1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
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:
22 @include asconfig.texi
27 @c Configure for the generation of man pages
64 @c common OR combinations of conditions
90 @set abnormal-separator
94 @settitle Using @value{AS}
97 @settitle Using @value{AS} (@value{TARGET})
99 @setchapternewpage odd
104 @c WARE! Some of the machine-dependent sections contain tables of machine
105 @c instructions. Except in multi-column format, these tables look silly.
106 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
107 @c the multi-col format is faked within @example sections.
109 @c Again unfortunately, the natural size that fits on a page, for these tables,
110 @c is different depending on whether or not smallbook is turned on.
111 @c This matters, because of order: text flow switches columns at each page
114 @c The format faked in this source works reasonably well for smallbook,
115 @c not well for the default large-page format. This manual expects that if you
116 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
117 @c tables in question. You can turn on one without the other at your
118 @c discretion, of course.
121 @c the insn tables look just as silly in info files regardless of smallbook,
122 @c might as well show 'em anyways.
128 * As: (as). The GNU assembler.
129 * Gas: (as). The GNU assembler.
138 This file documents the GNU Assembler "@value{AS}".
140 @c man begin COPYRIGHT
141 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002 Free Software Foundation, Inc.
143 Permission is granted to copy, distribute and/or modify this document
144 under the terms of the GNU Free Documentation License, Version 1.1
145 or any later version published by the Free Software Foundation;
146 with no Invariant Sections, with no Front-Cover Texts, and with no
147 Back-Cover Texts. A copy of the license is included in the
148 section entitled ``GNU Free Documentation License''.
153 Permission is granted to process this file through Tex and print the
154 results, provided the printed document carries copying permission
155 notice identical to this one except for the removal of this paragraph
156 (this paragraph not being relevant to the printed manual).
162 @title Using @value{AS}
163 @subtitle The @sc{gnu} Assembler
165 @subtitle for the @value{TARGET} family
168 @subtitle Version @value{VERSION}
171 The Free Software Foundation Inc. thanks The Nice Computer
172 Company of Australia for loaning Dean Elsner to write the
173 first (Vax) version of @command{as} for Project @sc{gnu}.
174 The proprietors, management and staff of TNCCA thank FSF for
175 distracting the boss while they got some work
178 @author Dean Elsner, Jay Fenlason & friends
182 \hfill {\it Using {\tt @value{AS}}}\par
183 \hfill Edited by Cygnus Support\par
185 %"boxit" macro for figures:
186 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
187 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
188 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
189 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
190 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
193 @vskip 0pt plus 1filll
194 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002 Free Software Foundation, Inc.
196 Permission is granted to copy, distribute and/or modify this document
197 under the terms of the GNU Free Documentation License, Version 1.1
198 or any later version published by the Free Software Foundation;
199 with no Invariant Sections, with no Front-Cover Texts, and with no
200 Back-Cover Texts. A copy of the license is included in the
201 section entitled ``GNU Free Documentation License''.
207 @top Using @value{AS}
209 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}} version
212 This version of the file describes @command{@value{AS}} configured to generate
213 code for @value{TARGET} architectures.
216 This document is distributed under the terms of the GNU Free
217 Documentation License. A copy of the license is included in the
218 section entitled ``GNU Free Documentation License''.
221 * Overview:: Overview
222 * Invoking:: Command-Line Options
224 * Sections:: Sections and Relocation
226 * Expressions:: Expressions
227 * Pseudo Ops:: Assembler Directives
228 * Machine Dependencies:: Machine Dependent Features
229 * Reporting Bugs:: Reporting Bugs
230 * Acknowledgements:: Who Did What
231 * GNU Free Documentation License:: GNU Free Documentation License
239 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
241 This version of the manual describes @command{@value{AS}} configured to generate
242 code for @value{TARGET} architectures.
246 @cindex invocation summary
247 @cindex option summary
248 @cindex summary of options
249 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
250 @pxref{Invoking,,Command-Line Options}.
252 @c man title AS the portable GNU assembler.
256 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
260 @c We don't use deffn and friends for the following because they seem
261 @c to be limited to one line for the header.
263 @c man begin SYNOPSIS
264 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{-D}] [@b{--defsym} @var{sym}=@var{val}]
265 [@b{-f}] [@b{--gstabs}] [@b{--gdwarf2}] [@b{--help}] [@b{-I} @var{dir}]
266 [@b{-J}] [@b{-K}] [@b{-L}]
267 [@b{--listing-lhs-width}=@var{NUM}] [@b{--listing-lhs-width2}=@var{NUM}]
268 [@b{--listing-rhs-width}=@var{NUM}] [@b{--listing-cont-lines}=@var{NUM}]
269 [@b{--keep-locals}] [@b{-o} @var{objfile}] [@b{-R}] [@b{--statistics}] [@b{-v}]
270 [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}] [@b{--fatal-warnings}]
271 [@b{-w}] [@b{-x}] [@b{-Z}] [@b{--target-help}] [@var{target-options}]
272 [@b{--}|@var{files} @dots{}]
274 @c Target dependent options are listed below. Keep the list sorted.
275 @c Add an empty line for separation.
277 @c am29k has no machine-dependent assembler options
281 @emph{Target Alpha options:}
283 [@b{-mdebug} | @b{-no-mdebug}]
284 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
285 [@b{-F}] [@b{-32addr}]
289 @emph{Target ARC options:}
295 @emph{Target ARM options:}
296 @c Don't document the deprecated options
297 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
298 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
299 [@b{-mfpu}=@var{floating-point-fromat}]
302 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
303 @b{-mapcs-reentrant}]
304 [@b{-mthumb-interwork}] [@b{-moabi}] [@b{-k}]
308 @emph{Target CRIS options:}
309 [@b{--underscore} | @b{--no-underscore}]
311 [@b{--emulation=criself} | @b{--emulation=crisaout}]
312 @c Deprecated -- deliberately not documented.
317 @emph{Target D10V options:}
322 @emph{Target D30V options:}
323 [@b{-O}|@b{-n}|@b{-N}]
326 @c Hitachi family chips have no machine-dependent assembler options
329 @c HPPA has no machine-dependent assembler options (yet).
333 @emph{Target i386 options:}
338 @emph{Target i960 options:}
339 @c see md_parse_option in tc-i960.c
340 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
342 [@b{-b}] [@b{-no-relax}]
346 @emph{Target IA-64 options:}
347 [@b{-mconstant-gp}|@b{-mauto-pic}]
348 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
350 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
354 @emph{Target IP2K options:}
355 [@b{-mip2022}|@b{-mip2022ext}]
359 @emph{Target M32R options:}
360 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
365 @emph{Target M680X0 options:}
366 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
370 @emph{Target M68HC11 options:}
371 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
372 [@b{-mshort}|@b{-mlong}]
373 [@b{-mshort-double}|@b{-mlong-double}]
374 [@b{--force-long-branchs}] [@b{--short-branchs}]
375 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
376 [@b{--print-opcodes}] [@b{--generate-example}]
380 @emph{Target MCORE options:}
381 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
382 [@b{-mcpu=[210|340]}]
386 @emph{Target MIPS options:}
387 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-n}] [@b{-O}[@var{optimization level}]]
388 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
389 [@b{-non_shared}] [@b{-xgot}] [@b{--membedded-pic}]
390 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
391 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
392 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
394 [@b{-construct-floats}] [@b{-no-construct-floats}]
395 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
396 [@b{-mfix7000}] [@b{-mno-fix7000}]
397 [@b{-mips16}] [@b{-no-mips16}]
398 [@b{-mips3d}] [@b{-no-mips3d}]
399 [@b{-mdmx}] [@b{-no-mdmx}]
400 [@b{-mdebug}] [@b{-no-mdebug}]
404 @emph{Target MMIX options:}
405 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
406 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
407 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
408 [@b{--linker-allocated-gregs}]
412 @emph{Target PDP11 options:}
413 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
414 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
415 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
419 @emph{Target picoJava options:}
424 @emph{Target PowerPC options:}
425 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
426 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
427 @b{-mbooke32}|@b{-mbooke64}]
428 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
429 [@b{-mregnames}|@b{-mno-regnames}]
430 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
431 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
432 [@b{-msolaris}|@b{-mno-solaris}]
436 @emph{Target SPARC options:}
437 @c The order here is important. See c-sparc.texi.
438 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
439 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
440 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
445 @emph{Target TIC54X options:}
446 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
447 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
450 @c Z8000 has no machine-dependent assembler options
459 Turn on listings, in any of a variety of ways:
463 omit false conditionals
466 omit debugging directives
469 include high-level source
475 include macro expansions
478 omit forms processing
484 set the name of the listing file
487 You may combine these options; for example, use @samp{-aln} for assembly
488 listing without forms processing. The @samp{=file} option, if used, must be
489 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
492 Ignored. This option is accepted for script compatibility with calls to
495 @item --defsym @var{sym}=@var{value}
496 Define the symbol @var{sym} to be @var{value} before assembling the input file.
497 @var{value} must be an integer constant. As in C, a leading @samp{0x}
498 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
501 ``fast''---skip whitespace and comment preprocessing (assume source is
505 Generate stabs debugging information for each assembler line. This
506 may help debugging assembler code, if the debugger can handle it.
509 Generate DWARF2 debugging information for each assembler line. This
510 may help debugging assembler code, if the debugger can handle it. Note---this
511 option is only supported by some targets, not all of them.
514 Print a summary of the command line options and exit.
517 Print a summary of all target specific options and exit.
520 Add directory @var{dir} to the search list for @code{.include} directives.
523 Don't warn about signed overflow.
526 @ifclear DIFF-TBL-KLUGE
527 This option is accepted but has no effect on the @value{TARGET} family.
529 @ifset DIFF-TBL-KLUGE
530 Issue warnings when difference tables altered for long displacements.
535 Keep (in the symbol table) local symbols. On traditional a.out systems
536 these start with @samp{L}, but different systems have different local
539 @item --listing-lhs-width=@var{number}
540 Set the maximum width, in words, of the output data column for an assembler
541 listing to @var{number}.
543 @item --listing-lhs-width2=@var{number}
544 Set the maximum width, in words, of the output data column for continuation
545 lines in an assembler listing to @var{number}.
547 @item --listing-rhs-width=@var{number}
548 Set the maximum width of an input source line, as displayed in a listing, to
551 @item --listing-cont-lines=@var{number}
552 Set the maximum number of lines printed in a listing for a single line of input
555 @item -o @var{objfile}
556 Name the object-file output from @command{@value{AS}} @var{objfile}.
559 Fold the data section into the text section.
562 Print the maximum space (in bytes) and total time (in seconds) used by
565 @item --strip-local-absolute
566 Remove local absolute symbols from the outgoing symbol table.
570 Print the @command{as} version.
573 Print the @command{as} version and exit.
577 Suppress warning messages.
579 @item --fatal-warnings
580 Treat warnings as errors.
583 Don't suppress warning messages or treat them as errors.
592 Generate an object file even after errors.
594 @item -- | @var{files} @dots{}
595 Standard input, or source files to assemble.
600 The following options are available when @value{AS} is configured for
605 This option selects the core processor variant.
607 Select either big-endian (-EB) or little-endian (-EL) output.
612 The following options are available when @value{AS} is configured for the ARM
616 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
617 Specify which ARM processor variant is the target.
618 @item -march=@var{architecture}[+@var{extension}@dots{}]
619 Specify which ARM architecture variant is used by the target.
620 @item -mfpu=@var{floating-point-format}
621 Select which Floating Point architecture is the target.
623 Enable Thumb only instruction decoding.
624 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant | -moabi
625 Select which procedure calling convention is in use.
627 Select either big-endian (-EB) or little-endian (-EL) output.
628 @item -mthumb-interwork
629 Specify that the code has been generated with interworking between Thumb and
632 Specify that PIC code has been generated.
637 See the info pages for documentation of the CRIS-specific options.
641 The following options are available when @value{AS} is configured for
644 @cindex D10V optimization
645 @cindex optimization, D10V
647 Optimize output by parallelizing instructions.
652 The following options are available when @value{AS} is configured for a D30V
655 @cindex D30V optimization
656 @cindex optimization, D30V
658 Optimize output by parallelizing instructions.
662 Warn when nops are generated.
664 @cindex D30V nops after 32-bit multiply
666 Warn when a nop after a 32-bit multiply instruction is generated.
671 The following options are available when @value{AS} is configured for the
672 Intel 80960 processor.
675 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
676 Specify which variant of the 960 architecture is the target.
679 Add code to collect statistics about branches taken.
682 Do not alter compare-and-branch instructions for long displacements;
689 The following options are available when @value{AS} is configured for the
695 Specifies that the extended IP2022 instructions are allowed.
698 Restores the default behaviour, which restricts the permitted instructions to
699 just the basic IP2022 ones.
705 The following options are available when @value{AS} is configured for the
706 Mitsubishi M32R series.
711 Specify which processor in the M32R family is the target. The default
712 is normally the M32R, but this option changes it to the M32RX.
714 @item --warn-explicit-parallel-conflicts or --Wp
715 Produce warning messages when questionable parallel constructs are
718 @item --no-warn-explicit-parallel-conflicts or --Wnp
719 Do not produce warning messages when questionable parallel constructs are
726 The following options are available when @value{AS} is configured for the
727 Motorola 68000 series.
732 Shorten references to undefined symbols, to one word instead of two.
734 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
735 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
736 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
737 Specify what processor in the 68000 family is the target. The default
738 is normally the 68020, but this can be changed at configuration time.
740 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
741 The target machine does (or does not) have a floating-point coprocessor.
742 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
743 the basic 68000 is not compatible with the 68881, a combination of the
744 two can be specified, since it's possible to do emulation of the
745 coprocessor instructions with the main processor.
747 @item -m68851 | -mno-68851
748 The target machine does (or does not) have a memory-management
749 unit coprocessor. The default is to assume an MMU for 68020 and up.
756 For details about the PDP-11 machine dependent features options,
757 see @ref{PDP-11-Options}.
760 @item -mpic | -mno-pic
761 Generate position-independent (or position-dependent) code. The
762 default is @option{-mpic}.
765 @itemx -mall-extensions
766 Enable all instruction set extensions. This is the default.
768 @item -mno-extensions
769 Disable all instruction set extensions.
771 @item -m@var{extension} | -mno-@var{extension}
772 Enable (or disable) a particular instruction set extension.
775 Enable the instruction set extensions supported by a particular CPU, and
776 disable all other extensions.
778 @item -m@var{machine}
779 Enable the instruction set extensions supported by a particular machine
780 model, and disable all other extensions.
786 The following options are available when @value{AS} is configured for
787 a picoJava processor.
791 @cindex PJ endianness
792 @cindex endianness, PJ
793 @cindex big endian output, PJ
795 Generate ``big endian'' format output.
797 @cindex little endian output, PJ
799 Generate ``little endian'' format output.
805 The following options are available when @value{AS} is configured for the
806 Motorola 68HC11 or 68HC12 series.
810 @item -m68hc11 | -m68hc12 | -m68hcs12
811 Specify what processor is the target. The default is
812 defined by the configuration option when building the assembler.
815 Specify to use the 16-bit integer ABI.
818 Specify to use the 32-bit integer ABI.
821 Specify to use the 32-bit double ABI.
824 Specify to use the 64-bit double ABI.
826 @item --force-long-branchs
827 Relative branches are turned into absolute ones. This concerns
828 conditional branches, unconditional branches and branches to a
831 @item -S | --short-branchs
832 Do not turn relative branchs into absolute ones
833 when the offset is out of range.
835 @item --strict-direct-mode
836 Do not turn the direct addressing mode into extended addressing mode
837 when the instruction does not support direct addressing mode.
839 @item --print-insn-syntax
840 Print the syntax of instruction in case of error.
842 @item --print-opcodes
843 print the list of instructions with syntax and then exit.
845 @item --generate-example
846 print an example of instruction for each possible instruction and then exit.
847 This option is only useful for testing @command{@value{AS}}.
853 The following options are available when @command{@value{AS}} is configured
854 for the SPARC architecture:
857 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
858 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
859 Explicitly select a variant of the SPARC architecture.
861 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
862 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
864 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
865 UltraSPARC extensions.
867 @item -xarch=v8plus | -xarch=v8plusa
868 For compatibility with the Solaris v9 assembler. These options are
869 equivalent to -Av8plus and -Av8plusa, respectively.
872 Warn when the assembler switches to another architecture.
877 The following options are available when @value{AS} is configured for the 'c54x
882 Enable extended addressing mode. All addresses and relocations will assume
883 extended addressing (usually 23 bits).
884 @item -mcpu=@var{CPU_VERSION}
885 Sets the CPU version being compiled for.
886 @item -merrors-to-file @var{FILENAME}
887 Redirect error output to a file, for broken systems which don't support such
888 behaviour in the shell.
893 The following options are available when @value{AS} is configured for
894 a @sc{mips} processor.
898 This option sets the largest size of an object that can be referenced
899 implicitly with the @code{gp} register. It is only accepted for targets that
900 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
902 @cindex MIPS endianness
903 @cindex endianness, MIPS
904 @cindex big endian output, MIPS
906 Generate ``big endian'' format output.
908 @cindex little endian output, MIPS
910 Generate ``little endian'' format output.
921 Generate code for a particular @sc{mips} Instruction Set Architecture level.
922 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
923 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
924 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
925 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, and @samp{-mips64}
926 correspond to generic
927 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, and
928 @samp{MIPS64} ISA processors,
931 @item -march=@var{CPU}
932 Generate code for a particular @sc{mips} cpu.
934 @item -mtune=@var{cpu}
935 Schedule and tune for a particular @sc{mips} cpu.
939 Cause nops to be inserted if the read of the destination register
940 of an mfhi or mflo instruction occurs in the following two instructions.
944 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
945 section instead of the standard ELF .stabs sections.
949 The register sizes are normally inferred from the ISA and ABI, but these
950 flags force a certain group of registers to be treated as 32 bits wide at
951 all times. @samp{-mgp32} controls the size of general-purpose registers
952 and @samp{-mfp32} controls the size of floating-point registers.
956 Generate code for the MIPS 16 processor. This is equivalent to putting
957 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
958 turns off this option.
962 Generate code for the MIPS-3D Application Specific Extension.
963 This tells the assembler to accept MIPS-3D instructions.
964 @samp{-no-mips3d} turns off this option.
968 Generate code for the MDMX Application Specific Extension.
969 This tells the assembler to accept MDMX instructions.
970 @samp{-no-mdmx} turns off this option.
972 @item --construct-floats
973 @itemx --no-construct-floats
974 The @samp{--no-construct-floats} option disables the construction of
975 double width floating point constants by loading the two halves of the
976 value into the two single width floating point registers that make up
977 the double width register. By default @samp{--construct-floats} is
978 selected, allowing construction of these floating point constants.
981 @item --emulation=@var{name}
982 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
983 for some other target, in all respects, including output format (choosing
984 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
985 debugging information or store symbol table information, and default
986 endianness. The available configuration names are: @samp{mipsecoff},
987 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
988 @samp{mipsbelf}. The first two do not alter the default endianness from that
989 of the primary target for which the assembler was configured; the others change
990 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
991 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
992 selection in any case.
994 This option is currently supported only when the primary target
995 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
996 Furthermore, the primary target or others specified with
997 @samp{--enable-targets=@dots{}} at configuration time must include support for
998 the other format, if both are to be available. For example, the Irix 5
999 configuration includes support for both.
1001 Eventually, this option will support more configurations, with more
1002 fine-grained control over the assembler's behavior, and will be supported for
1006 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1013 Control how to deal with multiplication overflow and division by zero.
1014 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1015 (and only work for Instruction Set Architecture level 2 and higher);
1016 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1020 When this option is used, @command{@value{AS}} will issue a warning every
1021 time it generates a nop instruction from a macro.
1026 The following options are available when @value{AS} is configured for
1032 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1033 The command line option @samp{-nojsri2bsr} can be used to disable it.
1037 Enable or disable the silicon filter behaviour. By default this is disabled.
1038 The default can be overridden by the @samp{-sifilter} command line option.
1041 Alter jump instructions for long displacements.
1043 @item -mcpu=[210|340]
1044 Select the cpu type on the target hardware. This controls which instructions
1048 Assemble for a big endian target.
1051 Assemble for a little endian target.
1057 See the info pages for documentation of the MMIX-specific options.
1063 * Manual:: Structure of this Manual
1064 * GNU Assembler:: The GNU Assembler
1065 * Object Formats:: Object File Formats
1066 * Command Line:: Command Line
1067 * Input Files:: Input Files
1068 * Object:: Output (Object) File
1069 * Errors:: Error and Warning Messages
1073 @section Structure of this Manual
1075 @cindex manual, structure and purpose
1076 This manual is intended to describe what you need to know to use
1077 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1078 notation for symbols, constants, and expressions; the directives that
1079 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1082 We also cover special features in the @value{TARGET}
1083 configuration of @command{@value{AS}}, including assembler directives.
1086 This manual also describes some of the machine-dependent features of
1087 various flavors of the assembler.
1090 @cindex machine instructions (not covered)
1091 On the other hand, this manual is @emph{not} intended as an introduction
1092 to programming in assembly language---let alone programming in general!
1093 In a similar vein, we make no attempt to introduce the machine
1094 architecture; we do @emph{not} describe the instruction set, standard
1095 mnemonics, registers or addressing modes that are standard to a
1096 particular architecture.
1098 You may want to consult the manufacturer's
1099 machine architecture manual for this information.
1103 For information on the H8/300 machine instruction set, see @cite{H8/300
1104 Series Programming Manual} (Hitachi ADE--602--025). For the H8/300H,
1105 see @cite{H8/300H Series Programming Manual} (Hitachi).
1108 For information on the H8/500 machine instruction set, see @cite{H8/500
1109 Series Programming Manual} (Hitachi M21T001).
1112 For information on the Hitachi SH machine instruction set, see
1113 @cite{SH-Microcomputer User's Manual} (Hitachi Micro Systems, Inc.).
1116 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1120 @c I think this is premature---doc@cygnus.com, 17jan1991
1122 Throughout this manual, we assume that you are running @dfn{GNU},
1123 the portable operating system from the @dfn{Free Software
1124 Foundation, Inc.}. This restricts our attention to certain kinds of
1125 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1126 once this assumption is granted examples and definitions need less
1129 @command{@value{AS}} is part of a team of programs that turn a high-level
1130 human-readable series of instructions into a low-level
1131 computer-readable series of instructions. Different versions of
1132 @command{@value{AS}} are used for different kinds of computer.
1135 @c There used to be a section "Terminology" here, which defined
1136 @c "contents", "byte", "word", and "long". Defining "word" to any
1137 @c particular size is confusing when the .word directive may generate 16
1138 @c bits on one machine and 32 bits on another; in general, for the user
1139 @c version of this manual, none of these terms seem essential to define.
1140 @c They were used very little even in the former draft of the manual;
1141 @c this draft makes an effort to avoid them (except in names of
1145 @section The GNU Assembler
1147 @c man begin DESCRIPTION
1149 @sc{gnu} @command{as} is really a family of assemblers.
1151 This manual describes @command{@value{AS}}, a member of that family which is
1152 configured for the @value{TARGET} architectures.
1154 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1155 should find a fairly similar environment when you use it on another
1156 architecture. Each version has much in common with the others,
1157 including object file formats, most assembler directives (often called
1158 @dfn{pseudo-ops}) and assembler syntax.@refill
1160 @cindex purpose of @sc{gnu} assembler
1161 @command{@value{AS}} is primarily intended to assemble the output of the
1162 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1163 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1164 assemble correctly everything that other assemblers for the same
1165 machine would assemble.
1167 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1170 @c This remark should appear in generic version of manual; assumption
1171 @c here is that generic version sets M680x0.
1172 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1173 assembler for the same architecture; for example, we know of several
1174 incompatible versions of 680x0 assembly language syntax.
1179 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1180 program in one pass of the source file. This has a subtle impact on the
1181 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1183 @node Object Formats
1184 @section Object File Formats
1186 @cindex object file format
1187 The @sc{gnu} assembler can be configured to produce several alternative
1188 object file formats. For the most part, this does not affect how you
1189 write assembly language programs; but directives for debugging symbols
1190 are typically different in different file formats. @xref{Symbol
1191 Attributes,,Symbol Attributes}.
1194 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1195 @value{OBJ-NAME} format object files.
1197 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1199 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1200 @code{a.out} or COFF format object files.
1203 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1204 @code{b.out} or COFF format object files.
1207 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1208 SOM or ELF format object files.
1213 @section Command Line
1215 @cindex command line conventions
1217 After the program name @command{@value{AS}}, the command line may contain
1218 options and file names. Options may appear in any order, and may be
1219 before, after, or between file names. The order of file names is
1222 @cindex standard input, as input file
1224 @file{--} (two hyphens) by itself names the standard input file
1225 explicitly, as one of the files for @command{@value{AS}} to assemble.
1227 @cindex options, command line
1228 Except for @samp{--} any command line argument that begins with a
1229 hyphen (@samp{-}) is an option. Each option changes the behavior of
1230 @command{@value{AS}}. No option changes the way another option works. An
1231 option is a @samp{-} followed by one or more letters; the case of
1232 the letter is important. All options are optional.
1234 Some options expect exactly one file name to follow them. The file
1235 name may either immediately follow the option's letter (compatible
1236 with older assemblers) or it may be the next command argument (@sc{gnu}
1237 standard). These two command lines are equivalent:
1240 @value{AS} -o my-object-file.o mumble.s
1241 @value{AS} -omy-object-file.o mumble.s
1245 @section Input Files
1248 @cindex source program
1249 @cindex files, input
1250 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1251 describe the program input to one run of @command{@value{AS}}. The program may
1252 be in one or more files; how the source is partitioned into files
1253 doesn't change the meaning of the source.
1255 @c I added "con" prefix to "catenation" just to prove I can overcome my
1256 @c APL training... doc@cygnus.com
1257 The source program is a concatenation of the text in all the files, in the
1260 @c man begin DESCRIPTION
1261 Each time you run @command{@value{AS}} it assembles exactly one source
1262 program. The source program is made up of one or more files.
1263 (The standard input is also a file.)
1265 You give @command{@value{AS}} a command line that has zero or more input file
1266 names. The input files are read (from left file name to right). A
1267 command line argument (in any position) that has no special meaning
1268 is taken to be an input file name.
1270 If you give @command{@value{AS}} no file names it attempts to read one input file
1271 from the @command{@value{AS}} standard input, which is normally your terminal. You
1272 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1275 Use @samp{--} if you need to explicitly name the standard input file
1276 in your command line.
1278 If the source is empty, @command{@value{AS}} produces a small, empty object
1283 @subheading Filenames and Line-numbers
1285 @cindex input file linenumbers
1286 @cindex line numbers, in input files
1287 There are two ways of locating a line in the input file (or files) and
1288 either may be used in reporting error messages. One way refers to a line
1289 number in a physical file; the other refers to a line number in a
1290 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1292 @dfn{Physical files} are those files named in the command line given
1293 to @command{@value{AS}}.
1295 @dfn{Logical files} are simply names declared explicitly by assembler
1296 directives; they bear no relation to physical files. Logical file names help
1297 error messages reflect the original source file, when @command{@value{AS}} source
1298 is itself synthesized from other files. @command{@value{AS}} understands the
1299 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1300 @ref{File,,@code{.file}}.
1303 @section Output (Object) File
1309 Every time you run @command{@value{AS}} it produces an output file, which is
1310 your assembly language program translated into numbers. This file
1311 is the object file. Its default name is
1319 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1321 You can give it another name by using the @option{-o} option. Conventionally,
1322 object file names end with @file{.o}. The default name is used for historical
1323 reasons: older assemblers were capable of assembling self-contained programs
1324 directly into a runnable program. (For some formats, this isn't currently
1325 possible, but it can be done for the @code{a.out} format.)
1329 The object file is meant for input to the linker @code{@value{LD}}. It contains
1330 assembled program code, information to help @code{@value{LD}} integrate
1331 the assembled program into a runnable file, and (optionally) symbolic
1332 information for the debugger.
1334 @c link above to some info file(s) like the description of a.out.
1335 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1338 @section Error and Warning Messages
1340 @c man begin DESCRIPTION
1342 @cindex error messages
1343 @cindex warning messages
1344 @cindex messages from assembler
1345 @command{@value{AS}} may write warnings and error messages to the standard error
1346 file (usually your terminal). This should not happen when a compiler
1347 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1348 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1349 grave problem that stops the assembly.
1353 @cindex format of warning messages
1354 Warning messages have the format
1357 file_name:@b{NNN}:Warning Message Text
1361 @cindex line numbers, in warnings/errors
1362 (where @b{NNN} is a line number). If a logical file name has been given
1363 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1364 the current input file is used. If a logical line number was given
1366 (@pxref{Line,,@code{.line}})
1370 (@pxref{Line,,@code{.line}})
1373 (@pxref{Ln,,@code{.ln}})
1376 then it is used to calculate the number printed,
1377 otherwise the actual line in the current source file is printed. The
1378 message text is intended to be self explanatory (in the grand Unix
1381 @cindex format of error messages
1382 Error messages have the format
1384 file_name:@b{NNN}:FATAL:Error Message Text
1386 The file name and line number are derived as for warning
1387 messages. The actual message text may be rather less explanatory
1388 because many of them aren't supposed to happen.
1391 @chapter Command-Line Options
1393 @cindex options, all versions of assembler
1394 This chapter describes command-line options available in @emph{all}
1395 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1397 to the @value{TARGET} target.
1400 to particular machine architectures.
1403 @c man begin DESCRIPTION
1405 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1406 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1407 The assembler arguments must be separated from each other (and the @samp{-Wa})
1408 by commas. For example:
1411 gcc -c -g -O -Wa,-alh,-L file.c
1415 This passes two options to the assembler: @samp{-alh} (emit a listing to
1416 standard output with high-level and assembly source) and @samp{-L} (retain
1417 local symbols in the symbol table).
1419 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1420 command-line options are automatically passed to the assembler by the compiler.
1421 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1422 precisely what options it passes to each compilation pass, including the
1428 * a:: -a[cdhlns] enable listings
1429 * D:: -D for compatibility
1430 * f:: -f to work faster
1431 * I:: -I for .include search path
1432 @ifclear DIFF-TBL-KLUGE
1433 * K:: -K for compatibility
1435 @ifset DIFF-TBL-KLUGE
1436 * K:: -K for difference tables
1439 * L:: -L to retain local labels
1440 * listing:: --listing-XXX to configure listing output
1441 * M:: -M or --mri to assemble in MRI compatibility mode
1442 * MD:: --MD for dependency tracking
1443 * o:: -o to name the object file
1444 * R:: -R to join data and text sections
1445 * statistics:: --statistics to see statistics about assembly
1446 * traditional-format:: --traditional-format for compatible output
1447 * v:: -v to announce version
1448 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1449 * Z:: -Z to make object file even after errors
1453 @section Enable Listings: @option{-a[cdhlns]}
1462 @cindex listings, enabling
1463 @cindex assembly listings, enabling
1465 These options enable listing output from the assembler. By itself,
1466 @samp{-a} requests high-level, assembly, and symbols listing.
1467 You can use other letters to select specific options for the list:
1468 @samp{-ah} requests a high-level language listing,
1469 @samp{-al} requests an output-program assembly listing, and
1470 @samp{-as} requests a symbol table listing.
1471 High-level listings require that a compiler debugging option like
1472 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1475 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1476 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1477 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1478 omitted from the listing.
1480 Use the @samp{-ad} option to omit debugging directives from the
1483 Once you have specified one of these options, you can further control
1484 listing output and its appearance using the directives @code{.list},
1485 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1487 The @samp{-an} option turns off all forms processing.
1488 If you do not request listing output with one of the @samp{-a} options, the
1489 listing-control directives have no effect.
1491 The letters after @samp{-a} may be combined into one option,
1492 @emph{e.g.}, @samp{-aln}.
1494 Note if the assembler source is coming from the standard input (eg because it
1495 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1496 is being used) then the listing will not contain any comments or preprocessor
1497 directives. This is because the listing code buffers input source lines from
1498 stdin only after they have been preprocessed by the assembler. This reduces
1499 memory usage and makes the code more efficient.
1502 @section @option{-D}
1505 This option has no effect whatsoever, but it is accepted to make it more
1506 likely that scripts written for other assemblers also work with
1507 @command{@value{AS}}.
1510 @section Work Faster: @option{-f}
1513 @cindex trusted compiler
1514 @cindex faster processing (@option{-f})
1515 @samp{-f} should only be used when assembling programs written by a
1516 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1517 and comment preprocessing on
1518 the input file(s) before assembling them. @xref{Preprocessing,
1522 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1523 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1528 @section @code{.include} Search Path: @option{-I} @var{path}
1530 @kindex -I @var{path}
1531 @cindex paths for @code{.include}
1532 @cindex search path for @code{.include}
1533 @cindex @code{include} directive search path
1534 Use this option to add a @var{path} to the list of directories
1535 @command{@value{AS}} searches for files specified in @code{.include}
1536 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1537 many times as necessary to include a variety of paths. The current
1538 working directory is always searched first; after that, @command{@value{AS}}
1539 searches any @samp{-I} directories in the same order as they were
1540 specified (left to right) on the command line.
1543 @section Difference Tables: @option{-K}
1546 @ifclear DIFF-TBL-KLUGE
1547 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1548 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1549 where it can be used to warn when the assembler alters the machine code
1550 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1551 family does not have the addressing limitations that sometimes lead to this
1552 alteration on other platforms.
1555 @ifset DIFF-TBL-KLUGE
1556 @cindex difference tables, warning
1557 @cindex warning for altered difference tables
1558 @command{@value{AS}} sometimes alters the code emitted for directives of the form
1559 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1560 You can use the @samp{-K} option if you want a warning issued when this
1565 @section Include Local Labels: @option{-L}
1568 @cindex local labels, retaining in output
1569 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1570 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1571 debugging, because they are intended for the use of programs (like
1572 compilers) that compose assembler programs, not for your notice.
1573 Normally both @command{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1574 normally debug with them.
1576 This option tells @command{@value{AS}} to retain those @samp{L@dots{}} symbols
1577 in the object file. Usually if you do this you also tell the linker
1578 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1580 By default, a local label is any label beginning with @samp{L}, but each
1581 target is allowed to redefine the local label prefix.
1583 On the HPPA local labels begin with @samp{L$}.
1587 @section Configuring listing output: @option{--listing}
1589 The listing feature of the assembler can be enabled via the command line switch
1590 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1591 hex dump of the corresponding locations in the output object file, and displays
1592 them as a listing file. The format of this listing can be controlled by pseudo
1593 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1594 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1597 @item --listing-lhs-width=@samp{number}
1598 @kindex --listing-lhs-width
1599 @cindex Width of first line disassembly output
1600 Sets the maximum width, in words, of the first line of the hex byte dump. This
1601 dump appears on the left hand side of the listing output.
1603 @item --listing-lhs-width2=@samp{number}
1604 @kindex --listing-lhs-width2
1605 @cindex Width of continuation lines of disassembly output
1606 Sets the maximum width, in words, of any further lines of the hex byte dump for
1607 a given input source line. If this value is not specified, it defaults to being
1608 the same as the value specified for @samp{--listing-lhs-width}. If neither
1609 switch is used the default is to one.
1611 @item --listing-rhs-width=@samp{number}
1612 @kindex --listing-rhs-width
1613 @cindex Width of source line output
1614 Sets the maximum width, in characters, of the source line that is displayed
1615 alongside the hex dump. The default value for this parameter is 100. The
1616 source line is displayed on the right hand side of the listing output.
1618 @item --listing-cont-lines=@samp{number}
1619 @kindex --listing-cont-lines
1620 @cindex Maximum number of continuation lines
1621 Sets the maximum number of continuation lines of hex dump that will be
1622 displayed for a given single line of source input. The default value is 4.
1626 @section Assemble in MRI Compatibility Mode: @option{-M}
1629 @cindex MRI compatibility mode
1630 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1631 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1632 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1633 configured target) assembler from Microtec Research. The exact nature of the
1634 MRI syntax will not be documented here; see the MRI manuals for more
1635 information. Note in particular that the handling of macros and macro
1636 arguments is somewhat different. The purpose of this option is to permit
1637 assembling existing MRI assembler code using @command{@value{AS}}.
1639 The MRI compatibility is not complete. Certain operations of the MRI assembler
1640 depend upon its object file format, and can not be supported using other object
1641 file formats. Supporting these would require enhancing each object file format
1642 individually. These are:
1645 @item global symbols in common section
1647 The m68k MRI assembler supports common sections which are merged by the linker.
1648 Other object file formats do not support this. @command{@value{AS}} handles
1649 common sections by treating them as a single common symbol. It permits local
1650 symbols to be defined within a common section, but it can not support global
1651 symbols, since it has no way to describe them.
1653 @item complex relocations
1655 The MRI assemblers support relocations against a negated section address, and
1656 relocations which combine the start addresses of two or more sections. These
1657 are not support by other object file formats.
1659 @item @code{END} pseudo-op specifying start address
1661 The MRI @code{END} pseudo-op permits the specification of a start address.
1662 This is not supported by other object file formats. The start address may
1663 instead be specified using the @option{-e} option to the linker, or in a linker
1666 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1668 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1669 name to the output file. This is not supported by other object file formats.
1671 @item @code{ORG} pseudo-op
1673 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1674 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1675 which changes the location within the current section. Absolute sections are
1676 not supported by other object file formats. The address of a section may be
1677 assigned within a linker script.
1680 There are some other features of the MRI assembler which are not supported by
1681 @command{@value{AS}}, typically either because they are difficult or because they
1682 seem of little consequence. Some of these may be supported in future releases.
1686 @item EBCDIC strings
1688 EBCDIC strings are not supported.
1690 @item packed binary coded decimal
1692 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1693 and @code{DCB.P} pseudo-ops are not supported.
1695 @item @code{FEQU} pseudo-op
1697 The m68k @code{FEQU} pseudo-op is not supported.
1699 @item @code{NOOBJ} pseudo-op
1701 The m68k @code{NOOBJ} pseudo-op is not supported.
1703 @item @code{OPT} branch control options
1705 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1706 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1707 relaxes all branches, whether forward or backward, to an appropriate size, so
1708 these options serve no purpose.
1710 @item @code{OPT} list control options
1712 The following m68k @code{OPT} list control options are ignored: @code{C},
1713 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1714 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1716 @item other @code{OPT} options
1718 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1719 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1721 @item @code{OPT} @code{D} option is default
1723 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1724 @code{OPT NOD} may be used to turn it off.
1726 @item @code{XREF} pseudo-op.
1728 The m68k @code{XREF} pseudo-op is ignored.
1730 @item @code{.debug} pseudo-op
1732 The i960 @code{.debug} pseudo-op is not supported.
1734 @item @code{.extended} pseudo-op
1736 The i960 @code{.extended} pseudo-op is not supported.
1738 @item @code{.list} pseudo-op.
1740 The various options of the i960 @code{.list} pseudo-op are not supported.
1742 @item @code{.optimize} pseudo-op
1744 The i960 @code{.optimize} pseudo-op is not supported.
1746 @item @code{.output} pseudo-op
1748 The i960 @code{.output} pseudo-op is not supported.
1750 @item @code{.setreal} pseudo-op
1752 The i960 @code{.setreal} pseudo-op is not supported.
1757 @section Dependency Tracking: @option{--MD}
1760 @cindex dependency tracking
1763 @command{@value{AS}} can generate a dependency file for the file it creates. This
1764 file consists of a single rule suitable for @code{make} describing the
1765 dependencies of the main source file.
1767 The rule is written to the file named in its argument.
1769 This feature is used in the automatic updating of makefiles.
1772 @section Name the Object File: @option{-o}
1775 @cindex naming object file
1776 @cindex object file name
1777 There is always one object file output when you run @command{@value{AS}}. By
1778 default it has the name
1781 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1795 You use this option (which takes exactly one filename) to give the
1796 object file a different name.
1798 Whatever the object file is called, @command{@value{AS}} overwrites any
1799 existing file of the same name.
1802 @section Join Data and Text Sections: @option{-R}
1805 @cindex data and text sections, joining
1806 @cindex text and data sections, joining
1807 @cindex joining text and data sections
1808 @cindex merging text and data sections
1809 @option{-R} tells @command{@value{AS}} to write the object file as if all
1810 data-section data lives in the text section. This is only done at
1811 the very last moment: your binary data are the same, but data
1812 section parts are relocated differently. The data section part of
1813 your object file is zero bytes long because all its bytes are
1814 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1816 When you specify @option{-R} it would be possible to generate shorter
1817 address displacements (because we do not have to cross between text and
1818 data section). We refrain from doing this simply for compatibility with
1819 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1822 When @command{@value{AS}} is configured for COFF or ELF output,
1823 this option is only useful if you use sections named @samp{.text} and
1828 @option{-R} is not supported for any of the HPPA targets. Using
1829 @option{-R} generates a warning from @command{@value{AS}}.
1833 @section Display Assembly Statistics: @option{--statistics}
1835 @kindex --statistics
1836 @cindex statistics, about assembly
1837 @cindex time, total for assembly
1838 @cindex space used, maximum for assembly
1839 Use @samp{--statistics} to display two statistics about the resources used by
1840 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1841 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1844 @node traditional-format
1845 @section Compatible Output: @option{--traditional-format}
1847 @kindex --traditional-format
1848 For some targets, the output of @command{@value{AS}} is different in some ways
1849 from the output of some existing assembler. This switch requests
1850 @command{@value{AS}} to use the traditional format instead.
1852 For example, it disables the exception frame optimizations which
1853 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
1856 @section Announce Version: @option{-v}
1860 @cindex assembler version
1861 @cindex version of assembler
1862 You can find out what version of as is running by including the
1863 option @samp{-v} (which you can also spell as @samp{-version}) on the
1867 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
1869 @command{@value{AS}} should never give a warning or error message when
1870 assembling compiler output. But programs written by people often
1871 cause @command{@value{AS}} to give a warning that a particular assumption was
1872 made. All such warnings are directed to the standard error file.
1876 @cindex suppressing warnings
1877 @cindex warnings, suppressing
1878 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
1879 This only affects the warning messages: it does not change any particular of
1880 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
1883 @kindex --fatal-warnings
1884 @cindex errors, caused by warnings
1885 @cindex warnings, causing error
1886 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
1887 files that generate warnings to be in error.
1890 @cindex warnings, switching on
1891 You can switch these options off again by specifying @option{--warn}, which
1892 causes warnings to be output as usual.
1895 @section Generate Object File in Spite of Errors: @option{-Z}
1896 @cindex object file, after errors
1897 @cindex errors, continuing after
1898 After an error message, @command{@value{AS}} normally produces no output. If for
1899 some reason you are interested in object file output even after
1900 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
1901 option. If there are any errors, @command{@value{AS}} continues anyways, and
1902 writes an object file after a final warning message of the form @samp{@var{n}
1903 errors, @var{m} warnings, generating bad object file.}
1908 @cindex machine-independent syntax
1909 @cindex syntax, machine-independent
1910 This chapter describes the machine-independent syntax allowed in a
1911 source file. @command{@value{AS}} syntax is similar to what many other
1912 assemblers use; it is inspired by the BSD 4.2
1917 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
1921 * Preprocessing:: Preprocessing
1922 * Whitespace:: Whitespace
1923 * Comments:: Comments
1924 * Symbol Intro:: Symbols
1925 * Statements:: Statements
1926 * Constants:: Constants
1930 @section Preprocessing
1932 @cindex preprocessing
1933 The @command{@value{AS}} internal preprocessor:
1935 @cindex whitespace, removed by preprocessor
1937 adjusts and removes extra whitespace. It leaves one space or tab before
1938 the keywords on a line, and turns any other whitespace on the line into
1941 @cindex comments, removed by preprocessor
1943 removes all comments, replacing them with a single space, or an
1944 appropriate number of newlines.
1946 @cindex constants, converted by preprocessor
1948 converts character constants into the appropriate numeric values.
1951 It does not do macro processing, include file handling, or
1952 anything else you may get from your C compiler's preprocessor. You can
1953 do include file processing with the @code{.include} directive
1954 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
1955 to get other ``CPP'' style preprocessing by giving the input file a
1956 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
1957 Output, gcc.info, Using GNU CC}.
1959 Excess whitespace, comments, and character constants
1960 cannot be used in the portions of the input text that are not
1963 @cindex turning preprocessing on and off
1964 @cindex preprocessing, turning on and off
1967 If the first line of an input file is @code{#NO_APP} or if you use the
1968 @samp{-f} option, whitespace and comments are not removed from the input file.
1969 Within an input file, you can ask for whitespace and comment removal in
1970 specific portions of the by putting a line that says @code{#APP} before the
1971 text that may contain whitespace or comments, and putting a line that says
1972 @code{#NO_APP} after this text. This feature is mainly intend to support
1973 @code{asm} statements in compilers whose output is otherwise free of comments
1980 @dfn{Whitespace} is one or more blanks or tabs, in any order.
1981 Whitespace is used to separate symbols, and to make programs neater for
1982 people to read. Unless within character constants
1983 (@pxref{Characters,,Character Constants}), any whitespace means the same
1984 as exactly one space.
1990 There are two ways of rendering comments to @command{@value{AS}}. In both
1991 cases the comment is equivalent to one space.
1993 Anything from @samp{/*} through the next @samp{*/} is a comment.
1994 This means you may not nest these comments.
1998 The only way to include a newline ('\n') in a comment
1999 is to use this sort of comment.
2002 /* This sort of comment does not nest. */
2005 @cindex line comment character
2006 Anything from the @dfn{line comment} character to the next newline
2007 is considered a comment and is ignored. The line comment character is
2009 @samp{;} for the AMD 29K family;
2012 @samp{;} on the ARC;
2015 @samp{@@} on the ARM;
2018 @samp{;} for the H8/300 family;
2021 @samp{!} for the H8/500 family;
2024 @samp{;} for the HPPA;
2027 @samp{#} on the i386 and x86-64;
2030 @samp{#} on the i960;
2033 @samp{;} for the PDP-11;
2036 @samp{;} for picoJava;
2039 @samp{;} for Motorola PowerPC;
2042 @samp{!} for the Hitachi SH;
2045 @samp{!} on the SPARC;
2048 @samp{#} on the ip2k;
2051 @samp{#} on the m32r;
2054 @samp{|} on the 680x0;
2057 @samp{#} on the 68HC11 and 68HC12;
2060 @samp{;} on the M880x0;
2063 @samp{#} on the Vax;
2066 @samp{!} for the Z8000;
2069 @samp{#} on the V850;
2071 see @ref{Machine Dependencies}. @refill
2072 @c FIXME What about i860?
2075 On some machines there are two different line comment characters. One
2076 character only begins a comment if it is the first non-whitespace character on
2077 a line, while the other always begins a comment.
2081 The V850 assembler also supports a double dash as starting a comment that
2082 extends to the end of the line.
2088 @cindex lines starting with @code{#}
2089 @cindex logical line numbers
2090 To be compatible with past assemblers, lines that begin with @samp{#} have a
2091 special interpretation. Following the @samp{#} should be an absolute
2092 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2093 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
2094 new logical file name. The rest of the line, if any, should be whitespace.
2096 If the first non-whitespace characters on the line are not numeric,
2097 the line is ignored. (Just like a comment.)
2100 # This is an ordinary comment.
2101 # 42-6 "new_file_name" # New logical file name
2102 # This is logical line # 36.
2104 This feature is deprecated, and may disappear from future versions
2105 of @command{@value{AS}}.
2110 @cindex characters used in symbols
2111 @ifclear SPECIAL-SYMS
2112 A @dfn{symbol} is one or more characters chosen from the set of all
2113 letters (both upper and lower case), digits and the three characters
2119 A @dfn{symbol} is one or more characters chosen from the set of all
2120 letters (both upper and lower case), digits and the three characters
2121 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2127 On most machines, you can also use @code{$} in symbol names; exceptions
2128 are noted in @ref{Machine Dependencies}.
2130 No symbol may begin with a digit. Case is significant.
2131 There is no length limit: all characters are significant. Symbols are
2132 delimited by characters not in that set, or by the beginning of a file
2133 (since the source program must end with a newline, the end of a file is
2134 not a possible symbol delimiter). @xref{Symbols}.
2135 @cindex length of symbols
2140 @cindex statements, structure of
2141 @cindex line separator character
2142 @cindex statement separator character
2144 @ifclear abnormal-separator
2145 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2146 semicolon (@samp{;}). The newline or semicolon is considered part of
2147 the preceding statement. Newlines and semicolons within character
2148 constants are an exception: they do not end statements.
2150 @ifset abnormal-separator
2152 A @dfn{statement} ends at a newline character (@samp{\n}) or an ``at''
2153 sign (@samp{@@}). The newline or at sign is considered part of the
2154 preceding statement. Newlines and at signs within character constants
2155 are an exception: they do not end statements.
2158 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2159 point (@samp{!}). The newline or exclamation point is considered part of the
2160 preceding statement. Newlines and exclamation points within character
2161 constants are an exception: they do not end statements.
2164 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2165 H8/300) a dollar sign (@samp{$}); or (for the
2168 (@samp{;}). The newline or separator character is considered part of
2169 the preceding statement. Newlines and separators within character
2170 constants are an exception: they do not end statements.
2175 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2176 separator character. (The line separator is usually @samp{;}, unless
2177 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
2178 newline or separator character is considered part of the preceding
2179 statement. Newlines and separators within character constants are an
2180 exception: they do not end statements.
2183 @cindex newline, required at file end
2184 @cindex EOF, newline must precede
2185 It is an error to end any statement with end-of-file: the last
2186 character of any input file should be a newline.@refill
2188 An empty statement is allowed, and may include whitespace. It is ignored.
2190 @cindex instructions and directives
2191 @cindex directives and instructions
2192 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2193 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2195 A statement begins with zero or more labels, optionally followed by a
2196 key symbol which determines what kind of statement it is. The key
2197 symbol determines the syntax of the rest of the statement. If the
2198 symbol begins with a dot @samp{.} then the statement is an assembler
2199 directive: typically valid for any computer. If the symbol begins with
2200 a letter the statement is an assembly language @dfn{instruction}: it
2201 assembles into a machine language instruction.
2203 Different versions of @command{@value{AS}} for different computers
2204 recognize different instructions. In fact, the same symbol may
2205 represent a different instruction in a different computer's assembly
2209 @cindex @code{:} (label)
2210 @cindex label (@code{:})
2211 A label is a symbol immediately followed by a colon (@code{:}).
2212 Whitespace before a label or after a colon is permitted, but you may not
2213 have whitespace between a label's symbol and its colon. @xref{Labels}.
2216 For HPPA targets, labels need not be immediately followed by a colon, but
2217 the definition of a label must begin in column zero. This also implies that
2218 only one label may be defined on each line.
2222 label: .directive followed by something
2223 another_label: # This is an empty statement.
2224 instruction operand_1, operand_2, @dots{}
2231 A constant is a number, written so that its value is known by
2232 inspection, without knowing any context. Like this:
2235 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2236 .ascii "Ring the bell\7" # A string constant.
2237 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2238 .float 0f-314159265358979323846264338327\
2239 95028841971.693993751E-40 # - pi, a flonum.
2244 * Characters:: Character Constants
2245 * Numbers:: Number Constants
2249 @subsection Character Constants
2251 @cindex character constants
2252 @cindex constants, character
2253 There are two kinds of character constants. A @dfn{character} stands
2254 for one character in one byte and its value may be used in
2255 numeric expressions. String constants (properly called string
2256 @emph{literals}) are potentially many bytes and their values may not be
2257 used in arithmetic expressions.
2261 * Chars:: Characters
2265 @subsubsection Strings
2267 @cindex string constants
2268 @cindex constants, string
2269 A @dfn{string} is written between double-quotes. It may contain
2270 double-quotes or null characters. The way to get special characters
2271 into a string is to @dfn{escape} these characters: precede them with
2272 a backslash @samp{\} character. For example @samp{\\} represents
2273 one backslash: the first @code{\} is an escape which tells
2274 @command{@value{AS}} to interpret the second character literally as a backslash
2275 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2276 escape character). The complete list of escapes follows.
2278 @cindex escape codes, character
2279 @cindex character escape codes
2282 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2284 @cindex @code{\b} (backspace character)
2285 @cindex backspace (@code{\b})
2287 Mnemonic for backspace; for ASCII this is octal code 010.
2290 @c Mnemonic for EOText; for ASCII this is octal code 004.
2292 @cindex @code{\f} (formfeed character)
2293 @cindex formfeed (@code{\f})
2295 Mnemonic for FormFeed; for ASCII this is octal code 014.
2297 @cindex @code{\n} (newline character)
2298 @cindex newline (@code{\n})
2300 Mnemonic for newline; for ASCII this is octal code 012.
2303 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2305 @cindex @code{\r} (carriage return character)
2306 @cindex carriage return (@code{\r})
2308 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2311 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2312 @c other assemblers.
2314 @cindex @code{\t} (tab)
2315 @cindex tab (@code{\t})
2317 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2320 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2321 @c @item \x @var{digit} @var{digit} @var{digit}
2322 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2324 @cindex @code{\@var{ddd}} (octal character code)
2325 @cindex octal character code (@code{\@var{ddd}})
2326 @item \ @var{digit} @var{digit} @var{digit}
2327 An octal character code. The numeric code is 3 octal digits.
2328 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2329 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2331 @cindex @code{\@var{xd...}} (hex character code)
2332 @cindex hex character code (@code{\@var{xd...}})
2333 @item \@code{x} @var{hex-digits...}
2334 A hex character code. All trailing hex digits are combined. Either upper or
2335 lower case @code{x} works.
2337 @cindex @code{\\} (@samp{\} character)
2338 @cindex backslash (@code{\\})
2340 Represents one @samp{\} character.
2343 @c Represents one @samp{'} (accent acute) character.
2344 @c This is needed in single character literals
2345 @c (@xref{Characters,,Character Constants}.) to represent
2348 @cindex @code{\"} (doublequote character)
2349 @cindex doublequote (@code{\"})
2351 Represents one @samp{"} character. Needed in strings to represent
2352 this character, because an unescaped @samp{"} would end the string.
2354 @item \ @var{anything-else}
2355 Any other character when escaped by @kbd{\} gives a warning, but
2356 assembles as if the @samp{\} was not present. The idea is that if
2357 you used an escape sequence you clearly didn't want the literal
2358 interpretation of the following character. However @command{@value{AS}} has no
2359 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2360 code and warns you of the fact.
2363 Which characters are escapable, and what those escapes represent,
2364 varies widely among assemblers. The current set is what we think
2365 the BSD 4.2 assembler recognizes, and is a subset of what most C
2366 compilers recognize. If you are in doubt, do not use an escape
2370 @subsubsection Characters
2372 @cindex single character constant
2373 @cindex character, single
2374 @cindex constant, single character
2375 A single character may be written as a single quote immediately
2376 followed by that character. The same escapes apply to characters as
2377 to strings. So if you want to write the character backslash, you
2378 must write @kbd{'\\} where the first @code{\} escapes the second
2379 @code{\}. As you can see, the quote is an acute accent, not a
2380 grave accent. A newline
2382 @ifclear abnormal-separator
2383 (or semicolon @samp{;})
2385 @ifset abnormal-separator
2387 (or at sign @samp{@@})
2390 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2396 immediately following an acute accent is taken as a literal character
2397 and does not count as the end of a statement. The value of a character
2398 constant in a numeric expression is the machine's byte-wide code for
2399 that character. @command{@value{AS}} assumes your character code is ASCII:
2400 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2403 @subsection Number Constants
2405 @cindex constants, number
2406 @cindex number constants
2407 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2408 are stored in the target machine. @emph{Integers} are numbers that
2409 would fit into an @code{int} in the C language. @emph{Bignums} are
2410 integers, but they are stored in more than 32 bits. @emph{Flonums}
2411 are floating point numbers, described below.
2414 * Integers:: Integers
2419 * Bit Fields:: Bit Fields
2425 @subsubsection Integers
2427 @cindex constants, integer
2429 @cindex binary integers
2430 @cindex integers, binary
2431 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2432 the binary digits @samp{01}.
2434 @cindex octal integers
2435 @cindex integers, octal
2436 An octal integer is @samp{0} followed by zero or more of the octal
2437 digits (@samp{01234567}).
2439 @cindex decimal integers
2440 @cindex integers, decimal
2441 A decimal integer starts with a non-zero digit followed by zero or
2442 more digits (@samp{0123456789}).
2444 @cindex hexadecimal integers
2445 @cindex integers, hexadecimal
2446 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2447 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2449 Integers have the usual values. To denote a negative integer, use
2450 the prefix operator @samp{-} discussed under expressions
2451 (@pxref{Prefix Ops,,Prefix Operators}).
2454 @subsubsection Bignums
2457 @cindex constants, bignum
2458 A @dfn{bignum} has the same syntax and semantics as an integer
2459 except that the number (or its negative) takes more than 32 bits to
2460 represent in binary. The distinction is made because in some places
2461 integers are permitted while bignums are not.
2464 @subsubsection Flonums
2466 @cindex floating point numbers
2467 @cindex constants, floating point
2469 @cindex precision, floating point
2470 A @dfn{flonum} represents a floating point number. The translation is
2471 indirect: a decimal floating point number from the text is converted by
2472 @command{@value{AS}} to a generic binary floating point number of more than
2473 sufficient precision. This generic floating point number is converted
2474 to a particular computer's floating point format (or formats) by a
2475 portion of @command{@value{AS}} specialized to that computer.
2477 A flonum is written by writing (in order)
2482 (@samp{0} is optional on the HPPA.)
2486 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2488 @kbd{e} is recommended. Case is not important.
2490 @c FIXME: verify if flonum syntax really this vague for most cases
2491 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2492 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2495 On the H8/300, H8/500,
2497 and AMD 29K architectures, the letter must be
2498 one of the letters @samp{DFPRSX} (in upper or lower case).
2500 On the ARC, the letter must be one of the letters @samp{DFRS}
2501 (in upper or lower case).
2503 On the Intel 960 architecture, the letter must be
2504 one of the letters @samp{DFT} (in upper or lower case).
2506 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2510 One of the letters @samp{DFPRSX} (in upper or lower case).
2513 One of the letters @samp{DFRS} (in upper or lower case).
2516 One of the letters @samp{DFPRSX} (in upper or lower case).
2519 The letter @samp{E} (upper case only).
2522 One of the letters @samp{DFT} (in upper or lower case).
2527 An optional sign: either @samp{+} or @samp{-}.
2530 An optional @dfn{integer part}: zero or more decimal digits.
2533 An optional @dfn{fractional part}: @samp{.} followed by zero
2534 or more decimal digits.
2537 An optional exponent, consisting of:
2541 An @samp{E} or @samp{e}.
2542 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2543 @c principle this can perfectly well be different on different targets.
2545 Optional sign: either @samp{+} or @samp{-}.
2547 One or more decimal digits.
2552 At least one of the integer part or the fractional part must be
2553 present. The floating point number has the usual base-10 value.
2555 @command{@value{AS}} does all processing using integers. Flonums are computed
2556 independently of any floating point hardware in the computer running
2557 @command{@value{AS}}.
2561 @c Bit fields are written as a general facility but are also controlled
2562 @c by a conditional-compilation flag---which is as of now (21mar91)
2563 @c turned on only by the i960 config of GAS.
2565 @subsubsection Bit Fields
2568 @cindex constants, bit field
2569 You can also define numeric constants as @dfn{bit fields}.
2570 specify two numbers separated by a colon---
2572 @var{mask}:@var{value}
2575 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2578 The resulting number is then packed
2580 @c this conditional paren in case bit fields turned on elsewhere than 960
2581 (in host-dependent byte order)
2583 into a field whose width depends on which assembler directive has the
2584 bit-field as its argument. Overflow (a result from the bitwise and
2585 requiring more binary digits to represent) is not an error; instead,
2586 more constants are generated, of the specified width, beginning with the
2587 least significant digits.@refill
2589 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2590 @code{.short}, and @code{.word} accept bit-field arguments.
2595 @chapter Sections and Relocation
2600 * Secs Background:: Background
2601 * Ld Sections:: Linker Sections
2602 * As Sections:: Assembler Internal Sections
2603 * Sub-Sections:: Sub-Sections
2607 @node Secs Background
2610 Roughly, a section is a range of addresses, with no gaps; all data
2611 ``in'' those addresses is treated the same for some particular purpose.
2612 For example there may be a ``read only'' section.
2614 @cindex linker, and assembler
2615 @cindex assembler, and linker
2616 The linker @code{@value{LD}} reads many object files (partial programs) and
2617 combines their contents to form a runnable program. When @command{@value{AS}}
2618 emits an object file, the partial program is assumed to start at address 0.
2619 @code{@value{LD}} assigns the final addresses for the partial program, so that
2620 different partial programs do not overlap. This is actually an
2621 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2624 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2625 addresses. These blocks slide to their run-time addresses as rigid
2626 units; their length does not change and neither does the order of bytes
2627 within them. Such a rigid unit is called a @emph{section}. Assigning
2628 run-time addresses to sections is called @dfn{relocation}. It includes
2629 the task of adjusting mentions of object-file addresses so they refer to
2630 the proper run-time addresses.
2632 For the H8/300 and H8/500,
2633 and for the Hitachi SH,
2634 @command{@value{AS}} pads sections if needed to
2635 ensure they end on a word (sixteen bit) boundary.
2638 @cindex standard assembler sections
2639 An object file written by @command{@value{AS}} has at least three sections, any
2640 of which may be empty. These are named @dfn{text}, @dfn{data} and
2645 When it generates COFF or ELF output,
2647 @command{@value{AS}} can also generate whatever other named sections you specify
2648 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2649 If you do not use any directives that place output in the @samp{.text}
2650 or @samp{.data} sections, these sections still exist, but are empty.
2655 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2657 @command{@value{AS}} can also generate whatever other named sections you
2658 specify using the @samp{.space} and @samp{.subspace} directives. See
2659 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2660 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2661 assembler directives.
2664 Additionally, @command{@value{AS}} uses different names for the standard
2665 text, data, and bss sections when generating SOM output. Program text
2666 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2667 BSS into @samp{$BSS$}.
2671 Within the object file, the text section starts at address @code{0}, the
2672 data section follows, and the bss section follows the data section.
2675 When generating either SOM or ELF output files on the HPPA, the text
2676 section starts at address @code{0}, the data section at address
2677 @code{0x4000000}, and the bss section follows the data section.
2680 To let @code{@value{LD}} know which data changes when the sections are
2681 relocated, and how to change that data, @command{@value{AS}} also writes to the
2682 object file details of the relocation needed. To perform relocation
2683 @code{@value{LD}} must know, each time an address in the object
2687 Where in the object file is the beginning of this reference to
2690 How long (in bytes) is this reference?
2692 Which section does the address refer to? What is the numeric value of
2694 (@var{address}) @minus{} (@var{start-address of section})?
2697 Is the reference to an address ``Program-Counter relative''?
2700 @cindex addresses, format of
2701 @cindex section-relative addressing
2702 In fact, every address @command{@value{AS}} ever uses is expressed as
2704 (@var{section}) + (@var{offset into section})
2707 Further, most expressions @command{@value{AS}} computes have this section-relative
2710 (For some object formats, such as SOM for the HPPA, some expressions are
2711 symbol-relative instead.)
2714 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2715 @var{N} into section @var{secname}.''
2717 Apart from text, data and bss sections you need to know about the
2718 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2719 addresses in the absolute section remain unchanged. For example, address
2720 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2721 @code{@value{LD}}. Although the linker never arranges two partial programs'
2722 data sections with overlapping addresses after linking, @emph{by definition}
2723 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2724 part of a program is always the same address when the program is running as
2725 address @code{@{absolute@ 239@}} in any other part of the program.
2727 The idea of sections is extended to the @dfn{undefined} section. Any
2728 address whose section is unknown at assembly time is by definition
2729 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2730 Since numbers are always defined, the only way to generate an undefined
2731 address is to mention an undefined symbol. A reference to a named
2732 common block would be such a symbol: its value is unknown at assembly
2733 time so it has section @emph{undefined}.
2735 By analogy the word @emph{section} is used to describe groups of sections in
2736 the linked program. @code{@value{LD}} puts all partial programs' text
2737 sections in contiguous addresses in the linked program. It is
2738 customary to refer to the @emph{text section} of a program, meaning all
2739 the addresses of all partial programs' text sections. Likewise for
2740 data and bss sections.
2742 Some sections are manipulated by @code{@value{LD}}; others are invented for
2743 use of @command{@value{AS}} and have no meaning except during assembly.
2746 @section Linker Sections
2747 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2752 @cindex named sections
2753 @cindex sections, named
2754 @item named sections
2757 @cindex text section
2758 @cindex data section
2762 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2763 separate but equal sections. Anything you can say of one section is
2766 When the program is running, however, it is
2767 customary for the text section to be unalterable. The
2768 text section is often shared among processes: it contains
2769 instructions, constants and the like. The data section of a running
2770 program is usually alterable: for example, C variables would be stored
2771 in the data section.
2776 This section contains zeroed bytes when your program begins running. It
2777 is used to hold uninitialized variables or common storage. The length of
2778 each partial program's bss section is important, but because it starts
2779 out containing zeroed bytes there is no need to store explicit zero
2780 bytes in the object file. The bss section was invented to eliminate
2781 those explicit zeros from object files.
2783 @cindex absolute section
2784 @item absolute section
2785 Address 0 of this section is always ``relocated'' to runtime address 0.
2786 This is useful if you want to refer to an address that @code{@value{LD}} must
2787 not change when relocating. In this sense we speak of absolute
2788 addresses being ``unrelocatable'': they do not change during relocation.
2790 @cindex undefined section
2791 @item undefined section
2792 This ``section'' is a catch-all for address references to objects not in
2793 the preceding sections.
2794 @c FIXME: ref to some other doc on obj-file formats could go here.
2797 @cindex relocation example
2798 An idealized example of three relocatable sections follows.
2800 The example uses the traditional section names @samp{.text} and @samp{.data}.
2802 Memory addresses are on the horizontal axis.
2806 @c END TEXI2ROFF-KILL
2809 partial program # 1: |ttttt|dddd|00|
2816 partial program # 2: |TTT|DDD|000|
2819 +--+---+-----+--+----+---+-----+~~
2820 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2821 +--+---+-----+--+----+---+-----+~~
2823 addresses: 0 @dots{}
2830 \line{\it Partial program \#1: \hfil}
2831 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2832 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2834 \line{\it Partial program \#2: \hfil}
2835 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2836 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2838 \line{\it linked program: \hfil}
2839 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2840 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2841 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2842 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2844 \line{\it addresses: \hfil}
2848 @c END TEXI2ROFF-KILL
2851 @section Assembler Internal Sections
2853 @cindex internal assembler sections
2854 @cindex sections in messages, internal
2855 These sections are meant only for the internal use of @command{@value{AS}}. They
2856 have no meaning at run-time. You do not really need to know about these
2857 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2858 warning messages, so it might be helpful to have an idea of their
2859 meanings to @command{@value{AS}}. These sections are used to permit the
2860 value of every expression in your assembly language program to be a
2861 section-relative address.
2864 @cindex assembler internal logic error
2865 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2866 An internal assembler logic error has been found. This means there is a
2867 bug in the assembler.
2869 @cindex expr (internal section)
2871 The assembler stores complex expression internally as combinations of
2872 symbols. When it needs to represent an expression as a symbol, it puts
2873 it in the expr section.
2875 @c FIXME item transfer[t] vector preload
2876 @c FIXME item transfer[t] vector postload
2877 @c FIXME item register
2881 @section Sub-Sections
2883 @cindex numbered subsections
2884 @cindex grouping data
2890 fall into two sections: text and data.
2892 You may have separate groups of
2894 data in named sections
2898 data in named sections
2904 that you want to end up near to each other in the object file, even though they
2905 are not contiguous in the assembler source. @command{@value{AS}} allows you to
2906 use @dfn{subsections} for this purpose. Within each section, there can be
2907 numbered subsections with values from 0 to 8192. Objects assembled into the
2908 same subsection go into the object file together with other objects in the same
2909 subsection. For example, a compiler might want to store constants in the text
2910 section, but might not want to have them interspersed with the program being
2911 assembled. In this case, the compiler could issue a @samp{.text 0} before each
2912 section of code being output, and a @samp{.text 1} before each group of
2913 constants being output.
2915 Subsections are optional. If you do not use subsections, everything
2916 goes in subsection number zero.
2919 Each subsection is zero-padded up to a multiple of four bytes.
2920 (Subsections may be padded a different amount on different flavors
2921 of @command{@value{AS}}.)
2925 On the H8/300 and H8/500 platforms, each subsection is zero-padded to a word
2926 boundary (two bytes).
2927 The same is true on the Hitachi SH.
2930 @c FIXME section padding (alignment)?
2931 @c Rich Pixley says padding here depends on target obj code format; that
2932 @c doesn't seem particularly useful to say without further elaboration,
2933 @c so for now I say nothing about it. If this is a generic BFD issue,
2934 @c these paragraphs might need to vanish from this manual, and be
2935 @c discussed in BFD chapter of binutils (or some such).
2938 On the AMD 29K family, no particular padding is added to section or
2939 subsection sizes; @value{AS} forces no alignment on this platform.
2943 Subsections appear in your object file in numeric order, lowest numbered
2944 to highest. (All this to be compatible with other people's assemblers.)
2945 The object file contains no representation of subsections; @code{@value{LD}} and
2946 other programs that manipulate object files see no trace of them.
2947 They just see all your text subsections as a text section, and all your
2948 data subsections as a data section.
2950 To specify which subsection you want subsequent statements assembled
2951 into, use a numeric argument to specify it, in a @samp{.text
2952 @var{expression}} or a @samp{.data @var{expression}} statement.
2955 When generating COFF or ELF output, you
2960 can also use an extra subsection
2961 argument with arbitrary named sections: @samp{.section @var{name},
2964 @var{Expression} should be an absolute expression.
2965 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
2966 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
2967 begins in @code{text 0}. For instance:
2969 .text 0 # The default subsection is text 0 anyway.
2970 .ascii "This lives in the first text subsection. *"
2972 .ascii "But this lives in the second text subsection."
2974 .ascii "This lives in the data section,"
2975 .ascii "in the first data subsection."
2977 .ascii "This lives in the first text section,"
2978 .ascii "immediately following the asterisk (*)."
2981 Each section has a @dfn{location counter} incremented by one for every byte
2982 assembled into that section. Because subsections are merely a convenience
2983 restricted to @command{@value{AS}} there is no concept of a subsection location
2984 counter. There is no way to directly manipulate a location counter---but the
2985 @code{.align} directive changes it, and any label definition captures its
2986 current value. The location counter of the section where statements are being
2987 assembled is said to be the @dfn{active} location counter.
2990 @section bss Section
2993 @cindex common variable storage
2994 The bss section is used for local common variable storage.
2995 You may allocate address space in the bss section, but you may
2996 not dictate data to load into it before your program executes. When
2997 your program starts running, all the contents of the bss
2998 section are zeroed bytes.
3000 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3001 @ref{Lcomm,,@code{.lcomm}}.
3003 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3004 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
3007 When assembling for a target which supports multiple sections, such as ELF or
3008 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3009 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3010 section. Typically the section will only contain symbol definitions and
3011 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3018 Symbols are a central concept: the programmer uses symbols to name
3019 things, the linker uses symbols to link, and the debugger uses symbols
3023 @cindex debuggers, and symbol order
3024 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3025 the same order they were declared. This may break some debuggers.
3030 * Setting Symbols:: Giving Symbols Other Values
3031 * Symbol Names:: Symbol Names
3032 * Dot:: The Special Dot Symbol
3033 * Symbol Attributes:: Symbol Attributes
3040 A @dfn{label} is written as a symbol immediately followed by a colon
3041 @samp{:}. The symbol then represents the current value of the
3042 active location counter, and is, for example, a suitable instruction
3043 operand. You are warned if you use the same symbol to represent two
3044 different locations: the first definition overrides any other
3048 On the HPPA, the usual form for a label need not be immediately followed by a
3049 colon, but instead must start in column zero. Only one label may be defined on
3050 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3051 provides a special directive @code{.label} for defining labels more flexibly.
3054 @node Setting Symbols
3055 @section Giving Symbols Other Values
3057 @cindex assigning values to symbols
3058 @cindex symbol values, assigning
3059 A symbol can be given an arbitrary value by writing a symbol, followed
3060 by an equals sign @samp{=}, followed by an expression
3061 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3062 directive. @xref{Set,,@code{.set}}.
3065 @section Symbol Names
3067 @cindex symbol names
3068 @cindex names, symbol
3069 @ifclear SPECIAL-SYMS
3070 Symbol names begin with a letter or with one of @samp{._}. On most
3071 machines, you can also use @code{$} in symbol names; exceptions are
3072 noted in @ref{Machine Dependencies}. That character may be followed by any
3073 string of digits, letters, dollar signs (unless otherwise noted in
3074 @ref{Machine Dependencies}), and underscores.
3077 For the AMD 29K family, @samp{?} is also allowed in the
3078 body of a symbol name, though not at its beginning.
3083 Symbol names begin with a letter or with one of @samp{._}. On the
3085 H8/500, you can also use @code{$} in symbol names. That character may
3086 be followed by any string of digits, letters, dollar signs (save on the
3087 H8/300), and underscores.
3091 Case of letters is significant: @code{foo} is a different symbol name
3094 Each symbol has exactly one name. Each name in an assembly language program
3095 refers to exactly one symbol. You may use that symbol name any number of times
3098 @subheading Local Symbol Names
3100 @cindex local symbol names
3101 @cindex symbol names, local
3102 @cindex temporary symbol names
3103 @cindex symbol names, temporary
3104 Local symbols help compilers and programmers use names temporarily.
3105 They create symbols which are guaranteed to be unique over the entire scope of
3106 the input source code and which can be referred to by a simple notation.
3107 To define a local symbol, write a label of the form @samp{@b{N}:} (where @b{N}
3108 represents any positive integer). To refer to the most recent previous
3109 definition of that symbol write @samp{@b{N}b}, using the same number as when
3110 you defined the label. To refer to the next definition of a local label, write
3111 @samp{@b{N}f}--- The @samp{b} stands for``backwards'' and the @samp{f} stands
3114 There is no restriction on how you can use these labels, and you can reuse them
3115 too. So that it is possible to repeatedly define the same local label (using
3116 the same number @samp{@b{N}}), although you can only refer to the most recently
3117 defined local label of that number (for a backwards reference) or the next
3118 definition of a specific local label for a forward reference. It is also worth
3119 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3120 implemented in a slightly more efficient manner than the others.
3131 Which is the equivalent of:
3134 label_1: branch label_3
3135 label_2: branch label_1
3136 label_3: branch label_4
3137 label_4: branch label_3
3140 Local symbol names are only a notational device. They are immediately
3141 transformed into more conventional symbol names before the assembler uses them.
3142 The symbol names stored in the symbol table, appearing in error messages and
3143 optionally emitted to the object file. The names are constructed using these
3148 All local labels begin with @samp{L}. Normally both @command{@value{AS}} and
3149 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
3150 used for symbols you are never intended to see. If you use the
3151 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3152 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3153 you may use them in debugging.
3156 This is the number that was used in the local label definition. So if the
3157 label is written @samp{55:} then the number is @samp{55}.
3160 This unusual character is included so you do not accidentally invent a symbol
3161 of the same name. The character has ASCII value of @samp{\002} (control-B).
3163 @item @emph{ordinal number}
3164 This is a serial number to keep the labels distinct. The first definition of
3165 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3166 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3167 the number @samp{1} and its 15th defintion gets @samp{15} as well.
3170 So for example, the first @code{1:} is named @code{L1@kbd{C-B}1}, the 44th
3171 @code{3:} is named @code{L3@kbd{C-B}44}.
3173 @subheading Dollar Local Labels
3174 @cindex dollar local symbols
3176 @code{@value{AS}} also supports an even more local form of local labels called
3177 dollar labels. These labels go out of scope (ie they become undefined) as soon
3178 as a non-local label is defined. Thus they remain valid for only a small
3179 region of the input source code. Normal local labels, by contrast, remain in
3180 scope for the entire file, or until they are redefined by another occurrence of
3181 the same local label.
3183 Dollar labels are defined in exactly the same way as ordinary local labels,
3184 except that instead of being terminated by a colon, they are terminated by a
3185 dollar sign. eg @samp{@b{55$}}.
3187 They can also be distinguished from ordinary local labels by their transformed
3188 name which uses ASCII character @samp{\001} (control-A) as the magic character
3189 to distinguish them from ordinary labels. Thus the 5th defintion of @samp{6$}
3190 is named @samp{L6@kbd{C-A}5}.
3193 @section The Special Dot Symbol
3195 @cindex dot (symbol)
3196 @cindex @code{.} (symbol)
3197 @cindex current address
3198 @cindex location counter
3199 The special symbol @samp{.} refers to the current address that
3200 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3201 .long .} defines @code{melvin} to contain its own address.
3202 Assigning a value to @code{.} is treated the same as a @code{.org}
3203 directive. Thus, the expression @samp{.=.+4} is the same as saying
3204 @ifclear no-space-dir
3213 @node Symbol Attributes
3214 @section Symbol Attributes
3216 @cindex symbol attributes
3217 @cindex attributes, symbol
3218 Every symbol has, as well as its name, the attributes ``Value'' and
3219 ``Type''. Depending on output format, symbols can also have auxiliary
3222 The detailed definitions are in @file{a.out.h}.
3225 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3226 all these attributes, and probably won't warn you. This makes the
3227 symbol an externally defined symbol, which is generally what you
3231 * Symbol Value:: Value
3232 * Symbol Type:: Type
3235 * a.out Symbols:: Symbol Attributes: @code{a.out}
3239 * a.out Symbols:: Symbol Attributes: @code{a.out}
3242 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3247 * COFF Symbols:: Symbol Attributes for COFF
3250 * SOM Symbols:: Symbol Attributes for SOM
3257 @cindex value of a symbol
3258 @cindex symbol value
3259 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3260 location in the text, data, bss or absolute sections the value is the
3261 number of addresses from the start of that section to the label.
3262 Naturally for text, data and bss sections the value of a symbol changes
3263 as @code{@value{LD}} changes section base addresses during linking. Absolute
3264 symbols' values do not change during linking: that is why they are
3267 The value of an undefined symbol is treated in a special way. If it is
3268 0 then the symbol is not defined in this assembler source file, and
3269 @code{@value{LD}} tries to determine its value from other files linked into the
3270 same program. You make this kind of symbol simply by mentioning a symbol
3271 name without defining it. A non-zero value represents a @code{.comm}
3272 common declaration. The value is how much common storage to reserve, in
3273 bytes (addresses). The symbol refers to the first address of the
3279 @cindex type of a symbol
3281 The type attribute of a symbol contains relocation (section)
3282 information, any flag settings indicating that a symbol is external, and
3283 (optionally), other information for linkers and debuggers. The exact
3284 format depends on the object-code output format in use.
3289 @c The following avoids a "widow" subsection title. @group would be
3290 @c better if it were available outside examples.
3293 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3295 @cindex @code{b.out} symbol attributes
3296 @cindex symbol attributes, @code{b.out}
3297 These symbol attributes appear only when @command{@value{AS}} is configured for
3298 one of the Berkeley-descended object output formats---@code{a.out} or
3304 @subsection Symbol Attributes: @code{a.out}
3306 @cindex @code{a.out} symbol attributes
3307 @cindex symbol attributes, @code{a.out}
3313 @subsection Symbol Attributes: @code{a.out}
3315 @cindex @code{a.out} symbol attributes
3316 @cindex symbol attributes, @code{a.out}
3320 * Symbol Desc:: Descriptor
3321 * Symbol Other:: Other
3325 @subsubsection Descriptor
3327 @cindex descriptor, of @code{a.out} symbol
3328 This is an arbitrary 16-bit value. You may establish a symbol's
3329 descriptor value by using a @code{.desc} statement
3330 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3331 @command{@value{AS}}.
3334 @subsubsection Other
3336 @cindex other attribute, of @code{a.out} symbol
3337 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3342 @subsection Symbol Attributes for COFF
3344 @cindex COFF symbol attributes
3345 @cindex symbol attributes, COFF
3347 The COFF format supports a multitude of auxiliary symbol attributes;
3348 like the primary symbol attributes, they are set between @code{.def} and
3349 @code{.endef} directives.
3351 @subsubsection Primary Attributes
3353 @cindex primary attributes, COFF symbols
3354 The symbol name is set with @code{.def}; the value and type,
3355 respectively, with @code{.val} and @code{.type}.
3357 @subsubsection Auxiliary Attributes
3359 @cindex auxiliary attributes, COFF symbols
3360 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3361 @code{.size}, and @code{.tag} can generate auxiliary symbol table
3362 information for COFF.
3367 @subsection Symbol Attributes for SOM
3369 @cindex SOM symbol attributes
3370 @cindex symbol attributes, SOM
3372 The SOM format for the HPPA supports a multitude of symbol attributes set with
3373 the @code{.EXPORT} and @code{.IMPORT} directives.
3375 The attributes are described in @cite{HP9000 Series 800 Assembly
3376 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3377 @code{EXPORT} assembler directive documentation.
3381 @chapter Expressions
3385 @cindex numeric values
3386 An @dfn{expression} specifies an address or numeric value.
3387 Whitespace may precede and/or follow an expression.
3389 The result of an expression must be an absolute number, or else an offset into
3390 a particular section. If an expression is not absolute, and there is not
3391 enough information when @command{@value{AS}} sees the expression to know its
3392 section, a second pass over the source program might be necessary to interpret
3393 the expression---but the second pass is currently not implemented.
3394 @command{@value{AS}} aborts with an error message in this situation.
3397 * Empty Exprs:: Empty Expressions
3398 * Integer Exprs:: Integer Expressions
3402 @section Empty Expressions
3404 @cindex empty expressions
3405 @cindex expressions, empty
3406 An empty expression has no value: it is just whitespace or null.
3407 Wherever an absolute expression is required, you may omit the
3408 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3409 is compatible with other assemblers.
3412 @section Integer Expressions
3414 @cindex integer expressions
3415 @cindex expressions, integer
3416 An @dfn{integer expression} is one or more @emph{arguments} delimited
3417 by @emph{operators}.
3420 * Arguments:: Arguments
3421 * Operators:: Operators
3422 * Prefix Ops:: Prefix Operators
3423 * Infix Ops:: Infix Operators
3427 @subsection Arguments
3429 @cindex expression arguments
3430 @cindex arguments in expressions
3431 @cindex operands in expressions
3432 @cindex arithmetic operands
3433 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3434 contexts arguments are sometimes called ``arithmetic operands''. In
3435 this manual, to avoid confusing them with the ``instruction operands'' of
3436 the machine language, we use the term ``argument'' to refer to parts of
3437 expressions only, reserving the word ``operand'' to refer only to machine
3438 instruction operands.
3440 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3441 @var{section} is one of text, data, bss, absolute,
3442 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3445 Numbers are usually integers.
3447 A number can be a flonum or bignum. In this case, you are warned
3448 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3449 these 32 bits are an integer. You may write integer-manipulating
3450 instructions that act on exotic constants, compatible with other
3453 @cindex subexpressions
3454 Subexpressions are a left parenthesis @samp{(} followed by an integer
3455 expression, followed by a right parenthesis @samp{)}; or a prefix
3456 operator followed by an argument.
3459 @subsection Operators
3461 @cindex operators, in expressions
3462 @cindex arithmetic functions
3463 @cindex functions, in expressions
3464 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3465 operators are followed by an argument. Infix operators appear
3466 between their arguments. Operators may be preceded and/or followed by
3470 @subsection Prefix Operator
3472 @cindex prefix operators
3473 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3474 one argument, which must be absolute.
3476 @c the tex/end tex stuff surrounding this small table is meant to make
3477 @c it align, on the printed page, with the similar table in the next
3478 @c section (which is inside an enumerate).
3480 \global\advance\leftskip by \itemindent
3485 @dfn{Negation}. Two's complement negation.
3487 @dfn{Complementation}. Bitwise not.
3491 \global\advance\leftskip by -\itemindent
3495 @subsection Infix Operators
3497 @cindex infix operators
3498 @cindex operators, permitted arguments
3499 @dfn{Infix operators} take two arguments, one on either side. Operators
3500 have precedence, but operations with equal precedence are performed left
3501 to right. Apart from @code{+} or @option{-}, both arguments must be
3502 absolute, and the result is absolute.
3505 @cindex operator precedence
3506 @cindex precedence of operators
3513 @dfn{Multiplication}.
3516 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3523 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3527 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3531 Intermediate precedence
3536 @dfn{Bitwise Inclusive Or}.
3542 @dfn{Bitwise Exclusive Or}.
3545 @dfn{Bitwise Or Not}.
3552 @cindex addition, permitted arguments
3553 @cindex plus, permitted arguments
3554 @cindex arguments for addition
3556 @dfn{Addition}. If either argument is absolute, the result has the section of
3557 the other argument. You may not add together arguments from different
3560 @cindex subtraction, permitted arguments
3561 @cindex minus, permitted arguments
3562 @cindex arguments for subtraction
3564 @dfn{Subtraction}. If the right argument is absolute, the
3565 result has the section of the left argument.
3566 If both arguments are in the same section, the result is absolute.
3567 You may not subtract arguments from different sections.
3568 @c FIXME is there still something useful to say about undefined - undefined ?
3570 @cindex comparison expressions
3571 @cindex expressions, comparison
3575 @dfn{Is Not Equal To}
3579 @dfn{Is Greater Than}
3581 @dfn{Is Greater Than Or Equal To}
3583 @dfn{Is Less Than Or Equal To}
3585 The comparison operators can be used as infix operators. A true results has a
3586 value of -1 whereas a false result has a value of 0. Note, these operators
3587 perform signed comparisons.
3590 @item Lowest Precedence
3599 These two logical operations can be used to combine the results of sub
3600 expressions. Note, unlike the comparison operators a true result returns a
3601 value of 1 but a false results does still return 0. Also note that the logical
3602 or operator has a slightly lower precedence than logical and.
3607 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3608 address; you can only have a defined section in one of the two arguments.
3611 @chapter Assembler Directives
3613 @cindex directives, machine independent
3614 @cindex pseudo-ops, machine independent
3615 @cindex machine independent directives
3616 All assembler directives have names that begin with a period (@samp{.}).
3617 The rest of the name is letters, usually in lower case.
3619 This chapter discusses directives that are available regardless of the
3620 target machine configuration for the @sc{gnu} assembler.
3622 Some machine configurations provide additional directives.
3623 @xref{Machine Dependencies}.
3626 @ifset machine-directives
3627 @xref{Machine Dependencies} for additional directives.
3632 * Abort:: @code{.abort}
3634 * ABORT:: @code{.ABORT}
3637 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3638 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3639 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3640 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3641 * Byte:: @code{.byte @var{expressions}}
3642 * Comm:: @code{.comm @var{symbol} , @var{length} }
3643 * Data:: @code{.data @var{subsection}}
3645 * Def:: @code{.def @var{name}}
3648 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3654 * Double:: @code{.double @var{flonums}}
3655 * Eject:: @code{.eject}
3656 * Else:: @code{.else}
3657 * Elseif:: @code{.elseif}
3660 * Endef:: @code{.endef}
3663 * Endfunc:: @code{.endfunc}
3664 * Endif:: @code{.endif}
3665 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3666 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3668 * Exitm:: @code{.exitm}
3669 * Extern:: @code{.extern}
3670 * Fail:: @code{.fail}
3671 @ifclear no-file-dir
3672 * File:: @code{.file @var{string}}
3675 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3676 * Float:: @code{.float @var{flonums}}
3677 * Func:: @code{.func}
3678 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3680 * Hidden:: @code{.hidden @var{names}}
3683 * hword:: @code{.hword @var{expressions}}
3684 * Ident:: @code{.ident}
3685 * If:: @code{.if @var{absolute expression}}
3686 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3687 * Include:: @code{.include "@var{file}"}
3688 * Int:: @code{.int @var{expressions}}
3690 * Internal:: @code{.internal @var{names}}
3693 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3694 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3695 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3696 * Lflags:: @code{.lflags}
3697 @ifclear no-line-dir
3698 * Line:: @code{.line @var{line-number}}
3701 * Ln:: @code{.ln @var{line-number}}
3702 * Linkonce:: @code{.linkonce [@var{type}]}
3703 * List:: @code{.list}
3704 * Long:: @code{.long @var{expressions}}
3706 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3709 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3710 * MRI:: @code{.mri @var{val}}
3711 * Nolist:: @code{.nolist}
3712 * Octa:: @code{.octa @var{bignums}}
3713 * Org:: @code{.org @var{new-lc} , @var{fill}}
3714 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3716 * PopSection:: @code{.popsection}
3717 * Previous:: @code{.previous}
3720 * Print:: @code{.print @var{string}}
3722 * Protected:: @code{.protected @var{names}}
3725 * Psize:: @code{.psize @var{lines}, @var{columns}}
3726 * Purgem:: @code{.purgem @var{name}}
3728 * PushSection:: @code{.pushsection @var{name}}
3731 * Quad:: @code{.quad @var{bignums}}
3732 * Rept:: @code{.rept @var{count}}
3733 * Sbttl:: @code{.sbttl "@var{subheading}"}
3735 * Scl:: @code{.scl @var{class}}
3738 * Section:: @code{.section @var{name}}
3741 * Set:: @code{.set @var{symbol}, @var{expression}}
3742 * Short:: @code{.short @var{expressions}}
3743 * Single:: @code{.single @var{flonums}}
3745 * Size:: @code{.size [@var{name} , @var{expression}]}
3748 * Skip:: @code{.skip @var{size} , @var{fill}}
3749 * Sleb128:: @code{.sleb128 @var{expressions}}
3750 * Space:: @code{.space @var{size} , @var{fill}}
3752 * Stab:: @code{.stabd, .stabn, .stabs}
3755 * String:: @code{.string "@var{str}"}
3756 * Struct:: @code{.struct @var{expression}}
3758 * SubSection:: @code{.subsection}
3759 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3763 * Tag:: @code{.tag @var{structname}}
3766 * Text:: @code{.text @var{subsection}}
3767 * Title:: @code{.title "@var{heading}"}
3769 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3772 * Uleb128:: @code{.uleb128 @var{expressions}}
3774 * Val:: @code{.val @var{addr}}
3778 * Version:: @code{.version "@var{string}"}
3779 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3780 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3781 * Weak:: @code{.weak @var{names}}
3784 * Word:: @code{.word @var{expressions}}
3785 * Deprecated:: Deprecated Directives
3789 @section @code{.abort}
3791 @cindex @code{abort} directive
3792 @cindex stopping the assembly
3793 This directive stops the assembly immediately. It is for
3794 compatibility with other assemblers. The original idea was that the
3795 assembly language source would be piped into the assembler. If the sender
3796 of the source quit, it could use this directive tells @command{@value{AS}} to
3797 quit also. One day @code{.abort} will not be supported.
3801 @section @code{.ABORT}
3803 @cindex @code{ABORT} directive
3804 When producing COFF output, @command{@value{AS}} accepts this directive as a
3805 synonym for @samp{.abort}.
3808 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3814 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3816 @cindex padding the location counter
3817 @cindex @code{align} directive
3818 Pad the location counter (in the current subsection) to a particular storage
3819 boundary. The first expression (which must be absolute) is the alignment
3820 required, as described below.
3822 The second expression (also absolute) gives the fill value to be stored in the
3823 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3824 padding bytes are normally zero. However, on some systems, if the section is
3825 marked as containing code and the fill value is omitted, the space is filled
3826 with no-op instructions.
3828 The third expression is also absolute, and is also optional. If it is present,
3829 it is the maximum number of bytes that should be skipped by this alignment
3830 directive. If doing the alignment would require skipping more bytes than the
3831 specified maximum, then the alignment is not done at all. You can omit the
3832 fill value (the second argument) entirely by simply using two commas after the
3833 required alignment; this can be useful if you want the alignment to be filled
3834 with no-op instructions when appropriate.
3836 The way the required alignment is specified varies from system to system.
3837 For the a29k, hppa, m68k, m88k, w65, sparc, and Hitachi SH, and i386 using ELF
3839 the first expression is the
3840 alignment request in bytes. For example @samp{.align 8} advances
3841 the location counter until it is a multiple of 8. If the location counter
3842 is already a multiple of 8, no change is needed.
3844 For other systems, including the i386 using a.out format, and the arm and
3845 strongarm, it is the
3846 number of low-order zero bits the location counter must have after
3847 advancement. For example @samp{.align 3} advances the location
3848 counter until it a multiple of 8. If the location counter is already a
3849 multiple of 8, no change is needed.
3851 This inconsistency is due to the different behaviors of the various
3852 native assemblers for these systems which GAS must emulate.
3853 GAS also provides @code{.balign} and @code{.p2align} directives,
3854 described later, which have a consistent behavior across all
3855 architectures (but are specific to GAS).
3858 @section @code{.ascii "@var{string}"}@dots{}
3860 @cindex @code{ascii} directive
3861 @cindex string literals
3862 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3863 separated by commas. It assembles each string (with no automatic
3864 trailing zero byte) into consecutive addresses.
3867 @section @code{.asciz "@var{string}"}@dots{}
3869 @cindex @code{asciz} directive
3870 @cindex zero-terminated strings
3871 @cindex null-terminated strings
3872 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3873 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3876 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3878 @cindex padding the location counter given number of bytes
3879 @cindex @code{balign} directive
3880 Pad the location counter (in the current subsection) to a particular
3881 storage boundary. The first expression (which must be absolute) is the
3882 alignment request in bytes. For example @samp{.balign 8} advances
3883 the location counter until it is a multiple of 8. If the location counter
3884 is already a multiple of 8, no change is needed.
3886 The second expression (also absolute) gives the fill value to be stored in the
3887 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3888 padding bytes are normally zero. However, on some systems, if the section is
3889 marked as containing code and the fill value is omitted, the space is filled
3890 with no-op instructions.
3892 The third expression is also absolute, and is also optional. If it is present,
3893 it is the maximum number of bytes that should be skipped by this alignment
3894 directive. If doing the alignment would require skipping more bytes than the
3895 specified maximum, then the alignment is not done at all. You can omit the
3896 fill value (the second argument) entirely by simply using two commas after the
3897 required alignment; this can be useful if you want the alignment to be filled
3898 with no-op instructions when appropriate.
3900 @cindex @code{balignw} directive
3901 @cindex @code{balignl} directive
3902 The @code{.balignw} and @code{.balignl} directives are variants of the
3903 @code{.balign} directive. The @code{.balignw} directive treats the fill
3904 pattern as a two byte word value. The @code{.balignl} directives treats the
3905 fill pattern as a four byte longword value. For example, @code{.balignw
3906 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
3907 filled in with the value 0x368d (the exact placement of the bytes depends upon
3908 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
3912 @section @code{.byte @var{expressions}}
3914 @cindex @code{byte} directive
3915 @cindex integers, one byte
3916 @code{.byte} expects zero or more expressions, separated by commas.
3917 Each expression is assembled into the next byte.
3920 @section @code{.comm @var{symbol} , @var{length} }
3922 @cindex @code{comm} directive
3923 @cindex symbol, common
3924 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
3925 common symbol in one object file may be merged with a defined or common symbol
3926 of the same name in another object file. If @code{@value{LD}} does not see a
3927 definition for the symbol--just one or more common symbols--then it will
3928 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
3929 absolute expression. If @code{@value{LD}} sees multiple common symbols with
3930 the same name, and they do not all have the same size, it will allocate space
3931 using the largest size.
3934 When using ELF, the @code{.comm} directive takes an optional third argument.
3935 This is the desired alignment of the symbol, specified as a byte boundary (for
3936 example, an alignment of 16 means that the least significant 4 bits of the
3937 address should be zero). The alignment must be an absolute expression, and it
3938 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
3939 for the common symbol, it will use the alignment when placing the symbol. If
3940 no alignment is specified, @command{@value{AS}} will set the alignment to the
3941 largest power of two less than or equal to the size of the symbol, up to a
3946 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
3947 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
3951 @section @code{.data @var{subsection}}
3953 @cindex @code{data} directive
3954 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
3955 end of the data subsection numbered @var{subsection} (which is an
3956 absolute expression). If @var{subsection} is omitted, it defaults
3961 @section @code{.def @var{name}}
3963 @cindex @code{def} directive
3964 @cindex COFF symbols, debugging
3965 @cindex debugging COFF symbols
3966 Begin defining debugging information for a symbol @var{name}; the
3967 definition extends until the @code{.endef} directive is encountered.
3970 This directive is only observed when @command{@value{AS}} is configured for COFF
3971 format output; when producing @code{b.out}, @samp{.def} is recognized,
3978 @section @code{.desc @var{symbol}, @var{abs-expression}}
3980 @cindex @code{desc} directive
3981 @cindex COFF symbol descriptor
3982 @cindex symbol descriptor, COFF
3983 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
3984 to the low 16 bits of an absolute expression.
3987 The @samp{.desc} directive is not available when @command{@value{AS}} is
3988 configured for COFF output; it is only for @code{a.out} or @code{b.out}
3989 object format. For the sake of compatibility, @command{@value{AS}} accepts
3990 it, but produces no output, when configured for COFF.
3996 @section @code{.dim}
3998 @cindex @code{dim} directive
3999 @cindex COFF auxiliary symbol information
4000 @cindex auxiliary symbol information, COFF
4001 This directive is generated by compilers to include auxiliary debugging
4002 information in the symbol table. It is only permitted inside
4003 @code{.def}/@code{.endef} pairs.
4006 @samp{.dim} is only meaningful when generating COFF format output; when
4007 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4013 @section @code{.double @var{flonums}}
4015 @cindex @code{double} directive
4016 @cindex floating point numbers (double)
4017 @code{.double} expects zero or more flonums, separated by commas. It
4018 assembles floating point numbers.
4020 The exact kind of floating point numbers emitted depends on how
4021 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4025 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4026 in @sc{ieee} format.
4031 @section @code{.eject}
4033 @cindex @code{eject} directive
4034 @cindex new page, in listings
4035 @cindex page, in listings
4036 @cindex listing control: new page
4037 Force a page break at this point, when generating assembly listings.
4040 @section @code{.else}
4042 @cindex @code{else} directive
4043 @code{.else} is part of the @command{@value{AS}} support for conditional
4044 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
4045 of code to be assembled if the condition for the preceding @code{.if}
4049 @section @code{.elseif}
4051 @cindex @code{elseif} directive
4052 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4053 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
4054 @code{.if} block that would otherwise fill the entire @code{.else} section.
4057 @section @code{.end}
4059 @cindex @code{end} directive
4060 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4061 process anything in the file past the @code{.end} directive.
4065 @section @code{.endef}
4067 @cindex @code{endef} directive
4068 This directive flags the end of a symbol definition begun with
4072 @samp{.endef} is only meaningful when generating COFF format output; if
4073 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4074 directive but ignores it.
4079 @section @code{.endfunc}
4080 @cindex @code{endfunc} directive
4081 @code{.endfunc} marks the end of a function specified with @code{.func}.
4084 @section @code{.endif}
4086 @cindex @code{endif} directive
4087 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4088 it marks the end of a block of code that is only assembled
4089 conditionally. @xref{If,,@code{.if}}.
4092 @section @code{.equ @var{symbol}, @var{expression}}
4094 @cindex @code{equ} directive
4095 @cindex assigning values to symbols
4096 @cindex symbols, assigning values to
4097 This directive sets the value of @var{symbol} to @var{expression}.
4098 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
4101 The syntax for @code{equ} on the HPPA is
4102 @samp{@var{symbol} .equ @var{expression}}.
4106 @section @code{.equiv @var{symbol}, @var{expression}}
4107 @cindex @code{equiv} directive
4108 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4109 the assembler will signal an error if @var{symbol} is already defined. Note a
4110 symbol which has been referenced but not actually defined is considered to be
4113 Except for the contents of the error message, this is roughly equivalent to
4122 @section @code{.err}
4123 @cindex @code{err} directive
4124 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4125 message and, unless the @option{-Z} option was used, it will not generate an
4126 object file. This can be used to signal error an conditionally compiled code.
4129 @section @code{.exitm}
4130 Exit early from the current macro definition. @xref{Macro}.
4133 @section @code{.extern}
4135 @cindex @code{extern} directive
4136 @code{.extern} is accepted in the source program---for compatibility
4137 with other assemblers---but it is ignored. @command{@value{AS}} treats
4138 all undefined symbols as external.
4141 @section @code{.fail @var{expression}}
4143 @cindex @code{fail} directive
4144 Generates an error or a warning. If the value of the @var{expression} is 500
4145 or more, @command{@value{AS}} will print a warning message. If the value is less
4146 than 500, @command{@value{AS}} will print an error message. The message will
4147 include the value of @var{expression}. This can occasionally be useful inside
4148 complex nested macros or conditional assembly.
4150 @ifclear no-file-dir
4152 @section @code{.file @var{string}}
4154 @cindex @code{file} directive
4155 @cindex logical file name
4156 @cindex file name, logical
4157 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4158 file. @var{string} is the new file name. In general, the filename is
4159 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4160 to specify an empty file name, you must give the quotes--@code{""}. This
4161 statement may go away in future: it is only recognized to be compatible with
4162 old @command{@value{AS}} programs.
4164 In some configurations of @command{@value{AS}}, @code{.file} has already been
4165 removed to avoid conflicts with other assemblers. @xref{Machine Dependencies}.
4170 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4172 @cindex @code{fill} directive
4173 @cindex writing patterns in memory
4174 @cindex patterns, writing in memory
4175 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4176 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4177 may be zero or more. @var{Size} may be zero or more, but if it is
4178 more than 8, then it is deemed to have the value 8, compatible with
4179 other people's assemblers. The contents of each @var{repeat} bytes
4180 is taken from an 8-byte number. The highest order 4 bytes are
4181 zero. The lowest order 4 bytes are @var{value} rendered in the
4182 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4183 Each @var{size} bytes in a repetition is taken from the lowest order
4184 @var{size} bytes of this number. Again, this bizarre behavior is
4185 compatible with other people's assemblers.
4187 @var{size} and @var{value} are optional.
4188 If the second comma and @var{value} are absent, @var{value} is
4189 assumed zero. If the first comma and following tokens are absent,
4190 @var{size} is assumed to be 1.
4193 @section @code{.float @var{flonums}}
4195 @cindex floating point numbers (single)
4196 @cindex @code{float} directive
4197 This directive assembles zero or more flonums, separated by commas. It
4198 has the same effect as @code{.single}.
4200 The exact kind of floating point numbers emitted depends on how
4201 @command{@value{AS}} is configured.
4202 @xref{Machine Dependencies}.
4206 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4207 in @sc{ieee} format.
4212 @section @code{.func @var{name}[,@var{label}]}
4213 @cindex @code{func} directive
4214 @code{.func} emits debugging information to denote function @var{name}, and
4215 is ignored unless the file is assembled with debugging enabled.
4216 Only @samp{--gstabs} is currently supported.
4217 @var{label} is the entry point of the function and if omitted @var{name}
4218 prepended with the @samp{leading char} is used.
4219 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4220 All functions are currently defined to have @code{void} return type.
4221 The function must be terminated with @code{.endfunc}.
4224 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4226 @cindex @code{global} directive
4227 @cindex symbol, making visible to linker
4228 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4229 @var{symbol} in your partial program, its value is made available to
4230 other partial programs that are linked with it. Otherwise,
4231 @var{symbol} takes its attributes from a symbol of the same name
4232 from another file linked into the same program.
4234 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4235 compatibility with other assemblers.
4238 On the HPPA, @code{.global} is not always enough to make it accessible to other
4239 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4240 @xref{HPPA Directives,, HPPA Assembler Directives}.
4245 @section @code{.hidden @var{names}}
4247 @cindex @code{hidden} directive
4249 This one of the ELF visibility directives. The other two are
4250 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4251 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4253 This directive overrides the named symbols default visibility (which is set by
4254 their binding: local, global or weak). The directive sets the visibility to
4255 @code{hidden} which means that the symbols are not visible to other components.
4256 Such symbols are always considered to be @code{protected} as well.
4260 @section @code{.hword @var{expressions}}
4262 @cindex @code{hword} directive
4263 @cindex integers, 16-bit
4264 @cindex numbers, 16-bit
4265 @cindex sixteen bit integers
4266 This expects zero or more @var{expressions}, and emits
4267 a 16 bit number for each.
4270 This directive is a synonym for @samp{.short}; depending on the target
4271 architecture, it may also be a synonym for @samp{.word}.
4275 This directive is a synonym for @samp{.short}.
4278 This directive is a synonym for both @samp{.short} and @samp{.word}.
4283 @section @code{.ident}
4285 @cindex @code{ident} directive
4286 This directive is used by some assemblers to place tags in object files.
4287 @command{@value{AS}} simply accepts the directive for source-file
4288 compatibility with such assemblers, but does not actually emit anything
4292 @section @code{.if @var{absolute expression}}
4294 @cindex conditional assembly
4295 @cindex @code{if} directive
4296 @code{.if} marks the beginning of a section of code which is only
4297 considered part of the source program being assembled if the argument
4298 (which must be an @var{absolute expression}) is non-zero. The end of
4299 the conditional section of code must be marked by @code{.endif}
4300 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4301 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4302 If you have several conditions to check, @code{.elseif} may be used to avoid
4303 nesting blocks if/else within each subsequent @code{.else} block.
4305 The following variants of @code{.if} are also supported:
4307 @cindex @code{ifdef} directive
4308 @item .ifdef @var{symbol}
4309 Assembles the following section of code if the specified @var{symbol}
4310 has been defined. Note a symbol which has been referenced but not yet defined
4311 is considered to be undefined.
4313 @cindex @code{ifc} directive
4314 @item .ifc @var{string1},@var{string2}
4315 Assembles the following section of code if the two strings are the same. The
4316 strings may be optionally quoted with single quotes. If they are not quoted,
4317 the first string stops at the first comma, and the second string stops at the
4318 end of the line. Strings which contain whitespace should be quoted. The
4319 string comparison is case sensitive.
4321 @cindex @code{ifeq} directive
4322 @item .ifeq @var{absolute expression}
4323 Assembles the following section of code if the argument is zero.
4325 @cindex @code{ifeqs} directive
4326 @item .ifeqs @var{string1},@var{string2}
4327 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4329 @cindex @code{ifge} directive
4330 @item .ifge @var{absolute expression}
4331 Assembles the following section of code if the argument is greater than or
4334 @cindex @code{ifgt} directive
4335 @item .ifgt @var{absolute expression}
4336 Assembles the following section of code if the argument is greater than zero.
4338 @cindex @code{ifle} directive
4339 @item .ifle @var{absolute expression}
4340 Assembles the following section of code if the argument is less than or equal
4343 @cindex @code{iflt} directive
4344 @item .iflt @var{absolute expression}
4345 Assembles the following section of code if the argument is less than zero.
4347 @cindex @code{ifnc} directive
4348 @item .ifnc @var{string1},@var{string2}.
4349 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4350 following section of code if the two strings are not the same.
4352 @cindex @code{ifndef} directive
4353 @cindex @code{ifnotdef} directive
4354 @item .ifndef @var{symbol}
4355 @itemx .ifnotdef @var{symbol}
4356 Assembles the following section of code if the specified @var{symbol}
4357 has not been defined. Both spelling variants are equivalent. Note a symbol
4358 which has been referenced but not yet defined is considered to be undefined.
4360 @cindex @code{ifne} directive
4361 @item .ifne @var{absolute expression}
4362 Assembles the following section of code if the argument is not equal to zero
4363 (in other words, this is equivalent to @code{.if}).
4365 @cindex @code{ifnes} directive
4366 @item .ifnes @var{string1},@var{string2}
4367 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4368 following section of code if the two strings are not the same.
4372 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4374 @cindex @code{incbin} directive
4375 @cindex binary files, including
4376 The @code{incbin} directive includes @var{file} verbatim at the current
4377 location. You can control the search paths used with the @samp{-I} command-line
4378 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4381 The @var{skip} argument skips a number of bytes from the start of the
4382 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4383 read. Note that the data is not aligned in any way, so it is the user's
4384 responsibility to make sure that proper alignment is provided both before and
4385 after the @code{incbin} directive.
4388 @section @code{.include "@var{file}"}
4390 @cindex @code{include} directive
4391 @cindex supporting files, including
4392 @cindex files, including
4393 This directive provides a way to include supporting files at specified
4394 points in your source program. The code from @var{file} is assembled as
4395 if it followed the point of the @code{.include}; when the end of the
4396 included file is reached, assembly of the original file continues. You
4397 can control the search paths used with the @samp{-I} command-line option
4398 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4402 @section @code{.int @var{expressions}}
4404 @cindex @code{int} directive
4405 @cindex integers, 32-bit
4406 Expect zero or more @var{expressions}, of any section, separated by commas.
4407 For each expression, emit a number that, at run time, is the value of that
4408 expression. The byte order and bit size of the number depends on what kind
4409 of target the assembly is for.
4413 On the H8/500 and most forms of the H8/300, @code{.int} emits 16-bit
4414 integers. On the H8/300H and the Hitachi SH, however, @code{.int} emits
4421 @section @code{.internal @var{names}}
4423 @cindex @code{internal} directive
4425 This one of the ELF visibility directives. The other two are
4426 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) 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{internal} which means that the symbols are considered to be @code{hidden}
4432 (i.e., not visible to other components), and that some extra, processor specific
4433 processing must also be performed upon the symbols as well.
4437 @section @code{.irp @var{symbol},@var{values}}@dots{}
4439 @cindex @code{irp} directive
4440 Evaluate a sequence of statements assigning different values to @var{symbol}.
4441 The sequence of statements starts at the @code{.irp} directive, and is
4442 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4443 set to @var{value}, and the sequence of statements is assembled. If no
4444 @var{value} is listed, the sequence of statements is assembled once, with
4445 @var{symbol} set to the null string. To refer to @var{symbol} within the
4446 sequence of statements, use @var{\symbol}.
4448 For example, assembling
4456 is equivalent to assembling
4465 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4467 @cindex @code{irpc} directive
4468 Evaluate a sequence of statements assigning different values to @var{symbol}.
4469 The sequence of statements starts at the @code{.irpc} directive, and is
4470 terminated by an @code{.endr} directive. For each character in @var{value},
4471 @var{symbol} is set to the character, and the sequence of statements is
4472 assembled. If no @var{value} is listed, the sequence of statements is
4473 assembled once, with @var{symbol} set to the null string. To refer to
4474 @var{symbol} within the sequence of statements, use @var{\symbol}.
4476 For example, assembling
4484 is equivalent to assembling
4493 @section @code{.lcomm @var{symbol} , @var{length}}
4495 @cindex @code{lcomm} directive
4496 @cindex local common symbols
4497 @cindex symbols, local common
4498 Reserve @var{length} (an absolute expression) bytes for a local common
4499 denoted by @var{symbol}. The section and value of @var{symbol} are
4500 those of the new local common. The addresses are allocated in the bss
4501 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4502 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4503 not visible to @code{@value{LD}}.
4506 Some targets permit a third argument to be used with @code{.lcomm}. This
4507 argument specifies the desired alignment of the symbol in the bss section.
4511 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4512 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4516 @section @code{.lflags}
4518 @cindex @code{lflags} directive (ignored)
4519 @command{@value{AS}} accepts this directive, for compatibility with other
4520 assemblers, but ignores it.
4522 @ifclear no-line-dir
4524 @section @code{.line @var{line-number}}
4526 @cindex @code{line} directive
4530 @section @code{.ln @var{line-number}}
4532 @cindex @code{ln} directive
4534 @cindex logical line number
4536 Change the logical line number. @var{line-number} must be an absolute
4537 expression. The next line has that logical line number. Therefore any other
4538 statements on the current line (after a statement separator character) are
4539 reported as on logical line number @var{line-number} @minus{} 1. One day
4540 @command{@value{AS}} will no longer support this directive: it is recognized only
4541 for compatibility with existing assembler programs.
4545 @emph{Warning:} In the AMD29K configuration of @value{AS}, this command is
4546 not available; use the synonym @code{.ln} in that context.
4551 @ifclear no-line-dir
4552 Even though this is a directive associated with the @code{a.out} or
4553 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4554 when producing COFF output, and treats @samp{.line} as though it
4555 were the COFF @samp{.ln} @emph{if} it is found outside a
4556 @code{.def}/@code{.endef} pair.
4558 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4559 used by compilers to generate auxiliary symbol information for
4564 @section @code{.linkonce [@var{type}]}
4566 @cindex @code{linkonce} directive
4567 @cindex common sections
4568 Mark the current section so that the linker only includes a single copy of it.
4569 This may be used to include the same section in several different object files,
4570 but ensure that the linker will only include it once in the final output file.
4571 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4572 Duplicate sections are detected based on the section name, so it should be
4575 This directive is only supported by a few object file formats; as of this
4576 writing, the only object file format which supports it is the Portable
4577 Executable format used on Windows NT.
4579 The @var{type} argument is optional. If specified, it must be one of the
4580 following strings. For example:
4584 Not all types may be supported on all object file formats.
4588 Silently discard duplicate sections. This is the default.
4591 Warn if there are duplicate sections, but still keep only one copy.
4594 Warn if any of the duplicates have different sizes.
4597 Warn if any of the duplicates do not have exactly the same contents.
4601 @section @code{.ln @var{line-number}}
4603 @cindex @code{ln} directive
4604 @ifclear no-line-dir
4605 @samp{.ln} is a synonym for @samp{.line}.
4608 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4609 must be an absolute expression. The next line has that logical
4610 line number, so any other statements on the current line (after a
4611 statement separator character @code{;}) are reported as on logical
4612 line number @var{line-number} @minus{} 1.
4615 This directive is accepted, but ignored, when @command{@value{AS}} is
4616 configured for @code{b.out}; its effect is only associated with COFF
4622 @section @code{.mri @var{val}}
4624 @cindex @code{mri} directive
4625 @cindex MRI mode, temporarily
4626 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4627 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4628 affects code assembled until the next @code{.mri} directive, or until the end
4629 of the file. @xref{M, MRI mode, MRI mode}.
4632 @section @code{.list}
4634 @cindex @code{list} directive
4635 @cindex listing control, turning on
4636 Control (in conjunction with the @code{.nolist} directive) whether or
4637 not assembly listings are generated. These two directives maintain an
4638 internal counter (which is zero initially). @code{.list} increments the
4639 counter, and @code{.nolist} decrements it. Assembly listings are
4640 generated whenever the counter is greater than zero.
4642 By default, listings are disabled. When you enable them (with the
4643 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4644 the initial value of the listing counter is one.
4647 @section @code{.long @var{expressions}}
4649 @cindex @code{long} directive
4650 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4653 @c no one seems to know what this is for or whether this description is
4654 @c what it really ought to do
4656 @section @code{.lsym @var{symbol}, @var{expression}}
4658 @cindex @code{lsym} directive
4659 @cindex symbol, not referenced in assembly
4660 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4661 the hash table, ensuring it cannot be referenced by name during the
4662 rest of the assembly. This sets the attributes of the symbol to be
4663 the same as the expression value:
4665 @var{other} = @var{descriptor} = 0
4666 @var{type} = @r{(section of @var{expression})}
4667 @var{value} = @var{expression}
4670 The new symbol is not flagged as external.
4674 @section @code{.macro}
4677 The commands @code{.macro} and @code{.endm} allow you to define macros that
4678 generate assembly output. For example, this definition specifies a macro
4679 @code{sum} that puts a sequence of numbers into memory:
4682 .macro sum from=0, to=5
4691 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4703 @item .macro @var{macname}
4704 @itemx .macro @var{macname} @var{macargs} @dots{}
4705 @cindex @code{macro} directive
4706 Begin the definition of a macro called @var{macname}. If your macro
4707 definition requires arguments, specify their names after the macro name,
4708 separated by commas or spaces. You can supply a default value for any
4709 macro argument by following the name with @samp{=@var{deflt}}. For
4710 example, these are all valid @code{.macro} statements:
4714 Begin the definition of a macro called @code{comm}, which takes no
4717 @item .macro plus1 p, p1
4718 @itemx .macro plus1 p p1
4719 Either statement begins the definition of a macro called @code{plus1},
4720 which takes two arguments; within the macro definition, write
4721 @samp{\p} or @samp{\p1} to evaluate the arguments.
4723 @item .macro reserve_str p1=0 p2
4724 Begin the definition of a macro called @code{reserve_str}, with two
4725 arguments. The first argument has a default value, but not the second.
4726 After the definition is complete, you can call the macro either as
4727 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
4728 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
4729 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
4730 @samp{0}, and @samp{\p2} evaluating to @var{b}).
4733 When you call a macro, you can specify the argument values either by
4734 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
4735 @samp{sum to=17, from=9}.
4738 @cindex @code{endm} directive
4739 Mark the end of a macro definition.
4742 @cindex @code{exitm} directive
4743 Exit early from the current macro definition.
4745 @cindex number of macros executed
4746 @cindex macros, count executed
4748 @command{@value{AS}} maintains a counter of how many macros it has
4749 executed in this pseudo-variable; you can copy that number to your
4750 output with @samp{\@@}, but @emph{only within a macro definition}.
4753 @item LOCAL @var{name} [ , @dots{} ]
4754 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
4755 macro syntax'' with @samp{-a} or @samp{--alternate}.} @xref{Alternate,,
4756 Alternate macro syntax}.
4758 Generate a string replacement for each of the @var{name} arguments, and
4759 replace any instances of @var{name} in each macro expansion. The
4760 replacement string is unique in the assembly, and different for each
4761 separate macro expansion. @code{LOCAL} allows you to write macros that
4762 define symbols, without fear of conflict between separate macro expansions.
4767 @section @code{.nolist}
4769 @cindex @code{nolist} directive
4770 @cindex listing control, turning off
4771 Control (in conjunction with the @code{.list} directive) whether or
4772 not assembly listings are generated. These two directives maintain an
4773 internal counter (which is zero initially). @code{.list} increments the
4774 counter, and @code{.nolist} decrements it. Assembly listings are
4775 generated whenever the counter is greater than zero.
4778 @section @code{.octa @var{bignums}}
4780 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
4781 @cindex @code{octa} directive
4782 @cindex integer, 16-byte
4783 @cindex sixteen byte integer
4784 This directive expects zero or more bignums, separated by commas. For each
4785 bignum, it emits a 16-byte integer.
4787 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
4788 hence @emph{octa}-word for 16 bytes.
4791 @section @code{.org @var{new-lc} , @var{fill}}
4793 @cindex @code{org} directive
4794 @cindex location counter, advancing
4795 @cindex advancing location counter
4796 @cindex current address, advancing
4797 Advance the location counter of the current section to
4798 @var{new-lc}. @var{new-lc} is either an absolute expression or an
4799 expression with the same section as the current subsection. That is,
4800 you can't use @code{.org} to cross sections: if @var{new-lc} has the
4801 wrong section, the @code{.org} directive is ignored. To be compatible
4802 with former assemblers, if the section of @var{new-lc} is absolute,
4803 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
4804 is the same as the current subsection.
4806 @code{.org} may only increase the location counter, or leave it
4807 unchanged; you cannot use @code{.org} to move the location counter
4810 @c double negative used below "not undefined" because this is a specific
4811 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
4812 @c section. doc@cygnus.com 18feb91
4813 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
4814 may not be undefined. If you really detest this restriction we eagerly await
4815 a chance to share your improved assembler.
4817 Beware that the origin is relative to the start of the section, not
4818 to the start of the subsection. This is compatible with other
4819 people's assemblers.
4821 When the location counter (of the current subsection) is advanced, the
4822 intervening bytes are filled with @var{fill} which should be an
4823 absolute expression. If the comma and @var{fill} are omitted,
4824 @var{fill} defaults to zero.
4827 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4829 @cindex padding the location counter given a power of two
4830 @cindex @code{p2align} directive
4831 Pad the location counter (in the current subsection) to a particular
4832 storage boundary. The first expression (which must be absolute) is the
4833 number of low-order zero bits the location counter must have after
4834 advancement. For example @samp{.p2align 3} advances the location
4835 counter until it a multiple of 8. If the location counter is already a
4836 multiple of 8, no change is needed.
4838 The second expression (also absolute) gives the fill value to be stored in the
4839 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4840 padding bytes are normally zero. However, on some systems, if the section is
4841 marked as containing code and the fill value is omitted, the space is filled
4842 with no-op instructions.
4844 The third expression is also absolute, and is also optional. If it is present,
4845 it is the maximum number of bytes that should be skipped by this alignment
4846 directive. If doing the alignment would require skipping more bytes than the
4847 specified maximum, then the alignment is not done at all. You can omit the
4848 fill value (the second argument) entirely by simply using two commas after the
4849 required alignment; this can be useful if you want the alignment to be filled
4850 with no-op instructions when appropriate.
4852 @cindex @code{p2alignw} directive
4853 @cindex @code{p2alignl} directive
4854 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
4855 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
4856 pattern as a two byte word value. The @code{.p2alignl} directives treats the
4857 fill pattern as a four byte longword value. For example, @code{.p2alignw
4858 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4859 filled in with the value 0x368d (the exact placement of the bytes depends upon
4860 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4865 @section @code{.previous}
4867 @cindex @code{previous} directive
4868 @cindex Section Stack
4869 This is one of the ELF section stack manipulation directives. The others are
4870 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
4871 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
4872 (@pxref{PopSection}).
4874 This directive swaps the current section (and subsection) with most recently
4875 referenced section (and subsection) prior to this one. Multiple
4876 @code{.previous} directives in a row will flip between two sections (and their
4879 In terms of the section stack, this directive swaps the current section with
4880 the top section on the section stack.
4885 @section @code{.popsection}
4887 @cindex @code{popsection} directive
4888 @cindex Section Stack
4889 This is one of the ELF section stack manipulation directives. The others are
4890 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
4891 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
4894 This directive replaces the current section (and subsection) with the top
4895 section (and subsection) on the section stack. This section is popped off the
4900 @section @code{.print @var{string}}
4902 @cindex @code{print} directive
4903 @command{@value{AS}} will print @var{string} on the standard output during
4904 assembly. You must put @var{string} in double quotes.
4908 @section @code{.protected @var{names}}
4910 @cindex @code{protected} directive
4912 This one of the ELF visibility directives. The other two are
4913 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
4915 This directive overrides the named symbols default visibility (which is set by
4916 their binding: local, global or weak). The directive sets the visibility to
4917 @code{protected} which means that any references to the symbols from within the
4918 components that defines them must be resolved to the definition in that
4919 component, even if a definition in another component would normally preempt
4924 @section @code{.psize @var{lines} , @var{columns}}
4926 @cindex @code{psize} directive
4927 @cindex listing control: paper size
4928 @cindex paper size, for listings
4929 Use this directive to declare the number of lines---and, optionally, the
4930 number of columns---to use for each page, when generating listings.
4932 If you do not use @code{.psize}, listings use a default line-count
4933 of 60. You may omit the comma and @var{columns} specification; the
4934 default width is 200 columns.
4936 @command{@value{AS}} generates formfeeds whenever the specified number of
4937 lines is exceeded (or whenever you explicitly request one, using
4940 If you specify @var{lines} as @code{0}, no formfeeds are generated save
4941 those explicitly specified with @code{.eject}.
4944 @section @code{.purgem @var{name}}
4946 @cindex @code{purgem} directive
4947 Undefine the macro @var{name}, so that later uses of the string will not be
4948 expanded. @xref{Macro}.
4952 @section @code{.pushsection @var{name} , @var{subsection}}
4954 @cindex @code{pushsection} directive
4955 @cindex Section Stack
4956 This is one of the ELF section stack manipulation directives. The others are
4957 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
4958 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
4961 This directive is a synonym for @code{.section}. It pushes the current section
4962 (and subsection) onto the top of the section stack, and then replaces the
4963 current section and subsection with @code{name} and @code{subsection}.
4967 @section @code{.quad @var{bignums}}
4969 @cindex @code{quad} directive
4970 @code{.quad} expects zero or more bignums, separated by commas. For
4971 each bignum, it emits
4973 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
4974 warning message; and just takes the lowest order 8 bytes of the bignum.
4975 @cindex eight-byte integer
4976 @cindex integer, 8-byte
4978 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
4979 hence @emph{quad}-word for 8 bytes.
4982 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
4983 warning message; and just takes the lowest order 16 bytes of the bignum.
4984 @cindex sixteen-byte integer
4985 @cindex integer, 16-byte
4989 @section @code{.rept @var{count}}
4991 @cindex @code{rept} directive
4992 Repeat the sequence of lines between the @code{.rept} directive and the next
4993 @code{.endr} directive @var{count} times.
4995 For example, assembling
5003 is equivalent to assembling
5012 @section @code{.sbttl "@var{subheading}"}
5014 @cindex @code{sbttl} directive
5015 @cindex subtitles for listings
5016 @cindex listing control: subtitle
5017 Use @var{subheading} as the title (third line, immediately after the
5018 title line) when generating assembly listings.
5020 This directive affects subsequent pages, as well as the current page if
5021 it appears within ten lines of the top of a page.
5025 @section @code{.scl @var{class}}
5027 @cindex @code{scl} directive
5028 @cindex symbol storage class (COFF)
5029 @cindex COFF symbol storage class
5030 Set the storage-class value for a symbol. This directive may only be
5031 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5032 whether a symbol is static or external, or it may record further
5033 symbolic debugging information.
5036 The @samp{.scl} directive is primarily associated with COFF output; when
5037 configured to generate @code{b.out} output format, @command{@value{AS}}
5038 accepts this directive but ignores it.
5044 @section @code{.section @var{name}}
5046 @cindex named section
5047 Use the @code{.section} directive to assemble the following code into a section
5050 This directive is only supported for targets that actually support arbitrarily
5051 named sections; on @code{a.out} targets, for example, it is not accepted, even
5052 with a standard @code{a.out} section name.
5056 @c only print the extra heading if both COFF and ELF are set
5057 @subheading COFF Version
5060 @cindex @code{section} directive (COFF version)
5061 For COFF targets, the @code{.section} directive is used in one of the following
5065 .section @var{name}[, "@var{flags}"]
5066 .section @var{name}[, @var{subsegment}]
5069 If the optional argument is quoted, it is taken as flags to use for the
5070 section. Each flag is a single character. The following flags are recognized:
5073 bss section (uninitialized data)
5075 section is not loaded
5085 shared section (meaningful for PE targets)
5087 ignored. (For compatibility with the ELF version)
5090 If no flags are specified, the default flags depend upon the section name. If
5091 the section name is not recognized, the default will be for the section to be
5092 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5093 from the section, rather than adding them, so if they are used on their own it
5094 will be as if no flags had been specified at all.
5096 If the optional argument to the @code{.section} directive is not quoted, it is
5097 taken as a subsegment number (@pxref{Sub-Sections}).
5102 @c only print the extra heading if both COFF and ELF are set
5103 @subheading ELF Version
5106 @cindex Section Stack
5107 This is one of the ELF section stack manipulation directives. The others are
5108 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5109 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5110 @code{.previous} (@pxref{Previous}).
5112 @cindex @code{section} directive (ELF version)
5113 For ELF targets, the @code{.section} directive is used like this:
5116 .section @var{name} [, "@var{flags}"[, @@@var{type}[, @@@var{entsize}]]]
5119 The optional @var{flags} argument is a quoted string which may contain any
5120 combination of the following characters:
5123 section is allocatable
5127 section is executable
5129 section is mergeable
5131 section contains zero terminated strings
5134 The optional @var{type} argument may contain one of the following constants:
5137 section contains data
5139 section does not contain data (i.e., section only occupies space)
5142 Note on targets where the @code{@@} character is the start of a comment (eg
5143 ARM) then another character is used instead. For example the ARM port uses the
5146 If @var{flags} contains @code{M} flag, @var{type} argument must be specified
5147 as well as @var{entsize} argument. Sections with @code{M} flag but not
5148 @code{S} flag must contain fixed size constants, each @var{entsize} octets
5149 long. Sections with both @code{M} and @code{S} must contain zero terminated
5150 strings where each character is @var{entsize} bytes long. The linker may remove
5151 duplicates within sections with the same name, same entity size and same flags.
5153 If no flags are specified, the default flags depend upon the section name. If
5154 the section name is not recognized, the default will be for the section to have
5155 none of the above flags: it will not be allocated in memory, nor writable, nor
5156 executable. The section will contain data.
5158 For ELF targets, the assembler supports another type of @code{.section}
5159 directive for compatibility with the Solaris assembler:
5162 .section "@var{name}"[, @var{flags}...]
5165 Note that the section name is quoted. There may be a sequence of comma
5169 section is allocatable
5173 section is executable
5176 This directive replaces the current section and subsection. The replaced
5177 section and subsection are pushed onto the section stack. See the contents of
5178 the gas testsuite directory @code{gas/testsuite/gas/elf} for some examples of
5179 how this directive and the other section stack directives work.
5184 @section @code{.set @var{symbol}, @var{expression}}
5186 @cindex @code{set} directive
5187 @cindex symbol value, setting
5188 Set the value of @var{symbol} to @var{expression}. This
5189 changes @var{symbol}'s value and type to conform to
5190 @var{expression}. If @var{symbol} was flagged as external, it remains
5191 flagged (@pxref{Symbol Attributes}).
5193 You may @code{.set} a symbol many times in the same assembly.
5195 If you @code{.set} a global symbol, the value stored in the object
5196 file is the last value stored into it.
5199 The syntax for @code{set} on the HPPA is
5200 @samp{@var{symbol} .set @var{expression}}.
5204 @section @code{.short @var{expressions}}
5206 @cindex @code{short} directive
5208 @code{.short} is normally the same as @samp{.word}.
5209 @xref{Word,,@code{.word}}.
5211 In some configurations, however, @code{.short} and @code{.word} generate
5212 numbers of different lengths; @pxref{Machine Dependencies}.
5216 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5219 This expects zero or more @var{expressions}, and emits
5220 a 16 bit number for each.
5225 @section @code{.single @var{flonums}}
5227 @cindex @code{single} directive
5228 @cindex floating point numbers (single)
5229 This directive assembles zero or more flonums, separated by commas. It
5230 has the same effect as @code{.float}.
5232 The exact kind of floating point numbers emitted depends on how
5233 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5237 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5238 numbers in @sc{ieee} format.
5244 @section @code{.size}
5246 This directive is used to set the size associated with a symbol.
5250 @c only print the extra heading if both COFF and ELF are set
5251 @subheading COFF Version
5254 @cindex @code{size} directive (COFF version)
5255 For COFF targets, the @code{.size} directive is only permitted inside
5256 @code{.def}/@code{.endef} pairs. It is used like this:
5259 .size @var{expression}
5263 @samp{.size} is only meaningful when generating COFF format output; when
5264 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5271 @c only print the extra heading if both COFF and ELF are set
5272 @subheading ELF Version
5275 @cindex @code{size} directive (ELF version)
5276 For ELF targets, the @code{.size} directive is used like this:
5279 .size @var{name} , @var{expression}
5282 This directive sets the size associated with a symbol @var{name}.
5283 The size in bytes is computed from @var{expression} which can make use of label
5284 arithmetic. This directive is typically used to set the size of function
5290 @section @code{.sleb128 @var{expressions}}
5292 @cindex @code{sleb128} directive
5293 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5294 compact, variable length representation of numbers used by the DWARF
5295 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
5297 @ifclear no-space-dir
5299 @section @code{.skip @var{size} , @var{fill}}
5301 @cindex @code{skip} directive
5302 @cindex filling memory
5303 This directive emits @var{size} bytes, each of value @var{fill}. Both
5304 @var{size} and @var{fill} are absolute expressions. If the comma and
5305 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5309 @section @code{.space @var{size} , @var{fill}}
5311 @cindex @code{space} directive
5312 @cindex filling memory
5313 This directive emits @var{size} bytes, each of value @var{fill}. Both
5314 @var{size} and @var{fill} are absolute expressions. If the comma
5315 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5320 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5321 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5322 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5323 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5332 @section @code{.space}
5333 @cindex @code{space} directive
5335 On the AMD 29K, this directive is ignored; it is accepted for
5336 compatibility with other AMD 29K assemblers.
5339 @emph{Warning:} In most versions of the @sc{gnu} assembler, the directive
5340 @code{.space} has the effect of @code{.block} @xref{Machine Dependencies}.
5346 @section @code{.stabd, .stabn, .stabs}
5348 @cindex symbolic debuggers, information for
5349 @cindex @code{stab@var{x}} directives
5350 There are three directives that begin @samp{.stab}.
5351 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5352 The symbols are not entered in the @command{@value{AS}} hash table: they
5353 cannot be referenced elsewhere in the source file.
5354 Up to five fields are required:
5358 This is the symbol's name. It may contain any character except
5359 @samp{\000}, so is more general than ordinary symbol names. Some
5360 debuggers used to code arbitrarily complex structures into symbol names
5364 An absolute expression. The symbol's type is set to the low 8 bits of
5365 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5366 and debuggers choke on silly bit patterns.
5369 An absolute expression. The symbol's ``other'' attribute is set to the
5370 low 8 bits of this expression.
5373 An absolute expression. The symbol's descriptor is set to the low 16
5374 bits of this expression.
5377 An absolute expression which becomes the symbol's value.
5380 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5381 or @code{.stabs} statement, the symbol has probably already been created;
5382 you get a half-formed symbol in your object file. This is
5383 compatible with earlier assemblers!
5386 @cindex @code{stabd} directive
5387 @item .stabd @var{type} , @var{other} , @var{desc}
5389 The ``name'' of the symbol generated is not even an empty string.
5390 It is a null pointer, for compatibility. Older assemblers used a
5391 null pointer so they didn't waste space in object files with empty
5394 The symbol's value is set to the location counter,
5395 relocatably. When your program is linked, the value of this symbol
5396 is the address of the location counter when the @code{.stabd} was
5399 @cindex @code{stabn} directive
5400 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5401 The name of the symbol is set to the empty string @code{""}.
5403 @cindex @code{stabs} directive
5404 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5405 All five fields are specified.
5411 @section @code{.string} "@var{str}"
5413 @cindex string, copying to object file
5414 @cindex @code{string} directive
5416 Copy the characters in @var{str} to the object file. You may specify more than
5417 one string to copy, separated by commas. Unless otherwise specified for a
5418 particular machine, the assembler marks the end of each string with a 0 byte.
5419 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5422 @section @code{.struct @var{expression}}
5424 @cindex @code{struct} directive
5425 Switch to the absolute section, and set the section offset to @var{expression},
5426 which must be an absolute expression. You might use this as follows:
5435 This would define the symbol @code{field1} to have the value 0, the symbol
5436 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5437 value 8. Assembly would be left in the absolute section, and you would need to
5438 use a @code{.section} directive of some sort to change to some other section
5439 before further assembly.
5443 @section @code{.subsection @var{name}}
5445 @cindex @code{subsection} directive
5446 @cindex Section Stack
5447 This is one of the ELF section stack manipulation directives. The others are
5448 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5449 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5452 This directive replaces the current subsection with @code{name}. The current
5453 section is not changed. The replaced subsection is put onto the section stack
5454 in place of the then current top of stack subsection.
5459 @section @code{.symver}
5460 @cindex @code{symver} directive
5461 @cindex symbol versioning
5462 @cindex versions of symbols
5463 Use the @code{.symver} directive to bind symbols to specific version nodes
5464 within a source file. This is only supported on ELF platforms, and is
5465 typically used when assembling files to be linked into a shared library.
5466 There are cases where it may make sense to use this in objects to be bound
5467 into an application itself so as to override a versioned symbol from a
5470 For ELF targets, the @code{.symver} directive can be used like this:
5472 .symver @var{name}, @var{name2@@nodename}
5474 If the symbol @var{name} is defined within the file
5475 being assembled, the @code{.symver} directive effectively creates a symbol
5476 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5477 just don't try and create a regular alias is that the @var{@@} character isn't
5478 permitted in symbol names. The @var{name2} part of the name is the actual name
5479 of the symbol by which it will be externally referenced. The name @var{name}
5480 itself is merely a name of convenience that is used so that it is possible to
5481 have definitions for multiple versions of a function within a single source
5482 file, and so that the compiler can unambiguously know which version of a
5483 function is being mentioned. The @var{nodename} portion of the alias should be
5484 the name of a node specified in the version script supplied to the linker when
5485 building a shared library. If you are attempting to override a versioned
5486 symbol from a shared library, then @var{nodename} should correspond to the
5487 nodename of the symbol you are trying to override.
5489 If the symbol @var{name} is not defined within the file being assembled, all
5490 references to @var{name} will be changed to @var{name2@@nodename}. If no
5491 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5494 Another usage of the @code{.symver} directive is:
5496 .symver @var{name}, @var{name2@@@@nodename}
5498 In this case, the symbol @var{name} must exist and be defined within
5499 the file being assembled. It is similar to @var{name2@@nodename}. The
5500 difference is @var{name2@@@@nodename} will also be used to resolve
5501 references to @var{name2} by the linker.
5503 The third usage of the @code{.symver} directive is:
5505 .symver @var{name}, @var{name2@@@@@@nodename}
5507 When @var{name} is not defined within the
5508 file being assembled, it is treated as @var{name2@@nodename}. When
5509 @var{name} is defined within the file being assembled, the symbol
5510 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5515 @section @code{.tag @var{structname}}
5517 @cindex COFF structure debugging
5518 @cindex structure debugging, COFF
5519 @cindex @code{tag} directive
5520 This directive is generated by compilers to include auxiliary debugging
5521 information in the symbol table. It is only permitted inside
5522 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5523 definitions in the symbol table with instances of those structures.
5526 @samp{.tag} is only used when generating COFF format output; when
5527 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5533 @section @code{.text @var{subsection}}
5535 @cindex @code{text} directive
5536 Tells @command{@value{AS}} to assemble the following statements onto the end of
5537 the text subsection numbered @var{subsection}, which is an absolute
5538 expression. If @var{subsection} is omitted, subsection number zero
5542 @section @code{.title "@var{heading}"}
5544 @cindex @code{title} directive
5545 @cindex listing control: title line
5546 Use @var{heading} as the title (second line, immediately after the
5547 source file name and pagenumber) when generating assembly listings.
5549 This directive affects subsequent pages, as well as the current page if
5550 it appears within ten lines of the top of a page.
5554 @section @code{.type}
5556 This directive is used to set the type of a symbol.
5560 @c only print the extra heading if both COFF and ELF are set
5561 @subheading COFF Version
5564 @cindex COFF symbol type
5565 @cindex symbol type, COFF
5566 @cindex @code{type} directive (COFF version)
5567 For COFF targets, this directive is permitted only within
5568 @code{.def}/@code{.endef} pairs. It is used like this:
5574 This records the integer @var{int} as the type attribute of a symbol table
5578 @samp{.type} is associated only with COFF format output; when
5579 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
5580 directive but ignores it.
5586 @c only print the extra heading if both COFF and ELF are set
5587 @subheading ELF Version
5590 @cindex ELF symbol type
5591 @cindex symbol type, ELF
5592 @cindex @code{type} directive (ELF version)
5593 For ELF targets, the @code{.type} directive is used like this:
5596 .type @var{name} , @var{type description}
5599 This sets the type of symbol @var{name} to be either a
5600 function symbol or an object symbol. There are five different syntaxes
5601 supported for the @var{type description} field, in order to provide
5602 compatibility with various other assemblers. The syntaxes supported are:
5605 .type <name>,#function
5606 .type <name>,#object
5608 .type <name>,@@function
5609 .type <name>,@@object
5611 .type <name>,%function
5612 .type <name>,%object
5614 .type <name>,"function"
5615 .type <name>,"object"
5617 .type <name> STT_FUNCTION
5618 .type <name> STT_OBJECT
5624 @section @code{.uleb128 @var{expressions}}
5626 @cindex @code{uleb128} directive
5627 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
5628 compact, variable length representation of numbers used by the DWARF
5629 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
5633 @section @code{.val @var{addr}}
5635 @cindex @code{val} directive
5636 @cindex COFF value attribute
5637 @cindex value attribute, COFF
5638 This directive, permitted only within @code{.def}/@code{.endef} pairs,
5639 records the address @var{addr} as the value attribute of a symbol table
5643 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
5644 configured for @code{b.out}, it accepts this directive but ignores it.
5650 @section @code{.version "@var{string}"}
5652 @cindex @code{version} directive
5653 This directive creates a @code{.note} section and places into it an ELF
5654 formatted note of type NT_VERSION. The note's name is set to @code{string}.
5659 @section @code{.vtable_entry @var{table}, @var{offset}}
5661 @cindex @code{vtable_entry}
5662 This directive finds or creates a symbol @code{table} and creates a
5663 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
5666 @section @code{.vtable_inherit @var{child}, @var{parent}}
5668 @cindex @code{vtable_inherit}
5669 This directive finds the symbol @code{child} and finds or creates the symbol
5670 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
5671 parent whose addend is the value of the child symbol. As a special case the
5672 parent name of @code{0} is treated as refering the @code{*ABS*} section.
5677 @section @code{.weak @var{names}}
5679 @cindex @code{weak} directive
5680 This directive sets the weak attribute on the comma separated list of symbol
5681 @code{names}. If the symbols do not already exist, they will be created.
5685 @section @code{.word @var{expressions}}
5687 @cindex @code{word} directive
5688 This directive expects zero or more @var{expressions}, of any section,
5689 separated by commas.
5692 For each expression, @command{@value{AS}} emits a 32-bit number.
5695 For each expression, @command{@value{AS}} emits a 16-bit number.
5700 The size of the number emitted, and its byte order,
5701 depend on what target computer the assembly is for.
5704 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
5705 @c happen---32-bit addressability, period; no long/short jumps.
5706 @ifset DIFF-TBL-KLUGE
5707 @cindex difference tables altered
5708 @cindex altered difference tables
5710 @emph{Warning: Special Treatment to support Compilers}
5714 Machines with a 32-bit address space, but that do less than 32-bit
5715 addressing, require the following special treatment. If the machine of
5716 interest to you does 32-bit addressing (or doesn't require it;
5717 @pxref{Machine Dependencies}), you can ignore this issue.
5720 In order to assemble compiler output into something that works,
5721 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
5722 Directives of the form @samp{.word sym1-sym2} are often emitted by
5723 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
5724 directive of the form @samp{.word sym1-sym2}, and the difference between
5725 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
5726 creates a @dfn{secondary jump table}, immediately before the next label.
5727 This secondary jump table is preceded by a short-jump to the
5728 first byte after the secondary table. This short-jump prevents the flow
5729 of control from accidentally falling into the new table. Inside the
5730 table is a long-jump to @code{sym2}. The original @samp{.word}
5731 contains @code{sym1} minus the address of the long-jump to
5734 If there were several occurrences of @samp{.word sym1-sym2} before the
5735 secondary jump table, all of them are adjusted. If there was a
5736 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
5737 long-jump to @code{sym4} is included in the secondary jump table,
5738 and the @code{.word} directives are adjusted to contain @code{sym3}
5739 minus the address of the long-jump to @code{sym4}; and so on, for as many
5740 entries in the original jump table as necessary.
5743 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
5744 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
5745 assembly language programmers.
5748 @c end DIFF-TBL-KLUGE
5751 @section Deprecated Directives
5753 @cindex deprecated directives
5754 @cindex obsolescent directives
5755 One day these directives won't work.
5756 They are included for compatibility with older assemblers.
5763 @node Machine Dependencies
5764 @chapter Machine Dependent Features
5766 @cindex machine dependencies
5767 The machine instruction sets are (almost by definition) different on
5768 each machine where @command{@value{AS}} runs. Floating point representations
5769 vary as well, and @command{@value{AS}} often supports a few additional
5770 directives or command-line options for compatibility with other
5771 assemblers on a particular platform. Finally, some versions of
5772 @command{@value{AS}} support special pseudo-instructions for branch
5775 This chapter discusses most of these differences, though it does not
5776 include details on any machine's instruction set. For details on that
5777 subject, see the hardware manufacturer's manual.
5781 * AMD29K-Dependent:: AMD 29K Dependent Features
5784 * Alpha-Dependent:: Alpha Dependent Features
5787 * ARC-Dependent:: ARC Dependent Features
5790 * ARM-Dependent:: ARM Dependent Features
5793 * CRIS-Dependent:: CRIS Dependent Features
5796 * D10V-Dependent:: D10V Dependent Features
5799 * D30V-Dependent:: D30V Dependent Features
5802 * H8/300-Dependent:: Hitachi H8/300 Dependent Features
5805 * H8/500-Dependent:: Hitachi H8/500 Dependent Features
5808 * HPPA-Dependent:: HPPA Dependent Features
5811 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
5814 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
5817 * i860-Dependent:: Intel 80860 Dependent Features
5820 * i960-Dependent:: Intel 80960 Dependent Features
5823 * IP2K-Dependent:: IP2K Dependent Features
5826 * M32R-Dependent:: M32R Dependent Features
5829 * M68K-Dependent:: M680x0 Dependent Features
5832 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
5835 * M88K-Dependent:: M880x0 Dependent Features
5838 * MIPS-Dependent:: MIPS Dependent Features
5841 * MMIX-Dependent:: MMIX Dependent Features
5844 * MSP430-Dependent:: MSP430 Dependent Features
5847 * SH-Dependent:: Hitachi SH Dependent Features
5848 * SH64-Dependent:: Hitachi SH64 Dependent Features
5851 * PDP-11-Dependent:: PDP-11 Dependent Features
5854 * PJ-Dependent:: picoJava Dependent Features
5857 * PPC-Dependent:: PowerPC Dependent Features
5860 * Sparc-Dependent:: SPARC Dependent Features
5863 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
5866 * V850-Dependent:: V850 Dependent Features
5869 * Z8000-Dependent:: Z8000 Dependent Features
5872 * Vax-Dependent:: VAX Dependent Features
5879 @c The following major nodes are *sections* in the GENERIC version, *chapters*
5880 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
5881 @c peculiarity: to preserve cross-references, there must be a node called
5882 @c "Machine Dependencies". Hence the conditional nodenames in each
5883 @c major node below. Node defaulting in makeinfo requires adjacency of
5884 @c node and sectioning commands; hence the repetition of @chapter BLAH
5885 @c in both conditional blocks.
5888 @include c-a29k.texi
5892 @include c-alpha.texi
5904 @include c-cris.texi
5909 @node Machine Dependencies
5910 @chapter Machine Dependent Features
5912 The machine instruction sets are different on each Hitachi chip family,
5913 and there are also some syntax differences among the families. This
5914 chapter describes the specific @command{@value{AS}} features for each
5918 * H8/300-Dependent:: Hitachi H8/300 Dependent Features
5919 * H8/500-Dependent:: Hitachi H8/500 Dependent Features
5920 * SH-Dependent:: Hitachi SH Dependent Features
5927 @include c-d10v.texi
5931 @include c-d30v.texi
5935 @include c-h8300.texi
5939 @include c-h8500.texi
5943 @include c-hppa.texi
5947 @include c-i370.texi
5951 @include c-i386.texi
5955 @include c-i860.texi
5959 @include c-i960.texi
5963 @include c-ia64.texi
5967 @include c-ip2k.texi
5971 @include c-m32r.texi
5975 @include c-m68k.texi
5979 @include c-m68hc11.texi
5983 @include c-m88k.texi
5987 @include c-mips.texi
5991 @include c-mmix.texi
5995 @include c-msp430.texi
5999 @include c-ns32k.texi
6003 @include c-pdp11.texi
6016 @include c-sh64.texi
6020 @include c-sparc.texi
6024 @include c-tic54x.texi
6036 @include c-v850.texi
6040 @c reverse effect of @down at top of generic Machine-Dep chapter
6044 @node Reporting Bugs
6045 @chapter Reporting Bugs
6046 @cindex bugs in assembler
6047 @cindex reporting bugs in assembler
6049 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6051 Reporting a bug may help you by bringing a solution to your problem, or it may
6052 not. But in any case the principal function of a bug report is to help the
6053 entire community by making the next version of @command{@value{AS}} work better.
6054 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6056 In order for a bug report to serve its purpose, you must include the
6057 information that enables us to fix the bug.
6060 * Bug Criteria:: Have you found a bug?
6061 * Bug Reporting:: How to report bugs
6065 @section Have You Found a Bug?
6066 @cindex bug criteria
6068 If you are not sure whether you have found a bug, here are some guidelines:
6071 @cindex fatal signal
6072 @cindex assembler crash
6073 @cindex crash of assembler
6075 If the assembler gets a fatal signal, for any input whatever, that is a
6076 @command{@value{AS}} bug. Reliable assemblers never crash.
6078 @cindex error on valid input
6080 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6082 @cindex invalid input
6084 If @command{@value{AS}} does not produce an error message for invalid input, that
6085 is a bug. However, you should note that your idea of ``invalid input'' might
6086 be our idea of ``an extension'' or ``support for traditional practice''.
6089 If you are an experienced user of assemblers, your suggestions for improvement
6090 of @command{@value{AS}} are welcome in any case.
6094 @section How to Report Bugs
6096 @cindex assembler bugs, reporting
6098 A number of companies and individuals offer support for @sc{gnu} products. If
6099 you obtained @command{@value{AS}} from a support organization, we recommend you
6100 contact that organization first.
6102 You can find contact information for many support companies and
6103 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6106 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6107 to @samp{bug-binutils@@gnu.org}.
6109 The fundamental principle of reporting bugs usefully is this:
6110 @strong{report all the facts}. If you are not sure whether to state a
6111 fact or leave it out, state it!
6113 Often people omit facts because they think they know what causes the problem
6114 and assume that some details do not matter. Thus, you might assume that the
6115 name of a symbol you use in an example does not matter. Well, probably it does
6116 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6117 happens to fetch from the location where that name is stored in memory;
6118 perhaps, if the name were different, the contents of that location would fool
6119 the assembler into doing the right thing despite the bug. Play it safe and
6120 give a specific, complete example. That is the easiest thing for you to do,
6121 and the most helpful.
6123 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6124 it is new to us. Therefore, always write your bug reports on the assumption
6125 that the bug has not been reported previously.
6127 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6128 bell?'' This cannot help us fix a bug, so it is basically useless. We
6129 respond by asking for enough details to enable us to investigate.
6130 You might as well expedite matters by sending them to begin with.
6132 To enable us to fix the bug, you should include all these things:
6136 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6137 it with the @samp{--version} argument.
6139 Without this, we will not know whether there is any point in looking for
6140 the bug in the current version of @command{@value{AS}}.
6143 Any patches you may have applied to the @command{@value{AS}} source.
6146 The type of machine you are using, and the operating system name and
6150 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6154 The command arguments you gave the assembler to assemble your example and
6155 observe the bug. To guarantee you will not omit something important, list them
6156 all. A copy of the Makefile (or the output from make) is sufficient.
6158 If we were to try to guess the arguments, we would probably guess wrong
6159 and then we might not encounter the bug.
6162 A complete input file that will reproduce the bug. If the bug is observed when
6163 the assembler is invoked via a compiler, send the assembler source, not the
6164 high level language source. Most compilers will produce the assembler source
6165 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6166 the options @samp{-v --save-temps}; this will save the assembler source in a
6167 file with an extension of @file{.s}, and also show you exactly how
6168 @command{@value{AS}} is being run.
6171 A description of what behavior you observe that you believe is
6172 incorrect. For example, ``It gets a fatal signal.''
6174 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6175 will certainly notice it. But if the bug is incorrect output, we might not
6176 notice unless it is glaringly wrong. You might as well not give us a chance to
6179 Even if the problem you experience is a fatal signal, you should still say so
6180 explicitly. Suppose something strange is going on, such as, your copy of
6181 @command{@value{AS}} is out of synch, or you have encountered a bug in the C
6182 library on your system. (This has happened!) Your copy might crash and ours
6183 would not. If you told us to expect a crash, then when ours fails to crash, we
6184 would know that the bug was not happening for us. If you had not told us to
6185 expect a crash, then we would not be able to draw any conclusion from our
6189 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6190 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6191 option. Always send diffs from the old file to the new file. If you even
6192 discuss something in the @command{@value{AS}} source, refer to it by context, not
6195 The line numbers in our development sources will not match those in your
6196 sources. Your line numbers would convey no useful information to us.
6199 Here are some things that are not necessary:
6203 A description of the envelope of the bug.
6205 Often people who encounter a bug spend a lot of time investigating
6206 which changes to the input file will make the bug go away and which
6207 changes will not affect it.
6209 This is often time consuming and not very useful, because the way we
6210 will find the bug is by running a single example under the debugger
6211 with breakpoints, not by pure deduction from a series of examples.
6212 We recommend that you save your time for something else.
6214 Of course, if you can find a simpler example to report @emph{instead}
6215 of the original one, that is a convenience for us. Errors in the
6216 output will be easier to spot, running under the debugger will take
6217 less time, and so on.
6219 However, simplification is not vital; if you do not want to do this,
6220 report the bug anyway and send us the entire test case you used.
6223 A patch for the bug.
6225 A patch for the bug does help us if it is a good one. But do not omit
6226 the necessary information, such as the test case, on the assumption that
6227 a patch is all we need. We might see problems with your patch and decide
6228 to fix the problem another way, or we might not understand it at all.
6230 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6231 construct an example that will make the program follow a certain path through
6232 the code. If you do not send us the example, we will not be able to construct
6233 one, so we will not be able to verify that the bug is fixed.
6235 And if we cannot understand what bug you are trying to fix, or why your
6236 patch should be an improvement, we will not install it. A test case will
6237 help us to understand.
6240 A guess about what the bug is or what it depends on.
6242 Such guesses are usually wrong. Even we cannot guess right about such
6243 things without first using the debugger to find the facts.
6246 @node Acknowledgements
6247 @chapter Acknowledgements
6249 If you have contributed to @command{@value{AS}} and your name isn't listed here,
6250 it is not meant as a slight. We just don't know about it. Send mail to the
6251 maintainer, and we'll correct the situation. Currently
6253 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6255 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6258 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6259 information and the 68k series machines, most of the preprocessing pass, and
6260 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6262 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6263 many bug fixes, including merging support for several processors, breaking GAS
6264 up to handle multiple object file format back ends (including heavy rewrite,
6265 testing, an integration of the coff and b.out back ends), adding configuration
6266 including heavy testing and verification of cross assemblers and file splits
6267 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6268 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6269 port (including considerable amounts of reverse engineering), a SPARC opcode
6270 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6271 assertions and made them work, much other reorganization, cleanup, and lint.
6273 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6274 in format-specific I/O modules.
6276 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6277 has done much work with it since.
6279 The Intel 80386 machine description was written by Eliot Dresselhaus.
6281 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6283 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6284 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6286 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6287 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6288 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6289 support a.out format.
6291 Support for the Zilog Z8k and Hitachi H8/300 and H8/500 processors (tc-z8k,
6292 tc-h8300, tc-h8500), and IEEE 695 object file format (obj-ieee), was written by
6293 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6294 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6297 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6298 simplified the configuration of which versions accept which directives. He
6299 updated the 68k machine description so that Motorola's opcodes always produced
6300 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6301 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6302 cross-compilation support, and one bug in relaxation that took a week and
6303 required the proverbial one-bit fix.
6305 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6306 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6307 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6308 PowerPC assembler, and made a few other minor patches.
6310 Steve Chamberlain made @command{@value{AS}} able to generate listings.
6312 Hewlett-Packard contributed support for the HP9000/300.
6314 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6315 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6316 formats). This work was supported by both the Center for Software Science at
6317 the University of Utah and Cygnus Support.
6319 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6320 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6321 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6322 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6323 and some initial 64-bit support).
6325 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6327 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6328 support for openVMS/Alpha.
6330 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6333 Several engineers at Cygnus Support have also provided many small bug fixes and
6334 configuration enhancements.
6336 Many others have contributed large or small bugfixes and enhancements. If
6337 you have contributed significant work and are not mentioned on this list, and
6338 want to be, let us know. Some of the history has been lost; we are not
6339 intentionally leaving anyone out.