1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright (C) 1991-2020 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{} 1991-2020 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{} 1991-2020 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-<N>}] [@b{--gdwarf-sections}]
234 [@b{--gdwarf-cie-version}=@var{VERSION}]
235 [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
236 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
237 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
238 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}]
239 [@b{--no-pad-sections}]
240 [@b{-o} @var{objfile}] [@b{-R}]
241 [@b{--hash-size}=@var{NUM}] [@b{--reduce-memory-overheads}]
243 [@b{-v}] [@b{-version}] [@b{--version}]
244 [@b{-W}] [@b{--warn}] [@b{--fatal-warnings}] [@b{-w}] [@b{-x}]
245 [@b{-Z}] [@b{@@@var{FILE}}]
246 [@b{--sectname-subst}] [@b{--size-check=[error|warning]}]
247 [@b{--elf-stt-common=[no|yes]}]
248 [@b{--generate-missing-build-notes=[no|yes]}]
249 [@b{--target-help}] [@var{target-options}]
250 [@b{--}|@var{files} @dots{}]
253 @c Target dependent options are listed below. Keep the list sorted.
254 @c Add an empty line for separation.
258 @emph{Target AArch64 options:}
260 [@b{-mabi}=@var{ABI}]
264 @emph{Target Alpha options:}
266 [@b{-mdebug} | @b{-no-mdebug}]
267 [@b{-replace} | @b{-noreplace}]
268 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
269 [@b{-F}] [@b{-32addr}]
273 @emph{Target ARC options:}
274 [@b{-mcpu=@var{cpu}}]
275 [@b{-mA6}|@b{-mARC600}|@b{-mARC601}|@b{-mA7}|@b{-mARC700}|@b{-mEM}|@b{-mHS}]
282 @emph{Target ARM options:}
283 @c Don't document the deprecated options
284 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
285 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
286 [@b{-mfpu}=@var{floating-point-format}]
287 [@b{-mfloat-abi}=@var{abi}]
288 [@b{-meabi}=@var{ver}]
291 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
292 @b{-mapcs-reentrant}]
293 [@b{-mthumb-interwork}] [@b{-k}]
297 @emph{Target Blackfin options:}
298 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
305 @emph{Target BPF options:}
310 @emph{Target CRIS options:}
311 [@b{--underscore} | @b{--no-underscore}]
313 [@b{--emulation=criself} | @b{--emulation=crisaout}]
314 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
315 @c Deprecated -- deliberately not documented.
320 @emph{Target C-SKY options:}
321 [@b{-march=@var{arch}}] [@b{-mcpu=@var{cpu}}]
322 [@b{-EL}] [@b{-mlittle-endian}] [@b{-EB}] [@b{-mbig-endian}]
323 [@b{-fpic}] [@b{-pic}]
324 [@b{-mljump}] [@b{-mno-ljump}]
325 [@b{-force2bsr}] [@b{-mforce2bsr}] [@b{-no-force2bsr}] [@b{-mno-force2bsr}]
326 [@b{-jsri2bsr}] [@b{-mjsri2bsr}] [@b{-no-jsri2bsr }] [@b{-mno-jsri2bsr}]
327 [@b{-mnolrw }] [@b{-mno-lrw}]
328 [@b{-melrw}] [@b{-mno-elrw}]
329 [@b{-mlaf }] [@b{-mliterals-after-func}]
330 [@b{-mno-laf}] [@b{-mno-literals-after-func}]
331 [@b{-mlabr}] [@b{-mliterals-after-br}]
332 [@b{-mno-labr}] [@b{-mnoliterals-after-br}]
333 [@b{-mistack}] [@b{-mno-istack}]
334 [@b{-mhard-float}] [@b{-mmp}] [@b{-mcp}] [@b{-mcache}]
335 [@b{-msecurity}] [@b{-mtrust}]
336 [@b{-mdsp}] [@b{-medsp}] [@b{-mvdsp}]
340 @emph{Target D10V options:}
345 @emph{Target D30V options:}
346 [@b{-O}|@b{-n}|@b{-N}]
350 @emph{Target EPIPHANY options:}
351 [@b{-mepiphany}|@b{-mepiphany16}]
355 @emph{Target H8/300 options:}
359 @c HPPA has no machine-dependent assembler options (yet).
363 @emph{Target i386 options:}
364 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
365 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
369 @emph{Target IA-64 options:}
370 [@b{-mconstant-gp}|@b{-mauto-pic}]
371 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
373 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
374 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
375 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
376 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
380 @emph{Target IP2K options:}
381 [@b{-mip2022}|@b{-mip2022ext}]
385 @emph{Target M32C options:}
386 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
390 @emph{Target M32R options:}
391 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
396 @emph{Target M680X0 options:}
397 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
401 @emph{Target M68HC11 options:}
402 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}|@b{-mm9s12x}|@b{-mm9s12xg}]
403 [@b{-mshort}|@b{-mlong}]
404 [@b{-mshort-double}|@b{-mlong-double}]
405 [@b{--force-long-branches}] [@b{--short-branches}]
406 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
407 [@b{--print-opcodes}] [@b{--generate-example}]
411 @emph{Target MCORE options:}
412 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
413 [@b{-mcpu=[210|340]}]
417 @emph{Target Meta options:}
418 [@b{-mcpu=@var{cpu}}] [@b{-mfpu=@var{cpu}}] [@b{-mdsp=@var{cpu}}]
421 @emph{Target MICROBLAZE options:}
422 @c MicroBlaze has no machine-dependent assembler options.
426 @emph{Target MIPS options:}
427 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
428 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
429 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
430 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
431 [@b{-mfp64}] [@b{-mgp64}] [@b{-mfpxx}]
432 [@b{-modd-spreg}] [@b{-mno-odd-spreg}]
433 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
434 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
435 [@b{-mips32r3}] [@b{-mips32r5}] [@b{-mips32r6}] [@b{-mips64}] [@b{-mips64r2}]
436 [@b{-mips64r3}] [@b{-mips64r5}] [@b{-mips64r6}]
437 [@b{-construct-floats}] [@b{-no-construct-floats}]
438 [@b{-mignore-branch-isa}] [@b{-mno-ignore-branch-isa}]
439 [@b{-mnan=@var{encoding}}]
440 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
441 [@b{-mips16}] [@b{-no-mips16}]
442 [@b{-mmips16e2}] [@b{-mno-mips16e2}]
443 [@b{-mmicromips}] [@b{-mno-micromips}]
444 [@b{-msmartmips}] [@b{-mno-smartmips}]
445 [@b{-mips3d}] [@b{-no-mips3d}]
446 [@b{-mdmx}] [@b{-no-mdmx}]
447 [@b{-mdsp}] [@b{-mno-dsp}]
448 [@b{-mdspr2}] [@b{-mno-dspr2}]
449 [@b{-mdspr3}] [@b{-mno-dspr3}]
450 [@b{-mmsa}] [@b{-mno-msa}]
451 [@b{-mxpa}] [@b{-mno-xpa}]
452 [@b{-mmt}] [@b{-mno-mt}]
453 [@b{-mmcu}] [@b{-mno-mcu}]
454 [@b{-mcrc}] [@b{-mno-crc}]
455 [@b{-mginv}] [@b{-mno-ginv}]
456 [@b{-mloongson-mmi}] [@b{-mno-loongson-mmi}]
457 [@b{-mloongson-cam}] [@b{-mno-loongson-cam}]
458 [@b{-mloongson-ext}] [@b{-mno-loongson-ext}]
459 [@b{-mloongson-ext2}] [@b{-mno-loongson-ext2}]
460 [@b{-minsn32}] [@b{-mno-insn32}]
461 [@b{-mfix7000}] [@b{-mno-fix7000}]
462 [@b{-mfix-rm7000}] [@b{-mno-fix-rm7000}]
463 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
464 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
465 [@b{-mfix-r5900}] [@b{-mno-fix-r5900}]
466 [@b{-mdebug}] [@b{-no-mdebug}]
467 [@b{-mpdr}] [@b{-mno-pdr}]
471 @emph{Target MMIX options:}
472 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
473 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
474 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
475 [@b{--linker-allocated-gregs}]
479 @emph{Target Nios II options:}
480 [@b{-relax-all}] [@b{-relax-section}] [@b{-no-relax}]
485 @emph{Target NDS32 options:}
486 [@b{-EL}] [@b{-EB}] [@b{-O}] [@b{-Os}] [@b{-mcpu=@var{cpu}}]
487 [@b{-misa=@var{isa}}] [@b{-mabi=@var{abi}}] [@b{-mall-ext}]
488 [@b{-m[no-]16-bit}] [@b{-m[no-]perf-ext}] [@b{-m[no-]perf2-ext}]
489 [@b{-m[no-]string-ext}] [@b{-m[no-]dsp-ext}] [@b{-m[no-]mac}] [@b{-m[no-]div}]
490 [@b{-m[no-]audio-isa-ext}] [@b{-m[no-]fpu-sp-ext}] [@b{-m[no-]fpu-dp-ext}]
491 [@b{-m[no-]fpu-fma}] [@b{-mfpu-freg=@var{FREG}}] [@b{-mreduced-regs}]
492 [@b{-mfull-regs}] [@b{-m[no-]dx-regs}] [@b{-mpic}] [@b{-mno-relax}]
496 @c OpenRISC has no machine-dependent assembler options.
500 @emph{Target PDP11 options:}
501 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
502 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
503 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
507 @emph{Target picoJava options:}
512 @emph{Target PowerPC options:}
514 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
515 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mgekko}|
516 @b{-mbroadway}|@b{-mppc64}|@b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|
517 @b{-me6500}|@b{-mppc64bridge}|@b{-mbooke}|@b{-mpower4}|@b{-mpwr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|
518 @b{-mpower6}|@b{-mpwr6}|@b{-mpower7}|@b{-mpwr7}|@b{-mpower8}|@b{-mpwr8}|@b{-mpower9}|@b{-mpwr9}@b{-ma2}|
519 @b{-mcell}|@b{-mspe}|@b{-mspe2}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
520 [@b{-many}] [@b{-maltivec}|@b{-mvsx}|@b{-mhtm}|@b{-mvle}]
521 [@b{-mregnames}|@b{-mno-regnames}]
522 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
523 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
524 [@b{-msolaris}|@b{-mno-solaris}]
525 [@b{-nops=@var{count}}]
529 @emph{Target PRU options:}
532 [@b{-mno-warn-regname-label}]
536 @emph{Target RISC-V options:}
537 [@b{-fpic}|@b{-fPIC}|@b{-fno-pic}]
538 [@b{-march}=@var{ISA}]
539 [@b{-mabi}=@var{ABI}]
543 @emph{Target RL78 options:}
545 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
549 @emph{Target RX options:}
550 [@b{-mlittle-endian}|@b{-mbig-endian}]
551 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
552 [@b{-muse-conventional-section-names}]
553 [@b{-msmall-data-limit}]
556 [@b{-mint-register=@var{number}}]
557 [@b{-mgcc-abi}|@b{-mrx-abi}]
561 @emph{Target s390 options:}
562 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
563 [@b{-mregnames}|@b{-mno-regnames}]
564 [@b{-mwarn-areg-zero}]
568 @emph{Target SCORE options:}
569 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
570 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
571 [@b{-march=score7}][@b{-march=score3}]
572 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
576 @emph{Target SPARC options:}
577 @c The order here is important. See c-sparc.texi.
578 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Aleon}|@b{-Asparclet}|@b{-Asparclite}
579 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av8plusb}|@b{-Av8plusc}|@b{-Av8plusd}
580 @b{-Av8plusv}|@b{-Av8plusm}|@b{-Av9}|@b{-Av9a}|@b{-Av9b}|@b{-Av9c}
581 @b{-Av9d}|@b{-Av9e}|@b{-Av9v}|@b{-Av9m}|@b{-Asparc}|@b{-Asparcvis}
582 @b{-Asparcvis2}|@b{-Asparcfmaf}|@b{-Asparcima}|@b{-Asparcvis3}
583 @b{-Asparcvisr}|@b{-Asparc5}]
584 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}]|@b{-xarch=v8plusb}|@b{-xarch=v8plusc}
585 @b{-xarch=v8plusd}|@b{-xarch=v8plusv}|@b{-xarch=v8plusm}|@b{-xarch=v9}
586 @b{-xarch=v9a}|@b{-xarch=v9b}|@b{-xarch=v9c}|@b{-xarch=v9d}|@b{-xarch=v9e}
587 @b{-xarch=v9v}|@b{-xarch=v9m}|@b{-xarch=sparc}|@b{-xarch=sparcvis}
588 @b{-xarch=sparcvis2}|@b{-xarch=sparcfmaf}|@b{-xarch=sparcima}
589 @b{-xarch=sparcvis3}|@b{-xarch=sparcvisr}|@b{-xarch=sparc5}
592 [@b{--enforce-aligned-data}][@b{--dcti-couples-detect}]
596 @emph{Target TIC54X options:}
597 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
598 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
602 @emph{Target TIC6X options:}
603 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
604 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
605 [@b{-mpic}|@b{-mno-pic}]
609 @emph{Target TILE-Gx options:}
610 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
613 @c TILEPro has no machine-dependent assembler options
617 @emph{Target Visium options:}
618 [@b{-mtune=@var{arch}}]
622 @emph{Target Xtensa options:}
623 [@b{--[no-]text-section-literals}] [@b{--[no-]auto-litpools}]
624 [@b{--[no-]absolute-literals}]
625 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
626 [@b{--[no-]transform}]
627 [@b{--rename-section} @var{oldname}=@var{newname}]
628 [@b{--[no-]trampolines}]
629 [@b{--abi-windowed}|@b{--abi-call0}]
633 @emph{Target Z80 options:}
634 [@b{-march=@var{CPU}@var{[-EXT]}@var{[+EXT]}}]
635 [@b{-local-prefix=}@var{PREFIX}]
638 [@b{-fp-s=}@var{FORMAT}]
639 [@b{-fp-d=}@var{FORMAT}]
643 @c Z8000 has no machine-dependent assembler options
652 @include at-file.texi
655 Turn on listings, in any of a variety of ways:
659 omit false conditionals
662 omit debugging directives
665 include general information, like @value{AS} version and options passed
668 include high-level source
674 include macro expansions
677 omit forms processing
683 set the name of the listing file
686 You may combine these options; for example, use @samp{-aln} for assembly
687 listing without forms processing. The @samp{=file} option, if used, must be
688 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
691 Begin in alternate macro mode.
693 @xref{Altmacro,,@code{.altmacro}}.
696 @item --compress-debug-sections
697 Compress DWARF debug sections using zlib with SHF_COMPRESSED from the
698 ELF ABI. The resulting object file may not be compatible with older
699 linkers and object file utilities. Note if compression would make a
700 given section @emph{larger} then it is not compressed.
703 @cindex @samp{--compress-debug-sections=} option
704 @item --compress-debug-sections=none
705 @itemx --compress-debug-sections=zlib
706 @itemx --compress-debug-sections=zlib-gnu
707 @itemx --compress-debug-sections=zlib-gabi
708 These options control how DWARF debug sections are compressed.
709 @option{--compress-debug-sections=none} is equivalent to
710 @option{--nocompress-debug-sections}.
711 @option{--compress-debug-sections=zlib} and
712 @option{--compress-debug-sections=zlib-gabi} are equivalent to
713 @option{--compress-debug-sections}.
714 @option{--compress-debug-sections=zlib-gnu} compresses DWARF debug
715 sections using zlib. The debug sections are renamed to begin with
716 @samp{.zdebug}. Note if compression would make a given section
717 @emph{larger} then it is not compressed nor renamed.
721 @item --nocompress-debug-sections
722 Do not compress DWARF debug sections. This is usually the default for all
723 targets except the x86/x86_64, but a configure time option can be used to
727 Ignored. This option is accepted for script compatibility with calls to
730 @item --debug-prefix-map @var{old}=@var{new}
731 When assembling files in directory @file{@var{old}}, record debugging
732 information describing them as in @file{@var{new}} instead.
734 @item --defsym @var{sym}=@var{value}
735 Define the symbol @var{sym} to be @var{value} before assembling the input file.
736 @var{value} must be an integer constant. As in C, a leading @samp{0x}
737 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
738 value. The value of the symbol can be overridden inside a source file via the
739 use of a @code{.set} pseudo-op.
742 ``fast''---skip whitespace and comment preprocessing (assume source is
747 Generate debugging information for each assembler source line using whichever
748 debug format is preferred by the target. This currently means either STABS,
752 Generate stabs debugging information for each assembler line. This
753 may help debugging assembler code, if the debugger can handle it.
756 Generate stabs debugging information for each assembler line, with GNU
757 extensions that probably only gdb can handle, and that could make other
758 debuggers crash or refuse to read your program. This
759 may help debugging assembler code. Currently the only GNU extension is
760 the location of the current working directory at assembling time.
763 Generate DWARF2 debugging information for each assembler line. This
764 may help debugging assembler code, if the debugger can handle it. Note---this
765 option is only supported by some targets, not all of them.
768 This option is the same as the @option{--gdwarf-2} option, except that it
769 allows for the possibility of the generation of extra debug information as per
770 version 3 of the DWARF specification. Note - enabling this option does not
771 guarantee the generation of any extra infortmation, the choice to do so is on a
775 This option is the same as the @option{--gdwarf-2} option, except that it
776 allows for the possibility of the generation of extra debug information as per
777 version 4 of the DWARF specification. Note - enabling this option does not
778 guarantee the generation of any extra infortmation, the choice to do so is on a
782 This option is the same as the @option{--gdwarf-2} option, except that it
783 allows for the possibility of the generation of extra debug information as per
784 version 5 of the DWARF specification. Note - enabling this option does not
785 guarantee the generation of any extra infortmation, the choice to do so is on a
788 @item --gdwarf-sections
789 Instead of creating a .debug_line section, create a series of
790 .debug_line.@var{foo} sections where @var{foo} is the name of the
791 corresponding code section. For example a code section called @var{.text.func}
792 will have its dwarf line number information placed into a section called
793 @var{.debug_line.text.func}. If the code section is just called @var{.text}
794 then debug line section will still be called just @var{.debug_line} without any
797 @item --gdwarf-cie-version=@var{version}
798 Control which version of DWARF Common Information Entries (CIEs) are produced.
799 When this flag is not specificed the default is version 1, though some targets
800 can modify this default. Other possible values for @var{version} are 3 or 4.
803 @item --size-check=error
804 @itemx --size-check=warning
805 Issue an error or warning for invalid ELF .size directive.
807 @item --elf-stt-common=no
808 @itemx --elf-stt-common=yes
809 These options control whether the ELF assembler should generate common
810 symbols with the @code{STT_COMMON} type. The default can be controlled
811 by a configure option @option{--enable-elf-stt-common}.
813 @item --generate-missing-build-notes=yes
814 @itemx --generate-missing-build-notes=no
815 These options control whether the ELF assembler should generate GNU Build
816 attribute notes if none are present in the input sources.
817 The default can be controlled by the @option{--enable-generate-build-notes}
823 Print a summary of the command-line options and exit.
826 Print a summary of all target specific options and exit.
829 Add directory @var{dir} to the search list for @code{.include} directives.
832 Don't warn about signed overflow.
835 @ifclear DIFF-TBL-KLUGE
836 This option is accepted but has no effect on the @value{TARGET} family.
838 @ifset DIFF-TBL-KLUGE
839 Issue warnings when difference tables altered for long displacements.
844 Keep (in the symbol table) local symbols. These symbols start with
845 system-specific local label prefixes, typically @samp{.L} for ELF systems
846 or @samp{L} for traditional a.out systems.
851 @item --listing-lhs-width=@var{number}
852 Set the maximum width, in words, of the output data column for an assembler
853 listing to @var{number}.
855 @item --listing-lhs-width2=@var{number}
856 Set the maximum width, in words, of the output data column for continuation
857 lines in an assembler listing to @var{number}.
859 @item --listing-rhs-width=@var{number}
860 Set the maximum width of an input source line, as displayed in a listing, to
863 @item --listing-cont-lines=@var{number}
864 Set the maximum number of lines printed in a listing for a single line of input
867 @item --no-pad-sections
868 Stop the assembler for padding the ends of output sections to the alignment
869 of that section. The default is to pad the sections, but this can waste space
870 which might be needed on targets which have tight memory constraints.
872 @item -o @var{objfile}
873 Name the object-file output from @command{@value{AS}} @var{objfile}.
876 Fold the data section into the text section.
878 @item --hash-size=@var{number}
879 Set the default size of GAS's hash tables to a prime number close to
880 @var{number}. Increasing this value can reduce the length of time it takes the
881 assembler to perform its tasks, at the expense of increasing the assembler's
882 memory requirements. Similarly reducing this value can reduce the memory
883 requirements at the expense of speed.
885 @item --reduce-memory-overheads
886 This option reduces GAS's memory requirements, at the expense of making the
887 assembly processes slower. Currently this switch is a synonym for
888 @samp{--hash-size=4051}, but in the future it may have other effects as well.
891 @item --sectname-subst
892 Honor substitution sequences in section names.
894 @xref{Section Name Substitutions,,@code{.section @var{name}}}.
899 Print the maximum space (in bytes) and total time (in seconds) used by
902 @item --strip-local-absolute
903 Remove local absolute symbols from the outgoing symbol table.
907 Print the @command{as} version.
910 Print the @command{as} version and exit.
914 Suppress warning messages.
916 @item --fatal-warnings
917 Treat warnings as errors.
920 Don't suppress warning messages or treat them as errors.
929 Generate an object file even after errors.
931 @item -- | @var{files} @dots{}
932 Standard input, or source files to assemble.
940 @xref{AArch64 Options}, for the options available when @value{AS} is configured
941 for the 64-bit mode of the ARM Architecture (AArch64).
946 The following options are available when @value{AS} is configured for the
947 64-bit mode of the ARM Architecture (AArch64).
950 @include c-aarch64.texi
951 @c ended inside the included file
959 @xref{Alpha Options}, for the options available when @value{AS} is configured
960 for an Alpha processor.
965 The following options are available when @value{AS} is configured for an Alpha
969 @include c-alpha.texi
970 @c ended inside the included file
977 The following options are available when @value{AS} is configured for an ARC
981 @item -mcpu=@var{cpu}
982 This option selects the core processor variant.
984 Select either big-endian (-EB) or little-endian (-EL) output.
986 Enable Code Density extenssion instructions.
991 The following options are available when @value{AS} is configured for the ARM
995 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
996 Specify which ARM processor variant is the target.
997 @item -march=@var{architecture}[+@var{extension}@dots{}]
998 Specify which ARM architecture variant is used by the target.
999 @item -mfpu=@var{floating-point-format}
1000 Select which Floating Point architecture is the target.
1001 @item -mfloat-abi=@var{abi}
1002 Select which floating point ABI is in use.
1004 Enable Thumb only instruction decoding.
1005 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
1006 Select which procedure calling convention is in use.
1008 Select either big-endian (-EB) or little-endian (-EL) output.
1009 @item -mthumb-interwork
1010 Specify that the code has been generated with interworking between Thumb and
1013 Turns on CodeComposer Studio assembly syntax compatibility mode.
1015 Specify that PIC code has been generated.
1023 @xref{Blackfin Options}, for the options available when @value{AS} is
1024 configured for the Blackfin processor family.
1028 @c man begin OPTIONS
1029 The following options are available when @value{AS} is configured for
1030 the Blackfin processor family.
1032 @c man begin INCLUDE
1033 @include c-bfin.texi
1034 @c ended inside the included file
1042 @xref{BPF Options}, for the options available when @value{AS} is
1043 configured for the Linux kernel BPF processor family.
1047 @c man begin OPTIONS
1048 The following options are available when @value{AS} is configured for
1049 the Linux kernel BPF processor family.
1051 @c man begin INCLUDE
1053 @c ended inside the included file
1058 @c man begin OPTIONS
1060 See the info pages for documentation of the CRIS-specific options.
1066 @xref{C-SKY Options}, for the options available when @value{AS} is
1067 configured for the C-SKY processor family.
1071 @c man begin OPTIONS
1072 The following options are available when @value{AS} is configured for
1073 the C-SKY processor family.
1075 @c man begin INCLUDE
1076 @include c-csky.texi
1077 @c ended inside the included file
1083 The following options are available when @value{AS} is configured for
1086 @cindex D10V optimization
1087 @cindex optimization, D10V
1089 Optimize output by parallelizing instructions.
1094 The following options are available when @value{AS} is configured for a D30V
1097 @cindex D30V optimization
1098 @cindex optimization, D30V
1100 Optimize output by parallelizing instructions.
1104 Warn when nops are generated.
1106 @cindex D30V nops after 32-bit multiply
1108 Warn when a nop after a 32-bit multiply instruction is generated.
1114 The following options are available when @value{AS} is configured for the
1115 Adapteva EPIPHANY series.
1118 @xref{Epiphany Options}, for the options available when @value{AS} is
1119 configured for an Epiphany processor.
1123 @c man begin OPTIONS
1124 The following options are available when @value{AS} is configured for
1125 an Epiphany processor.
1127 @c man begin INCLUDE
1128 @include c-epiphany.texi
1129 @c ended inside the included file
1137 @xref{H8/300 Options}, for the options available when @value{AS} is configured
1138 for an H8/300 processor.
1142 @c man begin OPTIONS
1143 The following options are available when @value{AS} is configured for an H8/300
1146 @c man begin INCLUDE
1147 @include c-h8300.texi
1148 @c ended inside the included file
1156 @xref{i386-Options}, for the options available when @value{AS} is
1157 configured for an i386 processor.
1161 @c man begin OPTIONS
1162 The following options are available when @value{AS} is configured for
1165 @c man begin INCLUDE
1166 @include c-i386.texi
1167 @c ended inside the included file
1172 @c man begin OPTIONS
1174 The following options are available when @value{AS} is configured for the
1180 Specifies that the extended IP2022 instructions are allowed.
1183 Restores the default behaviour, which restricts the permitted instructions to
1184 just the basic IP2022 ones.
1190 The following options are available when @value{AS} is configured for the
1191 Renesas M32C and M16C processors.
1196 Assemble M32C instructions.
1199 Assemble M16C instructions (the default).
1202 Enable support for link-time relaxations.
1205 Support H'00 style hex constants in addition to 0x00 style.
1211 The following options are available when @value{AS} is configured for the
1212 Renesas M32R (formerly Mitsubishi M32R) series.
1217 Specify which processor in the M32R family is the target. The default
1218 is normally the M32R, but this option changes it to the M32RX.
1220 @item --warn-explicit-parallel-conflicts or --Wp
1221 Produce warning messages when questionable parallel constructs are
1224 @item --no-warn-explicit-parallel-conflicts or --Wnp
1225 Do not produce warning messages when questionable parallel constructs are
1232 The following options are available when @value{AS} is configured for the
1233 Motorola 68000 series.
1238 Shorten references to undefined symbols, to one word instead of two.
1240 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1241 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1242 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1243 Specify what processor in the 68000 family is the target. The default
1244 is normally the 68020, but this can be changed at configuration time.
1246 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1247 The target machine does (or does not) have a floating-point coprocessor.
1248 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1249 the basic 68000 is not compatible with the 68881, a combination of the
1250 two can be specified, since it's possible to do emulation of the
1251 coprocessor instructions with the main processor.
1253 @item -m68851 | -mno-68851
1254 The target machine does (or does not) have a memory-management
1255 unit coprocessor. The default is to assume an MMU for 68020 and up.
1263 @xref{Nios II Options}, for the options available when @value{AS} is configured
1264 for an Altera Nios II processor.
1268 @c man begin OPTIONS
1269 The following options are available when @value{AS} is configured for an
1270 Altera Nios II processor.
1272 @c man begin INCLUDE
1273 @include c-nios2.texi
1274 @c ended inside the included file
1280 For details about the PDP-11 machine dependent features options,
1281 see @ref{PDP-11-Options}.
1284 @item -mpic | -mno-pic
1285 Generate position-independent (or position-dependent) code. The
1286 default is @option{-mpic}.
1289 @itemx -mall-extensions
1290 Enable all instruction set extensions. This is the default.
1292 @item -mno-extensions
1293 Disable all instruction set extensions.
1295 @item -m@var{extension} | -mno-@var{extension}
1296 Enable (or disable) a particular instruction set extension.
1299 Enable the instruction set extensions supported by a particular CPU, and
1300 disable all other extensions.
1302 @item -m@var{machine}
1303 Enable the instruction set extensions supported by a particular machine
1304 model, and disable all other extensions.
1310 The following options are available when @value{AS} is configured for
1311 a picoJava processor.
1315 @cindex PJ endianness
1316 @cindex endianness, PJ
1317 @cindex big endian output, PJ
1319 Generate ``big endian'' format output.
1321 @cindex little endian output, PJ
1323 Generate ``little endian'' format output.
1331 @xref{PRU Options}, for the options available when @value{AS} is configured
1332 for a PRU processor.
1336 @c man begin OPTIONS
1337 The following options are available when @value{AS} is configured for a
1340 @c man begin INCLUDE
1342 @c ended inside the included file
1347 The following options are available when @value{AS} is configured for the
1348 Motorola 68HC11 or 68HC12 series.
1352 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1353 Specify what processor is the target. The default is
1354 defined by the configuration option when building the assembler.
1356 @item --xgate-ramoffset
1357 Instruct the linker to offset RAM addresses from S12X address space into
1358 XGATE address space.
1361 Specify to use the 16-bit integer ABI.
1364 Specify to use the 32-bit integer ABI.
1366 @item -mshort-double
1367 Specify to use the 32-bit double ABI.
1370 Specify to use the 64-bit double ABI.
1372 @item --force-long-branches
1373 Relative branches are turned into absolute ones. This concerns
1374 conditional branches, unconditional branches and branches to a
1377 @item -S | --short-branches
1378 Do not turn relative branches into absolute ones
1379 when the offset is out of range.
1381 @item --strict-direct-mode
1382 Do not turn the direct addressing mode into extended addressing mode
1383 when the instruction does not support direct addressing mode.
1385 @item --print-insn-syntax
1386 Print the syntax of instruction in case of error.
1388 @item --print-opcodes
1389 Print the list of instructions with syntax and then exit.
1391 @item --generate-example
1392 Print an example of instruction for each possible instruction and then exit.
1393 This option is only useful for testing @command{@value{AS}}.
1399 The following options are available when @command{@value{AS}} is configured
1400 for the SPARC architecture:
1403 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1404 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1405 Explicitly select a variant of the SPARC architecture.
1407 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1408 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1410 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1411 UltraSPARC extensions.
1413 @item -xarch=v8plus | -xarch=v8plusa
1414 For compatibility with the Solaris v9 assembler. These options are
1415 equivalent to -Av8plus and -Av8plusa, respectively.
1418 Warn when the assembler switches to another architecture.
1423 The following options are available when @value{AS} is configured for the 'c54x
1428 Enable extended addressing mode. All addresses and relocations will assume
1429 extended addressing (usually 23 bits).
1430 @item -mcpu=@var{CPU_VERSION}
1431 Sets the CPU version being compiled for.
1432 @item -merrors-to-file @var{FILENAME}
1433 Redirect error output to a file, for broken systems which don't support such
1434 behaviour in the shell.
1439 @c man begin OPTIONS
1440 The following options are available when @value{AS} is configured for
1445 This option sets the largest size of an object that can be referenced
1446 implicitly with the @code{gp} register. It is only accepted for targets that
1447 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1449 @cindex MIPS endianness
1450 @cindex endianness, MIPS
1451 @cindex big endian output, MIPS
1453 Generate ``big endian'' format output.
1455 @cindex little endian output, MIPS
1457 Generate ``little endian'' format output.
1475 Generate code for a particular MIPS Instruction Set Architecture level.
1476 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1477 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1478 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1479 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips32r3},
1480 @samp{-mips32r5}, @samp{-mips32r6}, @samp{-mips64}, @samp{-mips64r2},
1481 @samp{-mips64r3}, @samp{-mips64r5}, and @samp{-mips64r6} correspond to generic
1482 MIPS V, MIPS32, MIPS32 Release 2, MIPS32 Release 3, MIPS32 Release 5, MIPS32
1483 Release 6, MIPS64, MIPS64 Release 2, MIPS64 Release 3, MIPS64 Release 5, and
1484 MIPS64 Release 6 ISA processors, respectively.
1486 @item -march=@var{cpu}
1487 Generate code for a particular MIPS CPU.
1489 @item -mtune=@var{cpu}
1490 Schedule and tune for a particular MIPS CPU.
1494 Cause nops to be inserted if the read of the destination register
1495 of an mfhi or mflo instruction occurs in the following two instructions.
1498 @itemx -mno-fix-rm7000
1499 Cause nops to be inserted if a dmult or dmultu instruction is
1500 followed by a load instruction.
1503 @itemx -mno-fix-r5900
1504 Do not attempt to schedule the preceding instruction into the delay slot
1505 of a branch instruction placed at the end of a short loop of six
1506 instructions or fewer and always schedule a @code{nop} instruction there
1507 instead. The short loop bug under certain conditions causes loops to
1508 execute only once or twice, due to a hardware bug in the R5900 chip.
1512 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1513 section instead of the standard ELF .stabs sections.
1517 Control generation of @code{.pdr} sections.
1521 The register sizes are normally inferred from the ISA and ABI, but these
1522 flags force a certain group of registers to be treated as 32 bits wide at
1523 all times. @samp{-mgp32} controls the size of general-purpose registers
1524 and @samp{-mfp32} controls the size of floating-point registers.
1528 The register sizes are normally inferred from the ISA and ABI, but these
1529 flags force a certain group of registers to be treated as 64 bits wide at
1530 all times. @samp{-mgp64} controls the size of general-purpose registers
1531 and @samp{-mfp64} controls the size of floating-point registers.
1534 The register sizes are normally inferred from the ISA and ABI, but using
1535 this flag in combination with @samp{-mabi=32} enables an ABI variant
1536 which will operate correctly with floating-point registers which are
1540 @itemx -mno-odd-spreg
1541 Enable use of floating-point operations on odd-numbered single-precision
1542 registers when supported by the ISA. @samp{-mfpxx} implies
1543 @samp{-mno-odd-spreg}, otherwise the default is @samp{-modd-spreg}.
1547 Generate code for the MIPS 16 processor. This is equivalent to putting
1548 @code{.module mips16} at the start of the assembly file. @samp{-no-mips16}
1549 turns off this option.
1552 @itemx -mno-mips16e2
1553 Enable the use of MIPS16e2 instructions in MIPS16 mode. This is equivalent
1554 to putting @code{.module mips16e2} at the start of the assembly file.
1555 @samp{-mno-mips16e2} turns off this option.
1558 @itemx -mno-micromips
1559 Generate code for the microMIPS processor. This is equivalent to putting
1560 @code{.module micromips} at the start of the assembly file.
1561 @samp{-mno-micromips} turns off this option. This is equivalent to putting
1562 @code{.module nomicromips} at the start of the assembly file.
1565 @itemx -mno-smartmips
1566 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1567 equivalent to putting @code{.module smartmips} at the start of the assembly
1568 file. @samp{-mno-smartmips} turns off this option.
1572 Generate code for the MIPS-3D Application Specific Extension.
1573 This tells the assembler to accept MIPS-3D instructions.
1574 @samp{-no-mips3d} turns off this option.
1578 Generate code for the MDMX Application Specific Extension.
1579 This tells the assembler to accept MDMX instructions.
1580 @samp{-no-mdmx} turns off this option.
1584 Generate code for the DSP Release 1 Application Specific Extension.
1585 This tells the assembler to accept DSP Release 1 instructions.
1586 @samp{-mno-dsp} turns off this option.
1590 Generate code for the DSP Release 2 Application Specific Extension.
1591 This option implies @samp{-mdsp}.
1592 This tells the assembler to accept DSP Release 2 instructions.
1593 @samp{-mno-dspr2} turns off this option.
1597 Generate code for the DSP Release 3 Application Specific Extension.
1598 This option implies @samp{-mdsp} and @samp{-mdspr2}.
1599 This tells the assembler to accept DSP Release 3 instructions.
1600 @samp{-mno-dspr3} turns off this option.
1604 Generate code for the MIPS SIMD Architecture Extension.
1605 This tells the assembler to accept MSA instructions.
1606 @samp{-mno-msa} turns off this option.
1610 Generate code for the MIPS eXtended Physical Address (XPA) Extension.
1611 This tells the assembler to accept XPA instructions.
1612 @samp{-mno-xpa} turns off this option.
1616 Generate code for the MT Application Specific Extension.
1617 This tells the assembler to accept MT instructions.
1618 @samp{-mno-mt} turns off this option.
1622 Generate code for the MCU Application Specific Extension.
1623 This tells the assembler to accept MCU instructions.
1624 @samp{-mno-mcu} turns off this option.
1628 Generate code for the MIPS cyclic redundancy check (CRC) Application
1629 Specific Extension. This tells the assembler to accept CRC instructions.
1630 @samp{-mno-crc} turns off this option.
1634 Generate code for the Global INValidate (GINV) Application Specific
1635 Extension. This tells the assembler to accept GINV instructions.
1636 @samp{-mno-ginv} turns off this option.
1638 @item -mloongson-mmi
1639 @itemx -mno-loongson-mmi
1640 Generate code for the Loongson MultiMedia extensions Instructions (MMI)
1641 Application Specific Extension. This tells the assembler to accept MMI
1643 @samp{-mno-loongson-mmi} turns off this option.
1645 @item -mloongson-cam
1646 @itemx -mno-loongson-cam
1647 Generate code for the Loongson Content Address Memory (CAM) instructions.
1648 This tells the assembler to accept Loongson CAM instructions.
1649 @samp{-mno-loongson-cam} turns off this option.
1651 @item -mloongson-ext
1652 @itemx -mno-loongson-ext
1653 Generate code for the Loongson EXTensions (EXT) instructions.
1654 This tells the assembler to accept Loongson EXT instructions.
1655 @samp{-mno-loongson-ext} turns off this option.
1657 @item -mloongson-ext2
1658 @itemx -mno-loongson-ext2
1659 Generate code for the Loongson EXTensions R2 (EXT2) instructions.
1660 This option implies @samp{-mloongson-ext}.
1661 This tells the assembler to accept Loongson EXT2 instructions.
1662 @samp{-mno-loongson-ext2} turns off this option.
1666 Only use 32-bit instruction encodings when generating code for the
1667 microMIPS processor. This option inhibits the use of any 16-bit
1668 instructions. This is equivalent to putting @code{.set insn32} at
1669 the start of the assembly file. @samp{-mno-insn32} turns off this
1670 option. This is equivalent to putting @code{.set noinsn32} at the
1671 start of the assembly file. By default @samp{-mno-insn32} is
1672 selected, allowing all instructions to be used.
1674 @item --construct-floats
1675 @itemx --no-construct-floats
1676 The @samp{--no-construct-floats} option disables the construction of
1677 double width floating point constants by loading the two halves of the
1678 value into the two single width floating point registers that make up
1679 the double width register. By default @samp{--construct-floats} is
1680 selected, allowing construction of these floating point constants.
1682 @item --relax-branch
1683 @itemx --no-relax-branch
1684 The @samp{--relax-branch} option enables the relaxation of out-of-range
1685 branches. By default @samp{--no-relax-branch} is selected, causing any
1686 out-of-range branches to produce an error.
1688 @item -mignore-branch-isa
1689 @itemx -mno-ignore-branch-isa
1690 Ignore branch checks for invalid transitions between ISA modes. The
1691 semantics of branches does not provide for an ISA mode switch, so in
1692 most cases the ISA mode a branch has been encoded for has to be the
1693 same as the ISA mode of the branch's target label. Therefore GAS has
1694 checks implemented that verify in branch assembly that the two ISA
1695 modes match. @samp{-mignore-branch-isa} disables these checks. By
1696 default @samp{-mno-ignore-branch-isa} is selected, causing any invalid
1697 branch requiring a transition between ISA modes to produce an error.
1699 @item -mnan=@var{encoding}
1700 Select between the IEEE 754-2008 (@option{-mnan=2008}) or the legacy
1701 (@option{-mnan=legacy}) NaN encoding format. The latter is the default.
1704 @item --emulation=@var{name}
1705 This option was formerly used to switch between ELF and ECOFF output
1706 on targets like IRIX 5 that supported both. MIPS ECOFF support was
1707 removed in GAS 2.24, so the option now serves little purpose.
1708 It is retained for backwards compatibility.
1710 The available configuration names are: @samp{mipself}, @samp{mipslelf} and
1711 @samp{mipsbelf}. Choosing @samp{mipself} now has no effect, since the output
1712 is always ELF. @samp{mipslelf} and @samp{mipsbelf} select little- and
1713 big-endian output respectively, but @samp{-EL} and @samp{-EB} are now the
1714 preferred options instead.
1717 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1724 Control how to deal with multiplication overflow and division by zero.
1725 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1726 (and only work for Instruction Set Architecture level 2 and higher);
1727 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1731 When this option is used, @command{@value{AS}} will issue a warning every
1732 time it generates a nop instruction from a macro.
1738 The following options are available when @value{AS} is configured for
1744 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1745 The command-line option @samp{-nojsri2bsr} can be used to disable it.
1749 Enable or disable the silicon filter behaviour. By default this is disabled.
1750 The default can be overridden by the @samp{-sifilter} command-line option.
1753 Alter jump instructions for long displacements.
1755 @item -mcpu=[210|340]
1756 Select the cpu type on the target hardware. This controls which instructions
1760 Assemble for a big endian target.
1763 Assemble for a little endian target.
1772 @xref{Meta Options}, for the options available when @value{AS} is configured
1773 for a Meta processor.
1777 @c man begin OPTIONS
1778 The following options are available when @value{AS} is configured for a
1781 @c man begin INCLUDE
1782 @include c-metag.texi
1783 @c ended inside the included file
1788 @c man begin OPTIONS
1790 See the info pages for documentation of the MMIX-specific options.
1796 @xref{NDS32 Options}, for the options available when @value{AS} is configured
1797 for a NDS32 processor.
1799 @c ended inside the included file
1803 @c man begin OPTIONS
1804 The following options are available when @value{AS} is configured for a
1807 @c man begin INCLUDE
1808 @include c-nds32.texi
1809 @c ended inside the included file
1816 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1817 for a PowerPC processor.
1821 @c man begin OPTIONS
1822 The following options are available when @value{AS} is configured for a
1825 @c man begin INCLUDE
1827 @c ended inside the included file
1835 @xref{RISC-V-Options}, for the options available when @value{AS} is configured
1836 for a RISC-V processor.
1840 @c man begin OPTIONS
1841 The following options are available when @value{AS} is configured for a
1844 @c man begin INCLUDE
1845 @include c-riscv.texi
1846 @c ended inside the included file
1851 @c man begin OPTIONS
1853 See the info pages for documentation of the RX-specific options.
1857 The following options are available when @value{AS} is configured for the s390
1863 Select the word size, either 31/32 bits or 64 bits.
1866 Select the architecture mode, either the Enterprise System
1867 Architecture (esa) or the z/Architecture mode (zarch).
1868 @item -march=@var{processor}
1869 Specify which s390 processor variant is the target, @samp{g5} (or
1870 @samp{arch3}), @samp{g6}, @samp{z900} (or @samp{arch5}), @samp{z990} (or
1871 @samp{arch6}), @samp{z9-109}, @samp{z9-ec} (or @samp{arch7}), @samp{z10} (or
1872 @samp{arch8}), @samp{z196} (or @samp{arch9}), @samp{zEC12} (or @samp{arch10}),
1873 @samp{z13} (or @samp{arch11}), @samp{z14} (or @samp{arch12}), or @samp{z15}
1876 @itemx -mno-regnames
1877 Allow or disallow symbolic names for registers.
1878 @item -mwarn-areg-zero
1879 Warn whenever the operand for a base or index register has been specified
1880 but evaluates to zero.
1888 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1889 for a TMS320C6000 processor.
1893 @c man begin OPTIONS
1894 The following options are available when @value{AS} is configured for a
1895 TMS320C6000 processor.
1897 @c man begin INCLUDE
1898 @include c-tic6x.texi
1899 @c ended inside the included file
1907 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1908 for a TILE-Gx processor.
1912 @c man begin OPTIONS
1913 The following options are available when @value{AS} is configured for a TILE-Gx
1916 @c man begin INCLUDE
1917 @include c-tilegx.texi
1918 @c ended inside the included file
1926 @xref{Visium Options}, for the options available when @value{AS} is configured
1927 for a Visium processor.
1931 @c man begin OPTIONS
1932 The following option is available when @value{AS} is configured for a Visium
1935 @c man begin INCLUDE
1936 @include c-visium.texi
1937 @c ended inside the included file
1945 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1946 for an Xtensa processor.
1950 @c man begin OPTIONS
1951 The following options are available when @value{AS} is configured for an
1954 @c man begin INCLUDE
1955 @include c-xtensa.texi
1956 @c ended inside the included file
1964 @xref{Z80 Options}, for the options available when @value{AS} is configured
1965 for an Z80 processor.
1969 @c man begin OPTIONS
1970 The following options are available when @value{AS} is configured for an
1973 @c man begin INCLUDE
1975 @c ended inside the included file
1981 * Manual:: Structure of this Manual
1982 * GNU Assembler:: The GNU Assembler
1983 * Object Formats:: Object File Formats
1984 * Command Line:: Command Line
1985 * Input Files:: Input Files
1986 * Object:: Output (Object) File
1987 * Errors:: Error and Warning Messages
1991 @section Structure of this Manual
1993 @cindex manual, structure and purpose
1994 This manual is intended to describe what you need to know to use
1995 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1996 notation for symbols, constants, and expressions; the directives that
1997 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
2000 We also cover special features in the @value{TARGET}
2001 configuration of @command{@value{AS}}, including assembler directives.
2004 This manual also describes some of the machine-dependent features of
2005 various flavors of the assembler.
2008 @cindex machine instructions (not covered)
2009 On the other hand, this manual is @emph{not} intended as an introduction
2010 to programming in assembly language---let alone programming in general!
2011 In a similar vein, we make no attempt to introduce the machine
2012 architecture; we do @emph{not} describe the instruction set, standard
2013 mnemonics, registers or addressing modes that are standard to a
2014 particular architecture.
2016 You may want to consult the manufacturer's
2017 machine architecture manual for this information.
2021 For information on the H8/300 machine instruction set, see @cite{H8/300
2022 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
2023 Programming Manual} (Renesas).
2026 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
2027 see @cite{SH-Microcomputer User's Manual} (Renesas) or
2028 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
2029 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
2032 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
2036 @c I think this is premature---doc@cygnus.com, 17jan1991
2038 Throughout this manual, we assume that you are running @dfn{GNU},
2039 the portable operating system from the @dfn{Free Software
2040 Foundation, Inc.}. This restricts our attention to certain kinds of
2041 computer (in particular, the kinds of computers that @sc{gnu} can run on);
2042 once this assumption is granted examples and definitions need less
2045 @command{@value{AS}} is part of a team of programs that turn a high-level
2046 human-readable series of instructions into a low-level
2047 computer-readable series of instructions. Different versions of
2048 @command{@value{AS}} are used for different kinds of computer.
2051 @c There used to be a section "Terminology" here, which defined
2052 @c "contents", "byte", "word", and "long". Defining "word" to any
2053 @c particular size is confusing when the .word directive may generate 16
2054 @c bits on one machine and 32 bits on another; in general, for the user
2055 @c version of this manual, none of these terms seem essential to define.
2056 @c They were used very little even in the former draft of the manual;
2057 @c this draft makes an effort to avoid them (except in names of
2061 @section The GNU Assembler
2063 @c man begin DESCRIPTION
2065 @sc{gnu} @command{as} is really a family of assemblers.
2067 This manual describes @command{@value{AS}}, a member of that family which is
2068 configured for the @value{TARGET} architectures.
2070 If you use (or have used) the @sc{gnu} assembler on one architecture, you
2071 should find a fairly similar environment when you use it on another
2072 architecture. Each version has much in common with the others,
2073 including object file formats, most assembler directives (often called
2074 @dfn{pseudo-ops}) and assembler syntax.@refill
2076 @cindex purpose of @sc{gnu} assembler
2077 @command{@value{AS}} is primarily intended to assemble the output of the
2078 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
2079 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
2080 assemble correctly everything that other assemblers for the same
2081 machine would assemble.
2083 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
2086 @c This remark should appear in generic version of manual; assumption
2087 @c here is that generic version sets M680x0.
2088 This doesn't mean @command{@value{AS}} always uses the same syntax as another
2089 assembler for the same architecture; for example, we know of several
2090 incompatible versions of 680x0 assembly language syntax.
2095 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
2096 program in one pass of the source file. This has a subtle impact on the
2097 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
2099 @node Object Formats
2100 @section Object File Formats
2102 @cindex object file format
2103 The @sc{gnu} assembler can be configured to produce several alternative
2104 object file formats. For the most part, this does not affect how you
2105 write assembly language programs; but directives for debugging symbols
2106 are typically different in different file formats. @xref{Symbol
2107 Attributes,,Symbol Attributes}.
2110 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
2111 @value{OBJ-NAME} format object files.
2113 @c The following should exhaust all configs that set MULTI-OBJ, ideally
2115 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
2116 SOM or ELF format object files.
2121 @section Command Line
2123 @cindex command line conventions
2125 After the program name @command{@value{AS}}, the command line may contain
2126 options and file names. Options may appear in any order, and may be
2127 before, after, or between file names. The order of file names is
2130 @cindex standard input, as input file
2132 @file{--} (two hyphens) by itself names the standard input file
2133 explicitly, as one of the files for @command{@value{AS}} to assemble.
2135 @cindex options, command line
2136 Except for @samp{--} any command-line argument that begins with a
2137 hyphen (@samp{-}) is an option. Each option changes the behavior of
2138 @command{@value{AS}}. No option changes the way another option works. An
2139 option is a @samp{-} followed by one or more letters; the case of
2140 the letter is important. All options are optional.
2142 Some options expect exactly one file name to follow them. The file
2143 name may either immediately follow the option's letter (compatible
2144 with older assemblers) or it may be the next command argument (@sc{gnu}
2145 standard). These two command lines are equivalent:
2148 @value{AS} -o my-object-file.o mumble.s
2149 @value{AS} -omy-object-file.o mumble.s
2153 @section Input Files
2156 @cindex source program
2157 @cindex files, input
2158 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
2159 describe the program input to one run of @command{@value{AS}}. The program may
2160 be in one or more files; how the source is partitioned into files
2161 doesn't change the meaning of the source.
2163 @c I added "con" prefix to "catenation" just to prove I can overcome my
2164 @c APL training... doc@cygnus.com
2165 The source program is a concatenation of the text in all the files, in the
2168 @c man begin DESCRIPTION
2169 Each time you run @command{@value{AS}} it assembles exactly one source
2170 program. The source program is made up of one or more files.
2171 (The standard input is also a file.)
2173 You give @command{@value{AS}} a command line that has zero or more input file
2174 names. The input files are read (from left file name to right). A
2175 command-line argument (in any position) that has no special meaning
2176 is taken to be an input file name.
2178 If you give @command{@value{AS}} no file names it attempts to read one input file
2179 from the @command{@value{AS}} standard input, which is normally your terminal. You
2180 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
2183 Use @samp{--} if you need to explicitly name the standard input file
2184 in your command line.
2186 If the source is empty, @command{@value{AS}} produces a small, empty object
2191 @subheading Filenames and Line-numbers
2193 @cindex input file linenumbers
2194 @cindex line numbers, in input files
2195 There are two ways of locating a line in the input file (or files) and
2196 either may be used in reporting error messages. One way refers to a line
2197 number in a physical file; the other refers to a line number in a
2198 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
2200 @dfn{Physical files} are those files named in the command line given
2201 to @command{@value{AS}}.
2203 @dfn{Logical files} are simply names declared explicitly by assembler
2204 directives; they bear no relation to physical files. Logical file names help
2205 error messages reflect the original source file, when @command{@value{AS}} source
2206 is itself synthesized from other files. @command{@value{AS}} understands the
2207 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
2208 @ref{File,,@code{.file}}.
2211 @section Output (Object) File
2217 Every time you run @command{@value{AS}} it produces an output file, which is
2218 your assembly language program translated into numbers. This file
2219 is the object file. Its default name is @code{a.out}.
2220 You can give it another name by using the @option{-o} option. Conventionally,
2221 object file names end with @file{.o}. The default name is used for historical
2222 reasons: older assemblers were capable of assembling self-contained programs
2223 directly into a runnable program. (For some formats, this isn't currently
2224 possible, but it can be done for the @code{a.out} format.)
2228 The object file is meant for input to the linker @code{@value{LD}}. It contains
2229 assembled program code, information to help @code{@value{LD}} integrate
2230 the assembled program into a runnable file, and (optionally) symbolic
2231 information for the debugger.
2233 @c link above to some info file(s) like the description of a.out.
2234 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
2237 @section Error and Warning Messages
2239 @c man begin DESCRIPTION
2241 @cindex error messages
2242 @cindex warning messages
2243 @cindex messages from assembler
2244 @command{@value{AS}} may write warnings and error messages to the standard error
2245 file (usually your terminal). This should not happen when a compiler
2246 runs @command{@value{AS}} automatically. Warnings report an assumption made so
2247 that @command{@value{AS}} could keep assembling a flawed program; errors report a
2248 grave problem that stops the assembly.
2252 @cindex format of warning messages
2253 Warning messages have the format
2256 file_name:@b{NNN}:Warning Message Text
2260 @cindex file names and line numbers, in warnings/errors
2261 (where @b{NNN} is a line number). If both a logical file name
2262 (@pxref{File,,@code{.file}}) and a logical line number
2264 (@pxref{Line,,@code{.line}})
2266 have been given then they will be used, otherwise the file name and line number
2267 in the current assembler source file will be used. The message text is
2268 intended to be self explanatory (in the grand Unix tradition).
2270 Note the file name must be set via the logical version of the @code{.file}
2271 directive, not the DWARF2 version of the @code{.file} directive. For example:
2275 error_assembler_source
2281 produces this output:
2285 asm.s:2: Error: no such instruction: `error_assembler_source'
2286 foo.c:31: Error: no such instruction: `error_c_source'
2289 @cindex format of error messages
2290 Error messages have the format
2293 file_name:@b{NNN}:FATAL:Error Message Text
2296 The file name and line number are derived as for warning
2297 messages. The actual message text may be rather less explanatory
2298 because many of them aren't supposed to happen.
2301 @chapter Command-Line Options
2303 @cindex options, all versions of assembler
2304 This chapter describes command-line options available in @emph{all}
2305 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
2306 for options specific
2308 to the @value{TARGET} target.
2311 to particular machine architectures.
2314 @c man begin DESCRIPTION
2316 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
2317 you can use the @samp{-Wa} option to pass arguments through to the assembler.
2318 The assembler arguments must be separated from each other (and the @samp{-Wa})
2319 by commas. For example:
2322 gcc -c -g -O -Wa,-alh,-L file.c
2326 This passes two options to the assembler: @samp{-alh} (emit a listing to
2327 standard output with high-level and assembly source) and @samp{-L} (retain
2328 local symbols in the symbol table).
2330 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
2331 command-line options are automatically passed to the assembler by the compiler.
2332 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
2333 precisely what options it passes to each compilation pass, including the
2339 * a:: -a[cdghlns] enable listings
2340 * alternate:: --alternate enable alternate macro syntax
2341 * D:: -D for compatibility
2342 * f:: -f to work faster
2343 * I:: -I for .include search path
2344 @ifclear DIFF-TBL-KLUGE
2345 * K:: -K for compatibility
2347 @ifset DIFF-TBL-KLUGE
2348 * K:: -K for difference tables
2351 * L:: -L to retain local symbols
2352 * listing:: --listing-XXX to configure listing output
2353 * M:: -M or --mri to assemble in MRI compatibility mode
2354 * MD:: --MD for dependency tracking
2355 * no-pad-sections:: --no-pad-sections to stop section padding
2356 * o:: -o to name the object file
2357 * R:: -R to join data and text sections
2358 * statistics:: --statistics to see statistics about assembly
2359 * traditional-format:: --traditional-format for compatible output
2360 * v:: -v to announce version
2361 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
2362 * Z:: -Z to make object file even after errors
2366 @section Enable Listings: @option{-a[cdghlns]}
2376 @cindex listings, enabling
2377 @cindex assembly listings, enabling
2379 These options enable listing output from the assembler. By itself,
2380 @samp{-a} requests high-level, assembly, and symbols listing.
2381 You can use other letters to select specific options for the list:
2382 @samp{-ah} requests a high-level language listing,
2383 @samp{-al} requests an output-program assembly listing, and
2384 @samp{-as} requests a symbol table listing.
2385 High-level listings require that a compiler debugging option like
2386 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
2389 Use the @samp{-ag} option to print a first section with general assembly
2390 information, like @value{AS} version, switches passed, or time stamp.
2392 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
2393 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
2394 other conditional), or a true @code{.if} followed by an @code{.else}, will be
2395 omitted from the listing.
2397 Use the @samp{-ad} option to omit debugging directives from the
2400 Once you have specified one of these options, you can further control
2401 listing output and its appearance using the directives @code{.list},
2402 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
2404 The @samp{-an} option turns off all forms processing.
2405 If you do not request listing output with one of the @samp{-a} options, the
2406 listing-control directives have no effect.
2408 The letters after @samp{-a} may be combined into one option,
2409 @emph{e.g.}, @samp{-aln}.
2411 Note if the assembler source is coming from the standard input (e.g.,
2413 is being created by @code{@value{GCC}} and the @samp{-pipe} command-line switch
2414 is being used) then the listing will not contain any comments or preprocessor
2415 directives. This is because the listing code buffers input source lines from
2416 stdin only after they have been preprocessed by the assembler. This reduces
2417 memory usage and makes the code more efficient.
2420 @section @option{--alternate}
2423 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
2426 @section @option{-D}
2429 This option has no effect whatsoever, but it is accepted to make it more
2430 likely that scripts written for other assemblers also work with
2431 @command{@value{AS}}.
2434 @section Work Faster: @option{-f}
2437 @cindex trusted compiler
2438 @cindex faster processing (@option{-f})
2439 @samp{-f} should only be used when assembling programs written by a
2440 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
2441 and comment preprocessing on
2442 the input file(s) before assembling them. @xref{Preprocessing,
2446 @emph{Warning:} if you use @samp{-f} when the files actually need to be
2447 preprocessed (if they contain comments, for example), @command{@value{AS}} does
2452 @section @code{.include} Search Path: @option{-I} @var{path}
2454 @kindex -I @var{path}
2455 @cindex paths for @code{.include}
2456 @cindex search path for @code{.include}
2457 @cindex @code{include} directive search path
2458 Use this option to add a @var{path} to the list of directories
2459 @command{@value{AS}} searches for files specified in @code{.include}
2460 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
2461 many times as necessary to include a variety of paths. The current
2462 working directory is always searched first; after that, @command{@value{AS}}
2463 searches any @samp{-I} directories in the same order as they were
2464 specified (left to right) on the command line.
2467 @section Difference Tables: @option{-K}
2470 @ifclear DIFF-TBL-KLUGE
2471 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2472 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2473 where it can be used to warn when the assembler alters the machine code
2474 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2475 family does not have the addressing limitations that sometimes lead to this
2476 alteration on other platforms.
2479 @ifset DIFF-TBL-KLUGE
2480 @cindex difference tables, warning
2481 @cindex warning for altered difference tables
2482 @command{@value{AS}} sometimes alters the code emitted for directives of the
2483 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2484 You can use the @samp{-K} option if you want a warning issued when this
2489 @section Include Local Symbols: @option{-L}
2492 @cindex local symbols, retaining in output
2493 Symbols beginning with system-specific local label prefixes, typically
2494 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2495 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2496 such symbols when debugging, because they are intended for the use of
2497 programs (like compilers) that compose assembler programs, not for your
2498 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2499 such symbols, so you do not normally debug with them.
2501 This option tells @command{@value{AS}} to retain those local symbols
2502 in the object file. Usually if you do this you also tell the linker
2503 @code{@value{LD}} to preserve those symbols.
2506 @section Configuring listing output: @option{--listing}
2508 The listing feature of the assembler can be enabled via the command-line switch
2509 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2510 hex dump of the corresponding locations in the output object file, and displays
2511 them as a listing file. The format of this listing can be controlled by
2512 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2513 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2514 @code{.psize} (@pxref{Psize}), and
2515 @code{.eject} (@pxref{Eject}) and also by the following switches:
2518 @item --listing-lhs-width=@samp{number}
2519 @kindex --listing-lhs-width
2520 @cindex Width of first line disassembly output
2521 Sets the maximum width, in words, of the first line of the hex byte dump. This
2522 dump appears on the left hand side of the listing output.
2524 @item --listing-lhs-width2=@samp{number}
2525 @kindex --listing-lhs-width2
2526 @cindex Width of continuation lines of disassembly output
2527 Sets the maximum width, in words, of any further lines of the hex byte dump for
2528 a given input source line. If this value is not specified, it defaults to being
2529 the same as the value specified for @samp{--listing-lhs-width}. If neither
2530 switch is used the default is to one.
2532 @item --listing-rhs-width=@samp{number}
2533 @kindex --listing-rhs-width
2534 @cindex Width of source line output
2535 Sets the maximum width, in characters, of the source line that is displayed
2536 alongside the hex dump. The default value for this parameter is 100. The
2537 source line is displayed on the right hand side of the listing output.
2539 @item --listing-cont-lines=@samp{number}
2540 @kindex --listing-cont-lines
2541 @cindex Maximum number of continuation lines
2542 Sets the maximum number of continuation lines of hex dump that will be
2543 displayed for a given single line of source input. The default value is 4.
2547 @section Assemble in MRI Compatibility Mode: @option{-M}
2550 @cindex MRI compatibility mode
2551 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2552 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2553 compatible with the @code{ASM68K} assembler from Microtec Research.
2554 The exact nature of the
2555 MRI syntax will not be documented here; see the MRI manuals for more
2556 information. Note in particular that the handling of macros and macro
2557 arguments is somewhat different. The purpose of this option is to permit
2558 assembling existing MRI assembler code using @command{@value{AS}}.
2560 The MRI compatibility is not complete. Certain operations of the MRI assembler
2561 depend upon its object file format, and can not be supported using other object
2562 file formats. Supporting these would require enhancing each object file format
2563 individually. These are:
2566 @item global symbols in common section
2568 The m68k MRI assembler supports common sections which are merged by the linker.
2569 Other object file formats do not support this. @command{@value{AS}} handles
2570 common sections by treating them as a single common symbol. It permits local
2571 symbols to be defined within a common section, but it can not support global
2572 symbols, since it has no way to describe them.
2574 @item complex relocations
2576 The MRI assemblers support relocations against a negated section address, and
2577 relocations which combine the start addresses of two or more sections. These
2578 are not support by other object file formats.
2580 @item @code{END} pseudo-op specifying start address
2582 The MRI @code{END} pseudo-op permits the specification of a start address.
2583 This is not supported by other object file formats. The start address may
2584 instead be specified using the @option{-e} option to the linker, or in a linker
2587 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2589 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2590 name to the output file. This is not supported by other object file formats.
2592 @item @code{ORG} pseudo-op
2594 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2595 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2596 which changes the location within the current section. Absolute sections are
2597 not supported by other object file formats. The address of a section may be
2598 assigned within a linker script.
2601 There are some other features of the MRI assembler which are not supported by
2602 @command{@value{AS}}, typically either because they are difficult or because they
2603 seem of little consequence. Some of these may be supported in future releases.
2607 @item EBCDIC strings
2609 EBCDIC strings are not supported.
2611 @item packed binary coded decimal
2613 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2614 and @code{DCB.P} pseudo-ops are not supported.
2616 @item @code{FEQU} pseudo-op
2618 The m68k @code{FEQU} pseudo-op is not supported.
2620 @item @code{NOOBJ} pseudo-op
2622 The m68k @code{NOOBJ} pseudo-op is not supported.
2624 @item @code{OPT} branch control options
2626 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2627 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2628 relaxes all branches, whether forward or backward, to an appropriate size, so
2629 these options serve no purpose.
2631 @item @code{OPT} list control options
2633 The following m68k @code{OPT} list control options are ignored: @code{C},
2634 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2635 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2637 @item other @code{OPT} options
2639 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2640 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2642 @item @code{OPT} @code{D} option is default
2644 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2645 @code{OPT NOD} may be used to turn it off.
2647 @item @code{XREF} pseudo-op.
2649 The m68k @code{XREF} pseudo-op is ignored.
2654 @section Dependency Tracking: @option{--MD}
2657 @cindex dependency tracking
2660 @command{@value{AS}} can generate a dependency file for the file it creates. This
2661 file consists of a single rule suitable for @code{make} describing the
2662 dependencies of the main source file.
2664 The rule is written to the file named in its argument.
2666 This feature is used in the automatic updating of makefiles.
2668 @node no-pad-sections
2669 @section Output Section Padding
2670 @kindex --no-pad-sections
2671 @cindex output section padding
2672 Normally the assembler will pad the end of each output section up to its
2673 alignment boundary. But this can waste space, which can be significant on
2674 memory constrained targets. So the @option{--no-pad-sections} option will
2675 disable this behaviour.
2678 @section Name the Object File: @option{-o}
2681 @cindex naming object file
2682 @cindex object file name
2683 There is always one object file output when you run @command{@value{AS}}. By
2684 default it has the name @file{a.out}.
2685 You use this option (which takes exactly one filename) to give the
2686 object file a different name.
2688 Whatever the object file is called, @command{@value{AS}} overwrites any
2689 existing file of the same name.
2692 @section Join Data and Text Sections: @option{-R}
2695 @cindex data and text sections, joining
2696 @cindex text and data sections, joining
2697 @cindex joining text and data sections
2698 @cindex merging text and data sections
2699 @option{-R} tells @command{@value{AS}} to write the object file as if all
2700 data-section data lives in the text section. This is only done at
2701 the very last moment: your binary data are the same, but data
2702 section parts are relocated differently. The data section part of
2703 your object file is zero bytes long because all its bytes are
2704 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2706 When you specify @option{-R} it would be possible to generate shorter
2707 address displacements (because we do not have to cross between text and
2708 data section). We refrain from doing this simply for compatibility with
2709 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2712 When @command{@value{AS}} is configured for COFF or ELF output,
2713 this option is only useful if you use sections named @samp{.text} and
2718 @option{-R} is not supported for any of the HPPA targets. Using
2719 @option{-R} generates a warning from @command{@value{AS}}.
2723 @section Display Assembly Statistics: @option{--statistics}
2725 @kindex --statistics
2726 @cindex statistics, about assembly
2727 @cindex time, total for assembly
2728 @cindex space used, maximum for assembly
2729 Use @samp{--statistics} to display two statistics about the resources used by
2730 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2731 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2734 @node traditional-format
2735 @section Compatible Output: @option{--traditional-format}
2737 @kindex --traditional-format
2738 For some targets, the output of @command{@value{AS}} is different in some ways
2739 from the output of some existing assembler. This switch requests
2740 @command{@value{AS}} to use the traditional format instead.
2742 For example, it disables the exception frame optimizations which
2743 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2746 @section Announce Version: @option{-v}
2750 @cindex assembler version
2751 @cindex version of assembler
2752 You can find out what version of as is running by including the
2753 option @samp{-v} (which you can also spell as @samp{-version}) on the
2757 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2759 @command{@value{AS}} should never give a warning or error message when
2760 assembling compiler output. But programs written by people often
2761 cause @command{@value{AS}} to give a warning that a particular assumption was
2762 made. All such warnings are directed to the standard error file.
2766 @cindex suppressing warnings
2767 @cindex warnings, suppressing
2768 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2769 This only affects the warning messages: it does not change any particular of
2770 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2773 @kindex --fatal-warnings
2774 @cindex errors, caused by warnings
2775 @cindex warnings, causing error
2776 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2777 files that generate warnings to be in error.
2780 @cindex warnings, switching on
2781 You can switch these options off again by specifying @option{--warn}, which
2782 causes warnings to be output as usual.
2785 @section Generate Object File in Spite of Errors: @option{-Z}
2786 @cindex object file, after errors
2787 @cindex errors, continuing after
2788 After an error message, @command{@value{AS}} normally produces no output. If for
2789 some reason you are interested in object file output even after
2790 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2791 option. If there are any errors, @command{@value{AS}} continues anyways, and
2792 writes an object file after a final warning message of the form @samp{@var{n}
2793 errors, @var{m} warnings, generating bad object file.}
2798 @cindex machine-independent syntax
2799 @cindex syntax, machine-independent
2800 This chapter describes the machine-independent syntax allowed in a
2801 source file. @command{@value{AS}} syntax is similar to what many other
2802 assemblers use; it is inspired by the BSD 4.2
2807 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2811 * Preprocessing:: Preprocessing
2812 * Whitespace:: Whitespace
2813 * Comments:: Comments
2814 * Symbol Intro:: Symbols
2815 * Statements:: Statements
2816 * Constants:: Constants
2820 @section Preprocessing
2822 @cindex preprocessing
2823 The @command{@value{AS}} internal preprocessor:
2825 @cindex whitespace, removed by preprocessor
2827 adjusts and removes extra whitespace. It leaves one space or tab before
2828 the keywords on a line, and turns any other whitespace on the line into
2831 @cindex comments, removed by preprocessor
2833 removes all comments, replacing them with a single space, or an
2834 appropriate number of newlines.
2836 @cindex constants, converted by preprocessor
2838 converts character constants into the appropriate numeric values.
2841 It does not do macro processing, include file handling, or
2842 anything else you may get from your C compiler's preprocessor. You can
2843 do include file processing with the @code{.include} directive
2844 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2845 to get other ``CPP'' style preprocessing by giving the input file a
2846 @samp{.S} suffix. @url{https://gcc.gnu.org/onlinedocs/gcc/Overall-Options.html#Overall-Options,
2847 See the 'Options Controlling the Kind of Output' section of the GCC manual for
2850 Excess whitespace, comments, and character constants
2851 cannot be used in the portions of the input text that are not
2854 @cindex turning preprocessing on and off
2855 @cindex preprocessing, turning on and off
2858 If the first line of an input file is @code{#NO_APP} or if you use the
2859 @samp{-f} option, whitespace and comments are not removed from the input file.
2860 Within an input file, you can ask for whitespace and comment removal in
2861 specific portions of the by putting a line that says @code{#APP} before the
2862 text that may contain whitespace or comments, and putting a line that says
2863 @code{#NO_APP} after this text. This feature is mainly intend to support
2864 @code{asm} statements in compilers whose output is otherwise free of comments
2871 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2872 Whitespace is used to separate symbols, and to make programs neater for
2873 people to read. Unless within character constants
2874 (@pxref{Characters,,Character Constants}), any whitespace means the same
2875 as exactly one space.
2881 There are two ways of rendering comments to @command{@value{AS}}. In both
2882 cases the comment is equivalent to one space.
2884 Anything from @samp{/*} through the next @samp{*/} is a comment.
2885 This means you may not nest these comments.
2889 The only way to include a newline ('\n') in a comment
2890 is to use this sort of comment.
2893 /* This sort of comment does not nest. */
2896 @cindex line comment character
2897 Anything from a @dfn{line comment} character up to the next newline is
2898 considered a comment and is ignored. The line comment character is target
2899 specific, and some targets multiple comment characters. Some targets also have
2900 line comment characters that only work if they are the first character on a
2901 line. Some targets use a sequence of two characters to introduce a line
2902 comment. Some targets can also change their line comment characters depending
2903 upon command-line options that have been used. For more details see the
2904 @emph{Syntax} section in the documentation for individual targets.
2906 If the line comment character is the hash sign (@samp{#}) then it still has the
2907 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2908 to specify logical line numbers:
2911 @cindex lines starting with @code{#}
2912 @cindex logical line numbers
2913 To be compatible with past assemblers, lines that begin with @samp{#} have a
2914 special interpretation. Following the @samp{#} should be an absolute
2915 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2916 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2917 new logical file name. The rest of the line, if any, should be whitespace.
2919 If the first non-whitespace characters on the line are not numeric,
2920 the line is ignored. (Just like a comment.)
2923 # This is an ordinary comment.
2924 # 42-6 "new_file_name" # New logical file name
2925 # This is logical line # 36.
2927 This feature is deprecated, and may disappear from future versions
2928 of @command{@value{AS}}.
2933 @cindex characters used in symbols
2934 @ifclear SPECIAL-SYMS
2935 A @dfn{symbol} is one or more characters chosen from the set of all
2936 letters (both upper and lower case), digits and the three characters
2942 A @dfn{symbol} is one or more characters chosen from the set of all
2943 letters (both upper and lower case), digits and the three characters
2944 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2950 On most machines, you can also use @code{$} in symbol names; exceptions
2951 are noted in @ref{Machine Dependencies}.
2953 No symbol may begin with a digit. Case is significant.
2954 There is no length limit; all characters are significant. Multibyte characters
2955 are supported. Symbols are delimited by characters not in that set, or by the
2956 beginning of a file (since the source program must end with a newline, the end
2957 of a file is not a possible symbol delimiter). @xref{Symbols}.
2959 Symbol names may also be enclosed in double quote @code{"} characters. In such
2960 cases any characters are allowed, except for the NUL character. If a double
2961 quote character is to be included in the symbol name it must be preceeded by a
2962 backslash @code{\} character.
2963 @cindex length of symbols
2968 @cindex statements, structure of
2969 @cindex line separator character
2970 @cindex statement separator character
2972 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2973 @dfn{line separator character}. The line separator character is target
2974 specific and described in the @emph{Syntax} section of each
2975 target's documentation. Not all targets support a line separator character.
2976 The newline or line separator character is considered to be part of the
2977 preceding statement. Newlines and separators within character constants are an
2978 exception: they do not end statements.
2980 @cindex newline, required at file end
2981 @cindex EOF, newline must precede
2982 It is an error to end any statement with end-of-file: the last
2983 character of any input file should be a newline.@refill
2985 An empty statement is allowed, and may include whitespace. It is ignored.
2987 @cindex instructions and directives
2988 @cindex directives and instructions
2989 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2990 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2992 A statement begins with zero or more labels, optionally followed by a
2993 key symbol which determines what kind of statement it is. The key
2994 symbol determines the syntax of the rest of the statement. If the
2995 symbol begins with a dot @samp{.} then the statement is an assembler
2996 directive: typically valid for any computer. If the symbol begins with
2997 a letter the statement is an assembly language @dfn{instruction}: it
2998 assembles into a machine language instruction.
3000 Different versions of @command{@value{AS}} for different computers
3001 recognize different instructions. In fact, the same symbol may
3002 represent a different instruction in a different computer's assembly
3006 @cindex @code{:} (label)
3007 @cindex label (@code{:})
3008 A label is a symbol immediately followed by a colon (@code{:}).
3009 Whitespace before a label or after a colon is permitted, but you may not
3010 have whitespace between a label's symbol and its colon. @xref{Labels}.
3013 For HPPA targets, labels need not be immediately followed by a colon, but
3014 the definition of a label must begin in column zero. This also implies that
3015 only one label may be defined on each line.
3019 label: .directive followed by something
3020 another_label: # This is an empty statement.
3021 instruction operand_1, operand_2, @dots{}
3028 A constant is a number, written so that its value is known by
3029 inspection, without knowing any context. Like this:
3032 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
3033 .ascii "Ring the bell\7" # A string constant.
3034 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
3035 .float 0f-314159265358979323846264338327\
3036 95028841971.693993751E-40 # - pi, a flonum.
3041 * Characters:: Character Constants
3042 * Numbers:: Number Constants
3046 @subsection Character Constants
3048 @cindex character constants
3049 @cindex constants, character
3050 There are two kinds of character constants. A @dfn{character} stands
3051 for one character in one byte and its value may be used in
3052 numeric expressions. String constants (properly called string
3053 @emph{literals}) are potentially many bytes and their values may not be
3054 used in arithmetic expressions.
3058 * Chars:: Characters
3062 @subsubsection Strings
3064 @cindex string constants
3065 @cindex constants, string
3066 A @dfn{string} is written between double-quotes. It may contain
3067 double-quotes or null characters. The way to get special characters
3068 into a string is to @dfn{escape} these characters: precede them with
3069 a backslash @samp{\} character. For example @samp{\\} represents
3070 one backslash: the first @code{\} is an escape which tells
3071 @command{@value{AS}} to interpret the second character literally as a backslash
3072 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
3073 escape character). The complete list of escapes follows.
3075 @cindex escape codes, character
3076 @cindex character escape codes
3077 @c NOTE: Cindex entries must not start with a backlash character.
3078 @c NOTE: This confuses the pdf2texi script when it is creating the
3079 @c NOTE: index based upon the first character and so it generates:
3080 @c NOTE: \initial {\\}
3081 @c NOTE: which then results in the error message:
3082 @c NOTE: Argument of \\ has an extra }.
3083 @c NOTE: So in the index entries below a space character has been
3084 @c NOTE: prepended to avoid this problem.
3087 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
3089 @cindex @code{ \b} (backspace character)
3090 @cindex backspace (@code{\b})
3092 Mnemonic for backspace; for ASCII this is octal code 010.
3095 @c Mnemonic for EOText; for ASCII this is octal code 004.
3097 @cindex @code{ \f} (formfeed character)
3098 @cindex formfeed (@code{\f})
3100 Mnemonic for FormFeed; for ASCII this is octal code 014.
3102 @cindex @code{ \n} (newline character)
3103 @cindex newline (@code{\n})
3105 Mnemonic for newline; for ASCII this is octal code 012.
3108 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
3110 @cindex @code{ \r} (carriage return character)
3111 @cindex carriage return (@code{backslash-r})
3113 Mnemonic for carriage-Return; for ASCII this is octal code 015.
3116 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
3117 @c other assemblers.
3119 @cindex @code{ \t} (tab)
3120 @cindex tab (@code{\t})
3122 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
3125 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
3126 @c @item \x @var{digit} @var{digit} @var{digit}
3127 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
3129 @cindex @code{ \@var{ddd}} (octal character code)
3130 @cindex octal character code (@code{\@var{ddd}})
3131 @item \ @var{digit} @var{digit} @var{digit}
3132 An octal character code. The numeric code is 3 octal digits.
3133 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
3134 for example, @code{\008} has the value 010, and @code{\009} the value 011.
3136 @cindex @code{ \@var{xd...}} (hex character code)
3137 @cindex hex character code (@code{\@var{xd...}})
3138 @item \@code{x} @var{hex-digits...}
3139 A hex character code. All trailing hex digits are combined. Either upper or
3140 lower case @code{x} works.
3142 @cindex @code{ \\} (@samp{\} character)
3143 @cindex backslash (@code{\\})
3145 Represents one @samp{\} character.
3148 @c Represents one @samp{'} (accent acute) character.
3149 @c This is needed in single character literals
3150 @c (@xref{Characters,,Character Constants}.) to represent
3153 @cindex @code{ \"} (doublequote character)
3154 @cindex doublequote (@code{\"})
3156 Represents one @samp{"} character. Needed in strings to represent
3157 this character, because an unescaped @samp{"} would end the string.
3159 @item \ @var{anything-else}
3160 Any other character when escaped by @kbd{\} gives a warning, but
3161 assembles as if the @samp{\} was not present. The idea is that if
3162 you used an escape sequence you clearly didn't want the literal
3163 interpretation of the following character. However @command{@value{AS}} has no
3164 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
3165 code and warns you of the fact.
3168 Which characters are escapable, and what those escapes represent,
3169 varies widely among assemblers. The current set is what we think
3170 the BSD 4.2 assembler recognizes, and is a subset of what most C
3171 compilers recognize. If you are in doubt, do not use an escape
3175 @subsubsection Characters
3177 @cindex single character constant
3178 @cindex character, single
3179 @cindex constant, single character
3180 A single character may be written as a single quote immediately followed by
3181 that character. Some backslash escapes apply to characters, @code{\b},
3182 @code{\f}, @code{\n}, @code{\r}, @code{\t}, and @code{\"} with the same meaning
3183 as for strings, plus @code{\'} for a single quote. So if you want to write the
3184 character backslash, you must write @kbd{'\\} where the first @code{\} escapes
3185 the second @code{\}. As you can see, the quote is an acute accent, not a grave
3188 @ifclear abnormal-separator
3189 (or semicolon @samp{;})
3191 @ifset abnormal-separator
3193 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
3198 immediately following an acute accent is taken as a literal character
3199 and does not count as the end of a statement. The value of a character
3200 constant in a numeric expression is the machine's byte-wide code for
3201 that character. @command{@value{AS}} assumes your character code is ASCII:
3202 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
3205 @subsection Number Constants
3207 @cindex constants, number
3208 @cindex number constants
3209 @command{@value{AS}} distinguishes three kinds of numbers according to how they
3210 are stored in the target machine. @emph{Integers} are numbers that
3211 would fit into an @code{int} in the C language. @emph{Bignums} are
3212 integers, but they are stored in more than 32 bits. @emph{Flonums}
3213 are floating point numbers, described below.
3216 * Integers:: Integers
3224 @subsubsection Integers
3226 @cindex constants, integer
3228 @cindex binary integers
3229 @cindex integers, binary
3230 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
3231 the binary digits @samp{01}.
3233 @cindex octal integers
3234 @cindex integers, octal
3235 An octal integer is @samp{0} followed by zero or more of the octal
3236 digits (@samp{01234567}).
3238 @cindex decimal integers
3239 @cindex integers, decimal
3240 A decimal integer starts with a non-zero digit followed by zero or
3241 more digits (@samp{0123456789}).
3243 @cindex hexadecimal integers
3244 @cindex integers, hexadecimal
3245 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
3246 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
3248 Integers have the usual values. To denote a negative integer, use
3249 the prefix operator @samp{-} discussed under expressions
3250 (@pxref{Prefix Ops,,Prefix Operators}).
3253 @subsubsection Bignums
3256 @cindex constants, bignum
3257 A @dfn{bignum} has the same syntax and semantics as an integer
3258 except that the number (or its negative) takes more than 32 bits to
3259 represent in binary. The distinction is made because in some places
3260 integers are permitted while bignums are not.
3263 @subsubsection Flonums
3265 @cindex floating point numbers
3266 @cindex constants, floating point
3268 @cindex precision, floating point
3269 A @dfn{flonum} represents a floating point number. The translation is
3270 indirect: a decimal floating point number from the text is converted by
3271 @command{@value{AS}} to a generic binary floating point number of more than
3272 sufficient precision. This generic floating point number is converted
3273 to a particular computer's floating point format (or formats) by a
3274 portion of @command{@value{AS}} specialized to that computer.
3276 A flonum is written by writing (in order)
3281 (@samp{0} is optional on the HPPA.)
3285 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
3287 @kbd{e} is recommended. Case is not important.
3289 @c FIXME: verify if flonum syntax really this vague for most cases
3290 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
3291 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
3294 On the H8/300 and Renesas / SuperH SH architectures, the letter must be
3295 one of the letters @samp{DFPRSX} (in upper or lower case).
3297 On the ARC, the letter must be one of the letters @samp{DFRS}
3298 (in upper or lower case).
3300 On the HPPA architecture, the letter must be @samp{E} (upper case only).
3304 One of the letters @samp{DFRS} (in upper or lower case).
3307 One of the letters @samp{DFPRSX} (in upper or lower case).
3310 The letter @samp{E} (upper case only).
3315 An optional sign: either @samp{+} or @samp{-}.
3318 An optional @dfn{integer part}: zero or more decimal digits.
3321 An optional @dfn{fractional part}: @samp{.} followed by zero
3322 or more decimal digits.
3325 An optional exponent, consisting of:
3329 An @samp{E} or @samp{e}.
3330 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
3331 @c principle this can perfectly well be different on different targets.
3333 Optional sign: either @samp{+} or @samp{-}.
3335 One or more decimal digits.
3340 At least one of the integer part or the fractional part must be
3341 present. The floating point number has the usual base-10 value.
3343 @command{@value{AS}} does all processing using integers. Flonums are computed
3344 independently of any floating point hardware in the computer running
3345 @command{@value{AS}}.
3348 @chapter Sections and Relocation
3353 * Secs Background:: Background
3354 * Ld Sections:: Linker Sections
3355 * As Sections:: Assembler Internal Sections
3356 * Sub-Sections:: Sub-Sections
3360 @node Secs Background
3363 Roughly, a section is a range of addresses, with no gaps; all data
3364 ``in'' those addresses is treated the same for some particular purpose.
3365 For example there may be a ``read only'' section.
3367 @cindex linker, and assembler
3368 @cindex assembler, and linker
3369 The linker @code{@value{LD}} reads many object files (partial programs) and
3370 combines their contents to form a runnable program. When @command{@value{AS}}
3371 emits an object file, the partial program is assumed to start at address 0.
3372 @code{@value{LD}} assigns the final addresses for the partial program, so that
3373 different partial programs do not overlap. This is actually an
3374 oversimplification, but it suffices to explain how @command{@value{AS}} uses
3377 @code{@value{LD}} moves blocks of bytes of your program to their run-time
3378 addresses. These blocks slide to their run-time addresses as rigid
3379 units; their length does not change and neither does the order of bytes
3380 within them. Such a rigid unit is called a @emph{section}. Assigning
3381 run-time addresses to sections is called @dfn{relocation}. It includes
3382 the task of adjusting mentions of object-file addresses so they refer to
3383 the proper run-time addresses.
3385 For the H8/300, and for the Renesas / SuperH SH,
3386 @command{@value{AS}} pads sections if needed to
3387 ensure they end on a word (sixteen bit) boundary.
3390 @cindex standard assembler sections
3391 An object file written by @command{@value{AS}} has at least three sections, any
3392 of which may be empty. These are named @dfn{text}, @dfn{data} and
3397 When it generates COFF or ELF output,
3399 @command{@value{AS}} can also generate whatever other named sections you specify
3400 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
3401 If you do not use any directives that place output in the @samp{.text}
3402 or @samp{.data} sections, these sections still exist, but are empty.
3407 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
3409 @command{@value{AS}} can also generate whatever other named sections you
3410 specify using the @samp{.space} and @samp{.subspace} directives. See
3411 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3412 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3413 assembler directives.
3416 Additionally, @command{@value{AS}} uses different names for the standard
3417 text, data, and bss sections when generating SOM output. Program text
3418 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3419 BSS into @samp{$BSS$}.
3423 Within the object file, the text section starts at address @code{0}, the
3424 data section follows, and the bss section follows the data section.
3427 When generating either SOM or ELF output files on the HPPA, the text
3428 section starts at address @code{0}, the data section at address
3429 @code{0x4000000}, and the bss section follows the data section.
3432 To let @code{@value{LD}} know which data changes when the sections are
3433 relocated, and how to change that data, @command{@value{AS}} also writes to the
3434 object file details of the relocation needed. To perform relocation
3435 @code{@value{LD}} must know, each time an address in the object
3439 Where in the object file is the beginning of this reference to
3442 How long (in bytes) is this reference?
3444 Which section does the address refer to? What is the numeric value of
3446 (@var{address}) @minus{} (@var{start-address of section})?
3449 Is the reference to an address ``Program-Counter relative''?
3452 @cindex addresses, format of
3453 @cindex section-relative addressing
3454 In fact, every address @command{@value{AS}} ever uses is expressed as
3456 (@var{section}) + (@var{offset into section})
3459 Further, most expressions @command{@value{AS}} computes have this section-relative
3462 (For some object formats, such as SOM for the HPPA, some expressions are
3463 symbol-relative instead.)
3466 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3467 @var{N} into section @var{secname}.''
3469 Apart from text, data and bss sections you need to know about the
3470 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3471 addresses in the absolute section remain unchanged. For example, address
3472 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3473 @code{@value{LD}}. Although the linker never arranges two partial programs'
3474 data sections with overlapping addresses after linking, @emph{by definition}
3475 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3476 part of a program is always the same address when the program is running as
3477 address @code{@{absolute@ 239@}} in any other part of the program.
3479 The idea of sections is extended to the @dfn{undefined} section. Any
3480 address whose section is unknown at assembly time is by definition
3481 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3482 Since numbers are always defined, the only way to generate an undefined
3483 address is to mention an undefined symbol. A reference to a named
3484 common block would be such a symbol: its value is unknown at assembly
3485 time so it has section @emph{undefined}.
3487 By analogy the word @emph{section} is used to describe groups of sections in
3488 the linked program. @code{@value{LD}} puts all partial programs' text
3489 sections in contiguous addresses in the linked program. It is
3490 customary to refer to the @emph{text section} of a program, meaning all
3491 the addresses of all partial programs' text sections. Likewise for
3492 data and bss sections.
3494 Some sections are manipulated by @code{@value{LD}}; others are invented for
3495 use of @command{@value{AS}} and have no meaning except during assembly.
3498 @section Linker Sections
3499 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3504 @cindex named sections
3505 @cindex sections, named
3506 @item named sections
3509 @cindex text section
3510 @cindex data section
3514 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3515 separate but equal sections. Anything you can say of one section is
3518 When the program is running, however, it is
3519 customary for the text section to be unalterable. The
3520 text section is often shared among processes: it contains
3521 instructions, constants and the like. The data section of a running
3522 program is usually alterable: for example, C variables would be stored
3523 in the data section.
3528 This section contains zeroed bytes when your program begins running. It
3529 is used to hold uninitialized variables or common storage. The length of
3530 each partial program's bss section is important, but because it starts
3531 out containing zeroed bytes there is no need to store explicit zero
3532 bytes in the object file. The bss section was invented to eliminate
3533 those explicit zeros from object files.
3535 @cindex absolute section
3536 @item absolute section
3537 Address 0 of this section is always ``relocated'' to runtime address 0.
3538 This is useful if you want to refer to an address that @code{@value{LD}} must
3539 not change when relocating. In this sense we speak of absolute
3540 addresses being ``unrelocatable'': they do not change during relocation.
3542 @cindex undefined section
3543 @item undefined section
3544 This ``section'' is a catch-all for address references to objects not in
3545 the preceding sections.
3546 @c FIXME: ref to some other doc on obj-file formats could go here.
3549 @cindex relocation example
3550 An idealized example of three relocatable sections follows.
3552 The example uses the traditional section names @samp{.text} and @samp{.data}.
3554 Memory addresses are on the horizontal axis.
3558 @c END TEXI2ROFF-KILL
3561 partial program # 1: |ttttt|dddd|00|
3568 partial program # 2: |TTT|DDD|000|
3571 +--+---+-----+--+----+---+-----+~~
3572 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3573 +--+---+-----+--+----+---+-----+~~
3575 addresses: 0 @dots{}
3582 \line{\it Partial program \#1: \hfil}
3583 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3584 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3586 \line{\it Partial program \#2: \hfil}
3587 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3588 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3590 \line{\it linked program: \hfil}
3591 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3592 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3593 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3594 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3596 \line{\it addresses: \hfil}
3600 @c END TEXI2ROFF-KILL
3603 @section Assembler Internal Sections
3605 @cindex internal assembler sections
3606 @cindex sections in messages, internal
3607 These sections are meant only for the internal use of @command{@value{AS}}. They
3608 have no meaning at run-time. You do not really need to know about these
3609 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3610 warning messages, so it might be helpful to have an idea of their
3611 meanings to @command{@value{AS}}. These sections are used to permit the
3612 value of every expression in your assembly language program to be a
3613 section-relative address.
3616 @cindex assembler internal logic error
3617 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3618 An internal assembler logic error has been found. This means there is a
3619 bug in the assembler.
3621 @cindex expr (internal section)
3623 The assembler stores complex expression internally as combinations of
3624 symbols. When it needs to represent an expression as a symbol, it puts
3625 it in the expr section.
3627 @c FIXME item transfer[t] vector preload
3628 @c FIXME item transfer[t] vector postload
3629 @c FIXME item register
3633 @section Sub-Sections
3635 @cindex numbered subsections
3636 @cindex grouping data
3642 fall into two sections: text and data.
3644 You may have separate groups of
3646 data in named sections
3650 data in named sections
3656 that you want to end up near to each other in the object file, even though they
3657 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3658 use @dfn{subsections} for this purpose. Within each section, there can be
3659 numbered subsections with values from 0 to 8192. Objects assembled into the
3660 same subsection go into the object file together with other objects in the same
3661 subsection. For example, a compiler might want to store constants in the text
3662 section, but might not want to have them interspersed with the program being
3663 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3664 section of code being output, and a @samp{.text 1} before each group of
3665 constants being output.
3667 Subsections are optional. If you do not use subsections, everything
3668 goes in subsection number zero.
3671 Each subsection is zero-padded up to a multiple of four bytes.
3672 (Subsections may be padded a different amount on different flavors
3673 of @command{@value{AS}}.)
3677 On the H8/300 platform, each subsection is zero-padded to a word
3678 boundary (two bytes).
3679 The same is true on the Renesas SH.
3683 Subsections appear in your object file in numeric order, lowest numbered
3684 to highest. (All this to be compatible with other people's assemblers.)
3685 The object file contains no representation of subsections; @code{@value{LD}} and
3686 other programs that manipulate object files see no trace of them.
3687 They just see all your text subsections as a text section, and all your
3688 data subsections as a data section.
3690 To specify which subsection you want subsequent statements assembled
3691 into, use a numeric argument to specify it, in a @samp{.text
3692 @var{expression}} or a @samp{.data @var{expression}} statement.
3695 When generating COFF output, you
3700 can also use an extra subsection
3701 argument with arbitrary named sections: @samp{.section @var{name},
3706 When generating ELF output, you
3711 can also use the @code{.subsection} directive (@pxref{SubSection})
3712 to specify a subsection: @samp{.subsection @var{expression}}.
3714 @var{Expression} should be an absolute expression
3715 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3716 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3717 begins in @code{text 0}. For instance:
3719 .text 0 # The default subsection is text 0 anyway.
3720 .ascii "This lives in the first text subsection. *"
3722 .ascii "But this lives in the second text subsection."
3724 .ascii "This lives in the data section,"
3725 .ascii "in the first data subsection."
3727 .ascii "This lives in the first text section,"
3728 .ascii "immediately following the asterisk (*)."
3731 Each section has a @dfn{location counter} incremented by one for every byte
3732 assembled into that section. Because subsections are merely a convenience
3733 restricted to @command{@value{AS}} there is no concept of a subsection location
3734 counter. There is no way to directly manipulate a location counter---but the
3735 @code{.align} directive changes it, and any label definition captures its
3736 current value. The location counter of the section where statements are being
3737 assembled is said to be the @dfn{active} location counter.
3740 @section bss Section
3743 @cindex common variable storage
3744 The bss section is used for local common variable storage.
3745 You may allocate address space in the bss section, but you may
3746 not dictate data to load into it before your program executes. When
3747 your program starts running, all the contents of the bss
3748 section are zeroed bytes.
3750 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3751 @ref{Lcomm,,@code{.lcomm}}.
3753 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3754 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3757 When assembling for a target which supports multiple sections, such as ELF or
3758 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3759 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3760 section. Typically the section will only contain symbol definitions and
3761 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3768 Symbols are a central concept: the programmer uses symbols to name
3769 things, the linker uses symbols to link, and the debugger uses symbols
3773 @cindex debuggers, and symbol order
3774 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3775 the same order they were declared. This may break some debuggers.
3780 * Setting Symbols:: Giving Symbols Other Values
3781 * Symbol Names:: Symbol Names
3782 * Dot:: The Special Dot Symbol
3783 * Symbol Attributes:: Symbol Attributes
3790 A @dfn{label} is written as a symbol immediately followed by a colon
3791 @samp{:}. The symbol then represents the current value of the
3792 active location counter, and is, for example, a suitable instruction
3793 operand. You are warned if you use the same symbol to represent two
3794 different locations: the first definition overrides any other
3798 On the HPPA, the usual form for a label need not be immediately followed by a
3799 colon, but instead must start in column zero. Only one label may be defined on
3800 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3801 provides a special directive @code{.label} for defining labels more flexibly.
3804 @node Setting Symbols
3805 @section Giving Symbols Other Values
3807 @cindex assigning values to symbols
3808 @cindex symbol values, assigning
3809 A symbol can be given an arbitrary value by writing a symbol, followed
3810 by an equals sign @samp{=}, followed by an expression
3811 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3812 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3813 equals sign @samp{=}@samp{=} here represents an equivalent of the
3814 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3817 Blackfin does not support symbol assignment with @samp{=}.
3821 @section Symbol Names
3823 @cindex symbol names
3824 @cindex names, symbol
3825 @ifclear SPECIAL-SYMS
3826 Symbol names begin with a letter or with one of @samp{._}. On most
3827 machines, you can also use @code{$} in symbol names; exceptions are
3828 noted in @ref{Machine Dependencies}. That character may be followed by any
3829 string of digits, letters, dollar signs (unless otherwise noted for a
3830 particular target machine), and underscores.
3834 Symbol names begin with a letter or with one of @samp{._}. On the
3835 Renesas SH you can also use @code{$} in symbol names. That
3836 character may be followed by any string of digits, letters, dollar signs (save
3837 on the H8/300), and underscores.
3841 Case of letters is significant: @code{foo} is a different symbol name
3844 Symbol names do not start with a digit. An exception to this rule is made for
3845 Local Labels. See below.
3847 Multibyte characters are supported. To generate a symbol name containing
3848 multibyte characters enclose it within double quotes and use escape codes. cf
3849 @xref{Strings}. Generating a multibyte symbol name from a label is not
3850 currently supported.
3852 Each symbol has exactly one name. Each name in an assembly language program
3853 refers to exactly one symbol. You may use that symbol name any number of times
3856 @subheading Local Symbol Names
3858 @cindex local symbol names
3859 @cindex symbol names, local
3860 A local symbol is any symbol beginning with certain local label prefixes.
3861 By default, the local label prefix is @samp{.L} for ELF systems or
3862 @samp{L} for traditional a.out systems, but each target may have its own
3863 set of local label prefixes.
3865 On the HPPA local symbols begin with @samp{L$}.
3868 Local symbols are defined and used within the assembler, but they are
3869 normally not saved in object files. Thus, they are not visible when debugging.
3870 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols})
3871 to retain the local symbols in the object files.
3873 @subheading Local Labels
3875 @cindex local labels
3876 @cindex temporary symbol names
3877 @cindex symbol names, temporary
3878 Local labels are different from local symbols. Local labels help compilers and
3879 programmers use names temporarily. They create symbols which are guaranteed to
3880 be unique over the entire scope of the input source code and which can be
3881 referred to by a simple notation. To define a local label, write a label of
3882 the form @samp{@b{N}:} (where @b{N} represents any non-negative integer).
3883 To refer to the most recent previous definition of that label write
3884 @samp{@b{N}b}, using the same number as when you defined the label. To refer
3885 to the next definition of a local label, write @samp{@b{N}f}. The @samp{b}
3886 stands for ``backwards'' and the @samp{f} stands for ``forwards''.
3888 There is no restriction on how you can use these labels, and you can reuse them
3889 too. So that it is possible to repeatedly define the same local label (using
3890 the same number @samp{@b{N}}), although you can only refer to the most recently
3891 defined local label of that number (for a backwards reference) or the next
3892 definition of a specific local label for a forward reference. It is also worth
3893 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3894 implemented in a slightly more efficient manner than the others.
3905 Which is the equivalent of:
3908 label_1: branch label_3
3909 label_2: branch label_1
3910 label_3: branch label_4
3911 label_4: branch label_3
3914 Local label names are only a notational device. They are immediately
3915 transformed into more conventional symbol names before the assembler uses them.
3916 The symbol names are stored in the symbol table, appear in error messages, and
3917 are optionally emitted to the object file. The names are constructed using
3921 @item @emph{local label prefix}
3922 All local symbols begin with the system-specific local label prefix.
3923 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3924 that start with the local label prefix. These labels are
3925 used for symbols you are never intended to see. If you use the
3926 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3927 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3928 you may use them in debugging.
3931 This is the number that was used in the local label definition. So if the
3932 label is written @samp{55:} then the number is @samp{55}.
3935 This unusual character is included so you do not accidentally invent a symbol
3936 of the same name. The character has ASCII value of @samp{\002} (control-B).
3938 @item @emph{ordinal number}
3939 This is a serial number to keep the labels distinct. The first definition of
3940 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3941 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3942 the number @samp{1} and its 15th definition gets @samp{15} as well.
3945 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3946 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3948 @subheading Dollar Local Labels
3949 @cindex dollar local symbols
3951 On some targets @code{@value{AS}} also supports an even more local form of
3952 local labels called dollar labels. These labels go out of scope (i.e., they
3953 become undefined) as soon as a non-local label is defined. Thus they remain
3954 valid for only a small region of the input source code. Normal local labels,
3955 by contrast, remain in scope for the entire file, or until they are redefined
3956 by another occurrence of the same local label.
3958 Dollar labels are defined in exactly the same way as ordinary local labels,
3959 except that they have a dollar sign suffix to their numeric value, e.g.,
3962 They can also be distinguished from ordinary local labels by their transformed
3963 names which use ASCII character @samp{\001} (control-A) as the magic character
3964 to distinguish them from ordinary labels. For example, the fifth definition of
3965 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3968 @section The Special Dot Symbol
3970 @cindex dot (symbol)
3971 @cindex @code{.} (symbol)
3972 @cindex current address
3973 @cindex location counter
3974 The special symbol @samp{.} refers to the current address that
3975 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3976 .long .} defines @code{melvin} to contain its own address.
3977 Assigning a value to @code{.} is treated the same as a @code{.org}
3979 @ifclear no-space-dir
3980 Thus, the expression @samp{.=.+4} is the same as saying
3984 @node Symbol Attributes
3985 @section Symbol Attributes
3987 @cindex symbol attributes
3988 @cindex attributes, symbol
3989 Every symbol has, as well as its name, the attributes ``Value'' and
3990 ``Type''. Depending on output format, symbols can also have auxiliary
3993 The detailed definitions are in @file{a.out.h}.
3996 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3997 all these attributes, and probably won't warn you. This makes the
3998 symbol an externally defined symbol, which is generally what you
4002 * Symbol Value:: Value
4003 * Symbol Type:: Type
4005 * a.out Symbols:: Symbol Attributes: @code{a.out}
4008 * COFF Symbols:: Symbol Attributes for COFF
4011 * SOM Symbols:: Symbol Attributes for SOM
4018 @cindex value of a symbol
4019 @cindex symbol value
4020 The value of a symbol is (usually) 32 bits. For a symbol which labels a
4021 location in the text, data, bss or absolute sections the value is the
4022 number of addresses from the start of that section to the label.
4023 Naturally for text, data and bss sections the value of a symbol changes
4024 as @code{@value{LD}} changes section base addresses during linking. Absolute
4025 symbols' values do not change during linking: that is why they are
4028 The value of an undefined symbol is treated in a special way. If it is
4029 0 then the symbol is not defined in this assembler source file, and
4030 @code{@value{LD}} tries to determine its value from other files linked into the
4031 same program. You make this kind of symbol simply by mentioning a symbol
4032 name without defining it. A non-zero value represents a @code{.comm}
4033 common declaration. The value is how much common storage to reserve, in
4034 bytes (addresses). The symbol refers to the first address of the
4040 @cindex type of a symbol
4042 The type attribute of a symbol contains relocation (section)
4043 information, any flag settings indicating that a symbol is external, and
4044 (optionally), other information for linkers and debuggers. The exact
4045 format depends on the object-code output format in use.
4049 @subsection Symbol Attributes: @code{a.out}
4051 @cindex @code{a.out} symbol attributes
4052 @cindex symbol attributes, @code{a.out}
4055 * Symbol Desc:: Descriptor
4056 * Symbol Other:: Other
4060 @subsubsection Descriptor
4062 @cindex descriptor, of @code{a.out} symbol
4063 This is an arbitrary 16-bit value. You may establish a symbol's
4064 descriptor value by using a @code{.desc} statement
4065 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
4066 @command{@value{AS}}.
4069 @subsubsection Other
4071 @cindex other attribute, of @code{a.out} symbol
4072 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
4077 @subsection Symbol Attributes for COFF
4079 @cindex COFF symbol attributes
4080 @cindex symbol attributes, COFF
4082 The COFF format supports a multitude of auxiliary symbol attributes;
4083 like the primary symbol attributes, they are set between @code{.def} and
4084 @code{.endef} directives.
4086 @subsubsection Primary Attributes
4088 @cindex primary attributes, COFF symbols
4089 The symbol name is set with @code{.def}; the value and type,
4090 respectively, with @code{.val} and @code{.type}.
4092 @subsubsection Auxiliary Attributes
4094 @cindex auxiliary attributes, COFF symbols
4095 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
4096 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
4097 table information for COFF.
4102 @subsection Symbol Attributes for SOM
4104 @cindex SOM symbol attributes
4105 @cindex symbol attributes, SOM
4107 The SOM format for the HPPA supports a multitude of symbol attributes set with
4108 the @code{.EXPORT} and @code{.IMPORT} directives.
4110 The attributes are described in @cite{HP9000 Series 800 Assembly
4111 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
4112 @code{EXPORT} assembler directive documentation.
4116 @chapter Expressions
4120 @cindex numeric values
4121 An @dfn{expression} specifies an address or numeric value.
4122 Whitespace may precede and/or follow an expression.
4124 The result of an expression must be an absolute number, or else an offset into
4125 a particular section. If an expression is not absolute, and there is not
4126 enough information when @command{@value{AS}} sees the expression to know its
4127 section, a second pass over the source program might be necessary to interpret
4128 the expression---but the second pass is currently not implemented.
4129 @command{@value{AS}} aborts with an error message in this situation.
4132 * Empty Exprs:: Empty Expressions
4133 * Integer Exprs:: Integer Expressions
4137 @section Empty Expressions
4139 @cindex empty expressions
4140 @cindex expressions, empty
4141 An empty expression has no value: it is just whitespace or null.
4142 Wherever an absolute expression is required, you may omit the
4143 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
4144 is compatible with other assemblers.
4147 @section Integer Expressions
4149 @cindex integer expressions
4150 @cindex expressions, integer
4151 An @dfn{integer expression} is one or more @emph{arguments} delimited
4152 by @emph{operators}.
4155 * Arguments:: Arguments
4156 * Operators:: Operators
4157 * Prefix Ops:: Prefix Operators
4158 * Infix Ops:: Infix Operators
4162 @subsection Arguments
4164 @cindex expression arguments
4165 @cindex arguments in expressions
4166 @cindex operands in expressions
4167 @cindex arithmetic operands
4168 @dfn{Arguments} are symbols, numbers or subexpressions. In other
4169 contexts arguments are sometimes called ``arithmetic operands''. In
4170 this manual, to avoid confusing them with the ``instruction operands'' of
4171 the machine language, we use the term ``argument'' to refer to parts of
4172 expressions only, reserving the word ``operand'' to refer only to machine
4173 instruction operands.
4175 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
4176 @var{section} is one of text, data, bss, absolute,
4177 or undefined. @var{NNN} is a signed, 2's complement 32 bit
4180 Numbers are usually integers.
4182 A number can be a flonum or bignum. In this case, you are warned
4183 that only the low order 32 bits are used, and @command{@value{AS}} pretends
4184 these 32 bits are an integer. You may write integer-manipulating
4185 instructions that act on exotic constants, compatible with other
4188 @cindex subexpressions
4189 Subexpressions are a left parenthesis @samp{(} followed by an integer
4190 expression, followed by a right parenthesis @samp{)}; or a prefix
4191 operator followed by an argument.
4194 @subsection Operators
4196 @cindex operators, in expressions
4197 @cindex arithmetic functions
4198 @cindex functions, in expressions
4199 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
4200 operators are followed by an argument. Infix operators appear
4201 between their arguments. Operators may be preceded and/or followed by
4205 @subsection Prefix Operator
4207 @cindex prefix operators
4208 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
4209 one argument, which must be absolute.
4211 @c the tex/end tex stuff surrounding this small table is meant to make
4212 @c it align, on the printed page, with the similar table in the next
4213 @c section (which is inside an enumerate).
4215 \global\advance\leftskip by \itemindent
4220 @dfn{Negation}. Two's complement negation.
4222 @dfn{Complementation}. Bitwise not.
4226 \global\advance\leftskip by -\itemindent
4230 @subsection Infix Operators
4232 @cindex infix operators
4233 @cindex operators, permitted arguments
4234 @dfn{Infix operators} take two arguments, one on either side. Operators
4235 have precedence, but operations with equal precedence are performed left
4236 to right. Apart from @code{+} or @option{-}, both arguments must be
4237 absolute, and the result is absolute.
4240 @cindex operator precedence
4241 @cindex precedence of operators
4248 @dfn{Multiplication}.
4251 @dfn{Division}. Truncation is the same as the C operator @samp{/}
4257 @dfn{Shift Left}. Same as the C operator @samp{<<}.
4260 @dfn{Shift Right}. Same as the C operator @samp{>>}.
4264 Intermediate precedence
4269 @dfn{Bitwise Inclusive Or}.
4275 @dfn{Bitwise Exclusive Or}.
4278 @dfn{Bitwise Or Not}.
4285 @cindex addition, permitted arguments
4286 @cindex plus, permitted arguments
4287 @cindex arguments for addition
4289 @dfn{Addition}. If either argument is absolute, the result has the section of
4290 the other argument. You may not add together arguments from different
4293 @cindex subtraction, permitted arguments
4294 @cindex minus, permitted arguments
4295 @cindex arguments for subtraction
4297 @dfn{Subtraction}. If the right argument is absolute, the
4298 result has the section of the left argument.
4299 If both arguments are in the same section, the result is absolute.
4300 You may not subtract arguments from different sections.
4301 @c FIXME is there still something useful to say about undefined - undefined ?
4303 @cindex comparison expressions
4304 @cindex expressions, comparison
4309 @dfn{Is Not Equal To}
4313 @dfn{Is Greater Than}
4315 @dfn{Is Greater Than Or Equal To}
4317 @dfn{Is Less Than Or Equal To}
4319 The comparison operators can be used as infix operators. A true results has a
4320 value of -1 whereas a false result has a value of 0. Note, these operators
4321 perform signed comparisons.
4324 @item Lowest Precedence
4333 These two logical operations can be used to combine the results of sub
4334 expressions. Note, unlike the comparison operators a true result returns a
4335 value of 1 but a false results does still return 0. Also note that the logical
4336 or operator has a slightly lower precedence than logical and.
4341 In short, it's only meaningful to add or subtract the @emph{offsets} in an
4342 address; you can only have a defined section in one of the two arguments.
4345 @chapter Assembler Directives
4347 @cindex directives, machine independent
4348 @cindex pseudo-ops, machine independent
4349 @cindex machine independent directives
4350 All assembler directives have names that begin with a period (@samp{.}).
4351 The names are case insensitive for most targets, and usually written
4354 This chapter discusses directives that are available regardless of the
4355 target machine configuration for the @sc{gnu} assembler.
4357 Some machine configurations provide additional directives.
4358 @xref{Machine Dependencies}.
4361 @ifset machine-directives
4362 @xref{Machine Dependencies}, for additional directives.
4367 * Abort:: @code{.abort}
4369 * ABORT (COFF):: @code{.ABORT}
4372 * Align:: @code{.align [@var{abs-expr}[, @var{abs-expr}[, @var{abs-expr}]]]}
4373 * Altmacro:: @code{.altmacro}
4374 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4375 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4376 * Balign:: @code{.balign [@var{abs-expr}[, @var{abs-expr}]]}
4377 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, etc
4378 * Byte:: @code{.byte @var{expressions}}
4379 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4380 * Comm:: @code{.comm @var{symbol} , @var{length} }
4381 * Data:: @code{.data @var{subsection}}
4382 * Dc:: @code{.dc[@var{size}] @var{expressions}}
4383 * Dcb:: @code{.dcb[@var{size}] @var{number} [,@var{fill}]}
4384 * Ds:: @code{.ds[@var{size}] @var{number} [,@var{fill}]}
4386 * Def:: @code{.def @var{name}}
4389 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4395 * Double:: @code{.double @var{flonums}}
4396 * Eject:: @code{.eject}
4397 * Else:: @code{.else}
4398 * Elseif:: @code{.elseif}
4401 * Endef:: @code{.endef}
4404 * Endfunc:: @code{.endfunc}
4405 * Endif:: @code{.endif}
4406 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4407 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4408 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4410 * Error:: @code{.error @var{string}}
4411 * Exitm:: @code{.exitm}
4412 * Extern:: @code{.extern}
4413 * Fail:: @code{.fail}
4414 * File:: @code{.file}
4415 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4416 * Float:: @code{.float @var{flonums}}
4417 * Func:: @code{.func}
4418 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4420 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4421 * Hidden:: @code{.hidden @var{names}}
4424 * hword:: @code{.hword @var{expressions}}
4425 * Ident:: @code{.ident}
4426 * If:: @code{.if @var{absolute expression}}
4427 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4428 * Include:: @code{.include "@var{file}"}
4429 * Int:: @code{.int @var{expressions}}
4431 * Internal:: @code{.internal @var{names}}
4434 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4435 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4436 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4437 * Lflags:: @code{.lflags}
4438 @ifclear no-line-dir
4439 * Line:: @code{.line @var{line-number}}
4442 * Linkonce:: @code{.linkonce [@var{type}]}
4443 * List:: @code{.list}
4444 * Ln:: @code{.ln @var{line-number}}
4445 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4446 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4448 * Local:: @code{.local @var{names}}
4451 * Long:: @code{.long @var{expressions}}
4453 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4456 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4457 * MRI:: @code{.mri @var{val}}
4458 * Noaltmacro:: @code{.noaltmacro}
4459 * Nolist:: @code{.nolist}
4460 * Nops:: @code{.nops @var{size}[, @var{control}]}
4461 * Octa:: @code{.octa @var{bignums}}
4462 * Offset:: @code{.offset @var{loc}}
4463 * Org:: @code{.org @var{new-lc}, @var{fill}}
4464 * P2align:: @code{.p2align [@var{abs-expr}[, @var{abs-expr}[, @var{abs-expr}]]]}
4466 * PopSection:: @code{.popsection}
4467 * Previous:: @code{.previous}
4470 * Print:: @code{.print @var{string}}
4472 * Protected:: @code{.protected @var{names}}
4475 * Psize:: @code{.psize @var{lines}, @var{columns}}
4476 * Purgem:: @code{.purgem @var{name}}
4478 * PushSection:: @code{.pushsection @var{name}}
4481 * Quad:: @code{.quad @var{bignums}}
4482 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4483 * Rept:: @code{.rept @var{count}}
4484 * Sbttl:: @code{.sbttl "@var{subheading}"}
4486 * Scl:: @code{.scl @var{class}}
4489 * Section:: @code{.section @var{name}[, @var{flags}]}
4492 * Set:: @code{.set @var{symbol}, @var{expression}}
4493 * Short:: @code{.short @var{expressions}}
4494 * Single:: @code{.single @var{flonums}}
4496 * Size:: @code{.size [@var{name} , @var{expression}]}
4498 @ifclear no-space-dir
4499 * Skip:: @code{.skip @var{size} [,@var{fill}]}
4502 * Sleb128:: @code{.sleb128 @var{expressions}}
4503 @ifclear no-space-dir
4504 * Space:: @code{.space @var{size} [,@var{fill}]}
4507 * Stab:: @code{.stabd, .stabn, .stabs}
4510 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4511 * Struct:: @code{.struct @var{expression}}
4513 * SubSection:: @code{.subsection}
4514 * Symver:: @code{.symver @var{name},@var{name2@@nodename}[,@var{visibility}]}
4518 * Tag:: @code{.tag @var{structname}}
4521 * Text:: @code{.text @var{subsection}}
4522 * Title:: @code{.title "@var{heading}"}
4524 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4527 * Uleb128:: @code{.uleb128 @var{expressions}}
4529 * Val:: @code{.val @var{addr}}
4533 * Version:: @code{.version "@var{string}"}
4534 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4535 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4538 * Warning:: @code{.warning @var{string}}
4539 * Weak:: @code{.weak @var{names}}
4540 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4541 * Word:: @code{.word @var{expressions}}
4542 @ifclear no-space-dir
4543 * Zero:: @code{.zero @var{size}}
4546 * 2byte:: @code{.2byte @var{expressions}}
4547 * 4byte:: @code{.4byte @var{expressions}}
4548 * 8byte:: @code{.8byte @var{bignums}}
4550 * Deprecated:: Deprecated Directives
4554 @section @code{.abort}
4556 @cindex @code{abort} directive
4557 @cindex stopping the assembly
4558 This directive stops the assembly immediately. It is for
4559 compatibility with other assemblers. The original idea was that the
4560 assembly language source would be piped into the assembler. If the sender
4561 of the source quit, it could use this directive tells @command{@value{AS}} to
4562 quit also. One day @code{.abort} will not be supported.
4566 @section @code{.ABORT} (COFF)
4568 @cindex @code{ABORT} directive
4569 When producing COFF output, @command{@value{AS}} accepts this directive as a
4570 synonym for @samp{.abort}.
4575 @section @code{.align [@var{abs-expr}[, @var{abs-expr}[, @var{abs-expr}]]]}
4577 @cindex padding the location counter
4578 @cindex @code{align} directive
4579 Pad the location counter (in the current subsection) to a particular storage
4580 boundary. The first expression (which must be absolute) is the alignment
4581 required, as described below. If this expression is omitted then a default
4582 value of 0 is used, effectively disabling alignment requirements.
4584 The second expression (also absolute) gives the fill value to be stored in the
4585 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4586 padding bytes are normally zero. However, on most systems, if the section is
4587 marked as containing code and the fill value is omitted, the space is filled
4588 with no-op instructions.
4590 The third expression is also absolute, and is also optional. If it is present,
4591 it is the maximum number of bytes that should be skipped by this alignment
4592 directive. If doing the alignment would require skipping more bytes than the
4593 specified maximum, then the alignment is not done at all. You can omit the
4594 fill value (the second argument) entirely by simply using two commas after the
4595 required alignment; this can be useful if you want the alignment to be filled
4596 with no-op instructions when appropriate.
4598 The way the required alignment is specified varies from system to system.
4599 For the arc, hppa, i386 using ELF, iq2000, m68k, or1k,
4600 s390, sparc, tic4x and xtensa, the first expression is the
4601 alignment request in bytes. For example @samp{.align 8} advances
4602 the location counter until it is a multiple of 8. If the location counter
4603 is already a multiple of 8, no change is needed. For the tic54x, the
4604 first expression is the alignment request in words.
4606 For other systems, including ppc, i386 using a.out format, arm and
4607 strongarm, it is the
4608 number of low-order zero bits the location counter must have after
4609 advancement. For example @samp{.align 3} advances the location
4610 counter until it is a multiple of 8. If the location counter is already a
4611 multiple of 8, no change is needed.
4613 This inconsistency is due to the different behaviors of the various
4614 native assemblers for these systems which GAS must emulate.
4615 GAS also provides @code{.balign} and @code{.p2align} directives,
4616 described later, which have a consistent behavior across all
4617 architectures (but are specific to GAS).
4620 @section @code{.altmacro}
4621 Enable alternate macro mode, enabling:
4624 @item LOCAL @var{name} [ , @dots{} ]
4625 One additional directive, @code{LOCAL}, is available. It is used to
4626 generate a string replacement for each of the @var{name} arguments, and
4627 replace any instances of @var{name} in each macro expansion. The
4628 replacement string is unique in the assembly, and different for each
4629 separate macro expansion. @code{LOCAL} allows you to write macros that
4630 define symbols, without fear of conflict between separate macro expansions.
4632 @item String delimiters
4633 You can write strings delimited in these other ways besides
4634 @code{"@var{string}"}:
4637 @item '@var{string}'
4638 You can delimit strings with single-quote characters.
4640 @item <@var{string}>
4641 You can delimit strings with matching angle brackets.
4644 @item single-character string escape
4645 To include any single character literally in a string (even if the
4646 character would otherwise have some special meaning), you can prefix the
4647 character with @samp{!} (an exclamation mark). For example, you can
4648 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4650 @item Expression results as strings
4651 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4652 and use the result as a string.
4656 @section @code{.ascii "@var{string}"}@dots{}
4658 @cindex @code{ascii} directive
4659 @cindex string literals
4660 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4661 separated by commas. It assembles each string (with no automatic
4662 trailing zero byte) into consecutive addresses.
4665 @section @code{.asciz "@var{string}"}@dots{}
4667 @cindex @code{asciz} directive
4668 @cindex zero-terminated strings
4669 @cindex null-terminated strings
4670 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4671 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4674 @section @code{.balign[wl] [@var{abs-expr}[, @var{abs-expr}[, @var{abs-expr}]]]}
4676 @cindex padding the location counter given number of bytes
4677 @cindex @code{balign} directive
4678 Pad the location counter (in the current subsection) to a particular
4679 storage boundary. The first expression (which must be absolute) is the
4680 alignment request in bytes. For example @samp{.balign 8} advances
4681 the location counter until it is a multiple of 8. If the location counter
4682 is already a multiple of 8, no change is needed. If the expression is omitted
4683 then a default value of 0 is used, effectively disabling alignment requirements.
4685 The second expression (also absolute) gives the fill value to be stored in the
4686 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4687 padding bytes are normally zero. However, on most systems, if the section is
4688 marked as containing code and the fill value is omitted, the space is filled
4689 with no-op instructions.
4691 The third expression is also absolute, and is also optional. If it is present,
4692 it is the maximum number of bytes that should be skipped by this alignment
4693 directive. If doing the alignment would require skipping more bytes than the
4694 specified maximum, then the alignment is not done at all. You can omit the
4695 fill value (the second argument) entirely by simply using two commas after the
4696 required alignment; this can be useful if you want the alignment to be filled
4697 with no-op instructions when appropriate.
4699 @cindex @code{balignw} directive
4700 @cindex @code{balignl} directive
4701 The @code{.balignw} and @code{.balignl} directives are variants of the
4702 @code{.balign} directive. The @code{.balignw} directive treats the fill
4703 pattern as a two byte word value. The @code{.balignl} directives treats the
4704 fill pattern as a four byte longword value. For example, @code{.balignw
4705 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4706 filled in with the value 0x368d (the exact placement of the bytes depends upon
4707 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4710 @node Bundle directives
4711 @section Bundle directives
4712 @subsection @code{.bundle_align_mode @var{abs-expr}}
4713 @cindex @code{bundle_align_mode} directive
4715 @cindex instruction bundle
4716 @cindex aligned instruction bundle
4717 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4718 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4719 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4720 disabled (which is the default state). If the argument it not zero, it
4721 gives the size of an instruction bundle as a power of two (as for the
4722 @code{.p2align} directive, @pxref{P2align}).
4724 For some targets, it's an ABI requirement that no instruction may span a
4725 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4726 instructions that starts on an aligned boundary. For example, if
4727 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4728 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4729 effect, no single instruction may span a boundary between bundles. If an
4730 instruction would start too close to the end of a bundle for the length of
4731 that particular instruction to fit within the bundle, then the space at the
4732 end of that bundle is filled with no-op instructions so the instruction
4733 starts in the next bundle. As a corollary, it's an error if any single
4734 instruction's encoding is longer than the bundle size.
4736 @subsection @code{.bundle_lock} and @code{.bundle_unlock}
4737 @cindex @code{bundle_lock} directive
4738 @cindex @code{bundle_unlock} directive
4739 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4740 allow explicit control over instruction bundle padding. These directives
4741 are only valid when @code{.bundle_align_mode} has been used to enable
4742 aligned instruction bundle mode. It's an error if they appear when
4743 @code{.bundle_align_mode} has not been used at all, or when the last
4744 directive was @w{@code{.bundle_align_mode 0}}.
4746 @cindex bundle-locked
4747 For some targets, it's an ABI requirement that certain instructions may
4748 appear only as part of specified permissible sequences of multiple
4749 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4750 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4751 instruction sequence. For purposes of aligned instruction bundle mode, a
4752 sequence starting with @code{.bundle_lock} and ending with
4753 @code{.bundle_unlock} is treated as a single instruction. That is, the
4754 entire sequence must fit into a single bundle and may not span a bundle
4755 boundary. If necessary, no-op instructions will be inserted before the
4756 first instruction of the sequence so that the whole sequence starts on an
4757 aligned bundle boundary. It's an error if the sequence is longer than the
4760 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4761 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4762 nested. That is, a second @code{.bundle_lock} directive before the next
4763 @code{.bundle_unlock} directive has no effect except that it must be
4764 matched by another closing @code{.bundle_unlock} so that there is the
4765 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4768 @section @code{.byte @var{expressions}}
4770 @cindex @code{byte} directive
4771 @cindex integers, one byte
4772 @code{.byte} expects zero or more expressions, separated by commas.
4773 Each expression is assembled into the next byte.
4775 @node CFI directives
4776 @section CFI directives
4777 @subsection @code{.cfi_sections @var{section_list}}
4778 @cindex @code{cfi_sections} directive
4779 @code{.cfi_sections} may be used to specify whether CFI directives
4780 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4781 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4782 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4783 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4784 directive is not used is @code{.cfi_sections .eh_frame}.
4786 On targets that support compact unwinding tables these can be generated
4787 by specifying @code{.eh_frame_entry} instead of @code{.eh_frame}.
4789 Some targets may support an additional name, such as @code{.c6xabi.exidx}
4790 which is used by the @value{TIC6X} target.
4792 The @code{.cfi_sections} directive can be repeated, with the same or different
4793 arguments, provided that CFI generation has not yet started. Once CFI
4794 generation has started however the section list is fixed and any attempts to
4795 redefine it will result in an error.
4797 @subsection @code{.cfi_startproc [simple]}
4798 @cindex @code{cfi_startproc} directive
4799 @code{.cfi_startproc} is used at the beginning of each function that
4800 should have an entry in @code{.eh_frame}. It initializes some internal
4801 data structures. Don't forget to close the function by
4802 @code{.cfi_endproc}.
4804 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4805 it also emits some architecture dependent initial CFI instructions.
4807 @subsection @code{.cfi_endproc}
4808 @cindex @code{cfi_endproc} directive
4809 @code{.cfi_endproc} is used at the end of a function where it closes its
4810 unwind entry previously opened by
4811 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4813 @subsection @code{.cfi_personality @var{encoding} [, @var{exp}]}
4814 @cindex @code{cfi_personality} directive
4815 @code{.cfi_personality} defines personality routine and its encoding.
4816 @var{encoding} must be a constant determining how the personality
4817 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4818 argument is not present, otherwise second argument should be
4819 a constant or a symbol name. When using indirect encodings,
4820 the symbol provided should be the location where personality
4821 can be loaded from, not the personality routine itself.
4822 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4823 no personality routine.
4825 @subsection @code{.cfi_personality_id @var{id}}
4826 @cindex @code{cfi_personality_id} directive
4827 @code{cfi_personality_id} defines a personality routine by its index as
4828 defined in a compact unwinding format.
4829 Only valid when generating compact EH frames (i.e.
4830 with @code{.cfi_sections eh_frame_entry}.
4832 @subsection @code{.cfi_fde_data [@var{opcode1} [, @dots{}]]}
4833 @cindex @code{cfi_fde_data} directive
4834 @code{cfi_fde_data} is used to describe the compact unwind opcodes to be
4835 used for the current function. These are emitted inline in the
4836 @code{.eh_frame_entry} section if small enough and there is no LSDA, or
4837 in the @code{.gnu.extab} section otherwise.
4838 Only valid when generating compact EH frames (i.e.
4839 with @code{.cfi_sections eh_frame_entry}.
4841 @subsection @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4842 @code{.cfi_lsda} defines LSDA and its encoding.
4843 @var{encoding} must be a constant determining how the LSDA
4844 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), the second
4845 argument is not present, otherwise the second argument should be a constant
4846 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4847 meaning that no LSDA is present.
4849 @subsection @code{.cfi_inline_lsda} [@var{align}]
4850 @code{.cfi_inline_lsda} marks the start of a LSDA data section and
4851 switches to the corresponding @code{.gnu.extab} section.
4852 Must be preceded by a CFI block containing a @code{.cfi_lsda} directive.
4853 Only valid when generating compact EH frames (i.e.
4854 with @code{.cfi_sections eh_frame_entry}.
4856 The table header and unwinding opcodes will be generated at this point,
4857 so that they are immediately followed by the LSDA data. The symbol
4858 referenced by the @code{.cfi_lsda} directive should still be defined
4859 in case a fallback FDE based encoding is used. The LSDA data is terminated
4860 by a section directive.
4862 The optional @var{align} argument specifies the alignment required.
4863 The alignment is specified as a power of two, as with the
4864 @code{.p2align} directive.
4866 @subsection @code{.cfi_def_cfa @var{register}, @var{offset}}
4867 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4868 address from @var{register} and add @var{offset} to it}.
4870 @subsection @code{.cfi_def_cfa_register @var{register}}
4871 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4872 now on @var{register} will be used instead of the old one. Offset
4875 @subsection @code{.cfi_def_cfa_offset @var{offset}}
4876 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4877 remains the same, but @var{offset} is new. Note that it is the
4878 absolute offset that will be added to a defined register to compute
4881 @subsection @code{.cfi_adjust_cfa_offset @var{offset}}
4882 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4883 value that is added/subtracted from the previous offset.
4885 @subsection @code{.cfi_offset @var{register}, @var{offset}}
4886 Previous value of @var{register} is saved at offset @var{offset} from
4889 @subsection @code{.cfi_val_offset @var{register}, @var{offset}}
4890 Previous value of @var{register} is CFA + @var{offset}.
4892 @subsection @code{.cfi_rel_offset @var{register}, @var{offset}}
4893 Previous value of @var{register} is saved at offset @var{offset} from
4894 the current CFA register. This is transformed to @code{.cfi_offset}
4895 using the known displacement of the CFA register from the CFA.
4896 This is often easier to use, because the number will match the
4897 code it's annotating.
4899 @subsection @code{.cfi_register @var{register1}, @var{register2}}
4900 Previous value of @var{register1} is saved in register @var{register2}.
4902 @subsection @code{.cfi_restore @var{register}}
4903 @code{.cfi_restore} says that the rule for @var{register} is now the
4904 same as it was at the beginning of the function, after all initial
4905 instruction added by @code{.cfi_startproc} were executed.
4907 @subsection @code{.cfi_undefined @var{register}}
4908 From now on the previous value of @var{register} can't be restored anymore.
4910 @subsection @code{.cfi_same_value @var{register}}
4911 Current value of @var{register} is the same like in the previous frame,
4912 i.e. no restoration needed.
4914 @subsection @code{.cfi_remember_state} and @code{.cfi_restore_state}
4915 @code{.cfi_remember_state} pushes the set of rules for every register onto an
4916 implicit stack, while @code{.cfi_restore_state} pops them off the stack and
4917 places them in the current row. This is useful for situations where you have
4918 multiple @code{.cfi_*} directives that need to be undone due to the control
4919 flow of the program. For example, we could have something like this (assuming
4920 the CFA is the value of @code{rbp}):
4930 .cfi_def_cfa %rsp, 8
4933 /* Do something else */
4936 Here, we want the @code{.cfi} directives to affect only the rows corresponding
4937 to the instructions before @code{label}. This means we'd have to add multiple
4938 @code{.cfi} directives after @code{label} to recreate the original save
4939 locations of the registers, as well as setting the CFA back to the value of
4940 @code{rbp}. This would be clumsy, and result in a larger binary size. Instead,
4952 .cfi_def_cfa %rsp, 8
4956 /* Do something else */
4959 That way, the rules for the instructions after @code{label} will be the same
4960 as before the first @code{.cfi_restore} without having to use multiple
4961 @code{.cfi} directives.
4963 @subsection @code{.cfi_return_column @var{register}}
4964 Change return column @var{register}, i.e. the return address is either
4965 directly in @var{register} or can be accessed by rules for @var{register}.
4967 @subsection @code{.cfi_signal_frame}
4968 Mark current function as signal trampoline.
4970 @subsection @code{.cfi_window_save}
4971 SPARC register window has been saved.
4973 @subsection @code{.cfi_escape} @var{expression}[, @dots{}]
4974 Allows the user to add arbitrary bytes to the unwind info. One
4975 might use this to add OS-specific CFI opcodes, or generic CFI
4976 opcodes that GAS does not yet support.
4978 @subsection @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4979 The current value of @var{register} is @var{label}. The value of @var{label}
4980 will be encoded in the output file according to @var{encoding}; see the
4981 description of @code{.cfi_personality} for details on this encoding.
4983 The usefulness of equating a register to a fixed label is probably
4984 limited to the return address register. Here, it can be useful to
4985 mark a code segment that has only one return address which is reached
4986 by a direct branch and no copy of the return address exists in memory
4987 or another register.
4990 @section @code{.comm @var{symbol} , @var{length} }
4992 @cindex @code{comm} directive
4993 @cindex symbol, common
4994 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4995 common symbol in one object file may be merged with a defined or common symbol
4996 of the same name in another object file. If @code{@value{LD}} does not see a
4997 definition for the symbol--just one or more common symbols--then it will
4998 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4999 absolute expression. If @code{@value{LD}} sees multiple common symbols with
5000 the same name, and they do not all have the same size, it will allocate space
5001 using the largest size.
5004 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
5005 an optional third argument. This is the desired alignment of the symbol,
5006 specified for ELF as a byte boundary (for example, an alignment of 16 means
5007 that the least significant 4 bits of the address should be zero), and for PE
5008 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
5009 boundary). The alignment must be an absolute expression, and it must be a
5010 power of two. If @code{@value{LD}} allocates uninitialized memory for the
5011 common symbol, it will use the alignment when placing the symbol. If no
5012 alignment is specified, @command{@value{AS}} will set the alignment to the
5013 largest power of two less than or equal to the size of the symbol, up to a
5014 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
5015 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
5016 @samp{--section-alignment} option; image file sections in PE are aligned to
5017 multiples of 4096, which is far too large an alignment for ordinary variables.
5018 It is rather the default alignment for (non-debug) sections within object
5019 (@samp{*.o}) files, which are less strictly aligned.}.
5023 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
5024 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
5028 @section @code{.data @var{subsection}}
5029 @cindex @code{data} directive
5031 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
5032 end of the data subsection numbered @var{subsection} (which is an
5033 absolute expression). If @var{subsection} is omitted, it defaults
5037 @section @code{.dc[@var{size}] @var{expressions}}
5038 @cindex @code{dc} directive
5040 The @code{.dc} directive expects zero or more @var{expressions} separated by
5041 commas. These expressions are evaluated and their values inserted into the
5042 current section. The size of the emitted value depends upon the suffix to the
5043 @code{.dc} directive:
5047 Emits N-bit values, where N is the size of an address on the target system.
5051 Emits double precision floating-point values.
5053 Emits 32-bit values.
5055 Emits single precision floating-point values.
5057 Emits 16-bit values.
5058 Note - this is true even on targets where the @code{.word} directive would emit
5061 Emits long double precision floating-point values.
5064 If no suffix is used then @samp{.w} is assumed.
5066 The byte ordering is target dependent, as is the size and format of floating
5070 @section @code{.dcb[@var{size}] @var{number} [,@var{fill}]}
5071 @cindex @code{dcb} directive
5072 This directive emits @var{number} copies of @var{fill}, each of @var{size}
5073 bytes. Both @var{number} and @var{fill} are absolute expressions. If the
5074 comma and @var{fill} are omitted, @var{fill} is assumed to be zero. The
5075 @var{size} suffix, if present, must be one of:
5079 Emits single byte values.
5081 Emits double-precision floating point values.
5083 Emits 4-byte values.
5085 Emits single-precision floating point values.
5087 Emits 2-byte values.
5089 Emits long double-precision floating point values.
5092 If the @var{size} suffix is omitted then @samp{.w} is assumed.
5094 The byte ordering is target dependent, as is the size and format of floating
5098 @section @code{.ds[@var{size}] @var{number} [,@var{fill}]}
5099 @cindex @code{ds} directive
5100 This directive emits @var{number} copies of @var{fill}, each of @var{size}
5101 bytes. Both @var{number} and @var{fill} are absolute expressions. If the
5102 comma and @var{fill} are omitted, @var{fill} is assumed to be zero. The
5103 @var{size} suffix, if present, must be one of:
5107 Emits single byte values.
5109 Emits 8-byte values.
5111 Emits 4-byte values.
5113 Emits 12-byte values.
5115 Emits 4-byte values.
5117 Emits 2-byte values.
5119 Emits 12-byte values.
5122 Note - unlike the @code{.dcb} directive the @samp{.d}, @samp{.s} and @samp{.x}
5123 suffixes do not indicate that floating-point values are to be inserted.
5125 If the @var{size} suffix is omitted then @samp{.w} is assumed.
5127 The byte ordering is target dependent.
5132 @section @code{.def @var{name}}
5134 @cindex @code{def} directive
5135 @cindex COFF symbols, debugging
5136 @cindex debugging COFF symbols
5137 Begin defining debugging information for a symbol @var{name}; the
5138 definition extends until the @code{.endef} directive is encountered.
5143 @section @code{.desc @var{symbol}, @var{abs-expression}}
5145 @cindex @code{desc} directive
5146 @cindex COFF symbol descriptor
5147 @cindex symbol descriptor, COFF
5148 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
5149 to the low 16 bits of an absolute expression.
5152 The @samp{.desc} directive is not available when @command{@value{AS}} is
5153 configured for COFF output; it is only for @code{a.out} or @code{b.out}
5154 object format. For the sake of compatibility, @command{@value{AS}} accepts
5155 it, but produces no output, when configured for COFF.
5161 @section @code{.dim}
5163 @cindex @code{dim} directive
5164 @cindex COFF auxiliary symbol information
5165 @cindex auxiliary symbol information, COFF
5166 This directive is generated by compilers to include auxiliary debugging
5167 information in the symbol table. It is only permitted inside
5168 @code{.def}/@code{.endef} pairs.
5172 @section @code{.double @var{flonums}}
5174 @cindex @code{double} directive
5175 @cindex floating point numbers (double)
5176 @code{.double} expects zero or more flonums, separated by commas. It
5177 assembles floating point numbers.
5179 The exact kind of floating point numbers emitted depends on how
5180 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5184 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
5185 in @sc{ieee} format.
5190 @section @code{.eject}
5192 @cindex @code{eject} directive
5193 @cindex new page, in listings
5194 @cindex page, in listings
5195 @cindex listing control: new page
5196 Force a page break at this point, when generating assembly listings.
5199 @section @code{.else}
5201 @cindex @code{else} directive
5202 @code{.else} is part of the @command{@value{AS}} support for conditional
5203 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
5204 of code to be assembled if the condition for the preceding @code{.if}
5208 @section @code{.elseif}
5210 @cindex @code{elseif} directive
5211 @code{.elseif} is part of the @command{@value{AS}} support for conditional
5212 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
5213 @code{.if} block that would otherwise fill the entire @code{.else} section.
5216 @section @code{.end}
5218 @cindex @code{end} directive
5219 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
5220 process anything in the file past the @code{.end} directive.
5224 @section @code{.endef}
5226 @cindex @code{endef} directive
5227 This directive flags the end of a symbol definition begun with
5232 @section @code{.endfunc}
5233 @cindex @code{endfunc} directive
5234 @code{.endfunc} marks the end of a function specified with @code{.func}.
5237 @section @code{.endif}
5239 @cindex @code{endif} directive
5240 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
5241 it marks the end of a block of code that is only assembled
5242 conditionally. @xref{If,,@code{.if}}.
5245 @section @code{.equ @var{symbol}, @var{expression}}
5247 @cindex @code{equ} directive
5248 @cindex assigning values to symbols
5249 @cindex symbols, assigning values to
5250 This directive sets the value of @var{symbol} to @var{expression}.
5251 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
5254 The syntax for @code{equ} on the HPPA is
5255 @samp{@var{symbol} .equ @var{expression}}.
5259 The syntax for @code{equ} on the Z80 is
5260 @samp{@var{symbol} equ @var{expression}}.
5261 On the Z80 it is an error if @var{symbol} is already defined,
5262 but the symbol is not protected from later redefinition.
5263 Compare @ref{Equiv}.
5267 @section @code{.equiv @var{symbol}, @var{expression}}
5268 @cindex @code{equiv} directive
5269 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
5270 the assembler will signal an error if @var{symbol} is already defined. Note a
5271 symbol which has been referenced but not actually defined is considered to be
5274 Except for the contents of the error message, this is roughly equivalent to
5281 plus it protects the symbol from later redefinition.
5284 @section @code{.eqv @var{symbol}, @var{expression}}
5285 @cindex @code{eqv} directive
5286 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
5287 evaluate the expression or any part of it immediately. Instead each time
5288 the resulting symbol is used in an expression, a snapshot of its current
5292 @section @code{.err}
5293 @cindex @code{err} directive
5294 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
5295 message and, unless the @option{-Z} option was used, it will not generate an
5296 object file. This can be used to signal an error in conditionally compiled code.
5299 @section @code{.error "@var{string}"}
5300 @cindex error directive
5302 Similarly to @code{.err}, this directive emits an error, but you can specify a
5303 string that will be emitted as the error message. If you don't specify the
5304 message, it defaults to @code{".error directive invoked in source file"}.
5305 @xref{Errors, ,Error and Warning Messages}.
5308 .error "This code has not been assembled and tested."
5312 @section @code{.exitm}
5313 Exit early from the current macro definition. @xref{Macro}.
5316 @section @code{.extern}
5318 @cindex @code{extern} directive
5319 @code{.extern} is accepted in the source program---for compatibility
5320 with other assemblers---but it is ignored. @command{@value{AS}} treats
5321 all undefined symbols as external.
5324 @section @code{.fail @var{expression}}
5326 @cindex @code{fail} directive
5327 Generates an error or a warning. If the value of the @var{expression} is 500
5328 or more, @command{@value{AS}} will print a warning message. If the value is less
5329 than 500, @command{@value{AS}} will print an error message. The message will
5330 include the value of @var{expression}. This can occasionally be useful inside
5331 complex nested macros or conditional assembly.
5334 @section @code{.file}
5335 @cindex @code{file} directive
5337 @ifclear no-file-dir
5338 There are two different versions of the @code{.file} directive. Targets
5339 that support DWARF2 line number information use the DWARF2 version of
5340 @code{.file}. Other targets use the default version.
5342 @subheading Default Version
5344 @cindex logical file name
5345 @cindex file name, logical
5346 This version of the @code{.file} directive tells @command{@value{AS}} that we
5347 are about to start a new logical file. The syntax is:
5353 @var{string} is the new file name. In general, the filename is
5354 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
5355 to specify an empty file name, you must give the quotes--@code{""}. This
5356 statement may go away in future: it is only recognized to be compatible with
5357 old @command{@value{AS}} programs.
5359 @subheading DWARF2 Version
5362 When emitting DWARF2 line number information, @code{.file} assigns filenames
5363 to the @code{.debug_line} file name table. The syntax is:
5366 .file @var{fileno} @var{filename}
5369 The @var{fileno} operand should be a unique positive integer to use as the
5370 index of the entry in the table. The @var{filename} operand is a C string
5371 literal enclosed in double quotes. The @var{filename} can include directory
5372 elements. If it does, then the directory will be added to the directory table
5373 and the basename will be added to the file table.
5375 The detail of filename indices is exposed to the user because the filename
5376 table is shared with the @code{.debug_info} section of the DWARF2 debugging
5377 information, and thus the user must know the exact indices that table
5380 If DWARF-5 support has been enabled via the @option{-gdwarf-5} option then
5381 an extended version of the @code{file} is also allowed:
5384 .file @var{fileno} [@var{dirname}] @var{filename} [md5 @var{value}]
5387 With this version a separate directory name is allowed, although if this is
5388 used then @var{filename} should not contain any directory components. In
5389 addtion an md5 hash value of the contents of @var{filename} can be provided.
5390 This will be stored in the the file table as well, and can be used by tools
5391 reading the debug information to verify that the contents of the source file
5392 match the contents of the compiled file.
5395 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
5397 @cindex @code{fill} directive
5398 @cindex writing patterns in memory
5399 @cindex patterns, writing in memory
5400 @var{repeat}, @var{size} and @var{value} are absolute expressions.
5401 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
5402 may be zero or more. @var{Size} may be zero or more, but if it is
5403 more than 8, then it is deemed to have the value 8, compatible with
5404 other people's assemblers. The contents of each @var{repeat} bytes
5405 is taken from an 8-byte number. The highest order 4 bytes are
5406 zero. The lowest order 4 bytes are @var{value} rendered in the
5407 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
5408 Each @var{size} bytes in a repetition is taken from the lowest order
5409 @var{size} bytes of this number. Again, this bizarre behavior is
5410 compatible with other people's assemblers.
5412 @var{size} and @var{value} are optional.
5413 If the second comma and @var{value} are absent, @var{value} is
5414 assumed zero. If the first comma and following tokens are absent,
5415 @var{size} is assumed to be 1.
5418 @section @code{.float @var{flonums}}
5420 @cindex floating point numbers (single)
5421 @cindex @code{float} directive
5422 This directive assembles zero or more flonums, separated by commas. It
5423 has the same effect as @code{.single}.
5425 The exact kind of floating point numbers emitted depends on how
5426 @command{@value{AS}} is configured.
5427 @xref{Machine Dependencies}.
5431 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
5432 in @sc{ieee} format.
5437 @section @code{.func @var{name}[,@var{label}]}
5438 @cindex @code{func} directive
5439 @code{.func} emits debugging information to denote function @var{name}, and
5440 is ignored unless the file is assembled with debugging enabled.
5441 Only @samp{--gstabs[+]} is currently supported.
5442 @var{label} is the entry point of the function and if omitted @var{name}
5443 prepended with the @samp{leading char} is used.
5444 @samp{leading char} is usually @code{_} or nothing, depending on the target.
5445 All functions are currently defined to have @code{void} return type.
5446 The function must be terminated with @code{.endfunc}.
5449 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
5451 @cindex @code{global} directive
5452 @cindex symbol, making visible to linker
5453 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
5454 @var{symbol} in your partial program, its value is made available to
5455 other partial programs that are linked with it. Otherwise,
5456 @var{symbol} takes its attributes from a symbol of the same name
5457 from another file linked into the same program.
5459 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
5460 compatibility with other assemblers.
5463 On the HPPA, @code{.global} is not always enough to make it accessible to other
5464 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
5465 @xref{HPPA Directives, ,HPPA Assembler Directives}.
5470 @section @code{.gnu_attribute @var{tag},@var{value}}
5471 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
5474 @section @code{.hidden @var{names}}
5476 @cindex @code{hidden} directive
5478 This is one of the ELF visibility directives. The other two are
5479 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
5480 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5482 This directive overrides the named symbols default visibility (which is set by
5483 their binding: local, global or weak). The directive sets the visibility to
5484 @code{hidden} which means that the symbols are not visible to other components.
5485 Such symbols are always considered to be @code{protected} as well.
5489 @section @code{.hword @var{expressions}}
5491 @cindex @code{hword} directive
5492 @cindex integers, 16-bit
5493 @cindex numbers, 16-bit
5494 @cindex sixteen bit integers
5495 This expects zero or more @var{expressions}, and emits
5496 a 16 bit number for each.
5499 This directive is a synonym for @samp{.short}; depending on the target
5500 architecture, it may also be a synonym for @samp{.word}.
5504 This directive is a synonym for @samp{.short}.
5507 This directive is a synonym for both @samp{.short} and @samp{.word}.
5512 @section @code{.ident}
5514 @cindex @code{ident} directive
5516 This directive is used by some assemblers to place tags in object files. The
5517 behavior of this directive varies depending on the target. When using the
5518 a.out object file format, @command{@value{AS}} simply accepts the directive for
5519 source-file compatibility with existing assemblers, but does not emit anything
5520 for it. When using COFF, comments are emitted to the @code{.comment} or
5521 @code{.rdata} section, depending on the target. When using ELF, comments are
5522 emitted to the @code{.comment} section.
5525 @section @code{.if @var{absolute expression}}
5527 @cindex conditional assembly
5528 @cindex @code{if} directive
5529 @code{.if} marks the beginning of a section of code which is only
5530 considered part of the source program being assembled if the argument
5531 (which must be an @var{absolute expression}) is non-zero. The end of
5532 the conditional section of code must be marked by @code{.endif}
5533 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
5534 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
5535 If you have several conditions to check, @code{.elseif} may be used to avoid
5536 nesting blocks if/else within each subsequent @code{.else} block.
5538 The following variants of @code{.if} are also supported:
5540 @cindex @code{ifdef} directive
5541 @item .ifdef @var{symbol}
5542 Assembles the following section of code if the specified @var{symbol}
5543 has been defined. Note a symbol which has been referenced but not yet defined
5544 is considered to be undefined.
5546 @cindex @code{ifb} directive
5547 @item .ifb @var{text}
5548 Assembles the following section of code if the operand is blank (empty).
5550 @cindex @code{ifc} directive
5551 @item .ifc @var{string1},@var{string2}
5552 Assembles the following section of code if the two strings are the same. The
5553 strings may be optionally quoted with single quotes. If they are not quoted,
5554 the first string stops at the first comma, and the second string stops at the
5555 end of the line. Strings which contain whitespace should be quoted. The
5556 string comparison is case sensitive.
5558 @cindex @code{ifeq} directive
5559 @item .ifeq @var{absolute expression}
5560 Assembles the following section of code if the argument is zero.
5562 @cindex @code{ifeqs} directive
5563 @item .ifeqs @var{string1},@var{string2}
5564 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5566 @cindex @code{ifge} directive
5567 @item .ifge @var{absolute expression}
5568 Assembles the following section of code if the argument is greater than or
5571 @cindex @code{ifgt} directive
5572 @item .ifgt @var{absolute expression}
5573 Assembles the following section of code if the argument is greater than zero.
5575 @cindex @code{ifle} directive
5576 @item .ifle @var{absolute expression}
5577 Assembles the following section of code if the argument is less than or equal
5580 @cindex @code{iflt} directive
5581 @item .iflt @var{absolute expression}
5582 Assembles the following section of code if the argument is less than zero.
5584 @cindex @code{ifnb} directive
5585 @item .ifnb @var{text}
5586 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5587 following section of code if the operand is non-blank (non-empty).
5589 @cindex @code{ifnc} directive
5590 @item .ifnc @var{string1},@var{string2}.
5591 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5592 following section of code if the two strings are not the same.
5594 @cindex @code{ifndef} directive
5595 @cindex @code{ifnotdef} directive
5596 @item .ifndef @var{symbol}
5597 @itemx .ifnotdef @var{symbol}
5598 Assembles the following section of code if the specified @var{symbol}
5599 has not been defined. Both spelling variants are equivalent. Note a symbol
5600 which has been referenced but not yet defined is considered to be undefined.
5602 @cindex @code{ifne} directive
5603 @item .ifne @var{absolute expression}
5604 Assembles the following section of code if the argument is not equal to zero
5605 (in other words, this is equivalent to @code{.if}).
5607 @cindex @code{ifnes} directive
5608 @item .ifnes @var{string1},@var{string2}
5609 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5610 following section of code if the two strings are not the same.
5614 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5616 @cindex @code{incbin} directive
5617 @cindex binary files, including
5618 The @code{incbin} directive includes @var{file} verbatim at the current
5619 location. You can control the search paths used with the @samp{-I} command-line
5620 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5623 The @var{skip} argument skips a number of bytes from the start of the
5624 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5625 read. Note that the data is not aligned in any way, so it is the user's
5626 responsibility to make sure that proper alignment is provided both before and
5627 after the @code{incbin} directive.
5630 @section @code{.include "@var{file}"}
5632 @cindex @code{include} directive
5633 @cindex supporting files, including
5634 @cindex files, including
5635 This directive provides a way to include supporting files at specified
5636 points in your source program. The code from @var{file} is assembled as
5637 if it followed the point of the @code{.include}; when the end of the
5638 included file is reached, assembly of the original file continues. You
5639 can control the search paths used with the @samp{-I} command-line option
5640 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5644 @section @code{.int @var{expressions}}
5646 @cindex @code{int} directive
5647 @cindex integers, 32-bit
5648 Expect zero or more @var{expressions}, of any section, separated by commas.
5649 For each expression, emit a number that, at run time, is the value of that
5650 expression. The byte order and bit size of the number depends on what kind
5651 of target the assembly is for.
5655 On most forms of the H8/300, @code{.int} emits 16-bit
5656 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5663 @section @code{.internal @var{names}}
5665 @cindex @code{internal} directive
5667 This is one of the ELF visibility directives. The other two are
5668 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5669 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5671 This directive overrides the named symbols default visibility (which is set by
5672 their binding: local, global or weak). The directive sets the visibility to
5673 @code{internal} which means that the symbols are considered to be @code{hidden}
5674 (i.e., not visible to other components), and that some extra, processor specific
5675 processing must also be performed upon the symbols as well.
5679 @section @code{.irp @var{symbol},@var{values}}@dots{}
5681 @cindex @code{irp} directive
5682 Evaluate a sequence of statements assigning different values to @var{symbol}.
5683 The sequence of statements starts at the @code{.irp} directive, and is
5684 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5685 set to @var{value}, and the sequence of statements is assembled. If no
5686 @var{value} is listed, the sequence of statements is assembled once, with
5687 @var{symbol} set to the null string. To refer to @var{symbol} within the
5688 sequence of statements, use @var{\symbol}.
5690 For example, assembling
5698 is equivalent to assembling
5706 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5709 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5711 @cindex @code{irpc} directive
5712 Evaluate a sequence of statements assigning different values to @var{symbol}.
5713 The sequence of statements starts at the @code{.irpc} directive, and is
5714 terminated by an @code{.endr} directive. For each character in @var{value},
5715 @var{symbol} is set to the character, and the sequence of statements is
5716 assembled. If no @var{value} is listed, the sequence of statements is
5717 assembled once, with @var{symbol} set to the null string. To refer to
5718 @var{symbol} within the sequence of statements, use @var{\symbol}.
5720 For example, assembling
5728 is equivalent to assembling
5736 For some caveats with the spelling of @var{symbol}, see also the discussion
5740 @section @code{.lcomm @var{symbol} , @var{length}}
5742 @cindex @code{lcomm} directive
5743 @cindex local common symbols
5744 @cindex symbols, local common
5745 Reserve @var{length} (an absolute expression) bytes for a local common
5746 denoted by @var{symbol}. The section and value of @var{symbol} are
5747 those of the new local common. The addresses are allocated in the bss
5748 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5749 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5750 not visible to @code{@value{LD}}.
5753 Some targets permit a third argument to be used with @code{.lcomm}. This
5754 argument specifies the desired alignment of the symbol in the bss section.
5758 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5759 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5763 @section @code{.lflags}
5765 @cindex @code{lflags} directive (ignored)
5766 @command{@value{AS}} accepts this directive, for compatibility with other
5767 assemblers, but ignores it.
5769 @ifclear no-line-dir
5771 @section @code{.line @var{line-number}}
5773 @cindex @code{line} directive
5774 @cindex logical line number
5776 Change the logical line number. @var{line-number} must be an absolute
5777 expression. The next line has that logical line number. Therefore any other
5778 statements on the current line (after a statement separator character) are
5779 reported as on logical line number @var{line-number} @minus{} 1. One day
5780 @command{@value{AS}} will no longer support this directive: it is recognized only
5781 for compatibility with existing assembler programs.
5784 Even though this is a directive associated with the @code{a.out} or
5785 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5786 when producing COFF output, and treats @samp{.line} as though it
5787 were the COFF @samp{.ln} @emph{if} it is found outside a
5788 @code{.def}/@code{.endef} pair.
5790 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5791 used by compilers to generate auxiliary symbol information for
5796 @section @code{.linkonce [@var{type}]}
5798 @cindex @code{linkonce} directive
5799 @cindex common sections
5800 Mark the current section so that the linker only includes a single copy of it.
5801 This may be used to include the same section in several different object files,
5802 but ensure that the linker will only include it once in the final output file.
5803 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5804 Duplicate sections are detected based on the section name, so it should be
5807 This directive is only supported by a few object file formats; as of this
5808 writing, the only object file format which supports it is the Portable
5809 Executable format used on Windows NT.
5811 The @var{type} argument is optional. If specified, it must be one of the
5812 following strings. For example:
5816 Not all types may be supported on all object file formats.
5820 Silently discard duplicate sections. This is the default.
5823 Warn if there are duplicate sections, but still keep only one copy.
5826 Warn if any of the duplicates have different sizes.
5829 Warn if any of the duplicates do not have exactly the same contents.
5833 @section @code{.list}
5835 @cindex @code{list} directive
5836 @cindex listing control, turning on
5837 Control (in conjunction with the @code{.nolist} directive) whether or
5838 not assembly listings are generated. These two directives maintain an
5839 internal counter (which is zero initially). @code{.list} increments the
5840 counter, and @code{.nolist} decrements it. Assembly listings are
5841 generated whenever the counter is greater than zero.
5843 By default, listings are disabled. When you enable them (with the
5844 @samp{-a} command-line option; @pxref{Invoking,,Command-Line Options}),
5845 the initial value of the listing counter is one.
5848 @section @code{.ln @var{line-number}}
5850 @cindex @code{ln} directive
5851 @ifclear no-line-dir
5852 @samp{.ln} is a synonym for @samp{.line}.
5855 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5856 must be an absolute expression. The next line has that logical
5857 line number, so any other statements on the current line (after a
5858 statement separator character @code{;}) are reported as on logical
5859 line number @var{line-number} @minus{} 1.
5863 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5864 @cindex @code{loc} directive
5865 When emitting DWARF2 line number information,
5866 the @code{.loc} directive will add a row to the @code{.debug_line} line
5867 number matrix corresponding to the immediately following assembly
5868 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5869 arguments will be applied to the @code{.debug_line} state machine before
5872 The @var{options} are a sequence of the following tokens in any order:
5876 This option will set the @code{basic_block} register in the
5877 @code{.debug_line} state machine to @code{true}.
5880 This option will set the @code{prologue_end} register in the
5881 @code{.debug_line} state machine to @code{true}.
5883 @item epilogue_begin
5884 This option will set the @code{epilogue_begin} register in the
5885 @code{.debug_line} state machine to @code{true}.
5887 @item is_stmt @var{value}
5888 This option will set the @code{is_stmt} register in the
5889 @code{.debug_line} state machine to @code{value}, which must be
5892 @item isa @var{value}
5893 This directive will set the @code{isa} register in the @code{.debug_line}
5894 state machine to @var{value}, which must be an unsigned integer.
5896 @item discriminator @var{value}
5897 This directive will set the @code{discriminator} register in the @code{.debug_line}
5898 state machine to @var{value}, which must be an unsigned integer.
5900 @item view @var{value}
5901 This option causes a row to be added to @code{.debug_line} in reference to the
5902 current address (which might not be the same as that of the following assembly
5903 instruction), and to associate @var{value} with the @code{view} register in the
5904 @code{.debug_line} state machine. If @var{value} is a label, both the
5905 @code{view} register and the label are set to the number of prior @code{.loc}
5906 directives at the same program location. If @var{value} is the literal
5907 @code{0}, the @code{view} register is set to zero, and the assembler asserts
5908 that there aren't any prior @code{.loc} directives at the same program
5909 location. If @var{value} is the literal @code{-0}, the assembler arrange for
5910 the @code{view} register to be reset in this row, even if there are prior
5911 @code{.loc} directives at the same program location.
5915 @node Loc_mark_labels
5916 @section @code{.loc_mark_labels @var{enable}}
5917 @cindex @code{loc_mark_labels} directive
5918 When emitting DWARF2 line number information,
5919 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5920 to the @code{.debug_line} line number matrix with the @code{basic_block}
5921 register in the state machine set whenever a code label is seen.
5922 The @var{enable} argument should be either 1 or 0, to enable or disable
5923 this function respectively.
5927 @section @code{.local @var{names}}
5929 @cindex @code{local} directive
5930 This directive, which is available for ELF targets, marks each symbol in
5931 the comma-separated list of @code{names} as a local symbol so that it
5932 will not be externally visible. If the symbols do not already exist,
5933 they will be created.
5935 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5936 accept an alignment argument, which is the case for most ELF targets,
5937 the @code{.local} directive can be used in combination with @code{.comm}
5938 (@pxref{Comm}) to define aligned local common data.
5942 @section @code{.long @var{expressions}}
5944 @cindex @code{long} directive
5945 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5948 @c no one seems to know what this is for or whether this description is
5949 @c what it really ought to do
5951 @section @code{.lsym @var{symbol}, @var{expression}}
5953 @cindex @code{lsym} directive
5954 @cindex symbol, not referenced in assembly
5955 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5956 the hash table, ensuring it cannot be referenced by name during the
5957 rest of the assembly. This sets the attributes of the symbol to be
5958 the same as the expression value:
5960 @var{other} = @var{descriptor} = 0
5961 @var{type} = @r{(section of @var{expression})}
5962 @var{value} = @var{expression}
5965 The new symbol is not flagged as external.
5969 @section @code{.macro}
5972 The commands @code{.macro} and @code{.endm} allow you to define macros that
5973 generate assembly output. For example, this definition specifies a macro
5974 @code{sum} that puts a sequence of numbers into memory:
5977 .macro sum from=0, to=5
5986 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5998 @item .macro @var{macname}
5999 @itemx .macro @var{macname} @var{macargs} @dots{}
6000 @cindex @code{macro} directive
6001 Begin the definition of a macro called @var{macname}. If your macro
6002 definition requires arguments, specify their names after the macro name,
6003 separated by commas or spaces. You can qualify the macro argument to
6004 indicate whether all invocations must specify a non-blank value (through
6005 @samp{:@code{req}}), or whether it takes all of the remaining arguments
6006 (through @samp{:@code{vararg}}). You can supply a default value for any
6007 macro argument by following the name with @samp{=@var{deflt}}. You
6008 cannot define two macros with the same @var{macname} unless it has been
6009 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
6010 definitions. For example, these are all valid @code{.macro} statements:
6014 Begin the definition of a macro called @code{comm}, which takes no
6017 @item .macro plus1 p, p1
6018 @itemx .macro plus1 p p1
6019 Either statement begins the definition of a macro called @code{plus1},
6020 which takes two arguments; within the macro definition, write
6021 @samp{\p} or @samp{\p1} to evaluate the arguments.
6023 @item .macro reserve_str p1=0 p2
6024 Begin the definition of a macro called @code{reserve_str}, with two
6025 arguments. The first argument has a default value, but not the second.
6026 After the definition is complete, you can call the macro either as
6027 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
6028 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
6029 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
6030 @samp{0}, and @samp{\p2} evaluating to @var{b}).
6032 @item .macro m p1:req, p2=0, p3:vararg
6033 Begin the definition of a macro called @code{m}, with at least three
6034 arguments. The first argument must always have a value specified, but
6035 not the second, which instead has a default value. The third formal
6036 will get assigned all remaining arguments specified at invocation time.
6038 When you call a macro, you can specify the argument values either by
6039 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
6040 @samp{sum to=17, from=9}.
6044 Note that since each of the @var{macargs} can be an identifier exactly
6045 as any other one permitted by the target architecture, there may be
6046 occasional problems if the target hand-crafts special meanings to certain
6047 characters when they occur in a special position. For example, if the colon
6048 (@code{:}) is generally permitted to be part of a symbol name, but the
6049 architecture specific code special-cases it when occurring as the final
6050 character of a symbol (to denote a label), then the macro parameter
6051 replacement code will have no way of knowing that and consider the whole
6052 construct (including the colon) an identifier, and check only this
6053 identifier for being the subject to parameter substitution. So for example
6054 this macro definition:
6062 might not work as expected. Invoking @samp{label foo} might not create a label
6063 called @samp{foo} but instead just insert the text @samp{\l:} into the
6064 assembler source, probably generating an error about an unrecognised
6067 Similarly problems might occur with the period character (@samp{.})
6068 which is often allowed inside opcode names (and hence identifier names). So
6069 for example constructing a macro to build an opcode from a base name and a
6070 length specifier like this:
6073 .macro opcode base length
6078 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
6079 instruction but instead generate some kind of error as the assembler tries to
6080 interpret the text @samp{\base.\length}.
6082 There are several possible ways around this problem:
6085 @item Insert white space
6086 If it is possible to use white space characters then this is the simplest
6095 @item Use @samp{\()}
6096 The string @samp{\()} can be used to separate the end of a macro argument from
6097 the following text. eg:
6100 .macro opcode base length
6105 @item Use the alternate macro syntax mode
6106 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
6107 used as a separator. eg:
6117 Note: this problem of correctly identifying string parameters to pseudo ops
6118 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
6119 and @code{.irpc} (@pxref{Irpc}) as well.
6122 @cindex @code{endm} directive
6123 Mark the end of a macro definition.
6126 @cindex @code{exitm} directive
6127 Exit early from the current macro definition.
6129 @cindex number of macros executed
6130 @cindex macros, count executed
6132 @command{@value{AS}} maintains a counter of how many macros it has
6133 executed in this pseudo-variable; you can copy that number to your
6134 output with @samp{\@@}, but @emph{only within a macro definition}.
6136 @item LOCAL @var{name} [ , @dots{} ]
6137 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
6138 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
6139 @xref{Altmacro,,@code{.altmacro}}.
6143 @section @code{.mri @var{val}}
6145 @cindex @code{mri} directive
6146 @cindex MRI mode, temporarily
6147 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
6148 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
6149 affects code assembled until the next @code{.mri} directive, or until the end
6150 of the file. @xref{M, MRI mode, MRI mode}.
6153 @section @code{.noaltmacro}
6154 Disable alternate macro mode. @xref{Altmacro}.
6157 @section @code{.nolist}
6159 @cindex @code{nolist} directive
6160 @cindex listing control, turning off
6161 Control (in conjunction with the @code{.list} directive) whether or
6162 not assembly listings are generated. These two directives maintain an
6163 internal counter (which is zero initially). @code{.list} increments the
6164 counter, and @code{.nolist} decrements it. Assembly listings are
6165 generated whenever the counter is greater than zero.
6168 @section @code{.nops @var{size}[, @var{control}]}
6170 @cindex @code{nops} directive
6171 @cindex filling memory with no-op instructions
6172 This directive emits @var{size} bytes filled with no-op instructions.
6173 @var{size} is absolute expression, which must be a positve value.
6174 @var{control} controls how no-op instructions should be generated. If
6175 the comma and @var{control} are omitted, @var{control} is assumed to be
6178 Note: For Intel 80386 and AMD x86-64 targets, @var{control} specifies
6179 the size limit of a no-op instruction. The valid values of @var{control}
6180 are between 0 and 4 in 16-bit mode, between 0 and 7 when tuning for
6181 older processors in 32-bit mode, between 0 and 11 in 64-bit mode or when
6182 tuning for newer processors in 32-bit mode. When 0 is used, the no-op
6183 instruction size limit is set to the maximum supported size.
6186 @section @code{.octa @var{bignums}}
6188 @c FIXME: double size emitted for "octa" on some? Or warn?
6189 @cindex @code{octa} directive
6190 @cindex integer, 16-byte
6191 @cindex sixteen byte integer
6192 This directive expects zero or more bignums, separated by commas. For each
6193 bignum, it emits a 16-byte integer.
6195 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
6196 hence @emph{octa}-word for 16 bytes.
6199 @section @code{.offset @var{loc}}
6201 @cindex @code{offset} directive
6202 Set the location counter to @var{loc} in the absolute section. @var{loc} must
6203 be an absolute expression. This directive may be useful for defining
6204 symbols with absolute values. Do not confuse it with the @code{.org}
6208 @section @code{.org @var{new-lc} , @var{fill}}
6210 @cindex @code{org} directive
6211 @cindex location counter, advancing
6212 @cindex advancing location counter
6213 @cindex current address, advancing
6214 Advance the location counter of the current section to
6215 @var{new-lc}. @var{new-lc} is either an absolute expression or an
6216 expression with the same section as the current subsection. That is,
6217 you can't use @code{.org} to cross sections: if @var{new-lc} has the
6218 wrong section, the @code{.org} directive is ignored. To be compatible
6219 with former assemblers, if the section of @var{new-lc} is absolute,
6220 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
6221 is the same as the current subsection.
6223 @code{.org} may only increase the location counter, or leave it
6224 unchanged; you cannot use @code{.org} to move the location counter
6227 @c double negative used below "not undefined" because this is a specific
6228 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
6229 @c section. doc@cygnus.com 18feb91
6230 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
6231 may not be undefined. If you really detest this restriction we eagerly await
6232 a chance to share your improved assembler.
6234 Beware that the origin is relative to the start of the section, not
6235 to the start of the subsection. This is compatible with other
6236 people's assemblers.
6238 When the location counter (of the current subsection) is advanced, the
6239 intervening bytes are filled with @var{fill} which should be an
6240 absolute expression. If the comma and @var{fill} are omitted,
6241 @var{fill} defaults to zero.
6244 @section @code{.p2align[wl] [@var{abs-expr}[, @var{abs-expr}[, @var{abs-expr}]]]}
6246 @cindex padding the location counter given a power of two
6247 @cindex @code{p2align} directive
6248 Pad the location counter (in the current subsection) to a particular
6249 storage boundary. The first expression (which must be absolute) is the
6250 number of low-order zero bits the location counter must have after
6251 advancement. For example @samp{.p2align 3} advances the location
6252 counter until it is a multiple of 8. If the location counter is already a
6253 multiple of 8, no change is needed. If the expression is omitted then a
6254 default value of 0 is used, effectively disabling alignment requirements.
6256 The second expression (also absolute) gives the fill value to be stored in the
6257 padding bytes. It (and the comma) may be omitted. If it is omitted, the
6258 padding bytes are normally zero. However, on most systems, if the section is
6259 marked as containing code and the fill value is omitted, the space is filled
6260 with no-op instructions.
6262 The third expression is also absolute, and is also optional. If it is present,
6263 it is the maximum number of bytes that should be skipped by this alignment
6264 directive. If doing the alignment would require skipping more bytes than the
6265 specified maximum, then the alignment is not done at all. You can omit the
6266 fill value (the second argument) entirely by simply using two commas after the
6267 required alignment; this can be useful if you want the alignment to be filled
6268 with no-op instructions when appropriate.
6270 @cindex @code{p2alignw} directive
6271 @cindex @code{p2alignl} directive
6272 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
6273 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
6274 pattern as a two byte word value. The @code{.p2alignl} directives treats the
6275 fill pattern as a four byte longword value. For example, @code{.p2alignw
6276 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
6277 filled in with the value 0x368d (the exact placement of the bytes depends upon
6278 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
6283 @section @code{.popsection}
6285 @cindex @code{popsection} directive
6286 @cindex Section Stack
6287 This is one of the ELF section stack manipulation directives. The others are
6288 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6289 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
6292 This directive replaces the current section (and subsection) with the top
6293 section (and subsection) on the section stack. This section is popped off the
6299 @section @code{.previous}
6301 @cindex @code{previous} directive
6302 @cindex Section Stack
6303 This is one of the ELF section stack manipulation directives. The others are
6304 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6305 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
6306 (@pxref{PopSection}).
6308 This directive swaps the current section (and subsection) with most recently
6309 referenced section/subsection pair prior to this one. Multiple
6310 @code{.previous} directives in a row will flip between two sections (and their
6311 subsections). For example:
6323 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
6329 # Now in section A subsection 1
6333 # Now in section B subsection 0
6336 # Now in section B subsection 1
6339 # Now in section B subsection 0
6343 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
6344 section B and 0x9abc into subsection 1 of section B.
6346 In terms of the section stack, this directive swaps the current section with
6347 the top section on the section stack.
6351 @section @code{.print @var{string}}
6353 @cindex @code{print} directive
6354 @command{@value{AS}} will print @var{string} on the standard output during
6355 assembly. You must put @var{string} in double quotes.
6359 @section @code{.protected @var{names}}
6361 @cindex @code{protected} directive
6363 This is one of the ELF visibility directives. The other two are
6364 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
6366 This directive overrides the named symbols default visibility (which is set by
6367 their binding: local, global or weak). The directive sets the visibility to
6368 @code{protected} which means that any references to the symbols from within the
6369 components that defines them must be resolved to the definition in that
6370 component, even if a definition in another component would normally preempt
6375 @section @code{.psize @var{lines} , @var{columns}}
6377 @cindex @code{psize} directive
6378 @cindex listing control: paper size
6379 @cindex paper size, for listings
6380 Use this directive to declare the number of lines---and, optionally, the
6381 number of columns---to use for each page, when generating listings.
6383 If you do not use @code{.psize}, listings use a default line-count
6384 of 60. You may omit the comma and @var{columns} specification; the
6385 default width is 200 columns.
6387 @command{@value{AS}} generates formfeeds whenever the specified number of
6388 lines is exceeded (or whenever you explicitly request one, using
6391 If you specify @var{lines} as @code{0}, no formfeeds are generated save
6392 those explicitly specified with @code{.eject}.
6395 @section @code{.purgem @var{name}}
6397 @cindex @code{purgem} directive
6398 Undefine the macro @var{name}, so that later uses of the string will not be
6399 expanded. @xref{Macro}.
6403 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
6405 @cindex @code{pushsection} directive
6406 @cindex Section Stack
6407 This is one of the ELF section stack manipulation directives. The others are
6408 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6409 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6412 This directive pushes the current section (and subsection) onto the
6413 top of the section stack, and then replaces the current section and
6414 subsection with @code{name} and @code{subsection}. The optional
6415 @code{flags}, @code{type} and @code{arguments} are treated the same
6416 as in the @code{.section} (@pxref{Section}) directive.
6420 @section @code{.quad @var{bignums}}
6422 @cindex @code{quad} directive
6423 @code{.quad} expects zero or more bignums, separated by commas. For
6424 each bignum, it emits
6426 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
6427 warning message; and just takes the lowest order 8 bytes of the bignum.
6428 @cindex eight-byte integer
6429 @cindex integer, 8-byte
6431 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
6432 hence @emph{quad}-word for 8 bytes.
6435 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
6436 warning message; and just takes the lowest order 16 bytes of the bignum.
6437 @cindex sixteen-byte integer
6438 @cindex integer, 16-byte
6442 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
6444 @cindex @code{reloc} directive
6445 Generate a relocation at @var{offset} of type @var{reloc_name} with value
6446 @var{expression}. If @var{offset} is a number, the relocation is generated in
6447 the current section. If @var{offset} is an expression that resolves to a
6448 symbol plus offset, the relocation is generated in the given symbol's section.
6449 @var{expression}, if present, must resolve to a symbol plus addend or to an
6450 absolute value, but note that not all targets support an addend. e.g. ELF REL
6451 targets such as i386 store an addend in the section contents rather than in the
6452 relocation. This low level interface does not support addends stored in the
6456 @section @code{.rept @var{count}}
6458 @cindex @code{rept} directive
6459 Repeat the sequence of lines between the @code{.rept} directive and the next
6460 @code{.endr} directive @var{count} times.
6462 For example, assembling
6470 is equivalent to assembling
6478 A count of zero is allowed, but nothing is generated. Negative counts are not
6479 allowed and if encountered will be treated as if they were zero.
6482 @section @code{.sbttl "@var{subheading}"}
6484 @cindex @code{sbttl} directive
6485 @cindex subtitles for listings
6486 @cindex listing control: subtitle
6487 Use @var{subheading} as the title (third line, immediately after the
6488 title line) when generating assembly listings.
6490 This directive affects subsequent pages, as well as the current page if
6491 it appears within ten lines of the top of a page.
6495 @section @code{.scl @var{class}}
6497 @cindex @code{scl} directive
6498 @cindex symbol storage class (COFF)
6499 @cindex COFF symbol storage class
6500 Set the storage-class value for a symbol. This directive may only be
6501 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
6502 whether a symbol is static or external, or it may record further
6503 symbolic debugging information.
6508 @section @code{.section @var{name}}
6510 @cindex named section
6511 Use the @code{.section} directive to assemble the following code into a section
6514 This directive is only supported for targets that actually support arbitrarily
6515 named sections; on @code{a.out} targets, for example, it is not accepted, even
6516 with a standard @code{a.out} section name.
6520 @c only print the extra heading if both COFF and ELF are set
6521 @subheading COFF Version
6524 @cindex @code{section} directive (COFF version)
6525 For COFF targets, the @code{.section} directive is used in one of the following
6529 .section @var{name}[, "@var{flags}"]
6530 .section @var{name}[, @var{subsection}]
6533 If the optional argument is quoted, it is taken as flags to use for the
6534 section. Each flag is a single character. The following flags are recognized:
6538 bss section (uninitialized data)
6540 section is not loaded
6546 exclude section from linking
6552 shared section (meaningful for PE targets)
6554 ignored. (For compatibility with the ELF version)
6556 section is not readable (meaningful for PE targets)
6558 single-digit power-of-two section alignment (GNU extension)
6561 If no flags are specified, the default flags depend upon the section name. If
6562 the section name is not recognized, the default will be for the section to be
6563 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
6564 from the section, rather than adding them, so if they are used on their own it
6565 will be as if no flags had been specified at all.
6567 If the optional argument to the @code{.section} directive is not quoted, it is
6568 taken as a subsection number (@pxref{Sub-Sections}).
6573 @c only print the extra heading if both COFF and ELF are set
6574 @subheading ELF Version
6577 @cindex Section Stack
6578 This is one of the ELF section stack manipulation directives. The others are
6579 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
6580 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
6581 @code{.previous} (@pxref{Previous}).
6583 @cindex @code{section} directive (ELF version)
6584 For ELF targets, the @code{.section} directive is used like this:
6587 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
6590 @anchor{Section Name Substitutions}
6591 @kindex --sectname-subst
6592 @cindex section name substitution
6593 If the @samp{--sectname-subst} command-line option is provided, the @var{name}
6594 argument may contain a substitution sequence. Only @code{%S} is supported
6595 at the moment, and substitutes the current section name. For example:
6598 .macro exception_code
6599 .section %S.exception
6600 [exception code here]
6615 The two @code{exception_code} invocations above would create the
6616 @code{.text.exception} and @code{.init.exception} sections respectively.
6617 This is useful e.g. to discriminate between ancillary sections that are
6618 tied to setup code to be discarded after use from ancillary sections that
6619 need to stay resident without having to define multiple @code{exception_code}
6620 macros just for that purpose.
6622 The optional @var{flags} argument is a quoted string which may contain any
6623 combination of the following characters:
6627 section is allocatable
6629 section is a GNU_MBIND section
6631 section is excluded from executable and shared library.
6633 section references a symbol defined in another section (the linked-to
6634 section) in the same file.
6638 section is executable
6640 section is mergeable
6642 section contains zero terminated strings
6644 section is a member of a section group
6646 section is used for thread-local-storage
6648 section is a member of the previously-current section's group, if any
6649 @item @code{<number>}
6650 a numeric value indicating the bits to be set in the ELF section header's flags
6651 field. Note - if one or more of the alphabetic characters described above is
6652 also included in the flags field, their bit values will be ORed into the
6654 @item @code{<target specific>}
6655 some targets extend this list with their own flag characters
6658 Note - once a section's flags have been set they cannot be changed. There are
6659 a few exceptions to this rule however. Processor and application specific
6660 flags can be added to an already defined section. The @code{.interp},
6661 @code{.strtab} and @code{.symtab} sections can have the allocate flag
6662 (@code{a}) set after they are initially defined, and the @code{.note-GNU-stack}
6663 section may have the executable (@code{x}) flag added.
6665 The optional @var{type} argument may contain one of the following constants:
6669 section contains data
6671 section does not contain data (i.e., section only occupies space)
6673 section contains data which is used by things other than the program
6675 section contains an array of pointers to init functions
6677 section contains an array of pointers to finish functions
6678 @item @@preinit_array
6679 section contains an array of pointers to pre-init functions
6680 @item @@@code{<number>}
6681 a numeric value to be set as the ELF section header's type field.
6682 @item @@@code{<target specific>}
6683 some targets extend this list with their own types
6686 Many targets only support the first three section types. The type may be
6687 enclosed in double quotes if necessary.
6689 Note on targets where the @code{@@} character is the start of a comment (eg
6690 ARM) then another character is used instead. For example the ARM port uses the
6693 Note - some sections, eg @code{.text} and @code{.data} are considered to be
6694 special and have fixed types. Any attempt to declare them with a different
6695 type will generate an error from the assembler.
6697 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6698 be specified as well as an extra argument---@var{entsize}---like this:
6701 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6704 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6705 constants, each @var{entsize} octets long. Sections with both @code{M} and
6706 @code{S} must contain zero terminated strings where each character is
6707 @var{entsize} bytes long. The linker may remove duplicates within sections with
6708 the same name, same entity size and same flags. @var{entsize} must be an
6709 absolute expression. For sections with both @code{M} and @code{S}, a string
6710 which is a suffix of a larger string is considered a duplicate. Thus
6711 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6712 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6714 If @var{flags} contains the @code{o} flag, then the @var{type} argument
6715 must be present along with an additional field like this:
6718 .section @var{name},"@var{flags}"o,@@@var{type},@var{SymbolName}
6721 The @var{SymbolName} field specifies the symbol name which the section
6724 Note: If both the @var{M} and @var{o} flags are present, then the fields
6725 for the Merge flag should come first, like this:
6728 .section @var{name},"@var{flags}"Mo,@@@var{type},@var{entsize},@var{SymbolName}
6731 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6732 be present along with an additional field like this:
6735 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6738 The @var{GroupName} field specifies the name of the section group to which this
6739 particular section belongs. The optional linkage field can contain:
6743 indicates that only one copy of this section should be retained
6748 Note: if both the @var{M} and @var{G} flags are present then the fields for
6749 the Merge flag should come first, like this:
6752 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6755 If both @code{o} flag and @code{G} flag are present, then the
6756 @var{SymbolName} field for @code{o} comes first, like this:
6759 .section @var{name},"@var{flags}"oG,@@@var{type},@var{SymbolName},@var{GroupName}[,@var{linkage}]
6762 If @var{flags} contains the @code{?} symbol then it may not also contain the
6763 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6764 present. Instead, @code{?} says to consider the section that's current before
6765 this directive. If that section used @code{G}, then the new section will use
6766 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6767 If not, then the @code{?} symbol has no effect.
6769 The optional @var{unique,@code{<number>}} argument must come last. It
6770 assigns @var{@code{<number>}} as a unique section ID to distinguish
6771 different sections with the same section name like these:
6774 .section @var{name},"@var{flags}",@@@var{type},@var{unique,@code{<number>}}
6775 .section @var{name},"@var{flags}"G,@@@var{type},@var{GroupName},[@var{linkage}],@var{unique,@code{<number>}}
6776 .section @var{name},"@var{flags}"MG,@@@var{type},@var{entsize},@var{GroupName}[,@var{linkage}],@var{unique,@code{<number>}}
6779 The valid values of @var{@code{<number>}} are between 0 and 4294967295.
6781 If no flags are specified, the default flags depend upon the section name. If
6782 the section name is not recognized, the default will be for the section to have
6783 none of the above flags: it will not be allocated in memory, nor writable, nor
6784 executable. The section will contain data.
6786 For ELF targets, the assembler supports another type of @code{.section}
6787 directive for compatibility with the Solaris assembler:
6790 .section "@var{name}"[, @var{flags}...]
6793 Note that the section name is quoted. There may be a sequence of comma
6798 section is allocatable
6802 section is executable
6804 section is excluded from executable and shared library.
6806 section is used for thread local storage
6809 This directive replaces the current section and subsection. See the
6810 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6811 some examples of how this directive and the other section stack directives
6817 @section @code{.set @var{symbol}, @var{expression}}
6819 @cindex @code{set} directive
6820 @cindex symbol value, setting
6821 Set the value of @var{symbol} to @var{expression}. This
6822 changes @var{symbol}'s value and type to conform to
6823 @var{expression}. If @var{symbol} was flagged as external, it remains
6824 flagged (@pxref{Symbol Attributes}).
6826 You may @code{.set} a symbol many times in the same assembly provided that the
6827 values given to the symbol are constants. Values that are based on expressions
6828 involving other symbols are allowed, but some targets may restrict this to only
6829 being done once per assembly. This is because those targets do not set the
6830 addresses of symbols at assembly time, but rather delay the assignment until a
6831 final link is performed. This allows the linker a chance to change the code in
6832 the files, changing the location of, and the relative distance between, various
6835 If you @code{.set} a global symbol, the value stored in the object
6836 file is the last value stored into it.
6839 On Z80 @code{set} is a real instruction, use @code{.set} or
6840 @samp{@var{symbol} defl @var{expression}} instead.
6844 @section @code{.short @var{expressions}}
6846 @cindex @code{short} directive
6848 @code{.short} is normally the same as @samp{.word}.
6849 @xref{Word,,@code{.word}}.
6851 In some configurations, however, @code{.short} and @code{.word} generate
6852 numbers of different lengths. @xref{Machine Dependencies}.
6856 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6859 This expects zero or more @var{expressions}, and emits
6860 a 16 bit number for each.
6865 @section @code{.single @var{flonums}}
6867 @cindex @code{single} directive
6868 @cindex floating point numbers (single)
6869 This directive assembles zero or more flonums, separated by commas. It
6870 has the same effect as @code{.float}.
6872 The exact kind of floating point numbers emitted depends on how
6873 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6877 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6878 numbers in @sc{ieee} format.
6884 @section @code{.size}
6886 This directive is used to set the size associated with a symbol.
6890 @c only print the extra heading if both COFF and ELF are set
6891 @subheading COFF Version
6894 @cindex @code{size} directive (COFF version)
6895 For COFF targets, the @code{.size} directive is only permitted inside
6896 @code{.def}/@code{.endef} pairs. It is used like this:
6899 .size @var{expression}
6906 @c only print the extra heading if both COFF and ELF are set
6907 @subheading ELF Version
6910 @cindex @code{size} directive (ELF version)
6911 For ELF targets, the @code{.size} directive is used like this:
6914 .size @var{name} , @var{expression}
6917 This directive sets the size associated with a symbol @var{name}.
6918 The size in bytes is computed from @var{expression} which can make use of label
6919 arithmetic. This directive is typically used to set the size of function
6924 @ifclear no-space-dir
6926 @section @code{.skip @var{size} [,@var{fill}]}
6928 @cindex @code{skip} directive
6929 @cindex filling memory
6930 This directive emits @var{size} bytes, each of value @var{fill}. Both
6931 @var{size} and @var{fill} are absolute expressions. If the comma and
6932 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6937 @section @code{.sleb128 @var{expressions}}
6939 @cindex @code{sleb128} directive
6940 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6941 compact, variable length representation of numbers used by the DWARF
6942 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6944 @ifclear no-space-dir
6946 @section @code{.space @var{size} [,@var{fill}]}
6948 @cindex @code{space} directive
6949 @cindex filling memory
6950 This directive emits @var{size} bytes, each of value @var{fill}. Both
6951 @var{size} and @var{fill} are absolute expressions. If the comma
6952 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6957 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6958 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6959 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6960 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6968 @section @code{.stabd, .stabn, .stabs}
6970 @cindex symbolic debuggers, information for
6971 @cindex @code{stab@var{x}} directives
6972 There are three directives that begin @samp{.stab}.
6973 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6974 The symbols are not entered in the @command{@value{AS}} hash table: they
6975 cannot be referenced elsewhere in the source file.
6976 Up to five fields are required:
6980 This is the symbol's name. It may contain any character except
6981 @samp{\000}, so is more general than ordinary symbol names. Some
6982 debuggers used to code arbitrarily complex structures into symbol names
6986 An absolute expression. The symbol's type is set to the low 8 bits of
6987 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6988 and debuggers choke on silly bit patterns.
6991 An absolute expression. The symbol's ``other'' attribute is set to the
6992 low 8 bits of this expression.
6995 An absolute expression. The symbol's descriptor is set to the low 16
6996 bits of this expression.
6999 An absolute expression which becomes the symbol's value.
7002 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
7003 or @code{.stabs} statement, the symbol has probably already been created;
7004 you get a half-formed symbol in your object file. This is
7005 compatible with earlier assemblers!
7008 @cindex @code{stabd} directive
7009 @item .stabd @var{type} , @var{other} , @var{desc}
7011 The ``name'' of the symbol generated is not even an empty string.
7012 It is a null pointer, for compatibility. Older assemblers used a
7013 null pointer so they didn't waste space in object files with empty
7016 The symbol's value is set to the location counter,
7017 relocatably. When your program is linked, the value of this symbol
7018 is the address of the location counter when the @code{.stabd} was
7021 @cindex @code{stabn} directive
7022 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
7023 The name of the symbol is set to the empty string @code{""}.
7025 @cindex @code{stabs} directive
7026 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
7027 All five fields are specified.
7033 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
7034 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
7036 @cindex string, copying to object file
7037 @cindex string8, copying to object file
7038 @cindex string16, copying to object file
7039 @cindex string32, copying to object file
7040 @cindex string64, copying to object file
7041 @cindex @code{string} directive
7042 @cindex @code{string8} directive
7043 @cindex @code{string16} directive
7044 @cindex @code{string32} directive
7045 @cindex @code{string64} directive
7047 Copy the characters in @var{str} to the object file. You may specify more than
7048 one string to copy, separated by commas. Unless otherwise specified for a
7049 particular machine, the assembler marks the end of each string with a 0 byte.
7050 You can use any of the escape sequences described in @ref{Strings,,Strings}.
7052 The variants @code{string16}, @code{string32} and @code{string64} differ from
7053 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
7054 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
7055 are stored in target endianness byte order.
7061 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
7062 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
7067 @section @code{.struct @var{expression}}
7069 @cindex @code{struct} directive
7070 Switch to the absolute section, and set the section offset to @var{expression},
7071 which must be an absolute expression. You might use this as follows:
7080 This would define the symbol @code{field1} to have the value 0, the symbol
7081 @code{field2} to have the value 4, and the symbol @code{field3} to have the
7082 value 8. Assembly would be left in the absolute section, and you would need to
7083 use a @code{.section} directive of some sort to change to some other section
7084 before further assembly.
7088 @section @code{.subsection @var{name}}
7090 @cindex @code{subsection} directive
7091 @cindex Section Stack
7092 This is one of the ELF section stack manipulation directives. The others are
7093 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
7094 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
7097 This directive replaces the current subsection with @code{name}. The current
7098 section is not changed. The replaced subsection is put onto the section stack
7099 in place of the then current top of stack subsection.
7104 @section @code{.symver}
7105 @cindex @code{symver} directive
7106 @cindex symbol versioning
7107 @cindex versions of symbols
7108 Use the @code{.symver} directive to bind symbols to specific version nodes
7109 within a source file. This is only supported on ELF platforms, and is
7110 typically used when assembling files to be linked into a shared library.
7111 There are cases where it may make sense to use this in objects to be bound
7112 into an application itself so as to override a versioned symbol from a
7115 For ELF targets, the @code{.symver} directive can be used like this:
7117 .symver @var{name}, @var{name2@@nodename}[ ,@var{visibility}]
7119 If the original symbol @var{name} is defined within the file
7120 being assembled, the @code{.symver} directive effectively creates a symbol
7121 alias with the name @var{name2@@nodename}, and in fact the main reason that we
7122 just don't try and create a regular alias is that the @var{@@} character isn't
7123 permitted in symbol names. The @var{name2} part of the name is the actual name
7124 of the symbol by which it will be externally referenced. The name @var{name}
7125 itself is merely a name of convenience that is used so that it is possible to
7126 have definitions for multiple versions of a function within a single source
7127 file, and so that the compiler can unambiguously know which version of a
7128 function is being mentioned. The @var{nodename} portion of the alias should be
7129 the name of a node specified in the version script supplied to the linker when
7130 building a shared library. If you are attempting to override a versioned
7131 symbol from a shared library, then @var{nodename} should correspond to the
7132 nodename of the symbol you are trying to override. The optional argument
7133 @var{visibility} updates the visibility of the original symbol. The valid
7134 visibilities are @code{local}, @code{hidden}, and @code{remove}. The
7135 @code{local} visibility makes the original symbol a local symbol
7136 (@pxref{Local}). The @code{hidden} visibility sets the visibility of the
7137 original symbol to @code{hidden} (@pxref{Hidden}). The @code{remove}
7138 visibility removes the original symbol from the symbol table. If visibility
7139 isn't specified, the original symbol is unchanged.
7141 If the symbol @var{name} is not defined within the file being assembled, all
7142 references to @var{name} will be changed to @var{name2@@nodename}. If no
7143 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
7146 Another usage of the @code{.symver} directive is:
7148 .symver @var{name}, @var{name2@@@@nodename}
7150 In this case, the symbol @var{name} must exist and be defined within
7151 the file being assembled. It is similar to @var{name2@@nodename}. The
7152 difference is @var{name2@@@@nodename} will also be used to resolve
7153 references to @var{name2} by the linker.
7155 The third usage of the @code{.symver} directive is:
7157 .symver @var{name}, @var{name2@@@@@@nodename}
7159 When @var{name} is not defined within the
7160 file being assembled, it is treated as @var{name2@@nodename}. When
7161 @var{name} is defined within the file being assembled, the symbol
7162 name, @var{name}, will be changed to @var{name2@@@@nodename}.
7167 @section @code{.tag @var{structname}}
7169 @cindex COFF structure debugging
7170 @cindex structure debugging, COFF
7171 @cindex @code{tag} directive
7172 This directive is generated by compilers to include auxiliary debugging
7173 information in the symbol table. It is only permitted inside
7174 @code{.def}/@code{.endef} pairs. Tags are used to link structure
7175 definitions in the symbol table with instances of those structures.
7179 @section @code{.text @var{subsection}}
7181 @cindex @code{text} directive
7182 Tells @command{@value{AS}} to assemble the following statements onto the end of
7183 the text subsection numbered @var{subsection}, which is an absolute
7184 expression. If @var{subsection} is omitted, subsection number zero
7188 @section @code{.title "@var{heading}"}
7190 @cindex @code{title} directive
7191 @cindex listing control: title line
7192 Use @var{heading} as the title (second line, immediately after the
7193 source file name and pagenumber) when generating assembly listings.
7195 This directive affects subsequent pages, as well as the current page if
7196 it appears within ten lines of the top of a page.
7200 @section @code{.type}
7202 This directive is used to set the type of a symbol.
7206 @c only print the extra heading if both COFF and ELF are set
7207 @subheading COFF Version
7210 @cindex COFF symbol type
7211 @cindex symbol type, COFF
7212 @cindex @code{type} directive (COFF version)
7213 For COFF targets, this directive is permitted only within
7214 @code{.def}/@code{.endef} pairs. It is used like this:
7220 This records the integer @var{int} as the type attribute of a symbol table
7227 @c only print the extra heading if both COFF and ELF are set
7228 @subheading ELF Version
7231 @cindex ELF symbol type
7232 @cindex symbol type, ELF
7233 @cindex @code{type} directive (ELF version)
7234 For ELF targets, the @code{.type} directive is used like this:
7237 .type @var{name} , @var{type description}
7240 This sets the type of symbol @var{name} to be either a
7241 function symbol or an object symbol. There are five different syntaxes
7242 supported for the @var{type description} field, in order to provide
7243 compatibility with various other assemblers.
7245 Because some of the characters used in these syntaxes (such as @samp{@@} and
7246 @samp{#}) are comment characters for some architectures, some of the syntaxes
7247 below do not work on all architectures. The first variant will be accepted by
7248 the GNU assembler on all architectures so that variant should be used for
7249 maximum portability, if you do not need to assemble your code with other
7252 The syntaxes supported are:
7255 .type <name> STT_<TYPE_IN_UPPER_CASE>
7256 .type <name>,#<type>
7257 .type <name>,@@<type>
7258 .type <name>,%<type>
7259 .type <name>,"<type>"
7262 The types supported are:
7267 Mark the symbol as being a function name.
7270 @itemx gnu_indirect_function
7271 Mark the symbol as an indirect function when evaluated during reloc
7272 processing. (This is only supported on assemblers targeting GNU systems).
7276 Mark the symbol as being a data object.
7280 Mark the symbol as being a thread-local data object.
7284 Mark the symbol as being a common data object.
7288 Does not mark the symbol in any way. It is supported just for completeness.
7290 @item gnu_unique_object
7291 Marks the symbol as being a globally unique data object. The dynamic linker
7292 will make sure that in the entire process there is just one symbol with this
7293 name and type in use. (This is only supported on assemblers targeting GNU
7298 Changing between incompatible types other than from/to STT_NOTYPE will
7299 result in a diagnostic. An intermediate change to STT_NOTYPE will silence
7302 Note: Some targets support extra types in addition to those listed above.
7308 @section @code{.uleb128 @var{expressions}}
7310 @cindex @code{uleb128} directive
7311 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
7312 compact, variable length representation of numbers used by the DWARF
7313 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
7317 @section @code{.val @var{addr}}
7319 @cindex @code{val} directive
7320 @cindex COFF value attribute
7321 @cindex value attribute, COFF
7322 This directive, permitted only within @code{.def}/@code{.endef} pairs,
7323 records the address @var{addr} as the value attribute of a symbol table
7329 @section @code{.version "@var{string}"}
7331 @cindex @code{version} directive
7332 This directive creates a @code{.note} section and places into it an ELF
7333 formatted note of type NT_VERSION. The note's name is set to @code{string}.
7338 @section @code{.vtable_entry @var{table}, @var{offset}}
7340 @cindex @code{vtable_entry} directive
7341 This directive finds or creates a symbol @code{table} and creates a
7342 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
7345 @section @code{.vtable_inherit @var{child}, @var{parent}}
7347 @cindex @code{vtable_inherit} directive
7348 This directive finds the symbol @code{child} and finds or creates the symbol
7349 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
7350 parent whose addend is the value of the child symbol. As a special case the
7351 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
7355 @section @code{.warning "@var{string}"}
7356 @cindex warning directive
7357 Similar to the directive @code{.error}
7358 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
7361 @section @code{.weak @var{names}}
7363 @cindex @code{weak} directive
7364 This directive sets the weak attribute on the comma separated list of symbol
7365 @code{names}. If the symbols do not already exist, they will be created.
7367 On COFF targets other than PE, weak symbols are a GNU extension. This
7368 directive sets the weak attribute on the comma separated list of symbol
7369 @code{names}. If the symbols do not already exist, they will be created.
7371 On the PE target, weak symbols are supported natively as weak aliases.
7372 When a weak symbol is created that is not an alias, GAS creates an
7373 alternate symbol to hold the default value.
7376 @section @code{.weakref @var{alias}, @var{target}}
7378 @cindex @code{weakref} directive
7379 This directive creates an alias to the target symbol that enables the symbol to
7380 be referenced with weak-symbol semantics, but without actually making it weak.
7381 If direct references or definitions of the symbol are present, then the symbol
7382 will not be weak, but if all references to it are through weak references, the
7383 symbol will be marked as weak in the symbol table.
7385 The effect is equivalent to moving all references to the alias to a separate
7386 assembly source file, renaming the alias to the symbol in it, declaring the
7387 symbol as weak there, and running a reloadable link to merge the object files
7388 resulting from the assembly of the new source file and the old source file that
7389 had the references to the alias removed.
7391 The alias itself never makes to the symbol table, and is entirely handled
7392 within the assembler.
7395 @section @code{.word @var{expressions}}
7397 @cindex @code{word} directive
7398 This directive expects zero or more @var{expressions}, of any section,
7399 separated by commas.
7402 For each expression, @command{@value{AS}} emits a 32-bit number.
7405 For each expression, @command{@value{AS}} emits a 16-bit number.
7410 The size of the number emitted, and its byte order,
7411 depend on what target computer the assembly is for.
7414 @c on sparc the "special treatment to support compilers" doesn't
7415 @c happen---32-bit addressability, period; no long/short jumps.
7416 @ifset DIFF-TBL-KLUGE
7417 @cindex difference tables altered
7418 @cindex altered difference tables
7420 @emph{Warning: Special Treatment to support Compilers}
7424 Machines with a 32-bit address space, but that do less than 32-bit
7425 addressing, require the following special treatment. If the machine of
7426 interest to you does 32-bit addressing (or doesn't require it;
7427 @pxref{Machine Dependencies}), you can ignore this issue.
7430 In order to assemble compiler output into something that works,
7431 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
7432 Directives of the form @samp{.word sym1-sym2} are often emitted by
7433 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
7434 directive of the form @samp{.word sym1-sym2}, and the difference between
7435 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
7436 creates a @dfn{secondary jump table}, immediately before the next label.
7437 This secondary jump table is preceded by a short-jump to the
7438 first byte after the secondary table. This short-jump prevents the flow
7439 of control from accidentally falling into the new table. Inside the
7440 table is a long-jump to @code{sym2}. The original @samp{.word}
7441 contains @code{sym1} minus the address of the long-jump to
7444 If there were several occurrences of @samp{.word sym1-sym2} before the
7445 secondary jump table, all of them are adjusted. If there was a
7446 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
7447 long-jump to @code{sym4} is included in the secondary jump table,
7448 and the @code{.word} directives are adjusted to contain @code{sym3}
7449 minus the address of the long-jump to @code{sym4}; and so on, for as many
7450 entries in the original jump table as necessary.
7453 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
7454 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
7455 assembly language programmers.
7458 @c end DIFF-TBL-KLUGE
7460 @ifclear no-space-dir
7462 @section @code{.zero @var{size}}
7464 @cindex @code{zero} directive
7465 @cindex filling memory with zero bytes
7466 This directive emits @var{size} 0-valued bytes. @var{size} must be an absolute
7467 expression. This directive is actually an alias for the @samp{.skip} directive
7468 so it can take an optional second argument of the value to store in the bytes
7469 instead of zero. Using @samp{.zero} in this way would be confusing however.
7474 @section @code{.2byte @var{expression} [, @var{expression}]*}
7475 @cindex @code{2byte} directive
7476 @cindex two-byte integer
7477 @cindex integer, 2-byte
7479 This directive expects zero or more expressions, separated by commas. If there
7480 are no expressions then the directive does nothing. Otherwise each expression
7481 is evaluated in turn and placed in the next two bytes of the current output
7482 section, using the endian model of the target. If an expression will not fit
7483 in two bytes, a warning message is displayed and the least significant two
7484 bytes of the expression's value are used. If an expression cannot be evaluated
7485 at assembly time then relocations will be generated in order to compute the
7488 This directive does not apply any alignment before or after inserting the
7489 values. As a result of this, if relocations are generated, they may be
7490 different from those used for inserting values with a guaranteed alignment.
7492 This directive is only available for ELF targets,
7495 @section @code{.4byte @var{expression} [, @var{expression}]*}
7496 @cindex @code{4byte} directive
7497 @cindex four-byte integer
7498 @cindex integer, 4-byte
7500 Like the @option{.2byte} directive, except that it inserts unaligned, four byte
7501 long values into the output.
7504 @section @code{.8byte @var{expression} [, @var{expression}]*}
7505 @cindex @code{8byte} directive
7506 @cindex eight-byte integer
7507 @cindex integer, 8-byte
7509 Like the @option{.2byte} directive, except that it inserts unaligned, eight
7510 byte long bignum values into the output.
7515 @section Deprecated Directives
7517 @cindex deprecated directives
7518 @cindex obsolescent directives
7519 One day these directives won't work.
7520 They are included for compatibility with older assemblers.
7527 @node Object Attributes
7528 @chapter Object Attributes
7529 @cindex object attributes
7531 @command{@value{AS}} assembles source files written for a specific architecture
7532 into object files for that architecture. But not all object files are alike.
7533 Many architectures support incompatible variations. For instance, floating
7534 point arguments might be passed in floating point registers if the object file
7535 requires hardware floating point support---or floating point arguments might be
7536 passed in integer registers if the object file supports processors with no
7537 hardware floating point unit. Or, if two objects are built for different
7538 generations of the same architecture, the combination may require the
7539 newer generation at run-time.
7541 This information is useful during and after linking. At link time,
7542 @command{@value{LD}} can warn about incompatible object files. After link
7543 time, tools like @command{gdb} can use it to process the linked file
7546 Compatibility information is recorded as a series of object attributes. Each
7547 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
7548 string, and indicates who sets the meaning of the tag. The tag is an integer,
7549 and indicates what property the attribute describes. The value may be a string
7550 or an integer, and indicates how the property affects this object. Missing
7551 attributes are the same as attributes with a zero value or empty string value.
7553 Object attributes were developed as part of the ABI for the ARM Architecture.
7554 The file format is documented in @cite{ELF for the ARM Architecture}.
7557 * GNU Object Attributes:: @sc{gnu} Object Attributes
7558 * Defining New Object Attributes:: Defining New Object Attributes
7561 @node GNU Object Attributes
7562 @section @sc{gnu} Object Attributes
7564 The @code{.gnu_attribute} directive records an object attribute
7565 with vendor @samp{gnu}.
7567 Except for @samp{Tag_compatibility}, which has both an integer and a string for
7568 its value, @sc{gnu} attributes have a string value if the tag number is odd and
7569 an integer value if the tag number is even. The second bit (@code{@var{tag} &
7570 2} is set for architecture-independent attributes and clear for
7571 architecture-dependent ones.
7573 @subsection Common @sc{gnu} attributes
7575 These attributes are valid on all architectures.
7578 @item Tag_compatibility (32)
7579 The compatibility attribute takes an integer flag value and a vendor name. If
7580 the flag value is 0, the file is compatible with other toolchains. If it is 1,
7581 then the file is only compatible with the named toolchain. If it is greater
7582 than 1, the file can only be processed by other toolchains under some private
7583 arrangement indicated by the flag value and the vendor name.
7586 @subsection MIPS Attributes
7589 @item Tag_GNU_MIPS_ABI_FP (4)
7590 The floating-point ABI used by this object file. The value will be:
7594 0 for files not affected by the floating-point ABI.
7596 1 for files using the hardware floating-point ABI with a standard
7597 double-precision FPU.
7599 2 for files using the hardware floating-point ABI with a single-precision FPU.
7601 3 for files using the software floating-point ABI.
7603 4 for files using the deprecated hardware floating-point ABI which used 64-bit
7604 floating-point registers, 32-bit general-purpose registers and increased the
7605 number of callee-saved floating-point registers.
7607 5 for files using the hardware floating-point ABI with a double-precision FPU
7608 with either 32-bit or 64-bit floating-point registers and 32-bit
7609 general-purpose registers.
7611 6 for files using the hardware floating-point ABI with 64-bit floating-point
7612 registers and 32-bit general-purpose registers.
7614 7 for files using the hardware floating-point ABI with 64-bit floating-point
7615 registers, 32-bit general-purpose registers and a rule that forbids the
7616 direct use of odd-numbered single-precision floating-point registers.
7620 @subsection PowerPC Attributes
7623 @item Tag_GNU_Power_ABI_FP (4)
7624 The floating-point ABI used by this object file. The value will be:
7628 0 for files not affected by the floating-point ABI.
7630 1 for files using double-precision hardware floating-point ABI.
7632 2 for files using the software floating-point ABI.
7634 3 for files using single-precision hardware floating-point ABI.
7637 @item Tag_GNU_Power_ABI_Vector (8)
7638 The vector ABI used by this object file. The value will be:
7642 0 for files not affected by the vector ABI.
7644 1 for files using general purpose registers to pass vectors.
7646 2 for files using AltiVec registers to pass vectors.
7648 3 for files using SPE registers to pass vectors.
7652 @subsection IBM z Systems Attributes
7655 @item Tag_GNU_S390_ABI_Vector (8)
7656 The vector ABI used by this object file. The value will be:
7660 0 for files not affected by the vector ABI.
7662 1 for files using software vector ABI.
7664 2 for files using hardware vector ABI.
7668 @subsection MSP430 Attributes
7671 @item Tag_GNU_MSP430_Data_Region (4)
7672 The data region used by this object file. The value will be:
7676 0 for files not using the large memory model.
7678 1 for files which have been compiled with the condition that all
7679 data is in the lower memory region, i.e. below address 0x10000.
7681 2 for files which allow data to be placed in the full 20-bit memory range.
7685 @node Defining New Object Attributes
7686 @section Defining New Object Attributes
7688 If you want to define a new @sc{gnu} object attribute, here are the places you
7689 will need to modify. New attributes should be discussed on the @samp{binutils}
7694 This manual, which is the official register of attributes.
7696 The header for your architecture @file{include/elf}, to define the tag.
7698 The @file{bfd} support file for your architecture, to merge the attribute
7699 and issue any appropriate link warnings.
7701 Test cases in @file{ld/testsuite} for merging and link warnings.
7703 @file{binutils/readelf.c} to display your attribute.
7705 GCC, if you want the compiler to mark the attribute automatically.
7711 @node Machine Dependencies
7712 @chapter Machine Dependent Features
7714 @cindex machine dependencies
7715 The machine instruction sets are (almost by definition) different on
7716 each machine where @command{@value{AS}} runs. Floating point representations
7717 vary as well, and @command{@value{AS}} often supports a few additional
7718 directives or command-line options for compatibility with other
7719 assemblers on a particular platform. Finally, some versions of
7720 @command{@value{AS}} support special pseudo-instructions for branch
7723 This chapter discusses most of these differences, though it does not
7724 include details on any machine's instruction set. For details on that
7725 subject, see the hardware manufacturer's manual.
7729 * AArch64-Dependent:: AArch64 Dependent Features
7732 * Alpha-Dependent:: Alpha Dependent Features
7735 * ARC-Dependent:: ARC Dependent Features
7738 * ARM-Dependent:: ARM Dependent Features
7741 * AVR-Dependent:: AVR Dependent Features
7744 * Blackfin-Dependent:: Blackfin Dependent Features
7747 * BPF-Dependent:: BPF Dependent Features
7750 * CR16-Dependent:: CR16 Dependent Features
7753 * CRIS-Dependent:: CRIS Dependent Features
7756 * C-SKY-Dependent:: C-SKY Dependent Features
7759 * D10V-Dependent:: D10V Dependent Features
7762 * D30V-Dependent:: D30V Dependent Features
7765 * Epiphany-Dependent:: EPIPHANY Dependent Features
7768 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7771 * HPPA-Dependent:: HPPA Dependent Features
7774 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
7777 * IA-64-Dependent:: Intel IA-64 Dependent Features
7780 * IP2K-Dependent:: IP2K Dependent Features
7783 * LM32-Dependent:: LM32 Dependent Features
7786 * M32C-Dependent:: M32C Dependent Features
7789 * M32R-Dependent:: M32R Dependent Features
7792 * M68K-Dependent:: M680x0 Dependent Features
7795 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7798 * S12Z-Dependent:: S12Z Dependent Features
7801 * Meta-Dependent :: Meta Dependent Features
7804 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7807 * MIPS-Dependent:: MIPS Dependent Features
7810 * MMIX-Dependent:: MMIX Dependent Features
7813 * MSP430-Dependent:: MSP430 Dependent Features
7816 * NDS32-Dependent:: Andes NDS32 Dependent Features
7819 * NiosII-Dependent:: Altera Nios II Dependent Features
7822 * NS32K-Dependent:: NS32K Dependent Features
7825 * OpenRISC-Dependent:: OpenRISC 1000 Features
7828 * PDP-11-Dependent:: PDP-11 Dependent Features
7831 * PJ-Dependent:: picoJava Dependent Features
7834 * PPC-Dependent:: PowerPC Dependent Features
7837 * PRU-Dependent:: PRU Dependent Features
7840 * RISC-V-Dependent:: RISC-V Dependent Features
7843 * RL78-Dependent:: RL78 Dependent Features
7846 * RX-Dependent:: RX Dependent Features
7849 * S/390-Dependent:: IBM S/390 Dependent Features
7852 * SCORE-Dependent:: SCORE Dependent Features
7855 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7858 * Sparc-Dependent:: SPARC Dependent Features
7861 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7864 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7867 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7870 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7873 * V850-Dependent:: V850 Dependent Features
7876 * Vax-Dependent:: VAX Dependent Features
7879 * Visium-Dependent:: Visium Dependent Features
7882 * WebAssembly-Dependent:: WebAssembly Dependent Features
7885 * XGATE-Dependent:: XGATE Dependent Features
7888 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7891 * Xtensa-Dependent:: Xtensa Dependent Features
7894 * Z80-Dependent:: Z80 Dependent Features
7897 * Z8000-Dependent:: Z8000 Dependent Features
7904 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7905 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7906 @c peculiarity: to preserve cross-references, there must be a node called
7907 @c "Machine Dependencies". Hence the conditional nodenames in each
7908 @c major node below. Node defaulting in makeinfo requires adjacency of
7909 @c node and sectioning commands; hence the repetition of @chapter BLAH
7910 @c in both conditional blocks.
7913 @include c-aarch64.texi
7917 @include c-alpha.texi
7933 @include c-bfin.texi
7941 @include c-cr16.texi
7945 @include c-cris.texi
7949 @include c-csky.texi
7954 @node Machine Dependencies
7955 @chapter Machine Dependent Features
7957 The machine instruction sets are different on each Renesas chip family,
7958 and there are also some syntax differences among the families. This
7959 chapter describes the specific @command{@value{AS}} features for each
7963 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7964 * SH-Dependent:: Renesas SH Dependent Features
7971 @include c-d10v.texi
7975 @include c-d30v.texi
7979 @include c-epiphany.texi
7983 @include c-h8300.texi
7987 @include c-hppa.texi
7991 @include c-i386.texi
7995 @include c-ia64.texi
7999 @include c-ip2k.texi
8003 @include c-lm32.texi
8007 @include c-m32c.texi
8011 @include c-m32r.texi
8015 @include c-m68k.texi
8019 @include c-m68hc11.texi
8023 @include c-s12z.texi
8027 @include c-metag.texi
8031 @include c-microblaze.texi
8035 @include c-mips.texi
8039 @include c-mmix.texi
8043 @include c-msp430.texi
8047 @include c-nds32.texi
8051 @include c-nios2.texi
8055 @include c-ns32k.texi
8059 @include c-or1k.texi
8063 @include c-pdp11.texi
8079 @include c-riscv.texi
8083 @include c-rl78.texi
8091 @include c-s390.texi
8095 @include c-score.texi
8103 @include c-sparc.texi
8107 @include c-tic54x.texi
8111 @include c-tic6x.texi
8115 @include c-tilegx.texi
8119 @include c-tilepro.texi
8123 @include c-v850.texi
8131 @include c-visium.texi
8135 @include c-wasm32.texi
8139 @include c-xgate.texi
8143 @include c-xstormy16.texi
8147 @include c-xtensa.texi
8159 @c reverse effect of @down at top of generic Machine-Dep chapter
8163 @node Reporting Bugs
8164 @chapter Reporting Bugs
8165 @cindex bugs in assembler
8166 @cindex reporting bugs in assembler
8168 Your bug reports play an essential role in making @command{@value{AS}} reliable.
8170 Reporting a bug may help you by bringing a solution to your problem, or it may
8171 not. But in any case the principal function of a bug report is to help the
8172 entire community by making the next version of @command{@value{AS}} work better.
8173 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
8175 In order for a bug report to serve its purpose, you must include the
8176 information that enables us to fix the bug.
8179 * Bug Criteria:: Have you found a bug?
8180 * Bug Reporting:: How to report bugs
8184 @section Have You Found a Bug?
8185 @cindex bug criteria
8187 If you are not sure whether you have found a bug, here are some guidelines:
8190 @cindex fatal signal
8191 @cindex assembler crash
8192 @cindex crash of assembler
8194 If the assembler gets a fatal signal, for any input whatever, that is a
8195 @command{@value{AS}} bug. Reliable assemblers never crash.
8197 @cindex error on valid input
8199 If @command{@value{AS}} produces an error message for valid input, that is a bug.
8201 @cindex invalid input
8203 If @command{@value{AS}} does not produce an error message for invalid input, that
8204 is a bug. However, you should note that your idea of ``invalid input'' might
8205 be our idea of ``an extension'' or ``support for traditional practice''.
8208 If you are an experienced user of assemblers, your suggestions for improvement
8209 of @command{@value{AS}} are welcome in any case.
8213 @section How to Report Bugs
8215 @cindex assembler bugs, reporting
8217 A number of companies and individuals offer support for @sc{gnu} products. If
8218 you obtained @command{@value{AS}} from a support organization, we recommend you
8219 contact that organization first.
8221 You can find contact information for many support companies and
8222 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
8226 In any event, we also recommend that you send bug reports for @command{@value{AS}}
8230 The fundamental principle of reporting bugs usefully is this:
8231 @strong{report all the facts}. If you are not sure whether to state a
8232 fact or leave it out, state it!
8234 Often people omit facts because they think they know what causes the problem
8235 and assume that some details do not matter. Thus, you might assume that the
8236 name of a symbol you use in an example does not matter. Well, probably it does
8237 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
8238 happens to fetch from the location where that name is stored in memory;
8239 perhaps, if the name were different, the contents of that location would fool
8240 the assembler into doing the right thing despite the bug. Play it safe and
8241 give a specific, complete example. That is the easiest thing for you to do,
8242 and the most helpful.
8244 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
8245 it is new to us. Therefore, always write your bug reports on the assumption
8246 that the bug has not been reported previously.
8248 Sometimes people give a few sketchy facts and ask, ``Does this ring a
8249 bell?'' This cannot help us fix a bug, so it is basically useless. We
8250 respond by asking for enough details to enable us to investigate.
8251 You might as well expedite matters by sending them to begin with.
8253 To enable us to fix the bug, you should include all these things:
8257 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
8258 it with the @samp{--version} argument.
8260 Without this, we will not know whether there is any point in looking for
8261 the bug in the current version of @command{@value{AS}}.
8264 Any patches you may have applied to the @command{@value{AS}} source.
8267 The type of machine you are using, and the operating system name and
8271 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
8275 The command arguments you gave the assembler to assemble your example and
8276 observe the bug. To guarantee you will not omit something important, list them
8277 all. A copy of the Makefile (or the output from make) is sufficient.
8279 If we were to try to guess the arguments, we would probably guess wrong
8280 and then we might not encounter the bug.
8283 A complete input file that will reproduce the bug. If the bug is observed when
8284 the assembler is invoked via a compiler, send the assembler source, not the
8285 high level language source. Most compilers will produce the assembler source
8286 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
8287 the options @samp{-v --save-temps}; this will save the assembler source in a
8288 file with an extension of @file{.s}, and also show you exactly how
8289 @command{@value{AS}} is being run.
8292 A description of what behavior you observe that you believe is
8293 incorrect. For example, ``It gets a fatal signal.''
8295 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
8296 will certainly notice it. But if the bug is incorrect output, we might not
8297 notice unless it is glaringly wrong. You might as well not give us a chance to
8300 Even if the problem you experience is a fatal signal, you should still say so
8301 explicitly. Suppose something strange is going on, such as, your copy of
8302 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
8303 library on your system. (This has happened!) Your copy might crash and ours
8304 would not. If you told us to expect a crash, then when ours fails to crash, we
8305 would know that the bug was not happening for us. If you had not told us to
8306 expect a crash, then we would not be able to draw any conclusion from our
8310 If you wish to suggest changes to the @command{@value{AS}} source, send us context
8311 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
8312 option. Always send diffs from the old file to the new file. If you even
8313 discuss something in the @command{@value{AS}} source, refer to it by context, not
8316 The line numbers in our development sources will not match those in your
8317 sources. Your line numbers would convey no useful information to us.
8320 Here are some things that are not necessary:
8324 A description of the envelope of the bug.
8326 Often people who encounter a bug spend a lot of time investigating
8327 which changes to the input file will make the bug go away and which
8328 changes will not affect it.
8330 This is often time consuming and not very useful, because the way we
8331 will find the bug is by running a single example under the debugger
8332 with breakpoints, not by pure deduction from a series of examples.
8333 We recommend that you save your time for something else.
8335 Of course, if you can find a simpler example to report @emph{instead}
8336 of the original one, that is a convenience for us. Errors in the
8337 output will be easier to spot, running under the debugger will take
8338 less time, and so on.
8340 However, simplification is not vital; if you do not want to do this,
8341 report the bug anyway and send us the entire test case you used.
8344 A patch for the bug.
8346 A patch for the bug does help us if it is a good one. But do not omit
8347 the necessary information, such as the test case, on the assumption that
8348 a patch is all we need. We might see problems with your patch and decide
8349 to fix the problem another way, or we might not understand it at all.
8351 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
8352 construct an example that will make the program follow a certain path through
8353 the code. If you do not send us the example, we will not be able to construct
8354 one, so we will not be able to verify that the bug is fixed.
8356 And if we cannot understand what bug you are trying to fix, or why your
8357 patch should be an improvement, we will not install it. A test case will
8358 help us to understand.
8361 A guess about what the bug is or what it depends on.
8363 Such guesses are usually wrong. Even we cannot guess right about such
8364 things without first using the debugger to find the facts.
8367 @node Acknowledgements
8368 @chapter Acknowledgements
8370 If you have contributed to GAS and your name isn't listed here,
8371 it is not meant as a slight. We just don't know about it. Send mail to the
8372 maintainer, and we'll correct the situation. Currently
8374 the maintainer is Nick Clifton (email address @code{nickc@@redhat.com}).
8376 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
8379 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
8380 information and the 68k series machines, most of the preprocessing pass, and
8381 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
8383 K. Richard Pixley maintained GAS for a while, adding various enhancements and
8384 many bug fixes, including merging support for several processors, breaking GAS
8385 up to handle multiple object file format back ends (including heavy rewrite,
8386 testing, an integration of the coff and b.out back ends), adding configuration
8387 including heavy testing and verification of cross assemblers and file splits
8388 and renaming, converted GAS to strictly ANSI C including full prototypes, added
8389 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
8390 port (including considerable amounts of reverse engineering), a SPARC opcode
8391 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
8392 assertions and made them work, much other reorganization, cleanup, and lint.
8394 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
8395 in format-specific I/O modules.
8397 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
8398 has done much work with it since.
8400 The Intel 80386 machine description was written by Eliot Dresselhaus.
8402 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
8404 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
8405 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
8407 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
8408 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
8409 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
8410 support a.out format.
8412 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
8413 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
8414 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
8415 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
8418 John Gilmore built the AMD 29000 support, added @code{.include} support, and
8419 simplified the configuration of which versions accept which directives. He
8420 updated the 68k machine description so that Motorola's opcodes always produced
8421 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
8422 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
8423 cross-compilation support, and one bug in relaxation that took a week and
8424 required the proverbial one-bit fix.
8426 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
8427 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
8428 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
8429 PowerPC assembler, and made a few other minor patches.
8431 Steve Chamberlain made GAS able to generate listings.
8433 Hewlett-Packard contributed support for the HP9000/300.
8435 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
8436 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
8437 formats). This work was supported by both the Center for Software Science at
8438 the University of Utah and Cygnus Support.
8440 Support for ELF format files has been worked on by Mark Eichin of Cygnus
8441 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
8442 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
8443 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
8444 and some initial 64-bit support).
8446 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
8448 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
8449 support for openVMS/Alpha.
8451 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
8454 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
8455 Inc.@: added support for Xtensa processors.
8457 Several engineers at Cygnus Support have also provided many small bug fixes and
8458 configuration enhancements.
8460 Jon Beniston added support for the Lattice Mico32 architecture.
8462 Many others have contributed large or small bugfixes and enhancements. If
8463 you have contributed significant work and are not mentioned on this list, and
8464 want to be, let us know. Some of the history has been lost; we are not
8465 intentionally leaving anyone out.
8467 @node GNU Free Documentation License
8468 @appendix GNU Free Documentation License
8472 @unnumbered AS Index