}} to give a warning that a particular assumption was
1998 made. All such warnings are directed to the standard error file.
2002 @cindex suppressing warnings
2003 @cindex warnings, suppressing
2004 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2005 This only affects the warning messages: it does not change any particular of
2006 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2009 @kindex --fatal-warnings
2010 @cindex errors, caused by warnings
2011 @cindex warnings, causing error
2012 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2013 files that generate warnings to be in error.
2016 @cindex warnings, switching on
2017 You can switch these options off again by specifying @option{--warn}, which
2018 causes warnings to be output as usual.
2021 @section Generate Object File in Spite of Errors: @option{-Z}
2022 @cindex object file, after errors
2023 @cindex errors, continuing after
2024 After an error message, @command{@value{AS}} normally produces no output. If for
2025 some reason you are interested in object file output even after
2026 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2027 option. If there are any errors, @command{@value{AS}} continues anyways, and
2028 writes an object file after a final warning message of the form @samp{@var{n}
2029 errors, @var{m} warnings, generating bad object file.}
2034 @cindex machine-independent syntax
2035 @cindex syntax, machine-independent
2036 This chapter describes the machine-independent syntax allowed in a
2037 source file. @command{@value{AS}} syntax is similar to what many other
2038 assemblers use; it is inspired by the BSD 4.2
2043 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2047 * Preprocessing:: Preprocessing
2048 * Whitespace:: Whitespace
2049 * Comments:: Comments
2050 * Symbol Intro:: Symbols
2051 * Statements:: Statements
2052 * Constants:: Constants
2056 @section Preprocessing
2058 @cindex preprocessing
2059 The @command{@value{AS}} internal preprocessor:
2061 @cindex whitespace, removed by preprocessor
2063 adjusts and removes extra whitespace. It leaves one space or tab before
2064 the keywords on a line, and turns any other whitespace on the line into
2067 @cindex comments, removed by preprocessor
2069 removes all comments, replacing them with a single space, or an
2070 appropriate number of newlines.
2072 @cindex constants, converted by preprocessor
2074 converts character constants into the appropriate numeric values.
2077 It does not do macro processing, include file handling, or
2078 anything else you may get from your C compiler's preprocessor. You can
2079 do include file processing with the @code{.include} directive
2080 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2081 to get other ``CPP'' style preprocessing by giving the input file a
2082 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
2083 Output, gcc.info, Using GNU CC}.
2085 Excess whitespace, comments, and character constants
2086 cannot be used in the portions of the input text that are not
2089 @cindex turning preprocessing on and off
2090 @cindex preprocessing, turning on and off
2093 If the first line of an input file is @code{#NO_APP} or if you use the
2094 @samp{-f} option, whitespace and comments are not removed from the input file.
2095 Within an input file, you can ask for whitespace and comment removal in
2096 specific portions of the by putting a line that says @code{#APP} before the
2097 text that may contain whitespace or comments, and putting a line that says
2098 @code{#NO_APP} after this text. This feature is mainly intend to support
2099 @code{asm} statements in compilers whose output is otherwise free of comments
2106 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2107 Whitespace is used to separate symbols, and to make programs neater for
2108 people to read. Unless within character constants
2109 (@pxref{Characters,,Character Constants}), any whitespace means the same
2110 as exactly one space.
2116 There are two ways of rendering comments to @command{@value{AS}}. In both
2117 cases the comment is equivalent to one space.
2119 Anything from @samp{/*} through the next @samp{*/} is a comment.
2120 This means you may not nest these comments.
2124 The only way to include a newline ('\n') in a comment
2125 is to use this sort of comment.
2128 /* This sort of comment does not nest. */
2131 @cindex line comment character
2132 Anything from the @dfn{line comment} character to the next newline
2133 is considered a comment and is ignored. The line comment character is
2135 @samp{;} on the ARC;
2138 @samp{@@} on the ARM;
2141 @samp{;} for the H8/300 family;
2144 @samp{;} for the HPPA;
2147 @samp{#} on the i386 and x86-64;
2150 @samp{#} on the i960;
2153 @samp{;} for the PDP-11;
2156 @samp{;} for picoJava;
2159 @samp{#} for Motorola PowerPC;
2162 @samp{!} for the Renesas / SuperH SH;
2165 @samp{!} on the SPARC;
2168 @samp{#} on the ip2k;
2171 @samp{#} on the m32c;
2174 @samp{#} on the m32r;
2177 @samp{|} on the 680x0;
2180 @samp{#} on the 68HC11 and 68HC12;
2183 @samp{#} on the Vax;
2186 @samp{;} for the Z80;
2189 @samp{!} for the Z8000;
2192 @samp{#} on the V850;
2195 @samp{#} for Xtensa systems;
2197 see @ref{Machine Dependencies}. @refill
2198 @c FIXME What about i860?
2201 On some machines there are two different line comment characters. One
2202 character only begins a comment if it is the first non-whitespace character on
2203 a line, while the other always begins a comment.
2207 The V850 assembler also supports a double dash as starting a comment that
2208 extends to the end of the line.
2214 @cindex lines starting with @code{#}
2215 @cindex logical line numbers
2216 To be compatible with past assemblers, lines that begin with @samp{#} have a
2217 special interpretation. Following the @samp{#} should be an absolute
2218 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2219 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
2220 new logical file name. The rest of the line, if any, should be whitespace.
2222 If the first non-whitespace characters on the line are not numeric,
2223 the line is ignored. (Just like a comment.)
2226 # This is an ordinary comment.
2227 # 42-6 "new_file_name" # New logical file name
2228 # This is logical line # 36.
2230 This feature is deprecated, and may disappear from future versions
2231 of @command{@value{AS}}.
2236 @cindex characters used in symbols
2237 @ifclear SPECIAL-SYMS
2238 A @dfn{symbol} is one or more characters chosen from the set of all
2239 letters (both upper and lower case), digits and the three characters
2245 A @dfn{symbol} is one or more characters chosen from the set of all
2246 letters (both upper and lower case), digits and the three characters
2247 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2253 On most machines, you can also use @code{$} in symbol names; exceptions
2254 are noted in @ref{Machine Dependencies}.
2256 No symbol may begin with a digit. Case is significant.
2257 There is no length limit: all characters are significant. Symbols are
2258 delimited by characters not in that set, or by the beginning of a file
2259 (since the source program must end with a newline, the end of a file is
2260 not a possible symbol delimiter). @xref{Symbols}.
2261 @cindex length of symbols
2266 @cindex statements, structure of
2267 @cindex line separator character
2268 @cindex statement separator character
2270 @ifclear abnormal-separator
2271 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2272 semicolon (@samp{;}). The newline or semicolon is considered part of
2273 the preceding statement. Newlines and semicolons within character
2274 constants are an exception: they do not end statements.
2276 @ifset abnormal-separator
2278 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2279 point (@samp{!}). The newline or exclamation point is considered part of the
2280 preceding statement. Newlines and exclamation points within character
2281 constants are an exception: they do not end statements.
2284 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2285 H8/300) a dollar sign (@samp{$}); or (for the Renesas-SH) a semicolon
2286 (@samp{;}). The newline or separator character is considered part of
2287 the preceding statement. Newlines and separators within character
2288 constants are an exception: they do not end statements.
2293 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2294 separator character. (The line separator is usually @samp{;}, unless
2295 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
2296 newline or separator character is considered part of the preceding
2297 statement. Newlines and separators within character constants are an
2298 exception: they do not end statements.
2301 @cindex newline, required at file end
2302 @cindex EOF, newline must precede
2303 It is an error to end any statement with end-of-file: the last
2304 character of any input file should be a newline.@refill
2306 An empty statement is allowed, and may include whitespace. It is ignored.
2308 @cindex instructions and directives
2309 @cindex directives and instructions
2310 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2311 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2313 A statement begins with zero or more labels, optionally followed by a
2314 key symbol which determines what kind of statement it is. The key
2315 symbol determines the syntax of the rest of the statement. If the
2316 symbol begins with a dot @samp{.} then the statement is an assembler
2317 directive: typically valid for any computer. If the symbol begins with
2318 a letter the statement is an assembly language @dfn{instruction}: it
2319 assembles into a machine language instruction.
2321 Different versions of @command{@value{AS}} for different computers
2322 recognize different instructions. In fact, the same symbol may
2323 represent a different instruction in a different computer's assembly
2327 @cindex @code{:} (label)
2328 @cindex label (@code{:})
2329 A label is a symbol immediately followed by a colon (@code{:}).
2330 Whitespace before a label or after a colon is permitted, but you may not
2331 have whitespace between a label's symbol and its colon. @xref{Labels}.
2334 For HPPA targets, labels need not be immediately followed by a colon, but
2335 the definition of a label must begin in column zero. This also implies that
2336 only one label may be defined on each line.
2340 label: .directive followed by something
2341 another_label: # This is an empty statement.
2342 instruction operand_1, operand_2, @dots{}
2349 A constant is a number, written so that its value is known by
2350 inspection, without knowing any context. Like this:
2353 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2354 .ascii "Ring the bell\7" # A string constant.
2355 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2356 .float 0f-314159265358979323846264338327\
2357 95028841971.693993751E-40 # - pi, a flonum.
2362 * Characters:: Character Constants
2363 * Numbers:: Number Constants
2367 @subsection Character Constants
2369 @cindex character constants
2370 @cindex constants, character
2371 There are two kinds of character constants. A @dfn{character} stands
2372 for one character in one byte and its value may be used in
2373 numeric expressions. String constants (properly called string
2374 @emph{literals}) are potentially many bytes and their values may not be
2375 used in arithmetic expressions.
2379 * Chars:: Characters
2383 @subsubsection Strings
2385 @cindex string constants
2386 @cindex constants, string
2387 A @dfn{string} is written between double-quotes. It may contain
2388 double-quotes or null characters. The way to get special characters
2389 into a string is to @dfn{escape} these characters: precede them with
2390 a backslash @samp{\} character. For example @samp{\\} represents
2391 one backslash: the first @code{\} is an escape which tells
2392 @command{@value{AS}} to interpret the second character literally as a backslash
2393 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2394 escape character). The complete list of escapes follows.
2396 @cindex escape codes, character
2397 @cindex character escape codes
2400 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2402 @cindex @code{\b} (backspace character)
2403 @cindex backspace (@code{\b})
2405 Mnemonic for backspace; for ASCII this is octal code 010.
2408 @c Mnemonic for EOText; for ASCII this is octal code 004.
2410 @cindex @code{\f} (formfeed character)
2411 @cindex formfeed (@code{\f})
2413 Mnemonic for FormFeed; for ASCII this is octal code 014.
2415 @cindex @code{\n} (newline character)
2416 @cindex newline (@code{\n})
2418 Mnemonic for newline; for ASCII this is octal code 012.
2421 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2423 @cindex @code{\r} (carriage return character)
2424 @cindex carriage return (@code{\r})
2426 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2429 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2430 @c other assemblers.
2432 @cindex @code{\t} (tab)
2433 @cindex tab (@code{\t})
2435 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2438 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2439 @c @item \x @var{digit} @var{digit} @var{digit}
2440 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2442 @cindex @code{\@var{ddd}} (octal character code)
2443 @cindex octal character code (@code{\@var{ddd}})
2444 @item \ @var{digit} @var{digit} @var{digit}
2445 An octal character code. The numeric code is 3 octal digits.
2446 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2447 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2449 @cindex @code{\@var{xd...}} (hex character code)
2450 @cindex hex character code (@code{\@var{xd...}})
2451 @item \@code{x} @var{hex-digits...}
2452 A hex character code. All trailing hex digits are combined. Either upper or
2453 lower case @code{x} works.
2455 @cindex @code{\\} (@samp{\} character)
2456 @cindex backslash (@code{\\})
2458 Represents one @samp{\} character.
2461 @c Represents one @samp{'} (accent acute) character.
2462 @c This is needed in single character literals
2463 @c (@xref{Characters,,Character Constants}.) to represent
2466 @cindex @code{\"} (doublequote character)
2467 @cindex doublequote (@code{\"})
2469 Represents one @samp{"} character. Needed in strings to represent
2470 this character, because an unescaped @samp{"} would end the string.
2472 @item \ @var{anything-else}
2473 Any other character when escaped by @kbd{\} gives a warning, but
2474 assembles as if the @samp{\} was not present. The idea is that if
2475 you used an escape sequence you clearly didn't want the literal
2476 interpretation of the following character. However @command{@value{AS}} has no
2477 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2478 code and warns you of the fact.
2481 Which characters are escapable, and what those escapes represent,
2482 varies widely among assemblers. The current set is what we think
2483 the BSD 4.2 assembler recognizes, and is a subset of what most C
2484 compilers recognize. If you are in doubt, do not use an escape
2488 @subsubsection Characters
2490 @cindex single character constant
2491 @cindex character, single
2492 @cindex constant, single character
2493 A single character may be written as a single quote immediately
2494 followed by that character. The same escapes apply to characters as
2495 to strings. So if you want to write the character backslash, you
2496 must write @kbd{'\\} where the first @code{\} escapes the second
2497 @code{\}. As you can see, the quote is an acute accent, not a
2498 grave accent. A newline
2500 @ifclear abnormal-separator
2501 (or semicolon @samp{;})
2503 @ifset abnormal-separator
2505 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2510 immediately following an acute accent is taken as a literal character
2511 and does not count as the end of a statement. The value of a character
2512 constant in a numeric expression is the machine's byte-wide code for
2513 that character. @command{@value{AS}} assumes your character code is ASCII:
2514 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2517 @subsection Number Constants
2519 @cindex constants, number
2520 @cindex number constants
2521 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2522 are stored in the target machine. @emph{Integers} are numbers that
2523 would fit into an @code{int} in the C language. @emph{Bignums} are
2524 integers, but they are stored in more than 32 bits. @emph{Flonums}
2525 are floating point numbers, described below.
2528 * Integers:: Integers
2533 * Bit Fields:: Bit Fields
2539 @subsubsection Integers
2541 @cindex constants, integer
2543 @cindex binary integers
2544 @cindex integers, binary
2545 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2546 the binary digits @samp{01}.
2548 @cindex octal integers
2549 @cindex integers, octal
2550 An octal integer is @samp{0} followed by zero or more of the octal
2551 digits (@samp{01234567}).
2553 @cindex decimal integers
2554 @cindex integers, decimal
2555 A decimal integer starts with a non-zero digit followed by zero or
2556 more digits (@samp{0123456789}).
2558 @cindex hexadecimal integers
2559 @cindex integers, hexadecimal
2560 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2561 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2563 Integers have the usual values. To denote a negative integer, use
2564 the prefix operator @samp{-} discussed under expressions
2565 (@pxref{Prefix Ops,,Prefix Operators}).
2568 @subsubsection Bignums
2571 @cindex constants, bignum
2572 A @dfn{bignum} has the same syntax and semantics as an integer
2573 except that the number (or its negative) takes more than 32 bits to
2574 represent in binary. The distinction is made because in some places
2575 integers are permitted while bignums are not.
2578 @subsubsection Flonums
2580 @cindex floating point numbers
2581 @cindex constants, floating point
2583 @cindex precision, floating point
2584 A @dfn{flonum} represents a floating point number. The translation is
2585 indirect: a decimal floating point number from the text is converted by
2586 @command{@value{AS}} to a generic binary floating point number of more than
2587 sufficient precision. This generic floating point number is converted
2588 to a particular computer's floating point format (or formats) by a
2589 portion of @command{@value{AS}} specialized to that computer.
2591 A flonum is written by writing (in order)
2596 (@samp{0} is optional on the HPPA.)
2600 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2602 @kbd{e} is recommended. Case is not important.
2604 @c FIXME: verify if flonum syntax really this vague for most cases
2605 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2606 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2609 On the H8/300, Renesas / SuperH SH,
2610 and AMD 29K architectures, the letter must be
2611 one of the letters @samp{DFPRSX} (in upper or lower case).
2613 On the ARC, the letter must be one of the letters @samp{DFRS}
2614 (in upper or lower case).
2616 On the Intel 960 architecture, the letter must be
2617 one of the letters @samp{DFT} (in upper or lower case).
2619 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2623 One of the letters @samp{DFRS} (in upper or lower case).
2626 One of the letters @samp{DFPRSX} (in upper or lower case).
2629 The letter @samp{E} (upper case only).
2632 One of the letters @samp{DFT} (in upper or lower case).
2637 An optional sign: either @samp{+} or @samp{-}.
2640 An optional @dfn{integer part}: zero or more decimal digits.
2643 An optional @dfn{fractional part}: @samp{.} followed by zero
2644 or more decimal digits.
2647 An optional exponent, consisting of:
2651 An @samp{E} or @samp{e}.
2652 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2653 @c principle this can perfectly well be different on different targets.
2655 Optional sign: either @samp{+} or @samp{-}.
2657 One or more decimal digits.
2662 At least one of the integer part or the fractional part must be
2663 present. The floating point number has the usual base-10 value.
2665 @command{@value{AS}} does all processing using integers. Flonums are computed
2666 independently of any floating point hardware in the computer running
2667 @command{@value{AS}}.
2671 @c Bit fields are written as a general facility but are also controlled
2672 @c by a conditional-compilation flag---which is as of now (21mar91)
2673 @c turned on only by the i960 config of GAS.
2675 @subsubsection Bit Fields
2678 @cindex constants, bit field
2679 You can also define numeric constants as @dfn{bit fields}.
2680 specify two numbers separated by a colon---
2682 @var{mask}:@var{value}
2685 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2688 The resulting number is then packed
2690 @c this conditional paren in case bit fields turned on elsewhere than 960
2691 (in host-dependent byte order)
2693 into a field whose width depends on which assembler directive has the
2694 bit-field as its argument. Overflow (a result from the bitwise and
2695 requiring more binary digits to represent) is not an error; instead,
2696 more constants are generated, of the specified width, beginning with the
2697 least significant digits.@refill
2699 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2700 @code{.short}, and @code{.word} accept bit-field arguments.
2705 @chapter Sections and Relocation
2710 * Secs Background:: Background
2711 * Ld Sections:: Linker Sections
2712 * As Sections:: Assembler Internal Sections
2713 * Sub-Sections:: Sub-Sections
2717 @node Secs Background
2720 Roughly, a section is a range of addresses, with no gaps; all data
2721 ``in'' those addresses is treated the same for some particular purpose.
2722 For example there may be a ``read only'' section.
2724 @cindex linker, and assembler
2725 @cindex assembler, and linker
2726 The linker @code{@value{LD}} reads many object files (partial programs) and
2727 combines their contents to form a runnable program. When @command{@value{AS}}
2728 emits an object file, the partial program is assumed to start at address 0.
2729 @code{@value{LD}} assigns the final addresses for the partial program, so that
2730 different partial programs do not overlap. This is actually an
2731 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2734 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2735 addresses. These blocks slide to their run-time addresses as rigid
2736 units; their length does not change and neither does the order of bytes
2737 within them. Such a rigid unit is called a @emph{section}. Assigning
2738 run-time addresses to sections is called @dfn{relocation}. It includes
2739 the task of adjusting mentions of object-file addresses so they refer to
2740 the proper run-time addresses.
2742 For the H8/300, and for the Renesas / SuperH SH,
2743 @command{@value{AS}} pads sections if needed to
2744 ensure they end on a word (sixteen bit) boundary.
2747 @cindex standard assembler sections
2748 An object file written by @command{@value{AS}} has at least three sections, any
2749 of which may be empty. These are named @dfn{text}, @dfn{data} and
2754 When it generates COFF or ELF output,
2756 @command{@value{AS}} can also generate whatever other named sections you specify
2757 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2758 If you do not use any directives that place output in the @samp{.text}
2759 or @samp{.data} sections, these sections still exist, but are empty.
2764 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2766 @command{@value{AS}} can also generate whatever other named sections you
2767 specify using the @samp{.space} and @samp{.subspace} directives. See
2768 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2769 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2770 assembler directives.
2773 Additionally, @command{@value{AS}} uses different names for the standard
2774 text, data, and bss sections when generating SOM output. Program text
2775 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2776 BSS into @samp{$BSS$}.
2780 Within the object file, the text section starts at address @code{0}, the
2781 data section follows, and the bss section follows the data section.
2784 When generating either SOM or ELF output files on the HPPA, the text
2785 section starts at address @code{0}, the data section at address
2786 @code{0x4000000}, and the bss section follows the data section.
2789 To let @code{@value{LD}} know which data changes when the sections are
2790 relocated, and how to change that data, @command{@value{AS}} also writes to the
2791 object file details of the relocation needed. To perform relocation
2792 @code{@value{LD}} must know, each time an address in the object
2796 Where in the object file is the beginning of this reference to
2799 How long (in bytes) is this reference?
2801 Which section does the address refer to? What is the numeric value of
2803 (@var{address}) @minus{} (@var{start-address of section})?
2806 Is the reference to an address ``Program-Counter relative''?
2809 @cindex addresses, format of
2810 @cindex section-relative addressing
2811 In fact, every address @command{@value{AS}} ever uses is expressed as
2813 (@var{section}) + (@var{offset into section})
2816 Further, most expressions @command{@value{AS}} computes have this section-relative
2819 (For some object formats, such as SOM for the HPPA, some expressions are
2820 symbol-relative instead.)
2823 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2824 @var{N} into section @var{secname}.''
2826 Apart from text, data and bss sections you need to know about the
2827 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2828 addresses in the absolute section remain unchanged. For example, address
2829 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2830 @code{@value{LD}}. Although the linker never arranges two partial programs'
2831 data sections with overlapping addresses after linking, @emph{by definition}
2832 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2833 part of a program is always the same address when the program is running as
2834 address @code{@{absolute@ 239@}} in any other part of the program.
2836 The idea of sections is extended to the @dfn{undefined} section. Any
2837 address whose section is unknown at assembly time is by definition
2838 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2839 Since numbers are always defined, the only way to generate an undefined
2840 address is to mention an undefined symbol. A reference to a named
2841 common block would be such a symbol: its value is unknown at assembly
2842 time so it has section @emph{undefined}.
2844 By analogy the word @emph{section} is used to describe groups of sections in
2845 the linked program. @code{@value{LD}} puts all partial programs' text
2846 sections in contiguous addresses in the linked program. It is
2847 customary to refer to the @emph{text section} of a program, meaning all
2848 the addresses of all partial programs' text sections. Likewise for
2849 data and bss sections.
2851 Some sections are manipulated by @code{@value{LD}}; others are invented for
2852 use of @command{@value{AS}} and have no meaning except during assembly.
2855 @section Linker Sections
2856 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2861 @cindex named sections
2862 @cindex sections, named
2863 @item named sections
2866 @cindex text section
2867 @cindex data section
2871 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2872 separate but equal sections. Anything you can say of one section is
2875 When the program is running, however, it is
2876 customary for the text section to be unalterable. The
2877 text section is often shared among processes: it contains
2878 instructions, constants and the like. The data section of a running
2879 program is usually alterable: for example, C variables would be stored
2880 in the data section.
2885 This section contains zeroed bytes when your program begins running. It
2886 is used to hold uninitialized variables or common storage. The length of
2887 each partial program's bss section is important, but because it starts
2888 out containing zeroed bytes there is no need to store explicit zero
2889 bytes in the object file. The bss section was invented to eliminate
2890 those explicit zeros from object files.
2892 @cindex absolute section
2893 @item absolute section
2894 Address 0 of this section is always ``relocated'' to runtime address 0.
2895 This is useful if you want to refer to an address that @code{@value{LD}} must
2896 not change when relocating. In this sense we speak of absolute
2897 addresses being ``unrelocatable'': they do not change during relocation.
2899 @cindex undefined section
2900 @item undefined section
2901 This ``section'' is a catch-all for address references to objects not in
2902 the preceding sections.
2903 @c FIXME: ref to some other doc on obj-file formats could go here.
2906 @cindex relocation example
2907 An idealized example of three relocatable sections follows.
2909 The example uses the traditional section names @samp{.text} and @samp{.data}.
2911 Memory addresses are on the horizontal axis.
2915 @c END TEXI2ROFF-KILL
2918 partial program # 1: |ttttt|dddd|00|
2925 partial program # 2: |TTT|DDD|000|
2928 +--+---+-----+--+----+---+-----+~~
2929 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2930 +--+---+-----+--+----+---+-----+~~
2932 addresses: 0 @dots{}
2939 \line{\it Partial program \#1: \hfil}
2940 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2941 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2943 \line{\it Partial program \#2: \hfil}
2944 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2945 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2947 \line{\it linked program: \hfil}
2948 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2949 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2950 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2951 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2953 \line{\it addresses: \hfil}
2957 @c END TEXI2ROFF-KILL
2960 @section Assembler Internal Sections
2962 @cindex internal assembler sections
2963 @cindex sections in messages, internal
2964 These sections are meant only for the internal use of @command{@value{AS}}. They
2965 have no meaning at run-time. You do not really need to know about these
2966 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2967 warning messages, so it might be helpful to have an idea of their
2968 meanings to @command{@value{AS}}. These sections are used to permit the
2969 value of every expression in your assembly language program to be a
2970 section-relative address.
2973 @cindex assembler internal logic error
2974 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2975 An internal assembler logic error has been found. This means there is a
2976 bug in the assembler.
2978 @cindex expr (internal section)
2980 The assembler stores complex expression internally as combinations of
2981 symbols. When it needs to represent an expression as a symbol, it puts
2982 it in the expr section.
2984 @c FIXME item transfer[t] vector preload
2985 @c FIXME item transfer[t] vector postload
2986 @c FIXME item register
2990 @section Sub-Sections
2992 @cindex numbered subsections
2993 @cindex grouping data
2999 fall into two sections: text and data.
3001 You may have separate groups of
3003 data in named sections
3007 data in named sections
3013 that you want to end up near to each other in the object file, even though they
3014 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3015 use @dfn{subsections} for this purpose. Within each section, there can be
3016 numbered subsections with values from 0 to 8192. Objects assembled into the
3017 same subsection go into the object file together with other objects in the same
3018 subsection. For example, a compiler might want to store constants in the text
3019 section, but might not want to have them interspersed with the program being
3020 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3021 section of code being output, and a @samp{.text 1} before each group of
3022 constants being output.
3024 Subsections are optional. If you do not use subsections, everything
3025 goes in subsection number zero.
3028 Each subsection is zero-padded up to a multiple of four bytes.
3029 (Subsections may be padded a different amount on different flavors
3030 of @command{@value{AS}}.)
3034 On the H8/300 platform, each subsection is zero-padded to a word
3035 boundary (two bytes).
3036 The same is true on the Renesas SH.
3039 @c FIXME section padding (alignment)?
3040 @c Rich Pixley says padding here depends on target obj code format; that
3041 @c doesn't seem particularly useful to say without further elaboration,
3042 @c so for now I say nothing about it. If this is a generic BFD issue,
3043 @c these paragraphs might need to vanish from this manual, and be
3044 @c discussed in BFD chapter of binutils (or some such).
3048 Subsections appear in your object file in numeric order, lowest numbered
3049 to highest. (All this to be compatible with other people's assemblers.)
3050 The object file contains no representation of subsections; @code{@value{LD}} and
3051 other programs that manipulate object files see no trace of them.
3052 They just see all your text subsections as a text section, and all your
3053 data subsections as a data section.
3055 To specify which subsection you want subsequent statements assembled
3056 into, use a numeric argument to specify it, in a @samp{.text
3057 @var{expression}} or a @samp{.data @var{expression}} statement.
3060 When generating COFF output, you
3065 can also use an extra subsection
3066 argument with arbitrary named sections: @samp{.section @var{name},
3071 When generating ELF output, you
3076 can also use the @code{.subsection} directive (@pxref{SubSection})
3077 to specify a subsection: @samp{.subsection @var{expression}}.
3079 @var{Expression} should be an absolute expression.
3080 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
3081 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3082 begins in @code{text 0}. For instance:
3084 .text 0 # The default subsection is text 0 anyway.
3085 .ascii "This lives in the first text subsection. *"
3087 .ascii "But this lives in the second text subsection."
3089 .ascii "This lives in the data section,"
3090 .ascii "in the first data subsection."
3092 .ascii "This lives in the first text section,"
3093 .ascii "immediately following the asterisk (*)."
3096 Each section has a @dfn{location counter} incremented by one for every byte
3097 assembled into that section. Because subsections are merely a convenience
3098 restricted to @command{@value{AS}} there is no concept of a subsection location
3099 counter. There is no way to directly manipulate a location counter---but the
3100 @code{.align} directive changes it, and any label definition captures its
3101 current value. The location counter of the section where statements are being
3102 assembled is said to be the @dfn{active} location counter.
3105 @section bss Section
3108 @cindex common variable storage
3109 The bss section is used for local common variable storage.
3110 You may allocate address space in the bss section, but you may
3111 not dictate data to load into it before your program executes. When
3112 your program starts running, all the contents of the bss
3113 section are zeroed bytes.
3115 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3116 @ref{Lcomm,,@code{.lcomm}}.
3118 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3119 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
3122 When assembling for a target which supports multiple sections, such as ELF or
3123 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3124 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3125 section. Typically the section will only contain symbol definitions and
3126 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3133 Symbols are a central concept: the programmer uses symbols to name
3134 things, the linker uses symbols to link, and the debugger uses symbols
3138 @cindex debuggers, and symbol order
3139 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3140 the same order they were declared. This may break some debuggers.
3145 * Setting Symbols:: Giving Symbols Other Values
3146 * Symbol Names:: Symbol Names
3147 * Dot:: The Special Dot Symbol
3148 * Symbol Attributes:: Symbol Attributes
3155 A @dfn{label} is written as a symbol immediately followed by a colon
3156 @samp{:}. The symbol then represents the current value of the
3157 active location counter, and is, for example, a suitable instruction
3158 operand. You are warned if you use the same symbol to represent two
3159 different locations: the first definition overrides any other
3163 On the HPPA, the usual form for a label need not be immediately followed by a
3164 colon, but instead must start in column zero. Only one label may be defined on
3165 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3166 provides a special directive @code{.label} for defining labels more flexibly.
3169 @node Setting Symbols
3170 @section Giving Symbols Other Values
3172 @cindex assigning values to symbols
3173 @cindex symbol values, assigning
3174 A symbol can be given an arbitrary value by writing a symbol, followed
3175 by an equals sign @samp{=}, followed by an expression
3176 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3177 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3178 equals sign @samp{=}@samp{=} here represents an equivalent of the
3179 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3182 @section Symbol Names
3184 @cindex symbol names
3185 @cindex names, symbol
3186 @ifclear SPECIAL-SYMS
3187 Symbol names begin with a letter or with one of @samp{._}. On most
3188 machines, you can also use @code{$} in symbol names; exceptions are
3189 noted in @ref{Machine Dependencies}. That character may be followed by any
3190 string of digits, letters, dollar signs (unless otherwise noted in
3191 @ref{Machine Dependencies}), and underscores.
3195 Symbol names begin with a letter or with one of @samp{._}. On the
3196 Renesas SH you can also use @code{$} in symbol names. That
3197 character may be followed by any string of digits, letters, dollar signs (save
3198 on the H8/300), and underscores.
3202 Case of letters is significant: @code{foo} is a different symbol name
3205 Each symbol has exactly one name. Each name in an assembly language program
3206 refers to exactly one symbol. You may use that symbol name any number of times
3209 @subheading Local Symbol Names
3211 @cindex local symbol names
3212 @cindex symbol names, local
3213 @cindex temporary symbol names
3214 @cindex symbol names, temporary
3215 Local symbols help compilers and programmers use names temporarily.
3216 They create symbols which are guaranteed to be unique over the entire scope of
3217 the input source code and which can be referred to by a simple notation.
3218 To define a local symbol, write a label of the form @samp{@b{N}:} (where @b{N}
3219 represents any positive integer). To refer to the most recent previous
3220 definition of that symbol write @samp{@b{N}b}, using the same number as when
3221 you defined the label. To refer to the next definition of a local label, write
3222 @samp{@b{N}f}--- The @samp{b} stands for``backwards'' and the @samp{f} stands
3225 There is no restriction on how you can use these labels, and you can reuse them
3226 too. So that it is possible to repeatedly define the same local label (using
3227 the same number @samp{@b{N}}), although you can only refer to the most recently
3228 defined local label of that number (for a backwards reference) or the next
3229 definition of a specific local label for a forward reference. It is also worth
3230 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3231 implemented in a slightly more efficient manner than the others.
3242 Which is the equivalent of:
3245 label_1: branch label_3
3246 label_2: branch label_1
3247 label_3: branch label_4
3248 label_4: branch label_3
3251 Local symbol names are only a notational device. They are immediately
3252 transformed into more conventional symbol names before the assembler uses them.
3253 The symbol names stored in the symbol table, appearing in error messages and
3254 optionally emitted to the object file. The names are constructed using these
3259 All local labels begin with @samp{L}. Normally both @command{@value{AS}} and
3260 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
3261 used for symbols you are never intended to see. If you use the
3262 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3263 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3264 you may use them in debugging.
3267 This is the number that was used in the local label definition. So if the
3268 label is written @samp{55:} then the number is @samp{55}.
3271 This unusual character is included so you do not accidentally invent a symbol
3272 of the same name. The character has ASCII value of @samp{\002} (control-B).
3274 @item @emph{ordinal number}
3275 This is a serial number to keep the labels distinct. The first definition of
3276 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3277 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3278 the number @samp{1} and its 15th defintion gets @samp{15} as well.
3281 So for example, the first @code{1:} is named @code{L1@kbd{C-B}1}, the 44th
3282 @code{3:} is named @code{L3@kbd{C-B}44}.
3284 @subheading Dollar Local Labels
3285 @cindex dollar local symbols
3287 @code{@value{AS}} also supports an even more local form of local labels called
3288 dollar labels. These labels go out of scope (ie they become undefined) as soon
3289 as a non-local label is defined. Thus they remain valid for only a small
3290 region of the input source code. Normal local labels, by contrast, remain in
3291 scope for the entire file, or until they are redefined by another occurrence of
3292 the same local label.
3294 Dollar labels are defined in exactly the same way as ordinary local labels,
3295 except that instead of being terminated by a colon, they are terminated by a
3296 dollar sign. eg @samp{@b{55$}}.
3298 They can also be distinguished from ordinary local labels by their transformed
3299 name which uses ASCII character @samp{\001} (control-A) as the magic character
3300 to distinguish them from ordinary labels. Thus the 5th defintion of @samp{6$}
3301 is named @samp{L6@kbd{C-A}5}.
3304 @section The Special Dot Symbol
3306 @cindex dot (symbol)
3307 @cindex @code{.} (symbol)
3308 @cindex current address
3309 @cindex location counter
3310 The special symbol @samp{.} refers to the current address that
3311 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3312 .long .} defines @code{melvin} to contain its own address.
3313 Assigning a value to @code{.} is treated the same as a @code{.org}
3314 directive. Thus, the expression @samp{.=.+4} is the same as saying
3315 @ifclear no-space-dir
3319 @node Symbol Attributes
3320 @section Symbol Attributes
3322 @cindex symbol attributes
3323 @cindex attributes, symbol
3324 Every symbol has, as well as its name, the attributes ``Value'' and
3325 ``Type''. Depending on output format, symbols can also have auxiliary
3328 The detailed definitions are in @file{a.out.h}.
3331 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3332 all these attributes, and probably won't warn you. This makes the
3333 symbol an externally defined symbol, which is generally what you
3337 * Symbol Value:: Value
3338 * Symbol Type:: Type
3341 * a.out Symbols:: Symbol Attributes: @code{a.out}
3345 * a.out Symbols:: Symbol Attributes: @code{a.out}
3348 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3353 * COFF Symbols:: Symbol Attributes for COFF
3356 * SOM Symbols:: Symbol Attributes for SOM
3363 @cindex value of a symbol
3364 @cindex symbol value
3365 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3366 location in the text, data, bss or absolute sections the value is the
3367 number of addresses from the start of that section to the label.
3368 Naturally for text, data and bss sections the value of a symbol changes
3369 as @code{@value{LD}} changes section base addresses during linking. Absolute
3370 symbols' values do not change during linking: that is why they are
3373 The value of an undefined symbol is treated in a special way. If it is
3374 0 then the symbol is not defined in this assembler source file, and
3375 @code{@value{LD}} tries to determine its value from other files linked into the
3376 same program. You make this kind of symbol simply by mentioning a symbol
3377 name without defining it. A non-zero value represents a @code{.comm}
3378 common declaration. The value is how much common storage to reserve, in
3379 bytes (addresses). The symbol refers to the first address of the
3385 @cindex type of a symbol
3387 The type attribute of a symbol contains relocation (section)
3388 information, any flag settings indicating that a symbol is external, and
3389 (optionally), other information for linkers and debuggers. The exact
3390 format depends on the object-code output format in use.
3395 @c The following avoids a "widow" subsection title. @group would be
3396 @c better if it were available outside examples.
3399 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3401 @cindex @code{b.out} symbol attributes
3402 @cindex symbol attributes, @code{b.out}
3403 These symbol attributes appear only when @command{@value{AS}} is configured for
3404 one of the Berkeley-descended object output formats---@code{a.out} or
3410 @subsection Symbol Attributes: @code{a.out}
3412 @cindex @code{a.out} symbol attributes
3413 @cindex symbol attributes, @code{a.out}
3419 @subsection Symbol Attributes: @code{a.out}
3421 @cindex @code{a.out} symbol attributes
3422 @cindex symbol attributes, @code{a.out}
3426 * Symbol Desc:: Descriptor
3427 * Symbol Other:: Other
3431 @subsubsection Descriptor
3433 @cindex descriptor, of @code{a.out} symbol
3434 This is an arbitrary 16-bit value. You may establish a symbol's
3435 descriptor value by using a @code{.desc} statement
3436 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3437 @command{@value{AS}}.
3440 @subsubsection Other
3442 @cindex other attribute, of @code{a.out} symbol
3443 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3448 @subsection Symbol Attributes for COFF
3450 @cindex COFF symbol attributes
3451 @cindex symbol attributes, COFF
3453 The COFF format supports a multitude of auxiliary symbol attributes;
3454 like the primary symbol attributes, they are set between @code{.def} and
3455 @code{.endef} directives.
3457 @subsubsection Primary Attributes
3459 @cindex primary attributes, COFF symbols
3460 The symbol name is set with @code{.def}; the value and type,
3461 respectively, with @code{.val} and @code{.type}.
3463 @subsubsection Auxiliary Attributes
3465 @cindex auxiliary attributes, COFF symbols
3466 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3467 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3468 table information for COFF.
3473 @subsection Symbol Attributes for SOM
3475 @cindex SOM symbol attributes
3476 @cindex symbol attributes, SOM
3478 The SOM format for the HPPA supports a multitude of symbol attributes set with
3479 the @code{.EXPORT} and @code{.IMPORT} directives.
3481 The attributes are described in @cite{HP9000 Series 800 Assembly
3482 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3483 @code{EXPORT} assembler directive documentation.
3487 @chapter Expressions
3491 @cindex numeric values
3492 An @dfn{expression} specifies an address or numeric value.
3493 Whitespace may precede and/or follow an expression.
3495 The result of an expression must be an absolute number, or else an offset into
3496 a particular section. If an expression is not absolute, and there is not
3497 enough information when @command{@value{AS}} sees the expression to know its
3498 section, a second pass over the source program might be necessary to interpret
3499 the expression---but the second pass is currently not implemented.
3500 @command{@value{AS}} aborts with an error message in this situation.
3503 * Empty Exprs:: Empty Expressions
3504 * Integer Exprs:: Integer Expressions
3508 @section Empty Expressions
3510 @cindex empty expressions
3511 @cindex expressions, empty
3512 An empty expression has no value: it is just whitespace or null.
3513 Wherever an absolute expression is required, you may omit the
3514 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3515 is compatible with other assemblers.
3518 @section Integer Expressions
3520 @cindex integer expressions
3521 @cindex expressions, integer
3522 An @dfn{integer expression} is one or more @emph{arguments} delimited
3523 by @emph{operators}.
3526 * Arguments:: Arguments
3527 * Operators:: Operators
3528 * Prefix Ops:: Prefix Operators
3529 * Infix Ops:: Infix Operators
3533 @subsection Arguments
3535 @cindex expression arguments
3536 @cindex arguments in expressions
3537 @cindex operands in expressions
3538 @cindex arithmetic operands
3539 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3540 contexts arguments are sometimes called ``arithmetic operands''. In
3541 this manual, to avoid confusing them with the ``instruction operands'' of
3542 the machine language, we use the term ``argument'' to refer to parts of
3543 expressions only, reserving the word ``operand'' to refer only to machine
3544 instruction operands.
3546 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3547 @var{section} is one of text, data, bss, absolute,
3548 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3551 Numbers are usually integers.
3553 A number can be a flonum or bignum. In this case, you are warned
3554 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3555 these 32 bits are an integer. You may write integer-manipulating
3556 instructions that act on exotic constants, compatible with other
3559 @cindex subexpressions
3560 Subexpressions are a left parenthesis @samp{(} followed by an integer
3561 expression, followed by a right parenthesis @samp{)}; or a prefix
3562 operator followed by an argument.
3565 @subsection Operators
3567 @cindex operators, in expressions
3568 @cindex arithmetic functions
3569 @cindex functions, in expressions
3570 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3571 operators are followed by an argument. Infix operators appear
3572 between their arguments. Operators may be preceded and/or followed by
3576 @subsection Prefix Operator
3578 @cindex prefix operators
3579 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3580 one argument, which must be absolute.
3582 @c the tex/end tex stuff surrounding this small table is meant to make
3583 @c it align, on the printed page, with the similar table in the next
3584 @c section (which is inside an enumerate).
3586 \global\advance\leftskip by \itemindent
3591 @dfn{Negation}. Two's complement negation.
3593 @dfn{Complementation}. Bitwise not.
3597 \global\advance\leftskip by -\itemindent
3601 @subsection Infix Operators
3603 @cindex infix operators
3604 @cindex operators, permitted arguments
3605 @dfn{Infix operators} take two arguments, one on either side. Operators
3606 have precedence, but operations with equal precedence are performed left
3607 to right. Apart from @code{+} or @option{-}, both arguments must be
3608 absolute, and the result is absolute.
3611 @cindex operator precedence
3612 @cindex precedence of operators
3619 @dfn{Multiplication}.
3622 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3628 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3631 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3635 Intermediate precedence
3640 @dfn{Bitwise Inclusive Or}.
3646 @dfn{Bitwise Exclusive Or}.
3649 @dfn{Bitwise Or Not}.
3656 @cindex addition, permitted arguments
3657 @cindex plus, permitted arguments
3658 @cindex arguments for addition
3660 @dfn{Addition}. If either argument is absolute, the result has the section of
3661 the other argument. You may not add together arguments from different
3664 @cindex subtraction, permitted arguments
3665 @cindex minus, permitted arguments
3666 @cindex arguments for subtraction
3668 @dfn{Subtraction}. If the right argument is absolute, the
3669 result has the section of the left argument.
3670 If both arguments are in the same section, the result is absolute.
3671 You may not subtract arguments from different sections.
3672 @c FIXME is there still something useful to say about undefined - undefined ?
3674 @cindex comparison expressions
3675 @cindex expressions, comparison
3680 @dfn{Is Not Equal To}
3684 @dfn{Is Greater Than}
3686 @dfn{Is Greater Than Or Equal To}
3688 @dfn{Is Less Than Or Equal To}
3690 The comparison operators can be used as infix operators. A true results has a
3691 value of -1 whereas a false result has a value of 0. Note, these operators
3692 perform signed comparisons.
3695 @item Lowest Precedence
3704 These two logical operations can be used to combine the results of sub
3705 expressions. Note, unlike the comparison operators a true result returns a
3706 value of 1 but a false results does still return 0. Also note that the logical
3707 or operator has a slightly lower precedence than logical and.
3712 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3713 address; you can only have a defined section in one of the two arguments.
3716 @chapter Assembler Directives
3718 @cindex directives, machine independent
3719 @cindex pseudo-ops, machine independent
3720 @cindex machine independent directives
3721 All assembler directives have names that begin with a period (@samp{.}).
3722 The rest of the name is letters, usually in lower case.
3724 This chapter discusses directives that are available regardless of the
3725 target machine configuration for the @sc{gnu} assembler.
3727 Some machine configurations provide additional directives.
3728 @xref{Machine Dependencies}.
3731 @ifset machine-directives
3732 @xref{Machine Dependencies} for additional directives.
3737 * Abort:: @code{.abort}
3739 * ABORT (COFF):: @code{.ABORT}
3742 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3743 * Altmacro:: @code{.altmacro}
3744 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3745 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3746 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3747 * Byte:: @code{.byte @var{expressions}}
3748 * Comm:: @code{.comm @var{symbol} , @var{length} }
3750 * CFI directives:: @code{.cfi_startproc}, @code{.cfi_endproc}, etc.
3752 * Data:: @code{.data @var{subsection}}
3754 * Def:: @code{.def @var{name}}
3757 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3763 * Double:: @code{.double @var{flonums}}
3764 * Eject:: @code{.eject}
3765 * Else:: @code{.else}
3766 * Elseif:: @code{.elseif}
3769 * Endef:: @code{.endef}
3772 * Endfunc:: @code{.endfunc}
3773 * Endif:: @code{.endif}
3774 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3775 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3776 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
3778 * Error:: @code{.error @var{string}}
3779 * Exitm:: @code{.exitm}
3780 * Extern:: @code{.extern}
3781 * Fail:: @code{.fail}
3782 @ifclear no-file-dir
3783 * File:: @code{.file @var{string}}
3786 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3787 * Float:: @code{.float @var{flonums}}
3788 * Func:: @code{.func}
3789 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3791 * Hidden:: @code{.hidden @var{names}}
3794 * hword:: @code{.hword @var{expressions}}
3795 * Ident:: @code{.ident}
3796 * If:: @code{.if @var{absolute expression}}
3797 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3798 * Include:: @code{.include "@var{file}"}
3799 * Int:: @code{.int @var{expressions}}
3801 * Internal:: @code{.internal @var{names}}
3804 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3805 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3806 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3807 * Lflags:: @code{.lflags}
3808 @ifclear no-line-dir
3809 * Line:: @code{.line @var{line-number}}
3812 * Linkonce:: @code{.linkonce [@var{type}]}
3813 * List:: @code{.list}
3814 * Ln:: @code{.ln @var{line-number}}
3816 * LNS directives:: @code{.file}, @code{.loc}, etc.
3818 * Long:: @code{.long @var{expressions}}
3820 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3823 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3824 * MRI:: @code{.mri @var{val}}
3825 * Noaltmacro:: @code{.noaltmacro}
3826 * Nolist:: @code{.nolist}
3827 * Octa:: @code{.octa @var{bignums}}
3828 * Org:: @code{.org @var{new-lc} , @var{fill}}
3829 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3831 * PopSection:: @code{.popsection}
3832 * Previous:: @code{.previous}
3835 * Print:: @code{.print @var{string}}
3837 * Protected:: @code{.protected @var{names}}
3840 * Psize:: @code{.psize @var{lines}, @var{columns}}
3841 * Purgem:: @code{.purgem @var{name}}
3843 * PushSection:: @code{.pushsection @var{name}}
3846 * Quad:: @code{.quad @var{bignums}}
3847 * Rept:: @code{.rept @var{count}}
3848 * Sbttl:: @code{.sbttl "@var{subheading}"}
3850 * Scl:: @code{.scl @var{class}}
3853 * Section:: @code{.section @var{name}}
3856 * Set:: @code{.set @var{symbol}, @var{expression}}
3857 * Short:: @code{.short @var{expressions}}
3858 * Single:: @code{.single @var{flonums}}
3860 * Size:: @code{.size [@var{name} , @var{expression}]}
3863 * Skip:: @code{.skip @var{size} , @var{fill}}
3864 * Sleb128:: @code{.sleb128 @var{expressions}}
3865 * Space:: @code{.space @var{size} , @var{fill}}
3867 * Stab:: @code{.stabd, .stabn, .stabs}
3870 * String:: @code{.string "@var{str}"}
3871 * Struct:: @code{.struct @var{expression}}
3873 * SubSection:: @code{.subsection}
3874 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3878 * Tag:: @code{.tag @var{structname}}
3881 * Text:: @code{.text @var{subsection}}
3882 * Title:: @code{.title "@var{heading}"}
3884 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3887 * Uleb128:: @code{.uleb128 @var{expressions}}
3889 * Val:: @code{.val @var{addr}}
3893 * Version:: @code{.version "@var{string}"}
3894 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3895 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3898 * Warning:: @code{.warning @var{string}}
3899 * Weak:: @code{.weak @var{names}}
3900 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
3901 * Word:: @code{.word @var{expressions}}
3902 * Deprecated:: Deprecated Directives
3906 @section @code{.abort}
3908 @cindex @code{abort} directive
3909 @cindex stopping the assembly
3910 This directive stops the assembly immediately. It is for
3911 compatibility with other assemblers. The original idea was that the
3912 assembly language source would be piped into the assembler. If the sender
3913 of the source quit, it could use this directive tells @command{@value{AS}} to
3914 quit also. One day @code{.abort} will not be supported.
3918 @section @code{.ABORT} (COFF)
3920 @cindex @code{ABORT} directive
3921 When producing COFF output, @command{@value{AS}} accepts this directive as a
3922 synonym for @samp{.abort}.
3925 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3931 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3933 @cindex padding the location counter
3934 @cindex @code{align} directive
3935 Pad the location counter (in the current subsection) to a particular storage
3936 boundary. The first expression (which must be absolute) is the alignment
3937 required, as described below.
3939 The second expression (also absolute) gives the fill value to be stored in the
3940 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3941 padding bytes are normally zero. However, on some systems, if the section is
3942 marked as containing code and the fill value is omitted, the space is filled
3943 with no-op instructions.
3945 The third expression is also absolute, and is also optional. If it is present,
3946 it is the maximum number of bytes that should be skipped by this alignment
3947 directive. If doing the alignment would require skipping more bytes than the
3948 specified maximum, then the alignment is not done at all. You can omit the
3949 fill value (the second argument) entirely by simply using two commas after the
3950 required alignment; this can be useful if you want the alignment to be filled
3951 with no-op instructions when appropriate.
3953 The way the required alignment is specified varies from system to system.
3954 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
3955 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
3956 alignment request in bytes. For example @samp{.align 8} advances
3957 the location counter until it is a multiple of 8. If the location counter
3958 is already a multiple of 8, no change is needed. For the tic54x, the
3959 first expression is the alignment request in words.
3961 For other systems, including the i386 using a.out format, and the arm and
3962 strongarm, it is the
3963 number of low-order zero bits the location counter must have after
3964 advancement. For example @samp{.align 3} advances the location
3965 counter until it a multiple of 8. If the location counter is already a
3966 multiple of 8, no change is needed.
3968 This inconsistency is due to the different behaviors of the various
3969 native assemblers for these systems which GAS must emulate.
3970 GAS also provides @code{.balign} and @code{.p2align} directives,
3971 described later, which have a consistent behavior across all
3972 architectures (but are specific to GAS).
3975 @section @code{.ascii "@var{string}"}@dots{}
3977 @cindex @code{ascii} directive
3978 @cindex string literals
3979 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3980 separated by commas. It assembles each string (with no automatic
3981 trailing zero byte) into consecutive addresses.
3984 @section @code{.asciz "@var{string}"}@dots{}
3986 @cindex @code{asciz} directive
3987 @cindex zero-terminated strings
3988 @cindex null-terminated strings
3989 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3990 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3993 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3995 @cindex padding the location counter given number of bytes
3996 @cindex @code{balign} directive
3997 Pad the location counter (in the current subsection) to a particular
3998 storage boundary. The first expression (which must be absolute) is the
3999 alignment request in bytes. For example @samp{.balign 8} advances
4000 the location counter until it is a multiple of 8. If the location counter
4001 is already a multiple of 8, no change is needed.
4003 The second expression (also absolute) gives the fill value to be stored in the
4004 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4005 padding bytes are normally zero. However, on some systems, if the section is
4006 marked as containing code and the fill value is omitted, the space is filled
4007 with no-op instructions.
4009 The third expression is also absolute, and is also optional. If it is present,
4010 it is the maximum number of bytes that should be skipped by this alignment
4011 directive. If doing the alignment would require skipping more bytes than the
4012 specified maximum, then the alignment is not done at all. You can omit the
4013 fill value (the second argument) entirely by simply using two commas after the
4014 required alignment; this can be useful if you want the alignment to be filled
4015 with no-op instructions when appropriate.
4017 @cindex @code{balignw} directive
4018 @cindex @code{balignl} directive
4019 The @code{.balignw} and @code{.balignl} directives are variants of the
4020 @code{.balign} directive. The @code{.balignw} directive treats the fill
4021 pattern as a two byte word value. The @code{.balignl} directives treats the
4022 fill pattern as a four byte longword value. For example, @code{.balignw
4023 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4024 filled in with the value 0x368d (the exact placement of the bytes depends upon
4025 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4029 @section @code{.byte @var{expressions}}
4031 @cindex @code{byte} directive
4032 @cindex integers, one byte
4033 @code{.byte} expects zero or more expressions, separated by commas.
4034 Each expression is assembled into the next byte.
4037 @section @code{.comm @var{symbol} , @var{length} }
4039 @cindex @code{comm} directive
4040 @cindex symbol, common
4041 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4042 common symbol in one object file may be merged with a defined or common symbol
4043 of the same name in another object file. If @code{@value{LD}} does not see a
4044 definition for the symbol--just one or more common symbols--then it will
4045 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4046 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4047 the same name, and they do not all have the same size, it will allocate space
4048 using the largest size.
4051 When using ELF, the @code{.comm} directive takes an optional third argument.
4052 This is the desired alignment of the symbol, specified as a byte boundary (for
4053 example, an alignment of 16 means that the least significant 4 bits of the
4054 address should be zero). The alignment must be an absolute expression, and it
4055 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
4056 for the common symbol, it will use the alignment when placing the symbol. If
4057 no alignment is specified, @command{@value{AS}} will set the alignment to the
4058 largest power of two less than or equal to the size of the symbol, up to a
4063 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4064 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4067 @node CFI directives
4068 @section @code{.cfi_startproc}
4069 @cindex @code{cfi_startproc} directive
4070 @code{.cfi_startproc} is used at the beginning of each function that
4071 should have an entry in @code{.eh_frame}. It initializes some internal
4072 data structures and emits architecture dependent initial CFI instructions.
4073 Don't forget to close the function by
4074 @code{.cfi_endproc}.
4076 @section @code{.cfi_endproc}
4077 @cindex @code{cfi_endproc} directive
4078 @code{.cfi_endproc} is used at the end of a function where it closes its
4079 unwind entry previously opened by
4080 @code{.cfi_startproc}. and emits it to @code{.eh_frame}.
4082 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4083 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4084 address from @var{register} and add @var{offset} to it}.
4086 @section @code{.cfi_def_cfa_register @var{register}}
4087 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4088 now on @var{register} will be used instead of the old one. Offset
4091 @section @code{.cfi_def_cfa_offset @var{offset}}
4092 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4093 remains the same, but @var{offset} is new. Note that it is the
4094 absolute offset that will be added to a defined register to compute
4097 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4098 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4099 value that is added/substracted from the previous offset.
4101 @section @code{.cfi_offset @var{register}, @var{offset}}
4102 Previous value of @var{register} is saved at offset @var{offset} from
4105 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4106 Previous value of @var{register} is saved at offset @var{offset} from
4107 the current CFA register. This is transformed to @code{.cfi_offset}
4108 using the known displacement of the CFA register from the CFA.
4109 This is often easier to use, because the number will match the
4110 code it's annotating.
4112 @section @code{.cfi_signal_frame}
4113 Mark current function as signal trampoline.
4115 @section @code{.cfi_window_save}
4116 SPARC register window has been saved.
4118 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4119 Allows the user to add arbitrary bytes to the unwind info. One
4120 might use this to add OS-specific CFI opcodes, or generic CFI
4121 opcodes that GAS does not yet support.
4123 @node LNS directives
4124 @section @code{.file @var{fileno} @var{filename}}
4125 @cindex @code{file} directive
4126 When emitting dwarf2 line number information @code{.file} assigns filenames
4127 to the @code{.debug_line} file name table. The @var{fileno} operand should
4128 be a unique positive integer to use as the index of the entry in the table.
4129 The @var{filename} operand is a C string literal.
4131 The detail of filename indicies is exposed to the user because the filename
4132 table is shared with the @code{.debug_info} section of the dwarf2 debugging
4133 information, and thus the user must know the exact indicies that table
4136 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
4137 @cindex @code{loc} directive
4138 The @code{.loc} directive will add row to the @code{.debug_line} line
4139 number matrix corresponding to the immediately following assembly
4140 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
4141 arguments will be applied to the @code{.debug_line} state machine before
4144 The @var{options} are a sequence of the following tokens in any order:
4148 This option will set the @code{basic_block} register in the
4149 @code{.debug_line} state machine to @code{true}.
4152 This option will set the @code{prologue_end} register in the
4153 @code{.debug_line} state machine to @code{true}.
4155 @item epilogue_begin
4156 This option will set the @code{epilogue_begin} register in the
4157 @code{.debug_line} state machine to @code{true}.
4159 @item is_stmt @var{value}
4160 This option will set the @code{is_stmt} register in the
4161 @code{.debug_line} state machine to @code{value}, which must be
4164 @item isa @var{value}
4165 This directive will set the @code{isa} register in the @code{.debug_line}
4166 state machine to @var{value}, which must be an unsigned integer.
4170 @section @code{.loc_mark_blocks @var{enable}}
4171 @cindex @code{loc_mark_blocks} directive
4172 The @code{.loc_mark_blocks} directive makes the assembler emit an entry
4173 to the @code{.debug_line} line number matrix with the @code{basic_block}
4174 register in the state machine set whenever a code label is seen.
4175 The @var{enable} argument should be either 1 or 0, to enable or disable
4176 this function respectively.
4179 @section @code{.data @var{subsection}}
4181 @cindex @code{data} directive
4182 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4183 end of the data subsection numbered @var{subsection} (which is an
4184 absolute expression). If @var{subsection} is omitted, it defaults
4189 @section @code{.def @var{name}}
4191 @cindex @code{def} directive
4192 @cindex COFF symbols, debugging
4193 @cindex debugging COFF symbols
4194 Begin defining debugging information for a symbol @var{name}; the
4195 definition extends until the @code{.endef} directive is encountered.
4198 This directive is only observed when @command{@value{AS}} is configured for COFF
4199 format output; when producing @code{b.out}, @samp{.def} is recognized,
4206 @section @code{.desc @var{symbol}, @var{abs-expression}}
4208 @cindex @code{desc} directive
4209 @cindex COFF symbol descriptor
4210 @cindex symbol descriptor, COFF
4211 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4212 to the low 16 bits of an absolute expression.
4215 The @samp{.desc} directive is not available when @command{@value{AS}} is
4216 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4217 object format. For the sake of compatibility, @command{@value{AS}} accepts
4218 it, but produces no output, when configured for COFF.
4224 @section @code{.dim}
4226 @cindex @code{dim} directive
4227 @cindex COFF auxiliary symbol information
4228 @cindex auxiliary symbol information, COFF
4229 This directive is generated by compilers to include auxiliary debugging
4230 information in the symbol table. It is only permitted inside
4231 @code{.def}/@code{.endef} pairs.
4234 @samp{.dim} is only meaningful when generating COFF format output; when
4235 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4241 @section @code{.double @var{flonums}}
4243 @cindex @code{double} directive
4244 @cindex floating point numbers (double)
4245 @code{.double} expects zero or more flonums, separated by commas. It
4246 assembles floating point numbers.
4248 The exact kind of floating point numbers emitted depends on how
4249 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4253 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4254 in @sc{ieee} format.
4259 @section @code{.eject}
4261 @cindex @code{eject} directive
4262 @cindex new page, in listings
4263 @cindex page, in listings
4264 @cindex listing control: new page
4265 Force a page break at this point, when generating assembly listings.
4268 @section @code{.else}
4270 @cindex @code{else} directive
4271 @code{.else} is part of the @command{@value{AS}} support for conditional
4272 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
4273 of code to be assembled if the condition for the preceding @code{.if}
4277 @section @code{.elseif}
4279 @cindex @code{elseif} directive
4280 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4281 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
4282 @code{.if} block that would otherwise fill the entire @code{.else} section.
4285 @section @code{.end}
4287 @cindex @code{end} directive
4288 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4289 process anything in the file past the @code{.end} directive.
4293 @section @code{.endef}
4295 @cindex @code{endef} directive
4296 This directive flags the end of a symbol definition begun with
4300 @samp{.endef} is only meaningful when generating COFF format output; if
4301 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4302 directive but ignores it.
4307 @section @code{.endfunc}
4308 @cindex @code{endfunc} directive
4309 @code{.endfunc} marks the end of a function specified with @code{.func}.
4312 @section @code{.endif}
4314 @cindex @code{endif} directive
4315 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4316 it marks the end of a block of code that is only assembled
4317 conditionally. @xref{If,,@code{.if}}.
4320 @section @code{.equ @var{symbol}, @var{expression}}
4322 @cindex @code{equ} directive
4323 @cindex assigning values to symbols
4324 @cindex symbols, assigning values to
4325 This directive sets the value of @var{symbol} to @var{expression}.
4326 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
4329 The syntax for @code{equ} on the HPPA is
4330 @samp{@var{symbol} .equ @var{expression}}.
4334 The syntax for @code{equ} on the Z80 is
4335 @samp{@var{symbol} equ @var{expression}}.
4336 On the Z80 it is an eror if @var{symbol} is already defined,
4337 but the symbol is not protected from later redefinition,
4338 compare @xref{Equiv}.
4342 @section @code{.equiv @var{symbol}, @var{expression}}
4343 @cindex @code{equiv} directive
4344 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4345 the assembler will signal an error if @var{symbol} is already defined. Note a
4346 symbol which has been referenced but not actually defined is considered to be
4349 Except for the contents of the error message, this is roughly equivalent to
4356 plus it protects the symbol from later redefinition.
4359 @section @code{.eqv @var{symbol}, @var{expression}}
4360 @cindex @code{eqv} directive
4361 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4362 evaluate the expression or any part of it immediately. Instead each time
4363 the resulting symbol is used in an expression, a snapshot of its current
4367 @section @code{.err}
4368 @cindex @code{err} directive
4369 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4370 message and, unless the @option{-Z} option was used, it will not generate an
4371 object file. This can be used to signal an error in conditionally compiled code.
4374 @section @code{.error "@var{string}"}
4375 @cindex error directive
4377 Similarly to @code{.err}, this directive emits an error, but you can specify a
4378 string that will be emitted as the error message. If you don't specify the
4379 message, it defaults to @code{".error directive invoked in source file"}.
4380 @xref{Errors, ,Error and Warning Messages}.
4383 .error "This code has not been assembled and tested."
4387 @section @code{.exitm}
4388 Exit early from the current macro definition. @xref{Macro}.
4391 @section @code{.extern}
4393 @cindex @code{extern} directive
4394 @code{.extern} is accepted in the source program---for compatibility
4395 with other assemblers---but it is ignored. @command{@value{AS}} treats
4396 all undefined symbols as external.
4399 @section @code{.fail @var{expression}}
4401 @cindex @code{fail} directive
4402 Generates an error or a warning. If the value of the @var{expression} is 500
4403 or more, @command{@value{AS}} will print a warning message. If the value is less
4404 than 500, @command{@value{AS}} will print an error message. The message will
4405 include the value of @var{expression}. This can occasionally be useful inside
4406 complex nested macros or conditional assembly.
4408 @ifclear no-file-dir
4410 @section @code{.file @var{string}}
4412 @cindex @code{file} directive
4413 @cindex logical file name
4414 @cindex file name, logical
4415 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4416 file. @var{string} is the new file name. In general, the filename is
4417 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4418 to specify an empty file name, you must give the quotes--@code{""}. This
4419 statement may go away in future: it is only recognized to be compatible with
4420 old @command{@value{AS}} programs.
4424 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4426 @cindex @code{fill} directive
4427 @cindex writing patterns in memory
4428 @cindex patterns, writing in memory
4429 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4430 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4431 may be zero or more. @var{Size} may be zero or more, but if it is
4432 more than 8, then it is deemed to have the value 8, compatible with
4433 other people's assemblers. The contents of each @var{repeat} bytes
4434 is taken from an 8-byte number. The highest order 4 bytes are
4435 zero. The lowest order 4 bytes are @var{value} rendered in the
4436 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4437 Each @var{size} bytes in a repetition is taken from the lowest order
4438 @var{size} bytes of this number. Again, this bizarre behavior is
4439 compatible with other people's assemblers.
4441 @var{size} and @var{value} are optional.
4442 If the second comma and @var{value} are absent, @var{value} is
4443 assumed zero. If the first comma and following tokens are absent,
4444 @var{size} is assumed to be 1.
4447 @section @code{.float @var{flonums}}
4449 @cindex floating point numbers (single)
4450 @cindex @code{float} directive
4451 This directive assembles zero or more flonums, separated by commas. It
4452 has the same effect as @code{.single}.
4454 The exact kind of floating point numbers emitted depends on how
4455 @command{@value{AS}} is configured.
4456 @xref{Machine Dependencies}.
4460 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4461 in @sc{ieee} format.
4466 @section @code{.func @var{name}[,@var{label}]}
4467 @cindex @code{func} directive
4468 @code{.func} emits debugging information to denote function @var{name}, and
4469 is ignored unless the file is assembled with debugging enabled.
4470 Only @samp{--gstabs[+]} is currently supported.
4471 @var{label} is the entry point of the function and if omitted @var{name}
4472 prepended with the @samp{leading char} is used.
4473 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4474 All functions are currently defined to have @code{void} return type.
4475 The function must be terminated with @code{.endfunc}.
4478 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4480 @cindex @code{global} directive
4481 @cindex symbol, making visible to linker
4482 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4483 @var{symbol} in your partial program, its value is made available to
4484 other partial programs that are linked with it. Otherwise,
4485 @var{symbol} takes its attributes from a symbol of the same name
4486 from another file linked into the same program.
4488 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4489 compatibility with other assemblers.
4492 On the HPPA, @code{.global} is not always enough to make it accessible to other
4493 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4494 @xref{HPPA Directives,, HPPA Assembler Directives}.
4499 @section @code{.hidden @var{names}}
4501 @cindex @code{hidden} directive
4503 This is one of the ELF visibility directives. The other two are
4504 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4505 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4507 This directive overrides the named symbols default visibility (which is set by
4508 their binding: local, global or weak). The directive sets the visibility to
4509 @code{hidden} which means that the symbols are not visible to other components.
4510 Such symbols are always considered to be @code{protected} as well.
4514 @section @code{.hword @var{expressions}}
4516 @cindex @code{hword} directive
4517 @cindex integers, 16-bit
4518 @cindex numbers, 16-bit
4519 @cindex sixteen bit integers
4520 This expects zero or more @var{expressions}, and emits
4521 a 16 bit number for each.
4524 This directive is a synonym for @samp{.short}; depending on the target
4525 architecture, it may also be a synonym for @samp{.word}.
4529 This directive is a synonym for @samp{.short}.
4532 This directive is a synonym for both @samp{.short} and @samp{.word}.
4537 @section @code{.ident}
4539 @cindex @code{ident} directive
4541 This directive is used by some assemblers to place tags in object files. The
4542 behavior of this directive varies depending on the target. When using the
4543 a.out object file format, @command{@value{AS}} simply accepts the directive for
4544 source-file compatibility with existing assemblers, but does not emit anything
4545 for it. When using COFF, comments are emitted to the @code{.comment} or
4546 @code{.rdata} section, depending on the target. When using ELF, comments are
4547 emitted to the @code{.comment} section.
4550 @section @code{.if @var{absolute expression}}
4552 @cindex conditional assembly
4553 @cindex @code{if} directive
4554 @code{.if} marks the beginning of a section of code which is only
4555 considered part of the source program being assembled if the argument
4556 (which must be an @var{absolute expression}) is non-zero. The end of
4557 the conditional section of code must be marked by @code{.endif}
4558 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4559 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4560 If you have several conditions to check, @code{.elseif} may be used to avoid
4561 nesting blocks if/else within each subsequent @code{.else} block.
4563 The following variants of @code{.if} are also supported:
4565 @cindex @code{ifdef} directive
4566 @item .ifdef @var{symbol}
4567 Assembles the following section of code if the specified @var{symbol}
4568 has been defined. Note a symbol which has been referenced but not yet defined
4569 is considered to be undefined.
4571 @cindex @code{ifb} directive
4572 @item .ifb @var{text}
4573 Assembles the following section of code if the operand is blank (empty).
4575 @cindex @code{ifc} directive
4576 @item .ifc @var{string1},@var{string2}
4577 Assembles the following section of code if the two strings are the same. The
4578 strings may be optionally quoted with single quotes. If they are not quoted,
4579 the first string stops at the first comma, and the second string stops at the
4580 end of the line. Strings which contain whitespace should be quoted. The
4581 string comparison is case sensitive.
4583 @cindex @code{ifeq} directive
4584 @item .ifeq @var{absolute expression}
4585 Assembles the following section of code if the argument is zero.
4587 @cindex @code{ifeqs} directive
4588 @item .ifeqs @var{string1},@var{string2}
4589 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4591 @cindex @code{ifge} directive
4592 @item .ifge @var{absolute expression}
4593 Assembles the following section of code if the argument is greater than or
4596 @cindex @code{ifgt} directive
4597 @item .ifgt @var{absolute expression}
4598 Assembles the following section of code if the argument is greater than zero.
4600 @cindex @code{ifle} directive
4601 @item .ifle @var{absolute expression}
4602 Assembles the following section of code if the argument is less than or equal
4605 @cindex @code{iflt} directive
4606 @item .iflt @var{absolute expression}
4607 Assembles the following section of code if the argument is less than zero.
4609 @cindex @code{ifnb} directive
4610 @item .ifnb @var{text}
4611 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
4612 following section of code if the operand is non-blank (non-empty).
4614 @cindex @code{ifnc} directive
4615 @item .ifnc @var{string1},@var{string2}.
4616 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4617 following section of code if the two strings are not the same.
4619 @cindex @code{ifndef} directive
4620 @cindex @code{ifnotdef} directive
4621 @item .ifndef @var{symbol}
4622 @itemx .ifnotdef @var{symbol}
4623 Assembles the following section of code if the specified @var{symbol}
4624 has not been defined. Both spelling variants are equivalent. Note a symbol
4625 which has been referenced but not yet defined is considered to be undefined.
4627 @cindex @code{ifne} directive
4628 @item .ifne @var{absolute expression}
4629 Assembles the following section of code if the argument is not equal to zero
4630 (in other words, this is equivalent to @code{.if}).
4632 @cindex @code{ifnes} directive
4633 @item .ifnes @var{string1},@var{string2}
4634 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4635 following section of code if the two strings are not the same.
4639 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4641 @cindex @code{incbin} directive
4642 @cindex binary files, including
4643 The @code{incbin} directive includes @var{file} verbatim at the current
4644 location. You can control the search paths used with the @samp{-I} command-line
4645 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4648 The @var{skip} argument skips a number of bytes from the start of the
4649 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4650 read. Note that the data is not aligned in any way, so it is the user's
4651 responsibility to make sure that proper alignment is provided both before and
4652 after the @code{incbin} directive.
4655 @section @code{.include "@var{file}"}
4657 @cindex @code{include} directive
4658 @cindex supporting files, including
4659 @cindex files, including
4660 This directive provides a way to include supporting files at specified
4661 points in your source program. The code from @var{file} is assembled as
4662 if it followed the point of the @code{.include}; when the end of the
4663 included file is reached, assembly of the original file continues. You
4664 can control the search paths used with the @samp{-I} command-line option
4665 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4669 @section @code{.int @var{expressions}}
4671 @cindex @code{int} directive
4672 @cindex integers, 32-bit
4673 Expect zero or more @var{expressions}, of any section, separated by commas.
4674 For each expression, emit a number that, at run time, is the value of that
4675 expression. The byte order and bit size of the number depends on what kind
4676 of target the assembly is for.
4680 On most forms of the H8/300, @code{.int} emits 16-bit
4681 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4688 @section @code{.internal @var{names}}
4690 @cindex @code{internal} directive
4692 This is one of the ELF visibility directives. The other two are
4693 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4694 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4696 This directive overrides the named symbols default visibility (which is set by
4697 their binding: local, global or weak). The directive sets the visibility to
4698 @code{internal} which means that the symbols are considered to be @code{hidden}
4699 (i.e., not visible to other components), and that some extra, processor specific
4700 processing must also be performed upon the symbols as well.
4704 @section @code{.irp @var{symbol},@var{values}}@dots{}
4706 @cindex @code{irp} directive
4707 Evaluate a sequence of statements assigning different values to @var{symbol}.
4708 The sequence of statements starts at the @code{.irp} directive, and is
4709 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4710 set to @var{value}, and the sequence of statements is assembled. If no
4711 @var{value} is listed, the sequence of statements is assembled once, with
4712 @var{symbol} set to the null string. To refer to @var{symbol} within the
4713 sequence of statements, use @var{\symbol}.
4715 For example, assembling
4723 is equivalent to assembling
4731 For some caveats with the spelling of @var{symbol}, see also the discussion
4735 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4737 @cindex @code{irpc} directive
4738 Evaluate a sequence of statements assigning different values to @var{symbol}.
4739 The sequence of statements starts at the @code{.irpc} directive, and is
4740 terminated by an @code{.endr} directive. For each character in @var{value},
4741 @var{symbol} is set to the character, and the sequence of statements is
4742 assembled. If no @var{value} is listed, the sequence of statements is
4743 assembled once, with @var{symbol} set to the null string. To refer to
4744 @var{symbol} within the sequence of statements, use @var{\symbol}.
4746 For example, assembling
4754 is equivalent to assembling
4762 For some caveats with the spelling of @var{symbol}, see also the discussion
4766 @section @code{.lcomm @var{symbol} , @var{length}}
4768 @cindex @code{lcomm} directive
4769 @cindex local common symbols
4770 @cindex symbols, local common
4771 Reserve @var{length} (an absolute expression) bytes for a local common
4772 denoted by @var{symbol}. The section and value of @var{symbol} are
4773 those of the new local common. The addresses are allocated in the bss
4774 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4775 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4776 not visible to @code{@value{LD}}.
4779 Some targets permit a third argument to be used with @code{.lcomm}. This
4780 argument specifies the desired alignment of the symbol in the bss section.
4784 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4785 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4789 @section @code{.lflags}
4791 @cindex @code{lflags} directive (ignored)
4792 @command{@value{AS}} accepts this directive, for compatibility with other
4793 assemblers, but ignores it.
4795 @ifclear no-line-dir
4797 @section @code{.line @var{line-number}}
4799 @cindex @code{line} directive
4803 @section @code{.ln @var{line-number}}
4805 @cindex @code{ln} directive
4807 @cindex logical line number
4809 Change the logical line number. @var{line-number} must be an absolute
4810 expression. The next line has that logical line number. Therefore any other
4811 statements on the current line (after a statement separator character) are
4812 reported as on logical line number @var{line-number} @minus{} 1. One day
4813 @command{@value{AS}} will no longer support this directive: it is recognized only
4814 for compatibility with existing assembler programs.
4818 @ifclear no-line-dir
4819 Even though this is a directive associated with the @code{a.out} or
4820 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4821 when producing COFF output, and treats @samp{.line} as though it
4822 were the COFF @samp{.ln} @emph{if} it is found outside a
4823 @code{.def}/@code{.endef} pair.
4825 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4826 used by compilers to generate auxiliary symbol information for
4831 @section @code{.linkonce [@var{type}]}
4833 @cindex @code{linkonce} directive
4834 @cindex common sections
4835 Mark the current section so that the linker only includes a single copy of it.
4836 This may be used to include the same section in several different object files,
4837 but ensure that the linker will only include it once in the final output file.
4838 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4839 Duplicate sections are detected based on the section name, so it should be
4842 This directive is only supported by a few object file formats; as of this
4843 writing, the only object file format which supports it is the Portable
4844 Executable format used on Windows NT.
4846 The @var{type} argument is optional. If specified, it must be one of the
4847 following strings. For example:
4851 Not all types may be supported on all object file formats.
4855 Silently discard duplicate sections. This is the default.
4858 Warn if there are duplicate sections, but still keep only one copy.
4861 Warn if any of the duplicates have different sizes.
4864 Warn if any of the duplicates do not have exactly the same contents.
4868 @section @code{.ln @var{line-number}}
4870 @cindex @code{ln} directive
4871 @ifclear no-line-dir
4872 @samp{.ln} is a synonym for @samp{.line}.
4875 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4876 must be an absolute expression. The next line has that logical
4877 line number, so any other statements on the current line (after a
4878 statement separator character @code{;}) are reported as on logical
4879 line number @var{line-number} @minus{} 1.
4882 This directive is accepted, but ignored, when @command{@value{AS}} is
4883 configured for @code{b.out}; its effect is only associated with COFF
4889 @section @code{.mri @var{val}}
4891 @cindex @code{mri} directive
4892 @cindex MRI mode, temporarily
4893 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4894 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4895 affects code assembled until the next @code{.mri} directive, or until the end
4896 of the file. @xref{M, MRI mode, MRI mode}.
4899 @section @code{.list}
4901 @cindex @code{list} directive
4902 @cindex listing control, turning on
4903 Control (in conjunction with the @code{.nolist} directive) whether or
4904 not assembly listings are generated. These two directives maintain an
4905 internal counter (which is zero initially). @code{.list} increments the
4906 counter, and @code{.nolist} decrements it. Assembly listings are
4907 generated whenever the counter is greater than zero.
4909 By default, listings are disabled. When you enable them (with the
4910 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4911 the initial value of the listing counter is one.
4914 @section @code{.long @var{expressions}}
4916 @cindex @code{long} directive
4917 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4920 @c no one seems to know what this is for or whether this description is
4921 @c what it really ought to do
4923 @section @code{.lsym @var{symbol}, @var{expression}}
4925 @cindex @code{lsym} directive
4926 @cindex symbol, not referenced in assembly
4927 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4928 the hash table, ensuring it cannot be referenced by name during the
4929 rest of the assembly. This sets the attributes of the symbol to be
4930 the same as the expression value:
4932 @var{other} = @var{descriptor} = 0
4933 @var{type} = @r{(section of @var{expression})}
4934 @var{value} = @var{expression}
4937 The new symbol is not flagged as external.
4941 @section @code{.macro}
4944 The commands @code{.macro} and @code{.endm} allow you to define macros that
4945 generate assembly output. For example, this definition specifies a macro
4946 @code{sum} that puts a sequence of numbers into memory:
4949 .macro sum from=0, to=5
4958 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4970 @item .macro @var{macname}
4971 @itemx .macro @var{macname} @var{macargs} @dots{}
4972 @cindex @code{macro} directive
4973 Begin the definition of a macro called @var{macname}. If your macro
4974 definition requires arguments, specify their names after the macro name,
4975 separated by commas or spaces. You can qualify the macro argument to
4976 indicate whether all invocations must specify a non-blank value (through
4977 @samp{:@code{req}}), or whether it takes all of the remaining arguments
4978 (through @samp{:@code{vararg}}). You can supply a default value for any
4979 macro argument by following the name with @samp{=@var{deflt}}. You
4980 cannot define two macros with the same @var{macname} unless it has been
4981 subject to the @code{.purgem} directive (@xref{Purgem}.) between the two
4982 definitions. For example, these are all valid @code{.macro} statements:
4986 Begin the definition of a macro called @code{comm}, which takes no
4989 @item .macro plus1 p, p1
4990 @itemx .macro plus1 p p1
4991 Either statement begins the definition of a macro called @code{plus1},
4992 which takes two arguments; within the macro definition, write
4993 @samp{\p} or @samp{\p1} to evaluate the arguments.
4995 @item .macro reserve_str p1=0 p2
4996 Begin the definition of a macro called @code{reserve_str}, with two
4997 arguments. The first argument has a default value, but not the second.
4998 After the definition is complete, you can call the macro either as
4999 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5000 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5001 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5002 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5005 @item .macro m p1:req, p2=0, p3:vararg
5006 Begin the definition of a macro called @code{m}, with at least three
5007 arguments. The first argument must always have a value specified, but
5008 not the second, which instead has a default value. The third formal
5009 will get assigned all remaining arguments specified at invocation time.
5011 When you call a macro, you can specify the argument values either by
5012 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5013 @samp{sum to=17, from=9}.
5015 Note that since each of the @var{macargs} can be an identifier exactly
5016 as any other one permitted by the target architecture, there may be
5017 occasional problems if the target hand-crafts special meanings to certain
5018 characters when they occur in a special position. For example, if colon
5019 (@code{:}) is generally permitted to be part of a symbol name, but the
5020 architecture specific code special-cases it when occuring as the final
5021 character of a symbol (to denote a label), then the macro parameter
5022 replacement code will have no way of knowing that and consider the whole
5023 construct (including the colon) an identifier, and check only this
5024 identifier for being the subject to parameter substitution. In this
5025 example, besides the potential of just separating identifier and colon
5026 by white space, using alternate macro syntax (@xref{Altmacro}.) and
5027 ampersand (@code{&}) as the character to separate literal text from macro
5028 parameters (or macro parameters from one another) would provide a way to
5029 achieve the same effect:
5038 This applies identically to the identifiers used in @code{.irp} (@xref{Irp}.)
5039 and @code{.irpc} (@xref{Irpc}.).
5042 @cindex @code{endm} directive
5043 Mark the end of a macro definition.
5046 @cindex @code{exitm} directive
5047 Exit early from the current macro definition.
5049 @cindex number of macros executed
5050 @cindex macros, count executed
5052 @command{@value{AS}} maintains a counter of how many macros it has
5053 executed in this pseudo-variable; you can copy that number to your
5054 output with @samp{\@@}, but @emph{only within a macro definition}.
5056 @item LOCAL @var{name} [ , @dots{} ]
5057 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5058 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5059 @xref{Altmacro,,@code{.altmacro}}.
5063 @section @code{.altmacro}
5064 Enable alternate macro mode, enabling:
5067 @item LOCAL @var{name} [ , @dots{} ]
5068 One additional directive, @code{LOCAL}, is available. It is used to
5069 generate a string replacement for each of the @var{name} arguments, and
5070 replace any instances of @var{name} in each macro expansion. The
5071 replacement string is unique in the assembly, and different for each
5072 separate macro expansion. @code{LOCAL} allows you to write macros that
5073 define symbols, without fear of conflict between separate macro expansions.
5075 @item String delimiters
5076 You can write strings delimited in these other ways besides
5077 @code{"@var{string}"}:
5080 @item '@var{string}'
5081 You can delimit strings with single-quote charaters.
5083 @item <@var{string}>
5084 You can delimit strings with matching angle brackets.
5087 @item single-character string escape
5088 To include any single character literally in a string (even if the
5089 character would otherwise have some special meaning), you can prefix the
5090 character with @samp{!} (an exclamation mark). For example, you can
5091 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
5093 @item Expression results as strings
5094 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
5095 and use the result as a string.
5099 @section @code{.noaltmacro}
5100 Disable alternate macro mode. @ref{Altmacro}
5103 @section @code{.nolist}
5105 @cindex @code{nolist} directive
5106 @cindex listing control, turning off
5107 Control (in conjunction with the @code{.list} directive) whether or
5108 not assembly listings are generated. These two directives maintain an
5109 internal counter (which is zero initially). @code{.list} increments the
5110 counter, and @code{.nolist} decrements it. Assembly listings are
5111 generated whenever the counter is greater than zero.
5114 @section @code{.octa @var{bignums}}
5116 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5117 @cindex @code{octa} directive
5118 @cindex integer, 16-byte
5119 @cindex sixteen byte integer
5120 This directive expects zero or more bignums, separated by commas. For each
5121 bignum, it emits a 16-byte integer.
5123 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5124 hence @emph{octa}-word for 16 bytes.
5127 @section @code{.org @var{new-lc} , @var{fill}}
5129 @cindex @code{org} directive
5130 @cindex location counter, advancing
5131 @cindex advancing location counter
5132 @cindex current address, advancing
5133 Advance the location counter of the current section to
5134 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5135 expression with the same section as the current subsection. That is,
5136 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5137 wrong section, the @code{.org} directive is ignored. To be compatible
5138 with former assemblers, if the section of @var{new-lc} is absolute,
5139 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5140 is the same as the current subsection.
5142 @code{.org} may only increase the location counter, or leave it
5143 unchanged; you cannot use @code{.org} to move the location counter
5146 @c double negative used below "not undefined" because this is a specific
5147 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5148 @c section. doc@cygnus.com 18feb91
5149 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5150 may not be undefined. If you really detest this restriction we eagerly await
5151 a chance to share your improved assembler.
5153 Beware that the origin is relative to the start of the section, not
5154 to the start of the subsection. This is compatible with other
5155 people's assemblers.
5157 When the location counter (of the current subsection) is advanced, the
5158 intervening bytes are filled with @var{fill} which should be an
5159 absolute expression. If the comma and @var{fill} are omitted,
5160 @var{fill} defaults to zero.
5163 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5165 @cindex padding the location counter given a power of two
5166 @cindex @code{p2align} directive
5167 Pad the location counter (in the current subsection) to a particular
5168 storage boundary. The first expression (which must be absolute) is the
5169 number of low-order zero bits the location counter must have after
5170 advancement. For example @samp{.p2align 3} advances the location
5171 counter until it a multiple of 8. If the location counter is already a
5172 multiple of 8, no change is needed.
5174 The second expression (also absolute) gives the fill value to be stored in the
5175 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5176 padding bytes are normally zero. However, on some systems, if the section is
5177 marked as containing code and the fill value is omitted, the space is filled
5178 with no-op instructions.
5180 The third expression is also absolute, and is also optional. If it is present,
5181 it is the maximum number of bytes that should be skipped by this alignment
5182 directive. If doing the alignment would require skipping more bytes than the
5183 specified maximum, then the alignment is not done at all. You can omit the
5184 fill value (the second argument) entirely by simply using two commas after the
5185 required alignment; this can be useful if you want the alignment to be filled
5186 with no-op instructions when appropriate.
5188 @cindex @code{p2alignw} directive
5189 @cindex @code{p2alignl} directive
5190 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5191 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5192 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5193 fill pattern as a four byte longword value. For example, @code{.p2alignw
5194 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5195 filled in with the value 0x368d (the exact placement of the bytes depends upon
5196 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5201 @section @code{.previous}
5203 @cindex @code{previous} directive
5204 @cindex Section Stack
5205 This is one of the ELF section stack manipulation directives. The others are
5206 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5207 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5208 (@pxref{PopSection}).
5210 This directive swaps the current section (and subsection) with most recently
5211 referenced section (and subsection) prior to this one. Multiple
5212 @code{.previous} directives in a row will flip between two sections (and their
5215 In terms of the section stack, this directive swaps the current section with
5216 the top section on the section stack.
5221 @section @code{.popsection}
5223 @cindex @code{popsection} directive
5224 @cindex Section Stack
5225 This is one of the ELF section stack manipulation directives. The others are
5226 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5227 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5230 This directive replaces the current section (and subsection) with the top
5231 section (and subsection) on the section stack. This section is popped off the
5236 @section @code{.print @var{string}}
5238 @cindex @code{print} directive
5239 @command{@value{AS}} will print @var{string} on the standard output during
5240 assembly. You must put @var{string} in double quotes.
5244 @section @code{.protected @var{names}}
5246 @cindex @code{protected} directive
5248 This is one of the ELF visibility directives. The other two are
5249 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5251 This directive overrides the named symbols default visibility (which is set by
5252 their binding: local, global or weak). The directive sets the visibility to
5253 @code{protected} which means that any references to the symbols from within the
5254 components that defines them must be resolved to the definition in that
5255 component, even if a definition in another component would normally preempt
5260 @section @code{.psize @var{lines} , @var{columns}}
5262 @cindex @code{psize} directive
5263 @cindex listing control: paper size
5264 @cindex paper size, for listings
5265 Use this directive to declare the number of lines---and, optionally, the
5266 number of columns---to use for each page, when generating listings.
5268 If you do not use @code{.psize}, listings use a default line-count
5269 of 60. You may omit the comma and @var{columns} specification; the
5270 default width is 200 columns.
5272 @command{@value{AS}} generates formfeeds whenever the specified number of
5273 lines is exceeded (or whenever you explicitly request one, using
5276 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5277 those explicitly specified with @code{.eject}.
5280 @section @code{.purgem @var{name}}
5282 @cindex @code{purgem} directive
5283 Undefine the macro @var{name}, so that later uses of the string will not be
5284 expanded. @xref{Macro}.
5288 @section @code{.pushsection @var{name} , @var{subsection}}
5290 @cindex @code{pushsection} directive
5291 @cindex Section Stack
5292 This is one of the ELF section stack manipulation directives. The others are
5293 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5294 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5297 This directive pushes the current section (and subsection) onto the
5298 top of the section stack, and then replaces the current section and
5299 subsection with @code{name} and @code{subsection}.
5303 @section @code{.quad @var{bignums}}
5305 @cindex @code{quad} directive
5306 @code{.quad} expects zero or more bignums, separated by commas. For
5307 each bignum, it emits
5309 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5310 warning message; and just takes the lowest order 8 bytes of the bignum.
5311 @cindex eight-byte integer
5312 @cindex integer, 8-byte
5314 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5315 hence @emph{quad}-word for 8 bytes.
5318 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5319 warning message; and just takes the lowest order 16 bytes of the bignum.
5320 @cindex sixteen-byte integer
5321 @cindex integer, 16-byte
5325 @section @code{.rept @var{count}}
5327 @cindex @code{rept} directive
5328 Repeat the sequence of lines between the @code{.rept} directive and the next
5329 @code{.endr} directive @var{count} times.
5331 For example, assembling
5339 is equivalent to assembling
5348 @section @code{.sbttl "@var{subheading}"}
5350 @cindex @code{sbttl} directive
5351 @cindex subtitles for listings
5352 @cindex listing control: subtitle
5353 Use @var{subheading} as the title (third line, immediately after the
5354 title line) when generating assembly listings.
5356 This directive affects subsequent pages, as well as the current page if
5357 it appears within ten lines of the top of a page.
5361 @section @code{.scl @var{class}}
5363 @cindex @code{scl} directive
5364 @cindex symbol storage class (COFF)
5365 @cindex COFF symbol storage class
5366 Set the storage-class value for a symbol. This directive may only be
5367 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5368 whether a symbol is static or external, or it may record further
5369 symbolic debugging information.
5372 The @samp{.scl} directive is primarily associated with COFF output; when
5373 configured to generate @code{b.out} output format, @command{@value{AS}}
5374 accepts this directive but ignores it.
5380 @section @code{.section @var{name}}
5382 @cindex named section
5383 Use the @code{.section} directive to assemble the following code into a section
5386 This directive is only supported for targets that actually support arbitrarily
5387 named sections; on @code{a.out} targets, for example, it is not accepted, even
5388 with a standard @code{a.out} section name.
5392 @c only print the extra heading if both COFF and ELF are set
5393 @subheading COFF Version
5396 @cindex @code{section} directive (COFF version)
5397 For COFF targets, the @code{.section} directive is used in one of the following
5401 .section @var{name}[, "@var{flags}"]
5402 .section @var{name}[, @var{subsegment}]
5405 If the optional argument is quoted, it is taken as flags to use for the
5406 section. Each flag is a single character. The following flags are recognized:
5409 bss section (uninitialized data)
5411 section is not loaded
5421 shared section (meaningful for PE targets)
5423 ignored. (For compatibility with the ELF version)
5426 If no flags are specified, the default flags depend upon the section name. If
5427 the section name is not recognized, the default will be for the section to be
5428 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5429 from the section, rather than adding them, so if they are used on their own it
5430 will be as if no flags had been specified at all.
5432 If the optional argument to the @code{.section} directive is not quoted, it is
5433 taken as a subsegment number (@pxref{Sub-Sections}).
5438 @c only print the extra heading if both COFF and ELF are set
5439 @subheading ELF Version
5442 @cindex Section Stack
5443 This is one of the ELF section stack manipulation directives. The others are
5444 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5445 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5446 @code{.previous} (@pxref{Previous}).
5448 @cindex @code{section} directive (ELF version)
5449 For ELF targets, the @code{.section} directive is used like this:
5452 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
5455 The optional @var{flags} argument is a quoted string which may contain any
5456 combination of the following characters:
5459 section is allocatable
5463 section is executable
5465 section is mergeable
5467 section contains zero terminated strings
5469 section is a member of a section group
5471 section is used for thread-local-storage
5474 The optional @var{type} argument may contain one of the following constants:
5477 section contains data
5479 section does not contain data (i.e., section only occupies space)
5481 section contains data which is used by things other than the program
5483 section contains an array of pointers to init functions
5485 section contains an array of pointers to finish functions
5486 @item @@preinit_array
5487 section contains an array of pointers to pre-init functions
5490 Many targets only support the first three section types.
5492 Note on targets where the @code{@@} character is the start of a comment (eg
5493 ARM) then another character is used instead. For example the ARM port uses the
5496 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
5497 be specified as well as an extra argument - @var{entsize} - like this:
5500 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
5503 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
5504 constants, each @var{entsize} octets long. Sections with both @code{M} and
5505 @code{S} must contain zero terminated strings where each character is
5506 @var{entsize} bytes long. The linker may remove duplicates within sections with
5507 the same name, same entity size and same flags. @var{entsize} must be an
5508 absolute expression.
5510 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
5511 be present along with an additional field like this:
5514 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
5517 The @var{GroupName} field specifies the name of the section group to which this
5518 particular section belongs. The optional linkage field can contain:
5521 indicates that only one copy of this section should be retained
5526 Note - if both the @var{M} and @var{G} flags are present then the fields for
5527 the Merge flag should come first, like this:
5530 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
5533 If no flags are specified, the default flags depend upon the section name. If
5534 the section name is not recognized, the default will be for the section to have
5535 none of the above flags: it will not be allocated in memory, nor writable, nor
5536 executable. The section will contain data.
5538 For ELF targets, the assembler supports another type of @code{.section}
5539 directive for compatibility with the Solaris assembler:
5542 .section "@var{name}"[, @var{flags}...]
5545 Note that the section name is quoted. There may be a sequence of comma
5549 section is allocatable
5553 section is executable
5555 section is used for thread local storage
5558 This directive replaces the current section and subsection. See the
5559 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
5560 some examples of how this directive and the other section stack directives
5566 @section @code{.set @var{symbol}, @var{expression}}
5568 @cindex @code{set} directive
5569 @cindex symbol value, setting
5570 Set the value of @var{symbol} to @var{expression}. This
5571 changes @var{symbol}'s value and type to conform to
5572 @var{expression}. If @var{symbol} was flagged as external, it remains
5573 flagged (@pxref{Symbol Attributes}).
5575 You may @code{.set} a symbol many times in the same assembly.
5577 If you @code{.set} a global symbol, the value stored in the object
5578 file is the last value stored into it.
5581 The syntax for @code{set} on the HPPA is
5582 @samp{@var{symbol} .set @var{expression}}.
5586 On Z80 @code{set} is a real instruction, use
5587 @samp{@var{symbol} defl @var{expression}} instead.
5591 @section @code{.short @var{expressions}}
5593 @cindex @code{short} directive
5595 @code{.short} is normally the same as @samp{.word}.
5596 @xref{Word,,@code{.word}}.
5598 In some configurations, however, @code{.short} and @code{.word} generate
5599 numbers of different lengths; @pxref{Machine Dependencies}.
5603 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5606 This expects zero or more @var{expressions}, and emits
5607 a 16 bit number for each.
5612 @section @code{.single @var{flonums}}
5614 @cindex @code{single} directive
5615 @cindex floating point numbers (single)
5616 This directive assembles zero or more flonums, separated by commas. It
5617 has the same effect as @code{.float}.
5619 The exact kind of floating point numbers emitted depends on how
5620 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5624 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5625 numbers in @sc{ieee} format.
5631 @section @code{.size}
5633 This directive is used to set the size associated with a symbol.
5637 @c only print the extra heading if both COFF and ELF are set
5638 @subheading COFF Version
5641 @cindex @code{size} directive (COFF version)
5642 For COFF targets, the @code{.size} directive is only permitted inside
5643 @code{.def}/@code{.endef} pairs. It is used like this:
5646 .size @var{expression}
5650 @samp{.size} is only meaningful when generating COFF format output; when
5651 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5658 @c only print the extra heading if both COFF and ELF are set
5659 @subheading ELF Version
5662 @cindex @code{size} directive (ELF version)
5663 For ELF targets, the @code{.size} directive is used like this:
5666 .size @var{name} , @var{expression}
5669 This directive sets the size associated with a symbol @var{name}.
5670 The size in bytes is computed from @var{expression} which can make use of label
5671 arithmetic. This directive is typically used to set the size of function
5677 @section @code{.sleb128 @var{expressions}}
5679 @cindex @code{sleb128} directive
5680 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5681 compact, variable length representation of numbers used by the DWARF
5682 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
5684 @ifclear no-space-dir
5686 @section @code{.skip @var{size} , @var{fill}}
5688 @cindex @code{skip} directive
5689 @cindex filling memory
5690 This directive emits @var{size} bytes, each of value @var{fill}. Both
5691 @var{size} and @var{fill} are absolute expressions. If the comma and
5692 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5696 @section @code{.space @var{size} , @var{fill}}
5698 @cindex @code{space} directive
5699 @cindex filling memory
5700 This directive emits @var{size} bytes, each of value @var{fill}. Both
5701 @var{size} and @var{fill} are absolute expressions. If the comma
5702 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5707 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5708 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5709 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5710 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5718 @section @code{.stabd, .stabn, .stabs}
5720 @cindex symbolic debuggers, information for
5721 @cindex @code{stab@var{x}} directives
5722 There are three directives that begin @samp{.stab}.
5723 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5724 The symbols are not entered in the @command{@value{AS}} hash table: they
5725 cannot be referenced elsewhere in the source file.
5726 Up to five fields are required:
5730 This is the symbol's name. It may contain any character except
5731 @samp{\000}, so is more general than ordinary symbol names. Some
5732 debuggers used to code arbitrarily complex structures into symbol names
5736 An absolute expression. The symbol's type is set to the low 8 bits of
5737 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5738 and debuggers choke on silly bit patterns.
5741 An absolute expression. The symbol's ``other'' attribute is set to the
5742 low 8 bits of this expression.
5745 An absolute expression. The symbol's descriptor is set to the low 16
5746 bits of this expression.
5749 An absolute expression which becomes the symbol's value.
5752 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5753 or @code{.stabs} statement, the symbol has probably already been created;
5754 you get a half-formed symbol in your object file. This is
5755 compatible with earlier assemblers!
5758 @cindex @code{stabd} directive
5759 @item .stabd @var{type} , @var{other} , @var{desc}
5761 The ``name'' of the symbol generated is not even an empty string.
5762 It is a null pointer, for compatibility. Older assemblers used a
5763 null pointer so they didn't waste space in object files with empty
5766 The symbol's value is set to the location counter,
5767 relocatably. When your program is linked, the value of this symbol
5768 is the address of the location counter when the @code{.stabd} was
5771 @cindex @code{stabn} directive
5772 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5773 The name of the symbol is set to the empty string @code{""}.
5775 @cindex @code{stabs} directive
5776 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5777 All five fields are specified.
5783 @section @code{.string} "@var{str}"
5785 @cindex string, copying to object file
5786 @cindex @code{string} directive
5788 Copy the characters in @var{str} to the object file. You may specify more than
5789 one string to copy, separated by commas. Unless otherwise specified for a
5790 particular machine, the assembler marks the end of each string with a 0 byte.
5791 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5794 @section @code{.struct @var{expression}}
5796 @cindex @code{struct} directive
5797 Switch to the absolute section, and set the section offset to @var{expression},
5798 which must be an absolute expression. You might use this as follows:
5807 This would define the symbol @code{field1} to have the value 0, the symbol
5808 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5809 value 8. Assembly would be left in the absolute section, and you would need to
5810 use a @code{.section} directive of some sort to change to some other section
5811 before further assembly.
5815 @section @code{.subsection @var{name}}
5817 @cindex @code{subsection} directive
5818 @cindex Section Stack
5819 This is one of the ELF section stack manipulation directives. The others are
5820 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5821 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5824 This directive replaces the current subsection with @code{name}. The current
5825 section is not changed. The replaced subsection is put onto the section stack
5826 in place of the then current top of stack subsection.
5831 @section @code{.symver}
5832 @cindex @code{symver} directive
5833 @cindex symbol versioning
5834 @cindex versions of symbols
5835 Use the @code{.symver} directive to bind symbols to specific version nodes
5836 within a source file. This is only supported on ELF platforms, and is
5837 typically used when assembling files to be linked into a shared library.
5838 There are cases where it may make sense to use this in objects to be bound
5839 into an application itself so as to override a versioned symbol from a
5842 For ELF targets, the @code{.symver} directive can be used like this:
5844 .symver @var{name}, @var{name2@@nodename}
5846 If the symbol @var{name} is defined within the file
5847 being assembled, the @code{.symver} directive effectively creates a symbol
5848 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5849 just don't try and create a regular alias is that the @var{@@} character isn't
5850 permitted in symbol names. The @var{name2} part of the name is the actual name
5851 of the symbol by which it will be externally referenced. The name @var{name}
5852 itself is merely a name of convenience that is used so that it is possible to
5853 have definitions for multiple versions of a function within a single source
5854 file, and so that the compiler can unambiguously know which version of a
5855 function is being mentioned. The @var{nodename} portion of the alias should be
5856 the name of a node specified in the version script supplied to the linker when
5857 building a shared library. If you are attempting to override a versioned
5858 symbol from a shared library, then @var{nodename} should correspond to the
5859 nodename of the symbol you are trying to override.
5861 If the symbol @var{name} is not defined within the file being assembled, all
5862 references to @var{name} will be changed to @var{name2@@nodename}. If no
5863 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5866 Another usage of the @code{.symver} directive is:
5868 .symver @var{name}, @var{name2@@@@nodename}
5870 In this case, the symbol @var{name} must exist and be defined within
5871 the file being assembled. It is similar to @var{name2@@nodename}. The
5872 difference is @var{name2@@@@nodename} will also be used to resolve
5873 references to @var{name2} by the linker.
5875 The third usage of the @code{.symver} directive is:
5877 .symver @var{name}, @var{name2@@@@@@nodename}
5879 When @var{name} is not defined within the
5880 file being assembled, it is treated as @var{name2@@nodename}. When
5881 @var{name} is defined within the file being assembled, the symbol
5882 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5887 @section @code{.tag @var{structname}}
5889 @cindex COFF structure debugging
5890 @cindex structure debugging, COFF
5891 @cindex @code{tag} directive
5892 This directive is generated by compilers to include auxiliary debugging
5893 information in the symbol table. It is only permitted inside
5894 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5895 definitions in the symbol table with instances of those structures.
5898 @samp{.tag} is only used when generating COFF format output; when
5899 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5905 @section @code{.text @var{subsection}}
5907 @cindex @code{text} directive
5908 Tells @command{@value{AS}} to assemble the following statements onto the end of
5909 the text subsection numbered @var{subsection}, which is an absolute
5910 expression. If @var{subsection} is omitted, subsection number zero
5914 @section @code{.title "@var{heading}"}
5916 @cindex @code{title} directive
5917 @cindex listing control: title line
5918 Use @var{heading} as the title (second line, immediately after the
5919 source file name and pagenumber) when generating assembly listings.
5921 This directive affects subsequent pages, as well as the current page if
5922 it appears within ten lines of the top of a page.
5926 @section @code{.type}
5928 This directive is used to set the type of a symbol.
5932 @c only print the extra heading if both COFF and ELF are set
5933 @subheading COFF Version
5936 @cindex COFF symbol type
5937 @cindex symbol type, COFF
5938 @cindex @code{type} directive (COFF version)
5939 For COFF targets, this directive is permitted only within
5940 @code{.def}/@code{.endef} pairs. It is used like this:
5946 This records the integer @var{int} as the type attribute of a symbol table
5950 @samp{.type} is associated only with COFF format output; when
5951 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
5952 directive but ignores it.
5958 @c only print the extra heading if both COFF and ELF are set
5959 @subheading ELF Version
5962 @cindex ELF symbol type
5963 @cindex symbol type, ELF
5964 @cindex @code{type} directive (ELF version)
5965 For ELF targets, the @code{.type} directive is used like this:
5968 .type @var{name} , @var{type description}
5971 This sets the type of symbol @var{name} to be either a
5972 function symbol or an object symbol. There are five different syntaxes
5973 supported for the @var{type description} field, in order to provide
5974 compatibility with various other assemblers.
5976 Because some of the characters used in these syntaxes (such as @samp{@@} and
5977 @samp{#}) are comment characters for some architectures, some of the syntaxes
5978 below do not work on all architectures. The first variant will be accepted by
5979 the GNU assembler on all architectures so that variant should be used for
5980 maximum portability, if you do not need to assemble your code with other
5983 The syntaxes supported are:
5986 .type <name> STT_FUNCTION
5987 .type <name> STT_OBJECT
5989 .type <name>,#function
5990 .type <name>,#object
5992 .type <name>,@@function
5993 .type <name>,@@object
5995 .type <name>,%function
5996 .type <name>,%object
5998 .type <name>,"function"
5999 .type <name>,"object"
6005 @section @code{.uleb128 @var{expressions}}
6007 @cindex @code{uleb128} directive
6008 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6009 compact, variable length representation of numbers used by the DWARF
6010 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
6014 @section @code{.val @var{addr}}
6016 @cindex @code{val} directive
6017 @cindex COFF value attribute
6018 @cindex value attribute, COFF
6019 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6020 records the address @var{addr} as the value attribute of a symbol table
6024 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6025 configured for @code{b.out}, it accepts this directive but ignores it.
6031 @section @code{.version "@var{string}"}
6033 @cindex @code{version} directive
6034 This directive creates a @code{.note} section and places into it an ELF
6035 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6040 @section @code{.vtable_entry @var{table}, @var{offset}}
6042 @cindex @code{vtable_entry} directive
6043 This directive finds or creates a symbol @code{table} and creates a
6044 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6047 @section @code{.vtable_inherit @var{child}, @var{parent}}
6049 @cindex @code{vtable_inherit} directive
6050 This directive finds the symbol @code{child} and finds or creates the symbol
6051 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6052 parent whose addend is the value of the child symbol. As a special case the
6053 parent name of @code{0} is treated as refering the @code{*ABS*} section.
6057 @section @code{.warning "@var{string}"}
6058 @cindex warning directive
6059 Similar to the directive @code{.error}
6060 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6063 @section @code{.weak @var{names}}
6065 @cindex @code{weak} directive
6066 This directive sets the weak attribute on the comma separated list of symbol
6067 @code{names}. If the symbols do not already exist, they will be created.
6069 On COFF targets other than PE, weak symbols are a GNU extension. This
6070 directive sets the weak attribute on the comma separated list of symbol
6071 @code{names}. If the symbols do not already exist, they will be created.
6073 On the PE target, weak symbols are supported natively as weak aliases.
6074 When a weak symbol is created that is not an alias, GAS creates an
6075 alternate symbol to hold the default value.
6078 @section @code{.weakref @var{alias}, @var{target}}
6080 @cindex @code{weakref} directive
6081 This directive creates an alias to the target symbol that enables the symbol to
6082 be referenced with weak-symbol semantics, but without actually making it weak.
6083 If direct references or definitions of the symbol are present, then the symbol
6084 will not be weak, but if all references to it are through weak references, the
6085 symbol will be marked as weak in the symbol table.
6087 The effect is equivalent to moving all references to the alias to a separate
6088 assembly source file, renaming the alias to the symbol in it, declaring the
6089 symbol as weak there, and running a reloadable link to merge the object files
6090 resulting from the assembly of the new source file and the old source file that
6091 had the references to the alias removed.
6093 The alias itself never makes to the symbol table, and is entirely handled
6094 within the assembler.
6097 @section @code{.word @var{expressions}}
6099 @cindex @code{word} directive
6100 This directive expects zero or more @var{expressions}, of any section,
6101 separated by commas.
6104 For each expression, @command{@value{AS}} emits a 32-bit number.
6107 For each expression, @command{@value{AS}} emits a 16-bit number.
6112 The size of the number emitted, and its byte order,
6113 depend on what target computer the assembly is for.
6116 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6117 @c happen---32-bit addressability, period; no long/short jumps.
6118 @ifset DIFF-TBL-KLUGE
6119 @cindex difference tables altered
6120 @cindex altered difference tables
6122 @emph{Warning: Special Treatment to support Compilers}
6126 Machines with a 32-bit address space, but that do less than 32-bit
6127 addressing, require the following special treatment. If the machine of
6128 interest to you does 32-bit addressing (or doesn't require it;
6129 @pxref{Machine Dependencies}), you can ignore this issue.
6132 In order to assemble compiler output into something that works,
6133 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6134 Directives of the form @samp{.word sym1-sym2} are often emitted by
6135 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6136 directive of the form @samp{.word sym1-sym2}, and the difference between
6137 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6138 creates a @dfn{secondary jump table}, immediately before the next label.
6139 This secondary jump table is preceded by a short-jump to the
6140 first byte after the secondary table. This short-jump prevents the flow
6141 of control from accidentally falling into the new table. Inside the
6142 table is a long-jump to @code{sym2}. The original @samp{.word}
6143 contains @code{sym1} minus the address of the long-jump to
6146 If there were several occurrences of @samp{.word sym1-sym2} before the
6147 secondary jump table, all of them are adjusted. If there was a
6148 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6149 long-jump to @code{sym4} is included in the secondary jump table,
6150 and the @code{.word} directives are adjusted to contain @code{sym3}
6151 minus the address of the long-jump to @code{sym4}; and so on, for as many
6152 entries in the original jump table as necessary.
6155 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6156 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6157 assembly language programmers.
6160 @c end DIFF-TBL-KLUGE
6163 @section Deprecated Directives
6165 @cindex deprecated directives
6166 @cindex obsolescent directives
6167 One day these directives won't work.
6168 They are included for compatibility with older assemblers.
6175 @node Machine Dependencies
6176 @chapter Machine Dependent Features
6178 @cindex machine dependencies
6179 The machine instruction sets are (almost by definition) different on
6180 each machine where @command{@value{AS}} runs. Floating point representations
6181 vary as well, and @command{@value{AS}} often supports a few additional
6182 directives or command-line options for compatibility with other
6183 assemblers on a particular platform. Finally, some versions of
6184 @command{@value{AS}} support special pseudo-instructions for branch
6187 This chapter discusses most of these differences, though it does not
6188 include details on any machine's instruction set. For details on that
6189 subject, see the hardware manufacturer's manual.
6193 * Alpha-Dependent:: Alpha Dependent Features
6196 * ARC-Dependent:: ARC Dependent Features
6199 * ARM-Dependent:: ARM Dependent Features
6202 * BFIN-Dependent:: BFIN Dependent Features
6205 * CRIS-Dependent:: CRIS Dependent Features
6208 * D10V-Dependent:: D10V Dependent Features
6211 * D30V-Dependent:: D30V Dependent Features
6214 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6217 * HPPA-Dependent:: HPPA Dependent Features
6220 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6223 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6226 * i860-Dependent:: Intel 80860 Dependent Features
6229 * i960-Dependent:: Intel 80960 Dependent Features
6232 * IA-64-Dependent:: Intel IA-64 Dependent Features
6235 * IP2K-Dependent:: IP2K Dependent Features
6238 * M32C-Dependent:: M32C Dependent Features
6241 * M32R-Dependent:: M32R Dependent Features
6244 * M68K-Dependent:: M680x0 Dependent Features
6247 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6250 * MIPS-Dependent:: MIPS Dependent Features
6253 * MMIX-Dependent:: MMIX Dependent Features
6256 * MSP430-Dependent:: MSP430 Dependent Features
6259 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6260 * SH64-Dependent:: SuperH SH64 Dependent Features
6263 * PDP-11-Dependent:: PDP-11 Dependent Features
6266 * PJ-Dependent:: picoJava Dependent Features
6269 * PPC-Dependent:: PowerPC Dependent Features
6272 * Sparc-Dependent:: SPARC Dependent Features
6275 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
6278 * V850-Dependent:: V850 Dependent Features
6281 * Xtensa-Dependent:: Xtensa Dependent Features
6284 * Z80-Dependent:: Z80 Dependent Features
6287 * Z8000-Dependent:: Z8000 Dependent Features
6290 * Vax-Dependent:: VAX Dependent Features
6297 @c The following major nodes are *sections* in the GENERIC version, *chapters*
6298 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
6299 @c peculiarity: to preserve cross-references, there must be a node called
6300 @c "Machine Dependencies". Hence the conditional nodenames in each
6301 @c major node below. Node defaulting in makeinfo requires adjacency of
6302 @c node and sectioning commands; hence the repetition of @chapter BLAH
6303 @c in both conditional blocks.
6306 @include c-alpha.texi
6318 @include c-bfin.texi
6322 @include c-cris.texi
6327 @node Machine Dependencies
6328 @chapter Machine Dependent Features
6330 The machine instruction sets are different on each Renesas chip family,
6331 and there are also some syntax differences among the families. This
6332 chapter describes the specific @command{@value{AS}} features for each
6336 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6337 * SH-Dependent:: Renesas SH Dependent Features
6344 @include c-d10v.texi
6348 @include c-d30v.texi
6352 @include c-h8300.texi
6356 @include c-hppa.texi
6360 @include c-i370.texi
6364 @include c-i386.texi
6368 @include c-i860.texi
6372 @include c-i960.texi
6376 @include c-ia64.texi
6380 @include c-ip2k.texi
6384 @include c-m32c.texi
6388 @include c-m32r.texi
6392 @include c-m68k.texi
6396 @include c-m68hc11.texi
6400 @include c-mips.texi
6404 @include c-mmix.texi
6408 @include c-msp430.texi
6412 @include c-ns32k.texi
6416 @include c-pdp11.texi
6429 @include c-sh64.texi
6433 @include c-sparc.texi
6437 @include c-tic54x.texi
6453 @include c-v850.texi
6457 @include c-xtensa.texi
6461 @c reverse effect of @down at top of generic Machine-Dep chapter
6465 @node Reporting Bugs
6466 @chapter Reporting Bugs
6467 @cindex bugs in assembler
6468 @cindex reporting bugs in assembler
6470 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6472 Reporting a bug may help you by bringing a solution to your problem, or it may
6473 not. But in any case the principal function of a bug report is to help the
6474 entire community by making the next version of @command{@value{AS}} work better.
6475 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6477 In order for a bug report to serve its purpose, you must include the
6478 information that enables us to fix the bug.
6481 * Bug Criteria:: Have you found a bug?
6482 * Bug Reporting:: How to report bugs
6486 @section Have You Found a Bug?
6487 @cindex bug criteria
6489 If you are not sure whether you have found a bug, here are some guidelines:
6492 @cindex fatal signal
6493 @cindex assembler crash
6494 @cindex crash of assembler
6496 If the assembler gets a fatal signal, for any input whatever, that is a
6497 @command{@value{AS}} bug. Reliable assemblers never crash.
6499 @cindex error on valid input
6501 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6503 @cindex invalid input
6505 If @command{@value{AS}} does not produce an error message for invalid input, that
6506 is a bug. However, you should note that your idea of ``invalid input'' might
6507 be our idea of ``an extension'' or ``support for traditional practice''.
6510 If you are an experienced user of assemblers, your suggestions for improvement
6511 of @command{@value{AS}} are welcome in any case.
6515 @section How to Report Bugs
6517 @cindex assembler bugs, reporting
6519 A number of companies and individuals offer support for @sc{gnu} products. If
6520 you obtained @command{@value{AS}} from a support organization, we recommend you
6521 contact that organization first.
6523 You can find contact information for many support companies and
6524 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6527 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6528 to @samp{bug-binutils@@gnu.org}.
6530 The fundamental principle of reporting bugs usefully is this:
6531 @strong{report all the facts}. If you are not sure whether to state a
6532 fact or leave it out, state it!
6534 Often people omit facts because they think they know what causes the problem
6535 and assume that some details do not matter. Thus, you might assume that the
6536 name of a symbol you use in an example does not matter. Well, probably it does
6537 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6538 happens to fetch from the location where that name is stored in memory;
6539 perhaps, if the name were different, the contents of that location would fool
6540 the assembler into doing the right thing despite the bug. Play it safe and
6541 give a specific, complete example. That is the easiest thing for you to do,
6542 and the most helpful.
6544 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6545 it is new to us. Therefore, always write your bug reports on the assumption
6546 that the bug has not been reported previously.
6548 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6549 bell?'' This cannot help us fix a bug, so it is basically useless. We
6550 respond by asking for enough details to enable us to investigate.
6551 You might as well expedite matters by sending them to begin with.
6553 To enable us to fix the bug, you should include all these things:
6557 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6558 it with the @samp{--version} argument.
6560 Without this, we will not know whether there is any point in looking for
6561 the bug in the current version of @command{@value{AS}}.
6564 Any patches you may have applied to the @command{@value{AS}} source.
6567 The type of machine you are using, and the operating system name and
6571 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6575 The command arguments you gave the assembler to assemble your example and
6576 observe the bug. To guarantee you will not omit something important, list them
6577 all. A copy of the Makefile (or the output from make) is sufficient.
6579 If we were to try to guess the arguments, we would probably guess wrong
6580 and then we might not encounter the bug.
6583 A complete input file that will reproduce the bug. If the bug is observed when
6584 the assembler is invoked via a compiler, send the assembler source, not the
6585 high level language source. Most compilers will produce the assembler source
6586 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6587 the options @samp{-v --save-temps}; this will save the assembler source in a
6588 file with an extension of @file{.s}, and also show you exactly how
6589 @command{@value{AS}} is being run.
6592 A description of what behavior you observe that you believe is
6593 incorrect. For example, ``It gets a fatal signal.''
6595 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6596 will certainly notice it. But if the bug is incorrect output, we might not
6597 notice unless it is glaringly wrong. You might as well not give us a chance to
6600 Even if the problem you experience is a fatal signal, you should still say so
6601 explicitly. Suppose something strange is going on, such as, your copy of
6602 @command{@value{AS}} is out of synch, or you have encountered a bug in the C
6603 library on your system. (This has happened!) Your copy might crash and ours
6604 would not. If you told us to expect a crash, then when ours fails to crash, we
6605 would know that the bug was not happening for us. If you had not told us to
6606 expect a crash, then we would not be able to draw any conclusion from our
6610 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6611 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6612 option. Always send diffs from the old file to the new file. If you even
6613 discuss something in the @command{@value{AS}} source, refer to it by context, not
6616 The line numbers in our development sources will not match those in your
6617 sources. Your line numbers would convey no useful information to us.
6620 Here are some things that are not necessary:
6624 A description of the envelope of the bug.
6626 Often people who encounter a bug spend a lot of time investigating
6627 which changes to the input file will make the bug go away and which
6628 changes will not affect it.
6630 This is often time consuming and not very useful, because the way we
6631 will find the bug is by running a single example under the debugger
6632 with breakpoints, not by pure deduction from a series of examples.
6633 We recommend that you save your time for something else.
6635 Of course, if you can find a simpler example to report @emph{instead}
6636 of the original one, that is a convenience for us. Errors in the
6637 output will be easier to spot, running under the debugger will take
6638 less time, and so on.
6640 However, simplification is not vital; if you do not want to do this,
6641 report the bug anyway and send us the entire test case you used.
6644 A patch for the bug.
6646 A patch for the bug does help us if it is a good one. But do not omit
6647 the necessary information, such as the test case, on the assumption that
6648 a patch is all we need. We might see problems with your patch and decide
6649 to fix the problem another way, or we might not understand it at all.
6651 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6652 construct an example that will make the program follow a certain path through
6653 the code. If you do not send us the example, we will not be able to construct
6654 one, so we will not be able to verify that the bug is fixed.
6656 And if we cannot understand what bug you are trying to fix, or why your
6657 patch should be an improvement, we will not install it. A test case will
6658 help us to understand.
6661 A guess about what the bug is or what it depends on.
6663 Such guesses are usually wrong. Even we cannot guess right about such
6664 things without first using the debugger to find the facts.
6667 @node Acknowledgements
6668 @chapter Acknowledgements
6670 If you have contributed to GAS and your name isn't listed here,
6671 it is not meant as a slight. We just don't know about it. Send mail to the
6672 maintainer, and we'll correct the situation. Currently
6674 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6676 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6679 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6680 information and the 68k series machines, most of the preprocessing pass, and
6681 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6683 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6684 many bug fixes, including merging support for several processors, breaking GAS
6685 up to handle multiple object file format back ends (including heavy rewrite,
6686 testing, an integration of the coff and b.out back ends), adding configuration
6687 including heavy testing and verification of cross assemblers and file splits
6688 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6689 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6690 port (including considerable amounts of reverse engineering), a SPARC opcode
6691 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6692 assertions and made them work, much other reorganization, cleanup, and lint.
6694 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6695 in format-specific I/O modules.
6697 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6698 has done much work with it since.
6700 The Intel 80386 machine description was written by Eliot Dresselhaus.
6702 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6704 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6705 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6707 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6708 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6709 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6710 support a.out format.
6712 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
6713 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
6714 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6715 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6718 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6719 simplified the configuration of which versions accept which directives. He
6720 updated the 68k machine description so that Motorola's opcodes always produced
6721 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6722 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6723 cross-compilation support, and one bug in relaxation that took a week and
6724 required the proverbial one-bit fix.
6726 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6727 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6728 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6729 PowerPC assembler, and made a few other minor patches.
6731 Steve Chamberlain made GAS able to generate listings.
6733 Hewlett-Packard contributed support for the HP9000/300.
6735 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6736 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6737 formats). This work was supported by both the Center for Software Science at
6738 the University of Utah and Cygnus Support.
6740 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6741 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6742 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6743 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6744 and some initial 64-bit support).
6746 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6748 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6749 support for openVMS/Alpha.
6751 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6754 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6755 Inc. added support for Xtensa processors.
6757 Several engineers at Cygnus Support have also provided many small bug fixes and
6758 configuration enhancements.
6760 Many others have contributed large or small bugfixes and enhancements. If
6761 you have contributed significant work and are not mentioned on this list, and
6762 want to be, let us know. Some of the history has been lost; we are not
6763 intentionally leaving anyone out.
6768 @unnumbered AS Index