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
65 @c common OR combinations of conditions
91 @set abnormal-separator
95 @settitle Using @value{AS}
98 @settitle Using @value{AS} (@value{TARGET})
100 @setchapternewpage odd
105 @c WARE! Some of the machine-dependent sections contain tables of machine
106 @c instructions. Except in multi-column format, these tables look silly.
107 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
108 @c the multi-col format is faked within @example sections.
110 @c Again unfortunately, the natural size that fits on a page, for these tables,
111 @c is different depending on whether or not smallbook is turned on.
112 @c This matters, because of order: text flow switches columns at each page
115 @c The format faked in this source works reasonably well for smallbook,
116 @c not well for the default large-page format. This manual expects that if you
117 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
118 @c tables in question. You can turn on one without the other at your
119 @c discretion, of course.
122 @c the insn tables look just as silly in info files regardless of smallbook,
123 @c might as well show 'em anyways.
129 * As: (as). The GNU assembler.
130 * Gas: (as). The GNU assembler.
139 This file documents the GNU Assembler "@value{AS}".
141 @c man begin COPYRIGHT
142 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002 Free Software Foundation, Inc.
144 Permission is granted to copy, distribute and/or modify this document
145 under the terms of the GNU Free Documentation License, Version 1.1
146 or any later version published by the Free Software Foundation;
147 with no Invariant Sections, with no Front-Cover Texts, and with no
148 Back-Cover Texts. A copy of the license is included in the
149 section entitled ``GNU Free Documentation License''.
154 Permission is granted to process this file through Tex and print the
155 results, provided the printed document carries copying permission
156 notice identical to this one except for the removal of this paragraph
157 (this paragraph not being relevant to the printed manual).
163 @title Using @value{AS}
164 @subtitle The @sc{gnu} Assembler
166 @subtitle for the @value{TARGET} family
169 @subtitle Version @value{VERSION}
172 The Free Software Foundation Inc. thanks The Nice Computer
173 Company of Australia for loaning Dean Elsner to write the
174 first (Vax) version of @command{as} for Project @sc{gnu}.
175 The proprietors, management and staff of TNCCA thank FSF for
176 distracting the boss while they got some work
179 @author Dean Elsner, Jay Fenlason & friends
183 \hfill {\it Using {\tt @value{AS}}}\par
184 \hfill Edited by Cygnus Support\par
186 %"boxit" macro for figures:
187 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
188 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
189 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
190 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
191 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
194 @vskip 0pt plus 1filll
195 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002 Free Software Foundation, Inc.
197 Permission is granted to copy, distribute and/or modify this document
198 under the terms of the GNU Free Documentation License, Version 1.1
199 or any later version published by the Free Software Foundation;
200 with no Invariant Sections, with no Front-Cover Texts, and with no
201 Back-Cover Texts. A copy of the license is included in the
202 section entitled ``GNU Free Documentation License''.
208 @top Using @value{AS}
210 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}} version
213 This version of the file describes @command{@value{AS}} configured to generate
214 code for @value{TARGET} architectures.
217 This document is distributed under the terms of the GNU Free
218 Documentation License. A copy of the license is included in the
219 section entitled ``GNU Free Documentation License''.
222 * Overview:: Overview
223 * Invoking:: Command-Line Options
225 * Sections:: Sections and Relocation
227 * Expressions:: Expressions
228 * Pseudo Ops:: Assembler Directives
229 * Machine Dependencies:: Machine Dependent Features
230 * Reporting Bugs:: Reporting Bugs
231 * Acknowledgements:: Who Did What
232 * GNU Free Documentation License:: GNU Free Documentation License
240 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
242 This version of the manual describes @command{@value{AS}} configured to generate
243 code for @value{TARGET} architectures.
247 @cindex invocation summary
248 @cindex option summary
249 @cindex summary of options
250 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
251 @pxref{Invoking,,Command-Line Options}.
253 @c man title AS the portable GNU assembler.
257 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
261 @c We don't use deffn and friends for the following because they seem
262 @c to be limited to one line for the header.
264 @c man begin SYNOPSIS
265 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{-D}] [@b{--defsym} @var{sym}=@var{val}]
266 [@b{-f}] [@b{--gstabs}] [@b{--gdwarf2}] [@b{--help}] [@b{-I} @var{dir}]
267 [@b{-J}] [@b{-K}] [@b{-L}]
268 [@b{--listing-lhs-width}=@var{NUM}] [@b{--listing-lhs-width2}=@var{NUM}]
269 [@b{--listing-rhs-width}=@var{NUM}] [@b{--listing-cont-lines}=@var{NUM}]
270 [@b{--keep-locals}] [@b{-o} @var{objfile}] [@b{-R}] [@b{--statistics}] [@b{-v}]
271 [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}] [@b{--fatal-warnings}]
272 [@b{-w}] [@b{-x}] [@b{-Z}] [@b{--target-help}] [@var{target-options}]
273 [@b{--}|@var{files} @dots{}]
275 @c Target dependent options are listed below. Keep the list sorted.
276 @c Add an empty line for separation.
278 @c am29k has no machine-dependent assembler options
282 @emph{Target Alpha options:}
284 [@b{-mdebug} | @b{-no-mdebug}]
285 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
286 [@b{-F}] [@b{-32addr}]
290 @emph{Target ARC options:}
296 @emph{Target ARM options:}
297 @c Don't document the deprecated options
298 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
299 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
300 [@b{-mfpu}=@var{floating-point-fromat}]
303 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
304 @b{-mapcs-reentrant}]
305 [@b{-mthumb-interwork}] [@b{-moabi}] [@b{-k}]
309 @emph{Target CRIS options:}
310 [@b{--underscore} | @b{--no-underscore}]
312 [@b{--emulation=criself} | @b{--emulation=crisaout}]
313 @c Deprecated -- deliberately not documented.
318 @emph{Target D10V options:}
323 @emph{Target D30V options:}
324 [@b{-O}|@b{-n}|@b{-N}]
327 @c Renesas family chips have no machine-dependent assembler options
330 @c HPPA has no machine-dependent assembler options (yet).
334 @emph{Target i386 options:}
335 [@b{--32}|@b{--64}] [@b{-n}]
339 @emph{Target i960 options:}
340 @c see md_parse_option in tc-i960.c
341 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
343 [@b{-b}] [@b{-no-relax}]
347 @emph{Target IA-64 options:}
348 [@b{-mconstant-gp}|@b{-mauto-pic}]
349 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
351 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
355 @emph{Target IP2K options:}
356 [@b{-mip2022}|@b{-mip2022ext}]
360 @emph{Target M32R options:}
361 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
366 @emph{Target M680X0 options:}
367 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
371 @emph{Target M68HC11 options:}
372 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
373 [@b{-mshort}|@b{-mlong}]
374 [@b{-mshort-double}|@b{-mlong-double}]
375 [@b{--force-long-branchs}] [@b{--short-branchs}]
376 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
377 [@b{--print-opcodes}] [@b{--generate-example}]
381 @emph{Target MCORE options:}
382 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
383 [@b{-mcpu=[210|340]}]
387 @emph{Target MIPS options:}
388 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-n}] [@b{-O}[@var{optimization level}]]
389 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
390 [@b{-non_shared}] [@b{-xgot}] [@b{--membedded-pic}]
391 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
392 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
393 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
394 [@b{-mips64}] [@b{-mips64r2}]
395 [@b{-construct-floats}] [@b{-no-construct-floats}]
396 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
397 [@b{-mfix7000}] [@b{-mno-fix7000}]
398 [@b{-mips16}] [@b{-no-mips16}]
399 [@b{-mips3d}] [@b{-no-mips3d}]
400 [@b{-mdmx}] [@b{-no-mdmx}]
401 [@b{-mdebug}] [@b{-no-mdebug}]
402 [@b{-mpdr}] [@b{-mno-pdr}]
406 @emph{Target MMIX options:}
407 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
408 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
409 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
410 [@b{--linker-allocated-gregs}]
414 @emph{Target PDP11 options:}
415 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
416 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
417 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
421 @emph{Target picoJava options:}
426 @emph{Target PowerPC options:}
427 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
428 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
429 @b{-mbooke32}|@b{-mbooke64}]
430 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
431 [@b{-mregnames}|@b{-mno-regnames}]
432 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
433 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
434 [@b{-msolaris}|@b{-mno-solaris}]
438 @emph{Target SPARC options:}
439 @c The order here is important. See c-sparc.texi.
440 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
441 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
442 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
447 @emph{Target TIC54X options:}
448 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
449 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
452 @c Z8000 has no machine-dependent assembler options
456 @emph{Target Xtensa options:}
457 [@b{--[no-]density}] [@b{--[no-]relax}] [@b{--[no-]generics}]
458 [@b{--[no-]text-section-literals}]
459 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
468 Turn on listings, in any of a variety of ways:
472 omit false conditionals
475 omit debugging directives
478 include high-level source
484 include macro expansions
487 omit forms processing
493 set the name of the listing file
496 You may combine these options; for example, use @samp{-aln} for assembly
497 listing without forms processing. The @samp{=file} option, if used, must be
498 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
501 Ignored. This option is accepted for script compatibility with calls to
504 @item --defsym @var{sym}=@var{value}
505 Define the symbol @var{sym} to be @var{value} before assembling the input file.
506 @var{value} must be an integer constant. As in C, a leading @samp{0x}
507 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
510 ``fast''---skip whitespace and comment preprocessing (assume source is
514 Generate stabs debugging information for each assembler line. This
515 may help debugging assembler code, if the debugger can handle it.
518 Generate DWARF2 debugging information for each assembler line. This
519 may help debugging assembler code, if the debugger can handle it. Note---this
520 option is only supported by some targets, not all of them.
523 Print a summary of the command line options and exit.
526 Print a summary of all target specific options and exit.
529 Add directory @var{dir} to the search list for @code{.include} directives.
532 Don't warn about signed overflow.
535 @ifclear DIFF-TBL-KLUGE
536 This option is accepted but has no effect on the @value{TARGET} family.
538 @ifset DIFF-TBL-KLUGE
539 Issue warnings when difference tables altered for long displacements.
544 Keep (in the symbol table) local symbols. On traditional a.out systems
545 these start with @samp{L}, but different systems have different local
548 @item --listing-lhs-width=@var{number}
549 Set the maximum width, in words, of the output data column for an assembler
550 listing to @var{number}.
552 @item --listing-lhs-width2=@var{number}
553 Set the maximum width, in words, of the output data column for continuation
554 lines in an assembler listing to @var{number}.
556 @item --listing-rhs-width=@var{number}
557 Set the maximum width of an input source line, as displayed in a listing, to
560 @item --listing-cont-lines=@var{number}
561 Set the maximum number of lines printed in a listing for a single line of input
564 @item -o @var{objfile}
565 Name the object-file output from @command{@value{AS}} @var{objfile}.
568 Fold the data section into the text section.
571 Print the maximum space (in bytes) and total time (in seconds) used by
574 @item --strip-local-absolute
575 Remove local absolute symbols from the outgoing symbol table.
579 Print the @command{as} version.
582 Print the @command{as} version and exit.
586 Suppress warning messages.
588 @item --fatal-warnings
589 Treat warnings as errors.
592 Don't suppress warning messages or treat them as errors.
601 Generate an object file even after errors.
603 @item -- | @var{files} @dots{}
604 Standard input, or source files to assemble.
609 The following options are available when @value{AS} is configured for
614 This option selects the core processor variant.
616 Select either big-endian (-EB) or little-endian (-EL) output.
621 The following options are available when @value{AS} is configured for the ARM
625 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
626 Specify which ARM processor variant is the target.
627 @item -march=@var{architecture}[+@var{extension}@dots{}]
628 Specify which ARM architecture variant is used by the target.
629 @item -mfpu=@var{floating-point-format}
630 Select which Floating Point architecture is the target.
632 Enable Thumb only instruction decoding.
633 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant | -moabi
634 Select which procedure calling convention is in use.
636 Select either big-endian (-EB) or little-endian (-EL) output.
637 @item -mthumb-interwork
638 Specify that the code has been generated with interworking between Thumb and
641 Specify that PIC code has been generated.
646 See the info pages for documentation of the CRIS-specific options.
650 The following options are available when @value{AS} is configured for
653 @cindex D10V optimization
654 @cindex optimization, D10V
656 Optimize output by parallelizing instructions.
661 The following options are available when @value{AS} is configured for a D30V
664 @cindex D30V optimization
665 @cindex optimization, D30V
667 Optimize output by parallelizing instructions.
671 Warn when nops are generated.
673 @cindex D30V nops after 32-bit multiply
675 Warn when a nop after a 32-bit multiply instruction is generated.
680 The following options are available when @value{AS} is configured for the
681 Intel 80960 processor.
684 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
685 Specify which variant of the 960 architecture is the target.
688 Add code to collect statistics about branches taken.
691 Do not alter compare-and-branch instructions for long displacements;
698 The following options are available when @value{AS} is configured for the
704 Specifies that the extended IP2022 instructions are allowed.
707 Restores the default behaviour, which restricts the permitted instructions to
708 just the basic IP2022 ones.
714 The following options are available when @value{AS} is configured for the
715 Renesas M32R (formerly Mitsubishi M32R) series.
720 Specify which processor in the M32R family is the target. The default
721 is normally the M32R, but this option changes it to the M32RX.
723 @item --warn-explicit-parallel-conflicts or --Wp
724 Produce warning messages when questionable parallel constructs are
727 @item --no-warn-explicit-parallel-conflicts or --Wnp
728 Do not produce warning messages when questionable parallel constructs are
735 The following options are available when @value{AS} is configured for the
736 Motorola 68000 series.
741 Shorten references to undefined symbols, to one word instead of two.
743 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
744 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
745 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
746 Specify what processor in the 68000 family is the target. The default
747 is normally the 68020, but this can be changed at configuration time.
749 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
750 The target machine does (or does not) have a floating-point coprocessor.
751 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
752 the basic 68000 is not compatible with the 68881, a combination of the
753 two can be specified, since it's possible to do emulation of the
754 coprocessor instructions with the main processor.
756 @item -m68851 | -mno-68851
757 The target machine does (or does not) have a memory-management
758 unit coprocessor. The default is to assume an MMU for 68020 and up.
765 For details about the PDP-11 machine dependent features options,
766 see @ref{PDP-11-Options}.
769 @item -mpic | -mno-pic
770 Generate position-independent (or position-dependent) code. The
771 default is @option{-mpic}.
774 @itemx -mall-extensions
775 Enable all instruction set extensions. This is the default.
777 @item -mno-extensions
778 Disable all instruction set extensions.
780 @item -m@var{extension} | -mno-@var{extension}
781 Enable (or disable) a particular instruction set extension.
784 Enable the instruction set extensions supported by a particular CPU, and
785 disable all other extensions.
787 @item -m@var{machine}
788 Enable the instruction set extensions supported by a particular machine
789 model, and disable all other extensions.
795 The following options are available when @value{AS} is configured for
796 a picoJava processor.
800 @cindex PJ endianness
801 @cindex endianness, PJ
802 @cindex big endian output, PJ
804 Generate ``big endian'' format output.
806 @cindex little endian output, PJ
808 Generate ``little endian'' format output.
814 The following options are available when @value{AS} is configured for the
815 Motorola 68HC11 or 68HC12 series.
819 @item -m68hc11 | -m68hc12 | -m68hcs12
820 Specify what processor is the target. The default is
821 defined by the configuration option when building the assembler.
824 Specify to use the 16-bit integer ABI.
827 Specify to use the 32-bit integer ABI.
830 Specify to use the 32-bit double ABI.
833 Specify to use the 64-bit double ABI.
835 @item --force-long-branchs
836 Relative branches are turned into absolute ones. This concerns
837 conditional branches, unconditional branches and branches to a
840 @item -S | --short-branchs
841 Do not turn relative branchs into absolute ones
842 when the offset is out of range.
844 @item --strict-direct-mode
845 Do not turn the direct addressing mode into extended addressing mode
846 when the instruction does not support direct addressing mode.
848 @item --print-insn-syntax
849 Print the syntax of instruction in case of error.
851 @item --print-opcodes
852 print the list of instructions with syntax and then exit.
854 @item --generate-example
855 print an example of instruction for each possible instruction and then exit.
856 This option is only useful for testing @command{@value{AS}}.
862 The following options are available when @command{@value{AS}} is configured
863 for the SPARC architecture:
866 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
867 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
868 Explicitly select a variant of the SPARC architecture.
870 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
871 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
873 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
874 UltraSPARC extensions.
876 @item -xarch=v8plus | -xarch=v8plusa
877 For compatibility with the Solaris v9 assembler. These options are
878 equivalent to -Av8plus and -Av8plusa, respectively.
881 Warn when the assembler switches to another architecture.
886 The following options are available when @value{AS} is configured for the 'c54x
891 Enable extended addressing mode. All addresses and relocations will assume
892 extended addressing (usually 23 bits).
893 @item -mcpu=@var{CPU_VERSION}
894 Sets the CPU version being compiled for.
895 @item -merrors-to-file @var{FILENAME}
896 Redirect error output to a file, for broken systems which don't support such
897 behaviour in the shell.
902 The following options are available when @value{AS} is configured for
903 a @sc{mips} processor.
907 This option sets the largest size of an object that can be referenced
908 implicitly with the @code{gp} register. It is only accepted for targets that
909 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
911 @cindex MIPS endianness
912 @cindex endianness, MIPS
913 @cindex big endian output, MIPS
915 Generate ``big endian'' format output.
917 @cindex little endian output, MIPS
919 Generate ``little endian'' format output.
931 Generate code for a particular @sc{mips} Instruction Set Architecture level.
932 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
933 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
934 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
935 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
937 correspond to generic
938 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
939 and @samp{MIPS64 Release 2}
940 ISA processors, respectively.
942 @item -march=@var{CPU}
943 Generate code for a particular @sc{mips} cpu.
945 @item -mtune=@var{cpu}
946 Schedule and tune for a particular @sc{mips} cpu.
950 Cause nops to be inserted if the read of the destination register
951 of an mfhi or mflo instruction occurs in the following two instructions.
955 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
956 section instead of the standard ELF .stabs sections.
960 Control generation of @code{.pdr} sections.
964 The register sizes are normally inferred from the ISA and ABI, but these
965 flags force a certain group of registers to be treated as 32 bits wide at
966 all times. @samp{-mgp32} controls the size of general-purpose registers
967 and @samp{-mfp32} controls the size of floating-point registers.
971 Generate code for the MIPS 16 processor. This is equivalent to putting
972 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
973 turns off this option.
977 Generate code for the MIPS-3D Application Specific Extension.
978 This tells the assembler to accept MIPS-3D instructions.
979 @samp{-no-mips3d} turns off this option.
983 Generate code for the MDMX Application Specific Extension.
984 This tells the assembler to accept MDMX instructions.
985 @samp{-no-mdmx} turns off this option.
987 @item --construct-floats
988 @itemx --no-construct-floats
989 The @samp{--no-construct-floats} option disables the construction of
990 double width floating point constants by loading the two halves of the
991 value into the two single width floating point registers that make up
992 the double width register. By default @samp{--construct-floats} is
993 selected, allowing construction of these floating point constants.
996 @item --emulation=@var{name}
997 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
998 for some other target, in all respects, including output format (choosing
999 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1000 debugging information or store symbol table information, and default
1001 endianness. The available configuration names are: @samp{mipsecoff},
1002 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1003 @samp{mipsbelf}. The first two do not alter the default endianness from that
1004 of the primary target for which the assembler was configured; the others change
1005 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1006 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1007 selection in any case.
1009 This option is currently supported only when the primary target
1010 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1011 Furthermore, the primary target or others specified with
1012 @samp{--enable-targets=@dots{}} at configuration time must include support for
1013 the other format, if both are to be available. For example, the Irix 5
1014 configuration includes support for both.
1016 Eventually, this option will support more configurations, with more
1017 fine-grained control over the assembler's behavior, and will be supported for
1021 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1028 Control how to deal with multiplication overflow and division by zero.
1029 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1030 (and only work for Instruction Set Architecture level 2 and higher);
1031 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1035 When this option is used, @command{@value{AS}} will issue a warning every
1036 time it generates a nop instruction from a macro.
1041 The following options are available when @value{AS} is configured for
1047 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1048 The command line option @samp{-nojsri2bsr} can be used to disable it.
1052 Enable or disable the silicon filter behaviour. By default this is disabled.
1053 The default can be overridden by the @samp{-sifilter} command line option.
1056 Alter jump instructions for long displacements.
1058 @item -mcpu=[210|340]
1059 Select the cpu type on the target hardware. This controls which instructions
1063 Assemble for a big endian target.
1066 Assemble for a little endian target.
1072 See the info pages for documentation of the MMIX-specific options.
1076 The following options are available when @value{AS} is configured for
1077 an Xtensa processor.
1080 @item --density | --no-density
1081 Enable or disable use of instructions from the Xtensa code density
1082 option. This is enabled by default when the Xtensa processor supports
1083 the code density option.
1085 @item --relax | --no-relax
1086 Enable or disable instruction relaxation. This is enabled by default.
1087 Note: In the current implementation, these options also control whether
1088 assembler optimizations are performed, making these options equivalent
1089 to @option{--generics} and @option{--no-generics}.
1091 @item --generics | --no-generics
1092 Enable or disable all assembler transformations of Xtensa instructions.
1093 The default is @option{--generics};
1094 @option{--no-generics} should be used only in the rare cases when the
1095 instructions must be exactly as specified in the assembly source.
1097 @item --text-section-literals | --no-text-section-literals
1098 With @option{--text-@-section-@-literals}, literal pools are interspersed
1099 in the text section. The default is
1100 @option{--no-@-text-@-section-@-literals}, which places literals in a
1101 separate section in the output file.
1103 @item --target-align | --no-target-align
1104 Enable or disable automatic alignment to reduce branch penalties at the
1105 expense of some code density. The default is @option{--target-@-align}.
1107 @item --longcalls | --no-longcalls
1108 Enable or disable transformation of call instructions to allow calls
1109 across a greater range of addresses. The default is
1110 @option{--no-@-longcalls}.
1117 * Manual:: Structure of this Manual
1118 * GNU Assembler:: The GNU Assembler
1119 * Object Formats:: Object File Formats
1120 * Command Line:: Command Line
1121 * Input Files:: Input Files
1122 * Object:: Output (Object) File
1123 * Errors:: Error and Warning Messages
1127 @section Structure of this Manual
1129 @cindex manual, structure and purpose
1130 This manual is intended to describe what you need to know to use
1131 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1132 notation for symbols, constants, and expressions; the directives that
1133 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1136 We also cover special features in the @value{TARGET}
1137 configuration of @command{@value{AS}}, including assembler directives.
1140 This manual also describes some of the machine-dependent features of
1141 various flavors of the assembler.
1144 @cindex machine instructions (not covered)
1145 On the other hand, this manual is @emph{not} intended as an introduction
1146 to programming in assembly language---let alone programming in general!
1147 In a similar vein, we make no attempt to introduce the machine
1148 architecture; we do @emph{not} describe the instruction set, standard
1149 mnemonics, registers or addressing modes that are standard to a
1150 particular architecture.
1152 You may want to consult the manufacturer's
1153 machine architecture manual for this information.
1157 For information on the H8/300 machine instruction set, see @cite{H8/300
1158 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1159 Programming Manual} (Renesas).
1162 For information on the H8/500 machine instruction set, see @cite{H8/500
1163 Series Programming Manual} (Renesas M21T001).
1166 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1167 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1168 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1169 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1172 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1176 @c I think this is premature---doc@cygnus.com, 17jan1991
1178 Throughout this manual, we assume that you are running @dfn{GNU},
1179 the portable operating system from the @dfn{Free Software
1180 Foundation, Inc.}. This restricts our attention to certain kinds of
1181 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1182 once this assumption is granted examples and definitions need less
1185 @command{@value{AS}} is part of a team of programs that turn a high-level
1186 human-readable series of instructions into a low-level
1187 computer-readable series of instructions. Different versions of
1188 @command{@value{AS}} are used for different kinds of computer.
1191 @c There used to be a section "Terminology" here, which defined
1192 @c "contents", "byte", "word", and "long". Defining "word" to any
1193 @c particular size is confusing when the .word directive may generate 16
1194 @c bits on one machine and 32 bits on another; in general, for the user
1195 @c version of this manual, none of these terms seem essential to define.
1196 @c They were used very little even in the former draft of the manual;
1197 @c this draft makes an effort to avoid them (except in names of
1201 @section The GNU Assembler
1203 @c man begin DESCRIPTION
1205 @sc{gnu} @command{as} is really a family of assemblers.
1207 This manual describes @command{@value{AS}}, a member of that family which is
1208 configured for the @value{TARGET} architectures.
1210 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1211 should find a fairly similar environment when you use it on another
1212 architecture. Each version has much in common with the others,
1213 including object file formats, most assembler directives (often called
1214 @dfn{pseudo-ops}) and assembler syntax.@refill
1216 @cindex purpose of @sc{gnu} assembler
1217 @command{@value{AS}} is primarily intended to assemble the output of the
1218 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1219 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1220 assemble correctly everything that other assemblers for the same
1221 machine would assemble.
1223 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1226 @c This remark should appear in generic version of manual; assumption
1227 @c here is that generic version sets M680x0.
1228 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1229 assembler for the same architecture; for example, we know of several
1230 incompatible versions of 680x0 assembly language syntax.
1235 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1236 program in one pass of the source file. This has a subtle impact on the
1237 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1239 @node Object Formats
1240 @section Object File Formats
1242 @cindex object file format
1243 The @sc{gnu} assembler can be configured to produce several alternative
1244 object file formats. For the most part, this does not affect how you
1245 write assembly language programs; but directives for debugging symbols
1246 are typically different in different file formats. @xref{Symbol
1247 Attributes,,Symbol Attributes}.
1250 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1251 @value{OBJ-NAME} format object files.
1253 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1255 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1256 @code{a.out} or COFF format object files.
1259 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1260 @code{b.out} or COFF format object files.
1263 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1264 SOM or ELF format object files.
1269 @section Command Line
1271 @cindex command line conventions
1273 After the program name @command{@value{AS}}, the command line may contain
1274 options and file names. Options may appear in any order, and may be
1275 before, after, or between file names. The order of file names is
1278 @cindex standard input, as input file
1280 @file{--} (two hyphens) by itself names the standard input file
1281 explicitly, as one of the files for @command{@value{AS}} to assemble.
1283 @cindex options, command line
1284 Except for @samp{--} any command line argument that begins with a
1285 hyphen (@samp{-}) is an option. Each option changes the behavior of
1286 @command{@value{AS}}. No option changes the way another option works. An
1287 option is a @samp{-} followed by one or more letters; the case of
1288 the letter is important. All options are optional.
1290 Some options expect exactly one file name to follow them. The file
1291 name may either immediately follow the option's letter (compatible
1292 with older assemblers) or it may be the next command argument (@sc{gnu}
1293 standard). These two command lines are equivalent:
1296 @value{AS} -o my-object-file.o mumble.s
1297 @value{AS} -omy-object-file.o mumble.s
1301 @section Input Files
1304 @cindex source program
1305 @cindex files, input
1306 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1307 describe the program input to one run of @command{@value{AS}}. The program may
1308 be in one or more files; how the source is partitioned into files
1309 doesn't change the meaning of the source.
1311 @c I added "con" prefix to "catenation" just to prove I can overcome my
1312 @c APL training... doc@cygnus.com
1313 The source program is a concatenation of the text in all the files, in the
1316 @c man begin DESCRIPTION
1317 Each time you run @command{@value{AS}} it assembles exactly one source
1318 program. The source program is made up of one or more files.
1319 (The standard input is also a file.)
1321 You give @command{@value{AS}} a command line that has zero or more input file
1322 names. The input files are read (from left file name to right). A
1323 command line argument (in any position) that has no special meaning
1324 is taken to be an input file name.
1326 If you give @command{@value{AS}} no file names it attempts to read one input file
1327 from the @command{@value{AS}} standard input, which is normally your terminal. You
1328 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1331 Use @samp{--} if you need to explicitly name the standard input file
1332 in your command line.
1334 If the source is empty, @command{@value{AS}} produces a small, empty object
1339 @subheading Filenames and Line-numbers
1341 @cindex input file linenumbers
1342 @cindex line numbers, in input files
1343 There are two ways of locating a line in the input file (or files) and
1344 either may be used in reporting error messages. One way refers to a line
1345 number in a physical file; the other refers to a line number in a
1346 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1348 @dfn{Physical files} are those files named in the command line given
1349 to @command{@value{AS}}.
1351 @dfn{Logical files} are simply names declared explicitly by assembler
1352 directives; they bear no relation to physical files. Logical file names help
1353 error messages reflect the original source file, when @command{@value{AS}} source
1354 is itself synthesized from other files. @command{@value{AS}} understands the
1355 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1356 @ref{File,,@code{.file}}.
1359 @section Output (Object) File
1365 Every time you run @command{@value{AS}} it produces an output file, which is
1366 your assembly language program translated into numbers. This file
1367 is the object file. Its default name is
1375 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1377 You can give it another name by using the @option{-o} option. Conventionally,
1378 object file names end with @file{.o}. The default name is used for historical
1379 reasons: older assemblers were capable of assembling self-contained programs
1380 directly into a runnable program. (For some formats, this isn't currently
1381 possible, but it can be done for the @code{a.out} format.)
1385 The object file is meant for input to the linker @code{@value{LD}}. It contains
1386 assembled program code, information to help @code{@value{LD}} integrate
1387 the assembled program into a runnable file, and (optionally) symbolic
1388 information for the debugger.
1390 @c link above to some info file(s) like the description of a.out.
1391 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1394 @section Error and Warning Messages
1396 @c man begin DESCRIPTION
1398 @cindex error messages
1399 @cindex warning messages
1400 @cindex messages from assembler
1401 @command{@value{AS}} may write warnings and error messages to the standard error
1402 file (usually your terminal). This should not happen when a compiler
1403 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1404 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1405 grave problem that stops the assembly.
1409 @cindex format of warning messages
1410 Warning messages have the format
1413 file_name:@b{NNN}:Warning Message Text
1417 @cindex line numbers, in warnings/errors
1418 (where @b{NNN} is a line number). If a logical file name has been given
1419 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1420 the current input file is used. If a logical line number was given
1422 (@pxref{Line,,@code{.line}})
1426 (@pxref{Line,,@code{.line}})
1429 (@pxref{Ln,,@code{.ln}})
1432 then it is used to calculate the number printed,
1433 otherwise the actual line in the current source file is printed. The
1434 message text is intended to be self explanatory (in the grand Unix
1437 @cindex format of error messages
1438 Error messages have the format
1440 file_name:@b{NNN}:FATAL:Error Message Text
1442 The file name and line number are derived as for warning
1443 messages. The actual message text may be rather less explanatory
1444 because many of them aren't supposed to happen.
1447 @chapter Command-Line Options
1449 @cindex options, all versions of assembler
1450 This chapter describes command-line options available in @emph{all}
1451 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1453 to the @value{TARGET} target.
1456 to particular machine architectures.
1459 @c man begin DESCRIPTION
1461 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1462 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1463 The assembler arguments must be separated from each other (and the @samp{-Wa})
1464 by commas. For example:
1467 gcc -c -g -O -Wa,-alh,-L file.c
1471 This passes two options to the assembler: @samp{-alh} (emit a listing to
1472 standard output with high-level and assembly source) and @samp{-L} (retain
1473 local symbols in the symbol table).
1475 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1476 command-line options are automatically passed to the assembler by the compiler.
1477 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1478 precisely what options it passes to each compilation pass, including the
1484 * a:: -a[cdhlns] enable listings
1485 * D:: -D for compatibility
1486 * f:: -f to work faster
1487 * I:: -I for .include search path
1488 @ifclear DIFF-TBL-KLUGE
1489 * K:: -K for compatibility
1491 @ifset DIFF-TBL-KLUGE
1492 * K:: -K for difference tables
1495 * L:: -L to retain local labels
1496 * listing:: --listing-XXX to configure listing output
1497 * M:: -M or --mri to assemble in MRI compatibility mode
1498 * MD:: --MD for dependency tracking
1499 * o:: -o to name the object file
1500 * R:: -R to join data and text sections
1501 * statistics:: --statistics to see statistics about assembly
1502 * traditional-format:: --traditional-format for compatible output
1503 * v:: -v to announce version
1504 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1505 * Z:: -Z to make object file even after errors
1509 @section Enable Listings: @option{-a[cdhlns]}
1518 @cindex listings, enabling
1519 @cindex assembly listings, enabling
1521 These options enable listing output from the assembler. By itself,
1522 @samp{-a} requests high-level, assembly, and symbols listing.
1523 You can use other letters to select specific options for the list:
1524 @samp{-ah} requests a high-level language listing,
1525 @samp{-al} requests an output-program assembly listing, and
1526 @samp{-as} requests a symbol table listing.
1527 High-level listings require that a compiler debugging option like
1528 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1531 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1532 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1533 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1534 omitted from the listing.
1536 Use the @samp{-ad} option to omit debugging directives from the
1539 Once you have specified one of these options, you can further control
1540 listing output and its appearance using the directives @code{.list},
1541 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1543 The @samp{-an} option turns off all forms processing.
1544 If you do not request listing output with one of the @samp{-a} options, the
1545 listing-control directives have no effect.
1547 The letters after @samp{-a} may be combined into one option,
1548 @emph{e.g.}, @samp{-aln}.
1550 Note if the assembler source is coming from the standard input (eg because it
1551 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1552 is being used) then the listing will not contain any comments or preprocessor
1553 directives. This is because the listing code buffers input source lines from
1554 stdin only after they have been preprocessed by the assembler. This reduces
1555 memory usage and makes the code more efficient.
1558 @section @option{-D}
1561 This option has no effect whatsoever, but it is accepted to make it more
1562 likely that scripts written for other assemblers also work with
1563 @command{@value{AS}}.
1566 @section Work Faster: @option{-f}
1569 @cindex trusted compiler
1570 @cindex faster processing (@option{-f})
1571 @samp{-f} should only be used when assembling programs written by a
1572 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1573 and comment preprocessing on
1574 the input file(s) before assembling them. @xref{Preprocessing,
1578 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1579 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1584 @section @code{.include} Search Path: @option{-I} @var{path}
1586 @kindex -I @var{path}
1587 @cindex paths for @code{.include}
1588 @cindex search path for @code{.include}
1589 @cindex @code{include} directive search path
1590 Use this option to add a @var{path} to the list of directories
1591 @command{@value{AS}} searches for files specified in @code{.include}
1592 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1593 many times as necessary to include a variety of paths. The current
1594 working directory is always searched first; after that, @command{@value{AS}}
1595 searches any @samp{-I} directories in the same order as they were
1596 specified (left to right) on the command line.
1599 @section Difference Tables: @option{-K}
1602 @ifclear DIFF-TBL-KLUGE
1603 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1604 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1605 where it can be used to warn when the assembler alters the machine code
1606 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1607 family does not have the addressing limitations that sometimes lead to this
1608 alteration on other platforms.
1611 @ifset DIFF-TBL-KLUGE
1612 @cindex difference tables, warning
1613 @cindex warning for altered difference tables
1614 @command{@value{AS}} sometimes alters the code emitted for directives of the form
1615 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1616 You can use the @samp{-K} option if you want a warning issued when this
1621 @section Include Local Labels: @option{-L}
1624 @cindex local labels, retaining in output
1625 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1626 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1627 debugging, because they are intended for the use of programs (like
1628 compilers) that compose assembler programs, not for your notice.
1629 Normally both @command{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1630 normally debug with them.
1632 This option tells @command{@value{AS}} to retain those @samp{L@dots{}} symbols
1633 in the object file. Usually if you do this you also tell the linker
1634 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1636 By default, a local label is any label beginning with @samp{L}, but each
1637 target is allowed to redefine the local label prefix.
1639 On the HPPA local labels begin with @samp{L$}.
1643 @section Configuring listing output: @option{--listing}
1645 The listing feature of the assembler can be enabled via the command line switch
1646 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1647 hex dump of the corresponding locations in the output object file, and displays
1648 them as a listing file. The format of this listing can be controlled by pseudo
1649 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1650 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1653 @item --listing-lhs-width=@samp{number}
1654 @kindex --listing-lhs-width
1655 @cindex Width of first line disassembly output
1656 Sets the maximum width, in words, of the first line of the hex byte dump. This
1657 dump appears on the left hand side of the listing output.
1659 @item --listing-lhs-width2=@samp{number}
1660 @kindex --listing-lhs-width2
1661 @cindex Width of continuation lines of disassembly output
1662 Sets the maximum width, in words, of any further lines of the hex byte dump for
1663 a given input source line. If this value is not specified, it defaults to being
1664 the same as the value specified for @samp{--listing-lhs-width}. If neither
1665 switch is used the default is to one.
1667 @item --listing-rhs-width=@samp{number}
1668 @kindex --listing-rhs-width
1669 @cindex Width of source line output
1670 Sets the maximum width, in characters, of the source line that is displayed
1671 alongside the hex dump. The default value for this parameter is 100. The
1672 source line is displayed on the right hand side of the listing output.
1674 @item --listing-cont-lines=@samp{number}
1675 @kindex --listing-cont-lines
1676 @cindex Maximum number of continuation lines
1677 Sets the maximum number of continuation lines of hex dump that will be
1678 displayed for a given single line of source input. The default value is 4.
1682 @section Assemble in MRI Compatibility Mode: @option{-M}
1685 @cindex MRI compatibility mode
1686 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1687 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1688 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1689 configured target) assembler from Microtec Research. The exact nature of the
1690 MRI syntax will not be documented here; see the MRI manuals for more
1691 information. Note in particular that the handling of macros and macro
1692 arguments is somewhat different. The purpose of this option is to permit
1693 assembling existing MRI assembler code using @command{@value{AS}}.
1695 The MRI compatibility is not complete. Certain operations of the MRI assembler
1696 depend upon its object file format, and can not be supported using other object
1697 file formats. Supporting these would require enhancing each object file format
1698 individually. These are:
1701 @item global symbols in common section
1703 The m68k MRI assembler supports common sections which are merged by the linker.
1704 Other object file formats do not support this. @command{@value{AS}} handles
1705 common sections by treating them as a single common symbol. It permits local
1706 symbols to be defined within a common section, but it can not support global
1707 symbols, since it has no way to describe them.
1709 @item complex relocations
1711 The MRI assemblers support relocations against a negated section address, and
1712 relocations which combine the start addresses of two or more sections. These
1713 are not support by other object file formats.
1715 @item @code{END} pseudo-op specifying start address
1717 The MRI @code{END} pseudo-op permits the specification of a start address.
1718 This is not supported by other object file formats. The start address may
1719 instead be specified using the @option{-e} option to the linker, or in a linker
1722 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1724 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1725 name to the output file. This is not supported by other object file formats.
1727 @item @code{ORG} pseudo-op
1729 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1730 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1731 which changes the location within the current section. Absolute sections are
1732 not supported by other object file formats. The address of a section may be
1733 assigned within a linker script.
1736 There are some other features of the MRI assembler which are not supported by
1737 @command{@value{AS}}, typically either because they are difficult or because they
1738 seem of little consequence. Some of these may be supported in future releases.
1742 @item EBCDIC strings
1744 EBCDIC strings are not supported.
1746 @item packed binary coded decimal
1748 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1749 and @code{DCB.P} pseudo-ops are not supported.
1751 @item @code{FEQU} pseudo-op
1753 The m68k @code{FEQU} pseudo-op is not supported.
1755 @item @code{NOOBJ} pseudo-op
1757 The m68k @code{NOOBJ} pseudo-op is not supported.
1759 @item @code{OPT} branch control options
1761 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1762 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1763 relaxes all branches, whether forward or backward, to an appropriate size, so
1764 these options serve no purpose.
1766 @item @code{OPT} list control options
1768 The following m68k @code{OPT} list control options are ignored: @code{C},
1769 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1770 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1772 @item other @code{OPT} options
1774 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1775 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1777 @item @code{OPT} @code{D} option is default
1779 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1780 @code{OPT NOD} may be used to turn it off.
1782 @item @code{XREF} pseudo-op.
1784 The m68k @code{XREF} pseudo-op is ignored.
1786 @item @code{.debug} pseudo-op
1788 The i960 @code{.debug} pseudo-op is not supported.
1790 @item @code{.extended} pseudo-op
1792 The i960 @code{.extended} pseudo-op is not supported.
1794 @item @code{.list} pseudo-op.
1796 The various options of the i960 @code{.list} pseudo-op are not supported.
1798 @item @code{.optimize} pseudo-op
1800 The i960 @code{.optimize} pseudo-op is not supported.
1802 @item @code{.output} pseudo-op
1804 The i960 @code{.output} pseudo-op is not supported.
1806 @item @code{.setreal} pseudo-op
1808 The i960 @code{.setreal} pseudo-op is not supported.
1813 @section Dependency Tracking: @option{--MD}
1816 @cindex dependency tracking
1819 @command{@value{AS}} can generate a dependency file for the file it creates. This
1820 file consists of a single rule suitable for @code{make} describing the
1821 dependencies of the main source file.
1823 The rule is written to the file named in its argument.
1825 This feature is used in the automatic updating of makefiles.
1828 @section Name the Object File: @option{-o}
1831 @cindex naming object file
1832 @cindex object file name
1833 There is always one object file output when you run @command{@value{AS}}. By
1834 default it has the name
1837 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1851 You use this option (which takes exactly one filename) to give the
1852 object file a different name.
1854 Whatever the object file is called, @command{@value{AS}} overwrites any
1855 existing file of the same name.
1858 @section Join Data and Text Sections: @option{-R}
1861 @cindex data and text sections, joining
1862 @cindex text and data sections, joining
1863 @cindex joining text and data sections
1864 @cindex merging text and data sections
1865 @option{-R} tells @command{@value{AS}} to write the object file as if all
1866 data-section data lives in the text section. This is only done at
1867 the very last moment: your binary data are the same, but data
1868 section parts are relocated differently. The data section part of
1869 your object file is zero bytes long because all its bytes are
1870 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1872 When you specify @option{-R} it would be possible to generate shorter
1873 address displacements (because we do not have to cross between text and
1874 data section). We refrain from doing this simply for compatibility with
1875 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1878 When @command{@value{AS}} is configured for COFF or ELF output,
1879 this option is only useful if you use sections named @samp{.text} and
1884 @option{-R} is not supported for any of the HPPA targets. Using
1885 @option{-R} generates a warning from @command{@value{AS}}.
1889 @section Display Assembly Statistics: @option{--statistics}
1891 @kindex --statistics
1892 @cindex statistics, about assembly
1893 @cindex time, total for assembly
1894 @cindex space used, maximum for assembly
1895 Use @samp{--statistics} to display two statistics about the resources used by
1896 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1897 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1900 @node traditional-format
1901 @section Compatible Output: @option{--traditional-format}
1903 @kindex --traditional-format
1904 For some targets, the output of @command{@value{AS}} is different in some ways
1905 from the output of some existing assembler. This switch requests
1906 @command{@value{AS}} to use the traditional format instead.
1908 For example, it disables the exception frame optimizations which
1909 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
1912 @section Announce Version: @option{-v}
1916 @cindex assembler version
1917 @cindex version of assembler
1918 You can find out what version of as is running by including the
1919 option @samp{-v} (which you can also spell as @samp{-version}) on the
1923 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
1925 @command{@value{AS}} should never give a warning or error message when
1926 assembling compiler output. But programs written by people often
1927 cause @command{@value{AS}} to give a warning that a particular assumption was
1928 made. All such warnings are directed to the standard error file.
1932 @cindex suppressing warnings
1933 @cindex warnings, suppressing
1934 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
1935 This only affects the warning messages: it does not change any particular of
1936 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
1939 @kindex --fatal-warnings
1940 @cindex errors, caused by warnings
1941 @cindex warnings, causing error
1942 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
1943 files that generate warnings to be in error.
1946 @cindex warnings, switching on
1947 You can switch these options off again by specifying @option{--warn}, which
1948 causes warnings to be output as usual.
1951 @section Generate Object File in Spite of Errors: @option{-Z}
1952 @cindex object file, after errors
1953 @cindex errors, continuing after
1954 After an error message, @command{@value{AS}} normally produces no output. If for
1955 some reason you are interested in object file output even after
1956 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
1957 option. If there are any errors, @command{@value{AS}} continues anyways, and
1958 writes an object file after a final warning message of the form @samp{@var{n}
1959 errors, @var{m} warnings, generating bad object file.}
1964 @cindex machine-independent syntax
1965 @cindex syntax, machine-independent
1966 This chapter describes the machine-independent syntax allowed in a
1967 source file. @command{@value{AS}} syntax is similar to what many other
1968 assemblers use; it is inspired by the BSD 4.2
1973 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
1977 * Preprocessing:: Preprocessing
1978 * Whitespace:: Whitespace
1979 * Comments:: Comments
1980 * Symbol Intro:: Symbols
1981 * Statements:: Statements
1982 * Constants:: Constants
1986 @section Preprocessing
1988 @cindex preprocessing
1989 The @command{@value{AS}} internal preprocessor:
1991 @cindex whitespace, removed by preprocessor
1993 adjusts and removes extra whitespace. It leaves one space or tab before
1994 the keywords on a line, and turns any other whitespace on the line into
1997 @cindex comments, removed by preprocessor
1999 removes all comments, replacing them with a single space, or an
2000 appropriate number of newlines.
2002 @cindex constants, converted by preprocessor
2004 converts character constants into the appropriate numeric values.
2007 It does not do macro processing, include file handling, or
2008 anything else you may get from your C compiler's preprocessor. You can
2009 do include file processing with the @code{.include} directive
2010 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2011 to get other ``CPP'' style preprocessing by giving the input file a
2012 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
2013 Output, gcc.info, Using GNU CC}.
2015 Excess whitespace, comments, and character constants
2016 cannot be used in the portions of the input text that are not
2019 @cindex turning preprocessing on and off
2020 @cindex preprocessing, turning on and off
2023 If the first line of an input file is @code{#NO_APP} or if you use the
2024 @samp{-f} option, whitespace and comments are not removed from the input file.
2025 Within an input file, you can ask for whitespace and comment removal in
2026 specific portions of the by putting a line that says @code{#APP} before the
2027 text that may contain whitespace or comments, and putting a line that says
2028 @code{#NO_APP} after this text. This feature is mainly intend to support
2029 @code{asm} statements in compilers whose output is otherwise free of comments
2036 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2037 Whitespace is used to separate symbols, and to make programs neater for
2038 people to read. Unless within character constants
2039 (@pxref{Characters,,Character Constants}), any whitespace means the same
2040 as exactly one space.
2046 There are two ways of rendering comments to @command{@value{AS}}. In both
2047 cases the comment is equivalent to one space.
2049 Anything from @samp{/*} through the next @samp{*/} is a comment.
2050 This means you may not nest these comments.
2054 The only way to include a newline ('\n') in a comment
2055 is to use this sort of comment.
2058 /* This sort of comment does not nest. */
2061 @cindex line comment character
2062 Anything from the @dfn{line comment} character to the next newline
2063 is considered a comment and is ignored. The line comment character is
2065 @samp{;} for the AMD 29K family;
2068 @samp{;} on the ARC;
2071 @samp{@@} on the ARM;
2074 @samp{;} for the H8/300 family;
2077 @samp{!} for the H8/500 family;
2080 @samp{;} for the HPPA;
2083 @samp{#} on the i386 and x86-64;
2086 @samp{#} on the i960;
2089 @samp{;} for the PDP-11;
2092 @samp{;} for picoJava;
2095 @samp{#} for Motorola PowerPC;
2098 @samp{!} for the Renesas / SuperH SH;
2101 @samp{!} on the SPARC;
2104 @samp{#} on the ip2k;
2107 @samp{#} on the m32r;
2110 @samp{|} on the 680x0;
2113 @samp{#} on the 68HC11 and 68HC12;
2116 @samp{;} on the M880x0;
2119 @samp{#} on the Vax;
2122 @samp{!} for the Z8000;
2125 @samp{#} on the V850;
2128 @samp{#} for Xtensa systems;
2130 see @ref{Machine Dependencies}. @refill
2131 @c FIXME What about i860?
2134 On some machines there are two different line comment characters. One
2135 character only begins a comment if it is the first non-whitespace character on
2136 a line, while the other always begins a comment.
2140 The V850 assembler also supports a double dash as starting a comment that
2141 extends to the end of the line.
2147 @cindex lines starting with @code{#}
2148 @cindex logical line numbers
2149 To be compatible with past assemblers, lines that begin with @samp{#} have a
2150 special interpretation. Following the @samp{#} should be an absolute
2151 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2152 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
2153 new logical file name. The rest of the line, if any, should be whitespace.
2155 If the first non-whitespace characters on the line are not numeric,
2156 the line is ignored. (Just like a comment.)
2159 # This is an ordinary comment.
2160 # 42-6 "new_file_name" # New logical file name
2161 # This is logical line # 36.
2163 This feature is deprecated, and may disappear from future versions
2164 of @command{@value{AS}}.
2169 @cindex characters used in symbols
2170 @ifclear SPECIAL-SYMS
2171 A @dfn{symbol} is one or more characters chosen from the set of all
2172 letters (both upper and lower case), digits and the three characters
2178 A @dfn{symbol} is one or more characters chosen from the set of all
2179 letters (both upper and lower case), digits and the three characters
2180 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2186 On most machines, you can also use @code{$} in symbol names; exceptions
2187 are noted in @ref{Machine Dependencies}.
2189 No symbol may begin with a digit. Case is significant.
2190 There is no length limit: all characters are significant. Symbols are
2191 delimited by characters not in that set, or by the beginning of a file
2192 (since the source program must end with a newline, the end of a file is
2193 not a possible symbol delimiter). @xref{Symbols}.
2194 @cindex length of symbols
2199 @cindex statements, structure of
2200 @cindex line separator character
2201 @cindex statement separator character
2203 @ifclear abnormal-separator
2204 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2205 semicolon (@samp{;}). The newline or semicolon is considered part of
2206 the preceding statement. Newlines and semicolons within character
2207 constants are an exception: they do not end statements.
2209 @ifset abnormal-separator
2211 A @dfn{statement} ends at a newline character (@samp{\n}) or an ``at''
2212 sign (@samp{@@}). The newline or at sign is considered part of the
2213 preceding statement. Newlines and at signs within character constants
2214 are an exception: they do not end statements.
2217 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2218 point (@samp{!}). The newline or exclamation point is considered part of the
2219 preceding statement. Newlines and exclamation points within character
2220 constants are an exception: they do not end statements.
2223 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2224 H8/300) a dollar sign (@samp{$}); or (for the
2227 (@samp{;}). The newline or separator character is considered part of
2228 the preceding statement. Newlines and separators within character
2229 constants are an exception: they do not end statements.
2234 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2235 separator character. (The line separator is usually @samp{;}, unless
2236 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
2237 newline or separator character is considered part of the preceding
2238 statement. Newlines and separators within character constants are an
2239 exception: they do not end statements.
2242 @cindex newline, required at file end
2243 @cindex EOF, newline must precede
2244 It is an error to end any statement with end-of-file: the last
2245 character of any input file should be a newline.@refill
2247 An empty statement is allowed, and may include whitespace. It is ignored.
2249 @cindex instructions and directives
2250 @cindex directives and instructions
2251 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2252 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2254 A statement begins with zero or more labels, optionally followed by a
2255 key symbol which determines what kind of statement it is. The key
2256 symbol determines the syntax of the rest of the statement. If the
2257 symbol begins with a dot @samp{.} then the statement is an assembler
2258 directive: typically valid for any computer. If the symbol begins with
2259 a letter the statement is an assembly language @dfn{instruction}: it
2260 assembles into a machine language instruction.
2262 Different versions of @command{@value{AS}} for different computers
2263 recognize different instructions. In fact, the same symbol may
2264 represent a different instruction in a different computer's assembly
2268 @cindex @code{:} (label)
2269 @cindex label (@code{:})
2270 A label is a symbol immediately followed by a colon (@code{:}).
2271 Whitespace before a label or after a colon is permitted, but you may not
2272 have whitespace between a label's symbol and its colon. @xref{Labels}.
2275 For HPPA targets, labels need not be immediately followed by a colon, but
2276 the definition of a label must begin in column zero. This also implies that
2277 only one label may be defined on each line.
2281 label: .directive followed by something
2282 another_label: # This is an empty statement.
2283 instruction operand_1, operand_2, @dots{}
2290 A constant is a number, written so that its value is known by
2291 inspection, without knowing any context. Like this:
2294 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2295 .ascii "Ring the bell\7" # A string constant.
2296 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2297 .float 0f-314159265358979323846264338327\
2298 95028841971.693993751E-40 # - pi, a flonum.
2303 * Characters:: Character Constants
2304 * Numbers:: Number Constants
2308 @subsection Character Constants
2310 @cindex character constants
2311 @cindex constants, character
2312 There are two kinds of character constants. A @dfn{character} stands
2313 for one character in one byte and its value may be used in
2314 numeric expressions. String constants (properly called string
2315 @emph{literals}) are potentially many bytes and their values may not be
2316 used in arithmetic expressions.
2320 * Chars:: Characters
2324 @subsubsection Strings
2326 @cindex string constants
2327 @cindex constants, string
2328 A @dfn{string} is written between double-quotes. It may contain
2329 double-quotes or null characters. The way to get special characters
2330 into a string is to @dfn{escape} these characters: precede them with
2331 a backslash @samp{\} character. For example @samp{\\} represents
2332 one backslash: the first @code{\} is an escape which tells
2333 @command{@value{AS}} to interpret the second character literally as a backslash
2334 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2335 escape character). The complete list of escapes follows.
2337 @cindex escape codes, character
2338 @cindex character escape codes
2341 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2343 @cindex @code{\b} (backspace character)
2344 @cindex backspace (@code{\b})
2346 Mnemonic for backspace; for ASCII this is octal code 010.
2349 @c Mnemonic for EOText; for ASCII this is octal code 004.
2351 @cindex @code{\f} (formfeed character)
2352 @cindex formfeed (@code{\f})
2354 Mnemonic for FormFeed; for ASCII this is octal code 014.
2356 @cindex @code{\n} (newline character)
2357 @cindex newline (@code{\n})
2359 Mnemonic for newline; for ASCII this is octal code 012.
2362 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2364 @cindex @code{\r} (carriage return character)
2365 @cindex carriage return (@code{\r})
2367 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2370 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2371 @c other assemblers.
2373 @cindex @code{\t} (tab)
2374 @cindex tab (@code{\t})
2376 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2379 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2380 @c @item \x @var{digit} @var{digit} @var{digit}
2381 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2383 @cindex @code{\@var{ddd}} (octal character code)
2384 @cindex octal character code (@code{\@var{ddd}})
2385 @item \ @var{digit} @var{digit} @var{digit}
2386 An octal character code. The numeric code is 3 octal digits.
2387 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2388 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2390 @cindex @code{\@var{xd...}} (hex character code)
2391 @cindex hex character code (@code{\@var{xd...}})
2392 @item \@code{x} @var{hex-digits...}
2393 A hex character code. All trailing hex digits are combined. Either upper or
2394 lower case @code{x} works.
2396 @cindex @code{\\} (@samp{\} character)
2397 @cindex backslash (@code{\\})
2399 Represents one @samp{\} character.
2402 @c Represents one @samp{'} (accent acute) character.
2403 @c This is needed in single character literals
2404 @c (@xref{Characters,,Character Constants}.) to represent
2407 @cindex @code{\"} (doublequote character)
2408 @cindex doublequote (@code{\"})
2410 Represents one @samp{"} character. Needed in strings to represent
2411 this character, because an unescaped @samp{"} would end the string.
2413 @item \ @var{anything-else}
2414 Any other character when escaped by @kbd{\} gives a warning, but
2415 assembles as if the @samp{\} was not present. The idea is that if
2416 you used an escape sequence you clearly didn't want the literal
2417 interpretation of the following character. However @command{@value{AS}} has no
2418 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2419 code and warns you of the fact.
2422 Which characters are escapable, and what those escapes represent,
2423 varies widely among assemblers. The current set is what we think
2424 the BSD 4.2 assembler recognizes, and is a subset of what most C
2425 compilers recognize. If you are in doubt, do not use an escape
2429 @subsubsection Characters
2431 @cindex single character constant
2432 @cindex character, single
2433 @cindex constant, single character
2434 A single character may be written as a single quote immediately
2435 followed by that character. The same escapes apply to characters as
2436 to strings. So if you want to write the character backslash, you
2437 must write @kbd{'\\} where the first @code{\} escapes the second
2438 @code{\}. As you can see, the quote is an acute accent, not a
2439 grave accent. A newline
2441 @ifclear abnormal-separator
2442 (or semicolon @samp{;})
2444 @ifset abnormal-separator
2446 (or at sign @samp{@@})
2449 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2450 Renesas SH or H8/500)
2454 immediately following an acute accent is taken as a literal character
2455 and does not count as the end of a statement. The value of a character
2456 constant in a numeric expression is the machine's byte-wide code for
2457 that character. @command{@value{AS}} assumes your character code is ASCII:
2458 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2461 @subsection Number Constants
2463 @cindex constants, number
2464 @cindex number constants
2465 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2466 are stored in the target machine. @emph{Integers} are numbers that
2467 would fit into an @code{int} in the C language. @emph{Bignums} are
2468 integers, but they are stored in more than 32 bits. @emph{Flonums}
2469 are floating point numbers, described below.
2472 * Integers:: Integers
2477 * Bit Fields:: Bit Fields
2483 @subsubsection Integers
2485 @cindex constants, integer
2487 @cindex binary integers
2488 @cindex integers, binary
2489 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2490 the binary digits @samp{01}.
2492 @cindex octal integers
2493 @cindex integers, octal
2494 An octal integer is @samp{0} followed by zero or more of the octal
2495 digits (@samp{01234567}).
2497 @cindex decimal integers
2498 @cindex integers, decimal
2499 A decimal integer starts with a non-zero digit followed by zero or
2500 more digits (@samp{0123456789}).
2502 @cindex hexadecimal integers
2503 @cindex integers, hexadecimal
2504 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2505 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2507 Integers have the usual values. To denote a negative integer, use
2508 the prefix operator @samp{-} discussed under expressions
2509 (@pxref{Prefix Ops,,Prefix Operators}).
2512 @subsubsection Bignums
2515 @cindex constants, bignum
2516 A @dfn{bignum} has the same syntax and semantics as an integer
2517 except that the number (or its negative) takes more than 32 bits to
2518 represent in binary. The distinction is made because in some places
2519 integers are permitted while bignums are not.
2522 @subsubsection Flonums
2524 @cindex floating point numbers
2525 @cindex constants, floating point
2527 @cindex precision, floating point
2528 A @dfn{flonum} represents a floating point number. The translation is
2529 indirect: a decimal floating point number from the text is converted by
2530 @command{@value{AS}} to a generic binary floating point number of more than
2531 sufficient precision. This generic floating point number is converted
2532 to a particular computer's floating point format (or formats) by a
2533 portion of @command{@value{AS}} specialized to that computer.
2535 A flonum is written by writing (in order)
2540 (@samp{0} is optional on the HPPA.)
2544 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2546 @kbd{e} is recommended. Case is not important.
2548 @c FIXME: verify if flonum syntax really this vague for most cases
2549 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2550 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2553 On the H8/300, H8/500,
2554 Renesas / SuperH SH,
2555 and AMD 29K architectures, the letter must be
2556 one of the letters @samp{DFPRSX} (in upper or lower case).
2558 On the ARC, the letter must be one of the letters @samp{DFRS}
2559 (in upper or lower case).
2561 On the Intel 960 architecture, the letter must be
2562 one of the letters @samp{DFT} (in upper or lower case).
2564 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2568 One of the letters @samp{DFPRSX} (in upper or lower case).
2571 One of the letters @samp{DFRS} (in upper or lower case).
2574 One of the letters @samp{DFPRSX} (in upper or lower case).
2577 The letter @samp{E} (upper case only).
2580 One of the letters @samp{DFT} (in upper or lower case).
2585 An optional sign: either @samp{+} or @samp{-}.
2588 An optional @dfn{integer part}: zero or more decimal digits.
2591 An optional @dfn{fractional part}: @samp{.} followed by zero
2592 or more decimal digits.
2595 An optional exponent, consisting of:
2599 An @samp{E} or @samp{e}.
2600 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2601 @c principle this can perfectly well be different on different targets.
2603 Optional sign: either @samp{+} or @samp{-}.
2605 One or more decimal digits.
2610 At least one of the integer part or the fractional part must be
2611 present. The floating point number has the usual base-10 value.
2613 @command{@value{AS}} does all processing using integers. Flonums are computed
2614 independently of any floating point hardware in the computer running
2615 @command{@value{AS}}.
2619 @c Bit fields are written as a general facility but are also controlled
2620 @c by a conditional-compilation flag---which is as of now (21mar91)
2621 @c turned on only by the i960 config of GAS.
2623 @subsubsection Bit Fields
2626 @cindex constants, bit field
2627 You can also define numeric constants as @dfn{bit fields}.
2628 specify two numbers separated by a colon---
2630 @var{mask}:@var{value}
2633 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2636 The resulting number is then packed
2638 @c this conditional paren in case bit fields turned on elsewhere than 960
2639 (in host-dependent byte order)
2641 into a field whose width depends on which assembler directive has the
2642 bit-field as its argument. Overflow (a result from the bitwise and
2643 requiring more binary digits to represent) is not an error; instead,
2644 more constants are generated, of the specified width, beginning with the
2645 least significant digits.@refill
2647 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2648 @code{.short}, and @code{.word} accept bit-field arguments.
2653 @chapter Sections and Relocation
2658 * Secs Background:: Background
2659 * Ld Sections:: Linker Sections
2660 * As Sections:: Assembler Internal Sections
2661 * Sub-Sections:: Sub-Sections
2665 @node Secs Background
2668 Roughly, a section is a range of addresses, with no gaps; all data
2669 ``in'' those addresses is treated the same for some particular purpose.
2670 For example there may be a ``read only'' section.
2672 @cindex linker, and assembler
2673 @cindex assembler, and linker
2674 The linker @code{@value{LD}} reads many object files (partial programs) and
2675 combines their contents to form a runnable program. When @command{@value{AS}}
2676 emits an object file, the partial program is assumed to start at address 0.
2677 @code{@value{LD}} assigns the final addresses for the partial program, so that
2678 different partial programs do not overlap. This is actually an
2679 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2682 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2683 addresses. These blocks slide to their run-time addresses as rigid
2684 units; their length does not change and neither does the order of bytes
2685 within them. Such a rigid unit is called a @emph{section}. Assigning
2686 run-time addresses to sections is called @dfn{relocation}. It includes
2687 the task of adjusting mentions of object-file addresses so they refer to
2688 the proper run-time addresses.
2690 For the H8/300 and H8/500,
2691 and for the Renesas / SuperH SH,
2692 @command{@value{AS}} pads sections if needed to
2693 ensure they end on a word (sixteen bit) boundary.
2696 @cindex standard assembler sections
2697 An object file written by @command{@value{AS}} has at least three sections, any
2698 of which may be empty. These are named @dfn{text}, @dfn{data} and
2703 When it generates COFF or ELF output,
2705 @command{@value{AS}} can also generate whatever other named sections you specify
2706 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2707 If you do not use any directives that place output in the @samp{.text}
2708 or @samp{.data} sections, these sections still exist, but are empty.
2713 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2715 @command{@value{AS}} can also generate whatever other named sections you
2716 specify using the @samp{.space} and @samp{.subspace} directives. See
2717 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2718 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2719 assembler directives.
2722 Additionally, @command{@value{AS}} uses different names for the standard
2723 text, data, and bss sections when generating SOM output. Program text
2724 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2725 BSS into @samp{$BSS$}.
2729 Within the object file, the text section starts at address @code{0}, the
2730 data section follows, and the bss section follows the data section.
2733 When generating either SOM or ELF output files on the HPPA, the text
2734 section starts at address @code{0}, the data section at address
2735 @code{0x4000000}, and the bss section follows the data section.
2738 To let @code{@value{LD}} know which data changes when the sections are
2739 relocated, and how to change that data, @command{@value{AS}} also writes to the
2740 object file details of the relocation needed. To perform relocation
2741 @code{@value{LD}} must know, each time an address in the object
2745 Where in the object file is the beginning of this reference to
2748 How long (in bytes) is this reference?
2750 Which section does the address refer to? What is the numeric value of
2752 (@var{address}) @minus{} (@var{start-address of section})?
2755 Is the reference to an address ``Program-Counter relative''?
2758 @cindex addresses, format of
2759 @cindex section-relative addressing
2760 In fact, every address @command{@value{AS}} ever uses is expressed as
2762 (@var{section}) + (@var{offset into section})
2765 Further, most expressions @command{@value{AS}} computes have this section-relative
2768 (For some object formats, such as SOM for the HPPA, some expressions are
2769 symbol-relative instead.)
2772 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2773 @var{N} into section @var{secname}.''
2775 Apart from text, data and bss sections you need to know about the
2776 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2777 addresses in the absolute section remain unchanged. For example, address
2778 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2779 @code{@value{LD}}. Although the linker never arranges two partial programs'
2780 data sections with overlapping addresses after linking, @emph{by definition}
2781 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2782 part of a program is always the same address when the program is running as
2783 address @code{@{absolute@ 239@}} in any other part of the program.
2785 The idea of sections is extended to the @dfn{undefined} section. Any
2786 address whose section is unknown at assembly time is by definition
2787 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2788 Since numbers are always defined, the only way to generate an undefined
2789 address is to mention an undefined symbol. A reference to a named
2790 common block would be such a symbol: its value is unknown at assembly
2791 time so it has section @emph{undefined}.
2793 By analogy the word @emph{section} is used to describe groups of sections in
2794 the linked program. @code{@value{LD}} puts all partial programs' text
2795 sections in contiguous addresses in the linked program. It is
2796 customary to refer to the @emph{text section} of a program, meaning all
2797 the addresses of all partial programs' text sections. Likewise for
2798 data and bss sections.
2800 Some sections are manipulated by @code{@value{LD}}; others are invented for
2801 use of @command{@value{AS}} and have no meaning except during assembly.
2804 @section Linker Sections
2805 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2810 @cindex named sections
2811 @cindex sections, named
2812 @item named sections
2815 @cindex text section
2816 @cindex data section
2820 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2821 separate but equal sections. Anything you can say of one section is
2824 When the program is running, however, it is
2825 customary for the text section to be unalterable. The
2826 text section is often shared among processes: it contains
2827 instructions, constants and the like. The data section of a running
2828 program is usually alterable: for example, C variables would be stored
2829 in the data section.
2834 This section contains zeroed bytes when your program begins running. It
2835 is used to hold uninitialized variables or common storage. The length of
2836 each partial program's bss section is important, but because it starts
2837 out containing zeroed bytes there is no need to store explicit zero
2838 bytes in the object file. The bss section was invented to eliminate
2839 those explicit zeros from object files.
2841 @cindex absolute section
2842 @item absolute section
2843 Address 0 of this section is always ``relocated'' to runtime address 0.
2844 This is useful if you want to refer to an address that @code{@value{LD}} must
2845 not change when relocating. In this sense we speak of absolute
2846 addresses being ``unrelocatable'': they do not change during relocation.
2848 @cindex undefined section
2849 @item undefined section
2850 This ``section'' is a catch-all for address references to objects not in
2851 the preceding sections.
2852 @c FIXME: ref to some other doc on obj-file formats could go here.
2855 @cindex relocation example
2856 An idealized example of three relocatable sections follows.
2858 The example uses the traditional section names @samp{.text} and @samp{.data}.
2860 Memory addresses are on the horizontal axis.
2864 @c END TEXI2ROFF-KILL
2867 partial program # 1: |ttttt|dddd|00|
2874 partial program # 2: |TTT|DDD|000|
2877 +--+---+-----+--+----+---+-----+~~
2878 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2879 +--+---+-----+--+----+---+-----+~~
2881 addresses: 0 @dots{}
2888 \line{\it Partial program \#1: \hfil}
2889 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2890 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2892 \line{\it Partial program \#2: \hfil}
2893 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2894 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2896 \line{\it linked program: \hfil}
2897 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2898 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2899 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2900 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2902 \line{\it addresses: \hfil}
2906 @c END TEXI2ROFF-KILL
2909 @section Assembler Internal Sections
2911 @cindex internal assembler sections
2912 @cindex sections in messages, internal
2913 These sections are meant only for the internal use of @command{@value{AS}}. They
2914 have no meaning at run-time. You do not really need to know about these
2915 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2916 warning messages, so it might be helpful to have an idea of their
2917 meanings to @command{@value{AS}}. These sections are used to permit the
2918 value of every expression in your assembly language program to be a
2919 section-relative address.
2922 @cindex assembler internal logic error
2923 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2924 An internal assembler logic error has been found. This means there is a
2925 bug in the assembler.
2927 @cindex expr (internal section)
2929 The assembler stores complex expression internally as combinations of
2930 symbols. When it needs to represent an expression as a symbol, it puts
2931 it in the expr section.
2933 @c FIXME item transfer[t] vector preload
2934 @c FIXME item transfer[t] vector postload
2935 @c FIXME item register
2939 @section Sub-Sections
2941 @cindex numbered subsections
2942 @cindex grouping data
2948 fall into two sections: text and data.
2950 You may have separate groups of
2952 data in named sections
2956 data in named sections
2962 that you want to end up near to each other in the object file, even though they
2963 are not contiguous in the assembler source. @command{@value{AS}} allows you to
2964 use @dfn{subsections} for this purpose. Within each section, there can be
2965 numbered subsections with values from 0 to 8192. Objects assembled into the
2966 same subsection go into the object file together with other objects in the same
2967 subsection. For example, a compiler might want to store constants in the text
2968 section, but might not want to have them interspersed with the program being
2969 assembled. In this case, the compiler could issue a @samp{.text 0} before each
2970 section of code being output, and a @samp{.text 1} before each group of
2971 constants being output.
2973 Subsections are optional. If you do not use subsections, everything
2974 goes in subsection number zero.
2977 Each subsection is zero-padded up to a multiple of four bytes.
2978 (Subsections may be padded a different amount on different flavors
2979 of @command{@value{AS}}.)
2983 On the H8/300 and H8/500 platforms, each subsection is zero-padded to a word
2984 boundary (two bytes).
2985 The same is true on the Renesas SH.
2988 @c FIXME section padding (alignment)?
2989 @c Rich Pixley says padding here depends on target obj code format; that
2990 @c doesn't seem particularly useful to say without further elaboration,
2991 @c so for now I say nothing about it. If this is a generic BFD issue,
2992 @c these paragraphs might need to vanish from this manual, and be
2993 @c discussed in BFD chapter of binutils (or some such).
2996 On the AMD 29K family, no particular padding is added to section or
2997 subsection sizes; @value{AS} forces no alignment on this platform.
3001 Subsections appear in your object file in numeric order, lowest numbered
3002 to highest. (All this to be compatible with other people's assemblers.)
3003 The object file contains no representation of subsections; @code{@value{LD}} and
3004 other programs that manipulate object files see no trace of them.
3005 They just see all your text subsections as a text section, and all your
3006 data subsections as a data section.
3008 To specify which subsection you want subsequent statements assembled
3009 into, use a numeric argument to specify it, in a @samp{.text
3010 @var{expression}} or a @samp{.data @var{expression}} statement.
3013 When generating COFF or ELF output, you
3018 can also use an extra subsection
3019 argument with arbitrary named sections: @samp{.section @var{name},
3022 @var{Expression} should be an absolute expression.
3023 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
3024 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3025 begins in @code{text 0}. For instance:
3027 .text 0 # The default subsection is text 0 anyway.
3028 .ascii "This lives in the first text subsection. *"
3030 .ascii "But this lives in the second text subsection."
3032 .ascii "This lives in the data section,"
3033 .ascii "in the first data subsection."
3035 .ascii "This lives in the first text section,"
3036 .ascii "immediately following the asterisk (*)."
3039 Each section has a @dfn{location counter} incremented by one for every byte
3040 assembled into that section. Because subsections are merely a convenience
3041 restricted to @command{@value{AS}} there is no concept of a subsection location
3042 counter. There is no way to directly manipulate a location counter---but the
3043 @code{.align} directive changes it, and any label definition captures its
3044 current value. The location counter of the section where statements are being
3045 assembled is said to be the @dfn{active} location counter.
3048 @section bss Section
3051 @cindex common variable storage
3052 The bss section is used for local common variable storage.
3053 You may allocate address space in the bss section, but you may
3054 not dictate data to load into it before your program executes. When
3055 your program starts running, all the contents of the bss
3056 section are zeroed bytes.
3058 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3059 @ref{Lcomm,,@code{.lcomm}}.
3061 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3062 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
3065 When assembling for a target which supports multiple sections, such as ELF or
3066 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3067 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3068 section. Typically the section will only contain symbol definitions and
3069 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3076 Symbols are a central concept: the programmer uses symbols to name
3077 things, the linker uses symbols to link, and the debugger uses symbols
3081 @cindex debuggers, and symbol order
3082 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3083 the same order they were declared. This may break some debuggers.
3088 * Setting Symbols:: Giving Symbols Other Values
3089 * Symbol Names:: Symbol Names
3090 * Dot:: The Special Dot Symbol
3091 * Symbol Attributes:: Symbol Attributes
3098 A @dfn{label} is written as a symbol immediately followed by a colon
3099 @samp{:}. The symbol then represents the current value of the
3100 active location counter, and is, for example, a suitable instruction
3101 operand. You are warned if you use the same symbol to represent two
3102 different locations: the first definition overrides any other
3106 On the HPPA, the usual form for a label need not be immediately followed by a
3107 colon, but instead must start in column zero. Only one label may be defined on
3108 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3109 provides a special directive @code{.label} for defining labels more flexibly.
3112 @node Setting Symbols
3113 @section Giving Symbols Other Values
3115 @cindex assigning values to symbols
3116 @cindex symbol values, assigning
3117 A symbol can be given an arbitrary value by writing a symbol, followed
3118 by an equals sign @samp{=}, followed by an expression
3119 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3120 directive. @xref{Set,,@code{.set}}.
3123 @section Symbol Names
3125 @cindex symbol names
3126 @cindex names, symbol
3127 @ifclear SPECIAL-SYMS
3128 Symbol names begin with a letter or with one of @samp{._}. On most
3129 machines, you can also use @code{$} in symbol names; exceptions are
3130 noted in @ref{Machine Dependencies}. That character may be followed by any
3131 string of digits, letters, dollar signs (unless otherwise noted in
3132 @ref{Machine Dependencies}), and underscores.
3135 For the AMD 29K family, @samp{?} is also allowed in the
3136 body of a symbol name, though not at its beginning.
3141 Symbol names begin with a letter or with one of @samp{._}. On the
3142 Renesas SH or the H8/500, you can also use @code{$} in symbol names. That
3143 character may be followed by any string of digits, letters, dollar signs (save
3144 on the H8/300), and underscores.
3148 Case of letters is significant: @code{foo} is a different symbol name
3151 Each symbol has exactly one name. Each name in an assembly language program
3152 refers to exactly one symbol. You may use that symbol name any number of times
3155 @subheading Local Symbol Names
3157 @cindex local symbol names
3158 @cindex symbol names, local
3159 @cindex temporary symbol names
3160 @cindex symbol names, temporary
3161 Local symbols help compilers and programmers use names temporarily.
3162 They create symbols which are guaranteed to be unique over the entire scope of
3163 the input source code and which can be referred to by a simple notation.
3164 To define a local symbol, write a label of the form @samp{@b{N}:} (where @b{N}
3165 represents any positive integer). To refer to the most recent previous
3166 definition of that symbol write @samp{@b{N}b}, using the same number as when
3167 you defined the label. To refer to the next definition of a local label, write
3168 @samp{@b{N}f}--- The @samp{b} stands for``backwards'' and the @samp{f} stands
3171 There is no restriction on how you can use these labels, and you can reuse them
3172 too. So that it is possible to repeatedly define the same local label (using
3173 the same number @samp{@b{N}}), although you can only refer to the most recently
3174 defined local label of that number (for a backwards reference) or the next
3175 definition of a specific local label for a forward reference. It is also worth
3176 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3177 implemented in a slightly more efficient manner than the others.
3188 Which is the equivalent of:
3191 label_1: branch label_3
3192 label_2: branch label_1
3193 label_3: branch label_4
3194 label_4: branch label_3
3197 Local symbol names are only a notational device. They are immediately
3198 transformed into more conventional symbol names before the assembler uses them.
3199 The symbol names stored in the symbol table, appearing in error messages and
3200 optionally emitted to the object file. The names are constructed using these
3205 All local labels begin with @samp{L}. Normally both @command{@value{AS}} and
3206 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
3207 used for symbols you are never intended to see. If you use the
3208 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3209 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3210 you may use them in debugging.
3213 This is the number that was used in the local label definition. So if the
3214 label is written @samp{55:} then the number is @samp{55}.
3217 This unusual character is included so you do not accidentally invent a symbol
3218 of the same name. The character has ASCII value of @samp{\002} (control-B).
3220 @item @emph{ordinal number}
3221 This is a serial number to keep the labels distinct. The first definition of
3222 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3223 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3224 the number @samp{1} and its 15th defintion gets @samp{15} as well.
3227 So for example, the first @code{1:} is named @code{L1@kbd{C-B}1}, the 44th
3228 @code{3:} is named @code{L3@kbd{C-B}44}.
3230 @subheading Dollar Local Labels
3231 @cindex dollar local symbols
3233 @code{@value{AS}} also supports an even more local form of local labels called
3234 dollar labels. These labels go out of scope (ie they become undefined) as soon
3235 as a non-local label is defined. Thus they remain valid for only a small
3236 region of the input source code. Normal local labels, by contrast, remain in
3237 scope for the entire file, or until they are redefined by another occurrence of
3238 the same local label.
3240 Dollar labels are defined in exactly the same way as ordinary local labels,
3241 except that instead of being terminated by a colon, they are terminated by a
3242 dollar sign. eg @samp{@b{55$}}.
3244 They can also be distinguished from ordinary local labels by their transformed
3245 name which uses ASCII character @samp{\001} (control-A) as the magic character
3246 to distinguish them from ordinary labels. Thus the 5th defintion of @samp{6$}
3247 is named @samp{L6@kbd{C-A}5}.
3250 @section The Special Dot Symbol
3252 @cindex dot (symbol)
3253 @cindex @code{.} (symbol)
3254 @cindex current address
3255 @cindex location counter
3256 The special symbol @samp{.} refers to the current address that
3257 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3258 .long .} defines @code{melvin} to contain its own address.
3259 Assigning a value to @code{.} is treated the same as a @code{.org}
3260 directive. Thus, the expression @samp{.=.+4} is the same as saying
3261 @ifclear no-space-dir
3270 @node Symbol Attributes
3271 @section Symbol Attributes
3273 @cindex symbol attributes
3274 @cindex attributes, symbol
3275 Every symbol has, as well as its name, the attributes ``Value'' and
3276 ``Type''. Depending on output format, symbols can also have auxiliary
3279 The detailed definitions are in @file{a.out.h}.
3282 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3283 all these attributes, and probably won't warn you. This makes the
3284 symbol an externally defined symbol, which is generally what you
3288 * Symbol Value:: Value
3289 * Symbol Type:: Type
3292 * a.out Symbols:: Symbol Attributes: @code{a.out}
3296 * a.out Symbols:: Symbol Attributes: @code{a.out}
3299 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3304 * COFF Symbols:: Symbol Attributes for COFF
3307 * SOM Symbols:: Symbol Attributes for SOM
3314 @cindex value of a symbol
3315 @cindex symbol value
3316 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3317 location in the text, data, bss or absolute sections the value is the
3318 number of addresses from the start of that section to the label.
3319 Naturally for text, data and bss sections the value of a symbol changes
3320 as @code{@value{LD}} changes section base addresses during linking. Absolute
3321 symbols' values do not change during linking: that is why they are
3324 The value of an undefined symbol is treated in a special way. If it is
3325 0 then the symbol is not defined in this assembler source file, and
3326 @code{@value{LD}} tries to determine its value from other files linked into the
3327 same program. You make this kind of symbol simply by mentioning a symbol
3328 name without defining it. A non-zero value represents a @code{.comm}
3329 common declaration. The value is how much common storage to reserve, in
3330 bytes (addresses). The symbol refers to the first address of the
3336 @cindex type of a symbol
3338 The type attribute of a symbol contains relocation (section)
3339 information, any flag settings indicating that a symbol is external, and
3340 (optionally), other information for linkers and debuggers. The exact
3341 format depends on the object-code output format in use.
3346 @c The following avoids a "widow" subsection title. @group would be
3347 @c better if it were available outside examples.
3350 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3352 @cindex @code{b.out} symbol attributes
3353 @cindex symbol attributes, @code{b.out}
3354 These symbol attributes appear only when @command{@value{AS}} is configured for
3355 one of the Berkeley-descended object output formats---@code{a.out} or
3361 @subsection Symbol Attributes: @code{a.out}
3363 @cindex @code{a.out} symbol attributes
3364 @cindex symbol attributes, @code{a.out}
3370 @subsection Symbol Attributes: @code{a.out}
3372 @cindex @code{a.out} symbol attributes
3373 @cindex symbol attributes, @code{a.out}
3377 * Symbol Desc:: Descriptor
3378 * Symbol Other:: Other
3382 @subsubsection Descriptor
3384 @cindex descriptor, of @code{a.out} symbol
3385 This is an arbitrary 16-bit value. You may establish a symbol's
3386 descriptor value by using a @code{.desc} statement
3387 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3388 @command{@value{AS}}.
3391 @subsubsection Other
3393 @cindex other attribute, of @code{a.out} symbol
3394 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3399 @subsection Symbol Attributes for COFF
3401 @cindex COFF symbol attributes
3402 @cindex symbol attributes, COFF
3404 The COFF format supports a multitude of auxiliary symbol attributes;
3405 like the primary symbol attributes, they are set between @code{.def} and
3406 @code{.endef} directives.
3408 @subsubsection Primary Attributes
3410 @cindex primary attributes, COFF symbols
3411 The symbol name is set with @code{.def}; the value and type,
3412 respectively, with @code{.val} and @code{.type}.
3414 @subsubsection Auxiliary Attributes
3416 @cindex auxiliary attributes, COFF symbols
3417 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3418 @code{.size}, and @code{.tag} can generate auxiliary symbol table
3419 information for COFF.
3424 @subsection Symbol Attributes for SOM
3426 @cindex SOM symbol attributes
3427 @cindex symbol attributes, SOM
3429 The SOM format for the HPPA supports a multitude of symbol attributes set with
3430 the @code{.EXPORT} and @code{.IMPORT} directives.
3432 The attributes are described in @cite{HP9000 Series 800 Assembly
3433 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3434 @code{EXPORT} assembler directive documentation.
3438 @chapter Expressions
3442 @cindex numeric values
3443 An @dfn{expression} specifies an address or numeric value.
3444 Whitespace may precede and/or follow an expression.
3446 The result of an expression must be an absolute number, or else an offset into
3447 a particular section. If an expression is not absolute, and there is not
3448 enough information when @command{@value{AS}} sees the expression to know its
3449 section, a second pass over the source program might be necessary to interpret
3450 the expression---but the second pass is currently not implemented.
3451 @command{@value{AS}} aborts with an error message in this situation.
3454 * Empty Exprs:: Empty Expressions
3455 * Integer Exprs:: Integer Expressions
3459 @section Empty Expressions
3461 @cindex empty expressions
3462 @cindex expressions, empty
3463 An empty expression has no value: it is just whitespace or null.
3464 Wherever an absolute expression is required, you may omit the
3465 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3466 is compatible with other assemblers.
3469 @section Integer Expressions
3471 @cindex integer expressions
3472 @cindex expressions, integer
3473 An @dfn{integer expression} is one or more @emph{arguments} delimited
3474 by @emph{operators}.
3477 * Arguments:: Arguments
3478 * Operators:: Operators
3479 * Prefix Ops:: Prefix Operators
3480 * Infix Ops:: Infix Operators
3484 @subsection Arguments
3486 @cindex expression arguments
3487 @cindex arguments in expressions
3488 @cindex operands in expressions
3489 @cindex arithmetic operands
3490 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3491 contexts arguments are sometimes called ``arithmetic operands''. In
3492 this manual, to avoid confusing them with the ``instruction operands'' of
3493 the machine language, we use the term ``argument'' to refer to parts of
3494 expressions only, reserving the word ``operand'' to refer only to machine
3495 instruction operands.
3497 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3498 @var{section} is one of text, data, bss, absolute,
3499 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3502 Numbers are usually integers.
3504 A number can be a flonum or bignum. In this case, you are warned
3505 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3506 these 32 bits are an integer. You may write integer-manipulating
3507 instructions that act on exotic constants, compatible with other
3510 @cindex subexpressions
3511 Subexpressions are a left parenthesis @samp{(} followed by an integer
3512 expression, followed by a right parenthesis @samp{)}; or a prefix
3513 operator followed by an argument.
3516 @subsection Operators
3518 @cindex operators, in expressions
3519 @cindex arithmetic functions
3520 @cindex functions, in expressions
3521 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3522 operators are followed by an argument. Infix operators appear
3523 between their arguments. Operators may be preceded and/or followed by
3527 @subsection Prefix Operator
3529 @cindex prefix operators
3530 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3531 one argument, which must be absolute.
3533 @c the tex/end tex stuff surrounding this small table is meant to make
3534 @c it align, on the printed page, with the similar table in the next
3535 @c section (which is inside an enumerate).
3537 \global\advance\leftskip by \itemindent
3542 @dfn{Negation}. Two's complement negation.
3544 @dfn{Complementation}. Bitwise not.
3548 \global\advance\leftskip by -\itemindent
3552 @subsection Infix Operators
3554 @cindex infix operators
3555 @cindex operators, permitted arguments
3556 @dfn{Infix operators} take two arguments, one on either side. Operators
3557 have precedence, but operations with equal precedence are performed left
3558 to right. Apart from @code{+} or @option{-}, both arguments must be
3559 absolute, and the result is absolute.
3562 @cindex operator precedence
3563 @cindex precedence of operators
3570 @dfn{Multiplication}.
3573 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3580 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3584 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3588 Intermediate precedence
3593 @dfn{Bitwise Inclusive Or}.
3599 @dfn{Bitwise Exclusive Or}.
3602 @dfn{Bitwise Or Not}.
3609 @cindex addition, permitted arguments
3610 @cindex plus, permitted arguments
3611 @cindex arguments for addition
3613 @dfn{Addition}. If either argument is absolute, the result has the section of
3614 the other argument. You may not add together arguments from different
3617 @cindex subtraction, permitted arguments
3618 @cindex minus, permitted arguments
3619 @cindex arguments for subtraction
3621 @dfn{Subtraction}. If the right argument is absolute, the
3622 result has the section of the left argument.
3623 If both arguments are in the same section, the result is absolute.
3624 You may not subtract arguments from different sections.
3625 @c FIXME is there still something useful to say about undefined - undefined ?
3627 @cindex comparison expressions
3628 @cindex expressions, comparison
3632 @dfn{Is Not Equal To}
3636 @dfn{Is Greater Than}
3638 @dfn{Is Greater Than Or Equal To}
3640 @dfn{Is Less Than Or Equal To}
3642 The comparison operators can be used as infix operators. A true results has a
3643 value of -1 whereas a false result has a value of 0. Note, these operators
3644 perform signed comparisons.
3647 @item Lowest Precedence
3656 These two logical operations can be used to combine the results of sub
3657 expressions. Note, unlike the comparison operators a true result returns a
3658 value of 1 but a false results does still return 0. Also note that the logical
3659 or operator has a slightly lower precedence than logical and.
3664 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3665 address; you can only have a defined section in one of the two arguments.
3668 @chapter Assembler Directives
3670 @cindex directives, machine independent
3671 @cindex pseudo-ops, machine independent
3672 @cindex machine independent directives
3673 All assembler directives have names that begin with a period (@samp{.}).
3674 The rest of the name is letters, usually in lower case.
3676 This chapter discusses directives that are available regardless of the
3677 target machine configuration for the @sc{gnu} assembler.
3679 Some machine configurations provide additional directives.
3680 @xref{Machine Dependencies}.
3683 @ifset machine-directives
3684 @xref{Machine Dependencies} for additional directives.
3689 * Abort:: @code{.abort}
3691 * ABORT:: @code{.ABORT}
3694 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3695 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3696 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3697 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3698 * Byte:: @code{.byte @var{expressions}}
3699 * Comm:: @code{.comm @var{symbol} , @var{length} }
3701 * CFI directives:: @code{.cfi_startproc}, @code{.cfi_endproc}, etc.
3703 * Data:: @code{.data @var{subsection}}
3705 * Def:: @code{.def @var{name}}
3708 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3714 * Double:: @code{.double @var{flonums}}
3715 * Eject:: @code{.eject}
3716 * Else:: @code{.else}
3717 * Elseif:: @code{.elseif}
3720 * Endef:: @code{.endef}
3723 * Endfunc:: @code{.endfunc}
3724 * Endif:: @code{.endif}
3725 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3726 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3728 * Exitm:: @code{.exitm}
3729 * Extern:: @code{.extern}
3730 * Fail:: @code{.fail}
3731 @ifclear no-file-dir
3732 * File:: @code{.file @var{string}}
3735 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3736 * Float:: @code{.float @var{flonums}}
3737 * Func:: @code{.func}
3738 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3740 * Hidden:: @code{.hidden @var{names}}
3743 * hword:: @code{.hword @var{expressions}}
3744 * Ident:: @code{.ident}
3745 * If:: @code{.if @var{absolute expression}}
3746 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3747 * Include:: @code{.include "@var{file}"}
3748 * Int:: @code{.int @var{expressions}}
3750 * Internal:: @code{.internal @var{names}}
3753 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3754 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3755 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3756 * Lflags:: @code{.lflags}
3757 @ifclear no-line-dir
3758 * Line:: @code{.line @var{line-number}}
3761 * Ln:: @code{.ln @var{line-number}}
3762 * Linkonce:: @code{.linkonce [@var{type}]}
3763 * List:: @code{.list}
3764 * Long:: @code{.long @var{expressions}}
3766 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3769 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3770 * MRI:: @code{.mri @var{val}}
3771 * Nolist:: @code{.nolist}
3772 * Octa:: @code{.octa @var{bignums}}
3773 * Org:: @code{.org @var{new-lc} , @var{fill}}
3774 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3776 * PopSection:: @code{.popsection}
3777 * Previous:: @code{.previous}
3780 * Print:: @code{.print @var{string}}
3782 * Protected:: @code{.protected @var{names}}
3785 * Psize:: @code{.psize @var{lines}, @var{columns}}
3786 * Purgem:: @code{.purgem @var{name}}
3788 * PushSection:: @code{.pushsection @var{name}}
3791 * Quad:: @code{.quad @var{bignums}}
3792 * Rept:: @code{.rept @var{count}}
3793 * Sbttl:: @code{.sbttl "@var{subheading}"}
3795 * Scl:: @code{.scl @var{class}}
3798 * Section:: @code{.section @var{name}}
3801 * Set:: @code{.set @var{symbol}, @var{expression}}
3802 * Short:: @code{.short @var{expressions}}
3803 * Single:: @code{.single @var{flonums}}
3805 * Size:: @code{.size [@var{name} , @var{expression}]}
3808 * Skip:: @code{.skip @var{size} , @var{fill}}
3809 * Sleb128:: @code{.sleb128 @var{expressions}}
3810 * Space:: @code{.space @var{size} , @var{fill}}
3812 * Stab:: @code{.stabd, .stabn, .stabs}
3815 * String:: @code{.string "@var{str}"}
3816 * Struct:: @code{.struct @var{expression}}
3818 * SubSection:: @code{.subsection}
3819 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3823 * Tag:: @code{.tag @var{structname}}
3826 * Text:: @code{.text @var{subsection}}
3827 * Title:: @code{.title "@var{heading}"}
3829 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3832 * Uleb128:: @code{.uleb128 @var{expressions}}
3834 * Val:: @code{.val @var{addr}}
3838 * Version:: @code{.version "@var{string}"}
3839 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3840 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3841 * Weak:: @code{.weak @var{names}}
3844 * Word:: @code{.word @var{expressions}}
3845 * Deprecated:: Deprecated Directives
3849 @section @code{.abort}
3851 @cindex @code{abort} directive
3852 @cindex stopping the assembly
3853 This directive stops the assembly immediately. It is for
3854 compatibility with other assemblers. The original idea was that the
3855 assembly language source would be piped into the assembler. If the sender
3856 of the source quit, it could use this directive tells @command{@value{AS}} to
3857 quit also. One day @code{.abort} will not be supported.
3861 @section @code{.ABORT}
3863 @cindex @code{ABORT} directive
3864 When producing COFF output, @command{@value{AS}} accepts this directive as a
3865 synonym for @samp{.abort}.
3868 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3874 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3876 @cindex padding the location counter
3877 @cindex @code{align} directive
3878 Pad the location counter (in the current subsection) to a particular storage
3879 boundary. The first expression (which must be absolute) is the alignment
3880 required, as described below.
3882 The second expression (also absolute) gives the fill value to be stored in the
3883 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3884 padding bytes are normally zero. However, on some systems, if the section is
3885 marked as containing code and the fill value is omitted, the space is filled
3886 with no-op instructions.
3888 The third expression is also absolute, and is also optional. If it is present,
3889 it is the maximum number of bytes that should be skipped by this alignment
3890 directive. If doing the alignment would require skipping more bytes than the
3891 specified maximum, then the alignment is not done at all. You can omit the
3892 fill value (the second argument) entirely by simply using two commas after the
3893 required alignment; this can be useful if you want the alignment to be filled
3894 with no-op instructions when appropriate.
3896 The way the required alignment is specified varies from system to system.
3897 For the a29k, hppa, m68k, m88k, w65, sparc, Xtensa, and Renesas / SuperH SH,
3898 and i386 using ELF format,
3899 the first expression is the
3900 alignment request in bytes. For example @samp{.align 8} advances
3901 the location counter until it is a multiple of 8. If the location counter
3902 is already a multiple of 8, no change is needed.
3904 For other systems, including the i386 using a.out format, and the arm and
3905 strongarm, it is the
3906 number of low-order zero bits the location counter must have after
3907 advancement. For example @samp{.align 3} advances the location
3908 counter until it a multiple of 8. If the location counter is already a
3909 multiple of 8, no change is needed.
3911 This inconsistency is due to the different behaviors of the various
3912 native assemblers for these systems which GAS must emulate.
3913 GAS also provides @code{.balign} and @code{.p2align} directives,
3914 described later, which have a consistent behavior across all
3915 architectures (but are specific to GAS).
3918 @section @code{.ascii "@var{string}"}@dots{}
3920 @cindex @code{ascii} directive
3921 @cindex string literals
3922 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3923 separated by commas. It assembles each string (with no automatic
3924 trailing zero byte) into consecutive addresses.
3927 @section @code{.asciz "@var{string}"}@dots{}
3929 @cindex @code{asciz} directive
3930 @cindex zero-terminated strings
3931 @cindex null-terminated strings
3932 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3933 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3936 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3938 @cindex padding the location counter given number of bytes
3939 @cindex @code{balign} directive
3940 Pad the location counter (in the current subsection) to a particular
3941 storage boundary. The first expression (which must be absolute) is the
3942 alignment request in bytes. For example @samp{.balign 8} advances
3943 the location counter until it is a multiple of 8. If the location counter
3944 is already a multiple of 8, no change is needed.
3946 The second expression (also absolute) gives the fill value to be stored in the
3947 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3948 padding bytes are normally zero. However, on some systems, if the section is
3949 marked as containing code and the fill value is omitted, the space is filled
3950 with no-op instructions.
3952 The third expression is also absolute, and is also optional. If it is present,
3953 it is the maximum number of bytes that should be skipped by this alignment
3954 directive. If doing the alignment would require skipping more bytes than the
3955 specified maximum, then the alignment is not done at all. You can omit the
3956 fill value (the second argument) entirely by simply using two commas after the
3957 required alignment; this can be useful if you want the alignment to be filled
3958 with no-op instructions when appropriate.
3960 @cindex @code{balignw} directive
3961 @cindex @code{balignl} directive
3962 The @code{.balignw} and @code{.balignl} directives are variants of the
3963 @code{.balign} directive. The @code{.balignw} directive treats the fill
3964 pattern as a two byte word value. The @code{.balignl} directives treats the
3965 fill pattern as a four byte longword value. For example, @code{.balignw
3966 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
3967 filled in with the value 0x368d (the exact placement of the bytes depends upon
3968 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
3972 @section @code{.byte @var{expressions}}
3974 @cindex @code{byte} directive
3975 @cindex integers, one byte
3976 @code{.byte} expects zero or more expressions, separated by commas.
3977 Each expression is assembled into the next byte.
3980 @section @code{.comm @var{symbol} , @var{length} }
3982 @cindex @code{comm} directive
3983 @cindex symbol, common
3984 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
3985 common symbol in one object file may be merged with a defined or common symbol
3986 of the same name in another object file. If @code{@value{LD}} does not see a
3987 definition for the symbol--just one or more common symbols--then it will
3988 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
3989 absolute expression. If @code{@value{LD}} sees multiple common symbols with
3990 the same name, and they do not all have the same size, it will allocate space
3991 using the largest size.
3994 When using ELF, the @code{.comm} directive takes an optional third argument.
3995 This is the desired alignment of the symbol, specified as a byte boundary (for
3996 example, an alignment of 16 means that the least significant 4 bits of the
3997 address should be zero). The alignment must be an absolute expression, and it
3998 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
3999 for the common symbol, it will use the alignment when placing the symbol. If
4000 no alignment is specified, @command{@value{AS}} will set the alignment to the
4001 largest power of two less than or equal to the size of the symbol, up to a
4006 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4007 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4010 @node CFI directives
4011 @section @code{.cfi_startproc}
4012 @cindex @code{cfi_startproc} directive
4013 @code{.cfi_startproc} is used at the beginning of each function that
4014 should have an entry in @code{.eh_frame}. It initializes some internal
4015 data structures and emits architecture dependent initial CFI instructions.
4016 Don't forget to close the function by
4017 @code{.cfi_endproc}.
4019 @section @code{.cfi_endproc}
4020 @cindex @code{cfi_endproc} directive
4021 @code{.cfi_endproc} is used at the end of a function where it closes its
4022 unwind entry previously opened by
4023 @code{.cfi_startproc}. and emits it to @code{.eh_frame}.
4025 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4026 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4027 address from @var{register} and add @var{offset} to it}.
4029 @section @code{.cfi_def_cfa_register @var{register}}
4030 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4031 now on @var{register} will be used instead of the old one. Offset
4034 @section @code{.cfi_def_cfa_offset @var{offset}}
4035 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4036 remains the same, but @var{offset} is new. Note that it is the
4037 absolute offset that will be added to a defined register to compute
4040 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4041 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4042 value that is added/substracted from the previous offset.
4044 @section @code{.cfi_offset @var{register}, @var{offset}}
4045 Previous value of @var{register} is saved at offset @var{offset} from
4048 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4049 Previous value of @var{register} is saved at offset @var{offset} from
4050 the current CFA register. This is transformed to @code{.cfi_offset}
4051 using the known displacement of the CFA register from the CFA.
4052 This is often easier to use, because the number will match the
4053 code it's annotating.
4055 @section @code{.cfi_window_save}
4056 SPARC register window has been saved.
4058 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4059 Allows the user to add arbitrary bytes to the unwind info. One
4060 might use this to add OS-specific CFI opcodes, or generic CFI
4061 opcodes that GAS does not yet support.
4064 @section @code{.data @var{subsection}}
4066 @cindex @code{data} directive
4067 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4068 end of the data subsection numbered @var{subsection} (which is an
4069 absolute expression). If @var{subsection} is omitted, it defaults
4074 @section @code{.def @var{name}}
4076 @cindex @code{def} directive
4077 @cindex COFF symbols, debugging
4078 @cindex debugging COFF symbols
4079 Begin defining debugging information for a symbol @var{name}; the
4080 definition extends until the @code{.endef} directive is encountered.
4083 This directive is only observed when @command{@value{AS}} is configured for COFF
4084 format output; when producing @code{b.out}, @samp{.def} is recognized,
4091 @section @code{.desc @var{symbol}, @var{abs-expression}}
4093 @cindex @code{desc} directive
4094 @cindex COFF symbol descriptor
4095 @cindex symbol descriptor, COFF
4096 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4097 to the low 16 bits of an absolute expression.
4100 The @samp{.desc} directive is not available when @command{@value{AS}} is
4101 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4102 object format. For the sake of compatibility, @command{@value{AS}} accepts
4103 it, but produces no output, when configured for COFF.
4109 @section @code{.dim}
4111 @cindex @code{dim} directive
4112 @cindex COFF auxiliary symbol information
4113 @cindex auxiliary symbol information, COFF
4114 This directive is generated by compilers to include auxiliary debugging
4115 information in the symbol table. It is only permitted inside
4116 @code{.def}/@code{.endef} pairs.
4119 @samp{.dim} is only meaningful when generating COFF format output; when
4120 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4126 @section @code{.double @var{flonums}}
4128 @cindex @code{double} directive
4129 @cindex floating point numbers (double)
4130 @code{.double} expects zero or more flonums, separated by commas. It
4131 assembles floating point numbers.
4133 The exact kind of floating point numbers emitted depends on how
4134 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4138 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4139 in @sc{ieee} format.
4144 @section @code{.eject}
4146 @cindex @code{eject} directive
4147 @cindex new page, in listings
4148 @cindex page, in listings
4149 @cindex listing control: new page
4150 Force a page break at this point, when generating assembly listings.
4153 @section @code{.else}
4155 @cindex @code{else} directive
4156 @code{.else} is part of the @command{@value{AS}} support for conditional
4157 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
4158 of code to be assembled if the condition for the preceding @code{.if}
4162 @section @code{.elseif}
4164 @cindex @code{elseif} directive
4165 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4166 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
4167 @code{.if} block that would otherwise fill the entire @code{.else} section.
4170 @section @code{.end}
4172 @cindex @code{end} directive
4173 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4174 process anything in the file past the @code{.end} directive.
4178 @section @code{.endef}
4180 @cindex @code{endef} directive
4181 This directive flags the end of a symbol definition begun with
4185 @samp{.endef} is only meaningful when generating COFF format output; if
4186 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4187 directive but ignores it.
4192 @section @code{.endfunc}
4193 @cindex @code{endfunc} directive
4194 @code{.endfunc} marks the end of a function specified with @code{.func}.
4197 @section @code{.endif}
4199 @cindex @code{endif} directive
4200 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4201 it marks the end of a block of code that is only assembled
4202 conditionally. @xref{If,,@code{.if}}.
4205 @section @code{.equ @var{symbol}, @var{expression}}
4207 @cindex @code{equ} directive
4208 @cindex assigning values to symbols
4209 @cindex symbols, assigning values to
4210 This directive sets the value of @var{symbol} to @var{expression}.
4211 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
4214 The syntax for @code{equ} on the HPPA is
4215 @samp{@var{symbol} .equ @var{expression}}.
4219 @section @code{.equiv @var{symbol}, @var{expression}}
4220 @cindex @code{equiv} directive
4221 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4222 the assembler will signal an error if @var{symbol} is already defined. Note a
4223 symbol which has been referenced but not actually defined is considered to be
4226 Except for the contents of the error message, this is roughly equivalent to
4235 @section @code{.err}
4236 @cindex @code{err} directive
4237 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4238 message and, unless the @option{-Z} option was used, it will not generate an
4239 object file. This can be used to signal error an conditionally compiled code.
4242 @section @code{.exitm}
4243 Exit early from the current macro definition. @xref{Macro}.
4246 @section @code{.extern}
4248 @cindex @code{extern} directive
4249 @code{.extern} is accepted in the source program---for compatibility
4250 with other assemblers---but it is ignored. @command{@value{AS}} treats
4251 all undefined symbols as external.
4254 @section @code{.fail @var{expression}}
4256 @cindex @code{fail} directive
4257 Generates an error or a warning. If the value of the @var{expression} is 500
4258 or more, @command{@value{AS}} will print a warning message. If the value is less
4259 than 500, @command{@value{AS}} will print an error message. The message will
4260 include the value of @var{expression}. This can occasionally be useful inside
4261 complex nested macros or conditional assembly.
4263 @ifclear no-file-dir
4265 @section @code{.file @var{string}}
4267 @cindex @code{file} directive
4268 @cindex logical file name
4269 @cindex file name, logical
4270 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4271 file. @var{string} is the new file name. In general, the filename is
4272 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4273 to specify an empty file name, you must give the quotes--@code{""}. This
4274 statement may go away in future: it is only recognized to be compatible with
4275 old @command{@value{AS}} programs.
4277 In some configurations of @command{@value{AS}}, @code{.file} has already been
4278 removed to avoid conflicts with other assemblers. @xref{Machine Dependencies}.
4283 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4285 @cindex @code{fill} directive
4286 @cindex writing patterns in memory
4287 @cindex patterns, writing in memory
4288 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4289 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4290 may be zero or more. @var{Size} may be zero or more, but if it is
4291 more than 8, then it is deemed to have the value 8, compatible with
4292 other people's assemblers. The contents of each @var{repeat} bytes
4293 is taken from an 8-byte number. The highest order 4 bytes are
4294 zero. The lowest order 4 bytes are @var{value} rendered in the
4295 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4296 Each @var{size} bytes in a repetition is taken from the lowest order
4297 @var{size} bytes of this number. Again, this bizarre behavior is
4298 compatible with other people's assemblers.
4300 @var{size} and @var{value} are optional.
4301 If the second comma and @var{value} are absent, @var{value} is
4302 assumed zero. If the first comma and following tokens are absent,
4303 @var{size} is assumed to be 1.
4306 @section @code{.float @var{flonums}}
4308 @cindex floating point numbers (single)
4309 @cindex @code{float} directive
4310 This directive assembles zero or more flonums, separated by commas. It
4311 has the same effect as @code{.single}.
4313 The exact kind of floating point numbers emitted depends on how
4314 @command{@value{AS}} is configured.
4315 @xref{Machine Dependencies}.
4319 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4320 in @sc{ieee} format.
4325 @section @code{.func @var{name}[,@var{label}]}
4326 @cindex @code{func} directive
4327 @code{.func} emits debugging information to denote function @var{name}, and
4328 is ignored unless the file is assembled with debugging enabled.
4329 Only @samp{--gstabs} is currently supported.
4330 @var{label} is the entry point of the function and if omitted @var{name}
4331 prepended with the @samp{leading char} is used.
4332 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4333 All functions are currently defined to have @code{void} return type.
4334 The function must be terminated with @code{.endfunc}.
4337 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4339 @cindex @code{global} directive
4340 @cindex symbol, making visible to linker
4341 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4342 @var{symbol} in your partial program, its value is made available to
4343 other partial programs that are linked with it. Otherwise,
4344 @var{symbol} takes its attributes from a symbol of the same name
4345 from another file linked into the same program.
4347 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4348 compatibility with other assemblers.
4351 On the HPPA, @code{.global} is not always enough to make it accessible to other
4352 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4353 @xref{HPPA Directives,, HPPA Assembler Directives}.
4358 @section @code{.hidden @var{names}}
4360 @cindex @code{hidden} directive
4362 This one of the ELF visibility directives. The other two are
4363 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4364 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4366 This directive overrides the named symbols default visibility (which is set by
4367 their binding: local, global or weak). The directive sets the visibility to
4368 @code{hidden} which means that the symbols are not visible to other components.
4369 Such symbols are always considered to be @code{protected} as well.
4373 @section @code{.hword @var{expressions}}
4375 @cindex @code{hword} directive
4376 @cindex integers, 16-bit
4377 @cindex numbers, 16-bit
4378 @cindex sixteen bit integers
4379 This expects zero or more @var{expressions}, and emits
4380 a 16 bit number for each.
4383 This directive is a synonym for @samp{.short}; depending on the target
4384 architecture, it may also be a synonym for @samp{.word}.
4388 This directive is a synonym for @samp{.short}.
4391 This directive is a synonym for both @samp{.short} and @samp{.word}.
4396 @section @code{.ident}
4398 @cindex @code{ident} directive
4399 This directive is used by some assemblers to place tags in object files.
4400 @command{@value{AS}} simply accepts the directive for source-file
4401 compatibility with such assemblers, but does not actually emit anything
4405 @section @code{.if @var{absolute expression}}
4407 @cindex conditional assembly
4408 @cindex @code{if} directive
4409 @code{.if} marks the beginning of a section of code which is only
4410 considered part of the source program being assembled if the argument
4411 (which must be an @var{absolute expression}) is non-zero. The end of
4412 the conditional section of code must be marked by @code{.endif}
4413 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4414 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4415 If you have several conditions to check, @code{.elseif} may be used to avoid
4416 nesting blocks if/else within each subsequent @code{.else} block.
4418 The following variants of @code{.if} are also supported:
4420 @cindex @code{ifdef} directive
4421 @item .ifdef @var{symbol}
4422 Assembles the following section of code if the specified @var{symbol}
4423 has been defined. Note a symbol which has been referenced but not yet defined
4424 is considered to be undefined.
4426 @cindex @code{ifc} directive
4427 @item .ifc @var{string1},@var{string2}
4428 Assembles the following section of code if the two strings are the same. The
4429 strings may be optionally quoted with single quotes. If they are not quoted,
4430 the first string stops at the first comma, and the second string stops at the
4431 end of the line. Strings which contain whitespace should be quoted. The
4432 string comparison is case sensitive.
4434 @cindex @code{ifeq} directive
4435 @item .ifeq @var{absolute expression}
4436 Assembles the following section of code if the argument is zero.
4438 @cindex @code{ifeqs} directive
4439 @item .ifeqs @var{string1},@var{string2}
4440 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4442 @cindex @code{ifge} directive
4443 @item .ifge @var{absolute expression}
4444 Assembles the following section of code if the argument is greater than or
4447 @cindex @code{ifgt} directive
4448 @item .ifgt @var{absolute expression}
4449 Assembles the following section of code if the argument is greater than zero.
4451 @cindex @code{ifle} directive
4452 @item .ifle @var{absolute expression}
4453 Assembles the following section of code if the argument is less than or equal
4456 @cindex @code{iflt} directive
4457 @item .iflt @var{absolute expression}
4458 Assembles the following section of code if the argument is less than zero.
4460 @cindex @code{ifnc} directive
4461 @item .ifnc @var{string1},@var{string2}.
4462 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4463 following section of code if the two strings are not the same.
4465 @cindex @code{ifndef} directive
4466 @cindex @code{ifnotdef} directive
4467 @item .ifndef @var{symbol}
4468 @itemx .ifnotdef @var{symbol}
4469 Assembles the following section of code if the specified @var{symbol}
4470 has not been defined. Both spelling variants are equivalent. Note a symbol
4471 which has been referenced but not yet defined is considered to be undefined.
4473 @cindex @code{ifne} directive
4474 @item .ifne @var{absolute expression}
4475 Assembles the following section of code if the argument is not equal to zero
4476 (in other words, this is equivalent to @code{.if}).
4478 @cindex @code{ifnes} directive
4479 @item .ifnes @var{string1},@var{string2}
4480 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4481 following section of code if the two strings are not the same.
4485 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4487 @cindex @code{incbin} directive
4488 @cindex binary files, including
4489 The @code{incbin} directive includes @var{file} verbatim at the current
4490 location. You can control the search paths used with the @samp{-I} command-line
4491 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4494 The @var{skip} argument skips a number of bytes from the start of the
4495 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4496 read. Note that the data is not aligned in any way, so it is the user's
4497 responsibility to make sure that proper alignment is provided both before and
4498 after the @code{incbin} directive.
4501 @section @code{.include "@var{file}"}
4503 @cindex @code{include} directive
4504 @cindex supporting files, including
4505 @cindex files, including
4506 This directive provides a way to include supporting files at specified
4507 points in your source program. The code from @var{file} is assembled as
4508 if it followed the point of the @code{.include}; when the end of the
4509 included file is reached, assembly of the original file continues. You
4510 can control the search paths used with the @samp{-I} command-line option
4511 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4515 @section @code{.int @var{expressions}}
4517 @cindex @code{int} directive
4518 @cindex integers, 32-bit
4519 Expect zero or more @var{expressions}, of any section, separated by commas.
4520 For each expression, emit a number that, at run time, is the value of that
4521 expression. The byte order and bit size of the number depends on what kind
4522 of target the assembly is for.
4526 On the H8/500 and most forms of the H8/300, @code{.int} emits 16-bit
4527 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4534 @section @code{.internal @var{names}}
4536 @cindex @code{internal} directive
4538 This one of the ELF visibility directives. The other two are
4539 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4540 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4542 This directive overrides the named symbols default visibility (which is set by
4543 their binding: local, global or weak). The directive sets the visibility to
4544 @code{internal} which means that the symbols are considered to be @code{hidden}
4545 (i.e., not visible to other components), and that some extra, processor specific
4546 processing must also be performed upon the symbols as well.
4550 @section @code{.irp @var{symbol},@var{values}}@dots{}
4552 @cindex @code{irp} directive
4553 Evaluate a sequence of statements assigning different values to @var{symbol}.
4554 The sequence of statements starts at the @code{.irp} directive, and is
4555 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4556 set to @var{value}, and the sequence of statements is assembled. If no
4557 @var{value} is listed, the sequence of statements is assembled once, with
4558 @var{symbol} set to the null string. To refer to @var{symbol} within the
4559 sequence of statements, use @var{\symbol}.
4561 For example, assembling
4569 is equivalent to assembling
4578 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4580 @cindex @code{irpc} directive
4581 Evaluate a sequence of statements assigning different values to @var{symbol}.
4582 The sequence of statements starts at the @code{.irpc} directive, and is
4583 terminated by an @code{.endr} directive. For each character in @var{value},
4584 @var{symbol} is set to the character, and the sequence of statements is
4585 assembled. If no @var{value} is listed, the sequence of statements is
4586 assembled once, with @var{symbol} set to the null string. To refer to
4587 @var{symbol} within the sequence of statements, use @var{\symbol}.
4589 For example, assembling
4597 is equivalent to assembling
4606 @section @code{.lcomm @var{symbol} , @var{length}}
4608 @cindex @code{lcomm} directive
4609 @cindex local common symbols
4610 @cindex symbols, local common
4611 Reserve @var{length} (an absolute expression) bytes for a local common
4612 denoted by @var{symbol}. The section and value of @var{symbol} are
4613 those of the new local common. The addresses are allocated in the bss
4614 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4615 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4616 not visible to @code{@value{LD}}.
4619 Some targets permit a third argument to be used with @code{.lcomm}. This
4620 argument specifies the desired alignment of the symbol in the bss section.
4624 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4625 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4629 @section @code{.lflags}
4631 @cindex @code{lflags} directive (ignored)
4632 @command{@value{AS}} accepts this directive, for compatibility with other
4633 assemblers, but ignores it.
4635 @ifclear no-line-dir
4637 @section @code{.line @var{line-number}}
4639 @cindex @code{line} directive
4643 @section @code{.ln @var{line-number}}
4645 @cindex @code{ln} directive
4647 @cindex logical line number
4649 Change the logical line number. @var{line-number} must be an absolute
4650 expression. The next line has that logical line number. Therefore any other
4651 statements on the current line (after a statement separator character) are
4652 reported as on logical line number @var{line-number} @minus{} 1. One day
4653 @command{@value{AS}} will no longer support this directive: it is recognized only
4654 for compatibility with existing assembler programs.
4658 @emph{Warning:} In the AMD29K configuration of @value{AS}, this command is
4659 not available; use the synonym @code{.ln} in that context.
4664 @ifclear no-line-dir
4665 Even though this is a directive associated with the @code{a.out} or
4666 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4667 when producing COFF output, and treats @samp{.line} as though it
4668 were the COFF @samp{.ln} @emph{if} it is found outside a
4669 @code{.def}/@code{.endef} pair.
4671 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4672 used by compilers to generate auxiliary symbol information for
4677 @section @code{.linkonce [@var{type}]}
4679 @cindex @code{linkonce} directive
4680 @cindex common sections
4681 Mark the current section so that the linker only includes a single copy of it.
4682 This may be used to include the same section in several different object files,
4683 but ensure that the linker will only include it once in the final output file.
4684 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4685 Duplicate sections are detected based on the section name, so it should be
4688 This directive is only supported by a few object file formats; as of this
4689 writing, the only object file format which supports it is the Portable
4690 Executable format used on Windows NT.
4692 The @var{type} argument is optional. If specified, it must be one of the
4693 following strings. For example:
4697 Not all types may be supported on all object file formats.
4701 Silently discard duplicate sections. This is the default.
4704 Warn if there are duplicate sections, but still keep only one copy.
4707 Warn if any of the duplicates have different sizes.
4710 Warn if any of the duplicates do not have exactly the same contents.
4714 @section @code{.ln @var{line-number}}
4716 @cindex @code{ln} directive
4717 @ifclear no-line-dir
4718 @samp{.ln} is a synonym for @samp{.line}.
4721 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4722 must be an absolute expression. The next line has that logical
4723 line number, so any other statements on the current line (after a
4724 statement separator character @code{;}) are reported as on logical
4725 line number @var{line-number} @minus{} 1.
4728 This directive is accepted, but ignored, when @command{@value{AS}} is
4729 configured for @code{b.out}; its effect is only associated with COFF
4735 @section @code{.mri @var{val}}
4737 @cindex @code{mri} directive
4738 @cindex MRI mode, temporarily
4739 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4740 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4741 affects code assembled until the next @code{.mri} directive, or until the end
4742 of the file. @xref{M, MRI mode, MRI mode}.
4745 @section @code{.list}
4747 @cindex @code{list} directive
4748 @cindex listing control, turning on
4749 Control (in conjunction with the @code{.nolist} directive) whether or
4750 not assembly listings are generated. These two directives maintain an
4751 internal counter (which is zero initially). @code{.list} increments the
4752 counter, and @code{.nolist} decrements it. Assembly listings are
4753 generated whenever the counter is greater than zero.
4755 By default, listings are disabled. When you enable them (with the
4756 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4757 the initial value of the listing counter is one.
4760 @section @code{.long @var{expressions}}
4762 @cindex @code{long} directive
4763 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4766 @c no one seems to know what this is for or whether this description is
4767 @c what it really ought to do
4769 @section @code{.lsym @var{symbol}, @var{expression}}
4771 @cindex @code{lsym} directive
4772 @cindex symbol, not referenced in assembly
4773 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4774 the hash table, ensuring it cannot be referenced by name during the
4775 rest of the assembly. This sets the attributes of the symbol to be
4776 the same as the expression value:
4778 @var{other} = @var{descriptor} = 0
4779 @var{type} = @r{(section of @var{expression})}
4780 @var{value} = @var{expression}
4783 The new symbol is not flagged as external.
4787 @section @code{.macro}
4790 The commands @code{.macro} and @code{.endm} allow you to define macros that
4791 generate assembly output. For example, this definition specifies a macro
4792 @code{sum} that puts a sequence of numbers into memory:
4795 .macro sum from=0, to=5
4804 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4816 @item .macro @var{macname}
4817 @itemx .macro @var{macname} @var{macargs} @dots{}
4818 @cindex @code{macro} directive
4819 Begin the definition of a macro called @var{macname}. If your macro
4820 definition requires arguments, specify their names after the macro name,
4821 separated by commas or spaces. You can supply a default value for any
4822 macro argument by following the name with @samp{=@var{deflt}}. For
4823 example, these are all valid @code{.macro} statements:
4827 Begin the definition of a macro called @code{comm}, which takes no
4830 @item .macro plus1 p, p1
4831 @itemx .macro plus1 p p1
4832 Either statement begins the definition of a macro called @code{plus1},
4833 which takes two arguments; within the macro definition, write
4834 @samp{\p} or @samp{\p1} to evaluate the arguments.
4836 @item .macro reserve_str p1=0 p2
4837 Begin the definition of a macro called @code{reserve_str}, with two
4838 arguments. The first argument has a default value, but not the second.
4839 After the definition is complete, you can call the macro either as
4840 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
4841 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
4842 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
4843 @samp{0}, and @samp{\p2} evaluating to @var{b}).
4846 When you call a macro, you can specify the argument values either by
4847 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
4848 @samp{sum to=17, from=9}.
4851 @cindex @code{endm} directive
4852 Mark the end of a macro definition.
4855 @cindex @code{exitm} directive
4856 Exit early from the current macro definition.
4858 @cindex number of macros executed
4859 @cindex macros, count executed
4861 @command{@value{AS}} maintains a counter of how many macros it has
4862 executed in this pseudo-variable; you can copy that number to your
4863 output with @samp{\@@}, but @emph{only within a macro definition}.
4866 @item LOCAL @var{name} [ , @dots{} ]
4867 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
4868 macro syntax'' with @samp{-a} or @samp{--alternate}.} @xref{Alternate,,
4869 Alternate macro syntax}.
4871 Generate a string replacement for each of the @var{name} arguments, and
4872 replace any instances of @var{name} in each macro expansion. The
4873 replacement string is unique in the assembly, and different for each
4874 separate macro expansion. @code{LOCAL} allows you to write macros that
4875 define symbols, without fear of conflict between separate macro expansions.
4880 @section @code{.nolist}
4882 @cindex @code{nolist} directive
4883 @cindex listing control, turning off
4884 Control (in conjunction with the @code{.list} directive) whether or
4885 not assembly listings are generated. These two directives maintain an
4886 internal counter (which is zero initially). @code{.list} increments the
4887 counter, and @code{.nolist} decrements it. Assembly listings are
4888 generated whenever the counter is greater than zero.
4891 @section @code{.octa @var{bignums}}
4893 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
4894 @cindex @code{octa} directive
4895 @cindex integer, 16-byte
4896 @cindex sixteen byte integer
4897 This directive expects zero or more bignums, separated by commas. For each
4898 bignum, it emits a 16-byte integer.
4900 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
4901 hence @emph{octa}-word for 16 bytes.
4904 @section @code{.org @var{new-lc} , @var{fill}}
4906 @cindex @code{org} directive
4907 @cindex location counter, advancing
4908 @cindex advancing location counter
4909 @cindex current address, advancing
4910 Advance the location counter of the current section to
4911 @var{new-lc}. @var{new-lc} is either an absolute expression or an
4912 expression with the same section as the current subsection. That is,
4913 you can't use @code{.org} to cross sections: if @var{new-lc} has the
4914 wrong section, the @code{.org} directive is ignored. To be compatible
4915 with former assemblers, if the section of @var{new-lc} is absolute,
4916 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
4917 is the same as the current subsection.
4919 @code{.org} may only increase the location counter, or leave it
4920 unchanged; you cannot use @code{.org} to move the location counter
4923 @c double negative used below "not undefined" because this is a specific
4924 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
4925 @c section. doc@cygnus.com 18feb91
4926 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
4927 may not be undefined. If you really detest this restriction we eagerly await
4928 a chance to share your improved assembler.
4930 Beware that the origin is relative to the start of the section, not
4931 to the start of the subsection. This is compatible with other
4932 people's assemblers.
4934 When the location counter (of the current subsection) is advanced, the
4935 intervening bytes are filled with @var{fill} which should be an
4936 absolute expression. If the comma and @var{fill} are omitted,
4937 @var{fill} defaults to zero.
4940 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4942 @cindex padding the location counter given a power of two
4943 @cindex @code{p2align} directive
4944 Pad the location counter (in the current subsection) to a particular
4945 storage boundary. The first expression (which must be absolute) is the
4946 number of low-order zero bits the location counter must have after
4947 advancement. For example @samp{.p2align 3} advances the location
4948 counter until it a multiple of 8. If the location counter is already a
4949 multiple of 8, no change is needed.
4951 The second expression (also absolute) gives the fill value to be stored in the
4952 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4953 padding bytes are normally zero. However, on some systems, if the section is
4954 marked as containing code and the fill value is omitted, the space is filled
4955 with no-op instructions.
4957 The third expression is also absolute, and is also optional. If it is present,
4958 it is the maximum number of bytes that should be skipped by this alignment
4959 directive. If doing the alignment would require skipping more bytes than the
4960 specified maximum, then the alignment is not done at all. You can omit the
4961 fill value (the second argument) entirely by simply using two commas after the
4962 required alignment; this can be useful if you want the alignment to be filled
4963 with no-op instructions when appropriate.
4965 @cindex @code{p2alignw} directive
4966 @cindex @code{p2alignl} directive
4967 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
4968 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
4969 pattern as a two byte word value. The @code{.p2alignl} directives treats the
4970 fill pattern as a four byte longword value. For example, @code{.p2alignw
4971 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4972 filled in with the value 0x368d (the exact placement of the bytes depends upon
4973 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4978 @section @code{.previous}
4980 @cindex @code{previous} directive
4981 @cindex Section Stack
4982 This is one of the ELF section stack manipulation directives. The others are
4983 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
4984 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
4985 (@pxref{PopSection}).
4987 This directive swaps the current section (and subsection) with most recently
4988 referenced section (and subsection) prior to this one. Multiple
4989 @code{.previous} directives in a row will flip between two sections (and their
4992 In terms of the section stack, this directive swaps the current section with
4993 the top section on the section stack.
4998 @section @code{.popsection}
5000 @cindex @code{popsection} directive
5001 @cindex Section Stack
5002 This is one of the ELF section stack manipulation directives. The others are
5003 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5004 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5007 This directive replaces the current section (and subsection) with the top
5008 section (and subsection) on the section stack. This section is popped off the
5013 @section @code{.print @var{string}}
5015 @cindex @code{print} directive
5016 @command{@value{AS}} will print @var{string} on the standard output during
5017 assembly. You must put @var{string} in double quotes.
5021 @section @code{.protected @var{names}}
5023 @cindex @code{protected} directive
5025 This one of the ELF visibility directives. The other two are
5026 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5028 This directive overrides the named symbols default visibility (which is set by
5029 their binding: local, global or weak). The directive sets the visibility to
5030 @code{protected} which means that any references to the symbols from within the
5031 components that defines them must be resolved to the definition in that
5032 component, even if a definition in another component would normally preempt
5037 @section @code{.psize @var{lines} , @var{columns}}
5039 @cindex @code{psize} directive
5040 @cindex listing control: paper size
5041 @cindex paper size, for listings
5042 Use this directive to declare the number of lines---and, optionally, the
5043 number of columns---to use for each page, when generating listings.
5045 If you do not use @code{.psize}, listings use a default line-count
5046 of 60. You may omit the comma and @var{columns} specification; the
5047 default width is 200 columns.
5049 @command{@value{AS}} generates formfeeds whenever the specified number of
5050 lines is exceeded (or whenever you explicitly request one, using
5053 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5054 those explicitly specified with @code{.eject}.
5057 @section @code{.purgem @var{name}}
5059 @cindex @code{purgem} directive
5060 Undefine the macro @var{name}, so that later uses of the string will not be
5061 expanded. @xref{Macro}.
5065 @section @code{.pushsection @var{name} , @var{subsection}}
5067 @cindex @code{pushsection} directive
5068 @cindex Section Stack
5069 This is one of the ELF section stack manipulation directives. The others are
5070 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5071 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5074 This directive is a synonym for @code{.section}. It pushes the current section
5075 (and subsection) onto the top of the section stack, and then replaces the
5076 current section and subsection with @code{name} and @code{subsection}.
5080 @section @code{.quad @var{bignums}}
5082 @cindex @code{quad} directive
5083 @code{.quad} expects zero or more bignums, separated by commas. For
5084 each bignum, it emits
5086 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5087 warning message; and just takes the lowest order 8 bytes of the bignum.
5088 @cindex eight-byte integer
5089 @cindex integer, 8-byte
5091 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5092 hence @emph{quad}-word for 8 bytes.
5095 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5096 warning message; and just takes the lowest order 16 bytes of the bignum.
5097 @cindex sixteen-byte integer
5098 @cindex integer, 16-byte
5102 @section @code{.rept @var{count}}
5104 @cindex @code{rept} directive
5105 Repeat the sequence of lines between the @code{.rept} directive and the next
5106 @code{.endr} directive @var{count} times.
5108 For example, assembling
5116 is equivalent to assembling
5125 @section @code{.sbttl "@var{subheading}"}
5127 @cindex @code{sbttl} directive
5128 @cindex subtitles for listings
5129 @cindex listing control: subtitle
5130 Use @var{subheading} as the title (third line, immediately after the
5131 title line) when generating assembly listings.
5133 This directive affects subsequent pages, as well as the current page if
5134 it appears within ten lines of the top of a page.
5138 @section @code{.scl @var{class}}
5140 @cindex @code{scl} directive
5141 @cindex symbol storage class (COFF)
5142 @cindex COFF symbol storage class
5143 Set the storage-class value for a symbol. This directive may only be
5144 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5145 whether a symbol is static or external, or it may record further
5146 symbolic debugging information.
5149 The @samp{.scl} directive is primarily associated with COFF output; when
5150 configured to generate @code{b.out} output format, @command{@value{AS}}
5151 accepts this directive but ignores it.
5157 @section @code{.section @var{name}}
5159 @cindex named section
5160 Use the @code{.section} directive to assemble the following code into a section
5163 This directive is only supported for targets that actually support arbitrarily
5164 named sections; on @code{a.out} targets, for example, it is not accepted, even
5165 with a standard @code{a.out} section name.
5169 @c only print the extra heading if both COFF and ELF are set
5170 @subheading COFF Version
5173 @cindex @code{section} directive (COFF version)
5174 For COFF targets, the @code{.section} directive is used in one of the following
5178 .section @var{name}[, "@var{flags}"]
5179 .section @var{name}[, @var{subsegment}]
5182 If the optional argument is quoted, it is taken as flags to use for the
5183 section. Each flag is a single character. The following flags are recognized:
5186 bss section (uninitialized data)
5188 section is not loaded
5198 shared section (meaningful for PE targets)
5200 ignored. (For compatibility with the ELF version)
5203 If no flags are specified, the default flags depend upon the section name. If
5204 the section name is not recognized, the default will be for the section to be
5205 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5206 from the section, rather than adding them, so if they are used on their own it
5207 will be as if no flags had been specified at all.
5209 If the optional argument to the @code{.section} directive is not quoted, it is
5210 taken as a subsegment number (@pxref{Sub-Sections}).
5215 @c only print the extra heading if both COFF and ELF are set
5216 @subheading ELF Version
5219 @cindex Section Stack
5220 This is one of the ELF section stack manipulation directives. The others are
5221 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5222 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5223 @code{.previous} (@pxref{Previous}).
5225 @cindex @code{section} directive (ELF version)
5226 For ELF targets, the @code{.section} directive is used like this:
5229 .section @var{name} [, "@var{flags}"[, @@@var{type}[, @@@var{entsize}]]]
5232 The optional @var{flags} argument is a quoted string which may contain any
5233 combination of the following characters:
5236 section is allocatable
5240 section is executable
5242 section is mergeable
5244 section contains zero terminated strings
5247 The optional @var{type} argument may contain one of the following constants:
5250 section contains data
5252 section does not contain data (i.e., section only occupies space)
5255 Note on targets where the @code{@@} character is the start of a comment (eg
5256 ARM) then another character is used instead. For example the ARM port uses the
5259 If @var{flags} contains @code{M} flag, @var{type} argument must be specified
5260 as well as @var{entsize} argument. Sections with @code{M} flag but not
5261 @code{S} flag must contain fixed size constants, each @var{entsize} octets
5262 long. Sections with both @code{M} and @code{S} must contain zero terminated
5263 strings where each character is @var{entsize} bytes long. The linker may remove
5264 duplicates within sections with the same name, same entity size and same flags.
5266 If no flags are specified, the default flags depend upon the section name. If
5267 the section name is not recognized, the default will be for the section to have
5268 none of the above flags: it will not be allocated in memory, nor writable, nor
5269 executable. The section will contain data.
5271 For ELF targets, the assembler supports another type of @code{.section}
5272 directive for compatibility with the Solaris assembler:
5275 .section "@var{name}"[, @var{flags}...]
5278 Note that the section name is quoted. There may be a sequence of comma
5282 section is allocatable
5286 section is executable
5289 This directive replaces the current section and subsection. The replaced
5290 section and subsection are pushed onto the section stack. See the contents of
5291 the gas testsuite directory @code{gas/testsuite/gas/elf} for some examples of
5292 how this directive and the other section stack directives work.
5297 @section @code{.set @var{symbol}, @var{expression}}
5299 @cindex @code{set} directive
5300 @cindex symbol value, setting
5301 Set the value of @var{symbol} to @var{expression}. This
5302 changes @var{symbol}'s value and type to conform to
5303 @var{expression}. If @var{symbol} was flagged as external, it remains
5304 flagged (@pxref{Symbol Attributes}).
5306 You may @code{.set} a symbol many times in the same assembly.
5308 If you @code{.set} a global symbol, the value stored in the object
5309 file is the last value stored into it.
5312 The syntax for @code{set} on the HPPA is
5313 @samp{@var{symbol} .set @var{expression}}.
5317 @section @code{.short @var{expressions}}
5319 @cindex @code{short} directive
5321 @code{.short} is normally the same as @samp{.word}.
5322 @xref{Word,,@code{.word}}.
5324 In some configurations, however, @code{.short} and @code{.word} generate
5325 numbers of different lengths; @pxref{Machine Dependencies}.
5329 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5332 This expects zero or more @var{expressions}, and emits
5333 a 16 bit number for each.
5338 @section @code{.single @var{flonums}}
5340 @cindex @code{single} directive
5341 @cindex floating point numbers (single)
5342 This directive assembles zero or more flonums, separated by commas. It
5343 has the same effect as @code{.float}.
5345 The exact kind of floating point numbers emitted depends on how
5346 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5350 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5351 numbers in @sc{ieee} format.
5357 @section @code{.size}
5359 This directive is used to set the size associated with a symbol.
5363 @c only print the extra heading if both COFF and ELF are set
5364 @subheading COFF Version
5367 @cindex @code{size} directive (COFF version)
5368 For COFF targets, the @code{.size} directive is only permitted inside
5369 @code{.def}/@code{.endef} pairs. It is used like this:
5372 .size @var{expression}
5376 @samp{.size} is only meaningful when generating COFF format output; when
5377 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5384 @c only print the extra heading if both COFF and ELF are set
5385 @subheading ELF Version
5388 @cindex @code{size} directive (ELF version)
5389 For ELF targets, the @code{.size} directive is used like this:
5392 .size @var{name} , @var{expression}
5395 This directive sets the size associated with a symbol @var{name}.
5396 The size in bytes is computed from @var{expression} which can make use of label
5397 arithmetic. This directive is typically used to set the size of function
5403 @section @code{.sleb128 @var{expressions}}
5405 @cindex @code{sleb128} directive
5406 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5407 compact, variable length representation of numbers used by the DWARF
5408 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
5410 @ifclear no-space-dir
5412 @section @code{.skip @var{size} , @var{fill}}
5414 @cindex @code{skip} directive
5415 @cindex filling memory
5416 This directive emits @var{size} bytes, each of value @var{fill}. Both
5417 @var{size} and @var{fill} are absolute expressions. If the comma and
5418 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5422 @section @code{.space @var{size} , @var{fill}}
5424 @cindex @code{space} directive
5425 @cindex filling memory
5426 This directive emits @var{size} bytes, each of value @var{fill}. Both
5427 @var{size} and @var{fill} are absolute expressions. If the comma
5428 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5433 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5434 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5435 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5436 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5445 @section @code{.space}
5446 @cindex @code{space} directive
5448 On the AMD 29K, this directive is ignored; it is accepted for
5449 compatibility with other AMD 29K assemblers.
5452 @emph{Warning:} In most versions of the @sc{gnu} assembler, the directive
5453 @code{.space} has the effect of @code{.block} @xref{Machine Dependencies}.
5459 @section @code{.stabd, .stabn, .stabs}
5461 @cindex symbolic debuggers, information for
5462 @cindex @code{stab@var{x}} directives
5463 There are three directives that begin @samp{.stab}.
5464 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5465 The symbols are not entered in the @command{@value{AS}} hash table: they
5466 cannot be referenced elsewhere in the source file.
5467 Up to five fields are required:
5471 This is the symbol's name. It may contain any character except
5472 @samp{\000}, so is more general than ordinary symbol names. Some
5473 debuggers used to code arbitrarily complex structures into symbol names
5477 An absolute expression. The symbol's type is set to the low 8 bits of
5478 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5479 and debuggers choke on silly bit patterns.
5482 An absolute expression. The symbol's ``other'' attribute is set to the
5483 low 8 bits of this expression.
5486 An absolute expression. The symbol's descriptor is set to the low 16
5487 bits of this expression.
5490 An absolute expression which becomes the symbol's value.
5493 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5494 or @code{.stabs} statement, the symbol has probably already been created;
5495 you get a half-formed symbol in your object file. This is
5496 compatible with earlier assemblers!
5499 @cindex @code{stabd} directive
5500 @item .stabd @var{type} , @var{other} , @var{desc}
5502 The ``name'' of the symbol generated is not even an empty string.
5503 It is a null pointer, for compatibility. Older assemblers used a
5504 null pointer so they didn't waste space in object files with empty
5507 The symbol's value is set to the location counter,
5508 relocatably. When your program is linked, the value of this symbol
5509 is the address of the location counter when the @code{.stabd} was
5512 @cindex @code{stabn} directive
5513 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5514 The name of the symbol is set to the empty string @code{""}.
5516 @cindex @code{stabs} directive
5517 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5518 All five fields are specified.
5524 @section @code{.string} "@var{str}"
5526 @cindex string, copying to object file
5527 @cindex @code{string} directive
5529 Copy the characters in @var{str} to the object file. You may specify more than
5530 one string to copy, separated by commas. Unless otherwise specified for a
5531 particular machine, the assembler marks the end of each string with a 0 byte.
5532 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5535 @section @code{.struct @var{expression}}
5537 @cindex @code{struct} directive
5538 Switch to the absolute section, and set the section offset to @var{expression},
5539 which must be an absolute expression. You might use this as follows:
5548 This would define the symbol @code{field1} to have the value 0, the symbol
5549 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5550 value 8. Assembly would be left in the absolute section, and you would need to
5551 use a @code{.section} directive of some sort to change to some other section
5552 before further assembly.
5556 @section @code{.subsection @var{name}}
5558 @cindex @code{subsection} directive
5559 @cindex Section Stack
5560 This is one of the ELF section stack manipulation directives. The others are
5561 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5562 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5565 This directive replaces the current subsection with @code{name}. The current
5566 section is not changed. The replaced subsection is put onto the section stack
5567 in place of the then current top of stack subsection.
5572 @section @code{.symver}
5573 @cindex @code{symver} directive
5574 @cindex symbol versioning
5575 @cindex versions of symbols
5576 Use the @code{.symver} directive to bind symbols to specific version nodes
5577 within a source file. This is only supported on ELF platforms, and is
5578 typically used when assembling files to be linked into a shared library.
5579 There are cases where it may make sense to use this in objects to be bound
5580 into an application itself so as to override a versioned symbol from a
5583 For ELF targets, the @code{.symver} directive can be used like this:
5585 .symver @var{name}, @var{name2@@nodename}
5587 If the symbol @var{name} is defined within the file
5588 being assembled, the @code{.symver} directive effectively creates a symbol
5589 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5590 just don't try and create a regular alias is that the @var{@@} character isn't
5591 permitted in symbol names. The @var{name2} part of the name is the actual name
5592 of the symbol by which it will be externally referenced. The name @var{name}
5593 itself is merely a name of convenience that is used so that it is possible to
5594 have definitions for multiple versions of a function within a single source
5595 file, and so that the compiler can unambiguously know which version of a
5596 function is being mentioned. The @var{nodename} portion of the alias should be
5597 the name of a node specified in the version script supplied to the linker when
5598 building a shared library. If you are attempting to override a versioned
5599 symbol from a shared library, then @var{nodename} should correspond to the
5600 nodename of the symbol you are trying to override.
5602 If the symbol @var{name} is not defined within the file being assembled, all
5603 references to @var{name} will be changed to @var{name2@@nodename}. If no
5604 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5607 Another usage of the @code{.symver} directive is:
5609 .symver @var{name}, @var{name2@@@@nodename}
5611 In this case, the symbol @var{name} must exist and be defined within
5612 the file being assembled. It is similar to @var{name2@@nodename}. The
5613 difference is @var{name2@@@@nodename} will also be used to resolve
5614 references to @var{name2} by the linker.
5616 The third usage of the @code{.symver} directive is:
5618 .symver @var{name}, @var{name2@@@@@@nodename}
5620 When @var{name} is not defined within the
5621 file being assembled, it is treated as @var{name2@@nodename}. When
5622 @var{name} is defined within the file being assembled, the symbol
5623 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5628 @section @code{.tag @var{structname}}
5630 @cindex COFF structure debugging
5631 @cindex structure debugging, COFF
5632 @cindex @code{tag} directive
5633 This directive is generated by compilers to include auxiliary debugging
5634 information in the symbol table. It is only permitted inside
5635 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5636 definitions in the symbol table with instances of those structures.
5639 @samp{.tag} is only used when generating COFF format output; when
5640 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5646 @section @code{.text @var{subsection}}
5648 @cindex @code{text} directive
5649 Tells @command{@value{AS}} to assemble the following statements onto the end of
5650 the text subsection numbered @var{subsection}, which is an absolute
5651 expression. If @var{subsection} is omitted, subsection number zero
5655 @section @code{.title "@var{heading}"}
5657 @cindex @code{title} directive
5658 @cindex listing control: title line
5659 Use @var{heading} as the title (second line, immediately after the
5660 source file name and pagenumber) when generating assembly listings.
5662 This directive affects subsequent pages, as well as the current page if
5663 it appears within ten lines of the top of a page.
5667 @section @code{.type}
5669 This directive is used to set the type of a symbol.
5673 @c only print the extra heading if both COFF and ELF are set
5674 @subheading COFF Version
5677 @cindex COFF symbol type
5678 @cindex symbol type, COFF
5679 @cindex @code{type} directive (COFF version)
5680 For COFF targets, this directive is permitted only within
5681 @code{.def}/@code{.endef} pairs. It is used like this:
5687 This records the integer @var{int} as the type attribute of a symbol table
5691 @samp{.type} is associated only with COFF format output; when
5692 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
5693 directive but ignores it.
5699 @c only print the extra heading if both COFF and ELF are set
5700 @subheading ELF Version
5703 @cindex ELF symbol type
5704 @cindex symbol type, ELF
5705 @cindex @code{type} directive (ELF version)
5706 For ELF targets, the @code{.type} directive is used like this:
5709 .type @var{name} , @var{type description}
5712 This sets the type of symbol @var{name} to be either a
5713 function symbol or an object symbol. There are five different syntaxes
5714 supported for the @var{type description} field, in order to provide
5715 compatibility with various other assemblers. The syntaxes supported are:
5718 .type <name>,#function
5719 .type <name>,#object
5721 .type <name>,@@function
5722 .type <name>,@@object
5724 .type <name>,%function
5725 .type <name>,%object
5727 .type <name>,"function"
5728 .type <name>,"object"
5730 .type <name> STT_FUNCTION
5731 .type <name> STT_OBJECT
5737 @section @code{.uleb128 @var{expressions}}
5739 @cindex @code{uleb128} directive
5740 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
5741 compact, variable length representation of numbers used by the DWARF
5742 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
5746 @section @code{.val @var{addr}}
5748 @cindex @code{val} directive
5749 @cindex COFF value attribute
5750 @cindex value attribute, COFF
5751 This directive, permitted only within @code{.def}/@code{.endef} pairs,
5752 records the address @var{addr} as the value attribute of a symbol table
5756 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
5757 configured for @code{b.out}, it accepts this directive but ignores it.
5763 @section @code{.version "@var{string}"}
5765 @cindex @code{version} directive
5766 This directive creates a @code{.note} section and places into it an ELF
5767 formatted note of type NT_VERSION. The note's name is set to @code{string}.
5772 @section @code{.vtable_entry @var{table}, @var{offset}}
5774 @cindex @code{vtable_entry}
5775 This directive finds or creates a symbol @code{table} and creates a
5776 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
5779 @section @code{.vtable_inherit @var{child}, @var{parent}}
5781 @cindex @code{vtable_inherit}
5782 This directive finds the symbol @code{child} and finds or creates the symbol
5783 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
5784 parent whose addend is the value of the child symbol. As a special case the
5785 parent name of @code{0} is treated as refering the @code{*ABS*} section.
5790 @section @code{.weak @var{names}}
5792 @cindex @code{weak} directive
5793 This directive sets the weak attribute on the comma separated list of symbol
5794 @code{names}. If the symbols do not already exist, they will be created.
5798 @section @code{.word @var{expressions}}
5800 @cindex @code{word} directive
5801 This directive expects zero or more @var{expressions}, of any section,
5802 separated by commas.
5805 For each expression, @command{@value{AS}} emits a 32-bit number.
5808 For each expression, @command{@value{AS}} emits a 16-bit number.
5813 The size of the number emitted, and its byte order,
5814 depend on what target computer the assembly is for.
5817 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
5818 @c happen---32-bit addressability, period; no long/short jumps.
5819 @ifset DIFF-TBL-KLUGE
5820 @cindex difference tables altered
5821 @cindex altered difference tables
5823 @emph{Warning: Special Treatment to support Compilers}
5827 Machines with a 32-bit address space, but that do less than 32-bit
5828 addressing, require the following special treatment. If the machine of
5829 interest to you does 32-bit addressing (or doesn't require it;
5830 @pxref{Machine Dependencies}), you can ignore this issue.
5833 In order to assemble compiler output into something that works,
5834 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
5835 Directives of the form @samp{.word sym1-sym2} are often emitted by
5836 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
5837 directive of the form @samp{.word sym1-sym2}, and the difference between
5838 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
5839 creates a @dfn{secondary jump table}, immediately before the next label.
5840 This secondary jump table is preceded by a short-jump to the
5841 first byte after the secondary table. This short-jump prevents the flow
5842 of control from accidentally falling into the new table. Inside the
5843 table is a long-jump to @code{sym2}. The original @samp{.word}
5844 contains @code{sym1} minus the address of the long-jump to
5847 If there were several occurrences of @samp{.word sym1-sym2} before the
5848 secondary jump table, all of them are adjusted. If there was a
5849 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
5850 long-jump to @code{sym4} is included in the secondary jump table,
5851 and the @code{.word} directives are adjusted to contain @code{sym3}
5852 minus the address of the long-jump to @code{sym4}; and so on, for as many
5853 entries in the original jump table as necessary.
5856 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
5857 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
5858 assembly language programmers.
5861 @c end DIFF-TBL-KLUGE
5864 @section Deprecated Directives
5866 @cindex deprecated directives
5867 @cindex obsolescent directives
5868 One day these directives won't work.
5869 They are included for compatibility with older assemblers.
5876 @node Machine Dependencies
5877 @chapter Machine Dependent Features
5879 @cindex machine dependencies
5880 The machine instruction sets are (almost by definition) different on
5881 each machine where @command{@value{AS}} runs. Floating point representations
5882 vary as well, and @command{@value{AS}} often supports a few additional
5883 directives or command-line options for compatibility with other
5884 assemblers on a particular platform. Finally, some versions of
5885 @command{@value{AS}} support special pseudo-instructions for branch
5888 This chapter discusses most of these differences, though it does not
5889 include details on any machine's instruction set. For details on that
5890 subject, see the hardware manufacturer's manual.
5894 * AMD29K-Dependent:: AMD 29K Dependent Features
5897 * Alpha-Dependent:: Alpha Dependent Features
5900 * ARC-Dependent:: ARC Dependent Features
5903 * ARM-Dependent:: ARM Dependent Features
5906 * CRIS-Dependent:: CRIS Dependent Features
5909 * D10V-Dependent:: D10V Dependent Features
5912 * D30V-Dependent:: D30V Dependent Features
5915 * H8/300-Dependent:: Renesas H8/300 Dependent Features
5918 * H8/500-Dependent:: Renesas H8/500 Dependent Features
5921 * HPPA-Dependent:: HPPA Dependent Features
5924 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
5927 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
5930 * i860-Dependent:: Intel 80860 Dependent Features
5933 * i960-Dependent:: Intel 80960 Dependent Features
5936 * IP2K-Dependent:: IP2K Dependent Features
5939 * M32R-Dependent:: M32R Dependent Features
5942 * M68K-Dependent:: M680x0 Dependent Features
5945 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
5948 * M88K-Dependent:: M880x0 Dependent Features
5951 * MIPS-Dependent:: MIPS Dependent Features
5954 * MMIX-Dependent:: MMIX Dependent Features
5957 * MSP430-Dependent:: MSP430 Dependent Features
5960 * SH-Dependent:: Renesas / SuperH SH Dependent Features
5961 * SH64-Dependent:: SuperH SH64 Dependent Features
5964 * PDP-11-Dependent:: PDP-11 Dependent Features
5967 * PJ-Dependent:: picoJava Dependent Features
5970 * PPC-Dependent:: PowerPC Dependent Features
5973 * Sparc-Dependent:: SPARC Dependent Features
5976 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
5979 * V850-Dependent:: V850 Dependent Features
5982 * Xtensa-Dependent:: Xtensa Dependent Features
5985 * Z8000-Dependent:: Z8000 Dependent Features
5988 * Vax-Dependent:: VAX Dependent Features
5995 @c The following major nodes are *sections* in the GENERIC version, *chapters*
5996 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
5997 @c peculiarity: to preserve cross-references, there must be a node called
5998 @c "Machine Dependencies". Hence the conditional nodenames in each
5999 @c major node below. Node defaulting in makeinfo requires adjacency of
6000 @c node and sectioning commands; hence the repetition of @chapter BLAH
6001 @c in both conditional blocks.
6004 @include c-a29k.texi
6008 @include c-alpha.texi
6020 @include c-cris.texi
6025 @node Machine Dependencies
6026 @chapter Machine Dependent Features
6028 The machine instruction sets are different on each Renesas chip family,
6029 and there are also some syntax differences among the families. This
6030 chapter describes the specific @command{@value{AS}} features for each
6034 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6035 * H8/500-Dependent:: Renesas H8/500 Dependent Features
6036 * SH-Dependent:: Renesas SH Dependent Features
6043 @include c-d10v.texi
6047 @include c-d30v.texi
6051 @include c-h8300.texi
6055 @include c-h8500.texi
6059 @include c-hppa.texi
6063 @include c-i370.texi
6067 @include c-i386.texi
6071 @include c-i860.texi
6075 @include c-i960.texi
6079 @include c-ia64.texi
6083 @include c-ip2k.texi
6087 @include c-m32r.texi
6091 @include c-m68k.texi
6095 @include c-m68hc11.texi
6099 @include c-m88k.texi
6103 @include c-mips.texi
6107 @include c-mmix.texi
6111 @include c-msp430.texi
6115 @include c-ns32k.texi
6119 @include c-pdp11.texi
6132 @include c-sh64.texi
6136 @include c-sparc.texi
6140 @include c-tic54x.texi
6152 @include c-v850.texi
6156 @include c-xtensa.texi
6160 @c reverse effect of @down at top of generic Machine-Dep chapter
6164 @node Reporting Bugs
6165 @chapter Reporting Bugs
6166 @cindex bugs in assembler
6167 @cindex reporting bugs in assembler
6169 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6171 Reporting a bug may help you by bringing a solution to your problem, or it may
6172 not. But in any case the principal function of a bug report is to help the
6173 entire community by making the next version of @command{@value{AS}} work better.
6174 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6176 In order for a bug report to serve its purpose, you must include the
6177 information that enables us to fix the bug.
6180 * Bug Criteria:: Have you found a bug?
6181 * Bug Reporting:: How to report bugs
6185 @section Have You Found a Bug?
6186 @cindex bug criteria
6188 If you are not sure whether you have found a bug, here are some guidelines:
6191 @cindex fatal signal
6192 @cindex assembler crash
6193 @cindex crash of assembler
6195 If the assembler gets a fatal signal, for any input whatever, that is a
6196 @command{@value{AS}} bug. Reliable assemblers never crash.
6198 @cindex error on valid input
6200 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6202 @cindex invalid input
6204 If @command{@value{AS}} does not produce an error message for invalid input, that
6205 is a bug. However, you should note that your idea of ``invalid input'' might
6206 be our idea of ``an extension'' or ``support for traditional practice''.
6209 If you are an experienced user of assemblers, your suggestions for improvement
6210 of @command{@value{AS}} are welcome in any case.
6214 @section How to Report Bugs
6216 @cindex assembler bugs, reporting
6218 A number of companies and individuals offer support for @sc{gnu} products. If
6219 you obtained @command{@value{AS}} from a support organization, we recommend you
6220 contact that organization first.
6222 You can find contact information for many support companies and
6223 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6226 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6227 to @samp{bug-binutils@@gnu.org}.
6229 The fundamental principle of reporting bugs usefully is this:
6230 @strong{report all the facts}. If you are not sure whether to state a
6231 fact or leave it out, state it!
6233 Often people omit facts because they think they know what causes the problem
6234 and assume that some details do not matter. Thus, you might assume that the
6235 name of a symbol you use in an example does not matter. Well, probably it does
6236 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6237 happens to fetch from the location where that name is stored in memory;
6238 perhaps, if the name were different, the contents of that location would fool
6239 the assembler into doing the right thing despite the bug. Play it safe and
6240 give a specific, complete example. That is the easiest thing for you to do,
6241 and the most helpful.
6243 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6244 it is new to us. Therefore, always write your bug reports on the assumption
6245 that the bug has not been reported previously.
6247 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6248 bell?'' This cannot help us fix a bug, so it is basically useless. We
6249 respond by asking for enough details to enable us to investigate.
6250 You might as well expedite matters by sending them to begin with.
6252 To enable us to fix the bug, you should include all these things:
6256 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6257 it with the @samp{--version} argument.
6259 Without this, we will not know whether there is any point in looking for
6260 the bug in the current version of @command{@value{AS}}.
6263 Any patches you may have applied to the @command{@value{AS}} source.
6266 The type of machine you are using, and the operating system name and
6270 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6274 The command arguments you gave the assembler to assemble your example and
6275 observe the bug. To guarantee you will not omit something important, list them
6276 all. A copy of the Makefile (or the output from make) is sufficient.
6278 If we were to try to guess the arguments, we would probably guess wrong
6279 and then we might not encounter the bug.
6282 A complete input file that will reproduce the bug. If the bug is observed when
6283 the assembler is invoked via a compiler, send the assembler source, not the
6284 high level language source. Most compilers will produce the assembler source
6285 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6286 the options @samp{-v --save-temps}; this will save the assembler source in a
6287 file with an extension of @file{.s}, and also show you exactly how
6288 @command{@value{AS}} is being run.
6291 A description of what behavior you observe that you believe is
6292 incorrect. For example, ``It gets a fatal signal.''
6294 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6295 will certainly notice it. But if the bug is incorrect output, we might not
6296 notice unless it is glaringly wrong. You might as well not give us a chance to
6299 Even if the problem you experience is a fatal signal, you should still say so
6300 explicitly. Suppose something strange is going on, such as, your copy of
6301 @command{@value{AS}} is out of synch, or you have encountered a bug in the C
6302 library on your system. (This has happened!) Your copy might crash and ours
6303 would not. If you told us to expect a crash, then when ours fails to crash, we
6304 would know that the bug was not happening for us. If you had not told us to
6305 expect a crash, then we would not be able to draw any conclusion from our
6309 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6310 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6311 option. Always send diffs from the old file to the new file. If you even
6312 discuss something in the @command{@value{AS}} source, refer to it by context, not
6315 The line numbers in our development sources will not match those in your
6316 sources. Your line numbers would convey no useful information to us.
6319 Here are some things that are not necessary:
6323 A description of the envelope of the bug.
6325 Often people who encounter a bug spend a lot of time investigating
6326 which changes to the input file will make the bug go away and which
6327 changes will not affect it.
6329 This is often time consuming and not very useful, because the way we
6330 will find the bug is by running a single example under the debugger
6331 with breakpoints, not by pure deduction from a series of examples.
6332 We recommend that you save your time for something else.
6334 Of course, if you can find a simpler example to report @emph{instead}
6335 of the original one, that is a convenience for us. Errors in the
6336 output will be easier to spot, running under the debugger will take
6337 less time, and so on.
6339 However, simplification is not vital; if you do not want to do this,
6340 report the bug anyway and send us the entire test case you used.
6343 A patch for the bug.
6345 A patch for the bug does help us if it is a good one. But do not omit
6346 the necessary information, such as the test case, on the assumption that
6347 a patch is all we need. We might see problems with your patch and decide
6348 to fix the problem another way, or we might not understand it at all.
6350 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6351 construct an example that will make the program follow a certain path through
6352 the code. If you do not send us the example, we will not be able to construct
6353 one, so we will not be able to verify that the bug is fixed.
6355 And if we cannot understand what bug you are trying to fix, or why your
6356 patch should be an improvement, we will not install it. A test case will
6357 help us to understand.
6360 A guess about what the bug is or what it depends on.
6362 Such guesses are usually wrong. Even we cannot guess right about such
6363 things without first using the debugger to find the facts.
6366 @node Acknowledgements
6367 @chapter Acknowledgements
6369 If you have contributed to @command{@value{AS}} and your name isn't listed here,
6370 it is not meant as a slight. We just don't know about it. Send mail to the
6371 maintainer, and we'll correct the situation. Currently
6373 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6375 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6378 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6379 information and the 68k series machines, most of the preprocessing pass, and
6380 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6382 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6383 many bug fixes, including merging support for several processors, breaking GAS
6384 up to handle multiple object file format back ends (including heavy rewrite,
6385 testing, an integration of the coff and b.out back ends), adding configuration
6386 including heavy testing and verification of cross assemblers and file splits
6387 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6388 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6389 port (including considerable amounts of reverse engineering), a SPARC opcode
6390 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6391 assertions and made them work, much other reorganization, cleanup, and lint.
6393 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6394 in format-specific I/O modules.
6396 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6397 has done much work with it since.
6399 The Intel 80386 machine description was written by Eliot Dresselhaus.
6401 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6403 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6404 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6406 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6407 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6408 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6409 support a.out format.
6411 Support for the Zilog Z8k and Renesas H8/300 and H8/500 processors (tc-z8k,
6412 tc-h8300, tc-h8500), and IEEE 695 object file format (obj-ieee), was written by
6413 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6414 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6417 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6418 simplified the configuration of which versions accept which directives. He
6419 updated the 68k machine description so that Motorola's opcodes always produced
6420 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6421 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6422 cross-compilation support, and one bug in relaxation that took a week and
6423 required the proverbial one-bit fix.
6425 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6426 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6427 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6428 PowerPC assembler, and made a few other minor patches.
6430 Steve Chamberlain made @command{@value{AS}} able to generate listings.
6432 Hewlett-Packard contributed support for the HP9000/300.
6434 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6435 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6436 formats). This work was supported by both the Center for Software Science at
6437 the University of Utah and Cygnus Support.
6439 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6440 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6441 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6442 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6443 and some initial 64-bit support).
6445 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6447 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6448 support for openVMS/Alpha.
6450 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6453 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6454 Inc. added support for Xtensa processors.
6456 Several engineers at Cygnus Support have also provided many small bug fixes and
6457 configuration enhancements.
6459 Many others have contributed large or small bugfixes and enhancements. If
6460 you have contributed significant work and are not mentioned on this list, and
6461 want to be, let us know. Some of the history has been lost; we are not
6462 intentionally leaving anyone out.