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{--target-help}] [@var{target-options}]
246 [@b{--}|@var{files} @dots{}]
248 @c Target dependent options are listed below. Keep the list sorted.
249 @c Add an empty line for separation.
252 @emph{Target AArch64 options:}
254 [@b{-mabi}=@var{ABI}]
258 @emph{Target Alpha options:}
260 [@b{-mdebug} | @b{-no-mdebug}]
261 [@b{-replace} | @b{-noreplace}]
262 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
263 [@b{-F}] [@b{-32addr}]
267 @emph{Target ARC options:}
268 [@b{-mcpu=@var{cpu}}]
269 [@b{-mA6}|@b{-mARC600}|@b{-mARC601}|@b{-mA7}|@b{-mARC700}|@b{-mEM}|@b{-mHS}]
276 @emph{Target ARM options:}
277 @c Don't document the deprecated options
278 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
279 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
280 [@b{-mfpu}=@var{floating-point-format}]
281 [@b{-mfloat-abi}=@var{abi}]
282 [@b{-meabi}=@var{ver}]
285 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
286 @b{-mapcs-reentrant}]
287 [@b{-mthumb-interwork}] [@b{-k}]
291 @emph{Target Blackfin options:}
292 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
299 @emph{Target CRIS options:}
300 [@b{--underscore} | @b{--no-underscore}]
302 [@b{--emulation=criself} | @b{--emulation=crisaout}]
303 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
304 @c Deprecated -- deliberately not documented.
309 @emph{Target D10V options:}
314 @emph{Target D30V options:}
315 [@b{-O}|@b{-n}|@b{-N}]
319 @emph{Target EPIPHANY options:}
320 [@b{-mepiphany}|@b{-mepiphany16}]
324 @emph{Target H8/300 options:}
328 @c HPPA has no machine-dependent assembler options (yet).
332 @emph{Target i386 options:}
333 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
334 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
338 @emph{Target i960 options:}
339 @c see md_parse_option in tc-i960.c
340 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
342 [@b{-b}] [@b{-no-relax}]
346 @emph{Target IA-64 options:}
347 [@b{-mconstant-gp}|@b{-mauto-pic}]
348 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
350 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
351 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
352 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
353 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
357 @emph{Target IP2K options:}
358 [@b{-mip2022}|@b{-mip2022ext}]
362 @emph{Target M32C options:}
363 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
367 @emph{Target M32R options:}
368 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
373 @emph{Target M680X0 options:}
374 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
378 @emph{Target M68HC11 options:}
379 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}|@b{-mm9s12x}|@b{-mm9s12xg}]
380 [@b{-mshort}|@b{-mlong}]
381 [@b{-mshort-double}|@b{-mlong-double}]
382 [@b{--force-long-branches}] [@b{--short-branches}]
383 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
384 [@b{--print-opcodes}] [@b{--generate-example}]
388 @emph{Target MCORE options:}
389 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
390 [@b{-mcpu=[210|340]}]
394 @emph{Target Meta options:}
395 [@b{-mcpu=@var{cpu}}] [@b{-mfpu=@var{cpu}}] [@b{-mdsp=@var{cpu}}]
398 @emph{Target MICROBLAZE options:}
399 @c MicroBlaze has no machine-dependent assembler options.
403 @emph{Target MIPS options:}
404 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
405 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
406 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
407 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
408 [@b{-mfp64}] [@b{-mgp64}] [@b{-mfpxx}]
409 [@b{-modd-spreg}] [@b{-mno-odd-spreg}]
410 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
411 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
412 [@b{-mips32r3}] [@b{-mips32r5}] [@b{-mips32r6}] [@b{-mips64}] [@b{-mips64r2}]
413 [@b{-mips64r3}] [@b{-mips64r5}] [@b{-mips64r6}]
414 [@b{-construct-floats}] [@b{-no-construct-floats}]
415 [@b{-mnan=@var{encoding}}]
416 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
417 [@b{-mips16}] [@b{-no-mips16}]
418 [@b{-mmicromips}] [@b{-mno-micromips}]
419 [@b{-msmartmips}] [@b{-mno-smartmips}]
420 [@b{-mips3d}] [@b{-no-mips3d}]
421 [@b{-mdmx}] [@b{-no-mdmx}]
422 [@b{-mdsp}] [@b{-mno-dsp}]
423 [@b{-mdspr2}] [@b{-mno-dspr2}]
424 [@b{-mmsa}] [@b{-mno-msa}]
425 [@b{-mxpa}] [@b{-mno-xpa}]
426 [@b{-mmt}] [@b{-mno-mt}]
427 [@b{-mmcu}] [@b{-mno-mcu}]
428 [@b{-minsn32}] [@b{-mno-insn32}]
429 [@b{-mfix7000}] [@b{-mno-fix7000}]
430 [@b{-mfix-rm7000}] [@b{-mno-fix-rm7000}]
431 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
432 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
433 [@b{-mdebug}] [@b{-no-mdebug}]
434 [@b{-mpdr}] [@b{-mno-pdr}]
438 @emph{Target MMIX options:}
439 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
440 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
441 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
442 [@b{--linker-allocated-gregs}]
446 @emph{Target Nios II options:}
447 [@b{-relax-all}] [@b{-relax-section}] [@b{-no-relax}]
452 @emph{Target NDS32 options:}
453 [@b{-EL}] [@b{-EB}] [@b{-O}] [@b{-Os}] [@b{-mcpu=@var{cpu}}]
454 [@b{-misa=@var{isa}}] [@b{-mabi=@var{abi}}] [@b{-mall-ext}]
455 [@b{-m[no-]16-bit}] [@b{-m[no-]perf-ext}] [@b{-m[no-]perf2-ext}]
456 [@b{-m[no-]string-ext}] [@b{-m[no-]dsp-ext}] [@b{-m[no-]mac}] [@b{-m[no-]div}]
457 [@b{-m[no-]audio-isa-ext}] [@b{-m[no-]fpu-sp-ext}] [@b{-m[no-]fpu-dp-ext}]
458 [@b{-m[no-]fpu-fma}] [@b{-mfpu-freg=@var{FREG}}] [@b{-mreduced-regs}]
459 [@b{-mfull-regs}] [@b{-m[no-]dx-regs}] [@b{-mpic}] [@b{-mno-relax}]
464 @emph{Target PDP11 options:}
465 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
466 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
467 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
471 @emph{Target picoJava options:}
476 @emph{Target PowerPC options:}
478 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
479 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
480 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
481 @b{-mbooke}|@b{-mpower4}|@b{-mpwr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
482 @b{-mpower7}|@b{-mpwr7}|@b{-mpower8}|@b{-mpwr8}|@b{-mpower9}|@b{-mpwr9}@b{-ma2}|
483 @b{-mcell}|@b{-mspe}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
484 [@b{-many}] [@b{-maltivec}|@b{-mvsx}|@b{-mhtm}|@b{-mvle}]
485 [@b{-mregnames}|@b{-mno-regnames}]
486 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
487 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
488 [@b{-msolaris}|@b{-mno-solaris}]
489 [@b{-nops=@var{count}}]
493 @emph{Target RL78 options:}
495 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
499 @emph{Target RX options:}
500 [@b{-mlittle-endian}|@b{-mbig-endian}]
501 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
502 [@b{-muse-conventional-section-names}]
503 [@b{-msmall-data-limit}]
506 [@b{-mint-register=@var{number}}]
507 [@b{-mgcc-abi}|@b{-mrx-abi}]
511 @emph{Target s390 options:}
512 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
513 [@b{-mregnames}|@b{-mno-regnames}]
514 [@b{-mwarn-areg-zero}]
518 @emph{Target SCORE options:}
519 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
520 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
521 [@b{-march=score7}][@b{-march=score3}]
522 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
526 @emph{Target SPARC options:}
527 @c The order here is important. See c-sparc.texi.
528 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
529 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
530 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
535 @emph{Target TIC54X options:}
536 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
537 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
541 @emph{Target TIC6X options:}
542 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
543 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
544 [@b{-mpic}|@b{-mno-pic}]
548 @emph{Target TILE-Gx options:}
549 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
552 @c TILEPro has no machine-dependent assembler options
556 @emph{Target Visium options:}
557 [@b{-mtune=@var{arch}}]
561 @emph{Target Xtensa options:}
562 [@b{--[no-]text-section-literals}] [@b{--[no-]auto-litpools}]
563 [@b{--[no-]absolute-literals}]
564 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
565 [@b{--[no-]transform}]
566 [@b{--rename-section} @var{oldname}=@var{newname}]
567 [@b{--[no-]trampolines}]
571 @emph{Target Z80 options:}
572 [@b{-z80}] [@b{-r800}]
573 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
574 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
575 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
576 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
577 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
578 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
582 @c Z8000 has no machine-dependent assembler options
591 @include at-file.texi
594 Turn on listings, in any of a variety of ways:
598 omit false conditionals
601 omit debugging directives
604 include general information, like @value{AS} version and options passed
607 include high-level source
613 include macro expansions
616 omit forms processing
622 set the name of the listing file
625 You may combine these options; for example, use @samp{-aln} for assembly
626 listing without forms processing. The @samp{=file} option, if used, must be
627 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
630 Begin in alternate macro mode.
632 @xref{Altmacro,,@code{.altmacro}}.
635 @item --compress-debug-sections
636 Compress DWARF debug sections using zlib with SHF_COMPRESSED from the
637 ELF ABI. The resulting object file may not be compatible with older
638 linkers and object file utilities. Note if compression would make a
639 given section @emph{larger} then it is not compressed.
642 @cindex @samp{--compress-debug-sections=} option
643 @item --compress-debug-sections=none
644 @itemx --compress-debug-sections=zlib
645 @itemx --compress-debug-sections=zlib-gnu
646 @itemx --compress-debug-sections=zlib-gabi
647 These options control how DWARF debug sections are compressed.
648 @option{--compress-debug-sections=none} is equivalent to
649 @option{--nocompress-debug-sections}.
650 @option{--compress-debug-sections=zlib} and
651 @option{--compress-debug-sections=zlib-gabi} are equivalent to
652 @option{--compress-debug-sections}.
653 @option{--compress-debug-sections=zlib-gnu} compresses DWARF debug
654 sections using zlib. The debug sections are renamed to begin with
655 @samp{.zdebug}. Note if compression would make a given section
656 @emph{larger} then it is not compressed nor renamed.
660 @item --nocompress-debug-sections
661 Do not compress DWARF debug sections. This is usually the default for all
662 targets except the x86/x86_64, but a configure time option can be used to
666 Ignored. This option is accepted for script compatibility with calls to
669 @item --debug-prefix-map @var{old}=@var{new}
670 When assembling files in directory @file{@var{old}}, record debugging
671 information describing them as in @file{@var{new}} instead.
673 @item --defsym @var{sym}=@var{value}
674 Define the symbol @var{sym} to be @var{value} before assembling the input file.
675 @var{value} must be an integer constant. As in C, a leading @samp{0x}
676 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
677 value. The value of the symbol can be overridden inside a source file via the
678 use of a @code{.set} pseudo-op.
681 ``fast''---skip whitespace and comment preprocessing (assume source is
686 Generate debugging information for each assembler source line using whichever
687 debug format is preferred by the target. This currently means either STABS,
691 Generate stabs debugging information for each assembler line. This
692 may help debugging assembler code, if the debugger can handle it.
695 Generate stabs debugging information for each assembler line, with GNU
696 extensions that probably only gdb can handle, and that could make other
697 debuggers crash or refuse to read your program. This
698 may help debugging assembler code. Currently the only GNU extension is
699 the location of the current working directory at assembling time.
702 Generate DWARF2 debugging information for each assembler line. This
703 may help debugging assembler code, if the debugger can handle it. Note---this
704 option is only supported by some targets, not all of them.
706 @item --gdwarf-sections
707 Instead of creating a .debug_line section, create a series of
708 .debug_line.@var{foo} sections where @var{foo} is the name of the
709 corresponding code section. For example a code section called @var{.text.func}
710 will have its dwarf line number information placed into a section called
711 @var{.debug_line.text.func}. If the code section is just called @var{.text}
712 then debug line section will still be called just @var{.debug_line} without any
715 @item --size-check=error
716 @itemx --size-check=warning
717 Issue an error or warning for invalid ELF .size directive.
720 Print a summary of the command line options and exit.
723 Print a summary of all target specific options and exit.
726 Add directory @var{dir} to the search list for @code{.include} directives.
729 Don't warn about signed overflow.
732 @ifclear DIFF-TBL-KLUGE
733 This option is accepted but has no effect on the @value{TARGET} family.
735 @ifset DIFF-TBL-KLUGE
736 Issue warnings when difference tables altered for long displacements.
741 Keep (in the symbol table) local symbols. These symbols start with
742 system-specific local label prefixes, typically @samp{.L} for ELF systems
743 or @samp{L} for traditional a.out systems.
748 @item --listing-lhs-width=@var{number}
749 Set the maximum width, in words, of the output data column for an assembler
750 listing to @var{number}.
752 @item --listing-lhs-width2=@var{number}
753 Set the maximum width, in words, of the output data column for continuation
754 lines in an assembler listing to @var{number}.
756 @item --listing-rhs-width=@var{number}
757 Set the maximum width of an input source line, as displayed in a listing, to
760 @item --listing-cont-lines=@var{number}
761 Set the maximum number of lines printed in a listing for a single line of input
764 @item -o @var{objfile}
765 Name the object-file output from @command{@value{AS}} @var{objfile}.
768 Fold the data section into the text section.
770 @item --hash-size=@var{number}
771 Set the default size of GAS's hash tables to a prime number close to
772 @var{number}. Increasing this value can reduce the length of time it takes the
773 assembler to perform its tasks, at the expense of increasing the assembler's
774 memory requirements. Similarly reducing this value can reduce the memory
775 requirements at the expense of speed.
777 @item --reduce-memory-overheads
778 This option reduces GAS's memory requirements, at the expense of making the
779 assembly processes slower. Currently this switch is a synonym for
780 @samp{--hash-size=4051}, but in the future it may have other effects as well.
783 @item --sectname-subst
784 Honor substitution sequences in section names.
786 @xref{Section Name Substitutions,,@code{.section @var{name}}}.
791 Print the maximum space (in bytes) and total time (in seconds) used by
794 @item --strip-local-absolute
795 Remove local absolute symbols from the outgoing symbol table.
799 Print the @command{as} version.
802 Print the @command{as} version and exit.
806 Suppress warning messages.
808 @item --fatal-warnings
809 Treat warnings as errors.
812 Don't suppress warning messages or treat them as errors.
821 Generate an object file even after errors.
823 @item -- | @var{files} @dots{}
824 Standard input, or source files to assemble.
832 @xref{AArch64 Options}, for the options available when @value{AS} is configured
833 for the 64-bit mode of the ARM Architecture (AArch64).
838 The following options are available when @value{AS} is configured for the
839 64-bit mode of the ARM Architecture (AArch64).
842 @include c-aarch64.texi
843 @c ended inside the included file
851 @xref{Alpha Options}, for the options available when @value{AS} is configured
852 for an Alpha processor.
857 The following options are available when @value{AS} is configured for an Alpha
861 @include c-alpha.texi
862 @c ended inside the included file
869 The following options are available when @value{AS} is configured for an ARC
873 @item -mcpu=@var{cpu}
874 This option selects the core processor variant.
876 Select either big-endian (-EB) or little-endian (-EL) output.
878 Enable Code Density extenssion instructions.
883 The following options are available when @value{AS} is configured for the ARM
887 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
888 Specify which ARM processor variant is the target.
889 @item -march=@var{architecture}[+@var{extension}@dots{}]
890 Specify which ARM architecture variant is used by the target.
891 @item -mfpu=@var{floating-point-format}
892 Select which Floating Point architecture is the target.
893 @item -mfloat-abi=@var{abi}
894 Select which floating point ABI is in use.
896 Enable Thumb only instruction decoding.
897 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
898 Select which procedure calling convention is in use.
900 Select either big-endian (-EB) or little-endian (-EL) output.
901 @item -mthumb-interwork
902 Specify that the code has been generated with interworking between Thumb and
905 Turns on CodeComposer Studio assembly syntax compatibility mode.
907 Specify that PIC code has been generated.
915 @xref{Blackfin Options}, for the options available when @value{AS} is
916 configured for the Blackfin processor family.
921 The following options are available when @value{AS} is configured for
922 the Blackfin processor family.
926 @c ended inside the included file
933 See the info pages for documentation of the CRIS-specific options.
937 The following options are available when @value{AS} is configured for
940 @cindex D10V optimization
941 @cindex optimization, D10V
943 Optimize output by parallelizing instructions.
948 The following options are available when @value{AS} is configured for a D30V
951 @cindex D30V optimization
952 @cindex optimization, D30V
954 Optimize output by parallelizing instructions.
958 Warn when nops are generated.
960 @cindex D30V nops after 32-bit multiply
962 Warn when a nop after a 32-bit multiply instruction is generated.
968 The following options are available when @value{AS} is configured for the
969 Adapteva EPIPHANY series.
972 @xref{Epiphany Options}, for the options available when @value{AS} is
973 configured for an Epiphany processor.
978 The following options are available when @value{AS} is configured for
979 an Epiphany processor.
982 @include c-epiphany.texi
983 @c ended inside the included file
991 @xref{H8/300 Options}, for the options available when @value{AS} is configured
992 for an H8/300 processor.
997 The following options are available when @value{AS} is configured for an H8/300
1000 @c man begin INCLUDE
1001 @include c-h8300.texi
1002 @c ended inside the included file
1010 @xref{i386-Options}, for the options available when @value{AS} is
1011 configured for an i386 processor.
1015 @c man begin OPTIONS
1016 The following options are available when @value{AS} is configured for
1019 @c man begin INCLUDE
1020 @include c-i386.texi
1021 @c ended inside the included file
1026 @c man begin OPTIONS
1028 The following options are available when @value{AS} is configured for the
1029 Intel 80960 processor.
1032 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
1033 Specify which variant of the 960 architecture is the target.
1036 Add code to collect statistics about branches taken.
1039 Do not alter compare-and-branch instructions for long displacements;
1046 The following options are available when @value{AS} is configured for the
1052 Specifies that the extended IP2022 instructions are allowed.
1055 Restores the default behaviour, which restricts the permitted instructions to
1056 just the basic IP2022 ones.
1062 The following options are available when @value{AS} is configured for the
1063 Renesas M32C and M16C processors.
1068 Assemble M32C instructions.
1071 Assemble M16C instructions (the default).
1074 Enable support for link-time relaxations.
1077 Support H'00 style hex constants in addition to 0x00 style.
1083 The following options are available when @value{AS} is configured for the
1084 Renesas M32R (formerly Mitsubishi M32R) series.
1089 Specify which processor in the M32R family is the target. The default
1090 is normally the M32R, but this option changes it to the M32RX.
1092 @item --warn-explicit-parallel-conflicts or --Wp
1093 Produce warning messages when questionable parallel constructs are
1096 @item --no-warn-explicit-parallel-conflicts or --Wnp
1097 Do not produce warning messages when questionable parallel constructs are
1104 The following options are available when @value{AS} is configured for the
1105 Motorola 68000 series.
1110 Shorten references to undefined symbols, to one word instead of two.
1112 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1113 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1114 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1115 Specify what processor in the 68000 family is the target. The default
1116 is normally the 68020, but this can be changed at configuration time.
1118 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1119 The target machine does (or does not) have a floating-point coprocessor.
1120 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1121 the basic 68000 is not compatible with the 68881, a combination of the
1122 two can be specified, since it's possible to do emulation of the
1123 coprocessor instructions with the main processor.
1125 @item -m68851 | -mno-68851
1126 The target machine does (or does not) have a memory-management
1127 unit coprocessor. The default is to assume an MMU for 68020 and up.
1135 @xref{Nios II Options}, for the options available when @value{AS} is configured
1136 for an Altera Nios II processor.
1140 @c man begin OPTIONS
1141 The following options are available when @value{AS} is configured for an
1142 Altera Nios II processor.
1144 @c man begin INCLUDE
1145 @include c-nios2.texi
1146 @c ended inside the included file
1152 For details about the PDP-11 machine dependent features options,
1153 see @ref{PDP-11-Options}.
1156 @item -mpic | -mno-pic
1157 Generate position-independent (or position-dependent) code. The
1158 default is @option{-mpic}.
1161 @itemx -mall-extensions
1162 Enable all instruction set extensions. This is the default.
1164 @item -mno-extensions
1165 Disable all instruction set extensions.
1167 @item -m@var{extension} | -mno-@var{extension}
1168 Enable (or disable) a particular instruction set extension.
1171 Enable the instruction set extensions supported by a particular CPU, and
1172 disable all other extensions.
1174 @item -m@var{machine}
1175 Enable the instruction set extensions supported by a particular machine
1176 model, and disable all other extensions.
1182 The following options are available when @value{AS} is configured for
1183 a picoJava processor.
1187 @cindex PJ endianness
1188 @cindex endianness, PJ
1189 @cindex big endian output, PJ
1191 Generate ``big endian'' format output.
1193 @cindex little endian output, PJ
1195 Generate ``little endian'' format output.
1201 The following options are available when @value{AS} is configured for the
1202 Motorola 68HC11 or 68HC12 series.
1206 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1207 Specify what processor is the target. The default is
1208 defined by the configuration option when building the assembler.
1210 @item --xgate-ramoffset
1211 Instruct the linker to offset RAM addresses from S12X address space into
1212 XGATE address space.
1215 Specify to use the 16-bit integer ABI.
1218 Specify to use the 32-bit integer ABI.
1220 @item -mshort-double
1221 Specify to use the 32-bit double ABI.
1224 Specify to use the 64-bit double ABI.
1226 @item --force-long-branches
1227 Relative branches are turned into absolute ones. This concerns
1228 conditional branches, unconditional branches and branches to a
1231 @item -S | --short-branches
1232 Do not turn relative branches into absolute ones
1233 when the offset is out of range.
1235 @item --strict-direct-mode
1236 Do not turn the direct addressing mode into extended addressing mode
1237 when the instruction does not support direct addressing mode.
1239 @item --print-insn-syntax
1240 Print the syntax of instruction in case of error.
1242 @item --print-opcodes
1243 Print the list of instructions with syntax and then exit.
1245 @item --generate-example
1246 Print an example of instruction for each possible instruction and then exit.
1247 This option is only useful for testing @command{@value{AS}}.
1253 The following options are available when @command{@value{AS}} is configured
1254 for the SPARC architecture:
1257 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1258 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1259 Explicitly select a variant of the SPARC architecture.
1261 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1262 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1264 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1265 UltraSPARC extensions.
1267 @item -xarch=v8plus | -xarch=v8plusa
1268 For compatibility with the Solaris v9 assembler. These options are
1269 equivalent to -Av8plus and -Av8plusa, respectively.
1272 Warn when the assembler switches to another architecture.
1277 The following options are available when @value{AS} is configured for the 'c54x
1282 Enable extended addressing mode. All addresses and relocations will assume
1283 extended addressing (usually 23 bits).
1284 @item -mcpu=@var{CPU_VERSION}
1285 Sets the CPU version being compiled for.
1286 @item -merrors-to-file @var{FILENAME}
1287 Redirect error output to a file, for broken systems which don't support such
1288 behaviour in the shell.
1293 The following options are available when @value{AS} is configured for
1298 This option sets the largest size of an object that can be referenced
1299 implicitly with the @code{gp} register. It is only accepted for targets that
1300 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1302 @cindex MIPS endianness
1303 @cindex endianness, MIPS
1304 @cindex big endian output, MIPS
1306 Generate ``big endian'' format output.
1308 @cindex little endian output, MIPS
1310 Generate ``little endian'' format output.
1328 Generate code for a particular MIPS Instruction Set Architecture level.
1329 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1330 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1331 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1332 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips32r3},
1333 @samp{-mips32r5}, @samp{-mips32r6}, @samp{-mips64}, @samp{-mips64r2},
1334 @samp{-mips64r3}, @samp{-mips64r5}, and @samp{-mips64r6} correspond to generic
1335 MIPS V, MIPS32, MIPS32 Release 2, MIPS32 Release 3, MIPS32 Release 5, MIPS32
1336 Release 6, MIPS64, MIPS64 Release 2, MIPS64 Release 3, MIPS64 Release 5, and
1337 MIPS64 Release 6 ISA processors, respectively.
1339 @item -march=@var{cpu}
1340 Generate code for a particular MIPS CPU.
1342 @item -mtune=@var{cpu}
1343 Schedule and tune for a particular MIPS CPU.
1347 Cause nops to be inserted if the read of the destination register
1348 of an mfhi or mflo instruction occurs in the following two instructions.
1351 @itemx -mno-fix-rm7000
1352 Cause nops to be inserted if a dmult or dmultu instruction is
1353 followed by a load instruction.
1357 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1358 section instead of the standard ELF .stabs sections.
1362 Control generation of @code{.pdr} sections.
1366 The register sizes are normally inferred from the ISA and ABI, but these
1367 flags force a certain group of registers to be treated as 32 bits wide at
1368 all times. @samp{-mgp32} controls the size of general-purpose registers
1369 and @samp{-mfp32} controls the size of floating-point registers.
1373 The register sizes are normally inferred from the ISA and ABI, but these
1374 flags force a certain group of registers to be treated as 64 bits wide at
1375 all times. @samp{-mgp64} controls the size of general-purpose registers
1376 and @samp{-mfp64} controls the size of floating-point registers.
1379 The register sizes are normally inferred from the ISA and ABI, but using
1380 this flag in combination with @samp{-mabi=32} enables an ABI variant
1381 which will operate correctly with floating-point registers which are
1385 @itemx -mno-odd-spreg
1386 Enable use of floating-point operations on odd-numbered single-precision
1387 registers when supported by the ISA. @samp{-mfpxx} implies
1388 @samp{-mno-odd-spreg}, otherwise the default is @samp{-modd-spreg}.
1392 Generate code for the MIPS 16 processor. This is equivalent to putting
1393 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1394 turns off this option.
1397 @itemx -mno-micromips
1398 Generate code for the microMIPS processor. This is equivalent to putting
1399 @code{.set micromips} at the start of the assembly file. @samp{-mno-micromips}
1400 turns off this option. This is equivalent to putting @code{.set nomicromips}
1401 at the start of the assembly file.
1404 @itemx -mno-smartmips
1405 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1406 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1407 @samp{-mno-smartmips} turns off this option.
1411 Generate code for the MIPS-3D Application Specific Extension.
1412 This tells the assembler to accept MIPS-3D instructions.
1413 @samp{-no-mips3d} turns off this option.
1417 Generate code for the MDMX Application Specific Extension.
1418 This tells the assembler to accept MDMX instructions.
1419 @samp{-no-mdmx} turns off this option.
1423 Generate code for the DSP Release 1 Application Specific Extension.
1424 This tells the assembler to accept DSP Release 1 instructions.
1425 @samp{-mno-dsp} turns off this option.
1429 Generate code for the DSP Release 2 Application Specific Extension.
1430 This option implies -mdsp.
1431 This tells the assembler to accept DSP Release 2 instructions.
1432 @samp{-mno-dspr2} turns off this option.
1436 Generate code for the MIPS SIMD Architecture Extension.
1437 This tells the assembler to accept MSA instructions.
1438 @samp{-mno-msa} turns off this option.
1442 Generate code for the MIPS eXtended Physical Address (XPA) Extension.
1443 This tells the assembler to accept XPA instructions.
1444 @samp{-mno-xpa} turns off this option.
1448 Generate code for the MT Application Specific Extension.
1449 This tells the assembler to accept MT instructions.
1450 @samp{-mno-mt} turns off this option.
1454 Generate code for the MCU Application Specific Extension.
1455 This tells the assembler to accept MCU instructions.
1456 @samp{-mno-mcu} turns off this option.
1460 Only use 32-bit instruction encodings when generating code for the
1461 microMIPS processor. This option inhibits the use of any 16-bit
1462 instructions. This is equivalent to putting @code{.set insn32} at
1463 the start of the assembly file. @samp{-mno-insn32} turns off this
1464 option. This is equivalent to putting @code{.set noinsn32} at the
1465 start of the assembly file. By default @samp{-mno-insn32} is
1466 selected, allowing all instructions to be used.
1468 @item --construct-floats
1469 @itemx --no-construct-floats
1470 The @samp{--no-construct-floats} option disables the construction of
1471 double width floating point constants by loading the two halves of the
1472 value into the two single width floating point registers that make up
1473 the double width register. By default @samp{--construct-floats} is
1474 selected, allowing construction of these floating point constants.
1476 @item --relax-branch
1477 @itemx --no-relax-branch
1478 The @samp{--relax-branch} option enables the relaxation of out-of-range
1479 branches. By default @samp{--no-relax-branch} is selected, causing any
1480 out-of-range branches to produce an error.
1482 @item -mnan=@var{encoding}
1483 Select between the IEEE 754-2008 (@option{-mnan=2008}) or the legacy
1484 (@option{-mnan=legacy}) NaN encoding format. The latter is the default.
1487 @item --emulation=@var{name}
1488 This option was formerly used to switch between ELF and ECOFF output
1489 on targets like IRIX 5 that supported both. MIPS ECOFF support was
1490 removed in GAS 2.24, so the option now serves little purpose.
1491 It is retained for backwards compatibility.
1493 The available configuration names are: @samp{mipself}, @samp{mipslelf} and
1494 @samp{mipsbelf}. Choosing @samp{mipself} now has no effect, since the output
1495 is always ELF. @samp{mipslelf} and @samp{mipsbelf} select little- and
1496 big-endian output respectively, but @samp{-EL} and @samp{-EB} are now the
1497 preferred options instead.
1500 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1507 Control how to deal with multiplication overflow and division by zero.
1508 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1509 (and only work for Instruction Set Architecture level 2 and higher);
1510 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1514 When this option is used, @command{@value{AS}} will issue a warning every
1515 time it generates a nop instruction from a macro.
1520 The following options are available when @value{AS} is configured for
1526 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1527 The command line option @samp{-nojsri2bsr} can be used to disable it.
1531 Enable or disable the silicon filter behaviour. By default this is disabled.
1532 The default can be overridden by the @samp{-sifilter} command line option.
1535 Alter jump instructions for long displacements.
1537 @item -mcpu=[210|340]
1538 Select the cpu type on the target hardware. This controls which instructions
1542 Assemble for a big endian target.
1545 Assemble for a little endian target.
1554 @xref{Meta Options}, for the options available when @value{AS} is configured
1555 for a Meta processor.
1559 @c man begin OPTIONS
1560 The following options are available when @value{AS} is configured for a
1563 @c man begin INCLUDE
1564 @include c-metag.texi
1565 @c ended inside the included file
1570 @c man begin OPTIONS
1572 See the info pages for documentation of the MMIX-specific options.
1578 @xref{NDS32 Options}, for the options available when @value{AS} is configured
1579 for a NDS32 processor.
1581 @c ended inside the included file
1585 @c man begin OPTIONS
1586 The following options are available when @value{AS} is configured for a
1589 @c man begin INCLUDE
1590 @include c-nds32.texi
1591 @c ended inside the included file
1598 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1599 for a PowerPC processor.
1603 @c man begin OPTIONS
1604 The following options are available when @value{AS} is configured for a
1607 @c man begin INCLUDE
1609 @c ended inside the included file
1614 @c man begin OPTIONS
1616 See the info pages for documentation of the RX-specific options.
1620 The following options are available when @value{AS} is configured for the s390
1626 Select the word size, either 31/32 bits or 64 bits.
1629 Select the architecture mode, either the Enterprise System
1630 Architecture (esa) or the z/Architecture mode (zarch).
1631 @item -march=@var{processor}
1632 Specify which s390 processor variant is the target, @samp{g6}, @samp{g6},
1633 @samp{z900}, @samp{z990}, @samp{z9-109}, @samp{z9-ec}, @samp{z10},
1634 @samp{z196}, @samp{zEC12}, or @samp{z13}.
1636 @itemx -mno-regnames
1637 Allow or disallow symbolic names for registers.
1638 @item -mwarn-areg-zero
1639 Warn whenever the operand for a base or index register has been specified
1640 but evaluates to zero.
1648 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1649 for a TMS320C6000 processor.
1653 @c man begin OPTIONS
1654 The following options are available when @value{AS} is configured for a
1655 TMS320C6000 processor.
1657 @c man begin INCLUDE
1658 @include c-tic6x.texi
1659 @c ended inside the included file
1667 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1668 for a TILE-Gx processor.
1672 @c man begin OPTIONS
1673 The following options are available when @value{AS} is configured for a TILE-Gx
1676 @c man begin INCLUDE
1677 @include c-tilegx.texi
1678 @c ended inside the included file
1686 @xref{Visium Options}, for the options available when @value{AS} is configured
1687 for a Visium processor.
1691 @c man begin OPTIONS
1692 The following option is available when @value{AS} is configured for a Visium
1695 @c man begin INCLUDE
1696 @include c-visium.texi
1697 @c ended inside the included file
1705 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1706 for an Xtensa processor.
1710 @c man begin OPTIONS
1711 The following options are available when @value{AS} is configured for an
1714 @c man begin INCLUDE
1715 @include c-xtensa.texi
1716 @c ended inside the included file
1721 @c man begin OPTIONS
1724 The following options are available when @value{AS} is configured for
1725 a Z80 family processor.
1728 Assemble for Z80 processor.
1730 Assemble for R800 processor.
1731 @item -ignore-undocumented-instructions
1733 Assemble undocumented Z80 instructions that also work on R800 without warning.
1734 @item -ignore-unportable-instructions
1736 Assemble all undocumented Z80 instructions without warning.
1737 @item -warn-undocumented-instructions
1739 Issue a warning for undocumented Z80 instructions that also work on R800.
1740 @item -warn-unportable-instructions
1742 Issue a warning for undocumented Z80 instructions that do not work on R800.
1743 @item -forbid-undocumented-instructions
1745 Treat all undocumented instructions as errors.
1746 @item -forbid-unportable-instructions
1748 Treat undocumented Z80 instructions that do not work on R800 as errors.
1755 * Manual:: Structure of this Manual
1756 * GNU Assembler:: The GNU Assembler
1757 * Object Formats:: Object File Formats
1758 * Command Line:: Command Line
1759 * Input Files:: Input Files
1760 * Object:: Output (Object) File
1761 * Errors:: Error and Warning Messages
1765 @section Structure of this Manual
1767 @cindex manual, structure and purpose
1768 This manual is intended to describe what you need to know to use
1769 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1770 notation for symbols, constants, and expressions; the directives that
1771 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1774 We also cover special features in the @value{TARGET}
1775 configuration of @command{@value{AS}}, including assembler directives.
1778 This manual also describes some of the machine-dependent features of
1779 various flavors of the assembler.
1782 @cindex machine instructions (not covered)
1783 On the other hand, this manual is @emph{not} intended as an introduction
1784 to programming in assembly language---let alone programming in general!
1785 In a similar vein, we make no attempt to introduce the machine
1786 architecture; we do @emph{not} describe the instruction set, standard
1787 mnemonics, registers or addressing modes that are standard to a
1788 particular architecture.
1790 You may want to consult the manufacturer's
1791 machine architecture manual for this information.
1795 For information on the H8/300 machine instruction set, see @cite{H8/300
1796 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1797 Programming Manual} (Renesas).
1800 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1801 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1802 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1803 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1806 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1810 @c I think this is premature---doc@cygnus.com, 17jan1991
1812 Throughout this manual, we assume that you are running @dfn{GNU},
1813 the portable operating system from the @dfn{Free Software
1814 Foundation, Inc.}. This restricts our attention to certain kinds of
1815 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1816 once this assumption is granted examples and definitions need less
1819 @command{@value{AS}} is part of a team of programs that turn a high-level
1820 human-readable series of instructions into a low-level
1821 computer-readable series of instructions. Different versions of
1822 @command{@value{AS}} are used for different kinds of computer.
1825 @c There used to be a section "Terminology" here, which defined
1826 @c "contents", "byte", "word", and "long". Defining "word" to any
1827 @c particular size is confusing when the .word directive may generate 16
1828 @c bits on one machine and 32 bits on another; in general, for the user
1829 @c version of this manual, none of these terms seem essential to define.
1830 @c They were used very little even in the former draft of the manual;
1831 @c this draft makes an effort to avoid them (except in names of
1835 @section The GNU Assembler
1837 @c man begin DESCRIPTION
1839 @sc{gnu} @command{as} is really a family of assemblers.
1841 This manual describes @command{@value{AS}}, a member of that family which is
1842 configured for the @value{TARGET} architectures.
1844 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1845 should find a fairly similar environment when you use it on another
1846 architecture. Each version has much in common with the others,
1847 including object file formats, most assembler directives (often called
1848 @dfn{pseudo-ops}) and assembler syntax.@refill
1850 @cindex purpose of @sc{gnu} assembler
1851 @command{@value{AS}} is primarily intended to assemble the output of the
1852 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1853 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1854 assemble correctly everything that other assemblers for the same
1855 machine would assemble.
1857 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1860 @c This remark should appear in generic version of manual; assumption
1861 @c here is that generic version sets M680x0.
1862 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1863 assembler for the same architecture; for example, we know of several
1864 incompatible versions of 680x0 assembly language syntax.
1869 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1870 program in one pass of the source file. This has a subtle impact on the
1871 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1873 @node Object Formats
1874 @section Object File Formats
1876 @cindex object file format
1877 The @sc{gnu} assembler can be configured to produce several alternative
1878 object file formats. For the most part, this does not affect how you
1879 write assembly language programs; but directives for debugging symbols
1880 are typically different in different file formats. @xref{Symbol
1881 Attributes,,Symbol Attributes}.
1884 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1885 @value{OBJ-NAME} format object files.
1887 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1889 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1890 @code{b.out} or COFF format object files.
1893 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1894 SOM or ELF format object files.
1899 @section Command Line
1901 @cindex command line conventions
1903 After the program name @command{@value{AS}}, the command line may contain
1904 options and file names. Options may appear in any order, and may be
1905 before, after, or between file names. The order of file names is
1908 @cindex standard input, as input file
1910 @file{--} (two hyphens) by itself names the standard input file
1911 explicitly, as one of the files for @command{@value{AS}} to assemble.
1913 @cindex options, command line
1914 Except for @samp{--} any command line argument that begins with a
1915 hyphen (@samp{-}) is an option. Each option changes the behavior of
1916 @command{@value{AS}}. No option changes the way another option works. An
1917 option is a @samp{-} followed by one or more letters; the case of
1918 the letter is important. All options are optional.
1920 Some options expect exactly one file name to follow them. The file
1921 name may either immediately follow the option's letter (compatible
1922 with older assemblers) or it may be the next command argument (@sc{gnu}
1923 standard). These two command lines are equivalent:
1926 @value{AS} -o my-object-file.o mumble.s
1927 @value{AS} -omy-object-file.o mumble.s
1931 @section Input Files
1934 @cindex source program
1935 @cindex files, input
1936 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1937 describe the program input to one run of @command{@value{AS}}. The program may
1938 be in one or more files; how the source is partitioned into files
1939 doesn't change the meaning of the source.
1941 @c I added "con" prefix to "catenation" just to prove I can overcome my
1942 @c APL training... doc@cygnus.com
1943 The source program is a concatenation of the text in all the files, in the
1946 @c man begin DESCRIPTION
1947 Each time you run @command{@value{AS}} it assembles exactly one source
1948 program. The source program is made up of one or more files.
1949 (The standard input is also a file.)
1951 You give @command{@value{AS}} a command line that has zero or more input file
1952 names. The input files are read (from left file name to right). A
1953 command line argument (in any position) that has no special meaning
1954 is taken to be an input file name.
1956 If you give @command{@value{AS}} no file names it attempts to read one input file
1957 from the @command{@value{AS}} standard input, which is normally your terminal. You
1958 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1961 Use @samp{--} if you need to explicitly name the standard input file
1962 in your command line.
1964 If the source is empty, @command{@value{AS}} produces a small, empty object
1969 @subheading Filenames and Line-numbers
1971 @cindex input file linenumbers
1972 @cindex line numbers, in input files
1973 There are two ways of locating a line in the input file (or files) and
1974 either may be used in reporting error messages. One way refers to a line
1975 number in a physical file; the other refers to a line number in a
1976 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1978 @dfn{Physical files} are those files named in the command line given
1979 to @command{@value{AS}}.
1981 @dfn{Logical files} are simply names declared explicitly by assembler
1982 directives; they bear no relation to physical files. Logical file names help
1983 error messages reflect the original source file, when @command{@value{AS}} source
1984 is itself synthesized from other files. @command{@value{AS}} understands the
1985 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1986 @ref{File,,@code{.file}}.
1989 @section Output (Object) File
1995 Every time you run @command{@value{AS}} it produces an output file, which is
1996 your assembly language program translated into numbers. This file
1997 is the object file. Its default name is
2005 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
2007 You can give it another name by using the @option{-o} option. Conventionally,
2008 object file names end with @file{.o}. The default name is used for historical
2009 reasons: older assemblers were capable of assembling self-contained programs
2010 directly into a runnable program. (For some formats, this isn't currently
2011 possible, but it can be done for the @code{a.out} format.)
2015 The object file is meant for input to the linker @code{@value{LD}}. It contains
2016 assembled program code, information to help @code{@value{LD}} integrate
2017 the assembled program into a runnable file, and (optionally) symbolic
2018 information for the debugger.
2020 @c link above to some info file(s) like the description of a.out.
2021 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
2024 @section Error and Warning Messages
2026 @c man begin DESCRIPTION
2028 @cindex error messages
2029 @cindex warning messages
2030 @cindex messages from assembler
2031 @command{@value{AS}} may write warnings and error messages to the standard error
2032 file (usually your terminal). This should not happen when a compiler
2033 runs @command{@value{AS}} automatically. Warnings report an assumption made so
2034 that @command{@value{AS}} could keep assembling a flawed program; errors report a
2035 grave problem that stops the assembly.
2039 @cindex format of warning messages
2040 Warning messages have the format
2043 file_name:@b{NNN}:Warning Message Text
2047 @cindex file names and line numbers, in warnings/errors
2048 (where @b{NNN} is a line number). If both a logical file name
2049 (@pxref{File,,@code{.file}}) and a logical line number
2051 (@pxref{Line,,@code{.line}})
2053 have been given then they will be used, otherwise the file name and line number
2054 in the current assembler source file will be used. The message text is
2055 intended to be self explanatory (in the grand Unix tradition).
2057 Note the file name must be set via the logical version of the @code{.file}
2058 directive, not the DWARF2 version of the @code{.file} directive. For example:
2062 error_assembler_source
2068 produces this output:
2072 asm.s:2: Error: no such instruction: `error_assembler_source'
2073 foo.c:31: Error: no such instruction: `error_c_source'
2076 @cindex format of error messages
2077 Error messages have the format
2080 file_name:@b{NNN}:FATAL:Error Message Text
2083 The file name and line number are derived as for warning
2084 messages. The actual message text may be rather less explanatory
2085 because many of them aren't supposed to happen.
2088 @chapter Command-Line Options
2090 @cindex options, all versions of assembler
2091 This chapter describes command-line options available in @emph{all}
2092 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
2093 for options specific
2095 to the @value{TARGET} target.
2098 to particular machine architectures.
2101 @c man begin DESCRIPTION
2103 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
2104 you can use the @samp{-Wa} option to pass arguments through to the assembler.
2105 The assembler arguments must be separated from each other (and the @samp{-Wa})
2106 by commas. For example:
2109 gcc -c -g -O -Wa,-alh,-L file.c
2113 This passes two options to the assembler: @samp{-alh} (emit a listing to
2114 standard output with high-level and assembly source) and @samp{-L} (retain
2115 local symbols in the symbol table).
2117 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
2118 command-line options are automatically passed to the assembler by the compiler.
2119 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
2120 precisely what options it passes to each compilation pass, including the
2126 * a:: -a[cdghlns] enable listings
2127 * alternate:: --alternate enable alternate macro syntax
2128 * D:: -D for compatibility
2129 * f:: -f to work faster
2130 * I:: -I for .include search path
2131 @ifclear DIFF-TBL-KLUGE
2132 * K:: -K for compatibility
2134 @ifset DIFF-TBL-KLUGE
2135 * K:: -K for difference tables
2138 * L:: -L to retain local symbols
2139 * listing:: --listing-XXX to configure listing output
2140 * M:: -M or --mri to assemble in MRI compatibility mode
2141 * MD:: --MD for dependency tracking
2142 * o:: -o to name the object file
2143 * R:: -R to join data and text sections
2144 * statistics:: --statistics to see statistics about assembly
2145 * traditional-format:: --traditional-format for compatible output
2146 * v:: -v to announce version
2147 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
2148 * Z:: -Z to make object file even after errors
2152 @section Enable Listings: @option{-a[cdghlns]}
2162 @cindex listings, enabling
2163 @cindex assembly listings, enabling
2165 These options enable listing output from the assembler. By itself,
2166 @samp{-a} requests high-level, assembly, and symbols listing.
2167 You can use other letters to select specific options for the list:
2168 @samp{-ah} requests a high-level language listing,
2169 @samp{-al} requests an output-program assembly listing, and
2170 @samp{-as} requests a symbol table listing.
2171 High-level listings require that a compiler debugging option like
2172 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
2175 Use the @samp{-ag} option to print a first section with general assembly
2176 information, like @value{AS} version, switches passed, or time stamp.
2178 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
2179 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
2180 other conditional), or a true @code{.if} followed by an @code{.else}, will be
2181 omitted from the listing.
2183 Use the @samp{-ad} option to omit debugging directives from the
2186 Once you have specified one of these options, you can further control
2187 listing output and its appearance using the directives @code{.list},
2188 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
2190 The @samp{-an} option turns off all forms processing.
2191 If you do not request listing output with one of the @samp{-a} options, the
2192 listing-control directives have no effect.
2194 The letters after @samp{-a} may be combined into one option,
2195 @emph{e.g.}, @samp{-aln}.
2197 Note if the assembler source is coming from the standard input (e.g.,
2199 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
2200 is being used) then the listing will not contain any comments or preprocessor
2201 directives. This is because the listing code buffers input source lines from
2202 stdin only after they have been preprocessed by the assembler. This reduces
2203 memory usage and makes the code more efficient.
2206 @section @option{--alternate}
2209 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
2212 @section @option{-D}
2215 This option has no effect whatsoever, but it is accepted to make it more
2216 likely that scripts written for other assemblers also work with
2217 @command{@value{AS}}.
2220 @section Work Faster: @option{-f}
2223 @cindex trusted compiler
2224 @cindex faster processing (@option{-f})
2225 @samp{-f} should only be used when assembling programs written by a
2226 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
2227 and comment preprocessing on
2228 the input file(s) before assembling them. @xref{Preprocessing,
2232 @emph{Warning:} if you use @samp{-f} when the files actually need to be
2233 preprocessed (if they contain comments, for example), @command{@value{AS}} does
2238 @section @code{.include} Search Path: @option{-I} @var{path}
2240 @kindex -I @var{path}
2241 @cindex paths for @code{.include}
2242 @cindex search path for @code{.include}
2243 @cindex @code{include} directive search path
2244 Use this option to add a @var{path} to the list of directories
2245 @command{@value{AS}} searches for files specified in @code{.include}
2246 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
2247 many times as necessary to include a variety of paths. The current
2248 working directory is always searched first; after that, @command{@value{AS}}
2249 searches any @samp{-I} directories in the same order as they were
2250 specified (left to right) on the command line.
2253 @section Difference Tables: @option{-K}
2256 @ifclear DIFF-TBL-KLUGE
2257 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2258 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2259 where it can be used to warn when the assembler alters the machine code
2260 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2261 family does not have the addressing limitations that sometimes lead to this
2262 alteration on other platforms.
2265 @ifset DIFF-TBL-KLUGE
2266 @cindex difference tables, warning
2267 @cindex warning for altered difference tables
2268 @command{@value{AS}} sometimes alters the code emitted for directives of the
2269 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2270 You can use the @samp{-K} option if you want a warning issued when this
2275 @section Include Local Symbols: @option{-L}
2278 @cindex local symbols, retaining in output
2279 Symbols beginning with system-specific local label prefixes, typically
2280 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2281 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2282 such symbols when debugging, because they are intended for the use of
2283 programs (like compilers) that compose assembler programs, not for your
2284 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2285 such symbols, so you do not normally debug with them.
2287 This option tells @command{@value{AS}} to retain those local symbols
2288 in the object file. Usually if you do this you also tell the linker
2289 @code{@value{LD}} to preserve those symbols.
2292 @section Configuring listing output: @option{--listing}
2294 The listing feature of the assembler can be enabled via the command line switch
2295 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2296 hex dump of the corresponding locations in the output object file, and displays
2297 them as a listing file. The format of this listing can be controlled by
2298 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2299 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2300 @code{.psize} (@pxref{Psize}), and
2301 @code{.eject} (@pxref{Eject}) and also by the following switches:
2304 @item --listing-lhs-width=@samp{number}
2305 @kindex --listing-lhs-width
2306 @cindex Width of first line disassembly output
2307 Sets the maximum width, in words, of the first line of the hex byte dump. This
2308 dump appears on the left hand side of the listing output.
2310 @item --listing-lhs-width2=@samp{number}
2311 @kindex --listing-lhs-width2
2312 @cindex Width of continuation lines of disassembly output
2313 Sets the maximum width, in words, of any further lines of the hex byte dump for
2314 a given input source line. If this value is not specified, it defaults to being
2315 the same as the value specified for @samp{--listing-lhs-width}. If neither
2316 switch is used the default is to one.
2318 @item --listing-rhs-width=@samp{number}
2319 @kindex --listing-rhs-width
2320 @cindex Width of source line output
2321 Sets the maximum width, in characters, of the source line that is displayed
2322 alongside the hex dump. The default value for this parameter is 100. The
2323 source line is displayed on the right hand side of the listing output.
2325 @item --listing-cont-lines=@samp{number}
2326 @kindex --listing-cont-lines
2327 @cindex Maximum number of continuation lines
2328 Sets the maximum number of continuation lines of hex dump that will be
2329 displayed for a given single line of source input. The default value is 4.
2333 @section Assemble in MRI Compatibility Mode: @option{-M}
2336 @cindex MRI compatibility mode
2337 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2338 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2339 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
2340 configured target) assembler from Microtec Research. The exact nature of the
2341 MRI syntax will not be documented here; see the MRI manuals for more
2342 information. Note in particular that the handling of macros and macro
2343 arguments is somewhat different. The purpose of this option is to permit
2344 assembling existing MRI assembler code using @command{@value{AS}}.
2346 The MRI compatibility is not complete. Certain operations of the MRI assembler
2347 depend upon its object file format, and can not be supported using other object
2348 file formats. Supporting these would require enhancing each object file format
2349 individually. These are:
2352 @item global symbols in common section
2354 The m68k MRI assembler supports common sections which are merged by the linker.
2355 Other object file formats do not support this. @command{@value{AS}} handles
2356 common sections by treating them as a single common symbol. It permits local
2357 symbols to be defined within a common section, but it can not support global
2358 symbols, since it has no way to describe them.
2360 @item complex relocations
2362 The MRI assemblers support relocations against a negated section address, and
2363 relocations which combine the start addresses of two or more sections. These
2364 are not support by other object file formats.
2366 @item @code{END} pseudo-op specifying start address
2368 The MRI @code{END} pseudo-op permits the specification of a start address.
2369 This is not supported by other object file formats. The start address may
2370 instead be specified using the @option{-e} option to the linker, or in a linker
2373 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2375 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2376 name to the output file. This is not supported by other object file formats.
2378 @item @code{ORG} pseudo-op
2380 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2381 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2382 which changes the location within the current section. Absolute sections are
2383 not supported by other object file formats. The address of a section may be
2384 assigned within a linker script.
2387 There are some other features of the MRI assembler which are not supported by
2388 @command{@value{AS}}, typically either because they are difficult or because they
2389 seem of little consequence. Some of these may be supported in future releases.
2393 @item EBCDIC strings
2395 EBCDIC strings are not supported.
2397 @item packed binary coded decimal
2399 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2400 and @code{DCB.P} pseudo-ops are not supported.
2402 @item @code{FEQU} pseudo-op
2404 The m68k @code{FEQU} pseudo-op is not supported.
2406 @item @code{NOOBJ} pseudo-op
2408 The m68k @code{NOOBJ} pseudo-op is not supported.
2410 @item @code{OPT} branch control options
2412 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2413 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2414 relaxes all branches, whether forward or backward, to an appropriate size, so
2415 these options serve no purpose.
2417 @item @code{OPT} list control options
2419 The following m68k @code{OPT} list control options are ignored: @code{C},
2420 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2421 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2423 @item other @code{OPT} options
2425 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2426 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2428 @item @code{OPT} @code{D} option is default
2430 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2431 @code{OPT NOD} may be used to turn it off.
2433 @item @code{XREF} pseudo-op.
2435 The m68k @code{XREF} pseudo-op is ignored.
2437 @item @code{.debug} pseudo-op
2439 The i960 @code{.debug} pseudo-op is not supported.
2441 @item @code{.extended} pseudo-op
2443 The i960 @code{.extended} pseudo-op is not supported.
2445 @item @code{.list} pseudo-op.
2447 The various options of the i960 @code{.list} pseudo-op are not supported.
2449 @item @code{.optimize} pseudo-op
2451 The i960 @code{.optimize} pseudo-op is not supported.
2453 @item @code{.output} pseudo-op
2455 The i960 @code{.output} pseudo-op is not supported.
2457 @item @code{.setreal} pseudo-op
2459 The i960 @code{.setreal} pseudo-op is not supported.
2464 @section Dependency Tracking: @option{--MD}
2467 @cindex dependency tracking
2470 @command{@value{AS}} can generate a dependency file for the file it creates. This
2471 file consists of a single rule suitable for @code{make} describing the
2472 dependencies of the main source file.
2474 The rule is written to the file named in its argument.
2476 This feature is used in the automatic updating of makefiles.
2479 @section Name the Object File: @option{-o}
2482 @cindex naming object file
2483 @cindex object file name
2484 There is always one object file output when you run @command{@value{AS}}. By
2485 default it has the name
2488 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
2502 You use this option (which takes exactly one filename) to give the
2503 object file a different name.
2505 Whatever the object file is called, @command{@value{AS}} overwrites any
2506 existing file of the same name.
2509 @section Join Data and Text Sections: @option{-R}
2512 @cindex data and text sections, joining
2513 @cindex text and data sections, joining
2514 @cindex joining text and data sections
2515 @cindex merging text and data sections
2516 @option{-R} tells @command{@value{AS}} to write the object file as if all
2517 data-section data lives in the text section. This is only done at
2518 the very last moment: your binary data are the same, but data
2519 section parts are relocated differently. The data section part of
2520 your object file is zero bytes long because all its bytes are
2521 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2523 When you specify @option{-R} it would be possible to generate shorter
2524 address displacements (because we do not have to cross between text and
2525 data section). We refrain from doing this simply for compatibility with
2526 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2529 When @command{@value{AS}} is configured for COFF or ELF output,
2530 this option is only useful if you use sections named @samp{.text} and
2535 @option{-R} is not supported for any of the HPPA targets. Using
2536 @option{-R} generates a warning from @command{@value{AS}}.
2540 @section Display Assembly Statistics: @option{--statistics}
2542 @kindex --statistics
2543 @cindex statistics, about assembly
2544 @cindex time, total for assembly
2545 @cindex space used, maximum for assembly
2546 Use @samp{--statistics} to display two statistics about the resources used by
2547 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2548 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2551 @node traditional-format
2552 @section Compatible Output: @option{--traditional-format}
2554 @kindex --traditional-format
2555 For some targets, the output of @command{@value{AS}} is different in some ways
2556 from the output of some existing assembler. This switch requests
2557 @command{@value{AS}} to use the traditional format instead.
2559 For example, it disables the exception frame optimizations which
2560 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2563 @section Announce Version: @option{-v}
2567 @cindex assembler version
2568 @cindex version of assembler
2569 You can find out what version of as is running by including the
2570 option @samp{-v} (which you can also spell as @samp{-version}) on the
2574 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2576 @command{@value{AS}} should never give a warning or error message when
2577 assembling compiler output. But programs written by people often
2578 cause @command{@value{AS}} to give a warning that a particular assumption was
2579 made. All such warnings are directed to the standard error file.
2583 @cindex suppressing warnings
2584 @cindex warnings, suppressing
2585 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2586 This only affects the warning messages: it does not change any particular of
2587 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2590 @kindex --fatal-warnings
2591 @cindex errors, caused by warnings
2592 @cindex warnings, causing error
2593 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2594 files that generate warnings to be in error.
2597 @cindex warnings, switching on
2598 You can switch these options off again by specifying @option{--warn}, which
2599 causes warnings to be output as usual.
2602 @section Generate Object File in Spite of Errors: @option{-Z}
2603 @cindex object file, after errors
2604 @cindex errors, continuing after
2605 After an error message, @command{@value{AS}} normally produces no output. If for
2606 some reason you are interested in object file output even after
2607 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2608 option. If there are any errors, @command{@value{AS}} continues anyways, and
2609 writes an object file after a final warning message of the form @samp{@var{n}
2610 errors, @var{m} warnings, generating bad object file.}
2615 @cindex machine-independent syntax
2616 @cindex syntax, machine-independent
2617 This chapter describes the machine-independent syntax allowed in a
2618 source file. @command{@value{AS}} syntax is similar to what many other
2619 assemblers use; it is inspired by the BSD 4.2
2624 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2628 * Preprocessing:: Preprocessing
2629 * Whitespace:: Whitespace
2630 * Comments:: Comments
2631 * Symbol Intro:: Symbols
2632 * Statements:: Statements
2633 * Constants:: Constants
2637 @section Preprocessing
2639 @cindex preprocessing
2640 The @command{@value{AS}} internal preprocessor:
2642 @cindex whitespace, removed by preprocessor
2644 adjusts and removes extra whitespace. It leaves one space or tab before
2645 the keywords on a line, and turns any other whitespace on the line into
2648 @cindex comments, removed by preprocessor
2650 removes all comments, replacing them with a single space, or an
2651 appropriate number of newlines.
2653 @cindex constants, converted by preprocessor
2655 converts character constants into the appropriate numeric values.
2658 It does not do macro processing, include file handling, or
2659 anything else you may get from your C compiler's preprocessor. You can
2660 do include file processing with the @code{.include} directive
2661 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2662 to get other ``CPP'' style preprocessing by giving the input file a
2663 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2664 Output, gcc.info, Using GNU CC}.
2666 Excess whitespace, comments, and character constants
2667 cannot be used in the portions of the input text that are not
2670 @cindex turning preprocessing on and off
2671 @cindex preprocessing, turning on and off
2674 If the first line of an input file is @code{#NO_APP} or if you use the
2675 @samp{-f} option, whitespace and comments are not removed from the input file.
2676 Within an input file, you can ask for whitespace and comment removal in
2677 specific portions of the by putting a line that says @code{#APP} before the
2678 text that may contain whitespace or comments, and putting a line that says
2679 @code{#NO_APP} after this text. This feature is mainly intend to support
2680 @code{asm} statements in compilers whose output is otherwise free of comments
2687 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2688 Whitespace is used to separate symbols, and to make programs neater for
2689 people to read. Unless within character constants
2690 (@pxref{Characters,,Character Constants}), any whitespace means the same
2691 as exactly one space.
2697 There are two ways of rendering comments to @command{@value{AS}}. In both
2698 cases the comment is equivalent to one space.
2700 Anything from @samp{/*} through the next @samp{*/} is a comment.
2701 This means you may not nest these comments.
2705 The only way to include a newline ('\n') in a comment
2706 is to use this sort of comment.
2709 /* This sort of comment does not nest. */
2712 @cindex line comment character
2713 Anything from a @dfn{line comment} character up to the next newline is
2714 considered a comment and is ignored. The line comment character is target
2715 specific, and some targets multiple comment characters. Some targets also have
2716 line comment characters that only work if they are the first character on a
2717 line. Some targets use a sequence of two characters to introduce a line
2718 comment. Some targets can also change their line comment characters depending
2719 upon command line options that have been used. For more details see the
2720 @emph{Syntax} section in the documentation for individual targets.
2722 If the line comment character is the hash sign (@samp{#}) then it still has the
2723 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2724 to specify logical line numbers:
2727 @cindex lines starting with @code{#}
2728 @cindex logical line numbers
2729 To be compatible with past assemblers, lines that begin with @samp{#} have a
2730 special interpretation. Following the @samp{#} should be an absolute
2731 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2732 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2733 new logical file name. The rest of the line, if any, should be whitespace.
2735 If the first non-whitespace characters on the line are not numeric,
2736 the line is ignored. (Just like a comment.)
2739 # This is an ordinary comment.
2740 # 42-6 "new_file_name" # New logical file name
2741 # This is logical line # 36.
2743 This feature is deprecated, and may disappear from future versions
2744 of @command{@value{AS}}.
2749 @cindex characters used in symbols
2750 @ifclear SPECIAL-SYMS
2751 A @dfn{symbol} is one or more characters chosen from the set of all
2752 letters (both upper and lower case), digits and the three characters
2758 A @dfn{symbol} is one or more characters chosen from the set of all
2759 letters (both upper and lower case), digits and the three characters
2760 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2766 On most machines, you can also use @code{$} in symbol names; exceptions
2767 are noted in @ref{Machine Dependencies}.
2769 No symbol may begin with a digit. Case is significant.
2770 There is no length limit; all characters are significant. Multibyte characters
2771 are supported. Symbols are delimited by characters not in that set, or by the
2772 beginning of a file (since the source program must end with a newline, the end
2773 of a file is not a possible symbol delimiter). @xref{Symbols}.
2775 Symbol names may also be enclosed in double quote @code{"} characters. In such
2776 cases any characters are allowed, except for the NUL character. If a double
2777 quote character is to be included in the symbol name it must be preceeded by a
2778 backslash @code{\} character.
2779 @cindex length of symbols
2784 @cindex statements, structure of
2785 @cindex line separator character
2786 @cindex statement separator character
2788 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2789 @dfn{line separator character}. The line separator character is target
2790 specific and described in the @emph{Syntax} section of each
2791 target's documentation. Not all targets support a line separator character.
2792 The newline or line separator character is considered to be part of the
2793 preceding statement. Newlines and separators within character constants are an
2794 exception: they do not end statements.
2796 @cindex newline, required at file end
2797 @cindex EOF, newline must precede
2798 It is an error to end any statement with end-of-file: the last
2799 character of any input file should be a newline.@refill
2801 An empty statement is allowed, and may include whitespace. It is ignored.
2803 @cindex instructions and directives
2804 @cindex directives and instructions
2805 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2806 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2808 A statement begins with zero or more labels, optionally followed by a
2809 key symbol which determines what kind of statement it is. The key
2810 symbol determines the syntax of the rest of the statement. If the
2811 symbol begins with a dot @samp{.} then the statement is an assembler
2812 directive: typically valid for any computer. If the symbol begins with
2813 a letter the statement is an assembly language @dfn{instruction}: it
2814 assembles into a machine language instruction.
2816 Different versions of @command{@value{AS}} for different computers
2817 recognize different instructions. In fact, the same symbol may
2818 represent a different instruction in a different computer's assembly
2822 @cindex @code{:} (label)
2823 @cindex label (@code{:})
2824 A label is a symbol immediately followed by a colon (@code{:}).
2825 Whitespace before a label or after a colon is permitted, but you may not
2826 have whitespace between a label's symbol and its colon. @xref{Labels}.
2829 For HPPA targets, labels need not be immediately followed by a colon, but
2830 the definition of a label must begin in column zero. This also implies that
2831 only one label may be defined on each line.
2835 label: .directive followed by something
2836 another_label: # This is an empty statement.
2837 instruction operand_1, operand_2, @dots{}
2844 A constant is a number, written so that its value is known by
2845 inspection, without knowing any context. Like this:
2848 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2849 .ascii "Ring the bell\7" # A string constant.
2850 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2851 .float 0f-314159265358979323846264338327\
2852 95028841971.693993751E-40 # - pi, a flonum.
2857 * Characters:: Character Constants
2858 * Numbers:: Number Constants
2862 @subsection Character Constants
2864 @cindex character constants
2865 @cindex constants, character
2866 There are two kinds of character constants. A @dfn{character} stands
2867 for one character in one byte and its value may be used in
2868 numeric expressions. String constants (properly called string
2869 @emph{literals}) are potentially many bytes and their values may not be
2870 used in arithmetic expressions.
2874 * Chars:: Characters
2878 @subsubsection Strings
2880 @cindex string constants
2881 @cindex constants, string
2882 A @dfn{string} is written between double-quotes. It may contain
2883 double-quotes or null characters. The way to get special characters
2884 into a string is to @dfn{escape} these characters: precede them with
2885 a backslash @samp{\} character. For example @samp{\\} represents
2886 one backslash: the first @code{\} is an escape which tells
2887 @command{@value{AS}} to interpret the second character literally as a backslash
2888 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2889 escape character). The complete list of escapes follows.
2891 @cindex escape codes, character
2892 @cindex character escape codes
2893 @c NOTE: Cindex entries must not start with a backlash character.
2894 @c NOTE: This confuses the pdf2texi script when it is creating the
2895 @c NOTE: index based upon the first character and so it generates:
2896 @c NOTE: \initial {\\}
2897 @c NOTE: which then results in the error message:
2898 @c NOTE: Argument of \\ has an extra }.
2899 @c NOTE: So in the index entries below a space character has been
2900 @c NOTE: prepended to avoid this problem.
2903 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2905 @cindex @code{ \b} (backspace character)
2906 @cindex backspace (@code{\b})
2908 Mnemonic for backspace; for ASCII this is octal code 010.
2911 @c Mnemonic for EOText; for ASCII this is octal code 004.
2913 @cindex @code{ \f} (formfeed character)
2914 @cindex formfeed (@code{\f})
2916 Mnemonic for FormFeed; for ASCII this is octal code 014.
2918 @cindex @code{ \n} (newline character)
2919 @cindex newline (@code{\n})
2921 Mnemonic for newline; for ASCII this is octal code 012.
2924 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2926 @cindex @code{ \r} (carriage return character)
2927 @cindex carriage return (@code{backslash-r})
2929 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2932 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2933 @c other assemblers.
2935 @cindex @code{ \t} (tab)
2936 @cindex tab (@code{\t})
2938 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2941 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2942 @c @item \x @var{digit} @var{digit} @var{digit}
2943 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2945 @cindex @code{ \@var{ddd}} (octal character code)
2946 @cindex octal character code (@code{\@var{ddd}})
2947 @item \ @var{digit} @var{digit} @var{digit}
2948 An octal character code. The numeric code is 3 octal digits.
2949 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2950 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2952 @cindex @code{ \@var{xd...}} (hex character code)
2953 @cindex hex character code (@code{\@var{xd...}})
2954 @item \@code{x} @var{hex-digits...}
2955 A hex character code. All trailing hex digits are combined. Either upper or
2956 lower case @code{x} works.
2958 @cindex @code{ \\} (@samp{\} character)
2959 @cindex backslash (@code{\\})
2961 Represents one @samp{\} character.
2964 @c Represents one @samp{'} (accent acute) character.
2965 @c This is needed in single character literals
2966 @c (@xref{Characters,,Character Constants}.) to represent
2969 @cindex @code{ \"} (doublequote character)
2970 @cindex doublequote (@code{\"})
2972 Represents one @samp{"} character. Needed in strings to represent
2973 this character, because an unescaped @samp{"} would end the string.
2975 @item \ @var{anything-else}
2976 Any other character when escaped by @kbd{\} gives a warning, but
2977 assembles as if the @samp{\} was not present. The idea is that if
2978 you used an escape sequence you clearly didn't want the literal
2979 interpretation of the following character. However @command{@value{AS}} has no
2980 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2981 code and warns you of the fact.
2984 Which characters are escapable, and what those escapes represent,
2985 varies widely among assemblers. The current set is what we think
2986 the BSD 4.2 assembler recognizes, and is a subset of what most C
2987 compilers recognize. If you are in doubt, do not use an escape
2991 @subsubsection Characters
2993 @cindex single character constant
2994 @cindex character, single
2995 @cindex constant, single character
2996 A single character may be written as a single quote immediately
2997 followed by that character. The same escapes apply to characters as
2998 to strings. So if you want to write the character backslash, you
2999 must write @kbd{'\\} where the first @code{\} escapes the second
3000 @code{\}. As you can see, the quote is an acute accent, not a
3001 grave accent. A newline
3003 @ifclear abnormal-separator
3004 (or semicolon @samp{;})
3006 @ifset abnormal-separator
3008 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
3013 immediately following an acute accent is taken as a literal character
3014 and does not count as the end of a statement. The value of a character
3015 constant in a numeric expression is the machine's byte-wide code for
3016 that character. @command{@value{AS}} assumes your character code is ASCII:
3017 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
3020 @subsection Number Constants
3022 @cindex constants, number
3023 @cindex number constants
3024 @command{@value{AS}} distinguishes three kinds of numbers according to how they
3025 are stored in the target machine. @emph{Integers} are numbers that
3026 would fit into an @code{int} in the C language. @emph{Bignums} are
3027 integers, but they are stored in more than 32 bits. @emph{Flonums}
3028 are floating point numbers, described below.
3031 * Integers:: Integers
3036 * Bit Fields:: Bit Fields
3042 @subsubsection Integers
3044 @cindex constants, integer
3046 @cindex binary integers
3047 @cindex integers, binary
3048 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
3049 the binary digits @samp{01}.
3051 @cindex octal integers
3052 @cindex integers, octal
3053 An octal integer is @samp{0} followed by zero or more of the octal
3054 digits (@samp{01234567}).
3056 @cindex decimal integers
3057 @cindex integers, decimal
3058 A decimal integer starts with a non-zero digit followed by zero or
3059 more digits (@samp{0123456789}).
3061 @cindex hexadecimal integers
3062 @cindex integers, hexadecimal
3063 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
3064 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
3066 Integers have the usual values. To denote a negative integer, use
3067 the prefix operator @samp{-} discussed under expressions
3068 (@pxref{Prefix Ops,,Prefix Operators}).
3071 @subsubsection Bignums
3074 @cindex constants, bignum
3075 A @dfn{bignum} has the same syntax and semantics as an integer
3076 except that the number (or its negative) takes more than 32 bits to
3077 represent in binary. The distinction is made because in some places
3078 integers are permitted while bignums are not.
3081 @subsubsection Flonums
3083 @cindex floating point numbers
3084 @cindex constants, floating point
3086 @cindex precision, floating point
3087 A @dfn{flonum} represents a floating point number. The translation is
3088 indirect: a decimal floating point number from the text is converted by
3089 @command{@value{AS}} to a generic binary floating point number of more than
3090 sufficient precision. This generic floating point number is converted
3091 to a particular computer's floating point format (or formats) by a
3092 portion of @command{@value{AS}} specialized to that computer.
3094 A flonum is written by writing (in order)
3099 (@samp{0} is optional on the HPPA.)
3103 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
3105 @kbd{e} is recommended. Case is not important.
3107 @c FIXME: verify if flonum syntax really this vague for most cases
3108 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
3109 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
3112 On the H8/300, Renesas / SuperH SH,
3113 and AMD 29K architectures, the letter must be
3114 one of the letters @samp{DFPRSX} (in upper or lower case).
3116 On the ARC, the letter must be one of the letters @samp{DFRS}
3117 (in upper or lower case).
3119 On the Intel 960 architecture, the letter must be
3120 one of the letters @samp{DFT} (in upper or lower case).
3122 On the HPPA architecture, the letter must be @samp{E} (upper case only).
3126 One of the letters @samp{DFRS} (in upper or lower case).
3129 One of the letters @samp{DFPRSX} (in upper or lower case).
3132 The letter @samp{E} (upper case only).
3135 One of the letters @samp{DFT} (in upper or lower case).
3140 An optional sign: either @samp{+} or @samp{-}.
3143 An optional @dfn{integer part}: zero or more decimal digits.
3146 An optional @dfn{fractional part}: @samp{.} followed by zero
3147 or more decimal digits.
3150 An optional exponent, consisting of:
3154 An @samp{E} or @samp{e}.
3155 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
3156 @c principle this can perfectly well be different on different targets.
3158 Optional sign: either @samp{+} or @samp{-}.
3160 One or more decimal digits.
3165 At least one of the integer part or the fractional part must be
3166 present. The floating point number has the usual base-10 value.
3168 @command{@value{AS}} does all processing using integers. Flonums are computed
3169 independently of any floating point hardware in the computer running
3170 @command{@value{AS}}.
3174 @c Bit fields are written as a general facility but are also controlled
3175 @c by a conditional-compilation flag---which is as of now (21mar91)
3176 @c turned on only by the i960 config of GAS.
3178 @subsubsection Bit Fields
3181 @cindex constants, bit field
3182 You can also define numeric constants as @dfn{bit fields}.
3183 Specify two numbers separated by a colon---
3185 @var{mask}:@var{value}
3188 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
3191 The resulting number is then packed
3193 @c this conditional paren in case bit fields turned on elsewhere than 960
3194 (in host-dependent byte order)
3196 into a field whose width depends on which assembler directive has the
3197 bit-field as its argument. Overflow (a result from the bitwise and
3198 requiring more binary digits to represent) is not an error; instead,
3199 more constants are generated, of the specified width, beginning with the
3200 least significant digits.@refill
3202 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
3203 @code{.short}, and @code{.word} accept bit-field arguments.
3208 @chapter Sections and Relocation
3213 * Secs Background:: Background
3214 * Ld Sections:: Linker Sections
3215 * As Sections:: Assembler Internal Sections
3216 * Sub-Sections:: Sub-Sections
3220 @node Secs Background
3223 Roughly, a section is a range of addresses, with no gaps; all data
3224 ``in'' those addresses is treated the same for some particular purpose.
3225 For example there may be a ``read only'' section.
3227 @cindex linker, and assembler
3228 @cindex assembler, and linker
3229 The linker @code{@value{LD}} reads many object files (partial programs) and
3230 combines their contents to form a runnable program. When @command{@value{AS}}
3231 emits an object file, the partial program is assumed to start at address 0.
3232 @code{@value{LD}} assigns the final addresses for the partial program, so that
3233 different partial programs do not overlap. This is actually an
3234 oversimplification, but it suffices to explain how @command{@value{AS}} uses
3237 @code{@value{LD}} moves blocks of bytes of your program to their run-time
3238 addresses. These blocks slide to their run-time addresses as rigid
3239 units; their length does not change and neither does the order of bytes
3240 within them. Such a rigid unit is called a @emph{section}. Assigning
3241 run-time addresses to sections is called @dfn{relocation}. It includes
3242 the task of adjusting mentions of object-file addresses so they refer to
3243 the proper run-time addresses.
3245 For the H8/300, and for the Renesas / SuperH SH,
3246 @command{@value{AS}} pads sections if needed to
3247 ensure they end on a word (sixteen bit) boundary.
3250 @cindex standard assembler sections
3251 An object file written by @command{@value{AS}} has at least three sections, any
3252 of which may be empty. These are named @dfn{text}, @dfn{data} and
3257 When it generates COFF or ELF output,
3259 @command{@value{AS}} can also generate whatever other named sections you specify
3260 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
3261 If you do not use any directives that place output in the @samp{.text}
3262 or @samp{.data} sections, these sections still exist, but are empty.
3267 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
3269 @command{@value{AS}} can also generate whatever other named sections you
3270 specify using the @samp{.space} and @samp{.subspace} directives. See
3271 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3272 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3273 assembler directives.
3276 Additionally, @command{@value{AS}} uses different names for the standard
3277 text, data, and bss sections when generating SOM output. Program text
3278 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3279 BSS into @samp{$BSS$}.
3283 Within the object file, the text section starts at address @code{0}, the
3284 data section follows, and the bss section follows the data section.
3287 When generating either SOM or ELF output files on the HPPA, the text
3288 section starts at address @code{0}, the data section at address
3289 @code{0x4000000}, and the bss section follows the data section.
3292 To let @code{@value{LD}} know which data changes when the sections are
3293 relocated, and how to change that data, @command{@value{AS}} also writes to the
3294 object file details of the relocation needed. To perform relocation
3295 @code{@value{LD}} must know, each time an address in the object
3299 Where in the object file is the beginning of this reference to
3302 How long (in bytes) is this reference?
3304 Which section does the address refer to? What is the numeric value of
3306 (@var{address}) @minus{} (@var{start-address of section})?
3309 Is the reference to an address ``Program-Counter relative''?
3312 @cindex addresses, format of
3313 @cindex section-relative addressing
3314 In fact, every address @command{@value{AS}} ever uses is expressed as
3316 (@var{section}) + (@var{offset into section})
3319 Further, most expressions @command{@value{AS}} computes have this section-relative
3322 (For some object formats, such as SOM for the HPPA, some expressions are
3323 symbol-relative instead.)
3326 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3327 @var{N} into section @var{secname}.''
3329 Apart from text, data and bss sections you need to know about the
3330 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3331 addresses in the absolute section remain unchanged. For example, address
3332 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3333 @code{@value{LD}}. Although the linker never arranges two partial programs'
3334 data sections with overlapping addresses after linking, @emph{by definition}
3335 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3336 part of a program is always the same address when the program is running as
3337 address @code{@{absolute@ 239@}} in any other part of the program.
3339 The idea of sections is extended to the @dfn{undefined} section. Any
3340 address whose section is unknown at assembly time is by definition
3341 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3342 Since numbers are always defined, the only way to generate an undefined
3343 address is to mention an undefined symbol. A reference to a named
3344 common block would be such a symbol: its value is unknown at assembly
3345 time so it has section @emph{undefined}.
3347 By analogy the word @emph{section} is used to describe groups of sections in
3348 the linked program. @code{@value{LD}} puts all partial programs' text
3349 sections in contiguous addresses in the linked program. It is
3350 customary to refer to the @emph{text section} of a program, meaning all
3351 the addresses of all partial programs' text sections. Likewise for
3352 data and bss sections.
3354 Some sections are manipulated by @code{@value{LD}}; others are invented for
3355 use of @command{@value{AS}} and have no meaning except during assembly.
3358 @section Linker Sections
3359 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3364 @cindex named sections
3365 @cindex sections, named
3366 @item named sections
3369 @cindex text section
3370 @cindex data section
3374 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3375 separate but equal sections. Anything you can say of one section is
3378 When the program is running, however, it is
3379 customary for the text section to be unalterable. The
3380 text section is often shared among processes: it contains
3381 instructions, constants and the like. The data section of a running
3382 program is usually alterable: for example, C variables would be stored
3383 in the data section.
3388 This section contains zeroed bytes when your program begins running. It
3389 is used to hold uninitialized variables or common storage. The length of
3390 each partial program's bss section is important, but because it starts
3391 out containing zeroed bytes there is no need to store explicit zero
3392 bytes in the object file. The bss section was invented to eliminate
3393 those explicit zeros from object files.
3395 @cindex absolute section
3396 @item absolute section
3397 Address 0 of this section is always ``relocated'' to runtime address 0.
3398 This is useful if you want to refer to an address that @code{@value{LD}} must
3399 not change when relocating. In this sense we speak of absolute
3400 addresses being ``unrelocatable'': they do not change during relocation.
3402 @cindex undefined section
3403 @item undefined section
3404 This ``section'' is a catch-all for address references to objects not in
3405 the preceding sections.
3406 @c FIXME: ref to some other doc on obj-file formats could go here.
3409 @cindex relocation example
3410 An idealized example of three relocatable sections follows.
3412 The example uses the traditional section names @samp{.text} and @samp{.data}.
3414 Memory addresses are on the horizontal axis.
3418 @c END TEXI2ROFF-KILL
3421 partial program # 1: |ttttt|dddd|00|
3428 partial program # 2: |TTT|DDD|000|
3431 +--+---+-----+--+----+---+-----+~~
3432 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3433 +--+---+-----+--+----+---+-----+~~
3435 addresses: 0 @dots{}
3442 \line{\it Partial program \#1: \hfil}
3443 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3444 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3446 \line{\it Partial program \#2: \hfil}
3447 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3448 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3450 \line{\it linked program: \hfil}
3451 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3452 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3453 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3454 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3456 \line{\it addresses: \hfil}
3460 @c END TEXI2ROFF-KILL
3463 @section Assembler Internal Sections
3465 @cindex internal assembler sections
3466 @cindex sections in messages, internal
3467 These sections are meant only for the internal use of @command{@value{AS}}. They
3468 have no meaning at run-time. You do not really need to know about these
3469 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3470 warning messages, so it might be helpful to have an idea of their
3471 meanings to @command{@value{AS}}. These sections are used to permit the
3472 value of every expression in your assembly language program to be a
3473 section-relative address.
3476 @cindex assembler internal logic error
3477 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3478 An internal assembler logic error has been found. This means there is a
3479 bug in the assembler.
3481 @cindex expr (internal section)
3483 The assembler stores complex expression internally as combinations of
3484 symbols. When it needs to represent an expression as a symbol, it puts
3485 it in the expr section.
3487 @c FIXME item transfer[t] vector preload
3488 @c FIXME item transfer[t] vector postload
3489 @c FIXME item register
3493 @section Sub-Sections
3495 @cindex numbered subsections
3496 @cindex grouping data
3502 fall into two sections: text and data.
3504 You may have separate groups of
3506 data in named sections
3510 data in named sections
3516 that you want to end up near to each other in the object file, even though they
3517 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3518 use @dfn{subsections} for this purpose. Within each section, there can be
3519 numbered subsections with values from 0 to 8192. Objects assembled into the
3520 same subsection go into the object file together with other objects in the same
3521 subsection. For example, a compiler might want to store constants in the text
3522 section, but might not want to have them interspersed with the program being
3523 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3524 section of code being output, and a @samp{.text 1} before each group of
3525 constants being output.
3527 Subsections are optional. If you do not use subsections, everything
3528 goes in subsection number zero.
3531 Each subsection is zero-padded up to a multiple of four bytes.
3532 (Subsections may be padded a different amount on different flavors
3533 of @command{@value{AS}}.)
3537 On the H8/300 platform, each subsection is zero-padded to a word
3538 boundary (two bytes).
3539 The same is true on the Renesas SH.
3542 @c FIXME section padding (alignment)?
3543 @c Rich Pixley says padding here depends on target obj code format; that
3544 @c doesn't seem particularly useful to say without further elaboration,
3545 @c so for now I say nothing about it. If this is a generic BFD issue,
3546 @c these paragraphs might need to vanish from this manual, and be
3547 @c discussed in BFD chapter of binutils (or some such).
3551 Subsections appear in your object file in numeric order, lowest numbered
3552 to highest. (All this to be compatible with other people's assemblers.)
3553 The object file contains no representation of subsections; @code{@value{LD}} and
3554 other programs that manipulate object files see no trace of them.
3555 They just see all your text subsections as a text section, and all your
3556 data subsections as a data section.
3558 To specify which subsection you want subsequent statements assembled
3559 into, use a numeric argument to specify it, in a @samp{.text
3560 @var{expression}} or a @samp{.data @var{expression}} statement.
3563 When generating COFF output, you
3568 can also use an extra subsection
3569 argument with arbitrary named sections: @samp{.section @var{name},
3574 When generating ELF output, you
3579 can also use the @code{.subsection} directive (@pxref{SubSection})
3580 to specify a subsection: @samp{.subsection @var{expression}}.
3582 @var{Expression} should be an absolute expression
3583 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3584 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3585 begins in @code{text 0}. For instance:
3587 .text 0 # The default subsection is text 0 anyway.
3588 .ascii "This lives in the first text subsection. *"
3590 .ascii "But this lives in the second text subsection."
3592 .ascii "This lives in the data section,"
3593 .ascii "in the first data subsection."
3595 .ascii "This lives in the first text section,"
3596 .ascii "immediately following the asterisk (*)."
3599 Each section has a @dfn{location counter} incremented by one for every byte
3600 assembled into that section. Because subsections are merely a convenience
3601 restricted to @command{@value{AS}} there is no concept of a subsection location
3602 counter. There is no way to directly manipulate a location counter---but the
3603 @code{.align} directive changes it, and any label definition captures its
3604 current value. The location counter of the section where statements are being
3605 assembled is said to be the @dfn{active} location counter.
3608 @section bss Section
3611 @cindex common variable storage
3612 The bss section is used for local common variable storage.
3613 You may allocate address space in the bss section, but you may
3614 not dictate data to load into it before your program executes. When
3615 your program starts running, all the contents of the bss
3616 section are zeroed bytes.
3618 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3619 @ref{Lcomm,,@code{.lcomm}}.
3621 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3622 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3625 When assembling for a target which supports multiple sections, such as ELF or
3626 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3627 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3628 section. Typically the section will only contain symbol definitions and
3629 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3636 Symbols are a central concept: the programmer uses symbols to name
3637 things, the linker uses symbols to link, and the debugger uses symbols
3641 @cindex debuggers, and symbol order
3642 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3643 the same order they were declared. This may break some debuggers.
3648 * Setting Symbols:: Giving Symbols Other Values
3649 * Symbol Names:: Symbol Names
3650 * Dot:: The Special Dot Symbol
3651 * Symbol Attributes:: Symbol Attributes
3658 A @dfn{label} is written as a symbol immediately followed by a colon
3659 @samp{:}. The symbol then represents the current value of the
3660 active location counter, and is, for example, a suitable instruction
3661 operand. You are warned if you use the same symbol to represent two
3662 different locations: the first definition overrides any other
3666 On the HPPA, the usual form for a label need not be immediately followed by a
3667 colon, but instead must start in column zero. Only one label may be defined on
3668 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3669 provides a special directive @code{.label} for defining labels more flexibly.
3672 @node Setting Symbols
3673 @section Giving Symbols Other Values
3675 @cindex assigning values to symbols
3676 @cindex symbol values, assigning
3677 A symbol can be given an arbitrary value by writing a symbol, followed
3678 by an equals sign @samp{=}, followed by an expression
3679 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3680 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3681 equals sign @samp{=}@samp{=} here represents an equivalent of the
3682 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3685 Blackfin does not support symbol assignment with @samp{=}.
3689 @section Symbol Names
3691 @cindex symbol names
3692 @cindex names, symbol
3693 @ifclear SPECIAL-SYMS
3694 Symbol names begin with a letter or with one of @samp{._}. On most
3695 machines, you can also use @code{$} in symbol names; exceptions are
3696 noted in @ref{Machine Dependencies}. That character may be followed by any
3697 string of digits, letters, dollar signs (unless otherwise noted for a
3698 particular target machine), and underscores.
3702 Symbol names begin with a letter or with one of @samp{._}. On the
3703 Renesas SH you can also use @code{$} in symbol names. That
3704 character may be followed by any string of digits, letters, dollar signs (save
3705 on the H8/300), and underscores.
3709 Case of letters is significant: @code{foo} is a different symbol name
3712 Symbol names do not start with a digit. An exception to this rule is made for
3713 Local Labels. See below.
3715 Multibyte characters are supported. To generate a symbol name containing
3716 multibyte characters enclose it within double quotes and use escape codes. cf
3717 @xref{Strings}. Generating a multibyte symbol name from a label is not
3718 currently supported.
3720 Each symbol has exactly one name. Each name in an assembly language program
3721 refers to exactly one symbol. You may use that symbol name any number of times
3724 @subheading Local Symbol Names
3726 @cindex local symbol names
3727 @cindex symbol names, local
3728 A local symbol is any symbol beginning with certain local label prefixes.
3729 By default, the local label prefix is @samp{.L} for ELF systems or
3730 @samp{L} for traditional a.out systems, but each target may have its own
3731 set of local label prefixes.
3733 On the HPPA local symbols begin with @samp{L$}.
3736 Local symbols are defined and used within the assembler, but they are
3737 normally not saved in object files. Thus, they are not visible when debugging.
3738 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols})
3739 to retain the local symbols in the object files.
3741 @subheading Local Labels
3743 @cindex local labels
3744 @cindex temporary symbol names
3745 @cindex symbol names, temporary
3746 Local labels are different from local symbols. Local labels help compilers and
3747 programmers use names temporarily. They create symbols which are guaranteed to
3748 be unique over the entire scope of the input source code and which can be
3749 referred to by a simple notation. To define a local label, write a label of
3750 the form @samp{@b{N}:} (where @b{N} represents any non-negative integer).
3751 To refer to the most recent previous definition of that label write
3752 @samp{@b{N}b}, using the same number as when you defined the label. To refer
3753 to the next definition of a local label, write @samp{@b{N}f}. The @samp{b}
3754 stands for ``backwards'' and the @samp{f} stands for ``forwards''.
3756 There is no restriction on how you can use these labels, and you can reuse them
3757 too. So that it is possible to repeatedly define the same local label (using
3758 the same number @samp{@b{N}}), although you can only refer to the most recently
3759 defined local label of that number (for a backwards reference) or the next
3760 definition of a specific local label for a forward reference. It is also worth
3761 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3762 implemented in a slightly more efficient manner than the others.
3773 Which is the equivalent of:
3776 label_1: branch label_3
3777 label_2: branch label_1
3778 label_3: branch label_4
3779 label_4: branch label_3
3782 Local label names are only a notational device. They are immediately
3783 transformed into more conventional symbol names before the assembler uses them.
3784 The symbol names are stored in the symbol table, appear in error messages, and
3785 are optionally emitted to the object file. The names are constructed using
3789 @item @emph{local label prefix}
3790 All local symbols begin with the system-specific local label prefix.
3791 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3792 that start with the local label prefix. These labels are
3793 used for symbols you are never intended to see. If you use the
3794 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3795 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3796 you may use them in debugging.
3799 This is the number that was used in the local label definition. So if the
3800 label is written @samp{55:} then the number is @samp{55}.
3803 This unusual character is included so you do not accidentally invent a symbol
3804 of the same name. The character has ASCII value of @samp{\002} (control-B).
3806 @item @emph{ordinal number}
3807 This is a serial number to keep the labels distinct. The first definition of
3808 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3809 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3810 the number @samp{1} and its 15th definition gets @samp{15} as well.
3813 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3814 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3816 @subheading Dollar Local Labels
3817 @cindex dollar local symbols
3819 On some targets @code{@value{AS}} also supports an even more local form of
3820 local labels called dollar labels. These labels go out of scope (i.e., they
3821 become undefined) as soon as a non-local label is defined. Thus they remain
3822 valid for only a small region of the input source code. Normal local labels,
3823 by contrast, remain in scope for the entire file, or until they are redefined
3824 by another occurrence of the same local label.
3826 Dollar labels are defined in exactly the same way as ordinary local labels,
3827 except that they have a dollar sign suffix to their numeric value, e.g.,
3830 They can also be distinguished from ordinary local labels by their transformed
3831 names which use ASCII character @samp{\001} (control-A) as the magic character
3832 to distinguish them from ordinary labels. For example, the fifth definition of
3833 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3836 @section The Special Dot Symbol
3838 @cindex dot (symbol)
3839 @cindex @code{.} (symbol)
3840 @cindex current address
3841 @cindex location counter
3842 The special symbol @samp{.} refers to the current address that
3843 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3844 .long .} defines @code{melvin} to contain its own address.
3845 Assigning a value to @code{.} is treated the same as a @code{.org}
3847 @ifclear no-space-dir
3848 Thus, the expression @samp{.=.+4} is the same as saying
3852 @node Symbol Attributes
3853 @section Symbol Attributes
3855 @cindex symbol attributes
3856 @cindex attributes, symbol
3857 Every symbol has, as well as its name, the attributes ``Value'' and
3858 ``Type''. Depending on output format, symbols can also have auxiliary
3861 The detailed definitions are in @file{a.out.h}.
3864 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3865 all these attributes, and probably won't warn you. This makes the
3866 symbol an externally defined symbol, which is generally what you
3870 * Symbol Value:: Value
3871 * Symbol Type:: Type
3874 * a.out Symbols:: Symbol Attributes: @code{a.out}
3878 * a.out Symbols:: Symbol Attributes: @code{a.out}
3881 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3886 * COFF Symbols:: Symbol Attributes for COFF
3889 * SOM Symbols:: Symbol Attributes for SOM
3896 @cindex value of a symbol
3897 @cindex symbol value
3898 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3899 location in the text, data, bss or absolute sections the value is the
3900 number of addresses from the start of that section to the label.
3901 Naturally for text, data and bss sections the value of a symbol changes
3902 as @code{@value{LD}} changes section base addresses during linking. Absolute
3903 symbols' values do not change during linking: that is why they are
3906 The value of an undefined symbol is treated in a special way. If it is
3907 0 then the symbol is not defined in this assembler source file, and
3908 @code{@value{LD}} tries to determine its value from other files linked into the
3909 same program. You make this kind of symbol simply by mentioning a symbol
3910 name without defining it. A non-zero value represents a @code{.comm}
3911 common declaration. The value is how much common storage to reserve, in
3912 bytes (addresses). The symbol refers to the first address of the
3918 @cindex type of a symbol
3920 The type attribute of a symbol contains relocation (section)
3921 information, any flag settings indicating that a symbol is external, and
3922 (optionally), other information for linkers and debuggers. The exact
3923 format depends on the object-code output format in use.
3928 @c The following avoids a "widow" subsection title. @group would be
3929 @c better if it were available outside examples.
3932 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3934 @cindex @code{b.out} symbol attributes
3935 @cindex symbol attributes, @code{b.out}
3936 These symbol attributes appear only when @command{@value{AS}} is configured for
3937 one of the Berkeley-descended object output formats---@code{a.out} or
3943 @subsection Symbol Attributes: @code{a.out}
3945 @cindex @code{a.out} symbol attributes
3946 @cindex symbol attributes, @code{a.out}
3952 @subsection Symbol Attributes: @code{a.out}
3954 @cindex @code{a.out} symbol attributes
3955 @cindex symbol attributes, @code{a.out}
3959 * Symbol Desc:: Descriptor
3960 * Symbol Other:: Other
3964 @subsubsection Descriptor
3966 @cindex descriptor, of @code{a.out} symbol
3967 This is an arbitrary 16-bit value. You may establish a symbol's
3968 descriptor value by using a @code{.desc} statement
3969 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3970 @command{@value{AS}}.
3973 @subsubsection Other
3975 @cindex other attribute, of @code{a.out} symbol
3976 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3981 @subsection Symbol Attributes for COFF
3983 @cindex COFF symbol attributes
3984 @cindex symbol attributes, COFF
3986 The COFF format supports a multitude of auxiliary symbol attributes;
3987 like the primary symbol attributes, they are set between @code{.def} and
3988 @code{.endef} directives.
3990 @subsubsection Primary Attributes
3992 @cindex primary attributes, COFF symbols
3993 The symbol name is set with @code{.def}; the value and type,
3994 respectively, with @code{.val} and @code{.type}.
3996 @subsubsection Auxiliary Attributes
3998 @cindex auxiliary attributes, COFF symbols
3999 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
4000 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
4001 table information for COFF.
4006 @subsection Symbol Attributes for SOM
4008 @cindex SOM symbol attributes
4009 @cindex symbol attributes, SOM
4011 The SOM format for the HPPA supports a multitude of symbol attributes set with
4012 the @code{.EXPORT} and @code{.IMPORT} directives.
4014 The attributes are described in @cite{HP9000 Series 800 Assembly
4015 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
4016 @code{EXPORT} assembler directive documentation.
4020 @chapter Expressions
4024 @cindex numeric values
4025 An @dfn{expression} specifies an address or numeric value.
4026 Whitespace may precede and/or follow an expression.
4028 The result of an expression must be an absolute number, or else an offset into
4029 a particular section. If an expression is not absolute, and there is not
4030 enough information when @command{@value{AS}} sees the expression to know its
4031 section, a second pass over the source program might be necessary to interpret
4032 the expression---but the second pass is currently not implemented.
4033 @command{@value{AS}} aborts with an error message in this situation.
4036 * Empty Exprs:: Empty Expressions
4037 * Integer Exprs:: Integer Expressions
4041 @section Empty Expressions
4043 @cindex empty expressions
4044 @cindex expressions, empty
4045 An empty expression has no value: it is just whitespace or null.
4046 Wherever an absolute expression is required, you may omit the
4047 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
4048 is compatible with other assemblers.
4051 @section Integer Expressions
4053 @cindex integer expressions
4054 @cindex expressions, integer
4055 An @dfn{integer expression} is one or more @emph{arguments} delimited
4056 by @emph{operators}.
4059 * Arguments:: Arguments
4060 * Operators:: Operators
4061 * Prefix Ops:: Prefix Operators
4062 * Infix Ops:: Infix Operators
4066 @subsection Arguments
4068 @cindex expression arguments
4069 @cindex arguments in expressions
4070 @cindex operands in expressions
4071 @cindex arithmetic operands
4072 @dfn{Arguments} are symbols, numbers or subexpressions. In other
4073 contexts arguments are sometimes called ``arithmetic operands''. In
4074 this manual, to avoid confusing them with the ``instruction operands'' of
4075 the machine language, we use the term ``argument'' to refer to parts of
4076 expressions only, reserving the word ``operand'' to refer only to machine
4077 instruction operands.
4079 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
4080 @var{section} is one of text, data, bss, absolute,
4081 or undefined. @var{NNN} is a signed, 2's complement 32 bit
4084 Numbers are usually integers.
4086 A number can be a flonum or bignum. In this case, you are warned
4087 that only the low order 32 bits are used, and @command{@value{AS}} pretends
4088 these 32 bits are an integer. You may write integer-manipulating
4089 instructions that act on exotic constants, compatible with other
4092 @cindex subexpressions
4093 Subexpressions are a left parenthesis @samp{(} followed by an integer
4094 expression, followed by a right parenthesis @samp{)}; or a prefix
4095 operator followed by an argument.
4098 @subsection Operators
4100 @cindex operators, in expressions
4101 @cindex arithmetic functions
4102 @cindex functions, in expressions
4103 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
4104 operators are followed by an argument. Infix operators appear
4105 between their arguments. Operators may be preceded and/or followed by
4109 @subsection Prefix Operator
4111 @cindex prefix operators
4112 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
4113 one argument, which must be absolute.
4115 @c the tex/end tex stuff surrounding this small table is meant to make
4116 @c it align, on the printed page, with the similar table in the next
4117 @c section (which is inside an enumerate).
4119 \global\advance\leftskip by \itemindent
4124 @dfn{Negation}. Two's complement negation.
4126 @dfn{Complementation}. Bitwise not.
4130 \global\advance\leftskip by -\itemindent
4134 @subsection Infix Operators
4136 @cindex infix operators
4137 @cindex operators, permitted arguments
4138 @dfn{Infix operators} take two arguments, one on either side. Operators
4139 have precedence, but operations with equal precedence are performed left
4140 to right. Apart from @code{+} or @option{-}, both arguments must be
4141 absolute, and the result is absolute.
4144 @cindex operator precedence
4145 @cindex precedence of operators
4152 @dfn{Multiplication}.
4155 @dfn{Division}. Truncation is the same as the C operator @samp{/}
4161 @dfn{Shift Left}. Same as the C operator @samp{<<}.
4164 @dfn{Shift Right}. Same as the C operator @samp{>>}.
4168 Intermediate precedence
4173 @dfn{Bitwise Inclusive Or}.
4179 @dfn{Bitwise Exclusive Or}.
4182 @dfn{Bitwise Or Not}.
4189 @cindex addition, permitted arguments
4190 @cindex plus, permitted arguments
4191 @cindex arguments for addition
4193 @dfn{Addition}. If either argument is absolute, the result has the section of
4194 the other argument. You may not add together arguments from different
4197 @cindex subtraction, permitted arguments
4198 @cindex minus, permitted arguments
4199 @cindex arguments for subtraction
4201 @dfn{Subtraction}. If the right argument is absolute, the
4202 result has the section of the left argument.
4203 If both arguments are in the same section, the result is absolute.
4204 You may not subtract arguments from different sections.
4205 @c FIXME is there still something useful to say about undefined - undefined ?
4207 @cindex comparison expressions
4208 @cindex expressions, comparison
4213 @dfn{Is Not Equal To}
4217 @dfn{Is Greater Than}
4219 @dfn{Is Greater Than Or Equal To}
4221 @dfn{Is Less Than Or Equal To}
4223 The comparison operators can be used as infix operators. A true results has a
4224 value of -1 whereas a false result has a value of 0. Note, these operators
4225 perform signed comparisons.
4228 @item Lowest Precedence
4237 These two logical operations can be used to combine the results of sub
4238 expressions. Note, unlike the comparison operators a true result returns a
4239 value of 1 but a false results does still return 0. Also note that the logical
4240 or operator has a slightly lower precedence than logical and.
4245 In short, it's only meaningful to add or subtract the @emph{offsets} in an
4246 address; you can only have a defined section in one of the two arguments.
4249 @chapter Assembler Directives
4251 @cindex directives, machine independent
4252 @cindex pseudo-ops, machine independent
4253 @cindex machine independent directives
4254 All assembler directives have names that begin with a period (@samp{.}).
4255 The names are case insensitive for most targets, and usually written
4258 This chapter discusses directives that are available regardless of the
4259 target machine configuration for the @sc{gnu} assembler.
4261 Some machine configurations provide additional directives.
4262 @xref{Machine Dependencies}.
4265 @ifset machine-directives
4266 @xref{Machine Dependencies}, for additional directives.
4271 * Abort:: @code{.abort}
4273 * ABORT (COFF):: @code{.ABORT}
4276 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
4277 * Altmacro:: @code{.altmacro}
4278 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4279 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4280 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
4281 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, etc
4282 * Byte:: @code{.byte @var{expressions}}
4283 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4284 * Comm:: @code{.comm @var{symbol} , @var{length} }
4285 * Data:: @code{.data @var{subsection}}
4287 * Def:: @code{.def @var{name}}
4290 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4296 * Double:: @code{.double @var{flonums}}
4297 * Eject:: @code{.eject}
4298 * Else:: @code{.else}
4299 * Elseif:: @code{.elseif}
4302 * Endef:: @code{.endef}
4305 * Endfunc:: @code{.endfunc}
4306 * Endif:: @code{.endif}
4307 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4308 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4309 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4311 * Error:: @code{.error @var{string}}
4312 * Exitm:: @code{.exitm}
4313 * Extern:: @code{.extern}
4314 * Fail:: @code{.fail}
4315 * File:: @code{.file}
4316 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4317 * Float:: @code{.float @var{flonums}}
4318 * Func:: @code{.func}
4319 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4321 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4322 * Hidden:: @code{.hidden @var{names}}
4325 * hword:: @code{.hword @var{expressions}}
4326 * Ident:: @code{.ident}
4327 * If:: @code{.if @var{absolute expression}}
4328 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4329 * Include:: @code{.include "@var{file}"}
4330 * Int:: @code{.int @var{expressions}}
4332 * Internal:: @code{.internal @var{names}}
4335 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4336 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4337 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4338 * Lflags:: @code{.lflags}
4339 @ifclear no-line-dir
4340 * Line:: @code{.line @var{line-number}}
4343 * Linkonce:: @code{.linkonce [@var{type}]}
4344 * List:: @code{.list}
4345 * Ln:: @code{.ln @var{line-number}}
4346 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4347 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4349 * Local:: @code{.local @var{names}}
4352 * Long:: @code{.long @var{expressions}}
4354 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4357 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4358 * MRI:: @code{.mri @var{val}}
4359 * Noaltmacro:: @code{.noaltmacro}
4360 * Nolist:: @code{.nolist}
4361 * Octa:: @code{.octa @var{bignums}}
4362 * Offset:: @code{.offset @var{loc}}
4363 * Org:: @code{.org @var{new-lc}, @var{fill}}
4364 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4366 * PopSection:: @code{.popsection}
4367 * Previous:: @code{.previous}
4370 * Print:: @code{.print @var{string}}
4372 * Protected:: @code{.protected @var{names}}
4375 * Psize:: @code{.psize @var{lines}, @var{columns}}
4376 * Purgem:: @code{.purgem @var{name}}
4378 * PushSection:: @code{.pushsection @var{name}}
4381 * Quad:: @code{.quad @var{bignums}}
4382 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4383 * Rept:: @code{.rept @var{count}}
4384 * Sbttl:: @code{.sbttl "@var{subheading}"}
4386 * Scl:: @code{.scl @var{class}}
4389 * Section:: @code{.section @var{name}[, @var{flags}]}
4392 * Set:: @code{.set @var{symbol}, @var{expression}}
4393 * Short:: @code{.short @var{expressions}}
4394 * Single:: @code{.single @var{flonums}}
4396 * Size:: @code{.size [@var{name} , @var{expression}]}
4398 @ifclear no-space-dir
4399 * Skip:: @code{.skip @var{size} , @var{fill}}
4402 * Sleb128:: @code{.sleb128 @var{expressions}}
4403 @ifclear no-space-dir
4404 * Space:: @code{.space @var{size} , @var{fill}}
4407 * Stab:: @code{.stabd, .stabn, .stabs}
4410 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4411 * Struct:: @code{.struct @var{expression}}
4413 * SubSection:: @code{.subsection}
4414 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4418 * Tag:: @code{.tag @var{structname}}
4421 * Text:: @code{.text @var{subsection}}
4422 * Title:: @code{.title "@var{heading}"}
4424 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4427 * Uleb128:: @code{.uleb128 @var{expressions}}
4429 * Val:: @code{.val @var{addr}}
4433 * Version:: @code{.version "@var{string}"}
4434 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4435 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4438 * Warning:: @code{.warning @var{string}}
4439 * Weak:: @code{.weak @var{names}}
4440 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4441 * Word:: @code{.word @var{expressions}}
4442 @ifclear no-space-dir
4443 * Zero:: @code{.zero @var{size}}
4445 * Deprecated:: Deprecated Directives
4449 @section @code{.abort}
4451 @cindex @code{abort} directive
4452 @cindex stopping the assembly
4453 This directive stops the assembly immediately. It is for
4454 compatibility with other assemblers. The original idea was that the
4455 assembly language source would be piped into the assembler. If the sender
4456 of the source quit, it could use this directive tells @command{@value{AS}} to
4457 quit also. One day @code{.abort} will not be supported.
4461 @section @code{.ABORT} (COFF)
4463 @cindex @code{ABORT} directive
4464 When producing COFF output, @command{@value{AS}} accepts this directive as a
4465 synonym for @samp{.abort}.
4468 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
4474 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4476 @cindex padding the location counter
4477 @cindex @code{align} directive
4478 Pad the location counter (in the current subsection) to a particular storage
4479 boundary. The first expression (which must be absolute) is the alignment
4480 required, as described below.
4482 The second expression (also absolute) gives the fill value to be stored in the
4483 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4484 padding bytes are normally zero. However, on some systems, if the section is
4485 marked as containing code and the fill value is omitted, the space is filled
4486 with no-op instructions.
4488 The third expression is also absolute, and is also optional. If it is present,
4489 it is the maximum number of bytes that should be skipped by this alignment
4490 directive. If doing the alignment would require skipping more bytes than the
4491 specified maximum, then the alignment is not done at all. You can omit the
4492 fill value (the second argument) entirely by simply using two commas after the
4493 required alignment; this can be useful if you want the alignment to be filled
4494 with no-op instructions when appropriate.
4496 The way the required alignment is specified varies from system to system.
4497 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or1k,
4498 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4499 alignment request in bytes. For example @samp{.align 8} advances
4500 the location counter until it is a multiple of 8. If the location counter
4501 is already a multiple of 8, no change is needed. For the tic54x, the
4502 first expression is the alignment request in words.
4504 For other systems, including ppc, i386 using a.out format, arm and
4505 strongarm, it is the
4506 number of low-order zero bits the location counter must have after
4507 advancement. For example @samp{.align 3} advances the location
4508 counter until it a multiple of 8. If the location counter is already a
4509 multiple of 8, no change is needed.
4511 This inconsistency is due to the different behaviors of the various
4512 native assemblers for these systems which GAS must emulate.
4513 GAS also provides @code{.balign} and @code{.p2align} directives,
4514 described later, which have a consistent behavior across all
4515 architectures (but are specific to GAS).
4518 @section @code{.altmacro}
4519 Enable alternate macro mode, enabling:
4522 @item LOCAL @var{name} [ , @dots{} ]
4523 One additional directive, @code{LOCAL}, is available. It is used to
4524 generate a string replacement for each of the @var{name} arguments, and
4525 replace any instances of @var{name} in each macro expansion. The
4526 replacement string is unique in the assembly, and different for each
4527 separate macro expansion. @code{LOCAL} allows you to write macros that
4528 define symbols, without fear of conflict between separate macro expansions.
4530 @item String delimiters
4531 You can write strings delimited in these other ways besides
4532 @code{"@var{string}"}:
4535 @item '@var{string}'
4536 You can delimit strings with single-quote characters.
4538 @item <@var{string}>
4539 You can delimit strings with matching angle brackets.
4542 @item single-character string escape
4543 To include any single character literally in a string (even if the
4544 character would otherwise have some special meaning), you can prefix the
4545 character with @samp{!} (an exclamation mark). For example, you can
4546 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4548 @item Expression results as strings
4549 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4550 and use the result as a string.
4554 @section @code{.ascii "@var{string}"}@dots{}
4556 @cindex @code{ascii} directive
4557 @cindex string literals
4558 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4559 separated by commas. It assembles each string (with no automatic
4560 trailing zero byte) into consecutive addresses.
4563 @section @code{.asciz "@var{string}"}@dots{}
4565 @cindex @code{asciz} directive
4566 @cindex zero-terminated strings
4567 @cindex null-terminated strings
4568 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4569 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4572 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4574 @cindex padding the location counter given number of bytes
4575 @cindex @code{balign} directive
4576 Pad the location counter (in the current subsection) to a particular
4577 storage boundary. The first expression (which must be absolute) is the
4578 alignment request in bytes. For example @samp{.balign 8} advances
4579 the location counter until it is a multiple of 8. If the location counter
4580 is already a multiple of 8, no change is needed.
4582 The second expression (also absolute) gives the fill value to be stored in the
4583 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4584 padding bytes are normally zero. However, on some systems, if the section is
4585 marked as containing code and the fill value is omitted, the space is filled
4586 with no-op instructions.
4588 The third expression is also absolute, and is also optional. If it is present,
4589 it is the maximum number of bytes that should be skipped by this alignment
4590 directive. If doing the alignment would require skipping more bytes than the
4591 specified maximum, then the alignment is not done at all. You can omit the
4592 fill value (the second argument) entirely by simply using two commas after the
4593 required alignment; this can be useful if you want the alignment to be filled
4594 with no-op instructions when appropriate.
4596 @cindex @code{balignw} directive
4597 @cindex @code{balignl} directive
4598 The @code{.balignw} and @code{.balignl} directives are variants of the
4599 @code{.balign} directive. The @code{.balignw} directive treats the fill
4600 pattern as a two byte word value. The @code{.balignl} directives treats the
4601 fill pattern as a four byte longword value. For example, @code{.balignw
4602 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4603 filled in with the value 0x368d (the exact placement of the bytes depends upon
4604 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4607 @node Bundle directives
4608 @section Bundle directives
4609 @subsection @code{.bundle_align_mode @var{abs-expr}}
4610 @cindex @code{bundle_align_mode} directive
4612 @cindex instruction bundle
4613 @cindex aligned instruction bundle
4614 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4615 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4616 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4617 disabled (which is the default state). If the argument it not zero, it
4618 gives the size of an instruction bundle as a power of two (as for the
4619 @code{.p2align} directive, @pxref{P2align}).
4621 For some targets, it's an ABI requirement that no instruction may span a
4622 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4623 instructions that starts on an aligned boundary. For example, if
4624 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4625 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4626 effect, no single instruction may span a boundary between bundles. If an
4627 instruction would start too close to the end of a bundle for the length of
4628 that particular instruction to fit within the bundle, then the space at the
4629 end of that bundle is filled with no-op instructions so the instruction
4630 starts in the next bundle. As a corollary, it's an error if any single
4631 instruction's encoding is longer than the bundle size.
4633 @subsection @code{.bundle_lock} and @code{.bundle_unlock}
4634 @cindex @code{bundle_lock} directive
4635 @cindex @code{bundle_unlock} directive
4636 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4637 allow explicit control over instruction bundle padding. These directives
4638 are only valid when @code{.bundle_align_mode} has been used to enable
4639 aligned instruction bundle mode. It's an error if they appear when
4640 @code{.bundle_align_mode} has not been used at all, or when the last
4641 directive was @w{@code{.bundle_align_mode 0}}.
4643 @cindex bundle-locked
4644 For some targets, it's an ABI requirement that certain instructions may
4645 appear only as part of specified permissible sequences of multiple
4646 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4647 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4648 instruction sequence. For purposes of aligned instruction bundle mode, a
4649 sequence starting with @code{.bundle_lock} and ending with
4650 @code{.bundle_unlock} is treated as a single instruction. That is, the
4651 entire sequence must fit into a single bundle and may not span a bundle
4652 boundary. If necessary, no-op instructions will be inserted before the
4653 first instruction of the sequence so that the whole sequence starts on an
4654 aligned bundle boundary. It's an error if the sequence is longer than the
4657 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4658 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4659 nested. That is, a second @code{.bundle_lock} directive before the next
4660 @code{.bundle_unlock} directive has no effect except that it must be
4661 matched by another closing @code{.bundle_unlock} so that there is the
4662 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4665 @section @code{.byte @var{expressions}}
4667 @cindex @code{byte} directive
4668 @cindex integers, one byte
4669 @code{.byte} expects zero or more expressions, separated by commas.
4670 Each expression is assembled into the next byte.
4672 @node CFI directives
4673 @section CFI directives
4674 @subsection @code{.cfi_sections @var{section_list}}
4675 @cindex @code{cfi_sections} directive
4676 @code{.cfi_sections} may be used to specify whether CFI directives
4677 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4678 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4679 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4680 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4681 directive is not used is @code{.cfi_sections .eh_frame}.
4683 On targets that support compact unwinding tables these can be generated
4684 by specifying @code{.eh_frame_entry} instead of @code{.eh_frame}.
4686 Some targets may support an additional name, such as @code{.c6xabi.exidx}
4687 which is used by the @value{TIC6X} target.
4689 The @code{.cfi_sections} directive can be repeated, with the same or different
4690 arguments, provided that CFI generation has not yet started. Once CFI
4691 generation has started however the section list is fixed and any attempts to
4692 redefine it will result in an error.
4694 @subsection @code{.cfi_startproc [simple]}
4695 @cindex @code{cfi_startproc} directive
4696 @code{.cfi_startproc} is used at the beginning of each function that
4697 should have an entry in @code{.eh_frame}. It initializes some internal
4698 data structures. Don't forget to close the function by
4699 @code{.cfi_endproc}.
4701 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4702 it also emits some architecture dependent initial CFI instructions.
4704 @subsection @code{.cfi_endproc}
4705 @cindex @code{cfi_endproc} directive
4706 @code{.cfi_endproc} is used at the end of a function where it closes its
4707 unwind entry previously opened by
4708 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4710 @subsection @code{.cfi_personality @var{encoding} [, @var{exp}]}
4711 @cindex @code{cfi_personality} directive
4712 @code{.cfi_personality} defines personality routine and its encoding.
4713 @var{encoding} must be a constant determining how the personality
4714 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4715 argument is not present, otherwise second argument should be
4716 a constant or a symbol name. When using indirect encodings,
4717 the symbol provided should be the location where personality
4718 can be loaded from, not the personality routine itself.
4719 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4720 no personality routine.
4722 @subsection @code{.cfi_personality_id @var{id}}
4723 @cindex @code{cfi_personality_id} directive
4724 @code{cfi_personality_id} defines a personality routine by its index as
4725 defined in a compact unwinding format.
4726 Only valid when generating compact EH frames (i.e.
4727 with @code{.cfi_sections eh_frame_entry}.
4729 @subsection @code{.cfi_fde_data [@var{opcode1} [, @dots{}]]}
4730 @cindex @code{cfi_fde_data} directive
4731 @code{cfi_fde_data} is used to describe the compact unwind opcodes to be
4732 used for the current function. These are emitted inline in the
4733 @code{.eh_frame_entry} section if small enough and there is no LSDA, or
4734 in the @code{.gnu.extab} section otherwise.
4735 Only valid when generating compact EH frames (i.e.
4736 with @code{.cfi_sections eh_frame_entry}.
4738 @subsection @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4739 @code{.cfi_lsda} defines LSDA and its encoding.
4740 @var{encoding} must be a constant determining how the LSDA
4741 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), the second
4742 argument is not present, otherwise the second argument should be a constant
4743 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4744 meaning that no LSDA is present.
4746 @subsection @code{.cfi_inline_lsda} [@var{align}]
4747 @code{.cfi_inline_lsda} marks the start of a LSDA data section and
4748 switches to the corresponding @code{.gnu.extab} section.
4749 Must be preceded by a CFI block containing a @code{.cfi_lsda} directive.
4750 Only valid when generating compact EH frames (i.e.
4751 with @code{.cfi_sections eh_frame_entry}.
4753 The table header and unwinding opcodes will be generated at this point,
4754 so that they are immediately followed by the LSDA data. The symbol
4755 referenced by the @code{.cfi_lsda} directive should still be defined
4756 in case a fallback FDE based encoding is used. The LSDA data is terminated
4757 by a section directive.
4759 The optional @var{align} argument specifies the alignment required.
4760 The alignment is specified as a power of two, as with the
4761 @code{.p2align} directive.
4763 @subsection @code{.cfi_def_cfa @var{register}, @var{offset}}
4764 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4765 address from @var{register} and add @var{offset} to it}.
4767 @subsection @code{.cfi_def_cfa_register @var{register}}
4768 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4769 now on @var{register} will be used instead of the old one. Offset
4772 @subsection @code{.cfi_def_cfa_offset @var{offset}}
4773 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4774 remains the same, but @var{offset} is new. Note that it is the
4775 absolute offset that will be added to a defined register to compute
4778 @subsection @code{.cfi_adjust_cfa_offset @var{offset}}
4779 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4780 value that is added/substracted from the previous offset.
4782 @subsection @code{.cfi_offset @var{register}, @var{offset}}
4783 Previous value of @var{register} is saved at offset @var{offset} from
4786 @subsection @code{.cfi_rel_offset @var{register}, @var{offset}}
4787 Previous value of @var{register} is saved at offset @var{offset} from
4788 the current CFA register. This is transformed to @code{.cfi_offset}
4789 using the known displacement of the CFA register from the CFA.
4790 This is often easier to use, because the number will match the
4791 code it's annotating.
4793 @subsection @code{.cfi_register @var{register1}, @var{register2}}
4794 Previous value of @var{register1} is saved in register @var{register2}.
4796 @subsection @code{.cfi_restore @var{register}}
4797 @code{.cfi_restore} says that the rule for @var{register} is now the
4798 same as it was at the beginning of the function, after all initial
4799 instruction added by @code{.cfi_startproc} were executed.
4801 @subsection @code{.cfi_undefined @var{register}}
4802 From now on the previous value of @var{register} can't be restored anymore.
4804 @subsection @code{.cfi_same_value @var{register}}
4805 Current value of @var{register} is the same like in the previous frame,
4806 i.e. no restoration needed.
4808 @subsection @code{.cfi_remember_state},
4809 First save all current rules for all registers by @code{.cfi_remember_state},
4810 then totally screw them up by subsequent @code{.cfi_*} directives and when
4811 everything is hopelessly bad, use @code{.cfi_restore_state} to restore
4812 the previous saved state.
4814 @subsection @code{.cfi_return_column @var{register}}
4815 Change return column @var{register}, i.e. the return address is either
4816 directly in @var{register} or can be accessed by rules for @var{register}.
4818 @subsection @code{.cfi_signal_frame}
4819 Mark current function as signal trampoline.
4821 @subsection @code{.cfi_window_save}
4822 SPARC register window has been saved.
4824 @subsection @code{.cfi_escape} @var{expression}[, @dots{}]
4825 Allows the user to add arbitrary bytes to the unwind info. One
4826 might use this to add OS-specific CFI opcodes, or generic CFI
4827 opcodes that GAS does not yet support.
4829 @subsection @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4830 The current value of @var{register} is @var{label}. The value of @var{label}
4831 will be encoded in the output file according to @var{encoding}; see the
4832 description of @code{.cfi_personality} for details on this encoding.
4834 The usefulness of equating a register to a fixed label is probably
4835 limited to the return address register. Here, it can be useful to
4836 mark a code segment that has only one return address which is reached
4837 by a direct branch and no copy of the return address exists in memory
4838 or another register.
4841 @section @code{.comm @var{symbol} , @var{length} }
4843 @cindex @code{comm} directive
4844 @cindex symbol, common
4845 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4846 common symbol in one object file may be merged with a defined or common symbol
4847 of the same name in another object file. If @code{@value{LD}} does not see a
4848 definition for the symbol--just one or more common symbols--then it will
4849 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4850 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4851 the same name, and they do not all have the same size, it will allocate space
4852 using the largest size.
4855 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4856 an optional third argument. This is the desired alignment of the symbol,
4857 specified for ELF as a byte boundary (for example, an alignment of 16 means
4858 that the least significant 4 bits of the address should be zero), and for PE
4859 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4860 boundary). The alignment must be an absolute expression, and it must be a
4861 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4862 common symbol, it will use the alignment when placing the symbol. If no
4863 alignment is specified, @command{@value{AS}} will set the alignment to the
4864 largest power of two less than or equal to the size of the symbol, up to a
4865 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4866 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4867 @samp{--section-alignment} option; image file sections in PE are aligned to
4868 multiples of 4096, which is far too large an alignment for ordinary variables.
4869 It is rather the default alignment for (non-debug) sections within object
4870 (@samp{*.o}) files, which are less strictly aligned.}.
4874 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4875 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4879 @section @code{.data @var{subsection}}
4881 @cindex @code{data} directive
4882 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4883 end of the data subsection numbered @var{subsection} (which is an
4884 absolute expression). If @var{subsection} is omitted, it defaults
4889 @section @code{.def @var{name}}
4891 @cindex @code{def} directive
4892 @cindex COFF symbols, debugging
4893 @cindex debugging COFF symbols
4894 Begin defining debugging information for a symbol @var{name}; the
4895 definition extends until the @code{.endef} directive is encountered.
4898 This directive is only observed when @command{@value{AS}} is configured for COFF
4899 format output; when producing @code{b.out}, @samp{.def} is recognized,
4906 @section @code{.desc @var{symbol}, @var{abs-expression}}
4908 @cindex @code{desc} directive
4909 @cindex COFF symbol descriptor
4910 @cindex symbol descriptor, COFF
4911 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4912 to the low 16 bits of an absolute expression.
4915 The @samp{.desc} directive is not available when @command{@value{AS}} is
4916 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4917 object format. For the sake of compatibility, @command{@value{AS}} accepts
4918 it, but produces no output, when configured for COFF.
4924 @section @code{.dim}
4926 @cindex @code{dim} directive
4927 @cindex COFF auxiliary symbol information
4928 @cindex auxiliary symbol information, COFF
4929 This directive is generated by compilers to include auxiliary debugging
4930 information in the symbol table. It is only permitted inside
4931 @code{.def}/@code{.endef} pairs.
4934 @samp{.dim} is only meaningful when generating COFF format output; when
4935 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4941 @section @code{.double @var{flonums}}
4943 @cindex @code{double} directive
4944 @cindex floating point numbers (double)
4945 @code{.double} expects zero or more flonums, separated by commas. It
4946 assembles floating point numbers.
4948 The exact kind of floating point numbers emitted depends on how
4949 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4953 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4954 in @sc{ieee} format.
4959 @section @code{.eject}
4961 @cindex @code{eject} directive
4962 @cindex new page, in listings
4963 @cindex page, in listings
4964 @cindex listing control: new page
4965 Force a page break at this point, when generating assembly listings.
4968 @section @code{.else}
4970 @cindex @code{else} directive
4971 @code{.else} is part of the @command{@value{AS}} support for conditional
4972 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
4973 of code to be assembled if the condition for the preceding @code{.if}
4977 @section @code{.elseif}
4979 @cindex @code{elseif} directive
4980 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4981 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
4982 @code{.if} block that would otherwise fill the entire @code{.else} section.
4985 @section @code{.end}
4987 @cindex @code{end} directive
4988 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4989 process anything in the file past the @code{.end} directive.
4993 @section @code{.endef}
4995 @cindex @code{endef} directive
4996 This directive flags the end of a symbol definition begun with
5000 @samp{.endef} is only meaningful when generating COFF format output; if
5001 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
5002 directive but ignores it.
5007 @section @code{.endfunc}
5008 @cindex @code{endfunc} directive
5009 @code{.endfunc} marks the end of a function specified with @code{.func}.
5012 @section @code{.endif}
5014 @cindex @code{endif} directive
5015 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
5016 it marks the end of a block of code that is only assembled
5017 conditionally. @xref{If,,@code{.if}}.
5020 @section @code{.equ @var{symbol}, @var{expression}}
5022 @cindex @code{equ} directive
5023 @cindex assigning values to symbols
5024 @cindex symbols, assigning values to
5025 This directive sets the value of @var{symbol} to @var{expression}.
5026 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
5029 The syntax for @code{equ} on the HPPA is
5030 @samp{@var{symbol} .equ @var{expression}}.
5034 The syntax for @code{equ} on the Z80 is
5035 @samp{@var{symbol} equ @var{expression}}.
5036 On the Z80 it is an eror if @var{symbol} is already defined,
5037 but the symbol is not protected from later redefinition.
5038 Compare @ref{Equiv}.
5042 @section @code{.equiv @var{symbol}, @var{expression}}
5043 @cindex @code{equiv} directive
5044 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
5045 the assembler will signal an error if @var{symbol} is already defined. Note a
5046 symbol which has been referenced but not actually defined is considered to be
5049 Except for the contents of the error message, this is roughly equivalent to
5056 plus it protects the symbol from later redefinition.
5059 @section @code{.eqv @var{symbol}, @var{expression}}
5060 @cindex @code{eqv} directive
5061 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
5062 evaluate the expression or any part of it immediately. Instead each time
5063 the resulting symbol is used in an expression, a snapshot of its current
5067 @section @code{.err}
5068 @cindex @code{err} directive
5069 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
5070 message and, unless the @option{-Z} option was used, it will not generate an
5071 object file. This can be used to signal an error in conditionally compiled code.
5074 @section @code{.error "@var{string}"}
5075 @cindex error directive
5077 Similarly to @code{.err}, this directive emits an error, but you can specify a
5078 string that will be emitted as the error message. If you don't specify the
5079 message, it defaults to @code{".error directive invoked in source file"}.
5080 @xref{Errors, ,Error and Warning Messages}.
5083 .error "This code has not been assembled and tested."
5087 @section @code{.exitm}
5088 Exit early from the current macro definition. @xref{Macro}.
5091 @section @code{.extern}
5093 @cindex @code{extern} directive
5094 @code{.extern} is accepted in the source program---for compatibility
5095 with other assemblers---but it is ignored. @command{@value{AS}} treats
5096 all undefined symbols as external.
5099 @section @code{.fail @var{expression}}
5101 @cindex @code{fail} directive
5102 Generates an error or a warning. If the value of the @var{expression} is 500
5103 or more, @command{@value{AS}} will print a warning message. If the value is less
5104 than 500, @command{@value{AS}} will print an error message. The message will
5105 include the value of @var{expression}. This can occasionally be useful inside
5106 complex nested macros or conditional assembly.
5109 @section @code{.file}
5110 @cindex @code{file} directive
5112 @ifclear no-file-dir
5113 There are two different versions of the @code{.file} directive. Targets
5114 that support DWARF2 line number information use the DWARF2 version of
5115 @code{.file}. Other targets use the default version.
5117 @subheading Default Version
5119 @cindex logical file name
5120 @cindex file name, logical
5121 This version of the @code{.file} directive tells @command{@value{AS}} that we
5122 are about to start a new logical file. The syntax is:
5128 @var{string} is the new file name. In general, the filename is
5129 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
5130 to specify an empty file name, you must give the quotes--@code{""}. This
5131 statement may go away in future: it is only recognized to be compatible with
5132 old @command{@value{AS}} programs.
5134 @subheading DWARF2 Version
5137 When emitting DWARF2 line number information, @code{.file} assigns filenames
5138 to the @code{.debug_line} file name table. The syntax is:
5141 .file @var{fileno} @var{filename}
5144 The @var{fileno} operand should be a unique positive integer to use as the
5145 index of the entry in the table. The @var{filename} operand is a C string
5148 The detail of filename indices is exposed to the user because the filename
5149 table is shared with the @code{.debug_info} section of the DWARF2 debugging
5150 information, and thus the user must know the exact indices that table
5154 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
5156 @cindex @code{fill} directive
5157 @cindex writing patterns in memory
5158 @cindex patterns, writing in memory
5159 @var{repeat}, @var{size} and @var{value} are absolute expressions.
5160 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
5161 may be zero or more. @var{Size} may be zero or more, but if it is
5162 more than 8, then it is deemed to have the value 8, compatible with
5163 other people's assemblers. The contents of each @var{repeat} bytes
5164 is taken from an 8-byte number. The highest order 4 bytes are
5165 zero. The lowest order 4 bytes are @var{value} rendered in the
5166 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
5167 Each @var{size} bytes in a repetition is taken from the lowest order
5168 @var{size} bytes of this number. Again, this bizarre behavior is
5169 compatible with other people's assemblers.
5171 @var{size} and @var{value} are optional.
5172 If the second comma and @var{value} are absent, @var{value} is
5173 assumed zero. If the first comma and following tokens are absent,
5174 @var{size} is assumed to be 1.
5177 @section @code{.float @var{flonums}}
5179 @cindex floating point numbers (single)
5180 @cindex @code{float} directive
5181 This directive assembles zero or more flonums, separated by commas. It
5182 has the same effect as @code{.single}.
5184 The exact kind of floating point numbers emitted depends on how
5185 @command{@value{AS}} is configured.
5186 @xref{Machine Dependencies}.
5190 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
5191 in @sc{ieee} format.
5196 @section @code{.func @var{name}[,@var{label}]}
5197 @cindex @code{func} directive
5198 @code{.func} emits debugging information to denote function @var{name}, and
5199 is ignored unless the file is assembled with debugging enabled.
5200 Only @samp{--gstabs[+]} is currently supported.
5201 @var{label} is the entry point of the function and if omitted @var{name}
5202 prepended with the @samp{leading char} is used.
5203 @samp{leading char} is usually @code{_} or nothing, depending on the target.
5204 All functions are currently defined to have @code{void} return type.
5205 The function must be terminated with @code{.endfunc}.
5208 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
5210 @cindex @code{global} directive
5211 @cindex symbol, making visible to linker
5212 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
5213 @var{symbol} in your partial program, its value is made available to
5214 other partial programs that are linked with it. Otherwise,
5215 @var{symbol} takes its attributes from a symbol of the same name
5216 from another file linked into the same program.
5218 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
5219 compatibility with other assemblers.
5222 On the HPPA, @code{.global} is not always enough to make it accessible to other
5223 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
5224 @xref{HPPA Directives, ,HPPA Assembler Directives}.
5229 @section @code{.gnu_attribute @var{tag},@var{value}}
5230 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
5233 @section @code{.hidden @var{names}}
5235 @cindex @code{hidden} directive
5237 This is one of the ELF visibility directives. The other two are
5238 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
5239 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5241 This directive overrides the named symbols default visibility (which is set by
5242 their binding: local, global or weak). The directive sets the visibility to
5243 @code{hidden} which means that the symbols are not visible to other components.
5244 Such symbols are always considered to be @code{protected} as well.
5248 @section @code{.hword @var{expressions}}
5250 @cindex @code{hword} directive
5251 @cindex integers, 16-bit
5252 @cindex numbers, 16-bit
5253 @cindex sixteen bit integers
5254 This expects zero or more @var{expressions}, and emits
5255 a 16 bit number for each.
5258 This directive is a synonym for @samp{.short}; depending on the target
5259 architecture, it may also be a synonym for @samp{.word}.
5263 This directive is a synonym for @samp{.short}.
5266 This directive is a synonym for both @samp{.short} and @samp{.word}.
5271 @section @code{.ident}
5273 @cindex @code{ident} directive
5275 This directive is used by some assemblers to place tags in object files. The
5276 behavior of this directive varies depending on the target. When using the
5277 a.out object file format, @command{@value{AS}} simply accepts the directive for
5278 source-file compatibility with existing assemblers, but does not emit anything
5279 for it. When using COFF, comments are emitted to the @code{.comment} or
5280 @code{.rdata} section, depending on the target. When using ELF, comments are
5281 emitted to the @code{.comment} section.
5284 @section @code{.if @var{absolute expression}}
5286 @cindex conditional assembly
5287 @cindex @code{if} directive
5288 @code{.if} marks the beginning of a section of code which is only
5289 considered part of the source program being assembled if the argument
5290 (which must be an @var{absolute expression}) is non-zero. The end of
5291 the conditional section of code must be marked by @code{.endif}
5292 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
5293 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
5294 If you have several conditions to check, @code{.elseif} may be used to avoid
5295 nesting blocks if/else within each subsequent @code{.else} block.
5297 The following variants of @code{.if} are also supported:
5299 @cindex @code{ifdef} directive
5300 @item .ifdef @var{symbol}
5301 Assembles the following section of code if the specified @var{symbol}
5302 has been defined. Note a symbol which has been referenced but not yet defined
5303 is considered to be undefined.
5305 @cindex @code{ifb} directive
5306 @item .ifb @var{text}
5307 Assembles the following section of code if the operand is blank (empty).
5309 @cindex @code{ifc} directive
5310 @item .ifc @var{string1},@var{string2}
5311 Assembles the following section of code if the two strings are the same. The
5312 strings may be optionally quoted with single quotes. If they are not quoted,
5313 the first string stops at the first comma, and the second string stops at the
5314 end of the line. Strings which contain whitespace should be quoted. The
5315 string comparison is case sensitive.
5317 @cindex @code{ifeq} directive
5318 @item .ifeq @var{absolute expression}
5319 Assembles the following section of code if the argument is zero.
5321 @cindex @code{ifeqs} directive
5322 @item .ifeqs @var{string1},@var{string2}
5323 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5325 @cindex @code{ifge} directive
5326 @item .ifge @var{absolute expression}
5327 Assembles the following section of code if the argument is greater than or
5330 @cindex @code{ifgt} directive
5331 @item .ifgt @var{absolute expression}
5332 Assembles the following section of code if the argument is greater than zero.
5334 @cindex @code{ifle} directive
5335 @item .ifle @var{absolute expression}
5336 Assembles the following section of code if the argument is less than or equal
5339 @cindex @code{iflt} directive
5340 @item .iflt @var{absolute expression}
5341 Assembles the following section of code if the argument is less than zero.
5343 @cindex @code{ifnb} directive
5344 @item .ifnb @var{text}
5345 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5346 following section of code if the operand is non-blank (non-empty).
5348 @cindex @code{ifnc} directive
5349 @item .ifnc @var{string1},@var{string2}.
5350 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5351 following section of code if the two strings are not the same.
5353 @cindex @code{ifndef} directive
5354 @cindex @code{ifnotdef} directive
5355 @item .ifndef @var{symbol}
5356 @itemx .ifnotdef @var{symbol}
5357 Assembles the following section of code if the specified @var{symbol}
5358 has not been defined. Both spelling variants are equivalent. Note a symbol
5359 which has been referenced but not yet defined is considered to be undefined.
5361 @cindex @code{ifne} directive
5362 @item .ifne @var{absolute expression}
5363 Assembles the following section of code if the argument is not equal to zero
5364 (in other words, this is equivalent to @code{.if}).
5366 @cindex @code{ifnes} directive
5367 @item .ifnes @var{string1},@var{string2}
5368 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5369 following section of code if the two strings are not the same.
5373 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5375 @cindex @code{incbin} directive
5376 @cindex binary files, including
5377 The @code{incbin} directive includes @var{file} verbatim at the current
5378 location. You can control the search paths used with the @samp{-I} command-line
5379 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5382 The @var{skip} argument skips a number of bytes from the start of the
5383 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5384 read. Note that the data is not aligned in any way, so it is the user's
5385 responsibility to make sure that proper alignment is provided both before and
5386 after the @code{incbin} directive.
5389 @section @code{.include "@var{file}"}
5391 @cindex @code{include} directive
5392 @cindex supporting files, including
5393 @cindex files, including
5394 This directive provides a way to include supporting files at specified
5395 points in your source program. The code from @var{file} is assembled as
5396 if it followed the point of the @code{.include}; when the end of the
5397 included file is reached, assembly of the original file continues. You
5398 can control the search paths used with the @samp{-I} command-line option
5399 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5403 @section @code{.int @var{expressions}}
5405 @cindex @code{int} directive
5406 @cindex integers, 32-bit
5407 Expect zero or more @var{expressions}, of any section, separated by commas.
5408 For each expression, emit a number that, at run time, is the value of that
5409 expression. The byte order and bit size of the number depends on what kind
5410 of target the assembly is for.
5414 On most forms of the H8/300, @code{.int} emits 16-bit
5415 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5422 @section @code{.internal @var{names}}
5424 @cindex @code{internal} directive
5426 This is one of the ELF visibility directives. The other two are
5427 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5428 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5430 This directive overrides the named symbols default visibility (which is set by
5431 their binding: local, global or weak). The directive sets the visibility to
5432 @code{internal} which means that the symbols are considered to be @code{hidden}
5433 (i.e., not visible to other components), and that some extra, processor specific
5434 processing must also be performed upon the symbols as well.
5438 @section @code{.irp @var{symbol},@var{values}}@dots{}
5440 @cindex @code{irp} directive
5441 Evaluate a sequence of statements assigning different values to @var{symbol}.
5442 The sequence of statements starts at the @code{.irp} directive, and is
5443 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5444 set to @var{value}, and the sequence of statements is assembled. If no
5445 @var{value} is listed, the sequence of statements is assembled once, with
5446 @var{symbol} set to the null string. To refer to @var{symbol} within the
5447 sequence of statements, use @var{\symbol}.
5449 For example, assembling
5457 is equivalent to assembling
5465 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5468 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5470 @cindex @code{irpc} directive
5471 Evaluate a sequence of statements assigning different values to @var{symbol}.
5472 The sequence of statements starts at the @code{.irpc} directive, and is
5473 terminated by an @code{.endr} directive. For each character in @var{value},
5474 @var{symbol} is set to the character, and the sequence of statements is
5475 assembled. If no @var{value} is listed, the sequence of statements is
5476 assembled once, with @var{symbol} set to the null string. To refer to
5477 @var{symbol} within the sequence of statements, use @var{\symbol}.
5479 For example, assembling
5487 is equivalent to assembling
5495 For some caveats with the spelling of @var{symbol}, see also the discussion
5499 @section @code{.lcomm @var{symbol} , @var{length}}
5501 @cindex @code{lcomm} directive
5502 @cindex local common symbols
5503 @cindex symbols, local common
5504 Reserve @var{length} (an absolute expression) bytes for a local common
5505 denoted by @var{symbol}. The section and value of @var{symbol} are
5506 those of the new local common. The addresses are allocated in the bss
5507 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5508 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5509 not visible to @code{@value{LD}}.
5512 Some targets permit a third argument to be used with @code{.lcomm}. This
5513 argument specifies the desired alignment of the symbol in the bss section.
5517 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5518 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5522 @section @code{.lflags}
5524 @cindex @code{lflags} directive (ignored)
5525 @command{@value{AS}} accepts this directive, for compatibility with other
5526 assemblers, but ignores it.
5528 @ifclear no-line-dir
5530 @section @code{.line @var{line-number}}
5532 @cindex @code{line} directive
5533 @cindex logical line number
5535 Change the logical line number. @var{line-number} must be an absolute
5536 expression. The next line has that logical line number. Therefore any other
5537 statements on the current line (after a statement separator character) are
5538 reported as on logical line number @var{line-number} @minus{} 1. One day
5539 @command{@value{AS}} will no longer support this directive: it is recognized only
5540 for compatibility with existing assembler programs.
5543 Even though this is a directive associated with the @code{a.out} or
5544 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5545 when producing COFF output, and treats @samp{.line} as though it
5546 were the COFF @samp{.ln} @emph{if} it is found outside a
5547 @code{.def}/@code{.endef} pair.
5549 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5550 used by compilers to generate auxiliary symbol information for
5555 @section @code{.linkonce [@var{type}]}
5557 @cindex @code{linkonce} directive
5558 @cindex common sections
5559 Mark the current section so that the linker only includes a single copy of it.
5560 This may be used to include the same section in several different object files,
5561 but ensure that the linker will only include it once in the final output file.
5562 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5563 Duplicate sections are detected based on the section name, so it should be
5566 This directive is only supported by a few object file formats; as of this
5567 writing, the only object file format which supports it is the Portable
5568 Executable format used on Windows NT.
5570 The @var{type} argument is optional. If specified, it must be one of the
5571 following strings. For example:
5575 Not all types may be supported on all object file formats.
5579 Silently discard duplicate sections. This is the default.
5582 Warn if there are duplicate sections, but still keep only one copy.
5585 Warn if any of the duplicates have different sizes.
5588 Warn if any of the duplicates do not have exactly the same contents.
5592 @section @code{.list}
5594 @cindex @code{list} directive
5595 @cindex listing control, turning on
5596 Control (in conjunction with the @code{.nolist} directive) whether or
5597 not assembly listings are generated. These two directives maintain an
5598 internal counter (which is zero initially). @code{.list} increments the
5599 counter, and @code{.nolist} decrements it. Assembly listings are
5600 generated whenever the counter is greater than zero.
5602 By default, listings are disabled. When you enable them (with the
5603 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5604 the initial value of the listing counter is one.
5607 @section @code{.ln @var{line-number}}
5609 @cindex @code{ln} directive
5610 @ifclear no-line-dir
5611 @samp{.ln} is a synonym for @samp{.line}.
5614 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5615 must be an absolute expression. The next line has that logical
5616 line number, so any other statements on the current line (after a
5617 statement separator character @code{;}) are reported as on logical
5618 line number @var{line-number} @minus{} 1.
5621 This directive is accepted, but ignored, when @command{@value{AS}} is
5622 configured for @code{b.out}; its effect is only associated with COFF
5628 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5629 @cindex @code{loc} directive
5630 When emitting DWARF2 line number information,
5631 the @code{.loc} directive will add a row to the @code{.debug_line} line
5632 number matrix corresponding to the immediately following assembly
5633 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5634 arguments will be applied to the @code{.debug_line} state machine before
5637 The @var{options} are a sequence of the following tokens in any order:
5641 This option will set the @code{basic_block} register in the
5642 @code{.debug_line} state machine to @code{true}.
5645 This option will set the @code{prologue_end} register in the
5646 @code{.debug_line} state machine to @code{true}.
5648 @item epilogue_begin
5649 This option will set the @code{epilogue_begin} register in the
5650 @code{.debug_line} state machine to @code{true}.
5652 @item is_stmt @var{value}
5653 This option will set the @code{is_stmt} register in the
5654 @code{.debug_line} state machine to @code{value}, which must be
5657 @item isa @var{value}
5658 This directive will set the @code{isa} register in the @code{.debug_line}
5659 state machine to @var{value}, which must be an unsigned integer.
5661 @item discriminator @var{value}
5662 This directive will set the @code{discriminator} register in the @code{.debug_line}
5663 state machine to @var{value}, which must be an unsigned integer.
5667 @node Loc_mark_labels
5668 @section @code{.loc_mark_labels @var{enable}}
5669 @cindex @code{loc_mark_labels} directive
5670 When emitting DWARF2 line number information,
5671 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5672 to the @code{.debug_line} line number matrix with the @code{basic_block}
5673 register in the state machine set whenever a code label is seen.
5674 The @var{enable} argument should be either 1 or 0, to enable or disable
5675 this function respectively.
5679 @section @code{.local @var{names}}
5681 @cindex @code{local} directive
5682 This directive, which is available for ELF targets, marks each symbol in
5683 the comma-separated list of @code{names} as a local symbol so that it
5684 will not be externally visible. If the symbols do not already exist,
5685 they will be created.
5687 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5688 accept an alignment argument, which is the case for most ELF targets,
5689 the @code{.local} directive can be used in combination with @code{.comm}
5690 (@pxref{Comm}) to define aligned local common data.
5694 @section @code{.long @var{expressions}}
5696 @cindex @code{long} directive
5697 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5700 @c no one seems to know what this is for or whether this description is
5701 @c what it really ought to do
5703 @section @code{.lsym @var{symbol}, @var{expression}}
5705 @cindex @code{lsym} directive
5706 @cindex symbol, not referenced in assembly
5707 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5708 the hash table, ensuring it cannot be referenced by name during the
5709 rest of the assembly. This sets the attributes of the symbol to be
5710 the same as the expression value:
5712 @var{other} = @var{descriptor} = 0
5713 @var{type} = @r{(section of @var{expression})}
5714 @var{value} = @var{expression}
5717 The new symbol is not flagged as external.
5721 @section @code{.macro}
5724 The commands @code{.macro} and @code{.endm} allow you to define macros that
5725 generate assembly output. For example, this definition specifies a macro
5726 @code{sum} that puts a sequence of numbers into memory:
5729 .macro sum from=0, to=5
5738 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5750 @item .macro @var{macname}
5751 @itemx .macro @var{macname} @var{macargs} @dots{}
5752 @cindex @code{macro} directive
5753 Begin the definition of a macro called @var{macname}. If your macro
5754 definition requires arguments, specify their names after the macro name,
5755 separated by commas or spaces. You can qualify the macro argument to
5756 indicate whether all invocations must specify a non-blank value (through
5757 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5758 (through @samp{:@code{vararg}}). You can supply a default value for any
5759 macro argument by following the name with @samp{=@var{deflt}}. You
5760 cannot define two macros with the same @var{macname} unless it has been
5761 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5762 definitions. For example, these are all valid @code{.macro} statements:
5766 Begin the definition of a macro called @code{comm}, which takes no
5769 @item .macro plus1 p, p1
5770 @itemx .macro plus1 p p1
5771 Either statement begins the definition of a macro called @code{plus1},
5772 which takes two arguments; within the macro definition, write
5773 @samp{\p} or @samp{\p1} to evaluate the arguments.
5775 @item .macro reserve_str p1=0 p2
5776 Begin the definition of a macro called @code{reserve_str}, with two
5777 arguments. The first argument has a default value, but not the second.
5778 After the definition is complete, you can call the macro either as
5779 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5780 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5781 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5782 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5784 @item .macro m p1:req, p2=0, p3:vararg
5785 Begin the definition of a macro called @code{m}, with at least three
5786 arguments. The first argument must always have a value specified, but
5787 not the second, which instead has a default value. The third formal
5788 will get assigned all remaining arguments specified at invocation time.
5790 When you call a macro, you can specify the argument values either by
5791 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5792 @samp{sum to=17, from=9}.
5796 Note that since each of the @var{macargs} can be an identifier exactly
5797 as any other one permitted by the target architecture, there may be
5798 occasional problems if the target hand-crafts special meanings to certain
5799 characters when they occur in a special position. For example, if the colon
5800 (@code{:}) is generally permitted to be part of a symbol name, but the
5801 architecture specific code special-cases it when occurring as the final
5802 character of a symbol (to denote a label), then the macro parameter
5803 replacement code will have no way of knowing that and consider the whole
5804 construct (including the colon) an identifier, and check only this
5805 identifier for being the subject to parameter substitution. So for example
5806 this macro definition:
5814 might not work as expected. Invoking @samp{label foo} might not create a label
5815 called @samp{foo} but instead just insert the text @samp{\l:} into the
5816 assembler source, probably generating an error about an unrecognised
5819 Similarly problems might occur with the period character (@samp{.})
5820 which is often allowed inside opcode names (and hence identifier names). So
5821 for example constructing a macro to build an opcode from a base name and a
5822 length specifier like this:
5825 .macro opcode base length
5830 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5831 instruction but instead generate some kind of error as the assembler tries to
5832 interpret the text @samp{\base.\length}.
5834 There are several possible ways around this problem:
5837 @item Insert white space
5838 If it is possible to use white space characters then this is the simplest
5847 @item Use @samp{\()}
5848 The string @samp{\()} can be used to separate the end of a macro argument from
5849 the following text. eg:
5852 .macro opcode base length
5857 @item Use the alternate macro syntax mode
5858 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5859 used as a separator. eg:
5869 Note: this problem of correctly identifying string parameters to pseudo ops
5870 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5871 and @code{.irpc} (@pxref{Irpc}) as well.
5874 @cindex @code{endm} directive
5875 Mark the end of a macro definition.
5878 @cindex @code{exitm} directive
5879 Exit early from the current macro definition.
5881 @cindex number of macros executed
5882 @cindex macros, count executed
5884 @command{@value{AS}} maintains a counter of how many macros it has
5885 executed in this pseudo-variable; you can copy that number to your
5886 output with @samp{\@@}, but @emph{only within a macro definition}.
5888 @item LOCAL @var{name} [ , @dots{} ]
5889 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5890 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5891 @xref{Altmacro,,@code{.altmacro}}.
5895 @section @code{.mri @var{val}}
5897 @cindex @code{mri} directive
5898 @cindex MRI mode, temporarily
5899 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
5900 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
5901 affects code assembled until the next @code{.mri} directive, or until the end
5902 of the file. @xref{M, MRI mode, MRI mode}.
5905 @section @code{.noaltmacro}
5906 Disable alternate macro mode. @xref{Altmacro}.
5909 @section @code{.nolist}
5911 @cindex @code{nolist} directive
5912 @cindex listing control, turning off
5913 Control (in conjunction with the @code{.list} directive) whether or
5914 not assembly listings are generated. These two directives maintain an
5915 internal counter (which is zero initially). @code{.list} increments the
5916 counter, and @code{.nolist} decrements it. Assembly listings are
5917 generated whenever the counter is greater than zero.
5920 @section @code{.octa @var{bignums}}
5922 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5923 @cindex @code{octa} directive
5924 @cindex integer, 16-byte
5925 @cindex sixteen byte integer
5926 This directive expects zero or more bignums, separated by commas. For each
5927 bignum, it emits a 16-byte integer.
5929 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5930 hence @emph{octa}-word for 16 bytes.
5933 @section @code{.offset @var{loc}}
5935 @cindex @code{offset} directive
5936 Set the location counter to @var{loc} in the absolute section. @var{loc} must
5937 be an absolute expression. This directive may be useful for defining
5938 symbols with absolute values. Do not confuse it with the @code{.org}
5942 @section @code{.org @var{new-lc} , @var{fill}}
5944 @cindex @code{org} directive
5945 @cindex location counter, advancing
5946 @cindex advancing location counter
5947 @cindex current address, advancing
5948 Advance the location counter of the current section to
5949 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5950 expression with the same section as the current subsection. That is,
5951 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5952 wrong section, the @code{.org} directive is ignored. To be compatible
5953 with former assemblers, if the section of @var{new-lc} is absolute,
5954 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5955 is the same as the current subsection.
5957 @code{.org} may only increase the location counter, or leave it
5958 unchanged; you cannot use @code{.org} to move the location counter
5961 @c double negative used below "not undefined" because this is a specific
5962 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5963 @c section. doc@cygnus.com 18feb91
5964 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5965 may not be undefined. If you really detest this restriction we eagerly await
5966 a chance to share your improved assembler.
5968 Beware that the origin is relative to the start of the section, not
5969 to the start of the subsection. This is compatible with other
5970 people's assemblers.
5972 When the location counter (of the current subsection) is advanced, the
5973 intervening bytes are filled with @var{fill} which should be an
5974 absolute expression. If the comma and @var{fill} are omitted,
5975 @var{fill} defaults to zero.
5978 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5980 @cindex padding the location counter given a power of two
5981 @cindex @code{p2align} directive
5982 Pad the location counter (in the current subsection) to a particular
5983 storage boundary. The first expression (which must be absolute) is the
5984 number of low-order zero bits the location counter must have after
5985 advancement. For example @samp{.p2align 3} advances the location
5986 counter until it a multiple of 8. If the location counter is already a
5987 multiple of 8, no change is needed.
5989 The second expression (also absolute) gives the fill value to be stored in the
5990 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5991 padding bytes are normally zero. However, on some systems, if the section is
5992 marked as containing code and the fill value is omitted, the space is filled
5993 with no-op instructions.
5995 The third expression is also absolute, and is also optional. If it is present,
5996 it is the maximum number of bytes that should be skipped by this alignment
5997 directive. If doing the alignment would require skipping more bytes than the
5998 specified maximum, then the alignment is not done at all. You can omit the
5999 fill value (the second argument) entirely by simply using two commas after the
6000 required alignment; this can be useful if you want the alignment to be filled
6001 with no-op instructions when appropriate.
6003 @cindex @code{p2alignw} directive
6004 @cindex @code{p2alignl} directive
6005 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
6006 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
6007 pattern as a two byte word value. The @code{.p2alignl} directives treats the
6008 fill pattern as a four byte longword value. For example, @code{.p2alignw
6009 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
6010 filled in with the value 0x368d (the exact placement of the bytes depends upon
6011 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
6016 @section @code{.popsection}
6018 @cindex @code{popsection} directive
6019 @cindex Section Stack
6020 This is one of the ELF section stack manipulation directives. The others are
6021 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6022 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
6025 This directive replaces the current section (and subsection) with the top
6026 section (and subsection) on the section stack. This section is popped off the
6032 @section @code{.previous}
6034 @cindex @code{previous} directive
6035 @cindex Section Stack
6036 This is one of the ELF section stack manipulation directives. The others are
6037 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6038 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
6039 (@pxref{PopSection}).
6041 This directive swaps the current section (and subsection) with most recently
6042 referenced section/subsection pair prior to this one. Multiple
6043 @code{.previous} directives in a row will flip between two sections (and their
6044 subsections). For example:
6056 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
6062 # Now in section A subsection 1
6066 # Now in section B subsection 0
6069 # Now in section B subsection 1
6072 # Now in section B subsection 0
6076 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
6077 section B and 0x9abc into subsection 1 of section B.
6079 In terms of the section stack, this directive swaps the current section with
6080 the top section on the section stack.
6084 @section @code{.print @var{string}}
6086 @cindex @code{print} directive
6087 @command{@value{AS}} will print @var{string} on the standard output during
6088 assembly. You must put @var{string} in double quotes.
6092 @section @code{.protected @var{names}}
6094 @cindex @code{protected} directive
6096 This is one of the ELF visibility directives. The other two are
6097 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
6099 This directive overrides the named symbols default visibility (which is set by
6100 their binding: local, global or weak). The directive sets the visibility to
6101 @code{protected} which means that any references to the symbols from within the
6102 components that defines them must be resolved to the definition in that
6103 component, even if a definition in another component would normally preempt
6108 @section @code{.psize @var{lines} , @var{columns}}
6110 @cindex @code{psize} directive
6111 @cindex listing control: paper size
6112 @cindex paper size, for listings
6113 Use this directive to declare the number of lines---and, optionally, the
6114 number of columns---to use for each page, when generating listings.
6116 If you do not use @code{.psize}, listings use a default line-count
6117 of 60. You may omit the comma and @var{columns} specification; the
6118 default width is 200 columns.
6120 @command{@value{AS}} generates formfeeds whenever the specified number of
6121 lines is exceeded (or whenever you explicitly request one, using
6124 If you specify @var{lines} as @code{0}, no formfeeds are generated save
6125 those explicitly specified with @code{.eject}.
6128 @section @code{.purgem @var{name}}
6130 @cindex @code{purgem} directive
6131 Undefine the macro @var{name}, so that later uses of the string will not be
6132 expanded. @xref{Macro}.
6136 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
6138 @cindex @code{pushsection} directive
6139 @cindex Section Stack
6140 This is one of the ELF section stack manipulation directives. The others are
6141 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6142 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6145 This directive pushes the current section (and subsection) onto the
6146 top of the section stack, and then replaces the current section and
6147 subsection with @code{name} and @code{subsection}. The optional
6148 @code{flags}, @code{type} and @code{arguments} are treated the same
6149 as in the @code{.section} (@pxref{Section}) directive.
6153 @section @code{.quad @var{bignums}}
6155 @cindex @code{quad} directive
6156 @code{.quad} expects zero or more bignums, separated by commas. For
6157 each bignum, it emits
6159 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
6160 warning message; and just takes the lowest order 8 bytes of the bignum.
6161 @cindex eight-byte integer
6162 @cindex integer, 8-byte
6164 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
6165 hence @emph{quad}-word for 8 bytes.
6168 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
6169 warning message; and just takes the lowest order 16 bytes of the bignum.
6170 @cindex sixteen-byte integer
6171 @cindex integer, 16-byte
6175 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
6177 @cindex @code{reloc} directive
6178 Generate a relocation at @var{offset} of type @var{reloc_name} with value
6179 @var{expression}. If @var{offset} is a number, the relocation is generated in
6180 the current section. If @var{offset} is an expression that resolves to a
6181 symbol plus offset, the relocation is generated in the given symbol's section.
6182 @var{expression}, if present, must resolve to a symbol plus addend or to an
6183 absolute value, but note that not all targets support an addend. e.g. ELF REL
6184 targets such as i386 store an addend in the section contents rather than in the
6185 relocation. This low level interface does not support addends stored in the
6189 @section @code{.rept @var{count}}
6191 @cindex @code{rept} directive
6192 Repeat the sequence of lines between the @code{.rept} directive and the next
6193 @code{.endr} directive @var{count} times.
6195 For example, assembling
6203 is equivalent to assembling
6212 @section @code{.sbttl "@var{subheading}"}
6214 @cindex @code{sbttl} directive
6215 @cindex subtitles for listings
6216 @cindex listing control: subtitle
6217 Use @var{subheading} as the title (third line, immediately after the
6218 title line) when generating assembly listings.
6220 This directive affects subsequent pages, as well as the current page if
6221 it appears within ten lines of the top of a page.
6225 @section @code{.scl @var{class}}
6227 @cindex @code{scl} directive
6228 @cindex symbol storage class (COFF)
6229 @cindex COFF symbol storage class
6230 Set the storage-class value for a symbol. This directive may only be
6231 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
6232 whether a symbol is static or external, or it may record further
6233 symbolic debugging information.
6236 The @samp{.scl} directive is primarily associated with COFF output; when
6237 configured to generate @code{b.out} output format, @command{@value{AS}}
6238 accepts this directive but ignores it.
6244 @section @code{.section @var{name}}
6246 @cindex named section
6247 Use the @code{.section} directive to assemble the following code into a section
6250 This directive is only supported for targets that actually support arbitrarily
6251 named sections; on @code{a.out} targets, for example, it is not accepted, even
6252 with a standard @code{a.out} section name.
6256 @c only print the extra heading if both COFF and ELF are set
6257 @subheading COFF Version
6260 @cindex @code{section} directive (COFF version)
6261 For COFF targets, the @code{.section} directive is used in one of the following
6265 .section @var{name}[, "@var{flags}"]
6266 .section @var{name}[, @var{subsection}]
6269 If the optional argument is quoted, it is taken as flags to use for the
6270 section. Each flag is a single character. The following flags are recognized:
6273 bss section (uninitialized data)
6275 section is not loaded
6281 exclude section from linking
6287 shared section (meaningful for PE targets)
6289 ignored. (For compatibility with the ELF version)
6291 section is not readable (meaningful for PE targets)
6293 single-digit power-of-two section alignment (GNU extension)
6296 If no flags are specified, the default flags depend upon the section name. If
6297 the section name is not recognized, the default will be for the section to be
6298 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
6299 from the section, rather than adding them, so if they are used on their own it
6300 will be as if no flags had been specified at all.
6302 If the optional argument to the @code{.section} directive is not quoted, it is
6303 taken as a subsection number (@pxref{Sub-Sections}).
6308 @c only print the extra heading if both COFF and ELF are set
6309 @subheading ELF Version
6312 @cindex Section Stack
6313 This is one of the ELF section stack manipulation directives. The others are
6314 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
6315 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
6316 @code{.previous} (@pxref{Previous}).
6318 @cindex @code{section} directive (ELF version)
6319 For ELF targets, the @code{.section} directive is used like this:
6322 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
6325 @anchor{Section Name Substitutions}
6326 @kindex --sectname-subst
6327 @cindex section name substitution
6328 If the @samp{--sectname-subst} command-line option is provided, the @var{name}
6329 argument may contain a substitution sequence. Only @code{%S} is supported
6330 at the moment, and substitutes the current section name. For example:
6333 .macro exception_code
6334 .section %S.exception
6335 [exception code here]
6350 The two @code{exception_code} invocations above would create the
6351 @code{.text.exception} and @code{.init.exception} sections respectively.
6352 This is useful e.g. to discriminate between anciliary sections that are
6353 tied to setup code to be discarded after use from anciliary sections that
6354 need to stay resident without having to define multiple @code{exception_code}
6355 macros just for that purpose.
6357 The optional @var{flags} argument is a quoted string which may contain any
6358 combination of the following characters:
6361 section is allocatable
6363 section is excluded from executable and shared library.
6367 section is executable
6369 section is mergeable
6371 section contains zero terminated strings
6373 section is a member of a section group
6375 section is used for thread-local-storage
6377 section is a member of the previously-current section's group, if any
6379 a numeric value indicating the bits to be set in the ELF section header's flags
6380 field. Note - if one or more of the alphabetic characters described above is
6381 also included in the flags field, their bit values will be ORed into the
6383 @item @{target specific}
6384 some targets extend this list with their own types
6387 Note - once a section's flags have been set they cannot be changed. There are
6388 a few exceptions to this rule however. Processor and application specific
6389 flags can be added to an already defined section. The @code{.interp},
6390 @code{.strtab} and @code{.symtab} sections can have the allocate flag
6391 (@code{a}) set after they are initially defined, and the @code{.note-GNU-stack}
6392 section may have the executable (@code(x)) flag added.
6394 The optional @var{type} argument may contain one of the following constants:
6397 section contains data
6399 section does not contain data (i.e., section only occupies space)
6401 section contains data which is used by things other than the program
6403 section contains an array of pointers to init functions
6405 section contains an array of pointers to finish functions
6406 @item @@preinit_array
6407 section contains an array of pointers to pre-init functions
6408 @item @@@var{number}
6409 a numeric value to be set as the ELF section header's type field.
6410 @item @@@{target specific}
6411 some targets extend this list with their own types
6414 Many targets only support the first three section types. The type may be
6415 enclosed in double quotes if necessary.
6417 Note on targets where the @code{@@} character is the start of a comment (eg
6418 ARM) then another character is used instead. For example the ARM port uses the
6421 Note - some sections, eg @code{.text} and @code{.data} are considered to be
6422 special and have fixed types. Any attempt to declare them with a different
6423 type will generate an error from the assembler.
6425 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6426 be specified as well as an extra argument---@var{entsize}---like this:
6429 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6432 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6433 constants, each @var{entsize} octets long. Sections with both @code{M} and
6434 @code{S} must contain zero terminated strings where each character is
6435 @var{entsize} bytes long. The linker may remove duplicates within sections with
6436 the same name, same entity size and same flags. @var{entsize} must be an
6437 absolute expression. For sections with both @code{M} and @code{S}, a string
6438 which is a suffix of a larger string is considered a duplicate. Thus
6439 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6440 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6442 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6443 be present along with an additional field like this:
6446 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6449 The @var{GroupName} field specifies the name of the section group to which this
6450 particular section belongs. The optional linkage field can contain:
6453 indicates that only one copy of this section should be retained
6458 Note: if both the @var{M} and @var{G} flags are present then the fields for
6459 the Merge flag should come first, like this:
6462 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6465 If @var{flags} contains the @code{?} symbol then it may not also contain the
6466 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6467 present. Instead, @code{?} says to consider the section that's current before
6468 this directive. If that section used @code{G}, then the new section will use
6469 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6470 If not, then the @code{?} symbol has no effect.
6472 If no flags are specified, the default flags depend upon the section name. If
6473 the section name is not recognized, the default will be for the section to have
6474 none of the above flags: it will not be allocated in memory, nor writable, nor
6475 executable. The section will contain data.
6477 For ELF targets, the assembler supports another type of @code{.section}
6478 directive for compatibility with the Solaris assembler:
6481 .section "@var{name}"[, @var{flags}...]
6484 Note that the section name is quoted. There may be a sequence of comma
6488 section is allocatable
6492 section is executable
6494 section is excluded from executable and shared library.
6496 section is used for thread local storage
6499 This directive replaces the current section and subsection. See the
6500 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6501 some examples of how this directive and the other section stack directives
6507 @section @code{.set @var{symbol}, @var{expression}}
6509 @cindex @code{set} directive
6510 @cindex symbol value, setting
6511 Set the value of @var{symbol} to @var{expression}. This
6512 changes @var{symbol}'s value and type to conform to
6513 @var{expression}. If @var{symbol} was flagged as external, it remains
6514 flagged (@pxref{Symbol Attributes}).
6516 You may @code{.set} a symbol many times in the same assembly provided that the
6517 values given to the symbol are constants. Values that are based on expressions
6518 involving other symbols are allowed, but some targets may restrict this to only
6519 being done once per assembly. This is because those targets do not set the
6520 addresses of symbols at assembly time, but rather delay the assignment until a
6521 final link is performed. This allows the linker a chance to change the code in
6522 the files, changing the location of, and the relative distance between, various
6525 If you @code{.set} a global symbol, the value stored in the object
6526 file is the last value stored into it.
6529 On Z80 @code{set} is a real instruction, use
6530 @samp{@var{symbol} defl @var{expression}} instead.
6534 @section @code{.short @var{expressions}}
6536 @cindex @code{short} directive
6538 @code{.short} is normally the same as @samp{.word}.
6539 @xref{Word,,@code{.word}}.
6541 In some configurations, however, @code{.short} and @code{.word} generate
6542 numbers of different lengths. @xref{Machine Dependencies}.
6546 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6549 This expects zero or more @var{expressions}, and emits
6550 a 16 bit number for each.
6555 @section @code{.single @var{flonums}}
6557 @cindex @code{single} directive
6558 @cindex floating point numbers (single)
6559 This directive assembles zero or more flonums, separated by commas. It
6560 has the same effect as @code{.float}.
6562 The exact kind of floating point numbers emitted depends on how
6563 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6567 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6568 numbers in @sc{ieee} format.
6574 @section @code{.size}
6576 This directive is used to set the size associated with a symbol.
6580 @c only print the extra heading if both COFF and ELF are set
6581 @subheading COFF Version
6584 @cindex @code{size} directive (COFF version)
6585 For COFF targets, the @code{.size} directive is only permitted inside
6586 @code{.def}/@code{.endef} pairs. It is used like this:
6589 .size @var{expression}
6593 @samp{.size} is only meaningful when generating COFF format output; when
6594 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6601 @c only print the extra heading if both COFF and ELF are set
6602 @subheading ELF Version
6605 @cindex @code{size} directive (ELF version)
6606 For ELF targets, the @code{.size} directive is used like this:
6609 .size @var{name} , @var{expression}
6612 This directive sets the size associated with a symbol @var{name}.
6613 The size in bytes is computed from @var{expression} which can make use of label
6614 arithmetic. This directive is typically used to set the size of function
6619 @ifclear no-space-dir
6621 @section @code{.skip @var{size} , @var{fill}}
6623 @cindex @code{skip} directive
6624 @cindex filling memory
6625 This directive emits @var{size} bytes, each of value @var{fill}. Both
6626 @var{size} and @var{fill} are absolute expressions. If the comma and
6627 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6632 @section @code{.sleb128 @var{expressions}}
6634 @cindex @code{sleb128} directive
6635 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6636 compact, variable length representation of numbers used by the DWARF
6637 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6639 @ifclear no-space-dir
6641 @section @code{.space @var{size} , @var{fill}}
6643 @cindex @code{space} directive
6644 @cindex filling memory
6645 This directive emits @var{size} bytes, each of value @var{fill}. Both
6646 @var{size} and @var{fill} are absolute expressions. If the comma
6647 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6652 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6653 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6654 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6655 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6663 @section @code{.stabd, .stabn, .stabs}
6665 @cindex symbolic debuggers, information for
6666 @cindex @code{stab@var{x}} directives
6667 There are three directives that begin @samp{.stab}.
6668 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6669 The symbols are not entered in the @command{@value{AS}} hash table: they
6670 cannot be referenced elsewhere in the source file.
6671 Up to five fields are required:
6675 This is the symbol's name. It may contain any character except
6676 @samp{\000}, so is more general than ordinary symbol names. Some
6677 debuggers used to code arbitrarily complex structures into symbol names
6681 An absolute expression. The symbol's type is set to the low 8 bits of
6682 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6683 and debuggers choke on silly bit patterns.
6686 An absolute expression. The symbol's ``other'' attribute is set to the
6687 low 8 bits of this expression.
6690 An absolute expression. The symbol's descriptor is set to the low 16
6691 bits of this expression.
6694 An absolute expression which becomes the symbol's value.
6697 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6698 or @code{.stabs} statement, the symbol has probably already been created;
6699 you get a half-formed symbol in your object file. This is
6700 compatible with earlier assemblers!
6703 @cindex @code{stabd} directive
6704 @item .stabd @var{type} , @var{other} , @var{desc}
6706 The ``name'' of the symbol generated is not even an empty string.
6707 It is a null pointer, for compatibility. Older assemblers used a
6708 null pointer so they didn't waste space in object files with empty
6711 The symbol's value is set to the location counter,
6712 relocatably. When your program is linked, the value of this symbol
6713 is the address of the location counter when the @code{.stabd} was
6716 @cindex @code{stabn} directive
6717 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6718 The name of the symbol is set to the empty string @code{""}.
6720 @cindex @code{stabs} directive
6721 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6722 All five fields are specified.
6728 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6729 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6731 @cindex string, copying to object file
6732 @cindex string8, copying to object file
6733 @cindex string16, copying to object file
6734 @cindex string32, copying to object file
6735 @cindex string64, copying to object file
6736 @cindex @code{string} directive
6737 @cindex @code{string8} directive
6738 @cindex @code{string16} directive
6739 @cindex @code{string32} directive
6740 @cindex @code{string64} directive
6742 Copy the characters in @var{str} to the object file. You may specify more than
6743 one string to copy, separated by commas. Unless otherwise specified for a
6744 particular machine, the assembler marks the end of each string with a 0 byte.
6745 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6747 The variants @code{string16}, @code{string32} and @code{string64} differ from
6748 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6749 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6750 are stored in target endianness byte order.
6756 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6757 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6762 @section @code{.struct @var{expression}}
6764 @cindex @code{struct} directive
6765 Switch to the absolute section, and set the section offset to @var{expression},
6766 which must be an absolute expression. You might use this as follows:
6775 This would define the symbol @code{field1} to have the value 0, the symbol
6776 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6777 value 8. Assembly would be left in the absolute section, and you would need to
6778 use a @code{.section} directive of some sort to change to some other section
6779 before further assembly.
6783 @section @code{.subsection @var{name}}
6785 @cindex @code{subsection} directive
6786 @cindex Section Stack
6787 This is one of the ELF section stack manipulation directives. The others are
6788 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6789 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6792 This directive replaces the current subsection with @code{name}. The current
6793 section is not changed. The replaced subsection is put onto the section stack
6794 in place of the then current top of stack subsection.
6799 @section @code{.symver}
6800 @cindex @code{symver} directive
6801 @cindex symbol versioning
6802 @cindex versions of symbols
6803 Use the @code{.symver} directive to bind symbols to specific version nodes
6804 within a source file. This is only supported on ELF platforms, and is
6805 typically used when assembling files to be linked into a shared library.
6806 There are cases where it may make sense to use this in objects to be bound
6807 into an application itself so as to override a versioned symbol from a
6810 For ELF targets, the @code{.symver} directive can be used like this:
6812 .symver @var{name}, @var{name2@@nodename}
6814 If the symbol @var{name} is defined within the file
6815 being assembled, the @code{.symver} directive effectively creates a symbol
6816 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6817 just don't try and create a regular alias is that the @var{@@} character isn't
6818 permitted in symbol names. The @var{name2} part of the name is the actual name
6819 of the symbol by which it will be externally referenced. The name @var{name}
6820 itself is merely a name of convenience that is used so that it is possible to
6821 have definitions for multiple versions of a function within a single source
6822 file, and so that the compiler can unambiguously know which version of a
6823 function is being mentioned. The @var{nodename} portion of the alias should be
6824 the name of a node specified in the version script supplied to the linker when
6825 building a shared library. If you are attempting to override a versioned
6826 symbol from a shared library, then @var{nodename} should correspond to the
6827 nodename of the symbol you are trying to override.
6829 If the symbol @var{name} is not defined within the file being assembled, all
6830 references to @var{name} will be changed to @var{name2@@nodename}. If no
6831 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6834 Another usage of the @code{.symver} directive is:
6836 .symver @var{name}, @var{name2@@@@nodename}
6838 In this case, the symbol @var{name} must exist and be defined within
6839 the file being assembled. It is similar to @var{name2@@nodename}. The
6840 difference is @var{name2@@@@nodename} will also be used to resolve
6841 references to @var{name2} by the linker.
6843 The third usage of the @code{.symver} directive is:
6845 .symver @var{name}, @var{name2@@@@@@nodename}
6847 When @var{name} is not defined within the
6848 file being assembled, it is treated as @var{name2@@nodename}. When
6849 @var{name} is defined within the file being assembled, the symbol
6850 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6855 @section @code{.tag @var{structname}}
6857 @cindex COFF structure debugging
6858 @cindex structure debugging, COFF
6859 @cindex @code{tag} directive
6860 This directive is generated by compilers to include auxiliary debugging
6861 information in the symbol table. It is only permitted inside
6862 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6863 definitions in the symbol table with instances of those structures.
6866 @samp{.tag} is only used when generating COFF format output; when
6867 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6873 @section @code{.text @var{subsection}}
6875 @cindex @code{text} directive
6876 Tells @command{@value{AS}} to assemble the following statements onto the end of
6877 the text subsection numbered @var{subsection}, which is an absolute
6878 expression. If @var{subsection} is omitted, subsection number zero
6882 @section @code{.title "@var{heading}"}
6884 @cindex @code{title} directive
6885 @cindex listing control: title line
6886 Use @var{heading} as the title (second line, immediately after the
6887 source file name and pagenumber) when generating assembly listings.
6889 This directive affects subsequent pages, as well as the current page if
6890 it appears within ten lines of the top of a page.
6894 @section @code{.type}
6896 This directive is used to set the type of a symbol.
6900 @c only print the extra heading if both COFF and ELF are set
6901 @subheading COFF Version
6904 @cindex COFF symbol type
6905 @cindex symbol type, COFF
6906 @cindex @code{type} directive (COFF version)
6907 For COFF targets, this directive is permitted only within
6908 @code{.def}/@code{.endef} pairs. It is used like this:
6914 This records the integer @var{int} as the type attribute of a symbol table
6918 @samp{.type} is associated only with COFF format output; when
6919 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6920 directive but ignores it.
6926 @c only print the extra heading if both COFF and ELF are set
6927 @subheading ELF Version
6930 @cindex ELF symbol type
6931 @cindex symbol type, ELF
6932 @cindex @code{type} directive (ELF version)
6933 For ELF targets, the @code{.type} directive is used like this:
6936 .type @var{name} , @var{type description}
6939 This sets the type of symbol @var{name} to be either a
6940 function symbol or an object symbol. There are five different syntaxes
6941 supported for the @var{type description} field, in order to provide
6942 compatibility with various other assemblers.
6944 Because some of the characters used in these syntaxes (such as @samp{@@} and
6945 @samp{#}) are comment characters for some architectures, some of the syntaxes
6946 below do not work on all architectures. The first variant will be accepted by
6947 the GNU assembler on all architectures so that variant should be used for
6948 maximum portability, if you do not need to assemble your code with other
6951 The syntaxes supported are:
6954 .type <name> STT_<TYPE_IN_UPPER_CASE>
6955 .type <name>,#<type>
6956 .type <name>,@@<type>
6957 .type <name>,%<type>
6958 .type <name>,"<type>"
6961 The types supported are:
6966 Mark the symbol as being a function name.
6969 @itemx gnu_indirect_function
6970 Mark the symbol as an indirect function when evaluated during reloc
6971 processing. (This is only supported on assemblers targeting GNU systems).
6975 Mark the symbol as being a data object.
6979 Mark the symbol as being a thead-local data object.
6983 Mark the symbol as being a common data object.
6987 Does not mark the symbol in any way. It is supported just for completeness.
6989 @item gnu_unique_object
6990 Marks the symbol as being a globally unique data object. The dynamic linker
6991 will make sure that in the entire process there is just one symbol with this
6992 name and type in use. (This is only supported on assemblers targeting GNU
6997 Note: Some targets support extra types in addition to those listed above.
7003 @section @code{.uleb128 @var{expressions}}
7005 @cindex @code{uleb128} directive
7006 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
7007 compact, variable length representation of numbers used by the DWARF
7008 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
7012 @section @code{.val @var{addr}}
7014 @cindex @code{val} directive
7015 @cindex COFF value attribute
7016 @cindex value attribute, COFF
7017 This directive, permitted only within @code{.def}/@code{.endef} pairs,
7018 records the address @var{addr} as the value attribute of a symbol table
7022 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
7023 configured for @code{b.out}, it accepts this directive but ignores it.
7029 @section @code{.version "@var{string}"}
7031 @cindex @code{version} directive
7032 This directive creates a @code{.note} section and places into it an ELF
7033 formatted note of type NT_VERSION. The note's name is set to @code{string}.
7038 @section @code{.vtable_entry @var{table}, @var{offset}}
7040 @cindex @code{vtable_entry} directive
7041 This directive finds or creates a symbol @code{table} and creates a
7042 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
7045 @section @code{.vtable_inherit @var{child}, @var{parent}}
7047 @cindex @code{vtable_inherit} directive
7048 This directive finds the symbol @code{child} and finds or creates the symbol
7049 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
7050 parent whose addend is the value of the child symbol. As a special case the
7051 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
7055 @section @code{.warning "@var{string}"}
7056 @cindex warning directive
7057 Similar to the directive @code{.error}
7058 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
7061 @section @code{.weak @var{names}}
7063 @cindex @code{weak} directive
7064 This directive sets the weak attribute on the comma separated list of symbol
7065 @code{names}. If the symbols do not already exist, they will be created.
7067 On COFF targets other than PE, weak symbols are a GNU extension. This
7068 directive sets the weak attribute on the comma separated list of symbol
7069 @code{names}. If the symbols do not already exist, they will be created.
7071 On the PE target, weak symbols are supported natively as weak aliases.
7072 When a weak symbol is created that is not an alias, GAS creates an
7073 alternate symbol to hold the default value.
7076 @section @code{.weakref @var{alias}, @var{target}}
7078 @cindex @code{weakref} directive
7079 This directive creates an alias to the target symbol that enables the symbol to
7080 be referenced with weak-symbol semantics, but without actually making it weak.
7081 If direct references or definitions of the symbol are present, then the symbol
7082 will not be weak, but if all references to it are through weak references, the
7083 symbol will be marked as weak in the symbol table.
7085 The effect is equivalent to moving all references to the alias to a separate
7086 assembly source file, renaming the alias to the symbol in it, declaring the
7087 symbol as weak there, and running a reloadable link to merge the object files
7088 resulting from the assembly of the new source file and the old source file that
7089 had the references to the alias removed.
7091 The alias itself never makes to the symbol table, and is entirely handled
7092 within the assembler.
7095 @section @code{.word @var{expressions}}
7097 @cindex @code{word} directive
7098 This directive expects zero or more @var{expressions}, of any section,
7099 separated by commas.
7102 For each expression, @command{@value{AS}} emits a 32-bit number.
7105 For each expression, @command{@value{AS}} emits a 16-bit number.
7110 The size of the number emitted, and its byte order,
7111 depend on what target computer the assembly is for.
7114 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
7115 @c happen---32-bit addressability, period; no long/short jumps.
7116 @ifset DIFF-TBL-KLUGE
7117 @cindex difference tables altered
7118 @cindex altered difference tables
7120 @emph{Warning: Special Treatment to support Compilers}
7124 Machines with a 32-bit address space, but that do less than 32-bit
7125 addressing, require the following special treatment. If the machine of
7126 interest to you does 32-bit addressing (or doesn't require it;
7127 @pxref{Machine Dependencies}), you can ignore this issue.
7130 In order to assemble compiler output into something that works,
7131 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
7132 Directives of the form @samp{.word sym1-sym2} are often emitted by
7133 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
7134 directive of the form @samp{.word sym1-sym2}, and the difference between
7135 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
7136 creates a @dfn{secondary jump table}, immediately before the next label.
7137 This secondary jump table is preceded by a short-jump to the
7138 first byte after the secondary table. This short-jump prevents the flow
7139 of control from accidentally falling into the new table. Inside the
7140 table is a long-jump to @code{sym2}. The original @samp{.word}
7141 contains @code{sym1} minus the address of the long-jump to
7144 If there were several occurrences of @samp{.word sym1-sym2} before the
7145 secondary jump table, all of them are adjusted. If there was a
7146 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
7147 long-jump to @code{sym4} is included in the secondary jump table,
7148 and the @code{.word} directives are adjusted to contain @code{sym3}
7149 minus the address of the long-jump to @code{sym4}; and so on, for as many
7150 entries in the original jump table as necessary.
7153 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
7154 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
7155 assembly language programmers.
7158 @c end DIFF-TBL-KLUGE
7160 @ifclear no-space-dir
7162 @section @code{.zero @var{size}}
7164 @cindex @code{zero} directive
7165 @cindex filling memory with zero bytes
7166 This directive emits @var{size} 0-valued bytes. @var{size} must be an absolute
7167 expression. This directive is actually an alias for the @samp{.skip} directive
7168 so in can take an optional second argument of the value to store in the bytes
7169 instead of zero. Using @samp{.zero} in this way would be confusing however.
7173 @section Deprecated Directives
7175 @cindex deprecated directives
7176 @cindex obsolescent directives
7177 One day these directives won't work.
7178 They are included for compatibility with older assemblers.
7185 @node Object Attributes
7186 @chapter Object Attributes
7187 @cindex object attributes
7189 @command{@value{AS}} assembles source files written for a specific architecture
7190 into object files for that architecture. But not all object files are alike.
7191 Many architectures support incompatible variations. For instance, floating
7192 point arguments might be passed in floating point registers if the object file
7193 requires hardware floating point support---or floating point arguments might be
7194 passed in integer registers if the object file supports processors with no
7195 hardware floating point unit. Or, if two objects are built for different
7196 generations of the same architecture, the combination may require the
7197 newer generation at run-time.
7199 This information is useful during and after linking. At link time,
7200 @command{@value{LD}} can warn about incompatible object files. After link
7201 time, tools like @command{gdb} can use it to process the linked file
7204 Compatibility information is recorded as a series of object attributes. Each
7205 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
7206 string, and indicates who sets the meaning of the tag. The tag is an integer,
7207 and indicates what property the attribute describes. The value may be a string
7208 or an integer, and indicates how the property affects this object. Missing
7209 attributes are the same as attributes with a zero value or empty string value.
7211 Object attributes were developed as part of the ABI for the ARM Architecture.
7212 The file format is documented in @cite{ELF for the ARM Architecture}.
7215 * GNU Object Attributes:: @sc{gnu} Object Attributes
7216 * Defining New Object Attributes:: Defining New Object Attributes
7219 @node GNU Object Attributes
7220 @section @sc{gnu} Object Attributes
7222 The @code{.gnu_attribute} directive records an object attribute
7223 with vendor @samp{gnu}.
7225 Except for @samp{Tag_compatibility}, which has both an integer and a string for
7226 its value, @sc{gnu} attributes have a string value if the tag number is odd and
7227 an integer value if the tag number is even. The second bit (@code{@var{tag} &
7228 2} is set for architecture-independent attributes and clear for
7229 architecture-dependent ones.
7231 @subsection Common @sc{gnu} attributes
7233 These attributes are valid on all architectures.
7236 @item Tag_compatibility (32)
7237 The compatibility attribute takes an integer flag value and a vendor name. If
7238 the flag value is 0, the file is compatible with other toolchains. If it is 1,
7239 then the file is only compatible with the named toolchain. If it is greater
7240 than 1, the file can only be processed by other toolchains under some private
7241 arrangement indicated by the flag value and the vendor name.
7244 @subsection MIPS Attributes
7247 @item Tag_GNU_MIPS_ABI_FP (4)
7248 The floating-point ABI used by this object file. The value will be:
7252 0 for files not affected by the floating-point ABI.
7254 1 for files using the hardware floating-point ABI with a standard
7255 double-precision FPU.
7257 2 for files using the hardware floating-point ABI with a single-precision FPU.
7259 3 for files using the software floating-point ABI.
7261 4 for files using the deprecated hardware floating-point ABI which used 64-bit
7262 floating-point registers, 32-bit general-purpose registers and increased the
7263 number of callee-saved floating-point registers.
7265 5 for files using the hardware floating-point ABI with a double-precision FPU
7266 with either 32-bit or 64-bit floating-point registers and 32-bit
7267 general-purpose registers.
7269 6 for files using the hardware floating-point ABI with 64-bit floating-point
7270 registers and 32-bit general-purpose registers.
7272 7 for files using the hardware floating-point ABI with 64-bit floating-point
7273 registers, 32-bit general-purpose registers and a rule that forbids the
7274 direct use of odd-numbered single-precision floating-point registers.
7278 @subsection PowerPC Attributes
7281 @item Tag_GNU_Power_ABI_FP (4)
7282 The floating-point ABI used by this object file. The value will be:
7286 0 for files not affected by the floating-point ABI.
7288 1 for files using double-precision hardware floating-point ABI.
7290 2 for files using the software floating-point ABI.
7292 3 for files using single-precision hardware floating-point ABI.
7295 @item Tag_GNU_Power_ABI_Vector (8)
7296 The vector ABI used by this object file. The value will be:
7300 0 for files not affected by the vector ABI.
7302 1 for files using general purpose registers to pass vectors.
7304 2 for files using AltiVec registers to pass vectors.
7306 3 for files using SPE registers to pass vectors.
7310 @subsection IBM z Systems Attributes
7313 @item Tag_GNU_S390_ABI_Vector (8)
7314 The vector ABI used by this object file. The value will be:
7318 0 for files not affected by the vector ABI.
7320 1 for files using software vector ABI.
7322 2 for files using hardware vector ABI.
7326 @node Defining New Object Attributes
7327 @section Defining New Object Attributes
7329 If you want to define a new @sc{gnu} object attribute, here are the places you
7330 will need to modify. New attributes should be discussed on the @samp{binutils}
7335 This manual, which is the official register of attributes.
7337 The header for your architecture @file{include/elf}, to define the tag.
7339 The @file{bfd} support file for your architecture, to merge the attribute
7340 and issue any appropriate link warnings.
7342 Test cases in @file{ld/testsuite} for merging and link warnings.
7344 @file{binutils/readelf.c} to display your attribute.
7346 GCC, if you want the compiler to mark the attribute automatically.
7352 @node Machine Dependencies
7353 @chapter Machine Dependent Features
7355 @cindex machine dependencies
7356 The machine instruction sets are (almost by definition) different on
7357 each machine where @command{@value{AS}} runs. Floating point representations
7358 vary as well, and @command{@value{AS}} often supports a few additional
7359 directives or command-line options for compatibility with other
7360 assemblers on a particular platform. Finally, some versions of
7361 @command{@value{AS}} support special pseudo-instructions for branch
7364 This chapter discusses most of these differences, though it does not
7365 include details on any machine's instruction set. For details on that
7366 subject, see the hardware manufacturer's manual.
7370 * AArch64-Dependent:: AArch64 Dependent Features
7373 * Alpha-Dependent:: Alpha Dependent Features
7376 * ARC-Dependent:: ARC Dependent Features
7379 * ARM-Dependent:: ARM Dependent Features
7382 * AVR-Dependent:: AVR Dependent Features
7385 * Blackfin-Dependent:: Blackfin Dependent Features
7388 * CR16-Dependent:: CR16 Dependent Features
7391 * CRIS-Dependent:: CRIS Dependent Features
7394 * D10V-Dependent:: D10V Dependent Features
7397 * D30V-Dependent:: D30V Dependent Features
7400 * Epiphany-Dependent:: EPIPHANY Dependent Features
7403 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7406 * HPPA-Dependent:: HPPA Dependent Features
7409 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
7412 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
7415 * i860-Dependent:: Intel 80860 Dependent Features
7418 * i960-Dependent:: Intel 80960 Dependent Features
7421 * IA-64-Dependent:: Intel IA-64 Dependent Features
7424 * IP2K-Dependent:: IP2K Dependent Features
7427 * LM32-Dependent:: LM32 Dependent Features
7430 * M32C-Dependent:: M32C Dependent Features
7433 * M32R-Dependent:: M32R Dependent Features
7436 * M68K-Dependent:: M680x0 Dependent Features
7439 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7442 * Meta-Dependent :: Meta Dependent Features
7445 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7448 * MIPS-Dependent:: MIPS Dependent Features
7451 * MMIX-Dependent:: MMIX Dependent Features
7454 * MSP430-Dependent:: MSP430 Dependent Features
7457 * NDS32-Dependent:: Andes NDS32 Dependent Features
7460 * NiosII-Dependent:: Altera Nios II Dependent Features
7463 * NS32K-Dependent:: NS32K Dependent Features
7466 * PDP-11-Dependent:: PDP-11 Dependent Features
7469 * PJ-Dependent:: picoJava Dependent Features
7472 * PPC-Dependent:: PowerPC Dependent Features
7475 * RL78-Dependent:: RL78 Dependent Features
7478 * RX-Dependent:: RX Dependent Features
7481 * S/390-Dependent:: IBM S/390 Dependent Features
7484 * SCORE-Dependent:: SCORE Dependent Features
7487 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7488 * SH64-Dependent:: SuperH SH64 Dependent Features
7491 * Sparc-Dependent:: SPARC Dependent Features
7494 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7497 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7500 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7503 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7506 * V850-Dependent:: V850 Dependent Features
7509 * Vax-Dependent:: VAX Dependent Features
7512 * Visium-Dependent:: Visium Dependent Features
7515 * XGATE-Dependent:: XGATE Features
7518 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7521 * Xtensa-Dependent:: Xtensa Dependent Features
7524 * Z80-Dependent:: Z80 Dependent Features
7527 * Z8000-Dependent:: Z8000 Dependent Features
7534 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7535 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7536 @c peculiarity: to preserve cross-references, there must be a node called
7537 @c "Machine Dependencies". Hence the conditional nodenames in each
7538 @c major node below. Node defaulting in makeinfo requires adjacency of
7539 @c node and sectioning commands; hence the repetition of @chapter BLAH
7540 @c in both conditional blocks.
7543 @include c-aarch64.texi
7547 @include c-alpha.texi
7563 @include c-bfin.texi
7567 @include c-cr16.texi
7571 @include c-cris.texi
7576 @node Machine Dependencies
7577 @chapter Machine Dependent Features
7579 The machine instruction sets are different on each Renesas chip family,
7580 and there are also some syntax differences among the families. This
7581 chapter describes the specific @command{@value{AS}} features for each
7585 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7586 * SH-Dependent:: Renesas SH Dependent Features
7593 @include c-d10v.texi
7597 @include c-d30v.texi
7601 @include c-epiphany.texi
7605 @include c-h8300.texi
7609 @include c-hppa.texi
7613 @include c-i370.texi
7617 @include c-i386.texi
7621 @include c-i860.texi
7625 @include c-i960.texi
7629 @include c-ia64.texi
7633 @include c-ip2k.texi
7637 @include c-lm32.texi
7641 @include c-m32c.texi
7645 @include c-m32r.texi
7649 @include c-m68k.texi
7653 @include c-m68hc11.texi
7657 @include c-metag.texi
7661 @include c-microblaze.texi
7665 @include c-mips.texi
7669 @include c-mmix.texi
7673 @include c-msp430.texi
7677 @include c-nds32.texi
7681 @include c-nios2.texi
7685 @include c-ns32k.texi
7689 @include c-pdp11.texi
7701 @include c-rl78.texi
7709 @include c-s390.texi
7713 @include c-score.texi
7718 @include c-sh64.texi
7722 @include c-sparc.texi
7726 @include c-tic54x.texi
7730 @include c-tic6x.texi
7734 @include c-tilegx.texi
7738 @include c-tilepro.texi
7742 @include c-v850.texi
7750 @include c-visium.texi
7754 @include c-xgate.texi
7758 @include c-xstormy16.texi
7762 @include c-xtensa.texi
7774 @c reverse effect of @down at top of generic Machine-Dep chapter
7778 @node Reporting Bugs
7779 @chapter Reporting Bugs
7780 @cindex bugs in assembler
7781 @cindex reporting bugs in assembler
7783 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7785 Reporting a bug may help you by bringing a solution to your problem, or it may
7786 not. But in any case the principal function of a bug report is to help the
7787 entire community by making the next version of @command{@value{AS}} work better.
7788 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7790 In order for a bug report to serve its purpose, you must include the
7791 information that enables us to fix the bug.
7794 * Bug Criteria:: Have you found a bug?
7795 * Bug Reporting:: How to report bugs
7799 @section Have You Found a Bug?
7800 @cindex bug criteria
7802 If you are not sure whether you have found a bug, here are some guidelines:
7805 @cindex fatal signal
7806 @cindex assembler crash
7807 @cindex crash of assembler
7809 If the assembler gets a fatal signal, for any input whatever, that is a
7810 @command{@value{AS}} bug. Reliable assemblers never crash.
7812 @cindex error on valid input
7814 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7816 @cindex invalid input
7818 If @command{@value{AS}} does not produce an error message for invalid input, that
7819 is a bug. However, you should note that your idea of ``invalid input'' might
7820 be our idea of ``an extension'' or ``support for traditional practice''.
7823 If you are an experienced user of assemblers, your suggestions for improvement
7824 of @command{@value{AS}} are welcome in any case.
7828 @section How to Report Bugs
7830 @cindex assembler bugs, reporting
7832 A number of companies and individuals offer support for @sc{gnu} products. If
7833 you obtained @command{@value{AS}} from a support organization, we recommend you
7834 contact that organization first.
7836 You can find contact information for many support companies and
7837 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7841 In any event, we also recommend that you send bug reports for @command{@value{AS}}
7845 The fundamental principle of reporting bugs usefully is this:
7846 @strong{report all the facts}. If you are not sure whether to state a
7847 fact or leave it out, state it!
7849 Often people omit facts because they think they know what causes the problem
7850 and assume that some details do not matter. Thus, you might assume that the
7851 name of a symbol you use in an example does not matter. Well, probably it does
7852 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
7853 happens to fetch from the location where that name is stored in memory;
7854 perhaps, if the name were different, the contents of that location would fool
7855 the assembler into doing the right thing despite the bug. Play it safe and
7856 give a specific, complete example. That is the easiest thing for you to do,
7857 and the most helpful.
7859 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
7860 it is new to us. Therefore, always write your bug reports on the assumption
7861 that the bug has not been reported previously.
7863 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7864 bell?'' This cannot help us fix a bug, so it is basically useless. We
7865 respond by asking for enough details to enable us to investigate.
7866 You might as well expedite matters by sending them to begin with.
7868 To enable us to fix the bug, you should include all these things:
7872 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
7873 it with the @samp{--version} argument.
7875 Without this, we will not know whether there is any point in looking for
7876 the bug in the current version of @command{@value{AS}}.
7879 Any patches you may have applied to the @command{@value{AS}} source.
7882 The type of machine you are using, and the operating system name and
7886 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
7890 The command arguments you gave the assembler to assemble your example and
7891 observe the bug. To guarantee you will not omit something important, list them
7892 all. A copy of the Makefile (or the output from make) is sufficient.
7894 If we were to try to guess the arguments, we would probably guess wrong
7895 and then we might not encounter the bug.
7898 A complete input file that will reproduce the bug. If the bug is observed when
7899 the assembler is invoked via a compiler, send the assembler source, not the
7900 high level language source. Most compilers will produce the assembler source
7901 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
7902 the options @samp{-v --save-temps}; this will save the assembler source in a
7903 file with an extension of @file{.s}, and also show you exactly how
7904 @command{@value{AS}} is being run.
7907 A description of what behavior you observe that you believe is
7908 incorrect. For example, ``It gets a fatal signal.''
7910 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
7911 will certainly notice it. But if the bug is incorrect output, we might not
7912 notice unless it is glaringly wrong. You might as well not give us a chance to
7915 Even if the problem you experience is a fatal signal, you should still say so
7916 explicitly. Suppose something strange is going on, such as, your copy of
7917 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
7918 library on your system. (This has happened!) Your copy might crash and ours
7919 would not. If you told us to expect a crash, then when ours fails to crash, we
7920 would know that the bug was not happening for us. If you had not told us to
7921 expect a crash, then we would not be able to draw any conclusion from our
7925 If you wish to suggest changes to the @command{@value{AS}} source, send us context
7926 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
7927 option. Always send diffs from the old file to the new file. If you even
7928 discuss something in the @command{@value{AS}} source, refer to it by context, not
7931 The line numbers in our development sources will not match those in your
7932 sources. Your line numbers would convey no useful information to us.
7935 Here are some things that are not necessary:
7939 A description of the envelope of the bug.
7941 Often people who encounter a bug spend a lot of time investigating
7942 which changes to the input file will make the bug go away and which
7943 changes will not affect it.
7945 This is often time consuming and not very useful, because the way we
7946 will find the bug is by running a single example under the debugger
7947 with breakpoints, not by pure deduction from a series of examples.
7948 We recommend that you save your time for something else.
7950 Of course, if you can find a simpler example to report @emph{instead}
7951 of the original one, that is a convenience for us. Errors in the
7952 output will be easier to spot, running under the debugger will take
7953 less time, and so on.
7955 However, simplification is not vital; if you do not want to do this,
7956 report the bug anyway and send us the entire test case you used.
7959 A patch for the bug.
7961 A patch for the bug does help us if it is a good one. But do not omit
7962 the necessary information, such as the test case, on the assumption that
7963 a patch is all we need. We might see problems with your patch and decide
7964 to fix the problem another way, or we might not understand it at all.
7966 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
7967 construct an example that will make the program follow a certain path through
7968 the code. If you do not send us the example, we will not be able to construct
7969 one, so we will not be able to verify that the bug is fixed.
7971 And if we cannot understand what bug you are trying to fix, or why your
7972 patch should be an improvement, we will not install it. A test case will
7973 help us to understand.
7976 A guess about what the bug is or what it depends on.
7978 Such guesses are usually wrong. Even we cannot guess right about such
7979 things without first using the debugger to find the facts.
7982 @node Acknowledgements
7983 @chapter Acknowledgements
7985 If you have contributed to GAS and your name isn't listed here,
7986 it is not meant as a slight. We just don't know about it. Send mail to the
7987 maintainer, and we'll correct the situation. Currently
7989 the maintainer is Nick Clifton (email address @code{nickc@@redhat.com}).
7991 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
7994 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
7995 information and the 68k series machines, most of the preprocessing pass, and
7996 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
7998 K. Richard Pixley maintained GAS for a while, adding various enhancements and
7999 many bug fixes, including merging support for several processors, breaking GAS
8000 up to handle multiple object file format back ends (including heavy rewrite,
8001 testing, an integration of the coff and b.out back ends), adding configuration
8002 including heavy testing and verification of cross assemblers and file splits
8003 and renaming, converted GAS to strictly ANSI C including full prototypes, added
8004 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
8005 port (including considerable amounts of reverse engineering), a SPARC opcode
8006 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
8007 assertions and made them work, much other reorganization, cleanup, and lint.
8009 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
8010 in format-specific I/O modules.
8012 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
8013 has done much work with it since.
8015 The Intel 80386 machine description was written by Eliot Dresselhaus.
8017 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
8019 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
8020 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
8022 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
8023 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
8024 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
8025 support a.out format.
8027 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
8028 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
8029 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
8030 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
8033 John Gilmore built the AMD 29000 support, added @code{.include} support, and
8034 simplified the configuration of which versions accept which directives. He
8035 updated the 68k machine description so that Motorola's opcodes always produced
8036 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
8037 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
8038 cross-compilation support, and one bug in relaxation that took a week and
8039 required the proverbial one-bit fix.
8041 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
8042 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
8043 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
8044 PowerPC assembler, and made a few other minor patches.
8046 Steve Chamberlain made GAS able to generate listings.
8048 Hewlett-Packard contributed support for the HP9000/300.
8050 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
8051 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
8052 formats). This work was supported by both the Center for Software Science at
8053 the University of Utah and Cygnus Support.
8055 Support for ELF format files has been worked on by Mark Eichin of Cygnus
8056 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
8057 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
8058 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
8059 and some initial 64-bit support).
8061 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
8063 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
8064 support for openVMS/Alpha.
8066 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
8069 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
8070 Inc.@: added support for Xtensa processors.
8072 Several engineers at Cygnus Support have also provided many small bug fixes and
8073 configuration enhancements.
8075 Jon Beniston added support for the Lattice Mico32 architecture.
8077 Many others have contributed large or small bugfixes and enhancements. If
8078 you have contributed significant work and are not mentioned on this list, and
8079 want to be, let us know. Some of the history has been lost; we are not
8080 intentionally leaving anyone out.
8082 @node GNU Free Documentation License
8083 @appendix GNU Free Documentation License
8087 @unnumbered AS Index