1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
3 @c 2001, 2002, 2003, 2004, 2005, 2006, 2007
4 @c Free Software Foundation, Inc.
5 @c UPDATE!! On future updates--
6 @c (1) check for new machine-dep cmdline options in
7 @c md_parse_option definitions in config/tc-*.c
8 @c (2) for platform-specific directives, examine md_pseudo_op
10 @c (3) for object-format specific directives, examine obj_pseudo_op
12 @c (4) portable directives in potable[] in read.c
16 @macro gcctabopt{body}
19 @c defaults, config file may override:
24 @include asconfig.texi
29 @c common OR combinations of conditions
52 @set abnormal-separator
56 @settitle Using @value{AS}
59 @settitle Using @value{AS} (@value{TARGET})
61 @setchapternewpage odd
66 @c WARE! Some of the machine-dependent sections contain tables of machine
67 @c instructions. Except in multi-column format, these tables look silly.
68 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
69 @c the multi-col format is faked within @example sections.
71 @c Again unfortunately, the natural size that fits on a page, for these tables,
72 @c is different depending on whether or not smallbook is turned on.
73 @c This matters, because of order: text flow switches columns at each page
76 @c The format faked in this source works reasonably well for smallbook,
77 @c not well for the default large-page format. This manual expects that if you
78 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
79 @c tables in question. You can turn on one without the other at your
80 @c discretion, of course.
83 @c the insn tables look just as silly in info files regardless of smallbook,
84 @c might as well show 'em anyways.
90 * As: (as). The GNU assembler.
91 * Gas: (as). The GNU assembler.
100 This file documents the GNU Assembler "@value{AS}".
102 @c man begin COPYRIGHT
103 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002,
104 2006, 2007 Free Software Foundation, Inc.
106 Permission is granted to copy, distribute and/or modify this document
107 under the terms of the GNU Free Documentation License, Version 1.1
108 or any later version published by the Free Software Foundation;
109 with no Invariant Sections, with no Front-Cover Texts, and with no
110 Back-Cover Texts. A copy of the license is included in the
111 section entitled ``GNU Free Documentation License''.
117 @title Using @value{AS}
118 @subtitle The @sc{gnu} Assembler
120 @subtitle for the @value{TARGET} family
122 @ifset VERSION_PACKAGE
124 @subtitle @value{VERSION_PACKAGE}
127 @subtitle Version @value{VERSION}
130 The Free Software Foundation Inc.@: thanks The Nice Computer
131 Company of Australia for loaning Dean Elsner to write the
132 first (Vax) version of @command{as} for Project @sc{gnu}.
133 The proprietors, management and staff of TNCCA thank FSF for
134 distracting the boss while they got some work
137 @author Dean Elsner, Jay Fenlason & friends
141 \hfill {\it Using {\tt @value{AS}}}\par
142 \hfill Edited by Cygnus Support\par
144 %"boxit" macro for figures:
145 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
146 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
147 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
148 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
149 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
152 @vskip 0pt plus 1filll
153 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001, 2002,
154 2006, 2007 Free Software Foundation, Inc.
156 Permission is granted to copy, distribute and/or modify this document
157 under the terms of the GNU Free Documentation License, Version 1.1
158 or any later version published by the Free Software Foundation;
159 with no Invariant Sections, with no Front-Cover Texts, and with no
160 Back-Cover Texts. A copy of the license is included in the
161 section entitled ``GNU Free Documentation License''.
168 @top Using @value{AS}
170 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}}
171 @ifset VERSION_PACKAGE
172 @value{VERSION_PACKAGE}
174 version @value{VERSION}.
176 This version of the file describes @command{@value{AS}} configured to generate
177 code for @value{TARGET} architectures.
180 This document is distributed under the terms of the GNU Free
181 Documentation License. A copy of the license is included in the
182 section entitled ``GNU Free Documentation License''.
185 * Overview:: Overview
186 * Invoking:: Command-Line Options
188 * Sections:: Sections and Relocation
190 * Expressions:: Expressions
191 * Pseudo Ops:: Assembler Directives
192 * Machine Dependencies:: Machine Dependent Features
193 * Reporting Bugs:: Reporting Bugs
194 * Acknowledgements:: Who Did What
195 * GNU Free Documentation License:: GNU Free Documentation License
196 * AS Index:: AS Index
203 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
205 This version of the manual describes @command{@value{AS}} configured to generate
206 code for @value{TARGET} architectures.
210 @cindex invocation summary
211 @cindex option summary
212 @cindex summary of options
213 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
214 see @ref{Invoking,,Command-Line Options}.
216 @c man title AS the portable GNU assembler.
220 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
224 @c We don't use deffn and friends for the following because they seem
225 @c to be limited to one line for the header.
227 @c man begin SYNOPSIS
228 @value{AS} [@b{-a}[@b{cdhlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
229 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
230 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
231 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
232 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
233 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
234 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
235 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
236 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
237 [@b{--target-help}] [@var{target-options}]
238 [@b{--}|@var{files} @dots{}]
240 @c Target dependent options are listed below. Keep the list sorted.
241 @c Add an empty line for separation.
244 @emph{Target Alpha options:}
246 [@b{-mdebug} | @b{-no-mdebug}]
247 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
248 [@b{-F}] [@b{-32addr}]
252 @emph{Target ARC options:}
258 @emph{Target ARM options:}
259 @c Don't document the deprecated options
260 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
261 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
262 [@b{-mfpu}=@var{floating-point-format}]
263 [@b{-mfloat-abi}=@var{abi}]
264 [@b{-meabi}=@var{ver}]
267 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
268 @b{-mapcs-reentrant}]
269 [@b{-mthumb-interwork}] [@b{-k}]
273 @emph{Target CRIS options:}
274 [@b{--underscore} | @b{--no-underscore}]
276 [@b{--emulation=criself} | @b{--emulation=crisaout}]
277 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
278 @c Deprecated -- deliberately not documented.
283 @emph{Target D10V options:}
288 @emph{Target D30V options:}
289 [@b{-O}|@b{-n}|@b{-N}]
292 @c Renesas family chips have no machine-dependent assembler options
295 @c HPPA has no machine-dependent assembler options (yet).
299 @emph{Target i386 options:}
300 [@b{--32}|@b{--64}] [@b{-n}]
301 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}]
305 @emph{Target i960 options:}
306 @c see md_parse_option in tc-i960.c
307 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
309 [@b{-b}] [@b{-no-relax}]
313 @emph{Target IA-64 options:}
314 [@b{-mconstant-gp}|@b{-mauto-pic}]
315 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
317 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
318 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
319 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
320 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
324 @emph{Target IP2K options:}
325 [@b{-mip2022}|@b{-mip2022ext}]
329 @emph{Target M32C options:}
330 [@b{-m32c}|@b{-m16c}]
334 @emph{Target M32R options:}
335 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
340 @emph{Target M680X0 options:}
341 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
345 @emph{Target M68HC11 options:}
346 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
347 [@b{-mshort}|@b{-mlong}]
348 [@b{-mshort-double}|@b{-mlong-double}]
349 [@b{--force-long-branches}] [@b{--short-branches}]
350 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
351 [@b{--print-opcodes}] [@b{--generate-example}]
355 @emph{Target MCORE options:}
356 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
357 [@b{-mcpu=[210|340]}]
361 @emph{Target MIPS options:}
362 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
363 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
364 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
365 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
366 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
367 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
368 [@b{-mips64}] [@b{-mips64r2}]
369 [@b{-construct-floats}] [@b{-no-construct-floats}]
370 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
371 [@b{-mfix7000}] [@b{-mno-fix7000}]
372 [@b{-mips16}] [@b{-no-mips16}]
373 [@b{-msmartmips}] [@b{-mno-smartmips}]
374 [@b{-mips3d}] [@b{-no-mips3d}]
375 [@b{-mdmx}] [@b{-no-mdmx}]
376 [@b{-mdsp}] [@b{-mno-dsp}]
377 [@b{-mdspr2}] [@b{-mno-dspr2}]
378 [@b{-mmt}] [@b{-mno-mt}]
379 [@b{-mdebug}] [@b{-no-mdebug}]
380 [@b{-mpdr}] [@b{-mno-pdr}]
384 @emph{Target MMIX options:}
385 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
386 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
387 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
388 [@b{--linker-allocated-gregs}]
392 @emph{Target PDP11 options:}
393 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
394 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
395 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
399 @emph{Target picoJava options:}
404 @emph{Target PowerPC options:}
405 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|
406 @b{-m403}|@b{-m405}|@b{-mppc64}|@b{-m620}|@b{-mppc64bridge}|@b{-mbooke}|
407 @b{-mbooke32}|@b{-mbooke64}]
408 [@b{-mcom}|@b{-many}|@b{-maltivec}] [@b{-memb}]
409 [@b{-mregnames}|@b{-mno-regnames}]
410 [@b{-mrelocatable}|@b{-mrelocatable-lib}]
411 [@b{-mlittle}|@b{-mlittle-endian}|@b{-mbig}|@b{-mbig-endian}]
412 [@b{-msolaris}|@b{-mno-solaris}]
416 @emph{Target SPARC options:}
417 @c The order here is important. See c-sparc.texi.
418 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
419 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
420 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
425 @emph{Target TIC54X options:}
426 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
427 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
432 @emph{Target Z80 options:}
433 [@b{-z80}] [@b{-r800}]
434 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
435 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
436 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
437 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
438 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
439 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
443 @c Z8000 has no machine-dependent assembler options
447 @emph{Target Xtensa options:}
448 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
449 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
450 [@b{--[no-]transform}]
451 [@b{--rename-section} @var{oldname}=@var{newname}]
459 @include at-file.texi
462 Turn on listings, in any of a variety of ways:
466 omit false conditionals
469 omit debugging directives
472 include high-level source
478 include macro expansions
481 omit forms processing
487 set the name of the listing file
490 You may combine these options; for example, use @samp{-aln} for assembly
491 listing without forms processing. The @samp{=file} option, if used, must be
492 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
495 Begin in alternate macro mode.
497 @xref{Altmacro,,@code{.altmacro}}.
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
508 value. The value of the symbol can be overridden inside a source file via the
509 use of a @code{.set} pseudo-op.
512 ``fast''---skip whitespace and comment preprocessing (assume source is
517 Generate debugging information for each assembler source line using whichever
518 debug format is preferred by the target. This currently means either STABS,
522 Generate stabs debugging information for each assembler line. This
523 may help debugging assembler code, if the debugger can handle it.
526 Generate stabs debugging information for each assembler line, with GNU
527 extensions that probably only gdb can handle, and that could make other
528 debuggers crash or refuse to read your program. This
529 may help debugging assembler code. Currently the only GNU extension is
530 the location of the current working directory at assembling time.
533 Generate DWARF2 debugging information for each assembler line. This
534 may help debugging assembler code, if the debugger can handle it. Note---this
535 option is only supported by some targets, not all of them.
538 Print a summary of the command line options and exit.
541 Print a summary of all target specific options and exit.
544 Add directory @var{dir} to the search list for @code{.include} directives.
547 Don't warn about signed overflow.
550 @ifclear DIFF-TBL-KLUGE
551 This option is accepted but has no effect on the @value{TARGET} family.
553 @ifset DIFF-TBL-KLUGE
554 Issue warnings when difference tables altered for long displacements.
559 Keep (in the symbol table) local symbols. These symbols start with
560 system-specific local label prefixes, typically @samp{.L} for ELF systems
561 or @samp{L} for traditional a.out systems.
566 @item --listing-lhs-width=@var{number}
567 Set the maximum width, in words, of the output data column for an assembler
568 listing to @var{number}.
570 @item --listing-lhs-width2=@var{number}
571 Set the maximum width, in words, of the output data column for continuation
572 lines in an assembler listing to @var{number}.
574 @item --listing-rhs-width=@var{number}
575 Set the maximum width of an input source line, as displayed in a listing, to
578 @item --listing-cont-lines=@var{number}
579 Set the maximum number of lines printed in a listing for a single line of input
582 @item -o @var{objfile}
583 Name the object-file output from @command{@value{AS}} @var{objfile}.
586 Fold the data section into the text section.
588 @kindex --hash-size=@var{number}
589 Set the default size of GAS's hash tables to a prime number close to
590 @var{number}. Increasing this value can reduce the length of time it takes the
591 assembler to perform its tasks, at the expense of increasing the assembler's
592 memory requirements. Similarly reducing this value can reduce the memory
593 requirements at the expense of speed.
595 @item --reduce-memory-overheads
596 This option reduces GAS's memory requirements, at the expense of making the
597 assembly processes slower. Currently this switch is a synonym for
598 @samp{--hash-size=4051}, but in the future it may have other effects as well.
601 Print the maximum space (in bytes) and total time (in seconds) used by
604 @item --strip-local-absolute
605 Remove local absolute symbols from the outgoing symbol table.
609 Print the @command{as} version.
612 Print the @command{as} version and exit.
616 Suppress warning messages.
618 @item --fatal-warnings
619 Treat warnings as errors.
622 Don't suppress warning messages or treat them as errors.
631 Generate an object file even after errors.
633 @item -- | @var{files} @dots{}
634 Standard input, or source files to assemble.
639 The following options are available when @value{AS} is configured for
644 This option selects the core processor variant.
646 Select either big-endian (-EB) or little-endian (-EL) output.
651 The following options are available when @value{AS} is configured for the ARM
655 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
656 Specify which ARM processor variant is the target.
657 @item -march=@var{architecture}[+@var{extension}@dots{}]
658 Specify which ARM architecture variant is used by the target.
659 @item -mfpu=@var{floating-point-format}
660 Select which Floating Point architecture is the target.
661 @item -mfloat-abi=@var{abi}
662 Select which floating point ABI is in use.
664 Enable Thumb only instruction decoding.
665 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
666 Select which procedure calling convention is in use.
668 Select either big-endian (-EB) or little-endian (-EL) output.
669 @item -mthumb-interwork
670 Specify that the code has been generated with interworking between Thumb and
673 Specify that PIC code has been generated.
678 See the info pages for documentation of the CRIS-specific options.
682 The following options are available when @value{AS} is configured for
685 @cindex D10V optimization
686 @cindex optimization, D10V
688 Optimize output by parallelizing instructions.
693 The following options are available when @value{AS} is configured for a D30V
696 @cindex D30V optimization
697 @cindex optimization, D30V
699 Optimize output by parallelizing instructions.
703 Warn when nops are generated.
705 @cindex D30V nops after 32-bit multiply
707 Warn when a nop after a 32-bit multiply instruction is generated.
712 The following options are available when @value{AS} is configured for the
713 Intel 80960 processor.
716 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
717 Specify which variant of the 960 architecture is the target.
720 Add code to collect statistics about branches taken.
723 Do not alter compare-and-branch instructions for long displacements;
730 The following options are available when @value{AS} is configured for the
736 Specifies that the extended IP2022 instructions are allowed.
739 Restores the default behaviour, which restricts the permitted instructions to
740 just the basic IP2022 ones.
746 The following options are available when @value{AS} is configured for the
747 Renesas M32C and M16C processors.
752 Assemble M32C instructions.
755 Assemble M16C instructions (the default).
761 The following options are available when @value{AS} is configured for the
762 Renesas M32R (formerly Mitsubishi M32R) series.
767 Specify which processor in the M32R family is the target. The default
768 is normally the M32R, but this option changes it to the M32RX.
770 @item --warn-explicit-parallel-conflicts or --Wp
771 Produce warning messages when questionable parallel constructs are
774 @item --no-warn-explicit-parallel-conflicts or --Wnp
775 Do not produce warning messages when questionable parallel constructs are
782 The following options are available when @value{AS} is configured for the
783 Motorola 68000 series.
788 Shorten references to undefined symbols, to one word instead of two.
790 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
791 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
792 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
793 Specify what processor in the 68000 family is the target. The default
794 is normally the 68020, but this can be changed at configuration time.
796 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
797 The target machine does (or does not) have a floating-point coprocessor.
798 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
799 the basic 68000 is not compatible with the 68881, a combination of the
800 two can be specified, since it's possible to do emulation of the
801 coprocessor instructions with the main processor.
803 @item -m68851 | -mno-68851
804 The target machine does (or does not) have a memory-management
805 unit coprocessor. The default is to assume an MMU for 68020 and up.
812 For details about the PDP-11 machine dependent features options,
813 see @ref{PDP-11-Options}.
816 @item -mpic | -mno-pic
817 Generate position-independent (or position-dependent) code. The
818 default is @option{-mpic}.
821 @itemx -mall-extensions
822 Enable all instruction set extensions. This is the default.
824 @item -mno-extensions
825 Disable all instruction set extensions.
827 @item -m@var{extension} | -mno-@var{extension}
828 Enable (or disable) a particular instruction set extension.
831 Enable the instruction set extensions supported by a particular CPU, and
832 disable all other extensions.
834 @item -m@var{machine}
835 Enable the instruction set extensions supported by a particular machine
836 model, and disable all other extensions.
842 The following options are available when @value{AS} is configured for
843 a picoJava processor.
847 @cindex PJ endianness
848 @cindex endianness, PJ
849 @cindex big endian output, PJ
851 Generate ``big endian'' format output.
853 @cindex little endian output, PJ
855 Generate ``little endian'' format output.
861 The following options are available when @value{AS} is configured for the
862 Motorola 68HC11 or 68HC12 series.
866 @item -m68hc11 | -m68hc12 | -m68hcs12
867 Specify what processor is the target. The default is
868 defined by the configuration option when building the assembler.
871 Specify to use the 16-bit integer ABI.
874 Specify to use the 32-bit integer ABI.
877 Specify to use the 32-bit double ABI.
880 Specify to use the 64-bit double ABI.
882 @item --force-long-branches
883 Relative branches are turned into absolute ones. This concerns
884 conditional branches, unconditional branches and branches to a
887 @item -S | --short-branches
888 Do not turn relative branches into absolute ones
889 when the offset is out of range.
891 @item --strict-direct-mode
892 Do not turn the direct addressing mode into extended addressing mode
893 when the instruction does not support direct addressing mode.
895 @item --print-insn-syntax
896 Print the syntax of instruction in case of error.
898 @item --print-opcodes
899 print the list of instructions with syntax and then exit.
901 @item --generate-example
902 print an example of instruction for each possible instruction and then exit.
903 This option is only useful for testing @command{@value{AS}}.
909 The following options are available when @command{@value{AS}} is configured
910 for the SPARC architecture:
913 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
914 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
915 Explicitly select a variant of the SPARC architecture.
917 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
918 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
920 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
921 UltraSPARC extensions.
923 @item -xarch=v8plus | -xarch=v8plusa
924 For compatibility with the Solaris v9 assembler. These options are
925 equivalent to -Av8plus and -Av8plusa, respectively.
928 Warn when the assembler switches to another architecture.
933 The following options are available when @value{AS} is configured for the 'c54x
938 Enable extended addressing mode. All addresses and relocations will assume
939 extended addressing (usually 23 bits).
940 @item -mcpu=@var{CPU_VERSION}
941 Sets the CPU version being compiled for.
942 @item -merrors-to-file @var{FILENAME}
943 Redirect error output to a file, for broken systems which don't support such
944 behaviour in the shell.
949 The following options are available when @value{AS} is configured for
950 a @sc{mips} processor.
954 This option sets the largest size of an object that can be referenced
955 implicitly with the @code{gp} register. It is only accepted for targets that
956 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
958 @cindex MIPS endianness
959 @cindex endianness, MIPS
960 @cindex big endian output, MIPS
962 Generate ``big endian'' format output.
964 @cindex little endian output, MIPS
966 Generate ``little endian'' format output.
978 Generate code for a particular @sc{mips} Instruction Set Architecture level.
979 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
980 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
981 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
982 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
984 correspond to generic
985 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
986 and @samp{MIPS64 Release 2}
987 ISA processors, respectively.
989 @item -march=@var{CPU}
990 Generate code for a particular @sc{mips} cpu.
992 @item -mtune=@var{cpu}
993 Schedule and tune for a particular @sc{mips} cpu.
997 Cause nops to be inserted if the read of the destination register
998 of an mfhi or mflo instruction occurs in the following two instructions.
1002 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1003 section instead of the standard ELF .stabs sections.
1007 Control generation of @code{.pdr} sections.
1011 The register sizes are normally inferred from the ISA and ABI, but these
1012 flags force a certain group of registers to be treated as 32 bits wide at
1013 all times. @samp{-mgp32} controls the size of general-purpose registers
1014 and @samp{-mfp32} controls the size of floating-point registers.
1018 Generate code for the MIPS 16 processor. This is equivalent to putting
1019 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1020 turns off this option.
1023 @itemx -mno-smartmips
1024 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1025 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1026 @samp{-mno-smartmips} turns off this option.
1030 Generate code for the MIPS-3D Application Specific Extension.
1031 This tells the assembler to accept MIPS-3D instructions.
1032 @samp{-no-mips3d} turns off this option.
1036 Generate code for the MDMX Application Specific Extension.
1037 This tells the assembler to accept MDMX instructions.
1038 @samp{-no-mdmx} turns off this option.
1042 Generate code for the DSP Release 1 Application Specific Extension.
1043 This tells the assembler to accept DSP Release 1 instructions.
1044 @samp{-mno-dsp} turns off this option.
1048 Generate code for the DSP Release 2 Application Specific Extension.
1049 This option implies -mdsp.
1050 This tells the assembler to accept DSP Release 2 instructions.
1051 @samp{-mno-dspr2} turns off this option.
1055 Generate code for the MT Application Specific Extension.
1056 This tells the assembler to accept MT instructions.
1057 @samp{-mno-mt} turns off this option.
1059 @item --construct-floats
1060 @itemx --no-construct-floats
1061 The @samp{--no-construct-floats} option disables the construction of
1062 double width floating point constants by loading the two halves of the
1063 value into the two single width floating point registers that make up
1064 the double width register. By default @samp{--construct-floats} is
1065 selected, allowing construction of these floating point constants.
1068 @item --emulation=@var{name}
1069 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1070 for some other target, in all respects, including output format (choosing
1071 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1072 debugging information or store symbol table information, and default
1073 endianness. The available configuration names are: @samp{mipsecoff},
1074 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1075 @samp{mipsbelf}. The first two do not alter the default endianness from that
1076 of the primary target for which the assembler was configured; the others change
1077 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1078 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1079 selection in any case.
1081 This option is currently supported only when the primary target
1082 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1083 Furthermore, the primary target or others specified with
1084 @samp{--enable-targets=@dots{}} at configuration time must include support for
1085 the other format, if both are to be available. For example, the Irix 5
1086 configuration includes support for both.
1088 Eventually, this option will support more configurations, with more
1089 fine-grained control over the assembler's behavior, and will be supported for
1093 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1100 Control how to deal with multiplication overflow and division by zero.
1101 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1102 (and only work for Instruction Set Architecture level 2 and higher);
1103 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1107 When this option is used, @command{@value{AS}} will issue a warning every
1108 time it generates a nop instruction from a macro.
1113 The following options are available when @value{AS} is configured for
1119 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1120 The command line option @samp{-nojsri2bsr} can be used to disable it.
1124 Enable or disable the silicon filter behaviour. By default this is disabled.
1125 The default can be overridden by the @samp{-sifilter} command line option.
1128 Alter jump instructions for long displacements.
1130 @item -mcpu=[210|340]
1131 Select the cpu type on the target hardware. This controls which instructions
1135 Assemble for a big endian target.
1138 Assemble for a little endian target.
1144 See the info pages for documentation of the MMIX-specific options.
1148 The following options are available when @value{AS} is configured for
1149 an Xtensa processor.
1152 @item --text-section-literals | --no-text-section-literals
1153 With @option{--text-@-section-@-literals}, literal pools are interspersed
1154 in the text section. The default is
1155 @option{--no-@-text-@-section-@-literals}, which places literals in a
1156 separate section in the output file. These options only affect literals
1157 referenced via PC-relative @code{L32R} instructions; literals for
1158 absolute mode @code{L32R} instructions are handled separately.
1160 @item --absolute-literals | --no-absolute-literals
1161 Indicate to the assembler whether @code{L32R} instructions use absolute
1162 or PC-relative addressing. The default is to assume absolute addressing
1163 if the Xtensa processor includes the absolute @code{L32R} addressing
1164 option. Otherwise, only the PC-relative @code{L32R} mode can be used.
1166 @item --target-align | --no-target-align
1167 Enable or disable automatic alignment to reduce branch penalties at the
1168 expense of some code density. The default is @option{--target-@-align}.
1170 @item --longcalls | --no-longcalls
1171 Enable or disable transformation of call instructions to allow calls
1172 across a greater range of addresses. The default is
1173 @option{--no-@-longcalls}.
1175 @item --transform | --no-transform
1176 Enable or disable all assembler transformations of Xtensa instructions.
1177 The default is @option{--transform};
1178 @option{--no-transform} should be used only in the rare cases when the
1179 instructions must be exactly as specified in the assembly source.
1184 The following options are available when @value{AS} is configured for
1185 a Z80 family processor.
1188 Assemble for Z80 processor.
1190 Assemble for R800 processor.
1191 @item -ignore-undocumented-instructions
1193 Assemble undocumented Z80 instructions that also work on R800 without warning.
1194 @item -ignore-unportable-instructions
1196 Assemble all undocumented Z80 instructions without warning.
1197 @item -warn-undocumented-instructions
1199 Issue a warning for undocumented Z80 instructions that also work on R800.
1200 @item -warn-unportable-instructions
1202 Issue a warning for undocumented Z80 instructions that do not work on R800.
1203 @item -forbid-undocumented-instructions
1205 Treat all undocumented instructions as errors.
1206 @item -forbid-unportable-instructions
1208 Treat undocumented Z80 instructions that do not work on R800 as errors.
1215 * Manual:: Structure of this Manual
1216 * GNU Assembler:: The GNU Assembler
1217 * Object Formats:: Object File Formats
1218 * Command Line:: Command Line
1219 * Input Files:: Input Files
1220 * Object:: Output (Object) File
1221 * Errors:: Error and Warning Messages
1225 @section Structure of this Manual
1227 @cindex manual, structure and purpose
1228 This manual is intended to describe what you need to know to use
1229 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1230 notation for symbols, constants, and expressions; the directives that
1231 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1234 We also cover special features in the @value{TARGET}
1235 configuration of @command{@value{AS}}, including assembler directives.
1238 This manual also describes some of the machine-dependent features of
1239 various flavors of the assembler.
1242 @cindex machine instructions (not covered)
1243 On the other hand, this manual is @emph{not} intended as an introduction
1244 to programming in assembly language---let alone programming in general!
1245 In a similar vein, we make no attempt to introduce the machine
1246 architecture; we do @emph{not} describe the instruction set, standard
1247 mnemonics, registers or addressing modes that are standard to a
1248 particular architecture.
1250 You may want to consult the manufacturer's
1251 machine architecture manual for this information.
1255 For information on the H8/300 machine instruction set, see @cite{H8/300
1256 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1257 Programming Manual} (Renesas).
1260 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1261 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1262 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1263 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1266 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1270 @c I think this is premature---doc@cygnus.com, 17jan1991
1272 Throughout this manual, we assume that you are running @dfn{GNU},
1273 the portable operating system from the @dfn{Free Software
1274 Foundation, Inc.}. This restricts our attention to certain kinds of
1275 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1276 once this assumption is granted examples and definitions need less
1279 @command{@value{AS}} is part of a team of programs that turn a high-level
1280 human-readable series of instructions into a low-level
1281 computer-readable series of instructions. Different versions of
1282 @command{@value{AS}} are used for different kinds of computer.
1285 @c There used to be a section "Terminology" here, which defined
1286 @c "contents", "byte", "word", and "long". Defining "word" to any
1287 @c particular size is confusing when the .word directive may generate 16
1288 @c bits on one machine and 32 bits on another; in general, for the user
1289 @c version of this manual, none of these terms seem essential to define.
1290 @c They were used very little even in the former draft of the manual;
1291 @c this draft makes an effort to avoid them (except in names of
1295 @section The GNU Assembler
1297 @c man begin DESCRIPTION
1298 @sc{gnu} @command{as} will be removed from a future version of the
1299 FreeBSD base system. Users who require
1300 @sc{gnu} @command{as} are advised to install the binutils port or package.
1302 @sc{gnu} @command{as} is really a family of assemblers.
1304 This manual describes @command{@value{AS}}, a member of that family which is
1305 configured for the @value{TARGET} architectures.
1307 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1308 should find a fairly similar environment when you use it on another
1309 architecture. Each version has much in common with the others,
1310 including object file formats, most assembler directives (often called
1311 @dfn{pseudo-ops}) and assembler syntax.@refill
1313 @cindex purpose of @sc{gnu} assembler
1314 @command{@value{AS}} is primarily intended to assemble the output of the
1315 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1316 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1317 assemble correctly everything that other assemblers for the same
1318 machine would assemble.
1320 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1323 @c This remark should appear in generic version of manual; assumption
1324 @c here is that generic version sets M680x0.
1325 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1326 assembler for the same architecture; for example, we know of several
1327 incompatible versions of 680x0 assembly language syntax.
1332 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1333 program in one pass of the source file. This has a subtle impact on the
1334 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1336 @node Object Formats
1337 @section Object File Formats
1339 @cindex object file format
1340 The @sc{gnu} assembler can be configured to produce several alternative
1341 object file formats. For the most part, this does not affect how you
1342 write assembly language programs; but directives for debugging symbols
1343 are typically different in different file formats. @xref{Symbol
1344 Attributes,,Symbol Attributes}.
1347 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1348 @value{OBJ-NAME} format object files.
1350 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1352 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1353 @code{b.out} or COFF format object files.
1356 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1357 SOM or ELF format object files.
1362 @section Command Line
1364 @cindex command line conventions
1366 After the program name @command{@value{AS}}, the command line may contain
1367 options and file names. Options may appear in any order, and may be
1368 before, after, or between file names. The order of file names is
1371 @cindex standard input, as input file
1373 @file{--} (two hyphens) by itself names the standard input file
1374 explicitly, as one of the files for @command{@value{AS}} to assemble.
1376 @cindex options, command line
1377 Except for @samp{--} any command line argument that begins with a
1378 hyphen (@samp{-}) is an option. Each option changes the behavior of
1379 @command{@value{AS}}. No option changes the way another option works. An
1380 option is a @samp{-} followed by one or more letters; the case of
1381 the letter is important. All options are optional.
1383 Some options expect exactly one file name to follow them. The file
1384 name may either immediately follow the option's letter (compatible
1385 with older assemblers) or it may be the next command argument (@sc{gnu}
1386 standard). These two command lines are equivalent:
1389 @value{AS} -o my-object-file.o mumble.s
1390 @value{AS} -omy-object-file.o mumble.s
1394 @section Input Files
1397 @cindex source program
1398 @cindex files, input
1399 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1400 describe the program input to one run of @command{@value{AS}}. The program may
1401 be in one or more files; how the source is partitioned into files
1402 doesn't change the meaning of the source.
1404 @c I added "con" prefix to "catenation" just to prove I can overcome my
1405 @c APL training... doc@cygnus.com
1406 The source program is a concatenation of the text in all the files, in the
1409 @c man begin DESCRIPTION
1410 Each time you run @command{@value{AS}} it assembles exactly one source
1411 program. The source program is made up of one or more files.
1412 (The standard input is also a file.)
1414 You give @command{@value{AS}} a command line that has zero or more input file
1415 names. The input files are read (from left file name to right). A
1416 command line argument (in any position) that has no special meaning
1417 is taken to be an input file name.
1419 If you give @command{@value{AS}} no file names it attempts to read one input file
1420 from the @command{@value{AS}} standard input, which is normally your terminal. You
1421 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1424 Use @samp{--} if you need to explicitly name the standard input file
1425 in your command line.
1427 If the source is empty, @command{@value{AS}} produces a small, empty object
1432 @subheading Filenames and Line-numbers
1434 @cindex input file linenumbers
1435 @cindex line numbers, in input files
1436 There are two ways of locating a line in the input file (or files) and
1437 either may be used in reporting error messages. One way refers to a line
1438 number in a physical file; the other refers to a line number in a
1439 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1441 @dfn{Physical files} are those files named in the command line given
1442 to @command{@value{AS}}.
1444 @dfn{Logical files} are simply names declared explicitly by assembler
1445 directives; they bear no relation to physical files. Logical file names help
1446 error messages reflect the original source file, when @command{@value{AS}} source
1447 is itself synthesized from other files. @command{@value{AS}} understands the
1448 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1449 @ref{File,,@code{.file}}.
1452 @section Output (Object) File
1458 Every time you run @command{@value{AS}} it produces an output file, which is
1459 your assembly language program translated into numbers. This file
1460 is the object file. Its default name is
1468 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1470 You can give it another name by using the @option{-o} option. Conventionally,
1471 object file names end with @file{.o}. The default name is used for historical
1472 reasons: older assemblers were capable of assembling self-contained programs
1473 directly into a runnable program. (For some formats, this isn't currently
1474 possible, but it can be done for the @code{a.out} format.)
1478 The object file is meant for input to the linker @code{@value{LD}}. It contains
1479 assembled program code, information to help @code{@value{LD}} integrate
1480 the assembled program into a runnable file, and (optionally) symbolic
1481 information for the debugger.
1483 @c link above to some info file(s) like the description of a.out.
1484 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1487 @section Error and Warning Messages
1489 @c man begin DESCRIPTION
1491 @cindex error messages
1492 @cindex warning messages
1493 @cindex messages from assembler
1494 @command{@value{AS}} may write warnings and error messages to the standard error
1495 file (usually your terminal). This should not happen when a compiler
1496 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1497 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1498 grave problem that stops the assembly.
1502 @cindex format of warning messages
1503 Warning messages have the format
1506 file_name:@b{NNN}:Warning Message Text
1510 @cindex line numbers, in warnings/errors
1511 (where @b{NNN} is a line number). If a logical file name has been given
1512 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1513 the current input file is used. If a logical line number was given
1515 (@pxref{Line,,@code{.line}})
1517 then it is used to calculate the number printed,
1518 otherwise the actual line in the current source file is printed. The
1519 message text is intended to be self explanatory (in the grand Unix
1522 @cindex format of error messages
1523 Error messages have the format
1525 file_name:@b{NNN}:FATAL:Error Message Text
1527 The file name and line number are derived as for warning
1528 messages. The actual message text may be rather less explanatory
1529 because many of them aren't supposed to happen.
1532 @chapter Command-Line Options
1534 @cindex options, all versions of assembler
1535 This chapter describes command-line options available in @emph{all}
1536 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
1537 for options specific
1539 to the @value{TARGET} target.
1542 to particular machine architectures.
1545 @c man begin DESCRIPTION
1547 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1548 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1549 The assembler arguments must be separated from each other (and the @samp{-Wa})
1550 by commas. For example:
1553 gcc -c -g -O -Wa,-alh,-L file.c
1557 This passes two options to the assembler: @samp{-alh} (emit a listing to
1558 standard output with high-level and assembly source) and @samp{-L} (retain
1559 local symbols in the symbol table).
1561 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1562 command-line options are automatically passed to the assembler by the compiler.
1563 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1564 precisely what options it passes to each compilation pass, including the
1570 * a:: -a[cdhlns] enable listings
1571 * alternate:: --alternate enable alternate macro syntax
1572 * D:: -D for compatibility
1573 * f:: -f to work faster
1574 * I:: -I for .include search path
1575 @ifclear DIFF-TBL-KLUGE
1576 * K:: -K for compatibility
1578 @ifset DIFF-TBL-KLUGE
1579 * K:: -K for difference tables
1582 * L:: -L to retain local symbols
1583 * listing:: --listing-XXX to configure listing output
1584 * M:: -M or --mri to assemble in MRI compatibility mode
1585 * MD:: --MD for dependency tracking
1586 * o:: -o to name the object file
1587 * R:: -R to join data and text sections
1588 * statistics:: --statistics to see statistics about assembly
1589 * traditional-format:: --traditional-format for compatible output
1590 * v:: -v to announce version
1591 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1592 * Z:: -Z to make object file even after errors
1596 @section Enable Listings: @option{-a[cdhlns]}
1605 @cindex listings, enabling
1606 @cindex assembly listings, enabling
1608 These options enable listing output from the assembler. By itself,
1609 @samp{-a} requests high-level, assembly, and symbols listing.
1610 You can use other letters to select specific options for the list:
1611 @samp{-ah} requests a high-level language listing,
1612 @samp{-al} requests an output-program assembly listing, and
1613 @samp{-as} requests a symbol table listing.
1614 High-level listings require that a compiler debugging option like
1615 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1618 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1619 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1620 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1621 omitted from the listing.
1623 Use the @samp{-ad} option to omit debugging directives from the
1626 Once you have specified one of these options, you can further control
1627 listing output and its appearance using the directives @code{.list},
1628 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1630 The @samp{-an} option turns off all forms processing.
1631 If you do not request listing output with one of the @samp{-a} options, the
1632 listing-control directives have no effect.
1634 The letters after @samp{-a} may be combined into one option,
1635 @emph{e.g.}, @samp{-aln}.
1637 Note if the assembler source is coming from the standard input (e.g.,
1639 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1640 is being used) then the listing will not contain any comments or preprocessor
1641 directives. This is because the listing code buffers input source lines from
1642 stdin only after they have been preprocessed by the assembler. This reduces
1643 memory usage and makes the code more efficient.
1646 @section @option{--alternate}
1649 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1652 @section @option{-D}
1655 This option has no effect whatsoever, but it is accepted to make it more
1656 likely that scripts written for other assemblers also work with
1657 @command{@value{AS}}.
1660 @section Work Faster: @option{-f}
1663 @cindex trusted compiler
1664 @cindex faster processing (@option{-f})
1665 @samp{-f} should only be used when assembling programs written by a
1666 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1667 and comment preprocessing on
1668 the input file(s) before assembling them. @xref{Preprocessing,
1672 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1673 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1678 @section @code{.include} Search Path: @option{-I} @var{path}
1680 @kindex -I @var{path}
1681 @cindex paths for @code{.include}
1682 @cindex search path for @code{.include}
1683 @cindex @code{include} directive search path
1684 Use this option to add a @var{path} to the list of directories
1685 @command{@value{AS}} searches for files specified in @code{.include}
1686 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1687 many times as necessary to include a variety of paths. The current
1688 working directory is always searched first; after that, @command{@value{AS}}
1689 searches any @samp{-I} directories in the same order as they were
1690 specified (left to right) on the command line.
1693 @section Difference Tables: @option{-K}
1696 @ifclear DIFF-TBL-KLUGE
1697 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1698 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1699 where it can be used to warn when the assembler alters the machine code
1700 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1701 family does not have the addressing limitations that sometimes lead to this
1702 alteration on other platforms.
1705 @ifset DIFF-TBL-KLUGE
1706 @cindex difference tables, warning
1707 @cindex warning for altered difference tables
1708 @command{@value{AS}} sometimes alters the code emitted for directives of the
1709 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
1710 You can use the @samp{-K} option if you want a warning issued when this
1715 @section Include Local Symbols: @option{-L}
1718 @cindex local symbols, retaining in output
1719 Symbols beginning with system-specific local label prefixes, typically
1720 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
1721 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
1722 such symbols when debugging, because they are intended for the use of
1723 programs (like compilers) that compose assembler programs, not for your
1724 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
1725 such symbols, so you do not normally debug with them.
1727 This option tells @command{@value{AS}} to retain those local symbols
1728 in the object file. Usually if you do this you also tell the linker
1729 @code{@value{LD}} to preserve those symbols.
1732 @section Configuring listing output: @option{--listing}
1734 The listing feature of the assembler can be enabled via the command line switch
1735 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1736 hex dump of the corresponding locations in the output object file, and displays
1737 them as a listing file. The format of this listing can be controlled by
1738 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
1739 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
1740 @code{.psize} (@pxref{Psize}), and
1741 @code{.eject} (@pxref{Eject}) and also by the following switches:
1744 @item --listing-lhs-width=@samp{number}
1745 @kindex --listing-lhs-width
1746 @cindex Width of first line disassembly output
1747 Sets the maximum width, in words, of the first line of the hex byte dump. This
1748 dump appears on the left hand side of the listing output.
1750 @item --listing-lhs-width2=@samp{number}
1751 @kindex --listing-lhs-width2
1752 @cindex Width of continuation lines of disassembly output
1753 Sets the maximum width, in words, of any further lines of the hex byte dump for
1754 a given input source line. If this value is not specified, it defaults to being
1755 the same as the value specified for @samp{--listing-lhs-width}. If neither
1756 switch is used the default is to one.
1758 @item --listing-rhs-width=@samp{number}
1759 @kindex --listing-rhs-width
1760 @cindex Width of source line output
1761 Sets the maximum width, in characters, of the source line that is displayed
1762 alongside the hex dump. The default value for this parameter is 100. The
1763 source line is displayed on the right hand side of the listing output.
1765 @item --listing-cont-lines=@samp{number}
1766 @kindex --listing-cont-lines
1767 @cindex Maximum number of continuation lines
1768 Sets the maximum number of continuation lines of hex dump that will be
1769 displayed for a given single line of source input. The default value is 4.
1773 @section Assemble in MRI Compatibility Mode: @option{-M}
1776 @cindex MRI compatibility mode
1777 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
1778 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
1779 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1780 configured target) assembler from Microtec Research. The exact nature of the
1781 MRI syntax will not be documented here; see the MRI manuals for more
1782 information. Note in particular that the handling of macros and macro
1783 arguments is somewhat different. The purpose of this option is to permit
1784 assembling existing MRI assembler code using @command{@value{AS}}.
1786 The MRI compatibility is not complete. Certain operations of the MRI assembler
1787 depend upon its object file format, and can not be supported using other object
1788 file formats. Supporting these would require enhancing each object file format
1789 individually. These are:
1792 @item global symbols in common section
1794 The m68k MRI assembler supports common sections which are merged by the linker.
1795 Other object file formats do not support this. @command{@value{AS}} handles
1796 common sections by treating them as a single common symbol. It permits local
1797 symbols to be defined within a common section, but it can not support global
1798 symbols, since it has no way to describe them.
1800 @item complex relocations
1802 The MRI assemblers support relocations against a negated section address, and
1803 relocations which combine the start addresses of two or more sections. These
1804 are not support by other object file formats.
1806 @item @code{END} pseudo-op specifying start address
1808 The MRI @code{END} pseudo-op permits the specification of a start address.
1809 This is not supported by other object file formats. The start address may
1810 instead be specified using the @option{-e} option to the linker, or in a linker
1813 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1815 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1816 name to the output file. This is not supported by other object file formats.
1818 @item @code{ORG} pseudo-op
1820 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1821 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
1822 which changes the location within the current section. Absolute sections are
1823 not supported by other object file formats. The address of a section may be
1824 assigned within a linker script.
1827 There are some other features of the MRI assembler which are not supported by
1828 @command{@value{AS}}, typically either because they are difficult or because they
1829 seem of little consequence. Some of these may be supported in future releases.
1833 @item EBCDIC strings
1835 EBCDIC strings are not supported.
1837 @item packed binary coded decimal
1839 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1840 and @code{DCB.P} pseudo-ops are not supported.
1842 @item @code{FEQU} pseudo-op
1844 The m68k @code{FEQU} pseudo-op is not supported.
1846 @item @code{NOOBJ} pseudo-op
1848 The m68k @code{NOOBJ} pseudo-op is not supported.
1850 @item @code{OPT} branch control options
1852 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1853 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
1854 relaxes all branches, whether forward or backward, to an appropriate size, so
1855 these options serve no purpose.
1857 @item @code{OPT} list control options
1859 The following m68k @code{OPT} list control options are ignored: @code{C},
1860 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1861 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1863 @item other @code{OPT} options
1865 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1866 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1868 @item @code{OPT} @code{D} option is default
1870 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1871 @code{OPT NOD} may be used to turn it off.
1873 @item @code{XREF} pseudo-op.
1875 The m68k @code{XREF} pseudo-op is ignored.
1877 @item @code{.debug} pseudo-op
1879 The i960 @code{.debug} pseudo-op is not supported.
1881 @item @code{.extended} pseudo-op
1883 The i960 @code{.extended} pseudo-op is not supported.
1885 @item @code{.list} pseudo-op.
1887 The various options of the i960 @code{.list} pseudo-op are not supported.
1889 @item @code{.optimize} pseudo-op
1891 The i960 @code{.optimize} pseudo-op is not supported.
1893 @item @code{.output} pseudo-op
1895 The i960 @code{.output} pseudo-op is not supported.
1897 @item @code{.setreal} pseudo-op
1899 The i960 @code{.setreal} pseudo-op is not supported.
1904 @section Dependency Tracking: @option{--MD}
1907 @cindex dependency tracking
1910 @command{@value{AS}} can generate a dependency file for the file it creates. This
1911 file consists of a single rule suitable for @code{make} describing the
1912 dependencies of the main source file.
1914 The rule is written to the file named in its argument.
1916 This feature is used in the automatic updating of makefiles.
1919 @section Name the Object File: @option{-o}
1922 @cindex naming object file
1923 @cindex object file name
1924 There is always one object file output when you run @command{@value{AS}}. By
1925 default it has the name
1928 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1942 You use this option (which takes exactly one filename) to give the
1943 object file a different name.
1945 Whatever the object file is called, @command{@value{AS}} overwrites any
1946 existing file of the same name.
1949 @section Join Data and Text Sections: @option{-R}
1952 @cindex data and text sections, joining
1953 @cindex text and data sections, joining
1954 @cindex joining text and data sections
1955 @cindex merging text and data sections
1956 @option{-R} tells @command{@value{AS}} to write the object file as if all
1957 data-section data lives in the text section. This is only done at
1958 the very last moment: your binary data are the same, but data
1959 section parts are relocated differently. The data section part of
1960 your object file is zero bytes long because all its bytes are
1961 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1963 When you specify @option{-R} it would be possible to generate shorter
1964 address displacements (because we do not have to cross between text and
1965 data section). We refrain from doing this simply for compatibility with
1966 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
1969 When @command{@value{AS}} is configured for COFF or ELF output,
1970 this option is only useful if you use sections named @samp{.text} and
1975 @option{-R} is not supported for any of the HPPA targets. Using
1976 @option{-R} generates a warning from @command{@value{AS}}.
1980 @section Display Assembly Statistics: @option{--statistics}
1982 @kindex --statistics
1983 @cindex statistics, about assembly
1984 @cindex time, total for assembly
1985 @cindex space used, maximum for assembly
1986 Use @samp{--statistics} to display two statistics about the resources used by
1987 @command{@value{AS}}: the maximum amount of space allocated during the assembly
1988 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1991 @node traditional-format
1992 @section Compatible Output: @option{--traditional-format}
1994 @kindex --traditional-format
1995 For some targets, the output of @command{@value{AS}} is different in some ways
1996 from the output of some existing assembler. This switch requests
1997 @command{@value{AS}} to use the traditional format instead.
1999 For example, it disables the exception frame optimizations which
2000 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2003 @section Announce Version: @option{-v}
2007 @cindex assembler version
2008 @cindex version of assembler
2009 You can find out what version of as is running by including the
2010 option @samp{-v} (which you can also spell as @samp{-version}) on the
2014 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2016 @command{@value{AS}} should never give a warning or error message when
2017 assembling compiler output. But programs written by people often
2018 cause @command{@value{AS}} to give a warning that a particular assumption was
2019 made. All such warnings are directed to the standard error file.
2023 @cindex suppressing warnings
2024 @cindex warnings, suppressing
2025 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2026 This only affects the warning messages: it does not change any particular of
2027 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2030 @kindex --fatal-warnings
2031 @cindex errors, caused by warnings
2032 @cindex warnings, causing error
2033 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2034 files that generate warnings to be in error.
2037 @cindex warnings, switching on
2038 You can switch these options off again by specifying @option{--warn}, which
2039 causes warnings to be output as usual.
2042 @section Generate Object File in Spite of Errors: @option{-Z}
2043 @cindex object file, after errors
2044 @cindex errors, continuing after
2045 After an error message, @command{@value{AS}} normally produces no output. If for
2046 some reason you are interested in object file output even after
2047 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2048 option. If there are any errors, @command{@value{AS}} continues anyways, and
2049 writes an object file after a final warning message of the form @samp{@var{n}
2050 errors, @var{m} warnings, generating bad object file.}
2055 @cindex machine-independent syntax
2056 @cindex syntax, machine-independent
2057 This chapter describes the machine-independent syntax allowed in a
2058 source file. @command{@value{AS}} syntax is similar to what many other
2059 assemblers use; it is inspired by the BSD 4.2
2064 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2068 * Preprocessing:: Preprocessing
2069 * Whitespace:: Whitespace
2070 * Comments:: Comments
2071 * Symbol Intro:: Symbols
2072 * Statements:: Statements
2073 * Constants:: Constants
2077 @section Preprocessing
2079 @cindex preprocessing
2080 The @command{@value{AS}} internal preprocessor:
2082 @cindex whitespace, removed by preprocessor
2084 adjusts and removes extra whitespace. It leaves one space or tab before
2085 the keywords on a line, and turns any other whitespace on the line into
2088 @cindex comments, removed by preprocessor
2090 removes all comments, replacing them with a single space, or an
2091 appropriate number of newlines.
2093 @cindex constants, converted by preprocessor
2095 converts character constants into the appropriate numeric values.
2098 It does not do macro processing, include file handling, or
2099 anything else you may get from your C compiler's preprocessor. You can
2100 do include file processing with the @code{.include} directive
2101 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2102 to get other ``CPP'' style preprocessing by giving the input file a
2103 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2104 Output, gcc.info, Using GNU CC}.
2106 Excess whitespace, comments, and character constants
2107 cannot be used in the portions of the input text that are not
2110 @cindex turning preprocessing on and off
2111 @cindex preprocessing, turning on and off
2114 If the first line of an input file is @code{#NO_APP} or if you use the
2115 @samp{-f} option, whitespace and comments are not removed from the input file.
2116 Within an input file, you can ask for whitespace and comment removal in
2117 specific portions of the by putting a line that says @code{#APP} before the
2118 text that may contain whitespace or comments, and putting a line that says
2119 @code{#NO_APP} after this text. This feature is mainly intend to support
2120 @code{asm} statements in compilers whose output is otherwise free of comments
2127 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2128 Whitespace is used to separate symbols, and to make programs neater for
2129 people to read. Unless within character constants
2130 (@pxref{Characters,,Character Constants}), any whitespace means the same
2131 as exactly one space.
2137 There are two ways of rendering comments to @command{@value{AS}}. In both
2138 cases the comment is equivalent to one space.
2140 Anything from @samp{/*} through the next @samp{*/} is a comment.
2141 This means you may not nest these comments.
2145 The only way to include a newline ('\n') in a comment
2146 is to use this sort of comment.
2149 /* This sort of comment does not nest. */
2152 @cindex line comment character
2153 Anything from the @dfn{line comment} character to the next newline
2154 is considered a comment and is ignored. The line comment character is
2156 @samp{;} on the ARC;
2159 @samp{@@} on the ARM;
2162 @samp{;} for the H8/300 family;
2165 @samp{;} for the HPPA;
2168 @samp{#} on the i386 and x86-64;
2171 @samp{#} on the i960;
2174 @samp{;} for the PDP-11;
2177 @samp{;} for picoJava;
2180 @samp{#} for Motorola PowerPC;
2183 @samp{!} for the Renesas / SuperH SH;
2186 @samp{!} on the SPARC;
2189 @samp{#} on the ip2k;
2192 @samp{#} on the m32c;
2195 @samp{#} on the m32r;
2198 @samp{|} on the 680x0;
2201 @samp{#} on the 68HC11 and 68HC12;
2204 @samp{#} on the Vax;
2207 @samp{;} for the Z80;
2210 @samp{!} for the Z8000;
2213 @samp{#} on the V850;
2216 @samp{#} for Xtensa systems;
2218 see @ref{Machine Dependencies}. @refill
2219 @c FIXME What about i860?
2222 On some machines there are two different line comment characters. One
2223 character only begins a comment if it is the first non-whitespace character on
2224 a line, while the other always begins a comment.
2228 The V850 assembler also supports a double dash as starting a comment that
2229 extends to the end of the line.
2235 @cindex lines starting with @code{#}
2236 @cindex logical line numbers
2237 To be compatible with past assemblers, lines that begin with @samp{#} have a
2238 special interpretation. Following the @samp{#} should be an absolute
2239 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2240 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2241 new logical file name. The rest of the line, if any, should be whitespace.
2243 If the first non-whitespace characters on the line are not numeric,
2244 the line is ignored. (Just like a comment.)
2247 # This is an ordinary comment.
2248 # 42-6 "new_file_name" # New logical file name
2249 # This is logical line # 36.
2251 This feature is deprecated, and may disappear from future versions
2252 of @command{@value{AS}}.
2257 @cindex characters used in symbols
2258 @ifclear SPECIAL-SYMS
2259 A @dfn{symbol} is one or more characters chosen from the set of all
2260 letters (both upper and lower case), digits and the three characters
2266 A @dfn{symbol} is one or more characters chosen from the set of all
2267 letters (both upper and lower case), digits and the three characters
2268 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2274 On most machines, you can also use @code{$} in symbol names; exceptions
2275 are noted in @ref{Machine Dependencies}.
2277 No symbol may begin with a digit. Case is significant.
2278 There is no length limit: all characters are significant. Symbols are
2279 delimited by characters not in that set, or by the beginning of a file
2280 (since the source program must end with a newline, the end of a file is
2281 not a possible symbol delimiter). @xref{Symbols}.
2282 @cindex length of symbols
2287 @cindex statements, structure of
2288 @cindex line separator character
2289 @cindex statement separator character
2291 @ifclear abnormal-separator
2292 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
2293 semicolon (@samp{;}). The newline or semicolon is considered part of
2294 the preceding statement. Newlines and semicolons within character
2295 constants are an exception: they do not end statements.
2297 @ifset abnormal-separator
2299 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
2300 point (@samp{!}). The newline or exclamation point is considered part of the
2301 preceding statement. Newlines and exclamation points within character
2302 constants are an exception: they do not end statements.
2305 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
2306 H8/300) a dollar sign (@samp{$}); or (for the Renesas-SH) a semicolon
2307 (@samp{;}). The newline or separator character is considered part of
2308 the preceding statement. Newlines and separators within character
2309 constants are an exception: they do not end statements.
2314 A @dfn{statement} ends at a newline character (@samp{\n}) or line
2315 separator character. (The line separator is usually @samp{;}, unless this
2316 conflicts with the comment character; see @ref{Machine Dependencies}.) The
2317 newline or separator character is considered part of the preceding
2318 statement. Newlines and separators within character constants are an
2319 exception: they do not end statements.
2322 @cindex newline, required at file end
2323 @cindex EOF, newline must precede
2324 It is an error to end any statement with end-of-file: the last
2325 character of any input file should be a newline.@refill
2327 An empty statement is allowed, and may include whitespace. It is ignored.
2329 @cindex instructions and directives
2330 @cindex directives and instructions
2331 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2332 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2334 A statement begins with zero or more labels, optionally followed by a
2335 key symbol which determines what kind of statement it is. The key
2336 symbol determines the syntax of the rest of the statement. If the
2337 symbol begins with a dot @samp{.} then the statement is an assembler
2338 directive: typically valid for any computer. If the symbol begins with
2339 a letter the statement is an assembly language @dfn{instruction}: it
2340 assembles into a machine language instruction.
2342 Different versions of @command{@value{AS}} for different computers
2343 recognize different instructions. In fact, the same symbol may
2344 represent a different instruction in a different computer's assembly
2348 @cindex @code{:} (label)
2349 @cindex label (@code{:})
2350 A label is a symbol immediately followed by a colon (@code{:}).
2351 Whitespace before a label or after a colon is permitted, but you may not
2352 have whitespace between a label's symbol and its colon. @xref{Labels}.
2355 For HPPA targets, labels need not be immediately followed by a colon, but
2356 the definition of a label must begin in column zero. This also implies that
2357 only one label may be defined on each line.
2361 label: .directive followed by something
2362 another_label: # This is an empty statement.
2363 instruction operand_1, operand_2, @dots{}
2370 A constant is a number, written so that its value is known by
2371 inspection, without knowing any context. Like this:
2374 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2375 .ascii "Ring the bell\7" # A string constant.
2376 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2377 .float 0f-314159265358979323846264338327\
2378 95028841971.693993751E-40 # - pi, a flonum.
2383 * Characters:: Character Constants
2384 * Numbers:: Number Constants
2388 @subsection Character Constants
2390 @cindex character constants
2391 @cindex constants, character
2392 There are two kinds of character constants. A @dfn{character} stands
2393 for one character in one byte and its value may be used in
2394 numeric expressions. String constants (properly called string
2395 @emph{literals}) are potentially many bytes and their values may not be
2396 used in arithmetic expressions.
2400 * Chars:: Characters
2404 @subsubsection Strings
2406 @cindex string constants
2407 @cindex constants, string
2408 A @dfn{string} is written between double-quotes. It may contain
2409 double-quotes or null characters. The way to get special characters
2410 into a string is to @dfn{escape} these characters: precede them with
2411 a backslash @samp{\} character. For example @samp{\\} represents
2412 one backslash: the first @code{\} is an escape which tells
2413 @command{@value{AS}} to interpret the second character literally as a backslash
2414 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2415 escape character). The complete list of escapes follows.
2417 @cindex escape codes, character
2418 @cindex character escape codes
2421 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2423 @cindex @code{\b} (backspace character)
2424 @cindex backspace (@code{\b})
2426 Mnemonic for backspace; for ASCII this is octal code 010.
2429 @c Mnemonic for EOText; for ASCII this is octal code 004.
2431 @cindex @code{\f} (formfeed character)
2432 @cindex formfeed (@code{\f})
2434 Mnemonic for FormFeed; for ASCII this is octal code 014.
2436 @cindex @code{\n} (newline character)
2437 @cindex newline (@code{\n})
2439 Mnemonic for newline; for ASCII this is octal code 012.
2442 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2444 @cindex @code{\r} (carriage return character)
2445 @cindex carriage return (@code{\r})
2447 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2450 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2451 @c other assemblers.
2453 @cindex @code{\t} (tab)
2454 @cindex tab (@code{\t})
2456 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2459 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2460 @c @item \x @var{digit} @var{digit} @var{digit}
2461 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2463 @cindex @code{\@var{ddd}} (octal character code)
2464 @cindex octal character code (@code{\@var{ddd}})
2465 @item \ @var{digit} @var{digit} @var{digit}
2466 An octal character code. The numeric code is 3 octal digits.
2467 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2468 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2470 @cindex @code{\@var{xd...}} (hex character code)
2471 @cindex hex character code (@code{\@var{xd...}})
2472 @item \@code{x} @var{hex-digits...}
2473 A hex character code. All trailing hex digits are combined. Either upper or
2474 lower case @code{x} works.
2476 @cindex @code{\\} (@samp{\} character)
2477 @cindex backslash (@code{\\})
2479 Represents one @samp{\} character.
2482 @c Represents one @samp{'} (accent acute) character.
2483 @c This is needed in single character literals
2484 @c (@xref{Characters,,Character Constants}.) to represent
2487 @cindex @code{\"} (doublequote character)
2488 @cindex doublequote (@code{\"})
2490 Represents one @samp{"} character. Needed in strings to represent
2491 this character, because an unescaped @samp{"} would end the string.
2493 @item \ @var{anything-else}
2494 Any other character when escaped by @kbd{\} gives a warning, but
2495 assembles as if the @samp{\} was not present. The idea is that if
2496 you used an escape sequence you clearly didn't want the literal
2497 interpretation of the following character. However @command{@value{AS}} has no
2498 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2499 code and warns you of the fact.
2502 Which characters are escapable, and what those escapes represent,
2503 varies widely among assemblers. The current set is what we think
2504 the BSD 4.2 assembler recognizes, and is a subset of what most C
2505 compilers recognize. If you are in doubt, do not use an escape
2509 @subsubsection Characters
2511 @cindex single character constant
2512 @cindex character, single
2513 @cindex constant, single character
2514 A single character may be written as a single quote immediately
2515 followed by that character. The same escapes apply to characters as
2516 to strings. So if you want to write the character backslash, you
2517 must write @kbd{'\\} where the first @code{\} escapes the second
2518 @code{\}. As you can see, the quote is an acute accent, not a
2519 grave accent. A newline
2521 @ifclear abnormal-separator
2522 (or semicolon @samp{;})
2524 @ifset abnormal-separator
2526 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2531 immediately following an acute accent is taken as a literal character
2532 and does not count as the end of a statement. The value of a character
2533 constant in a numeric expression is the machine's byte-wide code for
2534 that character. @command{@value{AS}} assumes your character code is ASCII:
2535 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2538 @subsection Number Constants
2540 @cindex constants, number
2541 @cindex number constants
2542 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2543 are stored in the target machine. @emph{Integers} are numbers that
2544 would fit into an @code{int} in the C language. @emph{Bignums} are
2545 integers, but they are stored in more than 32 bits. @emph{Flonums}
2546 are floating point numbers, described below.
2549 * Integers:: Integers
2554 * Bit Fields:: Bit Fields
2560 @subsubsection Integers
2562 @cindex constants, integer
2564 @cindex binary integers
2565 @cindex integers, binary
2566 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2567 the binary digits @samp{01}.
2569 @cindex octal integers
2570 @cindex integers, octal
2571 An octal integer is @samp{0} followed by zero or more of the octal
2572 digits (@samp{01234567}).
2574 @cindex decimal integers
2575 @cindex integers, decimal
2576 A decimal integer starts with a non-zero digit followed by zero or
2577 more digits (@samp{0123456789}).
2579 @cindex hexadecimal integers
2580 @cindex integers, hexadecimal
2581 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2582 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2584 Integers have the usual values. To denote a negative integer, use
2585 the prefix operator @samp{-} discussed under expressions
2586 (@pxref{Prefix Ops,,Prefix Operators}).
2589 @subsubsection Bignums
2592 @cindex constants, bignum
2593 A @dfn{bignum} has the same syntax and semantics as an integer
2594 except that the number (or its negative) takes more than 32 bits to
2595 represent in binary. The distinction is made because in some places
2596 integers are permitted while bignums are not.
2599 @subsubsection Flonums
2601 @cindex floating point numbers
2602 @cindex constants, floating point
2604 @cindex precision, floating point
2605 A @dfn{flonum} represents a floating point number. The translation is
2606 indirect: a decimal floating point number from the text is converted by
2607 @command{@value{AS}} to a generic binary floating point number of more than
2608 sufficient precision. This generic floating point number is converted
2609 to a particular computer's floating point format (or formats) by a
2610 portion of @command{@value{AS}} specialized to that computer.
2612 A flonum is written by writing (in order)
2617 (@samp{0} is optional on the HPPA.)
2621 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2623 @kbd{e} is recommended. Case is not important.
2625 @c FIXME: verify if flonum syntax really this vague for most cases
2626 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2627 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2630 On the H8/300, Renesas / SuperH SH,
2631 and AMD 29K architectures, the letter must be
2632 one of the letters @samp{DFPRSX} (in upper or lower case).
2634 On the ARC, the letter must be one of the letters @samp{DFRS}
2635 (in upper or lower case).
2637 On the Intel 960 architecture, the letter must be
2638 one of the letters @samp{DFT} (in upper or lower case).
2640 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2644 One of the letters @samp{DFRS} (in upper or lower case).
2647 One of the letters @samp{DFPRSX} (in upper or lower case).
2650 The letter @samp{E} (upper case only).
2653 One of the letters @samp{DFT} (in upper or lower case).
2658 An optional sign: either @samp{+} or @samp{-}.
2661 An optional @dfn{integer part}: zero or more decimal digits.
2664 An optional @dfn{fractional part}: @samp{.} followed by zero
2665 or more decimal digits.
2668 An optional exponent, consisting of:
2672 An @samp{E} or @samp{e}.
2673 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2674 @c principle this can perfectly well be different on different targets.
2676 Optional sign: either @samp{+} or @samp{-}.
2678 One or more decimal digits.
2683 At least one of the integer part or the fractional part must be
2684 present. The floating point number has the usual base-10 value.
2686 @command{@value{AS}} does all processing using integers. Flonums are computed
2687 independently of any floating point hardware in the computer running
2688 @command{@value{AS}}.
2692 @c Bit fields are written as a general facility but are also controlled
2693 @c by a conditional-compilation flag---which is as of now (21mar91)
2694 @c turned on only by the i960 config of GAS.
2696 @subsubsection Bit Fields
2699 @cindex constants, bit field
2700 You can also define numeric constants as @dfn{bit fields}.
2701 Specify two numbers separated by a colon---
2703 @var{mask}:@var{value}
2706 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2709 The resulting number is then packed
2711 @c this conditional paren in case bit fields turned on elsewhere than 960
2712 (in host-dependent byte order)
2714 into a field whose width depends on which assembler directive has the
2715 bit-field as its argument. Overflow (a result from the bitwise and
2716 requiring more binary digits to represent) is not an error; instead,
2717 more constants are generated, of the specified width, beginning with the
2718 least significant digits.@refill
2720 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2721 @code{.short}, and @code{.word} accept bit-field arguments.
2726 @chapter Sections and Relocation
2731 * Secs Background:: Background
2732 * Ld Sections:: Linker Sections
2733 * As Sections:: Assembler Internal Sections
2734 * Sub-Sections:: Sub-Sections
2738 @node Secs Background
2741 Roughly, a section is a range of addresses, with no gaps; all data
2742 ``in'' those addresses is treated the same for some particular purpose.
2743 For example there may be a ``read only'' section.
2745 @cindex linker, and assembler
2746 @cindex assembler, and linker
2747 The linker @code{@value{LD}} reads many object files (partial programs) and
2748 combines their contents to form a runnable program. When @command{@value{AS}}
2749 emits an object file, the partial program is assumed to start at address 0.
2750 @code{@value{LD}} assigns the final addresses for the partial program, so that
2751 different partial programs do not overlap. This is actually an
2752 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2755 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2756 addresses. These blocks slide to their run-time addresses as rigid
2757 units; their length does not change and neither does the order of bytes
2758 within them. Such a rigid unit is called a @emph{section}. Assigning
2759 run-time addresses to sections is called @dfn{relocation}. It includes
2760 the task of adjusting mentions of object-file addresses so they refer to
2761 the proper run-time addresses.
2763 For the H8/300, and for the Renesas / SuperH SH,
2764 @command{@value{AS}} pads sections if needed to
2765 ensure they end on a word (sixteen bit) boundary.
2768 @cindex standard assembler sections
2769 An object file written by @command{@value{AS}} has at least three sections, any
2770 of which may be empty. These are named @dfn{text}, @dfn{data} and
2775 When it generates COFF or ELF output,
2777 @command{@value{AS}} can also generate whatever other named sections you specify
2778 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2779 If you do not use any directives that place output in the @samp{.text}
2780 or @samp{.data} sections, these sections still exist, but are empty.
2785 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2787 @command{@value{AS}} can also generate whatever other named sections you
2788 specify using the @samp{.space} and @samp{.subspace} directives. See
2789 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2790 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2791 assembler directives.
2794 Additionally, @command{@value{AS}} uses different names for the standard
2795 text, data, and bss sections when generating SOM output. Program text
2796 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2797 BSS into @samp{$BSS$}.
2801 Within the object file, the text section starts at address @code{0}, the
2802 data section follows, and the bss section follows the data section.
2805 When generating either SOM or ELF output files on the HPPA, the text
2806 section starts at address @code{0}, the data section at address
2807 @code{0x4000000}, and the bss section follows the data section.
2810 To let @code{@value{LD}} know which data changes when the sections are
2811 relocated, and how to change that data, @command{@value{AS}} also writes to the
2812 object file details of the relocation needed. To perform relocation
2813 @code{@value{LD}} must know, each time an address in the object
2817 Where in the object file is the beginning of this reference to
2820 How long (in bytes) is this reference?
2822 Which section does the address refer to? What is the numeric value of
2824 (@var{address}) @minus{} (@var{start-address of section})?
2827 Is the reference to an address ``Program-Counter relative''?
2830 @cindex addresses, format of
2831 @cindex section-relative addressing
2832 In fact, every address @command{@value{AS}} ever uses is expressed as
2834 (@var{section}) + (@var{offset into section})
2837 Further, most expressions @command{@value{AS}} computes have this section-relative
2840 (For some object formats, such as SOM for the HPPA, some expressions are
2841 symbol-relative instead.)
2844 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2845 @var{N} into section @var{secname}.''
2847 Apart from text, data and bss sections you need to know about the
2848 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2849 addresses in the absolute section remain unchanged. For example, address
2850 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2851 @code{@value{LD}}. Although the linker never arranges two partial programs'
2852 data sections with overlapping addresses after linking, @emph{by definition}
2853 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2854 part of a program is always the same address when the program is running as
2855 address @code{@{absolute@ 239@}} in any other part of the program.
2857 The idea of sections is extended to the @dfn{undefined} section. Any
2858 address whose section is unknown at assembly time is by definition
2859 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2860 Since numbers are always defined, the only way to generate an undefined
2861 address is to mention an undefined symbol. A reference to a named
2862 common block would be such a symbol: its value is unknown at assembly
2863 time so it has section @emph{undefined}.
2865 By analogy the word @emph{section} is used to describe groups of sections in
2866 the linked program. @code{@value{LD}} puts all partial programs' text
2867 sections in contiguous addresses in the linked program. It is
2868 customary to refer to the @emph{text section} of a program, meaning all
2869 the addresses of all partial programs' text sections. Likewise for
2870 data and bss sections.
2872 Some sections are manipulated by @code{@value{LD}}; others are invented for
2873 use of @command{@value{AS}} and have no meaning except during assembly.
2876 @section Linker Sections
2877 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2882 @cindex named sections
2883 @cindex sections, named
2884 @item named sections
2887 @cindex text section
2888 @cindex data section
2892 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
2893 separate but equal sections. Anything you can say of one section is
2896 When the program is running, however, it is
2897 customary for the text section to be unalterable. The
2898 text section is often shared among processes: it contains
2899 instructions, constants and the like. The data section of a running
2900 program is usually alterable: for example, C variables would be stored
2901 in the data section.
2906 This section contains zeroed bytes when your program begins running. It
2907 is used to hold uninitialized variables or common storage. The length of
2908 each partial program's bss section is important, but because it starts
2909 out containing zeroed bytes there is no need to store explicit zero
2910 bytes in the object file. The bss section was invented to eliminate
2911 those explicit zeros from object files.
2913 @cindex absolute section
2914 @item absolute section
2915 Address 0 of this section is always ``relocated'' to runtime address 0.
2916 This is useful if you want to refer to an address that @code{@value{LD}} must
2917 not change when relocating. In this sense we speak of absolute
2918 addresses being ``unrelocatable'': they do not change during relocation.
2920 @cindex undefined section
2921 @item undefined section
2922 This ``section'' is a catch-all for address references to objects not in
2923 the preceding sections.
2924 @c FIXME: ref to some other doc on obj-file formats could go here.
2927 @cindex relocation example
2928 An idealized example of three relocatable sections follows.
2930 The example uses the traditional section names @samp{.text} and @samp{.data}.
2932 Memory addresses are on the horizontal axis.
2936 @c END TEXI2ROFF-KILL
2939 partial program # 1: |ttttt|dddd|00|
2946 partial program # 2: |TTT|DDD|000|
2949 +--+---+-----+--+----+---+-----+~~
2950 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2951 +--+---+-----+--+----+---+-----+~~
2953 addresses: 0 @dots{}
2960 \line{\it Partial program \#1: \hfil}
2961 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2962 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2964 \line{\it Partial program \#2: \hfil}
2965 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2966 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2968 \line{\it linked program: \hfil}
2969 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2970 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2971 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2972 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2974 \line{\it addresses: \hfil}
2978 @c END TEXI2ROFF-KILL
2981 @section Assembler Internal Sections
2983 @cindex internal assembler sections
2984 @cindex sections in messages, internal
2985 These sections are meant only for the internal use of @command{@value{AS}}. They
2986 have no meaning at run-time. You do not really need to know about these
2987 sections for most purposes; but they can be mentioned in @command{@value{AS}}
2988 warning messages, so it might be helpful to have an idea of their
2989 meanings to @command{@value{AS}}. These sections are used to permit the
2990 value of every expression in your assembly language program to be a
2991 section-relative address.
2994 @cindex assembler internal logic error
2995 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2996 An internal assembler logic error has been found. This means there is a
2997 bug in the assembler.
2999 @cindex expr (internal section)
3001 The assembler stores complex expression internally as combinations of
3002 symbols. When it needs to represent an expression as a symbol, it puts
3003 it in the expr section.
3005 @c FIXME item transfer[t] vector preload
3006 @c FIXME item transfer[t] vector postload
3007 @c FIXME item register
3011 @section Sub-Sections
3013 @cindex numbered subsections
3014 @cindex grouping data
3020 fall into two sections: text and data.
3022 You may have separate groups of
3024 data in named sections
3028 data in named sections
3034 that you want to end up near to each other in the object file, even though they
3035 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3036 use @dfn{subsections} for this purpose. Within each section, there can be
3037 numbered subsections with values from 0 to 8192. Objects assembled into the
3038 same subsection go into the object file together with other objects in the same
3039 subsection. For example, a compiler might want to store constants in the text
3040 section, but might not want to have them interspersed with the program being
3041 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3042 section of code being output, and a @samp{.text 1} before each group of
3043 constants being output.
3045 Subsections are optional. If you do not use subsections, everything
3046 goes in subsection number zero.
3049 Each subsection is zero-padded up to a multiple of four bytes.
3050 (Subsections may be padded a different amount on different flavors
3051 of @command{@value{AS}}.)
3055 On the H8/300 platform, each subsection is zero-padded to a word
3056 boundary (two bytes).
3057 The same is true on the Renesas SH.
3060 @c FIXME section padding (alignment)?
3061 @c Rich Pixley says padding here depends on target obj code format; that
3062 @c doesn't seem particularly useful to say without further elaboration,
3063 @c so for now I say nothing about it. If this is a generic BFD issue,
3064 @c these paragraphs might need to vanish from this manual, and be
3065 @c discussed in BFD chapter of binutils (or some such).
3069 Subsections appear in your object file in numeric order, lowest numbered
3070 to highest. (All this to be compatible with other people's assemblers.)
3071 The object file contains no representation of subsections; @code{@value{LD}} and
3072 other programs that manipulate object files see no trace of them.
3073 They just see all your text subsections as a text section, and all your
3074 data subsections as a data section.
3076 To specify which subsection you want subsequent statements assembled
3077 into, use a numeric argument to specify it, in a @samp{.text
3078 @var{expression}} or a @samp{.data @var{expression}} statement.
3081 When generating COFF output, you
3086 can also use an extra subsection
3087 argument with arbitrary named sections: @samp{.section @var{name},
3092 When generating ELF output, you
3097 can also use the @code{.subsection} directive (@pxref{SubSection})
3098 to specify a subsection: @samp{.subsection @var{expression}}.
3100 @var{Expression} should be an absolute expression
3101 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3102 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3103 begins in @code{text 0}. For instance:
3105 .text 0 # The default subsection is text 0 anyway.
3106 .ascii "This lives in the first text subsection. *"
3108 .ascii "But this lives in the second text subsection."
3110 .ascii "This lives in the data section,"
3111 .ascii "in the first data subsection."
3113 .ascii "This lives in the first text section,"
3114 .ascii "immediately following the asterisk (*)."
3117 Each section has a @dfn{location counter} incremented by one for every byte
3118 assembled into that section. Because subsections are merely a convenience
3119 restricted to @command{@value{AS}} there is no concept of a subsection location
3120 counter. There is no way to directly manipulate a location counter---but the
3121 @code{.align} directive changes it, and any label definition captures its
3122 current value. The location counter of the section where statements are being
3123 assembled is said to be the @dfn{active} location counter.
3126 @section bss Section
3129 @cindex common variable storage
3130 The bss section is used for local common variable storage.
3131 You may allocate address space in the bss section, but you may
3132 not dictate data to load into it before your program executes. When
3133 your program starts running, all the contents of the bss
3134 section are zeroed bytes.
3136 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3137 @ref{Lcomm,,@code{.lcomm}}.
3139 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3140 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3143 When assembling for a target which supports multiple sections, such as ELF or
3144 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3145 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3146 section. Typically the section will only contain symbol definitions and
3147 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3154 Symbols are a central concept: the programmer uses symbols to name
3155 things, the linker uses symbols to link, and the debugger uses symbols
3159 @cindex debuggers, and symbol order
3160 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3161 the same order they were declared. This may break some debuggers.
3166 * Setting Symbols:: Giving Symbols Other Values
3167 * Symbol Names:: Symbol Names
3168 * Dot:: The Special Dot Symbol
3169 * Symbol Attributes:: Symbol Attributes
3176 A @dfn{label} is written as a symbol immediately followed by a colon
3177 @samp{:}. The symbol then represents the current value of the
3178 active location counter, and is, for example, a suitable instruction
3179 operand. You are warned if you use the same symbol to represent two
3180 different locations: the first definition overrides any other
3184 On the HPPA, the usual form for a label need not be immediately followed by a
3185 colon, but instead must start in column zero. Only one label may be defined on
3186 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3187 provides a special directive @code{.label} for defining labels more flexibly.
3190 @node Setting Symbols
3191 @section Giving Symbols Other Values
3193 @cindex assigning values to symbols
3194 @cindex symbol values, assigning
3195 A symbol can be given an arbitrary value by writing a symbol, followed
3196 by an equals sign @samp{=}, followed by an expression
3197 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3198 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3199 equals sign @samp{=}@samp{=} here represents an equivalent of the
3200 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3203 @section Symbol Names
3205 @cindex symbol names
3206 @cindex names, symbol
3207 @ifclear SPECIAL-SYMS
3208 Symbol names begin with a letter or with one of @samp{._}. On most
3209 machines, you can also use @code{$} in symbol names; exceptions are
3210 noted in @ref{Machine Dependencies}. That character may be followed by any
3211 string of digits, letters, dollar signs (unless otherwise noted for a
3212 particular target machine), and underscores.
3216 Symbol names begin with a letter or with one of @samp{._}. On the
3217 Renesas SH you can also use @code{$} in symbol names. That
3218 character may be followed by any string of digits, letters, dollar signs (save
3219 on the H8/300), and underscores.
3223 Case of letters is significant: @code{foo} is a different symbol name
3226 Each symbol has exactly one name. Each name in an assembly language program
3227 refers to exactly one symbol. You may use that symbol name any number of times
3230 @subheading Local Symbol Names
3232 @cindex local symbol names
3233 @cindex symbol names, local
3234 A local symbol is any symbol beginning with certain local label prefixes.
3235 By default, the local label prefix is @samp{.L} for ELF systems or
3236 @samp{L} for traditional a.out systems, but each target may have its own
3237 set of local label prefixes.
3239 On the HPPA local symbols begin with @samp{L$}.
3242 Local symbols are defined and used within the assembler, but they are
3243 normally not saved in object files. Thus, they are not visible when debugging.
3244 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols:
3245 @option{-L}}) to retain the local symbols in the object files.
3247 @subheading Local Labels
3249 @cindex local labels
3250 @cindex temporary symbol names
3251 @cindex symbol names, temporary
3252 Local labels help compilers and programmers use names temporarily.
3253 They create symbols which are guaranteed to be unique over the entire scope of
3254 the input source code and which can be referred to by a simple notation.
3255 To define a local label, write a label of the form @samp{@b{N}:} (where @b{N}
3256 represents any positive integer). To refer to the most recent previous
3257 definition of that label write @samp{@b{N}b}, using the same number as when
3258 you defined the label. To refer to the next definition of a local label, write
3259 @samp{@b{N}f}---the @samp{b} stands for ``backwards'' and the @samp{f} stands
3262 There is no restriction on how you can use these labels, and you can reuse them
3263 too. So that it is possible to repeatedly define the same local label (using
3264 the same number @samp{@b{N}}), although you can only refer to the most recently
3265 defined local label of that number (for a backwards reference) or the next
3266 definition of a specific local label for a forward reference. It is also worth
3267 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3268 implemented in a slightly more efficient manner than the others.
3279 Which is the equivalent of:
3282 label_1: branch label_3
3283 label_2: branch label_1
3284 label_3: branch label_4
3285 label_4: branch label_3
3288 Local label names are only a notational device. They are immediately
3289 transformed into more conventional symbol names before the assembler uses them.
3290 The symbol names are stored in the symbol table, appear in error messages, and
3291 are optionally emitted to the object file. The names are constructed using
3295 @item @emph{local label prefix}
3296 All local symbols begin with the system-specific local label prefix.
3297 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3298 that start with the local label prefix. These labels are
3299 used for symbols you are never intended to see. If you use the
3300 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3301 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3302 you may use them in debugging.
3305 This is the number that was used in the local label definition. So if the
3306 label is written @samp{55:} then the number is @samp{55}.
3309 This unusual character is included so you do not accidentally invent a symbol
3310 of the same name. The character has ASCII value of @samp{\002} (control-B).
3312 @item @emph{ordinal number}
3313 This is a serial number to keep the labels distinct. The first definition of
3314 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3315 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3316 the number @samp{1} and its 15th definition gets @samp{15} as well.
3319 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3320 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3322 @subheading Dollar Local Labels
3323 @cindex dollar local symbols
3325 @code{@value{AS}} also supports an even more local form of local labels called
3326 dollar labels. These labels go out of scope (i.e., they become undefined) as
3327 soon as a non-local label is defined. Thus they remain valid for only a small
3328 region of the input source code. Normal local labels, by contrast, remain in
3329 scope for the entire file, or until they are redefined by another occurrence of
3330 the same local label.
3332 Dollar labels are defined in exactly the same way as ordinary local labels,
3333 except that instead of being terminated by a colon, they are terminated by a
3334 dollar sign, e.g., @samp{@b{55$}}.
3336 They can also be distinguished from ordinary local labels by their transformed
3337 names which use ASCII character @samp{\001} (control-A) as the magic character
3338 to distinguish them from ordinary labels. For example, the fifth definition of
3339 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3342 @section The Special Dot Symbol
3344 @cindex dot (symbol)
3345 @cindex @code{.} (symbol)
3346 @cindex current address
3347 @cindex location counter
3348 The special symbol @samp{.} refers to the current address that
3349 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3350 .long .} defines @code{melvin} to contain its own address.
3351 Assigning a value to @code{.} is treated the same as a @code{.org}
3352 directive. Thus, the expression @samp{.=.+4} is the same as saying
3353 @ifclear no-space-dir
3357 @node Symbol Attributes
3358 @section Symbol Attributes
3360 @cindex symbol attributes
3361 @cindex attributes, symbol
3362 Every symbol has, as well as its name, the attributes ``Value'' and
3363 ``Type''. Depending on output format, symbols can also have auxiliary
3366 The detailed definitions are in @file{a.out.h}.
3369 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3370 all these attributes, and probably won't warn you. This makes the
3371 symbol an externally defined symbol, which is generally what you
3375 * Symbol Value:: Value
3376 * Symbol Type:: Type
3379 * a.out Symbols:: Symbol Attributes: @code{a.out}
3383 * a.out Symbols:: Symbol Attributes: @code{a.out}
3386 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3391 * COFF Symbols:: Symbol Attributes for COFF
3394 * SOM Symbols:: Symbol Attributes for SOM
3401 @cindex value of a symbol
3402 @cindex symbol value
3403 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3404 location in the text, data, bss or absolute sections the value is the
3405 number of addresses from the start of that section to the label.
3406 Naturally for text, data and bss sections the value of a symbol changes
3407 as @code{@value{LD}} changes section base addresses during linking. Absolute
3408 symbols' values do not change during linking: that is why they are
3411 The value of an undefined symbol is treated in a special way. If it is
3412 0 then the symbol is not defined in this assembler source file, and
3413 @code{@value{LD}} tries to determine its value from other files linked into the
3414 same program. You make this kind of symbol simply by mentioning a symbol
3415 name without defining it. A non-zero value represents a @code{.comm}
3416 common declaration. The value is how much common storage to reserve, in
3417 bytes (addresses). The symbol refers to the first address of the
3423 @cindex type of a symbol
3425 The type attribute of a symbol contains relocation (section)
3426 information, any flag settings indicating that a symbol is external, and
3427 (optionally), other information for linkers and debuggers. The exact
3428 format depends on the object-code output format in use.
3433 @c The following avoids a "widow" subsection title. @group would be
3434 @c better if it were available outside examples.
3437 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3439 @cindex @code{b.out} symbol attributes
3440 @cindex symbol attributes, @code{b.out}
3441 These symbol attributes appear only when @command{@value{AS}} is configured for
3442 one of the Berkeley-descended object output formats---@code{a.out} or
3448 @subsection Symbol Attributes: @code{a.out}
3450 @cindex @code{a.out} symbol attributes
3451 @cindex symbol attributes, @code{a.out}
3457 @subsection Symbol Attributes: @code{a.out}
3459 @cindex @code{a.out} symbol attributes
3460 @cindex symbol attributes, @code{a.out}
3464 * Symbol Desc:: Descriptor
3465 * Symbol Other:: Other
3469 @subsubsection Descriptor
3471 @cindex descriptor, of @code{a.out} symbol
3472 This is an arbitrary 16-bit value. You may establish a symbol's
3473 descriptor value by using a @code{.desc} statement
3474 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3475 @command{@value{AS}}.
3478 @subsubsection Other
3480 @cindex other attribute, of @code{a.out} symbol
3481 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3486 @subsection Symbol Attributes for COFF
3488 @cindex COFF symbol attributes
3489 @cindex symbol attributes, COFF
3491 The COFF format supports a multitude of auxiliary symbol attributes;
3492 like the primary symbol attributes, they are set between @code{.def} and
3493 @code{.endef} directives.
3495 @subsubsection Primary Attributes
3497 @cindex primary attributes, COFF symbols
3498 The symbol name is set with @code{.def}; the value and type,
3499 respectively, with @code{.val} and @code{.type}.
3501 @subsubsection Auxiliary Attributes
3503 @cindex auxiliary attributes, COFF symbols
3504 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3505 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3506 table information for COFF.
3511 @subsection Symbol Attributes for SOM
3513 @cindex SOM symbol attributes
3514 @cindex symbol attributes, SOM
3516 The SOM format for the HPPA supports a multitude of symbol attributes set with
3517 the @code{.EXPORT} and @code{.IMPORT} directives.
3519 The attributes are described in @cite{HP9000 Series 800 Assembly
3520 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3521 @code{EXPORT} assembler directive documentation.
3525 @chapter Expressions
3529 @cindex numeric values
3530 An @dfn{expression} specifies an address or numeric value.
3531 Whitespace may precede and/or follow an expression.
3533 The result of an expression must be an absolute number, or else an offset into
3534 a particular section. If an expression is not absolute, and there is not
3535 enough information when @command{@value{AS}} sees the expression to know its
3536 section, a second pass over the source program might be necessary to interpret
3537 the expression---but the second pass is currently not implemented.
3538 @command{@value{AS}} aborts with an error message in this situation.
3541 * Empty Exprs:: Empty Expressions
3542 * Integer Exprs:: Integer Expressions
3546 @section Empty Expressions
3548 @cindex empty expressions
3549 @cindex expressions, empty
3550 An empty expression has no value: it is just whitespace or null.
3551 Wherever an absolute expression is required, you may omit the
3552 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3553 is compatible with other assemblers.
3556 @section Integer Expressions
3558 @cindex integer expressions
3559 @cindex expressions, integer
3560 An @dfn{integer expression} is one or more @emph{arguments} delimited
3561 by @emph{operators}.
3564 * Arguments:: Arguments
3565 * Operators:: Operators
3566 * Prefix Ops:: Prefix Operators
3567 * Infix Ops:: Infix Operators
3571 @subsection Arguments
3573 @cindex expression arguments
3574 @cindex arguments in expressions
3575 @cindex operands in expressions
3576 @cindex arithmetic operands
3577 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3578 contexts arguments are sometimes called ``arithmetic operands''. In
3579 this manual, to avoid confusing them with the ``instruction operands'' of
3580 the machine language, we use the term ``argument'' to refer to parts of
3581 expressions only, reserving the word ``operand'' to refer only to machine
3582 instruction operands.
3584 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3585 @var{section} is one of text, data, bss, absolute,
3586 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3589 Numbers are usually integers.
3591 A number can be a flonum or bignum. In this case, you are warned
3592 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3593 these 32 bits are an integer. You may write integer-manipulating
3594 instructions that act on exotic constants, compatible with other
3597 @cindex subexpressions
3598 Subexpressions are a left parenthesis @samp{(} followed by an integer
3599 expression, followed by a right parenthesis @samp{)}; or a prefix
3600 operator followed by an argument.
3603 @subsection Operators
3605 @cindex operators, in expressions
3606 @cindex arithmetic functions
3607 @cindex functions, in expressions
3608 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3609 operators are followed by an argument. Infix operators appear
3610 between their arguments. Operators may be preceded and/or followed by
3614 @subsection Prefix Operator
3616 @cindex prefix operators
3617 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3618 one argument, which must be absolute.
3620 @c the tex/end tex stuff surrounding this small table is meant to make
3621 @c it align, on the printed page, with the similar table in the next
3622 @c section (which is inside an enumerate).
3624 \global\advance\leftskip by \itemindent
3629 @dfn{Negation}. Two's complement negation.
3631 @dfn{Complementation}. Bitwise not.
3635 \global\advance\leftskip by -\itemindent
3639 @subsection Infix Operators
3641 @cindex infix operators
3642 @cindex operators, permitted arguments
3643 @dfn{Infix operators} take two arguments, one on either side. Operators
3644 have precedence, but operations with equal precedence are performed left
3645 to right. Apart from @code{+} or @option{-}, both arguments must be
3646 absolute, and the result is absolute.
3649 @cindex operator precedence
3650 @cindex precedence of operators
3657 @dfn{Multiplication}.
3660 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3666 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3669 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3673 Intermediate precedence
3678 @dfn{Bitwise Inclusive Or}.
3684 @dfn{Bitwise Exclusive Or}.
3687 @dfn{Bitwise Or Not}.
3694 @cindex addition, permitted arguments
3695 @cindex plus, permitted arguments
3696 @cindex arguments for addition
3698 @dfn{Addition}. If either argument is absolute, the result has the section of
3699 the other argument. You may not add together arguments from different
3702 @cindex subtraction, permitted arguments
3703 @cindex minus, permitted arguments
3704 @cindex arguments for subtraction
3706 @dfn{Subtraction}. If the right argument is absolute, the
3707 result has the section of the left argument.
3708 If both arguments are in the same section, the result is absolute.
3709 You may not subtract arguments from different sections.
3710 @c FIXME is there still something useful to say about undefined - undefined ?
3712 @cindex comparison expressions
3713 @cindex expressions, comparison
3718 @dfn{Is Not Equal To}
3722 @dfn{Is Greater Than}
3724 @dfn{Is Greater Than Or Equal To}
3726 @dfn{Is Less Than Or Equal To}
3728 The comparison operators can be used as infix operators. A true results has a
3729 value of -1 whereas a false result has a value of 0. Note, these operators
3730 perform signed comparisons.
3733 @item Lowest Precedence
3742 These two logical operations can be used to combine the results of sub
3743 expressions. Note, unlike the comparison operators a true result returns a
3744 value of 1 but a false results does still return 0. Also note that the logical
3745 or operator has a slightly lower precedence than logical and.
3750 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3751 address; you can only have a defined section in one of the two arguments.
3754 @chapter Assembler Directives
3756 @cindex directives, machine independent
3757 @cindex pseudo-ops, machine independent
3758 @cindex machine independent directives
3759 All assembler directives have names that begin with a period (@samp{.}).
3760 The rest of the name is letters, usually in lower case.
3762 This chapter discusses directives that are available regardless of the
3763 target machine configuration for the @sc{gnu} assembler.
3765 Some machine configurations provide additional directives.
3766 @xref{Machine Dependencies}.
3769 @ifset machine-directives
3770 @xref{Machine Dependencies}, for additional directives.
3775 * Abort:: @code{.abort}
3777 * ABORT (COFF):: @code{.ABORT}
3780 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3781 * Altmacro:: @code{.altmacro}
3782 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3783 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3784 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3785 * Byte:: @code{.byte @var{expressions}}
3786 * Comm:: @code{.comm @var{symbol} , @var{length} }
3788 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
3790 * Data:: @code{.data @var{subsection}}
3792 * Def:: @code{.def @var{name}}
3795 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3801 * Double:: @code{.double @var{flonums}}
3802 * Eject:: @code{.eject}
3803 * Else:: @code{.else}
3804 * Elseif:: @code{.elseif}
3807 * Endef:: @code{.endef}
3810 * Endfunc:: @code{.endfunc}
3811 * Endif:: @code{.endif}
3812 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3813 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3814 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
3816 * Error:: @code{.error @var{string}}
3817 * Exitm:: @code{.exitm}
3818 * Extern:: @code{.extern}
3819 * Fail:: @code{.fail}
3820 @ifclear no-file-dir
3821 * File:: @code{.file @var{string}}
3824 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3825 * Float:: @code{.float @var{flonums}}
3826 * Func:: @code{.func}
3827 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3829 * Hidden:: @code{.hidden @var{names}}
3832 * hword:: @code{.hword @var{expressions}}
3833 * Ident:: @code{.ident}
3834 * If:: @code{.if @var{absolute expression}}
3835 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
3836 * Include:: @code{.include "@var{file}"}
3837 * Int:: @code{.int @var{expressions}}
3839 * Internal:: @code{.internal @var{names}}
3842 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3843 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3844 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3845 * Lflags:: @code{.lflags}
3846 @ifclear no-line-dir
3847 * Line:: @code{.line @var{line-number}}
3850 * Linkonce:: @code{.linkonce [@var{type}]}
3851 * List:: @code{.list}
3852 * Ln:: @code{.ln @var{line-number}}
3854 * LNS directives:: @code{.file}, @code{.loc}, etc.
3856 * Long:: @code{.long @var{expressions}}
3858 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3861 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3862 * MRI:: @code{.mri @var{val}}
3863 * Noaltmacro:: @code{.noaltmacro}
3864 * Nolist:: @code{.nolist}
3865 * Octa:: @code{.octa @var{bignums}}
3866 * Org:: @code{.org @var{new-lc}, @var{fill}}
3867 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3869 * PopSection:: @code{.popsection}
3870 * Previous:: @code{.previous}
3873 * Print:: @code{.print @var{string}}
3875 * Protected:: @code{.protected @var{names}}
3878 * Psize:: @code{.psize @var{lines}, @var{columns}}
3879 * Purgem:: @code{.purgem @var{name}}
3881 * PushSection:: @code{.pushsection @var{name}}
3884 * Quad:: @code{.quad @var{bignums}}
3885 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
3886 * Rept:: @code{.rept @var{count}}
3887 * Sbttl:: @code{.sbttl "@var{subheading}"}
3889 * Scl:: @code{.scl @var{class}}
3892 * Section:: @code{.section @var{name}}
3895 * Set:: @code{.set @var{symbol}, @var{expression}}
3896 * Short:: @code{.short @var{expressions}}
3897 * Single:: @code{.single @var{flonums}}
3899 * Size:: @code{.size [@var{name} , @var{expression}]}
3902 * Skip:: @code{.skip @var{size} , @var{fill}}
3903 * Sleb128:: @code{.sleb128 @var{expressions}}
3904 * Space:: @code{.space @var{size} , @var{fill}}
3906 * Stab:: @code{.stabd, .stabn, .stabs}
3909 * String:: @code{.string "@var{str}"}
3910 * Struct:: @code{.struct @var{expression}}
3912 * SubSection:: @code{.subsection}
3913 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3917 * Tag:: @code{.tag @var{structname}}
3920 * Text:: @code{.text @var{subsection}}
3921 * Title:: @code{.title "@var{heading}"}
3923 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3926 * Uleb128:: @code{.uleb128 @var{expressions}}
3928 * Val:: @code{.val @var{addr}}
3932 * Version:: @code{.version "@var{string}"}
3933 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3934 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3937 * Warning:: @code{.warning @var{string}}
3938 * Weak:: @code{.weak @var{names}}
3939 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
3940 * Word:: @code{.word @var{expressions}}
3941 * Deprecated:: Deprecated Directives
3945 @section @code{.abort}
3947 @cindex @code{abort} directive
3948 @cindex stopping the assembly
3949 This directive stops the assembly immediately. It is for
3950 compatibility with other assemblers. The original idea was that the
3951 assembly language source would be piped into the assembler. If the sender
3952 of the source quit, it could use this directive tells @command{@value{AS}} to
3953 quit also. One day @code{.abort} will not be supported.
3957 @section @code{.ABORT} (COFF)
3959 @cindex @code{ABORT} directive
3960 When producing COFF output, @command{@value{AS}} accepts this directive as a
3961 synonym for @samp{.abort}.
3964 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
3970 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3972 @cindex padding the location counter
3973 @cindex @code{align} directive
3974 Pad the location counter (in the current subsection) to a particular storage
3975 boundary. The first expression (which must be absolute) is the alignment
3976 required, as described below.
3978 The second expression (also absolute) gives the fill value to be stored in the
3979 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3980 padding bytes are normally zero. However, on some systems, if the section is
3981 marked as containing code and the fill value is omitted, the space is filled
3982 with no-op instructions.
3984 The third expression is also absolute, and is also optional. If it is present,
3985 it is the maximum number of bytes that should be skipped by this alignment
3986 directive. If doing the alignment would require skipping more bytes than the
3987 specified maximum, then the alignment is not done at all. You can omit the
3988 fill value (the second argument) entirely by simply using two commas after the
3989 required alignment; this can be useful if you want the alignment to be filled
3990 with no-op instructions when appropriate.
3992 The way the required alignment is specified varies from system to system.
3993 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
3994 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
3995 alignment request in bytes. For example @samp{.align 8} advances
3996 the location counter until it is a multiple of 8. If the location counter
3997 is already a multiple of 8, no change is needed. For the tic54x, the
3998 first expression is the alignment request in words.
4000 For other systems, including the i386 using a.out format, and the arm and
4001 strongarm, it is the
4002 number of low-order zero bits the location counter must have after
4003 advancement. For example @samp{.align 3} advances the location
4004 counter until it a multiple of 8. If the location counter is already a
4005 multiple of 8, no change is needed.
4007 This inconsistency is due to the different behaviors of the various
4008 native assemblers for these systems which GAS must emulate.
4009 GAS also provides @code{.balign} and @code{.p2align} directives,
4010 described later, which have a consistent behavior across all
4011 architectures (but are specific to GAS).
4014 @section @code{.ascii "@var{string}"}@dots{}
4016 @cindex @code{ascii} directive
4017 @cindex string literals
4018 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4019 separated by commas. It assembles each string (with no automatic
4020 trailing zero byte) into consecutive addresses.
4023 @section @code{.asciz "@var{string}"}@dots{}
4025 @cindex @code{asciz} directive
4026 @cindex zero-terminated strings
4027 @cindex null-terminated strings
4028 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4029 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4032 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4034 @cindex padding the location counter given number of bytes
4035 @cindex @code{balign} directive
4036 Pad the location counter (in the current subsection) to a particular
4037 storage boundary. The first expression (which must be absolute) is the
4038 alignment request in bytes. For example @samp{.balign 8} advances
4039 the location counter until it is a multiple of 8. If the location counter
4040 is already a multiple of 8, no change is needed.
4042 The second expression (also absolute) gives the fill value to be stored in the
4043 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4044 padding bytes are normally zero. However, on some systems, if the section is
4045 marked as containing code and the fill value is omitted, the space is filled
4046 with no-op instructions.
4048 The third expression is also absolute, and is also optional. If it is present,
4049 it is the maximum number of bytes that should be skipped by this alignment
4050 directive. If doing the alignment would require skipping more bytes than the
4051 specified maximum, then the alignment is not done at all. You can omit the
4052 fill value (the second argument) entirely by simply using two commas after the
4053 required alignment; this can be useful if you want the alignment to be filled
4054 with no-op instructions when appropriate.
4056 @cindex @code{balignw} directive
4057 @cindex @code{balignl} directive
4058 The @code{.balignw} and @code{.balignl} directives are variants of the
4059 @code{.balign} directive. The @code{.balignw} directive treats the fill
4060 pattern as a two byte word value. The @code{.balignl} directives treats the
4061 fill pattern as a four byte longword value. For example, @code{.balignw
4062 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4063 filled in with the value 0x368d (the exact placement of the bytes depends upon
4064 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4068 @section @code{.byte @var{expressions}}
4070 @cindex @code{byte} directive
4071 @cindex integers, one byte
4072 @code{.byte} expects zero or more expressions, separated by commas.
4073 Each expression is assembled into the next byte.
4076 @section @code{.comm @var{symbol} , @var{length} }
4078 @cindex @code{comm} directive
4079 @cindex symbol, common
4080 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4081 common symbol in one object file may be merged with a defined or common symbol
4082 of the same name in another object file. If @code{@value{LD}} does not see a
4083 definition for the symbol--just one or more common symbols--then it will
4084 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4085 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4086 the same name, and they do not all have the same size, it will allocate space
4087 using the largest size.
4090 When using ELF, the @code{.comm} directive takes an optional third argument.
4091 This is the desired alignment of the symbol, specified as a byte boundary (for
4092 example, an alignment of 16 means that the least significant 4 bits of the
4093 address should be zero). The alignment must be an absolute expression, and it
4094 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
4095 for the common symbol, it will use the alignment when placing the symbol. If
4096 no alignment is specified, @command{@value{AS}} will set the alignment to the
4097 largest power of two less than or equal to the size of the symbol, up to a
4102 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4103 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4106 @node CFI directives
4107 @section @code{.cfi_startproc [simple]}
4108 @cindex @code{cfi_startproc} directive
4109 @code{.cfi_startproc} is used at the beginning of each function that
4110 should have an entry in @code{.eh_frame}. It initializes some internal
4111 data structures. Don't forget to close the function by
4112 @code{.cfi_endproc}.
4114 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4115 it also emits some architecture dependent initial CFI instructions.
4117 @section @code{.cfi_endproc}
4118 @cindex @code{cfi_endproc} directive
4119 @code{.cfi_endproc} is used at the end of a function where it closes its
4120 unwind entry previously opened by
4121 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4123 @section @code{.cfi_personality @var{encoding} [, @var{exp}]}
4124 @code{.cfi_personality} defines personality routine and its encoding.
4125 @var{encoding} must be a constant determining how the personality
4126 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4127 argument is not present, otherwise second argument should be
4128 a constant or a symbol name. When using indirect encodings,
4129 the symbol provided should be the location where personality
4130 can be loaded from, not the personality routine itself.
4131 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4132 no personality routine.
4134 @section @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4135 @code{.cfi_lsda} defines LSDA and its encoding.
4136 @var{encoding} must be a constant determining how the LSDA
4137 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4138 argument is not present, otherwise second argument should be a constant
4139 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4142 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4143 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4144 address from @var{register} and add @var{offset} to it}.
4146 @section @code{.cfi_def_cfa_register @var{register}}
4147 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4148 now on @var{register} will be used instead of the old one. Offset
4151 @section @code{.cfi_def_cfa_offset @var{offset}}
4152 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4153 remains the same, but @var{offset} is new. Note that it is the
4154 absolute offset that will be added to a defined register to compute
4157 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4158 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4159 value that is added/substracted from the previous offset.
4161 @section @code{.cfi_offset @var{register}, @var{offset}}
4162 Previous value of @var{register} is saved at offset @var{offset} from
4165 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4166 Previous value of @var{register} is saved at offset @var{offset} from
4167 the current CFA register. This is transformed to @code{.cfi_offset}
4168 using the known displacement of the CFA register from the CFA.
4169 This is often easier to use, because the number will match the
4170 code it's annotating.
4172 @section @code{.cfi_register @var{register1}, @var{register2}}
4173 Previous value of @var{register1} is saved in register @var{register2}.
4175 @section @code{.cfi_restore @var{register}}
4176 @code{.cfi_restore} says that the rule for @var{register} is now the
4177 same as it was at the beginning of the function, after all initial
4178 instruction added by @code{.cfi_startproc} were executed.
4180 @section @code{.cfi_undefined @var{register}}
4181 From now on the previous value of @var{register} can't be restored anymore.
4183 @section @code{.cfi_same_value @var{register}}
4184 Current value of @var{register} is the same like in the previous frame,
4185 i.e. no restoration needed.
4187 @section @code{.cfi_remember_state},
4188 First save all current rules for all registers by @code{.cfi_remember_state},
4189 then totally screw them up by subsequent @code{.cfi_*} directives and when
4190 everything is hopelessly bad, use @code{.cfi_restore_state} to restore
4191 the previous saved state.
4193 @section @code{.cfi_return_column @var{register}}
4194 Change return column @var{register}, i.e. the return address is either
4195 directly in @var{register} or can be accessed by rules for @var{register}.
4197 @section @code{.cfi_signal_frame}
4198 Mark current function as signal trampoline.
4200 @section @code{.cfi_window_save}
4201 SPARC register window has been saved.
4203 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4204 Allows the user to add arbitrary bytes to the unwind info. One
4205 might use this to add OS-specific CFI opcodes, or generic CFI
4206 opcodes that GAS does not yet support.
4208 @node LNS directives
4209 @section @code{.file @var{fileno} @var{filename}}
4210 @cindex @code{file} directive
4211 When emitting dwarf2 line number information @code{.file} assigns filenames
4212 to the @code{.debug_line} file name table. The @var{fileno} operand should
4213 be a unique positive integer to use as the index of the entry in the table.
4214 The @var{filename} operand is a C string literal.
4216 The detail of filename indices is exposed to the user because the filename
4217 table is shared with the @code{.debug_info} section of the dwarf2 debugging
4218 information, and thus the user must know the exact indices that table
4221 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
4222 @cindex @code{loc} directive
4223 The @code{.loc} directive will add row to the @code{.debug_line} line
4224 number matrix corresponding to the immediately following assembly
4225 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
4226 arguments will be applied to the @code{.debug_line} state machine before
4229 The @var{options} are a sequence of the following tokens in any order:
4233 This option will set the @code{basic_block} register in the
4234 @code{.debug_line} state machine to @code{true}.
4237 This option will set the @code{prologue_end} register in the
4238 @code{.debug_line} state machine to @code{true}.
4240 @item epilogue_begin
4241 This option will set the @code{epilogue_begin} register in the
4242 @code{.debug_line} state machine to @code{true}.
4244 @item is_stmt @var{value}
4245 This option will set the @code{is_stmt} register in the
4246 @code{.debug_line} state machine to @code{value}, which must be
4249 @item isa @var{value}
4250 This directive will set the @code{isa} register in the @code{.debug_line}
4251 state machine to @var{value}, which must be an unsigned integer.
4255 @section @code{.loc_mark_blocks @var{enable}}
4256 @cindex @code{loc_mark_blocks} directive
4257 The @code{.loc_mark_blocks} directive makes the assembler emit an entry
4258 to the @code{.debug_line} line number matrix with the @code{basic_block}
4259 register in the state machine set whenever a code label is seen.
4260 The @var{enable} argument should be either 1 or 0, to enable or disable
4261 this function respectively.
4264 @section @code{.data @var{subsection}}
4266 @cindex @code{data} directive
4267 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4268 end of the data subsection numbered @var{subsection} (which is an
4269 absolute expression). If @var{subsection} is omitted, it defaults
4274 @section @code{.def @var{name}}
4276 @cindex @code{def} directive
4277 @cindex COFF symbols, debugging
4278 @cindex debugging COFF symbols
4279 Begin defining debugging information for a symbol @var{name}; the
4280 definition extends until the @code{.endef} directive is encountered.
4283 This directive is only observed when @command{@value{AS}} is configured for COFF
4284 format output; when producing @code{b.out}, @samp{.def} is recognized,
4291 @section @code{.desc @var{symbol}, @var{abs-expression}}
4293 @cindex @code{desc} directive
4294 @cindex COFF symbol descriptor
4295 @cindex symbol descriptor, COFF
4296 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4297 to the low 16 bits of an absolute expression.
4300 The @samp{.desc} directive is not available when @command{@value{AS}} is
4301 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4302 object format. For the sake of compatibility, @command{@value{AS}} accepts
4303 it, but produces no output, when configured for COFF.
4309 @section @code{.dim}
4311 @cindex @code{dim} directive
4312 @cindex COFF auxiliary symbol information
4313 @cindex auxiliary symbol information, COFF
4314 This directive is generated by compilers to include auxiliary debugging
4315 information in the symbol table. It is only permitted inside
4316 @code{.def}/@code{.endef} pairs.
4319 @samp{.dim} is only meaningful when generating COFF format output; when
4320 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4326 @section @code{.double @var{flonums}}
4328 @cindex @code{double} directive
4329 @cindex floating point numbers (double)
4330 @code{.double} expects zero or more flonums, separated by commas. It
4331 assembles floating point numbers.
4333 The exact kind of floating point numbers emitted depends on how
4334 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4338 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4339 in @sc{ieee} format.
4344 @section @code{.eject}
4346 @cindex @code{eject} directive
4347 @cindex new page, in listings
4348 @cindex page, in listings
4349 @cindex listing control: new page
4350 Force a page break at this point, when generating assembly listings.
4353 @section @code{.else}
4355 @cindex @code{else} directive
4356 @code{.else} is part of the @command{@value{AS}} support for conditional
4357 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
4358 of code to be assembled if the condition for the preceding @code{.if}
4362 @section @code{.elseif}
4364 @cindex @code{elseif} directive
4365 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4366 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
4367 @code{.if} block that would otherwise fill the entire @code{.else} section.
4370 @section @code{.end}
4372 @cindex @code{end} directive
4373 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4374 process anything in the file past the @code{.end} directive.
4378 @section @code{.endef}
4380 @cindex @code{endef} directive
4381 This directive flags the end of a symbol definition begun with
4385 @samp{.endef} is only meaningful when generating COFF format output; if
4386 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4387 directive but ignores it.
4392 @section @code{.endfunc}
4393 @cindex @code{endfunc} directive
4394 @code{.endfunc} marks the end of a function specified with @code{.func}.
4397 @section @code{.endif}
4399 @cindex @code{endif} directive
4400 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4401 it marks the end of a block of code that is only assembled
4402 conditionally. @xref{If,,@code{.if}}.
4405 @section @code{.equ @var{symbol}, @var{expression}}
4407 @cindex @code{equ} directive
4408 @cindex assigning values to symbols
4409 @cindex symbols, assigning values to
4410 This directive sets the value of @var{symbol} to @var{expression}.
4411 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
4414 The syntax for @code{equ} on the HPPA is
4415 @samp{@var{symbol} .equ @var{expression}}.
4419 The syntax for @code{equ} on the Z80 is
4420 @samp{@var{symbol} equ @var{expression}}.
4421 On the Z80 it is an eror if @var{symbol} is already defined,
4422 but the symbol is not protected from later redefinition.
4423 Compare @ref{Equiv}.
4427 @section @code{.equiv @var{symbol}, @var{expression}}
4428 @cindex @code{equiv} directive
4429 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4430 the assembler will signal an error if @var{symbol} is already defined. Note a
4431 symbol which has been referenced but not actually defined is considered to be
4434 Except for the contents of the error message, this is roughly equivalent to
4441 plus it protects the symbol from later redefinition.
4444 @section @code{.eqv @var{symbol}, @var{expression}}
4445 @cindex @code{eqv} directive
4446 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4447 evaluate the expression or any part of it immediately. Instead each time
4448 the resulting symbol is used in an expression, a snapshot of its current
4452 @section @code{.err}
4453 @cindex @code{err} directive
4454 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4455 message and, unless the @option{-Z} option was used, it will not generate an
4456 object file. This can be used to signal an error in conditionally compiled code.
4459 @section @code{.error "@var{string}"}
4460 @cindex error directive
4462 Similarly to @code{.err}, this directive emits an error, but you can specify a
4463 string that will be emitted as the error message. If you don't specify the
4464 message, it defaults to @code{".error directive invoked in source file"}.
4465 @xref{Errors, ,Error and Warning Messages}.
4468 .error "This code has not been assembled and tested."
4472 @section @code{.exitm}
4473 Exit early from the current macro definition. @xref{Macro}.
4476 @section @code{.extern}
4478 @cindex @code{extern} directive
4479 @code{.extern} is accepted in the source program---for compatibility
4480 with other assemblers---but it is ignored. @command{@value{AS}} treats
4481 all undefined symbols as external.
4484 @section @code{.fail @var{expression}}
4486 @cindex @code{fail} directive
4487 Generates an error or a warning. If the value of the @var{expression} is 500
4488 or more, @command{@value{AS}} will print a warning message. If the value is less
4489 than 500, @command{@value{AS}} will print an error message. The message will
4490 include the value of @var{expression}. This can occasionally be useful inside
4491 complex nested macros or conditional assembly.
4493 @ifclear no-file-dir
4495 @section @code{.file @var{string}}
4497 @cindex @code{file} directive
4498 @cindex logical file name
4499 @cindex file name, logical
4500 @code{.file} tells @command{@value{AS}} that we are about to start a new logical
4501 file. @var{string} is the new file name. In general, the filename is
4502 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4503 to specify an empty file name, you must give the quotes--@code{""}. This
4504 statement may go away in future: it is only recognized to be compatible with
4505 old @command{@value{AS}} programs.
4509 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4511 @cindex @code{fill} directive
4512 @cindex writing patterns in memory
4513 @cindex patterns, writing in memory
4514 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4515 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4516 may be zero or more. @var{Size} may be zero or more, but if it is
4517 more than 8, then it is deemed to have the value 8, compatible with
4518 other people's assemblers. The contents of each @var{repeat} bytes
4519 is taken from an 8-byte number. The highest order 4 bytes are
4520 zero. The lowest order 4 bytes are @var{value} rendered in the
4521 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4522 Each @var{size} bytes in a repetition is taken from the lowest order
4523 @var{size} bytes of this number. Again, this bizarre behavior is
4524 compatible with other people's assemblers.
4526 @var{size} and @var{value} are optional.
4527 If the second comma and @var{value} are absent, @var{value} is
4528 assumed zero. If the first comma and following tokens are absent,
4529 @var{size} is assumed to be 1.
4532 @section @code{.float @var{flonums}}
4534 @cindex floating point numbers (single)
4535 @cindex @code{float} directive
4536 This directive assembles zero or more flonums, separated by commas. It
4537 has the same effect as @code{.single}.
4539 The exact kind of floating point numbers emitted depends on how
4540 @command{@value{AS}} is configured.
4541 @xref{Machine Dependencies}.
4545 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4546 in @sc{ieee} format.
4551 @section @code{.func @var{name}[,@var{label}]}
4552 @cindex @code{func} directive
4553 @code{.func} emits debugging information to denote function @var{name}, and
4554 is ignored unless the file is assembled with debugging enabled.
4555 Only @samp{--gstabs[+]} is currently supported.
4556 @var{label} is the entry point of the function and if omitted @var{name}
4557 prepended with the @samp{leading char} is used.
4558 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4559 All functions are currently defined to have @code{void} return type.
4560 The function must be terminated with @code{.endfunc}.
4563 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4565 @cindex @code{global} directive
4566 @cindex symbol, making visible to linker
4567 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4568 @var{symbol} in your partial program, its value is made available to
4569 other partial programs that are linked with it. Otherwise,
4570 @var{symbol} takes its attributes from a symbol of the same name
4571 from another file linked into the same program.
4573 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4574 compatibility with other assemblers.
4577 On the HPPA, @code{.global} is not always enough to make it accessible to other
4578 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4579 @xref{HPPA Directives, ,HPPA Assembler Directives}.
4584 @section @code{.hidden @var{names}}
4586 @cindex @code{hidden} directive
4588 This is one of the ELF visibility directives. The other two are
4589 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4590 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4592 This directive overrides the named symbols default visibility (which is set by
4593 their binding: local, global or weak). The directive sets the visibility to
4594 @code{hidden} which means that the symbols are not visible to other components.
4595 Such symbols are always considered to be @code{protected} as well.
4599 @section @code{.hword @var{expressions}}
4601 @cindex @code{hword} directive
4602 @cindex integers, 16-bit
4603 @cindex numbers, 16-bit
4604 @cindex sixteen bit integers
4605 This expects zero or more @var{expressions}, and emits
4606 a 16 bit number for each.
4609 This directive is a synonym for @samp{.short}; depending on the target
4610 architecture, it may also be a synonym for @samp{.word}.
4614 This directive is a synonym for @samp{.short}.
4617 This directive is a synonym for both @samp{.short} and @samp{.word}.
4622 @section @code{.ident}
4624 @cindex @code{ident} directive
4626 This directive is used by some assemblers to place tags in object files. The
4627 behavior of this directive varies depending on the target. When using the
4628 a.out object file format, @command{@value{AS}} simply accepts the directive for
4629 source-file compatibility with existing assemblers, but does not emit anything
4630 for it. When using COFF, comments are emitted to the @code{.comment} or
4631 @code{.rdata} section, depending on the target. When using ELF, comments are
4632 emitted to the @code{.comment} section.
4635 @section @code{.if @var{absolute expression}}
4637 @cindex conditional assembly
4638 @cindex @code{if} directive
4639 @code{.if} marks the beginning of a section of code which is only
4640 considered part of the source program being assembled if the argument
4641 (which must be an @var{absolute expression}) is non-zero. The end of
4642 the conditional section of code must be marked by @code{.endif}
4643 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4644 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4645 If you have several conditions to check, @code{.elseif} may be used to avoid
4646 nesting blocks if/else within each subsequent @code{.else} block.
4648 The following variants of @code{.if} are also supported:
4650 @cindex @code{ifdef} directive
4651 @item .ifdef @var{symbol}
4652 Assembles the following section of code if the specified @var{symbol}
4653 has been defined. Note a symbol which has been referenced but not yet defined
4654 is considered to be undefined.
4656 @cindex @code{ifb} directive
4657 @item .ifb @var{text}
4658 Assembles the following section of code if the operand is blank (empty).
4660 @cindex @code{ifc} directive
4661 @item .ifc @var{string1},@var{string2}
4662 Assembles the following section of code if the two strings are the same. The
4663 strings may be optionally quoted with single quotes. If they are not quoted,
4664 the first string stops at the first comma, and the second string stops at the
4665 end of the line. Strings which contain whitespace should be quoted. The
4666 string comparison is case sensitive.
4668 @cindex @code{ifeq} directive
4669 @item .ifeq @var{absolute expression}
4670 Assembles the following section of code if the argument is zero.
4672 @cindex @code{ifeqs} directive
4673 @item .ifeqs @var{string1},@var{string2}
4674 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4676 @cindex @code{ifge} directive
4677 @item .ifge @var{absolute expression}
4678 Assembles the following section of code if the argument is greater than or
4681 @cindex @code{ifgt} directive
4682 @item .ifgt @var{absolute expression}
4683 Assembles the following section of code if the argument is greater than zero.
4685 @cindex @code{ifle} directive
4686 @item .ifle @var{absolute expression}
4687 Assembles the following section of code if the argument is less than or equal
4690 @cindex @code{iflt} directive
4691 @item .iflt @var{absolute expression}
4692 Assembles the following section of code if the argument is less than zero.
4694 @cindex @code{ifnb} directive
4695 @item .ifnb @var{text}
4696 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
4697 following section of code if the operand is non-blank (non-empty).
4699 @cindex @code{ifnc} directive
4700 @item .ifnc @var{string1},@var{string2}.
4701 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4702 following section of code if the two strings are not the same.
4704 @cindex @code{ifndef} directive
4705 @cindex @code{ifnotdef} directive
4706 @item .ifndef @var{symbol}
4707 @itemx .ifnotdef @var{symbol}
4708 Assembles the following section of code if the specified @var{symbol}
4709 has not been defined. Both spelling variants are equivalent. Note a symbol
4710 which has been referenced but not yet defined is considered to be undefined.
4712 @cindex @code{ifne} directive
4713 @item .ifne @var{absolute expression}
4714 Assembles the following section of code if the argument is not equal to zero
4715 (in other words, this is equivalent to @code{.if}).
4717 @cindex @code{ifnes} directive
4718 @item .ifnes @var{string1},@var{string2}
4719 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4720 following section of code if the two strings are not the same.
4724 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4726 @cindex @code{incbin} directive
4727 @cindex binary files, including
4728 The @code{incbin} directive includes @var{file} verbatim at the current
4729 location. You can control the search paths used with the @samp{-I} command-line
4730 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4733 The @var{skip} argument skips a number of bytes from the start of the
4734 @var{file}. The @var{count} argument indicates the maximum number of bytes to
4735 read. Note that the data is not aligned in any way, so it is the user's
4736 responsibility to make sure that proper alignment is provided both before and
4737 after the @code{incbin} directive.
4740 @section @code{.include "@var{file}"}
4742 @cindex @code{include} directive
4743 @cindex supporting files, including
4744 @cindex files, including
4745 This directive provides a way to include supporting files at specified
4746 points in your source program. The code from @var{file} is assembled as
4747 if it followed the point of the @code{.include}; when the end of the
4748 included file is reached, assembly of the original file continues. You
4749 can control the search paths used with the @samp{-I} command-line option
4750 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4754 @section @code{.int @var{expressions}}
4756 @cindex @code{int} directive
4757 @cindex integers, 32-bit
4758 Expect zero or more @var{expressions}, of any section, separated by commas.
4759 For each expression, emit a number that, at run time, is the value of that
4760 expression. The byte order and bit size of the number depends on what kind
4761 of target the assembly is for.
4765 On most forms of the H8/300, @code{.int} emits 16-bit
4766 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
4773 @section @code{.internal @var{names}}
4775 @cindex @code{internal} directive
4777 This is one of the ELF visibility directives. The other two are
4778 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4779 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4781 This directive overrides the named symbols default visibility (which is set by
4782 their binding: local, global or weak). The directive sets the visibility to
4783 @code{internal} which means that the symbols are considered to be @code{hidden}
4784 (i.e., not visible to other components), and that some extra, processor specific
4785 processing must also be performed upon the symbols as well.
4789 @section @code{.irp @var{symbol},@var{values}}@dots{}
4791 @cindex @code{irp} directive
4792 Evaluate a sequence of statements assigning different values to @var{symbol}.
4793 The sequence of statements starts at the @code{.irp} directive, and is
4794 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4795 set to @var{value}, and the sequence of statements is assembled. If no
4796 @var{value} is listed, the sequence of statements is assembled once, with
4797 @var{symbol} set to the null string. To refer to @var{symbol} within the
4798 sequence of statements, use @var{\symbol}.
4800 For example, assembling
4808 is equivalent to assembling
4816 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
4819 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4821 @cindex @code{irpc} directive
4822 Evaluate a sequence of statements assigning different values to @var{symbol}.
4823 The sequence of statements starts at the @code{.irpc} directive, and is
4824 terminated by an @code{.endr} directive. For each character in @var{value},
4825 @var{symbol} is set to the character, and the sequence of statements is
4826 assembled. If no @var{value} is listed, the sequence of statements is
4827 assembled once, with @var{symbol} set to the null string. To refer to
4828 @var{symbol} within the sequence of statements, use @var{\symbol}.
4830 For example, assembling
4838 is equivalent to assembling
4846 For some caveats with the spelling of @var{symbol}, see also the discussion
4850 @section @code{.lcomm @var{symbol} , @var{length}}
4852 @cindex @code{lcomm} directive
4853 @cindex local common symbols
4854 @cindex symbols, local common
4855 Reserve @var{length} (an absolute expression) bytes for a local common
4856 denoted by @var{symbol}. The section and value of @var{symbol} are
4857 those of the new local common. The addresses are allocated in the bss
4858 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4859 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4860 not visible to @code{@value{LD}}.
4863 Some targets permit a third argument to be used with @code{.lcomm}. This
4864 argument specifies the desired alignment of the symbol in the bss section.
4868 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4869 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4873 @section @code{.lflags}
4875 @cindex @code{lflags} directive (ignored)
4876 @command{@value{AS}} accepts this directive, for compatibility with other
4877 assemblers, but ignores it.
4879 @ifclear no-line-dir
4881 @section @code{.line @var{line-number}}
4883 @cindex @code{line} directive
4887 @section @code{.ln @var{line-number}}
4889 @cindex @code{ln} directive
4891 @cindex logical line number
4893 Change the logical line number. @var{line-number} must be an absolute
4894 expression. The next line has that logical line number. Therefore any other
4895 statements on the current line (after a statement separator character) are
4896 reported as on logical line number @var{line-number} @minus{} 1. One day
4897 @command{@value{AS}} will no longer support this directive: it is recognized only
4898 for compatibility with existing assembler programs.
4902 @ifclear no-line-dir
4903 Even though this is a directive associated with the @code{a.out} or
4904 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
4905 when producing COFF output, and treats @samp{.line} as though it
4906 were the COFF @samp{.ln} @emph{if} it is found outside a
4907 @code{.def}/@code{.endef} pair.
4909 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4910 used by compilers to generate auxiliary symbol information for
4915 @section @code{.linkonce [@var{type}]}
4917 @cindex @code{linkonce} directive
4918 @cindex common sections
4919 Mark the current section so that the linker only includes a single copy of it.
4920 This may be used to include the same section in several different object files,
4921 but ensure that the linker will only include it once in the final output file.
4922 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4923 Duplicate sections are detected based on the section name, so it should be
4926 This directive is only supported by a few object file formats; as of this
4927 writing, the only object file format which supports it is the Portable
4928 Executable format used on Windows NT.
4930 The @var{type} argument is optional. If specified, it must be one of the
4931 following strings. For example:
4935 Not all types may be supported on all object file formats.
4939 Silently discard duplicate sections. This is the default.
4942 Warn if there are duplicate sections, but still keep only one copy.
4945 Warn if any of the duplicates have different sizes.
4948 Warn if any of the duplicates do not have exactly the same contents.
4952 @section @code{.ln @var{line-number}}
4954 @cindex @code{ln} directive
4955 @ifclear no-line-dir
4956 @samp{.ln} is a synonym for @samp{.line}.
4959 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
4960 must be an absolute expression. The next line has that logical
4961 line number, so any other statements on the current line (after a
4962 statement separator character @code{;}) are reported as on logical
4963 line number @var{line-number} @minus{} 1.
4966 This directive is accepted, but ignored, when @command{@value{AS}} is
4967 configured for @code{b.out}; its effect is only associated with COFF
4973 @section @code{.mri @var{val}}
4975 @cindex @code{mri} directive
4976 @cindex MRI mode, temporarily
4977 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
4978 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
4979 affects code assembled until the next @code{.mri} directive, or until the end
4980 of the file. @xref{M, MRI mode, MRI mode}.
4983 @section @code{.list}
4985 @cindex @code{list} directive
4986 @cindex listing control, turning on
4987 Control (in conjunction with the @code{.nolist} directive) whether or
4988 not assembly listings are generated. These two directives maintain an
4989 internal counter (which is zero initially). @code{.list} increments the
4990 counter, and @code{.nolist} decrements it. Assembly listings are
4991 generated whenever the counter is greater than zero.
4993 By default, listings are disabled. When you enable them (with the
4994 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4995 the initial value of the listing counter is one.
4998 @section @code{.long @var{expressions}}
5000 @cindex @code{long} directive
5001 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5004 @c no one seems to know what this is for or whether this description is
5005 @c what it really ought to do
5007 @section @code{.lsym @var{symbol}, @var{expression}}
5009 @cindex @code{lsym} directive
5010 @cindex symbol, not referenced in assembly
5011 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5012 the hash table, ensuring it cannot be referenced by name during the
5013 rest of the assembly. This sets the attributes of the symbol to be
5014 the same as the expression value:
5016 @var{other} = @var{descriptor} = 0
5017 @var{type} = @r{(section of @var{expression})}
5018 @var{value} = @var{expression}
5021 The new symbol is not flagged as external.
5025 @section @code{.macro}
5028 The commands @code{.macro} and @code{.endm} allow you to define macros that
5029 generate assembly output. For example, this definition specifies a macro
5030 @code{sum} that puts a sequence of numbers into memory:
5033 .macro sum from=0, to=5
5042 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5054 @item .macro @var{macname}
5055 @itemx .macro @var{macname} @var{macargs} @dots{}
5056 @cindex @code{macro} directive
5057 Begin the definition of a macro called @var{macname}. If your macro
5058 definition requires arguments, specify their names after the macro name,
5059 separated by commas or spaces. You can qualify the macro argument to
5060 indicate whether all invocations must specify a non-blank value (through
5061 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5062 (through @samp{:@code{vararg}}). You can supply a default value for any
5063 macro argument by following the name with @samp{=@var{deflt}}. You
5064 cannot define two macros with the same @var{macname} unless it has been
5065 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5066 definitions. For example, these are all valid @code{.macro} statements:
5070 Begin the definition of a macro called @code{comm}, which takes no
5073 @item .macro plus1 p, p1
5074 @itemx .macro plus1 p p1
5075 Either statement begins the definition of a macro called @code{plus1},
5076 which takes two arguments; within the macro definition, write
5077 @samp{\p} or @samp{\p1} to evaluate the arguments.
5079 @item .macro reserve_str p1=0 p2
5080 Begin the definition of a macro called @code{reserve_str}, with two
5081 arguments. The first argument has a default value, but not the second.
5082 After the definition is complete, you can call the macro either as
5083 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5084 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5085 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5086 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5088 @item .macro m p1:req, p2=0, p3:vararg
5089 Begin the definition of a macro called @code{m}, with at least three
5090 arguments. The first argument must always have a value specified, but
5091 not the second, which instead has a default value. The third formal
5092 will get assigned all remaining arguments specified at invocation time.
5094 When you call a macro, you can specify the argument values either by
5095 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5096 @samp{sum to=17, from=9}.
5100 Note that since each of the @var{macargs} can be an identifier exactly
5101 as any other one permitted by the target architecture, there may be
5102 occasional problems if the target hand-crafts special meanings to certain
5103 characters when they occur in a special position. For example, if the colon
5104 (@code{:}) is generally permitted to be part of a symbol name, but the
5105 architecture specific code special-cases it when occurring as the final
5106 character of a symbol (to denote a label), then the macro parameter
5107 replacement code will have no way of knowing that and consider the whole
5108 construct (including the colon) an identifier, and check only this
5109 identifier for being the subject to parameter substitution. So for example
5110 this macro definition:
5118 might not work as expected. Invoking @samp{label foo} might not create a label
5119 called @samp{foo} but instead just insert the text @samp{\l:} into the
5120 assembler source, probably generating an error about an unrecognised
5123 Similarly problems might occur with the period character (@samp{.})
5124 which is often allowed inside opcode names (and hence identifier names). So
5125 for example constructing a macro to build an opcode from a base name and a
5126 length specifier like this:
5129 .macro opcode base length
5134 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5135 instruction but instead generate some kind of error as the assembler tries to
5136 interpret the text @samp{\base.\length}.
5138 There are several possible ways around this problem:
5141 @item Insert white space
5142 If it is possible to use white space characters then this is the simplest
5151 @item Use @samp{\()}
5152 The string @samp{\()} can be used to separate the end of a macro argument from
5153 the following text. eg:
5156 .macro opcode base length
5161 @item Use the alternate macro syntax mode
5162 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5163 used as a separator. eg:
5173 Note: this problem of correctly identifying string parameters to pseudo ops
5174 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5175 and @code{.irpc} (@pxref{Irpc}) as well.
5178 @cindex @code{endm} directive
5179 Mark the end of a macro definition.
5182 @cindex @code{exitm} directive
5183 Exit early from the current macro definition.
5185 @cindex number of macros executed
5186 @cindex macros, count executed
5188 @command{@value{AS}} maintains a counter of how many macros it has
5189 executed in this pseudo-variable; you can copy that number to your
5190 output with @samp{\@@}, but @emph{only within a macro definition}.
5192 @item LOCAL @var{name} [ , @dots{} ]
5193 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5194 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5195 @xref{Altmacro,,@code{.altmacro}}.
5199 @section @code{.altmacro}
5200 Enable alternate macro mode, enabling:
5203 @item LOCAL @var{name} [ , @dots{} ]
5204 One additional directive, @code{LOCAL}, is available. It is used to
5205 generate a string replacement for each of the @var{name} arguments, and
5206 replace any instances of @var{name} in each macro expansion. The
5207 replacement string is unique in the assembly, and different for each
5208 separate macro expansion. @code{LOCAL} allows you to write macros that
5209 define symbols, without fear of conflict between separate macro expansions.
5211 @item String delimiters
5212 You can write strings delimited in these other ways besides
5213 @code{"@var{string}"}:
5216 @item '@var{string}'
5217 You can delimit strings with single-quote characters.
5219 @item <@var{string}>
5220 You can delimit strings with matching angle brackets.
5223 @item single-character string escape
5224 To include any single character literally in a string (even if the
5225 character would otherwise have some special meaning), you can prefix the
5226 character with @samp{!} (an exclamation mark). For example, you can
5227 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
5229 @item Expression results as strings
5230 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
5231 and use the result as a string.
5235 @section @code{.noaltmacro}
5236 Disable alternate macro mode. @xref{Altmacro}.
5239 @section @code{.nolist}
5241 @cindex @code{nolist} directive
5242 @cindex listing control, turning off
5243 Control (in conjunction with the @code{.list} directive) whether or
5244 not assembly listings are generated. These two directives maintain an
5245 internal counter (which is zero initially). @code{.list} increments the
5246 counter, and @code{.nolist} decrements it. Assembly listings are
5247 generated whenever the counter is greater than zero.
5250 @section @code{.octa @var{bignums}}
5252 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5253 @cindex @code{octa} directive
5254 @cindex integer, 16-byte
5255 @cindex sixteen byte integer
5256 This directive expects zero or more bignums, separated by commas. For each
5257 bignum, it emits a 16-byte integer.
5259 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5260 hence @emph{octa}-word for 16 bytes.
5263 @section @code{.org @var{new-lc} , @var{fill}}
5265 @cindex @code{org} directive
5266 @cindex location counter, advancing
5267 @cindex advancing location counter
5268 @cindex current address, advancing
5269 Advance the location counter of the current section to
5270 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5271 expression with the same section as the current subsection. That is,
5272 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5273 wrong section, the @code{.org} directive is ignored. To be compatible
5274 with former assemblers, if the section of @var{new-lc} is absolute,
5275 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5276 is the same as the current subsection.
5278 @code{.org} may only increase the location counter, or leave it
5279 unchanged; you cannot use @code{.org} to move the location counter
5282 @c double negative used below "not undefined" because this is a specific
5283 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5284 @c section. doc@cygnus.com 18feb91
5285 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5286 may not be undefined. If you really detest this restriction we eagerly await
5287 a chance to share your improved assembler.
5289 Beware that the origin is relative to the start of the section, not
5290 to the start of the subsection. This is compatible with other
5291 people's assemblers.
5293 When the location counter (of the current subsection) is advanced, the
5294 intervening bytes are filled with @var{fill} which should be an
5295 absolute expression. If the comma and @var{fill} are omitted,
5296 @var{fill} defaults to zero.
5299 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5301 @cindex padding the location counter given a power of two
5302 @cindex @code{p2align} directive
5303 Pad the location counter (in the current subsection) to a particular
5304 storage boundary. The first expression (which must be absolute) is the
5305 number of low-order zero bits the location counter must have after
5306 advancement. For example @samp{.p2align 3} advances the location
5307 counter until it a multiple of 8. If the location counter is already a
5308 multiple of 8, no change is needed.
5310 The second expression (also absolute) gives the fill value to be stored in the
5311 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5312 padding bytes are normally zero. However, on some systems, if the section is
5313 marked as containing code and the fill value is omitted, the space is filled
5314 with no-op instructions.
5316 The third expression is also absolute, and is also optional. If it is present,
5317 it is the maximum number of bytes that should be skipped by this alignment
5318 directive. If doing the alignment would require skipping more bytes than the
5319 specified maximum, then the alignment is not done at all. You can omit the
5320 fill value (the second argument) entirely by simply using two commas after the
5321 required alignment; this can be useful if you want the alignment to be filled
5322 with no-op instructions when appropriate.
5324 @cindex @code{p2alignw} directive
5325 @cindex @code{p2alignl} directive
5326 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5327 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5328 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5329 fill pattern as a four byte longword value. For example, @code{.p2alignw
5330 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5331 filled in with the value 0x368d (the exact placement of the bytes depends upon
5332 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5337 @section @code{.previous}
5339 @cindex @code{previous} directive
5340 @cindex Section Stack
5341 This is one of the ELF section stack manipulation directives. The others are
5342 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5343 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5344 (@pxref{PopSection}).
5346 This directive swaps the current section (and subsection) with most recently
5347 referenced section (and subsection) prior to this one. Multiple
5348 @code{.previous} directives in a row will flip between two sections (and their
5351 In terms of the section stack, this directive swaps the current section with
5352 the top section on the section stack.
5357 @section @code{.popsection}
5359 @cindex @code{popsection} directive
5360 @cindex Section Stack
5361 This is one of the ELF section stack manipulation directives. The others are
5362 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5363 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5366 This directive replaces the current section (and subsection) with the top
5367 section (and subsection) on the section stack. This section is popped off the
5372 @section @code{.print @var{string}}
5374 @cindex @code{print} directive
5375 @command{@value{AS}} will print @var{string} on the standard output during
5376 assembly. You must put @var{string} in double quotes.
5380 @section @code{.protected @var{names}}
5382 @cindex @code{protected} directive
5384 This is one of the ELF visibility directives. The other two are
5385 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5387 This directive overrides the named symbols default visibility (which is set by
5388 their binding: local, global or weak). The directive sets the visibility to
5389 @code{protected} which means that any references to the symbols from within the
5390 components that defines them must be resolved to the definition in that
5391 component, even if a definition in another component would normally preempt
5396 @section @code{.psize @var{lines} , @var{columns}}
5398 @cindex @code{psize} directive
5399 @cindex listing control: paper size
5400 @cindex paper size, for listings
5401 Use this directive to declare the number of lines---and, optionally, the
5402 number of columns---to use for each page, when generating listings.
5404 If you do not use @code{.psize}, listings use a default line-count
5405 of 60. You may omit the comma and @var{columns} specification; the
5406 default width is 200 columns.
5408 @command{@value{AS}} generates formfeeds whenever the specified number of
5409 lines is exceeded (or whenever you explicitly request one, using
5412 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5413 those explicitly specified with @code{.eject}.
5416 @section @code{.purgem @var{name}}
5418 @cindex @code{purgem} directive
5419 Undefine the macro @var{name}, so that later uses of the string will not be
5420 expanded. @xref{Macro}.
5424 @section @code{.pushsection @var{name} , @var{subsection}}
5426 @cindex @code{pushsection} directive
5427 @cindex Section Stack
5428 This is one of the ELF section stack manipulation directives. The others are
5429 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5430 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5433 This directive pushes the current section (and subsection) onto the
5434 top of the section stack, and then replaces the current section and
5435 subsection with @code{name} and @code{subsection}.
5439 @section @code{.quad @var{bignums}}
5441 @cindex @code{quad} directive
5442 @code{.quad} expects zero or more bignums, separated by commas. For
5443 each bignum, it emits
5445 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5446 warning message; and just takes the lowest order 8 bytes of the bignum.
5447 @cindex eight-byte integer
5448 @cindex integer, 8-byte
5450 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5451 hence @emph{quad}-word for 8 bytes.
5454 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5455 warning message; and just takes the lowest order 16 bytes of the bignum.
5456 @cindex sixteen-byte integer
5457 @cindex integer, 16-byte
5461 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
5463 @cindex @code{reloc} directive
5464 Generate a relocation at @var{offset} of type @var{reloc_name} with value
5465 @var{expression}. If @var{offset} is a number, the relocation is generated in
5466 the current section. If @var{offset} is an expression that resolves to a
5467 symbol plus offset, the relocation is generated in the given symbol's section.
5468 @var{expression}, if present, must resolve to a symbol plus addend or to an
5469 absolute value, but note that not all targets support an addend. e.g. ELF REL
5470 targets such as i386 store an addend in the section contents rather than in the
5471 relocation. This low level interface does not support addends stored in the
5475 @section @code{.rept @var{count}}
5477 @cindex @code{rept} directive
5478 Repeat the sequence of lines between the @code{.rept} directive and the next
5479 @code{.endr} directive @var{count} times.
5481 For example, assembling
5489 is equivalent to assembling
5498 @section @code{.sbttl "@var{subheading}"}
5500 @cindex @code{sbttl} directive
5501 @cindex subtitles for listings
5502 @cindex listing control: subtitle
5503 Use @var{subheading} as the title (third line, immediately after the
5504 title line) when generating assembly listings.
5506 This directive affects subsequent pages, as well as the current page if
5507 it appears within ten lines of the top of a page.
5511 @section @code{.scl @var{class}}
5513 @cindex @code{scl} directive
5514 @cindex symbol storage class (COFF)
5515 @cindex COFF symbol storage class
5516 Set the storage-class value for a symbol. This directive may only be
5517 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5518 whether a symbol is static or external, or it may record further
5519 symbolic debugging information.
5522 The @samp{.scl} directive is primarily associated with COFF output; when
5523 configured to generate @code{b.out} output format, @command{@value{AS}}
5524 accepts this directive but ignores it.
5530 @section @code{.section @var{name}}
5532 @cindex named section
5533 Use the @code{.section} directive to assemble the following code into a section
5536 This directive is only supported for targets that actually support arbitrarily
5537 named sections; on @code{a.out} targets, for example, it is not accepted, even
5538 with a standard @code{a.out} section name.
5542 @c only print the extra heading if both COFF and ELF are set
5543 @subheading COFF Version
5546 @cindex @code{section} directive (COFF version)
5547 For COFF targets, the @code{.section} directive is used in one of the following
5551 .section @var{name}[, "@var{flags}"]
5552 .section @var{name}[, @var{subsegment}]
5555 If the optional argument is quoted, it is taken as flags to use for the
5556 section. Each flag is a single character. The following flags are recognized:
5559 bss section (uninitialized data)
5561 section is not loaded
5571 shared section (meaningful for PE targets)
5573 ignored. (For compatibility with the ELF version)
5576 If no flags are specified, the default flags depend upon the section name. If
5577 the section name is not recognized, the default will be for the section to be
5578 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5579 from the section, rather than adding them, so if they are used on their own it
5580 will be as if no flags had been specified at all.
5582 If the optional argument to the @code{.section} directive is not quoted, it is
5583 taken as a subsegment number (@pxref{Sub-Sections}).
5588 @c only print the extra heading if both COFF and ELF are set
5589 @subheading ELF Version
5592 @cindex Section Stack
5593 This is one of the ELF section stack manipulation directives. The others are
5594 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5595 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5596 @code{.previous} (@pxref{Previous}).
5598 @cindex @code{section} directive (ELF version)
5599 For ELF targets, the @code{.section} directive is used like this:
5602 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
5605 The optional @var{flags} argument is a quoted string which may contain any
5606 combination of the following characters:
5609 section is allocatable
5613 section is executable
5615 section is mergeable
5617 section contains zero terminated strings
5619 section is a member of a section group
5621 section is used for thread-local-storage
5624 The optional @var{type} argument may contain one of the following constants:
5627 section contains data
5629 section does not contain data (i.e., section only occupies space)
5631 section contains data which is used by things other than the program
5633 section contains an array of pointers to init functions
5635 section contains an array of pointers to finish functions
5636 @item @@preinit_array
5637 section contains an array of pointers to pre-init functions
5640 Many targets only support the first three section types.
5642 Note on targets where the @code{@@} character is the start of a comment (eg
5643 ARM) then another character is used instead. For example the ARM port uses the
5646 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
5647 be specified as well as an extra argument---@var{entsize}---like this:
5650 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
5653 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
5654 constants, each @var{entsize} octets long. Sections with both @code{M} and
5655 @code{S} must contain zero terminated strings where each character is
5656 @var{entsize} bytes long. The linker may remove duplicates within sections with
5657 the same name, same entity size and same flags. @var{entsize} must be an
5658 absolute expression.
5660 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
5661 be present along with an additional field like this:
5664 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
5667 The @var{GroupName} field specifies the name of the section group to which this
5668 particular section belongs. The optional linkage field can contain:
5671 indicates that only one copy of this section should be retained
5676 Note: if both the @var{M} and @var{G} flags are present then the fields for
5677 the Merge flag should come first, like this:
5680 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
5683 If no flags are specified, the default flags depend upon the section name. If
5684 the section name is not recognized, the default will be for the section to have
5685 none of the above flags: it will not be allocated in memory, nor writable, nor
5686 executable. The section will contain data.
5688 For ELF targets, the assembler supports another type of @code{.section}
5689 directive for compatibility with the Solaris assembler:
5692 .section "@var{name}"[, @var{flags}...]
5695 Note that the section name is quoted. There may be a sequence of comma
5699 section is allocatable
5703 section is executable
5705 section is used for thread local storage
5708 This directive replaces the current section and subsection. See the
5709 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
5710 some examples of how this directive and the other section stack directives
5716 @section @code{.set @var{symbol}, @var{expression}}
5718 @cindex @code{set} directive
5719 @cindex symbol value, setting
5720 Set the value of @var{symbol} to @var{expression}. This
5721 changes @var{symbol}'s value and type to conform to
5722 @var{expression}. If @var{symbol} was flagged as external, it remains
5723 flagged (@pxref{Symbol Attributes}).
5725 You may @code{.set} a symbol many times in the same assembly.
5727 If you @code{.set} a global symbol, the value stored in the object
5728 file is the last value stored into it.
5731 The syntax for @code{set} on the HPPA is
5732 @samp{@var{symbol} .set @var{expression}}.
5736 On Z80 @code{set} is a real instruction, use
5737 @samp{@var{symbol} defl @var{expression}} instead.
5741 @section @code{.short @var{expressions}}
5743 @cindex @code{short} directive
5745 @code{.short} is normally the same as @samp{.word}.
5746 @xref{Word,,@code{.word}}.
5748 In some configurations, however, @code{.short} and @code{.word} generate
5749 numbers of different lengths. @xref{Machine Dependencies}.
5753 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
5756 This expects zero or more @var{expressions}, and emits
5757 a 16 bit number for each.
5762 @section @code{.single @var{flonums}}
5764 @cindex @code{single} directive
5765 @cindex floating point numbers (single)
5766 This directive assembles zero or more flonums, separated by commas. It
5767 has the same effect as @code{.float}.
5769 The exact kind of floating point numbers emitted depends on how
5770 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5774 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
5775 numbers in @sc{ieee} format.
5781 @section @code{.size}
5783 This directive is used to set the size associated with a symbol.
5787 @c only print the extra heading if both COFF and ELF are set
5788 @subheading COFF Version
5791 @cindex @code{size} directive (COFF version)
5792 For COFF targets, the @code{.size} directive is only permitted inside
5793 @code{.def}/@code{.endef} pairs. It is used like this:
5796 .size @var{expression}
5800 @samp{.size} is only meaningful when generating COFF format output; when
5801 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
5808 @c only print the extra heading if both COFF and ELF are set
5809 @subheading ELF Version
5812 @cindex @code{size} directive (ELF version)
5813 For ELF targets, the @code{.size} directive is used like this:
5816 .size @var{name} , @var{expression}
5819 This directive sets the size associated with a symbol @var{name}.
5820 The size in bytes is computed from @var{expression} which can make use of label
5821 arithmetic. This directive is typically used to set the size of function
5827 @section @code{.sleb128 @var{expressions}}
5829 @cindex @code{sleb128} directive
5830 @var{sleb128} stands for ``signed little endian base 128.'' This is a
5831 compact, variable length representation of numbers used by the DWARF
5832 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
5834 @ifclear no-space-dir
5836 @section @code{.skip @var{size} , @var{fill}}
5838 @cindex @code{skip} directive
5839 @cindex filling memory
5840 This directive emits @var{size} bytes, each of value @var{fill}. Both
5841 @var{size} and @var{fill} are absolute expressions. If the comma and
5842 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
5846 @section @code{.space @var{size} , @var{fill}}
5848 @cindex @code{space} directive
5849 @cindex filling memory
5850 This directive emits @var{size} bytes, each of value @var{fill}. Both
5851 @var{size} and @var{fill} are absolute expressions. If the comma
5852 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
5857 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
5858 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
5859 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
5860 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
5868 @section @code{.stabd, .stabn, .stabs}
5870 @cindex symbolic debuggers, information for
5871 @cindex @code{stab@var{x}} directives
5872 There are three directives that begin @samp{.stab}.
5873 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5874 The symbols are not entered in the @command{@value{AS}} hash table: they
5875 cannot be referenced elsewhere in the source file.
5876 Up to five fields are required:
5880 This is the symbol's name. It may contain any character except
5881 @samp{\000}, so is more general than ordinary symbol names. Some
5882 debuggers used to code arbitrarily complex structures into symbol names
5886 An absolute expression. The symbol's type is set to the low 8 bits of
5887 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5888 and debuggers choke on silly bit patterns.
5891 An absolute expression. The symbol's ``other'' attribute is set to the
5892 low 8 bits of this expression.
5895 An absolute expression. The symbol's descriptor is set to the low 16
5896 bits of this expression.
5899 An absolute expression which becomes the symbol's value.
5902 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5903 or @code{.stabs} statement, the symbol has probably already been created;
5904 you get a half-formed symbol in your object file. This is
5905 compatible with earlier assemblers!
5908 @cindex @code{stabd} directive
5909 @item .stabd @var{type} , @var{other} , @var{desc}
5911 The ``name'' of the symbol generated is not even an empty string.
5912 It is a null pointer, for compatibility. Older assemblers used a
5913 null pointer so they didn't waste space in object files with empty
5916 The symbol's value is set to the location counter,
5917 relocatably. When your program is linked, the value of this symbol
5918 is the address of the location counter when the @code{.stabd} was
5921 @cindex @code{stabn} directive
5922 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5923 The name of the symbol is set to the empty string @code{""}.
5925 @cindex @code{stabs} directive
5926 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5927 All five fields are specified.
5933 @section @code{.string} "@var{str}"
5935 @cindex string, copying to object file
5936 @cindex @code{string} directive
5938 Copy the characters in @var{str} to the object file. You may specify more than
5939 one string to copy, separated by commas. Unless otherwise specified for a
5940 particular machine, the assembler marks the end of each string with a 0 byte.
5941 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5944 @section @code{.struct @var{expression}}
5946 @cindex @code{struct} directive
5947 Switch to the absolute section, and set the section offset to @var{expression},
5948 which must be an absolute expression. You might use this as follows:
5957 This would define the symbol @code{field1} to have the value 0, the symbol
5958 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5959 value 8. Assembly would be left in the absolute section, and you would need to
5960 use a @code{.section} directive of some sort to change to some other section
5961 before further assembly.
5965 @section @code{.subsection @var{name}}
5967 @cindex @code{subsection} directive
5968 @cindex Section Stack
5969 This is one of the ELF section stack manipulation directives. The others are
5970 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5971 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5974 This directive replaces the current subsection with @code{name}. The current
5975 section is not changed. The replaced subsection is put onto the section stack
5976 in place of the then current top of stack subsection.
5981 @section @code{.symver}
5982 @cindex @code{symver} directive
5983 @cindex symbol versioning
5984 @cindex versions of symbols
5985 Use the @code{.symver} directive to bind symbols to specific version nodes
5986 within a source file. This is only supported on ELF platforms, and is
5987 typically used when assembling files to be linked into a shared library.
5988 There are cases where it may make sense to use this in objects to be bound
5989 into an application itself so as to override a versioned symbol from a
5992 For ELF targets, the @code{.symver} directive can be used like this:
5994 .symver @var{name}, @var{name2@@nodename}
5996 If the symbol @var{name} is defined within the file
5997 being assembled, the @code{.symver} directive effectively creates a symbol
5998 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5999 just don't try and create a regular alias is that the @var{@@} character isn't
6000 permitted in symbol names. The @var{name2} part of the name is the actual name
6001 of the symbol by which it will be externally referenced. The name @var{name}
6002 itself is merely a name of convenience that is used so that it is possible to
6003 have definitions for multiple versions of a function within a single source
6004 file, and so that the compiler can unambiguously know which version of a
6005 function is being mentioned. The @var{nodename} portion of the alias should be
6006 the name of a node specified in the version script supplied to the linker when
6007 building a shared library. If you are attempting to override a versioned
6008 symbol from a shared library, then @var{nodename} should correspond to the
6009 nodename of the symbol you are trying to override.
6011 If the symbol @var{name} is not defined within the file being assembled, all
6012 references to @var{name} will be changed to @var{name2@@nodename}. If no
6013 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6016 Another usage of the @code{.symver} directive is:
6018 .symver @var{name}, @var{name2@@@@nodename}
6020 In this case, the symbol @var{name} must exist and be defined within
6021 the file being assembled. It is similar to @var{name2@@nodename}. The
6022 difference is @var{name2@@@@nodename} will also be used to resolve
6023 references to @var{name2} by the linker.
6025 The third usage of the @code{.symver} directive is:
6027 .symver @var{name}, @var{name2@@@@@@nodename}
6029 When @var{name} is not defined within the
6030 file being assembled, it is treated as @var{name2@@nodename}. When
6031 @var{name} is defined within the file being assembled, the symbol
6032 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6037 @section @code{.tag @var{structname}}
6039 @cindex COFF structure debugging
6040 @cindex structure debugging, COFF
6041 @cindex @code{tag} directive
6042 This directive is generated by compilers to include auxiliary debugging
6043 information in the symbol table. It is only permitted inside
6044 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6045 definitions in the symbol table with instances of those structures.
6048 @samp{.tag} is only used when generating COFF format output; when
6049 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6055 @section @code{.text @var{subsection}}
6057 @cindex @code{text} directive
6058 Tells @command{@value{AS}} to assemble the following statements onto the end of
6059 the text subsection numbered @var{subsection}, which is an absolute
6060 expression. If @var{subsection} is omitted, subsection number zero
6064 @section @code{.title "@var{heading}"}
6066 @cindex @code{title} directive
6067 @cindex listing control: title line
6068 Use @var{heading} as the title (second line, immediately after the
6069 source file name and pagenumber) when generating assembly listings.
6071 This directive affects subsequent pages, as well as the current page if
6072 it appears within ten lines of the top of a page.
6076 @section @code{.type}
6078 This directive is used to set the type of a symbol.
6082 @c only print the extra heading if both COFF and ELF are set
6083 @subheading COFF Version
6086 @cindex COFF symbol type
6087 @cindex symbol type, COFF
6088 @cindex @code{type} directive (COFF version)
6089 For COFF targets, this directive is permitted only within
6090 @code{.def}/@code{.endef} pairs. It is used like this:
6096 This records the integer @var{int} as the type attribute of a symbol table
6100 @samp{.type} is associated only with COFF format output; when
6101 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6102 directive but ignores it.
6108 @c only print the extra heading if both COFF and ELF are set
6109 @subheading ELF Version
6112 @cindex ELF symbol type
6113 @cindex symbol type, ELF
6114 @cindex @code{type} directive (ELF version)
6115 For ELF targets, the @code{.type} directive is used like this:
6118 .type @var{name} , @var{type description}
6121 This sets the type of symbol @var{name} to be either a
6122 function symbol or an object symbol. There are five different syntaxes
6123 supported for the @var{type description} field, in order to provide
6124 compatibility with various other assemblers.
6126 Because some of the characters used in these syntaxes (such as @samp{@@} and
6127 @samp{#}) are comment characters for some architectures, some of the syntaxes
6128 below do not work on all architectures. The first variant will be accepted by
6129 the GNU assembler on all architectures so that variant should be used for
6130 maximum portability, if you do not need to assemble your code with other
6133 The syntaxes supported are:
6136 .type <name> STT_FUNCTION
6137 .type <name> STT_OBJECT
6139 .type <name>,#function
6140 .type <name>,#object
6142 .type <name>,@@function
6143 .type <name>,@@object
6145 .type <name>,%function
6146 .type <name>,%object
6148 .type <name>,"function"
6149 .type <name>,"object"
6155 @section @code{.uleb128 @var{expressions}}
6157 @cindex @code{uleb128} directive
6158 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6159 compact, variable length representation of numbers used by the DWARF
6160 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
6164 @section @code{.val @var{addr}}
6166 @cindex @code{val} directive
6167 @cindex COFF value attribute
6168 @cindex value attribute, COFF
6169 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6170 records the address @var{addr} as the value attribute of a symbol table
6174 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6175 configured for @code{b.out}, it accepts this directive but ignores it.
6181 @section @code{.version "@var{string}"}
6183 @cindex @code{version} directive
6184 This directive creates a @code{.note} section and places into it an ELF
6185 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6190 @section @code{.vtable_entry @var{table}, @var{offset}}
6192 @cindex @code{vtable_entry} directive
6193 This directive finds or creates a symbol @code{table} and creates a
6194 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6197 @section @code{.vtable_inherit @var{child}, @var{parent}}
6199 @cindex @code{vtable_inherit} directive
6200 This directive finds the symbol @code{child} and finds or creates the symbol
6201 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6202 parent whose addend is the value of the child symbol. As a special case the
6203 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
6207 @section @code{.warning "@var{string}"}
6208 @cindex warning directive
6209 Similar to the directive @code{.error}
6210 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6213 @section @code{.weak @var{names}}
6215 @cindex @code{weak} directive
6216 This directive sets the weak attribute on the comma separated list of symbol
6217 @code{names}. If the symbols do not already exist, they will be created.
6219 On COFF targets other than PE, weak symbols are a GNU extension. This
6220 directive sets the weak attribute on the comma separated list of symbol
6221 @code{names}. If the symbols do not already exist, they will be created.
6223 On the PE target, weak symbols are supported natively as weak aliases.
6224 When a weak symbol is created that is not an alias, GAS creates an
6225 alternate symbol to hold the default value.
6228 @section @code{.weakref @var{alias}, @var{target}}
6230 @cindex @code{weakref} directive
6231 This directive creates an alias to the target symbol that enables the symbol to
6232 be referenced with weak-symbol semantics, but without actually making it weak.
6233 If direct references or definitions of the symbol are present, then the symbol
6234 will not be weak, but if all references to it are through weak references, the
6235 symbol will be marked as weak in the symbol table.
6237 The effect is equivalent to moving all references to the alias to a separate
6238 assembly source file, renaming the alias to the symbol in it, declaring the
6239 symbol as weak there, and running a reloadable link to merge the object files
6240 resulting from the assembly of the new source file and the old source file that
6241 had the references to the alias removed.
6243 The alias itself never makes to the symbol table, and is entirely handled
6244 within the assembler.
6247 @section @code{.word @var{expressions}}
6249 @cindex @code{word} directive
6250 This directive expects zero or more @var{expressions}, of any section,
6251 separated by commas.
6254 For each expression, @command{@value{AS}} emits a 32-bit number.
6257 For each expression, @command{@value{AS}} emits a 16-bit number.
6262 The size of the number emitted, and its byte order,
6263 depend on what target computer the assembly is for.
6266 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6267 @c happen---32-bit addressability, period; no long/short jumps.
6268 @ifset DIFF-TBL-KLUGE
6269 @cindex difference tables altered
6270 @cindex altered difference tables
6272 @emph{Warning: Special Treatment to support Compilers}
6276 Machines with a 32-bit address space, but that do less than 32-bit
6277 addressing, require the following special treatment. If the machine of
6278 interest to you does 32-bit addressing (or doesn't require it;
6279 @pxref{Machine Dependencies}), you can ignore this issue.
6282 In order to assemble compiler output into something that works,
6283 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6284 Directives of the form @samp{.word sym1-sym2} are often emitted by
6285 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6286 directive of the form @samp{.word sym1-sym2}, and the difference between
6287 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6288 creates a @dfn{secondary jump table}, immediately before the next label.
6289 This secondary jump table is preceded by a short-jump to the
6290 first byte after the secondary table. This short-jump prevents the flow
6291 of control from accidentally falling into the new table. Inside the
6292 table is a long-jump to @code{sym2}. The original @samp{.word}
6293 contains @code{sym1} minus the address of the long-jump to
6296 If there were several occurrences of @samp{.word sym1-sym2} before the
6297 secondary jump table, all of them are adjusted. If there was a
6298 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6299 long-jump to @code{sym4} is included in the secondary jump table,
6300 and the @code{.word} directives are adjusted to contain @code{sym3}
6301 minus the address of the long-jump to @code{sym4}; and so on, for as many
6302 entries in the original jump table as necessary.
6305 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6306 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6307 assembly language programmers.
6310 @c end DIFF-TBL-KLUGE
6313 @section Deprecated Directives
6315 @cindex deprecated directives
6316 @cindex obsolescent directives
6317 One day these directives won't work.
6318 They are included for compatibility with older assemblers.
6325 @node Machine Dependencies
6326 @chapter Machine Dependent Features
6328 @cindex machine dependencies
6329 The machine instruction sets are (almost by definition) different on
6330 each machine where @command{@value{AS}} runs. Floating point representations
6331 vary as well, and @command{@value{AS}} often supports a few additional
6332 directives or command-line options for compatibility with other
6333 assemblers on a particular platform. Finally, some versions of
6334 @command{@value{AS}} support special pseudo-instructions for branch
6337 This chapter discusses most of these differences, though it does not
6338 include details on any machine's instruction set. For details on that
6339 subject, see the hardware manufacturer's manual.
6343 * Alpha-Dependent:: Alpha Dependent Features
6346 * ARC-Dependent:: ARC Dependent Features
6349 * ARM-Dependent:: ARM Dependent Features
6352 * AVR-Dependent:: AVR Dependent Features
6355 * BFIN-Dependent:: BFIN Dependent Features
6358 * CR16-Dependent:: CR16 Dependent Features
6361 * CRIS-Dependent:: CRIS Dependent Features
6364 * D10V-Dependent:: D10V Dependent Features
6367 * D30V-Dependent:: D30V Dependent Features
6370 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6373 * HPPA-Dependent:: HPPA Dependent Features
6376 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6379 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6382 * i860-Dependent:: Intel 80860 Dependent Features
6385 * i960-Dependent:: Intel 80960 Dependent Features
6388 * IA-64-Dependent:: Intel IA-64 Dependent Features
6391 * IP2K-Dependent:: IP2K Dependent Features
6394 * M32C-Dependent:: M32C Dependent Features
6397 * M32R-Dependent:: M32R Dependent Features
6400 * M68K-Dependent:: M680x0 Dependent Features
6403 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6406 * MIPS-Dependent:: MIPS Dependent Features
6409 * MMIX-Dependent:: MMIX Dependent Features
6412 * MSP430-Dependent:: MSP430 Dependent Features
6415 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6416 * SH64-Dependent:: SuperH SH64 Dependent Features
6419 * PDP-11-Dependent:: PDP-11 Dependent Features
6422 * PJ-Dependent:: picoJava Dependent Features
6425 * PPC-Dependent:: PowerPC Dependent Features
6428 * Sparc-Dependent:: SPARC Dependent Features
6431 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
6434 * V850-Dependent:: V850 Dependent Features
6437 * Xtensa-Dependent:: Xtensa Dependent Features
6440 * Z80-Dependent:: Z80 Dependent Features
6443 * Z8000-Dependent:: Z8000 Dependent Features
6446 * Vax-Dependent:: VAX Dependent Features
6453 @c The following major nodes are *sections* in the GENERIC version, *chapters*
6454 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
6455 @c peculiarity: to preserve cross-references, there must be a node called
6456 @c "Machine Dependencies". Hence the conditional nodenames in each
6457 @c major node below. Node defaulting in makeinfo requires adjacency of
6458 @c node and sectioning commands; hence the repetition of @chapter BLAH
6459 @c in both conditional blocks.
6462 @include c-alpha.texi
6478 @include c-bfin.texi
6482 @include c-cr16.texi
6486 @include c-cris.texi
6491 @node Machine Dependencies
6492 @chapter Machine Dependent Features
6494 The machine instruction sets are different on each Renesas chip family,
6495 and there are also some syntax differences among the families. This
6496 chapter describes the specific @command{@value{AS}} features for each
6500 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6501 * SH-Dependent:: Renesas SH Dependent Features
6508 @include c-d10v.texi
6512 @include c-d30v.texi
6516 @include c-h8300.texi
6520 @include c-hppa.texi
6524 @include c-i370.texi
6528 @include c-i386.texi
6532 @include c-i860.texi
6536 @include c-i960.texi
6540 @include c-ia64.texi
6544 @include c-ip2k.texi
6548 @include c-m32c.texi
6552 @include c-m32r.texi
6556 @include c-m68k.texi
6560 @include c-m68hc11.texi
6564 @include c-mips.texi
6568 @include c-mmix.texi
6572 @include c-msp430.texi
6576 @include c-ns32k.texi
6580 @include c-pdp11.texi
6593 @include c-sh64.texi
6597 @include c-sparc.texi
6601 @include c-tic54x.texi
6617 @include c-v850.texi
6621 @include c-xtensa.texi
6625 @c reverse effect of @down at top of generic Machine-Dep chapter
6629 @node Reporting Bugs
6630 @chapter Reporting Bugs
6631 @cindex bugs in assembler
6632 @cindex reporting bugs in assembler
6634 Your bug reports play an essential role in making @command{@value{AS}} reliable.
6636 Reporting a bug may help you by bringing a solution to your problem, or it may
6637 not. But in any case the principal function of a bug report is to help the
6638 entire community by making the next version of @command{@value{AS}} work better.
6639 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
6641 In order for a bug report to serve its purpose, you must include the
6642 information that enables us to fix the bug.
6645 * Bug Criteria:: Have you found a bug?
6646 * Bug Reporting:: How to report bugs
6650 @section Have You Found a Bug?
6651 @cindex bug criteria
6653 If you are not sure whether you have found a bug, here are some guidelines:
6656 @cindex fatal signal
6657 @cindex assembler crash
6658 @cindex crash of assembler
6660 If the assembler gets a fatal signal, for any input whatever, that is a
6661 @command{@value{AS}} bug. Reliable assemblers never crash.
6663 @cindex error on valid input
6665 If @command{@value{AS}} produces an error message for valid input, that is a bug.
6667 @cindex invalid input
6669 If @command{@value{AS}} does not produce an error message for invalid input, that
6670 is a bug. However, you should note that your idea of ``invalid input'' might
6671 be our idea of ``an extension'' or ``support for traditional practice''.
6674 If you are an experienced user of assemblers, your suggestions for improvement
6675 of @command{@value{AS}} are welcome in any case.
6679 @section How to Report Bugs
6681 @cindex assembler bugs, reporting
6683 A number of companies and individuals offer support for @sc{gnu} products. If
6684 you obtained @command{@value{AS}} from a support organization, we recommend you
6685 contact that organization first.
6687 You can find contact information for many support companies and
6688 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
6692 In any event, we also recommend that you send bug reports for @command{@value{AS}}
6696 The fundamental principle of reporting bugs usefully is this:
6697 @strong{report all the facts}. If you are not sure whether to state a
6698 fact or leave it out, state it!
6700 Often people omit facts because they think they know what causes the problem
6701 and assume that some details do not matter. Thus, you might assume that the
6702 name of a symbol you use in an example does not matter. Well, probably it does
6703 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
6704 happens to fetch from the location where that name is stored in memory;
6705 perhaps, if the name were different, the contents of that location would fool
6706 the assembler into doing the right thing despite the bug. Play it safe and
6707 give a specific, complete example. That is the easiest thing for you to do,
6708 and the most helpful.
6710 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
6711 it is new to us. Therefore, always write your bug reports on the assumption
6712 that the bug has not been reported previously.
6714 Sometimes people give a few sketchy facts and ask, ``Does this ring a
6715 bell?'' This cannot help us fix a bug, so it is basically useless. We
6716 respond by asking for enough details to enable us to investigate.
6717 You might as well expedite matters by sending them to begin with.
6719 To enable us to fix the bug, you should include all these things:
6723 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
6724 it with the @samp{--version} argument.
6726 Without this, we will not know whether there is any point in looking for
6727 the bug in the current version of @command{@value{AS}}.
6730 Any patches you may have applied to the @command{@value{AS}} source.
6733 The type of machine you are using, and the operating system name and
6737 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
6741 The command arguments you gave the assembler to assemble your example and
6742 observe the bug. To guarantee you will not omit something important, list them
6743 all. A copy of the Makefile (or the output from make) is sufficient.
6745 If we were to try to guess the arguments, we would probably guess wrong
6746 and then we might not encounter the bug.
6749 A complete input file that will reproduce the bug. If the bug is observed when
6750 the assembler is invoked via a compiler, send the assembler source, not the
6751 high level language source. Most compilers will produce the assembler source
6752 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
6753 the options @samp{-v --save-temps}; this will save the assembler source in a
6754 file with an extension of @file{.s}, and also show you exactly how
6755 @command{@value{AS}} is being run.
6758 A description of what behavior you observe that you believe is
6759 incorrect. For example, ``It gets a fatal signal.''
6761 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
6762 will certainly notice it. But if the bug is incorrect output, we might not
6763 notice unless it is glaringly wrong. You might as well not give us a chance to
6766 Even if the problem you experience is a fatal signal, you should still say so
6767 explicitly. Suppose something strange is going on, such as, your copy of
6768 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
6769 library on your system. (This has happened!) Your copy might crash and ours
6770 would not. If you told us to expect a crash, then when ours fails to crash, we
6771 would know that the bug was not happening for us. If you had not told us to
6772 expect a crash, then we would not be able to draw any conclusion from our
6776 If you wish to suggest changes to the @command{@value{AS}} source, send us context
6777 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
6778 option. Always send diffs from the old file to the new file. If you even
6779 discuss something in the @command{@value{AS}} source, refer to it by context, not
6782 The line numbers in our development sources will not match those in your
6783 sources. Your line numbers would convey no useful information to us.
6786 Here are some things that are not necessary:
6790 A description of the envelope of the bug.
6792 Often people who encounter a bug spend a lot of time investigating
6793 which changes to the input file will make the bug go away and which
6794 changes will not affect it.
6796 This is often time consuming and not very useful, because the way we
6797 will find the bug is by running a single example under the debugger
6798 with breakpoints, not by pure deduction from a series of examples.
6799 We recommend that you save your time for something else.
6801 Of course, if you can find a simpler example to report @emph{instead}
6802 of the original one, that is a convenience for us. Errors in the
6803 output will be easier to spot, running under the debugger will take
6804 less time, and so on.
6806 However, simplification is not vital; if you do not want to do this,
6807 report the bug anyway and send us the entire test case you used.
6810 A patch for the bug.
6812 A patch for the bug does help us if it is a good one. But do not omit
6813 the necessary information, such as the test case, on the assumption that
6814 a patch is all we need. We might see problems with your patch and decide
6815 to fix the problem another way, or we might not understand it at all.
6817 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
6818 construct an example that will make the program follow a certain path through
6819 the code. If you do not send us the example, we will not be able to construct
6820 one, so we will not be able to verify that the bug is fixed.
6822 And if we cannot understand what bug you are trying to fix, or why your
6823 patch should be an improvement, we will not install it. A test case will
6824 help us to understand.
6827 A guess about what the bug is or what it depends on.
6829 Such guesses are usually wrong. Even we cannot guess right about such
6830 things without first using the debugger to find the facts.
6833 @node Acknowledgements
6834 @chapter Acknowledgements
6836 If you have contributed to GAS and your name isn't listed here,
6837 it is not meant as a slight. We just don't know about it. Send mail to the
6838 maintainer, and we'll correct the situation. Currently
6840 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
6842 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
6845 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
6846 information and the 68k series machines, most of the preprocessing pass, and
6847 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
6849 K. Richard Pixley maintained GAS for a while, adding various enhancements and
6850 many bug fixes, including merging support for several processors, breaking GAS
6851 up to handle multiple object file format back ends (including heavy rewrite,
6852 testing, an integration of the coff and b.out back ends), adding configuration
6853 including heavy testing and verification of cross assemblers and file splits
6854 and renaming, converted GAS to strictly ANSI C including full prototypes, added
6855 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
6856 port (including considerable amounts of reverse engineering), a SPARC opcode
6857 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
6858 assertions and made them work, much other reorganization, cleanup, and lint.
6860 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
6861 in format-specific I/O modules.
6863 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
6864 has done much work with it since.
6866 The Intel 80386 machine description was written by Eliot Dresselhaus.
6868 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
6870 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
6871 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
6873 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
6874 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
6875 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
6876 support a.out format.
6878 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
6879 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
6880 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
6881 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
6884 John Gilmore built the AMD 29000 support, added @code{.include} support, and
6885 simplified the configuration of which versions accept which directives. He
6886 updated the 68k machine description so that Motorola's opcodes always produced
6887 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
6888 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
6889 cross-compilation support, and one bug in relaxation that took a week and
6890 required the proverbial one-bit fix.
6892 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
6893 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
6894 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
6895 PowerPC assembler, and made a few other minor patches.
6897 Steve Chamberlain made GAS able to generate listings.
6899 Hewlett-Packard contributed support for the HP9000/300.
6901 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
6902 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
6903 formats). This work was supported by both the Center for Software Science at
6904 the University of Utah and Cygnus Support.
6906 Support for ELF format files has been worked on by Mark Eichin of Cygnus
6907 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
6908 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
6909 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
6910 and some initial 64-bit support).
6912 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
6914 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
6915 support for openVMS/Alpha.
6917 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
6920 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
6921 Inc.@: added support for Xtensa processors.
6923 Several engineers at Cygnus Support have also provided many small bug fixes and
6924 configuration enhancements.
6926 Many others have contributed large or small bugfixes and enhancements. If
6927 you have contributed significant work and are not mentioned on this list, and
6928 want to be, let us know. Some of the history has been lost; we are not
6929 intentionally leaving anyone out.
6934 @unnumbered AS Index
projects / FreeBSD / FreeBSD.git / blob