1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
3 @c 2001, 2002, 2003, 2004, 2005, 2006
4 @c Free Software Foundation, Inc.
5 @c UPDATE!! On future updates--
6 @c (1) check for new machine-dep cmdline options in
7 @c md_parse_option definitions in config/tc-*.c
8 @c (2) for platform-specific directives, examine md_pseudo_op
10 @c (3) for object-format specific directives, examine obj_pseudo_op
12 @c (4) portable directives in potable[] in read.c
16 @macro gcctabopt{body}
19 @c defaults, config file may override:
24 @include asconfig.texi
29 @c common OR combinations of conditions
52 @set abnormal-separator
56 @settitle Using @value{AS}
59 @settitle Using @value{AS} (@value{TARGET})
61 @setchapternewpage odd
66 @c WARE! Some of the machine-dependent sections contain tables of machine
67 @c instructions. Except in multi-column format, these tables look silly.
68 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
69 @c the multi-col format is faked within @example sections.
71 @c Again unfortunately, the natural size that fits on a page, for these tables,
72 @c is different depending on whether or not smallbook is turned on.
73 @c This matters, because of order: text flow switches columns at each page
76 @c The format faked in this source works reasonably well for smallbook,
77 @c not well for the default large-page format. This manual expects that if you
78 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
79 @c tables in question. You can turn on one without the other at your
80 @c discretion, of course.
83 @c the insn tables look just as silly in info files regardless of smallbook,
84 @c might as well show 'em anyways.
90 * As: (as). The GNU assembler.
91 * Gas: (as). The GNU assembler.
100 This file documents the GNU Assembler "@value{AS}".
102 @c man begin COPYRIGHT
103 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002, 2006 Free Software Foundation, Inc.
105 Permission is granted to copy, distribute and/or modify this document
106 under the terms of the GNU Free Documentation License, Version 1.1
107 or any later version published by the Free Software Foundation;
108 with no Invariant Sections, with no Front-Cover Texts, and with no
109 Back-Cover Texts. A copy of the license is included in the
110 section entitled ``GNU Free Documentation License''.
115 Permission is granted to process this file through Tex and print the
116 results, provided the printed document carries copying permission
117 notice identical to this one except for the removal of this paragraph
118 (this paragraph not being relevant to the printed manual).
124 @title Using @value{AS}
125 @subtitle The @sc{gnu} Assembler
127 @subtitle for the @value{TARGET} family
130 @subtitle Version @value{VERSION}
133 The Free Software Foundation Inc. thanks The Nice Computer
134 Company of Australia for loaning Dean Elsner to write the
135 first (Vax) version of @command{as} for Project @sc{gnu}.
136 The proprietors, management and staff of TNCCA thank FSF for
137 distracting the boss while they got some work
140 @author Dean Elsner, Jay Fenlason & friends
144 \hfill {\it Using {\tt @value{AS}}}\par
145 \hfill Edited by Cygnus Support\par
147 %"boxit" macro for figures:
148 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
149 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
150 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
151 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
152 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
155 @vskip 0pt plus 1filll
156 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002, 2006 Free Software Foundation, Inc.
158 Permission is granted to copy, distribute and/or modify this document
159 under the terms of the GNU Free Documentation License, Version 1.1
160 or any later version published by the Free Software Foundation;
161 with no Invariant Sections, with no Front-Cover Texts, and with no
162 Back-Cover Texts. A copy of the license is included in the
163 section entitled ``GNU Free Documentation License''.
169 @top Using @value{AS}
171 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}} version
174 This version of the file describes @command{@value{AS}} configured to generate
175 code for @value{TARGET} architectures.
178 This document is distributed under the terms of the GNU Free
179 Documentation License. A copy of the license is included in the
180 section entitled ``GNU Free Documentation License''.
183 * Overview:: Overview
184 * Invoking:: Command-Line Options
186 * Sections:: Sections and Relocation
188 * Expressions:: Expressions
189 * Pseudo Ops:: Assembler Directives
190 * Machine Dependencies:: Machine Dependent Features
191 * Reporting Bugs:: Reporting Bugs
192 * Acknowledgements:: Who Did What
193 * GNU Free Documentation License:: GNU Free Documentation License
194 * AS Index:: AS Index
201 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
203 This version of the manual describes @command{@value{AS}} configured to generate
204 code for @value{TARGET} architectures.
208 @cindex invocation summary
209 @cindex option summary
210 @cindex summary of options
211 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
212 @pxref{Invoking,,Command-Line Options}.
214 @c man title AS the portable GNU assembler.
218 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
222 @c We don't use deffn and friends for the following because they seem
223 @c to be limited to one line for the header.
225 @c man begin SYNOPSIS
226 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
227 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
228 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
229 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
230 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
231 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
232 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
233 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
234 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
235 [@b{--target-help}] [@var{target-options}]
236 [@b{--}|@var{files} @dots{}]
238 @c Target dependent options are listed below. Keep the list sorted.
239 @c Add an empty line for separation.
242 @emph{Target Alpha options:}
244 [@b{-mdebug} | @b{-no-mdebug}]
245 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
246 [@b{-F}] [@b{-32addr}]
250 @emph{Target ARC options:}
256 @emph{Target ARM options:}
257 @c Don't document the deprecated options
258 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
259 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
260 [@b{-mfpu}=@var{floating-point-format}]
261 [@b{-mfloat-abi}=@var{abi}]
262 [@b{-meabi}=@var{ver}]
265 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
266 @b{-mapcs-reentrant}]
267 [@b{-mthumb-interwork}] [@b{-k}]
271 @emph{Target CRIS options:}
272 [@b{--underscore} | @b{--no-underscore}]
274 [@b{--emulation=criself} | @b{--emulation=crisaout}]
275 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
276 @c Deprecated -- deliberately not documented.
281 @emph{Target D10V options:}
286 @emph{Target D30V options:}
287 [@b{-O}|@b{-n}|@b{-N}]
290 @c Renesas family chips have no machine-dependent assembler options
293 @c HPPA has no machine-dependent assembler options (yet).
297 @emph{Target i386 options:}
298 [@b{--32}|@b{--64}] [@b{-n}]
299 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}]
303 @emph{Target i960 options:}
304 @c see md_parse_option in tc-i960.c
305 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
307 [@b{-b}] [@b{-no-relax}]
311 @emph{Target IA-64 options:}
312 [@b{-mconstant-gp}|@b{-mauto-pic}]
313 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
315 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
316 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
317 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
318 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
322 @emph{Target IP2K options:}
323 [@b{-mip2022}|@b{-mip2022ext}]
327 @emph{Target M32C options:}
328 [@b{-m32c}|@b{-m16c}]
332 @emph{Target M32R options:}
333 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
338 @emph{Target M680X0 options:}
339 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
343 @emph{Target M68HC11 options:}
344 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
345 [@b{-mshort}|@b{-mlong}]
346 [@b{-mshort-double}|@b{-mlong-double}]
347 [@b{--force-long-branchs}] [@b{--short-branchs}]
348 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
349 [@b{--print-opcodes}] [@b{--generate-example}]
353 @emph{Target MCORE options:}
354 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
355 [@b{-mcpu=[210|340]}]
359 @emph{Target MIPS options:}
360 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
361 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
362 [@b{-non_shared}] [@b{-xgot}]
363 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
364 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
365 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
366 [@b{-mips64}] [@b{-mips64r2}]
367 [@b{-construct-floats}] [@b{-no-construct-floats}]
368 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
369 [@b{-mfix7000}] [@b{-mno-fix7000}]
370 [@b{-mips16}] [@b{-no-mips16}]
371 [@b{-msmartmips}] [@b{-mno-smartmips}]
372 [@b{-mips3d}] [@b{-no-mips3d}]
373 [@b{-mdmx}] [@b{-no-mdmx}]
374 [@b{-mdsp}] [@b{-mno-dsp}]
375 [@b{-mmt}] [@b{-mno-mt}]
376 [@b{-mdebug}] [@b{-no-mdebug}]
377 [@b{-mpdr}] [@b{-mno-pdr}]
381 @emph{Target MMIX options:}
382 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
383 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
384 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
385 [@b{--linker-allocated-gregs}]
389 @emph{Target PDP11 options:}
390 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
391 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
392 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
396 @emph{Target picoJava options:}
401 @emph{Target PowerPC options:}
402 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
403 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
404 @b{-mbooke32}|@b{-mbooke64}]
405 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
406 [@b{-mregnames}|@b{-mno-regnames}]
407 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
408 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
409 [@b{-msolaris}|@b{-mno-solaris}]
413 @emph{Target SPARC options:}
414 @c The order here is important. See c-sparc.texi.
415 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
416 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
417 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
422 @emph{Target TIC54X options:}
423 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
424 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
429 @emph{Target Z80 options:}
430 [@b{-z80}] [@b{-r800}]
431 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
432 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
433 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
434 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
435 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
436 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
440 @c Z8000 has no machine-dependent assembler options
444 @emph{Target Xtensa options:}
445 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
446 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
447 [@b{--[no-]transform}]
448 [@b{--rename-section} @var{oldname}=@var{newname}]
456 @include at-file.texi
459 Turn on listings, in any of a variety of ways:
463 omit false conditionals
466 omit debugging directives
469 include high-level source
475 include macro expansions
478 omit forms processing
484 set the name of the listing file
487 You may combine these options; for example, use @samp{-aln} for assembly
488 listing without forms processing. The @samp{=file} option, if used, must be
489 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
492 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
495 Ignored. This option is accepted for script compatibility with calls to
498 @item --defsym @var{sym}=@var{value}
499 Define the symbol @var{sym} to be @var{value} before assembling the input file.
500 @var{value} must be an integer constant. As in C, a leading @samp{0x}
501 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
504 ``fast''---skip whitespace and comment preprocessing (assume source is
509 Generate debugging information for each assembler source line using whichever
510 debug format is preferred by the target. This currently means either STABS,
514 Generate stabs debugging information for each assembler line. This
515 may help debugging assembler code, if the debugger can handle it.
518 Generate stabs debugging information for each assembler line, with GNU
519 extensions that probably only gdb can handle, and that could make other
520 debuggers crash or refuse to read your program. This
521 may help debugging assembler code. Currently the only GNU extension is
522 the location of the current working directory at assembling time.
525 Generate DWARF2 debugging information for each assembler line. This
526 may help debugging assembler code, if the debugger can handle it. Note---this
527 option is only supported by some targets, not all of them.
530 Print a summary of the command line options and exit.
533 Print a summary of all target specific options and exit.
536 Add directory @var{dir} to the search list for @code{.include} directives.
539 Don't warn about signed overflow.
542 @ifclear DIFF-TBL-KLUGE
543 This option is accepted but has no effect on the @value{TARGET} family.
545 @ifset DIFF-TBL-KLUGE
546 Issue warnings when difference tables altered for long displacements.
551 Keep (in the symbol table) local symbols. On traditional a.out systems
552 these start with @samp{L}, but different systems have different local
555 @item --listing-lhs-width=@var{number}
556 Set the maximum width, in words, of the output data column for an assembler
557 listing to @var{number}.
559 @item --listing-lhs-width2=@var{number}
560 Set the maximum width, in words, of the output data column for continuation
561 lines in an assembler listing to @var{number}.
563 @item --listing-rhs-width=@var{number}
564 Set the maximum width of an input source line, as displayed in a listing, to
567 @item --listing-cont-lines=@var{number}
568 Set the maximum number of lines printed in a listing for a single line of input
571 @item -o @var{objfile}
572 Name the object-file output from @command{@value{AS}} @var{objfile}.
575 Fold the data section into the text section.
577 @kindex --hash-size=@var{number}
578 Set the default size of GAS's hash tables to a prime number close to
579 @var{number}. Increasing this value can reduce the length of time it takes the
580 assembler to perform its tasks, at the expense of increasing the assembler's
581 memory requirements. Similarly reducing this value can reduce the memory
582 requirements at the expense of speed.
584 @item --reduce-memory-overheads
585 This option reduces GAS's memory requirements, at the expense of making the
586 assembly processes slower. Currently this switch is a synonym for
587 @samp{--hash-size=4051}, but in the future it may have other effects as well.
590 Print the maximum space (in bytes) and total time (in seconds) used by
593 @item --strip-local-absolute
594 Remove local absolute symbols from the outgoing symbol table.
598 Print the @command{as} version.
601 Print the @command{as} version and exit.
605 Suppress warning messages.
607 @item --fatal-warnings
608 Treat warnings as errors.
611 Don't suppress warning messages or treat them as errors.
620 Generate an object file even after errors.
622 @item -- | @var{files} @dots{}
623 Standard input, or source files to assemble.
628 The following options are available when @value{AS} is configured for
633 This option selects the core processor variant.
635 Select either big-endian (-EB) or little-endian (-EL) output.
640 The following options are available when @value{AS} is configured for the ARM
644 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
645 Specify which ARM processor variant is the target.
646 @item -march=@var{architecture}[+@var{extension}@dots{}]
647 Specify which ARM architecture variant is used by the target.
648 @item -mfpu=@var{floating-point-format}
649 Select which Floating Point architecture is the target.
650 @item -mfloat-abi=@var{abi}
651 Select which floating point ABI is in use.
653 Enable Thumb only instruction decoding.
654 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
655 Select which procedure calling convention is in use.
657 Select either big-endian (-EB) or little-endian (-EL) output.
658 @item -mthumb-interwork
659 Specify that the code has been generated with interworking between Thumb and
662 Specify that PIC code has been generated.
667 See the info pages for documentation of the CRIS-specific options.
671 The following options are available when @value{AS} is configured for
674 @cindex D10V optimization
675 @cindex optimization, D10V
677 Optimize output by parallelizing instructions.
682 The following options are available when @value{AS} is configured for a D30V
685 @cindex D30V optimization
686 @cindex optimization, D30V
688 Optimize output by parallelizing instructions.
692 Warn when nops are generated.
694 @cindex D30V nops after 32-bit multiply
696 Warn when a nop after a 32-bit multiply instruction is generated.
701 The following options are available when @value{AS} is configured for the
702 Intel 80960 processor.
705 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
706 Specify which variant of the 960 architecture is the target.
709 Add code to collect statistics about branches taken.
712 Do not alter compare-and-branch instructions for long displacements;
719 The following options are available when @value{AS} is configured for the
725 Specifies that the extended IP2022 instructions are allowed.
728 Restores the default behaviour, which restricts the permitted instructions to
729 just the basic IP2022 ones.
735 The following options are available when @value{AS} is configured for the
736 Renesas M32C and M16C processors.
741 Assemble M32C instructions.
744 Assemble M16C instructions (the default).
750 The following options are available when @value{AS} is configured for the
751 Renesas M32R (formerly Mitsubishi M32R) series.
756 Specify which processor in the M32R family is the target. The default
757 is normally the M32R, but this option changes it to the M32RX.
759 @item --warn-explicit-parallel-conflicts or --Wp
760 Produce warning messages when questionable parallel constructs are
763 @item --no-warn-explicit-parallel-conflicts or --Wnp
764 Do not produce warning messages when questionable parallel constructs are
771 The following options are available when @value{AS} is configured for the
772 Motorola 68000 series.
777 Shorten references to undefined symbols, to one word instead of two.
779 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
780 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
781 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
782 Specify what processor in the 68000 family is the target. The default
783 is normally the 68020, but this can be changed at configuration time.
785 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
786 The target machine does (or does not) have a floating-point coprocessor.
787 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
788 the basic 68000 is not compatible with the 68881, a combination of the
789 two can be specified, since it's possible to do emulation of the
790 coprocessor instructions with the main processor.
792 @item -m68851 | -mno-68851
793 The target machine does (or does not) have a memory-management
794 unit coprocessor. The default is to assume an MMU for 68020 and up.
801 For details about the PDP-11 machine dependent features options,
802 see @ref{PDP-11-Options}.
805 @item -mpic | -mno-pic
806 Generate position-independent (or position-dependent) code. The
807 default is @option{-mpic}.
810 @itemx -mall-extensions
811 Enable all instruction set extensions. This is the default.
813 @item -mno-extensions
814 Disable all instruction set extensions.
816 @item -m@var{extension} | -mno-@var{extension}
817 Enable (or disable) a particular instruction set extension.
820 Enable the instruction set extensions supported by a particular CPU, and
821 disable all other extensions.
823 @item -m@var{machine}
824 Enable the instruction set extensions supported by a particular machine
825 model, and disable all other extensions.
831 The following options are available when @value{AS} is configured for
832 a picoJava processor.
836 @cindex PJ endianness
837 @cindex endianness, PJ
838 @cindex big endian output, PJ
840 Generate ``big endian'' format output.
842 @cindex little endian output, PJ
844 Generate ``little endian'' format output.
850 The following options are available when @value{AS} is configured for the
851 Motorola 68HC11 or 68HC12 series.
855 @item -m68hc11 | -m68hc12 | -m68hcs12
856 Specify what processor is the target. The default is
857 defined by the configuration option when building the assembler.
860 Specify to use the 16-bit integer ABI.
863 Specify to use the 32-bit integer ABI.
866 Specify to use the 32-bit double ABI.
869 Specify to use the 64-bit double ABI.
871 @item --force-long-branchs
872 Relative branches are turned into absolute ones. This concerns
873 conditional branches, unconditional branches and branches to a
876 @item -S | --short-branchs
877 Do not turn relative branchs into absolute ones
878 when the offset is out of range.
880 @item --strict-direct-mode
881 Do not turn the direct addressing mode into extended addressing mode
882 when the instruction does not support direct addressing mode.
884 @item --print-insn-syntax
885 Print the syntax of instruction in case of error.
887 @item --print-opcodes
888 print the list of instructions with syntax and then exit.
890 @item --generate-example
891 print an example of instruction for each possible instruction and then exit.
892 This option is only useful for testing @command{@value{AS}}.
898 The following options are available when @command{@value{AS}} is configured
899 for the SPARC architecture:
902 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
903 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
904 Explicitly select a variant of the SPARC architecture.
906 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
907 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
909 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
910 UltraSPARC extensions.
912 @item -xarch=v8plus | -xarch=v8plusa
913 For compatibility with the Solaris v9 assembler. These options are
914 equivalent to -Av8plus and -Av8plusa, respectively.
917 Warn when the assembler switches to another architecture.
922 The following options are available when @value{AS} is configured for the 'c54x
927 Enable extended addressing mode. All addresses and relocations will assume
928 extended addressing (usually 23 bits).
929 @item -mcpu=@var{CPU_VERSION}
930 Sets the CPU version being compiled for.
931 @item -merrors-to-file @var{FILENAME}
932 Redirect error output to a file, for broken systems which don't support such
933 behaviour in the shell.
938 The following options are available when @value{AS} is configured for
939 a @sc{mips} processor.
943 This option sets the largest size of an object that can be referenced
944 implicitly with the @code{gp} register. It is only accepted for targets that
945 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
947 @cindex MIPS endianness
948 @cindex endianness, MIPS
949 @cindex big endian output, MIPS
951 Generate ``big endian'' format output.
953 @cindex little endian output, MIPS
955 Generate ``little endian'' format output.
967 Generate code for a particular @sc{mips} Instruction Set Architecture level.
968 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
969 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
970 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
971 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
973 correspond to generic
974 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
975 and @samp{MIPS64 Release 2}
976 ISA processors, respectively.
978 @item -march=@var{CPU}
979 Generate code for a particular @sc{mips} cpu.
981 @item -mtune=@var{cpu}
982 Schedule and tune for a particular @sc{mips} cpu.
986 Cause nops to be inserted if the read of the destination register
987 of an mfhi or mflo instruction occurs in the following two instructions.
991 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
992 section instead of the standard ELF .stabs sections.
996 Control generation of @code{.pdr} sections.
1000 The register sizes are normally inferred from the ISA and ABI, but these
1001 flags force a certain group of registers to be treated as 32 bits wide at
1002 all times. @samp{-mgp32} controls the size of general-purpose registers
1003 and @samp{-mfp32} controls the size of floating-point registers.
1007 Generate code for the MIPS 16 processor. This is equivalent to putting
1008 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1009 turns off this option.
1012 @itemx -mno-smartmips
1013 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1014 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1015 @samp{-mno-smartmips} turns off this option.
1019 Generate code for the MIPS-3D Application Specific Extension.
1020 This tells the assembler to accept MIPS-3D instructions.
1021 @samp{-no-mips3d} turns off this option.
1025 Generate code for the MDMX Application Specific Extension.
1026 This tells the assembler to accept MDMX instructions.
1027 @samp{-no-mdmx} turns off this option.
1031 Generate code for the DSP Application Specific Extension.
1032 This tells the assembler to accept DSP instructions.
1033 @samp{-mno-dsp} turns off this option.
1037 Generate code for the MT Application Specific Extension.
1038 This tells the assembler to accept MT instructions.
1039 @samp{-mno-mt} turns off this option.
1041 @item --construct-floats
1042 @itemx --no-construct-floats
1043 The @samp{--no-construct-floats} option disables the construction of
1044 double width floating point constants by loading the two halves of the
1045 value into the two single width floating point registers that make up
1046 the double width register. By default @samp{--construct-floats} is
1047 selected, allowing construction of these floating point constants.
1050 @item --emulation=@var{name}
1051 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1052 for some other target, in all respects, including output format (choosing
1053 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1054 debugging information or store symbol table information, and default
1055 endianness. The available configuration names are: @samp{mipsecoff},
1056 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1057 @samp{mipsbelf}. The first two do not alter the default endianness from that
1058 of the primary target for which the assembler was configured; the others change
1059 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1060 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1061 selection in any case.
1063 This option is currently supported only when the primary target
1064 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1065 Furthermore, the primary target or others specified with
1066 @samp{--enable-targets=@dots{}} at configuration time must include support for
1067 the other format, if both are to be available. For example, the Irix 5
1068 configuration includes support for both.
1070 Eventually, this option will support more configurations, with more
1071 fine-grained control over the assembler's behavior, and will be supported for
1075 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1082 Control how to deal with multiplication overflow and division by zero.
1083 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1084 (and only work for Instruction Set Architecture level 2 and higher);
1085 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1089 When this option is used, @command{@value{AS}} will issue a warning every
1090 time it generates a nop instruction from a macro.
1095 The following options are available when @value{AS} is configured for
1101 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1102 The command line option @samp{-nojsri2bsr} can be used to disable it.
1106 Enable or disable the silicon filter behaviour. By default this is disabled.
1107 The default can be overridden by the @samp{-sifilter} command line option.
1110 Alter jump instructions for long displacements.
1112 @item -mcpu=[210|340]
1113 Select the cpu type on the target hardware. This controls which instructions
1117 Assemble for a big endian target.
1120 Assemble for a little endian target.
1126 See the info pages for documentation of the MMIX-specific options.
1130 The following options are available when @value{AS} is configured for
1131 an Xtensa processor.
1134 @item --text-section-literals | --no-text-section-literals
1135 With @option{--text-@-section-@-literals}, literal pools are interspersed
1136 in the text section. The default is
1137 @option{--no-@-text-@-section-@-literals}, which places literals in a
1138 separate section in the output file. These options only affect literals
1139 referenced via PC-relative @code{L32R} instructions; literals for
1140 absolute mode @code{L32R} instructions are handled separately.
1142 @item --absolute-literals | --no-absolute-literals
1143 Indicate to the assembler whether @code{L32R} instructions use absolute
1144 or PC-relative addressing. The default is to assume absolute addressing
1145 if the Xtensa processor includes the absolute @code{L32R} addressing
1146 option. Otherwise, only the PC-relative @code{L32R} mode can be used.
1148 @item --target-align | --no-target-align
1149 Enable or disable automatic alignment to reduce branch penalties at the
1150 expense of some code density. The default is @option{--target-@-align}.
1152 @item --longcalls | --no-longcalls
1153 Enable or disable transformation of call instructions to allow calls
1154 across a greater range of addresses. The default is
1155 @option{--no-@-longcalls}.
1157 @item --transform | --no-transform
1158 Enable or disable all assembler transformations of Xtensa instructions.
1159 The default is @option{--transform};
1160 @option{--no-transform} should be used only in the rare cases when the
1161 instructions must be exactly as specified in the assembly source.
1166 The following options are available when @value{AS} is configured for
1167 a Z80 family processor.
1170 Assemble for Z80 processor.
1172 Assemble for R800 processor.
1173 @item -ignore-undocumented-instructions
1175 Assemble undocumented Z80 instructions that also work on R800 without warning.
1176 @item -ignore-unportable-instructions
1178 Assemble all undocumented Z80 instructions without warning.
1179 @item -warn-undocumented-instructions
1181 Issue a warning for undocumented Z80 instructions that also work on R800.
1182 @item -warn-unportable-instructions
1184 Issue a warning for undocumented Z80 instructions that do notwork on R800.
1185 @item -forbid-undocumented-instructions
1187 Treat all undocumented instructions as errors.
1188 @item -forbid-unportable-instructions
1190 Treat undocumented Z80 intructions that do notwork on R800 as errors.
1197 * Manual:: Structure of this Manual
1198 * GNU Assembler:: The GNU Assembler
1199 * Object Formats:: Object File Formats
1200 * Command Line:: Command Line
1201 * Input Files:: Input Files
1202 * Object:: Output (Object) File
1203 * Errors:: Error and Warning Messages
1207 @section Structure of this Manual
1209 @cindex manual, structure and purpose
1210 This manual is intended to describe what you need to know to use
1211 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1212 notation for symbols, constants, and expressions; the directives that
1213 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1216 We also cover special features in the @value{TARGET}
1217 configuration of @command{@value{AS}}, including assembler directives.
1220 This manual also describes some of the machine-dependent features of
1221 various flavors of the assembler.
1224 @cindex machine instructions (not covered)
1225 On the other hand, this manual is @emph{not} intended as an introduction
1226 to programming in assembly language---let alone programming in general!
1227 In a similar vein, we make no attempt to introduce the machine
1228 architecture; we do @emph{not} describe the instruction set, standard
1229 mnemonics, registers or addressing modes that are standard to a
1230 particular architecture.
1232 You may want to consult the manufacturer's
1233 machine architecture manual for this information.
1237 For information on the H8/300 machine instruction set, see @cite{H8/300
1238 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1239 Programming Manual} (Renesas).
1242 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1243 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1244 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1245 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1248 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1252 @c I think this is premature---doc@cygnus.com, 17jan1991
1254 Throughout this manual, we assume that you are running @dfn{GNU},
1255 the portable operating system from the @dfn{Free Software
1256 Foundation, Inc.}. This restricts our attention to certain kinds of
1257 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1258 once this assumption is granted examples and definitions need less
1261 @command{@value{AS}} is part of a team of programs that turn a high-level
1262 human-readable series of instructions into a low-level
1263 computer-readable series of instructions. Different versions of
1264 @command{@value{AS}} are used for different kinds of computer.
1267 @c There used to be a section "Terminology" here, which defined
1268 @c "contents", "byte", "word", and "long". Defining "word" to any
1269 @c particular size is confusing when the .word directive may generate 16
1270 @c bits on one machine and 32 bits on another; in general, for the user
1271 @c version of this manual, none of these terms seem essential to define.
1272 @c They were used very little even in the former draft of the manual;
1273 @c this draft makes an effort to avoid them (except in names of
1277 @section The GNU Assembler
1279 @c man begin DESCRIPTION
1281 @sc{gnu} @command{as} is really a family of assemblers.
1283 This manual describes @command{@value{AS}}, a member of that family which is
1284 configured for the @value{TARGET} architectures.
1286 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1287 should find a fairly similar environment when you use it on another
1288 architecture. Each version has much in common with the others,
1289 including object file formats, most assembler directives (often called
1290 @dfn{pseudo-ops}) and assembler syntax.@refill
1292 @cindex purpose of @sc{gnu} assembler
1293 @command{@value{AS}} is primarily intended to assemble the output of the
1294 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1295 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1296 assemble correctly everything that other assemblers for the same
1297 machine would assemble.
1299 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1302 @c This remark should appear in generic version of manual; assumption
1303 @c here is that generic version sets M680x0.
1304 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1305 assembler for the same architecture; for example, we know of several
1306 incompatible versions of 680x0 assembly language syntax.
1311 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1312 program in one pass of the source file. This has a subtle impact on the
1313 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1315 @node Object Formats
1316 @section Object File Formats
1318 @cindex object file format
1319 The @sc{gnu} assembler can be configured to produce several alternative
1320 object file formats. For the most part, this does not affect how you
1321 write assembly language programs; but directives for debugging symbols
1322 are typically different in different file formats. @xref{Symbol
1323 Attributes,,Symbol Attributes}.
1326 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1327 @value{OBJ-NAME} format object files.
1329 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1331 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1332 @code{b.out} or COFF format object files.
1335 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1336 SOM or ELF format object files.
1341 @section Command Line
1343 @cindex command line conventions
1345 After the program name @command{@value{AS}}, the command line may contain
1346 options and file names. Options may appear in any order, and may be
1347 before, after, or between file names. The order of file names is
1350 @cindex standard input, as input file
1352 @file{--} (two hyphens) by itself names the standard input file
1353 explicitly, as one of the files for @command{@value{AS}} to assemble.
1355 @cindex options, command line
1356 Except for @samp{--} any command line argument that begins with a
1357 hyphen (@samp{-}) is an option. Each option changes the behavior of
1358 @command{@value{AS}}. No option changes the way another option works. An
1359 option is a @samp{-} followed by one or more letters; the case of
1360 the letter is important. All options are optional.
1362 Some options expect exactly one file name to follow them. The file
1363 name may either immediately follow the option's letter (compatible
1364 with older assemblers) or it may be the next command argument (@sc{gnu}
1365 standard). These two command lines are equivalent:
1368 @value{AS} -o my-object-file.o mumble.s
1369 @value{AS} -omy-object-file.o mumble.s
1373 @section Input Files
1376 @cindex source program
1377 @cindex files, input
1378 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1379 describe the program input to one run of @command{@value{AS}}. The program may
1380 be in one or more files; how the source is partitioned into files
1381 doesn't change the meaning of the source.
1383 @c I added "con" prefix to "catenation" just to prove I can overcome my
1384 @c APL training... doc@cygnus.com
1385 The source program is a concatenation of the text in all the files, in the
1388 @c man begin DESCRIPTION
1389 Each time you run @command{@value{AS}} it assembles exactly one source
1390 program. The source program is made up of one or more files.
1391 (The standard input is also a file.)
1393 You give @command{@value{AS}} a command line that has zero or more input file
1394 names. The input files are read (from left file name to right). A
1395 command line argument (in any position) that has no special meaning
1396 is taken to be an input file name.
1398 If you give @command{@value{AS}} no file names it attempts to read one input file
1399 from the @command{@value{AS}} standard input, which is normally your terminal. You
1400 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1403 Use @samp{--} if you need to explicitly name the standard input file
1404 in your command line.
1406 If the source is empty, @command{@value{AS}} produces a small, empty object
1411 @subheading Filenames and Line-numbers
1413 @cindex input file linenumbers
1414 @cindex line numbers, in input files
1415 There are two ways of locating a line in the input file (or files) and
1416 either may be used in reporting error messages. One way refers to a line
1417 number in a physical file; the other refers to a line number in a
1418 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1420 @dfn{Physical files} are those files named in the command line given
1421 to @command{@value{AS}}.
1423 @dfn{Logical files} are simply names declared explicitly by assembler
1424 directives; they bear no relation to physical files. Logical file names help
1425 error messages reflect the original source file, when @command{@value{AS}} source
1426 is itself synthesized from other files. @command{@value{AS}} understands the
1427 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1428 @ref{File,,@code{.file}}.
1431 @section Output (Object) File
1437 Every time you run @command{@value{AS}} it produces an output file, which is
1438 your assembly language program translated into numbers. This file
1439 is the object file. Its default name is
1447 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1449 You can give it another name by using the @option{-o} option. Conventionally,
1450 object file names end with @file{.o}. The default name is used for historical
1451 reasons: older assemblers were capable of assembling self-contained programs
1452 directly into a runnable program. (For some formats, this isn't currently
1453 possible, but it can be done for the @code{a.out} format.)
1457 The object file is meant for input to the linker @code{@value{LD}}. It contains
1458 assembled program code, information to help @code{@value{LD}} integrate
1459 the assembled program into a runnable file, and (optionally) symbolic
1460 information for the debugger.
1462 @c link above to some info file(s) like the description of a.out.
1463 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1466 @section Error and Warning Messages
1468 @c man begin DESCRIPTION
1470 @cindex error messages
1471 @cindex warning messages
1472 @cindex messages from assembler
1473 @command{@value{AS}} may write warnings and error messages to the standard error
1474 file (usually your terminal). This should not happen when a compiler
1475 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1476 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1477 grave problem that stops the assembly.
1481 @cindex format of warning messages
1482 Warning messages have the format
1485 file_name:@b{NNN}:Warning Message Text
1489 @cindex line numbers, in warnings/errors
1490 (where @b{NNN} is a line number). If a logical file name has been given
1491 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1492 the current input file is used. If a logical line number was given
1494 (@pxref{Line,,@code{.line}})
1496 then it is used to calculate the number printed,
1497 otherwise the actual line in the current source file is printed. The
1498 message text is intended to be self explanatory (in the grand Unix
1501 @cindex format of error messages
1502 Error messages have the format
1504 file_name:@b{NNN}:FATAL:Error Message Text
1506 The file name and line number are derived as for warning
1507 messages. The actual message text may be rather less explanatory
1508 because many of them aren't supposed to happen.
1511 @chapter Command-Line Options
1513 @cindex options, all versions of assembler
1514 This chapter describes command-line options available in @emph{all}
1515 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1517 to the @value{TARGET} target.
1520 to particular machine architectures.
1523 @c man begin DESCRIPTION
1525 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1526 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1527 The assembler arguments must be separated from each other (and the @samp{-Wa})
1528 by commas. For example:
1531 gcc -c -g -O -Wa,-alh,-L file.c
1535 This passes two options to the assembler: @samp{-alh} (emit a listing to
1536 standard output with high-level and assembly source) and @samp{-L} (retain
1537 local symbols in the symbol table).
1539 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1540 command-line options are automatically passed to the assembler by the compiler.
1541 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1542 precisely what options it passes to each compilation pass, including the
1548 * a:: -a[cdhlns] enable listings
1549 * alternate:: --alternate enable alternate macro syntax
1550 * D:: -D for compatibility
1551 * f:: -f to work faster
1552 * I:: -I for .include search path
1553 @ifclear DIFF-TBL-KLUGE
1554 * K:: -K for compatibility
1556 @ifset DIFF-TBL-KLUGE
1557 * K:: -K for difference tables
1560 * L:: -L to retain local labels
1561 * listing:: --listing-XXX to configure listing output
1562 * M:: -M or --mri to assemble in MRI compatibility mode
1563 * MD:: --MD for dependency tracking
1564 * o:: -o to name the object file
1565 * R:: -R to join data and text sections
1566 * statistics:: --statistics to see statistics about assembly
1567 * traditional-format:: --traditional-format for compatible output
1568 * v:: -v to announce version
1569 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1570 * Z:: -Z to make object file even after errors
1574 @section Enable Listings: @option{-a[cdhlns]}
1583 @cindex listings, enabling
1584 @cindex assembly listings, enabling
1586 These options enable listing output from the assembler. By itself,
1587 @samp{-a} requests high-level, assembly, and symbols listing.
1588 You can use other letters to select specific options for the list:
1589 @samp{-ah} requests a high-level language listing,
1590 @samp{-al} requests an output-program assembly listing, and
1591 @samp{-as} requests a symbol table listing.
1592 High-level listings require that a compiler debugging option like
1593 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1596 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1597 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1598 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1599 omitted from the listing.
1601 Use the @samp{-ad} option to omit debugging directives from the
1604 Once you have specified one of these options, you can further control
1605 listing output and its appearance using the directives @code{.list},
1606 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1608 The @samp{-an} option turns off all forms processing.
1609 If you do not request listing output with one of the @samp{-a} options, the
1610 listing-control directives have no effect.
1612 The letters after @samp{-a} may be combined into one option,
1613 @emph{e.g.}, @samp{-aln}.
1615 Note if the assembler source is coming from the standard input (eg because it
1616 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1617 is being used) then the listing will not contain any comments or preprocessor
1618 directives. This is because the listing code buffers input source lines from
1619 stdin only after they have been preprocessed by the assembler. This reduces
1620 memory usage and makes the code more efficient.
1623 @section @option{--alternate}
1626 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1629 @section @option{-D}
1632 This option has no effect whatsoever, but it is accepted to make it more
1633 likely that scripts written for other assemblers also work with
1634 @command{@value{AS}}.
1637 @section Work Faster: @option{-f}
1640 @cindex trusted compiler
1641 @cindex faster processing (@option{-f})
1642 @samp{-f} should only be used when assembling programs written by a
1643 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1644 and comment preprocessing on
1645 the input file(s) before assembling them. @xref{Preprocessing,
1649 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1650 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1655 @section @code{.include} Search Path: @option{-I} @var{path}
1657 @kindex -I @var{path}
1658 @cindex paths for @code{.include}
1659 @cindex search path for @code{.include}
1660 @cindex @code{include} directive search path
1661 Use this option to add a @var{path} to the list of directories
1662 @command{@value{AS}} searches for files specified in @code{.include}
1663 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1664 many times as necessary to include a variety of paths. The current
1665 working directory is always searched first; after that, @command{@value{AS}}
1666 searches any @samp{-I} directories in the same order as they were
1667 specified (left to right) on the command line.
1670 @section Difference Tables: @option{-K}
1673 @ifclear DIFF-TBL-KLUGE
1674 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1675 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1676 where it can be used to warn when the assembler alters the machine code
1677 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1678 family does not have the addressing limitations that sometimes lead to this
1679 alteration on other platforms.
1682 @ifset DIFF-TBL-KLUGE
1683 @cindex difference tables, warning
1684 @cindex warning for altered difference tables
1685 @command{@value{AS}} sometimes alters the code emitted for directives of the form
1686 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1687 You can use the @samp{-K} option if you want a warning issued when this
1692 @section Include Local Labels: @option{-L}
1695 @cindex local labels, retaining in output
1696 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1697 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1698 debugging, because they are intended for the use of programs (like
1699 compilers) that compose assembler programs, not for your notice.
1700 Normally both @command{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1701 normally debug with them.
1703 This option tells @command{@value{AS}} to retain those @samp{L@dots{}} symbols
1704 in the object file. Usually if you do this you also tell the linker
1705 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1707 By default, a local label is any label beginning with @samp{L}, but each
1708 target is allowed to redefine the local label prefix.
1710 On the HPPA local labels begin with @samp{L$}.
1714 @section Configuring listing output: @option{--listing}
1716 The listing feature of the assembler can be enabled via the command line switch
1717 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1718 hex dump of the corresponding locations in the output object file, and displays
1719 them as a listing file. The format of this listing can be controlled by pseudo
1720 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1721 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1724 @item --listing-lhs-width=@samp{number}
1725 @kindex --listing-lhs-width
1726 @cindex Width of first line disassembly output
1727 Sets the maximum width, in words, of the first line of the hex byte dump. This
1728 dump appears on the left hand side of the listing output.
1730 @item --listing-lhs-width2=@samp{number}
1731 @kindex --listing-lhs-width2
1732 @cindex Width of continuation lines of disassembly output
1733 Sets the maximum width, in words, of any further lines of the hex byte dump for
1734 a given input source line. If this value is not specified, it defaults to being
1735 the same as the value specified for @samp{--listing-lhs-width}. If neither
1736 switch is used the default is to one.
1738 @item --listing-rhs-width=@samp{number}
1739 @kindex --listing-rhs-width
1740 @cindex Width of source line output
1741 Sets the maximum width, in characters, of the source line that is displayed
1742 alongside the hex dump. The default value for this parameter is 100. The
1743 source line is displayed on the right hand side of the listing output.
1745 @item --listing-cont-lines=@samp{number}
1746 @kindex --listing-cont-lines
1747 @cindex Maximum number of continuation lines
1748 Sets the maximum number of continuation lines of hex dump that will be
1749 displayed for a given single line of source input. The default value is 4.
1753 @section Assemble in MRI Compatibility Mode: @option{-M}
1756 @cindex MRI compatibility mode
1757 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1758 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1759 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1760 configured target) assembler from Microtec Research. The exact nature of the
1761 MRI syntax will not be documented here; see the MRI manuals for more
1762 information. Note in particular that the handling of macros and macro
1763 arguments is somewhat different. The purpose of this option is to permit
1764 assembling existing MRI assembler code using @command{@value{AS}}.
1766 The MRI compatibility is not complete. Certain operations of the MRI assembler
1767 depend upon its object file format, and can not be supported using other object
1768 file formats. Supporting these would require enhancing each object file format
1769 individually. These are:
1772 @item global symbols in common section
1774 The m68k MRI assembler supports common sections which are merged by the linker.
1775 Other object file formats do not support this. @command{@value{AS}} handles
1776 common sections by treating them as a single common symbol. It permits local
1777 symbols to be defined within a common section, but it can not support global
1778 symbols, since it has no way to describe them.
1780 @item complex relocations
1782 The MRI assemblers support relocations against a negated section address, and
1783 relocations which combine the start addresses of two or more sections. These
1784 are not support by other object file formats.
1786 @item @code{END} pseudo-op specifying start address
1788 The MRI @code{END} pseudo-op permits the specification of a start address.
1789 This is not supported by other object file formats. The start address may
1790 instead be specified using the @option{-e} option to the linker, or in a linker
1793 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1795 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1796 name to the output file. This is not supported by other object file formats.
1798 @item @code{ORG} pseudo-op
1800 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1801 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1802 which changes the location within the current section. Absolute sections are
1803 not supported by other object file formats. The address of a section may be
1804 assigned within a linker script.
1807 There are some other features of the MRI assembler which are not supported by
1808 @command{@value{AS}}, typically either because they are difficult or because they
1809 seem of little consequence. Some of these may be supported in future releases.
1813 @item EBCDIC strings
1815 EBCDIC strings are not supported.
1817 @item packed binary coded decimal
1819 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1820 and @code{DCB.P} pseudo-ops are not supported.
1822 @item @code{FEQU} pseudo-op
1824 The m68k @code{FEQU} pseudo-op is not supported.
1826 @item @code{NOOBJ} pseudo-op
1828 The m68k @code{NOOBJ} pseudo-op is not supported.
1830 @item @code{OPT} branch control options
1832 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1833 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1834 relaxes all branches, whether forward or backward, to an appropriate size, so
1835 these options serve no purpose.
1837 @item @code{OPT} list control options
1839 The following m68k @code{OPT} list control options are ignored: @code{C},
1840 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1841 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1843 @item other @code{OPT} options
1845 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1846 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1848 @item @code{OPT} @code{D} option is default
1850 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1851 @code{OPT NOD} may be used to turn it off.
1853 @item @code{XREF} pseudo-op.
1855 The m68k @code{XREF} pseudo-op is ignored.
1857 @item @code{.debug} pseudo-op
1859 The i960 @code{.debug} pseudo-op is not supported.
1861 @item @code{.extended} pseudo-op
1863 The i960 @code{.extended} pseudo-op is not supported.
1865 @item @code{.list} pseudo-op.
1867 The various options of the i960 @code{.list} pseudo-op are not supported.
1869 @item @code{.optimize} pseudo-op
1871 The i960 @code{.optimize} pseudo-op is not supported.
1873 @item @code{.output} pseudo-op
1875 The i960 @code{.output} pseudo-op is not supported.
1877 @item @code{.setreal} pseudo-op
1879 The i960 @code{.setreal} pseudo-op is not supported.
1884 @section Dependency Tracking: @option{--MD}
1887 @cindex dependency tracking
1890 @command{@value{AS}} can generate a dependency file for the file it creates. This
1891 file consists of a single rule suitable for @code{make} describing the
1892 dependencies of the main source file.
1894 The rule is written to the file named in its argument.
1896 This feature is used in the automatic updating of makefiles.
1899 @section Name the Object File: @option{-o}
1902 @cindex naming object file
1903 @cindex object file name
1904 There is always one object file output when you run @command{@value{AS}}. By
1905 default it has the name
1908 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1922 You use this option (which takes exactly one filename) to give the
1923 object file a different name.
1925 Whatever the object file is called, @command{@value{AS}} overwrites any
1926 existing file of the same name.
1929 @section Join Data and Text Sections: @option{-R}
1932 @cindex data and text sections, joining
1933 @cindex text and data sections, joining
1934 @cindex joining text and data sections
1935 @cindex merging text and data sections
1936 @option{-R} tells @command{@value{AS}} to write the object file as if all
1937 data-section data lives in the text section. This is only done at
1938 the very last moment: your binary data are the same, but data
1939 section parts are relocated differently. The data section part of
1940 your object file is zero bytes long because all its bytes are
1941 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1943 When you specify @option{-R} it would be possible to generate shorter
1944 address displacements (because we do not have to cross between text and
1945 data section). We refrain from doing this simply for compatibility with
1946 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1949 When @command{@value{AS}} is configured for COFF or ELF output,
1950 this option is only useful if you use sections named @samp{.text} and
1955 @option{-R} is not supported for any of the HPPA targets. Using
1956 @option{-R} generates a warning from @command{@value{AS}}.
1960 @section Display Assembly Statistics: @option{--statistics}
1962 @kindex --statistics
1963 @cindex statistics, about assembly
1964 @cindex time, total for assembly
1965 @cindex space used, maximum for assembly
1966 Use @samp{--statistics} to display two statistics about the resources used by
1967 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1968 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1971 @node traditional-format
1972 @section Compatible Output: @option{--traditional-format}
1974 @kindex --traditional-format
1975 For some targets, the output of @command{@value{AS}} is different in some ways
1976 from the output of some existing assembler. This switch requests
1977 @command{@value{AS}} to use the traditional format instead.
1979 For example, it disables the exception frame optimizations which
1980 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
1983 @section Announce Version: @option{-v}
1987 @cindex assembler version
1988 @cindex version of assembler
1989 You can find out what version of as is running by including the
1990 option @samp{-v} (which you can also spell as @samp{-version}) on the
1994 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
1996 @command{@value{AS}} should never give a warning or error message when
1997 assembling compiler output. But programs written by people often
1998 cause @command{@value{AS}} to give a warning that a particular assumption was
1999 made. All such warnings are directed to the standard error file.
2003 @cindex suppressing warnings
2004 @cindex warnings, suppressing
2005 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2006 This only affects the warning messages: it does not change any particular of
2007 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2010 @kindex --fatal-warnings
2011 @cindex errors, caused by warnings
2012 @cindex warnings, causing error
2013 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2014 files that generate warnings to be in error.
2017 @cindex warnings, switching on
2018 You can switch these options off again by specifying @option{--warn}, which
2019 causes warnings to be output as usual.
2022 @section Generate Object File in Spite of Errors: @option{-Z}
2023 @cindex object file, after errors
2024 @cindex errors, continuing after
2025 After an error message, @command{@value{AS}} normally produces no output. If for
2026 some reason you are interested in object file output even after
2027 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2028 option. If there are any errors, @command{@value{AS}} continues anyways, and
2029 writes an object file after a final warning message of the form @samp{@var{n}
2030 errors, @var{m} warnings, generating bad object file.}
2035 @cindex machine-independent syntax
2036 @cindex syntax, machine-independent
2037 This chapter describes the machine-independent syntax allowed in a
2038 source file. @command{@value{AS}} syntax is similar to what many other
2039 assemblers use; it is inspired by the BSD 4.2
2044 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2048 * Preprocessing:: Preprocessing
2049 * Whitespace:: Whitespace
2050 * Comments:: Comments
2051 * Symbol Intro:: Symbols
2052 * Statements:: Statements
2053 * Constants:: Constants
2057 @section Preprocessing
2059 @cindex preprocessing
2060 The @command{@value{AS}} internal preprocessor:
2062 @cindex whitespace, removed by preprocessor
2064 adjusts and removes extra whitespace. It leaves one space or tab before
2065 the keywords on a line, and turns any other whitespace on the line into
2068 @cindex comments, removed by preprocessor
2070 removes all comments, replacing them with a single space, or an
2071 appropriate number of newlines.
2073 @cindex constants, converted by preprocessor
2075 converts character constants into the appropriate numeric values.
2078 It does not do macro processing, include file handling, or
2079 anything else you may get from your C compiler's preprocessor. You can
2080 do include file processing with the @code{.include} directive
2081 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2082 to get other ``CPP'' style preprocessing by giving the input file a
2083 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
2084 Output, gcc.info, Using GNU CC}.
2086 Excess whitespace, comments, and character constants
2087 cannot be used in the portions of the input text that are not
2090 @cindex turning preprocessing on and off
2091 @cindex preprocessing, turning on and off
2094 If the first line of an input file is @code{#NO_APP} or if you use the
2095 @samp{-f} option, whitespace and comments are not removed from the input file.
2096 Within an input file, you can ask for whitespace and comment removal in
2097 specific portions of the by putting a line that says @code{#APP} before the
2098 text that may contain whitespace or comments, and putting a line that says
2099 @code{#NO_APP} after this text. This feature is mainly intend to support
2100 @code{asm} statements in compilers whose output is otherwise free of comments
2107 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2108 Whitespace is used to separate symbols, and to make programs neater for
2109 people to read. Unless within character constants
2110 (@pxref{Characters,,Character Constants}), any whitespace means the same
2111 as exactly one space.
2117 There are two ways of rendering comments to @command{@value{AS}}. In both
2118 cases the comment is equivalent to one space.
2120 Anything from @samp{/*} through the next @samp{*/} is a comment.
2121 This means you may not nest these comments.
2125 The only way to include a newline ('\n') in a comment
2126 is to use this sort of comment.
2129 /* This sort of comment does not nest. */
2132 @cindex line comment character
2133 Anything from the @dfn{line comment} character to the next newline
2134 is considered a comment and is ignored. The line comment character is
2136 @samp{;} on the ARC;
2139 @samp{@@} on the ARM;
2142 @samp{;} for the H8/300 family;
2145 @samp{;} for the HPPA;
2148 @samp{#} on the i386 and x86-64;
2151 @samp{#} on the i960;
2154 @samp{;} for the PDP-11;
2157 @samp{;} for picoJava;
2160 @samp{#} for Motorola PowerPC;
2163 @samp{!} for the Renesas / SuperH SH;
2166 @samp{!} on the SPARC;
2169 @samp{#} on the ip2k;
2172 @samp{#} on the m32c;
2175 @samp{#} on the m32r;
2178 @samp{|} on the 680x0;
2181 @samp{#} on the 68HC11 and 68HC12;
2184 @samp{#} on the Vax;
2187 @samp{;} for the Z80;
2190 @samp{!} for the Z8000;
2193 @samp{#} on the V850;
2196 @samp{#} for Xtensa systems;
2198 see @ref{Machine Dependencies}. @refill
2199 @c FIXME What about i860?
2202 On some machines there are two different line comment characters. One
2203 character only begins a comment if it is the first non-whitespace character on
2204 a line, while the other always begins a comment.
2208 The V850 assembler also supports a double dash as starting a comment that
2209 extends to the end of the line.
2215 @cindex lines starting with @code{#}
2216 @cindex logical line numbers
2217 To be compatible with past assemblers, lines that begin with @samp{#} have a
2218 special interpretation. Following the @samp{#} should be an absolute
2219 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2220 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
2221 new logical file name. The rest of the line, if any, should be whitespace.
2223 If the first non-whitespace characters on the line are not numeric,
2224 the line is ignored. (Just like a comment.)
2227 # This is an ordinary comment.
2228 # 42-6 "new_file_name" # New logical file name
2229 # This is logical line # 36.
2231 This feature is deprecated, and may disappear from future versions
2232 of @command{@value{AS}}.
2237 @cindex characters used in symbols
2238 @ifclear SPECIAL-SYMS
2239 A @dfn{symbol} is one or more characters chosen from the set of all
2240 letters (both upper and lower case), digits and the three characters
2246 A @dfn{symbol} is one or more characters chosen from the set of all
2247 letters (both upper and lower case), digits and the three characters
2248 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2254 On most machines, you can also use @code{$} in symbol names; exceptions
2255 are noted in @ref{Machine Dependencies}.
2257 No symbol may begin with a digit. Case is significant.
2258 There is no length limit: all characters are significant. Symbols are
2259 delimited by characters not in that set, or by the beginning of a file
2260 (since the source program must end with a newline, the end of a file is
2261 not a possible symbol delimiter). @xref{Symbols}.
2262 @cindex length of symbols
2267 @cindex statements, structure of
2268 @cindex line separator character
2269 @cindex statement separator character
2271 @ifclear abnormal-separator
2272 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2273 semicolon (@samp{;}). The newline or semicolon is considered part of
2274 the preceding statement. Newlines and semicolons within character
2275 constants are an exception: they do not end statements.
2277 @ifset abnormal-separator
2279 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2280 point (@samp{!}). The newline or exclamation point is considered part of the
2281 preceding statement. Newlines and exclamation points within character
2282 constants are an exception: they do not end statements.
2285 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2286 H8/300) a dollar sign (@samp{$}); or (for the Renesas-SH) a semicolon
2287 (@samp{;}). The newline or separator character is considered part of
2288 the preceding statement. Newlines and separators within character
2289 constants are an exception: they do not end statements.
2294 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2295 separator character. (The line separator is usually @samp{;}, unless
2296 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
2297 newline or separator character is considered part of the preceding
2298 statement. Newlines and separators within character constants are an
2299 exception: they do not end statements.
2302 @cindex newline, required at file end
2303 @cindex EOF, newline must precede
2304 It is an error to end any statement with end-of-file: the last
2305 character of any input file should be a newline.@refill
2307 An empty statement is allowed, and may include whitespace. It is ignored.
2309 @cindex instructions and directives
2310 @cindex directives and instructions
2311 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2312 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2314 A statement begins with zero or more labels, optionally followed by a
2315 key symbol which determines what kind of statement it is. The key
2316 symbol determines the syntax of the rest of the statement. If the
2317 symbol begins with a dot @samp{.} then the statement is an assembler
2318 directive: typically valid for any computer. If the symbol begins with
2319 a letter the statement is an assembly language @dfn{instruction}: it
2320 assembles into a machine language instruction.
2322 Different versions of @command{@value{AS}} for different computers
2323 recognize different instructions. In fact, the same symbol may
2324 represent a different instruction in a different computer's assembly
2328 @cindex @code{:} (label)
2329 @cindex label (@code{:})
2330 A label is a symbol immediately followed by a colon (@code{:}).
2331 Whitespace before a label or after a colon is permitted, but you may not
2332 have whitespace between a label's symbol and its colon. @xref{Labels}.
2335 For HPPA targets, labels need not be immediately followed by a colon, but
2336 the definition of a label must begin in column zero. This also implies that
2337 only one label may be defined on each line.
2341 label: .directive followed by something
2342 another_label: # This is an empty statement.
2343 instruction operand_1, operand_2, @dots{}
2350 A constant is a number, written so that its value is known by
2351 inspection, without knowing any context. Like this:
2354 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2355 .ascii "Ring the bell\7" # A string constant.
2356 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2357 .float 0f-314159265358979323846264338327\
2358 95028841971.693993751E-40 # - pi, a flonum.
2363 * Characters:: Character Constants
2364 * Numbers:: Number Constants
2368 @subsection Character Constants
2370 @cindex character constants
2371 @cindex constants, character
2372 There are two kinds of character constants. A @dfn{character} stands
2373 for one character in one byte and its value may be used in
2374 numeric expressions. String constants (properly called string
2375 @emph{literals}) are potentially many bytes and their values may not be
2376 used in arithmetic expressions.
2380 * Chars:: Characters
2384 @subsubsection Strings
2386 @cindex string constants
2387 @cindex constants, string
2388 A @dfn{string} is written between double-quotes. It may contain
2389 double-quotes or null characters. The way to get special characters
2390 into a string is to @dfn{escape} these characters: precede them with
2391 a backslash @samp{\} character. For example @samp{\\} represents
2392 one backslash: the first @code{\} is an escape which tells
2393 @command{@value{AS}} to interpret the second character literally as a backslash
2394 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2395 escape character). The complete list of escapes follows.
2397 @cindex escape codes, character
2398 @cindex character escape codes
2401 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2403 @cindex @code{\b} (backspace character)
2404 @cindex backspace (@code{\b})
2406 Mnemonic for backspace; for ASCII this is octal code 010.
2409 @c Mnemonic for EOText; for ASCII this is octal code 004.
2411 @cindex @code{\f} (formfeed character)
2412 @cindex formfeed (@code{\f})
2414 Mnemonic for FormFeed; for ASCII this is octal code 014.
2416 @cindex @code{\n} (newline character)
2417 @cindex newline (@code{\n})
2419 Mnemonic for newline; for ASCII this is octal code 012.
2422 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2424 @cindex @code{\r} (carriage return character)
2425 @cindex carriage return (@code{\r})
2427 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2430 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2431 @c other assemblers.
2433 @cindex @code{\t} (tab)
2434 @cindex tab (@code{\t})
2436 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2439 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2440 @c @item \x @var{digit} @var{digit} @var{digit}
2441 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2443 @cindex @code{\@var{ddd}} (octal character code)
2444 @cindex octal character code (@code{\@var{ddd}})
2445 @item \ @var{digit} @var{digit} @var{digit}
2446 An octal character code. The numeric code is 3 octal digits.
2447 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2448 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2450 @cindex @code{\@var{xd...}} (hex character code)
2451 @cindex hex character code (@code{\@var{xd...}})
2452 @item \@code{x} @var{hex-digits...}
2453 A hex character code. All trailing hex digits are combined. Either upper or
2454 lower case @code{x} works.
2456 @cindex @code{\\} (@samp{\} character)
2457 @cindex backslash (@code{\\})
2459 Represents one @samp{\} character.
2462 @c Represents one @samp{'} (accent acute) character.
2463 @c This is needed in single character literals
2464 @c (@xref{Characters,,Character Constants}.) to represent
2467 @cindex @code{\"} (doublequote character)
2468 @cindex doublequote (@code{\"})
2470 Represents one @samp{"} character. Needed in strings to represent
2471 this character, because an unescaped @samp{"} would end the string.
2473 @item \ @var{anything-else}
2474 Any other character when escaped by @kbd{\} gives a warning, but
2475 assembles as if the @samp{\} was not present. The idea is that if
2476 you used an escape sequence you clearly didn't want the literal
2477 interpretation of the following character. However @command{@value{AS}} has no
2478 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2479 code and warns you of the fact.
2482 Which characters are escapable, and what those escapes represent,
2483 varies widely among assemblers. The current set is what we think
2484 the BSD 4.2 assembler recognizes, and is a subset of what most C
2485 compilers recognize. If you are in doubt, do not use an escape
2489 @subsubsection Characters
2491 @cindex single character constant
2492 @cindex character, single
2493 @cindex constant, single character
2494 A single character may be written as a single quote immediately
2495 followed by that character. The same escapes apply to characters as
2496 to strings. So if you want to write the character backslash, you
2497 must write @kbd{'\\} where the first @code{\} escapes the second
2498 @code{\}. As you can see, the quote is an acute accent, not a
2499 grave accent. A newline
2501 @ifclear abnormal-separator
2502 (or semicolon @samp{;})
2504 @ifset abnormal-separator
2506 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2511 immediately following an acute accent is taken as a literal character
2512 and does not count as the end of a statement. The value of a character
2513 constant in a numeric expression is the machine's byte-wide code for
2514 that character. @command{@value{AS}} assumes your character code is ASCII:
2515 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2518 @subsection Number Constants
2520 @cindex constants, number
2521 @cindex number constants
2522 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2523 are stored in the target machine. @emph{Integers} are numbers that
2524 would fit into an @code{int} in the C language. @emph{Bignums} are
2525 integers, but they are stored in more than 32 bits. @emph{Flonums}
2526 are floating point numbers, described below.
2529 * Integers:: Integers
2534 * Bit Fields:: Bit Fields
2540 @subsubsection Integers
2542 @cindex constants, integer
2544 @cindex binary integers
2545 @cindex integers, binary
2546 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2547 the binary digits @samp{01}.
2549 @cindex octal integers
2550 @cindex integers, octal
2551 An octal integer is @samp{0} followed by zero or more of the octal
2552 digits (@samp{01234567}).
2554 @cindex decimal integers
2555 @cindex integers, decimal
2556 A decimal integer starts with a non-zero digit followed by zero or
2557 more digits (@samp{0123456789}).
2559 @cindex hexadecimal integers
2560 @cindex integers, hexadecimal
2561 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2562 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2564 Integers have the usual values. To denote a negative integer, use
2565 the prefix operator @samp{-} discussed under expressions
2566 (@pxref{Prefix Ops,,Prefix Operators}).
2569 @subsubsection Bignums
2572 @cindex constants, bignum
2573 A @dfn{bignum} has the same syntax and semantics as an integer
2574 except that the number (or its negative) takes more than 32 bits to
2575 represent in binary. The distinction is made because in some places
2576 integers are permitted while bignums are not.
2579 @subsubsection Flonums
2581 @cindex floating point numbers
2582 @cindex constants, floating point
2584 @cindex precision, floating point
2585 A @dfn{flonum} represents a floating point number. The translation is
2586 indirect: a decimal floating point number from the text is converted by
2587 @command{@value{AS}} to a generic binary floating point number of more than
2588 sufficient precision. This generic floating point number is converted
2589 to a particular computer's floating point format (or formats) by a
2590 portion of @command{@value{AS}} specialized to that computer.
2592 A flonum is written by writing (in order)
2597 (@samp{0} is optional on the HPPA.)
2601 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2603 @kbd{e} is recommended. Case is not important.
2605 @c FIXME: verify if flonum syntax really this vague for most cases
2606 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2607 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2610 On the H8/300, Renesas / SuperH SH,
2611 and AMD 29K architectures, the letter must be
2612 one of the letters @samp{DFPRSX} (in upper or lower case).
2614 On the ARC, the letter must be one of the letters @samp{DFRS}
2615 (in upper or lower case).
2617 On the Intel 960 architecture, the letter must be
2618 one of the letters @samp{DFT} (in upper or lower case).
2620 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2624 One of the letters @samp{DFRS} (in upper or lower case).
2627 One of the letters @samp{DFPRSX} (in upper or lower case).
2630 The letter @samp{E} (upper case only).
2633 One of the letters @samp{DFT} (in upper or lower case).
2638 An optional sign: either @samp{+} or @samp{-}.
2641 An optional @dfn{integer part}: zero or more decimal digits.
2644 An optional @dfn{fractional part}: @samp{.} followed by zero
2645 or more decimal digits.
2648 An optional exponent, consisting of:
2652 An @samp{E} or @samp{e}.
2653 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2654 @c principle this can perfectly well be different on different targets.
2656 Optional sign: either @samp{+} or @samp{-}.
2658 One or more decimal digits.
2663 At least one of the integer part or the fractional part must be
2664 present. The floating point number has the usual base-10 value.
2666 @command{@value{AS}} does all processing using integers. Flonums are computed
2667 independently of any floating point hardware in the computer running
2668 @command{@value{AS}}.
2672 @c Bit fields are written as a general facility but are also controlled
2673 @c by a conditional-compilation flag---which is as of now (21mar91)
2674 @c turned on only by the i960 config of GAS.
2676 @subsubsection Bit Fields
2679 @cindex constants, bit field
2680 You can also define numeric constants as @dfn{bit fields}.
2681 specify two numbers separated by a colon---
2683 @var{mask}:@var{value}
2686 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2689 The resulting number is then packed
2691 @c this conditional paren in case bit fields turned on elsewhere than 960
2692 (in host-dependent byte order)
2694 into a field whose width depends on which assembler directive has the
2695 bit-field as its argument. Overflow (a result from the bitwise and
2696 requiring more binary digits to represent) is not an error; instead,
2697 more constants are generated, of the specified width, beginning with the
2698 least significant digits.@refill
2700 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2701 @code{.short}, and @code{.word} accept bit-field arguments.
2706 @chapter Sections and Relocation
2711 * Secs Background:: Background
2712 * Ld Sections:: Linker Sections
2713 * As Sections:: Assembler Internal Sections
2714 * Sub-Sections:: Sub-Sections
2718 @node Secs Background
2721 Roughly, a section is a range of addresses, with no gaps; all data
2722 ``in'' those addresses is treated the same for some particular purpose.
2723 For example there may be a ``read only'' section.
2725 @cindex linker, and assembler
2726 @cindex assembler, and linker
2727 The linker @code{@value{LD}} reads many object files (partial programs) and
2728 combines their contents to form a runnable program. When @command{@value{AS}}
2729 emits an object file, the partial program is assumed to start at address 0.
2730 @code{@value{LD}} assigns the final addresses for the partial program, so that
2731 different partial programs do not overlap. This is actually an
2732 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2735 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2736 addresses. These blocks slide to their run-time addresses as rigid
2737 units; their length does not change and neither does the order of bytes
2738 within them. Such a rigid unit is called a @emph{section}. Assigning
2739 run-time addresses to sections is called @dfn{relocation}. It includes
2740 the task of adjusting mentions of object-file addresses so they refer to
2741 the proper run-time addresses.
2743 For the H8/300, and for the Renesas / SuperH SH,
2744 @command{@value{AS}} pads sections if needed to
2745 ensure they end on a word (sixteen bit) boundary.
2748 @cindex standard assembler sections
2749 An object file written by @command{@value{AS}} has at least three sections, any
2750 of which may be empty. These are named @dfn{text}, @dfn{data} and
2755 When it generates COFF or ELF output,
2757 @command{@value{AS}} can also generate whatever other named sections you specify
2758 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2759 If you do not use any directives that place output in the @samp{.text}
2760 or @samp{.data} sections, these sections still exist, but are empty.
2765 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2767 @command{@value{AS}} can also generate whatever other named sections you
2768 specify using the @samp{.space} and @samp{.subspace} directives. See
2769 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2770 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2771 assembler directives.
2774 Additionally, @command{@value{AS}} uses different names for the standard
2775 text, data, and bss sections when generating SOM output. Program text
2776 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2777 BSS into @samp{$BSS$}.
2781 Within the object file, the text section starts at address @code{0}, the
2782 data section follows, and the bss section follows the data section.
2785 When generating either SOM or ELF output files on the HPPA, the text
2786 section starts at address @code{0}, the data section at address
2787 @code{0x4000000}, and the bss section follows the data section.
2790 To let @code{@value{LD}} know which data changes when the sections are
2791 relocated, and how to change that data, @command{@value{AS}} also writes to the
2792 object file details of the relocation needed. To perform relocation
2793 @code{@value{LD}} must know, each time an address in the object
2797 Where in the object file is the beginning of this reference to
2800 How long (in bytes) is this reference?
2802 Which section does the address refer to? What is the numeric value of
2804 (@var{address}) @minus{} (@var{start-address of section})?
2807 Is the reference to an address ``Program-Counter relative''?
2810 @cindex addresses, format of
2811 @cindex section-relative addressing
2812 In fact, every address @command{@value{AS}} ever uses is expressed as
2814 (@var{section}) + (@var{offset into section})
2817 Further, most expressions @command{@value{AS}} computes have this section-relative
2820 (For some object formats, such as SOM for the HPPA, some expressions are
2821 symbol-relative instead.)
2824 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2825 @var{N} into section @var{secname}.''
2827 Apart from text, data and bss sections you need to know about the
2828 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2829 addresses in the absolute section remain unchanged. For example, address
2830 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2831 @code{@value{LD}}. Although the linker never arranges two partial programs'
2832 data sections with overlapping addresses after linking, @emph{by definition}
2833 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2834 part of a program is always the same address when the program is running as
2835 address @code{@{absolute@ 239@}} in any other part of the program.
2837 The idea of sections is extended to the @dfn{undefined} section. Any
2838 address whose section is unknown at assembly time is by definition
2839 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2840 Since numbers are always defined, the only way to generate an undefined
2841 address is to mention an undefined symbol. A reference to a named
2842 common block would be such a symbol: its value is unknown at assembly
2843 time so it has section @emph{undefined}.
2845 By analogy the word @emph{section} is used to describe groups of sections in
2846 the linked program. @code{@value{LD}} puts all partial programs' text
2847 sections in contiguous addresses in the linked program. It is
2848 customary to refer to the @emph{text section} of a program, meaning all
2849 the addresses of all partial programs' text sections. Likewise for
2850 data and bss sections.
2852 Some sections are manipulated by @code{@value{LD}}; others are invented for
2853 use of @command{@value{AS}} and have no meaning except during assembly.
2856 @section Linker Sections
2857 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2862 @cindex named sections
2863 @cindex sections, named
2864 @item named sections
2867 @cindex text section
2868 @cindex data section
2872 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2873 separate but equal sections. Anything you can say of one section is
2876 When the program is running, however, it is
2877 customary for the text section to be unalterable. The
2878 text section is often shared among processes: it contains
2879 instructions, constants and the like. The data section of a running
2880 program is usually alterable: for example, C variables would be stored
2881 in the data section.
2886 This section contains zeroed bytes when your program begins running. It
2887 is used to hold uninitialized variables or common storage. The length of
2888 each partial program's bss section is important, but because it starts
2889 out containing zeroed bytes there is no need to store explicit zero
2890 bytes in the object file. The bss section was invented to eliminate
2891 those explicit zeros from object files.
2893 @cindex absolute section
2894 @item absolute section
2895 Address 0 of this section is always ``relocated'' to runtime address 0.
2896 This is useful if you want to refer to an address that @code{@value{LD}} must
2897 not change when relocating. In this sense we speak of absolute
2898 addresses being ``unrelocatable'': they do not change during relocation.
2900 @cindex undefined section
2901 @item undefined section
2902 This ``section'' is a catch-all for address references to objects not in
2903 the preceding sections.
2904 @c FIXME: ref to some other doc on obj-file formats could go here.
2907 @cindex relocation example
2908 An idealized example of three relocatable sections follows.
2910 The example uses the traditional section names @samp{.text} and @samp{.data}.
2912 Memory addresses are on the horizontal axis.
2916 @c END TEXI2ROFF-KILL
2919 partial program # 1: |ttttt|dddd|00|
2926 partial program # 2: |TTT|DDD|000|
2929 +--+---+-----+--+----+---+-----+~~
2930 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2931 +--+---+-----+--+----+---+-----+~~
2933 addresses: 0 @dots{}
2940 \line{\it Partial program \#1: \hfil}
2941 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2942 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2944 \line{\it Partial program \#2: \hfil}
2945 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2946 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2948 \line{\it linked program: \hfil}
2949 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2950 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2951 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2952 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2954 \line{\it addresses: \hfil}
2958 @c END TEXI2ROFF-KILL
2961 @section Assembler Internal Sections
2963 @cindex internal assembler sections
2964 @cindex sections in messages, internal
2965 These sections are meant only for the internal use of @command{@value{AS}}. They
2966 have no meaning at run-time. You do not really need to know about these
2967 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2968 warning messages, so it might be helpful to have an idea of their
2969 meanings to @command{@value{AS}}. These sections are used to permit the
2970 value of every expression in your assembly language program to be a
2971 section-relative address.
2974 @cindex assembler internal logic error
2975 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2976 An internal assembler logic error has been found. This means there is a
2977 bug in the assembler.
2979 @cindex expr (internal section)
2981 The assembler stores complex expression internally as combinations of
2982 symbols. When it needs to represent an expression as a symbol, it puts
2983 it in the expr section.
2985 @c FIXME item transfer[t] vector preload
2986 @c FIXME item transfer[t] vector postload
2987 @c FIXME item register
2991 @section Sub-Sections
2993 @cindex numbered subsections
2994 @cindex grouping data
3000 fall into two sections: text and data.
3002 You may have separate groups of
3004 data in named sections
3008 data in named sections
3014 that you want to end up near to each other in the object file, even though they
3015 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3016 use @dfn{subsections} for this purpose. Within each section, there can be
3017 numbered subsections with values from 0 to 8192. Objects assembled into the
3018 same subsection go into the object file together with other objects in the same
3019 subsection. For example, a compiler might want to store constants in the text
3020 section, but might not want to have them interspersed with the program being
3021 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3022 section of code being output, and a @samp{.text 1} before each group of
3023 constants being output.
3025 Subsections are optional. If you do not use subsections, everything
3026 goes in subsection number zero.
3029 Each subsection is zero-padded up to a multiple of four bytes.
3030 (Subsections may be padded a different amount on different flavors
3031 of @command{@value{AS}}.)
3035 On the H8/300 platform, each subsection is zero-padded to a word
3036 boundary (two bytes).
3037 The same is true on the Renesas SH.
3040 @c FIXME section padding (alignment)?
3041 @c Rich Pixley says padding here depends on target obj code format; that
3042 @c doesn't seem particularly useful to say without further elaboration,
3043 @c so for now I say nothing about it. If this is a generic BFD issue,
3044 @c these paragraphs might need to vanish from this manual, and be
3045 @c discussed in BFD chapter of binutils (or some such).
3049 Subsections appear in your object file in numeric order, lowest numbered
3050 to highest. (All this to be compatible with other people's assemblers.)
3051 The object file contains no representation of subsections; @code{@value{LD}} and
3052 other programs that manipulate object files see no trace of them.
3053 They just see all your text subsections as a text section, and all your
3054 data subsections as a data section.
3056 To specify which subsection you want subsequent statements assembled
3057 into, use a numeric argument to specify it, in a @samp{.text
3058 @var{expression}} or a @samp{.data @var{expression}} statement.
3061 When generating COFF output, you
3066 can also use an extra subsection
3067 argument with arbitrary named sections: @samp{.section @var{name},
3072 When generating ELF output, you
3077 can also use the @code{.subsection} directive (@pxref{SubSection})
3078 to specify a subsection: @samp{.subsection @var{expression}}.
3080 @var{Expression} should be an absolute expression.
3081 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
3082 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3083 begins in @code{text 0}. For instance:
3085 .text 0 # The default subsection is text 0 anyway.
3086 .ascii "This lives in the first text subsection. *"
3088 .ascii "But this lives in the second text subsection."
3090 .ascii "This lives in the data section,"
3091 .ascii "in the first data subsection."
3093 .ascii "This lives in the first text section,"
3094 .ascii "immediately following the asterisk (*)."
3097 Each section has a @dfn{location counter} incremented by one for every byte
3098 assembled into that section. Because subsections are merely a convenience
3099 restricted to @command{@value{AS}} there is no concept of a subsection location
3100 counter. There is no way to directly manipulate a location counter---but the
3101 @code{.align} directive changes it, and any label definition captures its
3102 current value. The location counter of the section where statements are being
3103 assembled is said to be the @dfn{active} location counter.
3106 @section bss Section
3109 @cindex common variable storage
3110 The bss section is used for local common variable storage.
3111 You may allocate address space in the bss section, but you may
3112 not dictate data to load into it before your program executes. When
3113 your program starts running, all the contents of the bss
3114 section are zeroed bytes.
3116 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3117 @ref{Lcomm,,@code{.lcomm}}.
3119 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3120 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
3123 When assembling for a target which supports multiple sections, such as ELF or
3124 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3125 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3126 section. Typically the section will only contain symbol definitions and
3127 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3134 Symbols are a central concept: the programmer uses symbols to name
3135 things, the linker uses symbols to link, and the debugger uses symbols
3139 @cindex debuggers, and symbol order
3140 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3141 the same order they were declared. This may break some debuggers.
3146 * Setting Symbols:: Giving Symbols Other Values
3147 * Symbol Names:: Symbol Names
3148 * Dot:: The Special Dot Symbol
3149 * Symbol Attributes:: Symbol Attributes
3156 A @dfn{label} is written as a symbol immediately followed by a colon
3157 @samp{:}. The symbol then represents the current value of the
3158 active location counter, and is, for example, a suitable instruction
3159 operand. You are warned if you use the same symbol to represent two
3160 different locations: the first definition overrides any other
3164 On the HPPA, the usual form for a label need not be immediately followed by a
3165 colon, but instead must start in column zero. Only one label may be defined on
3166 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3167 provides a special directive @code{.label} for defining labels more flexibly.
3170 @node Setting Symbols
3171 @section Giving Symbols Other Values
3173 @cindex assigning values to symbols
3174 @cindex symbol values, assigning
3175 A symbol can be given an arbitrary value by writing a symbol, followed
3176 by an equals sign @samp{=}, followed by an expression
3177 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3178 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3179 equals sign @samp{=}@samp{=} here represents an equivalent of the
3180 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3183 @section Symbol Names
3185 @cindex symbol names
3186 @cindex names, symbol
3187 @ifclear SPECIAL-SYMS
3188 Symbol names begin with a letter or with one of @samp{._}. On most
3189 machines, you can also use @code{$} in symbol names; exceptions are
3190 noted in @ref{Machine Dependencies}. That character may be followed by any
3191 string of digits, letters, dollar signs (unless otherwise noted in
3192 @ref{Machine Dependencies}), and underscores.
3196 Symbol names begin with a letter or with one of @samp{._}. On the
3197 Renesas SH you can also use @code{$} in symbol names. That
3198 character may be followed by any string of digits, letters, dollar signs (save
3199 on the H8/300), and underscores.
3203 Case of letters is significant: @code{foo} is a different symbol name
3206 Each symbol has exactly one name. Each name in an assembly language program
3207 refers to exactly one symbol. You may use that symbol name any number of times
3210 @subheading Local Symbol Names
3212 @cindex local symbol names
3213 @cindex symbol names, local
3214 @cindex temporary symbol names
3215 @cindex symbol names, temporary
3216 Local symbols help compilers and programmers use names temporarily.
3217 They create symbols which are guaranteed to be unique over the entire scope of
3218 the input source code and which can be referred to by a simple notation.
3219 To define a local symbol, write a label of the form @samp{@b{N}:} (where @b{N}
3220 represents any positive integer). To refer to the most recent previous
3221 definition of that symbol write @samp{@b{N}b}, using the same number as when
3222 you defined the label. To refer to the next definition of a local label, write
3223 @samp{@b{N}f}--- The @samp{b} stands for``backwards'' and the @samp{f} stands
3226 There is no restriction on how you can use these labels, and you can reuse them
3227 too. So that it is possible to repeatedly define the same local label (using
3228 the same number @samp{@b{N}}), although you can only refer to the most recently
3229 defined local label of that number (for a backwards reference) or the next
3230 definition of a specific local label for a forward reference. It is also worth
3231 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3232 implemented in a slightly more efficient manner than the others.
3243 Which is the equivalent of:
3246 label_1: branch label_3
3247 label_2: branch label_1
3248 label_3: branch label_4
3249 label_4: branch label_3
3252 Local symbol names are only a notational device. They are immediately
3253 transformed into more conventional symbol names before the assembler uses them.
3254 The symbol names stored in the symbol table, appearing in error messages and
3255 optionally emitted to the object file. The names are constructed using these
3260 All local labels begin with @samp{L}. Normally both @command{@value{AS}} and
3261 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
3262 used for symbols you are never intended to see. If you use the
3263 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3264 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3265 you may use them in debugging.
3268 This is the number that was used in the local label definition. So if the
3269 label is written @samp{55:} then the number is @samp{55}.
3272 This unusual character is included so you do not accidentally invent a symbol
3273 of the same name. The character has ASCII value of @samp{\002} (control-B).
3275 @item @emph{ordinal number}
3276 This is a serial number to keep the labels distinct. The first definition of
3277 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3278 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3279 the number @samp{1} and its 15th defintion gets @samp{15} as well.
3282 So for example, the first @code{1:} is named @code{L1@kbd{C-B}1}, the 44th
3283 @code{3:} is named @code{L3@kbd{C-B}44}.
3285 @subheading Dollar Local Labels
3286 @cindex dollar local symbols
3288 @code{@value{AS}} also supports an even more local form of local labels called
3289 dollar labels. These labels go out of scope (ie they become undefined) as soon
3290 as a non-local label is defined. Thus they remain valid for only a small
3291 region of the input source code. Normal local labels, by contrast, remain in
3292 scope for the entire file, or until they are redefined by another occurrence of
3293 the same local label.
3295 Dollar labels are defined in exactly the same way as ordinary local labels,
3296 except that instead of being terminated by a colon, they are terminated by a
3297 dollar sign. eg @samp{@b{55$}}.
3299 They can also be distinguished from ordinary local labels by their transformed
3300 name which uses ASCII character @samp{\001} (control-A) as the magic character
3301 to distinguish them from ordinary labels. Thus the 5th defintion of @samp{6$}
3302 is named @samp{L6@kbd{C-A}5}.
3305 @section The Special Dot Symbol
3307 @cindex dot (symbol)
3308 @cindex @code{.} (symbol)
3309 @cindex current address
3310 @cindex location counter
3311 The special symbol @samp{.} refers to the current address that
3312 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3313 .long .} defines @code{melvin} to contain its own address.
3314 Assigning a value to @code{.} is treated the same as a @code{.org}
3315 directive. Thus, the expression @samp{.=.+4} is the same as saying
3316 @ifclear no-space-dir
3320 @node Symbol Attributes
3321 @section Symbol Attributes
3323 @cindex symbol attributes
3324 @cindex attributes, symbol
3325 Every symbol has, as well as its name, the attributes ``Value'' and
3326 ``Type''. Depending on output format, symbols can also have auxiliary
3329 The detailed definitions are in @file{a.out.h}.
3332 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3333 all these attributes, and probably won't warn you. This makes the
3334 symbol an externally defined symbol, which is generally what you
3338 * Symbol Value:: Value
3339 * Symbol Type:: Type
3342 * a.out Symbols:: Symbol Attributes: @code{a.out}
3346 * a.out Symbols:: Symbol Attributes: @code{a.out}
3349 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3354 * COFF Symbols:: Symbol Attributes for COFF
3357 * SOM Symbols:: Symbol Attributes for SOM
3364 @cindex value of a symbol
3365 @cindex symbol value
3366 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3367 location in the text, data, bss or absolute sections the value is the
3368 number of addresses from the start of that section to the label.
3369 Naturally for text, data and bss sections the value of a symbol changes
3370 as @code{@value{LD}} changes section base addresses during linking. Absolute
3371 symbols' values do not change during linking: that is why they are
3374 The value of an undefined symbol is treated in a special way. If it is
3375 0 then the symbol is not defined in this assembler source file, and
3376 @code{@value{LD}} tries to determine its value from other files linked into the
3377 same program. You make this kind of symbol simply by mentioning a symbol
3378 name without defining it. A non-zero value represents a @code{.comm}
3379 common declaration. The value is how much common storage to reserve, in
3380 bytes (addresses). The symbol refers to the first address of the
3386 @cindex type of a symbol
3388 The type attribute of a symbol contains relocation (section)
3389 information, any flag settings indicating that a symbol is external, and
3390 (optionally), other information for linkers and debuggers. The exact
3391 format depends on the object-code output format in use.
3396 @c The following avoids a "widow" subsection title. @group would be
3397 @c better if it were available outside examples.
3400 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3402 @cindex @code{b.out} symbol attributes
3403 @cindex symbol attributes, @code{b.out}
3404 These symbol attributes appear only when @command{@value{AS}} is configured for
3405 one of the Berkeley-descended object output formats---@code{a.out} or
3411 @subsection Symbol Attributes: @code{a.out}
3413 @cindex @code{a.out} symbol attributes
3414 @cindex symbol attributes, @code{a.out}
3420 @subsection Symbol Attributes: @code{a.out}
3422 @cindex @code{a.out} symbol attributes
3423 @cindex symbol attributes, @code{a.out}
3427 * Symbol Desc:: Descriptor
3428 * Symbol Other:: Other
3432 @subsubsection Descriptor
3434 @cindex descriptor, of @code{a.out} symbol
3435 This is an arbitrary 16-bit value. You may establish a symbol's
3436 descriptor value by using a @code{.desc} statement
3437 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3438 @command{@value{AS}}.
3441 @subsubsection Other
3443 @cindex other attribute, of @code{a.out} symbol
3444 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3449 @subsection Symbol Attributes for COFF
3451 @cindex COFF symbol attributes
3452 @cindex symbol attributes, COFF
3454 The COFF format supports a multitude of auxiliary symbol attributes;
3455 like the primary symbol attributes, they are set between @code{.def} and
3456 @code{.endef} directives.
3458 @subsubsection Primary Attributes
3460 @cindex primary attributes, COFF symbols
3461 The symbol name is set with @code{.def}; the value and type,
3462 respectively, with @code{.val} and @code{.type}.
3464 @subsubsection Auxiliary Attributes
3466 @cindex auxiliary attributes, COFF symbols
3467 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3468 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3469 table information for COFF.
3474 @subsection Symbol Attributes for SOM
3476 @cindex SOM symbol attributes
3477 @cindex symbol attributes, SOM
3479 The SOM format for the HPPA supports a multitude of symbol attributes set with
3480 the @code{.EXPORT} and @code{.IMPORT} directives.
3482 The attributes are described in @cite{HP9000 Series 800 Assembly
3483 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3484 @code{EXPORT} assembler directive documentation.
3488 @chapter Expressions
3492 @cindex numeric values
3493 An @dfn{expression} specifies an address or numeric value.
3494 Whitespace may precede and/or follow an expression.
3496 The result of an expression must be an absolute number, or else an offset into
3497 a particular section. If an expression is not absolute, and there is not
3498 enough information when @command{@value{AS}} sees the expression to know its
3499 section, a second pass over the source program might be necessary to interpret
3500 the expression---but the second pass is currently not implemented.
3501 @command{@value{AS}} aborts with an error message in this situation.
3504 * Empty Exprs:: Empty Expressions
3505 * Integer Exprs:: Integer Expressions
3509 @section Empty Expressions
3511 @cindex empty expressions
3512 @cindex expressions, empty
3513 An empty expression has no value: it is just whitespace or null.
3514 Wherever an absolute expression is required, you may omit the
3515 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3516 is compatible with other assemblers.
3519 @section Integer Expressions
3521 @cindex integer expressions
3522 @cindex expressions, integer
3523 An @dfn{integer expression} is one or more @emph{arguments} delimited
3524 by @emph{operators}.
3527 * Arguments:: Arguments
3528 * Operators:: Operators
3529 * Prefix Ops:: Prefix Operators
3530 * Infix Ops:: Infix Operators
3534 @subsection Arguments
3536 @cindex expression arguments
3537 @cindex arguments in expressions
3538 @cindex operands in expressions
3539 @cindex arithmetic operands
3540 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3541 contexts arguments are sometimes called ``arithmetic operands''. In
3542 this manual, to avoid confusing them with the ``instruction operands'' of
3543 the machine language, we use the term ``argument'' to refer to parts of
3544 expressions only, reserving the word ``operand'' to refer only to machine
3545 instruction operands.
3547 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3548 @var{section} is one of text, data, bss, absolute,
3549 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3552 Numbers are usually integers.
3554 A number can be a flonum or bignum. In this case, you are warned
3555 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3556 these 32 bits are an integer. You may write integer-manipulating
3557 instructions that act on exotic constants, compatible with other
3560 @cindex subexpressions
3561 Subexpressions are a left parenthesis @samp{(} followed by an integer
3562 expression, followed by a right parenthesis @samp{)}; or a prefix
3563 operator followed by an argument.
3566 @subsection Operators
3568 @cindex operators, in expressions
3569 @cindex arithmetic functions
3570 @cindex functions, in expressions
3571 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3572 operators are followed by an argument. Infix operators appear
3573 between their arguments. Operators may be preceded and/or followed by
3577 @subsection Prefix Operator
3579 @cindex prefix operators
3580 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3581 one argument, which must be absolute.
3583 @c the tex/end tex stuff surrounding this small table is meant to make
3584 @c it align, on the printed page, with the similar table in the next
3585 @c section (which is inside an enumerate).
3587 \global\advance\leftskip by \itemindent
3592 @dfn{Negation}. Two's complement negation.
3594 @dfn{Complementation}. Bitwise not.
3598 \global\advance\leftskip by -\itemindent
3602 @subsection Infix Operators
3604 @cindex infix operators
3605 @cindex operators, permitted arguments
3606 @dfn{Infix operators} take two arguments, one on either side. Operators
3607 have precedence, but operations with equal precedence are performed left
3608 to right. Apart from @code{+} or @option{-}, both arguments must be
3609 absolute, and the result is absolute.
3612 @cindex operator precedence
3613 @cindex precedence of operators
3620 @dfn{Multiplication}.
3623 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3629 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3632 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3636 Intermediate precedence
3641 @dfn{Bitwise Inclusive Or}.
3647 @dfn{Bitwise Exclusive Or}.
3650 @dfn{Bitwise Or Not}.
3657 @cindex addition, permitted arguments
3658 @cindex plus, permitted arguments
3659 @cindex arguments for addition
3661 @dfn{Addition}. If either argument is absolute, the result has the section of
3662 the other argument. You may not add together arguments from different
3665 @cindex subtraction, permitted arguments
3666 @cindex minus, permitted arguments
3667 @cindex arguments for subtraction
3669 @dfn{Subtraction}. If the right argument is absolute, the
3670 result has the section of the left argument.
3671 If both arguments are in the same section, the result is absolute.
3672 You may not subtract arguments from different sections.
3673 @c FIXME is there still something useful to say about undefined - undefined ?
3675 @cindex comparison expressions
3676 @cindex expressions, comparison
3681 @dfn{Is Not Equal To}
3685 @dfn{Is Greater Than}
3687 @dfn{Is Greater Than Or Equal To}
3689 @dfn{Is Less Than Or Equal To}
3691 The comparison operators can be used as infix operators. A true results has a
3692 value of -1 whereas a false result has a value of 0. Note, these operators
3693 perform signed comparisons.
3696 @item Lowest Precedence
3705 These two logical operations can be used to combine the results of sub
3706 expressions. Note, unlike the comparison operators a true result returns a
3707 value of 1 but a false results does still return 0. Also note that the logical
3708 or operator has a slightly lower precedence than logical and.
3713 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3714 address; you can only have a defined section in one of the two arguments.
3717 @chapter Assembler Directives
3719 @cindex directives, machine independent
3720 @cindex pseudo-ops, machine independent
3721 @cindex machine independent directives
3722 All assembler directives have names that begin with a period (@samp{.}).
3723 The rest of the name is letters, usually in lower case.
3725 This chapter discusses directives that are available regardless of the
3726 target machine configuration for the @sc{gnu} assembler.
3728 Some machine configurations provide additional directives.
3729 @xref{Machine Dependencies}.
3732 @ifset machine-directives
3733 @xref{Machine Dependencies} for additional directives.
3738 * Abort:: @code{.abort}
3740 * ABORT (COFF):: @code{.ABORT}
3743 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3744 * Altmacro:: @code{.altmacro}
3745 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3746 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3747 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3748 * Byte:: @code{.byte @var{expressions}}
3749 * Comm:: @code{.comm @var{symbol} , @var{length} }
3751 * CFI directives:: @code{.cfi_startproc}, @code{.cfi_endproc}, etc.
3753 * Data:: @code{.data @var{subsection}}
3755 * Def:: @code{.def @var{name}}
3758 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3764 * Double:: @code{.double @var{flonums}}
3765 * Eject:: @code{.eject}
3766 * Else:: @code{.else}
3767 * Elseif:: @code{.elseif}
3770 * Endef:: @code{.endef}
3773 * Endfunc:: @code{.endfunc}
3774 * Endif:: @code{.endif}
3775 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3776 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3777 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
3779 * Error:: @code{.error @var{string}}
3780 * Exitm:: @code{.exitm}
3781 * Extern:: @code{.extern}
3782 * Fail:: @code{.fail}
3783 @ifclear no-file-dir
3784 * File:: @code{.file @var{string}}
3787 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3788 * Float:: @code{.float @var{flonums}}
3789 * Func:: @code{.func}
3790 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3792 * Hidden:: @code{.hidden @var{names}}
3795 * hword:: @code{.hword @var{expressions}}
3796 * Ident:: @code{.ident}
3797 * If:: @code{.if @var{absolute expression}}
3798 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3799 * Include:: @code{.include "@var{file}"}
3800 * Int:: @code{.int @var{expressions}}
3802 * Internal:: @code{.internal @var{names}}
3805 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3806 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3807 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3808 * Lflags:: @code{.lflags}
3809 @ifclear no-line-dir
3810 * Line:: @code{.line @var{line-number}}
3813 * Linkonce:: @code{.linkonce [@var{type}]}
3814 * List:: @code{.list}
3815 * Ln:: @code{.ln @var{line-number}}
3817 * LNS directives:: @code{.file}, @code{.loc}, etc.
3819 * Long:: @code{.long @var{expressions}}
3821 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3824 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3825 * MRI:: @code{.mri @var{val}}
3826 * Noaltmacro:: @code{.noaltmacro}
3827 * Nolist:: @code{.nolist}
3828 * Octa:: @code{.octa @var{bignums}}
3829 * Org:: @code{.org @var{new-lc} , @var{fill}}
3830 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3832 * PopSection:: @code{.popsection}
3833 * Previous:: @code{.previous}
3836 * Print:: @code{.print @var{string}}
3838 * Protected:: @code{.protected @var{names}}
3841 * Psize:: @code{.psize @var{lines}, @var{columns}}
3842 * Purgem:: @code{.purgem @var{name}}
3844 * PushSection:: @code{.pushsection @var{name}}
3847 * Quad:: @code{.quad @var{bignums}}
3848 * Rept:: @code{.rept @var{count}}
3849 * Sbttl:: @code{.sbttl "@var{subheading}"}
3851 * Scl:: @code{.scl @var{class}}
3854 * Section:: @code{.section @var{name}}
3857 * Set:: @code{.set @var{symbol}, @var{expression}}
3858 * Short:: @code{.short @var{expressions}}
3859 * Single:: @code{.single @var{flonums}}
3861 * Size:: @code{.size [@var{name} , @var{expression}]}
3864 * Skip:: @code{.skip @var{size} , @var{fill}}
3865 * Sleb128:: @code{.sleb128 @var{expressions}}
3866 * Space:: @code{.space @var{size} , @var{fill}}
3868 * Stab:: @code{.stabd, .stabn, .stabs}
3871 * String:: @code{.string "@var{str}"}
3872 * Struct:: @code{.struct @var{expression}}
3874 * SubSection:: @code{.subsection}
3875 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3879 * Tag:: @code{.tag @var{structname}}
3882 * Text:: @code{.text @var{subsection}}
3883 * Title:: @code{.title "@var{heading}"}
3885 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3888 * Uleb128:: @code{.uleb128 @var{expressions}}
3890 * Val:: @code{.val @var{addr}}
3894 * Version:: @code{.version "@var{string}"}
3895 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3896 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3899 * Warning:: @code{.warning @var{string}}
3900 * Weak:: @code{.weak @var{names}}
3901 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
3902 * Word:: @code{.word @var{expressions}}
3903 * Deprecated:: Deprecated Directives
3907 @section @code{.abort}
3909 @cindex @code{abort} directive
3910 @cindex stopping the assembly
3911 This directive stops the assembly immediately. It is for
3912 compatibility with other assemblers. The original idea was that the
3913 assembly language source would be piped into the assembler. If the sender
3914 of the source quit, it could use this directive tells @command{@value{AS}} to
3915 quit also. One day @code{.abort} will not be supported.
3919 @section @code{.ABORT} (COFF)
3921 @cindex @code{ABORT} directive
3922 When producing COFF output, @command{@value{AS}} accepts this directive as a
3923 synonym for @samp{.abort}.
3926 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3932 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3934 @cindex padding the location counter
3935 @cindex @code{align} directive
3936 Pad the location counter (in the current subsection) to a particular storage
3937 boundary. The first expression (which must be absolute) is the alignment
3938 required, as described below.
3940 The second expression (also absolute) gives the fill value to be stored in the
3941 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3942 padding bytes are normally zero. However, on some systems, if the section is
3943 marked as containing code and the fill value is omitted, the space is filled
3944 with no-op instructions.
3946 The third expression is also absolute, and is also optional. If it is present,
3947 it is the maximum number of bytes that should be skipped by this alignment
3948 directive. If doing the alignment would require skipping more bytes than the
3949 specified maximum, then the alignment is not done at all. You can omit the
3950 fill value (the second argument) entirely by simply using two commas after the
3951 required alignment; this can be useful if you want the alignment to be filled
3952 with no-op instructions when appropriate.
3954 The way the required alignment is specified varies from system to system.
3955 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
3956 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
3957 alignment request in bytes. For example @samp{.align 8} advances
3958 the location counter until it is a multiple of 8. If the location counter
3959 is already a multiple of 8, no change is needed. For the tic54x, the
3960 first expression is the alignment request in words.
3962 For other systems, including the i386 using a.out format, and the arm and
3963 strongarm, it is the
3964 number of low-order zero bits the location counter must have after
3965 advancement. For example @samp{.align 3} advances the location
3966 counter until it a multiple of 8. If the location counter is already a
3967 multiple of 8, no change is needed.
3969 This inconsistency is due to the different behaviors of the various
3970 native assemblers for these systems which GAS must emulate.
3971 GAS also provides @code{.balign} and @code{.p2align} directives,
3972 described later, which have a consistent behavior across all
3973 architectures (but are specific to GAS).
3976 @section @code{.ascii "@var{string}"}@dots{}
3978 @cindex @code{ascii} directive
3979 @cindex string literals
3980 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3981 separated by commas. It assembles each string (with no automatic
3982 trailing zero byte) into consecutive addresses.
3985 @section @code{.asciz "@var{string}"}@dots{}
3987 @cindex @code{asciz} directive
3988 @cindex zero-terminated strings
3989 @cindex null-terminated strings
3990 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3991 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3994 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3996 @cindex padding the location counter given number of bytes
3997 @cindex @code{balign} directive
3998 Pad the location counter (in the current subsection) to a particular
3999 storage boundary. The first expression (which must be absolute) is the
4000 alignment request in bytes. For example @samp{.balign 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.
4004 The second expression (also absolute) gives the fill value to be stored in the
4005 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4006 padding bytes are normally zero. However, on some systems, if the section is
4007 marked as containing code and the fill value is omitted, the space is filled
4008 with no-op instructions.
4010 The third expression is also absolute, and is also optional. If it is present,
4011 it is the maximum number of bytes that should be skipped by this alignment
4012 directive. If doing the alignment would require skipping more bytes than the
4013 specified maximum, then the alignment is not done at all. You can omit the
4014 fill value (the second argument) entirely by simply using two commas after the
4015 required alignment; this can be useful if you want the alignment to be filled
4016 with no-op instructions when appropriate.
4018 @cindex @code{balignw} directive
4019 @cindex @code{balignl} directive
4020 The @code{.balignw} and @code{.balignl} directives are variants of the
4021 @code{.balign} directive. The @code{.balignw} directive treats the fill
4022 pattern as a two byte word value. The @code{.balignl} directives treats the
4023 fill pattern as a four byte longword value. For example, @code{.balignw
4024 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4025 filled in with the value 0x368d (the exact placement of the bytes depends upon
4026 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4030 @section @code{.byte @var{expressions}}
4032 @cindex @code{byte} directive
4033 @cindex integers, one byte
4034 @code{.byte} expects zero or more expressions, separated by commas.
4035 Each expression is assembled into the next byte.
4038 @section @code{.comm @var{symbol} , @var{length} }
4040 @cindex @code{comm} directive
4041 @cindex symbol, common
4042 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4043 common symbol in one object file may be merged with a defined or common symbol
4044 of the same name in another object file. If @code{@value{LD}} does not see a
4045 definition for the symbol--just one or more common symbols--then it will
4046 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4047 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4048 the same name, and they do not all have the same size, it will allocate space
4049 using the largest size.
4052 When using ELF, the @code{.comm} directive takes an optional third argument.
4053 This is the desired alignment of the symbol, specified as a byte boundary (for
4054 example, an alignment of 16 means that the least significant 4 bits of the
4055 address should be zero). The alignment must be an absolute expression, and it
4056 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
4057 for the common symbol, it will use the alignment when placing the symbol. If
4058 no alignment is specified, @command{@value{AS}} will set the alignment to the
4059 largest power of two less than or equal to the size of the symbol, up to a
4064 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4065 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4068 @node CFI directives
4069 @section @code{.cfi_startproc}
4070 @cindex @code{cfi_startproc} directive
4071 @code{.cfi_startproc} is used at the beginning of each function that
4072 should have an entry in @code{.eh_frame}. It initializes some internal
4073 data structures and emits architecture dependent initial CFI instructions.
4074 Don't forget to close the function by
4075 @code{.cfi_endproc}.
4077 @section @code{.cfi_endproc}
4078 @cindex @code{cfi_endproc} directive
4079 @code{.cfi_endproc} is used at the end of a function where it closes its
4080 unwind entry previously opened by
4081 @code{.cfi_startproc}. and emits it to @code{.eh_frame}.
4083 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4084 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4085 address from @var{register} and add @var{offset} to it}.
4087 @section @code{.cfi_def_cfa_register @var{register}}
4088 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4089 now on @var{register} will be used instead of the old one. Offset
4092 @section @code{.cfi_def_cfa_offset @var{offset}}
4093 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4094 remains the same, but @var{offset} is new. Note that it is the
4095 absolute offset that will be added to a defined register to compute
4098 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4099 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4100 value that is added/substracted from the previous offset.
4102 @section @code{.cfi_offset @var{register}, @var{offset}}
4103 Previous value of @var{register} is saved at offset @var{offset} from
4106 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4107 Previous value of @var{register} is saved at offset @var{offset} from
4108 the current CFA register. This is transformed to @code{.cfi_offset}
4109 using the known displacement of the CFA register from the CFA.
4110 This is often easier to use, because the number will match the
4111 code it's annotating.
4113 @section @code{.cfi_signal_frame}
4114 Mark current function as signal trampoline.
4116 @section @code{.cfi_window_save}
4117 SPARC register window has been saved.
4119 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4120 Allows the user to add arbitrary bytes to the unwind info. One
4121 might use this to add OS-specific CFI opcodes, or generic CFI
4122 opcodes that GAS does not yet support.
4124 @node LNS directives
4125 @section @code{.file @var{fileno} @var{filename}}
4126 @cindex @code{file} directive
4127 When emitting dwarf2 line number information @code{.file} assigns filenames
4128 to the @code{.debug_line} file name table. The @var{fileno} operand should
4129 be a unique positive integer to use as the index of the entry in the table.
4130 The @var{filename} operand is a C string literal.
4132 The detail of filename indicies is exposed to the user because the filename
4133 table is shared with the @code{.debug_info} section of the dwarf2 debugging
4134 information, and thus the user must know the exact indicies that table
4137 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
4138 @cindex @code{loc} directive
4139 The @code{.loc} directive will add row to the @code{.debug_line} line
4140 number matrix corresponding to the immediately following assembly
4141 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
4142 arguments will be applied to the @code{.debug_line} state machine before
4145 The @var{options} are a sequence of the following tokens in any order:
4149 This option will set the @code{basic_block} register in the
4150 @code{.debug_line} state machine to @code{true}.
4153 This option will set the @code{prologue_end} register in the
4154 @code{.debug_line} state machine to @code{true}.
4156 @item epilogue_begin
4157 This option will set the @code{epilogue_begin} register in the
4158 @code{.debug_line} state machine to @code{true}.
4160 @item is_stmt @var{value}
4161 This option will set the @code{is_stmt} register in the
4162 @code{.debug_line} state machine to @code{value}, which must be
4165 @item isa @var{value}
4166 This directive will set the @code{isa} register in the @code{.debug_line}
4167 state machine to @var{value}, which must be an unsigned integer.
4171 @section @code{.loc_mark_blocks @var{enable}}
4172 @cindex @code{loc_mark_blocks} directive
4173 The @code{.loc_mark_blocks} directive makes the assembler emit an entry
4174 to the @code{.debug_line} line number matrix with the @code{basic_block}
4175 register in the state machine set whenever a code label is seen.
4176 The @var{enable} argument should be either 1 or 0, to enable or disable
4177 this function respectively.
4180 @section @code{.data @var{subsection}}
4182 @cindex @code{data} directive
4183 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4184 end of the data subsection numbered @var{subsection} (which is an
4185 absolute expression). If @var{subsection} is omitted, it defaults
4190 @section @code{.def @var{name}}
4192 @cindex @code{def} directive
4193 @cindex COFF symbols, debugging
4194 @cindex debugging COFF symbols
4195 Begin defining debugging information for a symbol @var{name}; the
4196 definition extends until the @code{.endef} directive is encountered.
4199 This directive is only observed when @command{@value{AS}} is configured for COFF
4200 format output; when producing @code{b.out}, @samp{.def} is recognized,
4207 @section @code{.desc @var{symbol}, @var{abs-expression}}
4209 @cindex @code{desc} directive
4210 @cindex COFF symbol descriptor
4211 @cindex symbol descriptor, COFF
4212 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4213 to the low 16 bits of an absolute expression.
4216 The @samp{.desc} directive is not available when @command{@value{AS}} is
4217 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4218 object format. For the sake of compatibility, @command{@value{AS}} accepts
4219 it, but produces no output, when configured for COFF.
4225 @section @code{.dim}
4227 @cindex @code{dim} directive
4228 @cindex COFF auxiliary symbol information
4229 @cindex auxiliary symbol information, COFF
4230 This directive is generated by compilers to include auxiliary debugging
4231 information in the symbol table. It is only permitted inside
4232 @code{.def}/@code{.endef} pairs.
4235 @samp{.dim} is only meaningful when generating COFF format output; when
4236 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4242 @section @code{.double @var{flonums}}
4244 @cindex @code{double} directive
4245 @cindex floating point numbers (double)
4246 @code{.double} expects zero or more flonums, separated by commas. It
4247 assembles floating point numbers.
4249 The exact kind of floating point numbers emitted depends on how
4250 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4254 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4255 in @sc{ieee} format.
4260 @section @code{.eject}
4262 @cindex @code{eject} directive
4263 @cindex new page, in listings
4264 @cindex page, in listings
4265 @cindex listing control: new page
4266 Force a page break at this point, when generating assembly listings.
4269 @section @code{.else}
4271 @cindex @code{else} directive
4272 @code{.else} is part of the @command{@value{AS}} support for conditional
4273 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
4274 of code to be assembled if the condition for the preceding @code{.if}
4278 @section @code{.elseif}
4280 @cindex @code{elseif} directive
4281 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4282 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
4283 @code{.if} block that would otherwise fill the entire @code{.else} section.
4286 @section @code{.end}
4288 @cindex @code{end} directive
4289 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4290 process anything in the file past the @code{.end} directive.
4294 @section @code{.endef}
4296 @cindex @code{endef} directive
4297 This directive flags the end of a symbol definition begun with
4301 @samp{.endef} is only meaningful when generating COFF format output; if
4302 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4303 directive but ignores it.
4308 @section @code{.endfunc}
4309 @cindex @code{endfunc} directive
4310 @code{.endfunc} marks the end of a function specified with @code{.func}.
4313 @section @code{.endif}
4315 @cindex @code{endif} directive
4316 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4317 it marks the end of a block of code that is only assembled
4318 conditionally. @xref{If,,@code{.if}}.
4321 @section @code{.equ @var{symbol}, @var{expression}}
4323 @cindex @code{equ} directive
4324 @cindex assigning values to symbols
4325 @cindex symbols, assigning values to
4326 This directive sets the value of @var{symbol} to @var{expression}.
4327 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
4330 The syntax for @code{equ} on the HPPA is
4331 @samp{@var{symbol} .equ @var{expression}}.
4335 The syntax for @code{equ} on the Z80 is
4336 @samp{@var{symbol} equ @var{expression}}.
4337 On the Z80 it is an eror if @var{symbol} is already defined,
4338 but the symbol is not protected from later redefinition,
4339 compare @xref{Equiv}.
4343 @section @code{.equiv @var{symbol}, @var{expression}}
4344 @cindex @code{equiv} directive
4345 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4346 the assembler will signal an error if @var{symbol} is already defined. Note a
4347 symbol which has been referenced but not actually defined is considered to be
4350 Except for the contents of the error message, this is roughly equivalent to
4357 plus it protects the symbol from later redefinition.
4360 @section @code{.eqv @var{symbol}, @var{expression}}
4361 @cindex @code{eqv} directive
4362 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4363 evaluate the expression or any part of it immediately. Instead each time
4364 the resulting symbol is used in an expression, a snapshot of its current
4368 @section @code{.err}
4369 @cindex @code{err} directive
4370 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4371 message and, unless the @option{-Z} option was used, it will not generate an
4372 object file. This can be used to signal an error in conditionally compiled code.
4375 @section @code{.error "@var{string}"}
4376 @cindex error directive
4378 Similarly to @code{.err}, this directive emits an error, but you can specify a
4379 string that will be emitted as the error message. If you don't specify the
4380 message, it defaults to @code{".error directive invoked in source file"}.
4381 @xref{Errors, ,Error and Warning Messages}.
4384 .error "This code has not been assembled and tested."
4388 @section @code{.exitm}
4389 Exit early from the current macro definition. @xref{Macro}.
4392 @section @code{.extern}
4394 @cindex @code{extern} directive
4395 @code{.extern} is accepted in the source program---for compatibility
4396 with other assemblers---but it is ignored. @command{@value{AS}} treats
4397 all undefined symbols as external.
4400 @section @code{.fail @var{expression}}
4402 @cindex @code{fail} directive
4403 Generates an error or a warning. If the value of the @var{expression} is 500
4404 or more, @command{@value{AS}} will print a warning message. If the value is less
4405 than 500, @command{@value{AS}} will print an error message. The message will
4406 include the value of @var{expression}. This can occasionally be useful inside
4407 complex nested macros or conditional assembly.
4409 @ifclear no-file-dir
4411 @section @code{.file @var{string}}
4413 @cindex @code{file} directive
4414 @cindex logical file name
4415 @cindex file name, logical
4416 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4417 file. @var{string} is the new file name. In general, the filename is
4418 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4419 to specify an empty file name, you must give the quotes--@code{""}. This
4420 statement may go away in future: it is only recognized to be compatible with
4421 old @command{@value{AS}} programs.
4425 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4427 @cindex @code{fill} directive
4428 @cindex writing patterns in memory
4429 @cindex patterns, writing in memory
4430 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4431 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4432 may be zero or more. @var{Size} may be zero or more, but if it is
4433 more than 8, then it is deemed to have the value 8, compatible with
4434 other people's assemblers. The contents of each @var{repeat} bytes
4435 is taken from an 8-byte number. The highest order 4 bytes are
4436 zero. The lowest order 4 bytes are @var{value} rendered in the
4437 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4438 Each @var{size} bytes in a repetition is taken from the lowest order
4439 @var{size} bytes of this number. Again, this bizarre behavior is
4440 compatible with other people's assemblers.
4442 @var{size} and @var{value} are optional.
4443 If the second comma and @var{value} are absent, @var{value} is
4444 assumed zero. If the first comma and following tokens are absent,
4445 @var{size} is assumed to be 1.
4448 @section @code{.float @var{flonums}}
4450 @cindex floating point numbers (single)
4451 @cindex @code{float} directive
4452 This directive assembles zero or more flonums, separated by commas. It
4453 has the same effect as @code{.single}.
4455 The exact kind of floating point numbers emitted depends on how
4456 @command{@value{AS}} is configured.
4457 @xref{Machine Dependencies}.
4461 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4462 in @sc{ieee} format.
4467 @section @code{.func @var{name}[,@var{label}]}
4468 @cindex @code{func} directive
4469 @code{.func} emits debugging information to denote function @var{name}, and
4470 is ignored unless the file is assembled with debugging enabled.
4471 Only @samp{--gstabs[+]} is currently supported.
4472 @var{label} is the entry point of the function and if omitted @var{name}
4473 prepended with the @samp{leading char} is used.
4474 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4475 All functions are currently defined to have @code{void} return type.
4476 The function must be terminated with @code{.endfunc}.
4479 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4481 @cindex @code{global} directive
4482 @cindex symbol, making visible to linker
4483 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4484 @var{symbol} in your partial program, its value is made available to
4485 other partial programs that are linked with it. Otherwise,
4486 @var{symbol} takes its attributes from a symbol of the same name
4487 from another file linked into the same program.
4489 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4490 compatibility with other assemblers.
4493 On the HPPA, @code{.global} is not always enough to make it accessible to other
4494 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4495 @xref{HPPA Directives,, HPPA Assembler Directives}.
4500 @section @code{.hidden @var{names}}
4502 @cindex @code{hidden} directive
4504 This is one of the ELF visibility directives. The other two are
4505 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4506 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4508 This directive overrides the named symbols default visibility (which is set by
4509 their binding: local, global or weak). The directive sets the visibility to
4510 @code{hidden} which means that the symbols are not visible to other components.
4511 Such symbols are always considered to be @code{protected} as well.
4515 @section @code{.hword @var{expressions}}
4517 @cindex @code{hword} directive
4518 @cindex integers, 16-bit
4519 @cindex numbers, 16-bit
4520 @cindex sixteen bit integers
4521 This expects zero or more @var{expressions}, and emits
4522 a 16 bit number for each.
4525 This directive is a synonym for @samp{.short}; depending on the target
4526 architecture, it may also be a synonym for @samp{.word}.
4530 This directive is a synonym for @samp{.short}.
4533 This directive is a synonym for both @samp{.short} and @samp{.word}.
4538 @section @code{.ident}
4540 @cindex @code{ident} directive
4542 This directive is used by some assemblers to place tags in object files. The
4543 behavior of this directive varies depending on the target. When using the
4544 a.out object file format, @command{@value{AS}} simply accepts the directive for
4545 source-file compatibility with existing assemblers, but does not emit anything
4546 for it. When using COFF, comments are emitted to the @code{.comment} or
4547 @code{.rdata} section, depending on the target. When using ELF, comments are
4548 emitted to the @code{.comment} section.
4551 @section @code{.if @var{absolute expression}}
4553 @cindex conditional assembly
4554 @cindex @code{if} directive
4555 @code{.if} marks the beginning of a section of code which is only
4556 considered part of the source program being assembled if the argument
4557 (which must be an @var{absolute expression}) is non-zero. The end of
4558 the conditional section of code must be marked by @code{.endif}
4559 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4560 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4561 If you have several conditions to check, @code{.elseif} may be used to avoid
4562 nesting blocks if/else within each subsequent @code{.else} block.
4564 The following variants of @code{.if} are also supported:
4566 @cindex @code{ifdef} directive
4567 @item .ifdef @var{symbol}
4568 Assembles the following section of code if the specified @var{symbol}
4569 has been defined. Note a symbol which has been referenced but not yet defined
4570 is considered to be undefined.
4572 @cindex @code{ifb} directive
4573 @item .ifb @var{text}
4574 Assembles the following section of code if the operand is blank (empty).
4576 @cindex @code{ifc} directive
4577 @item .ifc @var{string1},@var{string2}
4578 Assembles the following section of code if the two strings are the same. The
4579 strings may be optionally quoted with single quotes. If they are not quoted,
4580 the first string stops at the first comma, and the second string stops at the
4581 end of the line. Strings which contain whitespace should be quoted. The
4582 string comparison is case sensitive.
4584 @cindex @code{ifeq} directive
4585 @item .ifeq @var{absolute expression}
4586 Assembles the following section of code if the argument is zero.
4588 @cindex @code{ifeqs} directive
4589 @item .ifeqs @var{string1},@var{string2}
4590 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4592 @cindex @code{ifge} directive
4593 @item .ifge @var{absolute expression}
4594 Assembles the following section of code if the argument is greater than or
4597 @cindex @code{ifgt} directive
4598 @item .ifgt @var{absolute expression}
4599 Assembles the following section of code if the argument is greater than zero.
4601 @cindex @code{ifle} directive
4602 @item .ifle @var{absolute expression}
4603 Assembles the following section of code if the argument is less than or equal
4606 @cindex @code{iflt} directive
4607 @item .iflt @var{absolute expression}
4608 Assembles the following section of code if the argument is less than zero.
4610 @cindex @code{ifnb} directive
4611 @item .ifnb @var{text}
4612 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
4613 following section of code if the operand is non-blank (non-empty).
4615 @cindex @code{ifnc} directive
4616 @item .ifnc @var{string1},@var{string2}.
4617 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4618 following section of code if the two strings are not the same.
4620 @cindex @code{ifndef} directive
4621 @cindex @code{ifnotdef} directive
4622 @item .ifndef @var{symbol}
4623 @itemx .ifnotdef @var{symbol}
4624 Assembles the following section of code if the specified @var{symbol}
4625 has not been defined. Both spelling variants are equivalent. Note a symbol
4626 which has been referenced but not yet defined is considered to be undefined.
4628 @cindex @code{ifne} directive
4629 @item .ifne @var{absolute expression}
4630 Assembles the following section of code if the argument is not equal to zero
4631 (in other words, this is equivalent to @code{.if}).
4633 @cindex @code{ifnes} directive
4634 @item .ifnes @var{string1},@var{string2}
4635 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4636 following section of code if the two strings are not the same.
4640 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4642 @cindex @code{incbin} directive
4643 @cindex binary files, including
4644 The @code{incbin} directive includes @var{file} verbatim at the current
4645 location. You can control the search paths used with the @samp{-I} command-line
4646 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4649 The @var{skip} argument skips a number of bytes from the start of the
4650 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4651 read. Note that the data is not aligned in any way, so it is the user's
4652 responsibility to make sure that proper alignment is provided both before and
4653 after the @code{incbin} directive.
4656 @section @code{.include "@var{file}"}
4658 @cindex @code{include} directive
4659 @cindex supporting files, including
4660 @cindex files, including
4661 This directive provides a way to include supporting files at specified
4662 points in your source program. The code from @var{file} is assembled as
4663 if it followed the point of the @code{.include}; when the end of the
4664 included file is reached, assembly of the original file continues. You
4665 can control the search paths used with the @samp{-I} command-line option
4666 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4670 @section @code{.int @var{expressions}}
4672 @cindex @code{int} directive
4673 @cindex integers, 32-bit
4674 Expect zero or more @var{expressions}, of any section, separated by commas.
4675 For each expression, emit a number that, at run time, is the value of that
4676 expression. The byte order and bit size of the number depends on what kind
4677 of target the assembly is for.
4681 On most forms of the H8/300, @code{.int} emits 16-bit
4682 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4689 @section @code{.internal @var{names}}
4691 @cindex @code{internal} directive
4693 This is one of the ELF visibility directives. The other two are
4694 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4695 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4697 This directive overrides the named symbols default visibility (which is set by
4698 their binding: local, global or weak). The directive sets the visibility to
4699 @code{internal} which means that the symbols are considered to be @code{hidden}
4700 (i.e., not visible to other components), and that some extra, processor specific
4701 processing must also be performed upon the symbols as well.
4705 @section @code{.irp @var{symbol},@var{values}}@dots{}
4707 @cindex @code{irp} directive
4708 Evaluate a sequence of statements assigning different values to @var{symbol}.
4709 The sequence of statements starts at the @code{.irp} directive, and is
4710 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4711 set to @var{value}, and the sequence of statements is assembled. If no
4712 @var{value} is listed, the sequence of statements is assembled once, with
4713 @var{symbol} set to the null string. To refer to @var{symbol} within the
4714 sequence of statements, use @var{\symbol}.
4716 For example, assembling
4724 is equivalent to assembling
4732 For some caveats with the spelling of @var{symbol}, see also the discussion
4736 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4738 @cindex @code{irpc} directive
4739 Evaluate a sequence of statements assigning different values to @var{symbol}.
4740 The sequence of statements starts at the @code{.irpc} directive, and is
4741 terminated by an @code{.endr} directive. For each character in @var{value},
4742 @var{symbol} is set to the character, and the sequence of statements is
4743 assembled. If no @var{value} is listed, the sequence of statements is
4744 assembled once, with @var{symbol} set to the null string. To refer to
4745 @var{symbol} within the sequence of statements, use @var{\symbol}.
4747 For example, assembling
4755 is equivalent to assembling
4763 For some caveats with the spelling of @var{symbol}, see also the discussion
4767 @section @code{.lcomm @var{symbol} , @var{length}}
4769 @cindex @code{lcomm} directive
4770 @cindex local common symbols
4771 @cindex symbols, local common
4772 Reserve @var{length} (an absolute expression) bytes for a local common
4773 denoted by @var{symbol}. The section and value of @var{symbol} are
4774 those of the new local common. The addresses are allocated in the bss
4775 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4776 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4777 not visible to @code{@value{LD}}.
4780 Some targets permit a third argument to be used with @code{.lcomm}. This
4781 argument specifies the desired alignment of the symbol in the bss section.
4785 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4786 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4790 @section @code{.lflags}
4792 @cindex @code{lflags} directive (ignored)
4793 @command{@value{AS}} accepts this directive, for compatibility with other
4794 assemblers, but ignores it.
4796 @ifclear no-line-dir
4798 @section @code{.line @var{line-number}}
4800 @cindex @code{line} directive
4804 @section @code{.ln @var{line-number}}
4806 @cindex @code{ln} directive
4808 @cindex logical line number
4810 Change the logical line number. @var{line-number} must be an absolute
4811 expression. The next line has that logical line number. Therefore any other
4812 statements on the current line (after a statement separator character) are
4813 reported as on logical line number @var{line-number} @minus{} 1. One day
4814 @command{@value{AS}} will no longer support this directive: it is recognized only
4815 for compatibility with existing assembler programs.
4819 @ifclear no-line-dir
4820 Even though this is a directive associated with the @code{a.out} or
4821 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4822 when producing COFF output, and treats @samp{.line} as though it
4823 were the COFF @samp{.ln} @emph{if} it is found outside a
4824 @code{.def}/@code{.endef} pair.
4826 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4827 used by compilers to generate auxiliary symbol information for
4832 @section @code{.linkonce [@var{type}]}
4834 @cindex @code{linkonce} directive
4835 @cindex common sections
4836 Mark the current section so that the linker only includes a single copy of it.
4837 This may be used to include the same section in several different object files,
4838 but ensure that the linker will only include it once in the final output file.
4839 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4840 Duplicate sections are detected based on the section name, so it should be
4843 This directive is only supported by a few object file formats; as of this
4844 writing, the only object file format which supports it is the Portable
4845 Executable format used on Windows NT.
4847 The @var{type} argument is optional. If specified, it must be one of the
4848 following strings. For example:
4852 Not all types may be supported on all object file formats.
4856 Silently discard duplicate sections. This is the default.
4859 Warn if there are duplicate sections, but still keep only one copy.
4862 Warn if any of the duplicates have different sizes.
4865 Warn if any of the duplicates do not have exactly the same contents.
4869 @section @code{.ln @var{line-number}}
4871 @cindex @code{ln} directive
4872 @ifclear no-line-dir
4873 @samp{.ln} is a synonym for @samp{.line}.
4876 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4877 must be an absolute expression. The next line has that logical
4878 line number, so any other statements on the current line (after a
4879 statement separator character @code{;}) are reported as on logical
4880 line number @var{line-number} @minus{} 1.
4883 This directive is accepted, but ignored, when @command{@value{AS}} is
4884 configured for @code{b.out}; its effect is only associated with COFF
4890 @section @code{.mri @var{val}}
4892 @cindex @code{mri} directive
4893 @cindex MRI mode, temporarily
4894 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4895 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4896 affects code assembled until the next @code{.mri} directive, or until the end
4897 of the file. @xref{M, MRI mode, MRI mode}.
4900 @section @code{.list}
4902 @cindex @code{list} directive
4903 @cindex listing control, turning on
4904 Control (in conjunction with the @code{.nolist} directive) whether or
4905 not assembly listings are generated. These two directives maintain an
4906 internal counter (which is zero initially). @code{.list} increments the
4907 counter, and @code{.nolist} decrements it. Assembly listings are
4908 generated whenever the counter is greater than zero.
4910 By default, listings are disabled. When you enable them (with the
4911 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4912 the initial value of the listing counter is one.
4915 @section @code{.long @var{expressions}}
4917 @cindex @code{long} directive
4918 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4921 @c no one seems to know what this is for or whether this description is
4922 @c what it really ought to do
4924 @section @code{.lsym @var{symbol}, @var{expression}}
4926 @cindex @code{lsym} directive
4927 @cindex symbol, not referenced in assembly
4928 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4929 the hash table, ensuring it cannot be referenced by name during the
4930 rest of the assembly. This sets the attributes of the symbol to be
4931 the same as the expression value:
4933 @var{other} = @var{descriptor} = 0
4934 @var{type} = @r{(section of @var{expression})}
4935 @var{value} = @var{expression}
4938 The new symbol is not flagged as external.
4942 @section @code{.macro}
4945 The commands @code{.macro} and @code{.endm} allow you to define macros that
4946 generate assembly output. For example, this definition specifies a macro
4947 @code{sum} that puts a sequence of numbers into memory:
4950 .macro sum from=0, to=5
4959 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4971 @item .macro @var{macname}
4972 @itemx .macro @var{macname} @var{macargs} @dots{}
4973 @cindex @code{macro} directive
4974 Begin the definition of a macro called @var{macname}. If your macro
4975 definition requires arguments, specify their names after the macro name,
4976 separated by commas or spaces. You can qualify the macro argument to
4977 indicate whether all invocations must specify a non-blank value (through
4978 @samp{:@code{req}}), or whether it takes all of the remaining arguments
4979 (through @samp{:@code{vararg}}). You can supply a default value for any
4980 macro argument by following the name with @samp{=@var{deflt}}. You
4981 cannot define two macros with the same @var{macname} unless it has been
4982 subject to the @code{.purgem} directive (@xref{Purgem}.) between the two
4983 definitions. For example, these are all valid @code{.macro} statements:
4987 Begin the definition of a macro called @code{comm}, which takes no
4990 @item .macro plus1 p, p1
4991 @itemx .macro plus1 p p1
4992 Either statement begins the definition of a macro called @code{plus1},
4993 which takes two arguments; within the macro definition, write
4994 @samp{\p} or @samp{\p1} to evaluate the arguments.
4996 @item .macro reserve_str p1=0 p2
4997 Begin the definition of a macro called @code{reserve_str}, with two
4998 arguments. The first argument has a default value, but not the second.
4999 After the definition is complete, you can call the macro either as
5000 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5001 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5002 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5003 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5006 @item .macro m p1:req, p2=0, p3:vararg
5007 Begin the definition of a macro called @code{m}, with at least three
5008 arguments. The first argument must always have a value specified, but
5009 not the second, which instead has a default value. The third formal
5010 will get assigned all remaining arguments specified at invocation time.
5012 When you call a macro, you can specify the argument values either by
5013 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5014 @samp{sum to=17, from=9}.
5016 Note that since each of the @var{macargs} can be an identifier exactly
5017 as any other one permitted by the target architecture, there may be
5018 occasional problems if the target hand-crafts special meanings to certain
5019 characters when they occur in a special position. For example, if colon
5020 (@code{:}) is generally permitted to be part of a symbol name, but the
5021 architecture specific code special-cases it when occuring as the final
5022 character of a symbol (to denote a label), then the macro parameter
5023 replacement code will have no way of knowing that and consider the whole
5024 construct (including the colon) an identifier, and check only this
5025 identifier for being the subject to parameter substitution. In this
5026 example, besides the potential of just separating identifier and colon
5027 by white space, using alternate macro syntax (@xref{Altmacro}.) and
5028 ampersand (@code{&}) as the character to separate literal text from macro
5029 parameters (or macro parameters from one another) would provide a way to
5030 achieve the same effect:
5039 This applies identically to the identifiers used in @code{.irp} (@xref{Irp}.)
5040 and @code{.irpc} (@xref{Irpc}.).
5043 @cindex @code{endm} directive
5044 Mark the end of a macro definition.
5047 @cindex @code{exitm} directive
5048 Exit early from the current macro definition.
5050 @cindex number of macros executed
5051 @cindex macros, count executed
5053 @command{@value{AS}} maintains a counter of how many macros it has
5054 executed in this pseudo-variable; you can copy that number to your
5055 output with @samp{\@@}, but @emph{only within a macro definition}.
5057 @item LOCAL @var{name} [ , @dots{} ]
5058 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5059 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5060 @xref{Altmacro,,@code{.altmacro}}.
5064 @section @code{.altmacro}
5065 Enable alternate macro mode, enabling:
5068 @item LOCAL @var{name} [ , @dots{} ]
5069 One additional directive, @code{LOCAL}, is available. It is used to
5070 generate a string replacement for each of the @var{name} arguments, and
5071 replace any instances of @var{name} in each macro expansion. The
5072 replacement string is unique in the assembly, and different for each
5073 separate macro expansion. @code{LOCAL} allows you to write macros that
5074 define symbols, without fear of conflict between separate macro expansions.
5076 @item String delimiters
5077 You can write strings delimited in these other ways besides
5078 @code{"@var{string}"}:
5081 @item '@var{string}'
5082 You can delimit strings with single-quote charaters.
5084 @item <@var{string}>
5085 You can delimit strings with matching angle brackets.
5088 @item single-character string escape
5089 To include any single character literally in a string (even if the
5090 character would otherwise have some special meaning), you can prefix the
5091 character with @samp{!} (an exclamation mark). For example, you can
5092 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
5094 @item Expression results as strings
5095 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
5096 and use the result as a string.
5100 @section @code{.noaltmacro}
5101 Disable alternate macro mode. @ref{Altmacro}
5104 @section @code{.nolist}
5106 @cindex @code{nolist} directive
5107 @cindex listing control, turning off
5108 Control (in conjunction with the @code{.list} directive) whether or
5109 not assembly listings are generated. These two directives maintain an
5110 internal counter (which is zero initially). @code{.list} increments the
5111 counter, and @code{.nolist} decrements it. Assembly listings are
5112 generated whenever the counter is greater than zero.
5115 @section @code{.octa @var{bignums}}
5117 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5118 @cindex @code{octa} directive
5119 @cindex integer, 16-byte
5120 @cindex sixteen byte integer
5121 This directive expects zero or more bignums, separated by commas. For each
5122 bignum, it emits a 16-byte integer.
5124 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5125 hence @emph{octa}-word for 16 bytes.
5128 @section @code{.org @var{new-lc} , @var{fill}}
5130 @cindex @code{org} directive
5131 @cindex location counter, advancing
5132 @cindex advancing location counter
5133 @cindex current address, advancing
5134 Advance the location counter of the current section to
5135 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5136 expression with the same section as the current subsection. That is,
5137 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5138 wrong section, the @code{.org} directive is ignored. To be compatible
5139 with former assemblers, if the section of @var{new-lc} is absolute,
5140 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5141 is the same as the current subsection.
5143 @code{.org} may only increase the location counter, or leave it
5144 unchanged; you cannot use @code{.org} to move the location counter
5147 @c double negative used below "not undefined" because this is a specific
5148 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5149 @c section. doc@cygnus.com 18feb91
5150 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5151 may not be undefined. If you really detest this restriction we eagerly await
5152 a chance to share your improved assembler.
5154 Beware that the origin is relative to the start of the section, not
5155 to the start of the subsection. This is compatible with other
5156 people's assemblers.
5158 When the location counter (of the current subsection) is advanced, the
5159 intervening bytes are filled with @var{fill} which should be an
5160 absolute expression. If the comma and @var{fill} are omitted,
5161 @var{fill} defaults to zero.
5164 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5166 @cindex padding the location counter given a power of two
5167 @cindex @code{p2align} directive
5168 Pad the location counter (in the current subsection) to a particular
5169 storage boundary. The first expression (which must be absolute) is the
5170 number of low-order zero bits the location counter must have after
5171 advancement. For example @samp{.p2align 3} advances the location
5172 counter until it a multiple of 8. If the location counter is already a
5173 multiple of 8, no change is needed.
5175 The second expression (also absolute) gives the fill value to be stored in the
5176 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5177 padding bytes are normally zero. However, on some systems, if the section is
5178 marked as containing code and the fill value is omitted, the space is filled
5179 with no-op instructions.
5181 The third expression is also absolute, and is also optional. If it is present,
5182 it is the maximum number of bytes that should be skipped by this alignment
5183 directive. If doing the alignment would require skipping more bytes than the
5184 specified maximum, then the alignment is not done at all. You can omit the
5185 fill value (the second argument) entirely by simply using two commas after the
5186 required alignment; this can be useful if you want the alignment to be filled
5187 with no-op instructions when appropriate.
5189 @cindex @code{p2alignw} directive
5190 @cindex @code{p2alignl} directive
5191 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5192 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5193 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5194 fill pattern as a four byte longword value. For example, @code{.p2alignw
5195 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5196 filled in with the value 0x368d (the exact placement of the bytes depends upon
5197 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5202 @section @code{.previous}
5204 @cindex @code{previous} directive
5205 @cindex Section Stack
5206 This is one of the ELF section stack manipulation directives. The others are
5207 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5208 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5209 (@pxref{PopSection}).
5211 This directive swaps the current section (and subsection) with most recently
5212 referenced section (and subsection) prior to this one. Multiple
5213 @code{.previous} directives in a row will flip between two sections (and their
5216 In terms of the section stack, this directive swaps the current section with
5217 the top section on the section stack.
5222 @section @code{.popsection}
5224 @cindex @code{popsection} directive
5225 @cindex Section Stack
5226 This is one of the ELF section stack manipulation directives. The others are
5227 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5228 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5231 This directive replaces the current section (and subsection) with the top
5232 section (and subsection) on the section stack. This section is popped off the
5237 @section @code{.print @var{string}}
5239 @cindex @code{print} directive
5240 @command{@value{AS}} will print @var{string} on the standard output during
5241 assembly. You must put @var{string} in double quotes.
5245 @section @code{.protected @var{names}}
5247 @cindex @code{protected} directive
5249 This is one of the ELF visibility directives. The other two are
5250 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5252 This directive overrides the named symbols default visibility (which is set by
5253 their binding: local, global or weak). The directive sets the visibility to
5254 @code{protected} which means that any references to the symbols from within the
5255 components that defines them must be resolved to the definition in that
5256 component, even if a definition in another component would normally preempt
5261 @section @code{.psize @var{lines} , @var{columns}}
5263 @cindex @code{psize} directive
5264 @cindex listing control: paper size
5265 @cindex paper size, for listings
5266 Use this directive to declare the number of lines---and, optionally, the
5267 number of columns---to use for each page, when generating listings.
5269 If you do not use @code{.psize}, listings use a default line-count
5270 of 60. You may omit the comma and @var{columns} specification; the
5271 default width is 200 columns.
5273 @command{@value{AS}} generates formfeeds whenever the specified number of
5274 lines is exceeded (or whenever you explicitly request one, using
5277 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5278 those explicitly specified with @code{.eject}.
5281 @section @code{.purgem @var{name}}
5283 @cindex @code{purgem} directive
5284 Undefine the macro @var{name}, so that later uses of the string will not be
5285 expanded. @xref{Macro}.
5289 @section @code{.pushsection @var{name} , @var{subsection}}
5291 @cindex @code{pushsection} directive
5292 @cindex Section Stack
5293 This is one of the ELF section stack manipulation directives. The others are
5294 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5295 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5298 This directive pushes the current section (and subsection) onto the
5299 top of the section stack, and then replaces the current section and
5300 subsection with @code{name} and @code{subsection}.
5304 @section @code{.quad @var{bignums}}
5306 @cindex @code{quad} directive
5307 @code{.quad} expects zero or more bignums, separated by commas. For
5308 each bignum, it emits
5310 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5311 warning message; and just takes the lowest order 8 bytes of the bignum.
5312 @cindex eight-byte integer
5313 @cindex integer, 8-byte
5315 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5316 hence @emph{quad}-word for 8 bytes.
5319 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5320 warning message; and just takes the lowest order 16 bytes of the bignum.
5321 @cindex sixteen-byte integer
5322 @cindex integer, 16-byte
5326 @section @code{.rept @var{count}}
5328 @cindex @code{rept} directive
5329 Repeat the sequence of lines between the @code{.rept} directive and the next
5330 @code{.endr} directive @var{count} times.
5332 For example, assembling
5340 is equivalent to assembling
5349 @section @code{.sbttl "@var{subheading}"}
5351 @cindex @code{sbttl} directive
5352 @cindex subtitles for listings
5353 @cindex listing control: subtitle
5354 Use @var{subheading} as the title (third line, immediately after the
5355 title line) when generating assembly listings.
5357 This directive affects subsequent pages, as well as the current page if
5358 it appears within ten lines of the top of a page.
5362 @section @code{.scl @var{class}}
5364 @cindex @code{scl} directive
5365 @cindex symbol storage class (COFF)
5366 @cindex COFF symbol storage class
5367 Set the storage-class value for a symbol. This directive may only be
5368 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5369 whether a symbol is static or external, or it may record further
5370 symbolic debugging information.
5373 The @samp{.scl} directive is primarily associated with COFF output; when
5374 configured to generate @code{b.out} output format, @command{@value{AS}}
5375 accepts this directive but ignores it.
5381 @section @code{.section @var{name}}
5383 @cindex named section
5384 Use the @code{.section} directive to assemble the following code into a section
5387 This directive is only supported for targets that actually support arbitrarily
5388 named sections; on @code{a.out} targets, for example, it is not accepted, even
5389 with a standard @code{a.out} section name.
5393 @c only print the extra heading if both COFF and ELF are set
5394 @subheading COFF Version
5397 @cindex @code{section} directive (COFF version)
5398 For COFF targets, the @code{.section} directive is used in one of the following
5402 .section @var{name}[, "@var{flags}"]
5403 .section @var{name}[, @var{subsegment}]
5406 If the optional argument is quoted, it is taken as flags to use for the
5407 section. Each flag is a single character. The following flags are recognized:
5410 bss section (uninitialized data)
5412 section is not loaded
5422 shared section (meaningful for PE targets)
5424 ignored. (For compatibility with the ELF version)
5427 If no flags are specified, the default flags depend upon the section name. If
5428 the section name is not recognized, the default will be for the section to be
5429 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5430 from the section, rather than adding them, so if they are used on their own it
5431 will be as if no flags had been specified at all.
5433 If the optional argument to the @code{.section} directive is not quoted, it is
5434 taken as a subsegment number (@pxref{Sub-Sections}).
5439 @c only print the extra heading if both COFF and ELF are set
5440 @subheading ELF Version
5443 @cindex Section Stack
5444 This is one of the ELF section stack manipulation directives. The others are
5445 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5446 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5447 @code{.previous} (@pxref{Previous}).
5449 @cindex @code{section} directive (ELF version)
5450 For ELF targets, the @code{.section} directive is used like this:
5453 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
5456 The optional @var{flags} argument is a quoted string which may contain any
5457 combination of the following characters:
5460 section is allocatable
5464 section is executable
5466 section is mergeable
5468 section contains zero terminated strings
5470 section is a member of a section group
5472 section is used for thread-local-storage
5475 The optional @var{type} argument may contain one of the following constants:
5478 section contains data
5480 section does not contain data (i.e., section only occupies space)
5482 section contains data which is used by things other than the program
5484 section contains an array of pointers to init functions
5486 section contains an array of pointers to finish functions
5487 @item @@preinit_array
5488 section contains an array of pointers to pre-init functions
5491 Many targets only support the first three section types.
5493 Note on targets where the @code{@@} character is the start of a comment (eg
5494 ARM) then another character is used instead. For example the ARM port uses the
5497 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
5498 be specified as well as an extra argument - @var{entsize} - like this:
5501 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
5504 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
5505 constants, each @var{entsize} octets long. Sections with both @code{M} and
5506 @code{S} must contain zero terminated strings where each character is
5507 @var{entsize} bytes long. The linker may remove duplicates within sections with
5508 the same name, same entity size and same flags. @var{entsize} must be an
5509 absolute expression.
5511 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
5512 be present along with an additional field like this:
5515 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
5518 The @var{GroupName} field specifies the name of the section group to which this
5519 particular section belongs. The optional linkage field can contain:
5522 indicates that only one copy of this section should be retained
5527 Note - if both the @var{M} and @var{G} flags are present then the fields for
5528 the Merge flag should come first, like this:
5531 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
5534 If no flags are specified, the default flags depend upon the section name. If
5535 the section name is not recognized, the default will be for the section to have
5536 none of the above flags: it will not be allocated in memory, nor writable, nor
5537 executable. The section will contain data.
5539 For ELF targets, the assembler supports another type of @code{.section}
5540 directive for compatibility with the Solaris assembler:
5543 .section "@var{name}"[, @var{flags}...]
5546 Note that the section name is quoted. There may be a sequence of comma
5550 section is allocatable
5554 section is executable
5556 section is used for thread local storage
5559 This directive replaces the current section and subsection. See the
5560 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
5561 some examples of how this directive and the other section stack directives
5567 @section @code{.set @var{symbol}, @var{expression}}
5569 @cindex @code{set} directive
5570 @cindex symbol value, setting
5571 Set the value of @var{symbol} to @var{expression}. This
5572 changes @var{symbol}'s value and type to conform to
5573 @var{expression}. If @var{symbol} was flagged as external, it remains
5574 flagged (@pxref{Symbol Attributes}).
5576 You may @code{.set} a symbol many times in the same assembly.
5578 If you @code{.set} a global symbol, the value stored in the object
5579 file is the last value stored into it.
5582 The syntax for @code{set} on the HPPA is
5583 @samp{@var{symbol} .set @var{expression}}.
5587 On Z80 @code{set} is a real instruction, use
5588 @samp{@var{symbol} defl @var{expression}} instead.
5592 @section @code{.short @var{expressions}}
5594 @cindex @code{short} directive
5596 @code{.short} is normally the same as @samp{.word}.
5597 @xref{Word,,@code{.word}}.
5599 In some configurations, however, @code{.short} and @code{.word} generate
5600 numbers of different lengths; @pxref{Machine Dependencies}.
5604 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5607 This expects zero or more @var{expressions}, and emits
5608 a 16 bit number for each.
5613 @section @code{.single @var{flonums}}
5615 @cindex @code{single} directive
5616 @cindex floating point numbers (single)
5617 This directive assembles zero or more flonums, separated by commas. It
5618 has the same effect as @code{.float}.
5620 The exact kind of floating point numbers emitted depends on how
5621 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5625 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5626 numbers in @sc{ieee} format.
5632 @section @code{.size}
5634 This directive is used to set the size associated with a symbol.
5638 @c only print the extra heading if both COFF and ELF are set
5639 @subheading COFF Version
5642 @cindex @code{size} directive (COFF version)
5643 For COFF targets, the @code{.size} directive is only permitted inside
5644 @code{.def}/@code{.endef} pairs. It is used like this:
5647 .size @var{expression}
5651 @samp{.size} is only meaningful when generating COFF format output; when
5652 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5659 @c only print the extra heading if both COFF and ELF are set
5660 @subheading ELF Version
5663 @cindex @code{size} directive (ELF version)
5664 For ELF targets, the @code{.size} directive is used like this:
5667 .size @var{name} , @var{expression}
5670 This directive sets the size associated with a symbol @var{name}.
5671 The size in bytes is computed from @var{expression} which can make use of label
5672 arithmetic. This directive is typically used to set the size of function
5678 @section @code{.sleb128 @var{expressions}}
5680 @cindex @code{sleb128} directive
5681 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5682 compact, variable length representation of numbers used by the DWARF
5683 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
5685 @ifclear no-space-dir
5687 @section @code{.skip @var{size} , @var{fill}}
5689 @cindex @code{skip} directive
5690 @cindex filling memory
5691 This directive emits @var{size} bytes, each of value @var{fill}. Both
5692 @var{size} and @var{fill} are absolute expressions. If the comma and
5693 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5697 @section @code{.space @var{size} , @var{fill}}
5699 @cindex @code{space} directive
5700 @cindex filling memory
5701 This directive emits @var{size} bytes, each of value @var{fill}. Both
5702 @var{size} and @var{fill} are absolute expressions. If the comma
5703 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5708 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5709 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5710 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5711 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5719 @section @code{.stabd, .stabn, .stabs}
5721 @cindex symbolic debuggers, information for
5722 @cindex @code{stab@var{x}} directives
5723 There are three directives that begin @samp{.stab}.
5724 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5725 The symbols are not entered in the @command{@value{AS}} hash table: they
5726 cannot be referenced elsewhere in the source file.
5727 Up to five fields are required:
5731 This is the symbol's name. It may contain any character except
5732 @samp{\000}, so is more general than ordinary symbol names. Some
5733 debuggers used to code arbitrarily complex structures into symbol names
5737 An absolute expression. The symbol's type is set to the low 8 bits of
5738 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5739 and debuggers choke on silly bit patterns.
5742 An absolute expression. The symbol's ``other'' attribute is set to the
5743 low 8 bits of this expression.
5746 An absolute expression. The symbol's descriptor is set to the low 16
5747 bits of this expression.
5750 An absolute expression which becomes the symbol's value.
5753 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5754 or @code{.stabs} statement, the symbol has probably already been created;
5755 you get a half-formed symbol in your object file. This is
5756 compatible with earlier assemblers!
5759 @cindex @code{stabd} directive
5760 @item .stabd @var{type} , @var{other} , @var{desc}
5762 The ``name'' of the symbol generated is not even an empty string.
5763 It is a null pointer, for compatibility. Older assemblers used a
5764 null pointer so they didn't waste space in object files with empty
5767 The symbol's value is set to the location counter,
5768 relocatably. When your program is linked, the value of this symbol
5769 is the address of the location counter when the @code{.stabd} was
5772 @cindex @code{stabn} directive
5773 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5774 The name of the symbol is set to the empty string @code{""}.
5776 @cindex @code{stabs} directive
5777 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5778 All five fields are specified.
5784 @section @code{.string} "@var{str}"
5786 @cindex string, copying to object file
5787 @cindex @code{string} directive
5789 Copy the characters in @var{str} to the object file. You may specify more than
5790 one string to copy, separated by commas. Unless otherwise specified for a
5791 particular machine, the assembler marks the end of each string with a 0 byte.
5792 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5795 @section @code{.struct @var{expression}}
5797 @cindex @code{struct} directive
5798 Switch to the absolute section, and set the section offset to @var{expression},
5799 which must be an absolute expression. You might use this as follows:
5808 This would define the symbol @code{field1} to have the value 0, the symbol
5809 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5810 value 8. Assembly would be left in the absolute section, and you would need to
5811 use a @code{.section} directive of some sort to change to some other section
5812 before further assembly.
5816 @section @code{.subsection @var{name}}
5818 @cindex @code{subsection} directive
5819 @cindex Section Stack
5820 This is one of the ELF section stack manipulation directives. The others are
5821 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5822 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5825 This directive replaces the current subsection with @code{name}. The current
5826 section is not changed. The replaced subsection is put onto the section stack
5827 in place of the then current top of stack subsection.
5832 @section @code{.symver}
5833 @cindex @code{symver} directive
5834 @cindex symbol versioning
5835 @cindex versions of symbols
5836 Use the @code{.symver} directive to bind symbols to specific version nodes
5837 within a source file. This is only supported on ELF platforms, and is
5838 typically used when assembling files to be linked into a shared library.
5839 There are cases where it may make sense to use this in objects to be bound
5840 into an application itself so as to override a versioned symbol from a
5843 For ELF targets, the @code{.symver} directive can be used like this:
5845 .symver @var{name}, @var{name2@@nodename}
5847 If the symbol @var{name} is defined within the file
5848 being assembled, the @code{.symver} directive effectively creates a symbol
5849 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5850 just don't try and create a regular alias is that the @var{@@} character isn't
5851 permitted in symbol names. The @var{name2} part of the name is the actual name
5852 of the symbol by which it will be externally referenced. The name @var{name}
5853 itself is merely a name of convenience that is used so that it is possible to
5854 have definitions for multiple versions of a function within a single source
5855 file, and so that the compiler can unambiguously know which version of a
5856 function is being mentioned. The @var{nodename} portion of the alias should be
5857 the name of a node specified in the version script supplied to the linker when
5858 building a shared library. If you are attempting to override a versioned
5859 symbol from a shared library, then @var{nodename} should correspond to the
5860 nodename of the symbol you are trying to override.
5862 If the symbol @var{name} is not defined within the file being assembled, all
5863 references to @var{name} will be changed to @var{name2@@nodename}. If no
5864 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5867 Another usage of the @code{.symver} directive is:
5869 .symver @var{name}, @var{name2@@@@nodename}
5871 In this case, the symbol @var{name} must exist and be defined within
5872 the file being assembled. It is similar to @var{name2@@nodename}. The
5873 difference is @var{name2@@@@nodename} will also be used to resolve
5874 references to @var{name2} by the linker.
5876 The third usage of the @code{.symver} directive is:
5878 .symver @var{name}, @var{name2@@@@@@nodename}
5880 When @var{name} is not defined within the
5881 file being assembled, it is treated as @var{name2@@nodename}. When
5882 @var{name} is defined within the file being assembled, the symbol
5883 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5888 @section @code{.tag @var{structname}}
5890 @cindex COFF structure debugging
5891 @cindex structure debugging, COFF
5892 @cindex @code{tag} directive
5893 This directive is generated by compilers to include auxiliary debugging
5894 information in the symbol table. It is only permitted inside
5895 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5896 definitions in the symbol table with instances of those structures.
5899 @samp{.tag} is only used when generating COFF format output; when
5900 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5906 @section @code{.text @var{subsection}}
5908 @cindex @code{text} directive
5909 Tells @command{@value{AS}} to assemble the following statements onto the end of
5910 the text subsection numbered @var{subsection}, which is an absolute
5911 expression. If @var{subsection} is omitted, subsection number zero
5915 @section @code{.title "@var{heading}"}
5917 @cindex @code{title} directive
5918 @cindex listing control: title line
5919 Use @var{heading} as the title (second line, immediately after the
5920 source file name and pagenumber) when generating assembly listings.
5922 This directive affects subsequent pages, as well as the current page if
5923 it appears within ten lines of the top of a page.
5927 @section @code{.type}
5929 This directive is used to set the type of a symbol.
5933 @c only print the extra heading if both COFF and ELF are set
5934 @subheading COFF Version
5937 @cindex COFF symbol type
5938 @cindex symbol type, COFF
5939 @cindex @code{type} directive (COFF version)
5940 For COFF targets, this directive is permitted only within
5941 @code{.def}/@code{.endef} pairs. It is used like this:
5947 This records the integer @var{int} as the type attribute of a symbol table
5951 @samp{.type} is associated only with COFF format output; when
5952 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
5953 directive but ignores it.
5959 @c only print the extra heading if both COFF and ELF are set
5960 @subheading ELF Version
5963 @cindex ELF symbol type
5964 @cindex symbol type, ELF
5965 @cindex @code{type} directive (ELF version)
5966 For ELF targets, the @code{.type} directive is used like this:
5969 .type @var{name} , @var{type description}
5972 This sets the type of symbol @var{name} to be either a
5973 function symbol or an object symbol. There are five different syntaxes
5974 supported for the @var{type description} field, in order to provide
5975 compatibility with various other assemblers.
5977 Because some of the characters used in these syntaxes (such as @samp{@@} and
5978 @samp{#}) are comment characters for some architectures, some of the syntaxes
5979 below do not work on all architectures. The first variant will be accepted by
5980 the GNU assembler on all architectures so that variant should be used for
5981 maximum portability, if you do not need to assemble your code with other
5984 The syntaxes supported are:
5987 .type <name> STT_FUNCTION
5988 .type <name> STT_OBJECT
5990 .type <name>,#function
5991 .type <name>,#object
5993 .type <name>,@@function
5994 .type <name>,@@object
5996 .type <name>,%function
5997 .type <name>,%object
5999 .type <name>,"function"
6000 .type <name>,"object"
6006 @section @code{.uleb128 @var{expressions}}
6008 @cindex @code{uleb128} directive
6009 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6010 compact, variable length representation of numbers used by the DWARF
6011 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
6015 @section @code{.val @var{addr}}
6017 @cindex @code{val} directive
6018 @cindex COFF value attribute
6019 @cindex value attribute, COFF
6020 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6021 records the address @var{addr} as the value attribute of a symbol table
6025 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6026 configured for @code{b.out}, it accepts this directive but ignores it.
6032 @section @code{.version "@var{string}"}
6034 @cindex @code{version} directive
6035 This directive creates a @code{.note} section and places into it an ELF
6036 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6041 @section @code{.vtable_entry @var{table}, @var{offset}}
6043 @cindex @code{vtable_entry} directive
6044 This directive finds or creates a symbol @code{table} and creates a
6045 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6048 @section @code{.vtable_inherit @var{child}, @var{parent}}
6050 @cindex @code{vtable_inherit} directive
6051 This directive finds the symbol @code{child} and finds or creates the symbol
6052 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6053 parent whose addend is the value of the child symbol. As a special case the
6054 parent name of @code{0} is treated as refering the @code{*ABS*} section.
6058 @section @code{.warning "@var{string}"}
6059 @cindex warning directive
6060 Similar to the directive @code{.error}
6061 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6064 @section @code{.weak @var{names}}
6066 @cindex @code{weak} directive
6067 This directive sets the weak attribute on the comma separated list of symbol
6068 @code{names}. If the symbols do not already exist, they will be created.
6070 On COFF targets other than PE, weak symbols are a GNU extension. This
6071 directive sets the weak attribute on the comma separated list of symbol
6072 @code{names}. If the symbols do not already exist, they will be created.
6074 On the PE target, weak symbols are supported natively as weak aliases.
6075 When a weak symbol is created that is not an alias, GAS creates an
6076 alternate symbol to hold the default value.
6079 @section @code{.weakref @var{alias}, @var{target}}
6081 @cindex @code{weakref} directive
6082 This directive creates an alias to the target symbol that enables the symbol to
6083 be referenced with weak-symbol semantics, but without actually making it weak.
6084 If direct references or definitions of the symbol are present, then the symbol
6085 will not be weak, but if all references to it are through weak references, the
6086 symbol will be marked as weak in the symbol table.
6088 The effect is equivalent to moving all references to the alias to a separate
6089 assembly source file, renaming the alias to the symbol in it, declaring the
6090 symbol as weak there, and running a reloadable link to merge the object files
6091 resulting from the assembly of the new source file and the old source file that
6092 had the references to the alias removed.
6094 The alias itself never makes to the symbol table, and is entirely handled
6095 within the assembler.
6098 @section @code{.word @var{expressions}}
6100 @cindex @code{word} directive
6101 This directive expects zero or more @var{expressions}, of any section,
6102 separated by commas.
6105 For each expression, @command{@value{AS}} emits a 32-bit number.
6108 For each expression, @command{@value{AS}} emits a 16-bit number.
6113 The size of the number emitted, and its byte order,
6114 depend on what target computer the assembly is for.
6117 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6118 @c happen---32-bit addressability, period; no long/short jumps.
6119 @ifset DIFF-TBL-KLUGE
6120 @cindex difference tables altered
6121 @cindex altered difference tables
6123 @emph{Warning: Special Treatment to support Compilers}
6127 Machines with a 32-bit address space, but that do less than 32-bit
6128 addressing, require the following special treatment. If the machine of
6129 interest to you does 32-bit addressing (or doesn't require it;
6130 @pxref{Machine Dependencies}), you can ignore this issue.
6133 In order to assemble compiler output into something that works,
6134 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6135 Directives of the form @samp{.word sym1-sym2} are often emitted by
6136 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6137 directive of the form @samp{.word sym1-sym2}, and the difference between
6138 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6139 creates a @dfn{secondary jump table}, immediately before the next label.
6140 This secondary jump table is preceded by a short-jump to the
6141 first byte after the secondary table. This short-jump prevents the flow
6142 of control from accidentally falling into the new table. Inside the
6143 table is a long-jump to @code{sym2}. The original @samp{.word}
6144 contains @code{sym1} minus the address of the long-jump to
6147 If there were several occurrences of @samp{.word sym1-sym2} before the
6148 secondary jump table, all of them are adjusted. If there was a
6149 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6150 long-jump to @code{sym4} is included in the secondary jump table,
6151 and the @code{.word} directives are adjusted to contain @code{sym3}
6152 minus the address of the long-jump to @code{sym4}; and so on, for as many
6153 entries in the original jump table as necessary.
6156 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6157 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6158 assembly language programmers.
6161 @c end DIFF-TBL-KLUGE
6164 @section Deprecated Directives
6166 @cindex deprecated directives
6167 @cindex obsolescent directives
6168 One day these directives won't work.
6169 They are included for compatibility with older assemblers.
6176 @node Machine Dependencies
6177 @chapter Machine Dependent Features
6179 @cindex machine dependencies
6180 The machine instruction sets are (almost by definition) different on
6181 each machine where @command{@value{AS}} runs. Floating point representations
6182 vary as well, and @command{@value{AS}} often supports a few additional
6183 directives or command-line options for compatibility with other
6184 assemblers on a particular platform. Finally, some versions of
6185 @command{@value{AS}} support special pseudo-instructions for branch
6188 This chapter discusses most of these differences, though it does not
6189 include details on any machine's instruction set. For details on that
6190 subject, see the hardware manufacturer's manual.
6194 * Alpha-Dependent:: Alpha Dependent Features
6197 * ARC-Dependent:: ARC Dependent Features
6200 * ARM-Dependent:: ARM Dependent Features
6203 * AVR-Dependent:: AVR Dependent Features
6206 * BFIN-Dependent:: BFIN Dependent Features
6209 * CRIS-Dependent:: CRIS Dependent Features
6212 * D10V-Dependent:: D10V Dependent Features
6215 * D30V-Dependent:: D30V Dependent Features
6218 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6221 * HPPA-Dependent:: HPPA Dependent Features
6224 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6227 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6230 * i860-Dependent:: Intel 80860 Dependent Features
6233 * i960-Dependent:: Intel 80960 Dependent Features
6236 * IA-64-Dependent:: Intel IA-64 Dependent Features
6239 * IP2K-Dependent:: IP2K Dependent Features
6242 * M32C-Dependent:: M32C Dependent Features
6245 * M32R-Dependent:: M32R Dependent Features
6248 * M68K-Dependent:: M680x0 Dependent Features
6251 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6254 * MIPS-Dependent:: MIPS Dependent Features
6257 * MMIX-Dependent:: MMIX Dependent Features
6260 * MSP430-Dependent:: MSP430 Dependent Features
6263 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6264 * SH64-Dependent:: SuperH SH64 Dependent Features
6267 * PDP-11-Dependent:: PDP-11 Dependent Features
6270 * PJ-Dependent:: picoJava Dependent Features
6273 * PPC-Dependent:: PowerPC Dependent Features
6276 * Sparc-Dependent:: SPARC Dependent Features
6279 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
6282 * V850-Dependent:: V850 Dependent Features
6285 * Xtensa-Dependent:: Xtensa Dependent Features
6288 * Z80-Dependent:: Z80 Dependent Features
6291 * Z8000-Dependent:: Z8000 Dependent Features
6294 * Vax-Dependent:: VAX Dependent Features
6301 @c The following major nodes are *sections* in the GENERIC version, *chapters*
6302 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
6303 @c peculiarity: to preserve cross-references, there must be a node called
6304 @c "Machine Dependencies". Hence the conditional nodenames in each
6305 @c major node below. Node defaulting in makeinfo requires adjacency of
6306 @c node and sectioning commands; hence the repetition of @chapter BLAH
6307 @c in both conditional blocks.
6310 @include c-alpha.texi
6326 @include c-bfin.texi
6330 @include c-cris.texi
6335 @node Machine Dependencies
6336 @chapter Machine Dependent Features
6338 The machine instruction sets are different on each Renesas chip family,
6339 and there are also some syntax differences among the families. This
6340 chapter describes the specific @command{@value{AS}} features for each
6344 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6345 * SH-Dependent:: Renesas SH Dependent Features
6352 @include c-d10v.texi
6356 @include c-d30v.texi
6360 @include c-h8300.texi
6364 @include c-hppa.texi
6368 @include c-i370.texi
6372 @include c-i386.texi
6376 @include c-i860.texi
6380 @include c-i960.texi
6384 @include c-ia64.texi
6388 @include c-ip2k.texi
6392 @include c-m32c.texi
6396 @include c-m32r.texi
6400 @include c-m68k.texi
6404 @include c-m68hc11.texi
6408 @include c-mips.texi
6412 @include c-mmix.texi
6416 @include c-msp430.texi
6420 @include c-ns32k.texi
6424 @include c-pdp11.texi
6437 @include c-sh64.texi
6441 @include c-sparc.texi
6445 @include c-tic54x.texi
6461 @include c-v850.texi
6465 @include c-xtensa.texi
6469 @c reverse effect of @down at top of generic Machine-Dep chapter
6473 @node Reporting Bugs
6474 @chapter Reporting Bugs
6475 @cindex bugs in assembler
6476 @cindex reporting bugs in assembler
6478 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6480 Reporting a bug may help you by bringing a solution to your problem, or it may
6481 not. But in any case the principal function of a bug report is to help the
6482 entire community by making the next version of @command{@value{AS}} work better.
6483 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6485 In order for a bug report to serve its purpose, you must include the
6486 information that enables us to fix the bug.
6489 * Bug Criteria:: Have you found a bug?
6490 * Bug Reporting:: How to report bugs
6494 @section Have You Found a Bug?
6495 @cindex bug criteria
6497 If you are not sure whether you have found a bug, here are some guidelines:
6500 @cindex fatal signal
6501 @cindex assembler crash
6502 @cindex crash of assembler
6504 If the assembler gets a fatal signal, for any input whatever, that is a
6505 @command{@value{AS}} bug. Reliable assemblers never crash.
6507 @cindex error on valid input
6509 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6511 @cindex invalid input
6513 If @command{@value{AS}} does not produce an error message for invalid input, that
6514 is a bug. However, you should note that your idea of ``invalid input'' might
6515 be our idea of ``an extension'' or ``support for traditional practice''.
6518 If you are an experienced user of assemblers, your suggestions for improvement
6519 of @command{@value{AS}} are welcome in any case.
6523 @section How to Report Bugs
6525 @cindex assembler bugs, reporting
6527 A number of companies and individuals offer support for @sc{gnu} products. If
6528 you obtained @command{@value{AS}} from a support organization, we recommend you
6529 contact that organization first.
6531 You can find contact information for many support companies and
6532 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6535 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6536 to @samp{bug-binutils@@gnu.org}.
6538 The fundamental principle of reporting bugs usefully is this:
6539 @strong{report all the facts}. If you are not sure whether to state a
6540 fact or leave it out, state it!
6542 Often people omit facts because they think they know what causes the problem
6543 and assume that some details do not matter. Thus, you might assume that the
6544 name of a symbol you use in an example does not matter. Well, probably it does
6545 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6546 happens to fetch from the location where that name is stored in memory;
6547 perhaps, if the name were different, the contents of that location would fool
6548 the assembler into doing the right thing despite the bug. Play it safe and
6549 give a specific, complete example. That is the easiest thing for you to do,
6550 and the most helpful.
6552 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6553 it is new to us. Therefore, always write your bug reports on the assumption
6554 that the bug has not been reported previously.
6556 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6557 bell?'' This cannot help us fix a bug, so it is basically useless. We
6558 respond by asking for enough details to enable us to investigate.
6559 You might as well expedite matters by sending them to begin with.
6561 To enable us to fix the bug, you should include all these things:
6565 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6566 it with the @samp{--version} argument.
6568 Without this, we will not know whether there is any point in looking for
6569 the bug in the current version of @command{@value{AS}}.
6572 Any patches you may have applied to the @command{@value{AS}} source.
6575 The type of machine you are using, and the operating system name and
6579 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6583 The command arguments you gave the assembler to assemble your example and
6584 observe the bug. To guarantee you will not omit something important, list them
6585 all. A copy of the Makefile (or the output from make) is sufficient.
6587 If we were to try to guess the arguments, we would probably guess wrong
6588 and then we might not encounter the bug.
6591 A complete input file that will reproduce the bug. If the bug is observed when
6592 the assembler is invoked via a compiler, send the assembler source, not the
6593 high level language source. Most compilers will produce the assembler source
6594 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6595 the options @samp{-v --save-temps}; this will save the assembler source in a
6596 file with an extension of @file{.s}, and also show you exactly how
6597 @command{@value{AS}} is being run.
6600 A description of what behavior you observe that you believe is
6601 incorrect. For example, ``It gets a fatal signal.''
6603 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6604 will certainly notice it. But if the bug is incorrect output, we might not
6605 notice unless it is glaringly wrong. You might as well not give us a chance to
6608 Even if the problem you experience is a fatal signal, you should still say so
6609 explicitly. Suppose something strange is going on, such as, your copy of
6610 @command{@value{AS}} is out of synch, or you have encountered a bug in the C
6611 library on your system. (This has happened!) Your copy might crash and ours
6612 would not. If you told us to expect a crash, then when ours fails to crash, we
6613 would know that the bug was not happening for us. If you had not told us to
6614 expect a crash, then we would not be able to draw any conclusion from our
6618 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6619 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6620 option. Always send diffs from the old file to the new file. If you even
6621 discuss something in the @command{@value{AS}} source, refer to it by context, not
6624 The line numbers in our development sources will not match those in your
6625 sources. Your line numbers would convey no useful information to us.
6628 Here are some things that are not necessary:
6632 A description of the envelope of the bug.
6634 Often people who encounter a bug spend a lot of time investigating
6635 which changes to the input file will make the bug go away and which
6636 changes will not affect it.
6638 This is often time consuming and not very useful, because the way we
6639 will find the bug is by running a single example under the debugger
6640 with breakpoints, not by pure deduction from a series of examples.
6641 We recommend that you save your time for something else.
6643 Of course, if you can find a simpler example to report @emph{instead}
6644 of the original one, that is a convenience for us. Errors in the
6645 output will be easier to spot, running under the debugger will take
6646 less time, and so on.
6648 However, simplification is not vital; if you do not want to do this,
6649 report the bug anyway and send us the entire test case you used.
6652 A patch for the bug.
6654 A patch for the bug does help us if it is a good one. But do not omit
6655 the necessary information, such as the test case, on the assumption that
6656 a patch is all we need. We might see problems with your patch and decide
6657 to fix the problem another way, or we might not understand it at all.
6659 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6660 construct an example that will make the program follow a certain path through
6661 the code. If you do not send us the example, we will not be able to construct
6662 one, so we will not be able to verify that the bug is fixed.
6664 And if we cannot understand what bug you are trying to fix, or why your
6665 patch should be an improvement, we will not install it. A test case will
6666 help us to understand.
6669 A guess about what the bug is or what it depends on.
6671 Such guesses are usually wrong. Even we cannot guess right about such
6672 things without first using the debugger to find the facts.
6675 @node Acknowledgements
6676 @chapter Acknowledgements
6678 If you have contributed to GAS and your name isn't listed here,
6679 it is not meant as a slight. We just don't know about it. Send mail to the
6680 maintainer, and we'll correct the situation. Currently
6682 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6684 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6687 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6688 information and the 68k series machines, most of the preprocessing pass, and
6689 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6691 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6692 many bug fixes, including merging support for several processors, breaking GAS
6693 up to handle multiple object file format back ends (including heavy rewrite,
6694 testing, an integration of the coff and b.out back ends), adding configuration
6695 including heavy testing and verification of cross assemblers and file splits
6696 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6697 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6698 port (including considerable amounts of reverse engineering), a SPARC opcode
6699 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6700 assertions and made them work, much other reorganization, cleanup, and lint.
6702 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6703 in format-specific I/O modules.
6705 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6706 has done much work with it since.
6708 The Intel 80386 machine description was written by Eliot Dresselhaus.
6710 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6712 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6713 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6715 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6716 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6717 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6718 support a.out format.
6720 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
6721 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
6722 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6723 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6726 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6727 simplified the configuration of which versions accept which directives. He
6728 updated the 68k machine description so that Motorola's opcodes always produced
6729 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6730 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6731 cross-compilation support, and one bug in relaxation that took a week and
6732 required the proverbial one-bit fix.
6734 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6735 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6736 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6737 PowerPC assembler, and made a few other minor patches.
6739 Steve Chamberlain made GAS able to generate listings.
6741 Hewlett-Packard contributed support for the HP9000/300.
6743 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6744 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6745 formats). This work was supported by both the Center for Software Science at
6746 the University of Utah and Cygnus Support.
6748 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6749 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6750 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6751 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6752 and some initial 64-bit support).
6754 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6756 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6757 support for openVMS/Alpha.
6759 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6762 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6763 Inc. added support for Xtensa processors.
6765 Several engineers at Cygnus Support have also provided many small bug fixes and
6766 configuration enhancements.
6768 Many others have contributed large or small bugfixes and enhancements. If
6769 you have contributed significant work and are not mentioned on this list, and
6770 want to be, let us know. Some of the history has been lost; we are not
6771 intentionally leaving anyone out.
6776 @unnumbered AS Index
projects / deliverable / binutils-gdb.git / blob