1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright (C) 1991-2016 Free Software Foundation, Inc.
3 @c UPDATE!! On future updates--
4 @c (1) check for new machine-dep cmdline options in
5 @c md_parse_option definitions in config/tc-*.c
6 @c (2) for platform-specific directives, examine md_pseudo_op
8 @c (3) for object-format specific directives, examine obj_pseudo_op
10 @c (4) portable directives in potable[] in read.c
14 @macro gcctabopt{body}
17 @c defaults, config file may override:
22 @include asconfig.texi
27 @c common OR combinations of conditions
53 @set abnormal-separator
57 @settitle Using @value{AS}
60 @settitle Using @value{AS} (@value{TARGET})
62 @setchapternewpage odd
67 @c WARE! Some of the machine-dependent sections contain tables of machine
68 @c instructions. Except in multi-column format, these tables look silly.
69 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
70 @c the multi-col format is faked within @example sections.
72 @c Again unfortunately, the natural size that fits on a page, for these tables,
73 @c is different depending on whether or not smallbook is turned on.
74 @c This matters, because of order: text flow switches columns at each page
77 @c The format faked in this source works reasonably well for smallbook,
78 @c not well for the default large-page format. This manual expects that if you
79 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
80 @c tables in question. You can turn on one without the other at your
81 @c discretion, of course.
84 @c the insn tables look just as silly in info files regardless of smallbook,
85 @c might as well show 'em anyways.
89 @dircategory Software development
91 * As: (as). The GNU assembler.
92 * Gas: (as). The GNU assembler.
100 This file documents the GNU Assembler "@value{AS}".
102 @c man begin COPYRIGHT
103 Copyright @copyright{} 1991-2016 Free Software Foundation, Inc.
105 Permission is granted to copy, distribute and/or modify this document
106 under the terms of the GNU Free Documentation License, Version 1.3
107 or any later version published by the Free Software Foundation;
108 with no Invariant Sections, with no Front-Cover Texts, and with no
109 Back-Cover Texts. A copy of the license is included in the
110 section entitled ``GNU Free Documentation License''.
116 @title Using @value{AS}
117 @subtitle The @sc{gnu} Assembler
119 @subtitle for the @value{TARGET} family
121 @ifset VERSION_PACKAGE
123 @subtitle @value{VERSION_PACKAGE}
126 @subtitle Version @value{VERSION}
129 The Free Software Foundation Inc.@: thanks The Nice Computer
130 Company of Australia for loaning Dean Elsner to write the
131 first (Vax) version of @command{as} for Project @sc{gnu}.
132 The proprietors, management and staff of TNCCA thank FSF for
133 distracting the boss while they got some work
136 @author Dean Elsner, Jay Fenlason & friends
140 \hfill {\it Using {\tt @value{AS}}}\par
141 \hfill Edited by Cygnus Support\par
143 %"boxit" macro for figures:
144 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
145 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
146 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
147 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
148 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
151 @vskip 0pt plus 1filll
152 Copyright @copyright{} 1991-2016 Free Software Foundation, Inc.
154 Permission is granted to copy, distribute and/or modify this document
155 under the terms of the GNU Free Documentation License, Version 1.3
156 or any later version published by the Free Software Foundation;
157 with no Invariant Sections, with no Front-Cover Texts, and with no
158 Back-Cover Texts. A copy of the license is included in the
159 section entitled ``GNU Free Documentation License''.
166 @top Using @value{AS}
168 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}}
169 @ifset VERSION_PACKAGE
170 @value{VERSION_PACKAGE}
172 version @value{VERSION}.
174 This version of the file describes @command{@value{AS}} configured to generate
175 code for @value{TARGET} architectures.
178 This document is distributed under the terms of the GNU Free
179 Documentation License. A copy of the license is included in the
180 section entitled ``GNU Free Documentation License''.
183 * Overview:: Overview
184 * Invoking:: Command-Line Options
186 * Sections:: Sections and Relocation
188 * Expressions:: Expressions
189 * Pseudo Ops:: Assembler Directives
191 * Object Attributes:: Object Attributes
193 * Machine Dependencies:: Machine Dependent Features
194 * Reporting Bugs:: Reporting Bugs
195 * Acknowledgements:: Who Did What
196 * GNU Free Documentation License:: GNU Free Documentation License
197 * AS Index:: AS Index
204 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
206 This version of the manual describes @command{@value{AS}} configured to generate
207 code for @value{TARGET} architectures.
211 @cindex invocation summary
212 @cindex option summary
213 @cindex summary of options
214 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
215 see @ref{Invoking,,Command-Line Options}.
217 @c man title AS the portable GNU assembler.
221 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
225 @c We don't use deffn and friends for the following because they seem
226 @c to be limited to one line for the header.
228 @c man begin SYNOPSIS
229 @value{AS} [@b{-a}[@b{cdghlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
230 [@b{--compress-debug-sections}] [@b{--nocompress-debug-sections}]
231 [@b{--debug-prefix-map} @var{old}=@var{new}]
232 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
233 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--gdwarf-sections}]
234 [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
235 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
236 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
237 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}]
238 [@b{-o} @var{objfile}] [@b{-R}]
239 [@b{--hash-size}=@var{NUM}] [@b{--reduce-memory-overheads}]
241 [@b{-v}] [@b{-version}] [@b{--version}]
242 [@b{-W}] [@b{--warn}] [@b{--fatal-warnings}] [@b{-w}] [@b{-x}]
243 [@b{-Z}] [@b{@@@var{FILE}}]
244 [@b{--sectname-subst}] [@b{--size-check=[error|warning]}]
245 [@b{--elf-stt-common=[no|yes]}]
246 [@b{--target-help}] [@var{target-options}]
247 [@b{--}|@var{files} @dots{}]
250 @c Target dependent options are listed below. Keep the list sorted.
251 @c Add an empty line for separation.
255 @emph{Target AArch64 options:}
257 [@b{-mabi}=@var{ABI}]
261 @emph{Target Alpha options:}
263 [@b{-mdebug} | @b{-no-mdebug}]
264 [@b{-replace} | @b{-noreplace}]
265 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
266 [@b{-F}] [@b{-32addr}]
270 @emph{Target ARC options:}
271 [@b{-mcpu=@var{cpu}}]
272 [@b{-mA6}|@b{-mARC600}|@b{-mARC601}|@b{-mA7}|@b{-mARC700}|@b{-mEM}|@b{-mHS}]
279 @emph{Target ARM options:}
280 @c Don't document the deprecated options
281 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
282 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
283 [@b{-mfpu}=@var{floating-point-format}]
284 [@b{-mfloat-abi}=@var{abi}]
285 [@b{-meabi}=@var{ver}]
288 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
289 @b{-mapcs-reentrant}]
290 [@b{-mthumb-interwork}] [@b{-k}]
294 @emph{Target Blackfin options:}
295 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
302 @emph{Target CRIS options:}
303 [@b{--underscore} | @b{--no-underscore}]
305 [@b{--emulation=criself} | @b{--emulation=crisaout}]
306 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
307 @c Deprecated -- deliberately not documented.
312 @emph{Target D10V options:}
317 @emph{Target D30V options:}
318 [@b{-O}|@b{-n}|@b{-N}]
322 @emph{Target EPIPHANY options:}
323 [@b{-mepiphany}|@b{-mepiphany16}]
327 @emph{Target H8/300 options:}
331 @c HPPA has no machine-dependent assembler options (yet).
335 @emph{Target i386 options:}
336 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
337 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
341 @emph{Target i960 options:}
342 @c see md_parse_option in tc-i960.c
343 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
345 [@b{-b}] [@b{-no-relax}]
349 @emph{Target IA-64 options:}
350 [@b{-mconstant-gp}|@b{-mauto-pic}]
351 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
353 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
354 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
355 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
356 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
360 @emph{Target IP2K options:}
361 [@b{-mip2022}|@b{-mip2022ext}]
365 @emph{Target M32C options:}
366 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
370 @emph{Target M32R options:}
371 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
376 @emph{Target M680X0 options:}
377 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
381 @emph{Target M68HC11 options:}
382 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}|@b{-mm9s12x}|@b{-mm9s12xg}]
383 [@b{-mshort}|@b{-mlong}]
384 [@b{-mshort-double}|@b{-mlong-double}]
385 [@b{--force-long-branches}] [@b{--short-branches}]
386 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
387 [@b{--print-opcodes}] [@b{--generate-example}]
391 @emph{Target MCORE options:}
392 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
393 [@b{-mcpu=[210|340]}]
397 @emph{Target Meta options:}
398 [@b{-mcpu=@var{cpu}}] [@b{-mfpu=@var{cpu}}] [@b{-mdsp=@var{cpu}}]
401 @emph{Target MICROBLAZE options:}
402 @c MicroBlaze has no machine-dependent assembler options.
406 @emph{Target MIPS options:}
407 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
408 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
409 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
410 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
411 [@b{-mfp64}] [@b{-mgp64}] [@b{-mfpxx}]
412 [@b{-modd-spreg}] [@b{-mno-odd-spreg}]
413 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
414 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
415 [@b{-mips32r3}] [@b{-mips32r5}] [@b{-mips32r6}] [@b{-mips64}] [@b{-mips64r2}]
416 [@b{-mips64r3}] [@b{-mips64r5}] [@b{-mips64r6}]
417 [@b{-construct-floats}] [@b{-no-construct-floats}]
418 [@b{-mnan=@var{encoding}}]
419 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
420 [@b{-mips16}] [@b{-no-mips16}]
421 [@b{-mmicromips}] [@b{-mno-micromips}]
422 [@b{-msmartmips}] [@b{-mno-smartmips}]
423 [@b{-mips3d}] [@b{-no-mips3d}]
424 [@b{-mdmx}] [@b{-no-mdmx}]
425 [@b{-mdsp}] [@b{-mno-dsp}]
426 [@b{-mdspr2}] [@b{-mno-dspr2}]
427 [@b{-mmsa}] [@b{-mno-msa}]
428 [@b{-mxpa}] [@b{-mno-xpa}]
429 [@b{-mmt}] [@b{-mno-mt}]
430 [@b{-mmcu}] [@b{-mno-mcu}]
431 [@b{-minsn32}] [@b{-mno-insn32}]
432 [@b{-mfix7000}] [@b{-mno-fix7000}]
433 [@b{-mfix-rm7000}] [@b{-mno-fix-rm7000}]
434 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
435 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
436 [@b{-mdebug}] [@b{-no-mdebug}]
437 [@b{-mpdr}] [@b{-mno-pdr}]
441 @emph{Target MMIX options:}
442 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
443 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
444 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
445 [@b{--linker-allocated-gregs}]
449 @emph{Target Nios II options:}
450 [@b{-relax-all}] [@b{-relax-section}] [@b{-no-relax}]
455 @emph{Target NDS32 options:}
456 [@b{-EL}] [@b{-EB}] [@b{-O}] [@b{-Os}] [@b{-mcpu=@var{cpu}}]
457 [@b{-misa=@var{isa}}] [@b{-mabi=@var{abi}}] [@b{-mall-ext}]
458 [@b{-m[no-]16-bit}] [@b{-m[no-]perf-ext}] [@b{-m[no-]perf2-ext}]
459 [@b{-m[no-]string-ext}] [@b{-m[no-]dsp-ext}] [@b{-m[no-]mac}] [@b{-m[no-]div}]
460 [@b{-m[no-]audio-isa-ext}] [@b{-m[no-]fpu-sp-ext}] [@b{-m[no-]fpu-dp-ext}]
461 [@b{-m[no-]fpu-fma}] [@b{-mfpu-freg=@var{FREG}}] [@b{-mreduced-regs}]
462 [@b{-mfull-regs}] [@b{-m[no-]dx-regs}] [@b{-mpic}] [@b{-mno-relax}]
467 @emph{Target PDP11 options:}
468 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
469 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
470 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
474 @emph{Target picoJava options:}
479 @emph{Target PowerPC options:}
481 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
482 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
483 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
484 @b{-mbooke}|@b{-mpower4}|@b{-mpwr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
485 @b{-mpower7}|@b{-mpwr7}|@b{-mpower8}|@b{-mpwr8}|@b{-mpower9}|@b{-mpwr9}@b{-ma2}|
486 @b{-mcell}|@b{-mspe}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
487 [@b{-many}] [@b{-maltivec}|@b{-mvsx}|@b{-mhtm}|@b{-mvle}]
488 [@b{-mregnames}|@b{-mno-regnames}]
489 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
490 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
491 [@b{-msolaris}|@b{-mno-solaris}]
492 [@b{-nops=@var{count}}]
496 @emph{Target RL78 options:}
498 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
502 @emph{Target RX options:}
503 [@b{-mlittle-endian}|@b{-mbig-endian}]
504 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
505 [@b{-muse-conventional-section-names}]
506 [@b{-msmall-data-limit}]
509 [@b{-mint-register=@var{number}}]
510 [@b{-mgcc-abi}|@b{-mrx-abi}]
514 @emph{Target s390 options:}
515 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
516 [@b{-mregnames}|@b{-mno-regnames}]
517 [@b{-mwarn-areg-zero}]
521 @emph{Target SCORE options:}
522 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
523 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
524 [@b{-march=score7}][@b{-march=score3}]
525 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
529 @emph{Target SPARC options:}
530 @c The order here is important. See c-sparc.texi.
531 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
532 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
533 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
538 @emph{Target TIC54X options:}
539 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
540 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
544 @emph{Target TIC6X options:}
545 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
546 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
547 [@b{-mpic}|@b{-mno-pic}]
551 @emph{Target TILE-Gx options:}
552 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
555 @c TILEPro has no machine-dependent assembler options
559 @emph{Target Visium options:}
560 [@b{-mtune=@var{arch}}]
564 @emph{Target Xtensa options:}
565 [@b{--[no-]text-section-literals}] [@b{--[no-]auto-litpools}]
566 [@b{--[no-]absolute-literals}]
567 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
568 [@b{--[no-]transform}]
569 [@b{--rename-section} @var{oldname}=@var{newname}]
570 [@b{--[no-]trampolines}]
574 @emph{Target Z80 options:}
575 [@b{-z80}] [@b{-r800}]
576 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
577 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
578 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
579 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
580 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
581 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
585 @c Z8000 has no machine-dependent assembler options
594 @include at-file.texi
597 Turn on listings, in any of a variety of ways:
601 omit false conditionals
604 omit debugging directives
607 include general information, like @value{AS} version and options passed
610 include high-level source
616 include macro expansions
619 omit forms processing
625 set the name of the listing file
628 You may combine these options; for example, use @samp{-aln} for assembly
629 listing without forms processing. The @samp{=file} option, if used, must be
630 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
633 Begin in alternate macro mode.
635 @xref{Altmacro,,@code{.altmacro}}.
638 @item --compress-debug-sections
639 Compress DWARF debug sections using zlib with SHF_COMPRESSED from the
640 ELF ABI. The resulting object file may not be compatible with older
641 linkers and object file utilities. Note if compression would make a
642 given section @emph{larger} then it is not compressed.
645 @cindex @samp{--compress-debug-sections=} option
646 @item --compress-debug-sections=none
647 @itemx --compress-debug-sections=zlib
648 @itemx --compress-debug-sections=zlib-gnu
649 @itemx --compress-debug-sections=zlib-gabi
650 These options control how DWARF debug sections are compressed.
651 @option{--compress-debug-sections=none} is equivalent to
652 @option{--nocompress-debug-sections}.
653 @option{--compress-debug-sections=zlib} and
654 @option{--compress-debug-sections=zlib-gabi} are equivalent to
655 @option{--compress-debug-sections}.
656 @option{--compress-debug-sections=zlib-gnu} compresses DWARF debug
657 sections using zlib. The debug sections are renamed to begin with
658 @samp{.zdebug}. Note if compression would make a given section
659 @emph{larger} then it is not compressed nor renamed.
663 @item --nocompress-debug-sections
664 Do not compress DWARF debug sections. This is usually the default for all
665 targets except the x86/x86_64, but a configure time option can be used to
669 Ignored. This option is accepted for script compatibility with calls to
672 @item --debug-prefix-map @var{old}=@var{new}
673 When assembling files in directory @file{@var{old}}, record debugging
674 information describing them as in @file{@var{new}} instead.
676 @item --defsym @var{sym}=@var{value}
677 Define the symbol @var{sym} to be @var{value} before assembling the input file.
678 @var{value} must be an integer constant. As in C, a leading @samp{0x}
679 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
680 value. The value of the symbol can be overridden inside a source file via the
681 use of a @code{.set} pseudo-op.
684 ``fast''---skip whitespace and comment preprocessing (assume source is
689 Generate debugging information for each assembler source line using whichever
690 debug format is preferred by the target. This currently means either STABS,
694 Generate stabs debugging information for each assembler line. This
695 may help debugging assembler code, if the debugger can handle it.
698 Generate stabs debugging information for each assembler line, with GNU
699 extensions that probably only gdb can handle, and that could make other
700 debuggers crash or refuse to read your program. This
701 may help debugging assembler code. Currently the only GNU extension is
702 the location of the current working directory at assembling time.
705 Generate DWARF2 debugging information for each assembler line. This
706 may help debugging assembler code, if the debugger can handle it. Note---this
707 option is only supported by some targets, not all of them.
709 @item --gdwarf-sections
710 Instead of creating a .debug_line section, create a series of
711 .debug_line.@var{foo} sections where @var{foo} is the name of the
712 corresponding code section. For example a code section called @var{.text.func}
713 will have its dwarf line number information placed into a section called
714 @var{.debug_line.text.func}. If the code section is just called @var{.text}
715 then debug line section will still be called just @var{.debug_line} without any
719 @item --size-check=error
720 @itemx --size-check=warning
721 Issue an error or warning for invalid ELF .size directive.
723 @item --elf-stt-common=no
724 @itemx --elf-stt-common=yes
725 These options control whether the ELF assembler should generate common
726 symbols with the @code{STT_COMMON} type. The default can be controlled
727 by a configure option @option{--enable-elf-stt-common}.
731 Print a summary of the command line options and exit.
734 Print a summary of all target specific options and exit.
737 Add directory @var{dir} to the search list for @code{.include} directives.
740 Don't warn about signed overflow.
743 @ifclear DIFF-TBL-KLUGE
744 This option is accepted but has no effect on the @value{TARGET} family.
746 @ifset DIFF-TBL-KLUGE
747 Issue warnings when difference tables altered for long displacements.
752 Keep (in the symbol table) local symbols. These symbols start with
753 system-specific local label prefixes, typically @samp{.L} for ELF systems
754 or @samp{L} for traditional a.out systems.
759 @item --listing-lhs-width=@var{number}
760 Set the maximum width, in words, of the output data column for an assembler
761 listing to @var{number}.
763 @item --listing-lhs-width2=@var{number}
764 Set the maximum width, in words, of the output data column for continuation
765 lines in an assembler listing to @var{number}.
767 @item --listing-rhs-width=@var{number}
768 Set the maximum width of an input source line, as displayed in a listing, to
771 @item --listing-cont-lines=@var{number}
772 Set the maximum number of lines printed in a listing for a single line of input
775 @item -o @var{objfile}
776 Name the object-file output from @command{@value{AS}} @var{objfile}.
779 Fold the data section into the text section.
781 @item --hash-size=@var{number}
782 Set the default size of GAS's hash tables to a prime number close to
783 @var{number}. Increasing this value can reduce the length of time it takes the
784 assembler to perform its tasks, at the expense of increasing the assembler's
785 memory requirements. Similarly reducing this value can reduce the memory
786 requirements at the expense of speed.
788 @item --reduce-memory-overheads
789 This option reduces GAS's memory requirements, at the expense of making the
790 assembly processes slower. Currently this switch is a synonym for
791 @samp{--hash-size=4051}, but in the future it may have other effects as well.
794 @item --sectname-subst
795 Honor substitution sequences in section names.
797 @xref{Section Name Substitutions,,@code{.section @var{name}}}.
802 Print the maximum space (in bytes) and total time (in seconds) used by
805 @item --strip-local-absolute
806 Remove local absolute symbols from the outgoing symbol table.
810 Print the @command{as} version.
813 Print the @command{as} version and exit.
817 Suppress warning messages.
819 @item --fatal-warnings
820 Treat warnings as errors.
823 Don't suppress warning messages or treat them as errors.
832 Generate an object file even after errors.
834 @item -- | @var{files} @dots{}
835 Standard input, or source files to assemble.
843 @xref{AArch64 Options}, for the options available when @value{AS} is configured
844 for the 64-bit mode of the ARM Architecture (AArch64).
849 The following options are available when @value{AS} is configured for the
850 64-bit mode of the ARM Architecture (AArch64).
853 @include c-aarch64.texi
854 @c ended inside the included file
862 @xref{Alpha Options}, for the options available when @value{AS} is configured
863 for an Alpha processor.
868 The following options are available when @value{AS} is configured for an Alpha
872 @include c-alpha.texi
873 @c ended inside the included file
880 The following options are available when @value{AS} is configured for an ARC
884 @item -mcpu=@var{cpu}
885 This option selects the core processor variant.
887 Select either big-endian (-EB) or little-endian (-EL) output.
889 Enable Code Density extenssion instructions.
894 The following options are available when @value{AS} is configured for the ARM
898 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
899 Specify which ARM processor variant is the target.
900 @item -march=@var{architecture}[+@var{extension}@dots{}]
901 Specify which ARM architecture variant is used by the target.
902 @item -mfpu=@var{floating-point-format}
903 Select which Floating Point architecture is the target.
904 @item -mfloat-abi=@var{abi}
905 Select which floating point ABI is in use.
907 Enable Thumb only instruction decoding.
908 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
909 Select which procedure calling convention is in use.
911 Select either big-endian (-EB) or little-endian (-EL) output.
912 @item -mthumb-interwork
913 Specify that the code has been generated with interworking between Thumb and
916 Turns on CodeComposer Studio assembly syntax compatibility mode.
918 Specify that PIC code has been generated.
926 @xref{Blackfin Options}, for the options available when @value{AS} is
927 configured for the Blackfin processor family.
932 The following options are available when @value{AS} is configured for
933 the Blackfin processor family.
937 @c ended inside the included file
944 See the info pages for documentation of the CRIS-specific options.
948 The following options are available when @value{AS} is configured for
951 @cindex D10V optimization
952 @cindex optimization, D10V
954 Optimize output by parallelizing instructions.
959 The following options are available when @value{AS} is configured for a D30V
962 @cindex D30V optimization
963 @cindex optimization, D30V
965 Optimize output by parallelizing instructions.
969 Warn when nops are generated.
971 @cindex D30V nops after 32-bit multiply
973 Warn when a nop after a 32-bit multiply instruction is generated.
979 The following options are available when @value{AS} is configured for the
980 Adapteva EPIPHANY series.
983 @xref{Epiphany Options}, for the options available when @value{AS} is
984 configured for an Epiphany processor.
989 The following options are available when @value{AS} is configured for
990 an Epiphany processor.
993 @include c-epiphany.texi
994 @c ended inside the included file
1002 @xref{H8/300 Options}, for the options available when @value{AS} is configured
1003 for an H8/300 processor.
1007 @c man begin OPTIONS
1008 The following options are available when @value{AS} is configured for an H8/300
1011 @c man begin INCLUDE
1012 @include c-h8300.texi
1013 @c ended inside the included file
1021 @xref{i386-Options}, for the options available when @value{AS} is
1022 configured for an i386 processor.
1026 @c man begin OPTIONS
1027 The following options are available when @value{AS} is configured for
1030 @c man begin INCLUDE
1031 @include c-i386.texi
1032 @c ended inside the included file
1037 @c man begin OPTIONS
1039 The following options are available when @value{AS} is configured for the
1040 Intel 80960 processor.
1043 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
1044 Specify which variant of the 960 architecture is the target.
1047 Add code to collect statistics about branches taken.
1050 Do not alter compare-and-branch instructions for long displacements;
1057 The following options are available when @value{AS} is configured for the
1063 Specifies that the extended IP2022 instructions are allowed.
1066 Restores the default behaviour, which restricts the permitted instructions to
1067 just the basic IP2022 ones.
1073 The following options are available when @value{AS} is configured for the
1074 Renesas M32C and M16C processors.
1079 Assemble M32C instructions.
1082 Assemble M16C instructions (the default).
1085 Enable support for link-time relaxations.
1088 Support H'00 style hex constants in addition to 0x00 style.
1094 The following options are available when @value{AS} is configured for the
1095 Renesas M32R (formerly Mitsubishi M32R) series.
1100 Specify which processor in the M32R family is the target. The default
1101 is normally the M32R, but this option changes it to the M32RX.
1103 @item --warn-explicit-parallel-conflicts or --Wp
1104 Produce warning messages when questionable parallel constructs are
1107 @item --no-warn-explicit-parallel-conflicts or --Wnp
1108 Do not produce warning messages when questionable parallel constructs are
1115 The following options are available when @value{AS} is configured for the
1116 Motorola 68000 series.
1121 Shorten references to undefined symbols, to one word instead of two.
1123 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1124 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1125 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1126 Specify what processor in the 68000 family is the target. The default
1127 is normally the 68020, but this can be changed at configuration time.
1129 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1130 The target machine does (or does not) have a floating-point coprocessor.
1131 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1132 the basic 68000 is not compatible with the 68881, a combination of the
1133 two can be specified, since it's possible to do emulation of the
1134 coprocessor instructions with the main processor.
1136 @item -m68851 | -mno-68851
1137 The target machine does (or does not) have a memory-management
1138 unit coprocessor. The default is to assume an MMU for 68020 and up.
1146 @xref{Nios II Options}, for the options available when @value{AS} is configured
1147 for an Altera Nios II processor.
1151 @c man begin OPTIONS
1152 The following options are available when @value{AS} is configured for an
1153 Altera Nios II processor.
1155 @c man begin INCLUDE
1156 @include c-nios2.texi
1157 @c ended inside the included file
1163 For details about the PDP-11 machine dependent features options,
1164 see @ref{PDP-11-Options}.
1167 @item -mpic | -mno-pic
1168 Generate position-independent (or position-dependent) code. The
1169 default is @option{-mpic}.
1172 @itemx -mall-extensions
1173 Enable all instruction set extensions. This is the default.
1175 @item -mno-extensions
1176 Disable all instruction set extensions.
1178 @item -m@var{extension} | -mno-@var{extension}
1179 Enable (or disable) a particular instruction set extension.
1182 Enable the instruction set extensions supported by a particular CPU, and
1183 disable all other extensions.
1185 @item -m@var{machine}
1186 Enable the instruction set extensions supported by a particular machine
1187 model, and disable all other extensions.
1193 The following options are available when @value{AS} is configured for
1194 a picoJava processor.
1198 @cindex PJ endianness
1199 @cindex endianness, PJ
1200 @cindex big endian output, PJ
1202 Generate ``big endian'' format output.
1204 @cindex little endian output, PJ
1206 Generate ``little endian'' format output.
1212 The following options are available when @value{AS} is configured for the
1213 Motorola 68HC11 or 68HC12 series.
1217 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1218 Specify what processor is the target. The default is
1219 defined by the configuration option when building the assembler.
1221 @item --xgate-ramoffset
1222 Instruct the linker to offset RAM addresses from S12X address space into
1223 XGATE address space.
1226 Specify to use the 16-bit integer ABI.
1229 Specify to use the 32-bit integer ABI.
1231 @item -mshort-double
1232 Specify to use the 32-bit double ABI.
1235 Specify to use the 64-bit double ABI.
1237 @item --force-long-branches
1238 Relative branches are turned into absolute ones. This concerns
1239 conditional branches, unconditional branches and branches to a
1242 @item -S | --short-branches
1243 Do not turn relative branches into absolute ones
1244 when the offset is out of range.
1246 @item --strict-direct-mode
1247 Do not turn the direct addressing mode into extended addressing mode
1248 when the instruction does not support direct addressing mode.
1250 @item --print-insn-syntax
1251 Print the syntax of instruction in case of error.
1253 @item --print-opcodes
1254 Print the list of instructions with syntax and then exit.
1256 @item --generate-example
1257 Print an example of instruction for each possible instruction and then exit.
1258 This option is only useful for testing @command{@value{AS}}.
1264 The following options are available when @command{@value{AS}} is configured
1265 for the SPARC architecture:
1268 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1269 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1270 Explicitly select a variant of the SPARC architecture.
1272 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1273 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1275 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1276 UltraSPARC extensions.
1278 @item -xarch=v8plus | -xarch=v8plusa
1279 For compatibility with the Solaris v9 assembler. These options are
1280 equivalent to -Av8plus and -Av8plusa, respectively.
1283 Warn when the assembler switches to another architecture.
1288 The following options are available when @value{AS} is configured for the 'c54x
1293 Enable extended addressing mode. All addresses and relocations will assume
1294 extended addressing (usually 23 bits).
1295 @item -mcpu=@var{CPU_VERSION}
1296 Sets the CPU version being compiled for.
1297 @item -merrors-to-file @var{FILENAME}
1298 Redirect error output to a file, for broken systems which don't support such
1299 behaviour in the shell.
1304 The following options are available when @value{AS} is configured for
1309 This option sets the largest size of an object that can be referenced
1310 implicitly with the @code{gp} register. It is only accepted for targets that
1311 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1313 @cindex MIPS endianness
1314 @cindex endianness, MIPS
1315 @cindex big endian output, MIPS
1317 Generate ``big endian'' format output.
1319 @cindex little endian output, MIPS
1321 Generate ``little endian'' format output.
1339 Generate code for a particular MIPS Instruction Set Architecture level.
1340 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1341 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1342 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1343 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips32r3},
1344 @samp{-mips32r5}, @samp{-mips32r6}, @samp{-mips64}, @samp{-mips64r2},
1345 @samp{-mips64r3}, @samp{-mips64r5}, and @samp{-mips64r6} correspond to generic
1346 MIPS V, MIPS32, MIPS32 Release 2, MIPS32 Release 3, MIPS32 Release 5, MIPS32
1347 Release 6, MIPS64, MIPS64 Release 2, MIPS64 Release 3, MIPS64 Release 5, and
1348 MIPS64 Release 6 ISA processors, respectively.
1350 @item -march=@var{cpu}
1351 Generate code for a particular MIPS CPU.
1353 @item -mtune=@var{cpu}
1354 Schedule and tune for a particular MIPS CPU.
1358 Cause nops to be inserted if the read of the destination register
1359 of an mfhi or mflo instruction occurs in the following two instructions.
1362 @itemx -mno-fix-rm7000
1363 Cause nops to be inserted if a dmult or dmultu instruction is
1364 followed by a load instruction.
1368 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1369 section instead of the standard ELF .stabs sections.
1373 Control generation of @code{.pdr} sections.
1377 The register sizes are normally inferred from the ISA and ABI, but these
1378 flags force a certain group of registers to be treated as 32 bits wide at
1379 all times. @samp{-mgp32} controls the size of general-purpose registers
1380 and @samp{-mfp32} controls the size of floating-point registers.
1384 The register sizes are normally inferred from the ISA and ABI, but these
1385 flags force a certain group of registers to be treated as 64 bits wide at
1386 all times. @samp{-mgp64} controls the size of general-purpose registers
1387 and @samp{-mfp64} controls the size of floating-point registers.
1390 The register sizes are normally inferred from the ISA and ABI, but using
1391 this flag in combination with @samp{-mabi=32} enables an ABI variant
1392 which will operate correctly with floating-point registers which are
1396 @itemx -mno-odd-spreg
1397 Enable use of floating-point operations on odd-numbered single-precision
1398 registers when supported by the ISA. @samp{-mfpxx} implies
1399 @samp{-mno-odd-spreg}, otherwise the default is @samp{-modd-spreg}.
1403 Generate code for the MIPS 16 processor. This is equivalent to putting
1404 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1405 turns off this option.
1408 @itemx -mno-micromips
1409 Generate code for the microMIPS processor. This is equivalent to putting
1410 @code{.set micromips} at the start of the assembly file. @samp{-mno-micromips}
1411 turns off this option. This is equivalent to putting @code{.set nomicromips}
1412 at the start of the assembly file.
1415 @itemx -mno-smartmips
1416 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1417 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1418 @samp{-mno-smartmips} turns off this option.
1422 Generate code for the MIPS-3D Application Specific Extension.
1423 This tells the assembler to accept MIPS-3D instructions.
1424 @samp{-no-mips3d} turns off this option.
1428 Generate code for the MDMX Application Specific Extension.
1429 This tells the assembler to accept MDMX instructions.
1430 @samp{-no-mdmx} turns off this option.
1434 Generate code for the DSP Release 1 Application Specific Extension.
1435 This tells the assembler to accept DSP Release 1 instructions.
1436 @samp{-mno-dsp} turns off this option.
1440 Generate code for the DSP Release 2 Application Specific Extension.
1441 This option implies -mdsp.
1442 This tells the assembler to accept DSP Release 2 instructions.
1443 @samp{-mno-dspr2} turns off this option.
1447 Generate code for the MIPS SIMD Architecture Extension.
1448 This tells the assembler to accept MSA instructions.
1449 @samp{-mno-msa} turns off this option.
1453 Generate code for the MIPS eXtended Physical Address (XPA) Extension.
1454 This tells the assembler to accept XPA instructions.
1455 @samp{-mno-xpa} turns off this option.
1459 Generate code for the MT Application Specific Extension.
1460 This tells the assembler to accept MT instructions.
1461 @samp{-mno-mt} turns off this option.
1465 Generate code for the MCU Application Specific Extension.
1466 This tells the assembler to accept MCU instructions.
1467 @samp{-mno-mcu} turns off this option.
1471 Only use 32-bit instruction encodings when generating code for the
1472 microMIPS processor. This option inhibits the use of any 16-bit
1473 instructions. This is equivalent to putting @code{.set insn32} at
1474 the start of the assembly file. @samp{-mno-insn32} turns off this
1475 option. This is equivalent to putting @code{.set noinsn32} at the
1476 start of the assembly file. By default @samp{-mno-insn32} is
1477 selected, allowing all instructions to be used.
1479 @item --construct-floats
1480 @itemx --no-construct-floats
1481 The @samp{--no-construct-floats} option disables the construction of
1482 double width floating point constants by loading the two halves of the
1483 value into the two single width floating point registers that make up
1484 the double width register. By default @samp{--construct-floats} is
1485 selected, allowing construction of these floating point constants.
1487 @item --relax-branch
1488 @itemx --no-relax-branch
1489 The @samp{--relax-branch} option enables the relaxation of out-of-range
1490 branches. By default @samp{--no-relax-branch} is selected, causing any
1491 out-of-range branches to produce an error.
1493 @item -mnan=@var{encoding}
1494 Select between the IEEE 754-2008 (@option{-mnan=2008}) or the legacy
1495 (@option{-mnan=legacy}) NaN encoding format. The latter is the default.
1498 @item --emulation=@var{name}
1499 This option was formerly used to switch between ELF and ECOFF output
1500 on targets like IRIX 5 that supported both. MIPS ECOFF support was
1501 removed in GAS 2.24, so the option now serves little purpose.
1502 It is retained for backwards compatibility.
1504 The available configuration names are: @samp{mipself}, @samp{mipslelf} and
1505 @samp{mipsbelf}. Choosing @samp{mipself} now has no effect, since the output
1506 is always ELF. @samp{mipslelf} and @samp{mipsbelf} select little- and
1507 big-endian output respectively, but @samp{-EL} and @samp{-EB} are now the
1508 preferred options instead.
1511 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1518 Control how to deal with multiplication overflow and division by zero.
1519 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1520 (and only work for Instruction Set Architecture level 2 and higher);
1521 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1525 When this option is used, @command{@value{AS}} will issue a warning every
1526 time it generates a nop instruction from a macro.
1531 The following options are available when @value{AS} is configured for
1537 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1538 The command line option @samp{-nojsri2bsr} can be used to disable it.
1542 Enable or disable the silicon filter behaviour. By default this is disabled.
1543 The default can be overridden by the @samp{-sifilter} command line option.
1546 Alter jump instructions for long displacements.
1548 @item -mcpu=[210|340]
1549 Select the cpu type on the target hardware. This controls which instructions
1553 Assemble for a big endian target.
1556 Assemble for a little endian target.
1565 @xref{Meta Options}, for the options available when @value{AS} is configured
1566 for a Meta processor.
1570 @c man begin OPTIONS
1571 The following options are available when @value{AS} is configured for a
1574 @c man begin INCLUDE
1575 @include c-metag.texi
1576 @c ended inside the included file
1581 @c man begin OPTIONS
1583 See the info pages for documentation of the MMIX-specific options.
1589 @xref{NDS32 Options}, for the options available when @value{AS} is configured
1590 for a NDS32 processor.
1592 @c ended inside the included file
1596 @c man begin OPTIONS
1597 The following options are available when @value{AS} is configured for a
1600 @c man begin INCLUDE
1601 @include c-nds32.texi
1602 @c ended inside the included file
1609 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1610 for a PowerPC processor.
1614 @c man begin OPTIONS
1615 The following options are available when @value{AS} is configured for a
1618 @c man begin INCLUDE
1620 @c ended inside the included file
1625 @c man begin OPTIONS
1627 See the info pages for documentation of the RX-specific options.
1631 The following options are available when @value{AS} is configured for the s390
1637 Select the word size, either 31/32 bits or 64 bits.
1640 Select the architecture mode, either the Enterprise System
1641 Architecture (esa) or the z/Architecture mode (zarch).
1642 @item -march=@var{processor}
1643 Specify which s390 processor variant is the target, @samp{g6}, @samp{g6},
1644 @samp{z900}, @samp{z990}, @samp{z9-109}, @samp{z9-ec}, @samp{z10},
1645 @samp{z196}, @samp{zEC12}, or @samp{z13}.
1647 @itemx -mno-regnames
1648 Allow or disallow symbolic names for registers.
1649 @item -mwarn-areg-zero
1650 Warn whenever the operand for a base or index register has been specified
1651 but evaluates to zero.
1659 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1660 for a TMS320C6000 processor.
1664 @c man begin OPTIONS
1665 The following options are available when @value{AS} is configured for a
1666 TMS320C6000 processor.
1668 @c man begin INCLUDE
1669 @include c-tic6x.texi
1670 @c ended inside the included file
1678 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1679 for a TILE-Gx processor.
1683 @c man begin OPTIONS
1684 The following options are available when @value{AS} is configured for a TILE-Gx
1687 @c man begin INCLUDE
1688 @include c-tilegx.texi
1689 @c ended inside the included file
1697 @xref{Visium Options}, for the options available when @value{AS} is configured
1698 for a Visium processor.
1702 @c man begin OPTIONS
1703 The following option is available when @value{AS} is configured for a Visium
1706 @c man begin INCLUDE
1707 @include c-visium.texi
1708 @c ended inside the included file
1716 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1717 for an Xtensa processor.
1721 @c man begin OPTIONS
1722 The following options are available when @value{AS} is configured for an
1725 @c man begin INCLUDE
1726 @include c-xtensa.texi
1727 @c ended inside the included file
1732 @c man begin OPTIONS
1735 The following options are available when @value{AS} is configured for
1736 a Z80 family processor.
1739 Assemble for Z80 processor.
1741 Assemble for R800 processor.
1742 @item -ignore-undocumented-instructions
1744 Assemble undocumented Z80 instructions that also work on R800 without warning.
1745 @item -ignore-unportable-instructions
1747 Assemble all undocumented Z80 instructions without warning.
1748 @item -warn-undocumented-instructions
1750 Issue a warning for undocumented Z80 instructions that also work on R800.
1751 @item -warn-unportable-instructions
1753 Issue a warning for undocumented Z80 instructions that do not work on R800.
1754 @item -forbid-undocumented-instructions
1756 Treat all undocumented instructions as errors.
1757 @item -forbid-unportable-instructions
1759 Treat undocumented Z80 instructions that do not work on R800 as errors.
1766 * Manual:: Structure of this Manual
1767 * GNU Assembler:: The GNU Assembler
1768 * Object Formats:: Object File Formats
1769 * Command Line:: Command Line
1770 * Input Files:: Input Files
1771 * Object:: Output (Object) File
1772 * Errors:: Error and Warning Messages
1776 @section Structure of this Manual
1778 @cindex manual, structure and purpose
1779 This manual is intended to describe what you need to know to use
1780 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1781 notation for symbols, constants, and expressions; the directives that
1782 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1785 We also cover special features in the @value{TARGET}
1786 configuration of @command{@value{AS}}, including assembler directives.
1789 This manual also describes some of the machine-dependent features of
1790 various flavors of the assembler.
1793 @cindex machine instructions (not covered)
1794 On the other hand, this manual is @emph{not} intended as an introduction
1795 to programming in assembly language---let alone programming in general!
1796 In a similar vein, we make no attempt to introduce the machine
1797 architecture; we do @emph{not} describe the instruction set, standard
1798 mnemonics, registers or addressing modes that are standard to a
1799 particular architecture.
1801 You may want to consult the manufacturer's
1802 machine architecture manual for this information.
1806 For information on the H8/300 machine instruction set, see @cite{H8/300
1807 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1808 Programming Manual} (Renesas).
1811 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1812 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1813 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1814 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1817 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1821 @c I think this is premature---doc@cygnus.com, 17jan1991
1823 Throughout this manual, we assume that you are running @dfn{GNU},
1824 the portable operating system from the @dfn{Free Software
1825 Foundation, Inc.}. This restricts our attention to certain kinds of
1826 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1827 once this assumption is granted examples and definitions need less
1830 @command{@value{AS}} is part of a team of programs that turn a high-level
1831 human-readable series of instructions into a low-level
1832 computer-readable series of instructions. Different versions of
1833 @command{@value{AS}} are used for different kinds of computer.
1836 @c There used to be a section "Terminology" here, which defined
1837 @c "contents", "byte", "word", and "long". Defining "word" to any
1838 @c particular size is confusing when the .word directive may generate 16
1839 @c bits on one machine and 32 bits on another; in general, for the user
1840 @c version of this manual, none of these terms seem essential to define.
1841 @c They were used very little even in the former draft of the manual;
1842 @c this draft makes an effort to avoid them (except in names of
1846 @section The GNU Assembler
1848 @c man begin DESCRIPTION
1850 @sc{gnu} @command{as} is really a family of assemblers.
1852 This manual describes @command{@value{AS}}, a member of that family which is
1853 configured for the @value{TARGET} architectures.
1855 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1856 should find a fairly similar environment when you use it on another
1857 architecture. Each version has much in common with the others,
1858 including object file formats, most assembler directives (often called
1859 @dfn{pseudo-ops}) and assembler syntax.@refill
1861 @cindex purpose of @sc{gnu} assembler
1862 @command{@value{AS}} is primarily intended to assemble the output of the
1863 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1864 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1865 assemble correctly everything that other assemblers for the same
1866 machine would assemble.
1868 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1871 @c This remark should appear in generic version of manual; assumption
1872 @c here is that generic version sets M680x0.
1873 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1874 assembler for the same architecture; for example, we know of several
1875 incompatible versions of 680x0 assembly language syntax.
1880 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1881 program in one pass of the source file. This has a subtle impact on the
1882 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1884 @node Object Formats
1885 @section Object File Formats
1887 @cindex object file format
1888 The @sc{gnu} assembler can be configured to produce several alternative
1889 object file formats. For the most part, this does not affect how you
1890 write assembly language programs; but directives for debugging symbols
1891 are typically different in different file formats. @xref{Symbol
1892 Attributes,,Symbol Attributes}.
1895 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1896 @value{OBJ-NAME} format object files.
1898 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1900 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1901 @code{b.out} or COFF format object files.
1904 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1905 SOM or ELF format object files.
1910 @section Command Line
1912 @cindex command line conventions
1914 After the program name @command{@value{AS}}, the command line may contain
1915 options and file names. Options may appear in any order, and may be
1916 before, after, or between file names. The order of file names is
1919 @cindex standard input, as input file
1921 @file{--} (two hyphens) by itself names the standard input file
1922 explicitly, as one of the files for @command{@value{AS}} to assemble.
1924 @cindex options, command line
1925 Except for @samp{--} any command line argument that begins with a
1926 hyphen (@samp{-}) is an option. Each option changes the behavior of
1927 @command{@value{AS}}. No option changes the way another option works. An
1928 option is a @samp{-} followed by one or more letters; the case of
1929 the letter is important. All options are optional.
1931 Some options expect exactly one file name to follow them. The file
1932 name may either immediately follow the option's letter (compatible
1933 with older assemblers) or it may be the next command argument (@sc{gnu}
1934 standard). These two command lines are equivalent:
1937 @value{AS} -o my-object-file.o mumble.s
1938 @value{AS} -omy-object-file.o mumble.s
1942 @section Input Files
1945 @cindex source program
1946 @cindex files, input
1947 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1948 describe the program input to one run of @command{@value{AS}}. The program may
1949 be in one or more files; how the source is partitioned into files
1950 doesn't change the meaning of the source.
1952 @c I added "con" prefix to "catenation" just to prove I can overcome my
1953 @c APL training... doc@cygnus.com
1954 The source program is a concatenation of the text in all the files, in the
1957 @c man begin DESCRIPTION
1958 Each time you run @command{@value{AS}} it assembles exactly one source
1959 program. The source program is made up of one or more files.
1960 (The standard input is also a file.)
1962 You give @command{@value{AS}} a command line that has zero or more input file
1963 names. The input files are read (from left file name to right). A
1964 command line argument (in any position) that has no special meaning
1965 is taken to be an input file name.
1967 If you give @command{@value{AS}} no file names it attempts to read one input file
1968 from the @command{@value{AS}} standard input, which is normally your terminal. You
1969 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1972 Use @samp{--} if you need to explicitly name the standard input file
1973 in your command line.
1975 If the source is empty, @command{@value{AS}} produces a small, empty object
1980 @subheading Filenames and Line-numbers
1982 @cindex input file linenumbers
1983 @cindex line numbers, in input files
1984 There are two ways of locating a line in the input file (or files) and
1985 either may be used in reporting error messages. One way refers to a line
1986 number in a physical file; the other refers to a line number in a
1987 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1989 @dfn{Physical files} are those files named in the command line given
1990 to @command{@value{AS}}.
1992 @dfn{Logical files} are simply names declared explicitly by assembler
1993 directives; they bear no relation to physical files. Logical file names help
1994 error messages reflect the original source file, when @command{@value{AS}} source
1995 is itself synthesized from other files. @command{@value{AS}} understands the
1996 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1997 @ref{File,,@code{.file}}.
2000 @section Output (Object) File
2006 Every time you run @command{@value{AS}} it produces an output file, which is
2007 your assembly language program translated into numbers. This file
2008 is the object file. Its default name is
2016 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
2018 You can give it another name by using the @option{-o} option. Conventionally,
2019 object file names end with @file{.o}. The default name is used for historical
2020 reasons: older assemblers were capable of assembling self-contained programs
2021 directly into a runnable program. (For some formats, this isn't currently
2022 possible, but it can be done for the @code{a.out} format.)
2026 The object file is meant for input to the linker @code{@value{LD}}. It contains
2027 assembled program code, information to help @code{@value{LD}} integrate
2028 the assembled program into a runnable file, and (optionally) symbolic
2029 information for the debugger.
2031 @c link above to some info file(s) like the description of a.out.
2032 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
2035 @section Error and Warning Messages
2037 @c man begin DESCRIPTION
2039 @cindex error messages
2040 @cindex warning messages
2041 @cindex messages from assembler
2042 @command{@value{AS}} may write warnings and error messages to the standard error
2043 file (usually your terminal). This should not happen when a compiler
2044 runs @command{@value{AS}} automatically. Warnings report an assumption made so
2045 that @command{@value{AS}} could keep assembling a flawed program; errors report a
2046 grave problem that stops the assembly.
2050 @cindex format of warning messages
2051 Warning messages have the format
2054 file_name:@b{NNN}:Warning Message Text
2058 @cindex file names and line numbers, in warnings/errors
2059 (where @b{NNN} is a line number). If both a logical file name
2060 (@pxref{File,,@code{.file}}) and a logical line number
2062 (@pxref{Line,,@code{.line}})
2064 have been given then they will be used, otherwise the file name and line number
2065 in the current assembler source file will be used. The message text is
2066 intended to be self explanatory (in the grand Unix tradition).
2068 Note the file name must be set via the logical version of the @code{.file}
2069 directive, not the DWARF2 version of the @code{.file} directive. For example:
2073 error_assembler_source
2079 produces this output:
2083 asm.s:2: Error: no such instruction: `error_assembler_source'
2084 foo.c:31: Error: no such instruction: `error_c_source'
2087 @cindex format of error messages
2088 Error messages have the format
2091 file_name:@b{NNN}:FATAL:Error Message Text
2094 The file name and line number are derived as for warning
2095 messages. The actual message text may be rather less explanatory
2096 because many of them aren't supposed to happen.
2099 @chapter Command-Line Options
2101 @cindex options, all versions of assembler
2102 This chapter describes command-line options available in @emph{all}
2103 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
2104 for options specific
2106 to the @value{TARGET} target.
2109 to particular machine architectures.
2112 @c man begin DESCRIPTION
2114 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
2115 you can use the @samp{-Wa} option to pass arguments through to the assembler.
2116 The assembler arguments must be separated from each other (and the @samp{-Wa})
2117 by commas. For example:
2120 gcc -c -g -O -Wa,-alh,-L file.c
2124 This passes two options to the assembler: @samp{-alh} (emit a listing to
2125 standard output with high-level and assembly source) and @samp{-L} (retain
2126 local symbols in the symbol table).
2128 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
2129 command-line options are automatically passed to the assembler by the compiler.
2130 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
2131 precisely what options it passes to each compilation pass, including the
2137 * a:: -a[cdghlns] enable listings
2138 * alternate:: --alternate enable alternate macro syntax
2139 * D:: -D for compatibility
2140 * f:: -f to work faster
2141 * I:: -I for .include search path
2142 @ifclear DIFF-TBL-KLUGE
2143 * K:: -K for compatibility
2145 @ifset DIFF-TBL-KLUGE
2146 * K:: -K for difference tables
2149 * L:: -L to retain local symbols
2150 * listing:: --listing-XXX to configure listing output
2151 * M:: -M or --mri to assemble in MRI compatibility mode
2152 * MD:: --MD for dependency tracking
2153 * o:: -o to name the object file
2154 * R:: -R to join data and text sections
2155 * statistics:: --statistics to see statistics about assembly
2156 * traditional-format:: --traditional-format for compatible output
2157 * v:: -v to announce version
2158 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
2159 * Z:: -Z to make object file even after errors
2163 @section Enable Listings: @option{-a[cdghlns]}
2173 @cindex listings, enabling
2174 @cindex assembly listings, enabling
2176 These options enable listing output from the assembler. By itself,
2177 @samp{-a} requests high-level, assembly, and symbols listing.
2178 You can use other letters to select specific options for the list:
2179 @samp{-ah} requests a high-level language listing,
2180 @samp{-al} requests an output-program assembly listing, and
2181 @samp{-as} requests a symbol table listing.
2182 High-level listings require that a compiler debugging option like
2183 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
2186 Use the @samp{-ag} option to print a first section with general assembly
2187 information, like @value{AS} version, switches passed, or time stamp.
2189 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
2190 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
2191 other conditional), or a true @code{.if} followed by an @code{.else}, will be
2192 omitted from the listing.
2194 Use the @samp{-ad} option to omit debugging directives from the
2197 Once you have specified one of these options, you can further control
2198 listing output and its appearance using the directives @code{.list},
2199 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
2201 The @samp{-an} option turns off all forms processing.
2202 If you do not request listing output with one of the @samp{-a} options, the
2203 listing-control directives have no effect.
2205 The letters after @samp{-a} may be combined into one option,
2206 @emph{e.g.}, @samp{-aln}.
2208 Note if the assembler source is coming from the standard input (e.g.,
2210 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
2211 is being used) then the listing will not contain any comments or preprocessor
2212 directives. This is because the listing code buffers input source lines from
2213 stdin only after they have been preprocessed by the assembler. This reduces
2214 memory usage and makes the code more efficient.
2217 @section @option{--alternate}
2220 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
2223 @section @option{-D}
2226 This option has no effect whatsoever, but it is accepted to make it more
2227 likely that scripts written for other assemblers also work with
2228 @command{@value{AS}}.
2231 @section Work Faster: @option{-f}
2234 @cindex trusted compiler
2235 @cindex faster processing (@option{-f})
2236 @samp{-f} should only be used when assembling programs written by a
2237 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
2238 and comment preprocessing on
2239 the input file(s) before assembling them. @xref{Preprocessing,
2243 @emph{Warning:} if you use @samp{-f} when the files actually need to be
2244 preprocessed (if they contain comments, for example), @command{@value{AS}} does
2249 @section @code{.include} Search Path: @option{-I} @var{path}
2251 @kindex -I @var{path}
2252 @cindex paths for @code{.include}
2253 @cindex search path for @code{.include}
2254 @cindex @code{include} directive search path
2255 Use this option to add a @var{path} to the list of directories
2256 @command{@value{AS}} searches for files specified in @code{.include}
2257 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
2258 many times as necessary to include a variety of paths. The current
2259 working directory is always searched first; after that, @command{@value{AS}}
2260 searches any @samp{-I} directories in the same order as they were
2261 specified (left to right) on the command line.
2264 @section Difference Tables: @option{-K}
2267 @ifclear DIFF-TBL-KLUGE
2268 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2269 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2270 where it can be used to warn when the assembler alters the machine code
2271 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2272 family does not have the addressing limitations that sometimes lead to this
2273 alteration on other platforms.
2276 @ifset DIFF-TBL-KLUGE
2277 @cindex difference tables, warning
2278 @cindex warning for altered difference tables
2279 @command{@value{AS}} sometimes alters the code emitted for directives of the
2280 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2281 You can use the @samp{-K} option if you want a warning issued when this
2286 @section Include Local Symbols: @option{-L}
2289 @cindex local symbols, retaining in output
2290 Symbols beginning with system-specific local label prefixes, typically
2291 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2292 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2293 such symbols when debugging, because they are intended for the use of
2294 programs (like compilers) that compose assembler programs, not for your
2295 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2296 such symbols, so you do not normally debug with them.
2298 This option tells @command{@value{AS}} to retain those local symbols
2299 in the object file. Usually if you do this you also tell the linker
2300 @code{@value{LD}} to preserve those symbols.
2303 @section Configuring listing output: @option{--listing}
2305 The listing feature of the assembler can be enabled via the command line switch
2306 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2307 hex dump of the corresponding locations in the output object file, and displays
2308 them as a listing file. The format of this listing can be controlled by
2309 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2310 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2311 @code{.psize} (@pxref{Psize}), and
2312 @code{.eject} (@pxref{Eject}) and also by the following switches:
2315 @item --listing-lhs-width=@samp{number}
2316 @kindex --listing-lhs-width
2317 @cindex Width of first line disassembly output
2318 Sets the maximum width, in words, of the first line of the hex byte dump. This
2319 dump appears on the left hand side of the listing output.
2321 @item --listing-lhs-width2=@samp{number}
2322 @kindex --listing-lhs-width2
2323 @cindex Width of continuation lines of disassembly output
2324 Sets the maximum width, in words, of any further lines of the hex byte dump for
2325 a given input source line. If this value is not specified, it defaults to being
2326 the same as the value specified for @samp{--listing-lhs-width}. If neither
2327 switch is used the default is to one.
2329 @item --listing-rhs-width=@samp{number}
2330 @kindex --listing-rhs-width
2331 @cindex Width of source line output
2332 Sets the maximum width, in characters, of the source line that is displayed
2333 alongside the hex dump. The default value for this parameter is 100. The
2334 source line is displayed on the right hand side of the listing output.
2336 @item --listing-cont-lines=@samp{number}
2337 @kindex --listing-cont-lines
2338 @cindex Maximum number of continuation lines
2339 Sets the maximum number of continuation lines of hex dump that will be
2340 displayed for a given single line of source input. The default value is 4.
2344 @section Assemble in MRI Compatibility Mode: @option{-M}
2347 @cindex MRI compatibility mode
2348 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2349 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2350 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
2351 configured target) assembler from Microtec Research. The exact nature of the
2352 MRI syntax will not be documented here; see the MRI manuals for more
2353 information. Note in particular that the handling of macros and macro
2354 arguments is somewhat different. The purpose of this option is to permit
2355 assembling existing MRI assembler code using @command{@value{AS}}.
2357 The MRI compatibility is not complete. Certain operations of the MRI assembler
2358 depend upon its object file format, and can not be supported using other object
2359 file formats. Supporting these would require enhancing each object file format
2360 individually. These are:
2363 @item global symbols in common section
2365 The m68k MRI assembler supports common sections which are merged by the linker.
2366 Other object file formats do not support this. @command{@value{AS}} handles
2367 common sections by treating them as a single common symbol. It permits local
2368 symbols to be defined within a common section, but it can not support global
2369 symbols, since it has no way to describe them.
2371 @item complex relocations
2373 The MRI assemblers support relocations against a negated section address, and
2374 relocations which combine the start addresses of two or more sections. These
2375 are not support by other object file formats.
2377 @item @code{END} pseudo-op specifying start address
2379 The MRI @code{END} pseudo-op permits the specification of a start address.
2380 This is not supported by other object file formats. The start address may
2381 instead be specified using the @option{-e} option to the linker, or in a linker
2384 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2386 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2387 name to the output file. This is not supported by other object file formats.
2389 @item @code{ORG} pseudo-op
2391 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2392 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2393 which changes the location within the current section. Absolute sections are
2394 not supported by other object file formats. The address of a section may be
2395 assigned within a linker script.
2398 There are some other features of the MRI assembler which are not supported by
2399 @command{@value{AS}}, typically either because they are difficult or because they
2400 seem of little consequence. Some of these may be supported in future releases.
2404 @item EBCDIC strings
2406 EBCDIC strings are not supported.
2408 @item packed binary coded decimal
2410 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2411 and @code{DCB.P} pseudo-ops are not supported.
2413 @item @code{FEQU} pseudo-op
2415 The m68k @code{FEQU} pseudo-op is not supported.
2417 @item @code{NOOBJ} pseudo-op
2419 The m68k @code{NOOBJ} pseudo-op is not supported.
2421 @item @code{OPT} branch control options
2423 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2424 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2425 relaxes all branches, whether forward or backward, to an appropriate size, so
2426 these options serve no purpose.
2428 @item @code{OPT} list control options
2430 The following m68k @code{OPT} list control options are ignored: @code{C},
2431 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2432 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2434 @item other @code{OPT} options
2436 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2437 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2439 @item @code{OPT} @code{D} option is default
2441 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2442 @code{OPT NOD} may be used to turn it off.
2444 @item @code{XREF} pseudo-op.
2446 The m68k @code{XREF} pseudo-op is ignored.
2448 @item @code{.debug} pseudo-op
2450 The i960 @code{.debug} pseudo-op is not supported.
2452 @item @code{.extended} pseudo-op
2454 The i960 @code{.extended} pseudo-op is not supported.
2456 @item @code{.list} pseudo-op.
2458 The various options of the i960 @code{.list} pseudo-op are not supported.
2460 @item @code{.optimize} pseudo-op
2462 The i960 @code{.optimize} pseudo-op is not supported.
2464 @item @code{.output} pseudo-op
2466 The i960 @code{.output} pseudo-op is not supported.
2468 @item @code{.setreal} pseudo-op
2470 The i960 @code{.setreal} pseudo-op is not supported.
2475 @section Dependency Tracking: @option{--MD}
2478 @cindex dependency tracking
2481 @command{@value{AS}} can generate a dependency file for the file it creates. This
2482 file consists of a single rule suitable for @code{make} describing the
2483 dependencies of the main source file.
2485 The rule is written to the file named in its argument.
2487 This feature is used in the automatic updating of makefiles.
2490 @section Name the Object File: @option{-o}
2493 @cindex naming object file
2494 @cindex object file name
2495 There is always one object file output when you run @command{@value{AS}}. By
2496 default it has the name
2499 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
2513 You use this option (which takes exactly one filename) to give the
2514 object file a different name.
2516 Whatever the object file is called, @command{@value{AS}} overwrites any
2517 existing file of the same name.
2520 @section Join Data and Text Sections: @option{-R}
2523 @cindex data and text sections, joining
2524 @cindex text and data sections, joining
2525 @cindex joining text and data sections
2526 @cindex merging text and data sections
2527 @option{-R} tells @command{@value{AS}} to write the object file as if all
2528 data-section data lives in the text section. This is only done at
2529 the very last moment: your binary data are the same, but data
2530 section parts are relocated differently. The data section part of
2531 your object file is zero bytes long because all its bytes are
2532 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2534 When you specify @option{-R} it would be possible to generate shorter
2535 address displacements (because we do not have to cross between text and
2536 data section). We refrain from doing this simply for compatibility with
2537 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2540 When @command{@value{AS}} is configured for COFF or ELF output,
2541 this option is only useful if you use sections named @samp{.text} and
2546 @option{-R} is not supported for any of the HPPA targets. Using
2547 @option{-R} generates a warning from @command{@value{AS}}.
2551 @section Display Assembly Statistics: @option{--statistics}
2553 @kindex --statistics
2554 @cindex statistics, about assembly
2555 @cindex time, total for assembly
2556 @cindex space used, maximum for assembly
2557 Use @samp{--statistics} to display two statistics about the resources used by
2558 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2559 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2562 @node traditional-format
2563 @section Compatible Output: @option{--traditional-format}
2565 @kindex --traditional-format
2566 For some targets, the output of @command{@value{AS}} is different in some ways
2567 from the output of some existing assembler. This switch requests
2568 @command{@value{AS}} to use the traditional format instead.
2570 For example, it disables the exception frame optimizations which
2571 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2574 @section Announce Version: @option{-v}
2578 @cindex assembler version
2579 @cindex version of assembler
2580 You can find out what version of as is running by including the
2581 option @samp{-v} (which you can also spell as @samp{-version}) on the
2585 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2587 @command{@value{AS}} should never give a warning or error message when
2588 assembling compiler output. But programs written by people often
2589 cause @command{@value{AS}} to give a warning that a particular assumption was
2590 made. All such warnings are directed to the standard error file.
2594 @cindex suppressing warnings
2595 @cindex warnings, suppressing
2596 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2597 This only affects the warning messages: it does not change any particular of
2598 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2601 @kindex --fatal-warnings
2602 @cindex errors, caused by warnings
2603 @cindex warnings, causing error
2604 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2605 files that generate warnings to be in error.
2608 @cindex warnings, switching on
2609 You can switch these options off again by specifying @option{--warn}, which
2610 causes warnings to be output as usual.
2613 @section Generate Object File in Spite of Errors: @option{-Z}
2614 @cindex object file, after errors
2615 @cindex errors, continuing after
2616 After an error message, @command{@value{AS}} normally produces no output. If for
2617 some reason you are interested in object file output even after
2618 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2619 option. If there are any errors, @command{@value{AS}} continues anyways, and
2620 writes an object file after a final warning message of the form @samp{@var{n}
2621 errors, @var{m} warnings, generating bad object file.}
2626 @cindex machine-independent syntax
2627 @cindex syntax, machine-independent
2628 This chapter describes the machine-independent syntax allowed in a
2629 source file. @command{@value{AS}} syntax is similar to what many other
2630 assemblers use; it is inspired by the BSD 4.2
2635 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2639 * Preprocessing:: Preprocessing
2640 * Whitespace:: Whitespace
2641 * Comments:: Comments
2642 * Symbol Intro:: Symbols
2643 * Statements:: Statements
2644 * Constants:: Constants
2648 @section Preprocessing
2650 @cindex preprocessing
2651 The @command{@value{AS}} internal preprocessor:
2653 @cindex whitespace, removed by preprocessor
2655 adjusts and removes extra whitespace. It leaves one space or tab before
2656 the keywords on a line, and turns any other whitespace on the line into
2659 @cindex comments, removed by preprocessor
2661 removes all comments, replacing them with a single space, or an
2662 appropriate number of newlines.
2664 @cindex constants, converted by preprocessor
2666 converts character constants into the appropriate numeric values.
2669 It does not do macro processing, include file handling, or
2670 anything else you may get from your C compiler's preprocessor. You can
2671 do include file processing with the @code{.include} directive
2672 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2673 to get other ``CPP'' style preprocessing by giving the input file a
2674 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2675 Output, gcc.info, Using GNU CC}.
2677 Excess whitespace, comments, and character constants
2678 cannot be used in the portions of the input text that are not
2681 @cindex turning preprocessing on and off
2682 @cindex preprocessing, turning on and off
2685 If the first line of an input file is @code{#NO_APP} or if you use the
2686 @samp{-f} option, whitespace and comments are not removed from the input file.
2687 Within an input file, you can ask for whitespace and comment removal in
2688 specific portions of the by putting a line that says @code{#APP} before the
2689 text that may contain whitespace or comments, and putting a line that says
2690 @code{#NO_APP} after this text. This feature is mainly intend to support
2691 @code{asm} statements in compilers whose output is otherwise free of comments
2698 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2699 Whitespace is used to separate symbols, and to make programs neater for
2700 people to read. Unless within character constants
2701 (@pxref{Characters,,Character Constants}), any whitespace means the same
2702 as exactly one space.
2708 There are two ways of rendering comments to @command{@value{AS}}. In both
2709 cases the comment is equivalent to one space.
2711 Anything from @samp{/*} through the next @samp{*/} is a comment.
2712 This means you may not nest these comments.
2716 The only way to include a newline ('\n') in a comment
2717 is to use this sort of comment.
2720 /* This sort of comment does not nest. */
2723 @cindex line comment character
2724 Anything from a @dfn{line comment} character up to the next newline is
2725 considered a comment and is ignored. The line comment character is target
2726 specific, and some targets multiple comment characters. Some targets also have
2727 line comment characters that only work if they are the first character on a
2728 line. Some targets use a sequence of two characters to introduce a line
2729 comment. Some targets can also change their line comment characters depending
2730 upon command line options that have been used. For more details see the
2731 @emph{Syntax} section in the documentation for individual targets.
2733 If the line comment character is the hash sign (@samp{#}) then it still has the
2734 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2735 to specify logical line numbers:
2738 @cindex lines starting with @code{#}
2739 @cindex logical line numbers
2740 To be compatible with past assemblers, lines that begin with @samp{#} have a
2741 special interpretation. Following the @samp{#} should be an absolute
2742 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2743 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2744 new logical file name. The rest of the line, if any, should be whitespace.
2746 If the first non-whitespace characters on the line are not numeric,
2747 the line is ignored. (Just like a comment.)
2750 # This is an ordinary comment.
2751 # 42-6 "new_file_name" # New logical file name
2752 # This is logical line # 36.
2754 This feature is deprecated, and may disappear from future versions
2755 of @command{@value{AS}}.
2760 @cindex characters used in symbols
2761 @ifclear SPECIAL-SYMS
2762 A @dfn{symbol} is one or more characters chosen from the set of all
2763 letters (both upper and lower case), digits and the three characters
2769 A @dfn{symbol} is one or more characters chosen from the set of all
2770 letters (both upper and lower case), digits and the three characters
2771 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2777 On most machines, you can also use @code{$} in symbol names; exceptions
2778 are noted in @ref{Machine Dependencies}.
2780 No symbol may begin with a digit. Case is significant.
2781 There is no length limit; all characters are significant. Multibyte characters
2782 are supported. Symbols are delimited by characters not in that set, or by the
2783 beginning of a file (since the source program must end with a newline, the end
2784 of a file is not a possible symbol delimiter). @xref{Symbols}.
2786 Symbol names may also be enclosed in double quote @code{"} characters. In such
2787 cases any characters are allowed, except for the NUL character. If a double
2788 quote character is to be included in the symbol name it must be preceeded by a
2789 backslash @code{\} character.
2790 @cindex length of symbols
2795 @cindex statements, structure of
2796 @cindex line separator character
2797 @cindex statement separator character
2799 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2800 @dfn{line separator character}. The line separator character is target
2801 specific and described in the @emph{Syntax} section of each
2802 target's documentation. Not all targets support a line separator character.
2803 The newline or line separator character is considered to be part of the
2804 preceding statement. Newlines and separators within character constants are an
2805 exception: they do not end statements.
2807 @cindex newline, required at file end
2808 @cindex EOF, newline must precede
2809 It is an error to end any statement with end-of-file: the last
2810 character of any input file should be a newline.@refill
2812 An empty statement is allowed, and may include whitespace. It is ignored.
2814 @cindex instructions and directives
2815 @cindex directives and instructions
2816 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2817 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2819 A statement begins with zero or more labels, optionally followed by a
2820 key symbol which determines what kind of statement it is. The key
2821 symbol determines the syntax of the rest of the statement. If the
2822 symbol begins with a dot @samp{.} then the statement is an assembler
2823 directive: typically valid for any computer. If the symbol begins with
2824 a letter the statement is an assembly language @dfn{instruction}: it
2825 assembles into a machine language instruction.
2827 Different versions of @command{@value{AS}} for different computers
2828 recognize different instructions. In fact, the same symbol may
2829 represent a different instruction in a different computer's assembly
2833 @cindex @code{:} (label)
2834 @cindex label (@code{:})
2835 A label is a symbol immediately followed by a colon (@code{:}).
2836 Whitespace before a label or after a colon is permitted, but you may not
2837 have whitespace between a label's symbol and its colon. @xref{Labels}.
2840 For HPPA targets, labels need not be immediately followed by a colon, but
2841 the definition of a label must begin in column zero. This also implies that
2842 only one label may be defined on each line.
2846 label: .directive followed by something
2847 another_label: # This is an empty statement.
2848 instruction operand_1, operand_2, @dots{}
2855 A constant is a number, written so that its value is known by
2856 inspection, without knowing any context. Like this:
2859 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2860 .ascii "Ring the bell\7" # A string constant.
2861 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2862 .float 0f-314159265358979323846264338327\
2863 95028841971.693993751E-40 # - pi, a flonum.
2868 * Characters:: Character Constants
2869 * Numbers:: Number Constants
2873 @subsection Character Constants
2875 @cindex character constants
2876 @cindex constants, character
2877 There are two kinds of character constants. A @dfn{character} stands
2878 for one character in one byte and its value may be used in
2879 numeric expressions. String constants (properly called string
2880 @emph{literals}) are potentially many bytes and their values may not be
2881 used in arithmetic expressions.
2885 * Chars:: Characters
2889 @subsubsection Strings
2891 @cindex string constants
2892 @cindex constants, string
2893 A @dfn{string} is written between double-quotes. It may contain
2894 double-quotes or null characters. The way to get special characters
2895 into a string is to @dfn{escape} these characters: precede them with
2896 a backslash @samp{\} character. For example @samp{\\} represents
2897 one backslash: the first @code{\} is an escape which tells
2898 @command{@value{AS}} to interpret the second character literally as a backslash
2899 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2900 escape character). The complete list of escapes follows.
2902 @cindex escape codes, character
2903 @cindex character escape codes
2904 @c NOTE: Cindex entries must not start with a backlash character.
2905 @c NOTE: This confuses the pdf2texi script when it is creating the
2906 @c NOTE: index based upon the first character and so it generates:
2907 @c NOTE: \initial {\\}
2908 @c NOTE: which then results in the error message:
2909 @c NOTE: Argument of \\ has an extra }.
2910 @c NOTE: So in the index entries below a space character has been
2911 @c NOTE: prepended to avoid this problem.
2914 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2916 @cindex @code{ \b} (backspace character)
2917 @cindex backspace (@code{\b})
2919 Mnemonic for backspace; for ASCII this is octal code 010.
2922 @c Mnemonic for EOText; for ASCII this is octal code 004.
2924 @cindex @code{ \f} (formfeed character)
2925 @cindex formfeed (@code{\f})
2927 Mnemonic for FormFeed; for ASCII this is octal code 014.
2929 @cindex @code{ \n} (newline character)
2930 @cindex newline (@code{\n})
2932 Mnemonic for newline; for ASCII this is octal code 012.
2935 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2937 @cindex @code{ \r} (carriage return character)
2938 @cindex carriage return (@code{backslash-r})
2940 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2943 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2944 @c other assemblers.
2946 @cindex @code{ \t} (tab)
2947 @cindex tab (@code{\t})
2949 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2952 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2953 @c @item \x @var{digit} @var{digit} @var{digit}
2954 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2956 @cindex @code{ \@var{ddd}} (octal character code)
2957 @cindex octal character code (@code{\@var{ddd}})
2958 @item \ @var{digit} @var{digit} @var{digit}
2959 An octal character code. The numeric code is 3 octal digits.
2960 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2961 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2963 @cindex @code{ \@var{xd...}} (hex character code)
2964 @cindex hex character code (@code{\@var{xd...}})
2965 @item \@code{x} @var{hex-digits...}
2966 A hex character code. All trailing hex digits are combined. Either upper or
2967 lower case @code{x} works.
2969 @cindex @code{ \\} (@samp{\} character)
2970 @cindex backslash (@code{\\})
2972 Represents one @samp{\} character.
2975 @c Represents one @samp{'} (accent acute) character.
2976 @c This is needed in single character literals
2977 @c (@xref{Characters,,Character Constants}.) to represent
2980 @cindex @code{ \"} (doublequote character)
2981 @cindex doublequote (@code{\"})
2983 Represents one @samp{"} character. Needed in strings to represent
2984 this character, because an unescaped @samp{"} would end the string.
2986 @item \ @var{anything-else}
2987 Any other character when escaped by @kbd{\} gives a warning, but
2988 assembles as if the @samp{\} was not present. The idea is that if
2989 you used an escape sequence you clearly didn't want the literal
2990 interpretation of the following character. However @command{@value{AS}} has no
2991 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2992 code and warns you of the fact.
2995 Which characters are escapable, and what those escapes represent,
2996 varies widely among assemblers. The current set is what we think
2997 the BSD 4.2 assembler recognizes, and is a subset of what most C
2998 compilers recognize. If you are in doubt, do not use an escape
3002 @subsubsection Characters
3004 @cindex single character constant
3005 @cindex character, single
3006 @cindex constant, single character
3007 A single character may be written as a single quote immediately
3008 followed by that character. The same escapes apply to characters as
3009 to strings. So if you want to write the character backslash, you
3010 must write @kbd{'\\} where the first @code{\} escapes the second
3011 @code{\}. As you can see, the quote is an acute accent, not a
3012 grave accent. A newline
3014 @ifclear abnormal-separator
3015 (or semicolon @samp{;})
3017 @ifset abnormal-separator
3019 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
3024 immediately following an acute accent is taken as a literal character
3025 and does not count as the end of a statement. The value of a character
3026 constant in a numeric expression is the machine's byte-wide code for
3027 that character. @command{@value{AS}} assumes your character code is ASCII:
3028 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
3031 @subsection Number Constants
3033 @cindex constants, number
3034 @cindex number constants
3035 @command{@value{AS}} distinguishes three kinds of numbers according to how they
3036 are stored in the target machine. @emph{Integers} are numbers that
3037 would fit into an @code{int} in the C language. @emph{Bignums} are
3038 integers, but they are stored in more than 32 bits. @emph{Flonums}
3039 are floating point numbers, described below.
3042 * Integers:: Integers
3047 * Bit Fields:: Bit Fields
3053 @subsubsection Integers
3055 @cindex constants, integer
3057 @cindex binary integers
3058 @cindex integers, binary
3059 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
3060 the binary digits @samp{01}.
3062 @cindex octal integers
3063 @cindex integers, octal
3064 An octal integer is @samp{0} followed by zero or more of the octal
3065 digits (@samp{01234567}).
3067 @cindex decimal integers
3068 @cindex integers, decimal
3069 A decimal integer starts with a non-zero digit followed by zero or
3070 more digits (@samp{0123456789}).
3072 @cindex hexadecimal integers
3073 @cindex integers, hexadecimal
3074 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
3075 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
3077 Integers have the usual values. To denote a negative integer, use
3078 the prefix operator @samp{-} discussed under expressions
3079 (@pxref{Prefix Ops,,Prefix Operators}).
3082 @subsubsection Bignums
3085 @cindex constants, bignum
3086 A @dfn{bignum} has the same syntax and semantics as an integer
3087 except that the number (or its negative) takes more than 32 bits to
3088 represent in binary. The distinction is made because in some places
3089 integers are permitted while bignums are not.
3092 @subsubsection Flonums
3094 @cindex floating point numbers
3095 @cindex constants, floating point
3097 @cindex precision, floating point
3098 A @dfn{flonum} represents a floating point number. The translation is
3099 indirect: a decimal floating point number from the text is converted by
3100 @command{@value{AS}} to a generic binary floating point number of more than
3101 sufficient precision. This generic floating point number is converted
3102 to a particular computer's floating point format (or formats) by a
3103 portion of @command{@value{AS}} specialized to that computer.
3105 A flonum is written by writing (in order)
3110 (@samp{0} is optional on the HPPA.)
3114 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
3116 @kbd{e} is recommended. Case is not important.
3118 @c FIXME: verify if flonum syntax really this vague for most cases
3119 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
3120 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
3123 On the H8/300, Renesas / SuperH SH,
3124 and AMD 29K architectures, the letter must be
3125 one of the letters @samp{DFPRSX} (in upper or lower case).
3127 On the ARC, the letter must be one of the letters @samp{DFRS}
3128 (in upper or lower case).
3130 On the Intel 960 architecture, the letter must be
3131 one of the letters @samp{DFT} (in upper or lower case).
3133 On the HPPA architecture, the letter must be @samp{E} (upper case only).
3137 One of the letters @samp{DFRS} (in upper or lower case).
3140 One of the letters @samp{DFPRSX} (in upper or lower case).
3143 The letter @samp{E} (upper case only).
3146 One of the letters @samp{DFT} (in upper or lower case).
3151 An optional sign: either @samp{+} or @samp{-}.
3154 An optional @dfn{integer part}: zero or more decimal digits.
3157 An optional @dfn{fractional part}: @samp{.} followed by zero
3158 or more decimal digits.
3161 An optional exponent, consisting of:
3165 An @samp{E} or @samp{e}.
3166 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
3167 @c principle this can perfectly well be different on different targets.
3169 Optional sign: either @samp{+} or @samp{-}.
3171 One or more decimal digits.
3176 At least one of the integer part or the fractional part must be
3177 present. The floating point number has the usual base-10 value.
3179 @command{@value{AS}} does all processing using integers. Flonums are computed
3180 independently of any floating point hardware in the computer running
3181 @command{@value{AS}}.
3185 @c Bit fields are written as a general facility but are also controlled
3186 @c by a conditional-compilation flag---which is as of now (21mar91)
3187 @c turned on only by the i960 config of GAS.
3189 @subsubsection Bit Fields
3192 @cindex constants, bit field
3193 You can also define numeric constants as @dfn{bit fields}.
3194 Specify two numbers separated by a colon---
3196 @var{mask}:@var{value}
3199 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
3202 The resulting number is then packed
3204 @c this conditional paren in case bit fields turned on elsewhere than 960
3205 (in host-dependent byte order)
3207 into a field whose width depends on which assembler directive has the
3208 bit-field as its argument. Overflow (a result from the bitwise and
3209 requiring more binary digits to represent) is not an error; instead,
3210 more constants are generated, of the specified width, beginning with the
3211 least significant digits.@refill
3213 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
3214 @code{.short}, and @code{.word} accept bit-field arguments.
3219 @chapter Sections and Relocation
3224 * Secs Background:: Background
3225 * Ld Sections:: Linker Sections
3226 * As Sections:: Assembler Internal Sections
3227 * Sub-Sections:: Sub-Sections
3231 @node Secs Background
3234 Roughly, a section is a range of addresses, with no gaps; all data
3235 ``in'' those addresses is treated the same for some particular purpose.
3236 For example there may be a ``read only'' section.
3238 @cindex linker, and assembler
3239 @cindex assembler, and linker
3240 The linker @code{@value{LD}} reads many object files (partial programs) and
3241 combines their contents to form a runnable program. When @command{@value{AS}}
3242 emits an object file, the partial program is assumed to start at address 0.
3243 @code{@value{LD}} assigns the final addresses for the partial program, so that
3244 different partial programs do not overlap. This is actually an
3245 oversimplification, but it suffices to explain how @command{@value{AS}} uses
3248 @code{@value{LD}} moves blocks of bytes of your program to their run-time
3249 addresses. These blocks slide to their run-time addresses as rigid
3250 units; their length does not change and neither does the order of bytes
3251 within them. Such a rigid unit is called a @emph{section}. Assigning
3252 run-time addresses to sections is called @dfn{relocation}. It includes
3253 the task of adjusting mentions of object-file addresses so they refer to
3254 the proper run-time addresses.
3256 For the H8/300, and for the Renesas / SuperH SH,
3257 @command{@value{AS}} pads sections if needed to
3258 ensure they end on a word (sixteen bit) boundary.
3261 @cindex standard assembler sections
3262 An object file written by @command{@value{AS}} has at least three sections, any
3263 of which may be empty. These are named @dfn{text}, @dfn{data} and
3268 When it generates COFF or ELF output,
3270 @command{@value{AS}} can also generate whatever other named sections you specify
3271 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
3272 If you do not use any directives that place output in the @samp{.text}
3273 or @samp{.data} sections, these sections still exist, but are empty.
3278 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
3280 @command{@value{AS}} can also generate whatever other named sections you
3281 specify using the @samp{.space} and @samp{.subspace} directives. See
3282 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3283 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3284 assembler directives.
3287 Additionally, @command{@value{AS}} uses different names for the standard
3288 text, data, and bss sections when generating SOM output. Program text
3289 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3290 BSS into @samp{$BSS$}.
3294 Within the object file, the text section starts at address @code{0}, the
3295 data section follows, and the bss section follows the data section.
3298 When generating either SOM or ELF output files on the HPPA, the text
3299 section starts at address @code{0}, the data section at address
3300 @code{0x4000000}, and the bss section follows the data section.
3303 To let @code{@value{LD}} know which data changes when the sections are
3304 relocated, and how to change that data, @command{@value{AS}} also writes to the
3305 object file details of the relocation needed. To perform relocation
3306 @code{@value{LD}} must know, each time an address in the object
3310 Where in the object file is the beginning of this reference to
3313 How long (in bytes) is this reference?
3315 Which section does the address refer to? What is the numeric value of
3317 (@var{address}) @minus{} (@var{start-address of section})?
3320 Is the reference to an address ``Program-Counter relative''?
3323 @cindex addresses, format of
3324 @cindex section-relative addressing
3325 In fact, every address @command{@value{AS}} ever uses is expressed as
3327 (@var{section}) + (@var{offset into section})
3330 Further, most expressions @command{@value{AS}} computes have this section-relative
3333 (For some object formats, such as SOM for the HPPA, some expressions are
3334 symbol-relative instead.)
3337 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3338 @var{N} into section @var{secname}.''
3340 Apart from text, data and bss sections you need to know about the
3341 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3342 addresses in the absolute section remain unchanged. For example, address
3343 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3344 @code{@value{LD}}. Although the linker never arranges two partial programs'
3345 data sections with overlapping addresses after linking, @emph{by definition}
3346 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3347 part of a program is always the same address when the program is running as
3348 address @code{@{absolute@ 239@}} in any other part of the program.
3350 The idea of sections is extended to the @dfn{undefined} section. Any
3351 address whose section is unknown at assembly time is by definition
3352 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3353 Since numbers are always defined, the only way to generate an undefined
3354 address is to mention an undefined symbol. A reference to a named
3355 common block would be such a symbol: its value is unknown at assembly
3356 time so it has section @emph{undefined}.
3358 By analogy the word @emph{section} is used to describe groups of sections in
3359 the linked program. @code{@value{LD}} puts all partial programs' text
3360 sections in contiguous addresses in the linked program. It is
3361 customary to refer to the @emph{text section} of a program, meaning all
3362 the addresses of all partial programs' text sections. Likewise for
3363 data and bss sections.
3365 Some sections are manipulated by @code{@value{LD}}; others are invented for
3366 use of @command{@value{AS}} and have no meaning except during assembly.
3369 @section Linker Sections
3370 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3375 @cindex named sections
3376 @cindex sections, named
3377 @item named sections
3380 @cindex text section
3381 @cindex data section
3385 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3386 separate but equal sections. Anything you can say of one section is
3389 When the program is running, however, it is
3390 customary for the text section to be unalterable. The
3391 text section is often shared among processes: it contains
3392 instructions, constants and the like. The data section of a running
3393 program is usually alterable: for example, C variables would be stored
3394 in the data section.
3399 This section contains zeroed bytes when your program begins running. It
3400 is used to hold uninitialized variables or common storage. The length of
3401 each partial program's bss section is important, but because it starts
3402 out containing zeroed bytes there is no need to store explicit zero
3403 bytes in the object file. The bss section was invented to eliminate
3404 those explicit zeros from object files.
3406 @cindex absolute section
3407 @item absolute section
3408 Address 0 of this section is always ``relocated'' to runtime address 0.
3409 This is useful if you want to refer to an address that @code{@value{LD}} must
3410 not change when relocating. In this sense we speak of absolute
3411 addresses being ``unrelocatable'': they do not change during relocation.
3413 @cindex undefined section
3414 @item undefined section
3415 This ``section'' is a catch-all for address references to objects not in
3416 the preceding sections.
3417 @c FIXME: ref to some other doc on obj-file formats could go here.
3420 @cindex relocation example
3421 An idealized example of three relocatable sections follows.
3423 The example uses the traditional section names @samp{.text} and @samp{.data}.
3425 Memory addresses are on the horizontal axis.
3429 @c END TEXI2ROFF-KILL
3432 partial program # 1: |ttttt|dddd|00|
3439 partial program # 2: |TTT|DDD|000|
3442 +--+---+-----+--+----+---+-----+~~
3443 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3444 +--+---+-----+--+----+---+-----+~~
3446 addresses: 0 @dots{}
3453 \line{\it Partial program \#1: \hfil}
3454 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3455 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3457 \line{\it Partial program \#2: \hfil}
3458 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3459 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3461 \line{\it linked program: \hfil}
3462 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3463 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3464 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3465 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3467 \line{\it addresses: \hfil}
3471 @c END TEXI2ROFF-KILL
3474 @section Assembler Internal Sections
3476 @cindex internal assembler sections
3477 @cindex sections in messages, internal
3478 These sections are meant only for the internal use of @command{@value{AS}}. They
3479 have no meaning at run-time. You do not really need to know about these
3480 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3481 warning messages, so it might be helpful to have an idea of their
3482 meanings to @command{@value{AS}}. These sections are used to permit the
3483 value of every expression in your assembly language program to be a
3484 section-relative address.
3487 @cindex assembler internal logic error
3488 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3489 An internal assembler logic error has been found. This means there is a
3490 bug in the assembler.
3492 @cindex expr (internal section)
3494 The assembler stores complex expression internally as combinations of
3495 symbols. When it needs to represent an expression as a symbol, it puts
3496 it in the expr section.
3498 @c FIXME item transfer[t] vector preload
3499 @c FIXME item transfer[t] vector postload
3500 @c FIXME item register
3504 @section Sub-Sections
3506 @cindex numbered subsections
3507 @cindex grouping data
3513 fall into two sections: text and data.
3515 You may have separate groups of
3517 data in named sections
3521 data in named sections
3527 that you want to end up near to each other in the object file, even though they
3528 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3529 use @dfn{subsections} for this purpose. Within each section, there can be
3530 numbered subsections with values from 0 to 8192. Objects assembled into the
3531 same subsection go into the object file together with other objects in the same
3532 subsection. For example, a compiler might want to store constants in the text
3533 section, but might not want to have them interspersed with the program being
3534 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3535 section of code being output, and a @samp{.text 1} before each group of
3536 constants being output.
3538 Subsections are optional. If you do not use subsections, everything
3539 goes in subsection number zero.
3542 Each subsection is zero-padded up to a multiple of four bytes.
3543 (Subsections may be padded a different amount on different flavors
3544 of @command{@value{AS}}.)
3548 On the H8/300 platform, each subsection is zero-padded to a word
3549 boundary (two bytes).
3550 The same is true on the Renesas SH.
3553 @c FIXME section padding (alignment)?
3554 @c Rich Pixley says padding here depends on target obj code format; that
3555 @c doesn't seem particularly useful to say without further elaboration,
3556 @c so for now I say nothing about it. If this is a generic BFD issue,
3557 @c these paragraphs might need to vanish from this manual, and be
3558 @c discussed in BFD chapter of binutils (or some such).
3562 Subsections appear in your object file in numeric order, lowest numbered
3563 to highest. (All this to be compatible with other people's assemblers.)
3564 The object file contains no representation of subsections; @code{@value{LD}} and
3565 other programs that manipulate object files see no trace of them.
3566 They just see all your text subsections as a text section, and all your
3567 data subsections as a data section.
3569 To specify which subsection you want subsequent statements assembled
3570 into, use a numeric argument to specify it, in a @samp{.text
3571 @var{expression}} or a @samp{.data @var{expression}} statement.
3574 When generating COFF output, you
3579 can also use an extra subsection
3580 argument with arbitrary named sections: @samp{.section @var{name},
3585 When generating ELF output, you
3590 can also use the @code{.subsection} directive (@pxref{SubSection})
3591 to specify a subsection: @samp{.subsection @var{expression}}.
3593 @var{Expression} should be an absolute expression
3594 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3595 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3596 begins in @code{text 0}. For instance:
3598 .text 0 # The default subsection is text 0 anyway.
3599 .ascii "This lives in the first text subsection. *"
3601 .ascii "But this lives in the second text subsection."
3603 .ascii "This lives in the data section,"
3604 .ascii "in the first data subsection."
3606 .ascii "This lives in the first text section,"
3607 .ascii "immediately following the asterisk (*)."
3610 Each section has a @dfn{location counter} incremented by one for every byte
3611 assembled into that section. Because subsections are merely a convenience
3612 restricted to @command{@value{AS}} there is no concept of a subsection location
3613 counter. There is no way to directly manipulate a location counter---but the
3614 @code{.align} directive changes it, and any label definition captures its
3615 current value. The location counter of the section where statements are being
3616 assembled is said to be the @dfn{active} location counter.
3619 @section bss Section
3622 @cindex common variable storage
3623 The bss section is used for local common variable storage.
3624 You may allocate address space in the bss section, but you may
3625 not dictate data to load into it before your program executes. When
3626 your program starts running, all the contents of the bss
3627 section are zeroed bytes.
3629 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3630 @ref{Lcomm,,@code{.lcomm}}.
3632 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3633 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3636 When assembling for a target which supports multiple sections, such as ELF or
3637 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3638 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3639 section. Typically the section will only contain symbol definitions and
3640 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3647 Symbols are a central concept: the programmer uses symbols to name
3648 things, the linker uses symbols to link, and the debugger uses symbols
3652 @cindex debuggers, and symbol order
3653 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3654 the same order they were declared. This may break some debuggers.
3659 * Setting Symbols:: Giving Symbols Other Values
3660 * Symbol Names:: Symbol Names
3661 * Dot:: The Special Dot Symbol
3662 * Symbol Attributes:: Symbol Attributes
3669 A @dfn{label} is written as a symbol immediately followed by a colon
3670 @samp{:}. The symbol then represents the current value of the
3671 active location counter, and is, for example, a suitable instruction
3672 operand. You are warned if you use the same symbol to represent two
3673 different locations: the first definition overrides any other
3677 On the HPPA, the usual form for a label need not be immediately followed by a
3678 colon, but instead must start in column zero. Only one label may be defined on
3679 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3680 provides a special directive @code{.label} for defining labels more flexibly.
3683 @node Setting Symbols
3684 @section Giving Symbols Other Values
3686 @cindex assigning values to symbols
3687 @cindex symbol values, assigning
3688 A symbol can be given an arbitrary value by writing a symbol, followed
3689 by an equals sign @samp{=}, followed by an expression
3690 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3691 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3692 equals sign @samp{=}@samp{=} here represents an equivalent of the
3693 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3696 Blackfin does not support symbol assignment with @samp{=}.
3700 @section Symbol Names
3702 @cindex symbol names
3703 @cindex names, symbol
3704 @ifclear SPECIAL-SYMS
3705 Symbol names begin with a letter or with one of @samp{._}. On most
3706 machines, you can also use @code{$} in symbol names; exceptions are
3707 noted in @ref{Machine Dependencies}. That character may be followed by any
3708 string of digits, letters, dollar signs (unless otherwise noted for a
3709 particular target machine), and underscores.
3713 Symbol names begin with a letter or with one of @samp{._}. On the
3714 Renesas SH you can also use @code{$} in symbol names. That
3715 character may be followed by any string of digits, letters, dollar signs (save
3716 on the H8/300), and underscores.
3720 Case of letters is significant: @code{foo} is a different symbol name
3723 Symbol names do not start with a digit. An exception to this rule is made for
3724 Local Labels. See below.
3726 Multibyte characters are supported. To generate a symbol name containing
3727 multibyte characters enclose it within double quotes and use escape codes. cf
3728 @xref{Strings}. Generating a multibyte symbol name from a label is not
3729 currently supported.
3731 Each symbol has exactly one name. Each name in an assembly language program
3732 refers to exactly one symbol. You may use that symbol name any number of times
3735 @subheading Local Symbol Names
3737 @cindex local symbol names
3738 @cindex symbol names, local
3739 A local symbol is any symbol beginning with certain local label prefixes.
3740 By default, the local label prefix is @samp{.L} for ELF systems or
3741 @samp{L} for traditional a.out systems, but each target may have its own
3742 set of local label prefixes.
3744 On the HPPA local symbols begin with @samp{L$}.
3747 Local symbols are defined and used within the assembler, but they are
3748 normally not saved in object files. Thus, they are not visible when debugging.
3749 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols})
3750 to retain the local symbols in the object files.
3752 @subheading Local Labels
3754 @cindex local labels
3755 @cindex temporary symbol names
3756 @cindex symbol names, temporary
3757 Local labels are different from local symbols. Local labels help compilers and
3758 programmers use names temporarily. They create symbols which are guaranteed to
3759 be unique over the entire scope of the input source code and which can be
3760 referred to by a simple notation. To define a local label, write a label of
3761 the form @samp{@b{N}:} (where @b{N} represents any non-negative integer).
3762 To refer to the most recent previous definition of that label write
3763 @samp{@b{N}b}, using the same number as when you defined the label. To refer
3764 to the next definition of a local label, write @samp{@b{N}f}. The @samp{b}
3765 stands for ``backwards'' and the @samp{f} stands for ``forwards''.
3767 There is no restriction on how you can use these labels, and you can reuse them
3768 too. So that it is possible to repeatedly define the same local label (using
3769 the same number @samp{@b{N}}), although you can only refer to the most recently
3770 defined local label of that number (for a backwards reference) or the next
3771 definition of a specific local label for a forward reference. It is also worth
3772 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3773 implemented in a slightly more efficient manner than the others.
3784 Which is the equivalent of:
3787 label_1: branch label_3
3788 label_2: branch label_1
3789 label_3: branch label_4
3790 label_4: branch label_3
3793 Local label names are only a notational device. They are immediately
3794 transformed into more conventional symbol names before the assembler uses them.
3795 The symbol names are stored in the symbol table, appear in error messages, and
3796 are optionally emitted to the object file. The names are constructed using
3800 @item @emph{local label prefix}
3801 All local symbols begin with the system-specific local label prefix.
3802 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3803 that start with the local label prefix. These labels are
3804 used for symbols you are never intended to see. If you use the
3805 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3806 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3807 you may use them in debugging.
3810 This is the number that was used in the local label definition. So if the
3811 label is written @samp{55:} then the number is @samp{55}.
3814 This unusual character is included so you do not accidentally invent a symbol
3815 of the same name. The character has ASCII value of @samp{\002} (control-B).
3817 @item @emph{ordinal number}
3818 This is a serial number to keep the labels distinct. The first definition of
3819 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3820 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3821 the number @samp{1} and its 15th definition gets @samp{15} as well.
3824 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3825 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3827 @subheading Dollar Local Labels
3828 @cindex dollar local symbols
3830 On some targets @code{@value{AS}} also supports an even more local form of
3831 local labels called dollar labels. These labels go out of scope (i.e., they
3832 become undefined) as soon as a non-local label is defined. Thus they remain
3833 valid for only a small region of the input source code. Normal local labels,
3834 by contrast, remain in scope for the entire file, or until they are redefined
3835 by another occurrence of the same local label.
3837 Dollar labels are defined in exactly the same way as ordinary local labels,
3838 except that they have a dollar sign suffix to their numeric value, e.g.,
3841 They can also be distinguished from ordinary local labels by their transformed
3842 names which use ASCII character @samp{\001} (control-A) as the magic character
3843 to distinguish them from ordinary labels. For example, the fifth definition of
3844 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3847 @section The Special Dot Symbol
3849 @cindex dot (symbol)
3850 @cindex @code{.} (symbol)
3851 @cindex current address
3852 @cindex location counter
3853 The special symbol @samp{.} refers to the current address that
3854 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3855 .long .} defines @code{melvin} to contain its own address.
3856 Assigning a value to @code{.} is treated the same as a @code{.org}
3858 @ifclear no-space-dir
3859 Thus, the expression @samp{.=.+4} is the same as saying
3863 @node Symbol Attributes
3864 @section Symbol Attributes
3866 @cindex symbol attributes
3867 @cindex attributes, symbol
3868 Every symbol has, as well as its name, the attributes ``Value'' and
3869 ``Type''. Depending on output format, symbols can also have auxiliary
3872 The detailed definitions are in @file{a.out.h}.
3875 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3876 all these attributes, and probably won't warn you. This makes the
3877 symbol an externally defined symbol, which is generally what you
3881 * Symbol Value:: Value
3882 * Symbol Type:: Type
3885 * a.out Symbols:: Symbol Attributes: @code{a.out}
3889 * a.out Symbols:: Symbol Attributes: @code{a.out}
3892 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3897 * COFF Symbols:: Symbol Attributes for COFF
3900 * SOM Symbols:: Symbol Attributes for SOM
3907 @cindex value of a symbol
3908 @cindex symbol value
3909 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3910 location in the text, data, bss or absolute sections the value is the
3911 number of addresses from the start of that section to the label.
3912 Naturally for text, data and bss sections the value of a symbol changes
3913 as @code{@value{LD}} changes section base addresses during linking. Absolute
3914 symbols' values do not change during linking: that is why they are
3917 The value of an undefined symbol is treated in a special way. If it is
3918 0 then the symbol is not defined in this assembler source file, and
3919 @code{@value{LD}} tries to determine its value from other files linked into the
3920 same program. You make this kind of symbol simply by mentioning a symbol
3921 name without defining it. A non-zero value represents a @code{.comm}
3922 common declaration. The value is how much common storage to reserve, in
3923 bytes (addresses). The symbol refers to the first address of the
3929 @cindex type of a symbol
3931 The type attribute of a symbol contains relocation (section)
3932 information, any flag settings indicating that a symbol is external, and
3933 (optionally), other information for linkers and debuggers. The exact
3934 format depends on the object-code output format in use.
3939 @c The following avoids a "widow" subsection title. @group would be
3940 @c better if it were available outside examples.
3943 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3945 @cindex @code{b.out} symbol attributes
3946 @cindex symbol attributes, @code{b.out}
3947 These symbol attributes appear only when @command{@value{AS}} is configured for
3948 one of the Berkeley-descended object output formats---@code{a.out} or
3954 @subsection Symbol Attributes: @code{a.out}
3956 @cindex @code{a.out} symbol attributes
3957 @cindex symbol attributes, @code{a.out}
3963 @subsection Symbol Attributes: @code{a.out}
3965 @cindex @code{a.out} symbol attributes
3966 @cindex symbol attributes, @code{a.out}
3970 * Symbol Desc:: Descriptor
3971 * Symbol Other:: Other
3975 @subsubsection Descriptor
3977 @cindex descriptor, of @code{a.out} symbol
3978 This is an arbitrary 16-bit value. You may establish a symbol's
3979 descriptor value by using a @code{.desc} statement
3980 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3981 @command{@value{AS}}.
3984 @subsubsection Other
3986 @cindex other attribute, of @code{a.out} symbol
3987 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3992 @subsection Symbol Attributes for COFF
3994 @cindex COFF symbol attributes
3995 @cindex symbol attributes, COFF
3997 The COFF format supports a multitude of auxiliary symbol attributes;
3998 like the primary symbol attributes, they are set between @code{.def} and
3999 @code{.endef} directives.
4001 @subsubsection Primary Attributes
4003 @cindex primary attributes, COFF symbols
4004 The symbol name is set with @code{.def}; the value and type,
4005 respectively, with @code{.val} and @code{.type}.
4007 @subsubsection Auxiliary Attributes
4009 @cindex auxiliary attributes, COFF symbols
4010 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
4011 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
4012 table information for COFF.
4017 @subsection Symbol Attributes for SOM
4019 @cindex SOM symbol attributes
4020 @cindex symbol attributes, SOM
4022 The SOM format for the HPPA supports a multitude of symbol attributes set with
4023 the @code{.EXPORT} and @code{.IMPORT} directives.
4025 The attributes are described in @cite{HP9000 Series 800 Assembly
4026 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
4027 @code{EXPORT} assembler directive documentation.
4031 @chapter Expressions
4035 @cindex numeric values
4036 An @dfn{expression} specifies an address or numeric value.
4037 Whitespace may precede and/or follow an expression.
4039 The result of an expression must be an absolute number, or else an offset into
4040 a particular section. If an expression is not absolute, and there is not
4041 enough information when @command{@value{AS}} sees the expression to know its
4042 section, a second pass over the source program might be necessary to interpret
4043 the expression---but the second pass is currently not implemented.
4044 @command{@value{AS}} aborts with an error message in this situation.
4047 * Empty Exprs:: Empty Expressions
4048 * Integer Exprs:: Integer Expressions
4052 @section Empty Expressions
4054 @cindex empty expressions
4055 @cindex expressions, empty
4056 An empty expression has no value: it is just whitespace or null.
4057 Wherever an absolute expression is required, you may omit the
4058 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
4059 is compatible with other assemblers.
4062 @section Integer Expressions
4064 @cindex integer expressions
4065 @cindex expressions, integer
4066 An @dfn{integer expression} is one or more @emph{arguments} delimited
4067 by @emph{operators}.
4070 * Arguments:: Arguments
4071 * Operators:: Operators
4072 * Prefix Ops:: Prefix Operators
4073 * Infix Ops:: Infix Operators
4077 @subsection Arguments
4079 @cindex expression arguments
4080 @cindex arguments in expressions
4081 @cindex operands in expressions
4082 @cindex arithmetic operands
4083 @dfn{Arguments} are symbols, numbers or subexpressions. In other
4084 contexts arguments are sometimes called ``arithmetic operands''. In
4085 this manual, to avoid confusing them with the ``instruction operands'' of
4086 the machine language, we use the term ``argument'' to refer to parts of
4087 expressions only, reserving the word ``operand'' to refer only to machine
4088 instruction operands.
4090 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
4091 @var{section} is one of text, data, bss, absolute,
4092 or undefined. @var{NNN} is a signed, 2's complement 32 bit
4095 Numbers are usually integers.
4097 A number can be a flonum or bignum. In this case, you are warned
4098 that only the low order 32 bits are used, and @command{@value{AS}} pretends
4099 these 32 bits are an integer. You may write integer-manipulating
4100 instructions that act on exotic constants, compatible with other
4103 @cindex subexpressions
4104 Subexpressions are a left parenthesis @samp{(} followed by an integer
4105 expression, followed by a right parenthesis @samp{)}; or a prefix
4106 operator followed by an argument.
4109 @subsection Operators
4111 @cindex operators, in expressions
4112 @cindex arithmetic functions
4113 @cindex functions, in expressions
4114 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
4115 operators are followed by an argument. Infix operators appear
4116 between their arguments. Operators may be preceded and/or followed by
4120 @subsection Prefix Operator
4122 @cindex prefix operators
4123 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
4124 one argument, which must be absolute.
4126 @c the tex/end tex stuff surrounding this small table is meant to make
4127 @c it align, on the printed page, with the similar table in the next
4128 @c section (which is inside an enumerate).
4130 \global\advance\leftskip by \itemindent
4135 @dfn{Negation}. Two's complement negation.
4137 @dfn{Complementation}. Bitwise not.
4141 \global\advance\leftskip by -\itemindent
4145 @subsection Infix Operators
4147 @cindex infix operators
4148 @cindex operators, permitted arguments
4149 @dfn{Infix operators} take two arguments, one on either side. Operators
4150 have precedence, but operations with equal precedence are performed left
4151 to right. Apart from @code{+} or @option{-}, both arguments must be
4152 absolute, and the result is absolute.
4155 @cindex operator precedence
4156 @cindex precedence of operators
4163 @dfn{Multiplication}.
4166 @dfn{Division}. Truncation is the same as the C operator @samp{/}
4172 @dfn{Shift Left}. Same as the C operator @samp{<<}.
4175 @dfn{Shift Right}. Same as the C operator @samp{>>}.
4179 Intermediate precedence
4184 @dfn{Bitwise Inclusive Or}.
4190 @dfn{Bitwise Exclusive Or}.
4193 @dfn{Bitwise Or Not}.
4200 @cindex addition, permitted arguments
4201 @cindex plus, permitted arguments
4202 @cindex arguments for addition
4204 @dfn{Addition}. If either argument is absolute, the result has the section of
4205 the other argument. You may not add together arguments from different
4208 @cindex subtraction, permitted arguments
4209 @cindex minus, permitted arguments
4210 @cindex arguments for subtraction
4212 @dfn{Subtraction}. If the right argument is absolute, the
4213 result has the section of the left argument.
4214 If both arguments are in the same section, the result is absolute.
4215 You may not subtract arguments from different sections.
4216 @c FIXME is there still something useful to say about undefined - undefined ?
4218 @cindex comparison expressions
4219 @cindex expressions, comparison
4224 @dfn{Is Not Equal To}
4228 @dfn{Is Greater Than}
4230 @dfn{Is Greater Than Or Equal To}
4232 @dfn{Is Less Than Or Equal To}
4234 The comparison operators can be used as infix operators. A true results has a
4235 value of -1 whereas a false result has a value of 0. Note, these operators
4236 perform signed comparisons.
4239 @item Lowest Precedence
4248 These two logical operations can be used to combine the results of sub
4249 expressions. Note, unlike the comparison operators a true result returns a
4250 value of 1 but a false results does still return 0. Also note that the logical
4251 or operator has a slightly lower precedence than logical and.
4256 In short, it's only meaningful to add or subtract the @emph{offsets} in an
4257 address; you can only have a defined section in one of the two arguments.
4260 @chapter Assembler Directives
4262 @cindex directives, machine independent
4263 @cindex pseudo-ops, machine independent
4264 @cindex machine independent directives
4265 All assembler directives have names that begin with a period (@samp{.}).
4266 The names are case insensitive for most targets, and usually written
4269 This chapter discusses directives that are available regardless of the
4270 target machine configuration for the @sc{gnu} assembler.
4272 Some machine configurations provide additional directives.
4273 @xref{Machine Dependencies}.
4276 @ifset machine-directives
4277 @xref{Machine Dependencies}, for additional directives.
4282 * Abort:: @code{.abort}
4284 * ABORT (COFF):: @code{.ABORT}
4287 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
4288 * Altmacro:: @code{.altmacro}
4289 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4290 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4291 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
4292 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, etc
4293 * Byte:: @code{.byte @var{expressions}}
4294 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4295 * Comm:: @code{.comm @var{symbol} , @var{length} }
4296 * Data:: @code{.data @var{subsection}}
4298 * Def:: @code{.def @var{name}}
4301 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4307 * Double:: @code{.double @var{flonums}}
4308 * Eject:: @code{.eject}
4309 * Else:: @code{.else}
4310 * Elseif:: @code{.elseif}
4313 * Endef:: @code{.endef}
4316 * Endfunc:: @code{.endfunc}
4317 * Endif:: @code{.endif}
4318 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4319 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4320 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4322 * Error:: @code{.error @var{string}}
4323 * Exitm:: @code{.exitm}
4324 * Extern:: @code{.extern}
4325 * Fail:: @code{.fail}
4326 * File:: @code{.file}
4327 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4328 * Float:: @code{.float @var{flonums}}
4329 * Func:: @code{.func}
4330 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4332 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4333 * Hidden:: @code{.hidden @var{names}}
4336 * hword:: @code{.hword @var{expressions}}
4337 * Ident:: @code{.ident}
4338 * If:: @code{.if @var{absolute expression}}
4339 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4340 * Include:: @code{.include "@var{file}"}
4341 * Int:: @code{.int @var{expressions}}
4343 * Internal:: @code{.internal @var{names}}
4346 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4347 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4348 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4349 * Lflags:: @code{.lflags}
4350 @ifclear no-line-dir
4351 * Line:: @code{.line @var{line-number}}
4354 * Linkonce:: @code{.linkonce [@var{type}]}
4355 * List:: @code{.list}
4356 * Ln:: @code{.ln @var{line-number}}
4357 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4358 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4360 * Local:: @code{.local @var{names}}
4363 * Long:: @code{.long @var{expressions}}
4365 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4368 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4369 * MRI:: @code{.mri @var{val}}
4370 * Noaltmacro:: @code{.noaltmacro}
4371 * Nolist:: @code{.nolist}
4372 * Octa:: @code{.octa @var{bignums}}
4373 * Offset:: @code{.offset @var{loc}}
4374 * Org:: @code{.org @var{new-lc}, @var{fill}}
4375 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4377 * PopSection:: @code{.popsection}
4378 * Previous:: @code{.previous}
4381 * Print:: @code{.print @var{string}}
4383 * Protected:: @code{.protected @var{names}}
4386 * Psize:: @code{.psize @var{lines}, @var{columns}}
4387 * Purgem:: @code{.purgem @var{name}}
4389 * PushSection:: @code{.pushsection @var{name}}
4392 * Quad:: @code{.quad @var{bignums}}
4393 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4394 * Rept:: @code{.rept @var{count}}
4395 * Sbttl:: @code{.sbttl "@var{subheading}"}
4397 * Scl:: @code{.scl @var{class}}
4400 * Section:: @code{.section @var{name}[, @var{flags}]}
4403 * Set:: @code{.set @var{symbol}, @var{expression}}
4404 * Short:: @code{.short @var{expressions}}
4405 * Single:: @code{.single @var{flonums}}
4407 * Size:: @code{.size [@var{name} , @var{expression}]}
4409 @ifclear no-space-dir
4410 * Skip:: @code{.skip @var{size} , @var{fill}}
4413 * Sleb128:: @code{.sleb128 @var{expressions}}
4414 @ifclear no-space-dir
4415 * Space:: @code{.space @var{size} , @var{fill}}
4418 * Stab:: @code{.stabd, .stabn, .stabs}
4421 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4422 * Struct:: @code{.struct @var{expression}}
4424 * SubSection:: @code{.subsection}
4425 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4429 * Tag:: @code{.tag @var{structname}}
4432 * Text:: @code{.text @var{subsection}}
4433 * Title:: @code{.title "@var{heading}"}
4435 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4438 * Uleb128:: @code{.uleb128 @var{expressions}}
4440 * Val:: @code{.val @var{addr}}
4444 * Version:: @code{.version "@var{string}"}
4445 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4446 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4449 * Warning:: @code{.warning @var{string}}
4450 * Weak:: @code{.weak @var{names}}
4451 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4452 * Word:: @code{.word @var{expressions}}
4453 @ifclear no-space-dir
4454 * Zero:: @code{.zero @var{size}}
4456 * Deprecated:: Deprecated Directives
4460 @section @code{.abort}
4462 @cindex @code{abort} directive
4463 @cindex stopping the assembly
4464 This directive stops the assembly immediately. It is for
4465 compatibility with other assemblers. The original idea was that the
4466 assembly language source would be piped into the assembler. If the sender
4467 of the source quit, it could use this directive tells @command{@value{AS}} to
4468 quit also. One day @code{.abort} will not be supported.
4472 @section @code{.ABORT} (COFF)
4474 @cindex @code{ABORT} directive
4475 When producing COFF output, @command{@value{AS}} accepts this directive as a
4476 synonym for @samp{.abort}.
4479 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
4485 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4487 @cindex padding the location counter
4488 @cindex @code{align} directive
4489 Pad the location counter (in the current subsection) to a particular storage
4490 boundary. The first expression (which must be absolute) is the alignment
4491 required, as described below.
4493 The second expression (also absolute) gives the fill value to be stored in the
4494 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4495 padding bytes are normally zero. However, on some systems, if the section is
4496 marked as containing code and the fill value is omitted, the space is filled
4497 with no-op instructions.
4499 The third expression is also absolute, and is also optional. If it is present,
4500 it is the maximum number of bytes that should be skipped by this alignment
4501 directive. If doing the alignment would require skipping more bytes than the
4502 specified maximum, then the alignment is not done at all. You can omit the
4503 fill value (the second argument) entirely by simply using two commas after the
4504 required alignment; this can be useful if you want the alignment to be filled
4505 with no-op instructions when appropriate.
4507 The way the required alignment is specified varies from system to system.
4508 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or1k,
4509 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4510 alignment request in bytes. For example @samp{.align 8} advances
4511 the location counter until it is a multiple of 8. If the location counter
4512 is already a multiple of 8, no change is needed. For the tic54x, the
4513 first expression is the alignment request in words.
4515 For other systems, including ppc, i386 using a.out format, arm and
4516 strongarm, it is the
4517 number of low-order zero bits the location counter must have after
4518 advancement. For example @samp{.align 3} advances the location
4519 counter until it a multiple of 8. If the location counter is already a
4520 multiple of 8, no change is needed.
4522 This inconsistency is due to the different behaviors of the various
4523 native assemblers for these systems which GAS must emulate.
4524 GAS also provides @code{.balign} and @code{.p2align} directives,
4525 described later, which have a consistent behavior across all
4526 architectures (but are specific to GAS).
4529 @section @code{.altmacro}
4530 Enable alternate macro mode, enabling:
4533 @item LOCAL @var{name} [ , @dots{} ]
4534 One additional directive, @code{LOCAL}, is available. It is used to
4535 generate a string replacement for each of the @var{name} arguments, and
4536 replace any instances of @var{name} in each macro expansion. The
4537 replacement string is unique in the assembly, and different for each
4538 separate macro expansion. @code{LOCAL} allows you to write macros that
4539 define symbols, without fear of conflict between separate macro expansions.
4541 @item String delimiters
4542 You can write strings delimited in these other ways besides
4543 @code{"@var{string}"}:
4546 @item '@var{string}'
4547 You can delimit strings with single-quote characters.
4549 @item <@var{string}>
4550 You can delimit strings with matching angle brackets.
4553 @item single-character string escape
4554 To include any single character literally in a string (even if the
4555 character would otherwise have some special meaning), you can prefix the
4556 character with @samp{!} (an exclamation mark). For example, you can
4557 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4559 @item Expression results as strings
4560 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4561 and use the result as a string.
4565 @section @code{.ascii "@var{string}"}@dots{}
4567 @cindex @code{ascii} directive
4568 @cindex string literals
4569 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4570 separated by commas. It assembles each string (with no automatic
4571 trailing zero byte) into consecutive addresses.
4574 @section @code{.asciz "@var{string}"}@dots{}
4576 @cindex @code{asciz} directive
4577 @cindex zero-terminated strings
4578 @cindex null-terminated strings
4579 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4580 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4583 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4585 @cindex padding the location counter given number of bytes
4586 @cindex @code{balign} directive
4587 Pad the location counter (in the current subsection) to a particular
4588 storage boundary. The first expression (which must be absolute) is the
4589 alignment request in bytes. For example @samp{.balign 8} advances
4590 the location counter until it is a multiple of 8. If the location counter
4591 is already a multiple of 8, no change is needed.
4593 The second expression (also absolute) gives the fill value to be stored in the
4594 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4595 padding bytes are normally zero. However, on some systems, if the section is
4596 marked as containing code and the fill value is omitted, the space is filled
4597 with no-op instructions.
4599 The third expression is also absolute, and is also optional. If it is present,
4600 it is the maximum number of bytes that should be skipped by this alignment
4601 directive. If doing the alignment would require skipping more bytes than the
4602 specified maximum, then the alignment is not done at all. You can omit the
4603 fill value (the second argument) entirely by simply using two commas after the
4604 required alignment; this can be useful if you want the alignment to be filled
4605 with no-op instructions when appropriate.
4607 @cindex @code{balignw} directive
4608 @cindex @code{balignl} directive
4609 The @code{.balignw} and @code{.balignl} directives are variants of the
4610 @code{.balign} directive. The @code{.balignw} directive treats the fill
4611 pattern as a two byte word value. The @code{.balignl} directives treats the
4612 fill pattern as a four byte longword value. For example, @code{.balignw
4613 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4614 filled in with the value 0x368d (the exact placement of the bytes depends upon
4615 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4618 @node Bundle directives
4619 @section Bundle directives
4620 @subsection @code{.bundle_align_mode @var{abs-expr}}
4621 @cindex @code{bundle_align_mode} directive
4623 @cindex instruction bundle
4624 @cindex aligned instruction bundle
4625 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4626 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4627 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4628 disabled (which is the default state). If the argument it not zero, it
4629 gives the size of an instruction bundle as a power of two (as for the
4630 @code{.p2align} directive, @pxref{P2align}).
4632 For some targets, it's an ABI requirement that no instruction may span a
4633 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4634 instructions that starts on an aligned boundary. For example, if
4635 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4636 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4637 effect, no single instruction may span a boundary between bundles. If an
4638 instruction would start too close to the end of a bundle for the length of
4639 that particular instruction to fit within the bundle, then the space at the
4640 end of that bundle is filled with no-op instructions so the instruction
4641 starts in the next bundle. As a corollary, it's an error if any single
4642 instruction's encoding is longer than the bundle size.
4644 @subsection @code{.bundle_lock} and @code{.bundle_unlock}
4645 @cindex @code{bundle_lock} directive
4646 @cindex @code{bundle_unlock} directive
4647 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4648 allow explicit control over instruction bundle padding. These directives
4649 are only valid when @code{.bundle_align_mode} has been used to enable
4650 aligned instruction bundle mode. It's an error if they appear when
4651 @code{.bundle_align_mode} has not been used at all, or when the last
4652 directive was @w{@code{.bundle_align_mode 0}}.
4654 @cindex bundle-locked
4655 For some targets, it's an ABI requirement that certain instructions may
4656 appear only as part of specified permissible sequences of multiple
4657 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4658 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4659 instruction sequence. For purposes of aligned instruction bundle mode, a
4660 sequence starting with @code{.bundle_lock} and ending with
4661 @code{.bundle_unlock} is treated as a single instruction. That is, the
4662 entire sequence must fit into a single bundle and may not span a bundle
4663 boundary. If necessary, no-op instructions will be inserted before the
4664 first instruction of the sequence so that the whole sequence starts on an
4665 aligned bundle boundary. It's an error if the sequence is longer than the
4668 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4669 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4670 nested. That is, a second @code{.bundle_lock} directive before the next
4671 @code{.bundle_unlock} directive has no effect except that it must be
4672 matched by another closing @code{.bundle_unlock} so that there is the
4673 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4676 @section @code{.byte @var{expressions}}
4678 @cindex @code{byte} directive
4679 @cindex integers, one byte
4680 @code{.byte} expects zero or more expressions, separated by commas.
4681 Each expression is assembled into the next byte.
4683 @node CFI directives
4684 @section CFI directives
4685 @subsection @code{.cfi_sections @var{section_list}}
4686 @cindex @code{cfi_sections} directive
4687 @code{.cfi_sections} may be used to specify whether CFI directives
4688 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4689 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4690 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4691 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4692 directive is not used is @code{.cfi_sections .eh_frame}.
4694 On targets that support compact unwinding tables these can be generated
4695 by specifying @code{.eh_frame_entry} instead of @code{.eh_frame}.
4697 Some targets may support an additional name, such as @code{.c6xabi.exidx}
4698 which is used by the @value{TIC6X} target.
4700 The @code{.cfi_sections} directive can be repeated, with the same or different
4701 arguments, provided that CFI generation has not yet started. Once CFI
4702 generation has started however the section list is fixed and any attempts to
4703 redefine it will result in an error.
4705 @subsection @code{.cfi_startproc [simple]}
4706 @cindex @code{cfi_startproc} directive
4707 @code{.cfi_startproc} is used at the beginning of each function that
4708 should have an entry in @code{.eh_frame}. It initializes some internal
4709 data structures. Don't forget to close the function by
4710 @code{.cfi_endproc}.
4712 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4713 it also emits some architecture dependent initial CFI instructions.
4715 @subsection @code{.cfi_endproc}
4716 @cindex @code{cfi_endproc} directive
4717 @code{.cfi_endproc} is used at the end of a function where it closes its
4718 unwind entry previously opened by
4719 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4721 @subsection @code{.cfi_personality @var{encoding} [, @var{exp}]}
4722 @cindex @code{cfi_personality} directive
4723 @code{.cfi_personality} defines personality routine and its encoding.
4724 @var{encoding} must be a constant determining how the personality
4725 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4726 argument is not present, otherwise second argument should be
4727 a constant or a symbol name. When using indirect encodings,
4728 the symbol provided should be the location where personality
4729 can be loaded from, not the personality routine itself.
4730 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4731 no personality routine.
4733 @subsection @code{.cfi_personality_id @var{id}}
4734 @cindex @code{cfi_personality_id} directive
4735 @code{cfi_personality_id} defines a personality routine by its index as
4736 defined in a compact unwinding format.
4737 Only valid when generating compact EH frames (i.e.
4738 with @code{.cfi_sections eh_frame_entry}.
4740 @subsection @code{.cfi_fde_data [@var{opcode1} [, @dots{}]]}
4741 @cindex @code{cfi_fde_data} directive
4742 @code{cfi_fde_data} is used to describe the compact unwind opcodes to be
4743 used for the current function. These are emitted inline in the
4744 @code{.eh_frame_entry} section if small enough and there is no LSDA, or
4745 in the @code{.gnu.extab} section otherwise.
4746 Only valid when generating compact EH frames (i.e.
4747 with @code{.cfi_sections eh_frame_entry}.
4749 @subsection @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4750 @code{.cfi_lsda} defines LSDA and its encoding.
4751 @var{encoding} must be a constant determining how the LSDA
4752 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), the second
4753 argument is not present, otherwise the second argument should be a constant
4754 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4755 meaning that no LSDA is present.
4757 @subsection @code{.cfi_inline_lsda} [@var{align}]
4758 @code{.cfi_inline_lsda} marks the start of a LSDA data section and
4759 switches to the corresponding @code{.gnu.extab} section.
4760 Must be preceded by a CFI block containing a @code{.cfi_lsda} directive.
4761 Only valid when generating compact EH frames (i.e.
4762 with @code{.cfi_sections eh_frame_entry}.
4764 The table header and unwinding opcodes will be generated at this point,
4765 so that they are immediately followed by the LSDA data. The symbol
4766 referenced by the @code{.cfi_lsda} directive should still be defined
4767 in case a fallback FDE based encoding is used. The LSDA data is terminated
4768 by a section directive.
4770 The optional @var{align} argument specifies the alignment required.
4771 The alignment is specified as a power of two, as with the
4772 @code{.p2align} directive.
4774 @subsection @code{.cfi_def_cfa @var{register}, @var{offset}}
4775 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4776 address from @var{register} and add @var{offset} to it}.
4778 @subsection @code{.cfi_def_cfa_register @var{register}}
4779 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4780 now on @var{register} will be used instead of the old one. Offset
4783 @subsection @code{.cfi_def_cfa_offset @var{offset}}
4784 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4785 remains the same, but @var{offset} is new. Note that it is the
4786 absolute offset that will be added to a defined register to compute
4789 @subsection @code{.cfi_adjust_cfa_offset @var{offset}}
4790 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4791 value that is added/substracted from the previous offset.
4793 @subsection @code{.cfi_offset @var{register}, @var{offset}}
4794 Previous value of @var{register} is saved at offset @var{offset} from
4797 @subsection @code{.cfi_rel_offset @var{register}, @var{offset}}
4798 Previous value of @var{register} is saved at offset @var{offset} from
4799 the current CFA register. This is transformed to @code{.cfi_offset}
4800 using the known displacement of the CFA register from the CFA.
4801 This is often easier to use, because the number will match the
4802 code it's annotating.
4804 @subsection @code{.cfi_register @var{register1}, @var{register2}}
4805 Previous value of @var{register1} is saved in register @var{register2}.
4807 @subsection @code{.cfi_restore @var{register}}
4808 @code{.cfi_restore} says that the rule for @var{register} is now the
4809 same as it was at the beginning of the function, after all initial
4810 instruction added by @code{.cfi_startproc} were executed.
4812 @subsection @code{.cfi_undefined @var{register}}
4813 From now on the previous value of @var{register} can't be restored anymore.
4815 @subsection @code{.cfi_same_value @var{register}}
4816 Current value of @var{register} is the same like in the previous frame,
4817 i.e. no restoration needed.
4819 @subsection @code{.cfi_remember_state} and @code{.cfi_restore_state}
4820 @code{.cfi_remember_state} pushes the set of rules for every register onto an
4821 implicit stack, while @code{.cfi_restore_state} pops them off the stack and
4822 places them in the current row. This is useful for situations where you have
4823 multiple @code{.cfi_*} directives that need to be undone due to the control
4824 flow of the program. For example, we could have something like this (assuming
4825 the CFA is the value of @code{rbp}):
4835 .cfi_def_cfa %rsp, 8
4838 /* Do something else */
4841 Here, we want the @code{.cfi} directives to affect only the rows corresponding
4842 to the instructions before @code{label}. This means we'd have to add multiple
4843 @code{.cfi} directives after @code{label} to recreate the original save
4844 locations of the registers, as well as setting the CFA back to the value of
4845 @code{rbp}. This would be clumsy, and result in a larger binary size. Instead,
4857 .cfi_def_cfa %rsp, 8
4861 /* Do something else */
4864 That way, the rules for the instructions after @code{label} will be the same
4865 as before the first @code{.cfi_restore} without having to use multiple
4866 @code{.cfi} directives.
4868 @subsection @code{.cfi_return_column @var{register}}
4869 Change return column @var{register}, i.e. the return address is either
4870 directly in @var{register} or can be accessed by rules for @var{register}.
4872 @subsection @code{.cfi_signal_frame}
4873 Mark current function as signal trampoline.
4875 @subsection @code{.cfi_window_save}
4876 SPARC register window has been saved.
4878 @subsection @code{.cfi_escape} @var{expression}[, @dots{}]
4879 Allows the user to add arbitrary bytes to the unwind info. One
4880 might use this to add OS-specific CFI opcodes, or generic CFI
4881 opcodes that GAS does not yet support.
4883 @subsection @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4884 The current value of @var{register} is @var{label}. The value of @var{label}
4885 will be encoded in the output file according to @var{encoding}; see the
4886 description of @code{.cfi_personality} for details on this encoding.
4888 The usefulness of equating a register to a fixed label is probably
4889 limited to the return address register. Here, it can be useful to
4890 mark a code segment that has only one return address which is reached
4891 by a direct branch and no copy of the return address exists in memory
4892 or another register.
4895 @section @code{.comm @var{symbol} , @var{length} }
4897 @cindex @code{comm} directive
4898 @cindex symbol, common
4899 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4900 common symbol in one object file may be merged with a defined or common symbol
4901 of the same name in another object file. If @code{@value{LD}} does not see a
4902 definition for the symbol--just one or more common symbols--then it will
4903 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4904 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4905 the same name, and they do not all have the same size, it will allocate space
4906 using the largest size.
4909 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4910 an optional third argument. This is the desired alignment of the symbol,
4911 specified for ELF as a byte boundary (for example, an alignment of 16 means
4912 that the least significant 4 bits of the address should be zero), and for PE
4913 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4914 boundary). The alignment must be an absolute expression, and it must be a
4915 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4916 common symbol, it will use the alignment when placing the symbol. If no
4917 alignment is specified, @command{@value{AS}} will set the alignment to the
4918 largest power of two less than or equal to the size of the symbol, up to a
4919 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4920 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4921 @samp{--section-alignment} option; image file sections in PE are aligned to
4922 multiples of 4096, which is far too large an alignment for ordinary variables.
4923 It is rather the default alignment for (non-debug) sections within object
4924 (@samp{*.o}) files, which are less strictly aligned.}.
4928 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4929 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4933 @section @code{.data @var{subsection}}
4935 @cindex @code{data} directive
4936 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4937 end of the data subsection numbered @var{subsection} (which is an
4938 absolute expression). If @var{subsection} is omitted, it defaults
4943 @section @code{.def @var{name}}
4945 @cindex @code{def} directive
4946 @cindex COFF symbols, debugging
4947 @cindex debugging COFF symbols
4948 Begin defining debugging information for a symbol @var{name}; the
4949 definition extends until the @code{.endef} directive is encountered.
4952 This directive is only observed when @command{@value{AS}} is configured for COFF
4953 format output; when producing @code{b.out}, @samp{.def} is recognized,
4960 @section @code{.desc @var{symbol}, @var{abs-expression}}
4962 @cindex @code{desc} directive
4963 @cindex COFF symbol descriptor
4964 @cindex symbol descriptor, COFF
4965 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4966 to the low 16 bits of an absolute expression.
4969 The @samp{.desc} directive is not available when @command{@value{AS}} is
4970 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4971 object format. For the sake of compatibility, @command{@value{AS}} accepts
4972 it, but produces no output, when configured for COFF.
4978 @section @code{.dim}
4980 @cindex @code{dim} directive
4981 @cindex COFF auxiliary symbol information
4982 @cindex auxiliary symbol information, COFF
4983 This directive is generated by compilers to include auxiliary debugging
4984 information in the symbol table. It is only permitted inside
4985 @code{.def}/@code{.endef} pairs.
4988 @samp{.dim} is only meaningful when generating COFF format output; when
4989 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4995 @section @code{.double @var{flonums}}
4997 @cindex @code{double} directive
4998 @cindex floating point numbers (double)
4999 @code{.double} expects zero or more flonums, separated by commas. It
5000 assembles floating point numbers.
5002 The exact kind of floating point numbers emitted depends on how
5003 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5007 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
5008 in @sc{ieee} format.
5013 @section @code{.eject}
5015 @cindex @code{eject} directive
5016 @cindex new page, in listings
5017 @cindex page, in listings
5018 @cindex listing control: new page
5019 Force a page break at this point, when generating assembly listings.
5022 @section @code{.else}
5024 @cindex @code{else} directive
5025 @code{.else} is part of the @command{@value{AS}} support for conditional
5026 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
5027 of code to be assembled if the condition for the preceding @code{.if}
5031 @section @code{.elseif}
5033 @cindex @code{elseif} directive
5034 @code{.elseif} is part of the @command{@value{AS}} support for conditional
5035 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
5036 @code{.if} block that would otherwise fill the entire @code{.else} section.
5039 @section @code{.end}
5041 @cindex @code{end} directive
5042 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
5043 process anything in the file past the @code{.end} directive.
5047 @section @code{.endef}
5049 @cindex @code{endef} directive
5050 This directive flags the end of a symbol definition begun with
5054 @samp{.endef} is only meaningful when generating COFF format output; if
5055 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
5056 directive but ignores it.
5061 @section @code{.endfunc}
5062 @cindex @code{endfunc} directive
5063 @code{.endfunc} marks the end of a function specified with @code{.func}.
5066 @section @code{.endif}
5068 @cindex @code{endif} directive
5069 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
5070 it marks the end of a block of code that is only assembled
5071 conditionally. @xref{If,,@code{.if}}.
5074 @section @code{.equ @var{symbol}, @var{expression}}
5076 @cindex @code{equ} directive
5077 @cindex assigning values to symbols
5078 @cindex symbols, assigning values to
5079 This directive sets the value of @var{symbol} to @var{expression}.
5080 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
5083 The syntax for @code{equ} on the HPPA is
5084 @samp{@var{symbol} .equ @var{expression}}.
5088 The syntax for @code{equ} on the Z80 is
5089 @samp{@var{symbol} equ @var{expression}}.
5090 On the Z80 it is an eror if @var{symbol} is already defined,
5091 but the symbol is not protected from later redefinition.
5092 Compare @ref{Equiv}.
5096 @section @code{.equiv @var{symbol}, @var{expression}}
5097 @cindex @code{equiv} directive
5098 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
5099 the assembler will signal an error if @var{symbol} is already defined. Note a
5100 symbol which has been referenced but not actually defined is considered to be
5103 Except for the contents of the error message, this is roughly equivalent to
5110 plus it protects the symbol from later redefinition.
5113 @section @code{.eqv @var{symbol}, @var{expression}}
5114 @cindex @code{eqv} directive
5115 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
5116 evaluate the expression or any part of it immediately. Instead each time
5117 the resulting symbol is used in an expression, a snapshot of its current
5121 @section @code{.err}
5122 @cindex @code{err} directive
5123 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
5124 message and, unless the @option{-Z} option was used, it will not generate an
5125 object file. This can be used to signal an error in conditionally compiled code.
5128 @section @code{.error "@var{string}"}
5129 @cindex error directive
5131 Similarly to @code{.err}, this directive emits an error, but you can specify a
5132 string that will be emitted as the error message. If you don't specify the
5133 message, it defaults to @code{".error directive invoked in source file"}.
5134 @xref{Errors, ,Error and Warning Messages}.
5137 .error "This code has not been assembled and tested."
5141 @section @code{.exitm}
5142 Exit early from the current macro definition. @xref{Macro}.
5145 @section @code{.extern}
5147 @cindex @code{extern} directive
5148 @code{.extern} is accepted in the source program---for compatibility
5149 with other assemblers---but it is ignored. @command{@value{AS}} treats
5150 all undefined symbols as external.
5153 @section @code{.fail @var{expression}}
5155 @cindex @code{fail} directive
5156 Generates an error or a warning. If the value of the @var{expression} is 500
5157 or more, @command{@value{AS}} will print a warning message. If the value is less
5158 than 500, @command{@value{AS}} will print an error message. The message will
5159 include the value of @var{expression}. This can occasionally be useful inside
5160 complex nested macros or conditional assembly.
5163 @section @code{.file}
5164 @cindex @code{file} directive
5166 @ifclear no-file-dir
5167 There are two different versions of the @code{.file} directive. Targets
5168 that support DWARF2 line number information use the DWARF2 version of
5169 @code{.file}. Other targets use the default version.
5171 @subheading Default Version
5173 @cindex logical file name
5174 @cindex file name, logical
5175 This version of the @code{.file} directive tells @command{@value{AS}} that we
5176 are about to start a new logical file. The syntax is:
5182 @var{string} is the new file name. In general, the filename is
5183 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
5184 to specify an empty file name, you must give the quotes--@code{""}. This
5185 statement may go away in future: it is only recognized to be compatible with
5186 old @command{@value{AS}} programs.
5188 @subheading DWARF2 Version
5191 When emitting DWARF2 line number information, @code{.file} assigns filenames
5192 to the @code{.debug_line} file name table. The syntax is:
5195 .file @var{fileno} @var{filename}
5198 The @var{fileno} operand should be a unique positive integer to use as the
5199 index of the entry in the table. The @var{filename} operand is a C string
5202 The detail of filename indices is exposed to the user because the filename
5203 table is shared with the @code{.debug_info} section of the DWARF2 debugging
5204 information, and thus the user must know the exact indices that table
5208 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
5210 @cindex @code{fill} directive
5211 @cindex writing patterns in memory
5212 @cindex patterns, writing in memory
5213 @var{repeat}, @var{size} and @var{value} are absolute expressions.
5214 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
5215 may be zero or more. @var{Size} may be zero or more, but if it is
5216 more than 8, then it is deemed to have the value 8, compatible with
5217 other people's assemblers. The contents of each @var{repeat} bytes
5218 is taken from an 8-byte number. The highest order 4 bytes are
5219 zero. The lowest order 4 bytes are @var{value} rendered in the
5220 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
5221 Each @var{size} bytes in a repetition is taken from the lowest order
5222 @var{size} bytes of this number. Again, this bizarre behavior is
5223 compatible with other people's assemblers.
5225 @var{size} and @var{value} are optional.
5226 If the second comma and @var{value} are absent, @var{value} is
5227 assumed zero. If the first comma and following tokens are absent,
5228 @var{size} is assumed to be 1.
5231 @section @code{.float @var{flonums}}
5233 @cindex floating point numbers (single)
5234 @cindex @code{float} directive
5235 This directive assembles zero or more flonums, separated by commas. It
5236 has the same effect as @code{.single}.
5238 The exact kind of floating point numbers emitted depends on how
5239 @command{@value{AS}} is configured.
5240 @xref{Machine Dependencies}.
5244 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
5245 in @sc{ieee} format.
5250 @section @code{.func @var{name}[,@var{label}]}
5251 @cindex @code{func} directive
5252 @code{.func} emits debugging information to denote function @var{name}, and
5253 is ignored unless the file is assembled with debugging enabled.
5254 Only @samp{--gstabs[+]} is currently supported.
5255 @var{label} is the entry point of the function and if omitted @var{name}
5256 prepended with the @samp{leading char} is used.
5257 @samp{leading char} is usually @code{_} or nothing, depending on the target.
5258 All functions are currently defined to have @code{void} return type.
5259 The function must be terminated with @code{.endfunc}.
5262 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
5264 @cindex @code{global} directive
5265 @cindex symbol, making visible to linker
5266 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
5267 @var{symbol} in your partial program, its value is made available to
5268 other partial programs that are linked with it. Otherwise,
5269 @var{symbol} takes its attributes from a symbol of the same name
5270 from another file linked into the same program.
5272 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
5273 compatibility with other assemblers.
5276 On the HPPA, @code{.global} is not always enough to make it accessible to other
5277 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
5278 @xref{HPPA Directives, ,HPPA Assembler Directives}.
5283 @section @code{.gnu_attribute @var{tag},@var{value}}
5284 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
5287 @section @code{.hidden @var{names}}
5289 @cindex @code{hidden} directive
5291 This is one of the ELF visibility directives. The other two are
5292 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
5293 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5295 This directive overrides the named symbols default visibility (which is set by
5296 their binding: local, global or weak). The directive sets the visibility to
5297 @code{hidden} which means that the symbols are not visible to other components.
5298 Such symbols are always considered to be @code{protected} as well.
5302 @section @code{.hword @var{expressions}}
5304 @cindex @code{hword} directive
5305 @cindex integers, 16-bit
5306 @cindex numbers, 16-bit
5307 @cindex sixteen bit integers
5308 This expects zero or more @var{expressions}, and emits
5309 a 16 bit number for each.
5312 This directive is a synonym for @samp{.short}; depending on the target
5313 architecture, it may also be a synonym for @samp{.word}.
5317 This directive is a synonym for @samp{.short}.
5320 This directive is a synonym for both @samp{.short} and @samp{.word}.
5325 @section @code{.ident}
5327 @cindex @code{ident} directive
5329 This directive is used by some assemblers to place tags in object files. The
5330 behavior of this directive varies depending on the target. When using the
5331 a.out object file format, @command{@value{AS}} simply accepts the directive for
5332 source-file compatibility with existing assemblers, but does not emit anything
5333 for it. When using COFF, comments are emitted to the @code{.comment} or
5334 @code{.rdata} section, depending on the target. When using ELF, comments are
5335 emitted to the @code{.comment} section.
5338 @section @code{.if @var{absolute expression}}
5340 @cindex conditional assembly
5341 @cindex @code{if} directive
5342 @code{.if} marks the beginning of a section of code which is only
5343 considered part of the source program being assembled if the argument
5344 (which must be an @var{absolute expression}) is non-zero. The end of
5345 the conditional section of code must be marked by @code{.endif}
5346 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
5347 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
5348 If you have several conditions to check, @code{.elseif} may be used to avoid
5349 nesting blocks if/else within each subsequent @code{.else} block.
5351 The following variants of @code{.if} are also supported:
5353 @cindex @code{ifdef} directive
5354 @item .ifdef @var{symbol}
5355 Assembles the following section of code if the specified @var{symbol}
5356 has been defined. Note a symbol which has been referenced but not yet defined
5357 is considered to be undefined.
5359 @cindex @code{ifb} directive
5360 @item .ifb @var{text}
5361 Assembles the following section of code if the operand is blank (empty).
5363 @cindex @code{ifc} directive
5364 @item .ifc @var{string1},@var{string2}
5365 Assembles the following section of code if the two strings are the same. The
5366 strings may be optionally quoted with single quotes. If they are not quoted,
5367 the first string stops at the first comma, and the second string stops at the
5368 end of the line. Strings which contain whitespace should be quoted. The
5369 string comparison is case sensitive.
5371 @cindex @code{ifeq} directive
5372 @item .ifeq @var{absolute expression}
5373 Assembles the following section of code if the argument is zero.
5375 @cindex @code{ifeqs} directive
5376 @item .ifeqs @var{string1},@var{string2}
5377 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5379 @cindex @code{ifge} directive
5380 @item .ifge @var{absolute expression}
5381 Assembles the following section of code if the argument is greater than or
5384 @cindex @code{ifgt} directive
5385 @item .ifgt @var{absolute expression}
5386 Assembles the following section of code if the argument is greater than zero.
5388 @cindex @code{ifle} directive
5389 @item .ifle @var{absolute expression}
5390 Assembles the following section of code if the argument is less than or equal
5393 @cindex @code{iflt} directive
5394 @item .iflt @var{absolute expression}
5395 Assembles the following section of code if the argument is less than zero.
5397 @cindex @code{ifnb} directive
5398 @item .ifnb @var{text}
5399 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5400 following section of code if the operand is non-blank (non-empty).
5402 @cindex @code{ifnc} directive
5403 @item .ifnc @var{string1},@var{string2}.
5404 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5405 following section of code if the two strings are not the same.
5407 @cindex @code{ifndef} directive
5408 @cindex @code{ifnotdef} directive
5409 @item .ifndef @var{symbol}
5410 @itemx .ifnotdef @var{symbol}
5411 Assembles the following section of code if the specified @var{symbol}
5412 has not been defined. Both spelling variants are equivalent. Note a symbol
5413 which has been referenced but not yet defined is considered to be undefined.
5415 @cindex @code{ifne} directive
5416 @item .ifne @var{absolute expression}
5417 Assembles the following section of code if the argument is not equal to zero
5418 (in other words, this is equivalent to @code{.if}).
5420 @cindex @code{ifnes} directive
5421 @item .ifnes @var{string1},@var{string2}
5422 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5423 following section of code if the two strings are not the same.
5427 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5429 @cindex @code{incbin} directive
5430 @cindex binary files, including
5431 The @code{incbin} directive includes @var{file} verbatim at the current
5432 location. You can control the search paths used with the @samp{-I} command-line
5433 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5436 The @var{skip} argument skips a number of bytes from the start of the
5437 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5438 read. Note that the data is not aligned in any way, so it is the user's
5439 responsibility to make sure that proper alignment is provided both before and
5440 after the @code{incbin} directive.
5443 @section @code{.include "@var{file}"}
5445 @cindex @code{include} directive
5446 @cindex supporting files, including
5447 @cindex files, including
5448 This directive provides a way to include supporting files at specified
5449 points in your source program. The code from @var{file} is assembled as
5450 if it followed the point of the @code{.include}; when the end of the
5451 included file is reached, assembly of the original file continues. You
5452 can control the search paths used with the @samp{-I} command-line option
5453 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5457 @section @code{.int @var{expressions}}
5459 @cindex @code{int} directive
5460 @cindex integers, 32-bit
5461 Expect zero or more @var{expressions}, of any section, separated by commas.
5462 For each expression, emit a number that, at run time, is the value of that
5463 expression. The byte order and bit size of the number depends on what kind
5464 of target the assembly is for.
5468 On most forms of the H8/300, @code{.int} emits 16-bit
5469 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5476 @section @code{.internal @var{names}}
5478 @cindex @code{internal} directive
5480 This is one of the ELF visibility directives. The other two are
5481 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5482 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5484 This directive overrides the named symbols default visibility (which is set by
5485 their binding: local, global or weak). The directive sets the visibility to
5486 @code{internal} which means that the symbols are considered to be @code{hidden}
5487 (i.e., not visible to other components), and that some extra, processor specific
5488 processing must also be performed upon the symbols as well.
5492 @section @code{.irp @var{symbol},@var{values}}@dots{}
5494 @cindex @code{irp} directive
5495 Evaluate a sequence of statements assigning different values to @var{symbol}.
5496 The sequence of statements starts at the @code{.irp} directive, and is
5497 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5498 set to @var{value}, and the sequence of statements is assembled. If no
5499 @var{value} is listed, the sequence of statements is assembled once, with
5500 @var{symbol} set to the null string. To refer to @var{symbol} within the
5501 sequence of statements, use @var{\symbol}.
5503 For example, assembling
5511 is equivalent to assembling
5519 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5522 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5524 @cindex @code{irpc} directive
5525 Evaluate a sequence of statements assigning different values to @var{symbol}.
5526 The sequence of statements starts at the @code{.irpc} directive, and is
5527 terminated by an @code{.endr} directive. For each character in @var{value},
5528 @var{symbol} is set to the character, and the sequence of statements is
5529 assembled. If no @var{value} is listed, the sequence of statements is
5530 assembled once, with @var{symbol} set to the null string. To refer to
5531 @var{symbol} within the sequence of statements, use @var{\symbol}.
5533 For example, assembling
5541 is equivalent to assembling
5549 For some caveats with the spelling of @var{symbol}, see also the discussion
5553 @section @code{.lcomm @var{symbol} , @var{length}}
5555 @cindex @code{lcomm} directive
5556 @cindex local common symbols
5557 @cindex symbols, local common
5558 Reserve @var{length} (an absolute expression) bytes for a local common
5559 denoted by @var{symbol}. The section and value of @var{symbol} are
5560 those of the new local common. The addresses are allocated in the bss
5561 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5562 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5563 not visible to @code{@value{LD}}.
5566 Some targets permit a third argument to be used with @code{.lcomm}. This
5567 argument specifies the desired alignment of the symbol in the bss section.
5571 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5572 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5576 @section @code{.lflags}
5578 @cindex @code{lflags} directive (ignored)
5579 @command{@value{AS}} accepts this directive, for compatibility with other
5580 assemblers, but ignores it.
5582 @ifclear no-line-dir
5584 @section @code{.line @var{line-number}}
5586 @cindex @code{line} directive
5587 @cindex logical line number
5589 Change the logical line number. @var{line-number} must be an absolute
5590 expression. The next line has that logical line number. Therefore any other
5591 statements on the current line (after a statement separator character) are
5592 reported as on logical line number @var{line-number} @minus{} 1. One day
5593 @command{@value{AS}} will no longer support this directive: it is recognized only
5594 for compatibility with existing assembler programs.
5597 Even though this is a directive associated with the @code{a.out} or
5598 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5599 when producing COFF output, and treats @samp{.line} as though it
5600 were the COFF @samp{.ln} @emph{if} it is found outside a
5601 @code{.def}/@code{.endef} pair.
5603 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5604 used by compilers to generate auxiliary symbol information for
5609 @section @code{.linkonce [@var{type}]}
5611 @cindex @code{linkonce} directive
5612 @cindex common sections
5613 Mark the current section so that the linker only includes a single copy of it.
5614 This may be used to include the same section in several different object files,
5615 but ensure that the linker will only include it once in the final output file.
5616 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5617 Duplicate sections are detected based on the section name, so it should be
5620 This directive is only supported by a few object file formats; as of this
5621 writing, the only object file format which supports it is the Portable
5622 Executable format used on Windows NT.
5624 The @var{type} argument is optional. If specified, it must be one of the
5625 following strings. For example:
5629 Not all types may be supported on all object file formats.
5633 Silently discard duplicate sections. This is the default.
5636 Warn if there are duplicate sections, but still keep only one copy.
5639 Warn if any of the duplicates have different sizes.
5642 Warn if any of the duplicates do not have exactly the same contents.
5646 @section @code{.list}
5648 @cindex @code{list} directive
5649 @cindex listing control, turning on
5650 Control (in conjunction with the @code{.nolist} directive) whether or
5651 not assembly listings are generated. These two directives maintain an
5652 internal counter (which is zero initially). @code{.list} increments the
5653 counter, and @code{.nolist} decrements it. Assembly listings are
5654 generated whenever the counter is greater than zero.
5656 By default, listings are disabled. When you enable them (with the
5657 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5658 the initial value of the listing counter is one.
5661 @section @code{.ln @var{line-number}}
5663 @cindex @code{ln} directive
5664 @ifclear no-line-dir
5665 @samp{.ln} is a synonym for @samp{.line}.
5668 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5669 must be an absolute expression. The next line has that logical
5670 line number, so any other statements on the current line (after a
5671 statement separator character @code{;}) are reported as on logical
5672 line number @var{line-number} @minus{} 1.
5675 This directive is accepted, but ignored, when @command{@value{AS}} is
5676 configured for @code{b.out}; its effect is only associated with COFF
5682 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5683 @cindex @code{loc} directive
5684 When emitting DWARF2 line number information,
5685 the @code{.loc} directive will add a row to the @code{.debug_line} line
5686 number matrix corresponding to the immediately following assembly
5687 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5688 arguments will be applied to the @code{.debug_line} state machine before
5691 The @var{options} are a sequence of the following tokens in any order:
5695 This option will set the @code{basic_block} register in the
5696 @code{.debug_line} state machine to @code{true}.
5699 This option will set the @code{prologue_end} register in the
5700 @code{.debug_line} state machine to @code{true}.
5702 @item epilogue_begin
5703 This option will set the @code{epilogue_begin} register in the
5704 @code{.debug_line} state machine to @code{true}.
5706 @item is_stmt @var{value}
5707 This option will set the @code{is_stmt} register in the
5708 @code{.debug_line} state machine to @code{value}, which must be
5711 @item isa @var{value}
5712 This directive will set the @code{isa} register in the @code{.debug_line}
5713 state machine to @var{value}, which must be an unsigned integer.
5715 @item discriminator @var{value}
5716 This directive will set the @code{discriminator} register in the @code{.debug_line}
5717 state machine to @var{value}, which must be an unsigned integer.
5721 @node Loc_mark_labels
5722 @section @code{.loc_mark_labels @var{enable}}
5723 @cindex @code{loc_mark_labels} directive
5724 When emitting DWARF2 line number information,
5725 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5726 to the @code{.debug_line} line number matrix with the @code{basic_block}
5727 register in the state machine set whenever a code label is seen.
5728 The @var{enable} argument should be either 1 or 0, to enable or disable
5729 this function respectively.
5733 @section @code{.local @var{names}}
5735 @cindex @code{local} directive
5736 This directive, which is available for ELF targets, marks each symbol in
5737 the comma-separated list of @code{names} as a local symbol so that it
5738 will not be externally visible. If the symbols do not already exist,
5739 they will be created.
5741 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5742 accept an alignment argument, which is the case for most ELF targets,
5743 the @code{.local} directive can be used in combination with @code{.comm}
5744 (@pxref{Comm}) to define aligned local common data.
5748 @section @code{.long @var{expressions}}
5750 @cindex @code{long} directive
5751 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5754 @c no one seems to know what this is for or whether this description is
5755 @c what it really ought to do
5757 @section @code{.lsym @var{symbol}, @var{expression}}
5759 @cindex @code{lsym} directive
5760 @cindex symbol, not referenced in assembly
5761 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5762 the hash table, ensuring it cannot be referenced by name during the
5763 rest of the assembly. This sets the attributes of the symbol to be
5764 the same as the expression value:
5766 @var{other} = @var{descriptor} = 0
5767 @var{type} = @r{(section of @var{expression})}
5768 @var{value} = @var{expression}
5771 The new symbol is not flagged as external.
5775 @section @code{.macro}
5778 The commands @code{.macro} and @code{.endm} allow you to define macros that
5779 generate assembly output. For example, this definition specifies a macro
5780 @code{sum} that puts a sequence of numbers into memory:
5783 .macro sum from=0, to=5
5792 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5804 @item .macro @var{macname}
5805 @itemx .macro @var{macname} @var{macargs} @dots{}
5806 @cindex @code{macro} directive
5807 Begin the definition of a macro called @var{macname}. If your macro
5808 definition requires arguments, specify their names after the macro name,
5809 separated by commas or spaces. You can qualify the macro argument to
5810 indicate whether all invocations must specify a non-blank value (through
5811 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5812 (through @samp{:@code{vararg}}). You can supply a default value for any
5813 macro argument by following the name with @samp{=@var{deflt}}. You
5814 cannot define two macros with the same @var{macname} unless it has been
5815 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5816 definitions. For example, these are all valid @code{.macro} statements:
5820 Begin the definition of a macro called @code{comm}, which takes no
5823 @item .macro plus1 p, p1
5824 @itemx .macro plus1 p p1
5825 Either statement begins the definition of a macro called @code{plus1},
5826 which takes two arguments; within the macro definition, write
5827 @samp{\p} or @samp{\p1} to evaluate the arguments.
5829 @item .macro reserve_str p1=0 p2
5830 Begin the definition of a macro called @code{reserve_str}, with two
5831 arguments. The first argument has a default value, but not the second.
5832 After the definition is complete, you can call the macro either as
5833 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5834 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5835 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5836 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5838 @item .macro m p1:req, p2=0, p3:vararg
5839 Begin the definition of a macro called @code{m}, with at least three
5840 arguments. The first argument must always have a value specified, but
5841 not the second, which instead has a default value. The third formal
5842 will get assigned all remaining arguments specified at invocation time.
5844 When you call a macro, you can specify the argument values either by
5845 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5846 @samp{sum to=17, from=9}.
5850 Note that since each of the @var{macargs} can be an identifier exactly
5851 as any other one permitted by the target architecture, there may be
5852 occasional problems if the target hand-crafts special meanings to certain
5853 characters when they occur in a special position. For example, if the colon
5854 (@code{:}) is generally permitted to be part of a symbol name, but the
5855 architecture specific code special-cases it when occurring as the final
5856 character of a symbol (to denote a label), then the macro parameter
5857 replacement code will have no way of knowing that and consider the whole
5858 construct (including the colon) an identifier, and check only this
5859 identifier for being the subject to parameter substitution. So for example
5860 this macro definition:
5868 might not work as expected. Invoking @samp{label foo} might not create a label
5869 called @samp{foo} but instead just insert the text @samp{\l:} into the
5870 assembler source, probably generating an error about an unrecognised
5873 Similarly problems might occur with the period character (@samp{.})
5874 which is often allowed inside opcode names (and hence identifier names). So
5875 for example constructing a macro to build an opcode from a base name and a
5876 length specifier like this:
5879 .macro opcode base length
5884 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5885 instruction but instead generate some kind of error as the assembler tries to
5886 interpret the text @samp{\base.\length}.
5888 There are several possible ways around this problem:
5891 @item Insert white space
5892 If it is possible to use white space characters then this is the simplest
5901 @item Use @samp{\()}
5902 The string @samp{\()} can be used to separate the end of a macro argument from
5903 the following text. eg:
5906 .macro opcode base length
5911 @item Use the alternate macro syntax mode
5912 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5913 used as a separator. eg:
5923 Note: this problem of correctly identifying string parameters to pseudo ops
5924 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5925 and @code{.irpc} (@pxref{Irpc}) as well.
5928 @cindex @code{endm} directive
5929 Mark the end of a macro definition.
5932 @cindex @code{exitm} directive
5933 Exit early from the current macro definition.
5935 @cindex number of macros executed
5936 @cindex macros, count executed
5938 @command{@value{AS}} maintains a counter of how many macros it has
5939 executed in this pseudo-variable; you can copy that number to your
5940 output with @samp{\@@}, but @emph{only within a macro definition}.
5942 @item LOCAL @var{name} [ , @dots{} ]
5943 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5944 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5945 @xref{Altmacro,,@code{.altmacro}}.
5949 @section @code{.mri @var{val}}
5951 @cindex @code{mri} directive
5952 @cindex MRI mode, temporarily
5953 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
5954 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
5955 affects code assembled until the next @code{.mri} directive, or until the end
5956 of the file. @xref{M, MRI mode, MRI mode}.
5959 @section @code{.noaltmacro}
5960 Disable alternate macro mode. @xref{Altmacro}.
5963 @section @code{.nolist}
5965 @cindex @code{nolist} directive
5966 @cindex listing control, turning off
5967 Control (in conjunction with the @code{.list} directive) whether or
5968 not assembly listings are generated. These two directives maintain an
5969 internal counter (which is zero initially). @code{.list} increments the
5970 counter, and @code{.nolist} decrements it. Assembly listings are
5971 generated whenever the counter is greater than zero.
5974 @section @code{.octa @var{bignums}}
5976 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5977 @cindex @code{octa} directive
5978 @cindex integer, 16-byte
5979 @cindex sixteen byte integer
5980 This directive expects zero or more bignums, separated by commas. For each
5981 bignum, it emits a 16-byte integer.
5983 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5984 hence @emph{octa}-word for 16 bytes.
5987 @section @code{.offset @var{loc}}
5989 @cindex @code{offset} directive
5990 Set the location counter to @var{loc} in the absolute section. @var{loc} must
5991 be an absolute expression. This directive may be useful for defining
5992 symbols with absolute values. Do not confuse it with the @code{.org}
5996 @section @code{.org @var{new-lc} , @var{fill}}
5998 @cindex @code{org} directive
5999 @cindex location counter, advancing
6000 @cindex advancing location counter
6001 @cindex current address, advancing
6002 Advance the location counter of the current section to
6003 @var{new-lc}. @var{new-lc} is either an absolute expression or an
6004 expression with the same section as the current subsection. That is,
6005 you can't use @code{.org} to cross sections: if @var{new-lc} has the
6006 wrong section, the @code{.org} directive is ignored. To be compatible
6007 with former assemblers, if the section of @var{new-lc} is absolute,
6008 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
6009 is the same as the current subsection.
6011 @code{.org} may only increase the location counter, or leave it
6012 unchanged; you cannot use @code{.org} to move the location counter
6015 @c double negative used below "not undefined" because this is a specific
6016 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
6017 @c section. doc@cygnus.com 18feb91
6018 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
6019 may not be undefined. If you really detest this restriction we eagerly await
6020 a chance to share your improved assembler.
6022 Beware that the origin is relative to the start of the section, not
6023 to the start of the subsection. This is compatible with other
6024 people's assemblers.
6026 When the location counter (of the current subsection) is advanced, the
6027 intervening bytes are filled with @var{fill} which should be an
6028 absolute expression. If the comma and @var{fill} are omitted,
6029 @var{fill} defaults to zero.
6032 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
6034 @cindex padding the location counter given a power of two
6035 @cindex @code{p2align} directive
6036 Pad the location counter (in the current subsection) to a particular
6037 storage boundary. The first expression (which must be absolute) is the
6038 number of low-order zero bits the location counter must have after
6039 advancement. For example @samp{.p2align 3} advances the location
6040 counter until it a multiple of 8. If the location counter is already a
6041 multiple of 8, no change is needed.
6043 The second expression (also absolute) gives the fill value to be stored in the
6044 padding bytes. It (and the comma) may be omitted. If it is omitted, the
6045 padding bytes are normally zero. However, on some systems, if the section is
6046 marked as containing code and the fill value is omitted, the space is filled
6047 with no-op instructions.
6049 The third expression is also absolute, and is also optional. If it is present,
6050 it is the maximum number of bytes that should be skipped by this alignment
6051 directive. If doing the alignment would require skipping more bytes than the
6052 specified maximum, then the alignment is not done at all. You can omit the
6053 fill value (the second argument) entirely by simply using two commas after the
6054 required alignment; this can be useful if you want the alignment to be filled
6055 with no-op instructions when appropriate.
6057 @cindex @code{p2alignw} directive
6058 @cindex @code{p2alignl} directive
6059 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
6060 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
6061 pattern as a two byte word value. The @code{.p2alignl} directives treats the
6062 fill pattern as a four byte longword value. For example, @code{.p2alignw
6063 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
6064 filled in with the value 0x368d (the exact placement of the bytes depends upon
6065 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
6070 @section @code{.popsection}
6072 @cindex @code{popsection} directive
6073 @cindex Section Stack
6074 This is one of the ELF section stack manipulation directives. The others are
6075 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6076 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
6079 This directive replaces the current section (and subsection) with the top
6080 section (and subsection) on the section stack. This section is popped off the
6086 @section @code{.previous}
6088 @cindex @code{previous} directive
6089 @cindex Section Stack
6090 This is one of the ELF section stack manipulation directives. The others are
6091 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6092 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
6093 (@pxref{PopSection}).
6095 This directive swaps the current section (and subsection) with most recently
6096 referenced section/subsection pair prior to this one. Multiple
6097 @code{.previous} directives in a row will flip between two sections (and their
6098 subsections). For example:
6110 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
6116 # Now in section A subsection 1
6120 # Now in section B subsection 0
6123 # Now in section B subsection 1
6126 # Now in section B subsection 0
6130 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
6131 section B and 0x9abc into subsection 1 of section B.
6133 In terms of the section stack, this directive swaps the current section with
6134 the top section on the section stack.
6138 @section @code{.print @var{string}}
6140 @cindex @code{print} directive
6141 @command{@value{AS}} will print @var{string} on the standard output during
6142 assembly. You must put @var{string} in double quotes.
6146 @section @code{.protected @var{names}}
6148 @cindex @code{protected} directive
6150 This is one of the ELF visibility directives. The other two are
6151 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
6153 This directive overrides the named symbols default visibility (which is set by
6154 their binding: local, global or weak). The directive sets the visibility to
6155 @code{protected} which means that any references to the symbols from within the
6156 components that defines them must be resolved to the definition in that
6157 component, even if a definition in another component would normally preempt
6162 @section @code{.psize @var{lines} , @var{columns}}
6164 @cindex @code{psize} directive
6165 @cindex listing control: paper size
6166 @cindex paper size, for listings
6167 Use this directive to declare the number of lines---and, optionally, the
6168 number of columns---to use for each page, when generating listings.
6170 If you do not use @code{.psize}, listings use a default line-count
6171 of 60. You may omit the comma and @var{columns} specification; the
6172 default width is 200 columns.
6174 @command{@value{AS}} generates formfeeds whenever the specified number of
6175 lines is exceeded (or whenever you explicitly request one, using
6178 If you specify @var{lines} as @code{0}, no formfeeds are generated save
6179 those explicitly specified with @code{.eject}.
6182 @section @code{.purgem @var{name}}
6184 @cindex @code{purgem} directive
6185 Undefine the macro @var{name}, so that later uses of the string will not be
6186 expanded. @xref{Macro}.
6190 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
6192 @cindex @code{pushsection} directive
6193 @cindex Section Stack
6194 This is one of the ELF section stack manipulation directives. The others are
6195 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6196 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6199 This directive pushes the current section (and subsection) onto the
6200 top of the section stack, and then replaces the current section and
6201 subsection with @code{name} and @code{subsection}. The optional
6202 @code{flags}, @code{type} and @code{arguments} are treated the same
6203 as in the @code{.section} (@pxref{Section}) directive.
6207 @section @code{.quad @var{bignums}}
6209 @cindex @code{quad} directive
6210 @code{.quad} expects zero or more bignums, separated by commas. For
6211 each bignum, it emits
6213 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
6214 warning message; and just takes the lowest order 8 bytes of the bignum.
6215 @cindex eight-byte integer
6216 @cindex integer, 8-byte
6218 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
6219 hence @emph{quad}-word for 8 bytes.
6222 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
6223 warning message; and just takes the lowest order 16 bytes of the bignum.
6224 @cindex sixteen-byte integer
6225 @cindex integer, 16-byte
6229 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
6231 @cindex @code{reloc} directive
6232 Generate a relocation at @var{offset} of type @var{reloc_name} with value
6233 @var{expression}. If @var{offset} is a number, the relocation is generated in
6234 the current section. If @var{offset} is an expression that resolves to a
6235 symbol plus offset, the relocation is generated in the given symbol's section.
6236 @var{expression}, if present, must resolve to a symbol plus addend or to an
6237 absolute value, but note that not all targets support an addend. e.g. ELF REL
6238 targets such as i386 store an addend in the section contents rather than in the
6239 relocation. This low level interface does not support addends stored in the
6243 @section @code{.rept @var{count}}
6245 @cindex @code{rept} directive
6246 Repeat the sequence of lines between the @code{.rept} directive and the next
6247 @code{.endr} directive @var{count} times.
6249 For example, assembling
6257 is equivalent to assembling
6266 @section @code{.sbttl "@var{subheading}"}
6268 @cindex @code{sbttl} directive
6269 @cindex subtitles for listings
6270 @cindex listing control: subtitle
6271 Use @var{subheading} as the title (third line, immediately after the
6272 title line) when generating assembly listings.
6274 This directive affects subsequent pages, as well as the current page if
6275 it appears within ten lines of the top of a page.
6279 @section @code{.scl @var{class}}
6281 @cindex @code{scl} directive
6282 @cindex symbol storage class (COFF)
6283 @cindex COFF symbol storage class
6284 Set the storage-class value for a symbol. This directive may only be
6285 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
6286 whether a symbol is static or external, or it may record further
6287 symbolic debugging information.
6290 The @samp{.scl} directive is primarily associated with COFF output; when
6291 configured to generate @code{b.out} output format, @command{@value{AS}}
6292 accepts this directive but ignores it.
6298 @section @code{.section @var{name}}
6300 @cindex named section
6301 Use the @code{.section} directive to assemble the following code into a section
6304 This directive is only supported for targets that actually support arbitrarily
6305 named sections; on @code{a.out} targets, for example, it is not accepted, even
6306 with a standard @code{a.out} section name.
6310 @c only print the extra heading if both COFF and ELF are set
6311 @subheading COFF Version
6314 @cindex @code{section} directive (COFF version)
6315 For COFF targets, the @code{.section} directive is used in one of the following
6319 .section @var{name}[, "@var{flags}"]
6320 .section @var{name}[, @var{subsection}]
6323 If the optional argument is quoted, it is taken as flags to use for the
6324 section. Each flag is a single character. The following flags are recognized:
6328 bss section (uninitialized data)
6330 section is not loaded
6336 exclude section from linking
6342 shared section (meaningful for PE targets)
6344 ignored. (For compatibility with the ELF version)
6346 section is not readable (meaningful for PE targets)
6348 single-digit power-of-two section alignment (GNU extension)
6351 If no flags are specified, the default flags depend upon the section name. If
6352 the section name is not recognized, the default will be for the section to be
6353 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
6354 from the section, rather than adding them, so if they are used on their own it
6355 will be as if no flags had been specified at all.
6357 If the optional argument to the @code{.section} directive is not quoted, it is
6358 taken as a subsection number (@pxref{Sub-Sections}).
6363 @c only print the extra heading if both COFF and ELF are set
6364 @subheading ELF Version
6367 @cindex Section Stack
6368 This is one of the ELF section stack manipulation directives. The others are
6369 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
6370 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
6371 @code{.previous} (@pxref{Previous}).
6373 @cindex @code{section} directive (ELF version)
6374 For ELF targets, the @code{.section} directive is used like this:
6377 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
6380 @anchor{Section Name Substitutions}
6381 @kindex --sectname-subst
6382 @cindex section name substitution
6383 If the @samp{--sectname-subst} command-line option is provided, the @var{name}
6384 argument may contain a substitution sequence. Only @code{%S} is supported
6385 at the moment, and substitutes the current section name. For example:
6388 .macro exception_code
6389 .section %S.exception
6390 [exception code here]
6405 The two @code{exception_code} invocations above would create the
6406 @code{.text.exception} and @code{.init.exception} sections respectively.
6407 This is useful e.g. to discriminate between anciliary sections that are
6408 tied to setup code to be discarded after use from anciliary sections that
6409 need to stay resident without having to define multiple @code{exception_code}
6410 macros just for that purpose.
6412 The optional @var{flags} argument is a quoted string which may contain any
6413 combination of the following characters:
6417 section is allocatable
6419 section is excluded from executable and shared library.
6423 section is executable
6425 section is mergeable
6427 section contains zero terminated strings
6429 section is a member of a section group
6431 section is used for thread-local-storage
6433 section is a member of the previously-current section's group, if any
6434 @item @code{<number>}
6435 a numeric value indicating the bits to be set in the ELF section header's flags
6436 field. Note - if one or more of the alphabetic characters described above is
6437 also included in the flags field, their bit values will be ORed into the
6439 @item @code{<target specific>}
6440 some targets extend this list with their own flag characters
6443 Note - once a section's flags have been set they cannot be changed. There are
6444 a few exceptions to this rule however. Processor and application specific
6445 flags can be added to an already defined section. The @code{.interp},
6446 @code{.strtab} and @code{.symtab} sections can have the allocate flag
6447 (@code{a}) set after they are initially defined, and the @code{.note-GNU-stack}
6448 section may have the executable (@code{x}) flag added.
6450 The optional @var{type} argument may contain one of the following constants:
6454 section contains data
6456 section does not contain data (i.e., section only occupies space)
6458 section contains data which is used by things other than the program
6460 section contains an array of pointers to init functions
6462 section contains an array of pointers to finish functions
6463 @item @@preinit_array
6464 section contains an array of pointers to pre-init functions
6465 @item @@@code{<number>}
6466 a numeric value to be set as the ELF section header's type field.
6467 @item @@@code{<target specific>}
6468 some targets extend this list with their own types
6471 Many targets only support the first three section types. The type may be
6472 enclosed in double quotes if necessary.
6474 Note on targets where the @code{@@} character is the start of a comment (eg
6475 ARM) then another character is used instead. For example the ARM port uses the
6478 Note - some sections, eg @code{.text} and @code{.data} are considered to be
6479 special and have fixed types. Any attempt to declare them with a different
6480 type will generate an error from the assembler.
6482 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6483 be specified as well as an extra argument---@var{entsize}---like this:
6486 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6489 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6490 constants, each @var{entsize} octets long. Sections with both @code{M} and
6491 @code{S} must contain zero terminated strings where each character is
6492 @var{entsize} bytes long. The linker may remove duplicates within sections with
6493 the same name, same entity size and same flags. @var{entsize} must be an
6494 absolute expression. For sections with both @code{M} and @code{S}, a string
6495 which is a suffix of a larger string is considered a duplicate. Thus
6496 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6497 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6499 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6500 be present along with an additional field like this:
6503 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6506 The @var{GroupName} field specifies the name of the section group to which this
6507 particular section belongs. The optional linkage field can contain:
6511 indicates that only one copy of this section should be retained
6516 Note: if both the @var{M} and @var{G} flags are present then the fields for
6517 the Merge flag should come first, like this:
6520 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6523 If @var{flags} contains the @code{?} symbol then it may not also contain the
6524 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6525 present. Instead, @code{?} says to consider the section that's current before
6526 this directive. If that section used @code{G}, then the new section will use
6527 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6528 If not, then the @code{?} symbol has no effect.
6530 If no flags are specified, the default flags depend upon the section name. If
6531 the section name is not recognized, the default will be for the section to have
6532 none of the above flags: it will not be allocated in memory, nor writable, nor
6533 executable. The section will contain data.
6535 For ELF targets, the assembler supports another type of @code{.section}
6536 directive for compatibility with the Solaris assembler:
6539 .section "@var{name}"[, @var{flags}...]
6542 Note that the section name is quoted. There may be a sequence of comma
6547 section is allocatable
6551 section is executable
6553 section is excluded from executable and shared library.
6555 section is used for thread local storage
6558 This directive replaces the current section and subsection. See the
6559 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6560 some examples of how this directive and the other section stack directives
6566 @section @code{.set @var{symbol}, @var{expression}}
6568 @cindex @code{set} directive
6569 @cindex symbol value, setting
6570 Set the value of @var{symbol} to @var{expression}. This
6571 changes @var{symbol}'s value and type to conform to
6572 @var{expression}. If @var{symbol} was flagged as external, it remains
6573 flagged (@pxref{Symbol Attributes}).
6575 You may @code{.set} a symbol many times in the same assembly provided that the
6576 values given to the symbol are constants. Values that are based on expressions
6577 involving other symbols are allowed, but some targets may restrict this to only
6578 being done once per assembly. This is because those targets do not set the
6579 addresses of symbols at assembly time, but rather delay the assignment until a
6580 final link is performed. This allows the linker a chance to change the code in
6581 the files, changing the location of, and the relative distance between, various
6584 If you @code{.set} a global symbol, the value stored in the object
6585 file is the last value stored into it.
6588 On Z80 @code{set} is a real instruction, use
6589 @samp{@var{symbol} defl @var{expression}} instead.
6593 @section @code{.short @var{expressions}}
6595 @cindex @code{short} directive
6597 @code{.short} is normally the same as @samp{.word}.
6598 @xref{Word,,@code{.word}}.
6600 In some configurations, however, @code{.short} and @code{.word} generate
6601 numbers of different lengths. @xref{Machine Dependencies}.
6605 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6608 This expects zero or more @var{expressions}, and emits
6609 a 16 bit number for each.
6614 @section @code{.single @var{flonums}}
6616 @cindex @code{single} directive
6617 @cindex floating point numbers (single)
6618 This directive assembles zero or more flonums, separated by commas. It
6619 has the same effect as @code{.float}.
6621 The exact kind of floating point numbers emitted depends on how
6622 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6626 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6627 numbers in @sc{ieee} format.
6633 @section @code{.size}
6635 This directive is used to set the size associated with a symbol.
6639 @c only print the extra heading if both COFF and ELF are set
6640 @subheading COFF Version
6643 @cindex @code{size} directive (COFF version)
6644 For COFF targets, the @code{.size} directive is only permitted inside
6645 @code{.def}/@code{.endef} pairs. It is used like this:
6648 .size @var{expression}
6652 @samp{.size} is only meaningful when generating COFF format output; when
6653 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6660 @c only print the extra heading if both COFF and ELF are set
6661 @subheading ELF Version
6664 @cindex @code{size} directive (ELF version)
6665 For ELF targets, the @code{.size} directive is used like this:
6668 .size @var{name} , @var{expression}
6671 This directive sets the size associated with a symbol @var{name}.
6672 The size in bytes is computed from @var{expression} which can make use of label
6673 arithmetic. This directive is typically used to set the size of function
6678 @ifclear no-space-dir
6680 @section @code{.skip @var{size} , @var{fill}}
6682 @cindex @code{skip} directive
6683 @cindex filling memory
6684 This directive emits @var{size} bytes, each of value @var{fill}. Both
6685 @var{size} and @var{fill} are absolute expressions. If the comma and
6686 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6691 @section @code{.sleb128 @var{expressions}}
6693 @cindex @code{sleb128} directive
6694 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6695 compact, variable length representation of numbers used by the DWARF
6696 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6698 @ifclear no-space-dir
6700 @section @code{.space @var{size} , @var{fill}}
6702 @cindex @code{space} directive
6703 @cindex filling memory
6704 This directive emits @var{size} bytes, each of value @var{fill}. Both
6705 @var{size} and @var{fill} are absolute expressions. If the comma
6706 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6711 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6712 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6713 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6714 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6722 @section @code{.stabd, .stabn, .stabs}
6724 @cindex symbolic debuggers, information for
6725 @cindex @code{stab@var{x}} directives
6726 There are three directives that begin @samp{.stab}.
6727 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6728 The symbols are not entered in the @command{@value{AS}} hash table: they
6729 cannot be referenced elsewhere in the source file.
6730 Up to five fields are required:
6734 This is the symbol's name. It may contain any character except
6735 @samp{\000}, so is more general than ordinary symbol names. Some
6736 debuggers used to code arbitrarily complex structures into symbol names
6740 An absolute expression. The symbol's type is set to the low 8 bits of
6741 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6742 and debuggers choke on silly bit patterns.
6745 An absolute expression. The symbol's ``other'' attribute is set to the
6746 low 8 bits of this expression.
6749 An absolute expression. The symbol's descriptor is set to the low 16
6750 bits of this expression.
6753 An absolute expression which becomes the symbol's value.
6756 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6757 or @code{.stabs} statement, the symbol has probably already been created;
6758 you get a half-formed symbol in your object file. This is
6759 compatible with earlier assemblers!
6762 @cindex @code{stabd} directive
6763 @item .stabd @var{type} , @var{other} , @var{desc}
6765 The ``name'' of the symbol generated is not even an empty string.
6766 It is a null pointer, for compatibility. Older assemblers used a
6767 null pointer so they didn't waste space in object files with empty
6770 The symbol's value is set to the location counter,
6771 relocatably. When your program is linked, the value of this symbol
6772 is the address of the location counter when the @code{.stabd} was
6775 @cindex @code{stabn} directive
6776 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6777 The name of the symbol is set to the empty string @code{""}.
6779 @cindex @code{stabs} directive
6780 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6781 All five fields are specified.
6787 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6788 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6790 @cindex string, copying to object file
6791 @cindex string8, copying to object file
6792 @cindex string16, copying to object file
6793 @cindex string32, copying to object file
6794 @cindex string64, copying to object file
6795 @cindex @code{string} directive
6796 @cindex @code{string8} directive
6797 @cindex @code{string16} directive
6798 @cindex @code{string32} directive
6799 @cindex @code{string64} directive
6801 Copy the characters in @var{str} to the object file. You may specify more than
6802 one string to copy, separated by commas. Unless otherwise specified for a
6803 particular machine, the assembler marks the end of each string with a 0 byte.
6804 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6806 The variants @code{string16}, @code{string32} and @code{string64} differ from
6807 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6808 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6809 are stored in target endianness byte order.
6815 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6816 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6821 @section @code{.struct @var{expression}}
6823 @cindex @code{struct} directive
6824 Switch to the absolute section, and set the section offset to @var{expression},
6825 which must be an absolute expression. You might use this as follows:
6834 This would define the symbol @code{field1} to have the value 0, the symbol
6835 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6836 value 8. Assembly would be left in the absolute section, and you would need to
6837 use a @code{.section} directive of some sort to change to some other section
6838 before further assembly.
6842 @section @code{.subsection @var{name}}
6844 @cindex @code{subsection} directive
6845 @cindex Section Stack
6846 This is one of the ELF section stack manipulation directives. The others are
6847 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6848 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6851 This directive replaces the current subsection with @code{name}. The current
6852 section is not changed. The replaced subsection is put onto the section stack
6853 in place of the then current top of stack subsection.
6858 @section @code{.symver}
6859 @cindex @code{symver} directive
6860 @cindex symbol versioning
6861 @cindex versions of symbols
6862 Use the @code{.symver} directive to bind symbols to specific version nodes
6863 within a source file. This is only supported on ELF platforms, and is
6864 typically used when assembling files to be linked into a shared library.
6865 There are cases where it may make sense to use this in objects to be bound
6866 into an application itself so as to override a versioned symbol from a
6869 For ELF targets, the @code{.symver} directive can be used like this:
6871 .symver @var{name}, @var{name2@@nodename}
6873 If the symbol @var{name} is defined within the file
6874 being assembled, the @code{.symver} directive effectively creates a symbol
6875 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6876 just don't try and create a regular alias is that the @var{@@} character isn't
6877 permitted in symbol names. The @var{name2} part of the name is the actual name
6878 of the symbol by which it will be externally referenced. The name @var{name}
6879 itself is merely a name of convenience that is used so that it is possible to
6880 have definitions for multiple versions of a function within a single source
6881 file, and so that the compiler can unambiguously know which version of a
6882 function is being mentioned. The @var{nodename} portion of the alias should be
6883 the name of a node specified in the version script supplied to the linker when
6884 building a shared library. If you are attempting to override a versioned
6885 symbol from a shared library, then @var{nodename} should correspond to the
6886 nodename of the symbol you are trying to override.
6888 If the symbol @var{name} is not defined within the file being assembled, all
6889 references to @var{name} will be changed to @var{name2@@nodename}. If no
6890 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6893 Another usage of the @code{.symver} directive is:
6895 .symver @var{name}, @var{name2@@@@nodename}
6897 In this case, the symbol @var{name} must exist and be defined within
6898 the file being assembled. It is similar to @var{name2@@nodename}. The
6899 difference is @var{name2@@@@nodename} will also be used to resolve
6900 references to @var{name2} by the linker.
6902 The third usage of the @code{.symver} directive is:
6904 .symver @var{name}, @var{name2@@@@@@nodename}
6906 When @var{name} is not defined within the
6907 file being assembled, it is treated as @var{name2@@nodename}. When
6908 @var{name} is defined within the file being assembled, the symbol
6909 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6914 @section @code{.tag @var{structname}}
6916 @cindex COFF structure debugging
6917 @cindex structure debugging, COFF
6918 @cindex @code{tag} directive
6919 This directive is generated by compilers to include auxiliary debugging
6920 information in the symbol table. It is only permitted inside
6921 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6922 definitions in the symbol table with instances of those structures.
6925 @samp{.tag} is only used when generating COFF format output; when
6926 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6932 @section @code{.text @var{subsection}}
6934 @cindex @code{text} directive
6935 Tells @command{@value{AS}} to assemble the following statements onto the end of
6936 the text subsection numbered @var{subsection}, which is an absolute
6937 expression. If @var{subsection} is omitted, subsection number zero
6941 @section @code{.title "@var{heading}"}
6943 @cindex @code{title} directive
6944 @cindex listing control: title line
6945 Use @var{heading} as the title (second line, immediately after the
6946 source file name and pagenumber) when generating assembly listings.
6948 This directive affects subsequent pages, as well as the current page if
6949 it appears within ten lines of the top of a page.
6953 @section @code{.type}
6955 This directive is used to set the type of a symbol.
6959 @c only print the extra heading if both COFF and ELF are set
6960 @subheading COFF Version
6963 @cindex COFF symbol type
6964 @cindex symbol type, COFF
6965 @cindex @code{type} directive (COFF version)
6966 For COFF targets, this directive is permitted only within
6967 @code{.def}/@code{.endef} pairs. It is used like this:
6973 This records the integer @var{int} as the type attribute of a symbol table
6977 @samp{.type} is associated only with COFF format output; when
6978 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6979 directive but ignores it.
6985 @c only print the extra heading if both COFF and ELF are set
6986 @subheading ELF Version
6989 @cindex ELF symbol type
6990 @cindex symbol type, ELF
6991 @cindex @code{type} directive (ELF version)
6992 For ELF targets, the @code{.type} directive is used like this:
6995 .type @var{name} , @var{type description}
6998 This sets the type of symbol @var{name} to be either a
6999 function symbol or an object symbol. There are five different syntaxes
7000 supported for the @var{type description} field, in order to provide
7001 compatibility with various other assemblers.
7003 Because some of the characters used in these syntaxes (such as @samp{@@} and
7004 @samp{#}) are comment characters for some architectures, some of the syntaxes
7005 below do not work on all architectures. The first variant will be accepted by
7006 the GNU assembler on all architectures so that variant should be used for
7007 maximum portability, if you do not need to assemble your code with other
7010 The syntaxes supported are:
7013 .type <name> STT_<TYPE_IN_UPPER_CASE>
7014 .type <name>,#<type>
7015 .type <name>,@@<type>
7016 .type <name>,%<type>
7017 .type <name>,"<type>"
7020 The types supported are:
7025 Mark the symbol as being a function name.
7028 @itemx gnu_indirect_function
7029 Mark the symbol as an indirect function when evaluated during reloc
7030 processing. (This is only supported on assemblers targeting GNU systems).
7034 Mark the symbol as being a data object.
7038 Mark the symbol as being a thead-local data object.
7042 Mark the symbol as being a common data object.
7046 Does not mark the symbol in any way. It is supported just for completeness.
7048 @item gnu_unique_object
7049 Marks the symbol as being a globally unique data object. The dynamic linker
7050 will make sure that in the entire process there is just one symbol with this
7051 name and type in use. (This is only supported on assemblers targeting GNU
7056 Note: Some targets support extra types in addition to those listed above.
7062 @section @code{.uleb128 @var{expressions}}
7064 @cindex @code{uleb128} directive
7065 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
7066 compact, variable length representation of numbers used by the DWARF
7067 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
7071 @section @code{.val @var{addr}}
7073 @cindex @code{val} directive
7074 @cindex COFF value attribute
7075 @cindex value attribute, COFF
7076 This directive, permitted only within @code{.def}/@code{.endef} pairs,
7077 records the address @var{addr} as the value attribute of a symbol table
7081 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
7082 configured for @code{b.out}, it accepts this directive but ignores it.
7088 @section @code{.version "@var{string}"}
7090 @cindex @code{version} directive
7091 This directive creates a @code{.note} section and places into it an ELF
7092 formatted note of type NT_VERSION. The note's name is set to @code{string}.
7097 @section @code{.vtable_entry @var{table}, @var{offset}}
7099 @cindex @code{vtable_entry} directive
7100 This directive finds or creates a symbol @code{table} and creates a
7101 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
7104 @section @code{.vtable_inherit @var{child}, @var{parent}}
7106 @cindex @code{vtable_inherit} directive
7107 This directive finds the symbol @code{child} and finds or creates the symbol
7108 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
7109 parent whose addend is the value of the child symbol. As a special case the
7110 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
7114 @section @code{.warning "@var{string}"}
7115 @cindex warning directive
7116 Similar to the directive @code{.error}
7117 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
7120 @section @code{.weak @var{names}}
7122 @cindex @code{weak} directive
7123 This directive sets the weak attribute on the comma separated list of symbol
7124 @code{names}. If the symbols do not already exist, they will be created.
7126 On COFF targets other than PE, weak symbols are a GNU extension. This
7127 directive sets the weak attribute on the comma separated list of symbol
7128 @code{names}. If the symbols do not already exist, they will be created.
7130 On the PE target, weak symbols are supported natively as weak aliases.
7131 When a weak symbol is created that is not an alias, GAS creates an
7132 alternate symbol to hold the default value.
7135 @section @code{.weakref @var{alias}, @var{target}}
7137 @cindex @code{weakref} directive
7138 This directive creates an alias to the target symbol that enables the symbol to
7139 be referenced with weak-symbol semantics, but without actually making it weak.
7140 If direct references or definitions of the symbol are present, then the symbol
7141 will not be weak, but if all references to it are through weak references, the
7142 symbol will be marked as weak in the symbol table.
7144 The effect is equivalent to moving all references to the alias to a separate
7145 assembly source file, renaming the alias to the symbol in it, declaring the
7146 symbol as weak there, and running a reloadable link to merge the object files
7147 resulting from the assembly of the new source file and the old source file that
7148 had the references to the alias removed.
7150 The alias itself never makes to the symbol table, and is entirely handled
7151 within the assembler.
7154 @section @code{.word @var{expressions}}
7156 @cindex @code{word} directive
7157 This directive expects zero or more @var{expressions}, of any section,
7158 separated by commas.
7161 For each expression, @command{@value{AS}} emits a 32-bit number.
7164 For each expression, @command{@value{AS}} emits a 16-bit number.
7169 The size of the number emitted, and its byte order,
7170 depend on what target computer the assembly is for.
7173 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
7174 @c happen---32-bit addressability, period; no long/short jumps.
7175 @ifset DIFF-TBL-KLUGE
7176 @cindex difference tables altered
7177 @cindex altered difference tables
7179 @emph{Warning: Special Treatment to support Compilers}
7183 Machines with a 32-bit address space, but that do less than 32-bit
7184 addressing, require the following special treatment. If the machine of
7185 interest to you does 32-bit addressing (or doesn't require it;
7186 @pxref{Machine Dependencies}), you can ignore this issue.
7189 In order to assemble compiler output into something that works,
7190 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
7191 Directives of the form @samp{.word sym1-sym2} are often emitted by
7192 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
7193 directive of the form @samp{.word sym1-sym2}, and the difference between
7194 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
7195 creates a @dfn{secondary jump table}, immediately before the next label.
7196 This secondary jump table is preceded by a short-jump to the
7197 first byte after the secondary table. This short-jump prevents the flow
7198 of control from accidentally falling into the new table. Inside the
7199 table is a long-jump to @code{sym2}. The original @samp{.word}
7200 contains @code{sym1} minus the address of the long-jump to
7203 If there were several occurrences of @samp{.word sym1-sym2} before the
7204 secondary jump table, all of them are adjusted. If there was a
7205 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
7206 long-jump to @code{sym4} is included in the secondary jump table,
7207 and the @code{.word} directives are adjusted to contain @code{sym3}
7208 minus the address of the long-jump to @code{sym4}; and so on, for as many
7209 entries in the original jump table as necessary.
7212 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
7213 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
7214 assembly language programmers.
7217 @c end DIFF-TBL-KLUGE
7219 @ifclear no-space-dir
7221 @section @code{.zero @var{size}}
7223 @cindex @code{zero} directive
7224 @cindex filling memory with zero bytes
7225 This directive emits @var{size} 0-valued bytes. @var{size} must be an absolute
7226 expression. This directive is actually an alias for the @samp{.skip} directive
7227 so in can take an optional second argument of the value to store in the bytes
7228 instead of zero. Using @samp{.zero} in this way would be confusing however.
7232 @section Deprecated Directives
7234 @cindex deprecated directives
7235 @cindex obsolescent directives
7236 One day these directives won't work.
7237 They are included for compatibility with older assemblers.
7244 @node Object Attributes
7245 @chapter Object Attributes
7246 @cindex object attributes
7248 @command{@value{AS}} assembles source files written for a specific architecture
7249 into object files for that architecture. But not all object files are alike.
7250 Many architectures support incompatible variations. For instance, floating
7251 point arguments might be passed in floating point registers if the object file
7252 requires hardware floating point support---or floating point arguments might be
7253 passed in integer registers if the object file supports processors with no
7254 hardware floating point unit. Or, if two objects are built for different
7255 generations of the same architecture, the combination may require the
7256 newer generation at run-time.
7258 This information is useful during and after linking. At link time,
7259 @command{@value{LD}} can warn about incompatible object files. After link
7260 time, tools like @command{gdb} can use it to process the linked file
7263 Compatibility information is recorded as a series of object attributes. Each
7264 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
7265 string, and indicates who sets the meaning of the tag. The tag is an integer,
7266 and indicates what property the attribute describes. The value may be a string
7267 or an integer, and indicates how the property affects this object. Missing
7268 attributes are the same as attributes with a zero value or empty string value.
7270 Object attributes were developed as part of the ABI for the ARM Architecture.
7271 The file format is documented in @cite{ELF for the ARM Architecture}.
7274 * GNU Object Attributes:: @sc{gnu} Object Attributes
7275 * Defining New Object Attributes:: Defining New Object Attributes
7278 @node GNU Object Attributes
7279 @section @sc{gnu} Object Attributes
7281 The @code{.gnu_attribute} directive records an object attribute
7282 with vendor @samp{gnu}.
7284 Except for @samp{Tag_compatibility}, which has both an integer and a string for
7285 its value, @sc{gnu} attributes have a string value if the tag number is odd and
7286 an integer value if the tag number is even. The second bit (@code{@var{tag} &
7287 2} is set for architecture-independent attributes and clear for
7288 architecture-dependent ones.
7290 @subsection Common @sc{gnu} attributes
7292 These attributes are valid on all architectures.
7295 @item Tag_compatibility (32)
7296 The compatibility attribute takes an integer flag value and a vendor name. If
7297 the flag value is 0, the file is compatible with other toolchains. If it is 1,
7298 then the file is only compatible with the named toolchain. If it is greater
7299 than 1, the file can only be processed by other toolchains under some private
7300 arrangement indicated by the flag value and the vendor name.
7303 @subsection MIPS Attributes
7306 @item Tag_GNU_MIPS_ABI_FP (4)
7307 The floating-point ABI used by this object file. The value will be:
7311 0 for files not affected by the floating-point ABI.
7313 1 for files using the hardware floating-point ABI with a standard
7314 double-precision FPU.
7316 2 for files using the hardware floating-point ABI with a single-precision FPU.
7318 3 for files using the software floating-point ABI.
7320 4 for files using the deprecated hardware floating-point ABI which used 64-bit
7321 floating-point registers, 32-bit general-purpose registers and increased the
7322 number of callee-saved floating-point registers.
7324 5 for files using the hardware floating-point ABI with a double-precision FPU
7325 with either 32-bit or 64-bit floating-point registers and 32-bit
7326 general-purpose registers.
7328 6 for files using the hardware floating-point ABI with 64-bit floating-point
7329 registers and 32-bit general-purpose registers.
7331 7 for files using the hardware floating-point ABI with 64-bit floating-point
7332 registers, 32-bit general-purpose registers and a rule that forbids the
7333 direct use of odd-numbered single-precision floating-point registers.
7337 @subsection PowerPC Attributes
7340 @item Tag_GNU_Power_ABI_FP (4)
7341 The floating-point ABI used by this object file. The value will be:
7345 0 for files not affected by the floating-point ABI.
7347 1 for files using double-precision hardware floating-point ABI.
7349 2 for files using the software floating-point ABI.
7351 3 for files using single-precision hardware floating-point ABI.
7354 @item Tag_GNU_Power_ABI_Vector (8)
7355 The vector ABI used by this object file. The value will be:
7359 0 for files not affected by the vector ABI.
7361 1 for files using general purpose registers to pass vectors.
7363 2 for files using AltiVec registers to pass vectors.
7365 3 for files using SPE registers to pass vectors.
7369 @subsection IBM z Systems Attributes
7372 @item Tag_GNU_S390_ABI_Vector (8)
7373 The vector ABI used by this object file. The value will be:
7377 0 for files not affected by the vector ABI.
7379 1 for files using software vector ABI.
7381 2 for files using hardware vector ABI.
7385 @node Defining New Object Attributes
7386 @section Defining New Object Attributes
7388 If you want to define a new @sc{gnu} object attribute, here are the places you
7389 will need to modify. New attributes should be discussed on the @samp{binutils}
7394 This manual, which is the official register of attributes.
7396 The header for your architecture @file{include/elf}, to define the tag.
7398 The @file{bfd} support file for your architecture, to merge the attribute
7399 and issue any appropriate link warnings.
7401 Test cases in @file{ld/testsuite} for merging and link warnings.
7403 @file{binutils/readelf.c} to display your attribute.
7405 GCC, if you want the compiler to mark the attribute automatically.
7411 @node Machine Dependencies
7412 @chapter Machine Dependent Features
7414 @cindex machine dependencies
7415 The machine instruction sets are (almost by definition) different on
7416 each machine where @command{@value{AS}} runs. Floating point representations
7417 vary as well, and @command{@value{AS}} often supports a few additional
7418 directives or command-line options for compatibility with other
7419 assemblers on a particular platform. Finally, some versions of
7420 @command{@value{AS}} support special pseudo-instructions for branch
7423 This chapter discusses most of these differences, though it does not
7424 include details on any machine's instruction set. For details on that
7425 subject, see the hardware manufacturer's manual.
7429 * AArch64-Dependent:: AArch64 Dependent Features
7432 * Alpha-Dependent:: Alpha Dependent Features
7435 * ARC-Dependent:: ARC Dependent Features
7438 * ARM-Dependent:: ARM Dependent Features
7441 * AVR-Dependent:: AVR Dependent Features
7444 * Blackfin-Dependent:: Blackfin Dependent Features
7447 * CR16-Dependent:: CR16 Dependent Features
7450 * CRIS-Dependent:: CRIS Dependent Features
7453 * D10V-Dependent:: D10V Dependent Features
7456 * D30V-Dependent:: D30V Dependent Features
7459 * Epiphany-Dependent:: EPIPHANY Dependent Features
7462 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7465 * HPPA-Dependent:: HPPA Dependent Features
7468 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
7471 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
7474 * i860-Dependent:: Intel 80860 Dependent Features
7477 * i960-Dependent:: Intel 80960 Dependent Features
7480 * IA-64-Dependent:: Intel IA-64 Dependent Features
7483 * IP2K-Dependent:: IP2K Dependent Features
7486 * LM32-Dependent:: LM32 Dependent Features
7489 * M32C-Dependent:: M32C Dependent Features
7492 * M32R-Dependent:: M32R Dependent Features
7495 * M68K-Dependent:: M680x0 Dependent Features
7498 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7501 * Meta-Dependent :: Meta Dependent Features
7504 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7507 * MIPS-Dependent:: MIPS Dependent Features
7510 * MMIX-Dependent:: MMIX Dependent Features
7513 * MSP430-Dependent:: MSP430 Dependent Features
7516 * NDS32-Dependent:: Andes NDS32 Dependent Features
7519 * NiosII-Dependent:: Altera Nios II Dependent Features
7522 * NS32K-Dependent:: NS32K Dependent Features
7525 * PDP-11-Dependent:: PDP-11 Dependent Features
7528 * PJ-Dependent:: picoJava Dependent Features
7531 * PPC-Dependent:: PowerPC Dependent Features
7534 * RL78-Dependent:: RL78 Dependent Features
7537 * RX-Dependent:: RX Dependent Features
7540 * S/390-Dependent:: IBM S/390 Dependent Features
7543 * SCORE-Dependent:: SCORE Dependent Features
7546 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7547 * SH64-Dependent:: SuperH SH64 Dependent Features
7550 * Sparc-Dependent:: SPARC Dependent Features
7553 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7556 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7559 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7562 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7565 * V850-Dependent:: V850 Dependent Features
7568 * Vax-Dependent:: VAX Dependent Features
7571 * Visium-Dependent:: Visium Dependent Features
7574 * XGATE-Dependent:: XGATE Features
7577 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7580 * Xtensa-Dependent:: Xtensa Dependent Features
7583 * Z80-Dependent:: Z80 Dependent Features
7586 * Z8000-Dependent:: Z8000 Dependent Features
7593 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7594 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7595 @c peculiarity: to preserve cross-references, there must be a node called
7596 @c "Machine Dependencies". Hence the conditional nodenames in each
7597 @c major node below. Node defaulting in makeinfo requires adjacency of
7598 @c node and sectioning commands; hence the repetition of @chapter BLAH
7599 @c in both conditional blocks.
7602 @include c-aarch64.texi
7606 @include c-alpha.texi
7622 @include c-bfin.texi
7626 @include c-cr16.texi
7630 @include c-cris.texi
7635 @node Machine Dependencies
7636 @chapter Machine Dependent Features
7638 The machine instruction sets are different on each Renesas chip family,
7639 and there are also some syntax differences among the families. This
7640 chapter describes the specific @command{@value{AS}} features for each
7644 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7645 * SH-Dependent:: Renesas SH Dependent Features
7652 @include c-d10v.texi
7656 @include c-d30v.texi
7660 @include c-epiphany.texi
7664 @include c-h8300.texi
7668 @include c-hppa.texi
7672 @include c-i370.texi
7676 @include c-i386.texi
7680 @include c-i860.texi
7684 @include c-i960.texi
7688 @include c-ia64.texi
7692 @include c-ip2k.texi
7696 @include c-lm32.texi
7700 @include c-m32c.texi
7704 @include c-m32r.texi
7708 @include c-m68k.texi
7712 @include c-m68hc11.texi
7716 @include c-metag.texi
7720 @include c-microblaze.texi
7724 @include c-mips.texi
7728 @include c-mmix.texi
7732 @include c-msp430.texi
7736 @include c-nds32.texi
7740 @include c-nios2.texi
7744 @include c-ns32k.texi
7748 @include c-pdp11.texi
7760 @include c-rl78.texi
7768 @include c-s390.texi
7772 @include c-score.texi
7777 @include c-sh64.texi
7781 @include c-sparc.texi
7785 @include c-tic54x.texi
7789 @include c-tic6x.texi
7793 @include c-tilegx.texi
7797 @include c-tilepro.texi
7801 @include c-v850.texi
7809 @include c-visium.texi
7813 @include c-xgate.texi
7817 @include c-xstormy16.texi
7821 @include c-xtensa.texi
7833 @c reverse effect of @down at top of generic Machine-Dep chapter
7837 @node Reporting Bugs
7838 @chapter Reporting Bugs
7839 @cindex bugs in assembler
7840 @cindex reporting bugs in assembler
7842 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7844 Reporting a bug may help you by bringing a solution to your problem, or it may
7845 not. But in any case the principal function of a bug report is to help the
7846 entire community by making the next version of @command{@value{AS}} work better.
7847 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7849 In order for a bug report to serve its purpose, you must include the
7850 information that enables us to fix the bug.
7853 * Bug Criteria:: Have you found a bug?
7854 * Bug Reporting:: How to report bugs
7858 @section Have You Found a Bug?
7859 @cindex bug criteria
7861 If you are not sure whether you have found a bug, here are some guidelines:
7864 @cindex fatal signal
7865 @cindex assembler crash
7866 @cindex crash of assembler
7868 If the assembler gets a fatal signal, for any input whatever, that is a
7869 @command{@value{AS}} bug. Reliable assemblers never crash.
7871 @cindex error on valid input
7873 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7875 @cindex invalid input
7877 If @command{@value{AS}} does not produce an error message for invalid input, that
7878 is a bug. However, you should note that your idea of ``invalid input'' might
7879 be our idea of ``an extension'' or ``support for traditional practice''.
7882 If you are an experienced user of assemblers, your suggestions for improvement
7883 of @command{@value{AS}} are welcome in any case.
7887 @section How to Report Bugs
7889 @cindex assembler bugs, reporting
7891 A number of companies and individuals offer support for @sc{gnu} products. If
7892 you obtained @command{@value{AS}} from a support organization, we recommend you
7893 contact that organization first.
7895 You can find contact information for many support companies and
7896 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7900 In any event, we also recommend that you send bug reports for @command{@value{AS}}
7904 The fundamental principle of reporting bugs usefully is this:
7905 @strong{report all the facts}. If you are not sure whether to state a
7906 fact or leave it out, state it!
7908 Often people omit facts because they think they know what causes the problem
7909 and assume that some details do not matter. Thus, you might assume that the
7910 name of a symbol you use in an example does not matter. Well, probably it does
7911 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
7912 happens to fetch from the location where that name is stored in memory;
7913 perhaps, if the name were different, the contents of that location would fool
7914 the assembler into doing the right thing despite the bug. Play it safe and
7915 give a specific, complete example. That is the easiest thing for you to do,
7916 and the most helpful.
7918 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
7919 it is new to us. Therefore, always write your bug reports on the assumption
7920 that the bug has not been reported previously.
7922 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7923 bell?'' This cannot help us fix a bug, so it is basically useless. We
7924 respond by asking for enough details to enable us to investigate.
7925 You might as well expedite matters by sending them to begin with.
7927 To enable us to fix the bug, you should include all these things:
7931 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
7932 it with the @samp{--version} argument.
7934 Without this, we will not know whether there is any point in looking for
7935 the bug in the current version of @command{@value{AS}}.
7938 Any patches you may have applied to the @command{@value{AS}} source.
7941 The type of machine you are using, and the operating system name and
7945 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
7949 The command arguments you gave the assembler to assemble your example and
7950 observe the bug. To guarantee you will not omit something important, list them
7951 all. A copy of the Makefile (or the output from make) is sufficient.
7953 If we were to try to guess the arguments, we would probably guess wrong
7954 and then we might not encounter the bug.
7957 A complete input file that will reproduce the bug. If the bug is observed when
7958 the assembler is invoked via a compiler, send the assembler source, not the
7959 high level language source. Most compilers will produce the assembler source
7960 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
7961 the options @samp{-v --save-temps}; this will save the assembler source in a
7962 file with an extension of @file{.s}, and also show you exactly how
7963 @command{@value{AS}} is being run.
7966 A description of what behavior you observe that you believe is
7967 incorrect. For example, ``It gets a fatal signal.''
7969 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
7970 will certainly notice it. But if the bug is incorrect output, we might not
7971 notice unless it is glaringly wrong. You might as well not give us a chance to
7974 Even if the problem you experience is a fatal signal, you should still say so
7975 explicitly. Suppose something strange is going on, such as, your copy of
7976 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
7977 library on your system. (This has happened!) Your copy might crash and ours
7978 would not. If you told us to expect a crash, then when ours fails to crash, we
7979 would know that the bug was not happening for us. If you had not told us to
7980 expect a crash, then we would not be able to draw any conclusion from our
7984 If you wish to suggest changes to the @command{@value{AS}} source, send us context
7985 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
7986 option. Always send diffs from the old file to the new file. If you even
7987 discuss something in the @command{@value{AS}} source, refer to it by context, not
7990 The line numbers in our development sources will not match those in your
7991 sources. Your line numbers would convey no useful information to us.
7994 Here are some things that are not necessary:
7998 A description of the envelope of the bug.
8000 Often people who encounter a bug spend a lot of time investigating
8001 which changes to the input file will make the bug go away and which
8002 changes will not affect it.
8004 This is often time consuming and not very useful, because the way we
8005 will find the bug is by running a single example under the debugger
8006 with breakpoints, not by pure deduction from a series of examples.
8007 We recommend that you save your time for something else.
8009 Of course, if you can find a simpler example to report @emph{instead}
8010 of the original one, that is a convenience for us. Errors in the
8011 output will be easier to spot, running under the debugger will take
8012 less time, and so on.
8014 However, simplification is not vital; if you do not want to do this,
8015 report the bug anyway and send us the entire test case you used.
8018 A patch for the bug.
8020 A patch for the bug does help us if it is a good one. But do not omit
8021 the necessary information, such as the test case, on the assumption that
8022 a patch is all we need. We might see problems with your patch and decide
8023 to fix the problem another way, or we might not understand it at all.
8025 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
8026 construct an example that will make the program follow a certain path through
8027 the code. If you do not send us the example, we will not be able to construct
8028 one, so we will not be able to verify that the bug is fixed.
8030 And if we cannot understand what bug you are trying to fix, or why your
8031 patch should be an improvement, we will not install it. A test case will
8032 help us to understand.
8035 A guess about what the bug is or what it depends on.
8037 Such guesses are usually wrong. Even we cannot guess right about such
8038 things without first using the debugger to find the facts.
8041 @node Acknowledgements
8042 @chapter Acknowledgements
8044 If you have contributed to GAS and your name isn't listed here,
8045 it is not meant as a slight. We just don't know about it. Send mail to the
8046 maintainer, and we'll correct the situation. Currently
8048 the maintainer is Nick Clifton (email address @code{nickc@@redhat.com}).
8050 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
8053 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
8054 information and the 68k series machines, most of the preprocessing pass, and
8055 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
8057 K. Richard Pixley maintained GAS for a while, adding various enhancements and
8058 many bug fixes, including merging support for several processors, breaking GAS
8059 up to handle multiple object file format back ends (including heavy rewrite,
8060 testing, an integration of the coff and b.out back ends), adding configuration
8061 including heavy testing and verification of cross assemblers and file splits
8062 and renaming, converted GAS to strictly ANSI C including full prototypes, added
8063 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
8064 port (including considerable amounts of reverse engineering), a SPARC opcode
8065 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
8066 assertions and made them work, much other reorganization, cleanup, and lint.
8068 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
8069 in format-specific I/O modules.
8071 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
8072 has done much work with it since.
8074 The Intel 80386 machine description was written by Eliot Dresselhaus.
8076 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
8078 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
8079 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
8081 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
8082 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
8083 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
8084 support a.out format.
8086 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
8087 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
8088 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
8089 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
8092 John Gilmore built the AMD 29000 support, added @code{.include} support, and
8093 simplified the configuration of which versions accept which directives. He
8094 updated the 68k machine description so that Motorola's opcodes always produced
8095 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
8096 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
8097 cross-compilation support, and one bug in relaxation that took a week and
8098 required the proverbial one-bit fix.
8100 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
8101 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
8102 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
8103 PowerPC assembler, and made a few other minor patches.
8105 Steve Chamberlain made GAS able to generate listings.
8107 Hewlett-Packard contributed support for the HP9000/300.
8109 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
8110 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
8111 formats). This work was supported by both the Center for Software Science at
8112 the University of Utah and Cygnus Support.
8114 Support for ELF format files has been worked on by Mark Eichin of Cygnus
8115 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
8116 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
8117 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
8118 and some initial 64-bit support).
8120 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
8122 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
8123 support for openVMS/Alpha.
8125 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
8128 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
8129 Inc.@: added support for Xtensa processors.
8131 Several engineers at Cygnus Support have also provided many small bug fixes and
8132 configuration enhancements.
8134 Jon Beniston added support for the Lattice Mico32 architecture.
8136 Many others have contributed large or small bugfixes and enhancements. If
8137 you have contributed significant work and are not mentioned on this list, and
8138 want to be, let us know. Some of the history has been lost; we are not
8139 intentionally leaving anyone out.
8141 @node GNU Free Documentation License
8142 @appendix GNU Free Documentation License
8146 @unnumbered AS Index