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, 2007
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,
104 2006, 2007 Free Software Foundation, Inc.
106 Permission is granted to copy, distribute and/or modify this document
107 under the terms of the GNU Free Documentation License, Version 1.1
108 or any later version published by the Free Software Foundation;
109 with no Invariant Sections, with no Front-Cover Texts, and with no
110 Back-Cover Texts. A copy of the license is included in the
111 section entitled ``GNU Free Documentation License''.
116 Permission is granted to process this file through Tex and print the
117 results, provided the printed document carries copying permission
118 notice identical to this one except for the removal of this paragraph
119 (this paragraph not being relevant to the printed manual).
125 @title Using @value{AS}
126 @subtitle The @sc{gnu} Assembler
128 @subtitle for the @value{TARGET} family
130 @ifset VERSION_PACKAGE
132 @subtitle @value{VERSION_PACKAGE}
135 @subtitle Version @value{VERSION}
138 The Free Software Foundation Inc.@: thanks The Nice Computer
139 Company of Australia for loaning Dean Elsner to write the
140 first (Vax) version of @command{as} for Project @sc{gnu}.
141 The proprietors, management and staff of TNCCA thank FSF for
142 distracting the boss while they got some work
145 @author Dean Elsner, Jay Fenlason & friends
149 \hfill {\it Using {\tt @value{AS}}}\par
150 \hfill Edited by Cygnus Support\par
152 %"boxit" macro for figures:
153 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
154 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
155 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
156 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
157 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
160 @vskip 0pt plus 1filll
161 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002,
162 2006, 2007 Free Software Foundation, Inc.
164 Permission is granted to copy, distribute and/or modify this document
165 under the terms of the GNU Free Documentation License, Version 1.1
166 or any later version published by the Free Software Foundation;
167 with no Invariant Sections, with no Front-Cover Texts, and with no
168 Back-Cover Texts. A copy of the license is included in the
169 section entitled ``GNU Free Documentation License''.
176 @top Using @value{AS}
178 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}}
179 @ifset VERSION_PACKAGE
180 @value{VERSION_PACKAGE}
182 version @value{VERSION}.
184 This version of the file describes @command{@value{AS}} configured to generate
185 code for @value{TARGET} architectures.
188 This document is distributed under the terms of the GNU Free
189 Documentation License. A copy of the license is included in the
190 section entitled ``GNU Free Documentation License''.
193 * Overview:: Overview
194 * Invoking:: Command-Line Options
196 * Sections:: Sections and Relocation
198 * Expressions:: Expressions
199 * Pseudo Ops:: Assembler Directives
200 * Machine Dependencies:: Machine Dependent Features
201 * Reporting Bugs:: Reporting Bugs
202 * Acknowledgements:: Who Did What
203 * GNU Free Documentation License:: GNU Free Documentation License
204 * AS Index:: AS Index
211 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
213 This version of the manual describes @command{@value{AS}} configured to generate
214 code for @value{TARGET} architectures.
218 @cindex invocation summary
219 @cindex option summary
220 @cindex summary of options
221 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
222 see @ref{Invoking,,Command-Line Options}.
224 @c man title AS the portable GNU assembler.
228 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
232 @c We don't use deffn and friends for the following because they seem
233 @c to be limited to one line for the header.
235 @c man begin SYNOPSIS
236 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
237 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
238 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
239 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
240 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
241 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
242 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
243 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
244 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
245 [@b{--target-help}] [@var{target-options}]
246 [@b{--}|@var{files} @dots{}]
248 @c Target dependent options are listed below. Keep the list sorted.
249 @c Add an empty line for separation.
252 @emph{Target Alpha options:}
254 [@b{-mdebug} | @b{-no-mdebug}]
255 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
256 [@b{-F}] [@b{-32addr}]
260 @emph{Target ARC options:}
266 @emph{Target ARM options:}
267 @c Don't document the deprecated options
268 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
269 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
270 [@b{-mfpu}=@var{floating-point-format}]
271 [@b{-mfloat-abi}=@var{abi}]
272 [@b{-meabi}=@var{ver}]
275 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
276 @b{-mapcs-reentrant}]
277 [@b{-mthumb-interwork}] [@b{-k}]
281 @emph{Target CRIS options:}
282 [@b{--underscore} | @b{--no-underscore}]
284 [@b{--emulation=criself} | @b{--emulation=crisaout}]
285 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
286 @c Deprecated -- deliberately not documented.
291 @emph{Target D10V options:}
296 @emph{Target D30V options:}
297 [@b{-O}|@b{-n}|@b{-N}]
300 @c Renesas family chips have no machine-dependent assembler options
303 @c HPPA has no machine-dependent assembler options (yet).
307 @emph{Target i386 options:}
308 [@b{--32}|@b{--64}] [@b{-n}]
309 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}]
313 @emph{Target i960 options:}
314 @c see md_parse_option in tc-i960.c
315 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
317 [@b{-b}] [@b{-no-relax}]
321 @emph{Target IA-64 options:}
322 [@b{-mconstant-gp}|@b{-mauto-pic}]
323 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
325 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
326 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
327 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
328 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
332 @emph{Target IP2K options:}
333 [@b{-mip2022}|@b{-mip2022ext}]
337 @emph{Target M32C options:}
338 [@b{-m32c}|@b{-m16c}]
342 @emph{Target M32R options:}
343 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
348 @emph{Target M680X0 options:}
349 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
353 @emph{Target M68HC11 options:}
354 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
355 [@b{-mshort}|@b{-mlong}]
356 [@b{-mshort-double}|@b{-mlong-double}]
357 [@b{--force-long-branches}] [@b{--short-branches}]
358 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
359 [@b{--print-opcodes}] [@b{--generate-example}]
363 @emph{Target MCORE options:}
364 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
365 [@b{-mcpu=[210|340]}]
369 @emph{Target MIPS options:}
370 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
371 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
372 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
373 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
374 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
375 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
376 [@b{-mips64}] [@b{-mips64r2}]
377 [@b{-construct-floats}] [@b{-no-construct-floats}]
378 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
379 [@b{-mfix7000}] [@b{-mno-fix7000}]
380 [@b{-mips16}] [@b{-no-mips16}]
381 [@b{-msmartmips}] [@b{-mno-smartmips}]
382 [@b{-mips3d}] [@b{-no-mips3d}]
383 [@b{-mdmx}] [@b{-no-mdmx}]
384 [@b{-mdsp}] [@b{-mno-dsp}]
385 [@b{-mdspr2}] [@b{-mno-dspr2}]
386 [@b{-mmt}] [@b{-mno-mt}]
387 [@b{-mdebug}] [@b{-no-mdebug}]
388 [@b{-mpdr}] [@b{-mno-pdr}]
392 @emph{Target MMIX options:}
393 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
394 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
395 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
396 [@b{--linker-allocated-gregs}]
400 @emph{Target PDP11 options:}
401 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
402 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
403 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
407 @emph{Target picoJava options:}
412 @emph{Target PowerPC options:}
413 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
414 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
415 @b{-mbooke32}|@b{-mbooke64}]
416 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
417 [@b{-mregnames}|@b{-mno-regnames}]
418 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
419 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
420 [@b{-msolaris}|@b{-mno-solaris}]
424 @emph{Target SPARC options:}
425 @c The order here is important. See c-sparc.texi.
426 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
427 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
428 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
433 @emph{Target TIC54X options:}
434 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
435 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
440 @emph{Target Z80 options:}
441 [@b{-z80}] [@b{-r800}]
442 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
443 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
444 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
445 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
446 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
447 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
451 @c Z8000 has no machine-dependent assembler options
455 @emph{Target Xtensa options:}
456 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
457 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
458 [@b{--[no-]transform}]
459 [@b{--rename-section} @var{oldname}=@var{newname}]
467 @include at-file.texi
470 Turn on listings, in any of a variety of ways:
474 omit false conditionals
477 omit debugging directives
480 include high-level source
486 include macro expansions
489 omit forms processing
495 set the name of the listing file
498 You may combine these options; for example, use @samp{-aln} for assembly
499 listing without forms processing. The @samp{=file} option, if used, must be
500 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
503 Begin in alternate macro mode.
505 @xref{Altmacro,,@code{.altmacro}}.
509 Ignored. This option is accepted for script compatibility with calls to
512 @item --defsym @var{sym}=@var{value}
513 Define the symbol @var{sym} to be @var{value} before assembling the input file.
514 @var{value} must be an integer constant. As in C, a leading @samp{0x}
515 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
516 value. The value of the symbol can be overridden inside a source file via the
517 use of a @code{.set} pseudo-op.
520 ``fast''---skip whitespace and comment preprocessing (assume source is
525 Generate debugging information for each assembler source line using whichever
526 debug format is preferred by the target. This currently means either STABS,
530 Generate stabs debugging information for each assembler line. This
531 may help debugging assembler code, if the debugger can handle it.
534 Generate stabs debugging information for each assembler line, with GNU
535 extensions that probably only gdb can handle, and that could make other
536 debuggers crash or refuse to read your program. This
537 may help debugging assembler code. Currently the only GNU extension is
538 the location of the current working directory at assembling time.
541 Generate DWARF2 debugging information for each assembler line. This
542 may help debugging assembler code, if the debugger can handle it. Note---this
543 option is only supported by some targets, not all of them.
546 Print a summary of the command line options and exit.
549 Print a summary of all target specific options and exit.
552 Add directory @var{dir} to the search list for @code{.include} directives.
555 Don't warn about signed overflow.
558 @ifclear DIFF-TBL-KLUGE
559 This option is accepted but has no effect on the @value{TARGET} family.
561 @ifset DIFF-TBL-KLUGE
562 Issue warnings when difference tables altered for long displacements.
567 Keep (in the symbol table) local symbols. These symbols start with
568 system-specific local label prefixes, typically @samp{.L} for ELF systems
569 or @samp{L} for traditional a.out systems.
574 @item --listing-lhs-width=@var{number}
575 Set the maximum width, in words, of the output data column for an assembler
576 listing to @var{number}.
578 @item --listing-lhs-width2=@var{number}
579 Set the maximum width, in words, of the output data column for continuation
580 lines in an assembler listing to @var{number}.
582 @item --listing-rhs-width=@var{number}
583 Set the maximum width of an input source line, as displayed in a listing, to
586 @item --listing-cont-lines=@var{number}
587 Set the maximum number of lines printed in a listing for a single line of input
590 @item -o @var{objfile}
591 Name the object-file output from @command{@value{AS}} @var{objfile}.
594 Fold the data section into the text section.
596 @kindex --hash-size=@var{number}
597 Set the default size of GAS's hash tables to a prime number close to
598 @var{number}. Increasing this value can reduce the length of time it takes the
599 assembler to perform its tasks, at the expense of increasing the assembler's
600 memory requirements. Similarly reducing this value can reduce the memory
601 requirements at the expense of speed.
603 @item --reduce-memory-overheads
604 This option reduces GAS's memory requirements, at the expense of making the
605 assembly processes slower. Currently this switch is a synonym for
606 @samp{--hash-size=4051}, but in the future it may have other effects as well.
609 Print the maximum space (in bytes) and total time (in seconds) used by
612 @item --strip-local-absolute
613 Remove local absolute symbols from the outgoing symbol table.
617 Print the @command{as} version.
620 Print the @command{as} version and exit.
624 Suppress warning messages.
626 @item --fatal-warnings
627 Treat warnings as errors.
630 Don't suppress warning messages or treat them as errors.
639 Generate an object file even after errors.
641 @item -- | @var{files} @dots{}
642 Standard input, or source files to assemble.
647 The following options are available when @value{AS} is configured for
652 This option selects the core processor variant.
654 Select either big-endian (-EB) or little-endian (-EL) output.
659 The following options are available when @value{AS} is configured for the ARM
663 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
664 Specify which ARM processor variant is the target.
665 @item -march=@var{architecture}[+@var{extension}@dots{}]
666 Specify which ARM architecture variant is used by the target.
667 @item -mfpu=@var{floating-point-format}
668 Select which Floating Point architecture is the target.
669 @item -mfloat-abi=@var{abi}
670 Select which floating point ABI is in use.
672 Enable Thumb only instruction decoding.
673 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
674 Select which procedure calling convention is in use.
676 Select either big-endian (-EB) or little-endian (-EL) output.
677 @item -mthumb-interwork
678 Specify that the code has been generated with interworking between Thumb and
681 Specify that PIC code has been generated.
686 See the info pages for documentation of the CRIS-specific options.
690 The following options are available when @value{AS} is configured for
693 @cindex D10V optimization
694 @cindex optimization, D10V
696 Optimize output by parallelizing instructions.
701 The following options are available when @value{AS} is configured for a D30V
704 @cindex D30V optimization
705 @cindex optimization, D30V
707 Optimize output by parallelizing instructions.
711 Warn when nops are generated.
713 @cindex D30V nops after 32-bit multiply
715 Warn when a nop after a 32-bit multiply instruction is generated.
720 The following options are available when @value{AS} is configured for the
721 Intel 80960 processor.
724 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
725 Specify which variant of the 960 architecture is the target.
728 Add code to collect statistics about branches taken.
731 Do not alter compare-and-branch instructions for long displacements;
738 The following options are available when @value{AS} is configured for the
744 Specifies that the extended IP2022 instructions are allowed.
747 Restores the default behaviour, which restricts the permitted instructions to
748 just the basic IP2022 ones.
754 The following options are available when @value{AS} is configured for the
755 Renesas M32C and M16C processors.
760 Assemble M32C instructions.
763 Assemble M16C instructions (the default).
769 The following options are available when @value{AS} is configured for the
770 Renesas M32R (formerly Mitsubishi M32R) series.
775 Specify which processor in the M32R family is the target. The default
776 is normally the M32R, but this option changes it to the M32RX.
778 @item --warn-explicit-parallel-conflicts or --Wp
779 Produce warning messages when questionable parallel constructs are
782 @item --no-warn-explicit-parallel-conflicts or --Wnp
783 Do not produce warning messages when questionable parallel constructs are
790 The following options are available when @value{AS} is configured for the
791 Motorola 68000 series.
796 Shorten references to undefined symbols, to one word instead of two.
798 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
799 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
800 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
801 Specify what processor in the 68000 family is the target. The default
802 is normally the 68020, but this can be changed at configuration time.
804 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
805 The target machine does (or does not) have a floating-point coprocessor.
806 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
807 the basic 68000 is not compatible with the 68881, a combination of the
808 two can be specified, since it's possible to do emulation of the
809 coprocessor instructions with the main processor.
811 @item -m68851 | -mno-68851
812 The target machine does (or does not) have a memory-management
813 unit coprocessor. The default is to assume an MMU for 68020 and up.
820 For details about the PDP-11 machine dependent features options,
821 see @ref{PDP-11-Options}.
824 @item -mpic | -mno-pic
825 Generate position-independent (or position-dependent) code. The
826 default is @option{-mpic}.
829 @itemx -mall-extensions
830 Enable all instruction set extensions. This is the default.
832 @item -mno-extensions
833 Disable all instruction set extensions.
835 @item -m@var{extension} | -mno-@var{extension}
836 Enable (or disable) a particular instruction set extension.
839 Enable the instruction set extensions supported by a particular CPU, and
840 disable all other extensions.
842 @item -m@var{machine}
843 Enable the instruction set extensions supported by a particular machine
844 model, and disable all other extensions.
850 The following options are available when @value{AS} is configured for
851 a picoJava processor.
855 @cindex PJ endianness
856 @cindex endianness, PJ
857 @cindex big endian output, PJ
859 Generate ``big endian'' format output.
861 @cindex little endian output, PJ
863 Generate ``little endian'' format output.
869 The following options are available when @value{AS} is configured for the
870 Motorola 68HC11 or 68HC12 series.
874 @item -m68hc11 | -m68hc12 | -m68hcs12
875 Specify what processor is the target. The default is
876 defined by the configuration option when building the assembler.
879 Specify to use the 16-bit integer ABI.
882 Specify to use the 32-bit integer ABI.
885 Specify to use the 32-bit double ABI.
888 Specify to use the 64-bit double ABI.
890 @item --force-long-branches
891 Relative branches are turned into absolute ones. This concerns
892 conditional branches, unconditional branches and branches to a
895 @item -S | --short-branches
896 Do not turn relative branches into absolute ones
897 when the offset is out of range.
899 @item --strict-direct-mode
900 Do not turn the direct addressing mode into extended addressing mode
901 when the instruction does not support direct addressing mode.
903 @item --print-insn-syntax
904 Print the syntax of instruction in case of error.
906 @item --print-opcodes
907 print the list of instructions with syntax and then exit.
909 @item --generate-example
910 print an example of instruction for each possible instruction and then exit.
911 This option is only useful for testing @command{@value{AS}}.
917 The following options are available when @command{@value{AS}} is configured
918 for the SPARC architecture:
921 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
922 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
923 Explicitly select a variant of the SPARC architecture.
925 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
926 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
928 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
929 UltraSPARC extensions.
931 @item -xarch=v8plus | -xarch=v8plusa
932 For compatibility with the Solaris v9 assembler. These options are
933 equivalent to -Av8plus and -Av8plusa, respectively.
936 Warn when the assembler switches to another architecture.
941 The following options are available when @value{AS} is configured for the 'c54x
946 Enable extended addressing mode. All addresses and relocations will assume
947 extended addressing (usually 23 bits).
948 @item -mcpu=@var{CPU_VERSION}
949 Sets the CPU version being compiled for.
950 @item -merrors-to-file @var{FILENAME}
951 Redirect error output to a file, for broken systems which don't support such
952 behaviour in the shell.
957 The following options are available when @value{AS} is configured for
958 a @sc{mips} processor.
962 This option sets the largest size of an object that can be referenced
963 implicitly with the @code{gp} register. It is only accepted for targets that
964 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
966 @cindex MIPS endianness
967 @cindex endianness, MIPS
968 @cindex big endian output, MIPS
970 Generate ``big endian'' format output.
972 @cindex little endian output, MIPS
974 Generate ``little endian'' format output.
986 Generate code for a particular @sc{mips} Instruction Set Architecture level.
987 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
988 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
989 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
990 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
992 correspond to generic
993 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
994 and @samp{MIPS64 Release 2}
995 ISA processors, respectively.
997 @item -march=@var{CPU}
998 Generate code for a particular @sc{mips} cpu.
1000 @item -mtune=@var{cpu}
1001 Schedule and tune for a particular @sc{mips} cpu.
1005 Cause nops to be inserted if the read of the destination register
1006 of an mfhi or mflo instruction occurs in the following two instructions.
1010 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1011 section instead of the standard ELF .stabs sections.
1015 Control generation of @code{.pdr} sections.
1019 The register sizes are normally inferred from the ISA and ABI, but these
1020 flags force a certain group of registers to be treated as 32 bits wide at
1021 all times. @samp{-mgp32} controls the size of general-purpose registers
1022 and @samp{-mfp32} controls the size of floating-point registers.
1026 Generate code for the MIPS 16 processor. This is equivalent to putting
1027 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1028 turns off this option.
1031 @itemx -mno-smartmips
1032 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1033 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1034 @samp{-mno-smartmips} turns off this option.
1038 Generate code for the MIPS-3D Application Specific Extension.
1039 This tells the assembler to accept MIPS-3D instructions.
1040 @samp{-no-mips3d} turns off this option.
1044 Generate code for the MDMX Application Specific Extension.
1045 This tells the assembler to accept MDMX instructions.
1046 @samp{-no-mdmx} turns off this option.
1050 Generate code for the DSP Release 1 Application Specific Extension.
1051 This tells the assembler to accept DSP Release 1 instructions.
1052 @samp{-mno-dsp} turns off this option.
1056 Generate code for the DSP Release 2 Application Specific Extension.
1057 This option implies -mdsp.
1058 This tells the assembler to accept DSP Release 2 instructions.
1059 @samp{-mno-dspr2} turns off this option.
1063 Generate code for the MT Application Specific Extension.
1064 This tells the assembler to accept MT instructions.
1065 @samp{-mno-mt} turns off this option.
1067 @item --construct-floats
1068 @itemx --no-construct-floats
1069 The @samp{--no-construct-floats} option disables the construction of
1070 double width floating point constants by loading the two halves of the
1071 value into the two single width floating point registers that make up
1072 the double width register. By default @samp{--construct-floats} is
1073 selected, allowing construction of these floating point constants.
1076 @item --emulation=@var{name}
1077 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1078 for some other target, in all respects, including output format (choosing
1079 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1080 debugging information or store symbol table information, and default
1081 endianness. The available configuration names are: @samp{mipsecoff},
1082 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1083 @samp{mipsbelf}. The first two do not alter the default endianness from that
1084 of the primary target for which the assembler was configured; the others change
1085 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1086 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1087 selection in any case.
1089 This option is currently supported only when the primary target
1090 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1091 Furthermore, the primary target or others specified with
1092 @samp{--enable-targets=@dots{}} at configuration time must include support for
1093 the other format, if both are to be available. For example, the Irix 5
1094 configuration includes support for both.
1096 Eventually, this option will support more configurations, with more
1097 fine-grained control over the assembler's behavior, and will be supported for
1101 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1108 Control how to deal with multiplication overflow and division by zero.
1109 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1110 (and only work for Instruction Set Architecture level 2 and higher);
1111 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1115 When this option is used, @command{@value{AS}} will issue a warning every
1116 time it generates a nop instruction from a macro.
1121 The following options are available when @value{AS} is configured for
1127 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1128 The command line option @samp{-nojsri2bsr} can be used to disable it.
1132 Enable or disable the silicon filter behaviour. By default this is disabled.
1133 The default can be overridden by the @samp{-sifilter} command line option.
1136 Alter jump instructions for long displacements.
1138 @item -mcpu=[210|340]
1139 Select the cpu type on the target hardware. This controls which instructions
1143 Assemble for a big endian target.
1146 Assemble for a little endian target.
1152 See the info pages for documentation of the MMIX-specific options.
1156 The following options are available when @value{AS} is configured for
1157 an Xtensa processor.
1160 @item --text-section-literals | --no-text-section-literals
1161 With @option{--text-@-section-@-literals}, literal pools are interspersed
1162 in the text section. The default is
1163 @option{--no-@-text-@-section-@-literals}, which places literals in a
1164 separate section in the output file. These options only affect literals
1165 referenced via PC-relative @code{L32R} instructions; literals for
1166 absolute mode @code{L32R} instructions are handled separately.
1168 @item --absolute-literals | --no-absolute-literals
1169 Indicate to the assembler whether @code{L32R} instructions use absolute
1170 or PC-relative addressing. The default is to assume absolute addressing
1171 if the Xtensa processor includes the absolute @code{L32R} addressing
1172 option. Otherwise, only the PC-relative @code{L32R} mode can be used.
1174 @item --target-align | --no-target-align
1175 Enable or disable automatic alignment to reduce branch penalties at the
1176 expense of some code density. The default is @option{--target-@-align}.
1178 @item --longcalls | --no-longcalls
1179 Enable or disable transformation of call instructions to allow calls
1180 across a greater range of addresses. The default is
1181 @option{--no-@-longcalls}.
1183 @item --transform | --no-transform
1184 Enable or disable all assembler transformations of Xtensa instructions.
1185 The default is @option{--transform};
1186 @option{--no-transform} should be used only in the rare cases when the
1187 instructions must be exactly as specified in the assembly source.
1192 The following options are available when @value{AS} is configured for
1193 a Z80 family processor.
1196 Assemble for Z80 processor.
1198 Assemble for R800 processor.
1199 @item -ignore-undocumented-instructions
1201 Assemble undocumented Z80 instructions that also work on R800 without warning.
1202 @item -ignore-unportable-instructions
1204 Assemble all undocumented Z80 instructions without warning.
1205 @item -warn-undocumented-instructions
1207 Issue a warning for undocumented Z80 instructions that also work on R800.
1208 @item -warn-unportable-instructions
1210 Issue a warning for undocumented Z80 instructions that do not work on R800.
1211 @item -forbid-undocumented-instructions
1213 Treat all undocumented instructions as errors.
1214 @item -forbid-unportable-instructions
1216 Treat undocumented Z80 instructions that do not work on R800 as errors.
1223 * Manual:: Structure of this Manual
1224 * GNU Assembler:: The GNU Assembler
1225 * Object Formats:: Object File Formats
1226 * Command Line:: Command Line
1227 * Input Files:: Input Files
1228 * Object:: Output (Object) File
1229 * Errors:: Error and Warning Messages
1233 @section Structure of this Manual
1235 @cindex manual, structure and purpose
1236 This manual is intended to describe what you need to know to use
1237 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1238 notation for symbols, constants, and expressions; the directives that
1239 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1242 We also cover special features in the @value{TARGET}
1243 configuration of @command{@value{AS}}, including assembler directives.
1246 This manual also describes some of the machine-dependent features of
1247 various flavors of the assembler.
1250 @cindex machine instructions (not covered)
1251 On the other hand, this manual is @emph{not} intended as an introduction
1252 to programming in assembly language---let alone programming in general!
1253 In a similar vein, we make no attempt to introduce the machine
1254 architecture; we do @emph{not} describe the instruction set, standard
1255 mnemonics, registers or addressing modes that are standard to a
1256 particular architecture.
1258 You may want to consult the manufacturer's
1259 machine architecture manual for this information.
1263 For information on the H8/300 machine instruction set, see @cite{H8/300
1264 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1265 Programming Manual} (Renesas).
1268 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1269 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1270 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1271 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1274 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1278 @c I think this is premature---doc@cygnus.com, 17jan1991
1280 Throughout this manual, we assume that you are running @dfn{GNU},
1281 the portable operating system from the @dfn{Free Software
1282 Foundation, Inc.}. This restricts our attention to certain kinds of
1283 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1284 once this assumption is granted examples and definitions need less
1287 @command{@value{AS}} is part of a team of programs that turn a high-level
1288 human-readable series of instructions into a low-level
1289 computer-readable series of instructions. Different versions of
1290 @command{@value{AS}} are used for different kinds of computer.
1293 @c There used to be a section "Terminology" here, which defined
1294 @c "contents", "byte", "word", and "long". Defining "word" to any
1295 @c particular size is confusing when the .word directive may generate 16
1296 @c bits on one machine and 32 bits on another; in general, for the user
1297 @c version of this manual, none of these terms seem essential to define.
1298 @c They were used very little even in the former draft of the manual;
1299 @c this draft makes an effort to avoid them (except in names of
1303 @section The GNU Assembler
1305 @c man begin DESCRIPTION
1307 @sc{gnu} @command{as} is really a family of assemblers.
1309 This manual describes @command{@value{AS}}, a member of that family which is
1310 configured for the @value{TARGET} architectures.
1312 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1313 should find a fairly similar environment when you use it on another
1314 architecture. Each version has much in common with the others,
1315 including object file formats, most assembler directives (often called
1316 @dfn{pseudo-ops}) and assembler syntax.@refill
1318 @cindex purpose of @sc{gnu} assembler
1319 @command{@value{AS}} is primarily intended to assemble the output of the
1320 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1321 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1322 assemble correctly everything that other assemblers for the same
1323 machine would assemble.
1325 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1328 @c This remark should appear in generic version of manual; assumption
1329 @c here is that generic version sets M680x0.
1330 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1331 assembler for the same architecture; for example, we know of several
1332 incompatible versions of 680x0 assembly language syntax.
1337 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1338 program in one pass of the source file. This has a subtle impact on the
1339 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1341 @node Object Formats
1342 @section Object File Formats
1344 @cindex object file format
1345 The @sc{gnu} assembler can be configured to produce several alternative
1346 object file formats. For the most part, this does not affect how you
1347 write assembly language programs; but directives for debugging symbols
1348 are typically different in different file formats. @xref{Symbol
1349 Attributes,,Symbol Attributes}.
1352 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1353 @value{OBJ-NAME} format object files.
1355 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1357 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1358 @code{b.out} or COFF format object files.
1361 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1362 SOM or ELF format object files.
1367 @section Command Line
1369 @cindex command line conventions
1371 After the program name @command{@value{AS}}, the command line may contain
1372 options and file names. Options may appear in any order, and may be
1373 before, after, or between file names. The order of file names is
1376 @cindex standard input, as input file
1378 @file{--} (two hyphens) by itself names the standard input file
1379 explicitly, as one of the files for @command{@value{AS}} to assemble.
1381 @cindex options, command line
1382 Except for @samp{--} any command line argument that begins with a
1383 hyphen (@samp{-}) is an option. Each option changes the behavior of
1384 @command{@value{AS}}. No option changes the way another option works. An
1385 option is a @samp{-} followed by one or more letters; the case of
1386 the letter is important. All options are optional.
1388 Some options expect exactly one file name to follow them. The file
1389 name may either immediately follow the option's letter (compatible
1390 with older assemblers) or it may be the next command argument (@sc{gnu}
1391 standard). These two command lines are equivalent:
1394 @value{AS} -o my-object-file.o mumble.s
1395 @value{AS} -omy-object-file.o mumble.s
1399 @section Input Files
1402 @cindex source program
1403 @cindex files, input
1404 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1405 describe the program input to one run of @command{@value{AS}}. The program may
1406 be in one or more files; how the source is partitioned into files
1407 doesn't change the meaning of the source.
1409 @c I added "con" prefix to "catenation" just to prove I can overcome my
1410 @c APL training... doc@cygnus.com
1411 The source program is a concatenation of the text in all the files, in the
1414 @c man begin DESCRIPTION
1415 Each time you run @command{@value{AS}} it assembles exactly one source
1416 program. The source program is made up of one or more files.
1417 (The standard input is also a file.)
1419 You give @command{@value{AS}} a command line that has zero or more input file
1420 names. The input files are read (from left file name to right). A
1421 command line argument (in any position) that has no special meaning
1422 is taken to be an input file name.
1424 If you give @command{@value{AS}} no file names it attempts to read one input file
1425 from the @command{@value{AS}} standard input, which is normally your terminal. You
1426 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1429 Use @samp{--} if you need to explicitly name the standard input file
1430 in your command line.
1432 If the source is empty, @command{@value{AS}} produces a small, empty object
1437 @subheading Filenames and Line-numbers
1439 @cindex input file linenumbers
1440 @cindex line numbers, in input files
1441 There are two ways of locating a line in the input file (or files) and
1442 either may be used in reporting error messages. One way refers to a line
1443 number in a physical file; the other refers to a line number in a
1444 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1446 @dfn{Physical files} are those files named in the command line given
1447 to @command{@value{AS}}.
1449 @dfn{Logical files} are simply names declared explicitly by assembler
1450 directives; they bear no relation to physical files. Logical file names help
1451 error messages reflect the original source file, when @command{@value{AS}} source
1452 is itself synthesized from other files. @command{@value{AS}} understands the
1453 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1454 @ref{File,,@code{.file}}.
1457 @section Output (Object) File
1463 Every time you run @command{@value{AS}} it produces an output file, which is
1464 your assembly language program translated into numbers. This file
1465 is the object file. Its default name is
1473 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1475 You can give it another name by using the @option{-o} option. Conventionally,
1476 object file names end with @file{.o}. The default name is used for historical
1477 reasons: older assemblers were capable of assembling self-contained programs
1478 directly into a runnable program. (For some formats, this isn't currently
1479 possible, but it can be done for the @code{a.out} format.)
1483 The object file is meant for input to the linker @code{@value{LD}}. It contains
1484 assembled program code, information to help @code{@value{LD}} integrate
1485 the assembled program into a runnable file, and (optionally) symbolic
1486 information for the debugger.
1488 @c link above to some info file(s) like the description of a.out.
1489 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1492 @section Error and Warning Messages
1494 @c man begin DESCRIPTION
1496 @cindex error messages
1497 @cindex warning messages
1498 @cindex messages from assembler
1499 @command{@value{AS}} may write warnings and error messages to the standard error
1500 file (usually your terminal). This should not happen when a compiler
1501 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1502 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1503 grave problem that stops the assembly.
1507 @cindex format of warning messages
1508 Warning messages have the format
1511 file_name:@b{NNN}:Warning Message Text
1515 @cindex line numbers, in warnings/errors
1516 (where @b{NNN} is a line number). If a logical file name has been given
1517 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1518 the current input file is used. If a logical line number was given
1520 (@pxref{Line,,@code{.line}})
1522 then it is used to calculate the number printed,
1523 otherwise the actual line in the current source file is printed. The
1524 message text is intended to be self explanatory (in the grand Unix
1527 @cindex format of error messages
1528 Error messages have the format
1530 file_name:@b{NNN}:FATAL:Error Message Text
1532 The file name and line number are derived as for warning
1533 messages. The actual message text may be rather less explanatory
1534 because many of them aren't supposed to happen.
1537 @chapter Command-Line Options
1539 @cindex options, all versions of assembler
1540 This chapter describes command-line options available in @emph{all}
1541 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
1542 for options specific
1544 to the @value{TARGET} target.
1547 to particular machine architectures.
1550 @c man begin DESCRIPTION
1552 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1553 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1554 The assembler arguments must be separated from each other (and the @samp{-Wa})
1555 by commas. For example:
1558 gcc -c -g -O -Wa,-alh,-L file.c
1562 This passes two options to the assembler: @samp{-alh} (emit a listing to
1563 standard output with high-level and assembly source) and @samp{-L} (retain
1564 local symbols in the symbol table).
1566 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1567 command-line options are automatically passed to the assembler by the compiler.
1568 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1569 precisely what options it passes to each compilation pass, including the
1575 * a:: -a[cdhlns] enable listings
1576 * alternate:: --alternate enable alternate macro syntax
1577 * D:: -D for compatibility
1578 * f:: -f to work faster
1579 * I:: -I for .include search path
1580 @ifclear DIFF-TBL-KLUGE
1581 * K:: -K for compatibility
1583 @ifset DIFF-TBL-KLUGE
1584 * K:: -K for difference tables
1587 * L:: -L to retain local symbols
1588 * listing:: --listing-XXX to configure listing output
1589 * M:: -M or --mri to assemble in MRI compatibility mode
1590 * MD:: --MD for dependency tracking
1591 * o:: -o to name the object file
1592 * R:: -R to join data and text sections
1593 * statistics:: --statistics to see statistics about assembly
1594 * traditional-format:: --traditional-format for compatible output
1595 * v:: -v to announce version
1596 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1597 * Z:: -Z to make object file even after errors
1601 @section Enable Listings: @option{-a[cdhlns]}
1610 @cindex listings, enabling
1611 @cindex assembly listings, enabling
1613 These options enable listing output from the assembler. By itself,
1614 @samp{-a} requests high-level, assembly, and symbols listing.
1615 You can use other letters to select specific options for the list:
1616 @samp{-ah} requests a high-level language listing,
1617 @samp{-al} requests an output-program assembly listing, and
1618 @samp{-as} requests a symbol table listing.
1619 High-level listings require that a compiler debugging option like
1620 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1623 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1624 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1625 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1626 omitted from the listing.
1628 Use the @samp{-ad} option to omit debugging directives from the
1631 Once you have specified one of these options, you can further control
1632 listing output and its appearance using the directives @code{.list},
1633 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1635 The @samp{-an} option turns off all forms processing.
1636 If you do not request listing output with one of the @samp{-a} options, the
1637 listing-control directives have no effect.
1639 The letters after @samp{-a} may be combined into one option,
1640 @emph{e.g.}, @samp{-aln}.
1642 Note if the assembler source is coming from the standard input (e.g.,
1644 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1645 is being used) then the listing will not contain any comments or preprocessor
1646 directives. This is because the listing code buffers input source lines from
1647 stdin only after they have been preprocessed by the assembler. This reduces
1648 memory usage and makes the code more efficient.
1651 @section @option{--alternate}
1654 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1657 @section @option{-D}
1660 This option has no effect whatsoever, but it is accepted to make it more
1661 likely that scripts written for other assemblers also work with
1662 @command{@value{AS}}.
1665 @section Work Faster: @option{-f}
1668 @cindex trusted compiler
1669 @cindex faster processing (@option{-f})
1670 @samp{-f} should only be used when assembling programs written by a
1671 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1672 and comment preprocessing on
1673 the input file(s) before assembling them. @xref{Preprocessing,
1677 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1678 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1683 @section @code{.include} Search Path: @option{-I} @var{path}
1685 @kindex -I @var{path}
1686 @cindex paths for @code{.include}
1687 @cindex search path for @code{.include}
1688 @cindex @code{include} directive search path
1689 Use this option to add a @var{path} to the list of directories
1690 @command{@value{AS}} searches for files specified in @code{.include}
1691 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1692 many times as necessary to include a variety of paths. The current
1693 working directory is always searched first; after that, @command{@value{AS}}
1694 searches any @samp{-I} directories in the same order as they were
1695 specified (left to right) on the command line.
1698 @section Difference Tables: @option{-K}
1701 @ifclear DIFF-TBL-KLUGE
1702 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1703 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1704 where it can be used to warn when the assembler alters the machine code
1705 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1706 family does not have the addressing limitations that sometimes lead to this
1707 alteration on other platforms.
1710 @ifset DIFF-TBL-KLUGE
1711 @cindex difference tables, warning
1712 @cindex warning for altered difference tables
1713 @command{@value{AS}} sometimes alters the code emitted for directives of the
1714 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
1715 You can use the @samp{-K} option if you want a warning issued when this
1720 @section Include Local Symbols: @option{-L}
1723 @cindex local symbols, retaining in output
1724 Symbols beginning with system-specific local label prefixes, typically
1725 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
1726 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
1727 such symbols when debugging, because they are intended for the use of
1728 programs (like compilers) that compose assembler programs, not for your
1729 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
1730 such symbols, so you do not normally debug with them.
1732 This option tells @command{@value{AS}} to retain those local symbols
1733 in the object file. Usually if you do this you also tell the linker
1734 @code{@value{LD}} to preserve those symbols.
1737 @section Configuring listing output: @option{--listing}
1739 The listing feature of the assembler can be enabled via the command line switch
1740 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1741 hex dump of the corresponding locations in the output object file, and displays
1742 them as a listing file. The format of this listing can be controlled by
1743 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
1744 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
1745 @code{.psize} (@pxref{Psize}), and
1746 @code{.eject} (@pxref{Eject}) and also by the following switches:
1749 @item --listing-lhs-width=@samp{number}
1750 @kindex --listing-lhs-width
1751 @cindex Width of first line disassembly output
1752 Sets the maximum width, in words, of the first line of the hex byte dump. This
1753 dump appears on the left hand side of the listing output.
1755 @item --listing-lhs-width2=@samp{number}
1756 @kindex --listing-lhs-width2
1757 @cindex Width of continuation lines of disassembly output
1758 Sets the maximum width, in words, of any further lines of the hex byte dump for
1759 a given input source line. If this value is not specified, it defaults to being
1760 the same as the value specified for @samp{--listing-lhs-width}. If neither
1761 switch is used the default is to one.
1763 @item --listing-rhs-width=@samp{number}
1764 @kindex --listing-rhs-width
1765 @cindex Width of source line output
1766 Sets the maximum width, in characters, of the source line that is displayed
1767 alongside the hex dump. The default value for this parameter is 100. The
1768 source line is displayed on the right hand side of the listing output.
1770 @item --listing-cont-lines=@samp{number}
1771 @kindex --listing-cont-lines
1772 @cindex Maximum number of continuation lines
1773 Sets the maximum number of continuation lines of hex dump that will be
1774 displayed for a given single line of source input. The default value is 4.
1778 @section Assemble in MRI Compatibility Mode: @option{-M}
1781 @cindex MRI compatibility mode
1782 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1783 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1784 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1785 configured target) assembler from Microtec Research. The exact nature of the
1786 MRI syntax will not be documented here; see the MRI manuals for more
1787 information. Note in particular that the handling of macros and macro
1788 arguments is somewhat different. The purpose of this option is to permit
1789 assembling existing MRI assembler code using @command{@value{AS}}.
1791 The MRI compatibility is not complete. Certain operations of the MRI assembler
1792 depend upon its object file format, and can not be supported using other object
1793 file formats. Supporting these would require enhancing each object file format
1794 individually. These are:
1797 @item global symbols in common section
1799 The m68k MRI assembler supports common sections which are merged by the linker.
1800 Other object file formats do not support this. @command{@value{AS}} handles
1801 common sections by treating them as a single common symbol. It permits local
1802 symbols to be defined within a common section, but it can not support global
1803 symbols, since it has no way to describe them.
1805 @item complex relocations
1807 The MRI assemblers support relocations against a negated section address, and
1808 relocations which combine the start addresses of two or more sections. These
1809 are not support by other object file formats.
1811 @item @code{END} pseudo-op specifying start address
1813 The MRI @code{END} pseudo-op permits the specification of a start address.
1814 This is not supported by other object file formats. The start address may
1815 instead be specified using the @option{-e} option to the linker, or in a linker
1818 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1820 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1821 name to the output file. This is not supported by other object file formats.
1823 @item @code{ORG} pseudo-op
1825 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1826 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1827 which changes the location within the current section. Absolute sections are
1828 not supported by other object file formats. The address of a section may be
1829 assigned within a linker script.
1832 There are some other features of the MRI assembler which are not supported by
1833 @command{@value{AS}}, typically either because they are difficult or because they
1834 seem of little consequence. Some of these may be supported in future releases.
1838 @item EBCDIC strings
1840 EBCDIC strings are not supported.
1842 @item packed binary coded decimal
1844 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1845 and @code{DCB.P} pseudo-ops are not supported.
1847 @item @code{FEQU} pseudo-op
1849 The m68k @code{FEQU} pseudo-op is not supported.
1851 @item @code{NOOBJ} pseudo-op
1853 The m68k @code{NOOBJ} pseudo-op is not supported.
1855 @item @code{OPT} branch control options
1857 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1858 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1859 relaxes all branches, whether forward or backward, to an appropriate size, so
1860 these options serve no purpose.
1862 @item @code{OPT} list control options
1864 The following m68k @code{OPT} list control options are ignored: @code{C},
1865 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1866 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1868 @item other @code{OPT} options
1870 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1871 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1873 @item @code{OPT} @code{D} option is default
1875 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1876 @code{OPT NOD} may be used to turn it off.
1878 @item @code{XREF} pseudo-op.
1880 The m68k @code{XREF} pseudo-op is ignored.
1882 @item @code{.debug} pseudo-op
1884 The i960 @code{.debug} pseudo-op is not supported.
1886 @item @code{.extended} pseudo-op
1888 The i960 @code{.extended} pseudo-op is not supported.
1890 @item @code{.list} pseudo-op.
1892 The various options of the i960 @code{.list} pseudo-op are not supported.
1894 @item @code{.optimize} pseudo-op
1896 The i960 @code{.optimize} pseudo-op is not supported.
1898 @item @code{.output} pseudo-op
1900 The i960 @code{.output} pseudo-op is not supported.
1902 @item @code{.setreal} pseudo-op
1904 The i960 @code{.setreal} pseudo-op is not supported.
1909 @section Dependency Tracking: @option{--MD}
1912 @cindex dependency tracking
1915 @command{@value{AS}} can generate a dependency file for the file it creates. This
1916 file consists of a single rule suitable for @code{make} describing the
1917 dependencies of the main source file.
1919 The rule is written to the file named in its argument.
1921 This feature is used in the automatic updating of makefiles.
1924 @section Name the Object File: @option{-o}
1927 @cindex naming object file
1928 @cindex object file name
1929 There is always one object file output when you run @command{@value{AS}}. By
1930 default it has the name
1933 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1947 You use this option (which takes exactly one filename) to give the
1948 object file a different name.
1950 Whatever the object file is called, @command{@value{AS}} overwrites any
1951 existing file of the same name.
1954 @section Join Data and Text Sections: @option{-R}
1957 @cindex data and text sections, joining
1958 @cindex text and data sections, joining
1959 @cindex joining text and data sections
1960 @cindex merging text and data sections
1961 @option{-R} tells @command{@value{AS}} to write the object file as if all
1962 data-section data lives in the text section. This is only done at
1963 the very last moment: your binary data are the same, but data
1964 section parts are relocated differently. The data section part of
1965 your object file is zero bytes long because all its bytes are
1966 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1968 When you specify @option{-R} it would be possible to generate shorter
1969 address displacements (because we do not have to cross between text and
1970 data section). We refrain from doing this simply for compatibility with
1971 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1974 When @command{@value{AS}} is configured for COFF or ELF output,
1975 this option is only useful if you use sections named @samp{.text} and
1980 @option{-R} is not supported for any of the HPPA targets. Using
1981 @option{-R} generates a warning from @command{@value{AS}}.
1985 @section Display Assembly Statistics: @option{--statistics}
1987 @kindex --statistics
1988 @cindex statistics, about assembly
1989 @cindex time, total for assembly
1990 @cindex space used, maximum for assembly
1991 Use @samp{--statistics} to display two statistics about the resources used by
1992 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1993 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1996 @node traditional-format
1997 @section Compatible Output: @option{--traditional-format}
1999 @kindex --traditional-format
2000 For some targets, the output of @command{@value{AS}} is different in some ways
2001 from the output of some existing assembler. This switch requests
2002 @command{@value{AS}} to use the traditional format instead.
2004 For example, it disables the exception frame optimizations which
2005 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2008 @section Announce Version: @option{-v}
2012 @cindex assembler version
2013 @cindex version of assembler
2014 You can find out what version of as is running by including the
2015 option @samp{-v} (which you can also spell as @samp{-version}) on the
2019 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2021 @command{@value{AS}} should never give a warning or error message when
2022 assembling compiler output. But programs written by people often
2023 cause @command{@value{AS}} to give a warning that a particular assumption was
2024 made. All such warnings are directed to the standard error file.
2028 @cindex suppressing warnings
2029 @cindex warnings, suppressing
2030 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2031 This only affects the warning messages: it does not change any particular of
2032 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2035 @kindex --fatal-warnings
2036 @cindex errors, caused by warnings
2037 @cindex warnings, causing error
2038 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2039 files that generate warnings to be in error.
2042 @cindex warnings, switching on
2043 You can switch these options off again by specifying @option{--warn}, which
2044 causes warnings to be output as usual.
2047 @section Generate Object File in Spite of Errors: @option{-Z}
2048 @cindex object file, after errors
2049 @cindex errors, continuing after
2050 After an error message, @command{@value{AS}} normally produces no output. If for
2051 some reason you are interested in object file output even after
2052 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2053 option. If there are any errors, @command{@value{AS}} continues anyways, and
2054 writes an object file after a final warning message of the form @samp{@var{n}
2055 errors, @var{m} warnings, generating bad object file.}
2060 @cindex machine-independent syntax
2061 @cindex syntax, machine-independent
2062 This chapter describes the machine-independent syntax allowed in a
2063 source file. @command{@value{AS}} syntax is similar to what many other
2064 assemblers use; it is inspired by the BSD 4.2
2069 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2073 * Preprocessing:: Preprocessing
2074 * Whitespace:: Whitespace
2075 * Comments:: Comments
2076 * Symbol Intro:: Symbols
2077 * Statements:: Statements
2078 * Constants:: Constants
2082 @section Preprocessing
2084 @cindex preprocessing
2085 The @command{@value{AS}} internal preprocessor:
2087 @cindex whitespace, removed by preprocessor
2089 adjusts and removes extra whitespace. It leaves one space or tab before
2090 the keywords on a line, and turns any other whitespace on the line into
2093 @cindex comments, removed by preprocessor
2095 removes all comments, replacing them with a single space, or an
2096 appropriate number of newlines.
2098 @cindex constants, converted by preprocessor
2100 converts character constants into the appropriate numeric values.
2103 It does not do macro processing, include file handling, or
2104 anything else you may get from your C compiler's preprocessor. You can
2105 do include file processing with the @code{.include} directive
2106 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2107 to get other ``CPP'' style preprocessing by giving the input file a
2108 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2109 Output, gcc.info, Using GNU CC}.
2111 Excess whitespace, comments, and character constants
2112 cannot be used in the portions of the input text that are not
2115 @cindex turning preprocessing on and off
2116 @cindex preprocessing, turning on and off
2119 If the first line of an input file is @code{#NO_APP} or if you use the
2120 @samp{-f} option, whitespace and comments are not removed from the input file.
2121 Within an input file, you can ask for whitespace and comment removal in
2122 specific portions of the by putting a line that says @code{#APP} before the
2123 text that may contain whitespace or comments, and putting a line that says
2124 @code{#NO_APP} after this text. This feature is mainly intend to support
2125 @code{asm} statements in compilers whose output is otherwise free of comments
2132 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2133 Whitespace is used to separate symbols, and to make programs neater for
2134 people to read. Unless within character constants
2135 (@pxref{Characters,,Character Constants}), any whitespace means the same
2136 as exactly one space.
2142 There are two ways of rendering comments to @command{@value{AS}}. In both
2143 cases the comment is equivalent to one space.
2145 Anything from @samp{/*} through the next @samp{*/} is a comment.
2146 This means you may not nest these comments.
2150 The only way to include a newline ('\n') in a comment
2151 is to use this sort of comment.
2154 /* This sort of comment does not nest. */
2157 @cindex line comment character
2158 Anything from the @dfn{line comment} character to the next newline
2159 is considered a comment and is ignored. The line comment character is
2161 @samp{;} on the ARC;
2164 @samp{@@} on the ARM;
2167 @samp{;} for the H8/300 family;
2170 @samp{;} for the HPPA;
2173 @samp{#} on the i386 and x86-64;
2176 @samp{#} on the i960;
2179 @samp{;} for the PDP-11;
2182 @samp{;} for picoJava;
2185 @samp{#} for Motorola PowerPC;
2188 @samp{!} for the Renesas / SuperH SH;
2191 @samp{!} on the SPARC;
2194 @samp{#} on the ip2k;
2197 @samp{#} on the m32c;
2200 @samp{#} on the m32r;
2203 @samp{|} on the 680x0;
2206 @samp{#} on the 68HC11 and 68HC12;
2209 @samp{#} on the Vax;
2212 @samp{;} for the Z80;
2215 @samp{!} for the Z8000;
2218 @samp{#} on the V850;
2221 @samp{#} for Xtensa systems;
2223 see @ref{Machine Dependencies}. @refill
2224 @c FIXME What about i860?
2227 On some machines there are two different line comment characters. One
2228 character only begins a comment if it is the first non-whitespace character on
2229 a line, while the other always begins a comment.
2233 The V850 assembler also supports a double dash as starting a comment that
2234 extends to the end of the line.
2240 @cindex lines starting with @code{#}
2241 @cindex logical line numbers
2242 To be compatible with past assemblers, lines that begin with @samp{#} have a
2243 special interpretation. Following the @samp{#} should be an absolute
2244 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2245 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2246 new logical file name. The rest of the line, if any, should be whitespace.
2248 If the first non-whitespace characters on the line are not numeric,
2249 the line is ignored. (Just like a comment.)
2252 # This is an ordinary comment.
2253 # 42-6 "new_file_name" # New logical file name
2254 # This is logical line # 36.
2256 This feature is deprecated, and may disappear from future versions
2257 of @command{@value{AS}}.
2262 @cindex characters used in symbols
2263 @ifclear SPECIAL-SYMS
2264 A @dfn{symbol} is one or more characters chosen from the set of all
2265 letters (both upper and lower case), digits and the three characters
2271 A @dfn{symbol} is one or more characters chosen from the set of all
2272 letters (both upper and lower case), digits and the three characters
2273 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2279 On most machines, you can also use @code{$} in symbol names; exceptions
2280 are noted in @ref{Machine Dependencies}.
2282 No symbol may begin with a digit. Case is significant.
2283 There is no length limit: all characters are significant. Symbols are
2284 delimited by characters not in that set, or by the beginning of a file
2285 (since the source program must end with a newline, the end of a file is
2286 not a possible symbol delimiter). @xref{Symbols}.
2287 @cindex length of symbols
2292 @cindex statements, structure of
2293 @cindex line separator character
2294 @cindex statement separator character
2296 @ifclear abnormal-separator
2297 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2298 semicolon (@samp{;}). The newline or semicolon is considered part of
2299 the preceding statement. Newlines and semicolons within character
2300 constants are an exception: they do not end statements.
2302 @ifset abnormal-separator
2304 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2305 point (@samp{!}). The newline or exclamation point is considered part of the
2306 preceding statement. Newlines and exclamation points within character
2307 constants are an exception: they do not end statements.
2310 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2311 H8/300) a dollar sign (@samp{$}); or (for the Renesas-SH) a semicolon
2312 (@samp{;}). The newline or separator character is considered part of
2313 the preceding statement. Newlines and separators within character
2314 constants are an exception: they do not end statements.
2319 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2320 separator character. (The line separator is usually @samp{;}, unless this
2321 conflicts with the comment character; see @ref{Machine Dependencies}.) The
2322 newline or separator character is considered part of the preceding
2323 statement. Newlines and separators within character constants are an
2324 exception: they do not end statements.
2327 @cindex newline, required at file end
2328 @cindex EOF, newline must precede
2329 It is an error to end any statement with end-of-file: the last
2330 character of any input file should be a newline.@refill
2332 An empty statement is allowed, and may include whitespace. It is ignored.
2334 @cindex instructions and directives
2335 @cindex directives and instructions
2336 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2337 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2339 A statement begins with zero or more labels, optionally followed by a
2340 key symbol which determines what kind of statement it is. The key
2341 symbol determines the syntax of the rest of the statement. If the
2342 symbol begins with a dot @samp{.} then the statement is an assembler
2343 directive: typically valid for any computer. If the symbol begins with
2344 a letter the statement is an assembly language @dfn{instruction}: it
2345 assembles into a machine language instruction.
2347 Different versions of @command{@value{AS}} for different computers
2348 recognize different instructions. In fact, the same symbol may
2349 represent a different instruction in a different computer's assembly
2353 @cindex @code{:} (label)
2354 @cindex label (@code{:})
2355 A label is a symbol immediately followed by a colon (@code{:}).
2356 Whitespace before a label or after a colon is permitted, but you may not
2357 have whitespace between a label's symbol and its colon. @xref{Labels}.
2360 For HPPA targets, labels need not be immediately followed by a colon, but
2361 the definition of a label must begin in column zero. This also implies that
2362 only one label may be defined on each line.
2366 label: .directive followed by something
2367 another_label: # This is an empty statement.
2368 instruction operand_1, operand_2, @dots{}
2375 A constant is a number, written so that its value is known by
2376 inspection, without knowing any context. Like this:
2379 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2380 .ascii "Ring the bell\7" # A string constant.
2381 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2382 .float 0f-314159265358979323846264338327\
2383 95028841971.693993751E-40 # - pi, a flonum.
2388 * Characters:: Character Constants
2389 * Numbers:: Number Constants
2393 @subsection Character Constants
2395 @cindex character constants
2396 @cindex constants, character
2397 There are two kinds of character constants. A @dfn{character} stands
2398 for one character in one byte and its value may be used in
2399 numeric expressions. String constants (properly called string
2400 @emph{literals}) are potentially many bytes and their values may not be
2401 used in arithmetic expressions.
2405 * Chars:: Characters
2409 @subsubsection Strings
2411 @cindex string constants
2412 @cindex constants, string
2413 A @dfn{string} is written between double-quotes. It may contain
2414 double-quotes or null characters. The way to get special characters
2415 into a string is to @dfn{escape} these characters: precede them with
2416 a backslash @samp{\} character. For example @samp{\\} represents
2417 one backslash: the first @code{\} is an escape which tells
2418 @command{@value{AS}} to interpret the second character literally as a backslash
2419 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2420 escape character). The complete list of escapes follows.
2422 @cindex escape codes, character
2423 @cindex character escape codes
2426 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2428 @cindex @code{\b} (backspace character)
2429 @cindex backspace (@code{\b})
2431 Mnemonic for backspace; for ASCII this is octal code 010.
2434 @c Mnemonic for EOText; for ASCII this is octal code 004.
2436 @cindex @code{\f} (formfeed character)
2437 @cindex formfeed (@code{\f})
2439 Mnemonic for FormFeed; for ASCII this is octal code 014.
2441 @cindex @code{\n} (newline character)
2442 @cindex newline (@code{\n})
2444 Mnemonic for newline; for ASCII this is octal code 012.
2447 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2449 @cindex @code{\r} (carriage return character)
2450 @cindex carriage return (@code{\r})
2452 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2455 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2456 @c other assemblers.
2458 @cindex @code{\t} (tab)
2459 @cindex tab (@code{\t})
2461 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2464 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2465 @c @item \x @var{digit} @var{digit} @var{digit}
2466 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2468 @cindex @code{\@var{ddd}} (octal character code)
2469 @cindex octal character code (@code{\@var{ddd}})
2470 @item \ @var{digit} @var{digit} @var{digit}
2471 An octal character code. The numeric code is 3 octal digits.
2472 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2473 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2475 @cindex @code{\@var{xd...}} (hex character code)
2476 @cindex hex character code (@code{\@var{xd...}})
2477 @item \@code{x} @var{hex-digits...}
2478 A hex character code. All trailing hex digits are combined. Either upper or
2479 lower case @code{x} works.
2481 @cindex @code{\\} (@samp{\} character)
2482 @cindex backslash (@code{\\})
2484 Represents one @samp{\} character.
2487 @c Represents one @samp{'} (accent acute) character.
2488 @c This is needed in single character literals
2489 @c (@xref{Characters,,Character Constants}.) to represent
2492 @cindex @code{\"} (doublequote character)
2493 @cindex doublequote (@code{\"})
2495 Represents one @samp{"} character. Needed in strings to represent
2496 this character, because an unescaped @samp{"} would end the string.
2498 @item \ @var{anything-else}
2499 Any other character when escaped by @kbd{\} gives a warning, but
2500 assembles as if the @samp{\} was not present. The idea is that if
2501 you used an escape sequence you clearly didn't want the literal
2502 interpretation of the following character. However @command{@value{AS}} has no
2503 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2504 code and warns you of the fact.
2507 Which characters are escapable, and what those escapes represent,
2508 varies widely among assemblers. The current set is what we think
2509 the BSD 4.2 assembler recognizes, and is a subset of what most C
2510 compilers recognize. If you are in doubt, do not use an escape
2514 @subsubsection Characters
2516 @cindex single character constant
2517 @cindex character, single
2518 @cindex constant, single character
2519 A single character may be written as a single quote immediately
2520 followed by that character. The same escapes apply to characters as
2521 to strings. So if you want to write the character backslash, you
2522 must write @kbd{'\\} where the first @code{\} escapes the second
2523 @code{\}. As you can see, the quote is an acute accent, not a
2524 grave accent. A newline
2526 @ifclear abnormal-separator
2527 (or semicolon @samp{;})
2529 @ifset abnormal-separator
2531 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2536 immediately following an acute accent is taken as a literal character
2537 and does not count as the end of a statement. The value of a character
2538 constant in a numeric expression is the machine's byte-wide code for
2539 that character. @command{@value{AS}} assumes your character code is ASCII:
2540 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2543 @subsection Number Constants
2545 @cindex constants, number
2546 @cindex number constants
2547 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2548 are stored in the target machine. @emph{Integers} are numbers that
2549 would fit into an @code{int} in the C language. @emph{Bignums} are
2550 integers, but they are stored in more than 32 bits. @emph{Flonums}
2551 are floating point numbers, described below.
2554 * Integers:: Integers
2559 * Bit Fields:: Bit Fields
2565 @subsubsection Integers
2567 @cindex constants, integer
2569 @cindex binary integers
2570 @cindex integers, binary
2571 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2572 the binary digits @samp{01}.
2574 @cindex octal integers
2575 @cindex integers, octal
2576 An octal integer is @samp{0} followed by zero or more of the octal
2577 digits (@samp{01234567}).
2579 @cindex decimal integers
2580 @cindex integers, decimal
2581 A decimal integer starts with a non-zero digit followed by zero or
2582 more digits (@samp{0123456789}).
2584 @cindex hexadecimal integers
2585 @cindex integers, hexadecimal
2586 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2587 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2589 Integers have the usual values. To denote a negative integer, use
2590 the prefix operator @samp{-} discussed under expressions
2591 (@pxref{Prefix Ops,,Prefix Operators}).
2594 @subsubsection Bignums
2597 @cindex constants, bignum
2598 A @dfn{bignum} has the same syntax and semantics as an integer
2599 except that the number (or its negative) takes more than 32 bits to
2600 represent in binary. The distinction is made because in some places
2601 integers are permitted while bignums are not.
2604 @subsubsection Flonums
2606 @cindex floating point numbers
2607 @cindex constants, floating point
2609 @cindex precision, floating point
2610 A @dfn{flonum} represents a floating point number. The translation is
2611 indirect: a decimal floating point number from the text is converted by
2612 @command{@value{AS}} to a generic binary floating point number of more than
2613 sufficient precision. This generic floating point number is converted
2614 to a particular computer's floating point format (or formats) by a
2615 portion of @command{@value{AS}} specialized to that computer.
2617 A flonum is written by writing (in order)
2622 (@samp{0} is optional on the HPPA.)
2626 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2628 @kbd{e} is recommended. Case is not important.
2630 @c FIXME: verify if flonum syntax really this vague for most cases
2631 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2632 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2635 On the H8/300, Renesas / SuperH SH,
2636 and AMD 29K architectures, the letter must be
2637 one of the letters @samp{DFPRSX} (in upper or lower case).
2639 On the ARC, the letter must be one of the letters @samp{DFRS}
2640 (in upper or lower case).
2642 On the Intel 960 architecture, the letter must be
2643 one of the letters @samp{DFT} (in upper or lower case).
2645 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2649 One of the letters @samp{DFRS} (in upper or lower case).
2652 One of the letters @samp{DFPRSX} (in upper or lower case).
2655 The letter @samp{E} (upper case only).
2658 One of the letters @samp{DFT} (in upper or lower case).
2663 An optional sign: either @samp{+} or @samp{-}.
2666 An optional @dfn{integer part}: zero or more decimal digits.
2669 An optional @dfn{fractional part}: @samp{.} followed by zero
2670 or more decimal digits.
2673 An optional exponent, consisting of:
2677 An @samp{E} or @samp{e}.
2678 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2679 @c principle this can perfectly well be different on different targets.
2681 Optional sign: either @samp{+} or @samp{-}.
2683 One or more decimal digits.
2688 At least one of the integer part or the fractional part must be
2689 present. The floating point number has the usual base-10 value.
2691 @command{@value{AS}} does all processing using integers. Flonums are computed
2692 independently of any floating point hardware in the computer running
2693 @command{@value{AS}}.
2697 @c Bit fields are written as a general facility but are also controlled
2698 @c by a conditional-compilation flag---which is as of now (21mar91)
2699 @c turned on only by the i960 config of GAS.
2701 @subsubsection Bit Fields
2704 @cindex constants, bit field
2705 You can also define numeric constants as @dfn{bit fields}.
2706 Specify two numbers separated by a colon---
2708 @var{mask}:@var{value}
2711 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2714 The resulting number is then packed
2716 @c this conditional paren in case bit fields turned on elsewhere than 960
2717 (in host-dependent byte order)
2719 into a field whose width depends on which assembler directive has the
2720 bit-field as its argument. Overflow (a result from the bitwise and
2721 requiring more binary digits to represent) is not an error; instead,
2722 more constants are generated, of the specified width, beginning with the
2723 least significant digits.@refill
2725 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2726 @code{.short}, and @code{.word} accept bit-field arguments.
2731 @chapter Sections and Relocation
2736 * Secs Background:: Background
2737 * Ld Sections:: Linker Sections
2738 * As Sections:: Assembler Internal Sections
2739 * Sub-Sections:: Sub-Sections
2743 @node Secs Background
2746 Roughly, a section is a range of addresses, with no gaps; all data
2747 ``in'' those addresses is treated the same for some particular purpose.
2748 For example there may be a ``read only'' section.
2750 @cindex linker, and assembler
2751 @cindex assembler, and linker
2752 The linker @code{@value{LD}} reads many object files (partial programs) and
2753 combines their contents to form a runnable program. When @command{@value{AS}}
2754 emits an object file, the partial program is assumed to start at address 0.
2755 @code{@value{LD}} assigns the final addresses for the partial program, so that
2756 different partial programs do not overlap. This is actually an
2757 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2760 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2761 addresses. These blocks slide to their run-time addresses as rigid
2762 units; their length does not change and neither does the order of bytes
2763 within them. Such a rigid unit is called a @emph{section}. Assigning
2764 run-time addresses to sections is called @dfn{relocation}. It includes
2765 the task of adjusting mentions of object-file addresses so they refer to
2766 the proper run-time addresses.
2768 For the H8/300, and for the Renesas / SuperH SH,
2769 @command{@value{AS}} pads sections if needed to
2770 ensure they end on a word (sixteen bit) boundary.
2773 @cindex standard assembler sections
2774 An object file written by @command{@value{AS}} has at least three sections, any
2775 of which may be empty. These are named @dfn{text}, @dfn{data} and
2780 When it generates COFF or ELF output,
2782 @command{@value{AS}} can also generate whatever other named sections you specify
2783 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2784 If you do not use any directives that place output in the @samp{.text}
2785 or @samp{.data} sections, these sections still exist, but are empty.
2790 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2792 @command{@value{AS}} can also generate whatever other named sections you
2793 specify using the @samp{.space} and @samp{.subspace} directives. See
2794 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2795 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2796 assembler directives.
2799 Additionally, @command{@value{AS}} uses different names for the standard
2800 text, data, and bss sections when generating SOM output. Program text
2801 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2802 BSS into @samp{$BSS$}.
2806 Within the object file, the text section starts at address @code{0}, the
2807 data section follows, and the bss section follows the data section.
2810 When generating either SOM or ELF output files on the HPPA, the text
2811 section starts at address @code{0}, the data section at address
2812 @code{0x4000000}, and the bss section follows the data section.
2815 To let @code{@value{LD}} know which data changes when the sections are
2816 relocated, and how to change that data, @command{@value{AS}} also writes to the
2817 object file details of the relocation needed. To perform relocation
2818 @code{@value{LD}} must know, each time an address in the object
2822 Where in the object file is the beginning of this reference to
2825 How long (in bytes) is this reference?
2827 Which section does the address refer to? What is the numeric value of
2829 (@var{address}) @minus{} (@var{start-address of section})?
2832 Is the reference to an address ``Program-Counter relative''?
2835 @cindex addresses, format of
2836 @cindex section-relative addressing
2837 In fact, every address @command{@value{AS}} ever uses is expressed as
2839 (@var{section}) + (@var{offset into section})
2842 Further, most expressions @command{@value{AS}} computes have this section-relative
2845 (For some object formats, such as SOM for the HPPA, some expressions are
2846 symbol-relative instead.)
2849 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2850 @var{N} into section @var{secname}.''
2852 Apart from text, data and bss sections you need to know about the
2853 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2854 addresses in the absolute section remain unchanged. For example, address
2855 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2856 @code{@value{LD}}. Although the linker never arranges two partial programs'
2857 data sections with overlapping addresses after linking, @emph{by definition}
2858 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2859 part of a program is always the same address when the program is running as
2860 address @code{@{absolute@ 239@}} in any other part of the program.
2862 The idea of sections is extended to the @dfn{undefined} section. Any
2863 address whose section is unknown at assembly time is by definition
2864 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2865 Since numbers are always defined, the only way to generate an undefined
2866 address is to mention an undefined symbol. A reference to a named
2867 common block would be such a symbol: its value is unknown at assembly
2868 time so it has section @emph{undefined}.
2870 By analogy the word @emph{section} is used to describe groups of sections in
2871 the linked program. @code{@value{LD}} puts all partial programs' text
2872 sections in contiguous addresses in the linked program. It is
2873 customary to refer to the @emph{text section} of a program, meaning all
2874 the addresses of all partial programs' text sections. Likewise for
2875 data and bss sections.
2877 Some sections are manipulated by @code{@value{LD}}; others are invented for
2878 use of @command{@value{AS}} and have no meaning except during assembly.
2881 @section Linker Sections
2882 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2887 @cindex named sections
2888 @cindex sections, named
2889 @item named sections
2892 @cindex text section
2893 @cindex data section
2897 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2898 separate but equal sections. Anything you can say of one section is
2901 When the program is running, however, it is
2902 customary for the text section to be unalterable. The
2903 text section is often shared among processes: it contains
2904 instructions, constants and the like. The data section of a running
2905 program is usually alterable: for example, C variables would be stored
2906 in the data section.
2911 This section contains zeroed bytes when your program begins running. It
2912 is used to hold uninitialized variables or common storage. The length of
2913 each partial program's bss section is important, but because it starts
2914 out containing zeroed bytes there is no need to store explicit zero
2915 bytes in the object file. The bss section was invented to eliminate
2916 those explicit zeros from object files.
2918 @cindex absolute section
2919 @item absolute section
2920 Address 0 of this section is always ``relocated'' to runtime address 0.
2921 This is useful if you want to refer to an address that @code{@value{LD}} must
2922 not change when relocating. In this sense we speak of absolute
2923 addresses being ``unrelocatable'': they do not change during relocation.
2925 @cindex undefined section
2926 @item undefined section
2927 This ``section'' is a catch-all for address references to objects not in
2928 the preceding sections.
2929 @c FIXME: ref to some other doc on obj-file formats could go here.
2932 @cindex relocation example
2933 An idealized example of three relocatable sections follows.
2935 The example uses the traditional section names @samp{.text} and @samp{.data}.
2937 Memory addresses are on the horizontal axis.
2941 @c END TEXI2ROFF-KILL
2944 partial program # 1: |ttttt|dddd|00|
2951 partial program # 2: |TTT|DDD|000|
2954 +--+---+-----+--+----+---+-----+~~
2955 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2956 +--+---+-----+--+----+---+-----+~~
2958 addresses: 0 @dots{}
2965 \line{\it Partial program \#1: \hfil}
2966 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2967 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2969 \line{\it Partial program \#2: \hfil}
2970 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2971 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2973 \line{\it linked program: \hfil}
2974 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2975 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2976 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2977 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2979 \line{\it addresses: \hfil}
2983 @c END TEXI2ROFF-KILL
2986 @section Assembler Internal Sections
2988 @cindex internal assembler sections
2989 @cindex sections in messages, internal
2990 These sections are meant only for the internal use of @command{@value{AS}}. They
2991 have no meaning at run-time. You do not really need to know about these
2992 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2993 warning messages, so it might be helpful to have an idea of their
2994 meanings to @command{@value{AS}}. These sections are used to permit the
2995 value of every expression in your assembly language program to be a
2996 section-relative address.
2999 @cindex assembler internal logic error
3000 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3001 An internal assembler logic error has been found. This means there is a
3002 bug in the assembler.
3004 @cindex expr (internal section)
3006 The assembler stores complex expression internally as combinations of
3007 symbols. When it needs to represent an expression as a symbol, it puts
3008 it in the expr section.
3010 @c FIXME item transfer[t] vector preload
3011 @c FIXME item transfer[t] vector postload
3012 @c FIXME item register
3016 @section Sub-Sections
3018 @cindex numbered subsections
3019 @cindex grouping data
3025 fall into two sections: text and data.
3027 You may have separate groups of
3029 data in named sections
3033 data in named sections
3039 that you want to end up near to each other in the object file, even though they
3040 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3041 use @dfn{subsections} for this purpose. Within each section, there can be
3042 numbered subsections with values from 0 to 8192. Objects assembled into the
3043 same subsection go into the object file together with other objects in the same
3044 subsection. For example, a compiler might want to store constants in the text
3045 section, but might not want to have them interspersed with the program being
3046 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3047 section of code being output, and a @samp{.text 1} before each group of
3048 constants being output.
3050 Subsections are optional. If you do not use subsections, everything
3051 goes in subsection number zero.
3054 Each subsection is zero-padded up to a multiple of four bytes.
3055 (Subsections may be padded a different amount on different flavors
3056 of @command{@value{AS}}.)
3060 On the H8/300 platform, each subsection is zero-padded to a word
3061 boundary (two bytes).
3062 The same is true on the Renesas SH.
3065 @c FIXME section padding (alignment)?
3066 @c Rich Pixley says padding here depends on target obj code format; that
3067 @c doesn't seem particularly useful to say without further elaboration,
3068 @c so for now I say nothing about it. If this is a generic BFD issue,
3069 @c these paragraphs might need to vanish from this manual, and be
3070 @c discussed in BFD chapter of binutils (or some such).
3074 Subsections appear in your object file in numeric order, lowest numbered
3075 to highest. (All this to be compatible with other people's assemblers.)
3076 The object file contains no representation of subsections; @code{@value{LD}} and
3077 other programs that manipulate object files see no trace of them.
3078 They just see all your text subsections as a text section, and all your
3079 data subsections as a data section.
3081 To specify which subsection you want subsequent statements assembled
3082 into, use a numeric argument to specify it, in a @samp{.text
3083 @var{expression}} or a @samp{.data @var{expression}} statement.
3086 When generating COFF output, you
3091 can also use an extra subsection
3092 argument with arbitrary named sections: @samp{.section @var{name},
3097 When generating ELF output, you
3102 can also use the @code{.subsection} directive (@pxref{SubSection})
3103 to specify a subsection: @samp{.subsection @var{expression}}.
3105 @var{Expression} should be an absolute expression
3106 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3107 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3108 begins in @code{text 0}. For instance:
3110 .text 0 # The default subsection is text 0 anyway.
3111 .ascii "This lives in the first text subsection. *"
3113 .ascii "But this lives in the second text subsection."
3115 .ascii "This lives in the data section,"
3116 .ascii "in the first data subsection."
3118 .ascii "This lives in the first text section,"
3119 .ascii "immediately following the asterisk (*)."
3122 Each section has a @dfn{location counter} incremented by one for every byte
3123 assembled into that section. Because subsections are merely a convenience
3124 restricted to @command{@value{AS}} there is no concept of a subsection location
3125 counter. There is no way to directly manipulate a location counter---but the
3126 @code{.align} directive changes it, and any label definition captures its
3127 current value. The location counter of the section where statements are being
3128 assembled is said to be the @dfn{active} location counter.
3131 @section bss Section
3134 @cindex common variable storage
3135 The bss section is used for local common variable storage.
3136 You may allocate address space in the bss section, but you may
3137 not dictate data to load into it before your program executes. When
3138 your program starts running, all the contents of the bss
3139 section are zeroed bytes.
3141 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3142 @ref{Lcomm,,@code{.lcomm}}.
3144 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3145 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3148 When assembling for a target which supports multiple sections, such as ELF or
3149 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3150 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3151 section. Typically the section will only contain symbol definitions and
3152 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3159 Symbols are a central concept: the programmer uses symbols to name
3160 things, the linker uses symbols to link, and the debugger uses symbols
3164 @cindex debuggers, and symbol order
3165 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3166 the same order they were declared. This may break some debuggers.
3171 * Setting Symbols:: Giving Symbols Other Values
3172 * Symbol Names:: Symbol Names
3173 * Dot:: The Special Dot Symbol
3174 * Symbol Attributes:: Symbol Attributes
3181 A @dfn{label} is written as a symbol immediately followed by a colon
3182 @samp{:}. The symbol then represents the current value of the
3183 active location counter, and is, for example, a suitable instruction
3184 operand. You are warned if you use the same symbol to represent two
3185 different locations: the first definition overrides any other
3189 On the HPPA, the usual form for a label need not be immediately followed by a
3190 colon, but instead must start in column zero. Only one label may be defined on
3191 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3192 provides a special directive @code{.label} for defining labels more flexibly.
3195 @node Setting Symbols
3196 @section Giving Symbols Other Values
3198 @cindex assigning values to symbols
3199 @cindex symbol values, assigning
3200 A symbol can be given an arbitrary value by writing a symbol, followed
3201 by an equals sign @samp{=}, followed by an expression
3202 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3203 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3204 equals sign @samp{=}@samp{=} here represents an equivalent of the
3205 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3208 @section Symbol Names
3210 @cindex symbol names
3211 @cindex names, symbol
3212 @ifclear SPECIAL-SYMS
3213 Symbol names begin with a letter or with one of @samp{._}. On most
3214 machines, you can also use @code{$} in symbol names; exceptions are
3215 noted in @ref{Machine Dependencies}. That character may be followed by any
3216 string of digits, letters, dollar signs (unless otherwise noted for a
3217 particular target machine), and underscores.
3221 Symbol names begin with a letter or with one of @samp{._}. On the
3222 Renesas SH you can also use @code{$} in symbol names. That
3223 character may be followed by any string of digits, letters, dollar signs (save
3224 on the H8/300), and underscores.
3228 Case of letters is significant: @code{foo} is a different symbol name
3231 Each symbol has exactly one name. Each name in an assembly language program
3232 refers to exactly one symbol. You may use that symbol name any number of times
3235 @subheading Local Symbol Names
3237 @cindex local symbol names
3238 @cindex symbol names, local
3239 A local symbol is any symbol beginning with certain local label prefixes.
3240 By default, the local label prefix is @samp{.L} for ELF systems or
3241 @samp{L} for traditional a.out systems, but each target may have its own
3242 set of local label prefixes.
3244 On the HPPA local symbols begin with @samp{L$}.
3247 Local symbols are defined and used within the assembler, but they are
3248 normally not saved in object files. Thus, they are not visible when debugging.
3249 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols:
3250 @option{-L}}) to retain the local symbols in the object files.
3252 @subheading Local Labels
3254 @cindex local labels
3255 @cindex temporary symbol names
3256 @cindex symbol names, temporary
3257 Local labels help compilers and programmers use names temporarily.
3258 They create symbols which are guaranteed to be unique over the entire scope of
3259 the input source code and which can be referred to by a simple notation.
3260 To define a local label, write a label of the form @samp{@b{N}:} (where @b{N}
3261 represents any positive integer). To refer to the most recent previous
3262 definition of that label write @samp{@b{N}b}, using the same number as when
3263 you defined the label. To refer to the next definition of a local label, write
3264 @samp{@b{N}f}---the @samp{b} stands for ``backwards'' and the @samp{f} stands
3267 There is no restriction on how you can use these labels, and you can reuse them
3268 too. So that it is possible to repeatedly define the same local label (using
3269 the same number @samp{@b{N}}), although you can only refer to the most recently
3270 defined local label of that number (for a backwards reference) or the next
3271 definition of a specific local label for a forward reference. It is also worth
3272 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3273 implemented in a slightly more efficient manner than the others.
3284 Which is the equivalent of:
3287 label_1: branch label_3
3288 label_2: branch label_1
3289 label_3: branch label_4
3290 label_4: branch label_3
3293 Local label names are only a notational device. They are immediately
3294 transformed into more conventional symbol names before the assembler uses them.
3295 The symbol names are stored in the symbol table, appear in error messages, and
3296 are optionally emitted to the object file. The names are constructed using
3300 @item @emph{local label prefix}
3301 All local symbols begin with the system-specific local label prefix.
3302 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3303 that start with the local label prefix. These labels are
3304 used for symbols you are never intended to see. If you use the
3305 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3306 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3307 you may use them in debugging.
3310 This is the number that was used in the local label definition. So if the
3311 label is written @samp{55:} then the number is @samp{55}.
3314 This unusual character is included so you do not accidentally invent a symbol
3315 of the same name. The character has ASCII value of @samp{\002} (control-B).
3317 @item @emph{ordinal number}
3318 This is a serial number to keep the labels distinct. The first definition of
3319 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3320 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3321 the number @samp{1} and its 15th definition gets @samp{15} as well.
3324 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3325 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3327 @subheading Dollar Local Labels
3328 @cindex dollar local symbols
3330 @code{@value{AS}} also supports an even more local form of local labels called
3331 dollar labels. These labels go out of scope (i.e., they become undefined) as
3332 soon as a non-local label is defined. Thus they remain valid for only a small
3333 region of the input source code. Normal local labels, by contrast, remain in
3334 scope for the entire file, or until they are redefined by another occurrence of
3335 the same local label.
3337 Dollar labels are defined in exactly the same way as ordinary local labels,
3338 except that instead of being terminated by a colon, they are terminated by a
3339 dollar sign, e.g., @samp{@b{55$}}.
3341 They can also be distinguished from ordinary local labels by their transformed
3342 names which use ASCII character @samp{\001} (control-A) as the magic character
3343 to distinguish them from ordinary labels. For example, the fifth definition of
3344 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3347 @section The Special Dot Symbol
3349 @cindex dot (symbol)
3350 @cindex @code{.} (symbol)
3351 @cindex current address
3352 @cindex location counter
3353 The special symbol @samp{.} refers to the current address that
3354 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3355 .long .} defines @code{melvin} to contain its own address.
3356 Assigning a value to @code{.} is treated the same as a @code{.org}
3357 directive. Thus, the expression @samp{.=.+4} is the same as saying
3358 @ifclear no-space-dir
3362 @node Symbol Attributes
3363 @section Symbol Attributes
3365 @cindex symbol attributes
3366 @cindex attributes, symbol
3367 Every symbol has, as well as its name, the attributes ``Value'' and
3368 ``Type''. Depending on output format, symbols can also have auxiliary
3371 The detailed definitions are in @file{a.out.h}.
3374 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3375 all these attributes, and probably won't warn you. This makes the
3376 symbol an externally defined symbol, which is generally what you
3380 * Symbol Value:: Value
3381 * Symbol Type:: Type
3384 * a.out Symbols:: Symbol Attributes: @code{a.out}
3388 * a.out Symbols:: Symbol Attributes: @code{a.out}
3391 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3396 * COFF Symbols:: Symbol Attributes for COFF
3399 * SOM Symbols:: Symbol Attributes for SOM
3406 @cindex value of a symbol
3407 @cindex symbol value
3408 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3409 location in the text, data, bss or absolute sections the value is the
3410 number of addresses from the start of that section to the label.
3411 Naturally for text, data and bss sections the value of a symbol changes
3412 as @code{@value{LD}} changes section base addresses during linking. Absolute
3413 symbols' values do not change during linking: that is why they are
3416 The value of an undefined symbol is treated in a special way. If it is
3417 0 then the symbol is not defined in this assembler source file, and
3418 @code{@value{LD}} tries to determine its value from other files linked into the
3419 same program. You make this kind of symbol simply by mentioning a symbol
3420 name without defining it. A non-zero value represents a @code{.comm}
3421 common declaration. The value is how much common storage to reserve, in
3422 bytes (addresses). The symbol refers to the first address of the
3428 @cindex type of a symbol
3430 The type attribute of a symbol contains relocation (section)
3431 information, any flag settings indicating that a symbol is external, and
3432 (optionally), other information for linkers and debuggers. The exact
3433 format depends on the object-code output format in use.
3438 @c The following avoids a "widow" subsection title. @group would be
3439 @c better if it were available outside examples.
3442 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3444 @cindex @code{b.out} symbol attributes
3445 @cindex symbol attributes, @code{b.out}
3446 These symbol attributes appear only when @command{@value{AS}} is configured for
3447 one of the Berkeley-descended object output formats---@code{a.out} or
3453 @subsection Symbol Attributes: @code{a.out}
3455 @cindex @code{a.out} symbol attributes
3456 @cindex symbol attributes, @code{a.out}
3462 @subsection Symbol Attributes: @code{a.out}
3464 @cindex @code{a.out} symbol attributes
3465 @cindex symbol attributes, @code{a.out}
3469 * Symbol Desc:: Descriptor
3470 * Symbol Other:: Other
3474 @subsubsection Descriptor
3476 @cindex descriptor, of @code{a.out} symbol
3477 This is an arbitrary 16-bit value. You may establish a symbol's
3478 descriptor value by using a @code{.desc} statement
3479 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3480 @command{@value{AS}}.
3483 @subsubsection Other
3485 @cindex other attribute, of @code{a.out} symbol
3486 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3491 @subsection Symbol Attributes for COFF
3493 @cindex COFF symbol attributes
3494 @cindex symbol attributes, COFF
3496 The COFF format supports a multitude of auxiliary symbol attributes;
3497 like the primary symbol attributes, they are set between @code{.def} and
3498 @code{.endef} directives.
3500 @subsubsection Primary Attributes
3502 @cindex primary attributes, COFF symbols
3503 The symbol name is set with @code{.def}; the value and type,
3504 respectively, with @code{.val} and @code{.type}.
3506 @subsubsection Auxiliary Attributes
3508 @cindex auxiliary attributes, COFF symbols
3509 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3510 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3511 table information for COFF.
3516 @subsection Symbol Attributes for SOM
3518 @cindex SOM symbol attributes
3519 @cindex symbol attributes, SOM
3521 The SOM format for the HPPA supports a multitude of symbol attributes set with
3522 the @code{.EXPORT} and @code{.IMPORT} directives.
3524 The attributes are described in @cite{HP9000 Series 800 Assembly
3525 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3526 @code{EXPORT} assembler directive documentation.
3530 @chapter Expressions
3534 @cindex numeric values
3535 An @dfn{expression} specifies an address or numeric value.
3536 Whitespace may precede and/or follow an expression.
3538 The result of an expression must be an absolute number, or else an offset into
3539 a particular section. If an expression is not absolute, and there is not
3540 enough information when @command{@value{AS}} sees the expression to know its
3541 section, a second pass over the source program might be necessary to interpret
3542 the expression---but the second pass is currently not implemented.
3543 @command{@value{AS}} aborts with an error message in this situation.
3546 * Empty Exprs:: Empty Expressions
3547 * Integer Exprs:: Integer Expressions
3551 @section Empty Expressions
3553 @cindex empty expressions
3554 @cindex expressions, empty
3555 An empty expression has no value: it is just whitespace or null.
3556 Wherever an absolute expression is required, you may omit the
3557 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3558 is compatible with other assemblers.
3561 @section Integer Expressions
3563 @cindex integer expressions
3564 @cindex expressions, integer
3565 An @dfn{integer expression} is one or more @emph{arguments} delimited
3566 by @emph{operators}.
3569 * Arguments:: Arguments
3570 * Operators:: Operators
3571 * Prefix Ops:: Prefix Operators
3572 * Infix Ops:: Infix Operators
3576 @subsection Arguments
3578 @cindex expression arguments
3579 @cindex arguments in expressions
3580 @cindex operands in expressions
3581 @cindex arithmetic operands
3582 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3583 contexts arguments are sometimes called ``arithmetic operands''. In
3584 this manual, to avoid confusing them with the ``instruction operands'' of
3585 the machine language, we use the term ``argument'' to refer to parts of
3586 expressions only, reserving the word ``operand'' to refer only to machine
3587 instruction operands.
3589 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3590 @var{section} is one of text, data, bss, absolute,
3591 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3594 Numbers are usually integers.
3596 A number can be a flonum or bignum. In this case, you are warned
3597 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3598 these 32 bits are an integer. You may write integer-manipulating
3599 instructions that act on exotic constants, compatible with other
3602 @cindex subexpressions
3603 Subexpressions are a left parenthesis @samp{(} followed by an integer
3604 expression, followed by a right parenthesis @samp{)}; or a prefix
3605 operator followed by an argument.
3608 @subsection Operators
3610 @cindex operators, in expressions
3611 @cindex arithmetic functions
3612 @cindex functions, in expressions
3613 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3614 operators are followed by an argument. Infix operators appear
3615 between their arguments. Operators may be preceded and/or followed by
3619 @subsection Prefix Operator
3621 @cindex prefix operators
3622 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3623 one argument, which must be absolute.
3625 @c the tex/end tex stuff surrounding this small table is meant to make
3626 @c it align, on the printed page, with the similar table in the next
3627 @c section (which is inside an enumerate).
3629 \global\advance\leftskip by \itemindent
3634 @dfn{Negation}. Two's complement negation.
3636 @dfn{Complementation}. Bitwise not.
3640 \global\advance\leftskip by -\itemindent
3644 @subsection Infix Operators
3646 @cindex infix operators
3647 @cindex operators, permitted arguments
3648 @dfn{Infix operators} take two arguments, one on either side. Operators
3649 have precedence, but operations with equal precedence are performed left
3650 to right. Apart from @code{+} or @option{-}, both arguments must be
3651 absolute, and the result is absolute.
3654 @cindex operator precedence
3655 @cindex precedence of operators
3662 @dfn{Multiplication}.
3665 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3671 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3674 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3678 Intermediate precedence
3683 @dfn{Bitwise Inclusive Or}.
3689 @dfn{Bitwise Exclusive Or}.
3692 @dfn{Bitwise Or Not}.
3699 @cindex addition, permitted arguments
3700 @cindex plus, permitted arguments
3701 @cindex arguments for addition
3703 @dfn{Addition}. If either argument is absolute, the result has the section of
3704 the other argument. You may not add together arguments from different
3707 @cindex subtraction, permitted arguments
3708 @cindex minus, permitted arguments
3709 @cindex arguments for subtraction
3711 @dfn{Subtraction}. If the right argument is absolute, the
3712 result has the section of the left argument.
3713 If both arguments are in the same section, the result is absolute.
3714 You may not subtract arguments from different sections.
3715 @c FIXME is there still something useful to say about undefined - undefined ?
3717 @cindex comparison expressions
3718 @cindex expressions, comparison
3723 @dfn{Is Not Equal To}
3727 @dfn{Is Greater Than}
3729 @dfn{Is Greater Than Or Equal To}
3731 @dfn{Is Less Than Or Equal To}
3733 The comparison operators can be used as infix operators. A true results has a
3734 value of -1 whereas a false result has a value of 0. Note, these operators
3735 perform signed comparisons.
3738 @item Lowest Precedence
3747 These two logical operations can be used to combine the results of sub
3748 expressions. Note, unlike the comparison operators a true result returns a
3749 value of 1 but a false results does still return 0. Also note that the logical
3750 or operator has a slightly lower precedence than logical and.
3755 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3756 address; you can only have a defined section in one of the two arguments.
3759 @chapter Assembler Directives
3761 @cindex directives, machine independent
3762 @cindex pseudo-ops, machine independent
3763 @cindex machine independent directives
3764 All assembler directives have names that begin with a period (@samp{.}).
3765 The rest of the name is letters, usually in lower case.
3767 This chapter discusses directives that are available regardless of the
3768 target machine configuration for the @sc{gnu} assembler.
3770 Some machine configurations provide additional directives.
3771 @xref{Machine Dependencies}.
3774 @ifset machine-directives
3775 @xref{Machine Dependencies}, for additional directives.
3780 * Abort:: @code{.abort}
3782 * ABORT (COFF):: @code{.ABORT}
3785 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3786 * Altmacro:: @code{.altmacro}
3787 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3788 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3789 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3790 * Byte:: @code{.byte @var{expressions}}
3791 * Comm:: @code{.comm @var{symbol} , @var{length} }
3793 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
3795 * Data:: @code{.data @var{subsection}}
3797 * Def:: @code{.def @var{name}}
3800 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3806 * Double:: @code{.double @var{flonums}}
3807 * Eject:: @code{.eject}
3808 * Else:: @code{.else}
3809 * Elseif:: @code{.elseif}
3812 * Endef:: @code{.endef}
3815 * Endfunc:: @code{.endfunc}
3816 * Endif:: @code{.endif}
3817 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3818 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3819 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
3821 * Error:: @code{.error @var{string}}
3822 * Exitm:: @code{.exitm}
3823 * Extern:: @code{.extern}
3824 * Fail:: @code{.fail}
3825 @ifclear no-file-dir
3826 * File:: @code{.file @var{string}}
3829 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3830 * Float:: @code{.float @var{flonums}}
3831 * Func:: @code{.func}
3832 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3834 * Hidden:: @code{.hidden @var{names}}
3837 * hword:: @code{.hword @var{expressions}}
3838 * Ident:: @code{.ident}
3839 * If:: @code{.if @var{absolute expression}}
3840 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3841 * Include:: @code{.include "@var{file}"}
3842 * Int:: @code{.int @var{expressions}}
3844 * Internal:: @code{.internal @var{names}}
3847 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3848 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3849 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3850 * Lflags:: @code{.lflags}
3851 @ifclear no-line-dir
3852 * Line:: @code{.line @var{line-number}}
3855 * Linkonce:: @code{.linkonce [@var{type}]}
3856 * List:: @code{.list}
3857 * Ln:: @code{.ln @var{line-number}}
3859 * LNS directives:: @code{.file}, @code{.loc}, etc.
3861 * Long:: @code{.long @var{expressions}}
3863 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3866 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3867 * MRI:: @code{.mri @var{val}}
3868 * Noaltmacro:: @code{.noaltmacro}
3869 * Nolist:: @code{.nolist}
3870 * Octa:: @code{.octa @var{bignums}}
3871 * Org:: @code{.org @var{new-lc}, @var{fill}}
3872 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3874 * PopSection:: @code{.popsection}
3875 * Previous:: @code{.previous}
3878 * Print:: @code{.print @var{string}}
3880 * Protected:: @code{.protected @var{names}}
3883 * Psize:: @code{.psize @var{lines}, @var{columns}}
3884 * Purgem:: @code{.purgem @var{name}}
3886 * PushSection:: @code{.pushsection @var{name}}
3889 * Quad:: @code{.quad @var{bignums}}
3890 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
3891 * Rept:: @code{.rept @var{count}}
3892 * Sbttl:: @code{.sbttl "@var{subheading}"}
3894 * Scl:: @code{.scl @var{class}}
3897 * Section:: @code{.section @var{name}}
3900 * Set:: @code{.set @var{symbol}, @var{expression}}
3901 * Short:: @code{.short @var{expressions}}
3902 * Single:: @code{.single @var{flonums}}
3904 * Size:: @code{.size [@var{name} , @var{expression}]}
3907 * Skip:: @code{.skip @var{size} , @var{fill}}
3908 * Sleb128:: @code{.sleb128 @var{expressions}}
3909 * Space:: @code{.space @var{size} , @var{fill}}
3911 * Stab:: @code{.stabd, .stabn, .stabs}
3914 * String:: @code{.string "@var{str}"}
3915 * Struct:: @code{.struct @var{expression}}
3917 * SubSection:: @code{.subsection}
3918 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3922 * Tag:: @code{.tag @var{structname}}
3925 * Text:: @code{.text @var{subsection}}
3926 * Title:: @code{.title "@var{heading}"}
3928 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3931 * Uleb128:: @code{.uleb128 @var{expressions}}
3933 * Val:: @code{.val @var{addr}}
3937 * Version:: @code{.version "@var{string}"}
3938 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3939 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3942 * Warning:: @code{.warning @var{string}}
3943 * Weak:: @code{.weak @var{names}}
3944 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
3945 * Word:: @code{.word @var{expressions}}
3946 * Deprecated:: Deprecated Directives
3950 @section @code{.abort}
3952 @cindex @code{abort} directive
3953 @cindex stopping the assembly
3954 This directive stops the assembly immediately. It is for
3955 compatibility with other assemblers. The original idea was that the
3956 assembly language source would be piped into the assembler. If the sender
3957 of the source quit, it could use this directive tells @command{@value{AS}} to
3958 quit also. One day @code{.abort} will not be supported.
3962 @section @code{.ABORT} (COFF)
3964 @cindex @code{ABORT} directive
3965 When producing COFF output, @command{@value{AS}} accepts this directive as a
3966 synonym for @samp{.abort}.
3969 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3975 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3977 @cindex padding the location counter
3978 @cindex @code{align} directive
3979 Pad the location counter (in the current subsection) to a particular storage
3980 boundary. The first expression (which must be absolute) is the alignment
3981 required, as described below.
3983 The second expression (also absolute) gives the fill value to be stored in the
3984 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3985 padding bytes are normally zero. However, on some systems, if the section is
3986 marked as containing code and the fill value is omitted, the space is filled
3987 with no-op instructions.
3989 The third expression is also absolute, and is also optional. If it is present,
3990 it is the maximum number of bytes that should be skipped by this alignment
3991 directive. If doing the alignment would require skipping more bytes than the
3992 specified maximum, then the alignment is not done at all. You can omit the
3993 fill value (the second argument) entirely by simply using two commas after the
3994 required alignment; this can be useful if you want the alignment to be filled
3995 with no-op instructions when appropriate.
3997 The way the required alignment is specified varies from system to system.
3998 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
3999 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4000 alignment request in bytes. For example @samp{.align 8} advances
4001 the location counter until it is a multiple of 8. If the location counter
4002 is already a multiple of 8, no change is needed. For the tic54x, the
4003 first expression is the alignment request in words.
4005 For other systems, including the i386 using a.out format, and the arm and
4006 strongarm, it is the
4007 number of low-order zero bits the location counter must have after
4008 advancement. For example @samp{.align 3} advances the location
4009 counter until it a multiple of 8. If the location counter is already a
4010 multiple of 8, no change is needed.
4012 This inconsistency is due to the different behaviors of the various
4013 native assemblers for these systems which GAS must emulate.
4014 GAS also provides @code{.balign} and @code{.p2align} directives,
4015 described later, which have a consistent behavior across all
4016 architectures (but are specific to GAS).
4019 @section @code{.ascii "@var{string}"}@dots{}
4021 @cindex @code{ascii} directive
4022 @cindex string literals
4023 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4024 separated by commas. It assembles each string (with no automatic
4025 trailing zero byte) into consecutive addresses.
4028 @section @code{.asciz "@var{string}"}@dots{}
4030 @cindex @code{asciz} directive
4031 @cindex zero-terminated strings
4032 @cindex null-terminated strings
4033 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4034 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4037 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4039 @cindex padding the location counter given number of bytes
4040 @cindex @code{balign} directive
4041 Pad the location counter (in the current subsection) to a particular
4042 storage boundary. The first expression (which must be absolute) is the
4043 alignment request in bytes. For example @samp{.balign 8} advances
4044 the location counter until it is a multiple of 8. If the location counter
4045 is already a multiple of 8, no change is needed.
4047 The second expression (also absolute) gives the fill value to be stored in the
4048 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4049 padding bytes are normally zero. However, on some systems, if the section is
4050 marked as containing code and the fill value is omitted, the space is filled
4051 with no-op instructions.
4053 The third expression is also absolute, and is also optional. If it is present,
4054 it is the maximum number of bytes that should be skipped by this alignment
4055 directive. If doing the alignment would require skipping more bytes than the
4056 specified maximum, then the alignment is not done at all. You can omit the
4057 fill value (the second argument) entirely by simply using two commas after the
4058 required alignment; this can be useful if you want the alignment to be filled
4059 with no-op instructions when appropriate.
4061 @cindex @code{balignw} directive
4062 @cindex @code{balignl} directive
4063 The @code{.balignw} and @code{.balignl} directives are variants of the
4064 @code{.balign} directive. The @code{.balignw} directive treats the fill
4065 pattern as a two byte word value. The @code{.balignl} directives treats the
4066 fill pattern as a four byte longword value. For example, @code{.balignw
4067 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4068 filled in with the value 0x368d (the exact placement of the bytes depends upon
4069 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4073 @section @code{.byte @var{expressions}}
4075 @cindex @code{byte} directive
4076 @cindex integers, one byte
4077 @code{.byte} expects zero or more expressions, separated by commas.
4078 Each expression is assembled into the next byte.
4081 @section @code{.comm @var{symbol} , @var{length} }
4083 @cindex @code{comm} directive
4084 @cindex symbol, common
4085 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4086 common symbol in one object file may be merged with a defined or common symbol
4087 of the same name in another object file. If @code{@value{LD}} does not see a
4088 definition for the symbol--just one or more common symbols--then it will
4089 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4090 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4091 the same name, and they do not all have the same size, it will allocate space
4092 using the largest size.
4095 When using ELF, the @code{.comm} directive takes an optional third argument.
4096 This is the desired alignment of the symbol, specified as a byte boundary (for
4097 example, an alignment of 16 means that the least significant 4 bits of the
4098 address should be zero). The alignment must be an absolute expression, and it
4099 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
4100 for the common symbol, it will use the alignment when placing the symbol. If
4101 no alignment is specified, @command{@value{AS}} will set the alignment to the
4102 largest power of two less than or equal to the size of the symbol, up to a
4107 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4108 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4111 @node CFI directives
4112 @section @code{.cfi_startproc [simple]}
4113 @cindex @code{cfi_startproc} directive
4114 @code{.cfi_startproc} is used at the beginning of each function that
4115 should have an entry in @code{.eh_frame}. It initializes some internal
4116 data structures. Don't forget to close the function by
4117 @code{.cfi_endproc}.
4119 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4120 it also emits some architecture dependent initial CFI instructions.
4122 @section @code{.cfi_endproc}
4123 @cindex @code{cfi_endproc} directive
4124 @code{.cfi_endproc} is used at the end of a function where it closes its
4125 unwind entry previously opened by
4126 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4128 @section @code{.cfi_personality @var{encoding} [, @var{exp}]}
4129 @code{.cfi_personality} defines personality routine and its encoding.
4130 @var{encoding} must be a constant determining how the personality
4131 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4132 argument is not present, otherwise second argument should be
4133 a constant or a symbol name. When using indirect encodings,
4134 the symbol provided should be the location where personality
4135 can be loaded from, not the personality routine itself.
4136 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4137 no personality routine.
4139 @section @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4140 @code{.cfi_lsda} defines LSDA and its encoding.
4141 @var{encoding} must be a constant determining how the LSDA
4142 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4143 argument is not present, otherwise second argument should be a constant
4144 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4147 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4148 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4149 address from @var{register} and add @var{offset} to it}.
4151 @section @code{.cfi_def_cfa_register @var{register}}
4152 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4153 now on @var{register} will be used instead of the old one. Offset
4156 @section @code{.cfi_def_cfa_offset @var{offset}}
4157 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4158 remains the same, but @var{offset} is new. Note that it is the
4159 absolute offset that will be added to a defined register to compute
4162 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4163 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4164 value that is added/substracted from the previous offset.
4166 @section @code{.cfi_offset @var{register}, @var{offset}}
4167 Previous value of @var{register} is saved at offset @var{offset} from
4170 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4171 Previous value of @var{register} is saved at offset @var{offset} from
4172 the current CFA register. This is transformed to @code{.cfi_offset}
4173 using the known displacement of the CFA register from the CFA.
4174 This is often easier to use, because the number will match the
4175 code it's annotating.
4177 @section @code{.cfi_register @var{register1}, @var{register2}}
4178 Previous value of @var{register1} is saved in register @var{register2}.
4180 @section @code{.cfi_restore @var{register}}
4181 @code{.cfi_restore} says that the rule for @var{register} is now the
4182 same as it was at the beginning of the function, after all initial
4183 instruction added by @code{.cfi_startproc} were executed.
4185 @section @code{.cfi_undefined @var{register}}
4186 From now on the previous value of @var{register} can't be restored anymore.
4188 @section @code{.cfi_same_value @var{register}}
4189 Current value of @var{register} is the same like in the previous frame,
4190 i.e. no restoration needed.
4192 @section @code{.cfi_remember_state},
4193 First save all current rules for all registers by @code{.cfi_remember_state},
4194 then totally screw them up by subsequent @code{.cfi_*} directives and when
4195 everything is hopelessly bad, use @code{.cfi_restore_state} to restore
4196 the previous saved state.
4198 @section @code{.cfi_return_column @var{register}}
4199 Change return column @var{register}, i.e. the return address is either
4200 directly in @var{register} or can be accessed by rules for @var{register}.
4202 @section @code{.cfi_signal_frame}
4203 Mark current function as signal trampoline.
4205 @section @code{.cfi_window_save}
4206 SPARC register window has been saved.
4208 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4209 Allows the user to add arbitrary bytes to the unwind info. One
4210 might use this to add OS-specific CFI opcodes, or generic CFI
4211 opcodes that GAS does not yet support.
4213 @node LNS directives
4214 @section @code{.file @var{fileno} @var{filename}}
4215 @cindex @code{file} directive
4216 When emitting dwarf2 line number information @code{.file} assigns filenames
4217 to the @code{.debug_line} file name table. The @var{fileno} operand should
4218 be a unique positive integer to use as the index of the entry in the table.
4219 The @var{filename} operand is a C string literal.
4221 The detail of filename indices is exposed to the user because the filename
4222 table is shared with the @code{.debug_info} section of the dwarf2 debugging
4223 information, and thus the user must know the exact indices that table
4226 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
4227 @cindex @code{loc} directive
4228 The @code{.loc} directive will add row to the @code{.debug_line} line
4229 number matrix corresponding to the immediately following assembly
4230 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
4231 arguments will be applied to the @code{.debug_line} state machine before
4234 The @var{options} are a sequence of the following tokens in any order:
4238 This option will set the @code{basic_block} register in the
4239 @code{.debug_line} state machine to @code{true}.
4242 This option will set the @code{prologue_end} register in the
4243 @code{.debug_line} state machine to @code{true}.
4245 @item epilogue_begin
4246 This option will set the @code{epilogue_begin} register in the
4247 @code{.debug_line} state machine to @code{true}.
4249 @item is_stmt @var{value}
4250 This option will set the @code{is_stmt} register in the
4251 @code{.debug_line} state machine to @code{value}, which must be
4254 @item isa @var{value}
4255 This directive will set the @code{isa} register in the @code{.debug_line}
4256 state machine to @var{value}, which must be an unsigned integer.
4260 @section @code{.loc_mark_blocks @var{enable}}
4261 @cindex @code{loc_mark_blocks} directive
4262 The @code{.loc_mark_blocks} directive makes the assembler emit an entry
4263 to the @code{.debug_line} line number matrix with the @code{basic_block}
4264 register in the state machine set whenever a code label is seen.
4265 The @var{enable} argument should be either 1 or 0, to enable or disable
4266 this function respectively.
4269 @section @code{.data @var{subsection}}
4271 @cindex @code{data} directive
4272 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4273 end of the data subsection numbered @var{subsection} (which is an
4274 absolute expression). If @var{subsection} is omitted, it defaults
4279 @section @code{.def @var{name}}
4281 @cindex @code{def} directive
4282 @cindex COFF symbols, debugging
4283 @cindex debugging COFF symbols
4284 Begin defining debugging information for a symbol @var{name}; the
4285 definition extends until the @code{.endef} directive is encountered.
4288 This directive is only observed when @command{@value{AS}} is configured for COFF
4289 format output; when producing @code{b.out}, @samp{.def} is recognized,
4296 @section @code{.desc @var{symbol}, @var{abs-expression}}
4298 @cindex @code{desc} directive
4299 @cindex COFF symbol descriptor
4300 @cindex symbol descriptor, COFF
4301 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4302 to the low 16 bits of an absolute expression.
4305 The @samp{.desc} directive is not available when @command{@value{AS}} is
4306 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4307 object format. For the sake of compatibility, @command{@value{AS}} accepts
4308 it, but produces no output, when configured for COFF.
4314 @section @code{.dim}
4316 @cindex @code{dim} directive
4317 @cindex COFF auxiliary symbol information
4318 @cindex auxiliary symbol information, COFF
4319 This directive is generated by compilers to include auxiliary debugging
4320 information in the symbol table. It is only permitted inside
4321 @code{.def}/@code{.endef} pairs.
4324 @samp{.dim} is only meaningful when generating COFF format output; when
4325 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4331 @section @code{.double @var{flonums}}
4333 @cindex @code{double} directive
4334 @cindex floating point numbers (double)
4335 @code{.double} expects zero or more flonums, separated by commas. It
4336 assembles floating point numbers.
4338 The exact kind of floating point numbers emitted depends on how
4339 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4343 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4344 in @sc{ieee} format.
4349 @section @code{.eject}
4351 @cindex @code{eject} directive
4352 @cindex new page, in listings
4353 @cindex page, in listings
4354 @cindex listing control: new page
4355 Force a page break at this point, when generating assembly listings.
4358 @section @code{.else}
4360 @cindex @code{else} directive
4361 @code{.else} is part of the @command{@value{AS}} support for conditional
4362 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
4363 of code to be assembled if the condition for the preceding @code{.if}
4367 @section @code{.elseif}
4369 @cindex @code{elseif} directive
4370 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4371 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
4372 @code{.if} block that would otherwise fill the entire @code{.else} section.
4375 @section @code{.end}
4377 @cindex @code{end} directive
4378 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4379 process anything in the file past the @code{.end} directive.
4383 @section @code{.endef}
4385 @cindex @code{endef} directive
4386 This directive flags the end of a symbol definition begun with
4390 @samp{.endef} is only meaningful when generating COFF format output; if
4391 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4392 directive but ignores it.
4397 @section @code{.endfunc}
4398 @cindex @code{endfunc} directive
4399 @code{.endfunc} marks the end of a function specified with @code{.func}.
4402 @section @code{.endif}
4404 @cindex @code{endif} directive
4405 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4406 it marks the end of a block of code that is only assembled
4407 conditionally. @xref{If,,@code{.if}}.
4410 @section @code{.equ @var{symbol}, @var{expression}}
4412 @cindex @code{equ} directive
4413 @cindex assigning values to symbols
4414 @cindex symbols, assigning values to
4415 This directive sets the value of @var{symbol} to @var{expression}.
4416 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
4419 The syntax for @code{equ} on the HPPA is
4420 @samp{@var{symbol} .equ @var{expression}}.
4424 The syntax for @code{equ} on the Z80 is
4425 @samp{@var{symbol} equ @var{expression}}.
4426 On the Z80 it is an eror if @var{symbol} is already defined,
4427 but the symbol is not protected from later redefinition.
4428 Compare @ref{Equiv}.
4432 @section @code{.equiv @var{symbol}, @var{expression}}
4433 @cindex @code{equiv} directive
4434 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4435 the assembler will signal an error if @var{symbol} is already defined. Note a
4436 symbol which has been referenced but not actually defined is considered to be
4439 Except for the contents of the error message, this is roughly equivalent to
4446 plus it protects the symbol from later redefinition.
4449 @section @code{.eqv @var{symbol}, @var{expression}}
4450 @cindex @code{eqv} directive
4451 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4452 evaluate the expression or any part of it immediately. Instead each time
4453 the resulting symbol is used in an expression, a snapshot of its current
4457 @section @code{.err}
4458 @cindex @code{err} directive
4459 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4460 message and, unless the @option{-Z} option was used, it will not generate an
4461 object file. This can be used to signal an error in conditionally compiled code.
4464 @section @code{.error "@var{string}"}
4465 @cindex error directive
4467 Similarly to @code{.err}, this directive emits an error, but you can specify a
4468 string that will be emitted as the error message. If you don't specify the
4469 message, it defaults to @code{".error directive invoked in source file"}.
4470 @xref{Errors, ,Error and Warning Messages}.
4473 .error "This code has not been assembled and tested."
4477 @section @code{.exitm}
4478 Exit early from the current macro definition. @xref{Macro}.
4481 @section @code{.extern}
4483 @cindex @code{extern} directive
4484 @code{.extern} is accepted in the source program---for compatibility
4485 with other assemblers---but it is ignored. @command{@value{AS}} treats
4486 all undefined symbols as external.
4489 @section @code{.fail @var{expression}}
4491 @cindex @code{fail} directive
4492 Generates an error or a warning. If the value of the @var{expression} is 500
4493 or more, @command{@value{AS}} will print a warning message. If the value is less
4494 than 500, @command{@value{AS}} will print an error message. The message will
4495 include the value of @var{expression}. This can occasionally be useful inside
4496 complex nested macros or conditional assembly.
4498 @ifclear no-file-dir
4500 @section @code{.file @var{string}}
4502 @cindex @code{file} directive
4503 @cindex logical file name
4504 @cindex file name, logical
4505 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4506 file. @var{string} is the new file name. In general, the filename is
4507 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4508 to specify an empty file name, you must give the quotes--@code{""}. This
4509 statement may go away in future: it is only recognized to be compatible with
4510 old @command{@value{AS}} programs.
4514 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4516 @cindex @code{fill} directive
4517 @cindex writing patterns in memory
4518 @cindex patterns, writing in memory
4519 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4520 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4521 may be zero or more. @var{Size} may be zero or more, but if it is
4522 more than 8, then it is deemed to have the value 8, compatible with
4523 other people's assemblers. The contents of each @var{repeat} bytes
4524 is taken from an 8-byte number. The highest order 4 bytes are
4525 zero. The lowest order 4 bytes are @var{value} rendered in the
4526 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4527 Each @var{size} bytes in a repetition is taken from the lowest order
4528 @var{size} bytes of this number. Again, this bizarre behavior is
4529 compatible with other people's assemblers.
4531 @var{size} and @var{value} are optional.
4532 If the second comma and @var{value} are absent, @var{value} is
4533 assumed zero. If the first comma and following tokens are absent,
4534 @var{size} is assumed to be 1.
4537 @section @code{.float @var{flonums}}
4539 @cindex floating point numbers (single)
4540 @cindex @code{float} directive
4541 This directive assembles zero or more flonums, separated by commas. It
4542 has the same effect as @code{.single}.
4544 The exact kind of floating point numbers emitted depends on how
4545 @command{@value{AS}} is configured.
4546 @xref{Machine Dependencies}.
4550 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4551 in @sc{ieee} format.
4556 @section @code{.func @var{name}[,@var{label}]}
4557 @cindex @code{func} directive
4558 @code{.func} emits debugging information to denote function @var{name}, and
4559 is ignored unless the file is assembled with debugging enabled.
4560 Only @samp{--gstabs[+]} is currently supported.
4561 @var{label} is the entry point of the function and if omitted @var{name}
4562 prepended with the @samp{leading char} is used.
4563 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4564 All functions are currently defined to have @code{void} return type.
4565 The function must be terminated with @code{.endfunc}.
4568 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4570 @cindex @code{global} directive
4571 @cindex symbol, making visible to linker
4572 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4573 @var{symbol} in your partial program, its value is made available to
4574 other partial programs that are linked with it. Otherwise,
4575 @var{symbol} takes its attributes from a symbol of the same name
4576 from another file linked into the same program.
4578 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4579 compatibility with other assemblers.
4582 On the HPPA, @code{.global} is not always enough to make it accessible to other
4583 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4584 @xref{HPPA Directives, ,HPPA Assembler Directives}.
4589 @section @code{.hidden @var{names}}
4591 @cindex @code{hidden} directive
4593 This is one of the ELF visibility directives. The other two are
4594 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4595 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4597 This directive overrides the named symbols default visibility (which is set by
4598 their binding: local, global or weak). The directive sets the visibility to
4599 @code{hidden} which means that the symbols are not visible to other components.
4600 Such symbols are always considered to be @code{protected} as well.
4604 @section @code{.hword @var{expressions}}
4606 @cindex @code{hword} directive
4607 @cindex integers, 16-bit
4608 @cindex numbers, 16-bit
4609 @cindex sixteen bit integers
4610 This expects zero or more @var{expressions}, and emits
4611 a 16 bit number for each.
4614 This directive is a synonym for @samp{.short}; depending on the target
4615 architecture, it may also be a synonym for @samp{.word}.
4619 This directive is a synonym for @samp{.short}.
4622 This directive is a synonym for both @samp{.short} and @samp{.word}.
4627 @section @code{.ident}
4629 @cindex @code{ident} directive
4631 This directive is used by some assemblers to place tags in object files. The
4632 behavior of this directive varies depending on the target. When using the
4633 a.out object file format, @command{@value{AS}} simply accepts the directive for
4634 source-file compatibility with existing assemblers, but does not emit anything
4635 for it. When using COFF, comments are emitted to the @code{.comment} or
4636 @code{.rdata} section, depending on the target. When using ELF, comments are
4637 emitted to the @code{.comment} section.
4640 @section @code{.if @var{absolute expression}}
4642 @cindex conditional assembly
4643 @cindex @code{if} directive
4644 @code{.if} marks the beginning of a section of code which is only
4645 considered part of the source program being assembled if the argument
4646 (which must be an @var{absolute expression}) is non-zero. The end of
4647 the conditional section of code must be marked by @code{.endif}
4648 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4649 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4650 If you have several conditions to check, @code{.elseif} may be used to avoid
4651 nesting blocks if/else within each subsequent @code{.else} block.
4653 The following variants of @code{.if} are also supported:
4655 @cindex @code{ifdef} directive
4656 @item .ifdef @var{symbol}
4657 Assembles the following section of code if the specified @var{symbol}
4658 has been defined. Note a symbol which has been referenced but not yet defined
4659 is considered to be undefined.
4661 @cindex @code{ifb} directive
4662 @item .ifb @var{text}
4663 Assembles the following section of code if the operand is blank (empty).
4665 @cindex @code{ifc} directive
4666 @item .ifc @var{string1},@var{string2}
4667 Assembles the following section of code if the two strings are the same. The
4668 strings may be optionally quoted with single quotes. If they are not quoted,
4669 the first string stops at the first comma, and the second string stops at the
4670 end of the line. Strings which contain whitespace should be quoted. The
4671 string comparison is case sensitive.
4673 @cindex @code{ifeq} directive
4674 @item .ifeq @var{absolute expression}
4675 Assembles the following section of code if the argument is zero.
4677 @cindex @code{ifeqs} directive
4678 @item .ifeqs @var{string1},@var{string2}
4679 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4681 @cindex @code{ifge} directive
4682 @item .ifge @var{absolute expression}
4683 Assembles the following section of code if the argument is greater than or
4686 @cindex @code{ifgt} directive
4687 @item .ifgt @var{absolute expression}
4688 Assembles the following section of code if the argument is greater than zero.
4690 @cindex @code{ifle} directive
4691 @item .ifle @var{absolute expression}
4692 Assembles the following section of code if the argument is less than or equal
4695 @cindex @code{iflt} directive
4696 @item .iflt @var{absolute expression}
4697 Assembles the following section of code if the argument is less than zero.
4699 @cindex @code{ifnb} directive
4700 @item .ifnb @var{text}
4701 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
4702 following section of code if the operand is non-blank (non-empty).
4704 @cindex @code{ifnc} directive
4705 @item .ifnc @var{string1},@var{string2}.
4706 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4707 following section of code if the two strings are not the same.
4709 @cindex @code{ifndef} directive
4710 @cindex @code{ifnotdef} directive
4711 @item .ifndef @var{symbol}
4712 @itemx .ifnotdef @var{symbol}
4713 Assembles the following section of code if the specified @var{symbol}
4714 has not been defined. Both spelling variants are equivalent. Note a symbol
4715 which has been referenced but not yet defined is considered to be undefined.
4717 @cindex @code{ifne} directive
4718 @item .ifne @var{absolute expression}
4719 Assembles the following section of code if the argument is not equal to zero
4720 (in other words, this is equivalent to @code{.if}).
4722 @cindex @code{ifnes} directive
4723 @item .ifnes @var{string1},@var{string2}
4724 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4725 following section of code if the two strings are not the same.
4729 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4731 @cindex @code{incbin} directive
4732 @cindex binary files, including
4733 The @code{incbin} directive includes @var{file} verbatim at the current
4734 location. You can control the search paths used with the @samp{-I} command-line
4735 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4738 The @var{skip} argument skips a number of bytes from the start of the
4739 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4740 read. Note that the data is not aligned in any way, so it is the user's
4741 responsibility to make sure that proper alignment is provided both before and
4742 after the @code{incbin} directive.
4745 @section @code{.include "@var{file}"}
4747 @cindex @code{include} directive
4748 @cindex supporting files, including
4749 @cindex files, including
4750 This directive provides a way to include supporting files at specified
4751 points in your source program. The code from @var{file} is assembled as
4752 if it followed the point of the @code{.include}; when the end of the
4753 included file is reached, assembly of the original file continues. You
4754 can control the search paths used with the @samp{-I} command-line option
4755 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4759 @section @code{.int @var{expressions}}
4761 @cindex @code{int} directive
4762 @cindex integers, 32-bit
4763 Expect zero or more @var{expressions}, of any section, separated by commas.
4764 For each expression, emit a number that, at run time, is the value of that
4765 expression. The byte order and bit size of the number depends on what kind
4766 of target the assembly is for.
4770 On most forms of the H8/300, @code{.int} emits 16-bit
4771 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4778 @section @code{.internal @var{names}}
4780 @cindex @code{internal} directive
4782 This is one of the ELF visibility directives. The other two are
4783 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4784 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4786 This directive overrides the named symbols default visibility (which is set by
4787 their binding: local, global or weak). The directive sets the visibility to
4788 @code{internal} which means that the symbols are considered to be @code{hidden}
4789 (i.e., not visible to other components), and that some extra, processor specific
4790 processing must also be performed upon the symbols as well.
4794 @section @code{.irp @var{symbol},@var{values}}@dots{}
4796 @cindex @code{irp} directive
4797 Evaluate a sequence of statements assigning different values to @var{symbol}.
4798 The sequence of statements starts at the @code{.irp} directive, and is
4799 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4800 set to @var{value}, and the sequence of statements is assembled. If no
4801 @var{value} is listed, the sequence of statements is assembled once, with
4802 @var{symbol} set to the null string. To refer to @var{symbol} within the
4803 sequence of statements, use @var{\symbol}.
4805 For example, assembling
4813 is equivalent to assembling
4821 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
4824 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4826 @cindex @code{irpc} directive
4827 Evaluate a sequence of statements assigning different values to @var{symbol}.
4828 The sequence of statements starts at the @code{.irpc} directive, and is
4829 terminated by an @code{.endr} directive. For each character in @var{value},
4830 @var{symbol} is set to the character, and the sequence of statements is
4831 assembled. If no @var{value} is listed, the sequence of statements is
4832 assembled once, with @var{symbol} set to the null string. To refer to
4833 @var{symbol} within the sequence of statements, use @var{\symbol}.
4835 For example, assembling
4843 is equivalent to assembling
4851 For some caveats with the spelling of @var{symbol}, see also the discussion
4855 @section @code{.lcomm @var{symbol} , @var{length}}
4857 @cindex @code{lcomm} directive
4858 @cindex local common symbols
4859 @cindex symbols, local common
4860 Reserve @var{length} (an absolute expression) bytes for a local common
4861 denoted by @var{symbol}. The section and value of @var{symbol} are
4862 those of the new local common. The addresses are allocated in the bss
4863 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4864 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4865 not visible to @code{@value{LD}}.
4868 Some targets permit a third argument to be used with @code{.lcomm}. This
4869 argument specifies the desired alignment of the symbol in the bss section.
4873 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4874 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4878 @section @code{.lflags}
4880 @cindex @code{lflags} directive (ignored)
4881 @command{@value{AS}} accepts this directive, for compatibility with other
4882 assemblers, but ignores it.
4884 @ifclear no-line-dir
4886 @section @code{.line @var{line-number}}
4888 @cindex @code{line} directive
4892 @section @code{.ln @var{line-number}}
4894 @cindex @code{ln} directive
4896 @cindex logical line number
4898 Change the logical line number. @var{line-number} must be an absolute
4899 expression. The next line has that logical line number. Therefore any other
4900 statements on the current line (after a statement separator character) are
4901 reported as on logical line number @var{line-number} @minus{} 1. One day
4902 @command{@value{AS}} will no longer support this directive: it is recognized only
4903 for compatibility with existing assembler programs.
4907 @ifclear no-line-dir
4908 Even though this is a directive associated with the @code{a.out} or
4909 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4910 when producing COFF output, and treats @samp{.line} as though it
4911 were the COFF @samp{.ln} @emph{if} it is found outside a
4912 @code{.def}/@code{.endef} pair.
4914 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4915 used by compilers to generate auxiliary symbol information for
4920 @section @code{.linkonce [@var{type}]}
4922 @cindex @code{linkonce} directive
4923 @cindex common sections
4924 Mark the current section so that the linker only includes a single copy of it.
4925 This may be used to include the same section in several different object files,
4926 but ensure that the linker will only include it once in the final output file.
4927 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4928 Duplicate sections are detected based on the section name, so it should be
4931 This directive is only supported by a few object file formats; as of this
4932 writing, the only object file format which supports it is the Portable
4933 Executable format used on Windows NT.
4935 The @var{type} argument is optional. If specified, it must be one of the
4936 following strings. For example:
4940 Not all types may be supported on all object file formats.
4944 Silently discard duplicate sections. This is the default.
4947 Warn if there are duplicate sections, but still keep only one copy.
4950 Warn if any of the duplicates have different sizes.
4953 Warn if any of the duplicates do not have exactly the same contents.
4957 @section @code{.ln @var{line-number}}
4959 @cindex @code{ln} directive
4960 @ifclear no-line-dir
4961 @samp{.ln} is a synonym for @samp{.line}.
4964 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4965 must be an absolute expression. The next line has that logical
4966 line number, so any other statements on the current line (after a
4967 statement separator character @code{;}) are reported as on logical
4968 line number @var{line-number} @minus{} 1.
4971 This directive is accepted, but ignored, when @command{@value{AS}} is
4972 configured for @code{b.out}; its effect is only associated with COFF
4978 @section @code{.mri @var{val}}
4980 @cindex @code{mri} directive
4981 @cindex MRI mode, temporarily
4982 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4983 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4984 affects code assembled until the next @code{.mri} directive, or until the end
4985 of the file. @xref{M, MRI mode, MRI mode}.
4988 @section @code{.list}
4990 @cindex @code{list} directive
4991 @cindex listing control, turning on
4992 Control (in conjunction with the @code{.nolist} directive) whether or
4993 not assembly listings are generated. These two directives maintain an
4994 internal counter (which is zero initially). @code{.list} increments the
4995 counter, and @code{.nolist} decrements it. Assembly listings are
4996 generated whenever the counter is greater than zero.
4998 By default, listings are disabled. When you enable them (with the
4999 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5000 the initial value of the listing counter is one.
5003 @section @code{.long @var{expressions}}
5005 @cindex @code{long} directive
5006 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5009 @c no one seems to know what this is for or whether this description is
5010 @c what it really ought to do
5012 @section @code{.lsym @var{symbol}, @var{expression}}
5014 @cindex @code{lsym} directive
5015 @cindex symbol, not referenced in assembly
5016 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5017 the hash table, ensuring it cannot be referenced by name during the
5018 rest of the assembly. This sets the attributes of the symbol to be
5019 the same as the expression value:
5021 @var{other} = @var{descriptor} = 0
5022 @var{type} = @r{(section of @var{expression})}
5023 @var{value} = @var{expression}
5026 The new symbol is not flagged as external.
5030 @section @code{.macro}
5033 The commands @code{.macro} and @code{.endm} allow you to define macros that
5034 generate assembly output. For example, this definition specifies a macro
5035 @code{sum} that puts a sequence of numbers into memory:
5038 .macro sum from=0, to=5
5047 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5059 @item .macro @var{macname}
5060 @itemx .macro @var{macname} @var{macargs} @dots{}
5061 @cindex @code{macro} directive
5062 Begin the definition of a macro called @var{macname}. If your macro
5063 definition requires arguments, specify their names after the macro name,
5064 separated by commas or spaces. You can qualify the macro argument to
5065 indicate whether all invocations must specify a non-blank value (through
5066 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5067 (through @samp{:@code{vararg}}). You can supply a default value for any
5068 macro argument by following the name with @samp{=@var{deflt}}. You
5069 cannot define two macros with the same @var{macname} unless it has been
5070 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5071 definitions. For example, these are all valid @code{.macro} statements:
5075 Begin the definition of a macro called @code{comm}, which takes no
5078 @item .macro plus1 p, p1
5079 @itemx .macro plus1 p p1
5080 Either statement begins the definition of a macro called @code{plus1},
5081 which takes two arguments; within the macro definition, write
5082 @samp{\p} or @samp{\p1} to evaluate the arguments.
5084 @item .macro reserve_str p1=0 p2
5085 Begin the definition of a macro called @code{reserve_str}, with two
5086 arguments. The first argument has a default value, but not the second.
5087 After the definition is complete, you can call the macro either as
5088 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5089 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5090 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5091 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5093 @item .macro m p1:req, p2=0, p3:vararg
5094 Begin the definition of a macro called @code{m}, with at least three
5095 arguments. The first argument must always have a value specified, but
5096 not the second, which instead has a default value. The third formal
5097 will get assigned all remaining arguments specified at invocation time.
5099 When you call a macro, you can specify the argument values either by
5100 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5101 @samp{sum to=17, from=9}.
5105 Note that since each of the @var{macargs} can be an identifier exactly
5106 as any other one permitted by the target architecture, there may be
5107 occasional problems if the target hand-crafts special meanings to certain
5108 characters when they occur in a special position. For example, if the colon
5109 (@code{:}) is generally permitted to be part of a symbol name, but the
5110 architecture specific code special-cases it when occurring as the final
5111 character of a symbol (to denote a label), then the macro parameter
5112 replacement code will have no way of knowing that and consider the whole
5113 construct (including the colon) an identifier, and check only this
5114 identifier for being the subject to parameter substitution. So for example
5115 this macro definition:
5123 might not work as expected. Invoking @samp{label foo} might not create a label
5124 called @samp{foo} but instead just insert the text @samp{\l:} into the
5125 assembler source, probably generating an error about an unrecognised
5128 Similarly problems might occur with the period character (@samp{.})
5129 which is often allowed inside opcode names (and hence identifier names). So
5130 for example constructing a macro to build an opcode from a base name and a
5131 length specifier like this:
5134 .macro opcode base length
5139 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5140 instruction but instead generate some kind of error as the assembler tries to
5141 interpret the text @samp{\base.\length}.
5143 There are several possible ways around this problem:
5146 @item Insert white space
5147 If it is possible to use white space characters then this is the simplest
5156 @item Use @samp{\()}
5157 The string @samp{\()} can be used to separate the end of a macro argument from
5158 the following text. eg:
5161 .macro opcode base length
5166 @item Use the alternate macro syntax mode
5167 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5168 used as a separator. eg:
5178 Note: this problem of correctly identifying string parameters to pseudo ops
5179 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5180 and @code{.irpc} (@pxref{Irpc}) as well.
5183 @cindex @code{endm} directive
5184 Mark the end of a macro definition.
5187 @cindex @code{exitm} directive
5188 Exit early from the current macro definition.
5190 @cindex number of macros executed
5191 @cindex macros, count executed
5193 @command{@value{AS}} maintains a counter of how many macros it has
5194 executed in this pseudo-variable; you can copy that number to your
5195 output with @samp{\@@}, but @emph{only within a macro definition}.
5197 @item LOCAL @var{name} [ , @dots{} ]
5198 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5199 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5200 @xref{Altmacro,,@code{.altmacro}}.
5204 @section @code{.altmacro}
5205 Enable alternate macro mode, enabling:
5208 @item LOCAL @var{name} [ , @dots{} ]
5209 One additional directive, @code{LOCAL}, is available. It is used to
5210 generate a string replacement for each of the @var{name} arguments, and
5211 replace any instances of @var{name} in each macro expansion. The
5212 replacement string is unique in the assembly, and different for each
5213 separate macro expansion. @code{LOCAL} allows you to write macros that
5214 define symbols, without fear of conflict between separate macro expansions.
5216 @item String delimiters
5217 You can write strings delimited in these other ways besides
5218 @code{"@var{string}"}:
5221 @item '@var{string}'
5222 You can delimit strings with single-quote characters.
5224 @item <@var{string}>
5225 You can delimit strings with matching angle brackets.
5228 @item single-character string escape
5229 To include any single character literally in a string (even if the
5230 character would otherwise have some special meaning), you can prefix the
5231 character with @samp{!} (an exclamation mark). For example, you can
5232 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
5234 @item Expression results as strings
5235 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
5236 and use the result as a string.
5240 @section @code{.noaltmacro}
5241 Disable alternate macro mode. @xref{Altmacro}.
5244 @section @code{.nolist}
5246 @cindex @code{nolist} directive
5247 @cindex listing control, turning off
5248 Control (in conjunction with the @code{.list} directive) whether or
5249 not assembly listings are generated. These two directives maintain an
5250 internal counter (which is zero initially). @code{.list} increments the
5251 counter, and @code{.nolist} decrements it. Assembly listings are
5252 generated whenever the counter is greater than zero.
5255 @section @code{.octa @var{bignums}}
5257 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5258 @cindex @code{octa} directive
5259 @cindex integer, 16-byte
5260 @cindex sixteen byte integer
5261 This directive expects zero or more bignums, separated by commas. For each
5262 bignum, it emits a 16-byte integer.
5264 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5265 hence @emph{octa}-word for 16 bytes.
5268 @section @code{.org @var{new-lc} , @var{fill}}
5270 @cindex @code{org} directive
5271 @cindex location counter, advancing
5272 @cindex advancing location counter
5273 @cindex current address, advancing
5274 Advance the location counter of the current section to
5275 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5276 expression with the same section as the current subsection. That is,
5277 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5278 wrong section, the @code{.org} directive is ignored. To be compatible
5279 with former assemblers, if the section of @var{new-lc} is absolute,
5280 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5281 is the same as the current subsection.
5283 @code{.org} may only increase the location counter, or leave it
5284 unchanged; you cannot use @code{.org} to move the location counter
5287 @c double negative used below "not undefined" because this is a specific
5288 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5289 @c section. doc@cygnus.com 18feb91
5290 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5291 may not be undefined. If you really detest this restriction we eagerly await
5292 a chance to share your improved assembler.
5294 Beware that the origin is relative to the start of the section, not
5295 to the start of the subsection. This is compatible with other
5296 people's assemblers.
5298 When the location counter (of the current subsection) is advanced, the
5299 intervening bytes are filled with @var{fill} which should be an
5300 absolute expression. If the comma and @var{fill} are omitted,
5301 @var{fill} defaults to zero.
5304 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5306 @cindex padding the location counter given a power of two
5307 @cindex @code{p2align} directive
5308 Pad the location counter (in the current subsection) to a particular
5309 storage boundary. The first expression (which must be absolute) is the
5310 number of low-order zero bits the location counter must have after
5311 advancement. For example @samp{.p2align 3} advances the location
5312 counter until it a multiple of 8. If the location counter is already a
5313 multiple of 8, no change is needed.
5315 The second expression (also absolute) gives the fill value to be stored in the
5316 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5317 padding bytes are normally zero. However, on some systems, if the section is
5318 marked as containing code and the fill value is omitted, the space is filled
5319 with no-op instructions.
5321 The third expression is also absolute, and is also optional. If it is present,
5322 it is the maximum number of bytes that should be skipped by this alignment
5323 directive. If doing the alignment would require skipping more bytes than the
5324 specified maximum, then the alignment is not done at all. You can omit the
5325 fill value (the second argument) entirely by simply using two commas after the
5326 required alignment; this can be useful if you want the alignment to be filled
5327 with no-op instructions when appropriate.
5329 @cindex @code{p2alignw} directive
5330 @cindex @code{p2alignl} directive
5331 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5332 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5333 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5334 fill pattern as a four byte longword value. For example, @code{.p2alignw
5335 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5336 filled in with the value 0x368d (the exact placement of the bytes depends upon
5337 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5342 @section @code{.previous}
5344 @cindex @code{previous} directive
5345 @cindex Section Stack
5346 This is one of the ELF section stack manipulation directives. The others are
5347 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5348 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5349 (@pxref{PopSection}).
5351 This directive swaps the current section (and subsection) with most recently
5352 referenced section (and subsection) prior to this one. Multiple
5353 @code{.previous} directives in a row will flip between two sections (and their
5356 In terms of the section stack, this directive swaps the current section with
5357 the top section on the section stack.
5362 @section @code{.popsection}
5364 @cindex @code{popsection} directive
5365 @cindex Section Stack
5366 This is one of the ELF section stack manipulation directives. The others are
5367 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5368 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5371 This directive replaces the current section (and subsection) with the top
5372 section (and subsection) on the section stack. This section is popped off the
5377 @section @code{.print @var{string}}
5379 @cindex @code{print} directive
5380 @command{@value{AS}} will print @var{string} on the standard output during
5381 assembly. You must put @var{string} in double quotes.
5385 @section @code{.protected @var{names}}
5387 @cindex @code{protected} directive
5389 This is one of the ELF visibility directives. The other two are
5390 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5392 This directive overrides the named symbols default visibility (which is set by
5393 their binding: local, global or weak). The directive sets the visibility to
5394 @code{protected} which means that any references to the symbols from within the
5395 components that defines them must be resolved to the definition in that
5396 component, even if a definition in another component would normally preempt
5401 @section @code{.psize @var{lines} , @var{columns}}
5403 @cindex @code{psize} directive
5404 @cindex listing control: paper size
5405 @cindex paper size, for listings
5406 Use this directive to declare the number of lines---and, optionally, the
5407 number of columns---to use for each page, when generating listings.
5409 If you do not use @code{.psize}, listings use a default line-count
5410 of 60. You may omit the comma and @var{columns} specification; the
5411 default width is 200 columns.
5413 @command{@value{AS}} generates formfeeds whenever the specified number of
5414 lines is exceeded (or whenever you explicitly request one, using
5417 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5418 those explicitly specified with @code{.eject}.
5421 @section @code{.purgem @var{name}}
5423 @cindex @code{purgem} directive
5424 Undefine the macro @var{name}, so that later uses of the string will not be
5425 expanded. @xref{Macro}.
5429 @section @code{.pushsection @var{name} , @var{subsection}}
5431 @cindex @code{pushsection} directive
5432 @cindex Section Stack
5433 This is one of the ELF section stack manipulation directives. The others are
5434 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5435 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5438 This directive pushes the current section (and subsection) onto the
5439 top of the section stack, and then replaces the current section and
5440 subsection with @code{name} and @code{subsection}.
5444 @section @code{.quad @var{bignums}}
5446 @cindex @code{quad} directive
5447 @code{.quad} expects zero or more bignums, separated by commas. For
5448 each bignum, it emits
5450 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5451 warning message; and just takes the lowest order 8 bytes of the bignum.
5452 @cindex eight-byte integer
5453 @cindex integer, 8-byte
5455 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5456 hence @emph{quad}-word for 8 bytes.
5459 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5460 warning message; and just takes the lowest order 16 bytes of the bignum.
5461 @cindex sixteen-byte integer
5462 @cindex integer, 16-byte
5466 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
5468 @cindex @code{reloc} directive
5469 Generate a relocation at @var{offset} of type @var{reloc_name} with value
5470 @var{expression}. If @var{offset} is a number, the relocation is generated in
5471 the current section. If @var{offset} is an expression that resolves to a
5472 symbol plus offset, the relocation is generated in the given symbol's section.
5473 @var{expression}, if present, must resolve to a symbol plus addend or to an
5474 absolute value, but note that not all targets support an addend. e.g. ELF REL
5475 targets such as i386 store an addend in the section contents rather than in the
5476 relocation. This low level interface does not support addends stored in the
5480 @section @code{.rept @var{count}}
5482 @cindex @code{rept} directive
5483 Repeat the sequence of lines between the @code{.rept} directive and the next
5484 @code{.endr} directive @var{count} times.
5486 For example, assembling
5494 is equivalent to assembling
5503 @section @code{.sbttl "@var{subheading}"}
5505 @cindex @code{sbttl} directive
5506 @cindex subtitles for listings
5507 @cindex listing control: subtitle
5508 Use @var{subheading} as the title (third line, immediately after the
5509 title line) when generating assembly listings.
5511 This directive affects subsequent pages, as well as the current page if
5512 it appears within ten lines of the top of a page.
5516 @section @code{.scl @var{class}}
5518 @cindex @code{scl} directive
5519 @cindex symbol storage class (COFF)
5520 @cindex COFF symbol storage class
5521 Set the storage-class value for a symbol. This directive may only be
5522 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5523 whether a symbol is static or external, or it may record further
5524 symbolic debugging information.
5527 The @samp{.scl} directive is primarily associated with COFF output; when
5528 configured to generate @code{b.out} output format, @command{@value{AS}}
5529 accepts this directive but ignores it.
5535 @section @code{.section @var{name}}
5537 @cindex named section
5538 Use the @code{.section} directive to assemble the following code into a section
5541 This directive is only supported for targets that actually support arbitrarily
5542 named sections; on @code{a.out} targets, for example, it is not accepted, even
5543 with a standard @code{a.out} section name.
5547 @c only print the extra heading if both COFF and ELF are set
5548 @subheading COFF Version
5551 @cindex @code{section} directive (COFF version)
5552 For COFF targets, the @code{.section} directive is used in one of the following
5556 .section @var{name}[, "@var{flags}"]
5557 .section @var{name}[, @var{subsegment}]
5560 If the optional argument is quoted, it is taken as flags to use for the
5561 section. Each flag is a single character. The following flags are recognized:
5564 bss section (uninitialized data)
5566 section is not loaded
5576 shared section (meaningful for PE targets)
5578 ignored. (For compatibility with the ELF version)
5581 If no flags are specified, the default flags depend upon the section name. If
5582 the section name is not recognized, the default will be for the section to be
5583 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5584 from the section, rather than adding them, so if they are used on their own it
5585 will be as if no flags had been specified at all.
5587 If the optional argument to the @code{.section} directive is not quoted, it is
5588 taken as a subsegment number (@pxref{Sub-Sections}).
5593 @c only print the extra heading if both COFF and ELF are set
5594 @subheading ELF Version
5597 @cindex Section Stack
5598 This is one of the ELF section stack manipulation directives. The others are
5599 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5600 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5601 @code{.previous} (@pxref{Previous}).
5603 @cindex @code{section} directive (ELF version)
5604 For ELF targets, the @code{.section} directive is used like this:
5607 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
5610 The optional @var{flags} argument is a quoted string which may contain any
5611 combination of the following characters:
5614 section is allocatable
5618 section is executable
5620 section is mergeable
5622 section contains zero terminated strings
5624 section is a member of a section group
5626 section is used for thread-local-storage
5629 The optional @var{type} argument may contain one of the following constants:
5632 section contains data
5634 section does not contain data (i.e., section only occupies space)
5636 section contains data which is used by things other than the program
5638 section contains an array of pointers to init functions
5640 section contains an array of pointers to finish functions
5641 @item @@preinit_array
5642 section contains an array of pointers to pre-init functions
5645 Many targets only support the first three section types.
5647 Note on targets where the @code{@@} character is the start of a comment (eg
5648 ARM) then another character is used instead. For example the ARM port uses the
5651 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
5652 be specified as well as an extra argument---@var{entsize}---like this:
5655 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
5658 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
5659 constants, each @var{entsize} octets long. Sections with both @code{M} and
5660 @code{S} must contain zero terminated strings where each character is
5661 @var{entsize} bytes long. The linker may remove duplicates within sections with
5662 the same name, same entity size and same flags. @var{entsize} must be an
5663 absolute expression.
5665 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
5666 be present along with an additional field like this:
5669 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
5672 The @var{GroupName} field specifies the name of the section group to which this
5673 particular section belongs. The optional linkage field can contain:
5676 indicates that only one copy of this section should be retained
5681 Note: if both the @var{M} and @var{G} flags are present then the fields for
5682 the Merge flag should come first, like this:
5685 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
5688 If no flags are specified, the default flags depend upon the section name. If
5689 the section name is not recognized, the default will be for the section to have
5690 none of the above flags: it will not be allocated in memory, nor writable, nor
5691 executable. The section will contain data.
5693 For ELF targets, the assembler supports another type of @code{.section}
5694 directive for compatibility with the Solaris assembler:
5697 .section "@var{name}"[, @var{flags}...]
5700 Note that the section name is quoted. There may be a sequence of comma
5704 section is allocatable
5708 section is executable
5710 section is used for thread local storage
5713 This directive replaces the current section and subsection. See the
5714 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
5715 some examples of how this directive and the other section stack directives
5721 @section @code{.set @var{symbol}, @var{expression}}
5723 @cindex @code{set} directive
5724 @cindex symbol value, setting
5725 Set the value of @var{symbol} to @var{expression}. This
5726 changes @var{symbol}'s value and type to conform to
5727 @var{expression}. If @var{symbol} was flagged as external, it remains
5728 flagged (@pxref{Symbol Attributes}).
5730 You may @code{.set} a symbol many times in the same assembly.
5732 If you @code{.set} a global symbol, the value stored in the object
5733 file is the last value stored into it.
5736 The syntax for @code{set} on the HPPA is
5737 @samp{@var{symbol} .set @var{expression}}.
5741 On Z80 @code{set} is a real instruction, use
5742 @samp{@var{symbol} defl @var{expression}} instead.
5746 @section @code{.short @var{expressions}}
5748 @cindex @code{short} directive
5750 @code{.short} is normally the same as @samp{.word}.
5751 @xref{Word,,@code{.word}}.
5753 In some configurations, however, @code{.short} and @code{.word} generate
5754 numbers of different lengths. @xref{Machine Dependencies}.
5758 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5761 This expects zero or more @var{expressions}, and emits
5762 a 16 bit number for each.
5767 @section @code{.single @var{flonums}}
5769 @cindex @code{single} directive
5770 @cindex floating point numbers (single)
5771 This directive assembles zero or more flonums, separated by commas. It
5772 has the same effect as @code{.float}.
5774 The exact kind of floating point numbers emitted depends on how
5775 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5779 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5780 numbers in @sc{ieee} format.
5786 @section @code{.size}
5788 This directive is used to set the size associated with a symbol.
5792 @c only print the extra heading if both COFF and ELF are set
5793 @subheading COFF Version
5796 @cindex @code{size} directive (COFF version)
5797 For COFF targets, the @code{.size} directive is only permitted inside
5798 @code{.def}/@code{.endef} pairs. It is used like this:
5801 .size @var{expression}
5805 @samp{.size} is only meaningful when generating COFF format output; when
5806 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5813 @c only print the extra heading if both COFF and ELF are set
5814 @subheading ELF Version
5817 @cindex @code{size} directive (ELF version)
5818 For ELF targets, the @code{.size} directive is used like this:
5821 .size @var{name} , @var{expression}
5824 This directive sets the size associated with a symbol @var{name}.
5825 The size in bytes is computed from @var{expression} which can make use of label
5826 arithmetic. This directive is typically used to set the size of function
5832 @section @code{.sleb128 @var{expressions}}
5834 @cindex @code{sleb128} directive
5835 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5836 compact, variable length representation of numbers used by the DWARF
5837 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
5839 @ifclear no-space-dir
5841 @section @code{.skip @var{size} , @var{fill}}
5843 @cindex @code{skip} directive
5844 @cindex filling memory
5845 This directive emits @var{size} bytes, each of value @var{fill}. Both
5846 @var{size} and @var{fill} are absolute expressions. If the comma and
5847 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5851 @section @code{.space @var{size} , @var{fill}}
5853 @cindex @code{space} directive
5854 @cindex filling memory
5855 This directive emits @var{size} bytes, each of value @var{fill}. Both
5856 @var{size} and @var{fill} are absolute expressions. If the comma
5857 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5862 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5863 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5864 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5865 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5873 @section @code{.stabd, .stabn, .stabs}
5875 @cindex symbolic debuggers, information for
5876 @cindex @code{stab@var{x}} directives
5877 There are three directives that begin @samp{.stab}.
5878 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5879 The symbols are not entered in the @command{@value{AS}} hash table: they
5880 cannot be referenced elsewhere in the source file.
5881 Up to five fields are required:
5885 This is the symbol's name. It may contain any character except
5886 @samp{\000}, so is more general than ordinary symbol names. Some
5887 debuggers used to code arbitrarily complex structures into symbol names
5891 An absolute expression. The symbol's type is set to the low 8 bits of
5892 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5893 and debuggers choke on silly bit patterns.
5896 An absolute expression. The symbol's ``other'' attribute is set to the
5897 low 8 bits of this expression.
5900 An absolute expression. The symbol's descriptor is set to the low 16
5901 bits of this expression.
5904 An absolute expression which becomes the symbol's value.
5907 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5908 or @code{.stabs} statement, the symbol has probably already been created;
5909 you get a half-formed symbol in your object file. This is
5910 compatible with earlier assemblers!
5913 @cindex @code{stabd} directive
5914 @item .stabd @var{type} , @var{other} , @var{desc}
5916 The ``name'' of the symbol generated is not even an empty string.
5917 It is a null pointer, for compatibility. Older assemblers used a
5918 null pointer so they didn't waste space in object files with empty
5921 The symbol's value is set to the location counter,
5922 relocatably. When your program is linked, the value of this symbol
5923 is the address of the location counter when the @code{.stabd} was
5926 @cindex @code{stabn} directive
5927 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5928 The name of the symbol is set to the empty string @code{""}.
5930 @cindex @code{stabs} directive
5931 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5932 All five fields are specified.
5938 @section @code{.string} "@var{str}"
5940 @cindex string, copying to object file
5941 @cindex @code{string} directive
5943 Copy the characters in @var{str} to the object file. You may specify more than
5944 one string to copy, separated by commas. Unless otherwise specified for a
5945 particular machine, the assembler marks the end of each string with a 0 byte.
5946 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5949 @section @code{.struct @var{expression}}
5951 @cindex @code{struct} directive
5952 Switch to the absolute section, and set the section offset to @var{expression},
5953 which must be an absolute expression. You might use this as follows:
5962 This would define the symbol @code{field1} to have the value 0, the symbol
5963 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5964 value 8. Assembly would be left in the absolute section, and you would need to
5965 use a @code{.section} directive of some sort to change to some other section
5966 before further assembly.
5970 @section @code{.subsection @var{name}}
5972 @cindex @code{subsection} directive
5973 @cindex Section Stack
5974 This is one of the ELF section stack manipulation directives. The others are
5975 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5976 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5979 This directive replaces the current subsection with @code{name}. The current
5980 section is not changed. The replaced subsection is put onto the section stack
5981 in place of the then current top of stack subsection.
5986 @section @code{.symver}
5987 @cindex @code{symver} directive
5988 @cindex symbol versioning
5989 @cindex versions of symbols
5990 Use the @code{.symver} directive to bind symbols to specific version nodes
5991 within a source file. This is only supported on ELF platforms, and is
5992 typically used when assembling files to be linked into a shared library.
5993 There are cases where it may make sense to use this in objects to be bound
5994 into an application itself so as to override a versioned symbol from a
5997 For ELF targets, the @code{.symver} directive can be used like this:
5999 .symver @var{name}, @var{name2@@nodename}
6001 If the symbol @var{name} is defined within the file
6002 being assembled, the @code{.symver} directive effectively creates a symbol
6003 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6004 just don't try and create a regular alias is that the @var{@@} character isn't
6005 permitted in symbol names. The @var{name2} part of the name is the actual name
6006 of the symbol by which it will be externally referenced. The name @var{name}
6007 itself is merely a name of convenience that is used so that it is possible to
6008 have definitions for multiple versions of a function within a single source
6009 file, and so that the compiler can unambiguously know which version of a
6010 function is being mentioned. The @var{nodename} portion of the alias should be
6011 the name of a node specified in the version script supplied to the linker when
6012 building a shared library. If you are attempting to override a versioned
6013 symbol from a shared library, then @var{nodename} should correspond to the
6014 nodename of the symbol you are trying to override.
6016 If the symbol @var{name} is not defined within the file being assembled, all
6017 references to @var{name} will be changed to @var{name2@@nodename}. If no
6018 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6021 Another usage of the @code{.symver} directive is:
6023 .symver @var{name}, @var{name2@@@@nodename}
6025 In this case, the symbol @var{name} must exist and be defined within
6026 the file being assembled. It is similar to @var{name2@@nodename}. The
6027 difference is @var{name2@@@@nodename} will also be used to resolve
6028 references to @var{name2} by the linker.
6030 The third usage of the @code{.symver} directive is:
6032 .symver @var{name}, @var{name2@@@@@@nodename}
6034 When @var{name} is not defined within the
6035 file being assembled, it is treated as @var{name2@@nodename}. When
6036 @var{name} is defined within the file being assembled, the symbol
6037 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6042 @section @code{.tag @var{structname}}
6044 @cindex COFF structure debugging
6045 @cindex structure debugging, COFF
6046 @cindex @code{tag} directive
6047 This directive is generated by compilers to include auxiliary debugging
6048 information in the symbol table. It is only permitted inside
6049 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6050 definitions in the symbol table with instances of those structures.
6053 @samp{.tag} is only used when generating COFF format output; when
6054 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6060 @section @code{.text @var{subsection}}
6062 @cindex @code{text} directive
6063 Tells @command{@value{AS}} to assemble the following statements onto the end of
6064 the text subsection numbered @var{subsection}, which is an absolute
6065 expression. If @var{subsection} is omitted, subsection number zero
6069 @section @code{.title "@var{heading}"}
6071 @cindex @code{title} directive
6072 @cindex listing control: title line
6073 Use @var{heading} as the title (second line, immediately after the
6074 source file name and pagenumber) when generating assembly listings.
6076 This directive affects subsequent pages, as well as the current page if
6077 it appears within ten lines of the top of a page.
6081 @section @code{.type}
6083 This directive is used to set the type of a symbol.
6087 @c only print the extra heading if both COFF and ELF are set
6088 @subheading COFF Version
6091 @cindex COFF symbol type
6092 @cindex symbol type, COFF
6093 @cindex @code{type} directive (COFF version)
6094 For COFF targets, this directive is permitted only within
6095 @code{.def}/@code{.endef} pairs. It is used like this:
6101 This records the integer @var{int} as the type attribute of a symbol table
6105 @samp{.type} is associated only with COFF format output; when
6106 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6107 directive but ignores it.
6113 @c only print the extra heading if both COFF and ELF are set
6114 @subheading ELF Version
6117 @cindex ELF symbol type
6118 @cindex symbol type, ELF
6119 @cindex @code{type} directive (ELF version)
6120 For ELF targets, the @code{.type} directive is used like this:
6123 .type @var{name} , @var{type description}
6126 This sets the type of symbol @var{name} to be either a
6127 function symbol or an object symbol. There are five different syntaxes
6128 supported for the @var{type description} field, in order to provide
6129 compatibility with various other assemblers.
6131 Because some of the characters used in these syntaxes (such as @samp{@@} and
6132 @samp{#}) are comment characters for some architectures, some of the syntaxes
6133 below do not work on all architectures. The first variant will be accepted by
6134 the GNU assembler on all architectures so that variant should be used for
6135 maximum portability, if you do not need to assemble your code with other
6138 The syntaxes supported are:
6141 .type <name> STT_FUNCTION
6142 .type <name> STT_OBJECT
6144 .type <name>,#function
6145 .type <name>,#object
6147 .type <name>,@@function
6148 .type <name>,@@object
6150 .type <name>,%function
6151 .type <name>,%object
6153 .type <name>,"function"
6154 .type <name>,"object"
6160 @section @code{.uleb128 @var{expressions}}
6162 @cindex @code{uleb128} directive
6163 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6164 compact, variable length representation of numbers used by the DWARF
6165 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
6169 @section @code{.val @var{addr}}
6171 @cindex @code{val} directive
6172 @cindex COFF value attribute
6173 @cindex value attribute, COFF
6174 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6175 records the address @var{addr} as the value attribute of a symbol table
6179 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6180 configured for @code{b.out}, it accepts this directive but ignores it.
6186 @section @code{.version "@var{string}"}
6188 @cindex @code{version} directive
6189 This directive creates a @code{.note} section and places into it an ELF
6190 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6195 @section @code{.vtable_entry @var{table}, @var{offset}}
6197 @cindex @code{vtable_entry} directive
6198 This directive finds or creates a symbol @code{table} and creates a
6199 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6202 @section @code{.vtable_inherit @var{child}, @var{parent}}
6204 @cindex @code{vtable_inherit} directive
6205 This directive finds the symbol @code{child} and finds or creates the symbol
6206 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6207 parent whose addend is the value of the child symbol. As a special case the
6208 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
6212 @section @code{.warning "@var{string}"}
6213 @cindex warning directive
6214 Similar to the directive @code{.error}
6215 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6218 @section @code{.weak @var{names}}
6220 @cindex @code{weak} directive
6221 This directive sets the weak attribute on the comma separated list of symbol
6222 @code{names}. If the symbols do not already exist, they will be created.
6224 On COFF targets other than PE, weak symbols are a GNU extension. This
6225 directive sets the weak attribute on the comma separated list of symbol
6226 @code{names}. If the symbols do not already exist, they will be created.
6228 On the PE target, weak symbols are supported natively as weak aliases.
6229 When a weak symbol is created that is not an alias, GAS creates an
6230 alternate symbol to hold the default value.
6233 @section @code{.weakref @var{alias}, @var{target}}
6235 @cindex @code{weakref} directive
6236 This directive creates an alias to the target symbol that enables the symbol to
6237 be referenced with weak-symbol semantics, but without actually making it weak.
6238 If direct references or definitions of the symbol are present, then the symbol
6239 will not be weak, but if all references to it are through weak references, the
6240 symbol will be marked as weak in the symbol table.
6242 The effect is equivalent to moving all references to the alias to a separate
6243 assembly source file, renaming the alias to the symbol in it, declaring the
6244 symbol as weak there, and running a reloadable link to merge the object files
6245 resulting from the assembly of the new source file and the old source file that
6246 had the references to the alias removed.
6248 The alias itself never makes to the symbol table, and is entirely handled
6249 within the assembler.
6252 @section @code{.word @var{expressions}}
6254 @cindex @code{word} directive
6255 This directive expects zero or more @var{expressions}, of any section,
6256 separated by commas.
6259 For each expression, @command{@value{AS}} emits a 32-bit number.
6262 For each expression, @command{@value{AS}} emits a 16-bit number.
6267 The size of the number emitted, and its byte order,
6268 depend on what target computer the assembly is for.
6271 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6272 @c happen---32-bit addressability, period; no long/short jumps.
6273 @ifset DIFF-TBL-KLUGE
6274 @cindex difference tables altered
6275 @cindex altered difference tables
6277 @emph{Warning: Special Treatment to support Compilers}
6281 Machines with a 32-bit address space, but that do less than 32-bit
6282 addressing, require the following special treatment. If the machine of
6283 interest to you does 32-bit addressing (or doesn't require it;
6284 @pxref{Machine Dependencies}), you can ignore this issue.
6287 In order to assemble compiler output into something that works,
6288 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6289 Directives of the form @samp{.word sym1-sym2} are often emitted by
6290 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6291 directive of the form @samp{.word sym1-sym2}, and the difference between
6292 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6293 creates a @dfn{secondary jump table}, immediately before the next label.
6294 This secondary jump table is preceded by a short-jump to the
6295 first byte after the secondary table. This short-jump prevents the flow
6296 of control from accidentally falling into the new table. Inside the
6297 table is a long-jump to @code{sym2}. The original @samp{.word}
6298 contains @code{sym1} minus the address of the long-jump to
6301 If there were several occurrences of @samp{.word sym1-sym2} before the
6302 secondary jump table, all of them are adjusted. If there was a
6303 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6304 long-jump to @code{sym4} is included in the secondary jump table,
6305 and the @code{.word} directives are adjusted to contain @code{sym3}
6306 minus the address of the long-jump to @code{sym4}; and so on, for as many
6307 entries in the original jump table as necessary.
6310 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6311 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6312 assembly language programmers.
6315 @c end DIFF-TBL-KLUGE
6318 @section Deprecated Directives
6320 @cindex deprecated directives
6321 @cindex obsolescent directives
6322 One day these directives won't work.
6323 They are included for compatibility with older assemblers.
6330 @node Machine Dependencies
6331 @chapter Machine Dependent Features
6333 @cindex machine dependencies
6334 The machine instruction sets are (almost by definition) different on
6335 each machine where @command{@value{AS}} runs. Floating point representations
6336 vary as well, and @command{@value{AS}} often supports a few additional
6337 directives or command-line options for compatibility with other
6338 assemblers on a particular platform. Finally, some versions of
6339 @command{@value{AS}} support special pseudo-instructions for branch
6342 This chapter discusses most of these differences, though it does not
6343 include details on any machine's instruction set. For details on that
6344 subject, see the hardware manufacturer's manual.
6348 * Alpha-Dependent:: Alpha Dependent Features
6351 * ARC-Dependent:: ARC Dependent Features
6354 * ARM-Dependent:: ARM Dependent Features
6357 * AVR-Dependent:: AVR Dependent Features
6360 * BFIN-Dependent:: BFIN Dependent Features
6363 * CRIS-Dependent:: CRIS Dependent Features
6366 * D10V-Dependent:: D10V Dependent Features
6369 * D30V-Dependent:: D30V Dependent Features
6372 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6375 * HPPA-Dependent:: HPPA Dependent Features
6378 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6381 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6384 * i860-Dependent:: Intel 80860 Dependent Features
6387 * i960-Dependent:: Intel 80960 Dependent Features
6390 * IA-64-Dependent:: Intel IA-64 Dependent Features
6393 * IP2K-Dependent:: IP2K Dependent Features
6396 * M32C-Dependent:: M32C Dependent Features
6399 * M32R-Dependent:: M32R Dependent Features
6402 * M68K-Dependent:: M680x0 Dependent Features
6405 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6408 * MIPS-Dependent:: MIPS Dependent Features
6411 * MMIX-Dependent:: MMIX Dependent Features
6414 * MSP430-Dependent:: MSP430 Dependent Features
6417 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6418 * SH64-Dependent:: SuperH SH64 Dependent Features
6421 * PDP-11-Dependent:: PDP-11 Dependent Features
6424 * PJ-Dependent:: picoJava Dependent Features
6427 * PPC-Dependent:: PowerPC Dependent Features
6430 * Sparc-Dependent:: SPARC Dependent Features
6433 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
6436 * V850-Dependent:: V850 Dependent Features
6439 * Xtensa-Dependent:: Xtensa Dependent Features
6442 * Z80-Dependent:: Z80 Dependent Features
6445 * Z8000-Dependent:: Z8000 Dependent Features
6448 * Vax-Dependent:: VAX Dependent Features
6455 @c The following major nodes are *sections* in the GENERIC version, *chapters*
6456 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
6457 @c peculiarity: to preserve cross-references, there must be a node called
6458 @c "Machine Dependencies". Hence the conditional nodenames in each
6459 @c major node below. Node defaulting in makeinfo requires adjacency of
6460 @c node and sectioning commands; hence the repetition of @chapter BLAH
6461 @c in both conditional blocks.
6464 @include c-alpha.texi
6480 @include c-bfin.texi
6484 @include c-cris.texi
6489 @node Machine Dependencies
6490 @chapter Machine Dependent Features
6492 The machine instruction sets are different on each Renesas chip family,
6493 and there are also some syntax differences among the families. This
6494 chapter describes the specific @command{@value{AS}} features for each
6498 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6499 * SH-Dependent:: Renesas SH Dependent Features
6506 @include c-d10v.texi
6510 @include c-d30v.texi
6514 @include c-h8300.texi
6518 @include c-hppa.texi
6522 @include c-i370.texi
6526 @include c-i386.texi
6530 @include c-i860.texi
6534 @include c-i960.texi
6538 @include c-ia64.texi
6542 @include c-ip2k.texi
6546 @include c-m32c.texi
6550 @include c-m32r.texi
6554 @include c-m68k.texi
6558 @include c-m68hc11.texi
6562 @include c-mips.texi
6566 @include c-mmix.texi
6570 @include c-msp430.texi
6574 @include c-ns32k.texi
6578 @include c-pdp11.texi
6591 @include c-sh64.texi
6595 @include c-sparc.texi
6599 @include c-tic54x.texi
6615 @include c-v850.texi
6619 @include c-xtensa.texi
6623 @c reverse effect of @down at top of generic Machine-Dep chapter
6627 @node Reporting Bugs
6628 @chapter Reporting Bugs
6629 @cindex bugs in assembler
6630 @cindex reporting bugs in assembler
6632 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6634 Reporting a bug may help you by bringing a solution to your problem, or it may
6635 not. But in any case the principal function of a bug report is to help the
6636 entire community by making the next version of @command{@value{AS}} work better.
6637 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6639 In order for a bug report to serve its purpose, you must include the
6640 information that enables us to fix the bug.
6643 * Bug Criteria:: Have you found a bug?
6644 * Bug Reporting:: How to report bugs
6648 @section Have You Found a Bug?
6649 @cindex bug criteria
6651 If you are not sure whether you have found a bug, here are some guidelines:
6654 @cindex fatal signal
6655 @cindex assembler crash
6656 @cindex crash of assembler
6658 If the assembler gets a fatal signal, for any input whatever, that is a
6659 @command{@value{AS}} bug. Reliable assemblers never crash.
6661 @cindex error on valid input
6663 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6665 @cindex invalid input
6667 If @command{@value{AS}} does not produce an error message for invalid input, that
6668 is a bug. However, you should note that your idea of ``invalid input'' might
6669 be our idea of ``an extension'' or ``support for traditional practice''.
6672 If you are an experienced user of assemblers, your suggestions for improvement
6673 of @command{@value{AS}} are welcome in any case.
6677 @section How to Report Bugs
6679 @cindex assembler bugs, reporting
6681 A number of companies and individuals offer support for @sc{gnu} products. If
6682 you obtained @command{@value{AS}} from a support organization, we recommend you
6683 contact that organization first.
6685 You can find contact information for many support companies and
6686 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6690 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6694 The fundamental principle of reporting bugs usefully is this:
6695 @strong{report all the facts}. If you are not sure whether to state a
6696 fact or leave it out, state it!
6698 Often people omit facts because they think they know what causes the problem
6699 and assume that some details do not matter. Thus, you might assume that the
6700 name of a symbol you use in an example does not matter. Well, probably it does
6701 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6702 happens to fetch from the location where that name is stored in memory;
6703 perhaps, if the name were different, the contents of that location would fool
6704 the assembler into doing the right thing despite the bug. Play it safe and
6705 give a specific, complete example. That is the easiest thing for you to do,
6706 and the most helpful.
6708 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6709 it is new to us. Therefore, always write your bug reports on the assumption
6710 that the bug has not been reported previously.
6712 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6713 bell?'' This cannot help us fix a bug, so it is basically useless. We
6714 respond by asking for enough details to enable us to investigate.
6715 You might as well expedite matters by sending them to begin with.
6717 To enable us to fix the bug, you should include all these things:
6721 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6722 it with the @samp{--version} argument.
6724 Without this, we will not know whether there is any point in looking for
6725 the bug in the current version of @command{@value{AS}}.
6728 Any patches you may have applied to the @command{@value{AS}} source.
6731 The type of machine you are using, and the operating system name and
6735 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6739 The command arguments you gave the assembler to assemble your example and
6740 observe the bug. To guarantee you will not omit something important, list them
6741 all. A copy of the Makefile (or the output from make) is sufficient.
6743 If we were to try to guess the arguments, we would probably guess wrong
6744 and then we might not encounter the bug.
6747 A complete input file that will reproduce the bug. If the bug is observed when
6748 the assembler is invoked via a compiler, send the assembler source, not the
6749 high level language source. Most compilers will produce the assembler source
6750 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6751 the options @samp{-v --save-temps}; this will save the assembler source in a
6752 file with an extension of @file{.s}, and also show you exactly how
6753 @command{@value{AS}} is being run.
6756 A description of what behavior you observe that you believe is
6757 incorrect. For example, ``It gets a fatal signal.''
6759 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6760 will certainly notice it. But if the bug is incorrect output, we might not
6761 notice unless it is glaringly wrong. You might as well not give us a chance to
6764 Even if the problem you experience is a fatal signal, you should still say so
6765 explicitly. Suppose something strange is going on, such as, your copy of
6766 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
6767 library on your system. (This has happened!) Your copy might crash and ours
6768 would not. If you told us to expect a crash, then when ours fails to crash, we
6769 would know that the bug was not happening for us. If you had not told us to
6770 expect a crash, then we would not be able to draw any conclusion from our
6774 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6775 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6776 option. Always send diffs from the old file to the new file. If you even
6777 discuss something in the @command{@value{AS}} source, refer to it by context, not
6780 The line numbers in our development sources will not match those in your
6781 sources. Your line numbers would convey no useful information to us.
6784 Here are some things that are not necessary:
6788 A description of the envelope of the bug.
6790 Often people who encounter a bug spend a lot of time investigating
6791 which changes to the input file will make the bug go away and which
6792 changes will not affect it.
6794 This is often time consuming and not very useful, because the way we
6795 will find the bug is by running a single example under the debugger
6796 with breakpoints, not by pure deduction from a series of examples.
6797 We recommend that you save your time for something else.
6799 Of course, if you can find a simpler example to report @emph{instead}
6800 of the original one, that is a convenience for us. Errors in the
6801 output will be easier to spot, running under the debugger will take
6802 less time, and so on.
6804 However, simplification is not vital; if you do not want to do this,
6805 report the bug anyway and send us the entire test case you used.
6808 A patch for the bug.
6810 A patch for the bug does help us if it is a good one. But do not omit
6811 the necessary information, such as the test case, on the assumption that
6812 a patch is all we need. We might see problems with your patch and decide
6813 to fix the problem another way, or we might not understand it at all.
6815 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6816 construct an example that will make the program follow a certain path through
6817 the code. If you do not send us the example, we will not be able to construct
6818 one, so we will not be able to verify that the bug is fixed.
6820 And if we cannot understand what bug you are trying to fix, or why your
6821 patch should be an improvement, we will not install it. A test case will
6822 help us to understand.
6825 A guess about what the bug is or what it depends on.
6827 Such guesses are usually wrong. Even we cannot guess right about such
6828 things without first using the debugger to find the facts.
6831 @node Acknowledgements
6832 @chapter Acknowledgements
6834 If you have contributed to GAS and your name isn't listed here,
6835 it is not meant as a slight. We just don't know about it. Send mail to the
6836 maintainer, and we'll correct the situation. Currently
6838 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6840 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6843 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6844 information and the 68k series machines, most of the preprocessing pass, and
6845 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6847 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6848 many bug fixes, including merging support for several processors, breaking GAS
6849 up to handle multiple object file format back ends (including heavy rewrite,
6850 testing, an integration of the coff and b.out back ends), adding configuration
6851 including heavy testing and verification of cross assemblers and file splits
6852 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6853 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6854 port (including considerable amounts of reverse engineering), a SPARC opcode
6855 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6856 assertions and made them work, much other reorganization, cleanup, and lint.
6858 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6859 in format-specific I/O modules.
6861 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6862 has done much work with it since.
6864 The Intel 80386 machine description was written by Eliot Dresselhaus.
6866 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6868 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6869 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6871 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6872 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6873 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6874 support a.out format.
6876 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
6877 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
6878 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6879 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6882 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6883 simplified the configuration of which versions accept which directives. He
6884 updated the 68k machine description so that Motorola's opcodes always produced
6885 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6886 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6887 cross-compilation support, and one bug in relaxation that took a week and
6888 required the proverbial one-bit fix.
6890 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6891 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6892 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6893 PowerPC assembler, and made a few other minor patches.
6895 Steve Chamberlain made GAS able to generate listings.
6897 Hewlett-Packard contributed support for the HP9000/300.
6899 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6900 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6901 formats). This work was supported by both the Center for Software Science at
6902 the University of Utah and Cygnus Support.
6904 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6905 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6906 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6907 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6908 and some initial 64-bit support).
6910 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6912 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6913 support for openVMS/Alpha.
6915 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6918 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6919 Inc.@: added support for Xtensa processors.
6921 Several engineers at Cygnus Support have also provided many small bug fixes and
6922 configuration enhancements.
6924 Many others have contributed large or small bugfixes and enhancements. If
6925 you have contributed significant work and are not mentioned on this list, and
6926 want to be, let us know. Some of the history has been lost; we are not
6927 intentionally leaving anyone out.
6932 @unnumbered AS Index
projects / deliverable / binutils-gdb.git / blob