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
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 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 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
201 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
203 This version of the manual describes @command{@value{AS}} configured to generate
204 code for @value{TARGET} architectures.
208 @cindex invocation summary
209 @cindex option summary
210 @cindex summary of options
211 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
212 @pxref{Invoking,,Command-Line Options}.
214 @c man title AS the portable GNU assembler.
218 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
222 @c We don't use deffn and friends for the following because they seem
223 @c to be limited to one line for the header.
225 @c man begin SYNOPSIS
226 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
227 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
228 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
229 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
230 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
231 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
232 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
233 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
234 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
235 [@b{--target-help}] [@var{target-options}]
236 [@b{--}|@var{files} @dots{}]
238 @c Target dependent options are listed below. Keep the list sorted.
239 @c Add an empty line for separation.
242 @emph{Target Alpha options:}
244 [@b{-mdebug} | @b{-no-mdebug}]
245 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
246 [@b{-F}] [@b{-32addr}]
250 @emph{Target ARC options:}
256 @emph{Target ARM options:}
257 @c Don't document the deprecated options
258 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
259 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
260 [@b{-mfpu}=@var{floating-point-format}]
261 [@b{-mfloat-abi}=@var{abi}]
262 [@b{-meabi}=@var{ver}]
265 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
266 @b{-mapcs-reentrant}]
267 [@b{-mthumb-interwork}] [@b{-k}]
271 @emph{Target CRIS options:}
272 [@b{--underscore} | @b{--no-underscore}]
274 [@b{--emulation=criself} | @b{--emulation=crisaout}]
275 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
276 @c Deprecated -- deliberately not documented.
281 @emph{Target D10V options:}
286 @emph{Target D30V options:}
287 [@b{-O}|@b{-n}|@b{-N}]
290 @c Renesas family chips have no machine-dependent assembler options
293 @c HPPA has no machine-dependent assembler options (yet).
297 @emph{Target i386 options:}
298 [@b{--32}|@b{--64}] [@b{-n}]
302 @emph{Target i960 options:}
303 @c see md_parse_option in tc-i960.c
304 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
306 [@b{-b}] [@b{-no-relax}]
310 @emph{Target IA-64 options:}
311 [@b{-mconstant-gp}|@b{-mauto-pic}]
312 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
314 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
315 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
316 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
317 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
321 @emph{Target IP2K options:}
322 [@b{-mip2022}|@b{-mip2022ext}]
326 @emph{Target M32C options:}
327 [@b{-m32c}|@b{-m16c}]
331 @emph{Target M32R options:}
332 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
337 @emph{Target M680X0 options:}
338 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
342 @emph{Target M68HC11 options:}
343 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
344 [@b{-mshort}|@b{-mlong}]
345 [@b{-mshort-double}|@b{-mlong-double}]
346 [@b{--force-long-branchs}] [@b{--short-branchs}]
347 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
348 [@b{--print-opcodes}] [@b{--generate-example}]
352 @emph{Target MCORE options:}
353 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
354 [@b{-mcpu=[210|340]}]
358 @emph{Target MIPS options:}
359 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
360 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
361 [@b{-non_shared}] [@b{-xgot}]
362 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
363 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
364 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
365 [@b{-mips64}] [@b{-mips64r2}]
366 [@b{-construct-floats}] [@b{-no-construct-floats}]
367 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
368 [@b{-mfix7000}] [@b{-mno-fix7000}]
369 [@b{-mips16}] [@b{-no-mips16}]
370 [@b{-mips3d}] [@b{-no-mips3d}]
371 [@b{-mdmx}] [@b{-no-mdmx}]
372 [@b{-mdsp}] [@b{-mno-dsp}]
373 [@b{-mmt}] [@b{-mno-mt}]
374 [@b{-mdebug}] [@b{-no-mdebug}]
375 [@b{-mpdr}] [@b{-mno-pdr}]
379 @emph{Target MMIX options:}
380 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
381 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
382 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
383 [@b{--linker-allocated-gregs}]
387 @emph{Target PDP11 options:}
388 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
389 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
390 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
394 @emph{Target picoJava options:}
399 @emph{Target PowerPC options:}
400 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
401 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
402 @b{-mbooke32}|@b{-mbooke64}]
403 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
404 [@b{-mregnames}|@b{-mno-regnames}]
405 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
406 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
407 [@b{-msolaris}|@b{-mno-solaris}]
411 @emph{Target SPARC options:}
412 @c The order here is important. See c-sparc.texi.
413 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
414 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
415 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
420 @emph{Target TIC54X options:}
421 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
422 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
427 @emph{Target Z80 options:}
428 [@b{-z80}] [@b{-r800}]
429 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
430 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
431 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
432 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
433 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
434 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
438 @c Z8000 has no machine-dependent assembler options
442 @emph{Target Xtensa options:}
443 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
444 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
445 [@b{--[no-]transform}]
446 [@b{--rename-section} @var{oldname}=@var{newname}]
454 @include at-file.texi
457 Turn on listings, in any of a variety of ways:
461 omit false conditionals
464 omit debugging directives
467 include high-level source
473 include macro expansions
476 omit forms processing
482 set the name of the listing file
485 You may combine these options; for example, use @samp{-aln} for assembly
486 listing without forms processing. The @samp{=file} option, if used, must be
487 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
490 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
493 Ignored. This option is accepted for script compatibility with calls to
496 @item --defsym @var{sym}=@var{value}
497 Define the symbol @var{sym} to be @var{value} before assembling the input file.
498 @var{value} must be an integer constant. As in C, a leading @samp{0x}
499 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
502 ``fast''---skip whitespace and comment preprocessing (assume source is
507 Generate debugging information for each assembler source line using whichever
508 debug format is preferred by the target. This currently means either STABS,
512 Generate stabs debugging information for each assembler line. This
513 may help debugging assembler code, if the debugger can handle it.
516 Generate stabs debugging information for each assembler line, with GNU
517 extensions that probably only gdb can handle, and that could make other
518 debuggers crash or refuse to read your program. This
519 may help debugging assembler code. Currently the only GNU extension is
520 the location of the current working directory at assembling time.
523 Generate DWARF2 debugging information for each assembler line. This
524 may help debugging assembler code, if the debugger can handle it. Note---this
525 option is only supported by some targets, not all of them.
528 Print a summary of the command line options and exit.
531 Print a summary of all target specific options and exit.
534 Add directory @var{dir} to the search list for @code{.include} directives.
537 Don't warn about signed overflow.
540 @ifclear DIFF-TBL-KLUGE
541 This option is accepted but has no effect on the @value{TARGET} family.
543 @ifset DIFF-TBL-KLUGE
544 Issue warnings when difference tables altered for long displacements.
549 Keep (in the symbol table) local symbols. On traditional a.out systems
550 these start with @samp{L}, but different systems have different local
553 @item --listing-lhs-width=@var{number}
554 Set the maximum width, in words, of the output data column for an assembler
555 listing to @var{number}.
557 @item --listing-lhs-width2=@var{number}
558 Set the maximum width, in words, of the output data column for continuation
559 lines in an assembler listing to @var{number}.
561 @item --listing-rhs-width=@var{number}
562 Set the maximum width of an input source line, as displayed in a listing, to
565 @item --listing-cont-lines=@var{number}
566 Set the maximum number of lines printed in a listing for a single line of input
569 @item -o @var{objfile}
570 Name the object-file output from @command{@value{AS}} @var{objfile}.
573 Fold the data section into the text section.
575 @kindex --hash-size=@var{number}
576 Set the default size of GAS's hash tables to a prime number close to
577 @var{number}. Increasing this value can reduce the length of time it takes the
578 assembler to perform its tasks, at the expense of increasing the assembler's
579 memory requirements. Similarly reducing this value can reduce the memory
580 requirements at the expense of speed.
582 @item --reduce-memory-overheads
583 This option reduces GAS's memory requirements, at the expense of making the
584 assembly processes slower. Currently this switch is a synonym for
585 @samp{--hash-size=4051}, but in the future it may have other effects as well.
588 Print the maximum space (in bytes) and total time (in seconds) used by
591 @item --strip-local-absolute
592 Remove local absolute symbols from the outgoing symbol table.
596 Print the @command{as} version.
599 Print the @command{as} version and exit.
603 Suppress warning messages.
605 @item --fatal-warnings
606 Treat warnings as errors.
609 Don't suppress warning messages or treat them as errors.
618 Generate an object file even after errors.
620 @item -- | @var{files} @dots{}
621 Standard input, or source files to assemble.
626 The following options are available when @value{AS} is configured for
631 This option selects the core processor variant.
633 Select either big-endian (-EB) or little-endian (-EL) output.
638 The following options are available when @value{AS} is configured for the ARM
642 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
643 Specify which ARM processor variant is the target.
644 @item -march=@var{architecture}[+@var{extension}@dots{}]
645 Specify which ARM architecture variant is used by the target.
646 @item -mfpu=@var{floating-point-format}
647 Select which Floating Point architecture is the target.
648 @item -mfloat-abi=@var{abi}
649 Select which floating point ABI is in use.
651 Enable Thumb only instruction decoding.
652 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
653 Select which procedure calling convention is in use.
655 Select either big-endian (-EB) or little-endian (-EL) output.
656 @item -mthumb-interwork
657 Specify that the code has been generated with interworking between Thumb and
660 Specify that PIC code has been generated.
665 See the info pages for documentation of the CRIS-specific options.
669 The following options are available when @value{AS} is configured for
672 @cindex D10V optimization
673 @cindex optimization, D10V
675 Optimize output by parallelizing instructions.
680 The following options are available when @value{AS} is configured for a D30V
683 @cindex D30V optimization
684 @cindex optimization, D30V
686 Optimize output by parallelizing instructions.
690 Warn when nops are generated.
692 @cindex D30V nops after 32-bit multiply
694 Warn when a nop after a 32-bit multiply instruction is generated.
699 The following options are available when @value{AS} is configured for the
700 Intel 80960 processor.
703 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
704 Specify which variant of the 960 architecture is the target.
707 Add code to collect statistics about branches taken.
710 Do not alter compare-and-branch instructions for long displacements;
717 The following options are available when @value{AS} is configured for the
723 Specifies that the extended IP2022 instructions are allowed.
726 Restores the default behaviour, which restricts the permitted instructions to
727 just the basic IP2022 ones.
733 The following options are available when @value{AS} is configured for the
734 Renesas M32C and M16C processors.
739 Assemble M32C instructions.
742 Assemble M16C instructions (the default).
748 The following options are available when @value{AS} is configured for the
749 Renesas M32R (formerly Mitsubishi M32R) series.
754 Specify which processor in the M32R family is the target. The default
755 is normally the M32R, but this option changes it to the M32RX.
757 @item --warn-explicit-parallel-conflicts or --Wp
758 Produce warning messages when questionable parallel constructs are
761 @item --no-warn-explicit-parallel-conflicts or --Wnp
762 Do not produce warning messages when questionable parallel constructs are
769 The following options are available when @value{AS} is configured for the
770 Motorola 68000 series.
775 Shorten references to undefined symbols, to one word instead of two.
777 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
778 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
779 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
780 Specify what processor in the 68000 family is the target. The default
781 is normally the 68020, but this can be changed at configuration time.
783 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
784 The target machine does (or does not) have a floating-point coprocessor.
785 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
786 the basic 68000 is not compatible with the 68881, a combination of the
787 two can be specified, since it's possible to do emulation of the
788 coprocessor instructions with the main processor.
790 @item -m68851 | -mno-68851
791 The target machine does (or does not) have a memory-management
792 unit coprocessor. The default is to assume an MMU for 68020 and up.
799 For details about the PDP-11 machine dependent features options,
800 see @ref{PDP-11-Options}.
803 @item -mpic | -mno-pic
804 Generate position-independent (or position-dependent) code. The
805 default is @option{-mpic}.
808 @itemx -mall-extensions
809 Enable all instruction set extensions. This is the default.
811 @item -mno-extensions
812 Disable all instruction set extensions.
814 @item -m@var{extension} | -mno-@var{extension}
815 Enable (or disable) a particular instruction set extension.
818 Enable the instruction set extensions supported by a particular CPU, and
819 disable all other extensions.
821 @item -m@var{machine}
822 Enable the instruction set extensions supported by a particular machine
823 model, and disable all other extensions.
829 The following options are available when @value{AS} is configured for
830 a picoJava processor.
834 @cindex PJ endianness
835 @cindex endianness, PJ
836 @cindex big endian output, PJ
838 Generate ``big endian'' format output.
840 @cindex little endian output, PJ
842 Generate ``little endian'' format output.
848 The following options are available when @value{AS} is configured for the
849 Motorola 68HC11 or 68HC12 series.
853 @item -m68hc11 | -m68hc12 | -m68hcs12
854 Specify what processor is the target. The default is
855 defined by the configuration option when building the assembler.
858 Specify to use the 16-bit integer ABI.
861 Specify to use the 32-bit integer ABI.
864 Specify to use the 32-bit double ABI.
867 Specify to use the 64-bit double ABI.
869 @item --force-long-branchs
870 Relative branches are turned into absolute ones. This concerns
871 conditional branches, unconditional branches and branches to a
874 @item -S | --short-branchs
875 Do not turn relative branchs into absolute ones
876 when the offset is out of range.
878 @item --strict-direct-mode
879 Do not turn the direct addressing mode into extended addressing mode
880 when the instruction does not support direct addressing mode.
882 @item --print-insn-syntax
883 Print the syntax of instruction in case of error.
885 @item --print-opcodes
886 print the list of instructions with syntax and then exit.
888 @item --generate-example
889 print an example of instruction for each possible instruction and then exit.
890 This option is only useful for testing @command{@value{AS}}.
896 The following options are available when @command{@value{AS}} is configured
897 for the SPARC architecture:
900 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
901 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
902 Explicitly select a variant of the SPARC architecture.
904 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
905 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
907 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
908 UltraSPARC extensions.
910 @item -xarch=v8plus | -xarch=v8plusa
911 For compatibility with the Solaris v9 assembler. These options are
912 equivalent to -Av8plus and -Av8plusa, respectively.
915 Warn when the assembler switches to another architecture.
920 The following options are available when @value{AS} is configured for the 'c54x
925 Enable extended addressing mode. All addresses and relocations will assume
926 extended addressing (usually 23 bits).
927 @item -mcpu=@var{CPU_VERSION}
928 Sets the CPU version being compiled for.
929 @item -merrors-to-file @var{FILENAME}
930 Redirect error output to a file, for broken systems which don't support such
931 behaviour in the shell.
936 The following options are available when @value{AS} is configured for
937 a @sc{mips} processor.
941 This option sets the largest size of an object that can be referenced
942 implicitly with the @code{gp} register. It is only accepted for targets that
943 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
945 @cindex MIPS endianness
946 @cindex endianness, MIPS
947 @cindex big endian output, MIPS
949 Generate ``big endian'' format output.
951 @cindex little endian output, MIPS
953 Generate ``little endian'' format output.
965 Generate code for a particular @sc{mips} Instruction Set Architecture level.
966 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
967 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
968 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
969 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
971 correspond to generic
972 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
973 and @samp{MIPS64 Release 2}
974 ISA processors, respectively.
976 @item -march=@var{CPU}
977 Generate code for a particular @sc{mips} cpu.
979 @item -mtune=@var{cpu}
980 Schedule and tune for a particular @sc{mips} cpu.
984 Cause nops to be inserted if the read of the destination register
985 of an mfhi or mflo instruction occurs in the following two instructions.
989 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
990 section instead of the standard ELF .stabs sections.
994 Control generation of @code{.pdr} sections.
998 The register sizes are normally inferred from the ISA and ABI, but these
999 flags force a certain group of registers to be treated as 32 bits wide at
1000 all times. @samp{-mgp32} controls the size of general-purpose registers
1001 and @samp{-mfp32} controls the size of floating-point registers.
1005 Generate code for the MIPS 16 processor. This is equivalent to putting
1006 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1007 turns off this option.
1011 Generate code for the MIPS-3D Application Specific Extension.
1012 This tells the assembler to accept MIPS-3D instructions.
1013 @samp{-no-mips3d} turns off this option.
1017 Generate code for the MDMX Application Specific Extension.
1018 This tells the assembler to accept MDMX instructions.
1019 @samp{-no-mdmx} turns off this option.
1023 Generate code for the DSP Application Specific Extension.
1024 This tells the assembler to accept DSP instructions.
1025 @samp{-mno-dsp} turns off this option.
1029 Generate code for the MT Application Specific Extension.
1030 This tells the assembler to accept MT instructions.
1031 @samp{-mno-mt} turns off this option.
1033 @item --construct-floats
1034 @itemx --no-construct-floats
1035 The @samp{--no-construct-floats} option disables the construction of
1036 double width floating point constants by loading the two halves of the
1037 value into the two single width floating point registers that make up
1038 the double width register. By default @samp{--construct-floats} is
1039 selected, allowing construction of these floating point constants.
1042 @item --emulation=@var{name}
1043 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1044 for some other target, in all respects, including output format (choosing
1045 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1046 debugging information or store symbol table information, and default
1047 endianness. The available configuration names are: @samp{mipsecoff},
1048 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1049 @samp{mipsbelf}. The first two do not alter the default endianness from that
1050 of the primary target for which the assembler was configured; the others change
1051 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1052 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1053 selection in any case.
1055 This option is currently supported only when the primary target
1056 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1057 Furthermore, the primary target or others specified with
1058 @samp{--enable-targets=@dots{}} at configuration time must include support for
1059 the other format, if both are to be available. For example, the Irix 5
1060 configuration includes support for both.
1062 Eventually, this option will support more configurations, with more
1063 fine-grained control over the assembler's behavior, and will be supported for
1067 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1074 Control how to deal with multiplication overflow and division by zero.
1075 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1076 (and only work for Instruction Set Architecture level 2 and higher);
1077 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1081 When this option is used, @command{@value{AS}} will issue a warning every
1082 time it generates a nop instruction from a macro.
1087 The following options are available when @value{AS} is configured for
1093 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1094 The command line option @samp{-nojsri2bsr} can be used to disable it.
1098 Enable or disable the silicon filter behaviour. By default this is disabled.
1099 The default can be overridden by the @samp{-sifilter} command line option.
1102 Alter jump instructions for long displacements.
1104 @item -mcpu=[210|340]
1105 Select the cpu type on the target hardware. This controls which instructions
1109 Assemble for a big endian target.
1112 Assemble for a little endian target.
1118 See the info pages for documentation of the MMIX-specific options.
1122 The following options are available when @value{AS} is configured for
1123 an Xtensa processor.
1126 @item --text-section-literals | --no-text-section-literals
1127 With @option{--text-@-section-@-literals}, literal pools are interspersed
1128 in the text section. The default is
1129 @option{--no-@-text-@-section-@-literals}, which places literals in a
1130 separate section in the output file. These options only affect literals
1131 referenced via PC-relative @code{L32R} instructions; literals for
1132 absolute mode @code{L32R} instructions are handled separately.
1134 @item --absolute-literals | --no-absolute-literals
1135 Indicate to the assembler whether @code{L32R} instructions use absolute
1136 or PC-relative addressing. The default is to assume absolute addressing
1137 if the Xtensa processor includes the absolute @code{L32R} addressing
1138 option. Otherwise, only the PC-relative @code{L32R} mode can be used.
1140 @item --target-align | --no-target-align
1141 Enable or disable automatic alignment to reduce branch penalties at the
1142 expense of some code density. The default is @option{--target-@-align}.
1144 @item --longcalls | --no-longcalls
1145 Enable or disable transformation of call instructions to allow calls
1146 across a greater range of addresses. The default is
1147 @option{--no-@-longcalls}.
1149 @item --transform | --no-transform
1150 Enable or disable all assembler transformations of Xtensa instructions.
1151 The default is @option{--transform};
1152 @option{--no-transform} should be used only in the rare cases when the
1153 instructions must be exactly as specified in the assembly source.
1158 The following options are available when @value{AS} is configured for
1159 a Z80 family processor.
1162 Assemble for Z80 processor.
1164 Assemble for R800 processor.
1165 @item -ignore-undocumented-instructions
1167 Assemble undocumented Z80 instructions that also work on R800 without warning.
1168 @item -ignore-unportable-instructions
1170 Assemble all undocumented Z80 instructions without warning.
1171 @item -warn-undocumented-instructions
1173 Issue a warning for undocumented Z80 instructions that also work on R800.
1174 @item -warn-unportable-instructions
1176 Issue a warning for undocumented Z80 instructions that do notwork on R800.
1177 @item -forbid-undocumented-instructions
1179 Treat all undocumented instructions as errors.
1180 @item -forbid-unportable-instructions
1182 Treat undocumented Z80 intructions that do notwork on R800 as errors.
1189 * Manual:: Structure of this Manual
1190 * GNU Assembler:: The GNU Assembler
1191 * Object Formats:: Object File Formats
1192 * Command Line:: Command Line
1193 * Input Files:: Input Files
1194 * Object:: Output (Object) File
1195 * Errors:: Error and Warning Messages
1199 @section Structure of this Manual
1201 @cindex manual, structure and purpose
1202 This manual is intended to describe what you need to know to use
1203 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1204 notation for symbols, constants, and expressions; the directives that
1205 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1208 We also cover special features in the @value{TARGET}
1209 configuration of @command{@value{AS}}, including assembler directives.
1212 This manual also describes some of the machine-dependent features of
1213 various flavors of the assembler.
1216 @cindex machine instructions (not covered)
1217 On the other hand, this manual is @emph{not} intended as an introduction
1218 to programming in assembly language---let alone programming in general!
1219 In a similar vein, we make no attempt to introduce the machine
1220 architecture; we do @emph{not} describe the instruction set, standard
1221 mnemonics, registers or addressing modes that are standard to a
1222 particular architecture.
1224 You may want to consult the manufacturer's
1225 machine architecture manual for this information.
1229 For information on the H8/300 machine instruction set, see @cite{H8/300
1230 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1231 Programming Manual} (Renesas).
1234 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1235 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1236 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1237 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1240 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1244 @c I think this is premature---doc@cygnus.com, 17jan1991
1246 Throughout this manual, we assume that you are running @dfn{GNU},
1247 the portable operating system from the @dfn{Free Software
1248 Foundation, Inc.}. This restricts our attention to certain kinds of
1249 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1250 once this assumption is granted examples and definitions need less
1253 @command{@value{AS}} is part of a team of programs that turn a high-level
1254 human-readable series of instructions into a low-level
1255 computer-readable series of instructions. Different versions of
1256 @command{@value{AS}} are used for different kinds of computer.
1259 @c There used to be a section "Terminology" here, which defined
1260 @c "contents", "byte", "word", and "long". Defining "word" to any
1261 @c particular size is confusing when the .word directive may generate 16
1262 @c bits on one machine and 32 bits on another; in general, for the user
1263 @c version of this manual, none of these terms seem essential to define.
1264 @c They were used very little even in the former draft of the manual;
1265 @c this draft makes an effort to avoid them (except in names of
1269 @section The GNU Assembler
1271 @c man begin DESCRIPTION
1273 @sc{gnu} @command{as} is really a family of assemblers.
1275 This manual describes @command{@value{AS}}, a member of that family which is
1276 configured for the @value{TARGET} architectures.
1278 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1279 should find a fairly similar environment when you use it on another
1280 architecture. Each version has much in common with the others,
1281 including object file formats, most assembler directives (often called
1282 @dfn{pseudo-ops}) and assembler syntax.@refill
1284 @cindex purpose of @sc{gnu} assembler
1285 @command{@value{AS}} is primarily intended to assemble the output of the
1286 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1287 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1288 assemble correctly everything that other assemblers for the same
1289 machine would assemble.
1291 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1294 @c This remark should appear in generic version of manual; assumption
1295 @c here is that generic version sets M680x0.
1296 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1297 assembler for the same architecture; for example, we know of several
1298 incompatible versions of 680x0 assembly language syntax.
1303 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1304 program in one pass of the source file. This has a subtle impact on the
1305 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1307 @node Object Formats
1308 @section Object File Formats
1310 @cindex object file format
1311 The @sc{gnu} assembler can be configured to produce several alternative
1312 object file formats. For the most part, this does not affect how you
1313 write assembly language programs; but directives for debugging symbols
1314 are typically different in different file formats. @xref{Symbol
1315 Attributes,,Symbol Attributes}.
1318 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1319 @value{OBJ-NAME} format object files.
1321 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1323 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1324 @code{b.out} or COFF format object files.
1327 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1328 SOM or ELF format object files.
1333 @section Command Line
1335 @cindex command line conventions
1337 After the program name @command{@value{AS}}, the command line may contain
1338 options and file names. Options may appear in any order, and may be
1339 before, after, or between file names. The order of file names is
1342 @cindex standard input, as input file
1344 @file{--} (two hyphens) by itself names the standard input file
1345 explicitly, as one of the files for @command{@value{AS}} to assemble.
1347 @cindex options, command line
1348 Except for @samp{--} any command line argument that begins with a
1349 hyphen (@samp{-}) is an option. Each option changes the behavior of
1350 @command{@value{AS}}. No option changes the way another option works. An
1351 option is a @samp{-} followed by one or more letters; the case of
1352 the letter is important. All options are optional.
1354 Some options expect exactly one file name to follow them. The file
1355 name may either immediately follow the option's letter (compatible
1356 with older assemblers) or it may be the next command argument (@sc{gnu}
1357 standard). These two command lines are equivalent:
1360 @value{AS} -o my-object-file.o mumble.s
1361 @value{AS} -omy-object-file.o mumble.s
1365 @section Input Files
1368 @cindex source program
1369 @cindex files, input
1370 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1371 describe the program input to one run of @command{@value{AS}}. The program may
1372 be in one or more files; how the source is partitioned into files
1373 doesn't change the meaning of the source.
1375 @c I added "con" prefix to "catenation" just to prove I can overcome my
1376 @c APL training... doc@cygnus.com
1377 The source program is a concatenation of the text in all the files, in the
1380 @c man begin DESCRIPTION
1381 Each time you run @command{@value{AS}} it assembles exactly one source
1382 program. The source program is made up of one or more files.
1383 (The standard input is also a file.)
1385 You give @command{@value{AS}} a command line that has zero or more input file
1386 names. The input files are read (from left file name to right). A
1387 command line argument (in any position) that has no special meaning
1388 is taken to be an input file name.
1390 If you give @command{@value{AS}} no file names it attempts to read one input file
1391 from the @command{@value{AS}} standard input, which is normally your terminal. You
1392 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1395 Use @samp{--} if you need to explicitly name the standard input file
1396 in your command line.
1398 If the source is empty, @command{@value{AS}} produces a small, empty object
1403 @subheading Filenames and Line-numbers
1405 @cindex input file linenumbers
1406 @cindex line numbers, in input files
1407 There are two ways of locating a line in the input file (or files) and
1408 either may be used in reporting error messages. One way refers to a line
1409 number in a physical file; the other refers to a line number in a
1410 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1412 @dfn{Physical files} are those files named in the command line given
1413 to @command{@value{AS}}.
1415 @dfn{Logical files} are simply names declared explicitly by assembler
1416 directives; they bear no relation to physical files. Logical file names help
1417 error messages reflect the original source file, when @command{@value{AS}} source
1418 is itself synthesized from other files. @command{@value{AS}} understands the
1419 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1420 @ref{File,,@code{.file}}.
1423 @section Output (Object) File
1429 Every time you run @command{@value{AS}} it produces an output file, which is
1430 your assembly language program translated into numbers. This file
1431 is the object file. Its default name is
1439 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1441 You can give it another name by using the @option{-o} option. Conventionally,
1442 object file names end with @file{.o}. The default name is used for historical
1443 reasons: older assemblers were capable of assembling self-contained programs
1444 directly into a runnable program. (For some formats, this isn't currently
1445 possible, but it can be done for the @code{a.out} format.)
1449 The object file is meant for input to the linker @code{@value{LD}}. It contains
1450 assembled program code, information to help @code{@value{LD}} integrate
1451 the assembled program into a runnable file, and (optionally) symbolic
1452 information for the debugger.
1454 @c link above to some info file(s) like the description of a.out.
1455 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1458 @section Error and Warning Messages
1460 @c man begin DESCRIPTION
1462 @cindex error messages
1463 @cindex warning messages
1464 @cindex messages from assembler
1465 @command{@value{AS}} may write warnings and error messages to the standard error
1466 file (usually your terminal). This should not happen when a compiler
1467 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1468 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1469 grave problem that stops the assembly.
1473 @cindex format of warning messages
1474 Warning messages have the format
1477 file_name:@b{NNN}:Warning Message Text
1481 @cindex line numbers, in warnings/errors
1482 (where @b{NNN} is a line number). If a logical file name has been given
1483 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1484 the current input file is used. If a logical line number was given
1486 (@pxref{Line,,@code{.line}})
1488 then it is used to calculate the number printed,
1489 otherwise the actual line in the current source file is printed. The
1490 message text is intended to be self explanatory (in the grand Unix
1493 @cindex format of error messages
1494 Error messages have the format
1496 file_name:@b{NNN}:FATAL:Error Message Text
1498 The file name and line number are derived as for warning
1499 messages. The actual message text may be rather less explanatory
1500 because many of them aren't supposed to happen.
1503 @chapter Command-Line Options
1505 @cindex options, all versions of assembler
1506 This chapter describes command-line options available in @emph{all}
1507 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1509 to the @value{TARGET} target.
1512 to particular machine architectures.
1515 @c man begin DESCRIPTION
1517 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1518 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1519 The assembler arguments must be separated from each other (and the @samp{-Wa})
1520 by commas. For example:
1523 gcc -c -g -O -Wa,-alh,-L file.c
1527 This passes two options to the assembler: @samp{-alh} (emit a listing to
1528 standard output with high-level and assembly source) and @samp{-L} (retain
1529 local symbols in the symbol table).
1531 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1532 command-line options are automatically passed to the assembler by the compiler.
1533 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1534 precisely what options it passes to each compilation pass, including the
1540 * a:: -a[cdhlns] enable listings
1541 * alternate:: --alternate enable alternate macro syntax
1542 * D:: -D for compatibility
1543 * f:: -f to work faster
1544 * I:: -I for .include search path
1545 @ifclear DIFF-TBL-KLUGE
1546 * K:: -K for compatibility
1548 @ifset DIFF-TBL-KLUGE
1549 * K:: -K for difference tables
1552 * L:: -L to retain local labels
1553 * listing:: --listing-XXX to configure listing output
1554 * M:: -M or --mri to assemble in MRI compatibility mode
1555 * MD:: --MD for dependency tracking
1556 * o:: -o to name the object file
1557 * R:: -R to join data and text sections
1558 * statistics:: --statistics to see statistics about assembly
1559 * traditional-format:: --traditional-format for compatible output
1560 * v:: -v to announce version
1561 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1562 * Z:: -Z to make object file even after errors
1566 @section Enable Listings: @option{-a[cdhlns]}
1575 @cindex listings, enabling
1576 @cindex assembly listings, enabling
1578 These options enable listing output from the assembler. By itself,
1579 @samp{-a} requests high-level, assembly, and symbols listing.
1580 You can use other letters to select specific options for the list:
1581 @samp{-ah} requests a high-level language listing,
1582 @samp{-al} requests an output-program assembly listing, and
1583 @samp{-as} requests a symbol table listing.
1584 High-level listings require that a compiler debugging option like
1585 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1588 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1589 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1590 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1591 omitted from the listing.
1593 Use the @samp{-ad} option to omit debugging directives from the
1596 Once you have specified one of these options, you can further control
1597 listing output and its appearance using the directives @code{.list},
1598 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1600 The @samp{-an} option turns off all forms processing.
1601 If you do not request listing output with one of the @samp{-a} options, the
1602 listing-control directives have no effect.
1604 The letters after @samp{-a} may be combined into one option,
1605 @emph{e.g.}, @samp{-aln}.
1607 Note if the assembler source is coming from the standard input (eg because it
1608 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1609 is being used) then the listing will not contain any comments or preprocessor
1610 directives. This is because the listing code buffers input source lines from
1611 stdin only after they have been preprocessed by the assembler. This reduces
1612 memory usage and makes the code more efficient.
1615 @section @option{--alternate}
1618 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1621 @section @option{-D}
1624 This option has no effect whatsoever, but it is accepted to make it more
1625 likely that scripts written for other assemblers also work with
1626 @command{@value{AS}}.
1629 @section Work Faster: @option{-f}
1632 @cindex trusted compiler
1633 @cindex faster processing (@option{-f})
1634 @samp{-f} should only be used when assembling programs written by a
1635 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1636 and comment preprocessing on
1637 the input file(s) before assembling them. @xref{Preprocessing,
1641 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1642 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1647 @section @code{.include} Search Path: @option{-I} @var{path}
1649 @kindex -I @var{path}
1650 @cindex paths for @code{.include}
1651 @cindex search path for @code{.include}
1652 @cindex @code{include} directive search path
1653 Use this option to add a @var{path} to the list of directories
1654 @command{@value{AS}} searches for files specified in @code{.include}
1655 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1656 many times as necessary to include a variety of paths. The current
1657 working directory is always searched first; after that, @command{@value{AS}}
1658 searches any @samp{-I} directories in the same order as they were
1659 specified (left to right) on the command line.
1662 @section Difference Tables: @option{-K}
1665 @ifclear DIFF-TBL-KLUGE
1666 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1667 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1668 where it can be used to warn when the assembler alters the machine code
1669 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1670 family does not have the addressing limitations that sometimes lead to this
1671 alteration on other platforms.
1674 @ifset DIFF-TBL-KLUGE
1675 @cindex difference tables, warning
1676 @cindex warning for altered difference tables
1677 @command{@value{AS}} sometimes alters the code emitted for directives of the form
1678 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1679 You can use the @samp{-K} option if you want a warning issued when this
1684 @section Include Local Labels: @option{-L}
1687 @cindex local labels, retaining in output
1688 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1689 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1690 debugging, because they are intended for the use of programs (like
1691 compilers) that compose assembler programs, not for your notice.
1692 Normally both @command{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1693 normally debug with them.
1695 This option tells @command{@value{AS}} to retain those @samp{L@dots{}} symbols
1696 in the object file. Usually if you do this you also tell the linker
1697 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1699 By default, a local label is any label beginning with @samp{L}, but each
1700 target is allowed to redefine the local label prefix.
1702 On the HPPA local labels begin with @samp{L$}.
1706 @section Configuring listing output: @option{--listing}
1708 The listing feature of the assembler can be enabled via the command line switch
1709 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1710 hex dump of the corresponding locations in the output object file, and displays
1711 them as a listing file. The format of this listing can be controlled by pseudo
1712 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1713 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1716 @item --listing-lhs-width=@samp{number}
1717 @kindex --listing-lhs-width
1718 @cindex Width of first line disassembly output
1719 Sets the maximum width, in words, of the first line of the hex byte dump. This
1720 dump appears on the left hand side of the listing output.
1722 @item --listing-lhs-width2=@samp{number}
1723 @kindex --listing-lhs-width2
1724 @cindex Width of continuation lines of disassembly output
1725 Sets the maximum width, in words, of any further lines of the hex byte dump for
1726 a given input source line. If this value is not specified, it defaults to being
1727 the same as the value specified for @samp{--listing-lhs-width}. If neither
1728 switch is used the default is to one.
1730 @item --listing-rhs-width=@samp{number}
1731 @kindex --listing-rhs-width
1732 @cindex Width of source line output
1733 Sets the maximum width, in characters, of the source line that is displayed
1734 alongside the hex dump. The default value for this parameter is 100. The
1735 source line is displayed on the right hand side of the listing output.
1737 @item --listing-cont-lines=@samp{number}
1738 @kindex --listing-cont-lines
1739 @cindex Maximum number of continuation lines
1740 Sets the maximum number of continuation lines of hex dump that will be
1741 displayed for a given single line of source input. The default value is 4.
1745 @section Assemble in MRI Compatibility Mode: @option{-M}
1748 @cindex MRI compatibility mode
1749 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1750 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1751 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1752 configured target) assembler from Microtec Research. The exact nature of the
1753 MRI syntax will not be documented here; see the MRI manuals for more
1754 information. Note in particular that the handling of macros and macro
1755 arguments is somewhat different. The purpose of this option is to permit
1756 assembling existing MRI assembler code using @command{@value{AS}}.
1758 The MRI compatibility is not complete. Certain operations of the MRI assembler
1759 depend upon its object file format, and can not be supported using other object
1760 file formats. Supporting these would require enhancing each object file format
1761 individually. These are:
1764 @item global symbols in common section
1766 The m68k MRI assembler supports common sections which are merged by the linker.
1767 Other object file formats do not support this. @command{@value{AS}} handles
1768 common sections by treating them as a single common symbol. It permits local
1769 symbols to be defined within a common section, but it can not support global
1770 symbols, since it has no way to describe them.
1772 @item complex relocations
1774 The MRI assemblers support relocations against a negated section address, and
1775 relocations which combine the start addresses of two or more sections. These
1776 are not support by other object file formats.
1778 @item @code{END} pseudo-op specifying start address
1780 The MRI @code{END} pseudo-op permits the specification of a start address.
1781 This is not supported by other object file formats. The start address may
1782 instead be specified using the @option{-e} option to the linker, or in a linker
1785 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1787 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1788 name to the output file. This is not supported by other object file formats.
1790 @item @code{ORG} pseudo-op
1792 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1793 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1794 which changes the location within the current section. Absolute sections are
1795 not supported by other object file formats. The address of a section may be
1796 assigned within a linker script.
1799 There are some other features of the MRI assembler which are not supported by
1800 @command{@value{AS}}, typically either because they are difficult or because they
1801 seem of little consequence. Some of these may be supported in future releases.
1805 @item EBCDIC strings
1807 EBCDIC strings are not supported.
1809 @item packed binary coded decimal
1811 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1812 and @code{DCB.P} pseudo-ops are not supported.
1814 @item @code{FEQU} pseudo-op
1816 The m68k @code{FEQU} pseudo-op is not supported.
1818 @item @code{NOOBJ} pseudo-op
1820 The m68k @code{NOOBJ} pseudo-op is not supported.
1822 @item @code{OPT} branch control options
1824 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1825 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1826 relaxes all branches, whether forward or backward, to an appropriate size, so
1827 these options serve no purpose.
1829 @item @code{OPT} list control options
1831 The following m68k @code{OPT} list control options are ignored: @code{C},
1832 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1833 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1835 @item other @code{OPT} options
1837 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1838 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1840 @item @code{OPT} @code{D} option is default
1842 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1843 @code{OPT NOD} may be used to turn it off.
1845 @item @code{XREF} pseudo-op.
1847 The m68k @code{XREF} pseudo-op is ignored.
1849 @item @code{.debug} pseudo-op
1851 The i960 @code{.debug} pseudo-op is not supported.
1853 @item @code{.extended} pseudo-op
1855 The i960 @code{.extended} pseudo-op is not supported.
1857 @item @code{.list} pseudo-op.
1859 The various options of the i960 @code{.list} pseudo-op are not supported.
1861 @item @code{.optimize} pseudo-op
1863 The i960 @code{.optimize} pseudo-op is not supported.
1865 @item @code{.output} pseudo-op
1867 The i960 @code{.output} pseudo-op is not supported.
1869 @item @code{.setreal} pseudo-op
1871 The i960 @code{.setreal} pseudo-op is not supported.
1876 @section Dependency Tracking: @option{--MD}
1879 @cindex dependency tracking
1882 @command{@value{AS}} can generate a dependency file for the file it creates. This
1883 file consists of a single rule suitable for @code{make} describing the
1884 dependencies of the main source file.
1886 The rule is written to the file named in its argument.
1888 This feature is used in the automatic updating of makefiles.
1891 @section Name the Object File: @option{-o}
1894 @cindex naming object file
1895 @cindex object file name
1896 There is always one object file output when you run @command{@value{AS}}. By
1897 default it has the name
1900 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1914 You use this option (which takes exactly one filename) to give the
1915 object file a different name.
1917 Whatever the object file is called, @command{@value{AS}} overwrites any
1918 existing file of the same name.
1921 @section Join Data and Text Sections: @option{-R}
1924 @cindex data and text sections, joining
1925 @cindex text and data sections, joining
1926 @cindex joining text and data sections
1927 @cindex merging text and data sections
1928 @option{-R} tells @command{@value{AS}} to write the object file as if all
1929 data-section data lives in the text section. This is only done at
1930 the very last moment: your binary data are the same, but data
1931 section parts are relocated differently. The data section part of
1932 your object file is zero bytes long because all its bytes are
1933 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1935 When you specify @option{-R} it would be possible to generate shorter
1936 address displacements (because we do not have to cross between text and
1937 data section). We refrain from doing this simply for compatibility with
1938 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1941 When @command{@value{AS}} is configured for COFF or ELF output,
1942 this option is only useful if you use sections named @samp{.text} and
1947 @option{-R} is not supported for any of the HPPA targets. Using
1948 @option{-R} generates a warning from @command{@value{AS}}.
1952 @section Display Assembly Statistics: @option{--statistics}
1954 @kindex --statistics
1955 @cindex statistics, about assembly
1956 @cindex time, total for assembly
1957 @cindex space used, maximum for assembly
1958 Use @samp{--statistics} to display two statistics about the resources used by
1959 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1960 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1963 @node traditional-format
1964 @section Compatible Output: @option{--traditional-format}
1966 @kindex --traditional-format
1967 For some targets, the output of @command{@value{AS}} is different in some ways
1968 from the output of some existing assembler. This switch requests
1969 @command{@value{AS}} to use the traditional format instead.
1971 For example, it disables the exception frame optimizations which
1972 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
1975 @section Announce Version: @option{-v}
1979 @cindex assembler version
1980 @cindex version of assembler
1981 You can find out what version of as is running by including the
1982 option @samp{-v} (which you can also spell as @samp{-version}) on the
1986 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
1988 @command{@value{AS}} should never give a warning or error message when
1989 assembling compiler output. But programs written by people often
1990 cause @command{@value{AS}} to give a warning that a particular assumption was
1991 made. All such warnings are directed to the standard error file.
1995 @cindex suppressing warnings
1996 @cindex warnings, suppressing
1997 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
1998 This only affects the warning messages: it does not change any particular of
1999 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2002 @kindex --fatal-warnings
2003 @cindex errors, caused by warnings
2004 @cindex warnings, causing error
2005 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2006 files that generate warnings to be in error.
2009 @cindex warnings, switching on
2010 You can switch these options off again by specifying @option{--warn}, which
2011 causes warnings to be output as usual.
2014 @section Generate Object File in Spite of Errors: @option{-Z}
2015 @cindex object file, after errors
2016 @cindex errors, continuing after
2017 After an error message, @command{@value{AS}} normally produces no output. If for
2018 some reason you are interested in object file output even after
2019 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2020 option. If there are any errors, @command{@value{AS}} continues anyways, and
2021 writes an object file after a final warning message of the form @samp{@var{n}
2022 errors, @var{m} warnings, generating bad object file.}
2027 @cindex machine-independent syntax
2028 @cindex syntax, machine-independent
2029 This chapter describes the machine-independent syntax allowed in a
2030 source file. @command{@value{AS}} syntax is similar to what many other
2031 assemblers use; it is inspired by the BSD 4.2
2036 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2040 * Preprocessing:: Preprocessing
2041 * Whitespace:: Whitespace
2042 * Comments:: Comments
2043 * Symbol Intro:: Symbols
2044 * Statements:: Statements
2045 * Constants:: Constants
2049 @section Preprocessing
2051 @cindex preprocessing
2052 The @command{@value{AS}} internal preprocessor:
2054 @cindex whitespace, removed by preprocessor
2056 adjusts and removes extra whitespace. It leaves one space or tab before
2057 the keywords on a line, and turns any other whitespace on the line into
2060 @cindex comments, removed by preprocessor
2062 removes all comments, replacing them with a single space, or an
2063 appropriate number of newlines.
2065 @cindex constants, converted by preprocessor
2067 converts character constants into the appropriate numeric values.
2070 It does not do macro processing, include file handling, or
2071 anything else you may get from your C compiler's preprocessor. You can
2072 do include file processing with the @code{.include} directive
2073 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2074 to get other ``CPP'' style preprocessing by giving the input file a
2075 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
2076 Output, gcc.info, Using GNU CC}.
2078 Excess whitespace, comments, and character constants
2079 cannot be used in the portions of the input text that are not
2082 @cindex turning preprocessing on and off
2083 @cindex preprocessing, turning on and off
2086 If the first line of an input file is @code{#NO_APP} or if you use the
2087 @samp{-f} option, whitespace and comments are not removed from the input file.
2088 Within an input file, you can ask for whitespace and comment removal in
2089 specific portions of the by putting a line that says @code{#APP} before the
2090 text that may contain whitespace or comments, and putting a line that says
2091 @code{#NO_APP} after this text. This feature is mainly intend to support
2092 @code{asm} statements in compilers whose output is otherwise free of comments
2099 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2100 Whitespace is used to separate symbols, and to make programs neater for
2101 people to read. Unless within character constants
2102 (@pxref{Characters,,Character Constants}), any whitespace means the same
2103 as exactly one space.
2109 There are two ways of rendering comments to @command{@value{AS}}. In both
2110 cases the comment is equivalent to one space.
2112 Anything from @samp{/*} through the next @samp{*/} is a comment.
2113 This means you may not nest these comments.
2117 The only way to include a newline ('\n') in a comment
2118 is to use this sort of comment.
2121 /* This sort of comment does not nest. */
2124 @cindex line comment character
2125 Anything from the @dfn{line comment} character to the next newline
2126 is considered a comment and is ignored. The line comment character is
2128 @samp{;} on the ARC;
2131 @samp{@@} on the ARM;
2134 @samp{;} for the H8/300 family;
2137 @samp{;} for the HPPA;
2140 @samp{#} on the i386 and x86-64;
2143 @samp{#} on the i960;
2146 @samp{;} for the PDP-11;
2149 @samp{;} for picoJava;
2152 @samp{#} for Motorola PowerPC;
2155 @samp{!} for the Renesas / SuperH SH;
2158 @samp{!} on the SPARC;
2161 @samp{#} on the ip2k;
2164 @samp{#} on the m32c;
2167 @samp{#} on the m32r;
2170 @samp{|} on the 680x0;
2173 @samp{#} on the 68HC11 and 68HC12;
2176 @samp{#} on the Vax;
2179 @samp{;} for the Z80;
2182 @samp{!} for the Z8000;
2185 @samp{#} on the V850;
2188 @samp{#} for Xtensa systems;
2190 see @ref{Machine Dependencies}. @refill
2191 @c FIXME What about i860?
2194 On some machines there are two different line comment characters. One
2195 character only begins a comment if it is the first non-whitespace character on
2196 a line, while the other always begins a comment.
2200 The V850 assembler also supports a double dash as starting a comment that
2201 extends to the end of the line.
2207 @cindex lines starting with @code{#}
2208 @cindex logical line numbers
2209 To be compatible with past assemblers, lines that begin with @samp{#} have a
2210 special interpretation. Following the @samp{#} should be an absolute
2211 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2212 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
2213 new logical file name. The rest of the line, if any, should be whitespace.
2215 If the first non-whitespace characters on the line are not numeric,
2216 the line is ignored. (Just like a comment.)
2219 # This is an ordinary comment.
2220 # 42-6 "new_file_name" # New logical file name
2221 # This is logical line # 36.
2223 This feature is deprecated, and may disappear from future versions
2224 of @command{@value{AS}}.
2229 @cindex characters used in symbols
2230 @ifclear SPECIAL-SYMS
2231 A @dfn{symbol} is one or more characters chosen from the set of all
2232 letters (both upper and lower case), digits and the three characters
2238 A @dfn{symbol} is one or more characters chosen from the set of all
2239 letters (both upper and lower case), digits and the three characters
2240 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2246 On most machines, you can also use @code{$} in symbol names; exceptions
2247 are noted in @ref{Machine Dependencies}.
2249 No symbol may begin with a digit. Case is significant.
2250 There is no length limit: all characters are significant. Symbols are
2251 delimited by characters not in that set, or by the beginning of a file
2252 (since the source program must end with a newline, the end of a file is
2253 not a possible symbol delimiter). @xref{Symbols}.
2254 @cindex length of symbols
2259 @cindex statements, structure of
2260 @cindex line separator character
2261 @cindex statement separator character
2263 @ifclear abnormal-separator
2264 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2265 semicolon (@samp{;}). The newline or semicolon is considered part of
2266 the preceding statement. Newlines and semicolons within character
2267 constants are an exception: they do not end statements.
2269 @ifset abnormal-separator
2271 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2272 point (@samp{!}). The newline or exclamation point is considered part of the
2273 preceding statement. Newlines and exclamation points within character
2274 constants are an exception: they do not end statements.
2277 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2278 H8/300) a dollar sign (@samp{$}); or (for the Renesas-SH) a semicolon
2279 (@samp{;}). The newline or separator character is considered part of
2280 the preceding statement. Newlines and separators within character
2281 constants are an exception: they do not end statements.
2286 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2287 separator character. (The line separator is usually @samp{;}, unless
2288 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
2289 newline or separator character is considered part of the preceding
2290 statement. Newlines and separators within character constants are an
2291 exception: they do not end statements.
2294 @cindex newline, required at file end
2295 @cindex EOF, newline must precede
2296 It is an error to end any statement with end-of-file: the last
2297 character of any input file should be a newline.@refill
2299 An empty statement is allowed, and may include whitespace. It is ignored.
2301 @cindex instructions and directives
2302 @cindex directives and instructions
2303 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2304 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2306 A statement begins with zero or more labels, optionally followed by a
2307 key symbol which determines what kind of statement it is. The key
2308 symbol determines the syntax of the rest of the statement. If the
2309 symbol begins with a dot @samp{.} then the statement is an assembler
2310 directive: typically valid for any computer. If the symbol begins with
2311 a letter the statement is an assembly language @dfn{instruction}: it
2312 assembles into a machine language instruction.
2314 Different versions of @command{@value{AS}} for different computers
2315 recognize different instructions. In fact, the same symbol may
2316 represent a different instruction in a different computer's assembly
2320 @cindex @code{:} (label)
2321 @cindex label (@code{:})
2322 A label is a symbol immediately followed by a colon (@code{:}).
2323 Whitespace before a label or after a colon is permitted, but you may not
2324 have whitespace between a label's symbol and its colon. @xref{Labels}.
2327 For HPPA targets, labels need not be immediately followed by a colon, but
2328 the definition of a label must begin in column zero. This also implies that
2329 only one label may be defined on each line.
2333 label: .directive followed by something
2334 another_label: # This is an empty statement.
2335 instruction operand_1, operand_2, @dots{}
2342 A constant is a number, written so that its value is known by
2343 inspection, without knowing any context. Like this:
2346 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2347 .ascii "Ring the bell\7" # A string constant.
2348 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2349 .float 0f-314159265358979323846264338327\
2350 95028841971.693993751E-40 # - pi, a flonum.
2355 * Characters:: Character Constants
2356 * Numbers:: Number Constants
2360 @subsection Character Constants
2362 @cindex character constants
2363 @cindex constants, character
2364 There are two kinds of character constants. A @dfn{character} stands
2365 for one character in one byte and its value may be used in
2366 numeric expressions. String constants (properly called string
2367 @emph{literals}) are potentially many bytes and their values may not be
2368 used in arithmetic expressions.
2372 * Chars:: Characters
2376 @subsubsection Strings
2378 @cindex string constants
2379 @cindex constants, string
2380 A @dfn{string} is written between double-quotes. It may contain
2381 double-quotes or null characters. The way to get special characters
2382 into a string is to @dfn{escape} these characters: precede them with
2383 a backslash @samp{\} character. For example @samp{\\} represents
2384 one backslash: the first @code{\} is an escape which tells
2385 @command{@value{AS}} to interpret the second character literally as a backslash
2386 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2387 escape character). The complete list of escapes follows.
2389 @cindex escape codes, character
2390 @cindex character escape codes
2393 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2395 @cindex @code{\b} (backspace character)
2396 @cindex backspace (@code{\b})
2398 Mnemonic for backspace; for ASCII this is octal code 010.
2401 @c Mnemonic for EOText; for ASCII this is octal code 004.
2403 @cindex @code{\f} (formfeed character)
2404 @cindex formfeed (@code{\f})
2406 Mnemonic for FormFeed; for ASCII this is octal code 014.
2408 @cindex @code{\n} (newline character)
2409 @cindex newline (@code{\n})
2411 Mnemonic for newline; for ASCII this is octal code 012.
2414 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2416 @cindex @code{\r} (carriage return character)
2417 @cindex carriage return (@code{\r})
2419 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2422 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2423 @c other assemblers.
2425 @cindex @code{\t} (tab)
2426 @cindex tab (@code{\t})
2428 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2431 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2432 @c @item \x @var{digit} @var{digit} @var{digit}
2433 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2435 @cindex @code{\@var{ddd}} (octal character code)
2436 @cindex octal character code (@code{\@var{ddd}})
2437 @item \ @var{digit} @var{digit} @var{digit}
2438 An octal character code. The numeric code is 3 octal digits.
2439 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2440 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2442 @cindex @code{\@var{xd...}} (hex character code)
2443 @cindex hex character code (@code{\@var{xd...}})
2444 @item \@code{x} @var{hex-digits...}
2445 A hex character code. All trailing hex digits are combined. Either upper or
2446 lower case @code{x} works.
2448 @cindex @code{\\} (@samp{\} character)
2449 @cindex backslash (@code{\\})
2451 Represents one @samp{\} character.
2454 @c Represents one @samp{'} (accent acute) character.
2455 @c This is needed in single character literals
2456 @c (@xref{Characters,,Character Constants}.) to represent
2459 @cindex @code{\"} (doublequote character)
2460 @cindex doublequote (@code{\"})
2462 Represents one @samp{"} character. Needed in strings to represent
2463 this character, because an unescaped @samp{"} would end the string.
2465 @item \ @var{anything-else}
2466 Any other character when escaped by @kbd{\} gives a warning, but
2467 assembles as if the @samp{\} was not present. The idea is that if
2468 you used an escape sequence you clearly didn't want the literal
2469 interpretation of the following character. However @command{@value{AS}} has no
2470 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2471 code and warns you of the fact.
2474 Which characters are escapable, and what those escapes represent,
2475 varies widely among assemblers. The current set is what we think
2476 the BSD 4.2 assembler recognizes, and is a subset of what most C
2477 compilers recognize. If you are in doubt, do not use an escape
2481 @subsubsection Characters
2483 @cindex single character constant
2484 @cindex character, single
2485 @cindex constant, single character
2486 A single character may be written as a single quote immediately
2487 followed by that character. The same escapes apply to characters as
2488 to strings. So if you want to write the character backslash, you
2489 must write @kbd{'\\} where the first @code{\} escapes the second
2490 @code{\}. As you can see, the quote is an acute accent, not a
2491 grave accent. A newline
2493 @ifclear abnormal-separator
2494 (or semicolon @samp{;})
2496 @ifset abnormal-separator
2498 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2503 immediately following an acute accent is taken as a literal character
2504 and does not count as the end of a statement. The value of a character
2505 constant in a numeric expression is the machine's byte-wide code for
2506 that character. @command{@value{AS}} assumes your character code is ASCII:
2507 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2510 @subsection Number Constants
2512 @cindex constants, number
2513 @cindex number constants
2514 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2515 are stored in the target machine. @emph{Integers} are numbers that
2516 would fit into an @code{int} in the C language. @emph{Bignums} are
2517 integers, but they are stored in more than 32 bits. @emph{Flonums}
2518 are floating point numbers, described below.
2521 * Integers:: Integers
2526 * Bit Fields:: Bit Fields
2532 @subsubsection Integers
2534 @cindex constants, integer
2536 @cindex binary integers
2537 @cindex integers, binary
2538 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2539 the binary digits @samp{01}.
2541 @cindex octal integers
2542 @cindex integers, octal
2543 An octal integer is @samp{0} followed by zero or more of the octal
2544 digits (@samp{01234567}).
2546 @cindex decimal integers
2547 @cindex integers, decimal
2548 A decimal integer starts with a non-zero digit followed by zero or
2549 more digits (@samp{0123456789}).
2551 @cindex hexadecimal integers
2552 @cindex integers, hexadecimal
2553 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2554 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2556 Integers have the usual values. To denote a negative integer, use
2557 the prefix operator @samp{-} discussed under expressions
2558 (@pxref{Prefix Ops,,Prefix Operators}).
2561 @subsubsection Bignums
2564 @cindex constants, bignum
2565 A @dfn{bignum} has the same syntax and semantics as an integer
2566 except that the number (or its negative) takes more than 32 bits to
2567 represent in binary. The distinction is made because in some places
2568 integers are permitted while bignums are not.
2571 @subsubsection Flonums
2573 @cindex floating point numbers
2574 @cindex constants, floating point
2576 @cindex precision, floating point
2577 A @dfn{flonum} represents a floating point number. The translation is
2578 indirect: a decimal floating point number from the text is converted by
2579 @command{@value{AS}} to a generic binary floating point number of more than
2580 sufficient precision. This generic floating point number is converted
2581 to a particular computer's floating point format (or formats) by a
2582 portion of @command{@value{AS}} specialized to that computer.
2584 A flonum is written by writing (in order)
2589 (@samp{0} is optional on the HPPA.)
2593 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2595 @kbd{e} is recommended. Case is not important.
2597 @c FIXME: verify if flonum syntax really this vague for most cases
2598 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2599 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2602 On the H8/300, Renesas / SuperH SH,
2603 and AMD 29K architectures, the letter must be
2604 one of the letters @samp{DFPRSX} (in upper or lower case).
2606 On the ARC, the letter must be one of the letters @samp{DFRS}
2607 (in upper or lower case).
2609 On the Intel 960 architecture, the letter must be
2610 one of the letters @samp{DFT} (in upper or lower case).
2612 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2616 One of the letters @samp{DFRS} (in upper or lower case).
2619 One of the letters @samp{DFPRSX} (in upper or lower case).
2622 The letter @samp{E} (upper case only).
2625 One of the letters @samp{DFT} (in upper or lower case).
2630 An optional sign: either @samp{+} or @samp{-}.
2633 An optional @dfn{integer part}: zero or more decimal digits.
2636 An optional @dfn{fractional part}: @samp{.} followed by zero
2637 or more decimal digits.
2640 An optional exponent, consisting of:
2644 An @samp{E} or @samp{e}.
2645 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2646 @c principle this can perfectly well be different on different targets.
2648 Optional sign: either @samp{+} or @samp{-}.
2650 One or more decimal digits.
2655 At least one of the integer part or the fractional part must be
2656 present. The floating point number has the usual base-10 value.
2658 @command{@value{AS}} does all processing using integers. Flonums are computed
2659 independently of any floating point hardware in the computer running
2660 @command{@value{AS}}.
2664 @c Bit fields are written as a general facility but are also controlled
2665 @c by a conditional-compilation flag---which is as of now (21mar91)
2666 @c turned on only by the i960 config of GAS.
2668 @subsubsection Bit Fields
2671 @cindex constants, bit field
2672 You can also define numeric constants as @dfn{bit fields}.
2673 specify two numbers separated by a colon---
2675 @var{mask}:@var{value}
2678 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2681 The resulting number is then packed
2683 @c this conditional paren in case bit fields turned on elsewhere than 960
2684 (in host-dependent byte order)
2686 into a field whose width depends on which assembler directive has the
2687 bit-field as its argument. Overflow (a result from the bitwise and
2688 requiring more binary digits to represent) is not an error; instead,
2689 more constants are generated, of the specified width, beginning with the
2690 least significant digits.@refill
2692 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2693 @code{.short}, and @code{.word} accept bit-field arguments.
2698 @chapter Sections and Relocation
2703 * Secs Background:: Background
2704 * Ld Sections:: Linker Sections
2705 * As Sections:: Assembler Internal Sections
2706 * Sub-Sections:: Sub-Sections
2710 @node Secs Background
2713 Roughly, a section is a range of addresses, with no gaps; all data
2714 ``in'' those addresses is treated the same for some particular purpose.
2715 For example there may be a ``read only'' section.
2717 @cindex linker, and assembler
2718 @cindex assembler, and linker
2719 The linker @code{@value{LD}} reads many object files (partial programs) and
2720 combines their contents to form a runnable program. When @command{@value{AS}}
2721 emits an object file, the partial program is assumed to start at address 0.
2722 @code{@value{LD}} assigns the final addresses for the partial program, so that
2723 different partial programs do not overlap. This is actually an
2724 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2727 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2728 addresses. These blocks slide to their run-time addresses as rigid
2729 units; their length does not change and neither does the order of bytes
2730 within them. Such a rigid unit is called a @emph{section}. Assigning
2731 run-time addresses to sections is called @dfn{relocation}. It includes
2732 the task of adjusting mentions of object-file addresses so they refer to
2733 the proper run-time addresses.
2735 For the H8/300, and for the Renesas / SuperH SH,
2736 @command{@value{AS}} pads sections if needed to
2737 ensure they end on a word (sixteen bit) boundary.
2740 @cindex standard assembler sections
2741 An object file written by @command{@value{AS}} has at least three sections, any
2742 of which may be empty. These are named @dfn{text}, @dfn{data} and
2747 When it generates COFF or ELF output,
2749 @command{@value{AS}} can also generate whatever other named sections you specify
2750 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2751 If you do not use any directives that place output in the @samp{.text}
2752 or @samp{.data} sections, these sections still exist, but are empty.
2757 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2759 @command{@value{AS}} can also generate whatever other named sections you
2760 specify using the @samp{.space} and @samp{.subspace} directives. See
2761 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2762 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2763 assembler directives.
2766 Additionally, @command{@value{AS}} uses different names for the standard
2767 text, data, and bss sections when generating SOM output. Program text
2768 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2769 BSS into @samp{$BSS$}.
2773 Within the object file, the text section starts at address @code{0}, the
2774 data section follows, and the bss section follows the data section.
2777 When generating either SOM or ELF output files on the HPPA, the text
2778 section starts at address @code{0}, the data section at address
2779 @code{0x4000000}, and the bss section follows the data section.
2782 To let @code{@value{LD}} know which data changes when the sections are
2783 relocated, and how to change that data, @command{@value{AS}} also writes to the
2784 object file details of the relocation needed. To perform relocation
2785 @code{@value{LD}} must know, each time an address in the object
2789 Where in the object file is the beginning of this reference to
2792 How long (in bytes) is this reference?
2794 Which section does the address refer to? What is the numeric value of
2796 (@var{address}) @minus{} (@var{start-address of section})?
2799 Is the reference to an address ``Program-Counter relative''?
2802 @cindex addresses, format of
2803 @cindex section-relative addressing
2804 In fact, every address @command{@value{AS}} ever uses is expressed as
2806 (@var{section}) + (@var{offset into section})
2809 Further, most expressions @command{@value{AS}} computes have this section-relative
2812 (For some object formats, such as SOM for the HPPA, some expressions are
2813 symbol-relative instead.)
2816 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2817 @var{N} into section @var{secname}.''
2819 Apart from text, data and bss sections you need to know about the
2820 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2821 addresses in the absolute section remain unchanged. For example, address
2822 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2823 @code{@value{LD}}. Although the linker never arranges two partial programs'
2824 data sections with overlapping addresses after linking, @emph{by definition}
2825 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2826 part of a program is always the same address when the program is running as
2827 address @code{@{absolute@ 239@}} in any other part of the program.
2829 The idea of sections is extended to the @dfn{undefined} section. Any
2830 address whose section is unknown at assembly time is by definition
2831 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2832 Since numbers are always defined, the only way to generate an undefined
2833 address is to mention an undefined symbol. A reference to a named
2834 common block would be such a symbol: its value is unknown at assembly
2835 time so it has section @emph{undefined}.
2837 By analogy the word @emph{section} is used to describe groups of sections in
2838 the linked program. @code{@value{LD}} puts all partial programs' text
2839 sections in contiguous addresses in the linked program. It is
2840 customary to refer to the @emph{text section} of a program, meaning all
2841 the addresses of all partial programs' text sections. Likewise for
2842 data and bss sections.
2844 Some sections are manipulated by @code{@value{LD}}; others are invented for
2845 use of @command{@value{AS}} and have no meaning except during assembly.
2848 @section Linker Sections
2849 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2854 @cindex named sections
2855 @cindex sections, named
2856 @item named sections
2859 @cindex text section
2860 @cindex data section
2864 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2865 separate but equal sections. Anything you can say of one section is
2868 When the program is running, however, it is
2869 customary for the text section to be unalterable. The
2870 text section is often shared among processes: it contains
2871 instructions, constants and the like. The data section of a running
2872 program is usually alterable: for example, C variables would be stored
2873 in the data section.
2878 This section contains zeroed bytes when your program begins running. It
2879 is used to hold uninitialized variables or common storage. The length of
2880 each partial program's bss section is important, but because it starts
2881 out containing zeroed bytes there is no need to store explicit zero
2882 bytes in the object file. The bss section was invented to eliminate
2883 those explicit zeros from object files.
2885 @cindex absolute section
2886 @item absolute section
2887 Address 0 of this section is always ``relocated'' to runtime address 0.
2888 This is useful if you want to refer to an address that @code{@value{LD}} must
2889 not change when relocating. In this sense we speak of absolute
2890 addresses being ``unrelocatable'': they do not change during relocation.
2892 @cindex undefined section
2893 @item undefined section
2894 This ``section'' is a catch-all for address references to objects not in
2895 the preceding sections.
2896 @c FIXME: ref to some other doc on obj-file formats could go here.
2899 @cindex relocation example
2900 An idealized example of three relocatable sections follows.
2902 The example uses the traditional section names @samp{.text} and @samp{.data}.
2904 Memory addresses are on the horizontal axis.
2908 @c END TEXI2ROFF-KILL
2911 partial program # 1: |ttttt|dddd|00|
2918 partial program # 2: |TTT|DDD|000|
2921 +--+---+-----+--+----+---+-----+~~
2922 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2923 +--+---+-----+--+----+---+-----+~~
2925 addresses: 0 @dots{}
2932 \line{\it Partial program \#1: \hfil}
2933 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2934 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2936 \line{\it Partial program \#2: \hfil}
2937 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2938 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2940 \line{\it linked program: \hfil}
2941 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2942 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2943 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2944 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2946 \line{\it addresses: \hfil}
2950 @c END TEXI2ROFF-KILL
2953 @section Assembler Internal Sections
2955 @cindex internal assembler sections
2956 @cindex sections in messages, internal
2957 These sections are meant only for the internal use of @command{@value{AS}}. They
2958 have no meaning at run-time. You do not really need to know about these
2959 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2960 warning messages, so it might be helpful to have an idea of their
2961 meanings to @command{@value{AS}}. These sections are used to permit the
2962 value of every expression in your assembly language program to be a
2963 section-relative address.
2966 @cindex assembler internal logic error
2967 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2968 An internal assembler logic error has been found. This means there is a
2969 bug in the assembler.
2971 @cindex expr (internal section)
2973 The assembler stores complex expression internally as combinations of
2974 symbols. When it needs to represent an expression as a symbol, it puts
2975 it in the expr section.
2977 @c FIXME item transfer[t] vector preload
2978 @c FIXME item transfer[t] vector postload
2979 @c FIXME item register
2983 @section Sub-Sections
2985 @cindex numbered subsections
2986 @cindex grouping data
2992 fall into two sections: text and data.
2994 You may have separate groups of
2996 data in named sections
3000 data in named sections
3006 that you want to end up near to each other in the object file, even though they
3007 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3008 use @dfn{subsections} for this purpose. Within each section, there can be
3009 numbered subsections with values from 0 to 8192. Objects assembled into the
3010 same subsection go into the object file together with other objects in the same
3011 subsection. For example, a compiler might want to store constants in the text
3012 section, but might not want to have them interspersed with the program being
3013 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3014 section of code being output, and a @samp{.text 1} before each group of
3015 constants being output.
3017 Subsections are optional. If you do not use subsections, everything
3018 goes in subsection number zero.
3021 Each subsection is zero-padded up to a multiple of four bytes.
3022 (Subsections may be padded a different amount on different flavors
3023 of @command{@value{AS}}.)
3027 On the H8/300 platform, each subsection is zero-padded to a word
3028 boundary (two bytes).
3029 The same is true on the Renesas SH.
3032 @c FIXME section padding (alignment)?
3033 @c Rich Pixley says padding here depends on target obj code format; that
3034 @c doesn't seem particularly useful to say without further elaboration,
3035 @c so for now I say nothing about it. If this is a generic BFD issue,
3036 @c these paragraphs might need to vanish from this manual, and be
3037 @c discussed in BFD chapter of binutils (or some such).
3041 Subsections appear in your object file in numeric order, lowest numbered
3042 to highest. (All this to be compatible with other people's assemblers.)
3043 The object file contains no representation of subsections; @code{@value{LD}} and
3044 other programs that manipulate object files see no trace of them.
3045 They just see all your text subsections as a text section, and all your
3046 data subsections as a data section.
3048 To specify which subsection you want subsequent statements assembled
3049 into, use a numeric argument to specify it, in a @samp{.text
3050 @var{expression}} or a @samp{.data @var{expression}} statement.
3053 When generating COFF output, you
3058 can also use an extra subsection
3059 argument with arbitrary named sections: @samp{.section @var{name},
3064 When generating ELF output, you
3069 can also use the @code{.subsection} directive (@pxref{SubSection})
3070 to specify a subsection: @samp{.subsection @var{expression}}.
3072 @var{Expression} should be an absolute expression.
3073 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
3074 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3075 begins in @code{text 0}. For instance:
3077 .text 0 # The default subsection is text 0 anyway.
3078 .ascii "This lives in the first text subsection. *"
3080 .ascii "But this lives in the second text subsection."
3082 .ascii "This lives in the data section,"
3083 .ascii "in the first data subsection."
3085 .ascii "This lives in the first text section,"
3086 .ascii "immediately following the asterisk (*)."
3089 Each section has a @dfn{location counter} incremented by one for every byte
3090 assembled into that section. Because subsections are merely a convenience
3091 restricted to @command{@value{AS}} there is no concept of a subsection location
3092 counter. There is no way to directly manipulate a location counter---but the
3093 @code{.align} directive changes it, and any label definition captures its
3094 current value. The location counter of the section where statements are being
3095 assembled is said to be the @dfn{active} location counter.
3098 @section bss Section
3101 @cindex common variable storage
3102 The bss section is used for local common variable storage.
3103 You may allocate address space in the bss section, but you may
3104 not dictate data to load into it before your program executes. When
3105 your program starts running, all the contents of the bss
3106 section are zeroed bytes.
3108 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3109 @ref{Lcomm,,@code{.lcomm}}.
3111 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3112 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
3115 When assembling for a target which supports multiple sections, such as ELF or
3116 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3117 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3118 section. Typically the section will only contain symbol definitions and
3119 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3126 Symbols are a central concept: the programmer uses symbols to name
3127 things, the linker uses symbols to link, and the debugger uses symbols
3131 @cindex debuggers, and symbol order
3132 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3133 the same order they were declared. This may break some debuggers.
3138 * Setting Symbols:: Giving Symbols Other Values
3139 * Symbol Names:: Symbol Names
3140 * Dot:: The Special Dot Symbol
3141 * Symbol Attributes:: Symbol Attributes
3148 A @dfn{label} is written as a symbol immediately followed by a colon
3149 @samp{:}. The symbol then represents the current value of the
3150 active location counter, and is, for example, a suitable instruction
3151 operand. You are warned if you use the same symbol to represent two
3152 different locations: the first definition overrides any other
3156 On the HPPA, the usual form for a label need not be immediately followed by a
3157 colon, but instead must start in column zero. Only one label may be defined on
3158 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3159 provides a special directive @code{.label} for defining labels more flexibly.
3162 @node Setting Symbols
3163 @section Giving Symbols Other Values
3165 @cindex assigning values to symbols
3166 @cindex symbol values, assigning
3167 A symbol can be given an arbitrary value by writing a symbol, followed
3168 by an equals sign @samp{=}, followed by an expression
3169 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3170 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3171 equals sign @samp{=}@samp{=} here represents an equivalent of the
3172 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3175 @section Symbol Names
3177 @cindex symbol names
3178 @cindex names, symbol
3179 @ifclear SPECIAL-SYMS
3180 Symbol names begin with a letter or with one of @samp{._}. On most
3181 machines, you can also use @code{$} in symbol names; exceptions are
3182 noted in @ref{Machine Dependencies}. That character may be followed by any
3183 string of digits, letters, dollar signs (unless otherwise noted in
3184 @ref{Machine Dependencies}), and underscores.
3188 Symbol names begin with a letter or with one of @samp{._}. On the
3189 Renesas SH you can also use @code{$} in symbol names. That
3190 character may be followed by any string of digits, letters, dollar signs (save
3191 on the H8/300), and underscores.
3195 Case of letters is significant: @code{foo} is a different symbol name
3198 Each symbol has exactly one name. Each name in an assembly language program
3199 refers to exactly one symbol. You may use that symbol name any number of times
3202 @subheading Local Symbol Names
3204 @cindex local symbol names
3205 @cindex symbol names, local
3206 @cindex temporary symbol names
3207 @cindex symbol names, temporary
3208 Local symbols help compilers and programmers use names temporarily.
3209 They create symbols which are guaranteed to be unique over the entire scope of
3210 the input source code and which can be referred to by a simple notation.
3211 To define a local symbol, write a label of the form @samp{@b{N}:} (where @b{N}
3212 represents any positive integer). To refer to the most recent previous
3213 definition of that symbol write @samp{@b{N}b}, using the same number as when
3214 you defined the label. To refer to the next definition of a local label, write
3215 @samp{@b{N}f}--- The @samp{b} stands for``backwards'' and the @samp{f} stands
3218 There is no restriction on how you can use these labels, and you can reuse them
3219 too. So that it is possible to repeatedly define the same local label (using
3220 the same number @samp{@b{N}}), although you can only refer to the most recently
3221 defined local label of that number (for a backwards reference) or the next
3222 definition of a specific local label for a forward reference. It is also worth
3223 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3224 implemented in a slightly more efficient manner than the others.
3235 Which is the equivalent of:
3238 label_1: branch label_3
3239 label_2: branch label_1
3240 label_3: branch label_4
3241 label_4: branch label_3
3244 Local symbol names are only a notational device. They are immediately
3245 transformed into more conventional symbol names before the assembler uses them.
3246 The symbol names stored in the symbol table, appearing in error messages and
3247 optionally emitted to the object file. The names are constructed using these
3252 All local labels begin with @samp{L}. Normally both @command{@value{AS}} and
3253 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
3254 used for symbols you are never intended to see. If you use the
3255 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3256 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3257 you may use them in debugging.
3260 This is the number that was used in the local label definition. So if the
3261 label is written @samp{55:} then the number is @samp{55}.
3264 This unusual character is included so you do not accidentally invent a symbol
3265 of the same name. The character has ASCII value of @samp{\002} (control-B).
3267 @item @emph{ordinal number}
3268 This is a serial number to keep the labels distinct. The first definition of
3269 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3270 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3271 the number @samp{1} and its 15th defintion gets @samp{15} as well.
3274 So for example, the first @code{1:} is named @code{L1@kbd{C-B}1}, the 44th
3275 @code{3:} is named @code{L3@kbd{C-B}44}.
3277 @subheading Dollar Local Labels
3278 @cindex dollar local symbols
3280 @code{@value{AS}} also supports an even more local form of local labels called
3281 dollar labels. These labels go out of scope (ie they become undefined) as soon
3282 as a non-local label is defined. Thus they remain valid for only a small
3283 region of the input source code. Normal local labels, by contrast, remain in
3284 scope for the entire file, or until they are redefined by another occurrence of
3285 the same local label.
3287 Dollar labels are defined in exactly the same way as ordinary local labels,
3288 except that instead of being terminated by a colon, they are terminated by a
3289 dollar sign. eg @samp{@b{55$}}.
3291 They can also be distinguished from ordinary local labels by their transformed
3292 name which uses ASCII character @samp{\001} (control-A) as the magic character
3293 to distinguish them from ordinary labels. Thus the 5th defintion of @samp{6$}
3294 is named @samp{L6@kbd{C-A}5}.
3297 @section The Special Dot Symbol
3299 @cindex dot (symbol)
3300 @cindex @code{.} (symbol)
3301 @cindex current address
3302 @cindex location counter
3303 The special symbol @samp{.} refers to the current address that
3304 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3305 .long .} defines @code{melvin} to contain its own address.
3306 Assigning a value to @code{.} is treated the same as a @code{.org}
3307 directive. Thus, the expression @samp{.=.+4} is the same as saying
3308 @ifclear no-space-dir
3312 @node Symbol Attributes
3313 @section Symbol Attributes
3315 @cindex symbol attributes
3316 @cindex attributes, symbol
3317 Every symbol has, as well as its name, the attributes ``Value'' and
3318 ``Type''. Depending on output format, symbols can also have auxiliary
3321 The detailed definitions are in @file{a.out.h}.
3324 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3325 all these attributes, and probably won't warn you. This makes the
3326 symbol an externally defined symbol, which is generally what you
3330 * Symbol Value:: Value
3331 * Symbol Type:: Type
3334 * a.out Symbols:: Symbol Attributes: @code{a.out}
3338 * a.out Symbols:: Symbol Attributes: @code{a.out}
3341 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3346 * COFF Symbols:: Symbol Attributes for COFF
3349 * SOM Symbols:: Symbol Attributes for SOM
3356 @cindex value of a symbol
3357 @cindex symbol value
3358 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3359 location in the text, data, bss or absolute sections the value is the
3360 number of addresses from the start of that section to the label.
3361 Naturally for text, data and bss sections the value of a symbol changes
3362 as @code{@value{LD}} changes section base addresses during linking. Absolute
3363 symbols' values do not change during linking: that is why they are
3366 The value of an undefined symbol is treated in a special way. If it is
3367 0 then the symbol is not defined in this assembler source file, and
3368 @code{@value{LD}} tries to determine its value from other files linked into the
3369 same program. You make this kind of symbol simply by mentioning a symbol
3370 name without defining it. A non-zero value represents a @code{.comm}
3371 common declaration. The value is how much common storage to reserve, in
3372 bytes (addresses). The symbol refers to the first address of the
3378 @cindex type of a symbol
3380 The type attribute of a symbol contains relocation (section)
3381 information, any flag settings indicating that a symbol is external, and
3382 (optionally), other information for linkers and debuggers. The exact
3383 format depends on the object-code output format in use.
3388 @c The following avoids a "widow" subsection title. @group would be
3389 @c better if it were available outside examples.
3392 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3394 @cindex @code{b.out} symbol attributes
3395 @cindex symbol attributes, @code{b.out}
3396 These symbol attributes appear only when @command{@value{AS}} is configured for
3397 one of the Berkeley-descended object output formats---@code{a.out} or
3403 @subsection Symbol Attributes: @code{a.out}
3405 @cindex @code{a.out} symbol attributes
3406 @cindex symbol attributes, @code{a.out}
3412 @subsection Symbol Attributes: @code{a.out}
3414 @cindex @code{a.out} symbol attributes
3415 @cindex symbol attributes, @code{a.out}
3419 * Symbol Desc:: Descriptor
3420 * Symbol Other:: Other
3424 @subsubsection Descriptor
3426 @cindex descriptor, of @code{a.out} symbol
3427 This is an arbitrary 16-bit value. You may establish a symbol's
3428 descriptor value by using a @code{.desc} statement
3429 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3430 @command{@value{AS}}.
3433 @subsubsection Other
3435 @cindex other attribute, of @code{a.out} symbol
3436 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3441 @subsection Symbol Attributes for COFF
3443 @cindex COFF symbol attributes
3444 @cindex symbol attributes, COFF
3446 The COFF format supports a multitude of auxiliary symbol attributes;
3447 like the primary symbol attributes, they are set between @code{.def} and
3448 @code{.endef} directives.
3450 @subsubsection Primary Attributes
3452 @cindex primary attributes, COFF symbols
3453 The symbol name is set with @code{.def}; the value and type,
3454 respectively, with @code{.val} and @code{.type}.
3456 @subsubsection Auxiliary Attributes
3458 @cindex auxiliary attributes, COFF symbols
3459 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3460 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3461 table information for COFF.
3466 @subsection Symbol Attributes for SOM
3468 @cindex SOM symbol attributes
3469 @cindex symbol attributes, SOM
3471 The SOM format for the HPPA supports a multitude of symbol attributes set with
3472 the @code{.EXPORT} and @code{.IMPORT} directives.
3474 The attributes are described in @cite{HP9000 Series 800 Assembly
3475 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3476 @code{EXPORT} assembler directive documentation.
3480 @chapter Expressions
3484 @cindex numeric values
3485 An @dfn{expression} specifies an address or numeric value.
3486 Whitespace may precede and/or follow an expression.
3488 The result of an expression must be an absolute number, or else an offset into
3489 a particular section. If an expression is not absolute, and there is not
3490 enough information when @command{@value{AS}} sees the expression to know its
3491 section, a second pass over the source program might be necessary to interpret
3492 the expression---but the second pass is currently not implemented.
3493 @command{@value{AS}} aborts with an error message in this situation.
3496 * Empty Exprs:: Empty Expressions
3497 * Integer Exprs:: Integer Expressions
3501 @section Empty Expressions
3503 @cindex empty expressions
3504 @cindex expressions, empty
3505 An empty expression has no value: it is just whitespace or null.
3506 Wherever an absolute expression is required, you may omit the
3507 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3508 is compatible with other assemblers.
3511 @section Integer Expressions
3513 @cindex integer expressions
3514 @cindex expressions, integer
3515 An @dfn{integer expression} is one or more @emph{arguments} delimited
3516 by @emph{operators}.
3519 * Arguments:: Arguments
3520 * Operators:: Operators
3521 * Prefix Ops:: Prefix Operators
3522 * Infix Ops:: Infix Operators
3526 @subsection Arguments
3528 @cindex expression arguments
3529 @cindex arguments in expressions
3530 @cindex operands in expressions
3531 @cindex arithmetic operands
3532 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3533 contexts arguments are sometimes called ``arithmetic operands''. In
3534 this manual, to avoid confusing them with the ``instruction operands'' of
3535 the machine language, we use the term ``argument'' to refer to parts of
3536 expressions only, reserving the word ``operand'' to refer only to machine
3537 instruction operands.
3539 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3540 @var{section} is one of text, data, bss, absolute,
3541 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3544 Numbers are usually integers.
3546 A number can be a flonum or bignum. In this case, you are warned
3547 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3548 these 32 bits are an integer. You may write integer-manipulating
3549 instructions that act on exotic constants, compatible with other
3552 @cindex subexpressions
3553 Subexpressions are a left parenthesis @samp{(} followed by an integer
3554 expression, followed by a right parenthesis @samp{)}; or a prefix
3555 operator followed by an argument.
3558 @subsection Operators
3560 @cindex operators, in expressions
3561 @cindex arithmetic functions
3562 @cindex functions, in expressions
3563 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3564 operators are followed by an argument. Infix operators appear
3565 between their arguments. Operators may be preceded and/or followed by
3569 @subsection Prefix Operator
3571 @cindex prefix operators
3572 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3573 one argument, which must be absolute.
3575 @c the tex/end tex stuff surrounding this small table is meant to make
3576 @c it align, on the printed page, with the similar table in the next
3577 @c section (which is inside an enumerate).
3579 \global\advance\leftskip by \itemindent
3584 @dfn{Negation}. Two's complement negation.
3586 @dfn{Complementation}. Bitwise not.
3590 \global\advance\leftskip by -\itemindent
3594 @subsection Infix Operators
3596 @cindex infix operators
3597 @cindex operators, permitted arguments
3598 @dfn{Infix operators} take two arguments, one on either side. Operators
3599 have precedence, but operations with equal precedence are performed left
3600 to right. Apart from @code{+} or @option{-}, both arguments must be
3601 absolute, and the result is absolute.
3604 @cindex operator precedence
3605 @cindex precedence of operators
3612 @dfn{Multiplication}.
3615 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3621 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3624 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3628 Intermediate precedence
3633 @dfn{Bitwise Inclusive Or}.
3639 @dfn{Bitwise Exclusive Or}.
3642 @dfn{Bitwise Or Not}.
3649 @cindex addition, permitted arguments
3650 @cindex plus, permitted arguments
3651 @cindex arguments for addition
3653 @dfn{Addition}. If either argument is absolute, the result has the section of
3654 the other argument. You may not add together arguments from different
3657 @cindex subtraction, permitted arguments
3658 @cindex minus, permitted arguments
3659 @cindex arguments for subtraction
3661 @dfn{Subtraction}. If the right argument is absolute, the
3662 result has the section of the left argument.
3663 If both arguments are in the same section, the result is absolute.
3664 You may not subtract arguments from different sections.
3665 @c FIXME is there still something useful to say about undefined - undefined ?
3667 @cindex comparison expressions
3668 @cindex expressions, comparison
3673 @dfn{Is Not Equal To}
3677 @dfn{Is Greater Than}
3679 @dfn{Is Greater Than Or Equal To}
3681 @dfn{Is Less Than Or Equal To}
3683 The comparison operators can be used as infix operators. A true results has a
3684 value of -1 whereas a false result has a value of 0. Note, these operators
3685 perform signed comparisons.
3688 @item Lowest Precedence
3697 These two logical operations can be used to combine the results of sub
3698 expressions. Note, unlike the comparison operators a true result returns a
3699 value of 1 but a false results does still return 0. Also note that the logical
3700 or operator has a slightly lower precedence than logical and.
3705 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3706 address; you can only have a defined section in one of the two arguments.
3709 @chapter Assembler Directives
3711 @cindex directives, machine independent
3712 @cindex pseudo-ops, machine independent
3713 @cindex machine independent directives
3714 All assembler directives have names that begin with a period (@samp{.}).
3715 The rest of the name is letters, usually in lower case.
3717 This chapter discusses directives that are available regardless of the
3718 target machine configuration for the @sc{gnu} assembler.
3720 Some machine configurations provide additional directives.
3721 @xref{Machine Dependencies}.
3724 @ifset machine-directives
3725 @xref{Machine Dependencies} for additional directives.
3730 * Abort:: @code{.abort}
3732 * ABORT:: @code{.ABORT}
3735 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3736 * Altmacro:: @code{.altmacro}
3737 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3738 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3739 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3740 * Byte:: @code{.byte @var{expressions}}
3741 * Comm:: @code{.comm @var{symbol} , @var{length} }
3743 * CFI directives:: @code{.cfi_startproc}, @code{.cfi_endproc}, etc.
3745 * Data:: @code{.data @var{subsection}}
3747 * Def:: @code{.def @var{name}}
3750 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3756 * Double:: @code{.double @var{flonums}}
3757 * Eject:: @code{.eject}
3758 * Else:: @code{.else}
3759 * Elseif:: @code{.elseif}
3762 * Endef:: @code{.endef}
3765 * Endfunc:: @code{.endfunc}
3766 * Endif:: @code{.endif}
3767 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3768 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3769 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
3771 * Error:: @code{.error @var{string}}
3772 * Exitm:: @code{.exitm}
3773 * Extern:: @code{.extern}
3774 * Fail:: @code{.fail}
3775 @ifclear no-file-dir
3776 * File:: @code{.file @var{string}}
3779 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3780 * Float:: @code{.float @var{flonums}}
3781 * Func:: @code{.func}
3782 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3784 * Hidden:: @code{.hidden @var{names}}
3787 * hword:: @code{.hword @var{expressions}}
3788 * Ident:: @code{.ident}
3789 * If:: @code{.if @var{absolute expression}}
3790 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3791 * Include:: @code{.include "@var{file}"}
3792 * Int:: @code{.int @var{expressions}}
3794 * Internal:: @code{.internal @var{names}}
3797 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3798 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3799 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3800 * Lflags:: @code{.lflags}
3801 @ifclear no-line-dir
3802 * Line:: @code{.line @var{line-number}}
3805 * Linkonce:: @code{.linkonce [@var{type}]}
3806 * List:: @code{.list}
3807 * Ln:: @code{.ln @var{line-number}}
3809 * LNS directives:: @code{.file}, @code{.loc}, etc.
3811 * Long:: @code{.long @var{expressions}}
3813 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3816 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3817 * MRI:: @code{.mri @var{val}}
3818 * Noaltmacro:: @code{.noaltmacro}
3819 * Nolist:: @code{.nolist}
3820 * Octa:: @code{.octa @var{bignums}}
3821 * Org:: @code{.org @var{new-lc} , @var{fill}}
3822 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3824 * PopSection:: @code{.popsection}
3825 * Previous:: @code{.previous}
3828 * Print:: @code{.print @var{string}}
3830 * Protected:: @code{.protected @var{names}}
3833 * Psize:: @code{.psize @var{lines}, @var{columns}}
3834 * Purgem:: @code{.purgem @var{name}}
3836 * PushSection:: @code{.pushsection @var{name}}
3839 * Quad:: @code{.quad @var{bignums}}
3840 * Rept:: @code{.rept @var{count}}
3841 * Sbttl:: @code{.sbttl "@var{subheading}"}
3843 * Scl:: @code{.scl @var{class}}
3846 * Section:: @code{.section @var{name}}
3849 * Set:: @code{.set @var{symbol}, @var{expression}}
3850 * Short:: @code{.short @var{expressions}}
3851 * Single:: @code{.single @var{flonums}}
3853 * Size:: @code{.size [@var{name} , @var{expression}]}
3856 * Skip:: @code{.skip @var{size} , @var{fill}}
3857 * Sleb128:: @code{.sleb128 @var{expressions}}
3858 * Space:: @code{.space @var{size} , @var{fill}}
3860 * Stab:: @code{.stabd, .stabn, .stabs}
3863 * String:: @code{.string "@var{str}"}
3864 * Struct:: @code{.struct @var{expression}}
3866 * SubSection:: @code{.subsection}
3867 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3871 * Tag:: @code{.tag @var{structname}}
3874 * Text:: @code{.text @var{subsection}}
3875 * Title:: @code{.title "@var{heading}"}
3877 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3880 * Uleb128:: @code{.uleb128 @var{expressions}}
3882 * Val:: @code{.val @var{addr}}
3886 * Version:: @code{.version "@var{string}"}
3887 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3888 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3891 * Warning:: @code{.warning @var{string}}
3892 * Weak:: @code{.weak @var{names}}
3893 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
3894 * Word:: @code{.word @var{expressions}}
3895 * Deprecated:: Deprecated Directives
3899 @section @code{.abort}
3901 @cindex @code{abort} directive
3902 @cindex stopping the assembly
3903 This directive stops the assembly immediately. It is for
3904 compatibility with other assemblers. The original idea was that the
3905 assembly language source would be piped into the assembler. If the sender
3906 of the source quit, it could use this directive tells @command{@value{AS}} to
3907 quit also. One day @code{.abort} will not be supported.
3911 @section @code{.ABORT}
3913 @cindex @code{ABORT} directive
3914 When producing COFF output, @command{@value{AS}} accepts this directive as a
3915 synonym for @samp{.abort}.
3918 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3924 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3926 @cindex padding the location counter
3927 @cindex @code{align} directive
3928 Pad the location counter (in the current subsection) to a particular storage
3929 boundary. The first expression (which must be absolute) is the alignment
3930 required, as described below.
3932 The second expression (also absolute) gives the fill value to be stored in the
3933 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3934 padding bytes are normally zero. However, on some systems, if the section is
3935 marked as containing code and the fill value is omitted, the space is filled
3936 with no-op instructions.
3938 The third expression is also absolute, and is also optional. If it is present,
3939 it is the maximum number of bytes that should be skipped by this alignment
3940 directive. If doing the alignment would require skipping more bytes than the
3941 specified maximum, then the alignment is not done at all. You can omit the
3942 fill value (the second argument) entirely by simply using two commas after the
3943 required alignment; this can be useful if you want the alignment to be filled
3944 with no-op instructions when appropriate.
3946 The way the required alignment is specified varies from system to system.
3947 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
3948 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
3949 alignment request in bytes. For example @samp{.align 8} advances
3950 the location counter until it is a multiple of 8. If the location counter
3951 is already a multiple of 8, no change is needed. For the tic54x, the
3952 first expression is the alignment request in words.
3954 For other systems, including the i386 using a.out format, and the arm and
3955 strongarm, it is the
3956 number of low-order zero bits the location counter must have after
3957 advancement. For example @samp{.align 3} advances the location
3958 counter until it a multiple of 8. If the location counter is already a
3959 multiple of 8, no change is needed.
3961 This inconsistency is due to the different behaviors of the various
3962 native assemblers for these systems which GAS must emulate.
3963 GAS also provides @code{.balign} and @code{.p2align} directives,
3964 described later, which have a consistent behavior across all
3965 architectures (but are specific to GAS).
3968 @section @code{.ascii "@var{string}"}@dots{}
3970 @cindex @code{ascii} directive
3971 @cindex string literals
3972 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3973 separated by commas. It assembles each string (with no automatic
3974 trailing zero byte) into consecutive addresses.
3977 @section @code{.asciz "@var{string}"}@dots{}
3979 @cindex @code{asciz} directive
3980 @cindex zero-terminated strings
3981 @cindex null-terminated strings
3982 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3983 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3986 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3988 @cindex padding the location counter given number of bytes
3989 @cindex @code{balign} directive
3990 Pad the location counter (in the current subsection) to a particular
3991 storage boundary. The first expression (which must be absolute) is the
3992 alignment request in bytes. For example @samp{.balign 8} advances
3993 the location counter until it is a multiple of 8. If the location counter
3994 is already a multiple of 8, no change is needed.
3996 The second expression (also absolute) gives the fill value to be stored in the
3997 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3998 padding bytes are normally zero. However, on some systems, if the section is
3999 marked as containing code and the fill value is omitted, the space is filled
4000 with no-op instructions.
4002 The third expression is also absolute, and is also optional. If it is present,
4003 it is the maximum number of bytes that should be skipped by this alignment
4004 directive. If doing the alignment would require skipping more bytes than the
4005 specified maximum, then the alignment is not done at all. You can omit the
4006 fill value (the second argument) entirely by simply using two commas after the
4007 required alignment; this can be useful if you want the alignment to be filled
4008 with no-op instructions when appropriate.
4010 @cindex @code{balignw} directive
4011 @cindex @code{balignl} directive
4012 The @code{.balignw} and @code{.balignl} directives are variants of the
4013 @code{.balign} directive. The @code{.balignw} directive treats the fill
4014 pattern as a two byte word value. The @code{.balignl} directives treats the
4015 fill pattern as a four byte longword value. For example, @code{.balignw
4016 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4017 filled in with the value 0x368d (the exact placement of the bytes depends upon
4018 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4022 @section @code{.byte @var{expressions}}
4024 @cindex @code{byte} directive
4025 @cindex integers, one byte
4026 @code{.byte} expects zero or more expressions, separated by commas.
4027 Each expression is assembled into the next byte.
4030 @section @code{.comm @var{symbol} , @var{length} }
4032 @cindex @code{comm} directive
4033 @cindex symbol, common
4034 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4035 common symbol in one object file may be merged with a defined or common symbol
4036 of the same name in another object file. If @code{@value{LD}} does not see a
4037 definition for the symbol--just one or more common symbols--then it will
4038 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4039 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4040 the same name, and they do not all have the same size, it will allocate space
4041 using the largest size.
4044 When using ELF, the @code{.comm} directive takes an optional third argument.
4045 This is the desired alignment of the symbol, specified as a byte boundary (for
4046 example, an alignment of 16 means that the least significant 4 bits of the
4047 address should be zero). The alignment must be an absolute expression, and it
4048 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
4049 for the common symbol, it will use the alignment when placing the symbol. If
4050 no alignment is specified, @command{@value{AS}} will set the alignment to the
4051 largest power of two less than or equal to the size of the symbol, up to a
4056 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4057 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4060 @node CFI directives
4061 @section @code{.cfi_startproc}
4062 @cindex @code{cfi_startproc} directive
4063 @code{.cfi_startproc} is used at the beginning of each function that
4064 should have an entry in @code{.eh_frame}. It initializes some internal
4065 data structures and emits architecture dependent initial CFI instructions.
4066 Don't forget to close the function by
4067 @code{.cfi_endproc}.
4069 @section @code{.cfi_endproc}
4070 @cindex @code{cfi_endproc} directive
4071 @code{.cfi_endproc} is used at the end of a function where it closes its
4072 unwind entry previously opened by
4073 @code{.cfi_startproc}. and emits it to @code{.eh_frame}.
4075 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4076 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4077 address from @var{register} and add @var{offset} to it}.
4079 @section @code{.cfi_def_cfa_register @var{register}}
4080 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4081 now on @var{register} will be used instead of the old one. Offset
4084 @section @code{.cfi_def_cfa_offset @var{offset}}
4085 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4086 remains the same, but @var{offset} is new. Note that it is the
4087 absolute offset that will be added to a defined register to compute
4090 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4091 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4092 value that is added/substracted from the previous offset.
4094 @section @code{.cfi_offset @var{register}, @var{offset}}
4095 Previous value of @var{register} is saved at offset @var{offset} from
4098 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4099 Previous value of @var{register} is saved at offset @var{offset} from
4100 the current CFA register. This is transformed to @code{.cfi_offset}
4101 using the known displacement of the CFA register from the CFA.
4102 This is often easier to use, because the number will match the
4103 code it's annotating.
4105 @section @code{.cfi_signal_frame}
4106 Mark current function as signal trampoline.
4108 @section @code{.cfi_window_save}
4109 SPARC register window has been saved.
4111 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4112 Allows the user to add arbitrary bytes to the unwind info. One
4113 might use this to add OS-specific CFI opcodes, or generic CFI
4114 opcodes that GAS does not yet support.
4116 @node LNS directives
4117 @section @code{.file @var{fileno} @var{filename}}
4118 @cindex @code{file} directive
4119 When emitting dwarf2 line number information @code{.file} assigns filenames
4120 to the @code{.debug_line} file name table. The @var{fileno} operand should
4121 be a unique positive integer to use as the index of the entry in the table.
4122 The @var{filename} operand is a C string literal.
4124 The detail of filename indicies is exposed to the user because the filename
4125 table is shared with the @code{.debug_info} section of the dwarf2 debugging
4126 information, and thus the user must know the exact indicies that table
4129 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
4130 @cindex @code{loc} directive
4131 The @code{.loc} directive will add row to the @code{.debug_line} line
4132 number matrix corresponding to the immediately following assembly
4133 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
4134 arguments will be applied to the @code{.debug_line} state machine before
4137 The @var{options} are a sequence of the following tokens in any order:
4141 This option will set the @code{basic_block} register in the
4142 @code{.debug_line} state machine to @code{true}.
4145 This option will set the @code{prologue_end} register in the
4146 @code{.debug_line} state machine to @code{true}.
4148 @item epilogue_begin
4149 This option will set the @code{epilogue_begin} register in the
4150 @code{.debug_line} state machine to @code{true}.
4152 @item is_stmt @var{value}
4153 This option will set the @code{is_stmt} register in the
4154 @code{.debug_line} state machine to @code{value}, which must be
4157 @item isa @var{value}
4158 This directive will set the @code{isa} register in the @code{.debug_line}
4159 state machine to @var{value}, which must be an unsigned integer.
4163 @section @code{.loc_mark_blocks @var{enable}}
4164 @cindex @code{loc_mark_blocks} directive
4165 The @code{.loc_mark_blocks} directive makes the assembler emit an entry
4166 to the @code{.debug_line} line number matrix with the @code{basic_block}
4167 register in the state machine set whenever a code label is seen.
4168 The @var{enable} argument should be either 1 or 0, to enable or disable
4169 this function respectively.
4172 @section @code{.data @var{subsection}}
4174 @cindex @code{data} directive
4175 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4176 end of the data subsection numbered @var{subsection} (which is an
4177 absolute expression). If @var{subsection} is omitted, it defaults
4182 @section @code{.def @var{name}}
4184 @cindex @code{def} directive
4185 @cindex COFF symbols, debugging
4186 @cindex debugging COFF symbols
4187 Begin defining debugging information for a symbol @var{name}; the
4188 definition extends until the @code{.endef} directive is encountered.
4191 This directive is only observed when @command{@value{AS}} is configured for COFF
4192 format output; when producing @code{b.out}, @samp{.def} is recognized,
4199 @section @code{.desc @var{symbol}, @var{abs-expression}}
4201 @cindex @code{desc} directive
4202 @cindex COFF symbol descriptor
4203 @cindex symbol descriptor, COFF
4204 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4205 to the low 16 bits of an absolute expression.
4208 The @samp{.desc} directive is not available when @command{@value{AS}} is
4209 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4210 object format. For the sake of compatibility, @command{@value{AS}} accepts
4211 it, but produces no output, when configured for COFF.
4217 @section @code{.dim}
4219 @cindex @code{dim} directive
4220 @cindex COFF auxiliary symbol information
4221 @cindex auxiliary symbol information, COFF
4222 This directive is generated by compilers to include auxiliary debugging
4223 information in the symbol table. It is only permitted inside
4224 @code{.def}/@code{.endef} pairs.
4227 @samp{.dim} is only meaningful when generating COFF format output; when
4228 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4234 @section @code{.double @var{flonums}}
4236 @cindex @code{double} directive
4237 @cindex floating point numbers (double)
4238 @code{.double} expects zero or more flonums, separated by commas. It
4239 assembles floating point numbers.
4241 The exact kind of floating point numbers emitted depends on how
4242 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4246 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4247 in @sc{ieee} format.
4252 @section @code{.eject}
4254 @cindex @code{eject} directive
4255 @cindex new page, in listings
4256 @cindex page, in listings
4257 @cindex listing control: new page
4258 Force a page break at this point, when generating assembly listings.
4261 @section @code{.else}
4263 @cindex @code{else} directive
4264 @code{.else} is part of the @command{@value{AS}} support for conditional
4265 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
4266 of code to be assembled if the condition for the preceding @code{.if}
4270 @section @code{.elseif}
4272 @cindex @code{elseif} directive
4273 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4274 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
4275 @code{.if} block that would otherwise fill the entire @code{.else} section.
4278 @section @code{.end}
4280 @cindex @code{end} directive
4281 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4282 process anything in the file past the @code{.end} directive.
4286 @section @code{.endef}
4288 @cindex @code{endef} directive
4289 This directive flags the end of a symbol definition begun with
4293 @samp{.endef} is only meaningful when generating COFF format output; if
4294 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4295 directive but ignores it.
4300 @section @code{.endfunc}
4301 @cindex @code{endfunc} directive
4302 @code{.endfunc} marks the end of a function specified with @code{.func}.
4305 @section @code{.endif}
4307 @cindex @code{endif} directive
4308 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4309 it marks the end of a block of code that is only assembled
4310 conditionally. @xref{If,,@code{.if}}.
4313 @section @code{.equ @var{symbol}, @var{expression}}
4315 @cindex @code{equ} directive
4316 @cindex assigning values to symbols
4317 @cindex symbols, assigning values to
4318 This directive sets the value of @var{symbol} to @var{expression}.
4319 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
4322 The syntax for @code{equ} on the HPPA is
4323 @samp{@var{symbol} .equ @var{expression}}.
4327 The syntax for @code{equ} on the Z80 is
4328 @samp{@var{symbol} equ @var{expression}}.
4329 On the Z80 it is an eror if @var{symbol} is already defined,
4330 but the symbol is not protected from later redefinition,
4331 compare @xref{Equiv}.
4335 @section @code{.equiv @var{symbol}, @var{expression}}
4336 @cindex @code{equiv} directive
4337 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4338 the assembler will signal an error if @var{symbol} is already defined. Note a
4339 symbol which has been referenced but not actually defined is considered to be
4342 Except for the contents of the error message, this is roughly equivalent to
4349 plus it protects the symbol from later redefinition.
4352 @section @code{.eqv @var{symbol}, @var{expression}}
4353 @cindex @code{eqv} directive
4354 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4355 evaluate the expression or any part of it immediately. Instead each time
4356 the resulting symbol is used in an expression, a snapshot of its current
4360 @section @code{.err}
4361 @cindex @code{err} directive
4362 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4363 message and, unless the @option{-Z} option was used, it will not generate an
4364 object file. This can be used to signal an error in conditionally compiled code.
4367 @section @code{.error "@var{string}"}
4368 @cindex error directive
4370 Similarly to @code{.err}, this directive emits an error, but you can specify a
4371 string that will be emitted as the error message. If you don't specify the
4372 message, it defaults to @code{".error directive invoked in source file"}.
4373 @xref{Errors, ,Error and Warning Messages}.
4376 .error "This code has not been assembled and tested."
4380 @section @code{.exitm}
4381 Exit early from the current macro definition. @xref{Macro}.
4384 @section @code{.extern}
4386 @cindex @code{extern} directive
4387 @code{.extern} is accepted in the source program---for compatibility
4388 with other assemblers---but it is ignored. @command{@value{AS}} treats
4389 all undefined symbols as external.
4392 @section @code{.fail @var{expression}}
4394 @cindex @code{fail} directive
4395 Generates an error or a warning. If the value of the @var{expression} is 500
4396 or more, @command{@value{AS}} will print a warning message. If the value is less
4397 than 500, @command{@value{AS}} will print an error message. The message will
4398 include the value of @var{expression}. This can occasionally be useful inside
4399 complex nested macros or conditional assembly.
4401 @ifclear no-file-dir
4403 @section @code{.file @var{string}}
4405 @cindex @code{file} directive
4406 @cindex logical file name
4407 @cindex file name, logical
4408 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4409 file. @var{string} is the new file name. In general, the filename is
4410 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4411 to specify an empty file name, you must give the quotes--@code{""}. This
4412 statement may go away in future: it is only recognized to be compatible with
4413 old @command{@value{AS}} programs.
4417 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4419 @cindex @code{fill} directive
4420 @cindex writing patterns in memory
4421 @cindex patterns, writing in memory
4422 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4423 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4424 may be zero or more. @var{Size} may be zero or more, but if it is
4425 more than 8, then it is deemed to have the value 8, compatible with
4426 other people's assemblers. The contents of each @var{repeat} bytes
4427 is taken from an 8-byte number. The highest order 4 bytes are
4428 zero. The lowest order 4 bytes are @var{value} rendered in the
4429 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4430 Each @var{size} bytes in a repetition is taken from the lowest order
4431 @var{size} bytes of this number. Again, this bizarre behavior is
4432 compatible with other people's assemblers.
4434 @var{size} and @var{value} are optional.
4435 If the second comma and @var{value} are absent, @var{value} is
4436 assumed zero. If the first comma and following tokens are absent,
4437 @var{size} is assumed to be 1.
4440 @section @code{.float @var{flonums}}
4442 @cindex floating point numbers (single)
4443 @cindex @code{float} directive
4444 This directive assembles zero or more flonums, separated by commas. It
4445 has the same effect as @code{.single}.
4447 The exact kind of floating point numbers emitted depends on how
4448 @command{@value{AS}} is configured.
4449 @xref{Machine Dependencies}.
4453 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4454 in @sc{ieee} format.
4459 @section @code{.func @var{name}[,@var{label}]}
4460 @cindex @code{func} directive
4461 @code{.func} emits debugging information to denote function @var{name}, and
4462 is ignored unless the file is assembled with debugging enabled.
4463 Only @samp{--gstabs[+]} is currently supported.
4464 @var{label} is the entry point of the function and if omitted @var{name}
4465 prepended with the @samp{leading char} is used.
4466 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4467 All functions are currently defined to have @code{void} return type.
4468 The function must be terminated with @code{.endfunc}.
4471 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4473 @cindex @code{global} directive
4474 @cindex symbol, making visible to linker
4475 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4476 @var{symbol} in your partial program, its value is made available to
4477 other partial programs that are linked with it. Otherwise,
4478 @var{symbol} takes its attributes from a symbol of the same name
4479 from another file linked into the same program.
4481 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4482 compatibility with other assemblers.
4485 On the HPPA, @code{.global} is not always enough to make it accessible to other
4486 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4487 @xref{HPPA Directives,, HPPA Assembler Directives}.
4492 @section @code{.hidden @var{names}}
4494 @cindex @code{hidden} directive
4496 This is one of the ELF visibility directives. The other two are
4497 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4498 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4500 This directive overrides the named symbols default visibility (which is set by
4501 their binding: local, global or weak). The directive sets the visibility to
4502 @code{hidden} which means that the symbols are not visible to other components.
4503 Such symbols are always considered to be @code{protected} as well.
4507 @section @code{.hword @var{expressions}}
4509 @cindex @code{hword} directive
4510 @cindex integers, 16-bit
4511 @cindex numbers, 16-bit
4512 @cindex sixteen bit integers
4513 This expects zero or more @var{expressions}, and emits
4514 a 16 bit number for each.
4517 This directive is a synonym for @samp{.short}; depending on the target
4518 architecture, it may also be a synonym for @samp{.word}.
4522 This directive is a synonym for @samp{.short}.
4525 This directive is a synonym for both @samp{.short} and @samp{.word}.
4530 @section @code{.ident}
4532 @cindex @code{ident} directive
4534 This directive is used by some assemblers to place tags in object files. The
4535 behavior of this directive varies depending on the target. When using the
4536 a.out object file format, @command{@value{AS}} simply accepts the directive for
4537 source-file compatibility with existing assemblers, but does not emit anything
4538 for it. When using COFF, comments are emitted to the @code{.comment} or
4539 @code{.rdata} section, depending on the target. When using ELF, comments are
4540 emitted to the @code{.comment} section.
4543 @section @code{.if @var{absolute expression}}
4545 @cindex conditional assembly
4546 @cindex @code{if} directive
4547 @code{.if} marks the beginning of a section of code which is only
4548 considered part of the source program being assembled if the argument
4549 (which must be an @var{absolute expression}) is non-zero. The end of
4550 the conditional section of code must be marked by @code{.endif}
4551 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4552 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4553 If you have several conditions to check, @code{.elseif} may be used to avoid
4554 nesting blocks if/else within each subsequent @code{.else} block.
4556 The following variants of @code{.if} are also supported:
4558 @cindex @code{ifdef} directive
4559 @item .ifdef @var{symbol}
4560 Assembles the following section of code if the specified @var{symbol}
4561 has been defined. Note a symbol which has been referenced but not yet defined
4562 is considered to be undefined.
4564 @cindex @code{ifb} directive
4565 @item .ifb @var{text}
4566 Assembles the following section of code if the operand is blank (empty).
4568 @cindex @code{ifc} directive
4569 @item .ifc @var{string1},@var{string2}
4570 Assembles the following section of code if the two strings are the same. The
4571 strings may be optionally quoted with single quotes. If they are not quoted,
4572 the first string stops at the first comma, and the second string stops at the
4573 end of the line. Strings which contain whitespace should be quoted. The
4574 string comparison is case sensitive.
4576 @cindex @code{ifeq} directive
4577 @item .ifeq @var{absolute expression}
4578 Assembles the following section of code if the argument is zero.
4580 @cindex @code{ifeqs} directive
4581 @item .ifeqs @var{string1},@var{string2}
4582 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4584 @cindex @code{ifge} directive
4585 @item .ifge @var{absolute expression}
4586 Assembles the following section of code if the argument is greater than or
4589 @cindex @code{ifgt} directive
4590 @item .ifgt @var{absolute expression}
4591 Assembles the following section of code if the argument is greater than zero.
4593 @cindex @code{ifle} directive
4594 @item .ifle @var{absolute expression}
4595 Assembles the following section of code if the argument is less than or equal
4598 @cindex @code{iflt} directive
4599 @item .iflt @var{absolute expression}
4600 Assembles the following section of code if the argument is less than zero.
4602 @cindex @code{ifnb} directive
4603 @item .ifnb @var{text}
4604 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
4605 following section of code if the operand is non-blank (non-empty).
4607 @cindex @code{ifnc} directive
4608 @item .ifnc @var{string1},@var{string2}.
4609 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4610 following section of code if the two strings are not the same.
4612 @cindex @code{ifndef} directive
4613 @cindex @code{ifnotdef} directive
4614 @item .ifndef @var{symbol}
4615 @itemx .ifnotdef @var{symbol}
4616 Assembles the following section of code if the specified @var{symbol}
4617 has not been defined. Both spelling variants are equivalent. Note a symbol
4618 which has been referenced but not yet defined is considered to be undefined.
4620 @cindex @code{ifne} directive
4621 @item .ifne @var{absolute expression}
4622 Assembles the following section of code if the argument is not equal to zero
4623 (in other words, this is equivalent to @code{.if}).
4625 @cindex @code{ifnes} directive
4626 @item .ifnes @var{string1},@var{string2}
4627 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4628 following section of code if the two strings are not the same.
4632 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4634 @cindex @code{incbin} directive
4635 @cindex binary files, including
4636 The @code{incbin} directive includes @var{file} verbatim at the current
4637 location. You can control the search paths used with the @samp{-I} command-line
4638 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4641 The @var{skip} argument skips a number of bytes from the start of the
4642 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4643 read. Note that the data is not aligned in any way, so it is the user's
4644 responsibility to make sure that proper alignment is provided both before and
4645 after the @code{incbin} directive.
4648 @section @code{.include "@var{file}"}
4650 @cindex @code{include} directive
4651 @cindex supporting files, including
4652 @cindex files, including
4653 This directive provides a way to include supporting files at specified
4654 points in your source program. The code from @var{file} is assembled as
4655 if it followed the point of the @code{.include}; when the end of the
4656 included file is reached, assembly of the original file continues. You
4657 can control the search paths used with the @samp{-I} command-line option
4658 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4662 @section @code{.int @var{expressions}}
4664 @cindex @code{int} directive
4665 @cindex integers, 32-bit
4666 Expect zero or more @var{expressions}, of any section, separated by commas.
4667 For each expression, emit a number that, at run time, is the value of that
4668 expression. The byte order and bit size of the number depends on what kind
4669 of target the assembly is for.
4673 On most forms of the H8/300, @code{.int} emits 16-bit
4674 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4681 @section @code{.internal @var{names}}
4683 @cindex @code{internal} directive
4685 This is one of the ELF visibility directives. The other two are
4686 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4687 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4689 This directive overrides the named symbols default visibility (which is set by
4690 their binding: local, global or weak). The directive sets the visibility to
4691 @code{internal} which means that the symbols are considered to be @code{hidden}
4692 (i.e., not visible to other components), and that some extra, processor specific
4693 processing must also be performed upon the symbols as well.
4697 @section @code{.irp @var{symbol},@var{values}}@dots{}
4699 @cindex @code{irp} directive
4700 Evaluate a sequence of statements assigning different values to @var{symbol}.
4701 The sequence of statements starts at the @code{.irp} directive, and is
4702 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4703 set to @var{value}, and the sequence of statements is assembled. If no
4704 @var{value} is listed, the sequence of statements is assembled once, with
4705 @var{symbol} set to the null string. To refer to @var{symbol} within the
4706 sequence of statements, use @var{\symbol}.
4708 For example, assembling
4716 is equivalent to assembling
4724 For some caveats with the spelling of @var{symbol}, see also the discussion
4728 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4730 @cindex @code{irpc} directive
4731 Evaluate a sequence of statements assigning different values to @var{symbol}.
4732 The sequence of statements starts at the @code{.irpc} directive, and is
4733 terminated by an @code{.endr} directive. For each character in @var{value},
4734 @var{symbol} is set to the character, and the sequence of statements is
4735 assembled. If no @var{value} is listed, the sequence of statements is
4736 assembled once, with @var{symbol} set to the null string. To refer to
4737 @var{symbol} within the sequence of statements, use @var{\symbol}.
4739 For example, assembling
4747 is equivalent to assembling
4755 For some caveats with the spelling of @var{symbol}, see also the discussion
4759 @section @code{.lcomm @var{symbol} , @var{length}}
4761 @cindex @code{lcomm} directive
4762 @cindex local common symbols
4763 @cindex symbols, local common
4764 Reserve @var{length} (an absolute expression) bytes for a local common
4765 denoted by @var{symbol}. The section and value of @var{symbol} are
4766 those of the new local common. The addresses are allocated in the bss
4767 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4768 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4769 not visible to @code{@value{LD}}.
4772 Some targets permit a third argument to be used with @code{.lcomm}. This
4773 argument specifies the desired alignment of the symbol in the bss section.
4777 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4778 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4782 @section @code{.lflags}
4784 @cindex @code{lflags} directive (ignored)
4785 @command{@value{AS}} accepts this directive, for compatibility with other
4786 assemblers, but ignores it.
4788 @ifclear no-line-dir
4790 @section @code{.line @var{line-number}}
4792 @cindex @code{line} directive
4796 @section @code{.ln @var{line-number}}
4798 @cindex @code{ln} directive
4800 @cindex logical line number
4802 Change the logical line number. @var{line-number} must be an absolute
4803 expression. The next line has that logical line number. Therefore any other
4804 statements on the current line (after a statement separator character) are
4805 reported as on logical line number @var{line-number} @minus{} 1. One day
4806 @command{@value{AS}} will no longer support this directive: it is recognized only
4807 for compatibility with existing assembler programs.
4811 @ifclear no-line-dir
4812 Even though this is a directive associated with the @code{a.out} or
4813 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4814 when producing COFF output, and treats @samp{.line} as though it
4815 were the COFF @samp{.ln} @emph{if} it is found outside a
4816 @code{.def}/@code{.endef} pair.
4818 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4819 used by compilers to generate auxiliary symbol information for
4824 @section @code{.linkonce [@var{type}]}
4826 @cindex @code{linkonce} directive
4827 @cindex common sections
4828 Mark the current section so that the linker only includes a single copy of it.
4829 This may be used to include the same section in several different object files,
4830 but ensure that the linker will only include it once in the final output file.
4831 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4832 Duplicate sections are detected based on the section name, so it should be
4835 This directive is only supported by a few object file formats; as of this
4836 writing, the only object file format which supports it is the Portable
4837 Executable format used on Windows NT.
4839 The @var{type} argument is optional. If specified, it must be one of the
4840 following strings. For example:
4844 Not all types may be supported on all object file formats.
4848 Silently discard duplicate sections. This is the default.
4851 Warn if there are duplicate sections, but still keep only one copy.
4854 Warn if any of the duplicates have different sizes.
4857 Warn if any of the duplicates do not have exactly the same contents.
4861 @section @code{.ln @var{line-number}}
4863 @cindex @code{ln} directive
4864 @ifclear no-line-dir
4865 @samp{.ln} is a synonym for @samp{.line}.
4868 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4869 must be an absolute expression. The next line has that logical
4870 line number, so any other statements on the current line (after a
4871 statement separator character @code{;}) are reported as on logical
4872 line number @var{line-number} @minus{} 1.
4875 This directive is accepted, but ignored, when @command{@value{AS}} is
4876 configured for @code{b.out}; its effect is only associated with COFF
4882 @section @code{.mri @var{val}}
4884 @cindex @code{mri} directive
4885 @cindex MRI mode, temporarily
4886 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4887 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4888 affects code assembled until the next @code{.mri} directive, or until the end
4889 of the file. @xref{M, MRI mode, MRI mode}.
4892 @section @code{.list}
4894 @cindex @code{list} directive
4895 @cindex listing control, turning on
4896 Control (in conjunction with the @code{.nolist} directive) whether or
4897 not assembly listings are generated. These two directives maintain an
4898 internal counter (which is zero initially). @code{.list} increments the
4899 counter, and @code{.nolist} decrements it. Assembly listings are
4900 generated whenever the counter is greater than zero.
4902 By default, listings are disabled. When you enable them (with the
4903 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4904 the initial value of the listing counter is one.
4907 @section @code{.long @var{expressions}}
4909 @cindex @code{long} directive
4910 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4913 @c no one seems to know what this is for or whether this description is
4914 @c what it really ought to do
4916 @section @code{.lsym @var{symbol}, @var{expression}}
4918 @cindex @code{lsym} directive
4919 @cindex symbol, not referenced in assembly
4920 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4921 the hash table, ensuring it cannot be referenced by name during the
4922 rest of the assembly. This sets the attributes of the symbol to be
4923 the same as the expression value:
4925 @var{other} = @var{descriptor} = 0
4926 @var{type} = @r{(section of @var{expression})}
4927 @var{value} = @var{expression}
4930 The new symbol is not flagged as external.
4934 @section @code{.macro}
4937 The commands @code{.macro} and @code{.endm} allow you to define macros that
4938 generate assembly output. For example, this definition specifies a macro
4939 @code{sum} that puts a sequence of numbers into memory:
4942 .macro sum from=0, to=5
4951 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4963 @item .macro @var{macname}
4964 @itemx .macro @var{macname} @var{macargs} @dots{}
4965 @cindex @code{macro} directive
4966 Begin the definition of a macro called @var{macname}. If your macro
4967 definition requires arguments, specify their names after the macro name,
4968 separated by commas or spaces. You can qualify the macro argument to
4969 indicate whether all invocations must specify a non-blank value (through
4970 @samp{:@code{req}}), or whether it takes all of the remaining arguments
4971 (through @samp{:@code{vararg}}). You can supply a default value for any
4972 macro argument by following the name with @samp{=@var{deflt}}. You
4973 cannot define two macros with the same @var{macname} unless it has been
4974 subject to the @code{.purgem} directive (@xref{Purgem}.) between the two
4975 definitions. For example, these are all valid @code{.macro} statements:
4979 Begin the definition of a macro called @code{comm}, which takes no
4982 @item .macro plus1 p, p1
4983 @itemx .macro plus1 p p1
4984 Either statement begins the definition of a macro called @code{plus1},
4985 which takes two arguments; within the macro definition, write
4986 @samp{\p} or @samp{\p1} to evaluate the arguments.
4988 @item .macro reserve_str p1=0 p2
4989 Begin the definition of a macro called @code{reserve_str}, with two
4990 arguments. The first argument has a default value, but not the second.
4991 After the definition is complete, you can call the macro either as
4992 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
4993 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
4994 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
4995 @samp{0}, and @samp{\p2} evaluating to @var{b}).
4998 @item .macro m p1:req, p2=0, p3:vararg
4999 Begin the definition of a macro called @code{m}, with at least three
5000 arguments. The first argument must always have a value specified, but
5001 not the second, which instead has a default value. The third formal
5002 will get assigned all remaining arguments specified at invocation time.
5004 When you call a macro, you can specify the argument values either by
5005 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5006 @samp{sum to=17, from=9}.
5008 Note that since each of the @var{macargs} can be an identifier exactly
5009 as any other one permitted by the target architecture, there may be
5010 occasional problems if the target hand-crafts special meanings to certain
5011 characters when they occur in a special position. For example, if colon
5012 (@code{:}) is generally permitted to be part of a symbol name, but the
5013 architecture specific code special-cases it when occuring as the final
5014 character of a symbol (to denote a label), then the macro parameter
5015 replacement code will have no way of knowing that and consider the whole
5016 construct (including the colon) an identifier, and check only this
5017 identifier for being the subject to parameter substitution. In this
5018 example, besides the potential of just separating identifier and colon
5019 by white space, using alternate macro syntax (@xref{Altmacro}.) and
5020 ampersand (@code{&}) as the character to separate literal text from macro
5021 parameters (or macro parameters from one another) would provide a way to
5022 achieve the same effect:
5031 This applies identically to the identifiers used in @code{.irp} (@xref{Irp}.)
5032 and @code{.irpc} (@xref{Irpc}.).
5035 @cindex @code{endm} directive
5036 Mark the end of a macro definition.
5039 @cindex @code{exitm} directive
5040 Exit early from the current macro definition.
5042 @cindex number of macros executed
5043 @cindex macros, count executed
5045 @command{@value{AS}} maintains a counter of how many macros it has
5046 executed in this pseudo-variable; you can copy that number to your
5047 output with @samp{\@@}, but @emph{only within a macro definition}.
5049 @item LOCAL @var{name} [ , @dots{} ]
5050 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5051 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5052 @xref{Altmacro,,@code{.altmacro}}.
5056 @section @code{.altmacro}
5057 Enable alternate macro mode, enabling:
5060 @item LOCAL @var{name} [ , @dots{} ]
5061 One additional directive, @code{LOCAL}, is available. It is used to
5062 generate a string replacement for each of the @var{name} arguments, and
5063 replace any instances of @var{name} in each macro expansion. The
5064 replacement string is unique in the assembly, and different for each
5065 separate macro expansion. @code{LOCAL} allows you to write macros that
5066 define symbols, without fear of conflict between separate macro expansions.
5068 @item String delimiters
5069 You can write strings delimited in these other ways besides
5070 @code{"@var{string}"}:
5073 @item '@var{string}'
5074 You can delimit strings with single-quote charaters.
5076 @item <@var{string}>
5077 You can delimit strings with matching angle brackets.
5080 @item single-character string escape
5081 To include any single character literally in a string (even if the
5082 character would otherwise have some special meaning), you can prefix the
5083 character with @samp{!} (an exclamation mark). For example, you can
5084 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
5086 @item Expression results as strings
5087 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
5088 and use the result as a string.
5092 @section @code{.noaltmacro}
5093 Disable alternate macro mode. @ref{Altmacro}
5096 @section @code{.nolist}
5098 @cindex @code{nolist} directive
5099 @cindex listing control, turning off
5100 Control (in conjunction with the @code{.list} directive) whether or
5101 not assembly listings are generated. These two directives maintain an
5102 internal counter (which is zero initially). @code{.list} increments the
5103 counter, and @code{.nolist} decrements it. Assembly listings are
5104 generated whenever the counter is greater than zero.
5107 @section @code{.octa @var{bignums}}
5109 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5110 @cindex @code{octa} directive
5111 @cindex integer, 16-byte
5112 @cindex sixteen byte integer
5113 This directive expects zero or more bignums, separated by commas. For each
5114 bignum, it emits a 16-byte integer.
5116 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5117 hence @emph{octa}-word for 16 bytes.
5120 @section @code{.org @var{new-lc} , @var{fill}}
5122 @cindex @code{org} directive
5123 @cindex location counter, advancing
5124 @cindex advancing location counter
5125 @cindex current address, advancing
5126 Advance the location counter of the current section to
5127 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5128 expression with the same section as the current subsection. That is,
5129 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5130 wrong section, the @code{.org} directive is ignored. To be compatible
5131 with former assemblers, if the section of @var{new-lc} is absolute,
5132 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5133 is the same as the current subsection.
5135 @code{.org} may only increase the location counter, or leave it
5136 unchanged; you cannot use @code{.org} to move the location counter
5139 @c double negative used below "not undefined" because this is a specific
5140 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5141 @c section. doc@cygnus.com 18feb91
5142 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5143 may not be undefined. If you really detest this restriction we eagerly await
5144 a chance to share your improved assembler.
5146 Beware that the origin is relative to the start of the section, not
5147 to the start of the subsection. This is compatible with other
5148 people's assemblers.
5150 When the location counter (of the current subsection) is advanced, the
5151 intervening bytes are filled with @var{fill} which should be an
5152 absolute expression. If the comma and @var{fill} are omitted,
5153 @var{fill} defaults to zero.
5156 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5158 @cindex padding the location counter given a power of two
5159 @cindex @code{p2align} directive
5160 Pad the location counter (in the current subsection) to a particular
5161 storage boundary. The first expression (which must be absolute) is the
5162 number of low-order zero bits the location counter must have after
5163 advancement. For example @samp{.p2align 3} advances the location
5164 counter until it a multiple of 8. If the location counter is already a
5165 multiple of 8, no change is needed.
5167 The second expression (also absolute) gives the fill value to be stored in the
5168 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5169 padding bytes are normally zero. However, on some systems, if the section is
5170 marked as containing code and the fill value is omitted, the space is filled
5171 with no-op instructions.
5173 The third expression is also absolute, and is also optional. If it is present,
5174 it is the maximum number of bytes that should be skipped by this alignment
5175 directive. If doing the alignment would require skipping more bytes than the
5176 specified maximum, then the alignment is not done at all. You can omit the
5177 fill value (the second argument) entirely by simply using two commas after the
5178 required alignment; this can be useful if you want the alignment to be filled
5179 with no-op instructions when appropriate.
5181 @cindex @code{p2alignw} directive
5182 @cindex @code{p2alignl} directive
5183 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5184 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5185 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5186 fill pattern as a four byte longword value. For example, @code{.p2alignw
5187 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5188 filled in with the value 0x368d (the exact placement of the bytes depends upon
5189 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5194 @section @code{.previous}
5196 @cindex @code{previous} directive
5197 @cindex Section Stack
5198 This is one of the ELF section stack manipulation directives. The others are
5199 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5200 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5201 (@pxref{PopSection}).
5203 This directive swaps the current section (and subsection) with most recently
5204 referenced section (and subsection) prior to this one. Multiple
5205 @code{.previous} directives in a row will flip between two sections (and their
5208 In terms of the section stack, this directive swaps the current section with
5209 the top section on the section stack.
5214 @section @code{.popsection}
5216 @cindex @code{popsection} directive
5217 @cindex Section Stack
5218 This is one of the ELF section stack manipulation directives. The others are
5219 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5220 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5223 This directive replaces the current section (and subsection) with the top
5224 section (and subsection) on the section stack. This section is popped off the
5229 @section @code{.print @var{string}}
5231 @cindex @code{print} directive
5232 @command{@value{AS}} will print @var{string} on the standard output during
5233 assembly. You must put @var{string} in double quotes.
5237 @section @code{.protected @var{names}}
5239 @cindex @code{protected} directive
5241 This is one of the ELF visibility directives. The other two are
5242 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5244 This directive overrides the named symbols default visibility (which is set by
5245 their binding: local, global or weak). The directive sets the visibility to
5246 @code{protected} which means that any references to the symbols from within the
5247 components that defines them must be resolved to the definition in that
5248 component, even if a definition in another component would normally preempt
5253 @section @code{.psize @var{lines} , @var{columns}}
5255 @cindex @code{psize} directive
5256 @cindex listing control: paper size
5257 @cindex paper size, for listings
5258 Use this directive to declare the number of lines---and, optionally, the
5259 number of columns---to use for each page, when generating listings.
5261 If you do not use @code{.psize}, listings use a default line-count
5262 of 60. You may omit the comma and @var{columns} specification; the
5263 default width is 200 columns.
5265 @command{@value{AS}} generates formfeeds whenever the specified number of
5266 lines is exceeded (or whenever you explicitly request one, using
5269 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5270 those explicitly specified with @code{.eject}.
5273 @section @code{.purgem @var{name}}
5275 @cindex @code{purgem} directive
5276 Undefine the macro @var{name}, so that later uses of the string will not be
5277 expanded. @xref{Macro}.
5281 @section @code{.pushsection @var{name} , @var{subsection}}
5283 @cindex @code{pushsection} directive
5284 @cindex Section Stack
5285 This is one of the ELF section stack manipulation directives. The others are
5286 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5287 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5290 This directive pushes the current section (and subsection) onto the
5291 top of the section stack, and then replaces the current section and
5292 subsection with @code{name} and @code{subsection}.
5296 @section @code{.quad @var{bignums}}
5298 @cindex @code{quad} directive
5299 @code{.quad} expects zero or more bignums, separated by commas. For
5300 each bignum, it emits
5302 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5303 warning message; and just takes the lowest order 8 bytes of the bignum.
5304 @cindex eight-byte integer
5305 @cindex integer, 8-byte
5307 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5308 hence @emph{quad}-word for 8 bytes.
5311 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5312 warning message; and just takes the lowest order 16 bytes of the bignum.
5313 @cindex sixteen-byte integer
5314 @cindex integer, 16-byte
5318 @section @code{.rept @var{count}}
5320 @cindex @code{rept} directive
5321 Repeat the sequence of lines between the @code{.rept} directive and the next
5322 @code{.endr} directive @var{count} times.
5324 For example, assembling
5332 is equivalent to assembling
5341 @section @code{.sbttl "@var{subheading}"}
5343 @cindex @code{sbttl} directive
5344 @cindex subtitles for listings
5345 @cindex listing control: subtitle
5346 Use @var{subheading} as the title (third line, immediately after the
5347 title line) when generating assembly listings.
5349 This directive affects subsequent pages, as well as the current page if
5350 it appears within ten lines of the top of a page.
5354 @section @code{.scl @var{class}}
5356 @cindex @code{scl} directive
5357 @cindex symbol storage class (COFF)
5358 @cindex COFF symbol storage class
5359 Set the storage-class value for a symbol. This directive may only be
5360 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5361 whether a symbol is static or external, or it may record further
5362 symbolic debugging information.
5365 The @samp{.scl} directive is primarily associated with COFF output; when
5366 configured to generate @code{b.out} output format, @command{@value{AS}}
5367 accepts this directive but ignores it.
5373 @section @code{.section @var{name}}
5375 @cindex named section
5376 Use the @code{.section} directive to assemble the following code into a section
5379 This directive is only supported for targets that actually support arbitrarily
5380 named sections; on @code{a.out} targets, for example, it is not accepted, even
5381 with a standard @code{a.out} section name.
5385 @c only print the extra heading if both COFF and ELF are set
5386 @subheading COFF Version
5389 @cindex @code{section} directive (COFF version)
5390 For COFF targets, the @code{.section} directive is used in one of the following
5394 .section @var{name}[, "@var{flags}"]
5395 .section @var{name}[, @var{subsegment}]
5398 If the optional argument is quoted, it is taken as flags to use for the
5399 section. Each flag is a single character. The following flags are recognized:
5402 bss section (uninitialized data)
5404 section is not loaded
5414 shared section (meaningful for PE targets)
5416 ignored. (For compatibility with the ELF version)
5419 If no flags are specified, the default flags depend upon the section name. If
5420 the section name is not recognized, the default will be for the section to be
5421 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5422 from the section, rather than adding them, so if they are used on their own it
5423 will be as if no flags had been specified at all.
5425 If the optional argument to the @code{.section} directive is not quoted, it is
5426 taken as a subsegment number (@pxref{Sub-Sections}).
5431 @c only print the extra heading if both COFF and ELF are set
5432 @subheading ELF Version
5435 @cindex Section Stack
5436 This is one of the ELF section stack manipulation directives. The others are
5437 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5438 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5439 @code{.previous} (@pxref{Previous}).
5441 @cindex @code{section} directive (ELF version)
5442 For ELF targets, the @code{.section} directive is used like this:
5445 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
5448 The optional @var{flags} argument is a quoted string which may contain any
5449 combination of the following characters:
5452 section is allocatable
5456 section is executable
5458 section is mergeable
5460 section contains zero terminated strings
5462 section is a member of a section group
5464 section is used for thread-local-storage
5467 The optional @var{type} argument may contain one of the following constants:
5470 section contains data
5472 section does not contain data (i.e., section only occupies space)
5474 section contains data which is used by things other than the program
5476 section contains an array of pointers to init functions
5478 section contains an array of pointers to finish functions
5479 @item @@preinit_array
5480 section contains an array of pointers to pre-init functions
5483 Many targets only support the first three section types.
5485 Note on targets where the @code{@@} character is the start of a comment (eg
5486 ARM) then another character is used instead. For example the ARM port uses the
5489 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
5490 be specified as well as an extra argument - @var{entsize} - like this:
5493 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
5496 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
5497 constants, each @var{entsize} octets long. Sections with both @code{M} and
5498 @code{S} must contain zero terminated strings where each character is
5499 @var{entsize} bytes long. The linker may remove duplicates within sections with
5500 the same name, same entity size and same flags. @var{entsize} must be an
5501 absolute expression.
5503 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
5504 be present along with an additional field like this:
5507 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
5510 The @var{GroupName} field specifies the name of the section group to which this
5511 particular section belongs. The optional linkage field can contain:
5514 indicates that only one copy of this section should be retained
5519 Note - if both the @var{M} and @var{G} flags are present then the fields for
5520 the Merge flag should come first, like this:
5523 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
5526 If no flags are specified, the default flags depend upon the section name. If
5527 the section name is not recognized, the default will be for the section to have
5528 none of the above flags: it will not be allocated in memory, nor writable, nor
5529 executable. The section will contain data.
5531 For ELF targets, the assembler supports another type of @code{.section}
5532 directive for compatibility with the Solaris assembler:
5535 .section "@var{name}"[, @var{flags}...]
5538 Note that the section name is quoted. There may be a sequence of comma
5542 section is allocatable
5546 section is executable
5548 section is used for thread local storage
5551 This directive replaces the current section and subsection. See the
5552 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
5553 some examples of how this directive and the other section stack directives
5559 @section @code{.set @var{symbol}, @var{expression}}
5561 @cindex @code{set} directive
5562 @cindex symbol value, setting
5563 Set the value of @var{symbol} to @var{expression}. This
5564 changes @var{symbol}'s value and type to conform to
5565 @var{expression}. If @var{symbol} was flagged as external, it remains
5566 flagged (@pxref{Symbol Attributes}).
5568 You may @code{.set} a symbol many times in the same assembly.
5570 If you @code{.set} a global symbol, the value stored in the object
5571 file is the last value stored into it.
5574 The syntax for @code{set} on the HPPA is
5575 @samp{@var{symbol} .set @var{expression}}.
5579 On Z80 @code{set} is a real instruction, use
5580 @samp{@var{symbol} defl @var{expression}} instead.
5584 @section @code{.short @var{expressions}}
5586 @cindex @code{short} directive
5588 @code{.short} is normally the same as @samp{.word}.
5589 @xref{Word,,@code{.word}}.
5591 In some configurations, however, @code{.short} and @code{.word} generate
5592 numbers of different lengths; @pxref{Machine Dependencies}.
5596 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5599 This expects zero or more @var{expressions}, and emits
5600 a 16 bit number for each.
5605 @section @code{.single @var{flonums}}
5607 @cindex @code{single} directive
5608 @cindex floating point numbers (single)
5609 This directive assembles zero or more flonums, separated by commas. It
5610 has the same effect as @code{.float}.
5612 The exact kind of floating point numbers emitted depends on how
5613 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5617 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5618 numbers in @sc{ieee} format.
5624 @section @code{.size}
5626 This directive is used to set the size associated with a symbol.
5630 @c only print the extra heading if both COFF and ELF are set
5631 @subheading COFF Version
5634 @cindex @code{size} directive (COFF version)
5635 For COFF targets, the @code{.size} directive is only permitted inside
5636 @code{.def}/@code{.endef} pairs. It is used like this:
5639 .size @var{expression}
5643 @samp{.size} is only meaningful when generating COFF format output; when
5644 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5651 @c only print the extra heading if both COFF and ELF are set
5652 @subheading ELF Version
5655 @cindex @code{size} directive (ELF version)
5656 For ELF targets, the @code{.size} directive is used like this:
5659 .size @var{name} , @var{expression}
5662 This directive sets the size associated with a symbol @var{name}.
5663 The size in bytes is computed from @var{expression} which can make use of label
5664 arithmetic. This directive is typically used to set the size of function
5670 @section @code{.sleb128 @var{expressions}}
5672 @cindex @code{sleb128} directive
5673 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5674 compact, variable length representation of numbers used by the DWARF
5675 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
5677 @ifclear no-space-dir
5679 @section @code{.skip @var{size} , @var{fill}}
5681 @cindex @code{skip} directive
5682 @cindex filling memory
5683 This directive emits @var{size} bytes, each of value @var{fill}. Both
5684 @var{size} and @var{fill} are absolute expressions. If the comma and
5685 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5689 @section @code{.space @var{size} , @var{fill}}
5691 @cindex @code{space} directive
5692 @cindex filling memory
5693 This directive emits @var{size} bytes, each of value @var{fill}. Both
5694 @var{size} and @var{fill} are absolute expressions. If the comma
5695 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5700 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5701 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5702 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5703 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5711 @section @code{.stabd, .stabn, .stabs}
5713 @cindex symbolic debuggers, information for
5714 @cindex @code{stab@var{x}} directives
5715 There are three directives that begin @samp{.stab}.
5716 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5717 The symbols are not entered in the @command{@value{AS}} hash table: they
5718 cannot be referenced elsewhere in the source file.
5719 Up to five fields are required:
5723 This is the symbol's name. It may contain any character except
5724 @samp{\000}, so is more general than ordinary symbol names. Some
5725 debuggers used to code arbitrarily complex structures into symbol names
5729 An absolute expression. The symbol's type is set to the low 8 bits of
5730 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5731 and debuggers choke on silly bit patterns.
5734 An absolute expression. The symbol's ``other'' attribute is set to the
5735 low 8 bits of this expression.
5738 An absolute expression. The symbol's descriptor is set to the low 16
5739 bits of this expression.
5742 An absolute expression which becomes the symbol's value.
5745 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5746 or @code{.stabs} statement, the symbol has probably already been created;
5747 you get a half-formed symbol in your object file. This is
5748 compatible with earlier assemblers!
5751 @cindex @code{stabd} directive
5752 @item .stabd @var{type} , @var{other} , @var{desc}
5754 The ``name'' of the symbol generated is not even an empty string.
5755 It is a null pointer, for compatibility. Older assemblers used a
5756 null pointer so they didn't waste space in object files with empty
5759 The symbol's value is set to the location counter,
5760 relocatably. When your program is linked, the value of this symbol
5761 is the address of the location counter when the @code{.stabd} was
5764 @cindex @code{stabn} directive
5765 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5766 The name of the symbol is set to the empty string @code{""}.
5768 @cindex @code{stabs} directive
5769 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5770 All five fields are specified.
5776 @section @code{.string} "@var{str}"
5778 @cindex string, copying to object file
5779 @cindex @code{string} directive
5781 Copy the characters in @var{str} to the object file. You may specify more than
5782 one string to copy, separated by commas. Unless otherwise specified for a
5783 particular machine, the assembler marks the end of each string with a 0 byte.
5784 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5787 @section @code{.struct @var{expression}}
5789 @cindex @code{struct} directive
5790 Switch to the absolute section, and set the section offset to @var{expression},
5791 which must be an absolute expression. You might use this as follows:
5800 This would define the symbol @code{field1} to have the value 0, the symbol
5801 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5802 value 8. Assembly would be left in the absolute section, and you would need to
5803 use a @code{.section} directive of some sort to change to some other section
5804 before further assembly.
5808 @section @code{.subsection @var{name}}
5810 @cindex @code{subsection} directive
5811 @cindex Section Stack
5812 This is one of the ELF section stack manipulation directives. The others are
5813 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5814 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5817 This directive replaces the current subsection with @code{name}. The current
5818 section is not changed. The replaced subsection is put onto the section stack
5819 in place of the then current top of stack subsection.
5824 @section @code{.symver}
5825 @cindex @code{symver} directive
5826 @cindex symbol versioning
5827 @cindex versions of symbols
5828 Use the @code{.symver} directive to bind symbols to specific version nodes
5829 within a source file. This is only supported on ELF platforms, and is
5830 typically used when assembling files to be linked into a shared library.
5831 There are cases where it may make sense to use this in objects to be bound
5832 into an application itself so as to override a versioned symbol from a
5835 For ELF targets, the @code{.symver} directive can be used like this:
5837 .symver @var{name}, @var{name2@@nodename}
5839 If the symbol @var{name} is defined within the file
5840 being assembled, the @code{.symver} directive effectively creates a symbol
5841 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5842 just don't try and create a regular alias is that the @var{@@} character isn't
5843 permitted in symbol names. The @var{name2} part of the name is the actual name
5844 of the symbol by which it will be externally referenced. The name @var{name}
5845 itself is merely a name of convenience that is used so that it is possible to
5846 have definitions for multiple versions of a function within a single source
5847 file, and so that the compiler can unambiguously know which version of a
5848 function is being mentioned. The @var{nodename} portion of the alias should be
5849 the name of a node specified in the version script supplied to the linker when
5850 building a shared library. If you are attempting to override a versioned
5851 symbol from a shared library, then @var{nodename} should correspond to the
5852 nodename of the symbol you are trying to override.
5854 If the symbol @var{name} is not defined within the file being assembled, all
5855 references to @var{name} will be changed to @var{name2@@nodename}. If no
5856 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5859 Another usage of the @code{.symver} directive is:
5861 .symver @var{name}, @var{name2@@@@nodename}
5863 In this case, the symbol @var{name} must exist and be defined within
5864 the file being assembled. It is similar to @var{name2@@nodename}. The
5865 difference is @var{name2@@@@nodename} will also be used to resolve
5866 references to @var{name2} by the linker.
5868 The third usage of the @code{.symver} directive is:
5870 .symver @var{name}, @var{name2@@@@@@nodename}
5872 When @var{name} is not defined within the
5873 file being assembled, it is treated as @var{name2@@nodename}. When
5874 @var{name} is defined within the file being assembled, the symbol
5875 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5880 @section @code{.tag @var{structname}}
5882 @cindex COFF structure debugging
5883 @cindex structure debugging, COFF
5884 @cindex @code{tag} directive
5885 This directive is generated by compilers to include auxiliary debugging
5886 information in the symbol table. It is only permitted inside
5887 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5888 definitions in the symbol table with instances of those structures.
5891 @samp{.tag} is only used when generating COFF format output; when
5892 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5898 @section @code{.text @var{subsection}}
5900 @cindex @code{text} directive
5901 Tells @command{@value{AS}} to assemble the following statements onto the end of
5902 the text subsection numbered @var{subsection}, which is an absolute
5903 expression. If @var{subsection} is omitted, subsection number zero
5907 @section @code{.title "@var{heading}"}
5909 @cindex @code{title} directive
5910 @cindex listing control: title line
5911 Use @var{heading} as the title (second line, immediately after the
5912 source file name and pagenumber) when generating assembly listings.
5914 This directive affects subsequent pages, as well as the current page if
5915 it appears within ten lines of the top of a page.
5919 @section @code{.type}
5921 This directive is used to set the type of a symbol.
5925 @c only print the extra heading if both COFF and ELF are set
5926 @subheading COFF Version
5929 @cindex COFF symbol type
5930 @cindex symbol type, COFF
5931 @cindex @code{type} directive (COFF version)
5932 For COFF targets, this directive is permitted only within
5933 @code{.def}/@code{.endef} pairs. It is used like this:
5939 This records the integer @var{int} as the type attribute of a symbol table
5943 @samp{.type} is associated only with COFF format output; when
5944 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
5945 directive but ignores it.
5951 @c only print the extra heading if both COFF and ELF are set
5952 @subheading ELF Version
5955 @cindex ELF symbol type
5956 @cindex symbol type, ELF
5957 @cindex @code{type} directive (ELF version)
5958 For ELF targets, the @code{.type} directive is used like this:
5961 .type @var{name} , @var{type description}
5964 This sets the type of symbol @var{name} to be either a
5965 function symbol or an object symbol. There are five different syntaxes
5966 supported for the @var{type description} field, in order to provide
5967 compatibility with various other assemblers. The syntaxes supported are:
5970 .type <name>,#function
5971 .type <name>,#object
5973 .type <name>,@@function
5974 .type <name>,@@object
5976 .type <name>,%function
5977 .type <name>,%object
5979 .type <name>,"function"
5980 .type <name>,"object"
5982 .type <name> STT_FUNCTION
5983 .type <name> STT_OBJECT
5989 @section @code{.uleb128 @var{expressions}}
5991 @cindex @code{uleb128} directive
5992 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
5993 compact, variable length representation of numbers used by the DWARF
5994 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
5998 @section @code{.val @var{addr}}
6000 @cindex @code{val} directive
6001 @cindex COFF value attribute
6002 @cindex value attribute, COFF
6003 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6004 records the address @var{addr} as the value attribute of a symbol table
6008 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6009 configured for @code{b.out}, it accepts this directive but ignores it.
6015 @section @code{.version "@var{string}"}
6017 @cindex @code{version} directive
6018 This directive creates a @code{.note} section and places into it an ELF
6019 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6024 @section @code{.vtable_entry @var{table}, @var{offset}}
6026 @cindex @code{vtable_entry} directive
6027 This directive finds or creates a symbol @code{table} and creates a
6028 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6031 @section @code{.vtable_inherit @var{child}, @var{parent}}
6033 @cindex @code{vtable_inherit} directive
6034 This directive finds the symbol @code{child} and finds or creates the symbol
6035 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6036 parent whose addend is the value of the child symbol. As a special case the
6037 parent name of @code{0} is treated as refering the @code{*ABS*} section.
6041 @section @code{.warning "@var{string}"}
6042 @cindex warning directive
6043 Similar to the directive @code{.error}
6044 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6047 @section @code{.weak @var{names}}
6049 @cindex @code{weak} directive
6050 This directive sets the weak attribute on the comma separated list of symbol
6051 @code{names}. If the symbols do not already exist, they will be created.
6053 On COFF targets other than PE, weak symbols are a GNU extension. This
6054 directive sets the weak attribute on the comma separated list of symbol
6055 @code{names}. If the symbols do not already exist, they will be created.
6057 On the PE target, weak symbols are supported natively as weak aliases.
6058 When a weak symbol is created that is not an alias, GAS creates an
6059 alternate symbol to hold the default value.
6062 @section @code{.weakref @var{alias}, @var{target}}
6064 @cindex @code{weakref} directive
6065 This directive creates an alias to the target symbol that enables the symbol to
6066 be referenced with weak-symbol semantics, but without actually making it weak.
6067 If direct references or definitions of the symbol are present, then the symbol
6068 will not be weak, but if all references to it are through weak references, the
6069 symbol will be marked as weak in the symbol table.
6071 The effect is equivalent to moving all references to the alias to a separate
6072 assembly source file, renaming the alias to the symbol in it, declaring the
6073 symbol as weak there, and running a reloadable link to merge the object files
6074 resulting from the assembly of the new source file and the old source file that
6075 had the references to the alias removed.
6077 The alias itself never makes to the symbol table, and is entirely handled
6078 within the assembler.
6081 @section @code{.word @var{expressions}}
6083 @cindex @code{word} directive
6084 This directive expects zero or more @var{expressions}, of any section,
6085 separated by commas.
6088 For each expression, @command{@value{AS}} emits a 32-bit number.
6091 For each expression, @command{@value{AS}} emits a 16-bit number.
6096 The size of the number emitted, and its byte order,
6097 depend on what target computer the assembly is for.
6100 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6101 @c happen---32-bit addressability, period; no long/short jumps.
6102 @ifset DIFF-TBL-KLUGE
6103 @cindex difference tables altered
6104 @cindex altered difference tables
6106 @emph{Warning: Special Treatment to support Compilers}
6110 Machines with a 32-bit address space, but that do less than 32-bit
6111 addressing, require the following special treatment. If the machine of
6112 interest to you does 32-bit addressing (or doesn't require it;
6113 @pxref{Machine Dependencies}), you can ignore this issue.
6116 In order to assemble compiler output into something that works,
6117 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6118 Directives of the form @samp{.word sym1-sym2} are often emitted by
6119 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6120 directive of the form @samp{.word sym1-sym2}, and the difference between
6121 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6122 creates a @dfn{secondary jump table}, immediately before the next label.
6123 This secondary jump table is preceded by a short-jump to the
6124 first byte after the secondary table. This short-jump prevents the flow
6125 of control from accidentally falling into the new table. Inside the
6126 table is a long-jump to @code{sym2}. The original @samp{.word}
6127 contains @code{sym1} minus the address of the long-jump to
6130 If there were several occurrences of @samp{.word sym1-sym2} before the
6131 secondary jump table, all of them are adjusted. If there was a
6132 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6133 long-jump to @code{sym4} is included in the secondary jump table,
6134 and the @code{.word} directives are adjusted to contain @code{sym3}
6135 minus the address of the long-jump to @code{sym4}; and so on, for as many
6136 entries in the original jump table as necessary.
6139 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6140 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6141 assembly language programmers.
6144 @c end DIFF-TBL-KLUGE
6147 @section Deprecated Directives
6149 @cindex deprecated directives
6150 @cindex obsolescent directives
6151 One day these directives won't work.
6152 They are included for compatibility with older assemblers.
6159 @node Machine Dependencies
6160 @chapter Machine Dependent Features
6162 @cindex machine dependencies
6163 The machine instruction sets are (almost by definition) different on
6164 each machine where @command{@value{AS}} runs. Floating point representations
6165 vary as well, and @command{@value{AS}} often supports a few additional
6166 directives or command-line options for compatibility with other
6167 assemblers on a particular platform. Finally, some versions of
6168 @command{@value{AS}} support special pseudo-instructions for branch
6171 This chapter discusses most of these differences, though it does not
6172 include details on any machine's instruction set. For details on that
6173 subject, see the hardware manufacturer's manual.
6177 * Alpha-Dependent:: Alpha Dependent Features
6180 * ARC-Dependent:: ARC Dependent Features
6183 * ARM-Dependent:: ARM Dependent Features
6186 * BFIN-Dependent:: BFIN Dependent Features
6189 * CRIS-Dependent:: CRIS Dependent Features
6192 * D10V-Dependent:: D10V Dependent Features
6195 * D30V-Dependent:: D30V Dependent Features
6198 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6201 * HPPA-Dependent:: HPPA Dependent Features
6204 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6207 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6210 * i860-Dependent:: Intel 80860 Dependent Features
6213 * i960-Dependent:: Intel 80960 Dependent Features
6216 * IA-64-Dependent:: Intel IA-64 Dependent Features
6219 * IP2K-Dependent:: IP2K Dependent Features
6222 * M32C-Dependent:: M32C Dependent Features
6225 * M32R-Dependent:: M32R Dependent Features
6228 * M68K-Dependent:: M680x0 Dependent Features
6231 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6234 * MIPS-Dependent:: MIPS Dependent Features
6237 * MMIX-Dependent:: MMIX Dependent Features
6240 * MSP430-Dependent:: MSP430 Dependent Features
6243 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6244 * SH64-Dependent:: SuperH SH64 Dependent Features
6247 * PDP-11-Dependent:: PDP-11 Dependent Features
6250 * PJ-Dependent:: picoJava Dependent Features
6253 * PPC-Dependent:: PowerPC Dependent Features
6256 * Sparc-Dependent:: SPARC Dependent Features
6259 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
6262 * V850-Dependent:: V850 Dependent Features
6265 * Xtensa-Dependent:: Xtensa Dependent Features
6268 * Z80-Dependent:: Z80 Dependent Features
6271 * Z8000-Dependent:: Z8000 Dependent Features
6274 * Vax-Dependent:: VAX Dependent Features
6281 @c The following major nodes are *sections* in the GENERIC version, *chapters*
6282 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
6283 @c peculiarity: to preserve cross-references, there must be a node called
6284 @c "Machine Dependencies". Hence the conditional nodenames in each
6285 @c major node below. Node defaulting in makeinfo requires adjacency of
6286 @c node and sectioning commands; hence the repetition of @chapter BLAH
6287 @c in both conditional blocks.
6290 @include c-alpha.texi
6302 @include c-bfin.texi
6306 @include c-cris.texi
6311 @node Machine Dependencies
6312 @chapter Machine Dependent Features
6314 The machine instruction sets are different on each Renesas chip family,
6315 and there are also some syntax differences among the families. This
6316 chapter describes the specific @command{@value{AS}} features for each
6320 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6321 * SH-Dependent:: Renesas SH Dependent Features
6328 @include c-d10v.texi
6332 @include c-d30v.texi
6336 @include c-h8300.texi
6340 @include c-hppa.texi
6344 @include c-i370.texi
6348 @include c-i386.texi
6352 @include c-i860.texi
6356 @include c-i960.texi
6360 @include c-ia64.texi
6364 @include c-ip2k.texi
6368 @include c-m32c.texi
6372 @include c-m32r.texi
6376 @include c-m68k.texi
6380 @include c-m68hc11.texi
6384 @include c-mips.texi
6388 @include c-mmix.texi
6392 @include c-msp430.texi
6396 @include c-ns32k.texi
6400 @include c-pdp11.texi
6413 @include c-sh64.texi
6417 @include c-sparc.texi
6421 @include c-tic54x.texi
6437 @include c-v850.texi
6441 @include c-xtensa.texi
6445 @c reverse effect of @down at top of generic Machine-Dep chapter
6449 @node Reporting Bugs
6450 @chapter Reporting Bugs
6451 @cindex bugs in assembler
6452 @cindex reporting bugs in assembler
6454 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6456 Reporting a bug may help you by bringing a solution to your problem, or it may
6457 not. But in any case the principal function of a bug report is to help the
6458 entire community by making the next version of @command{@value{AS}} work better.
6459 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6461 In order for a bug report to serve its purpose, you must include the
6462 information that enables us to fix the bug.
6465 * Bug Criteria:: Have you found a bug?
6466 * Bug Reporting:: How to report bugs
6470 @section Have You Found a Bug?
6471 @cindex bug criteria
6473 If you are not sure whether you have found a bug, here are some guidelines:
6476 @cindex fatal signal
6477 @cindex assembler crash
6478 @cindex crash of assembler
6480 If the assembler gets a fatal signal, for any input whatever, that is a
6481 @command{@value{AS}} bug. Reliable assemblers never crash.
6483 @cindex error on valid input
6485 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6487 @cindex invalid input
6489 If @command{@value{AS}} does not produce an error message for invalid input, that
6490 is a bug. However, you should note that your idea of ``invalid input'' might
6491 be our idea of ``an extension'' or ``support for traditional practice''.
6494 If you are an experienced user of assemblers, your suggestions for improvement
6495 of @command{@value{AS}} are welcome in any case.
6499 @section How to Report Bugs
6501 @cindex assembler bugs, reporting
6503 A number of companies and individuals offer support for @sc{gnu} products. If
6504 you obtained @command{@value{AS}} from a support organization, we recommend you
6505 contact that organization first.
6507 You can find contact information for many support companies and
6508 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6511 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6512 to @samp{bug-binutils@@gnu.org}.
6514 The fundamental principle of reporting bugs usefully is this:
6515 @strong{report all the facts}. If you are not sure whether to state a
6516 fact or leave it out, state it!
6518 Often people omit facts because they think they know what causes the problem
6519 and assume that some details do not matter. Thus, you might assume that the
6520 name of a symbol you use in an example does not matter. Well, probably it does
6521 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6522 happens to fetch from the location where that name is stored in memory;
6523 perhaps, if the name were different, the contents of that location would fool
6524 the assembler into doing the right thing despite the bug. Play it safe and
6525 give a specific, complete example. That is the easiest thing for you to do,
6526 and the most helpful.
6528 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6529 it is new to us. Therefore, always write your bug reports on the assumption
6530 that the bug has not been reported previously.
6532 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6533 bell?'' This cannot help us fix a bug, so it is basically useless. We
6534 respond by asking for enough details to enable us to investigate.
6535 You might as well expedite matters by sending them to begin with.
6537 To enable us to fix the bug, you should include all these things:
6541 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6542 it with the @samp{--version} argument.
6544 Without this, we will not know whether there is any point in looking for
6545 the bug in the current version of @command{@value{AS}}.
6548 Any patches you may have applied to the @command{@value{AS}} source.
6551 The type of machine you are using, and the operating system name and
6555 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6559 The command arguments you gave the assembler to assemble your example and
6560 observe the bug. To guarantee you will not omit something important, list them
6561 all. A copy of the Makefile (or the output from make) is sufficient.
6563 If we were to try to guess the arguments, we would probably guess wrong
6564 and then we might not encounter the bug.
6567 A complete input file that will reproduce the bug. If the bug is observed when
6568 the assembler is invoked via a compiler, send the assembler source, not the
6569 high level language source. Most compilers will produce the assembler source
6570 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6571 the options @samp{-v --save-temps}; this will save the assembler source in a
6572 file with an extension of @file{.s}, and also show you exactly how
6573 @command{@value{AS}} is being run.
6576 A description of what behavior you observe that you believe is
6577 incorrect. For example, ``It gets a fatal signal.''
6579 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6580 will certainly notice it. But if the bug is incorrect output, we might not
6581 notice unless it is glaringly wrong. You might as well not give us a chance to
6584 Even if the problem you experience is a fatal signal, you should still say so
6585 explicitly. Suppose something strange is going on, such as, your copy of
6586 @command{@value{AS}} is out of synch, or you have encountered a bug in the C
6587 library on your system. (This has happened!) Your copy might crash and ours
6588 would not. If you told us to expect a crash, then when ours fails to crash, we
6589 would know that the bug was not happening for us. If you had not told us to
6590 expect a crash, then we would not be able to draw any conclusion from our
6594 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6595 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6596 option. Always send diffs from the old file to the new file. If you even
6597 discuss something in the @command{@value{AS}} source, refer to it by context, not
6600 The line numbers in our development sources will not match those in your
6601 sources. Your line numbers would convey no useful information to us.
6604 Here are some things that are not necessary:
6608 A description of the envelope of the bug.
6610 Often people who encounter a bug spend a lot of time investigating
6611 which changes to the input file will make the bug go away and which
6612 changes will not affect it.
6614 This is often time consuming and not very useful, because the way we
6615 will find the bug is by running a single example under the debugger
6616 with breakpoints, not by pure deduction from a series of examples.
6617 We recommend that you save your time for something else.
6619 Of course, if you can find a simpler example to report @emph{instead}
6620 of the original one, that is a convenience for us. Errors in the
6621 output will be easier to spot, running under the debugger will take
6622 less time, and so on.
6624 However, simplification is not vital; if you do not want to do this,
6625 report the bug anyway and send us the entire test case you used.
6628 A patch for the bug.
6630 A patch for the bug does help us if it is a good one. But do not omit
6631 the necessary information, such as the test case, on the assumption that
6632 a patch is all we need. We might see problems with your patch and decide
6633 to fix the problem another way, or we might not understand it at all.
6635 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6636 construct an example that will make the program follow a certain path through
6637 the code. If you do not send us the example, we will not be able to construct
6638 one, so we will not be able to verify that the bug is fixed.
6640 And if we cannot understand what bug you are trying to fix, or why your
6641 patch should be an improvement, we will not install it. A test case will
6642 help us to understand.
6645 A guess about what the bug is or what it depends on.
6647 Such guesses are usually wrong. Even we cannot guess right about such
6648 things without first using the debugger to find the facts.
6651 @node Acknowledgements
6652 @chapter Acknowledgements
6654 If you have contributed to GAS and your name isn't listed here,
6655 it is not meant as a slight. We just don't know about it. Send mail to the
6656 maintainer, and we'll correct the situation. Currently
6658 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6660 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6663 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6664 information and the 68k series machines, most of the preprocessing pass, and
6665 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6667 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6668 many bug fixes, including merging support for several processors, breaking GAS
6669 up to handle multiple object file format back ends (including heavy rewrite,
6670 testing, an integration of the coff and b.out back ends), adding configuration
6671 including heavy testing and verification of cross assemblers and file splits
6672 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6673 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6674 port (including considerable amounts of reverse engineering), a SPARC opcode
6675 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6676 assertions and made them work, much other reorganization, cleanup, and lint.
6678 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6679 in format-specific I/O modules.
6681 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6682 has done much work with it since.
6684 The Intel 80386 machine description was written by Eliot Dresselhaus.
6686 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6688 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6689 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6691 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6692 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6693 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6694 support a.out format.
6696 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
6697 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
6698 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6699 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6702 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6703 simplified the configuration of which versions accept which directives. He
6704 updated the 68k machine description so that Motorola's opcodes always produced
6705 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6706 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6707 cross-compilation support, and one bug in relaxation that took a week and
6708 required the proverbial one-bit fix.
6710 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6711 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6712 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6713 PowerPC assembler, and made a few other minor patches.
6715 Steve Chamberlain made GAS able to generate listings.
6717 Hewlett-Packard contributed support for the HP9000/300.
6719 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6720 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6721 formats). This work was supported by both the Center for Software Science at
6722 the University of Utah and Cygnus Support.
6724 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6725 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6726 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6727 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6728 and some initial 64-bit support).
6730 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6732 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6733 support for openVMS/Alpha.
6735 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6738 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6739 Inc. added support for Xtensa processors.
6741 Several engineers at Cygnus Support have also provided many small bug fixes and
6742 configuration enhancements.
6744 Many others have contributed large or small bugfixes and enhancements. If
6745 you have contributed significant work and are not mentioned on this list, and
6746 want to be, let us know. Some of the history has been lost; we are not
6747 intentionally leaving anyone out.
projects / dragonfly.git / blob