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{--target-help}]
235 [@var{target-options}] [@b{--}|@var{files} @dots{}]
237 @c Target dependent options are listed below. Keep the list sorted.
238 @c Add an empty line for separation.
241 @emph{Target Alpha options:}
243 [@b{-mdebug} | @b{-no-mdebug}]
244 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
245 [@b{-F}] [@b{-32addr}]
249 @emph{Target ARC options:}
255 @emph{Target ARM options:}
256 @c Don't document the deprecated options
257 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
258 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
259 [@b{-mfpu}=@var{floating-point-format}]
260 [@b{-mfloat-abi}=@var{abi}]
261 [@b{-meabi}=@var{ver}]
264 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
265 @b{-mapcs-reentrant}]
266 [@b{-mthumb-interwork}] [@b{-k}]
270 @emph{Target CRIS options:}
271 [@b{--underscore} | @b{--no-underscore}]
273 [@b{--emulation=criself} | @b{--emulation=crisaout}]
274 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
275 @c Deprecated -- deliberately not documented.
280 @emph{Target D10V options:}
285 @emph{Target D30V options:}
286 [@b{-O}|@b{-n}|@b{-N}]
289 @c Renesas family chips have no machine-dependent assembler options
292 @c HPPA has no machine-dependent assembler options (yet).
296 @emph{Target i386 options:}
297 [@b{--32}|@b{--64}] [@b{-n}]
301 @emph{Target i960 options:}
302 @c see md_parse_option in tc-i960.c
303 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
305 [@b{-b}] [@b{-no-relax}]
309 @emph{Target IA-64 options:}
310 [@b{-mconstant-gp}|@b{-mauto-pic}]
311 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
313 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
314 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
315 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
316 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
320 @emph{Target IP2K options:}
321 [@b{-mip2022}|@b{-mip2022ext}]
325 @emph{Target M32C options:}
326 [@b{-m32c}|@b{-m16c}]
330 @emph{Target M32R options:}
331 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
336 @emph{Target M680X0 options:}
337 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
341 @emph{Target M68HC11 options:}
342 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
343 [@b{-mshort}|@b{-mlong}]
344 [@b{-mshort-double}|@b{-mlong-double}]
345 [@b{--force-long-branchs}] [@b{--short-branchs}]
346 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
347 [@b{--print-opcodes}] [@b{--generate-example}]
351 @emph{Target MCORE options:}
352 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
353 [@b{-mcpu=[210|340]}]
357 @emph{Target MIPS options:}
358 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
359 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
360 [@b{-non_shared}] [@b{-xgot}]
361 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
362 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
363 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
364 [@b{-mips64}] [@b{-mips64r2}]
365 [@b{-construct-floats}] [@b{-no-construct-floats}]
366 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
367 [@b{-mfix7000}] [@b{-mno-fix7000}]
368 [@b{-mips16}] [@b{-no-mips16}]
369 [@b{-mips3d}] [@b{-no-mips3d}]
370 [@b{-mdmx}] [@b{-no-mdmx}]
371 [@b{-mdsp}] [@b{-mno-dsp}]
372 [@b{-mmt}] [@b{-mno-mt}]
373 [@b{-mdebug}] [@b{-no-mdebug}]
374 [@b{-mpdr}] [@b{-mno-pdr}]
378 @emph{Target MMIX options:}
379 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
380 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
381 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
382 [@b{--linker-allocated-gregs}]
386 @emph{Target PDP11 options:}
387 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
388 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
389 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
393 @emph{Target picoJava options:}
398 @emph{Target PowerPC options:}
399 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
400 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
401 @b{-mbooke32}|@b{-mbooke64}]
402 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
403 [@b{-mregnames}|@b{-mno-regnames}]
404 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
405 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
406 [@b{-msolaris}|@b{-mno-solaris}]
410 @emph{Target SPARC options:}
411 @c The order here is important. See c-sparc.texi.
412 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
413 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
414 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
419 @emph{Target TIC54X options:}
420 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
421 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
424 @c Z8000 has no machine-dependent assembler options
428 @emph{Target Xtensa options:}
429 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
430 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
431 [@b{--[no-]transform}]
432 [@b{--rename-section} @var{oldname}=@var{newname}]
441 Turn on listings, in any of a variety of ways:
445 omit false conditionals
448 omit debugging directives
451 include high-level source
457 include macro expansions
460 omit forms processing
466 set the name of the listing file
469 You may combine these options; for example, use @samp{-aln} for assembly
470 listing without forms processing. The @samp{=file} option, if used, must be
471 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
474 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
477 Ignored. This option is accepted for script compatibility with calls to
480 @item --defsym @var{sym}=@var{value}
481 Define the symbol @var{sym} to be @var{value} before assembling the input file.
482 @var{value} must be an integer constant. As in C, a leading @samp{0x}
483 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
486 ``fast''---skip whitespace and comment preprocessing (assume source is
491 Generate debugging information for each assembler source line using whichever
492 debug format is preferred by the target. This currently means either STABS,
496 Generate stabs debugging information for each assembler line. This
497 may help debugging assembler code, if the debugger can handle it.
500 Generate stabs debugging information for each assembler line, with GNU
501 extensions that probably only gdb can handle, and that could make other
502 debuggers crash or refuse to read your program. This
503 may help debugging assembler code. Currently the only GNU extension is
504 the location of the current working directory at assembling time.
507 Generate DWARF2 debugging information for each assembler line. This
508 may help debugging assembler code, if the debugger can handle it. Note---this
509 option is only supported by some targets, not all of them.
512 Print a summary of the command line options and exit.
515 Print a summary of all target specific options and exit.
518 Add directory @var{dir} to the search list for @code{.include} directives.
521 Don't warn about signed overflow.
524 @ifclear DIFF-TBL-KLUGE
525 This option is accepted but has no effect on the @value{TARGET} family.
527 @ifset DIFF-TBL-KLUGE
528 Issue warnings when difference tables altered for long displacements.
533 Keep (in the symbol table) local symbols. On traditional a.out systems
534 these start with @samp{L}, but different systems have different local
537 @item --listing-lhs-width=@var{number}
538 Set the maximum width, in words, of the output data column for an assembler
539 listing to @var{number}.
541 @item --listing-lhs-width2=@var{number}
542 Set the maximum width, in words, of the output data column for continuation
543 lines in an assembler listing to @var{number}.
545 @item --listing-rhs-width=@var{number}
546 Set the maximum width of an input source line, as displayed in a listing, to
549 @item --listing-cont-lines=@var{number}
550 Set the maximum number of lines printed in a listing for a single line of input
553 @item -o @var{objfile}
554 Name the object-file output from @command{@value{AS}} @var{objfile}.
557 Fold the data section into the text section.
559 @kindex --hash-size=@var{number}
560 Set the default size of GAS's hash tables to a prime number close to
561 @var{number}. Increasing this value can reduce the length of time it takes the
562 assembler to perform its tasks, at the expense of increasing the assembler's
563 memory requirements. Similarly reducing this value can reduce the memory
564 requirements at the expense of speed.
566 @item --reduce-memory-overheads
567 This option reduces GAS's memory requirements, at the expense of making the
568 assembly processes slower. Currently this switch is a synonym for
569 @samp{--hash-size=4051}, but in the future it may have other effects as well.
572 Print the maximum space (in bytes) and total time (in seconds) used by
575 @item --strip-local-absolute
576 Remove local absolute symbols from the outgoing symbol table.
580 Print the @command{as} version.
583 Print the @command{as} version and exit.
587 Suppress warning messages.
589 @item --fatal-warnings
590 Treat warnings as errors.
593 Don't suppress warning messages or treat them as errors.
602 Generate an object file even after errors.
604 @item -- | @var{files} @dots{}
605 Standard input, or source files to assemble.
610 The following options are available when @value{AS} is configured for
615 This option selects the core processor variant.
617 Select either big-endian (-EB) or little-endian (-EL) output.
622 The following options are available when @value{AS} is configured for the ARM
626 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
627 Specify which ARM processor variant is the target.
628 @item -march=@var{architecture}[+@var{extension}@dots{}]
629 Specify which ARM architecture variant is used by the target.
630 @item -mfpu=@var{floating-point-format}
631 Select which Floating Point architecture is the target.
632 @item -mfloat-abi=@var{abi}
633 Select which floating point ABI is in use.
635 Enable Thumb only instruction decoding.
636 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
637 Select which procedure calling convention is in use.
639 Select either big-endian (-EB) or little-endian (-EL) output.
640 @item -mthumb-interwork
641 Specify that the code has been generated with interworking between Thumb and
644 Specify that PIC code has been generated.
649 See the info pages for documentation of the CRIS-specific options.
653 The following options are available when @value{AS} is configured for
656 @cindex D10V optimization
657 @cindex optimization, D10V
659 Optimize output by parallelizing instructions.
664 The following options are available when @value{AS} is configured for a D30V
667 @cindex D30V optimization
668 @cindex optimization, D30V
670 Optimize output by parallelizing instructions.
674 Warn when nops are generated.
676 @cindex D30V nops after 32-bit multiply
678 Warn when a nop after a 32-bit multiply instruction is generated.
683 The following options are available when @value{AS} is configured for the
684 Intel 80960 processor.
687 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
688 Specify which variant of the 960 architecture is the target.
691 Add code to collect statistics about branches taken.
694 Do not alter compare-and-branch instructions for long displacements;
701 The following options are available when @value{AS} is configured for the
707 Specifies that the extended IP2022 instructions are allowed.
710 Restores the default behaviour, which restricts the permitted instructions to
711 just the basic IP2022 ones.
717 The following options are available when @value{AS} is configured for the
718 Renesas M32C and M16C processors.
723 Assemble M32C instructions.
726 Assemble M16C instructions (the default).
732 The following options are available when @value{AS} is configured for the
733 Renesas M32R (formerly Mitsubishi M32R) series.
738 Specify which processor in the M32R family is the target. The default
739 is normally the M32R, but this option changes it to the M32RX.
741 @item --warn-explicit-parallel-conflicts or --Wp
742 Produce warning messages when questionable parallel constructs are
745 @item --no-warn-explicit-parallel-conflicts or --Wnp
746 Do not produce warning messages when questionable parallel constructs are
753 The following options are available when @value{AS} is configured for the
754 Motorola 68000 series.
759 Shorten references to undefined symbols, to one word instead of two.
761 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
762 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
763 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
764 Specify what processor in the 68000 family is the target. The default
765 is normally the 68020, but this can be changed at configuration time.
767 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
768 The target machine does (or does not) have a floating-point coprocessor.
769 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
770 the basic 68000 is not compatible with the 68881, a combination of the
771 two can be specified, since it's possible to do emulation of the
772 coprocessor instructions with the main processor.
774 @item -m68851 | -mno-68851
775 The target machine does (or does not) have a memory-management
776 unit coprocessor. The default is to assume an MMU for 68020 and up.
783 For details about the PDP-11 machine dependent features options,
784 see @ref{PDP-11-Options}.
787 @item -mpic | -mno-pic
788 Generate position-independent (or position-dependent) code. The
789 default is @option{-mpic}.
792 @itemx -mall-extensions
793 Enable all instruction set extensions. This is the default.
795 @item -mno-extensions
796 Disable all instruction set extensions.
798 @item -m@var{extension} | -mno-@var{extension}
799 Enable (or disable) a particular instruction set extension.
802 Enable the instruction set extensions supported by a particular CPU, and
803 disable all other extensions.
805 @item -m@var{machine}
806 Enable the instruction set extensions supported by a particular machine
807 model, and disable all other extensions.
813 The following options are available when @value{AS} is configured for
814 a picoJava processor.
818 @cindex PJ endianness
819 @cindex endianness, PJ
820 @cindex big endian output, PJ
822 Generate ``big endian'' format output.
824 @cindex little endian output, PJ
826 Generate ``little endian'' format output.
832 The following options are available when @value{AS} is configured for the
833 Motorola 68HC11 or 68HC12 series.
837 @item -m68hc11 | -m68hc12 | -m68hcs12
838 Specify what processor is the target. The default is
839 defined by the configuration option when building the assembler.
842 Specify to use the 16-bit integer ABI.
845 Specify to use the 32-bit integer ABI.
848 Specify to use the 32-bit double ABI.
851 Specify to use the 64-bit double ABI.
853 @item --force-long-branchs
854 Relative branches are turned into absolute ones. This concerns
855 conditional branches, unconditional branches and branches to a
858 @item -S | --short-branchs
859 Do not turn relative branchs into absolute ones
860 when the offset is out of range.
862 @item --strict-direct-mode
863 Do not turn the direct addressing mode into extended addressing mode
864 when the instruction does not support direct addressing mode.
866 @item --print-insn-syntax
867 Print the syntax of instruction in case of error.
869 @item --print-opcodes
870 print the list of instructions with syntax and then exit.
872 @item --generate-example
873 print an example of instruction for each possible instruction and then exit.
874 This option is only useful for testing @command{@value{AS}}.
880 The following options are available when @command{@value{AS}} is configured
881 for the SPARC architecture:
884 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
885 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
886 Explicitly select a variant of the SPARC architecture.
888 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
889 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
891 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
892 UltraSPARC extensions.
894 @item -xarch=v8plus | -xarch=v8plusa
895 For compatibility with the Solaris v9 assembler. These options are
896 equivalent to -Av8plus and -Av8plusa, respectively.
899 Warn when the assembler switches to another architecture.
904 The following options are available when @value{AS} is configured for the 'c54x
909 Enable extended addressing mode. All addresses and relocations will assume
910 extended addressing (usually 23 bits).
911 @item -mcpu=@var{CPU_VERSION}
912 Sets the CPU version being compiled for.
913 @item -merrors-to-file @var{FILENAME}
914 Redirect error output to a file, for broken systems which don't support such
915 behaviour in the shell.
920 The following options are available when @value{AS} is configured for
921 a @sc{mips} processor.
925 This option sets the largest size of an object that can be referenced
926 implicitly with the @code{gp} register. It is only accepted for targets that
927 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
929 @cindex MIPS endianness
930 @cindex endianness, MIPS
931 @cindex big endian output, MIPS
933 Generate ``big endian'' format output.
935 @cindex little endian output, MIPS
937 Generate ``little endian'' format output.
949 Generate code for a particular @sc{mips} Instruction Set Architecture level.
950 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
951 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
952 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
953 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
955 correspond to generic
956 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
957 and @samp{MIPS64 Release 2}
958 ISA processors, respectively.
960 @item -march=@var{CPU}
961 Generate code for a particular @sc{mips} cpu.
963 @item -mtune=@var{cpu}
964 Schedule and tune for a particular @sc{mips} cpu.
968 Cause nops to be inserted if the read of the destination register
969 of an mfhi or mflo instruction occurs in the following two instructions.
973 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
974 section instead of the standard ELF .stabs sections.
978 Control generation of @code{.pdr} sections.
982 The register sizes are normally inferred from the ISA and ABI, but these
983 flags force a certain group of registers to be treated as 32 bits wide at
984 all times. @samp{-mgp32} controls the size of general-purpose registers
985 and @samp{-mfp32} controls the size of floating-point registers.
989 Generate code for the MIPS 16 processor. This is equivalent to putting
990 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
991 turns off this option.
995 Generate code for the MIPS-3D Application Specific Extension.
996 This tells the assembler to accept MIPS-3D instructions.
997 @samp{-no-mips3d} turns off this option.
1001 Generate code for the MDMX Application Specific Extension.
1002 This tells the assembler to accept MDMX instructions.
1003 @samp{-no-mdmx} turns off this option.
1007 Generate code for the DSP Application Specific Extension.
1008 This tells the assembler to accept DSP instructions.
1009 @samp{-mno-dsp} turns off this option.
1013 Generate code for the MT Application Specific Extension.
1014 This tells the assembler to accept MT instructions.
1015 @samp{-mno-mt} turns off this option.
1017 @item --construct-floats
1018 @itemx --no-construct-floats
1019 The @samp{--no-construct-floats} option disables the construction of
1020 double width floating point constants by loading the two halves of the
1021 value into the two single width floating point registers that make up
1022 the double width register. By default @samp{--construct-floats} is
1023 selected, allowing construction of these floating point constants.
1026 @item --emulation=@var{name}
1027 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1028 for some other target, in all respects, including output format (choosing
1029 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1030 debugging information or store symbol table information, and default
1031 endianness. The available configuration names are: @samp{mipsecoff},
1032 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1033 @samp{mipsbelf}. The first two do not alter the default endianness from that
1034 of the primary target for which the assembler was configured; the others change
1035 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1036 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1037 selection in any case.
1039 This option is currently supported only when the primary target
1040 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1041 Furthermore, the primary target or others specified with
1042 @samp{--enable-targets=@dots{}} at configuration time must include support for
1043 the other format, if both are to be available. For example, the Irix 5
1044 configuration includes support for both.
1046 Eventually, this option will support more configurations, with more
1047 fine-grained control over the assembler's behavior, and will be supported for
1051 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1058 Control how to deal with multiplication overflow and division by zero.
1059 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1060 (and only work for Instruction Set Architecture level 2 and higher);
1061 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1065 When this option is used, @command{@value{AS}} will issue a warning every
1066 time it generates a nop instruction from a macro.
1071 The following options are available when @value{AS} is configured for
1077 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1078 The command line option @samp{-nojsri2bsr} can be used to disable it.
1082 Enable or disable the silicon filter behaviour. By default this is disabled.
1083 The default can be overridden by the @samp{-sifilter} command line option.
1086 Alter jump instructions for long displacements.
1088 @item -mcpu=[210|340]
1089 Select the cpu type on the target hardware. This controls which instructions
1093 Assemble for a big endian target.
1096 Assemble for a little endian target.
1102 See the info pages for documentation of the MMIX-specific options.
1106 The following options are available when @value{AS} is configured for
1107 an Xtensa processor.
1110 @item --text-section-literals | --no-text-section-literals
1111 With @option{--text-@-section-@-literals}, literal pools are interspersed
1112 in the text section. The default is
1113 @option{--no-@-text-@-section-@-literals}, which places literals in a
1114 separate section in the output file. These options only affect literals
1115 referenced via PC-relative @code{L32R} instructions; literals for
1116 absolute mode @code{L32R} instructions are handled separately.
1118 @item --absolute-literals | --no-absolute-literals
1119 Indicate to the assembler whether @code{L32R} instructions use absolute
1120 or PC-relative addressing. The default is to assume absolute addressing
1121 if the Xtensa processor includes the absolute @code{L32R} addressing
1122 option. Otherwise, only the PC-relative @code{L32R} mode can be used.
1124 @item --target-align | --no-target-align
1125 Enable or disable automatic alignment to reduce branch penalties at the
1126 expense of some code density. The default is @option{--target-@-align}.
1128 @item --longcalls | --no-longcalls
1129 Enable or disable transformation of call instructions to allow calls
1130 across a greater range of addresses. The default is
1131 @option{--no-@-longcalls}.
1133 @item --transform | --no-transform
1134 Enable or disable all assembler transformations of Xtensa instructions.
1135 The default is @option{--transform};
1136 @option{--no-transform} should be used only in the rare cases when the
1137 instructions must be exactly as specified in the assembly source.
1144 * Manual:: Structure of this Manual
1145 * GNU Assembler:: The GNU Assembler
1146 * Object Formats:: Object File Formats
1147 * Command Line:: Command Line
1148 * Input Files:: Input Files
1149 * Object:: Output (Object) File
1150 * Errors:: Error and Warning Messages
1154 @section Structure of this Manual
1156 @cindex manual, structure and purpose
1157 This manual is intended to describe what you need to know to use
1158 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1159 notation for symbols, constants, and expressions; the directives that
1160 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1163 We also cover special features in the @value{TARGET}
1164 configuration of @command{@value{AS}}, including assembler directives.
1167 This manual also describes some of the machine-dependent features of
1168 various flavors of the assembler.
1171 @cindex machine instructions (not covered)
1172 On the other hand, this manual is @emph{not} intended as an introduction
1173 to programming in assembly language---let alone programming in general!
1174 In a similar vein, we make no attempt to introduce the machine
1175 architecture; we do @emph{not} describe the instruction set, standard
1176 mnemonics, registers or addressing modes that are standard to a
1177 particular architecture.
1179 You may want to consult the manufacturer's
1180 machine architecture manual for this information.
1184 For information on the H8/300 machine instruction set, see @cite{H8/300
1185 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1186 Programming Manual} (Renesas).
1189 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1190 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1191 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1192 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1195 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1199 @c I think this is premature---doc@cygnus.com, 17jan1991
1201 Throughout this manual, we assume that you are running @dfn{GNU},
1202 the portable operating system from the @dfn{Free Software
1203 Foundation, Inc.}. This restricts our attention to certain kinds of
1204 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1205 once this assumption is granted examples and definitions need less
1208 @command{@value{AS}} is part of a team of programs that turn a high-level
1209 human-readable series of instructions into a low-level
1210 computer-readable series of instructions. Different versions of
1211 @command{@value{AS}} are used for different kinds of computer.
1214 @c There used to be a section "Terminology" here, which defined
1215 @c "contents", "byte", "word", and "long". Defining "word" to any
1216 @c particular size is confusing when the .word directive may generate 16
1217 @c bits on one machine and 32 bits on another; in general, for the user
1218 @c version of this manual, none of these terms seem essential to define.
1219 @c They were used very little even in the former draft of the manual;
1220 @c this draft makes an effort to avoid them (except in names of
1224 @section The GNU Assembler
1226 @c man begin DESCRIPTION
1228 @sc{gnu} @command{as} is really a family of assemblers.
1230 This manual describes @command{@value{AS}}, a member of that family which is
1231 configured for the @value{TARGET} architectures.
1233 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1234 should find a fairly similar environment when you use it on another
1235 architecture. Each version has much in common with the others,
1236 including object file formats, most assembler directives (often called
1237 @dfn{pseudo-ops}) and assembler syntax.@refill
1239 @cindex purpose of @sc{gnu} assembler
1240 @command{@value{AS}} is primarily intended to assemble the output of the
1241 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1242 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1243 assemble correctly everything that other assemblers for the same
1244 machine would assemble.
1246 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1249 @c This remark should appear in generic version of manual; assumption
1250 @c here is that generic version sets M680x0.
1251 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1252 assembler for the same architecture; for example, we know of several
1253 incompatible versions of 680x0 assembly language syntax.
1258 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1259 program in one pass of the source file. This has a subtle impact on the
1260 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1262 @node Object Formats
1263 @section Object File Formats
1265 @cindex object file format
1266 The @sc{gnu} assembler can be configured to produce several alternative
1267 object file formats. For the most part, this does not affect how you
1268 write assembly language programs; but directives for debugging symbols
1269 are typically different in different file formats. @xref{Symbol
1270 Attributes,,Symbol Attributes}.
1273 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1274 @value{OBJ-NAME} format object files.
1276 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1278 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1279 @code{b.out} or COFF format object files.
1282 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1283 SOM or ELF format object files.
1288 @section Command Line
1290 @cindex command line conventions
1292 After the program name @command{@value{AS}}, the command line may contain
1293 options and file names. Options may appear in any order, and may be
1294 before, after, or between file names. The order of file names is
1297 @cindex standard input, as input file
1299 @file{--} (two hyphens) by itself names the standard input file
1300 explicitly, as one of the files for @command{@value{AS}} to assemble.
1302 @cindex options, command line
1303 Except for @samp{--} any command line argument that begins with a
1304 hyphen (@samp{-}) is an option. Each option changes the behavior of
1305 @command{@value{AS}}. No option changes the way another option works. An
1306 option is a @samp{-} followed by one or more letters; the case of
1307 the letter is important. All options are optional.
1309 Some options expect exactly one file name to follow them. The file
1310 name may either immediately follow the option's letter (compatible
1311 with older assemblers) or it may be the next command argument (@sc{gnu}
1312 standard). These two command lines are equivalent:
1315 @value{AS} -o my-object-file.o mumble.s
1316 @value{AS} -omy-object-file.o mumble.s
1320 @section Input Files
1323 @cindex source program
1324 @cindex files, input
1325 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1326 describe the program input to one run of @command{@value{AS}}. The program may
1327 be in one or more files; how the source is partitioned into files
1328 doesn't change the meaning of the source.
1330 @c I added "con" prefix to "catenation" just to prove I can overcome my
1331 @c APL training... doc@cygnus.com
1332 The source program is a concatenation of the text in all the files, in the
1335 @c man begin DESCRIPTION
1336 Each time you run @command{@value{AS}} it assembles exactly one source
1337 program. The source program is made up of one or more files.
1338 (The standard input is also a file.)
1340 You give @command{@value{AS}} a command line that has zero or more input file
1341 names. The input files are read (from left file name to right). A
1342 command line argument (in any position) that has no special meaning
1343 is taken to be an input file name.
1345 If you give @command{@value{AS}} no file names it attempts to read one input file
1346 from the @command{@value{AS}} standard input, which is normally your terminal. You
1347 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1350 Use @samp{--} if you need to explicitly name the standard input file
1351 in your command line.
1353 If the source is empty, @command{@value{AS}} produces a small, empty object
1358 @subheading Filenames and Line-numbers
1360 @cindex input file linenumbers
1361 @cindex line numbers, in input files
1362 There are two ways of locating a line in the input file (or files) and
1363 either may be used in reporting error messages. One way refers to a line
1364 number in a physical file; the other refers to a line number in a
1365 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1367 @dfn{Physical files} are those files named in the command line given
1368 to @command{@value{AS}}.
1370 @dfn{Logical files} are simply names declared explicitly by assembler
1371 directives; they bear no relation to physical files. Logical file names help
1372 error messages reflect the original source file, when @command{@value{AS}} source
1373 is itself synthesized from other files. @command{@value{AS}} understands the
1374 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1375 @ref{File,,@code{.file}}.
1378 @section Output (Object) File
1384 Every time you run @command{@value{AS}} it produces an output file, which is
1385 your assembly language program translated into numbers. This file
1386 is the object file. Its default name is
1394 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1396 You can give it another name by using the @option{-o} option. Conventionally,
1397 object file names end with @file{.o}. The default name is used for historical
1398 reasons: older assemblers were capable of assembling self-contained programs
1399 directly into a runnable program. (For some formats, this isn't currently
1400 possible, but it can be done for the @code{a.out} format.)
1404 The object file is meant for input to the linker @code{@value{LD}}. It contains
1405 assembled program code, information to help @code{@value{LD}} integrate
1406 the assembled program into a runnable file, and (optionally) symbolic
1407 information for the debugger.
1409 @c link above to some info file(s) like the description of a.out.
1410 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1413 @section Error and Warning Messages
1415 @c man begin DESCRIPTION
1417 @cindex error messages
1418 @cindex warning messages
1419 @cindex messages from assembler
1420 @command{@value{AS}} may write warnings and error messages to the standard error
1421 file (usually your terminal). This should not happen when a compiler
1422 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1423 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1424 grave problem that stops the assembly.
1428 @cindex format of warning messages
1429 Warning messages have the format
1432 file_name:@b{NNN}:Warning Message Text
1436 @cindex line numbers, in warnings/errors
1437 (where @b{NNN} is a line number). If a logical file name has been given
1438 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1439 the current input file is used. If a logical line number was given
1441 (@pxref{Line,,@code{.line}})
1443 then it is used to calculate the number printed,
1444 otherwise the actual line in the current source file is printed. The
1445 message text is intended to be self explanatory (in the grand Unix
1448 @cindex format of error messages
1449 Error messages have the format
1451 file_name:@b{NNN}:FATAL:Error Message Text
1453 The file name and line number are derived as for warning
1454 messages. The actual message text may be rather less explanatory
1455 because many of them aren't supposed to happen.
1458 @chapter Command-Line Options
1460 @cindex options, all versions of assembler
1461 This chapter describes command-line options available in @emph{all}
1462 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1464 to the @value{TARGET} target.
1467 to particular machine architectures.
1470 @c man begin DESCRIPTION
1472 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1473 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1474 The assembler arguments must be separated from each other (and the @samp{-Wa})
1475 by commas. For example:
1478 gcc -c -g -O -Wa,-alh,-L file.c
1482 This passes two options to the assembler: @samp{-alh} (emit a listing to
1483 standard output with high-level and assembly source) and @samp{-L} (retain
1484 local symbols in the symbol table).
1486 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1487 command-line options are automatically passed to the assembler by the compiler.
1488 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1489 precisely what options it passes to each compilation pass, including the
1495 * a:: -a[cdhlns] enable listings
1496 * alternate:: --alternate enable alternate macro syntax
1497 * D:: -D for compatibility
1498 * f:: -f to work faster
1499 * I:: -I for .include search path
1500 @ifclear DIFF-TBL-KLUGE
1501 * K:: -K for compatibility
1503 @ifset DIFF-TBL-KLUGE
1504 * K:: -K for difference tables
1507 * L:: -L to retain local labels
1508 * listing:: --listing-XXX to configure listing output
1509 * M:: -M or --mri to assemble in MRI compatibility mode
1510 * MD:: --MD for dependency tracking
1511 * o:: -o to name the object file
1512 * R:: -R to join data and text sections
1513 * statistics:: --statistics to see statistics about assembly
1514 * traditional-format:: --traditional-format for compatible output
1515 * v:: -v to announce version
1516 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1517 * Z:: -Z to make object file even after errors
1521 @section Enable Listings: @option{-a[cdhlns]}
1530 @cindex listings, enabling
1531 @cindex assembly listings, enabling
1533 These options enable listing output from the assembler. By itself,
1534 @samp{-a} requests high-level, assembly, and symbols listing.
1535 You can use other letters to select specific options for the list:
1536 @samp{-ah} requests a high-level language listing,
1537 @samp{-al} requests an output-program assembly listing, and
1538 @samp{-as} requests a symbol table listing.
1539 High-level listings require that a compiler debugging option like
1540 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1543 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1544 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1545 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1546 omitted from the listing.
1548 Use the @samp{-ad} option to omit debugging directives from the
1551 Once you have specified one of these options, you can further control
1552 listing output and its appearance using the directives @code{.list},
1553 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1555 The @samp{-an} option turns off all forms processing.
1556 If you do not request listing output with one of the @samp{-a} options, the
1557 listing-control directives have no effect.
1559 The letters after @samp{-a} may be combined into one option,
1560 @emph{e.g.}, @samp{-aln}.
1562 Note if the assembler source is coming from the standard input (eg because it
1563 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1564 is being used) then the listing will not contain any comments or preprocessor
1565 directives. This is because the listing code buffers input source lines from
1566 stdin only after they have been preprocessed by the assembler. This reduces
1567 memory usage and makes the code more efficient.
1570 @section @option{--alternate}
1573 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1576 @section @option{-D}
1579 This option has no effect whatsoever, but it is accepted to make it more
1580 likely that scripts written for other assemblers also work with
1581 @command{@value{AS}}.
1584 @section Work Faster: @option{-f}
1587 @cindex trusted compiler
1588 @cindex faster processing (@option{-f})
1589 @samp{-f} should only be used when assembling programs written by a
1590 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1591 and comment preprocessing on
1592 the input file(s) before assembling them. @xref{Preprocessing,
1596 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1597 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1602 @section @code{.include} Search Path: @option{-I} @var{path}
1604 @kindex -I @var{path}
1605 @cindex paths for @code{.include}
1606 @cindex search path for @code{.include}
1607 @cindex @code{include} directive search path
1608 Use this option to add a @var{path} to the list of directories
1609 @command{@value{AS}} searches for files specified in @code{.include}
1610 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1611 many times as necessary to include a variety of paths. The current
1612 working directory is always searched first; after that, @command{@value{AS}}
1613 searches any @samp{-I} directories in the same order as they were
1614 specified (left to right) on the command line.
1617 @section Difference Tables: @option{-K}
1620 @ifclear DIFF-TBL-KLUGE
1621 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1622 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1623 where it can be used to warn when the assembler alters the machine code
1624 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1625 family does not have the addressing limitations that sometimes lead to this
1626 alteration on other platforms.
1629 @ifset DIFF-TBL-KLUGE
1630 @cindex difference tables, warning
1631 @cindex warning for altered difference tables
1632 @command{@value{AS}} sometimes alters the code emitted for directives of the form
1633 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1634 You can use the @samp{-K} option if you want a warning issued when this
1639 @section Include Local Labels: @option{-L}
1642 @cindex local labels, retaining in output
1643 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1644 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1645 debugging, because they are intended for the use of programs (like
1646 compilers) that compose assembler programs, not for your notice.
1647 Normally both @command{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1648 normally debug with them.
1650 This option tells @command{@value{AS}} to retain those @samp{L@dots{}} symbols
1651 in the object file. Usually if you do this you also tell the linker
1652 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1654 By default, a local label is any label beginning with @samp{L}, but each
1655 target is allowed to redefine the local label prefix.
1657 On the HPPA local labels begin with @samp{L$}.
1661 @section Configuring listing output: @option{--listing}
1663 The listing feature of the assembler can be enabled via the command line switch
1664 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1665 hex dump of the corresponding locations in the output object file, and displays
1666 them as a listing file. The format of this listing can be controlled by pseudo
1667 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1668 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1671 @item --listing-lhs-width=@samp{number}
1672 @kindex --listing-lhs-width
1673 @cindex Width of first line disassembly output
1674 Sets the maximum width, in words, of the first line of the hex byte dump. This
1675 dump appears on the left hand side of the listing output.
1677 @item --listing-lhs-width2=@samp{number}
1678 @kindex --listing-lhs-width2
1679 @cindex Width of continuation lines of disassembly output
1680 Sets the maximum width, in words, of any further lines of the hex byte dump for
1681 a given input source line. If this value is not specified, it defaults to being
1682 the same as the value specified for @samp{--listing-lhs-width}. If neither
1683 switch is used the default is to one.
1685 @item --listing-rhs-width=@samp{number}
1686 @kindex --listing-rhs-width
1687 @cindex Width of source line output
1688 Sets the maximum width, in characters, of the source line that is displayed
1689 alongside the hex dump. The default value for this parameter is 100. The
1690 source line is displayed on the right hand side of the listing output.
1692 @item --listing-cont-lines=@samp{number}
1693 @kindex --listing-cont-lines
1694 @cindex Maximum number of continuation lines
1695 Sets the maximum number of continuation lines of hex dump that will be
1696 displayed for a given single line of source input. The default value is 4.
1700 @section Assemble in MRI Compatibility Mode: @option{-M}
1703 @cindex MRI compatibility mode
1704 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1705 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1706 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1707 configured target) assembler from Microtec Research. The exact nature of the
1708 MRI syntax will not be documented here; see the MRI manuals for more
1709 information. Note in particular that the handling of macros and macro
1710 arguments is somewhat different. The purpose of this option is to permit
1711 assembling existing MRI assembler code using @command{@value{AS}}.
1713 The MRI compatibility is not complete. Certain operations of the MRI assembler
1714 depend upon its object file format, and can not be supported using other object
1715 file formats. Supporting these would require enhancing each object file format
1716 individually. These are:
1719 @item global symbols in common section
1721 The m68k MRI assembler supports common sections which are merged by the linker.
1722 Other object file formats do not support this. @command{@value{AS}} handles
1723 common sections by treating them as a single common symbol. It permits local
1724 symbols to be defined within a common section, but it can not support global
1725 symbols, since it has no way to describe them.
1727 @item complex relocations
1729 The MRI assemblers support relocations against a negated section address, and
1730 relocations which combine the start addresses of two or more sections. These
1731 are not support by other object file formats.
1733 @item @code{END} pseudo-op specifying start address
1735 The MRI @code{END} pseudo-op permits the specification of a start address.
1736 This is not supported by other object file formats. The start address may
1737 instead be specified using the @option{-e} option to the linker, or in a linker
1740 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1742 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1743 name to the output file. This is not supported by other object file formats.
1745 @item @code{ORG} pseudo-op
1747 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1748 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1749 which changes the location within the current section. Absolute sections are
1750 not supported by other object file formats. The address of a section may be
1751 assigned within a linker script.
1754 There are some other features of the MRI assembler which are not supported by
1755 @command{@value{AS}}, typically either because they are difficult or because they
1756 seem of little consequence. Some of these may be supported in future releases.
1760 @item EBCDIC strings
1762 EBCDIC strings are not supported.
1764 @item packed binary coded decimal
1766 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1767 and @code{DCB.P} pseudo-ops are not supported.
1769 @item @code{FEQU} pseudo-op
1771 The m68k @code{FEQU} pseudo-op is not supported.
1773 @item @code{NOOBJ} pseudo-op
1775 The m68k @code{NOOBJ} pseudo-op is not supported.
1777 @item @code{OPT} branch control options
1779 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1780 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1781 relaxes all branches, whether forward or backward, to an appropriate size, so
1782 these options serve no purpose.
1784 @item @code{OPT} list control options
1786 The following m68k @code{OPT} list control options are ignored: @code{C},
1787 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1788 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1790 @item other @code{OPT} options
1792 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1793 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1795 @item @code{OPT} @code{D} option is default
1797 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1798 @code{OPT NOD} may be used to turn it off.
1800 @item @code{XREF} pseudo-op.
1802 The m68k @code{XREF} pseudo-op is ignored.
1804 @item @code{.debug} pseudo-op
1806 The i960 @code{.debug} pseudo-op is not supported.
1808 @item @code{.extended} pseudo-op
1810 The i960 @code{.extended} pseudo-op is not supported.
1812 @item @code{.list} pseudo-op.
1814 The various options of the i960 @code{.list} pseudo-op are not supported.
1816 @item @code{.optimize} pseudo-op
1818 The i960 @code{.optimize} pseudo-op is not supported.
1820 @item @code{.output} pseudo-op
1822 The i960 @code{.output} pseudo-op is not supported.
1824 @item @code{.setreal} pseudo-op
1826 The i960 @code{.setreal} pseudo-op is not supported.
1831 @section Dependency Tracking: @option{--MD}
1834 @cindex dependency tracking
1837 @command{@value{AS}} can generate a dependency file for the file it creates. This
1838 file consists of a single rule suitable for @code{make} describing the
1839 dependencies of the main source file.
1841 The rule is written to the file named in its argument.
1843 This feature is used in the automatic updating of makefiles.
1846 @section Name the Object File: @option{-o}
1849 @cindex naming object file
1850 @cindex object file name
1851 There is always one object file output when you run @command{@value{AS}}. By
1852 default it has the name
1855 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1869 You use this option (which takes exactly one filename) to give the
1870 object file a different name.
1872 Whatever the object file is called, @command{@value{AS}} overwrites any
1873 existing file of the same name.
1876 @section Join Data and Text Sections: @option{-R}
1879 @cindex data and text sections, joining
1880 @cindex text and data sections, joining
1881 @cindex joining text and data sections
1882 @cindex merging text and data sections
1883 @option{-R} tells @command{@value{AS}} to write the object file as if all
1884 data-section data lives in the text section. This is only done at
1885 the very last moment: your binary data are the same, but data
1886 section parts are relocated differently. The data section part of
1887 your object file is zero bytes long because all its bytes are
1888 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1890 When you specify @option{-R} it would be possible to generate shorter
1891 address displacements (because we do not have to cross between text and
1892 data section). We refrain from doing this simply for compatibility with
1893 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1896 When @command{@value{AS}} is configured for COFF or ELF output,
1897 this option is only useful if you use sections named @samp{.text} and
1902 @option{-R} is not supported for any of the HPPA targets. Using
1903 @option{-R} generates a warning from @command{@value{AS}}.
1907 @section Display Assembly Statistics: @option{--statistics}
1909 @kindex --statistics
1910 @cindex statistics, about assembly
1911 @cindex time, total for assembly
1912 @cindex space used, maximum for assembly
1913 Use @samp{--statistics} to display two statistics about the resources used by
1914 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1915 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1918 @node traditional-format
1919 @section Compatible Output: @option{--traditional-format}
1921 @kindex --traditional-format
1922 For some targets, the output of @command{@value{AS}} is different in some ways
1923 from the output of some existing assembler. This switch requests
1924 @command{@value{AS}} to use the traditional format instead.
1926 For example, it disables the exception frame optimizations which
1927 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
1930 @section Announce Version: @option{-v}
1934 @cindex assembler version
1935 @cindex version of assembler
1936 You can find out what version of as is running by including the
1937 option @samp{-v} (which you can also spell as @samp{-version}) on the
1941 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
1943 @command{@value{AS}} should never give a warning or error message when
1944 assembling compiler output. But programs written by people often
1945 cause @command{@value{AS}} to give a warning that a particular assumption was
1946 made. All such warnings are directed to the standard error file.
1950 @cindex suppressing warnings
1951 @cindex warnings, suppressing
1952 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
1953 This only affects the warning messages: it does not change any particular of
1954 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
1957 @kindex --fatal-warnings
1958 @cindex errors, caused by warnings
1959 @cindex warnings, causing error
1960 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
1961 files that generate warnings to be in error.
1964 @cindex warnings, switching on
1965 You can switch these options off again by specifying @option{--warn}, which
1966 causes warnings to be output as usual.
1969 @section Generate Object File in Spite of Errors: @option{-Z}
1970 @cindex object file, after errors
1971 @cindex errors, continuing after
1972 After an error message, @command{@value{AS}} normally produces no output. If for
1973 some reason you are interested in object file output even after
1974 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
1975 option. If there are any errors, @command{@value{AS}} continues anyways, and
1976 writes an object file after a final warning message of the form @samp{@var{n}
1977 errors, @var{m} warnings, generating bad object file.}
1982 @cindex machine-independent syntax
1983 @cindex syntax, machine-independent
1984 This chapter describes the machine-independent syntax allowed in a
1985 source file. @command{@value{AS}} syntax is similar to what many other
1986 assemblers use; it is inspired by the BSD 4.2
1991 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
1995 * Preprocessing:: Preprocessing
1996 * Whitespace:: Whitespace
1997 * Comments:: Comments
1998 * Symbol Intro:: Symbols
1999 * Statements:: Statements
2000 * Constants:: Constants
2004 @section Preprocessing
2006 @cindex preprocessing
2007 The @command{@value{AS}} internal preprocessor:
2009 @cindex whitespace, removed by preprocessor
2011 adjusts and removes extra whitespace. It leaves one space or tab before
2012 the keywords on a line, and turns any other whitespace on the line into
2015 @cindex comments, removed by preprocessor
2017 removes all comments, replacing them with a single space, or an
2018 appropriate number of newlines.
2020 @cindex constants, converted by preprocessor
2022 converts character constants into the appropriate numeric values.
2025 It does not do macro processing, include file handling, or
2026 anything else you may get from your C compiler's preprocessor. You can
2027 do include file processing with the @code{.include} directive
2028 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2029 to get other ``CPP'' style preprocessing by giving the input file a
2030 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
2031 Output, gcc.info, Using GNU CC}.
2033 Excess whitespace, comments, and character constants
2034 cannot be used in the portions of the input text that are not
2037 @cindex turning preprocessing on and off
2038 @cindex preprocessing, turning on and off
2041 If the first line of an input file is @code{#NO_APP} or if you use the
2042 @samp{-f} option, whitespace and comments are not removed from the input file.
2043 Within an input file, you can ask for whitespace and comment removal in
2044 specific portions of the by putting a line that says @code{#APP} before the
2045 text that may contain whitespace or comments, and putting a line that says
2046 @code{#NO_APP} after this text. This feature is mainly intend to support
2047 @code{asm} statements in compilers whose output is otherwise free of comments
2054 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2055 Whitespace is used to separate symbols, and to make programs neater for
2056 people to read. Unless within character constants
2057 (@pxref{Characters,,Character Constants}), any whitespace means the same
2058 as exactly one space.
2064 There are two ways of rendering comments to @command{@value{AS}}. In both
2065 cases the comment is equivalent to one space.
2067 Anything from @samp{/*} through the next @samp{*/} is a comment.
2068 This means you may not nest these comments.
2072 The only way to include a newline ('\n') in a comment
2073 is to use this sort of comment.
2076 /* This sort of comment does not nest. */
2079 @cindex line comment character
2080 Anything from the @dfn{line comment} character to the next newline
2081 is considered a comment and is ignored. The line comment character is
2083 @samp{;} on the ARC;
2086 @samp{@@} on the ARM;
2089 @samp{;} for the H8/300 family;
2092 @samp{;} for the HPPA;
2095 @samp{#} on the i386 and x86-64;
2098 @samp{#} on the i960;
2101 @samp{;} for the PDP-11;
2104 @samp{;} for picoJava;
2107 @samp{#} for Motorola PowerPC;
2110 @samp{!} for the Renesas / SuperH SH;
2113 @samp{!} on the SPARC;
2116 @samp{#} on the ip2k;
2119 @samp{#} on the m32c;
2122 @samp{#} on the m32r;
2125 @samp{|} on the 680x0;
2128 @samp{#} on the 68HC11 and 68HC12;
2131 @samp{#} on the Vax;
2134 @samp{!} for the Z8000;
2137 @samp{#} on the V850;
2140 @samp{#} for Xtensa systems;
2142 see @ref{Machine Dependencies}. @refill
2143 @c FIXME What about i860?
2146 On some machines there are two different line comment characters. One
2147 character only begins a comment if it is the first non-whitespace character on
2148 a line, while the other always begins a comment.
2152 The V850 assembler also supports a double dash as starting a comment that
2153 extends to the end of the line.
2159 @cindex lines starting with @code{#}
2160 @cindex logical line numbers
2161 To be compatible with past assemblers, lines that begin with @samp{#} have a
2162 special interpretation. Following the @samp{#} should be an absolute
2163 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2164 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
2165 new logical file name. The rest of the line, if any, should be whitespace.
2167 If the first non-whitespace characters on the line are not numeric,
2168 the line is ignored. (Just like a comment.)
2171 # This is an ordinary comment.
2172 # 42-6 "new_file_name" # New logical file name
2173 # This is logical line # 36.
2175 This feature is deprecated, and may disappear from future versions
2176 of @command{@value{AS}}.
2181 @cindex characters used in symbols
2182 @ifclear SPECIAL-SYMS
2183 A @dfn{symbol} is one or more characters chosen from the set of all
2184 letters (both upper and lower case), digits and the three characters
2190 A @dfn{symbol} is one or more characters chosen from the set of all
2191 letters (both upper and lower case), digits and the three characters
2192 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2198 On most machines, you can also use @code{$} in symbol names; exceptions
2199 are noted in @ref{Machine Dependencies}.
2201 No symbol may begin with a digit. Case is significant.
2202 There is no length limit: all characters are significant. Symbols are
2203 delimited by characters not in that set, or by the beginning of a file
2204 (since the source program must end with a newline, the end of a file is
2205 not a possible symbol delimiter). @xref{Symbols}.
2206 @cindex length of symbols
2211 @cindex statements, structure of
2212 @cindex line separator character
2213 @cindex statement separator character
2215 @ifclear abnormal-separator
2216 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2217 semicolon (@samp{;}). The newline or semicolon is considered part of
2218 the preceding statement. Newlines and semicolons within character
2219 constants are an exception: they do not end statements.
2221 @ifset abnormal-separator
2223 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2224 point (@samp{!}). The newline or exclamation point is considered part of the
2225 preceding statement. Newlines and exclamation points within character
2226 constants are an exception: they do not end statements.
2229 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2230 H8/300) a dollar sign (@samp{$}); or (for the Renesas-SH) a semicolon
2231 (@samp{;}). The newline or separator character is considered part of
2232 the preceding statement. Newlines and separators within character
2233 constants are an exception: they do not end statements.
2238 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2239 separator character. (The line separator is usually @samp{;}, unless
2240 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
2241 newline or separator character is considered part of the preceding
2242 statement. Newlines and separators within character constants are an
2243 exception: they do not end statements.
2246 @cindex newline, required at file end
2247 @cindex EOF, newline must precede
2248 It is an error to end any statement with end-of-file: the last
2249 character of any input file should be a newline.@refill
2251 An empty statement is allowed, and may include whitespace. It is ignored.
2253 @cindex instructions and directives
2254 @cindex directives and instructions
2255 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2256 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2258 A statement begins with zero or more labels, optionally followed by a
2259 key symbol which determines what kind of statement it is. The key
2260 symbol determines the syntax of the rest of the statement. If the
2261 symbol begins with a dot @samp{.} then the statement is an assembler
2262 directive: typically valid for any computer. If the symbol begins with
2263 a letter the statement is an assembly language @dfn{instruction}: it
2264 assembles into a machine language instruction.
2266 Different versions of @command{@value{AS}} for different computers
2267 recognize different instructions. In fact, the same symbol may
2268 represent a different instruction in a different computer's assembly
2272 @cindex @code{:} (label)
2273 @cindex label (@code{:})
2274 A label is a symbol immediately followed by a colon (@code{:}).
2275 Whitespace before a label or after a colon is permitted, but you may not
2276 have whitespace between a label's symbol and its colon. @xref{Labels}.
2279 For HPPA targets, labels need not be immediately followed by a colon, but
2280 the definition of a label must begin in column zero. This also implies that
2281 only one label may be defined on each line.
2285 label: .directive followed by something
2286 another_label: # This is an empty statement.
2287 instruction operand_1, operand_2, @dots{}
2294 A constant is a number, written so that its value is known by
2295 inspection, without knowing any context. Like this:
2298 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2299 .ascii "Ring the bell\7" # A string constant.
2300 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2301 .float 0f-314159265358979323846264338327\
2302 95028841971.693993751E-40 # - pi, a flonum.
2307 * Characters:: Character Constants
2308 * Numbers:: Number Constants
2312 @subsection Character Constants
2314 @cindex character constants
2315 @cindex constants, character
2316 There are two kinds of character constants. A @dfn{character} stands
2317 for one character in one byte and its value may be used in
2318 numeric expressions. String constants (properly called string
2319 @emph{literals}) are potentially many bytes and their values may not be
2320 used in arithmetic expressions.
2324 * Chars:: Characters
2328 @subsubsection Strings
2330 @cindex string constants
2331 @cindex constants, string
2332 A @dfn{string} is written between double-quotes. It may contain
2333 double-quotes or null characters. The way to get special characters
2334 into a string is to @dfn{escape} these characters: precede them with
2335 a backslash @samp{\} character. For example @samp{\\} represents
2336 one backslash: the first @code{\} is an escape which tells
2337 @command{@value{AS}} to interpret the second character literally as a backslash
2338 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2339 escape character). The complete list of escapes follows.
2341 @cindex escape codes, character
2342 @cindex character escape codes
2345 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2347 @cindex @code{\b} (backspace character)
2348 @cindex backspace (@code{\b})
2350 Mnemonic for backspace; for ASCII this is octal code 010.
2353 @c Mnemonic for EOText; for ASCII this is octal code 004.
2355 @cindex @code{\f} (formfeed character)
2356 @cindex formfeed (@code{\f})
2358 Mnemonic for FormFeed; for ASCII this is octal code 014.
2360 @cindex @code{\n} (newline character)
2361 @cindex newline (@code{\n})
2363 Mnemonic for newline; for ASCII this is octal code 012.
2366 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2368 @cindex @code{\r} (carriage return character)
2369 @cindex carriage return (@code{\r})
2371 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2374 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2375 @c other assemblers.
2377 @cindex @code{\t} (tab)
2378 @cindex tab (@code{\t})
2380 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2383 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2384 @c @item \x @var{digit} @var{digit} @var{digit}
2385 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2387 @cindex @code{\@var{ddd}} (octal character code)
2388 @cindex octal character code (@code{\@var{ddd}})
2389 @item \ @var{digit} @var{digit} @var{digit}
2390 An octal character code. The numeric code is 3 octal digits.
2391 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2392 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2394 @cindex @code{\@var{xd...}} (hex character code)
2395 @cindex hex character code (@code{\@var{xd...}})
2396 @item \@code{x} @var{hex-digits...}
2397 A hex character code. All trailing hex digits are combined. Either upper or
2398 lower case @code{x} works.
2400 @cindex @code{\\} (@samp{\} character)
2401 @cindex backslash (@code{\\})
2403 Represents one @samp{\} character.
2406 @c Represents one @samp{'} (accent acute) character.
2407 @c This is needed in single character literals
2408 @c (@xref{Characters,,Character Constants}.) to represent
2411 @cindex @code{\"} (doublequote character)
2412 @cindex doublequote (@code{\"})
2414 Represents one @samp{"} character. Needed in strings to represent
2415 this character, because an unescaped @samp{"} would end the string.
2417 @item \ @var{anything-else}
2418 Any other character when escaped by @kbd{\} gives a warning, but
2419 assembles as if the @samp{\} was not present. The idea is that if
2420 you used an escape sequence you clearly didn't want the literal
2421 interpretation of the following character. However @command{@value{AS}} has no
2422 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2423 code and warns you of the fact.
2426 Which characters are escapable, and what those escapes represent,
2427 varies widely among assemblers. The current set is what we think
2428 the BSD 4.2 assembler recognizes, and is a subset of what most C
2429 compilers recognize. If you are in doubt, do not use an escape
2433 @subsubsection Characters
2435 @cindex single character constant
2436 @cindex character, single
2437 @cindex constant, single character
2438 A single character may be written as a single quote immediately
2439 followed by that character. The same escapes apply to characters as
2440 to strings. So if you want to write the character backslash, you
2441 must write @kbd{'\\} where the first @code{\} escapes the second
2442 @code{\}. As you can see, the quote is an acute accent, not a
2443 grave accent. A newline
2445 @ifclear abnormal-separator
2446 (or semicolon @samp{;})
2448 @ifset abnormal-separator
2450 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2455 immediately following an acute accent is taken as a literal character
2456 and does not count as the end of a statement. The value of a character
2457 constant in a numeric expression is the machine's byte-wide code for
2458 that character. @command{@value{AS}} assumes your character code is ASCII:
2459 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2462 @subsection Number Constants
2464 @cindex constants, number
2465 @cindex number constants
2466 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2467 are stored in the target machine. @emph{Integers} are numbers that
2468 would fit into an @code{int} in the C language. @emph{Bignums} are
2469 integers, but they are stored in more than 32 bits. @emph{Flonums}
2470 are floating point numbers, described below.
2473 * Integers:: Integers
2478 * Bit Fields:: Bit Fields
2484 @subsubsection Integers
2486 @cindex constants, integer
2488 @cindex binary integers
2489 @cindex integers, binary
2490 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2491 the binary digits @samp{01}.
2493 @cindex octal integers
2494 @cindex integers, octal
2495 An octal integer is @samp{0} followed by zero or more of the octal
2496 digits (@samp{01234567}).
2498 @cindex decimal integers
2499 @cindex integers, decimal
2500 A decimal integer starts with a non-zero digit followed by zero or
2501 more digits (@samp{0123456789}).
2503 @cindex hexadecimal integers
2504 @cindex integers, hexadecimal
2505 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2506 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2508 Integers have the usual values. To denote a negative integer, use
2509 the prefix operator @samp{-} discussed under expressions
2510 (@pxref{Prefix Ops,,Prefix Operators}).
2513 @subsubsection Bignums
2516 @cindex constants, bignum
2517 A @dfn{bignum} has the same syntax and semantics as an integer
2518 except that the number (or its negative) takes more than 32 bits to
2519 represent in binary. The distinction is made because in some places
2520 integers are permitted while bignums are not.
2523 @subsubsection Flonums
2525 @cindex floating point numbers
2526 @cindex constants, floating point
2528 @cindex precision, floating point
2529 A @dfn{flonum} represents a floating point number. The translation is
2530 indirect: a decimal floating point number from the text is converted by
2531 @command{@value{AS}} to a generic binary floating point number of more than
2532 sufficient precision. This generic floating point number is converted
2533 to a particular computer's floating point format (or formats) by a
2534 portion of @command{@value{AS}} specialized to that computer.
2536 A flonum is written by writing (in order)
2541 (@samp{0} is optional on the HPPA.)
2545 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2547 @kbd{e} is recommended. Case is not important.
2549 @c FIXME: verify if flonum syntax really this vague for most cases
2550 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2551 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2554 On the H8/300, Renesas / SuperH SH,
2555 and AMD 29K architectures, the letter must be
2556 one of the letters @samp{DFPRSX} (in upper or lower case).
2558 On the ARC, the letter must be one of the letters @samp{DFRS}
2559 (in upper or lower case).
2561 On the Intel 960 architecture, the letter must be
2562 one of the letters @samp{DFT} (in upper or lower case).
2564 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2568 One of the letters @samp{DFRS} (in upper or lower case).
2571 One of the letters @samp{DFPRSX} (in upper or lower case).
2574 The letter @samp{E} (upper case only).
2577 One of the letters @samp{DFT} (in upper or lower case).
2582 An optional sign: either @samp{+} or @samp{-}.
2585 An optional @dfn{integer part}: zero or more decimal digits.
2588 An optional @dfn{fractional part}: @samp{.} followed by zero
2589 or more decimal digits.
2592 An optional exponent, consisting of:
2596 An @samp{E} or @samp{e}.
2597 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2598 @c principle this can perfectly well be different on different targets.
2600 Optional sign: either @samp{+} or @samp{-}.
2602 One or more decimal digits.
2607 At least one of the integer part or the fractional part must be
2608 present. The floating point number has the usual base-10 value.
2610 @command{@value{AS}} does all processing using integers. Flonums are computed
2611 independently of any floating point hardware in the computer running
2612 @command{@value{AS}}.
2616 @c Bit fields are written as a general facility but are also controlled
2617 @c by a conditional-compilation flag---which is as of now (21mar91)
2618 @c turned on only by the i960 config of GAS.
2620 @subsubsection Bit Fields
2623 @cindex constants, bit field
2624 You can also define numeric constants as @dfn{bit fields}.
2625 specify two numbers separated by a colon---
2627 @var{mask}:@var{value}
2630 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2633 The resulting number is then packed
2635 @c this conditional paren in case bit fields turned on elsewhere than 960
2636 (in host-dependent byte order)
2638 into a field whose width depends on which assembler directive has the
2639 bit-field as its argument. Overflow (a result from the bitwise and
2640 requiring more binary digits to represent) is not an error; instead,
2641 more constants are generated, of the specified width, beginning with the
2642 least significant digits.@refill
2644 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2645 @code{.short}, and @code{.word} accept bit-field arguments.
2650 @chapter Sections and Relocation
2655 * Secs Background:: Background
2656 * Ld Sections:: Linker Sections
2657 * As Sections:: Assembler Internal Sections
2658 * Sub-Sections:: Sub-Sections
2662 @node Secs Background
2665 Roughly, a section is a range of addresses, with no gaps; all data
2666 ``in'' those addresses is treated the same for some particular purpose.
2667 For example there may be a ``read only'' section.
2669 @cindex linker, and assembler
2670 @cindex assembler, and linker
2671 The linker @code{@value{LD}} reads many object files (partial programs) and
2672 combines their contents to form a runnable program. When @command{@value{AS}}
2673 emits an object file, the partial program is assumed to start at address 0.
2674 @code{@value{LD}} assigns the final addresses for the partial program, so that
2675 different partial programs do not overlap. This is actually an
2676 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2679 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2680 addresses. These blocks slide to their run-time addresses as rigid
2681 units; their length does not change and neither does the order of bytes
2682 within them. Such a rigid unit is called a @emph{section}. Assigning
2683 run-time addresses to sections is called @dfn{relocation}. It includes
2684 the task of adjusting mentions of object-file addresses so they refer to
2685 the proper run-time addresses.
2687 For the H8/300, and for the Renesas / SuperH SH,
2688 @command{@value{AS}} pads sections if needed to
2689 ensure they end on a word (sixteen bit) boundary.
2692 @cindex standard assembler sections
2693 An object file written by @command{@value{AS}} has at least three sections, any
2694 of which may be empty. These are named @dfn{text}, @dfn{data} and
2699 When it generates COFF or ELF output,
2701 @command{@value{AS}} can also generate whatever other named sections you specify
2702 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2703 If you do not use any directives that place output in the @samp{.text}
2704 or @samp{.data} sections, these sections still exist, but are empty.
2709 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2711 @command{@value{AS}} can also generate whatever other named sections you
2712 specify using the @samp{.space} and @samp{.subspace} directives. See
2713 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2714 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2715 assembler directives.
2718 Additionally, @command{@value{AS}} uses different names for the standard
2719 text, data, and bss sections when generating SOM output. Program text
2720 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2721 BSS into @samp{$BSS$}.
2725 Within the object file, the text section starts at address @code{0}, the
2726 data section follows, and the bss section follows the data section.
2729 When generating either SOM or ELF output files on the HPPA, the text
2730 section starts at address @code{0}, the data section at address
2731 @code{0x4000000}, and the bss section follows the data section.
2734 To let @code{@value{LD}} know which data changes when the sections are
2735 relocated, and how to change that data, @command{@value{AS}} also writes to the
2736 object file details of the relocation needed. To perform relocation
2737 @code{@value{LD}} must know, each time an address in the object
2741 Where in the object file is the beginning of this reference to
2744 How long (in bytes) is this reference?
2746 Which section does the address refer to? What is the numeric value of
2748 (@var{address}) @minus{} (@var{start-address of section})?
2751 Is the reference to an address ``Program-Counter relative''?
2754 @cindex addresses, format of
2755 @cindex section-relative addressing
2756 In fact, every address @command{@value{AS}} ever uses is expressed as
2758 (@var{section}) + (@var{offset into section})
2761 Further, most expressions @command{@value{AS}} computes have this section-relative
2764 (For some object formats, such as SOM for the HPPA, some expressions are
2765 symbol-relative instead.)
2768 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2769 @var{N} into section @var{secname}.''
2771 Apart from text, data and bss sections you need to know about the
2772 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2773 addresses in the absolute section remain unchanged. For example, address
2774 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2775 @code{@value{LD}}. Although the linker never arranges two partial programs'
2776 data sections with overlapping addresses after linking, @emph{by definition}
2777 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2778 part of a program is always the same address when the program is running as
2779 address @code{@{absolute@ 239@}} in any other part of the program.
2781 The idea of sections is extended to the @dfn{undefined} section. Any
2782 address whose section is unknown at assembly time is by definition
2783 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2784 Since numbers are always defined, the only way to generate an undefined
2785 address is to mention an undefined symbol. A reference to a named
2786 common block would be such a symbol: its value is unknown at assembly
2787 time so it has section @emph{undefined}.
2789 By analogy the word @emph{section} is used to describe groups of sections in
2790 the linked program. @code{@value{LD}} puts all partial programs' text
2791 sections in contiguous addresses in the linked program. It is
2792 customary to refer to the @emph{text section} of a program, meaning all
2793 the addresses of all partial programs' text sections. Likewise for
2794 data and bss sections.
2796 Some sections are manipulated by @code{@value{LD}}; others are invented for
2797 use of @command{@value{AS}} and have no meaning except during assembly.
2800 @section Linker Sections
2801 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2806 @cindex named sections
2807 @cindex sections, named
2808 @item named sections
2811 @cindex text section
2812 @cindex data section
2816 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2817 separate but equal sections. Anything you can say of one section is
2820 When the program is running, however, it is
2821 customary for the text section to be unalterable. The
2822 text section is often shared among processes: it contains
2823 instructions, constants and the like. The data section of a running
2824 program is usually alterable: for example, C variables would be stored
2825 in the data section.
2830 This section contains zeroed bytes when your program begins running. It
2831 is used to hold uninitialized variables or common storage. The length of
2832 each partial program's bss section is important, but because it starts
2833 out containing zeroed bytes there is no need to store explicit zero
2834 bytes in the object file. The bss section was invented to eliminate
2835 those explicit zeros from object files.
2837 @cindex absolute section
2838 @item absolute section
2839 Address 0 of this section is always ``relocated'' to runtime address 0.
2840 This is useful if you want to refer to an address that @code{@value{LD}} must
2841 not change when relocating. In this sense we speak of absolute
2842 addresses being ``unrelocatable'': they do not change during relocation.
2844 @cindex undefined section
2845 @item undefined section
2846 This ``section'' is a catch-all for address references to objects not in
2847 the preceding sections.
2848 @c FIXME: ref to some other doc on obj-file formats could go here.
2851 @cindex relocation example
2852 An idealized example of three relocatable sections follows.
2854 The example uses the traditional section names @samp{.text} and @samp{.data}.
2856 Memory addresses are on the horizontal axis.
2860 @c END TEXI2ROFF-KILL
2863 partial program # 1: |ttttt|dddd|00|
2870 partial program # 2: |TTT|DDD|000|
2873 +--+---+-----+--+----+---+-----+~~
2874 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2875 +--+---+-----+--+----+---+-----+~~
2877 addresses: 0 @dots{}
2884 \line{\it Partial program \#1: \hfil}
2885 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2886 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2888 \line{\it Partial program \#2: \hfil}
2889 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2890 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2892 \line{\it linked program: \hfil}
2893 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2894 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2895 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2896 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2898 \line{\it addresses: \hfil}
2902 @c END TEXI2ROFF-KILL
2905 @section Assembler Internal Sections
2907 @cindex internal assembler sections
2908 @cindex sections in messages, internal
2909 These sections are meant only for the internal use of @command{@value{AS}}. They
2910 have no meaning at run-time. You do not really need to know about these
2911 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2912 warning messages, so it might be helpful to have an idea of their
2913 meanings to @command{@value{AS}}. These sections are used to permit the
2914 value of every expression in your assembly language program to be a
2915 section-relative address.
2918 @cindex assembler internal logic error
2919 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2920 An internal assembler logic error has been found. This means there is a
2921 bug in the assembler.
2923 @cindex expr (internal section)
2925 The assembler stores complex expression internally as combinations of
2926 symbols. When it needs to represent an expression as a symbol, it puts
2927 it in the expr section.
2929 @c FIXME item transfer[t] vector preload
2930 @c FIXME item transfer[t] vector postload
2931 @c FIXME item register
2935 @section Sub-Sections
2937 @cindex numbered subsections
2938 @cindex grouping data
2944 fall into two sections: text and data.
2946 You may have separate groups of
2948 data in named sections
2952 data in named sections
2958 that you want to end up near to each other in the object file, even though they
2959 are not contiguous in the assembler source. @command{@value{AS}} allows you to
2960 use @dfn{subsections} for this purpose. Within each section, there can be
2961 numbered subsections with values from 0 to 8192. Objects assembled into the
2962 same subsection go into the object file together with other objects in the same
2963 subsection. For example, a compiler might want to store constants in the text
2964 section, but might not want to have them interspersed with the program being
2965 assembled. In this case, the compiler could issue a @samp{.text 0} before each
2966 section of code being output, and a @samp{.text 1} before each group of
2967 constants being output.
2969 Subsections are optional. If you do not use subsections, everything
2970 goes in subsection number zero.
2973 Each subsection is zero-padded up to a multiple of four bytes.
2974 (Subsections may be padded a different amount on different flavors
2975 of @command{@value{AS}}.)
2979 On the H8/300 platform, each subsection is zero-padded to a word
2980 boundary (two bytes).
2981 The same is true on the Renesas SH.
2984 @c FIXME section padding (alignment)?
2985 @c Rich Pixley says padding here depends on target obj code format; that
2986 @c doesn't seem particularly useful to say without further elaboration,
2987 @c so for now I say nothing about it. If this is a generic BFD issue,
2988 @c these paragraphs might need to vanish from this manual, and be
2989 @c discussed in BFD chapter of binutils (or some such).
2993 Subsections appear in your object file in numeric order, lowest numbered
2994 to highest. (All this to be compatible with other people's assemblers.)
2995 The object file contains no representation of subsections; @code{@value{LD}} and
2996 other programs that manipulate object files see no trace of them.
2997 They just see all your text subsections as a text section, and all your
2998 data subsections as a data section.
3000 To specify which subsection you want subsequent statements assembled
3001 into, use a numeric argument to specify it, in a @samp{.text
3002 @var{expression}} or a @samp{.data @var{expression}} statement.
3005 When generating COFF output, you
3010 can also use an extra subsection
3011 argument with arbitrary named sections: @samp{.section @var{name},
3016 When generating ELF output, you
3021 can also use the @code{.subsection} directive (@pxref{SubSection})
3022 to specify a subsection: @samp{.subsection @var{expression}}.
3024 @var{Expression} should be an absolute expression.
3025 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
3026 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3027 begins in @code{text 0}. For instance:
3029 .text 0 # The default subsection is text 0 anyway.
3030 .ascii "This lives in the first text subsection. *"
3032 .ascii "But this lives in the second text subsection."
3034 .ascii "This lives in the data section,"
3035 .ascii "in the first data subsection."
3037 .ascii "This lives in the first text section,"
3038 .ascii "immediately following the asterisk (*)."
3041 Each section has a @dfn{location counter} incremented by one for every byte
3042 assembled into that section. Because subsections are merely a convenience
3043 restricted to @command{@value{AS}} there is no concept of a subsection location
3044 counter. There is no way to directly manipulate a location counter---but the
3045 @code{.align} directive changes it, and any label definition captures its
3046 current value. The location counter of the section where statements are being
3047 assembled is said to be the @dfn{active} location counter.
3050 @section bss Section
3053 @cindex common variable storage
3054 The bss section is used for local common variable storage.
3055 You may allocate address space in the bss section, but you may
3056 not dictate data to load into it before your program executes. When
3057 your program starts running, all the contents of the bss
3058 section are zeroed bytes.
3060 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3061 @ref{Lcomm,,@code{.lcomm}}.
3063 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3064 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
3067 When assembling for a target which supports multiple sections, such as ELF or
3068 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3069 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3070 section. Typically the section will only contain symbol definitions and
3071 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3078 Symbols are a central concept: the programmer uses symbols to name
3079 things, the linker uses symbols to link, and the debugger uses symbols
3083 @cindex debuggers, and symbol order
3084 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3085 the same order they were declared. This may break some debuggers.
3090 * Setting Symbols:: Giving Symbols Other Values
3091 * Symbol Names:: Symbol Names
3092 * Dot:: The Special Dot Symbol
3093 * Symbol Attributes:: Symbol Attributes
3100 A @dfn{label} is written as a symbol immediately followed by a colon
3101 @samp{:}. The symbol then represents the current value of the
3102 active location counter, and is, for example, a suitable instruction
3103 operand. You are warned if you use the same symbol to represent two
3104 different locations: the first definition overrides any other
3108 On the HPPA, the usual form for a label need not be immediately followed by a
3109 colon, but instead must start in column zero. Only one label may be defined on
3110 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3111 provides a special directive @code{.label} for defining labels more flexibly.
3114 @node Setting Symbols
3115 @section Giving Symbols Other Values
3117 @cindex assigning values to symbols
3118 @cindex symbol values, assigning
3119 A symbol can be given an arbitrary value by writing a symbol, followed
3120 by an equals sign @samp{=}, followed by an expression
3121 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3122 directive. @xref{Set,,@code{.set}}.
3125 @section Symbol Names
3127 @cindex symbol names
3128 @cindex names, symbol
3129 @ifclear SPECIAL-SYMS
3130 Symbol names begin with a letter or with one of @samp{._}. On most
3131 machines, you can also use @code{$} in symbol names; exceptions are
3132 noted in @ref{Machine Dependencies}. That character may be followed by any
3133 string of digits, letters, dollar signs (unless otherwise noted in
3134 @ref{Machine Dependencies}), and underscores.
3138 Symbol names begin with a letter or with one of @samp{._}. On the
3139 Renesas SH you can also use @code{$} in symbol names. That
3140 character may be followed by any string of digits, letters, dollar signs (save
3141 on the H8/300), and underscores.
3145 Case of letters is significant: @code{foo} is a different symbol name
3148 Each symbol has exactly one name. Each name in an assembly language program
3149 refers to exactly one symbol. You may use that symbol name any number of times
3152 @subheading Local Symbol Names
3154 @cindex local symbol names
3155 @cindex symbol names, local
3156 @cindex temporary symbol names
3157 @cindex symbol names, temporary
3158 Local symbols help compilers and programmers use names temporarily.
3159 They create symbols which are guaranteed to be unique over the entire scope of
3160 the input source code and which can be referred to by a simple notation.
3161 To define a local symbol, write a label of the form @samp{@b{N}:} (where @b{N}
3162 represents any positive integer). To refer to the most recent previous
3163 definition of that symbol write @samp{@b{N}b}, using the same number as when
3164 you defined the label. To refer to the next definition of a local label, write
3165 @samp{@b{N}f}--- The @samp{b} stands for``backwards'' and the @samp{f} stands
3168 There is no restriction on how you can use these labels, and you can reuse them
3169 too. So that it is possible to repeatedly define the same local label (using
3170 the same number @samp{@b{N}}), although you can only refer to the most recently
3171 defined local label of that number (for a backwards reference) or the next
3172 definition of a specific local label for a forward reference. It is also worth
3173 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3174 implemented in a slightly more efficient manner than the others.
3185 Which is the equivalent of:
3188 label_1: branch label_3
3189 label_2: branch label_1
3190 label_3: branch label_4
3191 label_4: branch label_3
3194 Local symbol names are only a notational device. They are immediately
3195 transformed into more conventional symbol names before the assembler uses them.
3196 The symbol names stored in the symbol table, appearing in error messages and
3197 optionally emitted to the object file. The names are constructed using these
3202 All local labels begin with @samp{L}. Normally both @command{@value{AS}} and
3203 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
3204 used for symbols you are never intended to see. If you use the
3205 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3206 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3207 you may use them in debugging.
3210 This is the number that was used in the local label definition. So if the
3211 label is written @samp{55:} then the number is @samp{55}.
3214 This unusual character is included so you do not accidentally invent a symbol
3215 of the same name. The character has ASCII value of @samp{\002} (control-B).
3217 @item @emph{ordinal number}
3218 This is a serial number to keep the labels distinct. The first definition of
3219 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3220 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3221 the number @samp{1} and its 15th defintion gets @samp{15} as well.
3224 So for example, the first @code{1:} is named @code{L1@kbd{C-B}1}, the 44th
3225 @code{3:} is named @code{L3@kbd{C-B}44}.
3227 @subheading Dollar Local Labels
3228 @cindex dollar local symbols
3230 @code{@value{AS}} also supports an even more local form of local labels called
3231 dollar labels. These labels go out of scope (ie they become undefined) as soon
3232 as a non-local label is defined. Thus they remain valid for only a small
3233 region of the input source code. Normal local labels, by contrast, remain in
3234 scope for the entire file, or until they are redefined by another occurrence of
3235 the same local label.
3237 Dollar labels are defined in exactly the same way as ordinary local labels,
3238 except that instead of being terminated by a colon, they are terminated by a
3239 dollar sign. eg @samp{@b{55$}}.
3241 They can also be distinguished from ordinary local labels by their transformed
3242 name which uses ASCII character @samp{\001} (control-A) as the magic character
3243 to distinguish them from ordinary labels. Thus the 5th defintion of @samp{6$}
3244 is named @samp{L6@kbd{C-A}5}.
3247 @section The Special Dot Symbol
3249 @cindex dot (symbol)
3250 @cindex @code{.} (symbol)
3251 @cindex current address
3252 @cindex location counter
3253 The special symbol @samp{.} refers to the current address that
3254 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3255 .long .} defines @code{melvin} to contain its own address.
3256 Assigning a value to @code{.} is treated the same as a @code{.org}
3257 directive. Thus, the expression @samp{.=.+4} is the same as saying
3258 @ifclear no-space-dir
3262 @node Symbol Attributes
3263 @section Symbol Attributes
3265 @cindex symbol attributes
3266 @cindex attributes, symbol
3267 Every symbol has, as well as its name, the attributes ``Value'' and
3268 ``Type''. Depending on output format, symbols can also have auxiliary
3271 The detailed definitions are in @file{a.out.h}.
3274 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3275 all these attributes, and probably won't warn you. This makes the
3276 symbol an externally defined symbol, which is generally what you
3280 * Symbol Value:: Value
3281 * Symbol Type:: Type
3284 * a.out Symbols:: Symbol Attributes: @code{a.out}
3288 * a.out Symbols:: Symbol Attributes: @code{a.out}
3291 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3296 * COFF Symbols:: Symbol Attributes for COFF
3299 * SOM Symbols:: Symbol Attributes for SOM
3306 @cindex value of a symbol
3307 @cindex symbol value
3308 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3309 location in the text, data, bss or absolute sections the value is the
3310 number of addresses from the start of that section to the label.
3311 Naturally for text, data and bss sections the value of a symbol changes
3312 as @code{@value{LD}} changes section base addresses during linking. Absolute
3313 symbols' values do not change during linking: that is why they are
3316 The value of an undefined symbol is treated in a special way. If it is
3317 0 then the symbol is not defined in this assembler source file, and
3318 @code{@value{LD}} tries to determine its value from other files linked into the
3319 same program. You make this kind of symbol simply by mentioning a symbol
3320 name without defining it. A non-zero value represents a @code{.comm}
3321 common declaration. The value is how much common storage to reserve, in
3322 bytes (addresses). The symbol refers to the first address of the
3328 @cindex type of a symbol
3330 The type attribute of a symbol contains relocation (section)
3331 information, any flag settings indicating that a symbol is external, and
3332 (optionally), other information for linkers and debuggers. The exact
3333 format depends on the object-code output format in use.
3338 @c The following avoids a "widow" subsection title. @group would be
3339 @c better if it were available outside examples.
3342 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3344 @cindex @code{b.out} symbol attributes
3345 @cindex symbol attributes, @code{b.out}
3346 These symbol attributes appear only when @command{@value{AS}} is configured for
3347 one of the Berkeley-descended object output formats---@code{a.out} or
3353 @subsection Symbol Attributes: @code{a.out}
3355 @cindex @code{a.out} symbol attributes
3356 @cindex symbol attributes, @code{a.out}
3362 @subsection Symbol Attributes: @code{a.out}
3364 @cindex @code{a.out} symbol attributes
3365 @cindex symbol attributes, @code{a.out}
3369 * Symbol Desc:: Descriptor
3370 * Symbol Other:: Other
3374 @subsubsection Descriptor
3376 @cindex descriptor, of @code{a.out} symbol
3377 This is an arbitrary 16-bit value. You may establish a symbol's
3378 descriptor value by using a @code{.desc} statement
3379 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3380 @command{@value{AS}}.
3383 @subsubsection Other
3385 @cindex other attribute, of @code{a.out} symbol
3386 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3391 @subsection Symbol Attributes for COFF
3393 @cindex COFF symbol attributes
3394 @cindex symbol attributes, COFF
3396 The COFF format supports a multitude of auxiliary symbol attributes;
3397 like the primary symbol attributes, they are set between @code{.def} and
3398 @code{.endef} directives.
3400 @subsubsection Primary Attributes
3402 @cindex primary attributes, COFF symbols
3403 The symbol name is set with @code{.def}; the value and type,
3404 respectively, with @code{.val} and @code{.type}.
3406 @subsubsection Auxiliary Attributes
3408 @cindex auxiliary attributes, COFF symbols
3409 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3410 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3411 table information for COFF.
3416 @subsection Symbol Attributes for SOM
3418 @cindex SOM symbol attributes
3419 @cindex symbol attributes, SOM
3421 The SOM format for the HPPA supports a multitude of symbol attributes set with
3422 the @code{.EXPORT} and @code{.IMPORT} directives.
3424 The attributes are described in @cite{HP9000 Series 800 Assembly
3425 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3426 @code{EXPORT} assembler directive documentation.
3430 @chapter Expressions
3434 @cindex numeric values
3435 An @dfn{expression} specifies an address or numeric value.
3436 Whitespace may precede and/or follow an expression.
3438 The result of an expression must be an absolute number, or else an offset into
3439 a particular section. If an expression is not absolute, and there is not
3440 enough information when @command{@value{AS}} sees the expression to know its
3441 section, a second pass over the source program might be necessary to interpret
3442 the expression---but the second pass is currently not implemented.
3443 @command{@value{AS}} aborts with an error message in this situation.
3446 * Empty Exprs:: Empty Expressions
3447 * Integer Exprs:: Integer Expressions
3451 @section Empty Expressions
3453 @cindex empty expressions
3454 @cindex expressions, empty
3455 An empty expression has no value: it is just whitespace or null.
3456 Wherever an absolute expression is required, you may omit the
3457 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3458 is compatible with other assemblers.
3461 @section Integer Expressions
3463 @cindex integer expressions
3464 @cindex expressions, integer
3465 An @dfn{integer expression} is one or more @emph{arguments} delimited
3466 by @emph{operators}.
3469 * Arguments:: Arguments
3470 * Operators:: Operators
3471 * Prefix Ops:: Prefix Operators
3472 * Infix Ops:: Infix Operators
3476 @subsection Arguments
3478 @cindex expression arguments
3479 @cindex arguments in expressions
3480 @cindex operands in expressions
3481 @cindex arithmetic operands
3482 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3483 contexts arguments are sometimes called ``arithmetic operands''. In
3484 this manual, to avoid confusing them with the ``instruction operands'' of
3485 the machine language, we use the term ``argument'' to refer to parts of
3486 expressions only, reserving the word ``operand'' to refer only to machine
3487 instruction operands.
3489 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3490 @var{section} is one of text, data, bss, absolute,
3491 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3494 Numbers are usually integers.
3496 A number can be a flonum or bignum. In this case, you are warned
3497 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3498 these 32 bits are an integer. You may write integer-manipulating
3499 instructions that act on exotic constants, compatible with other
3502 @cindex subexpressions
3503 Subexpressions are a left parenthesis @samp{(} followed by an integer
3504 expression, followed by a right parenthesis @samp{)}; or a prefix
3505 operator followed by an argument.
3508 @subsection Operators
3510 @cindex operators, in expressions
3511 @cindex arithmetic functions
3512 @cindex functions, in expressions
3513 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3514 operators are followed by an argument. Infix operators appear
3515 between their arguments. Operators may be preceded and/or followed by
3519 @subsection Prefix Operator
3521 @cindex prefix operators
3522 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3523 one argument, which must be absolute.
3525 @c the tex/end tex stuff surrounding this small table is meant to make
3526 @c it align, on the printed page, with the similar table in the next
3527 @c section (which is inside an enumerate).
3529 \global\advance\leftskip by \itemindent
3534 @dfn{Negation}. Two's complement negation.
3536 @dfn{Complementation}. Bitwise not.
3540 \global\advance\leftskip by -\itemindent
3544 @subsection Infix Operators
3546 @cindex infix operators
3547 @cindex operators, permitted arguments
3548 @dfn{Infix operators} take two arguments, one on either side. Operators
3549 have precedence, but operations with equal precedence are performed left
3550 to right. Apart from @code{+} or @option{-}, both arguments must be
3551 absolute, and the result is absolute.
3554 @cindex operator precedence
3555 @cindex precedence of operators
3562 @dfn{Multiplication}.
3565 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3572 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3576 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3580 Intermediate precedence
3585 @dfn{Bitwise Inclusive Or}.
3591 @dfn{Bitwise Exclusive Or}.
3594 @dfn{Bitwise Or Not}.
3601 @cindex addition, permitted arguments
3602 @cindex plus, permitted arguments
3603 @cindex arguments for addition
3605 @dfn{Addition}. If either argument is absolute, the result has the section of
3606 the other argument. You may not add together arguments from different
3609 @cindex subtraction, permitted arguments
3610 @cindex minus, permitted arguments
3611 @cindex arguments for subtraction
3613 @dfn{Subtraction}. If the right argument is absolute, the
3614 result has the section of the left argument.
3615 If both arguments are in the same section, the result is absolute.
3616 You may not subtract arguments from different sections.
3617 @c FIXME is there still something useful to say about undefined - undefined ?
3619 @cindex comparison expressions
3620 @cindex expressions, comparison
3624 @dfn{Is Not Equal To}
3628 @dfn{Is Greater Than}
3630 @dfn{Is Greater Than Or Equal To}
3632 @dfn{Is Less Than Or Equal To}
3634 The comparison operators can be used as infix operators. A true results has a
3635 value of -1 whereas a false result has a value of 0. Note, these operators
3636 perform signed comparisons.
3639 @item Lowest Precedence
3648 These two logical operations can be used to combine the results of sub
3649 expressions. Note, unlike the comparison operators a true result returns a
3650 value of 1 but a false results does still return 0. Also note that the logical
3651 or operator has a slightly lower precedence than logical and.
3656 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3657 address; you can only have a defined section in one of the two arguments.
3660 @chapter Assembler Directives
3662 @cindex directives, machine independent
3663 @cindex pseudo-ops, machine independent
3664 @cindex machine independent directives
3665 All assembler directives have names that begin with a period (@samp{.}).
3666 The rest of the name is letters, usually in lower case.
3668 This chapter discusses directives that are available regardless of the
3669 target machine configuration for the @sc{gnu} assembler.
3671 Some machine configurations provide additional directives.
3672 @xref{Machine Dependencies}.
3675 @ifset machine-directives
3676 @xref{Machine Dependencies} for additional directives.
3681 * Abort:: @code{.abort}
3683 * ABORT:: @code{.ABORT}
3686 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3687 * Altmacro:: @code{.altmacro}
3688 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3689 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3690 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3691 * Byte:: @code{.byte @var{expressions}}
3692 * Comm:: @code{.comm @var{symbol} , @var{length} }
3694 * CFI directives:: @code{.cfi_startproc}, @code{.cfi_endproc}, etc.
3696 * Data:: @code{.data @var{subsection}}
3698 * Def:: @code{.def @var{name}}
3701 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3707 * Double:: @code{.double @var{flonums}}
3708 * Eject:: @code{.eject}
3709 * Else:: @code{.else}
3710 * Elseif:: @code{.elseif}
3713 * Endef:: @code{.endef}
3716 * Endfunc:: @code{.endfunc}
3717 * Endif:: @code{.endif}
3718 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3719 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3721 * Error:: @code{.error @var{string}}
3722 * Exitm:: @code{.exitm}
3723 * Extern:: @code{.extern}
3724 * Fail:: @code{.fail}
3725 @ifclear no-file-dir
3726 * File:: @code{.file @var{string}}
3729 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3730 * Float:: @code{.float @var{flonums}}
3731 * Func:: @code{.func}
3732 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3734 * Hidden:: @code{.hidden @var{names}}
3737 * hword:: @code{.hword @var{expressions}}
3738 * Ident:: @code{.ident}
3739 * If:: @code{.if @var{absolute expression}}
3740 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3741 * Include:: @code{.include "@var{file}"}
3742 * Int:: @code{.int @var{expressions}}
3744 * Internal:: @code{.internal @var{names}}
3747 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3748 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3749 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3750 * Lflags:: @code{.lflags}
3751 @ifclear no-line-dir
3752 * Line:: @code{.line @var{line-number}}
3755 * Linkonce:: @code{.linkonce [@var{type}]}
3756 * List:: @code{.list}
3757 * Ln:: @code{.ln @var{line-number}}
3759 * LNS directives:: @code{.file}, @code{.loc}, etc.
3761 * Long:: @code{.long @var{expressions}}
3763 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3766 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3767 * MRI:: @code{.mri @var{val}}
3768 * Noaltmacro:: @code{.noaltmacro}
3769 * Nolist:: @code{.nolist}
3770 * Octa:: @code{.octa @var{bignums}}
3771 * Org:: @code{.org @var{new-lc} , @var{fill}}
3772 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3774 * PopSection:: @code{.popsection}
3775 * Previous:: @code{.previous}
3778 * Print:: @code{.print @var{string}}
3780 * Protected:: @code{.protected @var{names}}
3783 * Psize:: @code{.psize @var{lines}, @var{columns}}
3784 * Purgem:: @code{.purgem @var{name}}
3786 * PushSection:: @code{.pushsection @var{name}}
3789 * Quad:: @code{.quad @var{bignums}}
3790 * Rept:: @code{.rept @var{count}}
3791 * Sbttl:: @code{.sbttl "@var{subheading}"}
3793 * Scl:: @code{.scl @var{class}}
3796 * Section:: @code{.section @var{name}}
3799 * Set:: @code{.set @var{symbol}, @var{expression}}
3800 * Short:: @code{.short @var{expressions}}
3801 * Single:: @code{.single @var{flonums}}
3803 * Size:: @code{.size [@var{name} , @var{expression}]}
3806 * Skip:: @code{.skip @var{size} , @var{fill}}
3807 * Sleb128:: @code{.sleb128 @var{expressions}}
3808 * Space:: @code{.space @var{size} , @var{fill}}
3810 * Stab:: @code{.stabd, .stabn, .stabs}
3813 * String:: @code{.string "@var{str}"}
3814 * Struct:: @code{.struct @var{expression}}
3816 * SubSection:: @code{.subsection}
3817 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3821 * Tag:: @code{.tag @var{structname}}
3824 * Text:: @code{.text @var{subsection}}
3825 * Title:: @code{.title "@var{heading}"}
3827 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3830 * Uleb128:: @code{.uleb128 @var{expressions}}
3832 * Val:: @code{.val @var{addr}}
3836 * Version:: @code{.version "@var{string}"}
3837 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3838 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3841 * Warning:: @code{.warning @var{string}}
3842 * Weak:: @code{.weak @var{names}}
3843 * Word:: @code{.word @var{expressions}}
3844 * Deprecated:: Deprecated Directives
3848 @section @code{.abort}
3850 @cindex @code{abort} directive
3851 @cindex stopping the assembly
3852 This directive stops the assembly immediately. It is for
3853 compatibility with other assemblers. The original idea was that the
3854 assembly language source would be piped into the assembler. If the sender
3855 of the source quit, it could use this directive tells @command{@value{AS}} to
3856 quit also. One day @code{.abort} will not be supported.
3860 @section @code{.ABORT}
3862 @cindex @code{ABORT} directive
3863 When producing COFF output, @command{@value{AS}} accepts this directive as a
3864 synonym for @samp{.abort}.
3867 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3873 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3875 @cindex padding the location counter
3876 @cindex @code{align} directive
3877 Pad the location counter (in the current subsection) to a particular storage
3878 boundary. The first expression (which must be absolute) is the alignment
3879 required, as described below.
3881 The second expression (also absolute) gives the fill value to be stored in the
3882 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3883 padding bytes are normally zero. However, on some systems, if the section is
3884 marked as containing code and the fill value is omitted, the space is filled
3885 with no-op instructions.
3887 The third expression is also absolute, and is also optional. If it is present,
3888 it is the maximum number of bytes that should be skipped by this alignment
3889 directive. If doing the alignment would require skipping more bytes than the
3890 specified maximum, then the alignment is not done at all. You can omit the
3891 fill value (the second argument) entirely by simply using two commas after the
3892 required alignment; this can be useful if you want the alignment to be filled
3893 with no-op instructions when appropriate.
3895 The way the required alignment is specified varies from system to system.
3896 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
3897 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
3898 alignment request in bytes. For example @samp{.align 8} advances
3899 the location counter until it is a multiple of 8. If the location counter
3900 is already a multiple of 8, no change is needed. For the tic54x, the
3901 first expression is the alignment request in words.
3903 For other systems, including the i386 using a.out format, and the arm and
3904 strongarm, it is the
3905 number of low-order zero bits the location counter must have after
3906 advancement. For example @samp{.align 3} advances the location
3907 counter until it a multiple of 8. If the location counter is already a
3908 multiple of 8, no change is needed.
3910 This inconsistency is due to the different behaviors of the various
3911 native assemblers for these systems which GAS must emulate.
3912 GAS also provides @code{.balign} and @code{.p2align} directives,
3913 described later, which have a consistent behavior across all
3914 architectures (but are specific to GAS).
3917 @section @code{.ascii "@var{string}"}@dots{}
3919 @cindex @code{ascii} directive
3920 @cindex string literals
3921 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3922 separated by commas. It assembles each string (with no automatic
3923 trailing zero byte) into consecutive addresses.
3926 @section @code{.asciz "@var{string}"}@dots{}
3928 @cindex @code{asciz} directive
3929 @cindex zero-terminated strings
3930 @cindex null-terminated strings
3931 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3932 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3935 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3937 @cindex padding the location counter given number of bytes
3938 @cindex @code{balign} directive
3939 Pad the location counter (in the current subsection) to a particular
3940 storage boundary. The first expression (which must be absolute) is the
3941 alignment request in bytes. For example @samp{.balign 8} advances
3942 the location counter until it is a multiple of 8. If the location counter
3943 is already a multiple of 8, no change is needed.
3945 The second expression (also absolute) gives the fill value to be stored in the
3946 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3947 padding bytes are normally zero. However, on some systems, if the section is
3948 marked as containing code and the fill value is omitted, the space is filled
3949 with no-op instructions.
3951 The third expression is also absolute, and is also optional. If it is present,
3952 it is the maximum number of bytes that should be skipped by this alignment
3953 directive. If doing the alignment would require skipping more bytes than the
3954 specified maximum, then the alignment is not done at all. You can omit the
3955 fill value (the second argument) entirely by simply using two commas after the
3956 required alignment; this can be useful if you want the alignment to be filled
3957 with no-op instructions when appropriate.
3959 @cindex @code{balignw} directive
3960 @cindex @code{balignl} directive
3961 The @code{.balignw} and @code{.balignl} directives are variants of the
3962 @code{.balign} directive. The @code{.balignw} directive treats the fill
3963 pattern as a two byte word value. The @code{.balignl} directives treats the
3964 fill pattern as a four byte longword value. For example, @code{.balignw
3965 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
3966 filled in with the value 0x368d (the exact placement of the bytes depends upon
3967 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
3971 @section @code{.byte @var{expressions}}
3973 @cindex @code{byte} directive
3974 @cindex integers, one byte
3975 @code{.byte} expects zero or more expressions, separated by commas.
3976 Each expression is assembled into the next byte.
3979 @section @code{.comm @var{symbol} , @var{length} }
3981 @cindex @code{comm} directive
3982 @cindex symbol, common
3983 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
3984 common symbol in one object file may be merged with a defined or common symbol
3985 of the same name in another object file. If @code{@value{LD}} does not see a
3986 definition for the symbol--just one or more common symbols--then it will
3987 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
3988 absolute expression. If @code{@value{LD}} sees multiple common symbols with
3989 the same name, and they do not all have the same size, it will allocate space
3990 using the largest size.
3993 When using ELF, the @code{.comm} directive takes an optional third argument.
3994 This is the desired alignment of the symbol, specified as a byte boundary (for
3995 example, an alignment of 16 means that the least significant 4 bits of the
3996 address should be zero). The alignment must be an absolute expression, and it
3997 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
3998 for the common symbol, it will use the alignment when placing the symbol. If
3999 no alignment is specified, @command{@value{AS}} will set the alignment to the
4000 largest power of two less than or equal to the size of the symbol, up to a
4005 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4006 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4009 @node CFI directives
4010 @section @code{.cfi_startproc}
4011 @cindex @code{cfi_startproc} directive
4012 @code{.cfi_startproc} is used at the beginning of each function that
4013 should have an entry in @code{.eh_frame}. It initializes some internal
4014 data structures and emits architecture dependent initial CFI instructions.
4015 Don't forget to close the function by
4016 @code{.cfi_endproc}.
4018 @section @code{.cfi_endproc}
4019 @cindex @code{cfi_endproc} directive
4020 @code{.cfi_endproc} is used at the end of a function where it closes its
4021 unwind entry previously opened by
4022 @code{.cfi_startproc}. and emits it to @code{.eh_frame}.
4024 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4025 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4026 address from @var{register} and add @var{offset} to it}.
4028 @section @code{.cfi_def_cfa_register @var{register}}
4029 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4030 now on @var{register} will be used instead of the old one. Offset
4033 @section @code{.cfi_def_cfa_offset @var{offset}}
4034 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4035 remains the same, but @var{offset} is new. Note that it is the
4036 absolute offset that will be added to a defined register to compute
4039 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4040 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4041 value that is added/substracted from the previous offset.
4043 @section @code{.cfi_offset @var{register}, @var{offset}}
4044 Previous value of @var{register} is saved at offset @var{offset} from
4047 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4048 Previous value of @var{register} is saved at offset @var{offset} from
4049 the current CFA register. This is transformed to @code{.cfi_offset}
4050 using the known displacement of the CFA register from the CFA.
4051 This is often easier to use, because the number will match the
4052 code it's annotating.
4054 @section @code{.cfi_window_save}
4055 SPARC register window has been saved.
4057 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4058 Allows the user to add arbitrary bytes to the unwind info. One
4059 might use this to add OS-specific CFI opcodes, or generic CFI
4060 opcodes that GAS does not yet support.
4062 @node LNS directives
4063 @section @code{.file @var{fileno} @var{filename}}
4064 @cindex @code{file} directive
4065 When emitting dwarf2 line number information @code{.file} assigns filenames
4066 to the @code{.debug_line} file name table. The @var{fileno} operand should
4067 be a unique positive integer to use as the index of the entry in the table.
4068 The @var{filename} operand is a C string literal.
4070 The detail of filename indicies is exposed to the user because the filename
4071 table is shared with the @code{.debug_info} section of the dwarf2 debugging
4072 information, and thus the user must know the exact indicies that table
4075 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
4076 @cindex @code{loc} directive
4077 The @code{.loc} directive will add row to the @code{.debug_line} line
4078 number matrix corresponding to the immediately following assembly
4079 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
4080 arguments will be applied to the @code{.debug_line} state machine before
4083 The @var{options} are a sequence of the following tokens in any order:
4087 This option will set the @code{basic_block} register in the
4088 @code{.debug_line} state machine to @code{true}.
4091 This option will set the @code{prologue_end} register in the
4092 @code{.debug_line} state machine to @code{true}.
4094 @item epilogue_begin
4095 This option will set the @code{epilogue_begin} register in the
4096 @code{.debug_line} state machine to @code{true}.
4098 @item is_stmt @var{value}
4099 This option will set the @code{is_stmt} register in the
4100 @code{.debug_line} state machine to @code{value}, which must be
4103 @item isa @var{value}
4104 This directive will set the @code{isa} register in the @code{.debug_line}
4105 state machine to @var{value}, which must be an unsigned integer.
4107 @section @code{.loc_mark_blocks @var{enable}}
4108 @cindex @code{loc_mark_blocks} directive
4109 The @code{.loc_mark_blocks} directive makes the assembler emit an entry
4110 to the @code{.debug_line} line number matrix with the @code{basic_block}
4111 register in the state machine set whenever a code label is seen.
4112 The @var{enable} argument should be either 1 or 0, to enable or disable
4113 this function respectively.
4118 @section @code{.data @var{subsection}}
4120 @cindex @code{data} directive
4121 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4122 end of the data subsection numbered @var{subsection} (which is an
4123 absolute expression). If @var{subsection} is omitted, it defaults
4128 @section @code{.def @var{name}}
4130 @cindex @code{def} directive
4131 @cindex COFF symbols, debugging
4132 @cindex debugging COFF symbols
4133 Begin defining debugging information for a symbol @var{name}; the
4134 definition extends until the @code{.endef} directive is encountered.
4137 This directive is only observed when @command{@value{AS}} is configured for COFF
4138 format output; when producing @code{b.out}, @samp{.def} is recognized,
4145 @section @code{.desc @var{symbol}, @var{abs-expression}}
4147 @cindex @code{desc} directive
4148 @cindex COFF symbol descriptor
4149 @cindex symbol descriptor, COFF
4150 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4151 to the low 16 bits of an absolute expression.
4154 The @samp{.desc} directive is not available when @command{@value{AS}} is
4155 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4156 object format. For the sake of compatibility, @command{@value{AS}} accepts
4157 it, but produces no output, when configured for COFF.
4163 @section @code{.dim}
4165 @cindex @code{dim} directive
4166 @cindex COFF auxiliary symbol information
4167 @cindex auxiliary symbol information, COFF
4168 This directive is generated by compilers to include auxiliary debugging
4169 information in the symbol table. It is only permitted inside
4170 @code{.def}/@code{.endef} pairs.
4173 @samp{.dim} is only meaningful when generating COFF format output; when
4174 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4180 @section @code{.double @var{flonums}}
4182 @cindex @code{double} directive
4183 @cindex floating point numbers (double)
4184 @code{.double} expects zero or more flonums, separated by commas. It
4185 assembles floating point numbers.
4187 The exact kind of floating point numbers emitted depends on how
4188 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4192 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4193 in @sc{ieee} format.
4198 @section @code{.eject}
4200 @cindex @code{eject} directive
4201 @cindex new page, in listings
4202 @cindex page, in listings
4203 @cindex listing control: new page
4204 Force a page break at this point, when generating assembly listings.
4207 @section @code{.else}
4209 @cindex @code{else} directive
4210 @code{.else} is part of the @command{@value{AS}} support for conditional
4211 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
4212 of code to be assembled if the condition for the preceding @code{.if}
4216 @section @code{.elseif}
4218 @cindex @code{elseif} directive
4219 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4220 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
4221 @code{.if} block that would otherwise fill the entire @code{.else} section.
4224 @section @code{.end}
4226 @cindex @code{end} directive
4227 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4228 process anything in the file past the @code{.end} directive.
4232 @section @code{.endef}
4234 @cindex @code{endef} directive
4235 This directive flags the end of a symbol definition begun with
4239 @samp{.endef} is only meaningful when generating COFF format output; if
4240 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4241 directive but ignores it.
4246 @section @code{.endfunc}
4247 @cindex @code{endfunc} directive
4248 @code{.endfunc} marks the end of a function specified with @code{.func}.
4251 @section @code{.endif}
4253 @cindex @code{endif} directive
4254 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4255 it marks the end of a block of code that is only assembled
4256 conditionally. @xref{If,,@code{.if}}.
4259 @section @code{.equ @var{symbol}, @var{expression}}
4261 @cindex @code{equ} directive
4262 @cindex assigning values to symbols
4263 @cindex symbols, assigning values to
4264 This directive sets the value of @var{symbol} to @var{expression}.
4265 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
4268 The syntax for @code{equ} on the HPPA is
4269 @samp{@var{symbol} .equ @var{expression}}.
4273 @section @code{.equiv @var{symbol}, @var{expression}}
4274 @cindex @code{equiv} directive
4275 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4276 the assembler will signal an error if @var{symbol} is already defined. Note a
4277 symbol which has been referenced but not actually defined is considered to be
4280 Except for the contents of the error message, this is roughly equivalent to
4289 @section @code{.err}
4290 @cindex @code{err} directive
4291 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4292 message and, unless the @option{-Z} option was used, it will not generate an
4293 object file. This can be used to signal error an conditionally compiled code.
4296 @section @code{.error "@var{string}"}
4297 @cindex error directive
4299 Similarly to @code{.err}, this directive emits an error, but you can specify a
4300 string that will be emitted as the error message. If you don't specify the
4301 message, it defaults to @code{".error directive invoked in source file"}.
4302 @xref{Errors, ,Error and Warning Messages}.
4305 .error "This code has not been assembled and tested."
4309 @section @code{.exitm}
4310 Exit early from the current macro definition. @xref{Macro}.
4313 @section @code{.extern}
4315 @cindex @code{extern} directive
4316 @code{.extern} is accepted in the source program---for compatibility
4317 with other assemblers---but it is ignored. @command{@value{AS}} treats
4318 all undefined symbols as external.
4321 @section @code{.fail @var{expression}}
4323 @cindex @code{fail} directive
4324 Generates an error or a warning. If the value of the @var{expression} is 500
4325 or more, @command{@value{AS}} will print a warning message. If the value is less
4326 than 500, @command{@value{AS}} will print an error message. The message will
4327 include the value of @var{expression}. This can occasionally be useful inside
4328 complex nested macros or conditional assembly.
4330 @ifclear no-file-dir
4332 @section @code{.file @var{string}}
4334 @cindex @code{file} directive
4335 @cindex logical file name
4336 @cindex file name, logical
4337 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4338 file. @var{string} is the new file name. In general, the filename is
4339 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4340 to specify an empty file name, you must give the quotes--@code{""}. This
4341 statement may go away in future: it is only recognized to be compatible with
4342 old @command{@value{AS}} programs.
4346 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4348 @cindex @code{fill} directive
4349 @cindex writing patterns in memory
4350 @cindex patterns, writing in memory
4351 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4352 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4353 may be zero or more. @var{Size} may be zero or more, but if it is
4354 more than 8, then it is deemed to have the value 8, compatible with
4355 other people's assemblers. The contents of each @var{repeat} bytes
4356 is taken from an 8-byte number. The highest order 4 bytes are
4357 zero. The lowest order 4 bytes are @var{value} rendered in the
4358 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4359 Each @var{size} bytes in a repetition is taken from the lowest order
4360 @var{size} bytes of this number. Again, this bizarre behavior is
4361 compatible with other people's assemblers.
4363 @var{size} and @var{value} are optional.
4364 If the second comma and @var{value} are absent, @var{value} is
4365 assumed zero. If the first comma and following tokens are absent,
4366 @var{size} is assumed to be 1.
4369 @section @code{.float @var{flonums}}
4371 @cindex floating point numbers (single)
4372 @cindex @code{float} directive
4373 This directive assembles zero or more flonums, separated by commas. It
4374 has the same effect as @code{.single}.
4376 The exact kind of floating point numbers emitted depends on how
4377 @command{@value{AS}} is configured.
4378 @xref{Machine Dependencies}.
4382 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4383 in @sc{ieee} format.
4388 @section @code{.func @var{name}[,@var{label}]}
4389 @cindex @code{func} directive
4390 @code{.func} emits debugging information to denote function @var{name}, and
4391 is ignored unless the file is assembled with debugging enabled.
4392 Only @samp{--gstabs[+]} is currently supported.
4393 @var{label} is the entry point of the function and if omitted @var{name}
4394 prepended with the @samp{leading char} is used.
4395 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4396 All functions are currently defined to have @code{void} return type.
4397 The function must be terminated with @code{.endfunc}.
4400 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4402 @cindex @code{global} directive
4403 @cindex symbol, making visible to linker
4404 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4405 @var{symbol} in your partial program, its value is made available to
4406 other partial programs that are linked with it. Otherwise,
4407 @var{symbol} takes its attributes from a symbol of the same name
4408 from another file linked into the same program.
4410 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4411 compatibility with other assemblers.
4414 On the HPPA, @code{.global} is not always enough to make it accessible to other
4415 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4416 @xref{HPPA Directives,, HPPA Assembler Directives}.
4421 @section @code{.hidden @var{names}}
4423 @cindex @code{hidden} directive
4425 This is one of the ELF visibility directives. The other two are
4426 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4427 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4429 This directive overrides the named symbols default visibility (which is set by
4430 their binding: local, global or weak). The directive sets the visibility to
4431 @code{hidden} which means that the symbols are not visible to other components.
4432 Such symbols are always considered to be @code{protected} as well.
4436 @section @code{.hword @var{expressions}}
4438 @cindex @code{hword} directive
4439 @cindex integers, 16-bit
4440 @cindex numbers, 16-bit
4441 @cindex sixteen bit integers
4442 This expects zero or more @var{expressions}, and emits
4443 a 16 bit number for each.
4446 This directive is a synonym for @samp{.short}; depending on the target
4447 architecture, it may also be a synonym for @samp{.word}.
4451 This directive is a synonym for @samp{.short}.
4454 This directive is a synonym for both @samp{.short} and @samp{.word}.
4459 @section @code{.ident}
4461 @cindex @code{ident} directive
4462 This directive is used by some assemblers to place tags in object files.
4463 @command{@value{AS}} simply accepts the directive for source-file
4464 compatibility with such assemblers, but does not actually emit anything
4468 @section @code{.if @var{absolute expression}}
4470 @cindex conditional assembly
4471 @cindex @code{if} directive
4472 @code{.if} marks the beginning of a section of code which is only
4473 considered part of the source program being assembled if the argument
4474 (which must be an @var{absolute expression}) is non-zero. The end of
4475 the conditional section of code must be marked by @code{.endif}
4476 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4477 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4478 If you have several conditions to check, @code{.elseif} may be used to avoid
4479 nesting blocks if/else within each subsequent @code{.else} block.
4481 The following variants of @code{.if} are also supported:
4483 @cindex @code{ifdef} directive
4484 @item .ifdef @var{symbol}
4485 Assembles the following section of code if the specified @var{symbol}
4486 has been defined. Note a symbol which has been referenced but not yet defined
4487 is considered to be undefined.
4489 @cindex @code{ifb} directive
4490 @item .ifb @var{text}
4491 Assembles the following section of code if the operand is blank (empty).
4493 @cindex @code{ifc} directive
4494 @item .ifc @var{string1},@var{string2}
4495 Assembles the following section of code if the two strings are the same. The
4496 strings may be optionally quoted with single quotes. If they are not quoted,
4497 the first string stops at the first comma, and the second string stops at the
4498 end of the line. Strings which contain whitespace should be quoted. The
4499 string comparison is case sensitive.
4501 @cindex @code{ifeq} directive
4502 @item .ifeq @var{absolute expression}
4503 Assembles the following section of code if the argument is zero.
4505 @cindex @code{ifeqs} directive
4506 @item .ifeqs @var{string1},@var{string2}
4507 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4509 @cindex @code{ifge} directive
4510 @item .ifge @var{absolute expression}
4511 Assembles the following section of code if the argument is greater than or
4514 @cindex @code{ifgt} directive
4515 @item .ifgt @var{absolute expression}
4516 Assembles the following section of code if the argument is greater than zero.
4518 @cindex @code{ifle} directive
4519 @item .ifle @var{absolute expression}
4520 Assembles the following section of code if the argument is less than or equal
4523 @cindex @code{iflt} directive
4524 @item .iflt @var{absolute expression}
4525 Assembles the following section of code if the argument is less than zero.
4527 @cindex @code{ifnb} directive
4528 @item .ifnb @var{text}
4529 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
4530 following section of code if the operand is non-blank (non-empty).
4532 @cindex @code{ifnc} directive
4533 @item .ifnc @var{string1},@var{string2}.
4534 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4535 following section of code if the two strings are not the same.
4537 @cindex @code{ifndef} directive
4538 @cindex @code{ifnotdef} directive
4539 @item .ifndef @var{symbol}
4540 @itemx .ifnotdef @var{symbol}
4541 Assembles the following section of code if the specified @var{symbol}
4542 has not been defined. Both spelling variants are equivalent. Note a symbol
4543 which has been referenced but not yet defined is considered to be undefined.
4545 @cindex @code{ifne} directive
4546 @item .ifne @var{absolute expression}
4547 Assembles the following section of code if the argument is not equal to zero
4548 (in other words, this is equivalent to @code{.if}).
4550 @cindex @code{ifnes} directive
4551 @item .ifnes @var{string1},@var{string2}
4552 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4553 following section of code if the two strings are not the same.
4557 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4559 @cindex @code{incbin} directive
4560 @cindex binary files, including
4561 The @code{incbin} directive includes @var{file} verbatim at the current
4562 location. You can control the search paths used with the @samp{-I} command-line
4563 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4566 The @var{skip} argument skips a number of bytes from the start of the
4567 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4568 read. Note that the data is not aligned in any way, so it is the user's
4569 responsibility to make sure that proper alignment is provided both before and
4570 after the @code{incbin} directive.
4573 @section @code{.include "@var{file}"}
4575 @cindex @code{include} directive
4576 @cindex supporting files, including
4577 @cindex files, including
4578 This directive provides a way to include supporting files at specified
4579 points in your source program. The code from @var{file} is assembled as
4580 if it followed the point of the @code{.include}; when the end of the
4581 included file is reached, assembly of the original file continues. You
4582 can control the search paths used with the @samp{-I} command-line option
4583 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4587 @section @code{.int @var{expressions}}
4589 @cindex @code{int} directive
4590 @cindex integers, 32-bit
4591 Expect zero or more @var{expressions}, of any section, separated by commas.
4592 For each expression, emit a number that, at run time, is the value of that
4593 expression. The byte order and bit size of the number depends on what kind
4594 of target the assembly is for.
4598 On most forms of the H8/300, @code{.int} emits 16-bit
4599 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4606 @section @code{.internal @var{names}}
4608 @cindex @code{internal} directive
4610 This is one of the ELF visibility directives. The other two are
4611 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4612 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4614 This directive overrides the named symbols default visibility (which is set by
4615 their binding: local, global or weak). The directive sets the visibility to
4616 @code{internal} which means that the symbols are considered to be @code{hidden}
4617 (i.e., not visible to other components), and that some extra, processor specific
4618 processing must also be performed upon the symbols as well.
4622 @section @code{.irp @var{symbol},@var{values}}@dots{}
4624 @cindex @code{irp} directive
4625 Evaluate a sequence of statements assigning different values to @var{symbol}.
4626 The sequence of statements starts at the @code{.irp} directive, and is
4627 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4628 set to @var{value}, and the sequence of statements is assembled. If no
4629 @var{value} is listed, the sequence of statements is assembled once, with
4630 @var{symbol} set to the null string. To refer to @var{symbol} within the
4631 sequence of statements, use @var{\symbol}.
4633 For example, assembling
4641 is equivalent to assembling
4649 For some caveats with the spelling of @var{symbol}, see also the discussion
4653 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4655 @cindex @code{irpc} directive
4656 Evaluate a sequence of statements assigning different values to @var{symbol}.
4657 The sequence of statements starts at the @code{.irpc} directive, and is
4658 terminated by an @code{.endr} directive. For each character in @var{value},
4659 @var{symbol} is set to the character, and the sequence of statements is
4660 assembled. If no @var{value} is listed, the sequence of statements is
4661 assembled once, with @var{symbol} set to the null string. To refer to
4662 @var{symbol} within the sequence of statements, use @var{\symbol}.
4664 For example, assembling
4672 is equivalent to assembling
4680 For some caveats with the spelling of @var{symbol}, see also the discussion
4684 @section @code{.lcomm @var{symbol} , @var{length}}
4686 @cindex @code{lcomm} directive
4687 @cindex local common symbols
4688 @cindex symbols, local common
4689 Reserve @var{length} (an absolute expression) bytes for a local common
4690 denoted by @var{symbol}. The section and value of @var{symbol} are
4691 those of the new local common. The addresses are allocated in the bss
4692 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4693 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4694 not visible to @code{@value{LD}}.
4697 Some targets permit a third argument to be used with @code{.lcomm}. This
4698 argument specifies the desired alignment of the symbol in the bss section.
4702 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4703 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4707 @section @code{.lflags}
4709 @cindex @code{lflags} directive (ignored)
4710 @command{@value{AS}} accepts this directive, for compatibility with other
4711 assemblers, but ignores it.
4713 @ifclear no-line-dir
4715 @section @code{.line @var{line-number}}
4717 @cindex @code{line} directive
4721 @section @code{.ln @var{line-number}}
4723 @cindex @code{ln} directive
4725 @cindex logical line number
4727 Change the logical line number. @var{line-number} must be an absolute
4728 expression. The next line has that logical line number. Therefore any other
4729 statements on the current line (after a statement separator character) are
4730 reported as on logical line number @var{line-number} @minus{} 1. One day
4731 @command{@value{AS}} will no longer support this directive: it is recognized only
4732 for compatibility with existing assembler programs.
4736 @ifclear no-line-dir
4737 Even though this is a directive associated with the @code{a.out} or
4738 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4739 when producing COFF output, and treats @samp{.line} as though it
4740 were the COFF @samp{.ln} @emph{if} it is found outside a
4741 @code{.def}/@code{.endef} pair.
4743 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4744 used by compilers to generate auxiliary symbol information for
4749 @section @code{.linkonce [@var{type}]}
4751 @cindex @code{linkonce} directive
4752 @cindex common sections
4753 Mark the current section so that the linker only includes a single copy of it.
4754 This may be used to include the same section in several different object files,
4755 but ensure that the linker will only include it once in the final output file.
4756 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4757 Duplicate sections are detected based on the section name, so it should be
4760 This directive is only supported by a few object file formats; as of this
4761 writing, the only object file format which supports it is the Portable
4762 Executable format used on Windows NT.
4764 The @var{type} argument is optional. If specified, it must be one of the
4765 following strings. For example:
4769 Not all types may be supported on all object file formats.
4773 Silently discard duplicate sections. This is the default.
4776 Warn if there are duplicate sections, but still keep only one copy.
4779 Warn if any of the duplicates have different sizes.
4782 Warn if any of the duplicates do not have exactly the same contents.
4786 @section @code{.ln @var{line-number}}
4788 @cindex @code{ln} directive
4789 @ifclear no-line-dir
4790 @samp{.ln} is a synonym for @samp{.line}.
4793 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4794 must be an absolute expression. The next line has that logical
4795 line number, so any other statements on the current line (after a
4796 statement separator character @code{;}) are reported as on logical
4797 line number @var{line-number} @minus{} 1.
4800 This directive is accepted, but ignored, when @command{@value{AS}} is
4801 configured for @code{b.out}; its effect is only associated with COFF
4807 @section @code{.mri @var{val}}
4809 @cindex @code{mri} directive
4810 @cindex MRI mode, temporarily
4811 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4812 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4813 affects code assembled until the next @code{.mri} directive, or until the end
4814 of the file. @xref{M, MRI mode, MRI mode}.
4817 @section @code{.list}
4819 @cindex @code{list} directive
4820 @cindex listing control, turning on
4821 Control (in conjunction with the @code{.nolist} directive) whether or
4822 not assembly listings are generated. These two directives maintain an
4823 internal counter (which is zero initially). @code{.list} increments the
4824 counter, and @code{.nolist} decrements it. Assembly listings are
4825 generated whenever the counter is greater than zero.
4827 By default, listings are disabled. When you enable them (with the
4828 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4829 the initial value of the listing counter is one.
4832 @section @code{.long @var{expressions}}
4834 @cindex @code{long} directive
4835 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4838 @c no one seems to know what this is for or whether this description is
4839 @c what it really ought to do
4841 @section @code{.lsym @var{symbol}, @var{expression}}
4843 @cindex @code{lsym} directive
4844 @cindex symbol, not referenced in assembly
4845 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4846 the hash table, ensuring it cannot be referenced by name during the
4847 rest of the assembly. This sets the attributes of the symbol to be
4848 the same as the expression value:
4850 @var{other} = @var{descriptor} = 0
4851 @var{type} = @r{(section of @var{expression})}
4852 @var{value} = @var{expression}
4855 The new symbol is not flagged as external.
4859 @section @code{.macro}
4862 The commands @code{.macro} and @code{.endm} allow you to define macros that
4863 generate assembly output. For example, this definition specifies a macro
4864 @code{sum} that puts a sequence of numbers into memory:
4867 .macro sum from=0, to=5
4876 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4888 @item .macro @var{macname}
4889 @itemx .macro @var{macname} @var{macargs} @dots{}
4890 @cindex @code{macro} directive
4891 Begin the definition of a macro called @var{macname}. If your macro
4892 definition requires arguments, specify their names after the macro name,
4893 separated by commas or spaces. You can qualify the macro argument to
4894 indicate whether all invocations must specify a non-blank value (through
4895 @samp{:@code{req}}), or whether it takes all of the remaining arguments
4896 (through @samp{:@code{vararg}}). You can supply a default value for any
4897 macro argument by following the name with @samp{=@var{deflt}}. You
4898 cannot define two macros with the same @var{macname} unless it has been
4899 subject to the @code{.purgem} directive (@xref{Purgem}.) between the two
4900 definitions. For example, these are all valid @code{.macro} statements:
4904 Begin the definition of a macro called @code{comm}, which takes no
4907 @item .macro plus1 p, p1
4908 @itemx .macro plus1 p p1
4909 Either statement begins the definition of a macro called @code{plus1},
4910 which takes two arguments; within the macro definition, write
4911 @samp{\p} or @samp{\p1} to evaluate the arguments.
4913 @item .macro reserve_str p1=0 p2
4914 Begin the definition of a macro called @code{reserve_str}, with two
4915 arguments. The first argument has a default value, but not the second.
4916 After the definition is complete, you can call the macro either as
4917 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
4918 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
4919 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
4920 @samp{0}, and @samp{\p2} evaluating to @var{b}).
4923 @item .macro m p1:req, p2=0, p3:vararg
4924 Begin the definition of a macro called @code{m}, with at least three
4925 arguments. The first argument must always have a value specified, but
4926 not the second, which instead has a default value. The third formal
4927 will get assigned all remaining arguments specified at invocation time.
4929 When you call a macro, you can specify the argument values either by
4930 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
4931 @samp{sum to=17, from=9}.
4933 Note that since each of the @var{macargs} can be an identifier exactly
4934 as any other one permitted by the target architecture, there may be
4935 occasional problems if the target hand-crafts special meanings to certain
4936 characters when they occur in a special position. For example, if colon
4937 (@code{:}) is generally permitted to be part of a symbol name, but the
4938 architecture specific code special-cases it when occuring as the final
4939 character of a symbol (to denote a label), then the macro parameter
4940 replacement code will have no way of knowing that and consider the whole
4941 construct (including the colon) an identifier, and check only this
4942 identifier for being the subject to parameter substitution. In this
4943 example, besides the potential of just separating identifier and colon
4944 by white space, using alternate macro syntax (@xref{Altmacro}.) and
4945 ampersand (@code{&}) as the character to separate literal text from macro
4946 parameters (or macro parameters from one another) would provide a way to
4947 achieve the same effect:
4956 This applies identically to the identifiers used in @code{.irp} (@xref{Irp}.)
4957 and @code{.irpc} (@xref{Irpc}.).
4960 @cindex @code{endm} directive
4961 Mark the end of a macro definition.
4964 @cindex @code{exitm} directive
4965 Exit early from the current macro definition.
4967 @cindex number of macros executed
4968 @cindex macros, count executed
4970 @command{@value{AS}} maintains a counter of how many macros it has
4971 executed in this pseudo-variable; you can copy that number to your
4972 output with @samp{\@@}, but @emph{only within a macro definition}.
4974 @item LOCAL @var{name} [ , @dots{} ]
4975 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
4976 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
4977 @xref{Altmacro,,@code{.altmacro}}.
4981 @section @code{.altmacro}
4982 Enable alternate macro mode, enabling:
4985 @item LOCAL @var{name} [ , @dots{} ]
4986 One additional directive, @code{LOCAL}, is available. It is used to
4987 generate a string replacement for each of the @var{name} arguments, and
4988 replace any instances of @var{name} in each macro expansion. The
4989 replacement string is unique in the assembly, and different for each
4990 separate macro expansion. @code{LOCAL} allows you to write macros that
4991 define symbols, without fear of conflict between separate macro expansions.
4993 @item String delimiters
4994 You can write strings delimited in these other ways besides
4995 @code{"@var{string}"}:
4998 @item '@var{string}'
4999 You can delimit strings with single-quote charaters.
5001 @item <@var{string}>
5002 You can delimit strings with matching angle brackets.
5005 @item single-character string escape
5006 To include any single character literally in a string (even if the
5007 character would otherwise have some special meaning), you can prefix the
5008 character with @samp{!} (an exclamation mark). For example, you can
5009 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
5011 @item Expression results as strings
5012 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
5013 and use the result as a string.
5017 @section @code{.noaltmacro}
5018 Disable alternate macro mode. @ref{Altmacro}
5021 @section @code{.nolist}
5023 @cindex @code{nolist} directive
5024 @cindex listing control, turning off
5025 Control (in conjunction with the @code{.list} directive) whether or
5026 not assembly listings are generated. These two directives maintain an
5027 internal counter (which is zero initially). @code{.list} increments the
5028 counter, and @code{.nolist} decrements it. Assembly listings are
5029 generated whenever the counter is greater than zero.
5032 @section @code{.octa @var{bignums}}
5034 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5035 @cindex @code{octa} directive
5036 @cindex integer, 16-byte
5037 @cindex sixteen byte integer
5038 This directive expects zero or more bignums, separated by commas. For each
5039 bignum, it emits a 16-byte integer.
5041 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5042 hence @emph{octa}-word for 16 bytes.
5045 @section @code{.org @var{new-lc} , @var{fill}}
5047 @cindex @code{org} directive
5048 @cindex location counter, advancing
5049 @cindex advancing location counter
5050 @cindex current address, advancing
5051 Advance the location counter of the current section to
5052 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5053 expression with the same section as the current subsection. That is,
5054 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5055 wrong section, the @code{.org} directive is ignored. To be compatible
5056 with former assemblers, if the section of @var{new-lc} is absolute,
5057 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5058 is the same as the current subsection.
5060 @code{.org} may only increase the location counter, or leave it
5061 unchanged; you cannot use @code{.org} to move the location counter
5064 @c double negative used below "not undefined" because this is a specific
5065 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5066 @c section. doc@cygnus.com 18feb91
5067 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5068 may not be undefined. If you really detest this restriction we eagerly await
5069 a chance to share your improved assembler.
5071 Beware that the origin is relative to the start of the section, not
5072 to the start of the subsection. This is compatible with other
5073 people's assemblers.
5075 When the location counter (of the current subsection) is advanced, the
5076 intervening bytes are filled with @var{fill} which should be an
5077 absolute expression. If the comma and @var{fill} are omitted,
5078 @var{fill} defaults to zero.
5081 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5083 @cindex padding the location counter given a power of two
5084 @cindex @code{p2align} directive
5085 Pad the location counter (in the current subsection) to a particular
5086 storage boundary. The first expression (which must be absolute) is the
5087 number of low-order zero bits the location counter must have after
5088 advancement. For example @samp{.p2align 3} advances the location
5089 counter until it a multiple of 8. If the location counter is already a
5090 multiple of 8, no change is needed.
5092 The second expression (also absolute) gives the fill value to be stored in the
5093 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5094 padding bytes are normally zero. However, on some systems, if the section is
5095 marked as containing code and the fill value is omitted, the space is filled
5096 with no-op instructions.
5098 The third expression is also absolute, and is also optional. If it is present,
5099 it is the maximum number of bytes that should be skipped by this alignment
5100 directive. If doing the alignment would require skipping more bytes than the
5101 specified maximum, then the alignment is not done at all. You can omit the
5102 fill value (the second argument) entirely by simply using two commas after the
5103 required alignment; this can be useful if you want the alignment to be filled
5104 with no-op instructions when appropriate.
5106 @cindex @code{p2alignw} directive
5107 @cindex @code{p2alignl} directive
5108 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5109 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5110 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5111 fill pattern as a four byte longword value. For example, @code{.p2alignw
5112 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5113 filled in with the value 0x368d (the exact placement of the bytes depends upon
5114 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5119 @section @code{.previous}
5121 @cindex @code{previous} directive
5122 @cindex Section Stack
5123 This is one of the ELF section stack manipulation directives. The others are
5124 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5125 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5126 (@pxref{PopSection}).
5128 This directive swaps the current section (and subsection) with most recently
5129 referenced section (and subsection) prior to this one. Multiple
5130 @code{.previous} directives in a row will flip between two sections (and their
5133 In terms of the section stack, this directive swaps the current section with
5134 the top section on the section stack.
5139 @section @code{.popsection}
5141 @cindex @code{popsection} directive
5142 @cindex Section Stack
5143 This is one of the ELF section stack manipulation directives. The others are
5144 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5145 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5148 This directive replaces the current section (and subsection) with the top
5149 section (and subsection) on the section stack. This section is popped off the
5154 @section @code{.print @var{string}}
5156 @cindex @code{print} directive
5157 @command{@value{AS}} will print @var{string} on the standard output during
5158 assembly. You must put @var{string} in double quotes.
5162 @section @code{.protected @var{names}}
5164 @cindex @code{protected} directive
5166 This is one of the ELF visibility directives. The other two are
5167 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5169 This directive overrides the named symbols default visibility (which is set by
5170 their binding: local, global or weak). The directive sets the visibility to
5171 @code{protected} which means that any references to the symbols from within the
5172 components that defines them must be resolved to the definition in that
5173 component, even if a definition in another component would normally preempt
5178 @section @code{.psize @var{lines} , @var{columns}}
5180 @cindex @code{psize} directive
5181 @cindex listing control: paper size
5182 @cindex paper size, for listings
5183 Use this directive to declare the number of lines---and, optionally, the
5184 number of columns---to use for each page, when generating listings.
5186 If you do not use @code{.psize}, listings use a default line-count
5187 of 60. You may omit the comma and @var{columns} specification; the
5188 default width is 200 columns.
5190 @command{@value{AS}} generates formfeeds whenever the specified number of
5191 lines is exceeded (or whenever you explicitly request one, using
5194 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5195 those explicitly specified with @code{.eject}.
5198 @section @code{.purgem @var{name}}
5200 @cindex @code{purgem} directive
5201 Undefine the macro @var{name}, so that later uses of the string will not be
5202 expanded. @xref{Macro}.
5206 @section @code{.pushsection @var{name} , @var{subsection}}
5208 @cindex @code{pushsection} directive
5209 @cindex Section Stack
5210 This is one of the ELF section stack manipulation directives. The others are
5211 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5212 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5215 This directive pushes the current section (and subsection) onto the
5216 top of the section stack, and then replaces the current section and
5217 subsection with @code{name} and @code{subsection}.
5221 @section @code{.quad @var{bignums}}
5223 @cindex @code{quad} directive
5224 @code{.quad} expects zero or more bignums, separated by commas. For
5225 each bignum, it emits
5227 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5228 warning message; and just takes the lowest order 8 bytes of the bignum.
5229 @cindex eight-byte integer
5230 @cindex integer, 8-byte
5232 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5233 hence @emph{quad}-word for 8 bytes.
5236 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5237 warning message; and just takes the lowest order 16 bytes of the bignum.
5238 @cindex sixteen-byte integer
5239 @cindex integer, 16-byte
5243 @section @code{.rept @var{count}}
5245 @cindex @code{rept} directive
5246 Repeat the sequence of lines between the @code{.rept} directive and the next
5247 @code{.endr} directive @var{count} times.
5249 For example, assembling
5257 is equivalent to assembling
5266 @section @code{.sbttl "@var{subheading}"}
5268 @cindex @code{sbttl} directive
5269 @cindex subtitles for listings
5270 @cindex listing control: subtitle
5271 Use @var{subheading} as the title (third line, immediately after the
5272 title line) when generating assembly listings.
5274 This directive affects subsequent pages, as well as the current page if
5275 it appears within ten lines of the top of a page.
5279 @section @code{.scl @var{class}}
5281 @cindex @code{scl} directive
5282 @cindex symbol storage class (COFF)
5283 @cindex COFF symbol storage class
5284 Set the storage-class value for a symbol. This directive may only be
5285 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5286 whether a symbol is static or external, or it may record further
5287 symbolic debugging information.
5290 The @samp{.scl} directive is primarily associated with COFF output; when
5291 configured to generate @code{b.out} output format, @command{@value{AS}}
5292 accepts this directive but ignores it.
5298 @section @code{.section @var{name}}
5300 @cindex named section
5301 Use the @code{.section} directive to assemble the following code into a section
5304 This directive is only supported for targets that actually support arbitrarily
5305 named sections; on @code{a.out} targets, for example, it is not accepted, even
5306 with a standard @code{a.out} section name.
5310 @c only print the extra heading if both COFF and ELF are set
5311 @subheading COFF Version
5314 @cindex @code{section} directive (COFF version)
5315 For COFF targets, the @code{.section} directive is used in one of the following
5319 .section @var{name}[, "@var{flags}"]
5320 .section @var{name}[, @var{subsegment}]
5323 If the optional argument is quoted, it is taken as flags to use for the
5324 section. Each flag is a single character. The following flags are recognized:
5327 bss section (uninitialized data)
5329 section is not loaded
5339 shared section (meaningful for PE targets)
5341 ignored. (For compatibility with the ELF version)
5344 If no flags are specified, the default flags depend upon the section name. If
5345 the section name is not recognized, the default will be for the section to be
5346 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5347 from the section, rather than adding them, so if they are used on their own it
5348 will be as if no flags had been specified at all.
5350 If the optional argument to the @code{.section} directive is not quoted, it is
5351 taken as a subsegment number (@pxref{Sub-Sections}).
5356 @c only print the extra heading if both COFF and ELF are set
5357 @subheading ELF Version
5360 @cindex Section Stack
5361 This is one of the ELF section stack manipulation directives. The others are
5362 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5363 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5364 @code{.previous} (@pxref{Previous}).
5366 @cindex @code{section} directive (ELF version)
5367 For ELF targets, the @code{.section} directive is used like this:
5370 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]
5373 The optional @var{flags} argument is a quoted string which may contain any
5374 combination of the following characters:
5377 section is allocatable
5381 section is executable
5383 section is mergeable
5385 section contains zero terminated strings
5387 section is a member of a section group
5389 section is used for thread-local-storage
5392 The optional @var{type} argument may contain one of the following constants:
5395 section contains data
5397 section does not contain data (i.e., section only occupies space)
5399 section contains data which is used by things other than the program
5401 section contains an array of pointers to init functions
5403 section contains an array of pointers to finish functions
5404 @item @@preinit_array
5405 section contains an array of pointers to pre-init functions
5408 Many targets only support the first three section types.
5410 Note on targets where the @code{@@} character is the start of a comment (eg
5411 ARM) then another character is used instead. For example the ARM port uses the
5414 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
5415 be specified as well as an extra argument - @var{entsize} - like this:
5418 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
5421 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
5422 constants, each @var{entsize} octets long. Sections with both @code{M} and
5423 @code{S} must contain zero terminated strings where each character is
5424 @var{entsize} bytes long. The linker may remove duplicates within sections with
5425 the same name, same entity size and same flags. @var{entsize} must be an
5426 absolute expression.
5428 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
5429 be present along with an additional field like this:
5432 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
5435 The @var{GroupName} field specifies the name of the section group to which this
5436 particular section belongs. The optional linkage field can contain:
5439 indicates that only one copy of this section should be retained
5444 Note - if both the @var{M} and @var{G} flags are present then the fields for
5445 the Merge flag should come first, like this:
5448 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
5451 If no flags are specified, the default flags depend upon the section name. If
5452 the section name is not recognized, the default will be for the section to have
5453 none of the above flags: it will not be allocated in memory, nor writable, nor
5454 executable. The section will contain data.
5456 For ELF targets, the assembler supports another type of @code{.section}
5457 directive for compatibility with the Solaris assembler:
5460 .section "@var{name}"[, @var{flags}...]
5463 Note that the section name is quoted. There may be a sequence of comma
5467 section is allocatable
5471 section is executable
5473 section is used for thread local storage
5476 This directive replaces the current section and subsection. See the
5477 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
5478 some examples of how this directive and the other section stack directives
5484 @section @code{.set @var{symbol}, @var{expression}}
5486 @cindex @code{set} directive
5487 @cindex symbol value, setting
5488 Set the value of @var{symbol} to @var{expression}. This
5489 changes @var{symbol}'s value and type to conform to
5490 @var{expression}. If @var{symbol} was flagged as external, it remains
5491 flagged (@pxref{Symbol Attributes}).
5493 You may @code{.set} a symbol many times in the same assembly.
5495 If you @code{.set} a global symbol, the value stored in the object
5496 file is the last value stored into it.
5499 The syntax for @code{set} on the HPPA is
5500 @samp{@var{symbol} .set @var{expression}}.
5504 @section @code{.short @var{expressions}}
5506 @cindex @code{short} directive
5508 @code{.short} is normally the same as @samp{.word}.
5509 @xref{Word,,@code{.word}}.
5511 In some configurations, however, @code{.short} and @code{.word} generate
5512 numbers of different lengths; @pxref{Machine Dependencies}.
5516 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5519 This expects zero or more @var{expressions}, and emits
5520 a 16 bit number for each.
5525 @section @code{.single @var{flonums}}
5527 @cindex @code{single} directive
5528 @cindex floating point numbers (single)
5529 This directive assembles zero or more flonums, separated by commas. It
5530 has the same effect as @code{.float}.
5532 The exact kind of floating point numbers emitted depends on how
5533 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5537 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5538 numbers in @sc{ieee} format.
5544 @section @code{.size}
5546 This directive is used to set the size associated with a symbol.
5550 @c only print the extra heading if both COFF and ELF are set
5551 @subheading COFF Version
5554 @cindex @code{size} directive (COFF version)
5555 For COFF targets, the @code{.size} directive is only permitted inside
5556 @code{.def}/@code{.endef} pairs. It is used like this:
5559 .size @var{expression}
5563 @samp{.size} is only meaningful when generating COFF format output; when
5564 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5571 @c only print the extra heading if both COFF and ELF are set
5572 @subheading ELF Version
5575 @cindex @code{size} directive (ELF version)
5576 For ELF targets, the @code{.size} directive is used like this:
5579 .size @var{name} , @var{expression}
5582 This directive sets the size associated with a symbol @var{name}.
5583 The size in bytes is computed from @var{expression} which can make use of label
5584 arithmetic. This directive is typically used to set the size of function
5590 @section @code{.sleb128 @var{expressions}}
5592 @cindex @code{sleb128} directive
5593 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5594 compact, variable length representation of numbers used by the DWARF
5595 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
5597 @ifclear no-space-dir
5599 @section @code{.skip @var{size} , @var{fill}}
5601 @cindex @code{skip} directive
5602 @cindex filling memory
5603 This directive emits @var{size} bytes, each of value @var{fill}. Both
5604 @var{size} and @var{fill} are absolute expressions. If the comma and
5605 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5609 @section @code{.space @var{size} , @var{fill}}
5611 @cindex @code{space} directive
5612 @cindex filling memory
5613 This directive emits @var{size} bytes, each of value @var{fill}. Both
5614 @var{size} and @var{fill} are absolute expressions. If the comma
5615 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5620 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5621 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5622 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5623 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5631 @section @code{.stabd, .stabn, .stabs}
5633 @cindex symbolic debuggers, information for
5634 @cindex @code{stab@var{x}} directives
5635 There are three directives that begin @samp{.stab}.
5636 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5637 The symbols are not entered in the @command{@value{AS}} hash table: they
5638 cannot be referenced elsewhere in the source file.
5639 Up to five fields are required:
5643 This is the symbol's name. It may contain any character except
5644 @samp{\000}, so is more general than ordinary symbol names. Some
5645 debuggers used to code arbitrarily complex structures into symbol names
5649 An absolute expression. The symbol's type is set to the low 8 bits of
5650 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5651 and debuggers choke on silly bit patterns.
5654 An absolute expression. The symbol's ``other'' attribute is set to the
5655 low 8 bits of this expression.
5658 An absolute expression. The symbol's descriptor is set to the low 16
5659 bits of this expression.
5662 An absolute expression which becomes the symbol's value.
5665 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5666 or @code{.stabs} statement, the symbol has probably already been created;
5667 you get a half-formed symbol in your object file. This is
5668 compatible with earlier assemblers!
5671 @cindex @code{stabd} directive
5672 @item .stabd @var{type} , @var{other} , @var{desc}
5674 The ``name'' of the symbol generated is not even an empty string.
5675 It is a null pointer, for compatibility. Older assemblers used a
5676 null pointer so they didn't waste space in object files with empty
5679 The symbol's value is set to the location counter,
5680 relocatably. When your program is linked, the value of this symbol
5681 is the address of the location counter when the @code{.stabd} was
5684 @cindex @code{stabn} directive
5685 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5686 The name of the symbol is set to the empty string @code{""}.
5688 @cindex @code{stabs} directive
5689 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5690 All five fields are specified.
5696 @section @code{.string} "@var{str}"
5698 @cindex string, copying to object file
5699 @cindex @code{string} directive
5701 Copy the characters in @var{str} to the object file. You may specify more than
5702 one string to copy, separated by commas. Unless otherwise specified for a
5703 particular machine, the assembler marks the end of each string with a 0 byte.
5704 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5707 @section @code{.struct @var{expression}}
5709 @cindex @code{struct} directive
5710 Switch to the absolute section, and set the section offset to @var{expression},
5711 which must be an absolute expression. You might use this as follows:
5720 This would define the symbol @code{field1} to have the value 0, the symbol
5721 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5722 value 8. Assembly would be left in the absolute section, and you would need to
5723 use a @code{.section} directive of some sort to change to some other section
5724 before further assembly.
5728 @section @code{.subsection @var{name}}
5730 @cindex @code{subsection} directive
5731 @cindex Section Stack
5732 This is one of the ELF section stack manipulation directives. The others are
5733 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5734 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5737 This directive replaces the current subsection with @code{name}. The current
5738 section is not changed. The replaced subsection is put onto the section stack
5739 in place of the then current top of stack subsection.
5744 @section @code{.symver}
5745 @cindex @code{symver} directive
5746 @cindex symbol versioning
5747 @cindex versions of symbols
5748 Use the @code{.symver} directive to bind symbols to specific version nodes
5749 within a source file. This is only supported on ELF platforms, and is
5750 typically used when assembling files to be linked into a shared library.
5751 There are cases where it may make sense to use this in objects to be bound
5752 into an application itself so as to override a versioned symbol from a
5755 For ELF targets, the @code{.symver} directive can be used like this:
5757 .symver @var{name}, @var{name2@@nodename}
5759 If the symbol @var{name} is defined within the file
5760 being assembled, the @code{.symver} directive effectively creates a symbol
5761 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5762 just don't try and create a regular alias is that the @var{@@} character isn't
5763 permitted in symbol names. The @var{name2} part of the name is the actual name
5764 of the symbol by which it will be externally referenced. The name @var{name}
5765 itself is merely a name of convenience that is used so that it is possible to
5766 have definitions for multiple versions of a function within a single source
5767 file, and so that the compiler can unambiguously know which version of a
5768 function is being mentioned. The @var{nodename} portion of the alias should be
5769 the name of a node specified in the version script supplied to the linker when
5770 building a shared library. If you are attempting to override a versioned
5771 symbol from a shared library, then @var{nodename} should correspond to the
5772 nodename of the symbol you are trying to override.
5774 If the symbol @var{name} is not defined within the file being assembled, all
5775 references to @var{name} will be changed to @var{name2@@nodename}. If no
5776 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5779 Another usage of the @code{.symver} directive is:
5781 .symver @var{name}, @var{name2@@@@nodename}
5783 In this case, the symbol @var{name} must exist and be defined within
5784 the file being assembled. It is similar to @var{name2@@nodename}. The
5785 difference is @var{name2@@@@nodename} will also be used to resolve
5786 references to @var{name2} by the linker.
5788 The third usage of the @code{.symver} directive is:
5790 .symver @var{name}, @var{name2@@@@@@nodename}
5792 When @var{name} is not defined within the
5793 file being assembled, it is treated as @var{name2@@nodename}. When
5794 @var{name} is defined within the file being assembled, the symbol
5795 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5800 @section @code{.tag @var{structname}}
5802 @cindex COFF structure debugging
5803 @cindex structure debugging, COFF
5804 @cindex @code{tag} directive
5805 This directive is generated by compilers to include auxiliary debugging
5806 information in the symbol table. It is only permitted inside
5807 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5808 definitions in the symbol table with instances of those structures.
5811 @samp{.tag} is only used when generating COFF format output; when
5812 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5818 @section @code{.text @var{subsection}}
5820 @cindex @code{text} directive
5821 Tells @command{@value{AS}} to assemble the following statements onto the end of
5822 the text subsection numbered @var{subsection}, which is an absolute
5823 expression. If @var{subsection} is omitted, subsection number zero
5827 @section @code{.title "@var{heading}"}
5829 @cindex @code{title} directive
5830 @cindex listing control: title line
5831 Use @var{heading} as the title (second line, immediately after the
5832 source file name and pagenumber) when generating assembly listings.
5834 This directive affects subsequent pages, as well as the current page if
5835 it appears within ten lines of the top of a page.
5839 @section @code{.type}
5841 This directive is used to set the type of a symbol.
5845 @c only print the extra heading if both COFF and ELF are set
5846 @subheading COFF Version
5849 @cindex COFF symbol type
5850 @cindex symbol type, COFF
5851 @cindex @code{type} directive (COFF version)
5852 For COFF targets, this directive is permitted only within
5853 @code{.def}/@code{.endef} pairs. It is used like this:
5859 This records the integer @var{int} as the type attribute of a symbol table
5863 @samp{.type} is associated only with COFF format output; when
5864 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
5865 directive but ignores it.
5871 @c only print the extra heading if both COFF and ELF are set
5872 @subheading ELF Version
5875 @cindex ELF symbol type
5876 @cindex symbol type, ELF
5877 @cindex @code{type} directive (ELF version)
5878 For ELF targets, the @code{.type} directive is used like this:
5881 .type @var{name} , @var{type description}
5884 This sets the type of symbol @var{name} to be either a
5885 function symbol or an object symbol. There are five different syntaxes
5886 supported for the @var{type description} field, in order to provide
5887 compatibility with various other assemblers. The syntaxes supported are:
5890 .type <name>,#function
5891 .type <name>,#object
5893 .type <name>,@@function
5894 .type <name>,@@object
5896 .type <name>,%function
5897 .type <name>,%object
5899 .type <name>,"function"
5900 .type <name>,"object"
5902 .type <name> STT_FUNCTION
5903 .type <name> STT_OBJECT
5909 @section @code{.uleb128 @var{expressions}}
5911 @cindex @code{uleb128} directive
5912 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
5913 compact, variable length representation of numbers used by the DWARF
5914 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
5918 @section @code{.val @var{addr}}
5920 @cindex @code{val} directive
5921 @cindex COFF value attribute
5922 @cindex value attribute, COFF
5923 This directive, permitted only within @code{.def}/@code{.endef} pairs,
5924 records the address @var{addr} as the value attribute of a symbol table
5928 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
5929 configured for @code{b.out}, it accepts this directive but ignores it.
5935 @section @code{.version "@var{string}"}
5937 @cindex @code{version} directive
5938 This directive creates a @code{.note} section and places into it an ELF
5939 formatted note of type NT_VERSION. The note's name is set to @code{string}.
5944 @section @code{.vtable_entry @var{table}, @var{offset}}
5946 @cindex @code{vtable_entry} directive
5947 This directive finds or creates a symbol @code{table} and creates a
5948 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
5951 @section @code{.vtable_inherit @var{child}, @var{parent}}
5953 @cindex @code{vtable_inherit} directive
5954 This directive finds the symbol @code{child} and finds or creates the symbol
5955 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
5956 parent whose addend is the value of the child symbol. As a special case the
5957 parent name of @code{0} is treated as refering the @code{*ABS*} section.
5961 @section @code{.warning "@var{string}"}
5962 @cindex warning directive
5963 Similar to the directive @code{.error}
5964 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
5967 @section @code{.weak @var{names}}
5969 @cindex @code{weak} directive
5970 This directive sets the weak attribute on the comma separated list of symbol
5971 @code{names}. If the symbols do not already exist, they will be created.
5973 On COFF targets other than PE, weak symbols are a GNU extension. This
5974 directive sets the weak attribute on the comma separated list of symbol
5975 @code{names}. If the symbols do not already exist, they will be created.
5977 On the PE target, weak symbols are supported natively as weak aliases.
5978 When a weak symbol is created that is not an alias, GAS creates an
5979 alternate symbol to hold the default value.
5982 @section @code{.word @var{expressions}}
5984 @cindex @code{word} directive
5985 This directive expects zero or more @var{expressions}, of any section,
5986 separated by commas.
5989 For each expression, @command{@value{AS}} emits a 32-bit number.
5992 For each expression, @command{@value{AS}} emits a 16-bit number.
5997 The size of the number emitted, and its byte order,
5998 depend on what target computer the assembly is for.
6001 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6002 @c happen---32-bit addressability, period; no long/short jumps.
6003 @ifset DIFF-TBL-KLUGE
6004 @cindex difference tables altered
6005 @cindex altered difference tables
6007 @emph{Warning: Special Treatment to support Compilers}
6011 Machines with a 32-bit address space, but that do less than 32-bit
6012 addressing, require the following special treatment. If the machine of
6013 interest to you does 32-bit addressing (or doesn't require it;
6014 @pxref{Machine Dependencies}), you can ignore this issue.
6017 In order to assemble compiler output into something that works,
6018 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6019 Directives of the form @samp{.word sym1-sym2} are often emitted by
6020 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6021 directive of the form @samp{.word sym1-sym2}, and the difference between
6022 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6023 creates a @dfn{secondary jump table}, immediately before the next label.
6024 This secondary jump table is preceded by a short-jump to the
6025 first byte after the secondary table. This short-jump prevents the flow
6026 of control from accidentally falling into the new table. Inside the
6027 table is a long-jump to @code{sym2}. The original @samp{.word}
6028 contains @code{sym1} minus the address of the long-jump to
6031 If there were several occurrences of @samp{.word sym1-sym2} before the
6032 secondary jump table, all of them are adjusted. If there was a
6033 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6034 long-jump to @code{sym4} is included in the secondary jump table,
6035 and the @code{.word} directives are adjusted to contain @code{sym3}
6036 minus the address of the long-jump to @code{sym4}; and so on, for as many
6037 entries in the original jump table as necessary.
6040 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6041 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6042 assembly language programmers.
6045 @c end DIFF-TBL-KLUGE
6048 @section Deprecated Directives
6050 @cindex deprecated directives
6051 @cindex obsolescent directives
6052 One day these directives won't work.
6053 They are included for compatibility with older assemblers.
6060 @node Machine Dependencies
6061 @chapter Machine Dependent Features
6063 @cindex machine dependencies
6064 The machine instruction sets are (almost by definition) different on
6065 each machine where @command{@value{AS}} runs. Floating point representations
6066 vary as well, and @command{@value{AS}} often supports a few additional
6067 directives or command-line options for compatibility with other
6068 assemblers on a particular platform. Finally, some versions of
6069 @command{@value{AS}} support special pseudo-instructions for branch
6072 This chapter discusses most of these differences, though it does not
6073 include details on any machine's instruction set. For details on that
6074 subject, see the hardware manufacturer's manual.
6078 * Alpha-Dependent:: Alpha Dependent Features
6081 * ARC-Dependent:: ARC Dependent Features
6084 * ARM-Dependent:: ARM Dependent Features
6087 * CRIS-Dependent:: CRIS Dependent Features
6090 * D10V-Dependent:: D10V Dependent Features
6093 * D30V-Dependent:: D30V Dependent Features
6096 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6099 * HPPA-Dependent:: HPPA Dependent Features
6102 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6105 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6108 * i860-Dependent:: Intel 80860 Dependent Features
6111 * i960-Dependent:: Intel 80960 Dependent Features
6114 * IA-64-Dependent:: Intel IA-64 Dependent Features
6117 * IP2K-Dependent:: IP2K Dependent Features
6120 * M32C-Dependent:: M32C Dependent Features
6123 * M32R-Dependent:: M32R Dependent Features
6126 * M68K-Dependent:: M680x0 Dependent Features
6129 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6132 * MIPS-Dependent:: MIPS Dependent Features
6135 * MMIX-Dependent:: MMIX Dependent Features
6138 * MSP430-Dependent:: MSP430 Dependent Features
6141 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6142 * SH64-Dependent:: SuperH SH64 Dependent Features
6145 * PDP-11-Dependent:: PDP-11 Dependent Features
6148 * PJ-Dependent:: picoJava Dependent Features
6151 * PPC-Dependent:: PowerPC Dependent Features
6154 * Sparc-Dependent:: SPARC Dependent Features
6157 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
6160 * V850-Dependent:: V850 Dependent Features
6163 * Xtensa-Dependent:: Xtensa Dependent Features
6166 * Z8000-Dependent:: Z8000 Dependent Features
6169 * Vax-Dependent:: VAX Dependent Features
6176 @c The following major nodes are *sections* in the GENERIC version, *chapters*
6177 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
6178 @c peculiarity: to preserve cross-references, there must be a node called
6179 @c "Machine Dependencies". Hence the conditional nodenames in each
6180 @c major node below. Node defaulting in makeinfo requires adjacency of
6181 @c node and sectioning commands; hence the repetition of @chapter BLAH
6182 @c in both conditional blocks.
6185 @include c-alpha.texi
6197 @include c-cris.texi
6202 @node Machine Dependencies
6203 @chapter Machine Dependent Features
6205 The machine instruction sets are different on each Renesas chip family,
6206 and there are also some syntax differences among the families. This
6207 chapter describes the specific @command{@value{AS}} features for each
6211 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6212 * SH-Dependent:: Renesas SH Dependent Features
6219 @include c-d10v.texi
6223 @include c-d30v.texi
6227 @include c-h8300.texi
6231 @include c-hppa.texi
6235 @include c-i370.texi
6239 @include c-i386.texi
6243 @include c-i860.texi
6247 @include c-i960.texi
6251 @include c-ia64.texi
6255 @include c-ip2k.texi
6259 @include c-m32c.texi
6263 @include c-m32r.texi
6267 @include c-m68k.texi
6271 @include c-m68hc11.texi
6275 @include c-mips.texi
6279 @include c-mmix.texi
6283 @include c-msp430.texi
6287 @include c-ns32k.texi
6291 @include c-pdp11.texi
6304 @include c-sh64.texi
6308 @include c-sparc.texi
6312 @include c-tic54x.texi
6324 @include c-v850.texi
6328 @include c-xtensa.texi
6332 @c reverse effect of @down at top of generic Machine-Dep chapter
6336 @node Reporting Bugs
6337 @chapter Reporting Bugs
6338 @cindex bugs in assembler
6339 @cindex reporting bugs in assembler
6341 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6343 Reporting a bug may help you by bringing a solution to your problem, or it may
6344 not. But in any case the principal function of a bug report is to help the
6345 entire community by making the next version of @command{@value{AS}} work better.
6346 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6348 In order for a bug report to serve its purpose, you must include the
6349 information that enables us to fix the bug.
6352 * Bug Criteria:: Have you found a bug?
6353 * Bug Reporting:: How to report bugs
6357 @section Have You Found a Bug?
6358 @cindex bug criteria
6360 If you are not sure whether you have found a bug, here are some guidelines:
6363 @cindex fatal signal
6364 @cindex assembler crash
6365 @cindex crash of assembler
6367 If the assembler gets a fatal signal, for any input whatever, that is a
6368 @command{@value{AS}} bug. Reliable assemblers never crash.
6370 @cindex error on valid input
6372 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6374 @cindex invalid input
6376 If @command{@value{AS}} does not produce an error message for invalid input, that
6377 is a bug. However, you should note that your idea of ``invalid input'' might
6378 be our idea of ``an extension'' or ``support for traditional practice''.
6381 If you are an experienced user of assemblers, your suggestions for improvement
6382 of @command{@value{AS}} are welcome in any case.
6386 @section How to Report Bugs
6388 @cindex assembler bugs, reporting
6390 A number of companies and individuals offer support for @sc{gnu} products. If
6391 you obtained @command{@value{AS}} from a support organization, we recommend you
6392 contact that organization first.
6394 You can find contact information for many support companies and
6395 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6398 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6399 to @samp{bug-binutils@@gnu.org}.
6401 The fundamental principle of reporting bugs usefully is this:
6402 @strong{report all the facts}. If you are not sure whether to state a
6403 fact or leave it out, state it!
6405 Often people omit facts because they think they know what causes the problem
6406 and assume that some details do not matter. Thus, you might assume that the
6407 name of a symbol you use in an example does not matter. Well, probably it does
6408 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6409 happens to fetch from the location where that name is stored in memory;
6410 perhaps, if the name were different, the contents of that location would fool
6411 the assembler into doing the right thing despite the bug. Play it safe and
6412 give a specific, complete example. That is the easiest thing for you to do,
6413 and the most helpful.
6415 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6416 it is new to us. Therefore, always write your bug reports on the assumption
6417 that the bug has not been reported previously.
6419 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6420 bell?'' This cannot help us fix a bug, so it is basically useless. We
6421 respond by asking for enough details to enable us to investigate.
6422 You might as well expedite matters by sending them to begin with.
6424 To enable us to fix the bug, you should include all these things:
6428 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6429 it with the @samp{--version} argument.
6431 Without this, we will not know whether there is any point in looking for
6432 the bug in the current version of @command{@value{AS}}.
6435 Any patches you may have applied to the @command{@value{AS}} source.
6438 The type of machine you are using, and the operating system name and
6442 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6446 The command arguments you gave the assembler to assemble your example and
6447 observe the bug. To guarantee you will not omit something important, list them
6448 all. A copy of the Makefile (or the output from make) is sufficient.
6450 If we were to try to guess the arguments, we would probably guess wrong
6451 and then we might not encounter the bug.
6454 A complete input file that will reproduce the bug. If the bug is observed when
6455 the assembler is invoked via a compiler, send the assembler source, not the
6456 high level language source. Most compilers will produce the assembler source
6457 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6458 the options @samp{-v --save-temps}; this will save the assembler source in a
6459 file with an extension of @file{.s}, and also show you exactly how
6460 @command{@value{AS}} is being run.
6463 A description of what behavior you observe that you believe is
6464 incorrect. For example, ``It gets a fatal signal.''
6466 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6467 will certainly notice it. But if the bug is incorrect output, we might not
6468 notice unless it is glaringly wrong. You might as well not give us a chance to
6471 Even if the problem you experience is a fatal signal, you should still say so
6472 explicitly. Suppose something strange is going on, such as, your copy of
6473 @command{@value{AS}} is out of synch, or you have encountered a bug in the C
6474 library on your system. (This has happened!) Your copy might crash and ours
6475 would not. If you told us to expect a crash, then when ours fails to crash, we
6476 would know that the bug was not happening for us. If you had not told us to
6477 expect a crash, then we would not be able to draw any conclusion from our
6481 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6482 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6483 option. Always send diffs from the old file to the new file. If you even
6484 discuss something in the @command{@value{AS}} source, refer to it by context, not
6487 The line numbers in our development sources will not match those in your
6488 sources. Your line numbers would convey no useful information to us.
6491 Here are some things that are not necessary:
6495 A description of the envelope of the bug.
6497 Often people who encounter a bug spend a lot of time investigating
6498 which changes to the input file will make the bug go away and which
6499 changes will not affect it.
6501 This is often time consuming and not very useful, because the way we
6502 will find the bug is by running a single example under the debugger
6503 with breakpoints, not by pure deduction from a series of examples.
6504 We recommend that you save your time for something else.
6506 Of course, if you can find a simpler example to report @emph{instead}
6507 of the original one, that is a convenience for us. Errors in the
6508 output will be easier to spot, running under the debugger will take
6509 less time, and so on.
6511 However, simplification is not vital; if you do not want to do this,
6512 report the bug anyway and send us the entire test case you used.
6515 A patch for the bug.
6517 A patch for the bug does help us if it is a good one. But do not omit
6518 the necessary information, such as the test case, on the assumption that
6519 a patch is all we need. We might see problems with your patch and decide
6520 to fix the problem another way, or we might not understand it at all.
6522 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6523 construct an example that will make the program follow a certain path through
6524 the code. If you do not send us the example, we will not be able to construct
6525 one, so we will not be able to verify that the bug is fixed.
6527 And if we cannot understand what bug you are trying to fix, or why your
6528 patch should be an improvement, we will not install it. A test case will
6529 help us to understand.
6532 A guess about what the bug is or what it depends on.
6534 Such guesses are usually wrong. Even we cannot guess right about such
6535 things without first using the debugger to find the facts.
6538 @node Acknowledgements
6539 @chapter Acknowledgements
6541 If you have contributed to GAS and your name isn't listed here,
6542 it is not meant as a slight. We just don't know about it. Send mail to the
6543 maintainer, and we'll correct the situation. Currently
6545 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6547 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6550 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6551 information and the 68k series machines, most of the preprocessing pass, and
6552 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6554 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6555 many bug fixes, including merging support for several processors, breaking GAS
6556 up to handle multiple object file format back ends (including heavy rewrite,
6557 testing, an integration of the coff and b.out back ends), adding configuration
6558 including heavy testing and verification of cross assemblers and file splits
6559 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6560 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6561 port (including considerable amounts of reverse engineering), a SPARC opcode
6562 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6563 assertions and made them work, much other reorganization, cleanup, and lint.
6565 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6566 in format-specific I/O modules.
6568 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6569 has done much work with it since.
6571 The Intel 80386 machine description was written by Eliot Dresselhaus.
6573 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6575 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6576 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6578 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6579 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6580 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6581 support a.out format.
6583 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
6584 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
6585 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6586 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6589 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6590 simplified the configuration of which versions accept which directives. He
6591 updated the 68k machine description so that Motorola's opcodes always produced
6592 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6593 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6594 cross-compilation support, and one bug in relaxation that took a week and
6595 required the proverbial one-bit fix.
6597 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6598 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6599 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6600 PowerPC assembler, and made a few other minor patches.
6602 Steve Chamberlain made GAS able to generate listings.
6604 Hewlett-Packard contributed support for the HP9000/300.
6606 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6607 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6608 formats). This work was supported by both the Center for Software Science at
6609 the University of Utah and Cygnus Support.
6611 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6612 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6613 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6614 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6615 and some initial 64-bit support).
6617 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6619 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6620 support for openVMS/Alpha.
6622 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6625 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6626 Inc. added support for Xtensa processors.
6628 Several engineers at Cygnus Support have also provided many small bug fixes and
6629 configuration enhancements.
6631 Many others have contributed large or small bugfixes and enhancements. If
6632 you have contributed significant work and are not mentioned on this list, and
6633 want to be, let us know. Some of the history has been lost; we are not
6634 intentionally leaving anyone out.