1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright (C) 1991-2018 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
50 @set abnormal-separator
54 @settitle Using @value{AS}
57 @settitle Using @value{AS} (@value{TARGET})
59 @setchapternewpage odd
64 @c WARE! Some of the machine-dependent sections contain tables of machine
65 @c instructions. Except in multi-column format, these tables look silly.
66 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
67 @c the multi-col format is faked within @example sections.
69 @c Again unfortunately, the natural size that fits on a page, for these tables,
70 @c is different depending on whether or not smallbook is turned on.
71 @c This matters, because of order: text flow switches columns at each page
74 @c The format faked in this source works reasonably well for smallbook,
75 @c not well for the default large-page format. This manual expects that if you
76 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
77 @c tables in question. You can turn on one without the other at your
78 @c discretion, of course.
81 @c the insn tables look just as silly in info files regardless of smallbook,
82 @c might as well show 'em anyways.
86 @dircategory Software development
88 * As: (as). The GNU assembler.
89 * Gas: (as). The GNU assembler.
97 This file documents the GNU Assembler "@value{AS}".
99 @c man begin COPYRIGHT
100 Copyright @copyright{} 1991-2018 Free Software Foundation, Inc.
102 Permission is granted to copy, distribute and/or modify this document
103 under the terms of the GNU Free Documentation License, Version 1.3
104 or any later version published by the Free Software Foundation;
105 with no Invariant Sections, with no Front-Cover Texts, and with no
106 Back-Cover Texts. A copy of the license is included in the
107 section entitled ``GNU Free Documentation License''.
113 @title Using @value{AS}
114 @subtitle The @sc{gnu} Assembler
116 @subtitle for the @value{TARGET} family
118 @ifset VERSION_PACKAGE
120 @subtitle @value{VERSION_PACKAGE}
123 @subtitle Version @value{VERSION}
126 The Free Software Foundation Inc.@: thanks The Nice Computer
127 Company of Australia for loaning Dean Elsner to write the
128 first (Vax) version of @command{as} for Project @sc{gnu}.
129 The proprietors, management and staff of TNCCA thank FSF for
130 distracting the boss while they got some work
133 @author Dean Elsner, Jay Fenlason & friends
137 \hfill {\it Using {\tt @value{AS}}}\par
138 \hfill Edited by Cygnus Support\par
140 %"boxit" macro for figures:
141 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
142 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
143 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
144 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
145 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
148 @vskip 0pt plus 1filll
149 Copyright @copyright{} 1991-2018 Free Software Foundation, Inc.
151 Permission is granted to copy, distribute and/or modify this document
152 under the terms of the GNU Free Documentation License, Version 1.3
153 or any later version published by the Free Software Foundation;
154 with no Invariant Sections, with no Front-Cover Texts, and with no
155 Back-Cover Texts. A copy of the license is included in the
156 section entitled ``GNU Free Documentation License''.
163 @top Using @value{AS}
165 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}}
166 @ifset VERSION_PACKAGE
167 @value{VERSION_PACKAGE}
169 version @value{VERSION}.
171 This version of the file describes @command{@value{AS}} configured to generate
172 code for @value{TARGET} architectures.
175 This document is distributed under the terms of the GNU Free
176 Documentation License. A copy of the license is included in the
177 section entitled ``GNU Free Documentation License''.
180 * Overview:: Overview
181 * Invoking:: Command-Line Options
183 * Sections:: Sections and Relocation
185 * Expressions:: Expressions
186 * Pseudo Ops:: Assembler Directives
188 * Object Attributes:: Object Attributes
190 * Machine Dependencies:: Machine Dependent Features
191 * Reporting Bugs:: Reporting Bugs
192 * Acknowledgements:: Who Did What
193 * GNU Free Documentation License:: GNU Free Documentation License
194 * AS Index:: AS Index
201 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
203 This version of the manual describes @command{@value{AS}} configured to generate
204 code for @value{TARGET} architectures.
208 @cindex invocation summary
209 @cindex option summary
210 @cindex summary of options
211 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
212 see @ref{Invoking,,Command-Line Options}.
214 @c man title AS the portable GNU assembler.
218 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
222 @c We don't use deffn and friends for the following because they seem
223 @c to be limited to one line for the header.
225 @c man begin SYNOPSIS
226 @value{AS} [@b{-a}[@b{cdghlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
227 [@b{--compress-debug-sections}] [@b{--nocompress-debug-sections}]
228 [@b{--debug-prefix-map} @var{old}=@var{new}]
229 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
230 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--gdwarf-sections}]
231 [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
232 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
233 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
234 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}]
235 [@b{--no-pad-sections}]
236 [@b{-o} @var{objfile}] [@b{-R}]
237 [@b{--hash-size}=@var{NUM}] [@b{--reduce-memory-overheads}]
239 [@b{-v}] [@b{-version}] [@b{--version}]
240 [@b{-W}] [@b{--warn}] [@b{--fatal-warnings}] [@b{-w}] [@b{-x}]
241 [@b{-Z}] [@b{@@@var{FILE}}]
242 [@b{--sectname-subst}] [@b{--size-check=[error|warning]}]
243 [@b{--elf-stt-common=[no|yes]}]
244 [@b{--target-help}] [@var{target-options}]
245 [@b{--}|@var{files} @dots{}]
248 @c Target dependent options are listed below. Keep the list sorted.
249 @c Add an empty line for separation.
253 @emph{Target AArch64 options:}
255 [@b{-mabi}=@var{ABI}]
259 @emph{Target Alpha options:}
261 [@b{-mdebug} | @b{-no-mdebug}]
262 [@b{-replace} | @b{-noreplace}]
263 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
264 [@b{-F}] [@b{-32addr}]
268 @emph{Target ARC options:}
269 [@b{-mcpu=@var{cpu}}]
270 [@b{-mA6}|@b{-mARC600}|@b{-mARC601}|@b{-mA7}|@b{-mARC700}|@b{-mEM}|@b{-mHS}]
277 @emph{Target ARM options:}
278 @c Don't document the deprecated options
279 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
280 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
281 [@b{-mfpu}=@var{floating-point-format}]
282 [@b{-mfloat-abi}=@var{abi}]
283 [@b{-meabi}=@var{ver}]
286 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
287 @b{-mapcs-reentrant}]
288 [@b{-mthumb-interwork}] [@b{-k}]
292 @emph{Target Blackfin options:}
293 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
300 @emph{Target CRIS options:}
301 [@b{--underscore} | @b{--no-underscore}]
303 [@b{--emulation=criself} | @b{--emulation=crisaout}]
304 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
305 @c Deprecated -- deliberately not documented.
310 @emph{Target D10V options:}
315 @emph{Target D30V options:}
316 [@b{-O}|@b{-n}|@b{-N}]
320 @emph{Target EPIPHANY options:}
321 [@b{-mepiphany}|@b{-mepiphany16}]
325 @emph{Target H8/300 options:}
329 @c HPPA has no machine-dependent assembler options (yet).
333 @emph{Target i386 options:}
334 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
335 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
339 @emph{Target IA-64 options:}
340 [@b{-mconstant-gp}|@b{-mauto-pic}]
341 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
343 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
344 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
345 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
346 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
350 @emph{Target IP2K options:}
351 [@b{-mip2022}|@b{-mip2022ext}]
355 @emph{Target M32C options:}
356 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
360 @emph{Target M32R options:}
361 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
366 @emph{Target M680X0 options:}
367 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
371 @emph{Target M68HC11 options:}
372 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}|@b{-mm9s12x}|@b{-mm9s12xg}]
373 [@b{-mshort}|@b{-mlong}]
374 [@b{-mshort-double}|@b{-mlong-double}]
375 [@b{--force-long-branches}] [@b{--short-branches}]
376 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
377 [@b{--print-opcodes}] [@b{--generate-example}]
381 @emph{Target MCORE options:}
382 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
383 [@b{-mcpu=[210|340]}]
387 @emph{Target Meta options:}
388 [@b{-mcpu=@var{cpu}}] [@b{-mfpu=@var{cpu}}] [@b{-mdsp=@var{cpu}}]
391 @emph{Target MICROBLAZE options:}
392 @c MicroBlaze has no machine-dependent assembler options.
396 @emph{Target MIPS options:}
397 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
398 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
399 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
400 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
401 [@b{-mfp64}] [@b{-mgp64}] [@b{-mfpxx}]
402 [@b{-modd-spreg}] [@b{-mno-odd-spreg}]
403 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
404 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
405 [@b{-mips32r3}] [@b{-mips32r5}] [@b{-mips32r6}] [@b{-mips64}] [@b{-mips64r2}]
406 [@b{-mips64r3}] [@b{-mips64r5}] [@b{-mips64r6}]
407 [@b{-construct-floats}] [@b{-no-construct-floats}]
408 [@b{-mignore-branch-isa}] [@b{-mno-ignore-branch-isa}]
409 [@b{-mnan=@var{encoding}}]
410 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
411 [@b{-mips16}] [@b{-no-mips16}]
412 [@b{-mmips16e2}] [@b{-mno-mips16e2}]
413 [@b{-mmicromips}] [@b{-mno-micromips}]
414 [@b{-msmartmips}] [@b{-mno-smartmips}]
415 [@b{-mips3d}] [@b{-no-mips3d}]
416 [@b{-mdmx}] [@b{-no-mdmx}]
417 [@b{-mdsp}] [@b{-mno-dsp}]
418 [@b{-mdspr2}] [@b{-mno-dspr2}]
419 [@b{-mdspr3}] [@b{-mno-dspr3}]
420 [@b{-mmsa}] [@b{-mno-msa}]
421 [@b{-mxpa}] [@b{-mno-xpa}]
422 [@b{-mmt}] [@b{-mno-mt}]
423 [@b{-mmcu}] [@b{-mno-mcu}]
424 [@b{-minsn32}] [@b{-mno-insn32}]
425 [@b{-mfix7000}] [@b{-mno-fix7000}]
426 [@b{-mfix-rm7000}] [@b{-mno-fix-rm7000}]
427 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
428 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
429 [@b{-mdebug}] [@b{-no-mdebug}]
430 [@b{-mpdr}] [@b{-mno-pdr}]
434 @emph{Target MMIX options:}
435 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
436 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
437 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
438 [@b{--linker-allocated-gregs}]
442 @emph{Target Nios II options:}
443 [@b{-relax-all}] [@b{-relax-section}] [@b{-no-relax}]
448 @emph{Target NDS32 options:}
449 [@b{-EL}] [@b{-EB}] [@b{-O}] [@b{-Os}] [@b{-mcpu=@var{cpu}}]
450 [@b{-misa=@var{isa}}] [@b{-mabi=@var{abi}}] [@b{-mall-ext}]
451 [@b{-m[no-]16-bit}] [@b{-m[no-]perf-ext}] [@b{-m[no-]perf2-ext}]
452 [@b{-m[no-]string-ext}] [@b{-m[no-]dsp-ext}] [@b{-m[no-]mac}] [@b{-m[no-]div}]
453 [@b{-m[no-]audio-isa-ext}] [@b{-m[no-]fpu-sp-ext}] [@b{-m[no-]fpu-dp-ext}]
454 [@b{-m[no-]fpu-fma}] [@b{-mfpu-freg=@var{FREG}}] [@b{-mreduced-regs}]
455 [@b{-mfull-regs}] [@b{-m[no-]dx-regs}] [@b{-mpic}] [@b{-mno-relax}]
460 @emph{Target PDP11 options:}
461 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
462 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
463 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
467 @emph{Target picoJava options:}
472 @emph{Target PowerPC options:}
474 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
475 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
476 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
477 @b{-mbooke}|@b{-mpower4}|@b{-mpwr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
478 @b{-mpower7}|@b{-mpwr7}|@b{-mpower8}|@b{-mpwr8}|@b{-mpower9}|@b{-mpwr9}@b{-ma2}|
479 @b{-mcell}|@b{-mspe}|@b{-mspe2}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
480 [@b{-many}] [@b{-maltivec}|@b{-mvsx}|@b{-mhtm}|@b{-mvle}]
481 [@b{-mregnames}|@b{-mno-regnames}]
482 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
483 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
484 [@b{-msolaris}|@b{-mno-solaris}]
485 [@b{-nops=@var{count}}]
489 @emph{Target PRU options:}
492 [@b{-mno-warn-regname-label}]
496 @emph{Target RISC-V options:}
497 [@b{-fpic}|@b{-fPIC}|@b{-fno-pic}]
498 [@b{-march}=@var{ISA}]
499 [@b{-mabi}=@var{ABI}]
503 @emph{Target RL78 options:}
505 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
509 @emph{Target RX options:}
510 [@b{-mlittle-endian}|@b{-mbig-endian}]
511 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
512 [@b{-muse-conventional-section-names}]
513 [@b{-msmall-data-limit}]
516 [@b{-mint-register=@var{number}}]
517 [@b{-mgcc-abi}|@b{-mrx-abi}]
521 @emph{Target s390 options:}
522 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
523 [@b{-mregnames}|@b{-mno-regnames}]
524 [@b{-mwarn-areg-zero}]
528 @emph{Target SCORE options:}
529 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
530 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
531 [@b{-march=score7}][@b{-march=score3}]
532 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
536 @emph{Target SPARC options:}
537 @c The order here is important. See c-sparc.texi.
538 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Aleon}|@b{-Asparclet}|@b{-Asparclite}
539 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av8plusb}|@b{-Av8plusc}|@b{-Av8plusd}
540 @b{-Av8plusv}|@b{-Av8plusm}|@b{-Av9}|@b{-Av9a}|@b{-Av9b}|@b{-Av9c}
541 @b{-Av9d}|@b{-Av9e}|@b{-Av9v}|@b{-Av9m}|@b{-Asparc}|@b{-Asparcvis}
542 @b{-Asparcvis2}|@b{-Asparcfmaf}|@b{-Asparcima}|@b{-Asparcvis3}
543 @b{-Asparcvisr}|@b{-Asparc5}]
544 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}]|@b{-xarch=v8plusb}|@b{-xarch=v8plusc}
545 @b{-xarch=v8plusd}|@b{-xarch=v8plusv}|@b{-xarch=v8plusm}|@b{-xarch=v9}
546 @b{-xarch=v9a}|@b{-xarch=v9b}|@b{-xarch=v9c}|@b{-xarch=v9d}|@b{-xarch=v9e}
547 @b{-xarch=v9v}|@b{-xarch=v9m}|@b{-xarch=sparc}|@b{-xarch=sparcvis}
548 @b{-xarch=sparcvis2}|@b{-xarch=sparcfmaf}|@b{-xarch=sparcima}
549 @b{-xarch=sparcvis3}|@b{-xarch=sparcvisr}|@b{-xarch=sparc5}
552 [@b{--enforce-aligned-data}][@b{--dcti-couples-detect}]
556 @emph{Target TIC54X options:}
557 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
558 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
562 @emph{Target TIC6X options:}
563 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
564 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
565 [@b{-mpic}|@b{-mno-pic}]
569 @emph{Target TILE-Gx options:}
570 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
573 @c TILEPro has no machine-dependent assembler options
577 @emph{Target Visium options:}
578 [@b{-mtune=@var{arch}}]
582 @emph{Target Xtensa options:}
583 [@b{--[no-]text-section-literals}] [@b{--[no-]auto-litpools}]
584 [@b{--[no-]absolute-literals}]
585 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
586 [@b{--[no-]transform}]
587 [@b{--rename-section} @var{oldname}=@var{newname}]
588 [@b{--[no-]trampolines}]
592 @emph{Target Z80 options:}
593 [@b{-z80}] [@b{-r800}]
594 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
595 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
596 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
597 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
598 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
599 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
603 @c Z8000 has no machine-dependent assembler options
612 @include at-file.texi
615 Turn on listings, in any of a variety of ways:
619 omit false conditionals
622 omit debugging directives
625 include general information, like @value{AS} version and options passed
628 include high-level source
634 include macro expansions
637 omit forms processing
643 set the name of the listing file
646 You may combine these options; for example, use @samp{-aln} for assembly
647 listing without forms processing. The @samp{=file} option, if used, must be
648 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
651 Begin in alternate macro mode.
653 @xref{Altmacro,,@code{.altmacro}}.
656 @item --compress-debug-sections
657 Compress DWARF debug sections using zlib with SHF_COMPRESSED from the
658 ELF ABI. The resulting object file may not be compatible with older
659 linkers and object file utilities. Note if compression would make a
660 given section @emph{larger} then it is not compressed.
663 @cindex @samp{--compress-debug-sections=} option
664 @item --compress-debug-sections=none
665 @itemx --compress-debug-sections=zlib
666 @itemx --compress-debug-sections=zlib-gnu
667 @itemx --compress-debug-sections=zlib-gabi
668 These options control how DWARF debug sections are compressed.
669 @option{--compress-debug-sections=none} is equivalent to
670 @option{--nocompress-debug-sections}.
671 @option{--compress-debug-sections=zlib} and
672 @option{--compress-debug-sections=zlib-gabi} are equivalent to
673 @option{--compress-debug-sections}.
674 @option{--compress-debug-sections=zlib-gnu} compresses DWARF debug
675 sections using zlib. The debug sections are renamed to begin with
676 @samp{.zdebug}. Note if compression would make a given section
677 @emph{larger} then it is not compressed nor renamed.
681 @item --nocompress-debug-sections
682 Do not compress DWARF debug sections. This is usually the default for all
683 targets except the x86/x86_64, but a configure time option can be used to
687 Ignored. This option is accepted for script compatibility with calls to
690 @item --debug-prefix-map @var{old}=@var{new}
691 When assembling files in directory @file{@var{old}}, record debugging
692 information describing them as in @file{@var{new}} instead.
694 @item --defsym @var{sym}=@var{value}
695 Define the symbol @var{sym} to be @var{value} before assembling the input file.
696 @var{value} must be an integer constant. As in C, a leading @samp{0x}
697 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
698 value. The value of the symbol can be overridden inside a source file via the
699 use of a @code{.set} pseudo-op.
702 ``fast''---skip whitespace and comment preprocessing (assume source is
707 Generate debugging information for each assembler source line using whichever
708 debug format is preferred by the target. This currently means either STABS,
712 Generate stabs debugging information for each assembler line. This
713 may help debugging assembler code, if the debugger can handle it.
716 Generate stabs debugging information for each assembler line, with GNU
717 extensions that probably only gdb can handle, and that could make other
718 debuggers crash or refuse to read your program. This
719 may help debugging assembler code. Currently the only GNU extension is
720 the location of the current working directory at assembling time.
723 Generate DWARF2 debugging information for each assembler line. This
724 may help debugging assembler code, if the debugger can handle it. Note---this
725 option is only supported by some targets, not all of them.
727 @item --gdwarf-sections
728 Instead of creating a .debug_line section, create a series of
729 .debug_line.@var{foo} sections where @var{foo} is the name of the
730 corresponding code section. For example a code section called @var{.text.func}
731 will have its dwarf line number information placed into a section called
732 @var{.debug_line.text.func}. If the code section is just called @var{.text}
733 then debug line section will still be called just @var{.debug_line} without any
737 @item --size-check=error
738 @itemx --size-check=warning
739 Issue an error or warning for invalid ELF .size directive.
741 @item --elf-stt-common=no
742 @itemx --elf-stt-common=yes
743 These options control whether the ELF assembler should generate common
744 symbols with the @code{STT_COMMON} type. The default can be controlled
745 by a configure option @option{--enable-elf-stt-common}.
749 Print a summary of the command line options and exit.
752 Print a summary of all target specific options and exit.
755 Add directory @var{dir} to the search list for @code{.include} directives.
758 Don't warn about signed overflow.
761 @ifclear DIFF-TBL-KLUGE
762 This option is accepted but has no effect on the @value{TARGET} family.
764 @ifset DIFF-TBL-KLUGE
765 Issue warnings when difference tables altered for long displacements.
770 Keep (in the symbol table) local symbols. These symbols start with
771 system-specific local label prefixes, typically @samp{.L} for ELF systems
772 or @samp{L} for traditional a.out systems.
777 @item --listing-lhs-width=@var{number}
778 Set the maximum width, in words, of the output data column for an assembler
779 listing to @var{number}.
781 @item --listing-lhs-width2=@var{number}
782 Set the maximum width, in words, of the output data column for continuation
783 lines in an assembler listing to @var{number}.
785 @item --listing-rhs-width=@var{number}
786 Set the maximum width of an input source line, as displayed in a listing, to
789 @item --listing-cont-lines=@var{number}
790 Set the maximum number of lines printed in a listing for a single line of input
793 @item --no-pad-sections
794 Stop the assembler for padding the ends of output sections to the alignment
795 of that section. The default is to pad the sections, but this can waste space
796 which might be needed on targets which have tight memory constraints.
798 @item -o @var{objfile}
799 Name the object-file output from @command{@value{AS}} @var{objfile}.
802 Fold the data section into the text section.
804 @item --hash-size=@var{number}
805 Set the default size of GAS's hash tables to a prime number close to
806 @var{number}. Increasing this value can reduce the length of time it takes the
807 assembler to perform its tasks, at the expense of increasing the assembler's
808 memory requirements. Similarly reducing this value can reduce the memory
809 requirements at the expense of speed.
811 @item --reduce-memory-overheads
812 This option reduces GAS's memory requirements, at the expense of making the
813 assembly processes slower. Currently this switch is a synonym for
814 @samp{--hash-size=4051}, but in the future it may have other effects as well.
817 @item --sectname-subst
818 Honor substitution sequences in section names.
820 @xref{Section Name Substitutions,,@code{.section @var{name}}}.
825 Print the maximum space (in bytes) and total time (in seconds) used by
828 @item --strip-local-absolute
829 Remove local absolute symbols from the outgoing symbol table.
833 Print the @command{as} version.
836 Print the @command{as} version and exit.
840 Suppress warning messages.
842 @item --fatal-warnings
843 Treat warnings as errors.
846 Don't suppress warning messages or treat them as errors.
855 Generate an object file even after errors.
857 @item -- | @var{files} @dots{}
858 Standard input, or source files to assemble.
866 @xref{AArch64 Options}, for the options available when @value{AS} is configured
867 for the 64-bit mode of the ARM Architecture (AArch64).
872 The following options are available when @value{AS} is configured for the
873 64-bit mode of the ARM Architecture (AArch64).
876 @include c-aarch64.texi
877 @c ended inside the included file
885 @xref{Alpha Options}, for the options available when @value{AS} is configured
886 for an Alpha processor.
891 The following options are available when @value{AS} is configured for an Alpha
895 @include c-alpha.texi
896 @c ended inside the included file
903 The following options are available when @value{AS} is configured for an ARC
907 @item -mcpu=@var{cpu}
908 This option selects the core processor variant.
910 Select either big-endian (-EB) or little-endian (-EL) output.
912 Enable Code Density extenssion instructions.
917 The following options are available when @value{AS} is configured for the ARM
921 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
922 Specify which ARM processor variant is the target.
923 @item -march=@var{architecture}[+@var{extension}@dots{}]
924 Specify which ARM architecture variant is used by the target.
925 @item -mfpu=@var{floating-point-format}
926 Select which Floating Point architecture is the target.
927 @item -mfloat-abi=@var{abi}
928 Select which floating point ABI is in use.
930 Enable Thumb only instruction decoding.
931 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
932 Select which procedure calling convention is in use.
934 Select either big-endian (-EB) or little-endian (-EL) output.
935 @item -mthumb-interwork
936 Specify that the code has been generated with interworking between Thumb and
939 Turns on CodeComposer Studio assembly syntax compatibility mode.
941 Specify that PIC code has been generated.
949 @xref{Blackfin Options}, for the options available when @value{AS} is
950 configured for the Blackfin processor family.
955 The following options are available when @value{AS} is configured for
956 the Blackfin processor family.
960 @c ended inside the included file
967 See the info pages for documentation of the CRIS-specific options.
971 The following options are available when @value{AS} is configured for
974 @cindex D10V optimization
975 @cindex optimization, D10V
977 Optimize output by parallelizing instructions.
982 The following options are available when @value{AS} is configured for a D30V
985 @cindex D30V optimization
986 @cindex optimization, D30V
988 Optimize output by parallelizing instructions.
992 Warn when nops are generated.
994 @cindex D30V nops after 32-bit multiply
996 Warn when a nop after a 32-bit multiply instruction is generated.
1002 The following options are available when @value{AS} is configured for the
1003 Adapteva EPIPHANY series.
1006 @xref{Epiphany Options}, for the options available when @value{AS} is
1007 configured for an Epiphany processor.
1011 @c man begin OPTIONS
1012 The following options are available when @value{AS} is configured for
1013 an Epiphany processor.
1015 @c man begin INCLUDE
1016 @include c-epiphany.texi
1017 @c ended inside the included file
1025 @xref{H8/300 Options}, for the options available when @value{AS} is configured
1026 for an H8/300 processor.
1030 @c man begin OPTIONS
1031 The following options are available when @value{AS} is configured for an H8/300
1034 @c man begin INCLUDE
1035 @include c-h8300.texi
1036 @c ended inside the included file
1044 @xref{i386-Options}, for the options available when @value{AS} is
1045 configured for an i386 processor.
1049 @c man begin OPTIONS
1050 The following options are available when @value{AS} is configured for
1053 @c man begin INCLUDE
1054 @include c-i386.texi
1055 @c ended inside the included file
1060 @c man begin OPTIONS
1062 The following options are available when @value{AS} is configured for the
1068 Specifies that the extended IP2022 instructions are allowed.
1071 Restores the default behaviour, which restricts the permitted instructions to
1072 just the basic IP2022 ones.
1078 The following options are available when @value{AS} is configured for the
1079 Renesas M32C and M16C processors.
1084 Assemble M32C instructions.
1087 Assemble M16C instructions (the default).
1090 Enable support for link-time relaxations.
1093 Support H'00 style hex constants in addition to 0x00 style.
1099 The following options are available when @value{AS} is configured for the
1100 Renesas M32R (formerly Mitsubishi M32R) series.
1105 Specify which processor in the M32R family is the target. The default
1106 is normally the M32R, but this option changes it to the M32RX.
1108 @item --warn-explicit-parallel-conflicts or --Wp
1109 Produce warning messages when questionable parallel constructs are
1112 @item --no-warn-explicit-parallel-conflicts or --Wnp
1113 Do not produce warning messages when questionable parallel constructs are
1120 The following options are available when @value{AS} is configured for the
1121 Motorola 68000 series.
1126 Shorten references to undefined symbols, to one word instead of two.
1128 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1129 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1130 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1131 Specify what processor in the 68000 family is the target. The default
1132 is normally the 68020, but this can be changed at configuration time.
1134 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1135 The target machine does (or does not) have a floating-point coprocessor.
1136 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1137 the basic 68000 is not compatible with the 68881, a combination of the
1138 two can be specified, since it's possible to do emulation of the
1139 coprocessor instructions with the main processor.
1141 @item -m68851 | -mno-68851
1142 The target machine does (or does not) have a memory-management
1143 unit coprocessor. The default is to assume an MMU for 68020 and up.
1151 @xref{Nios II Options}, for the options available when @value{AS} is configured
1152 for an Altera Nios II processor.
1156 @c man begin OPTIONS
1157 The following options are available when @value{AS} is configured for an
1158 Altera Nios II processor.
1160 @c man begin INCLUDE
1161 @include c-nios2.texi
1162 @c ended inside the included file
1168 For details about the PDP-11 machine dependent features options,
1169 see @ref{PDP-11-Options}.
1172 @item -mpic | -mno-pic
1173 Generate position-independent (or position-dependent) code. The
1174 default is @option{-mpic}.
1177 @itemx -mall-extensions
1178 Enable all instruction set extensions. This is the default.
1180 @item -mno-extensions
1181 Disable all instruction set extensions.
1183 @item -m@var{extension} | -mno-@var{extension}
1184 Enable (or disable) a particular instruction set extension.
1187 Enable the instruction set extensions supported by a particular CPU, and
1188 disable all other extensions.
1190 @item -m@var{machine}
1191 Enable the instruction set extensions supported by a particular machine
1192 model, and disable all other extensions.
1198 The following options are available when @value{AS} is configured for
1199 a picoJava processor.
1203 @cindex PJ endianness
1204 @cindex endianness, PJ
1205 @cindex big endian output, PJ
1207 Generate ``big endian'' format output.
1209 @cindex little endian output, PJ
1211 Generate ``little endian'' format output.
1219 @xref{PRU Options}, for the options available when @value{AS} is configured
1220 for a PRU processor.
1224 @c man begin OPTIONS
1225 The following options are available when @value{AS} is configured for a
1228 @c man begin INCLUDE
1230 @c ended inside the included file
1235 The following options are available when @value{AS} is configured for the
1236 Motorola 68HC11 or 68HC12 series.
1240 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1241 Specify what processor is the target. The default is
1242 defined by the configuration option when building the assembler.
1244 @item --xgate-ramoffset
1245 Instruct the linker to offset RAM addresses from S12X address space into
1246 XGATE address space.
1249 Specify to use the 16-bit integer ABI.
1252 Specify to use the 32-bit integer ABI.
1254 @item -mshort-double
1255 Specify to use the 32-bit double ABI.
1258 Specify to use the 64-bit double ABI.
1260 @item --force-long-branches
1261 Relative branches are turned into absolute ones. This concerns
1262 conditional branches, unconditional branches and branches to a
1265 @item -S | --short-branches
1266 Do not turn relative branches into absolute ones
1267 when the offset is out of range.
1269 @item --strict-direct-mode
1270 Do not turn the direct addressing mode into extended addressing mode
1271 when the instruction does not support direct addressing mode.
1273 @item --print-insn-syntax
1274 Print the syntax of instruction in case of error.
1276 @item --print-opcodes
1277 Print the list of instructions with syntax and then exit.
1279 @item --generate-example
1280 Print an example of instruction for each possible instruction and then exit.
1281 This option is only useful for testing @command{@value{AS}}.
1287 The following options are available when @command{@value{AS}} is configured
1288 for the SPARC architecture:
1291 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1292 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1293 Explicitly select a variant of the SPARC architecture.
1295 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1296 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1298 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1299 UltraSPARC extensions.
1301 @item -xarch=v8plus | -xarch=v8plusa
1302 For compatibility with the Solaris v9 assembler. These options are
1303 equivalent to -Av8plus and -Av8plusa, respectively.
1306 Warn when the assembler switches to another architecture.
1311 The following options are available when @value{AS} is configured for the 'c54x
1316 Enable extended addressing mode. All addresses and relocations will assume
1317 extended addressing (usually 23 bits).
1318 @item -mcpu=@var{CPU_VERSION}
1319 Sets the CPU version being compiled for.
1320 @item -merrors-to-file @var{FILENAME}
1321 Redirect error output to a file, for broken systems which don't support such
1322 behaviour in the shell.
1327 @c man begin OPTIONS
1328 The following options are available when @value{AS} is configured for
1333 This option sets the largest size of an object that can be referenced
1334 implicitly with the @code{gp} register. It is only accepted for targets that
1335 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1337 @cindex MIPS endianness
1338 @cindex endianness, MIPS
1339 @cindex big endian output, MIPS
1341 Generate ``big endian'' format output.
1343 @cindex little endian output, MIPS
1345 Generate ``little endian'' format output.
1363 Generate code for a particular MIPS Instruction Set Architecture level.
1364 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1365 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1366 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1367 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips32r3},
1368 @samp{-mips32r5}, @samp{-mips32r6}, @samp{-mips64}, @samp{-mips64r2},
1369 @samp{-mips64r3}, @samp{-mips64r5}, and @samp{-mips64r6} correspond to generic
1370 MIPS V, MIPS32, MIPS32 Release 2, MIPS32 Release 3, MIPS32 Release 5, MIPS32
1371 Release 6, MIPS64, MIPS64 Release 2, MIPS64 Release 3, MIPS64 Release 5, and
1372 MIPS64 Release 6 ISA processors, respectively.
1374 @item -march=@var{cpu}
1375 Generate code for a particular MIPS CPU.
1377 @item -mtune=@var{cpu}
1378 Schedule and tune for a particular MIPS CPU.
1382 Cause nops to be inserted if the read of the destination register
1383 of an mfhi or mflo instruction occurs in the following two instructions.
1386 @itemx -mno-fix-rm7000
1387 Cause nops to be inserted if a dmult or dmultu instruction is
1388 followed by a load instruction.
1392 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1393 section instead of the standard ELF .stabs sections.
1397 Control generation of @code{.pdr} sections.
1401 The register sizes are normally inferred from the ISA and ABI, but these
1402 flags force a certain group of registers to be treated as 32 bits wide at
1403 all times. @samp{-mgp32} controls the size of general-purpose registers
1404 and @samp{-mfp32} controls the size of floating-point registers.
1408 The register sizes are normally inferred from the ISA and ABI, but these
1409 flags force a certain group of registers to be treated as 64 bits wide at
1410 all times. @samp{-mgp64} controls the size of general-purpose registers
1411 and @samp{-mfp64} controls the size of floating-point registers.
1414 The register sizes are normally inferred from the ISA and ABI, but using
1415 this flag in combination with @samp{-mabi=32} enables an ABI variant
1416 which will operate correctly with floating-point registers which are
1420 @itemx -mno-odd-spreg
1421 Enable use of floating-point operations on odd-numbered single-precision
1422 registers when supported by the ISA. @samp{-mfpxx} implies
1423 @samp{-mno-odd-spreg}, otherwise the default is @samp{-modd-spreg}.
1427 Generate code for the MIPS 16 processor. This is equivalent to putting
1428 @code{.module mips16} at the start of the assembly file. @samp{-no-mips16}
1429 turns off this option.
1432 @itemx -mno-mips16e2
1433 Enable the use of MIPS16e2 instructions in MIPS16 mode. This is equivalent
1434 to putting @code{.module mips16e2} at the start of the assembly file.
1435 @samp{-mno-mips16e2} turns off this option.
1438 @itemx -mno-micromips
1439 Generate code for the microMIPS processor. This is equivalent to putting
1440 @code{.module micromips} at the start of the assembly file.
1441 @samp{-mno-micromips} turns off this option. This is equivalent to putting
1442 @code{.module nomicromips} at the start of the assembly file.
1445 @itemx -mno-smartmips
1446 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1447 equivalent to putting @code{.module smartmips} at the start of the assembly
1448 file. @samp{-mno-smartmips} turns off this option.
1452 Generate code for the MIPS-3D Application Specific Extension.
1453 This tells the assembler to accept MIPS-3D instructions.
1454 @samp{-no-mips3d} turns off this option.
1458 Generate code for the MDMX Application Specific Extension.
1459 This tells the assembler to accept MDMX instructions.
1460 @samp{-no-mdmx} turns off this option.
1464 Generate code for the DSP Release 1 Application Specific Extension.
1465 This tells the assembler to accept DSP Release 1 instructions.
1466 @samp{-mno-dsp} turns off this option.
1470 Generate code for the DSP Release 2 Application Specific Extension.
1471 This option implies @samp{-mdsp}.
1472 This tells the assembler to accept DSP Release 2 instructions.
1473 @samp{-mno-dspr2} turns off this option.
1477 Generate code for the DSP Release 3 Application Specific Extension.
1478 This option implies @samp{-mdsp} and @samp{-mdspr2}.
1479 This tells the assembler to accept DSP Release 3 instructions.
1480 @samp{-mno-dspr3} turns off this option.
1484 Generate code for the MIPS SIMD Architecture Extension.
1485 This tells the assembler to accept MSA instructions.
1486 @samp{-mno-msa} turns off this option.
1490 Generate code for the MIPS eXtended Physical Address (XPA) Extension.
1491 This tells the assembler to accept XPA instructions.
1492 @samp{-mno-xpa} turns off this option.
1496 Generate code for the MT Application Specific Extension.
1497 This tells the assembler to accept MT instructions.
1498 @samp{-mno-mt} turns off this option.
1502 Generate code for the MCU Application Specific Extension.
1503 This tells the assembler to accept MCU instructions.
1504 @samp{-mno-mcu} turns off this option.
1508 Only use 32-bit instruction encodings when generating code for the
1509 microMIPS processor. This option inhibits the use of any 16-bit
1510 instructions. This is equivalent to putting @code{.set insn32} at
1511 the start of the assembly file. @samp{-mno-insn32} turns off this
1512 option. This is equivalent to putting @code{.set noinsn32} at the
1513 start of the assembly file. By default @samp{-mno-insn32} is
1514 selected, allowing all instructions to be used.
1516 @item --construct-floats
1517 @itemx --no-construct-floats
1518 The @samp{--no-construct-floats} option disables the construction of
1519 double width floating point constants by loading the two halves of the
1520 value into the two single width floating point registers that make up
1521 the double width register. By default @samp{--construct-floats} is
1522 selected, allowing construction of these floating point constants.
1524 @item --relax-branch
1525 @itemx --no-relax-branch
1526 The @samp{--relax-branch} option enables the relaxation of out-of-range
1527 branches. By default @samp{--no-relax-branch} is selected, causing any
1528 out-of-range branches to produce an error.
1530 @item -mignore-branch-isa
1531 @itemx -mno-ignore-branch-isa
1532 Ignore branch checks for invalid transitions between ISA modes. The
1533 semantics of branches does not provide for an ISA mode switch, so in
1534 most cases the ISA mode a branch has been encoded for has to be the
1535 same as the ISA mode of the branch's target label. Therefore GAS has
1536 checks implemented that verify in branch assembly that the two ISA
1537 modes match. @samp{-mignore-branch-isa} disables these checks. By
1538 default @samp{-mno-ignore-branch-isa} is selected, causing any invalid
1539 branch requiring a transition between ISA modes to produce an error.
1541 @item -mnan=@var{encoding}
1542 Select between the IEEE 754-2008 (@option{-mnan=2008}) or the legacy
1543 (@option{-mnan=legacy}) NaN encoding format. The latter is the default.
1546 @item --emulation=@var{name}
1547 This option was formerly used to switch between ELF and ECOFF output
1548 on targets like IRIX 5 that supported both. MIPS ECOFF support was
1549 removed in GAS 2.24, so the option now serves little purpose.
1550 It is retained for backwards compatibility.
1552 The available configuration names are: @samp{mipself}, @samp{mipslelf} and
1553 @samp{mipsbelf}. Choosing @samp{mipself} now has no effect, since the output
1554 is always ELF. @samp{mipslelf} and @samp{mipsbelf} select little- and
1555 big-endian output respectively, but @samp{-EL} and @samp{-EB} are now the
1556 preferred options instead.
1559 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1566 Control how to deal with multiplication overflow and division by zero.
1567 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1568 (and only work for Instruction Set Architecture level 2 and higher);
1569 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1573 When this option is used, @command{@value{AS}} will issue a warning every
1574 time it generates a nop instruction from a macro.
1580 The following options are available when @value{AS} is configured for
1586 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1587 The command line option @samp{-nojsri2bsr} can be used to disable it.
1591 Enable or disable the silicon filter behaviour. By default this is disabled.
1592 The default can be overridden by the @samp{-sifilter} command line option.
1595 Alter jump instructions for long displacements.
1597 @item -mcpu=[210|340]
1598 Select the cpu type on the target hardware. This controls which instructions
1602 Assemble for a big endian target.
1605 Assemble for a little endian target.
1614 @xref{Meta Options}, for the options available when @value{AS} is configured
1615 for a Meta processor.
1619 @c man begin OPTIONS
1620 The following options are available when @value{AS} is configured for a
1623 @c man begin INCLUDE
1624 @include c-metag.texi
1625 @c ended inside the included file
1630 @c man begin OPTIONS
1632 See the info pages for documentation of the MMIX-specific options.
1638 @xref{NDS32 Options}, for the options available when @value{AS} is configured
1639 for a NDS32 processor.
1641 @c ended inside the included file
1645 @c man begin OPTIONS
1646 The following options are available when @value{AS} is configured for a
1649 @c man begin INCLUDE
1650 @include c-nds32.texi
1651 @c ended inside the included file
1658 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1659 for a PowerPC processor.
1663 @c man begin OPTIONS
1664 The following options are available when @value{AS} is configured for a
1667 @c man begin INCLUDE
1669 @c ended inside the included file
1677 @xref{RISC-V-Options}, for the options available when @value{AS} is configured
1678 for a RISC-V processor.
1682 @c man begin OPTIONS
1683 The following options are available when @value{AS} is configured for a
1686 @c man begin INCLUDE
1687 @include c-riscv.texi
1688 @c ended inside the included file
1693 @c man begin OPTIONS
1695 See the info pages for documentation of the RX-specific options.
1699 The following options are available when @value{AS} is configured for the s390
1705 Select the word size, either 31/32 bits or 64 bits.
1708 Select the architecture mode, either the Enterprise System
1709 Architecture (esa) or the z/Architecture mode (zarch).
1710 @item -march=@var{processor}
1711 Specify which s390 processor variant is the target, @samp{g5} (or
1712 @samp{arch3}), @samp{g6}, @samp{z900} (or @samp{arch5}), @samp{z990} (or
1713 @samp{arch6}), @samp{z9-109}, @samp{z9-ec} (or @samp{arch7}), @samp{z10} (or
1714 @samp{arch8}), @samp{z196} (or @samp{arch9}), @samp{zEC12} (or @samp{arch10}),
1715 @samp{z13} (or @samp{arch11}), or @samp{z14} (or @samp{arch12}).
1717 @itemx -mno-regnames
1718 Allow or disallow symbolic names for registers.
1719 @item -mwarn-areg-zero
1720 Warn whenever the operand for a base or index register has been specified
1721 but evaluates to zero.
1729 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1730 for a TMS320C6000 processor.
1734 @c man begin OPTIONS
1735 The following options are available when @value{AS} is configured for a
1736 TMS320C6000 processor.
1738 @c man begin INCLUDE
1739 @include c-tic6x.texi
1740 @c ended inside the included file
1748 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1749 for a TILE-Gx processor.
1753 @c man begin OPTIONS
1754 The following options are available when @value{AS} is configured for a TILE-Gx
1757 @c man begin INCLUDE
1758 @include c-tilegx.texi
1759 @c ended inside the included file
1767 @xref{Visium Options}, for the options available when @value{AS} is configured
1768 for a Visium processor.
1772 @c man begin OPTIONS
1773 The following option is available when @value{AS} is configured for a Visium
1776 @c man begin INCLUDE
1777 @include c-visium.texi
1778 @c ended inside the included file
1786 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1787 for an Xtensa processor.
1791 @c man begin OPTIONS
1792 The following options are available when @value{AS} is configured for an
1795 @c man begin INCLUDE
1796 @include c-xtensa.texi
1797 @c ended inside the included file
1802 @c man begin OPTIONS
1805 The following options are available when @value{AS} is configured for
1806 a Z80 family processor.
1809 Assemble for Z80 processor.
1811 Assemble for R800 processor.
1812 @item -ignore-undocumented-instructions
1814 Assemble undocumented Z80 instructions that also work on R800 without warning.
1815 @item -ignore-unportable-instructions
1817 Assemble all undocumented Z80 instructions without warning.
1818 @item -warn-undocumented-instructions
1820 Issue a warning for undocumented Z80 instructions that also work on R800.
1821 @item -warn-unportable-instructions
1823 Issue a warning for undocumented Z80 instructions that do not work on R800.
1824 @item -forbid-undocumented-instructions
1826 Treat all undocumented instructions as errors.
1827 @item -forbid-unportable-instructions
1829 Treat undocumented Z80 instructions that do not work on R800 as errors.
1836 * Manual:: Structure of this Manual
1837 * GNU Assembler:: The GNU Assembler
1838 * Object Formats:: Object File Formats
1839 * Command Line:: Command Line
1840 * Input Files:: Input Files
1841 * Object:: Output (Object) File
1842 * Errors:: Error and Warning Messages
1846 @section Structure of this Manual
1848 @cindex manual, structure and purpose
1849 This manual is intended to describe what you need to know to use
1850 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1851 notation for symbols, constants, and expressions; the directives that
1852 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1855 We also cover special features in the @value{TARGET}
1856 configuration of @command{@value{AS}}, including assembler directives.
1859 This manual also describes some of the machine-dependent features of
1860 various flavors of the assembler.
1863 @cindex machine instructions (not covered)
1864 On the other hand, this manual is @emph{not} intended as an introduction
1865 to programming in assembly language---let alone programming in general!
1866 In a similar vein, we make no attempt to introduce the machine
1867 architecture; we do @emph{not} describe the instruction set, standard
1868 mnemonics, registers or addressing modes that are standard to a
1869 particular architecture.
1871 You may want to consult the manufacturer's
1872 machine architecture manual for this information.
1876 For information on the H8/300 machine instruction set, see @cite{H8/300
1877 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1878 Programming Manual} (Renesas).
1881 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1882 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1883 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1884 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1887 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1891 @c I think this is premature---doc@cygnus.com, 17jan1991
1893 Throughout this manual, we assume that you are running @dfn{GNU},
1894 the portable operating system from the @dfn{Free Software
1895 Foundation, Inc.}. This restricts our attention to certain kinds of
1896 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1897 once this assumption is granted examples and definitions need less
1900 @command{@value{AS}} is part of a team of programs that turn a high-level
1901 human-readable series of instructions into a low-level
1902 computer-readable series of instructions. Different versions of
1903 @command{@value{AS}} are used for different kinds of computer.
1906 @c There used to be a section "Terminology" here, which defined
1907 @c "contents", "byte", "word", and "long". Defining "word" to any
1908 @c particular size is confusing when the .word directive may generate 16
1909 @c bits on one machine and 32 bits on another; in general, for the user
1910 @c version of this manual, none of these terms seem essential to define.
1911 @c They were used very little even in the former draft of the manual;
1912 @c this draft makes an effort to avoid them (except in names of
1916 @section The GNU Assembler
1918 @c man begin DESCRIPTION
1920 @sc{gnu} @command{as} is really a family of assemblers.
1922 This manual describes @command{@value{AS}}, a member of that family which is
1923 configured for the @value{TARGET} architectures.
1925 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1926 should find a fairly similar environment when you use it on another
1927 architecture. Each version has much in common with the others,
1928 including object file formats, most assembler directives (often called
1929 @dfn{pseudo-ops}) and assembler syntax.@refill
1931 @cindex purpose of @sc{gnu} assembler
1932 @command{@value{AS}} is primarily intended to assemble the output of the
1933 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1934 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1935 assemble correctly everything that other assemblers for the same
1936 machine would assemble.
1938 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1941 @c This remark should appear in generic version of manual; assumption
1942 @c here is that generic version sets M680x0.
1943 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1944 assembler for the same architecture; for example, we know of several
1945 incompatible versions of 680x0 assembly language syntax.
1950 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1951 program in one pass of the source file. This has a subtle impact on the
1952 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1954 @node Object Formats
1955 @section Object File Formats
1957 @cindex object file format
1958 The @sc{gnu} assembler can be configured to produce several alternative
1959 object file formats. For the most part, this does not affect how you
1960 write assembly language programs; but directives for debugging symbols
1961 are typically different in different file formats. @xref{Symbol
1962 Attributes,,Symbol Attributes}.
1965 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1966 @value{OBJ-NAME} format object files.
1968 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1970 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1971 SOM or ELF format object files.
1976 @section Command Line
1978 @cindex command line conventions
1980 After the program name @command{@value{AS}}, the command line may contain
1981 options and file names. Options may appear in any order, and may be
1982 before, after, or between file names. The order of file names is
1985 @cindex standard input, as input file
1987 @file{--} (two hyphens) by itself names the standard input file
1988 explicitly, as one of the files for @command{@value{AS}} to assemble.
1990 @cindex options, command line
1991 Except for @samp{--} any command line argument that begins with a
1992 hyphen (@samp{-}) is an option. Each option changes the behavior of
1993 @command{@value{AS}}. No option changes the way another option works. An
1994 option is a @samp{-} followed by one or more letters; the case of
1995 the letter is important. All options are optional.
1997 Some options expect exactly one file name to follow them. The file
1998 name may either immediately follow the option's letter (compatible
1999 with older assemblers) or it may be the next command argument (@sc{gnu}
2000 standard). These two command lines are equivalent:
2003 @value{AS} -o my-object-file.o mumble.s
2004 @value{AS} -omy-object-file.o mumble.s
2008 @section Input Files
2011 @cindex source program
2012 @cindex files, input
2013 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
2014 describe the program input to one run of @command{@value{AS}}. The program may
2015 be in one or more files; how the source is partitioned into files
2016 doesn't change the meaning of the source.
2018 @c I added "con" prefix to "catenation" just to prove I can overcome my
2019 @c APL training... doc@cygnus.com
2020 The source program is a concatenation of the text in all the files, in the
2023 @c man begin DESCRIPTION
2024 Each time you run @command{@value{AS}} it assembles exactly one source
2025 program. The source program is made up of one or more files.
2026 (The standard input is also a file.)
2028 You give @command{@value{AS}} a command line that has zero or more input file
2029 names. The input files are read (from left file name to right). A
2030 command line argument (in any position) that has no special meaning
2031 is taken to be an input file name.
2033 If you give @command{@value{AS}} no file names it attempts to read one input file
2034 from the @command{@value{AS}} standard input, which is normally your terminal. You
2035 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
2038 Use @samp{--} if you need to explicitly name the standard input file
2039 in your command line.
2041 If the source is empty, @command{@value{AS}} produces a small, empty object
2046 @subheading Filenames and Line-numbers
2048 @cindex input file linenumbers
2049 @cindex line numbers, in input files
2050 There are two ways of locating a line in the input file (or files) and
2051 either may be used in reporting error messages. One way refers to a line
2052 number in a physical file; the other refers to a line number in a
2053 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
2055 @dfn{Physical files} are those files named in the command line given
2056 to @command{@value{AS}}.
2058 @dfn{Logical files} are simply names declared explicitly by assembler
2059 directives; they bear no relation to physical files. Logical file names help
2060 error messages reflect the original source file, when @command{@value{AS}} source
2061 is itself synthesized from other files. @command{@value{AS}} understands the
2062 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
2063 @ref{File,,@code{.file}}.
2066 @section Output (Object) File
2072 Every time you run @command{@value{AS}} it produces an output file, which is
2073 your assembly language program translated into numbers. This file
2074 is the object file. Its default name is @code{a.out}.
2075 You can give it another name by using the @option{-o} option. Conventionally,
2076 object file names end with @file{.o}. The default name is used for historical
2077 reasons: older assemblers were capable of assembling self-contained programs
2078 directly into a runnable program. (For some formats, this isn't currently
2079 possible, but it can be done for the @code{a.out} format.)
2083 The object file is meant for input to the linker @code{@value{LD}}. It contains
2084 assembled program code, information to help @code{@value{LD}} integrate
2085 the assembled program into a runnable file, and (optionally) symbolic
2086 information for the debugger.
2088 @c link above to some info file(s) like the description of a.out.
2089 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
2092 @section Error and Warning Messages
2094 @c man begin DESCRIPTION
2096 @cindex error messages
2097 @cindex warning messages
2098 @cindex messages from assembler
2099 @command{@value{AS}} may write warnings and error messages to the standard error
2100 file (usually your terminal). This should not happen when a compiler
2101 runs @command{@value{AS}} automatically. Warnings report an assumption made so
2102 that @command{@value{AS}} could keep assembling a flawed program; errors report a
2103 grave problem that stops the assembly.
2107 @cindex format of warning messages
2108 Warning messages have the format
2111 file_name:@b{NNN}:Warning Message Text
2115 @cindex file names and line numbers, in warnings/errors
2116 (where @b{NNN} is a line number). If both a logical file name
2117 (@pxref{File,,@code{.file}}) and a logical line number
2119 (@pxref{Line,,@code{.line}})
2121 have been given then they will be used, otherwise the file name and line number
2122 in the current assembler source file will be used. The message text is
2123 intended to be self explanatory (in the grand Unix tradition).
2125 Note the file name must be set via the logical version of the @code{.file}
2126 directive, not the DWARF2 version of the @code{.file} directive. For example:
2130 error_assembler_source
2136 produces this output:
2140 asm.s:2: Error: no such instruction: `error_assembler_source'
2141 foo.c:31: Error: no such instruction: `error_c_source'
2144 @cindex format of error messages
2145 Error messages have the format
2148 file_name:@b{NNN}:FATAL:Error Message Text
2151 The file name and line number are derived as for warning
2152 messages. The actual message text may be rather less explanatory
2153 because many of them aren't supposed to happen.
2156 @chapter Command-Line Options
2158 @cindex options, all versions of assembler
2159 This chapter describes command-line options available in @emph{all}
2160 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
2161 for options specific
2163 to the @value{TARGET} target.
2166 to particular machine architectures.
2169 @c man begin DESCRIPTION
2171 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
2172 you can use the @samp{-Wa} option to pass arguments through to the assembler.
2173 The assembler arguments must be separated from each other (and the @samp{-Wa})
2174 by commas. For example:
2177 gcc -c -g -O -Wa,-alh,-L file.c
2181 This passes two options to the assembler: @samp{-alh} (emit a listing to
2182 standard output with high-level and assembly source) and @samp{-L} (retain
2183 local symbols in the symbol table).
2185 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
2186 command-line options are automatically passed to the assembler by the compiler.
2187 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
2188 precisely what options it passes to each compilation pass, including the
2194 * a:: -a[cdghlns] enable listings
2195 * alternate:: --alternate enable alternate macro syntax
2196 * D:: -D for compatibility
2197 * f:: -f to work faster
2198 * I:: -I for .include search path
2199 @ifclear DIFF-TBL-KLUGE
2200 * K:: -K for compatibility
2202 @ifset DIFF-TBL-KLUGE
2203 * K:: -K for difference tables
2206 * L:: -L to retain local symbols
2207 * listing:: --listing-XXX to configure listing output
2208 * M:: -M or --mri to assemble in MRI compatibility mode
2209 * MD:: --MD for dependency tracking
2210 * no-pad-sections:: --no-pad-sections to stop section padding
2211 * o:: -o to name the object file
2212 * R:: -R to join data and text sections
2213 * statistics:: --statistics to see statistics about assembly
2214 * traditional-format:: --traditional-format for compatible output
2215 * v:: -v to announce version
2216 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
2217 * Z:: -Z to make object file even after errors
2221 @section Enable Listings: @option{-a[cdghlns]}
2231 @cindex listings, enabling
2232 @cindex assembly listings, enabling
2234 These options enable listing output from the assembler. By itself,
2235 @samp{-a} requests high-level, assembly, and symbols listing.
2236 You can use other letters to select specific options for the list:
2237 @samp{-ah} requests a high-level language listing,
2238 @samp{-al} requests an output-program assembly listing, and
2239 @samp{-as} requests a symbol table listing.
2240 High-level listings require that a compiler debugging option like
2241 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
2244 Use the @samp{-ag} option to print a first section with general assembly
2245 information, like @value{AS} version, switches passed, or time stamp.
2247 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
2248 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
2249 other conditional), or a true @code{.if} followed by an @code{.else}, will be
2250 omitted from the listing.
2252 Use the @samp{-ad} option to omit debugging directives from the
2255 Once you have specified one of these options, you can further control
2256 listing output and its appearance using the directives @code{.list},
2257 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
2259 The @samp{-an} option turns off all forms processing.
2260 If you do not request listing output with one of the @samp{-a} options, the
2261 listing-control directives have no effect.
2263 The letters after @samp{-a} may be combined into one option,
2264 @emph{e.g.}, @samp{-aln}.
2266 Note if the assembler source is coming from the standard input (e.g.,
2268 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
2269 is being used) then the listing will not contain any comments or preprocessor
2270 directives. This is because the listing code buffers input source lines from
2271 stdin only after they have been preprocessed by the assembler. This reduces
2272 memory usage and makes the code more efficient.
2275 @section @option{--alternate}
2278 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
2281 @section @option{-D}
2284 This option has no effect whatsoever, but it is accepted to make it more
2285 likely that scripts written for other assemblers also work with
2286 @command{@value{AS}}.
2289 @section Work Faster: @option{-f}
2292 @cindex trusted compiler
2293 @cindex faster processing (@option{-f})
2294 @samp{-f} should only be used when assembling programs written by a
2295 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
2296 and comment preprocessing on
2297 the input file(s) before assembling them. @xref{Preprocessing,
2301 @emph{Warning:} if you use @samp{-f} when the files actually need to be
2302 preprocessed (if they contain comments, for example), @command{@value{AS}} does
2307 @section @code{.include} Search Path: @option{-I} @var{path}
2309 @kindex -I @var{path}
2310 @cindex paths for @code{.include}
2311 @cindex search path for @code{.include}
2312 @cindex @code{include} directive search path
2313 Use this option to add a @var{path} to the list of directories
2314 @command{@value{AS}} searches for files specified in @code{.include}
2315 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
2316 many times as necessary to include a variety of paths. The current
2317 working directory is always searched first; after that, @command{@value{AS}}
2318 searches any @samp{-I} directories in the same order as they were
2319 specified (left to right) on the command line.
2322 @section Difference Tables: @option{-K}
2325 @ifclear DIFF-TBL-KLUGE
2326 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2327 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2328 where it can be used to warn when the assembler alters the machine code
2329 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2330 family does not have the addressing limitations that sometimes lead to this
2331 alteration on other platforms.
2334 @ifset DIFF-TBL-KLUGE
2335 @cindex difference tables, warning
2336 @cindex warning for altered difference tables
2337 @command{@value{AS}} sometimes alters the code emitted for directives of the
2338 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2339 You can use the @samp{-K} option if you want a warning issued when this
2344 @section Include Local Symbols: @option{-L}
2347 @cindex local symbols, retaining in output
2348 Symbols beginning with system-specific local label prefixes, typically
2349 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2350 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2351 such symbols when debugging, because they are intended for the use of
2352 programs (like compilers) that compose assembler programs, not for your
2353 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2354 such symbols, so you do not normally debug with them.
2356 This option tells @command{@value{AS}} to retain those local symbols
2357 in the object file. Usually if you do this you also tell the linker
2358 @code{@value{LD}} to preserve those symbols.
2361 @section Configuring listing output: @option{--listing}
2363 The listing feature of the assembler can be enabled via the command line switch
2364 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2365 hex dump of the corresponding locations in the output object file, and displays
2366 them as a listing file. The format of this listing can be controlled by
2367 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2368 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2369 @code{.psize} (@pxref{Psize}), and
2370 @code{.eject} (@pxref{Eject}) and also by the following switches:
2373 @item --listing-lhs-width=@samp{number}
2374 @kindex --listing-lhs-width
2375 @cindex Width of first line disassembly output
2376 Sets the maximum width, in words, of the first line of the hex byte dump. This
2377 dump appears on the left hand side of the listing output.
2379 @item --listing-lhs-width2=@samp{number}
2380 @kindex --listing-lhs-width2
2381 @cindex Width of continuation lines of disassembly output
2382 Sets the maximum width, in words, of any further lines of the hex byte dump for
2383 a given input source line. If this value is not specified, it defaults to being
2384 the same as the value specified for @samp{--listing-lhs-width}. If neither
2385 switch is used the default is to one.
2387 @item --listing-rhs-width=@samp{number}
2388 @kindex --listing-rhs-width
2389 @cindex Width of source line output
2390 Sets the maximum width, in characters, of the source line that is displayed
2391 alongside the hex dump. The default value for this parameter is 100. The
2392 source line is displayed on the right hand side of the listing output.
2394 @item --listing-cont-lines=@samp{number}
2395 @kindex --listing-cont-lines
2396 @cindex Maximum number of continuation lines
2397 Sets the maximum number of continuation lines of hex dump that will be
2398 displayed for a given single line of source input. The default value is 4.
2402 @section Assemble in MRI Compatibility Mode: @option{-M}
2405 @cindex MRI compatibility mode
2406 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2407 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2408 compatible with the @code{ASM68K} assembler from Microtec Research.
2409 The exact nature of the
2410 MRI syntax will not be documented here; see the MRI manuals for more
2411 information. Note in particular that the handling of macros and macro
2412 arguments is somewhat different. The purpose of this option is to permit
2413 assembling existing MRI assembler code using @command{@value{AS}}.
2415 The MRI compatibility is not complete. Certain operations of the MRI assembler
2416 depend upon its object file format, and can not be supported using other object
2417 file formats. Supporting these would require enhancing each object file format
2418 individually. These are:
2421 @item global symbols in common section
2423 The m68k MRI assembler supports common sections which are merged by the linker.
2424 Other object file formats do not support this. @command{@value{AS}} handles
2425 common sections by treating them as a single common symbol. It permits local
2426 symbols to be defined within a common section, but it can not support global
2427 symbols, since it has no way to describe them.
2429 @item complex relocations
2431 The MRI assemblers support relocations against a negated section address, and
2432 relocations which combine the start addresses of two or more sections. These
2433 are not support by other object file formats.
2435 @item @code{END} pseudo-op specifying start address
2437 The MRI @code{END} pseudo-op permits the specification of a start address.
2438 This is not supported by other object file formats. The start address may
2439 instead be specified using the @option{-e} option to the linker, or in a linker
2442 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2444 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2445 name to the output file. This is not supported by other object file formats.
2447 @item @code{ORG} pseudo-op
2449 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2450 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2451 which changes the location within the current section. Absolute sections are
2452 not supported by other object file formats. The address of a section may be
2453 assigned within a linker script.
2456 There are some other features of the MRI assembler which are not supported by
2457 @command{@value{AS}}, typically either because they are difficult or because they
2458 seem of little consequence. Some of these may be supported in future releases.
2462 @item EBCDIC strings
2464 EBCDIC strings are not supported.
2466 @item packed binary coded decimal
2468 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2469 and @code{DCB.P} pseudo-ops are not supported.
2471 @item @code{FEQU} pseudo-op
2473 The m68k @code{FEQU} pseudo-op is not supported.
2475 @item @code{NOOBJ} pseudo-op
2477 The m68k @code{NOOBJ} pseudo-op is not supported.
2479 @item @code{OPT} branch control options
2481 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2482 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2483 relaxes all branches, whether forward or backward, to an appropriate size, so
2484 these options serve no purpose.
2486 @item @code{OPT} list control options
2488 The following m68k @code{OPT} list control options are ignored: @code{C},
2489 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2490 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2492 @item other @code{OPT} options
2494 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2495 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2497 @item @code{OPT} @code{D} option is default
2499 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2500 @code{OPT NOD} may be used to turn it off.
2502 @item @code{XREF} pseudo-op.
2504 The m68k @code{XREF} pseudo-op is ignored.
2509 @section Dependency Tracking: @option{--MD}
2512 @cindex dependency tracking
2515 @command{@value{AS}} can generate a dependency file for the file it creates. This
2516 file consists of a single rule suitable for @code{make} describing the
2517 dependencies of the main source file.
2519 The rule is written to the file named in its argument.
2521 This feature is used in the automatic updating of makefiles.
2523 @node no-pad-sections
2524 @section Output Section Padding
2525 @kindex --no-pad-sections
2526 @cindex output section padding
2527 Normally the assembler will pad the end of each output section up to its
2528 alignment boundary. But this can waste space, which can be significant on
2529 memory constrained targets. So the @option{--no-pad-sections} option will
2530 disable this behaviour.
2533 @section Name the Object File: @option{-o}
2536 @cindex naming object file
2537 @cindex object file name
2538 There is always one object file output when you run @command{@value{AS}}. By
2539 default it has the name @file{a.out}.
2540 You use this option (which takes exactly one filename) to give the
2541 object file a different name.
2543 Whatever the object file is called, @command{@value{AS}} overwrites any
2544 existing file of the same name.
2547 @section Join Data and Text Sections: @option{-R}
2550 @cindex data and text sections, joining
2551 @cindex text and data sections, joining
2552 @cindex joining text and data sections
2553 @cindex merging text and data sections
2554 @option{-R} tells @command{@value{AS}} to write the object file as if all
2555 data-section data lives in the text section. This is only done at
2556 the very last moment: your binary data are the same, but data
2557 section parts are relocated differently. The data section part of
2558 your object file is zero bytes long because all its bytes are
2559 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2561 When you specify @option{-R} it would be possible to generate shorter
2562 address displacements (because we do not have to cross between text and
2563 data section). We refrain from doing this simply for compatibility with
2564 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2567 When @command{@value{AS}} is configured for COFF or ELF output,
2568 this option is only useful if you use sections named @samp{.text} and
2573 @option{-R} is not supported for any of the HPPA targets. Using
2574 @option{-R} generates a warning from @command{@value{AS}}.
2578 @section Display Assembly Statistics: @option{--statistics}
2580 @kindex --statistics
2581 @cindex statistics, about assembly
2582 @cindex time, total for assembly
2583 @cindex space used, maximum for assembly
2584 Use @samp{--statistics} to display two statistics about the resources used by
2585 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2586 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2589 @node traditional-format
2590 @section Compatible Output: @option{--traditional-format}
2592 @kindex --traditional-format
2593 For some targets, the output of @command{@value{AS}} is different in some ways
2594 from the output of some existing assembler. This switch requests
2595 @command{@value{AS}} to use the traditional format instead.
2597 For example, it disables the exception frame optimizations which
2598 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2601 @section Announce Version: @option{-v}
2605 @cindex assembler version
2606 @cindex version of assembler
2607 You can find out what version of as is running by including the
2608 option @samp{-v} (which you can also spell as @samp{-version}) on the
2612 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2614 @command{@value{AS}} should never give a warning or error message when
2615 assembling compiler output. But programs written by people often
2616 cause @command{@value{AS}} to give a warning that a particular assumption was
2617 made. All such warnings are directed to the standard error file.
2621 @cindex suppressing warnings
2622 @cindex warnings, suppressing
2623 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2624 This only affects the warning messages: it does not change any particular of
2625 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2628 @kindex --fatal-warnings
2629 @cindex errors, caused by warnings
2630 @cindex warnings, causing error
2631 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2632 files that generate warnings to be in error.
2635 @cindex warnings, switching on
2636 You can switch these options off again by specifying @option{--warn}, which
2637 causes warnings to be output as usual.
2640 @section Generate Object File in Spite of Errors: @option{-Z}
2641 @cindex object file, after errors
2642 @cindex errors, continuing after
2643 After an error message, @command{@value{AS}} normally produces no output. If for
2644 some reason you are interested in object file output even after
2645 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2646 option. If there are any errors, @command{@value{AS}} continues anyways, and
2647 writes an object file after a final warning message of the form @samp{@var{n}
2648 errors, @var{m} warnings, generating bad object file.}
2653 @cindex machine-independent syntax
2654 @cindex syntax, machine-independent
2655 This chapter describes the machine-independent syntax allowed in a
2656 source file. @command{@value{AS}} syntax is similar to what many other
2657 assemblers use; it is inspired by the BSD 4.2
2662 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2666 * Preprocessing:: Preprocessing
2667 * Whitespace:: Whitespace
2668 * Comments:: Comments
2669 * Symbol Intro:: Symbols
2670 * Statements:: Statements
2671 * Constants:: Constants
2675 @section Preprocessing
2677 @cindex preprocessing
2678 The @command{@value{AS}} internal preprocessor:
2680 @cindex whitespace, removed by preprocessor
2682 adjusts and removes extra whitespace. It leaves one space or tab before
2683 the keywords on a line, and turns any other whitespace on the line into
2686 @cindex comments, removed by preprocessor
2688 removes all comments, replacing them with a single space, or an
2689 appropriate number of newlines.
2691 @cindex constants, converted by preprocessor
2693 converts character constants into the appropriate numeric values.
2696 It does not do macro processing, include file handling, or
2697 anything else you may get from your C compiler's preprocessor. You can
2698 do include file processing with the @code{.include} directive
2699 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2700 to get other ``CPP'' style preprocessing by giving the input file a
2701 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2702 Output, gcc info, Using GNU CC}.
2704 Excess whitespace, comments, and character constants
2705 cannot be used in the portions of the input text that are not
2708 @cindex turning preprocessing on and off
2709 @cindex preprocessing, turning on and off
2712 If the first line of an input file is @code{#NO_APP} or if you use the
2713 @samp{-f} option, whitespace and comments are not removed from the input file.
2714 Within an input file, you can ask for whitespace and comment removal in
2715 specific portions of the by putting a line that says @code{#APP} before the
2716 text that may contain whitespace or comments, and putting a line that says
2717 @code{#NO_APP} after this text. This feature is mainly intend to support
2718 @code{asm} statements in compilers whose output is otherwise free of comments
2725 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2726 Whitespace is used to separate symbols, and to make programs neater for
2727 people to read. Unless within character constants
2728 (@pxref{Characters,,Character Constants}), any whitespace means the same
2729 as exactly one space.
2735 There are two ways of rendering comments to @command{@value{AS}}. In both
2736 cases the comment is equivalent to one space.
2738 Anything from @samp{/*} through the next @samp{*/} is a comment.
2739 This means you may not nest these comments.
2743 The only way to include a newline ('\n') in a comment
2744 is to use this sort of comment.
2747 /* This sort of comment does not nest. */
2750 @cindex line comment character
2751 Anything from a @dfn{line comment} character up to the next newline is
2752 considered a comment and is ignored. The line comment character is target
2753 specific, and some targets multiple comment characters. Some targets also have
2754 line comment characters that only work if they are the first character on a
2755 line. Some targets use a sequence of two characters to introduce a line
2756 comment. Some targets can also change their line comment characters depending
2757 upon command line options that have been used. For more details see the
2758 @emph{Syntax} section in the documentation for individual targets.
2760 If the line comment character is the hash sign (@samp{#}) then it still has the
2761 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2762 to specify logical line numbers:
2765 @cindex lines starting with @code{#}
2766 @cindex logical line numbers
2767 To be compatible with past assemblers, lines that begin with @samp{#} have a
2768 special interpretation. Following the @samp{#} should be an absolute
2769 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2770 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2771 new logical file name. The rest of the line, if any, should be whitespace.
2773 If the first non-whitespace characters on the line are not numeric,
2774 the line is ignored. (Just like a comment.)
2777 # This is an ordinary comment.
2778 # 42-6 "new_file_name" # New logical file name
2779 # This is logical line # 36.
2781 This feature is deprecated, and may disappear from future versions
2782 of @command{@value{AS}}.
2787 @cindex characters used in symbols
2788 @ifclear SPECIAL-SYMS
2789 A @dfn{symbol} is one or more characters chosen from the set of all
2790 letters (both upper and lower case), digits and the three characters
2796 A @dfn{symbol} is one or more characters chosen from the set of all
2797 letters (both upper and lower case), digits and the three characters
2798 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2804 On most machines, you can also use @code{$} in symbol names; exceptions
2805 are noted in @ref{Machine Dependencies}.
2807 No symbol may begin with a digit. Case is significant.
2808 There is no length limit; all characters are significant. Multibyte characters
2809 are supported. Symbols are delimited by characters not in that set, or by the
2810 beginning of a file (since the source program must end with a newline, the end
2811 of a file is not a possible symbol delimiter). @xref{Symbols}.
2813 Symbol names may also be enclosed in double quote @code{"} characters. In such
2814 cases any characters are allowed, except for the NUL character. If a double
2815 quote character is to be included in the symbol name it must be preceeded by a
2816 backslash @code{\} character.
2817 @cindex length of symbols
2822 @cindex statements, structure of
2823 @cindex line separator character
2824 @cindex statement separator character
2826 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2827 @dfn{line separator character}. The line separator character is target
2828 specific and described in the @emph{Syntax} section of each
2829 target's documentation. Not all targets support a line separator character.
2830 The newline or line separator character is considered to be part of the
2831 preceding statement. Newlines and separators within character constants are an
2832 exception: they do not end statements.
2834 @cindex newline, required at file end
2835 @cindex EOF, newline must precede
2836 It is an error to end any statement with end-of-file: the last
2837 character of any input file should be a newline.@refill
2839 An empty statement is allowed, and may include whitespace. It is ignored.
2841 @cindex instructions and directives
2842 @cindex directives and instructions
2843 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2844 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2846 A statement begins with zero or more labels, optionally followed by a
2847 key symbol which determines what kind of statement it is. The key
2848 symbol determines the syntax of the rest of the statement. If the
2849 symbol begins with a dot @samp{.} then the statement is an assembler
2850 directive: typically valid for any computer. If the symbol begins with
2851 a letter the statement is an assembly language @dfn{instruction}: it
2852 assembles into a machine language instruction.
2854 Different versions of @command{@value{AS}} for different computers
2855 recognize different instructions. In fact, the same symbol may
2856 represent a different instruction in a different computer's assembly
2860 @cindex @code{:} (label)
2861 @cindex label (@code{:})
2862 A label is a symbol immediately followed by a colon (@code{:}).
2863 Whitespace before a label or after a colon is permitted, but you may not
2864 have whitespace between a label's symbol and its colon. @xref{Labels}.
2867 For HPPA targets, labels need not be immediately followed by a colon, but
2868 the definition of a label must begin in column zero. This also implies that
2869 only one label may be defined on each line.
2873 label: .directive followed by something
2874 another_label: # This is an empty statement.
2875 instruction operand_1, operand_2, @dots{}
2882 A constant is a number, written so that its value is known by
2883 inspection, without knowing any context. Like this:
2886 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2887 .ascii "Ring the bell\7" # A string constant.
2888 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2889 .float 0f-314159265358979323846264338327\
2890 95028841971.693993751E-40 # - pi, a flonum.
2895 * Characters:: Character Constants
2896 * Numbers:: Number Constants
2900 @subsection Character Constants
2902 @cindex character constants
2903 @cindex constants, character
2904 There are two kinds of character constants. A @dfn{character} stands
2905 for one character in one byte and its value may be used in
2906 numeric expressions. String constants (properly called string
2907 @emph{literals}) are potentially many bytes and their values may not be
2908 used in arithmetic expressions.
2912 * Chars:: Characters
2916 @subsubsection Strings
2918 @cindex string constants
2919 @cindex constants, string
2920 A @dfn{string} is written between double-quotes. It may contain
2921 double-quotes or null characters. The way to get special characters
2922 into a string is to @dfn{escape} these characters: precede them with
2923 a backslash @samp{\} character. For example @samp{\\} represents
2924 one backslash: the first @code{\} is an escape which tells
2925 @command{@value{AS}} to interpret the second character literally as a backslash
2926 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2927 escape character). The complete list of escapes follows.
2929 @cindex escape codes, character
2930 @cindex character escape codes
2931 @c NOTE: Cindex entries must not start with a backlash character.
2932 @c NOTE: This confuses the pdf2texi script when it is creating the
2933 @c NOTE: index based upon the first character and so it generates:
2934 @c NOTE: \initial {\\}
2935 @c NOTE: which then results in the error message:
2936 @c NOTE: Argument of \\ has an extra }.
2937 @c NOTE: So in the index entries below a space character has been
2938 @c NOTE: prepended to avoid this problem.
2941 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2943 @cindex @code{ \b} (backspace character)
2944 @cindex backspace (@code{\b})
2946 Mnemonic for backspace; for ASCII this is octal code 010.
2949 @c Mnemonic for EOText; for ASCII this is octal code 004.
2951 @cindex @code{ \f} (formfeed character)
2952 @cindex formfeed (@code{\f})
2954 Mnemonic for FormFeed; for ASCII this is octal code 014.
2956 @cindex @code{ \n} (newline character)
2957 @cindex newline (@code{\n})
2959 Mnemonic for newline; for ASCII this is octal code 012.
2962 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2964 @cindex @code{ \r} (carriage return character)
2965 @cindex carriage return (@code{backslash-r})
2967 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2970 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2971 @c other assemblers.
2973 @cindex @code{ \t} (tab)
2974 @cindex tab (@code{\t})
2976 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2979 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2980 @c @item \x @var{digit} @var{digit} @var{digit}
2981 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2983 @cindex @code{ \@var{ddd}} (octal character code)
2984 @cindex octal character code (@code{\@var{ddd}})
2985 @item \ @var{digit} @var{digit} @var{digit}
2986 An octal character code. The numeric code is 3 octal digits.
2987 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2988 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2990 @cindex @code{ \@var{xd...}} (hex character code)
2991 @cindex hex character code (@code{\@var{xd...}})
2992 @item \@code{x} @var{hex-digits...}
2993 A hex character code. All trailing hex digits are combined. Either upper or
2994 lower case @code{x} works.
2996 @cindex @code{ \\} (@samp{\} character)
2997 @cindex backslash (@code{\\})
2999 Represents one @samp{\} character.
3002 @c Represents one @samp{'} (accent acute) character.
3003 @c This is needed in single character literals
3004 @c (@xref{Characters,,Character Constants}.) to represent
3007 @cindex @code{ \"} (doublequote character)
3008 @cindex doublequote (@code{\"})
3010 Represents one @samp{"} character. Needed in strings to represent
3011 this character, because an unescaped @samp{"} would end the string.
3013 @item \ @var{anything-else}
3014 Any other character when escaped by @kbd{\} gives a warning, but
3015 assembles as if the @samp{\} was not present. The idea is that if
3016 you used an escape sequence you clearly didn't want the literal
3017 interpretation of the following character. However @command{@value{AS}} has no
3018 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
3019 code and warns you of the fact.
3022 Which characters are escapable, and what those escapes represent,
3023 varies widely among assemblers. The current set is what we think
3024 the BSD 4.2 assembler recognizes, and is a subset of what most C
3025 compilers recognize. If you are in doubt, do not use an escape
3029 @subsubsection Characters
3031 @cindex single character constant
3032 @cindex character, single
3033 @cindex constant, single character
3034 A single character may be written as a single quote immediately followed by
3035 that character. Some backslash escapes apply to characters, @code{\b},
3036 @code{\f}, @code{\n}, @code{\r}, @code{\t}, and @code{\"} with the same meaning
3037 as for strings, plus @code{\'} for a single quote. So if you want to write the
3038 character backslash, you must write @kbd{'\\} where the first @code{\} escapes
3039 the second @code{\}. As you can see, the quote is an acute accent, not a grave
3042 @ifclear abnormal-separator
3043 (or semicolon @samp{;})
3045 @ifset abnormal-separator
3047 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
3052 immediately following an acute accent is taken as a literal character
3053 and does not count as the end of a statement. The value of a character
3054 constant in a numeric expression is the machine's byte-wide code for
3055 that character. @command{@value{AS}} assumes your character code is ASCII:
3056 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
3059 @subsection Number Constants
3061 @cindex constants, number
3062 @cindex number constants
3063 @command{@value{AS}} distinguishes three kinds of numbers according to how they
3064 are stored in the target machine. @emph{Integers} are numbers that
3065 would fit into an @code{int} in the C language. @emph{Bignums} are
3066 integers, but they are stored in more than 32 bits. @emph{Flonums}
3067 are floating point numbers, described below.
3070 * Integers:: Integers
3078 @subsubsection Integers
3080 @cindex constants, integer
3082 @cindex binary integers
3083 @cindex integers, binary
3084 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
3085 the binary digits @samp{01}.
3087 @cindex octal integers
3088 @cindex integers, octal
3089 An octal integer is @samp{0} followed by zero or more of the octal
3090 digits (@samp{01234567}).
3092 @cindex decimal integers
3093 @cindex integers, decimal
3094 A decimal integer starts with a non-zero digit followed by zero or
3095 more digits (@samp{0123456789}).
3097 @cindex hexadecimal integers
3098 @cindex integers, hexadecimal
3099 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
3100 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
3102 Integers have the usual values. To denote a negative integer, use
3103 the prefix operator @samp{-} discussed under expressions
3104 (@pxref{Prefix Ops,,Prefix Operators}).
3107 @subsubsection Bignums
3110 @cindex constants, bignum
3111 A @dfn{bignum} has the same syntax and semantics as an integer
3112 except that the number (or its negative) takes more than 32 bits to
3113 represent in binary. The distinction is made because in some places
3114 integers are permitted while bignums are not.
3117 @subsubsection Flonums
3119 @cindex floating point numbers
3120 @cindex constants, floating point
3122 @cindex precision, floating point
3123 A @dfn{flonum} represents a floating point number. The translation is
3124 indirect: a decimal floating point number from the text is converted by
3125 @command{@value{AS}} to a generic binary floating point number of more than
3126 sufficient precision. This generic floating point number is converted
3127 to a particular computer's floating point format (or formats) by a
3128 portion of @command{@value{AS}} specialized to that computer.
3130 A flonum is written by writing (in order)
3135 (@samp{0} is optional on the HPPA.)
3139 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
3141 @kbd{e} is recommended. Case is not important.
3143 @c FIXME: verify if flonum syntax really this vague for most cases
3144 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
3145 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
3148 On the H8/300 and Renesas / SuperH SH architectures, the letter must be
3149 one of the letters @samp{DFPRSX} (in upper or lower case).
3151 On the ARC, the letter must be one of the letters @samp{DFRS}
3152 (in upper or lower case).
3154 On the HPPA architecture, the letter must be @samp{E} (upper case only).
3158 One of the letters @samp{DFRS} (in upper or lower case).
3161 One of the letters @samp{DFPRSX} (in upper or lower case).
3164 The letter @samp{E} (upper case only).
3169 An optional sign: either @samp{+} or @samp{-}.
3172 An optional @dfn{integer part}: zero or more decimal digits.
3175 An optional @dfn{fractional part}: @samp{.} followed by zero
3176 or more decimal digits.
3179 An optional exponent, consisting of:
3183 An @samp{E} or @samp{e}.
3184 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
3185 @c principle this can perfectly well be different on different targets.
3187 Optional sign: either @samp{+} or @samp{-}.
3189 One or more decimal digits.
3194 At least one of the integer part or the fractional part must be
3195 present. The floating point number has the usual base-10 value.
3197 @command{@value{AS}} does all processing using integers. Flonums are computed
3198 independently of any floating point hardware in the computer running
3199 @command{@value{AS}}.
3202 @chapter Sections and Relocation
3207 * Secs Background:: Background
3208 * Ld Sections:: Linker Sections
3209 * As Sections:: Assembler Internal Sections
3210 * Sub-Sections:: Sub-Sections
3214 @node Secs Background
3217 Roughly, a section is a range of addresses, with no gaps; all data
3218 ``in'' those addresses is treated the same for some particular purpose.
3219 For example there may be a ``read only'' section.
3221 @cindex linker, and assembler
3222 @cindex assembler, and linker
3223 The linker @code{@value{LD}} reads many object files (partial programs) and
3224 combines their contents to form a runnable program. When @command{@value{AS}}
3225 emits an object file, the partial program is assumed to start at address 0.
3226 @code{@value{LD}} assigns the final addresses for the partial program, so that
3227 different partial programs do not overlap. This is actually an
3228 oversimplification, but it suffices to explain how @command{@value{AS}} uses
3231 @code{@value{LD}} moves blocks of bytes of your program to their run-time
3232 addresses. These blocks slide to their run-time addresses as rigid
3233 units; their length does not change and neither does the order of bytes
3234 within them. Such a rigid unit is called a @emph{section}. Assigning
3235 run-time addresses to sections is called @dfn{relocation}. It includes
3236 the task of adjusting mentions of object-file addresses so they refer to
3237 the proper run-time addresses.
3239 For the H8/300, and for the Renesas / SuperH SH,
3240 @command{@value{AS}} pads sections if needed to
3241 ensure they end on a word (sixteen bit) boundary.
3244 @cindex standard assembler sections
3245 An object file written by @command{@value{AS}} has at least three sections, any
3246 of which may be empty. These are named @dfn{text}, @dfn{data} and
3251 When it generates COFF or ELF output,
3253 @command{@value{AS}} can also generate whatever other named sections you specify
3254 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
3255 If you do not use any directives that place output in the @samp{.text}
3256 or @samp{.data} sections, these sections still exist, but are empty.
3261 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
3263 @command{@value{AS}} can also generate whatever other named sections you
3264 specify using the @samp{.space} and @samp{.subspace} directives. See
3265 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3266 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3267 assembler directives.
3270 Additionally, @command{@value{AS}} uses different names for the standard
3271 text, data, and bss sections when generating SOM output. Program text
3272 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3273 BSS into @samp{$BSS$}.
3277 Within the object file, the text section starts at address @code{0}, the
3278 data section follows, and the bss section follows the data section.
3281 When generating either SOM or ELF output files on the HPPA, the text
3282 section starts at address @code{0}, the data section at address
3283 @code{0x4000000}, and the bss section follows the data section.
3286 To let @code{@value{LD}} know which data changes when the sections are
3287 relocated, and how to change that data, @command{@value{AS}} also writes to the
3288 object file details of the relocation needed. To perform relocation
3289 @code{@value{LD}} must know, each time an address in the object
3293 Where in the object file is the beginning of this reference to
3296 How long (in bytes) is this reference?
3298 Which section does the address refer to? What is the numeric value of
3300 (@var{address}) @minus{} (@var{start-address of section})?
3303 Is the reference to an address ``Program-Counter relative''?
3306 @cindex addresses, format of
3307 @cindex section-relative addressing
3308 In fact, every address @command{@value{AS}} ever uses is expressed as
3310 (@var{section}) + (@var{offset into section})
3313 Further, most expressions @command{@value{AS}} computes have this section-relative
3316 (For some object formats, such as SOM for the HPPA, some expressions are
3317 symbol-relative instead.)
3320 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3321 @var{N} into section @var{secname}.''
3323 Apart from text, data and bss sections you need to know about the
3324 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3325 addresses in the absolute section remain unchanged. For example, address
3326 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3327 @code{@value{LD}}. Although the linker never arranges two partial programs'
3328 data sections with overlapping addresses after linking, @emph{by definition}
3329 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3330 part of a program is always the same address when the program is running as
3331 address @code{@{absolute@ 239@}} in any other part of the program.
3333 The idea of sections is extended to the @dfn{undefined} section. Any
3334 address whose section is unknown at assembly time is by definition
3335 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3336 Since numbers are always defined, the only way to generate an undefined
3337 address is to mention an undefined symbol. A reference to a named
3338 common block would be such a symbol: its value is unknown at assembly
3339 time so it has section @emph{undefined}.
3341 By analogy the word @emph{section} is used to describe groups of sections in
3342 the linked program. @code{@value{LD}} puts all partial programs' text
3343 sections in contiguous addresses in the linked program. It is
3344 customary to refer to the @emph{text section} of a program, meaning all
3345 the addresses of all partial programs' text sections. Likewise for
3346 data and bss sections.
3348 Some sections are manipulated by @code{@value{LD}}; others are invented for
3349 use of @command{@value{AS}} and have no meaning except during assembly.
3352 @section Linker Sections
3353 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3358 @cindex named sections
3359 @cindex sections, named
3360 @item named sections
3363 @cindex text section
3364 @cindex data section
3368 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3369 separate but equal sections. Anything you can say of one section is
3372 When the program is running, however, it is
3373 customary for the text section to be unalterable. The
3374 text section is often shared among processes: it contains
3375 instructions, constants and the like. The data section of a running
3376 program is usually alterable: for example, C variables would be stored
3377 in the data section.
3382 This section contains zeroed bytes when your program begins running. It
3383 is used to hold uninitialized variables or common storage. The length of
3384 each partial program's bss section is important, but because it starts
3385 out containing zeroed bytes there is no need to store explicit zero
3386 bytes in the object file. The bss section was invented to eliminate
3387 those explicit zeros from object files.
3389 @cindex absolute section
3390 @item absolute section
3391 Address 0 of this section is always ``relocated'' to runtime address 0.
3392 This is useful if you want to refer to an address that @code{@value{LD}} must
3393 not change when relocating. In this sense we speak of absolute
3394 addresses being ``unrelocatable'': they do not change during relocation.
3396 @cindex undefined section
3397 @item undefined section
3398 This ``section'' is a catch-all for address references to objects not in
3399 the preceding sections.
3400 @c FIXME: ref to some other doc on obj-file formats could go here.
3403 @cindex relocation example
3404 An idealized example of three relocatable sections follows.
3406 The example uses the traditional section names @samp{.text} and @samp{.data}.
3408 Memory addresses are on the horizontal axis.
3412 @c END TEXI2ROFF-KILL
3415 partial program # 1: |ttttt|dddd|00|
3422 partial program # 2: |TTT|DDD|000|
3425 +--+---+-----+--+----+---+-----+~~
3426 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3427 +--+---+-----+--+----+---+-----+~~
3429 addresses: 0 @dots{}
3436 \line{\it Partial program \#1: \hfil}
3437 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3438 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3440 \line{\it Partial program \#2: \hfil}
3441 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3442 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3444 \line{\it linked program: \hfil}
3445 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3446 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3447 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3448 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3450 \line{\it addresses: \hfil}
3454 @c END TEXI2ROFF-KILL
3457 @section Assembler Internal Sections
3459 @cindex internal assembler sections
3460 @cindex sections in messages, internal
3461 These sections are meant only for the internal use of @command{@value{AS}}. They
3462 have no meaning at run-time. You do not really need to know about these
3463 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3464 warning messages, so it might be helpful to have an idea of their
3465 meanings to @command{@value{AS}}. These sections are used to permit the
3466 value of every expression in your assembly language program to be a
3467 section-relative address.
3470 @cindex assembler internal logic error
3471 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3472 An internal assembler logic error has been found. This means there is a
3473 bug in the assembler.
3475 @cindex expr (internal section)
3477 The assembler stores complex expression internally as combinations of
3478 symbols. When it needs to represent an expression as a symbol, it puts
3479 it in the expr section.
3481 @c FIXME item transfer[t] vector preload
3482 @c FIXME item transfer[t] vector postload
3483 @c FIXME item register
3487 @section Sub-Sections
3489 @cindex numbered subsections
3490 @cindex grouping data
3496 fall into two sections: text and data.
3498 You may have separate groups of
3500 data in named sections
3504 data in named sections
3510 that you want to end up near to each other in the object file, even though they
3511 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3512 use @dfn{subsections} for this purpose. Within each section, there can be
3513 numbered subsections with values from 0 to 8192. Objects assembled into the
3514 same subsection go into the object file together with other objects in the same
3515 subsection. For example, a compiler might want to store constants in the text
3516 section, but might not want to have them interspersed with the program being
3517 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3518 section of code being output, and a @samp{.text 1} before each group of
3519 constants being output.
3521 Subsections are optional. If you do not use subsections, everything
3522 goes in subsection number zero.
3525 Each subsection is zero-padded up to a multiple of four bytes.
3526 (Subsections may be padded a different amount on different flavors
3527 of @command{@value{AS}}.)
3531 On the H8/300 platform, each subsection is zero-padded to a word
3532 boundary (two bytes).
3533 The same is true on the Renesas SH.
3537 Subsections appear in your object file in numeric order, lowest numbered
3538 to highest. (All this to be compatible with other people's assemblers.)
3539 The object file contains no representation of subsections; @code{@value{LD}} and
3540 other programs that manipulate object files see no trace of them.
3541 They just see all your text subsections as a text section, and all your
3542 data subsections as a data section.
3544 To specify which subsection you want subsequent statements assembled
3545 into, use a numeric argument to specify it, in a @samp{.text
3546 @var{expression}} or a @samp{.data @var{expression}} statement.
3549 When generating COFF output, you
3554 can also use an extra subsection
3555 argument with arbitrary named sections: @samp{.section @var{name},
3560 When generating ELF output, you
3565 can also use the @code{.subsection} directive (@pxref{SubSection})
3566 to specify a subsection: @samp{.subsection @var{expression}}.
3568 @var{Expression} should be an absolute expression
3569 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3570 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3571 begins in @code{text 0}. For instance:
3573 .text 0 # The default subsection is text 0 anyway.
3574 .ascii "This lives in the first text subsection. *"
3576 .ascii "But this lives in the second text subsection."
3578 .ascii "This lives in the data section,"
3579 .ascii "in the first data subsection."
3581 .ascii "This lives in the first text section,"
3582 .ascii "immediately following the asterisk (*)."
3585 Each section has a @dfn{location counter} incremented by one for every byte
3586 assembled into that section. Because subsections are merely a convenience
3587 restricted to @command{@value{AS}} there is no concept of a subsection location
3588 counter. There is no way to directly manipulate a location counter---but the
3589 @code{.align} directive changes it, and any label definition captures its
3590 current value. The location counter of the section where statements are being
3591 assembled is said to be the @dfn{active} location counter.
3594 @section bss Section
3597 @cindex common variable storage
3598 The bss section is used for local common variable storage.
3599 You may allocate address space in the bss section, but you may
3600 not dictate data to load into it before your program executes. When
3601 your program starts running, all the contents of the bss
3602 section are zeroed bytes.
3604 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3605 @ref{Lcomm,,@code{.lcomm}}.
3607 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3608 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3611 When assembling for a target which supports multiple sections, such as ELF or
3612 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3613 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3614 section. Typically the section will only contain symbol definitions and
3615 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3622 Symbols are a central concept: the programmer uses symbols to name
3623 things, the linker uses symbols to link, and the debugger uses symbols
3627 @cindex debuggers, and symbol order
3628 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3629 the same order they were declared. This may break some debuggers.
3634 * Setting Symbols:: Giving Symbols Other Values
3635 * Symbol Names:: Symbol Names
3636 * Dot:: The Special Dot Symbol
3637 * Symbol Attributes:: Symbol Attributes
3644 A @dfn{label} is written as a symbol immediately followed by a colon
3645 @samp{:}. The symbol then represents the current value of the
3646 active location counter, and is, for example, a suitable instruction
3647 operand. You are warned if you use the same symbol to represent two
3648 different locations: the first definition overrides any other
3652 On the HPPA, the usual form for a label need not be immediately followed by a
3653 colon, but instead must start in column zero. Only one label may be defined on
3654 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3655 provides a special directive @code{.label} for defining labels more flexibly.
3658 @node Setting Symbols
3659 @section Giving Symbols Other Values
3661 @cindex assigning values to symbols
3662 @cindex symbol values, assigning
3663 A symbol can be given an arbitrary value by writing a symbol, followed
3664 by an equals sign @samp{=}, followed by an expression
3665 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3666 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3667 equals sign @samp{=}@samp{=} here represents an equivalent of the
3668 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3671 Blackfin does not support symbol assignment with @samp{=}.
3675 @section Symbol Names
3677 @cindex symbol names
3678 @cindex names, symbol
3679 @ifclear SPECIAL-SYMS
3680 Symbol names begin with a letter or with one of @samp{._}. On most
3681 machines, you can also use @code{$} in symbol names; exceptions are
3682 noted in @ref{Machine Dependencies}. That character may be followed by any
3683 string of digits, letters, dollar signs (unless otherwise noted for a
3684 particular target machine), and underscores.
3688 Symbol names begin with a letter or with one of @samp{._}. On the
3689 Renesas SH you can also use @code{$} in symbol names. That
3690 character may be followed by any string of digits, letters, dollar signs (save
3691 on the H8/300), and underscores.
3695 Case of letters is significant: @code{foo} is a different symbol name
3698 Symbol names do not start with a digit. An exception to this rule is made for
3699 Local Labels. See below.
3701 Multibyte characters are supported. To generate a symbol name containing
3702 multibyte characters enclose it within double quotes and use escape codes. cf
3703 @xref{Strings}. Generating a multibyte symbol name from a label is not
3704 currently supported.
3706 Each symbol has exactly one name. Each name in an assembly language program
3707 refers to exactly one symbol. You may use that symbol name any number of times
3710 @subheading Local Symbol Names
3712 @cindex local symbol names
3713 @cindex symbol names, local
3714 A local symbol is any symbol beginning with certain local label prefixes.
3715 By default, the local label prefix is @samp{.L} for ELF systems or
3716 @samp{L} for traditional a.out systems, but each target may have its own
3717 set of local label prefixes.
3719 On the HPPA local symbols begin with @samp{L$}.
3722 Local symbols are defined and used within the assembler, but they are
3723 normally not saved in object files. Thus, they are not visible when debugging.
3724 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols})
3725 to retain the local symbols in the object files.
3727 @subheading Local Labels
3729 @cindex local labels
3730 @cindex temporary symbol names
3731 @cindex symbol names, temporary
3732 Local labels are different from local symbols. Local labels help compilers and
3733 programmers use names temporarily. They create symbols which are guaranteed to
3734 be unique over the entire scope of the input source code and which can be
3735 referred to by a simple notation. To define a local label, write a label of
3736 the form @samp{@b{N}:} (where @b{N} represents any non-negative integer).
3737 To refer to the most recent previous definition of that label write
3738 @samp{@b{N}b}, using the same number as when you defined the label. To refer
3739 to the next definition of a local label, write @samp{@b{N}f}. The @samp{b}
3740 stands for ``backwards'' and the @samp{f} stands for ``forwards''.
3742 There is no restriction on how you can use these labels, and you can reuse them
3743 too. So that it is possible to repeatedly define the same local label (using
3744 the same number @samp{@b{N}}), although you can only refer to the most recently
3745 defined local label of that number (for a backwards reference) or the next
3746 definition of a specific local label for a forward reference. It is also worth
3747 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3748 implemented in a slightly more efficient manner than the others.
3759 Which is the equivalent of:
3762 label_1: branch label_3
3763 label_2: branch label_1
3764 label_3: branch label_4
3765 label_4: branch label_3
3768 Local label names are only a notational device. They are immediately
3769 transformed into more conventional symbol names before the assembler uses them.
3770 The symbol names are stored in the symbol table, appear in error messages, and
3771 are optionally emitted to the object file. The names are constructed using
3775 @item @emph{local label prefix}
3776 All local symbols begin with the system-specific local label prefix.
3777 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3778 that start with the local label prefix. These labels are
3779 used for symbols you are never intended to see. If you use the
3780 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3781 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3782 you may use them in debugging.
3785 This is the number that was used in the local label definition. So if the
3786 label is written @samp{55:} then the number is @samp{55}.
3789 This unusual character is included so you do not accidentally invent a symbol
3790 of the same name. The character has ASCII value of @samp{\002} (control-B).
3792 @item @emph{ordinal number}
3793 This is a serial number to keep the labels distinct. The first definition of
3794 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3795 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3796 the number @samp{1} and its 15th definition gets @samp{15} as well.
3799 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3800 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3802 @subheading Dollar Local Labels
3803 @cindex dollar local symbols
3805 On some targets @code{@value{AS}} also supports an even more local form of
3806 local labels called dollar labels. These labels go out of scope (i.e., they
3807 become undefined) as soon as a non-local label is defined. Thus they remain
3808 valid for only a small region of the input source code. Normal local labels,
3809 by contrast, remain in scope for the entire file, or until they are redefined
3810 by another occurrence of the same local label.
3812 Dollar labels are defined in exactly the same way as ordinary local labels,
3813 except that they have a dollar sign suffix to their numeric value, e.g.,
3816 They can also be distinguished from ordinary local labels by their transformed
3817 names which use ASCII character @samp{\001} (control-A) as the magic character
3818 to distinguish them from ordinary labels. For example, the fifth definition of
3819 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3822 @section The Special Dot Symbol
3824 @cindex dot (symbol)
3825 @cindex @code{.} (symbol)
3826 @cindex current address
3827 @cindex location counter
3828 The special symbol @samp{.} refers to the current address that
3829 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3830 .long .} defines @code{melvin} to contain its own address.
3831 Assigning a value to @code{.} is treated the same as a @code{.org}
3833 @ifclear no-space-dir
3834 Thus, the expression @samp{.=.+4} is the same as saying
3838 @node Symbol Attributes
3839 @section Symbol Attributes
3841 @cindex symbol attributes
3842 @cindex attributes, symbol
3843 Every symbol has, as well as its name, the attributes ``Value'' and
3844 ``Type''. Depending on output format, symbols can also have auxiliary
3847 The detailed definitions are in @file{a.out.h}.
3850 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3851 all these attributes, and probably won't warn you. This makes the
3852 symbol an externally defined symbol, which is generally what you
3856 * Symbol Value:: Value
3857 * Symbol Type:: Type
3859 * a.out Symbols:: Symbol Attributes: @code{a.out}
3862 * COFF Symbols:: Symbol Attributes for COFF
3865 * SOM Symbols:: Symbol Attributes for SOM
3872 @cindex value of a symbol
3873 @cindex symbol value
3874 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3875 location in the text, data, bss or absolute sections the value is the
3876 number of addresses from the start of that section to the label.
3877 Naturally for text, data and bss sections the value of a symbol changes
3878 as @code{@value{LD}} changes section base addresses during linking. Absolute
3879 symbols' values do not change during linking: that is why they are
3882 The value of an undefined symbol is treated in a special way. If it is
3883 0 then the symbol is not defined in this assembler source file, and
3884 @code{@value{LD}} tries to determine its value from other files linked into the
3885 same program. You make this kind of symbol simply by mentioning a symbol
3886 name without defining it. A non-zero value represents a @code{.comm}
3887 common declaration. The value is how much common storage to reserve, in
3888 bytes (addresses). The symbol refers to the first address of the
3894 @cindex type of a symbol
3896 The type attribute of a symbol contains relocation (section)
3897 information, any flag settings indicating that a symbol is external, and
3898 (optionally), other information for linkers and debuggers. The exact
3899 format depends on the object-code output format in use.
3903 @subsection Symbol Attributes: @code{a.out}
3905 @cindex @code{a.out} symbol attributes
3906 @cindex symbol attributes, @code{a.out}
3909 * Symbol Desc:: Descriptor
3910 * Symbol Other:: Other
3914 @subsubsection Descriptor
3916 @cindex descriptor, of @code{a.out} symbol
3917 This is an arbitrary 16-bit value. You may establish a symbol's
3918 descriptor value by using a @code{.desc} statement
3919 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3920 @command{@value{AS}}.
3923 @subsubsection Other
3925 @cindex other attribute, of @code{a.out} symbol
3926 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3931 @subsection Symbol Attributes for COFF
3933 @cindex COFF symbol attributes
3934 @cindex symbol attributes, COFF
3936 The COFF format supports a multitude of auxiliary symbol attributes;
3937 like the primary symbol attributes, they are set between @code{.def} and
3938 @code{.endef} directives.
3940 @subsubsection Primary Attributes
3942 @cindex primary attributes, COFF symbols
3943 The symbol name is set with @code{.def}; the value and type,
3944 respectively, with @code{.val} and @code{.type}.
3946 @subsubsection Auxiliary Attributes
3948 @cindex auxiliary attributes, COFF symbols
3949 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3950 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3951 table information for COFF.
3956 @subsection Symbol Attributes for SOM
3958 @cindex SOM symbol attributes
3959 @cindex symbol attributes, SOM
3961 The SOM format for the HPPA supports a multitude of symbol attributes set with
3962 the @code{.EXPORT} and @code{.IMPORT} directives.
3964 The attributes are described in @cite{HP9000 Series 800 Assembly
3965 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3966 @code{EXPORT} assembler directive documentation.
3970 @chapter Expressions
3974 @cindex numeric values
3975 An @dfn{expression} specifies an address or numeric value.
3976 Whitespace may precede and/or follow an expression.
3978 The result of an expression must be an absolute number, or else an offset into
3979 a particular section. If an expression is not absolute, and there is not
3980 enough information when @command{@value{AS}} sees the expression to know its
3981 section, a second pass over the source program might be necessary to interpret
3982 the expression---but the second pass is currently not implemented.
3983 @command{@value{AS}} aborts with an error message in this situation.
3986 * Empty Exprs:: Empty Expressions
3987 * Integer Exprs:: Integer Expressions
3991 @section Empty Expressions
3993 @cindex empty expressions
3994 @cindex expressions, empty
3995 An empty expression has no value: it is just whitespace or null.
3996 Wherever an absolute expression is required, you may omit the
3997 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3998 is compatible with other assemblers.
4001 @section Integer Expressions
4003 @cindex integer expressions
4004 @cindex expressions, integer
4005 An @dfn{integer expression} is one or more @emph{arguments} delimited
4006 by @emph{operators}.
4009 * Arguments:: Arguments
4010 * Operators:: Operators
4011 * Prefix Ops:: Prefix Operators
4012 * Infix Ops:: Infix Operators
4016 @subsection Arguments
4018 @cindex expression arguments
4019 @cindex arguments in expressions
4020 @cindex operands in expressions
4021 @cindex arithmetic operands
4022 @dfn{Arguments} are symbols, numbers or subexpressions. In other
4023 contexts arguments are sometimes called ``arithmetic operands''. In
4024 this manual, to avoid confusing them with the ``instruction operands'' of
4025 the machine language, we use the term ``argument'' to refer to parts of
4026 expressions only, reserving the word ``operand'' to refer only to machine
4027 instruction operands.
4029 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
4030 @var{section} is one of text, data, bss, absolute,
4031 or undefined. @var{NNN} is a signed, 2's complement 32 bit
4034 Numbers are usually integers.
4036 A number can be a flonum or bignum. In this case, you are warned
4037 that only the low order 32 bits are used, and @command{@value{AS}} pretends
4038 these 32 bits are an integer. You may write integer-manipulating
4039 instructions that act on exotic constants, compatible with other
4042 @cindex subexpressions
4043 Subexpressions are a left parenthesis @samp{(} followed by an integer
4044 expression, followed by a right parenthesis @samp{)}; or a prefix
4045 operator followed by an argument.
4048 @subsection Operators
4050 @cindex operators, in expressions
4051 @cindex arithmetic functions
4052 @cindex functions, in expressions
4053 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
4054 operators are followed by an argument. Infix operators appear
4055 between their arguments. Operators may be preceded and/or followed by
4059 @subsection Prefix Operator
4061 @cindex prefix operators
4062 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
4063 one argument, which must be absolute.
4065 @c the tex/end tex stuff surrounding this small table is meant to make
4066 @c it align, on the printed page, with the similar table in the next
4067 @c section (which is inside an enumerate).
4069 \global\advance\leftskip by \itemindent
4074 @dfn{Negation}. Two's complement negation.
4076 @dfn{Complementation}. Bitwise not.
4080 \global\advance\leftskip by -\itemindent
4084 @subsection Infix Operators
4086 @cindex infix operators
4087 @cindex operators, permitted arguments
4088 @dfn{Infix operators} take two arguments, one on either side. Operators
4089 have precedence, but operations with equal precedence are performed left
4090 to right. Apart from @code{+} or @option{-}, both arguments must be
4091 absolute, and the result is absolute.
4094 @cindex operator precedence
4095 @cindex precedence of operators
4102 @dfn{Multiplication}.
4105 @dfn{Division}. Truncation is the same as the C operator @samp{/}
4111 @dfn{Shift Left}. Same as the C operator @samp{<<}.
4114 @dfn{Shift Right}. Same as the C operator @samp{>>}.
4118 Intermediate precedence
4123 @dfn{Bitwise Inclusive Or}.
4129 @dfn{Bitwise Exclusive Or}.
4132 @dfn{Bitwise Or Not}.
4139 @cindex addition, permitted arguments
4140 @cindex plus, permitted arguments
4141 @cindex arguments for addition
4143 @dfn{Addition}. If either argument is absolute, the result has the section of
4144 the other argument. You may not add together arguments from different
4147 @cindex subtraction, permitted arguments
4148 @cindex minus, permitted arguments
4149 @cindex arguments for subtraction
4151 @dfn{Subtraction}. If the right argument is absolute, the
4152 result has the section of the left argument.
4153 If both arguments are in the same section, the result is absolute.
4154 You may not subtract arguments from different sections.
4155 @c FIXME is there still something useful to say about undefined - undefined ?
4157 @cindex comparison expressions
4158 @cindex expressions, comparison
4163 @dfn{Is Not Equal To}
4167 @dfn{Is Greater Than}
4169 @dfn{Is Greater Than Or Equal To}
4171 @dfn{Is Less Than Or Equal To}
4173 The comparison operators can be used as infix operators. A true results has a
4174 value of -1 whereas a false result has a value of 0. Note, these operators
4175 perform signed comparisons.
4178 @item Lowest Precedence
4187 These two logical operations can be used to combine the results of sub
4188 expressions. Note, unlike the comparison operators a true result returns a
4189 value of 1 but a false results does still return 0. Also note that the logical
4190 or operator has a slightly lower precedence than logical and.
4195 In short, it's only meaningful to add or subtract the @emph{offsets} in an
4196 address; you can only have a defined section in one of the two arguments.
4199 @chapter Assembler Directives
4201 @cindex directives, machine independent
4202 @cindex pseudo-ops, machine independent
4203 @cindex machine independent directives
4204 All assembler directives have names that begin with a period (@samp{.}).
4205 The names are case insensitive for most targets, and usually written
4208 This chapter discusses directives that are available regardless of the
4209 target machine configuration for the @sc{gnu} assembler.
4211 Some machine configurations provide additional directives.
4212 @xref{Machine Dependencies}.
4215 @ifset machine-directives
4216 @xref{Machine Dependencies}, for additional directives.
4221 * Abort:: @code{.abort}
4223 * ABORT (COFF):: @code{.ABORT}
4226 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
4227 * Altmacro:: @code{.altmacro}
4228 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4229 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4230 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
4231 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, etc
4232 * Byte:: @code{.byte @var{expressions}}
4233 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4234 * Comm:: @code{.comm @var{symbol} , @var{length} }
4235 * Data:: @code{.data @var{subsection}}
4236 * Dc:: @code{.dc[@var{size}] @var{expressions}}
4237 * Dcb:: @code{.dcb[@var{size}] @var{number} [,@var{fill}]}
4238 * Ds:: @code{.ds[@var{size}] @var{number} [,@var{fill}]}
4240 * Def:: @code{.def @var{name}}
4243 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4249 * Double:: @code{.double @var{flonums}}
4250 * Eject:: @code{.eject}
4251 * Else:: @code{.else}
4252 * Elseif:: @code{.elseif}
4255 * Endef:: @code{.endef}
4258 * Endfunc:: @code{.endfunc}
4259 * Endif:: @code{.endif}
4260 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4261 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4262 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4264 * Error:: @code{.error @var{string}}
4265 * Exitm:: @code{.exitm}
4266 * Extern:: @code{.extern}
4267 * Fail:: @code{.fail}
4268 * File:: @code{.file}
4269 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4270 * Float:: @code{.float @var{flonums}}
4271 * Func:: @code{.func}
4272 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4274 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4275 * Hidden:: @code{.hidden @var{names}}
4278 * hword:: @code{.hword @var{expressions}}
4279 * Ident:: @code{.ident}
4280 * If:: @code{.if @var{absolute expression}}
4281 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4282 * Include:: @code{.include "@var{file}"}
4283 * Int:: @code{.int @var{expressions}}
4285 * Internal:: @code{.internal @var{names}}
4288 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4289 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4290 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4291 * Lflags:: @code{.lflags}
4292 @ifclear no-line-dir
4293 * Line:: @code{.line @var{line-number}}
4296 * Linkonce:: @code{.linkonce [@var{type}]}
4297 * List:: @code{.list}
4298 * Ln:: @code{.ln @var{line-number}}
4299 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4300 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4302 * Local:: @code{.local @var{names}}
4305 * Long:: @code{.long @var{expressions}}
4307 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4310 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4311 * MRI:: @code{.mri @var{val}}
4312 * Noaltmacro:: @code{.noaltmacro}
4313 * Nolist:: @code{.nolist}
4314 * Nops:: @code{.nops @var{size}[, @var{control}]}
4315 * Octa:: @code{.octa @var{bignums}}
4316 * Offset:: @code{.offset @var{loc}}
4317 * Org:: @code{.org @var{new-lc}, @var{fill}}
4318 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4320 * PopSection:: @code{.popsection}
4321 * Previous:: @code{.previous}
4324 * Print:: @code{.print @var{string}}
4326 * Protected:: @code{.protected @var{names}}
4329 * Psize:: @code{.psize @var{lines}, @var{columns}}
4330 * Purgem:: @code{.purgem @var{name}}
4332 * PushSection:: @code{.pushsection @var{name}}
4335 * Quad:: @code{.quad @var{bignums}}
4336 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4337 * Rept:: @code{.rept @var{count}}
4338 * Sbttl:: @code{.sbttl "@var{subheading}"}
4340 * Scl:: @code{.scl @var{class}}
4343 * Section:: @code{.section @var{name}[, @var{flags}]}
4346 * Set:: @code{.set @var{symbol}, @var{expression}}
4347 * Short:: @code{.short @var{expressions}}
4348 * Single:: @code{.single @var{flonums}}
4350 * Size:: @code{.size [@var{name} , @var{expression}]}
4352 @ifclear no-space-dir
4353 * Skip:: @code{.skip @var{size} [,@var{fill}]}
4356 * Sleb128:: @code{.sleb128 @var{expressions}}
4357 @ifclear no-space-dir
4358 * Space:: @code{.space @var{size} [,@var{fill}]}
4361 * Stab:: @code{.stabd, .stabn, .stabs}
4364 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4365 * Struct:: @code{.struct @var{expression}}
4367 * SubSection:: @code{.subsection}
4368 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4372 * Tag:: @code{.tag @var{structname}}
4375 * Text:: @code{.text @var{subsection}}
4376 * Title:: @code{.title "@var{heading}"}
4378 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4381 * Uleb128:: @code{.uleb128 @var{expressions}}
4383 * Val:: @code{.val @var{addr}}
4387 * Version:: @code{.version "@var{string}"}
4388 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4389 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4392 * Warning:: @code{.warning @var{string}}
4393 * Weak:: @code{.weak @var{names}}
4394 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4395 * Word:: @code{.word @var{expressions}}
4396 @ifclear no-space-dir
4397 * Zero:: @code{.zero @var{size}}
4400 * 2byte:: @code{.2byte @var{expressions}}
4401 * 4byte:: @code{.4byte @var{expressions}}
4402 * 8byte:: @code{.8byte @var{bignums}}
4404 * Deprecated:: Deprecated Directives
4408 @section @code{.abort}
4410 @cindex @code{abort} directive
4411 @cindex stopping the assembly
4412 This directive stops the assembly immediately. It is for
4413 compatibility with other assemblers. The original idea was that the
4414 assembly language source would be piped into the assembler. If the sender
4415 of the source quit, it could use this directive tells @command{@value{AS}} to
4416 quit also. One day @code{.abort} will not be supported.
4420 @section @code{.ABORT} (COFF)
4422 @cindex @code{ABORT} directive
4423 When producing COFF output, @command{@value{AS}} accepts this directive as a
4424 synonym for @samp{.abort}.
4429 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4431 @cindex padding the location counter
4432 @cindex @code{align} directive
4433 Pad the location counter (in the current subsection) to a particular storage
4434 boundary. The first expression (which must be absolute) is the alignment
4435 required, as described below.
4437 The second expression (also absolute) gives the fill value to be stored in the
4438 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4439 padding bytes are normally zero. However, on most systems, if the section is
4440 marked as containing code and the fill value is omitted, the space is filled
4441 with no-op instructions.
4443 The third expression is also absolute, and is also optional. If it is present,
4444 it is the maximum number of bytes that should be skipped by this alignment
4445 directive. If doing the alignment would require skipping more bytes than the
4446 specified maximum, then the alignment is not done at all. You can omit the
4447 fill value (the second argument) entirely by simply using two commas after the
4448 required alignment; this can be useful if you want the alignment to be filled
4449 with no-op instructions when appropriate.
4451 The way the required alignment is specified varies from system to system.
4452 For the arc, hppa, i386 using ELF, iq2000, m68k, or1k,
4453 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4454 alignment request in bytes. For example @samp{.align 8} advances
4455 the location counter until it is a multiple of 8. If the location counter
4456 is already a multiple of 8, no change is needed. For the tic54x, the
4457 first expression is the alignment request in words.
4459 For other systems, including ppc, i386 using a.out format, arm and
4460 strongarm, it is the
4461 number of low-order zero bits the location counter must have after
4462 advancement. For example @samp{.align 3} advances the location
4463 counter until it a multiple of 8. If the location counter is already a
4464 multiple of 8, no change is needed.
4466 This inconsistency is due to the different behaviors of the various
4467 native assemblers for these systems which GAS must emulate.
4468 GAS also provides @code{.balign} and @code{.p2align} directives,
4469 described later, which have a consistent behavior across all
4470 architectures (but are specific to GAS).
4473 @section @code{.altmacro}
4474 Enable alternate macro mode, enabling:
4477 @item LOCAL @var{name} [ , @dots{} ]
4478 One additional directive, @code{LOCAL}, is available. It is used to
4479 generate a string replacement for each of the @var{name} arguments, and
4480 replace any instances of @var{name} in each macro expansion. The
4481 replacement string is unique in the assembly, and different for each
4482 separate macro expansion. @code{LOCAL} allows you to write macros that
4483 define symbols, without fear of conflict between separate macro expansions.
4485 @item String delimiters
4486 You can write strings delimited in these other ways besides
4487 @code{"@var{string}"}:
4490 @item '@var{string}'
4491 You can delimit strings with single-quote characters.
4493 @item <@var{string}>
4494 You can delimit strings with matching angle brackets.
4497 @item single-character string escape
4498 To include any single character literally in a string (even if the
4499 character would otherwise have some special meaning), you can prefix the
4500 character with @samp{!} (an exclamation mark). For example, you can
4501 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4503 @item Expression results as strings
4504 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4505 and use the result as a string.
4509 @section @code{.ascii "@var{string}"}@dots{}
4511 @cindex @code{ascii} directive
4512 @cindex string literals
4513 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4514 separated by commas. It assembles each string (with no automatic
4515 trailing zero byte) into consecutive addresses.
4518 @section @code{.asciz "@var{string}"}@dots{}
4520 @cindex @code{asciz} directive
4521 @cindex zero-terminated strings
4522 @cindex null-terminated strings
4523 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4524 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4527 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4529 @cindex padding the location counter given number of bytes
4530 @cindex @code{balign} directive
4531 Pad the location counter (in the current subsection) to a particular
4532 storage boundary. The first expression (which must be absolute) is the
4533 alignment request in bytes. For example @samp{.balign 8} advances
4534 the location counter until it is a multiple of 8. If the location counter
4535 is already a multiple of 8, no change is needed.
4537 The second expression (also absolute) gives the fill value to be stored in the
4538 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4539 padding bytes are normally zero. However, on most systems, if the section is
4540 marked as containing code and the fill value is omitted, the space is filled
4541 with no-op instructions.
4543 The third expression is also absolute, and is also optional. If it is present,
4544 it is the maximum number of bytes that should be skipped by this alignment
4545 directive. If doing the alignment would require skipping more bytes than the
4546 specified maximum, then the alignment is not done at all. You can omit the
4547 fill value (the second argument) entirely by simply using two commas after the
4548 required alignment; this can be useful if you want the alignment to be filled
4549 with no-op instructions when appropriate.
4551 @cindex @code{balignw} directive
4552 @cindex @code{balignl} directive
4553 The @code{.balignw} and @code{.balignl} directives are variants of the
4554 @code{.balign} directive. The @code{.balignw} directive treats the fill
4555 pattern as a two byte word value. The @code{.balignl} directives treats the
4556 fill pattern as a four byte longword value. For example, @code{.balignw
4557 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4558 filled in with the value 0x368d (the exact placement of the bytes depends upon
4559 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4562 @node Bundle directives
4563 @section Bundle directives
4564 @subsection @code{.bundle_align_mode @var{abs-expr}}
4565 @cindex @code{bundle_align_mode} directive
4567 @cindex instruction bundle
4568 @cindex aligned instruction bundle
4569 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4570 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4571 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4572 disabled (which is the default state). If the argument it not zero, it
4573 gives the size of an instruction bundle as a power of two (as for the
4574 @code{.p2align} directive, @pxref{P2align}).
4576 For some targets, it's an ABI requirement that no instruction may span a
4577 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4578 instructions that starts on an aligned boundary. For example, if
4579 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4580 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4581 effect, no single instruction may span a boundary between bundles. If an
4582 instruction would start too close to the end of a bundle for the length of
4583 that particular instruction to fit within the bundle, then the space at the
4584 end of that bundle is filled with no-op instructions so the instruction
4585 starts in the next bundle. As a corollary, it's an error if any single
4586 instruction's encoding is longer than the bundle size.
4588 @subsection @code{.bundle_lock} and @code{.bundle_unlock}
4589 @cindex @code{bundle_lock} directive
4590 @cindex @code{bundle_unlock} directive
4591 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4592 allow explicit control over instruction bundle padding. These directives
4593 are only valid when @code{.bundle_align_mode} has been used to enable
4594 aligned instruction bundle mode. It's an error if they appear when
4595 @code{.bundle_align_mode} has not been used at all, or when the last
4596 directive was @w{@code{.bundle_align_mode 0}}.
4598 @cindex bundle-locked
4599 For some targets, it's an ABI requirement that certain instructions may
4600 appear only as part of specified permissible sequences of multiple
4601 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4602 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4603 instruction sequence. For purposes of aligned instruction bundle mode, a
4604 sequence starting with @code{.bundle_lock} and ending with
4605 @code{.bundle_unlock} is treated as a single instruction. That is, the
4606 entire sequence must fit into a single bundle and may not span a bundle
4607 boundary. If necessary, no-op instructions will be inserted before the
4608 first instruction of the sequence so that the whole sequence starts on an
4609 aligned bundle boundary. It's an error if the sequence is longer than the
4612 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4613 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4614 nested. That is, a second @code{.bundle_lock} directive before the next
4615 @code{.bundle_unlock} directive has no effect except that it must be
4616 matched by another closing @code{.bundle_unlock} so that there is the
4617 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4620 @section @code{.byte @var{expressions}}
4622 @cindex @code{byte} directive
4623 @cindex integers, one byte
4624 @code{.byte} expects zero or more expressions, separated by commas.
4625 Each expression is assembled into the next byte.
4627 @node CFI directives
4628 @section CFI directives
4629 @subsection @code{.cfi_sections @var{section_list}}
4630 @cindex @code{cfi_sections} directive
4631 @code{.cfi_sections} may be used to specify whether CFI directives
4632 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4633 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4634 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4635 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4636 directive is not used is @code{.cfi_sections .eh_frame}.
4638 On targets that support compact unwinding tables these can be generated
4639 by specifying @code{.eh_frame_entry} instead of @code{.eh_frame}.
4641 Some targets may support an additional name, such as @code{.c6xabi.exidx}
4642 which is used by the @value{TIC6X} target.
4644 The @code{.cfi_sections} directive can be repeated, with the same or different
4645 arguments, provided that CFI generation has not yet started. Once CFI
4646 generation has started however the section list is fixed and any attempts to
4647 redefine it will result in an error.
4649 @subsection @code{.cfi_startproc [simple]}
4650 @cindex @code{cfi_startproc} directive
4651 @code{.cfi_startproc} is used at the beginning of each function that
4652 should have an entry in @code{.eh_frame}. It initializes some internal
4653 data structures. Don't forget to close the function by
4654 @code{.cfi_endproc}.
4656 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4657 it also emits some architecture dependent initial CFI instructions.
4659 @subsection @code{.cfi_endproc}
4660 @cindex @code{cfi_endproc} directive
4661 @code{.cfi_endproc} is used at the end of a function where it closes its
4662 unwind entry previously opened by
4663 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4665 @subsection @code{.cfi_personality @var{encoding} [, @var{exp}]}
4666 @cindex @code{cfi_personality} directive
4667 @code{.cfi_personality} defines personality routine and its encoding.
4668 @var{encoding} must be a constant determining how the personality
4669 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4670 argument is not present, otherwise second argument should be
4671 a constant or a symbol name. When using indirect encodings,
4672 the symbol provided should be the location where personality
4673 can be loaded from, not the personality routine itself.
4674 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4675 no personality routine.
4677 @subsection @code{.cfi_personality_id @var{id}}
4678 @cindex @code{cfi_personality_id} directive
4679 @code{cfi_personality_id} defines a personality routine by its index as
4680 defined in a compact unwinding format.
4681 Only valid when generating compact EH frames (i.e.
4682 with @code{.cfi_sections eh_frame_entry}.
4684 @subsection @code{.cfi_fde_data [@var{opcode1} [, @dots{}]]}
4685 @cindex @code{cfi_fde_data} directive
4686 @code{cfi_fde_data} is used to describe the compact unwind opcodes to be
4687 used for the current function. These are emitted inline in the
4688 @code{.eh_frame_entry} section if small enough and there is no LSDA, or
4689 in the @code{.gnu.extab} section otherwise.
4690 Only valid when generating compact EH frames (i.e.
4691 with @code{.cfi_sections eh_frame_entry}.
4693 @subsection @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4694 @code{.cfi_lsda} defines LSDA and its encoding.
4695 @var{encoding} must be a constant determining how the LSDA
4696 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), the second
4697 argument is not present, otherwise the second argument should be a constant
4698 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4699 meaning that no LSDA is present.
4701 @subsection @code{.cfi_inline_lsda} [@var{align}]
4702 @code{.cfi_inline_lsda} marks the start of a LSDA data section and
4703 switches to the corresponding @code{.gnu.extab} section.
4704 Must be preceded by a CFI block containing a @code{.cfi_lsda} directive.
4705 Only valid when generating compact EH frames (i.e.
4706 with @code{.cfi_sections eh_frame_entry}.
4708 The table header and unwinding opcodes will be generated at this point,
4709 so that they are immediately followed by the LSDA data. The symbol
4710 referenced by the @code{.cfi_lsda} directive should still be defined
4711 in case a fallback FDE based encoding is used. The LSDA data is terminated
4712 by a section directive.
4714 The optional @var{align} argument specifies the alignment required.
4715 The alignment is specified as a power of two, as with the
4716 @code{.p2align} directive.
4718 @subsection @code{.cfi_def_cfa @var{register}, @var{offset}}
4719 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4720 address from @var{register} and add @var{offset} to it}.
4722 @subsection @code{.cfi_def_cfa_register @var{register}}
4723 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4724 now on @var{register} will be used instead of the old one. Offset
4727 @subsection @code{.cfi_def_cfa_offset @var{offset}}
4728 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4729 remains the same, but @var{offset} is new. Note that it is the
4730 absolute offset that will be added to a defined register to compute
4733 @subsection @code{.cfi_adjust_cfa_offset @var{offset}}
4734 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4735 value that is added/subtracted from the previous offset.
4737 @subsection @code{.cfi_offset @var{register}, @var{offset}}
4738 Previous value of @var{register} is saved at offset @var{offset} from
4741 @subsection @code{.cfi_val_offset @var{register}, @var{offset}}
4742 Previous value of @var{register} is CFA + @var{offset}.
4744 @subsection @code{.cfi_rel_offset @var{register}, @var{offset}}
4745 Previous value of @var{register} is saved at offset @var{offset} from
4746 the current CFA register. This is transformed to @code{.cfi_offset}
4747 using the known displacement of the CFA register from the CFA.
4748 This is often easier to use, because the number will match the
4749 code it's annotating.
4751 @subsection @code{.cfi_register @var{register1}, @var{register2}}
4752 Previous value of @var{register1} is saved in register @var{register2}.
4754 @subsection @code{.cfi_restore @var{register}}
4755 @code{.cfi_restore} says that the rule for @var{register} is now the
4756 same as it was at the beginning of the function, after all initial
4757 instruction added by @code{.cfi_startproc} were executed.
4759 @subsection @code{.cfi_undefined @var{register}}
4760 From now on the previous value of @var{register} can't be restored anymore.
4762 @subsection @code{.cfi_same_value @var{register}}
4763 Current value of @var{register} is the same like in the previous frame,
4764 i.e. no restoration needed.
4766 @subsection @code{.cfi_remember_state} and @code{.cfi_restore_state}
4767 @code{.cfi_remember_state} pushes the set of rules for every register onto an
4768 implicit stack, while @code{.cfi_restore_state} pops them off the stack and
4769 places them in the current row. This is useful for situations where you have
4770 multiple @code{.cfi_*} directives that need to be undone due to the control
4771 flow of the program. For example, we could have something like this (assuming
4772 the CFA is the value of @code{rbp}):
4782 .cfi_def_cfa %rsp, 8
4785 /* Do something else */
4788 Here, we want the @code{.cfi} directives to affect only the rows corresponding
4789 to the instructions before @code{label}. This means we'd have to add multiple
4790 @code{.cfi} directives after @code{label} to recreate the original save
4791 locations of the registers, as well as setting the CFA back to the value of
4792 @code{rbp}. This would be clumsy, and result in a larger binary size. Instead,
4804 .cfi_def_cfa %rsp, 8
4808 /* Do something else */
4811 That way, the rules for the instructions after @code{label} will be the same
4812 as before the first @code{.cfi_restore} without having to use multiple
4813 @code{.cfi} directives.
4815 @subsection @code{.cfi_return_column @var{register}}
4816 Change return column @var{register}, i.e. the return address is either
4817 directly in @var{register} or can be accessed by rules for @var{register}.
4819 @subsection @code{.cfi_signal_frame}
4820 Mark current function as signal trampoline.
4822 @subsection @code{.cfi_window_save}
4823 SPARC register window has been saved.
4825 @subsection @code{.cfi_escape} @var{expression}[, @dots{}]
4826 Allows the user to add arbitrary bytes to the unwind info. One
4827 might use this to add OS-specific CFI opcodes, or generic CFI
4828 opcodes that GAS does not yet support.
4830 @subsection @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4831 The current value of @var{register} is @var{label}. The value of @var{label}
4832 will be encoded in the output file according to @var{encoding}; see the
4833 description of @code{.cfi_personality} for details on this encoding.
4835 The usefulness of equating a register to a fixed label is probably
4836 limited to the return address register. Here, it can be useful to
4837 mark a code segment that has only one return address which is reached
4838 by a direct branch and no copy of the return address exists in memory
4839 or another register.
4842 @section @code{.comm @var{symbol} , @var{length} }
4844 @cindex @code{comm} directive
4845 @cindex symbol, common
4846 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4847 common symbol in one object file may be merged with a defined or common symbol
4848 of the same name in another object file. If @code{@value{LD}} does not see a
4849 definition for the symbol--just one or more common symbols--then it will
4850 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4851 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4852 the same name, and they do not all have the same size, it will allocate space
4853 using the largest size.
4856 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4857 an optional third argument. This is the desired alignment of the symbol,
4858 specified for ELF as a byte boundary (for example, an alignment of 16 means
4859 that the least significant 4 bits of the address should be zero), and for PE
4860 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4861 boundary). The alignment must be an absolute expression, and it must be a
4862 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4863 common symbol, it will use the alignment when placing the symbol. If no
4864 alignment is specified, @command{@value{AS}} will set the alignment to the
4865 largest power of two less than or equal to the size of the symbol, up to a
4866 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4867 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4868 @samp{--section-alignment} option; image file sections in PE are aligned to
4869 multiples of 4096, which is far too large an alignment for ordinary variables.
4870 It is rather the default alignment for (non-debug) sections within object
4871 (@samp{*.o}) files, which are less strictly aligned.}.
4875 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4876 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4880 @section @code{.data @var{subsection}}
4881 @cindex @code{data} directive
4883 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4884 end of the data subsection numbered @var{subsection} (which is an
4885 absolute expression). If @var{subsection} is omitted, it defaults
4889 @section @code{.dc[@var{size}] @var{expressions}}
4890 @cindex @code{dc} directive
4892 The @code{.dc} directive expects zero or more @var{expressions} separated by
4893 commas. These expressions are evaluated and their values inserted into the
4894 current section. The size of the emitted value depends upon the suffix to the
4895 @code{.dc} directive:
4899 Emits N-bit values, where N is the size of an address on the target system.
4903 Emits double precision floating-point values.
4905 Emits 32-bit values.
4907 Emits single precision floating-point values.
4909 Emits 16-bit values.
4910 Note - this is true even on targets where the @code{.word} directive would emit
4913 Emits long double precision floating-point values.
4916 If no suffix is used then @samp{.w} is assumed.
4918 The byte ordering is target dependent, as is the size and format of floating
4922 @section @code{.dcb[@var{size}] @var{number} [,@var{fill}]}
4923 @cindex @code{dcb} directive
4924 This directive emits @var{number} copies of @var{fill}, each of @var{size}
4925 bytes. Both @var{number} and @var{fill} are absolute expressions. If the
4926 comma and @var{fill} are omitted, @var{fill} is assumed to be zero. The
4927 @var{size} suffix, if present, must be one of:
4931 Emits single byte values.
4933 Emits double-precision floating point values.
4935 Emits 4-byte values.
4937 Emits single-precision floating point values.
4939 Emits 2-byte values.
4941 Emits long double-precision floating point values.
4944 If the @var{size} suffix is omitted then @samp{.w} is assumed.
4946 The byte ordering is target dependent, as is the size and format of floating
4950 @section @code{.ds[@var{size}] @var{number} [,@var{fill}]}
4951 @cindex @code{ds} directive
4952 This directive emits @var{number} copies of @var{fill}, each of @var{size}
4953 bytes. Both @var{number} and @var{fill} are absolute expressions. If the
4954 comma and @var{fill} are omitted, @var{fill} is assumed to be zero. The
4955 @var{size} suffix, if present, must be one of:
4959 Emits single byte values.
4961 Emits 8-byte values.
4963 Emits 4-byte values.
4965 Emits 12-byte values.
4967 Emits 4-byte values.
4969 Emits 2-byte values.
4971 Emits 12-byte values.
4974 Note - unlike the @code{.dcb} directive the @samp{.d}, @samp{.s} and @samp{.x}
4975 suffixes do not indicate that floating-point values are to be inserted.
4977 If the @var{size} suffix is omitted then @samp{.w} is assumed.
4979 The byte ordering is target dependent.
4984 @section @code{.def @var{name}}
4986 @cindex @code{def} directive
4987 @cindex COFF symbols, debugging
4988 @cindex debugging COFF symbols
4989 Begin defining debugging information for a symbol @var{name}; the
4990 definition extends until the @code{.endef} directive is encountered.
4995 @section @code{.desc @var{symbol}, @var{abs-expression}}
4997 @cindex @code{desc} directive
4998 @cindex COFF symbol descriptor
4999 @cindex symbol descriptor, COFF
5000 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
5001 to the low 16 bits of an absolute expression.
5004 The @samp{.desc} directive is not available when @command{@value{AS}} is
5005 configured for COFF output; it is only for @code{a.out} or @code{b.out}
5006 object format. For the sake of compatibility, @command{@value{AS}} accepts
5007 it, but produces no output, when configured for COFF.
5013 @section @code{.dim}
5015 @cindex @code{dim} directive
5016 @cindex COFF auxiliary symbol information
5017 @cindex auxiliary symbol information, COFF
5018 This directive is generated by compilers to include auxiliary debugging
5019 information in the symbol table. It is only permitted inside
5020 @code{.def}/@code{.endef} pairs.
5024 @section @code{.double @var{flonums}}
5026 @cindex @code{double} directive
5027 @cindex floating point numbers (double)
5028 @code{.double} expects zero or more flonums, separated by commas. It
5029 assembles floating point numbers.
5031 The exact kind of floating point numbers emitted depends on how
5032 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5036 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
5037 in @sc{ieee} format.
5042 @section @code{.eject}
5044 @cindex @code{eject} directive
5045 @cindex new page, in listings
5046 @cindex page, in listings
5047 @cindex listing control: new page
5048 Force a page break at this point, when generating assembly listings.
5051 @section @code{.else}
5053 @cindex @code{else} directive
5054 @code{.else} is part of the @command{@value{AS}} support for conditional
5055 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
5056 of code to be assembled if the condition for the preceding @code{.if}
5060 @section @code{.elseif}
5062 @cindex @code{elseif} directive
5063 @code{.elseif} is part of the @command{@value{AS}} support for conditional
5064 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
5065 @code{.if} block that would otherwise fill the entire @code{.else} section.
5068 @section @code{.end}
5070 @cindex @code{end} directive
5071 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
5072 process anything in the file past the @code{.end} directive.
5076 @section @code{.endef}
5078 @cindex @code{endef} directive
5079 This directive flags the end of a symbol definition begun with
5084 @section @code{.endfunc}
5085 @cindex @code{endfunc} directive
5086 @code{.endfunc} marks the end of a function specified with @code{.func}.
5089 @section @code{.endif}
5091 @cindex @code{endif} directive
5092 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
5093 it marks the end of a block of code that is only assembled
5094 conditionally. @xref{If,,@code{.if}}.
5097 @section @code{.equ @var{symbol}, @var{expression}}
5099 @cindex @code{equ} directive
5100 @cindex assigning values to symbols
5101 @cindex symbols, assigning values to
5102 This directive sets the value of @var{symbol} to @var{expression}.
5103 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
5106 The syntax for @code{equ} on the HPPA is
5107 @samp{@var{symbol} .equ @var{expression}}.
5111 The syntax for @code{equ} on the Z80 is
5112 @samp{@var{symbol} equ @var{expression}}.
5113 On the Z80 it is an error if @var{symbol} is already defined,
5114 but the symbol is not protected from later redefinition.
5115 Compare @ref{Equiv}.
5119 @section @code{.equiv @var{symbol}, @var{expression}}
5120 @cindex @code{equiv} directive
5121 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
5122 the assembler will signal an error if @var{symbol} is already defined. Note a
5123 symbol which has been referenced but not actually defined is considered to be
5126 Except for the contents of the error message, this is roughly equivalent to
5133 plus it protects the symbol from later redefinition.
5136 @section @code{.eqv @var{symbol}, @var{expression}}
5137 @cindex @code{eqv} directive
5138 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
5139 evaluate the expression or any part of it immediately. Instead each time
5140 the resulting symbol is used in an expression, a snapshot of its current
5144 @section @code{.err}
5145 @cindex @code{err} directive
5146 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
5147 message and, unless the @option{-Z} option was used, it will not generate an
5148 object file. This can be used to signal an error in conditionally compiled code.
5151 @section @code{.error "@var{string}"}
5152 @cindex error directive
5154 Similarly to @code{.err}, this directive emits an error, but you can specify a
5155 string that will be emitted as the error message. If you don't specify the
5156 message, it defaults to @code{".error directive invoked in source file"}.
5157 @xref{Errors, ,Error and Warning Messages}.
5160 .error "This code has not been assembled and tested."
5164 @section @code{.exitm}
5165 Exit early from the current macro definition. @xref{Macro}.
5168 @section @code{.extern}
5170 @cindex @code{extern} directive
5171 @code{.extern} is accepted in the source program---for compatibility
5172 with other assemblers---but it is ignored. @command{@value{AS}} treats
5173 all undefined symbols as external.
5176 @section @code{.fail @var{expression}}
5178 @cindex @code{fail} directive
5179 Generates an error or a warning. If the value of the @var{expression} is 500
5180 or more, @command{@value{AS}} will print a warning message. If the value is less
5181 than 500, @command{@value{AS}} will print an error message. The message will
5182 include the value of @var{expression}. This can occasionally be useful inside
5183 complex nested macros or conditional assembly.
5186 @section @code{.file}
5187 @cindex @code{file} directive
5189 @ifclear no-file-dir
5190 There are two different versions of the @code{.file} directive. Targets
5191 that support DWARF2 line number information use the DWARF2 version of
5192 @code{.file}. Other targets use the default version.
5194 @subheading Default Version
5196 @cindex logical file name
5197 @cindex file name, logical
5198 This version of the @code{.file} directive tells @command{@value{AS}} that we
5199 are about to start a new logical file. The syntax is:
5205 @var{string} is the new file name. In general, the filename is
5206 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
5207 to specify an empty file name, you must give the quotes--@code{""}. This
5208 statement may go away in future: it is only recognized to be compatible with
5209 old @command{@value{AS}} programs.
5211 @subheading DWARF2 Version
5214 When emitting DWARF2 line number information, @code{.file} assigns filenames
5215 to the @code{.debug_line} file name table. The syntax is:
5218 .file @var{fileno} @var{filename}
5221 The @var{fileno} operand should be a unique positive integer to use as the
5222 index of the entry in the table. The @var{filename} operand is a C string
5225 The detail of filename indices is exposed to the user because the filename
5226 table is shared with the @code{.debug_info} section of the DWARF2 debugging
5227 information, and thus the user must know the exact indices that table
5231 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
5233 @cindex @code{fill} directive
5234 @cindex writing patterns in memory
5235 @cindex patterns, writing in memory
5236 @var{repeat}, @var{size} and @var{value} are absolute expressions.
5237 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
5238 may be zero or more. @var{Size} may be zero or more, but if it is
5239 more than 8, then it is deemed to have the value 8, compatible with
5240 other people's assemblers. The contents of each @var{repeat} bytes
5241 is taken from an 8-byte number. The highest order 4 bytes are
5242 zero. The lowest order 4 bytes are @var{value} rendered in the
5243 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
5244 Each @var{size} bytes in a repetition is taken from the lowest order
5245 @var{size} bytes of this number. Again, this bizarre behavior is
5246 compatible with other people's assemblers.
5248 @var{size} and @var{value} are optional.
5249 If the second comma and @var{value} are absent, @var{value} is
5250 assumed zero. If the first comma and following tokens are absent,
5251 @var{size} is assumed to be 1.
5254 @section @code{.float @var{flonums}}
5256 @cindex floating point numbers (single)
5257 @cindex @code{float} directive
5258 This directive assembles zero or more flonums, separated by commas. It
5259 has the same effect as @code{.single}.
5261 The exact kind of floating point numbers emitted depends on how
5262 @command{@value{AS}} is configured.
5263 @xref{Machine Dependencies}.
5267 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
5268 in @sc{ieee} format.
5273 @section @code{.func @var{name}[,@var{label}]}
5274 @cindex @code{func} directive
5275 @code{.func} emits debugging information to denote function @var{name}, and
5276 is ignored unless the file is assembled with debugging enabled.
5277 Only @samp{--gstabs[+]} is currently supported.
5278 @var{label} is the entry point of the function and if omitted @var{name}
5279 prepended with the @samp{leading char} is used.
5280 @samp{leading char} is usually @code{_} or nothing, depending on the target.
5281 All functions are currently defined to have @code{void} return type.
5282 The function must be terminated with @code{.endfunc}.
5285 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
5287 @cindex @code{global} directive
5288 @cindex symbol, making visible to linker
5289 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
5290 @var{symbol} in your partial program, its value is made available to
5291 other partial programs that are linked with it. Otherwise,
5292 @var{symbol} takes its attributes from a symbol of the same name
5293 from another file linked into the same program.
5295 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
5296 compatibility with other assemblers.
5299 On the HPPA, @code{.global} is not always enough to make it accessible to other
5300 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
5301 @xref{HPPA Directives, ,HPPA Assembler Directives}.
5306 @section @code{.gnu_attribute @var{tag},@var{value}}
5307 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
5310 @section @code{.hidden @var{names}}
5312 @cindex @code{hidden} directive
5314 This is one of the ELF visibility directives. The other two are
5315 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
5316 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5318 This directive overrides the named symbols default visibility (which is set by
5319 their binding: local, global or weak). The directive sets the visibility to
5320 @code{hidden} which means that the symbols are not visible to other components.
5321 Such symbols are always considered to be @code{protected} as well.
5325 @section @code{.hword @var{expressions}}
5327 @cindex @code{hword} directive
5328 @cindex integers, 16-bit
5329 @cindex numbers, 16-bit
5330 @cindex sixteen bit integers
5331 This expects zero or more @var{expressions}, and emits
5332 a 16 bit number for each.
5335 This directive is a synonym for @samp{.short}; depending on the target
5336 architecture, it may also be a synonym for @samp{.word}.
5340 This directive is a synonym for @samp{.short}.
5343 This directive is a synonym for both @samp{.short} and @samp{.word}.
5348 @section @code{.ident}
5350 @cindex @code{ident} directive
5352 This directive is used by some assemblers to place tags in object files. The
5353 behavior of this directive varies depending on the target. When using the
5354 a.out object file format, @command{@value{AS}} simply accepts the directive for
5355 source-file compatibility with existing assemblers, but does not emit anything
5356 for it. When using COFF, comments are emitted to the @code{.comment} or
5357 @code{.rdata} section, depending on the target. When using ELF, comments are
5358 emitted to the @code{.comment} section.
5361 @section @code{.if @var{absolute expression}}
5363 @cindex conditional assembly
5364 @cindex @code{if} directive
5365 @code{.if} marks the beginning of a section of code which is only
5366 considered part of the source program being assembled if the argument
5367 (which must be an @var{absolute expression}) is non-zero. The end of
5368 the conditional section of code must be marked by @code{.endif}
5369 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
5370 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
5371 If you have several conditions to check, @code{.elseif} may be used to avoid
5372 nesting blocks if/else within each subsequent @code{.else} block.
5374 The following variants of @code{.if} are also supported:
5376 @cindex @code{ifdef} directive
5377 @item .ifdef @var{symbol}
5378 Assembles the following section of code if the specified @var{symbol}
5379 has been defined. Note a symbol which has been referenced but not yet defined
5380 is considered to be undefined.
5382 @cindex @code{ifb} directive
5383 @item .ifb @var{text}
5384 Assembles the following section of code if the operand is blank (empty).
5386 @cindex @code{ifc} directive
5387 @item .ifc @var{string1},@var{string2}
5388 Assembles the following section of code if the two strings are the same. The
5389 strings may be optionally quoted with single quotes. If they are not quoted,
5390 the first string stops at the first comma, and the second string stops at the
5391 end of the line. Strings which contain whitespace should be quoted. The
5392 string comparison is case sensitive.
5394 @cindex @code{ifeq} directive
5395 @item .ifeq @var{absolute expression}
5396 Assembles the following section of code if the argument is zero.
5398 @cindex @code{ifeqs} directive
5399 @item .ifeqs @var{string1},@var{string2}
5400 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5402 @cindex @code{ifge} directive
5403 @item .ifge @var{absolute expression}
5404 Assembles the following section of code if the argument is greater than or
5407 @cindex @code{ifgt} directive
5408 @item .ifgt @var{absolute expression}
5409 Assembles the following section of code if the argument is greater than zero.
5411 @cindex @code{ifle} directive
5412 @item .ifle @var{absolute expression}
5413 Assembles the following section of code if the argument is less than or equal
5416 @cindex @code{iflt} directive
5417 @item .iflt @var{absolute expression}
5418 Assembles the following section of code if the argument is less than zero.
5420 @cindex @code{ifnb} directive
5421 @item .ifnb @var{text}
5422 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5423 following section of code if the operand is non-blank (non-empty).
5425 @cindex @code{ifnc} directive
5426 @item .ifnc @var{string1},@var{string2}.
5427 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5428 following section of code if the two strings are not the same.
5430 @cindex @code{ifndef} directive
5431 @cindex @code{ifnotdef} directive
5432 @item .ifndef @var{symbol}
5433 @itemx .ifnotdef @var{symbol}
5434 Assembles the following section of code if the specified @var{symbol}
5435 has not been defined. Both spelling variants are equivalent. Note a symbol
5436 which has been referenced but not yet defined is considered to be undefined.
5438 @cindex @code{ifne} directive
5439 @item .ifne @var{absolute expression}
5440 Assembles the following section of code if the argument is not equal to zero
5441 (in other words, this is equivalent to @code{.if}).
5443 @cindex @code{ifnes} directive
5444 @item .ifnes @var{string1},@var{string2}
5445 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5446 following section of code if the two strings are not the same.
5450 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5452 @cindex @code{incbin} directive
5453 @cindex binary files, including
5454 The @code{incbin} directive includes @var{file} verbatim at the current
5455 location. You can control the search paths used with the @samp{-I} command-line
5456 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5459 The @var{skip} argument skips a number of bytes from the start of the
5460 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5461 read. Note that the data is not aligned in any way, so it is the user's
5462 responsibility to make sure that proper alignment is provided both before and
5463 after the @code{incbin} directive.
5466 @section @code{.include "@var{file}"}
5468 @cindex @code{include} directive
5469 @cindex supporting files, including
5470 @cindex files, including
5471 This directive provides a way to include supporting files at specified
5472 points in your source program. The code from @var{file} is assembled as
5473 if it followed the point of the @code{.include}; when the end of the
5474 included file is reached, assembly of the original file continues. You
5475 can control the search paths used with the @samp{-I} command-line option
5476 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5480 @section @code{.int @var{expressions}}
5482 @cindex @code{int} directive
5483 @cindex integers, 32-bit
5484 Expect zero or more @var{expressions}, of any section, separated by commas.
5485 For each expression, emit a number that, at run time, is the value of that
5486 expression. The byte order and bit size of the number depends on what kind
5487 of target the assembly is for.
5491 On most forms of the H8/300, @code{.int} emits 16-bit
5492 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5499 @section @code{.internal @var{names}}
5501 @cindex @code{internal} directive
5503 This is one of the ELF visibility directives. The other two are
5504 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5505 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5507 This directive overrides the named symbols default visibility (which is set by
5508 their binding: local, global or weak). The directive sets the visibility to
5509 @code{internal} which means that the symbols are considered to be @code{hidden}
5510 (i.e., not visible to other components), and that some extra, processor specific
5511 processing must also be performed upon the symbols as well.
5515 @section @code{.irp @var{symbol},@var{values}}@dots{}
5517 @cindex @code{irp} directive
5518 Evaluate a sequence of statements assigning different values to @var{symbol}.
5519 The sequence of statements starts at the @code{.irp} directive, and is
5520 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5521 set to @var{value}, and the sequence of statements is assembled. If no
5522 @var{value} is listed, the sequence of statements is assembled once, with
5523 @var{symbol} set to the null string. To refer to @var{symbol} within the
5524 sequence of statements, use @var{\symbol}.
5526 For example, assembling
5534 is equivalent to assembling
5542 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5545 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5547 @cindex @code{irpc} directive
5548 Evaluate a sequence of statements assigning different values to @var{symbol}.
5549 The sequence of statements starts at the @code{.irpc} directive, and is
5550 terminated by an @code{.endr} directive. For each character in @var{value},
5551 @var{symbol} is set to the character, and the sequence of statements is
5552 assembled. If no @var{value} is listed, the sequence of statements is
5553 assembled once, with @var{symbol} set to the null string. To refer to
5554 @var{symbol} within the sequence of statements, use @var{\symbol}.
5556 For example, assembling
5564 is equivalent to assembling
5572 For some caveats with the spelling of @var{symbol}, see also the discussion
5576 @section @code{.lcomm @var{symbol} , @var{length}}
5578 @cindex @code{lcomm} directive
5579 @cindex local common symbols
5580 @cindex symbols, local common
5581 Reserve @var{length} (an absolute expression) bytes for a local common
5582 denoted by @var{symbol}. The section and value of @var{symbol} are
5583 those of the new local common. The addresses are allocated in the bss
5584 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5585 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5586 not visible to @code{@value{LD}}.
5589 Some targets permit a third argument to be used with @code{.lcomm}. This
5590 argument specifies the desired alignment of the symbol in the bss section.
5594 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5595 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5599 @section @code{.lflags}
5601 @cindex @code{lflags} directive (ignored)
5602 @command{@value{AS}} accepts this directive, for compatibility with other
5603 assemblers, but ignores it.
5605 @ifclear no-line-dir
5607 @section @code{.line @var{line-number}}
5609 @cindex @code{line} directive
5610 @cindex logical line number
5612 Change the logical line number. @var{line-number} must be an absolute
5613 expression. The next line has that logical line number. Therefore any other
5614 statements on the current line (after a statement separator character) are
5615 reported as on logical line number @var{line-number} @minus{} 1. One day
5616 @command{@value{AS}} will no longer support this directive: it is recognized only
5617 for compatibility with existing assembler programs.
5620 Even though this is a directive associated with the @code{a.out} or
5621 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5622 when producing COFF output, and treats @samp{.line} as though it
5623 were the COFF @samp{.ln} @emph{if} it is found outside a
5624 @code{.def}/@code{.endef} pair.
5626 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5627 used by compilers to generate auxiliary symbol information for
5632 @section @code{.linkonce [@var{type}]}
5634 @cindex @code{linkonce} directive
5635 @cindex common sections
5636 Mark the current section so that the linker only includes a single copy of it.
5637 This may be used to include the same section in several different object files,
5638 but ensure that the linker will only include it once in the final output file.
5639 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5640 Duplicate sections are detected based on the section name, so it should be
5643 This directive is only supported by a few object file formats; as of this
5644 writing, the only object file format which supports it is the Portable
5645 Executable format used on Windows NT.
5647 The @var{type} argument is optional. If specified, it must be one of the
5648 following strings. For example:
5652 Not all types may be supported on all object file formats.
5656 Silently discard duplicate sections. This is the default.
5659 Warn if there are duplicate sections, but still keep only one copy.
5662 Warn if any of the duplicates have different sizes.
5665 Warn if any of the duplicates do not have exactly the same contents.
5669 @section @code{.list}
5671 @cindex @code{list} directive
5672 @cindex listing control, turning on
5673 Control (in conjunction with the @code{.nolist} directive) whether or
5674 not assembly listings are generated. These two directives maintain an
5675 internal counter (which is zero initially). @code{.list} increments the
5676 counter, and @code{.nolist} decrements it. Assembly listings are
5677 generated whenever the counter is greater than zero.
5679 By default, listings are disabled. When you enable them (with the
5680 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5681 the initial value of the listing counter is one.
5684 @section @code{.ln @var{line-number}}
5686 @cindex @code{ln} directive
5687 @ifclear no-line-dir
5688 @samp{.ln} is a synonym for @samp{.line}.
5691 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5692 must be an absolute expression. The next line has that logical
5693 line number, so any other statements on the current line (after a
5694 statement separator character @code{;}) are reported as on logical
5695 line number @var{line-number} @minus{} 1.
5699 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5700 @cindex @code{loc} directive
5701 When emitting DWARF2 line number information,
5702 the @code{.loc} directive will add a row to the @code{.debug_line} line
5703 number matrix corresponding to the immediately following assembly
5704 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5705 arguments will be applied to the @code{.debug_line} state machine before
5708 The @var{options} are a sequence of the following tokens in any order:
5712 This option will set the @code{basic_block} register in the
5713 @code{.debug_line} state machine to @code{true}.
5716 This option will set the @code{prologue_end} register in the
5717 @code{.debug_line} state machine to @code{true}.
5719 @item epilogue_begin
5720 This option will set the @code{epilogue_begin} register in the
5721 @code{.debug_line} state machine to @code{true}.
5723 @item is_stmt @var{value}
5724 This option will set the @code{is_stmt} register in the
5725 @code{.debug_line} state machine to @code{value}, which must be
5728 @item isa @var{value}
5729 This directive will set the @code{isa} register in the @code{.debug_line}
5730 state machine to @var{value}, which must be an unsigned integer.
5732 @item discriminator @var{value}
5733 This directive will set the @code{discriminator} register in the @code{.debug_line}
5734 state machine to @var{value}, which must be an unsigned integer.
5736 @item view @var{value}
5737 This option causes a row to be added to @code{.debug_line} in reference to the
5738 current address (which might not be the same as that of the following assembly
5739 instruction), and to associate @var{value} with the @code{view} register in the
5740 @code{.debug_line} state machine. If @var{value} is a label, both the
5741 @code{view} register and the label are set to the number of prior @code{.loc}
5742 directives at the same program location. If @var{value} is the literal
5743 @code{0}, the @code{view} register is set to zero, and the assembler asserts
5744 that there aren't any prior @code{.loc} directives at the same program
5745 location. If @var{value} is the literal @code{-0}, the assembler arrange for
5746 the @code{view} register to be reset in this row, even if there are prior
5747 @code{.loc} directives at the same program location.
5751 @node Loc_mark_labels
5752 @section @code{.loc_mark_labels @var{enable}}
5753 @cindex @code{loc_mark_labels} directive
5754 When emitting DWARF2 line number information,
5755 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5756 to the @code{.debug_line} line number matrix with the @code{basic_block}
5757 register in the state machine set whenever a code label is seen.
5758 The @var{enable} argument should be either 1 or 0, to enable or disable
5759 this function respectively.
5763 @section @code{.local @var{names}}
5765 @cindex @code{local} directive
5766 This directive, which is available for ELF targets, marks each symbol in
5767 the comma-separated list of @code{names} as a local symbol so that it
5768 will not be externally visible. If the symbols do not already exist,
5769 they will be created.
5771 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5772 accept an alignment argument, which is the case for most ELF targets,
5773 the @code{.local} directive can be used in combination with @code{.comm}
5774 (@pxref{Comm}) to define aligned local common data.
5778 @section @code{.long @var{expressions}}
5780 @cindex @code{long} directive
5781 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5784 @c no one seems to know what this is for or whether this description is
5785 @c what it really ought to do
5787 @section @code{.lsym @var{symbol}, @var{expression}}
5789 @cindex @code{lsym} directive
5790 @cindex symbol, not referenced in assembly
5791 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5792 the hash table, ensuring it cannot be referenced by name during the
5793 rest of the assembly. This sets the attributes of the symbol to be
5794 the same as the expression value:
5796 @var{other} = @var{descriptor} = 0
5797 @var{type} = @r{(section of @var{expression})}
5798 @var{value} = @var{expression}
5801 The new symbol is not flagged as external.
5805 @section @code{.macro}
5808 The commands @code{.macro} and @code{.endm} allow you to define macros that
5809 generate assembly output. For example, this definition specifies a macro
5810 @code{sum} that puts a sequence of numbers into memory:
5813 .macro sum from=0, to=5
5822 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5834 @item .macro @var{macname}
5835 @itemx .macro @var{macname} @var{macargs} @dots{}
5836 @cindex @code{macro} directive
5837 Begin the definition of a macro called @var{macname}. If your macro
5838 definition requires arguments, specify their names after the macro name,
5839 separated by commas or spaces. You can qualify the macro argument to
5840 indicate whether all invocations must specify a non-blank value (through
5841 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5842 (through @samp{:@code{vararg}}). You can supply a default value for any
5843 macro argument by following the name with @samp{=@var{deflt}}. You
5844 cannot define two macros with the same @var{macname} unless it has been
5845 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5846 definitions. For example, these are all valid @code{.macro} statements:
5850 Begin the definition of a macro called @code{comm}, which takes no
5853 @item .macro plus1 p, p1
5854 @itemx .macro plus1 p p1
5855 Either statement begins the definition of a macro called @code{plus1},
5856 which takes two arguments; within the macro definition, write
5857 @samp{\p} or @samp{\p1} to evaluate the arguments.
5859 @item .macro reserve_str p1=0 p2
5860 Begin the definition of a macro called @code{reserve_str}, with two
5861 arguments. The first argument has a default value, but not the second.
5862 After the definition is complete, you can call the macro either as
5863 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5864 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5865 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5866 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5868 @item .macro m p1:req, p2=0, p3:vararg
5869 Begin the definition of a macro called @code{m}, with at least three
5870 arguments. The first argument must always have a value specified, but
5871 not the second, which instead has a default value. The third formal
5872 will get assigned all remaining arguments specified at invocation time.
5874 When you call a macro, you can specify the argument values either by
5875 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5876 @samp{sum to=17, from=9}.
5880 Note that since each of the @var{macargs} can be an identifier exactly
5881 as any other one permitted by the target architecture, there may be
5882 occasional problems if the target hand-crafts special meanings to certain
5883 characters when they occur in a special position. For example, if the colon
5884 (@code{:}) is generally permitted to be part of a symbol name, but the
5885 architecture specific code special-cases it when occurring as the final
5886 character of a symbol (to denote a label), then the macro parameter
5887 replacement code will have no way of knowing that and consider the whole
5888 construct (including the colon) an identifier, and check only this
5889 identifier for being the subject to parameter substitution. So for example
5890 this macro definition:
5898 might not work as expected. Invoking @samp{label foo} might not create a label
5899 called @samp{foo} but instead just insert the text @samp{\l:} into the
5900 assembler source, probably generating an error about an unrecognised
5903 Similarly problems might occur with the period character (@samp{.})
5904 which is often allowed inside opcode names (and hence identifier names). So
5905 for example constructing a macro to build an opcode from a base name and a
5906 length specifier like this:
5909 .macro opcode base length
5914 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5915 instruction but instead generate some kind of error as the assembler tries to
5916 interpret the text @samp{\base.\length}.
5918 There are several possible ways around this problem:
5921 @item Insert white space
5922 If it is possible to use white space characters then this is the simplest
5931 @item Use @samp{\()}
5932 The string @samp{\()} can be used to separate the end of a macro argument from
5933 the following text. eg:
5936 .macro opcode base length
5941 @item Use the alternate macro syntax mode
5942 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5943 used as a separator. eg:
5953 Note: this problem of correctly identifying string parameters to pseudo ops
5954 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5955 and @code{.irpc} (@pxref{Irpc}) as well.
5958 @cindex @code{endm} directive
5959 Mark the end of a macro definition.
5962 @cindex @code{exitm} directive
5963 Exit early from the current macro definition.
5965 @cindex number of macros executed
5966 @cindex macros, count executed
5968 @command{@value{AS}} maintains a counter of how many macros it has
5969 executed in this pseudo-variable; you can copy that number to your
5970 output with @samp{\@@}, but @emph{only within a macro definition}.
5972 @item LOCAL @var{name} [ , @dots{} ]
5973 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5974 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5975 @xref{Altmacro,,@code{.altmacro}}.
5979 @section @code{.mri @var{val}}
5981 @cindex @code{mri} directive
5982 @cindex MRI mode, temporarily
5983 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
5984 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
5985 affects code assembled until the next @code{.mri} directive, or until the end
5986 of the file. @xref{M, MRI mode, MRI mode}.
5989 @section @code{.noaltmacro}
5990 Disable alternate macro mode. @xref{Altmacro}.
5993 @section @code{.nolist}
5995 @cindex @code{nolist} directive
5996 @cindex listing control, turning off
5997 Control (in conjunction with the @code{.list} directive) whether or
5998 not assembly listings are generated. These two directives maintain an
5999 internal counter (which is zero initially). @code{.list} increments the
6000 counter, and @code{.nolist} decrements it. Assembly listings are
6001 generated whenever the counter is greater than zero.
6004 @section @code{.nops @var{size}[, @var{control}]}
6006 @cindex @code{nops} directive
6007 @cindex filling memory with no-op instructions
6008 This directive emits @var{size} bytes filled with no-op instructions.
6009 @var{size} is absolute expression, which must be a positve value.
6010 @var{control} controls how no-op instructions should be generated. If
6011 the comma and @var{control} are omitted, @var{control} is assumed to be
6014 Note: For Intel 80386 and AMD x86-64 targets, @var{control} specifies
6015 the size limit of a no-op instruction. The valid values of @var{control}
6016 are between 0 and 4 in 16-bit mode, between 0 and 7 when tuning for
6017 older processors in 32-bit mode, between 0 and 11 in 64-bit mode or when
6018 tuning for newer processors in 32-bit mode. When 0 is used, the no-op
6019 instruction size limit is set to the maximum supported size.
6022 @section @code{.octa @var{bignums}}
6024 @c FIXME: double size emitted for "octa" on some? Or warn?
6025 @cindex @code{octa} directive
6026 @cindex integer, 16-byte
6027 @cindex sixteen byte integer
6028 This directive expects zero or more bignums, separated by commas. For each
6029 bignum, it emits a 16-byte integer.
6031 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
6032 hence @emph{octa}-word for 16 bytes.
6035 @section @code{.offset @var{loc}}
6037 @cindex @code{offset} directive
6038 Set the location counter to @var{loc} in the absolute section. @var{loc} must
6039 be an absolute expression. This directive may be useful for defining
6040 symbols with absolute values. Do not confuse it with the @code{.org}
6044 @section @code{.org @var{new-lc} , @var{fill}}
6046 @cindex @code{org} directive
6047 @cindex location counter, advancing
6048 @cindex advancing location counter
6049 @cindex current address, advancing
6050 Advance the location counter of the current section to
6051 @var{new-lc}. @var{new-lc} is either an absolute expression or an
6052 expression with the same section as the current subsection. That is,
6053 you can't use @code{.org} to cross sections: if @var{new-lc} has the
6054 wrong section, the @code{.org} directive is ignored. To be compatible
6055 with former assemblers, if the section of @var{new-lc} is absolute,
6056 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
6057 is the same as the current subsection.
6059 @code{.org} may only increase the location counter, or leave it
6060 unchanged; you cannot use @code{.org} to move the location counter
6063 @c double negative used below "not undefined" because this is a specific
6064 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
6065 @c section. doc@cygnus.com 18feb91
6066 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
6067 may not be undefined. If you really detest this restriction we eagerly await
6068 a chance to share your improved assembler.
6070 Beware that the origin is relative to the start of the section, not
6071 to the start of the subsection. This is compatible with other
6072 people's assemblers.
6074 When the location counter (of the current subsection) is advanced, the
6075 intervening bytes are filled with @var{fill} which should be an
6076 absolute expression. If the comma and @var{fill} are omitted,
6077 @var{fill} defaults to zero.
6080 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
6082 @cindex padding the location counter given a power of two
6083 @cindex @code{p2align} directive
6084 Pad the location counter (in the current subsection) to a particular
6085 storage boundary. The first expression (which must be absolute) is the
6086 number of low-order zero bits the location counter must have after
6087 advancement. For example @samp{.p2align 3} advances the location
6088 counter until it a multiple of 8. If the location counter is already a
6089 multiple of 8, no change is needed.
6091 The second expression (also absolute) gives the fill value to be stored in the
6092 padding bytes. It (and the comma) may be omitted. If it is omitted, the
6093 padding bytes are normally zero. However, on most systems, if the section is
6094 marked as containing code and the fill value is omitted, the space is filled
6095 with no-op instructions.
6097 The third expression is also absolute, and is also optional. If it is present,
6098 it is the maximum number of bytes that should be skipped by this alignment
6099 directive. If doing the alignment would require skipping more bytes than the
6100 specified maximum, then the alignment is not done at all. You can omit the
6101 fill value (the second argument) entirely by simply using two commas after the
6102 required alignment; this can be useful if you want the alignment to be filled
6103 with no-op instructions when appropriate.
6105 @cindex @code{p2alignw} directive
6106 @cindex @code{p2alignl} directive
6107 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
6108 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
6109 pattern as a two byte word value. The @code{.p2alignl} directives treats the
6110 fill pattern as a four byte longword value. For example, @code{.p2alignw
6111 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
6112 filled in with the value 0x368d (the exact placement of the bytes depends upon
6113 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
6118 @section @code{.popsection}
6120 @cindex @code{popsection} directive
6121 @cindex Section Stack
6122 This is one of the ELF section stack manipulation directives. The others are
6123 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6124 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
6127 This directive replaces the current section (and subsection) with the top
6128 section (and subsection) on the section stack. This section is popped off the
6134 @section @code{.previous}
6136 @cindex @code{previous} directive
6137 @cindex Section Stack
6138 This is one of the ELF section stack manipulation directives. The others are
6139 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6140 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
6141 (@pxref{PopSection}).
6143 This directive swaps the current section (and subsection) with most recently
6144 referenced section/subsection pair prior to this one. Multiple
6145 @code{.previous} directives in a row will flip between two sections (and their
6146 subsections). For example:
6158 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
6164 # Now in section A subsection 1
6168 # Now in section B subsection 0
6171 # Now in section B subsection 1
6174 # Now in section B subsection 0
6178 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
6179 section B and 0x9abc into subsection 1 of section B.
6181 In terms of the section stack, this directive swaps the current section with
6182 the top section on the section stack.
6186 @section @code{.print @var{string}}
6188 @cindex @code{print} directive
6189 @command{@value{AS}} will print @var{string} on the standard output during
6190 assembly. You must put @var{string} in double quotes.
6194 @section @code{.protected @var{names}}
6196 @cindex @code{protected} directive
6198 This is one of the ELF visibility directives. The other two are
6199 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
6201 This directive overrides the named symbols default visibility (which is set by
6202 their binding: local, global or weak). The directive sets the visibility to
6203 @code{protected} which means that any references to the symbols from within the
6204 components that defines them must be resolved to the definition in that
6205 component, even if a definition in another component would normally preempt
6210 @section @code{.psize @var{lines} , @var{columns}}
6212 @cindex @code{psize} directive
6213 @cindex listing control: paper size
6214 @cindex paper size, for listings
6215 Use this directive to declare the number of lines---and, optionally, the
6216 number of columns---to use for each page, when generating listings.
6218 If you do not use @code{.psize}, listings use a default line-count
6219 of 60. You may omit the comma and @var{columns} specification; the
6220 default width is 200 columns.
6222 @command{@value{AS}} generates formfeeds whenever the specified number of
6223 lines is exceeded (or whenever you explicitly request one, using
6226 If you specify @var{lines} as @code{0}, no formfeeds are generated save
6227 those explicitly specified with @code{.eject}.
6230 @section @code{.purgem @var{name}}
6232 @cindex @code{purgem} directive
6233 Undefine the macro @var{name}, so that later uses of the string will not be
6234 expanded. @xref{Macro}.
6238 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
6240 @cindex @code{pushsection} directive
6241 @cindex Section Stack
6242 This is one of the ELF section stack manipulation directives. The others are
6243 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6244 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6247 This directive pushes the current section (and subsection) onto the
6248 top of the section stack, and then replaces the current section and
6249 subsection with @code{name} and @code{subsection}. The optional
6250 @code{flags}, @code{type} and @code{arguments} are treated the same
6251 as in the @code{.section} (@pxref{Section}) directive.
6255 @section @code{.quad @var{bignums}}
6257 @cindex @code{quad} directive
6258 @code{.quad} expects zero or more bignums, separated by commas. For
6259 each bignum, it emits
6261 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
6262 warning message; and just takes the lowest order 8 bytes of the bignum.
6263 @cindex eight-byte integer
6264 @cindex integer, 8-byte
6266 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
6267 hence @emph{quad}-word for 8 bytes.
6270 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
6271 warning message; and just takes the lowest order 16 bytes of the bignum.
6272 @cindex sixteen-byte integer
6273 @cindex integer, 16-byte
6277 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
6279 @cindex @code{reloc} directive
6280 Generate a relocation at @var{offset} of type @var{reloc_name} with value
6281 @var{expression}. If @var{offset} is a number, the relocation is generated in
6282 the current section. If @var{offset} is an expression that resolves to a
6283 symbol plus offset, the relocation is generated in the given symbol's section.
6284 @var{expression}, if present, must resolve to a symbol plus addend or to an
6285 absolute value, but note that not all targets support an addend. e.g. ELF REL
6286 targets such as i386 store an addend in the section contents rather than in the
6287 relocation. This low level interface does not support addends stored in the
6291 @section @code{.rept @var{count}}
6293 @cindex @code{rept} directive
6294 Repeat the sequence of lines between the @code{.rept} directive and the next
6295 @code{.endr} directive @var{count} times.
6297 For example, assembling
6305 is equivalent to assembling
6313 A count of zero is allowed, but nothing is generated. Negative counts are not
6314 allowed and if encountered will be treated as if they were zero.
6317 @section @code{.sbttl "@var{subheading}"}
6319 @cindex @code{sbttl} directive
6320 @cindex subtitles for listings
6321 @cindex listing control: subtitle
6322 Use @var{subheading} as the title (third line, immediately after the
6323 title line) when generating assembly listings.
6325 This directive affects subsequent pages, as well as the current page if
6326 it appears within ten lines of the top of a page.
6330 @section @code{.scl @var{class}}
6332 @cindex @code{scl} directive
6333 @cindex symbol storage class (COFF)
6334 @cindex COFF symbol storage class
6335 Set the storage-class value for a symbol. This directive may only be
6336 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
6337 whether a symbol is static or external, or it may record further
6338 symbolic debugging information.
6343 @section @code{.section @var{name}}
6345 @cindex named section
6346 Use the @code{.section} directive to assemble the following code into a section
6349 This directive is only supported for targets that actually support arbitrarily
6350 named sections; on @code{a.out} targets, for example, it is not accepted, even
6351 with a standard @code{a.out} section name.
6355 @c only print the extra heading if both COFF and ELF are set
6356 @subheading COFF Version
6359 @cindex @code{section} directive (COFF version)
6360 For COFF targets, the @code{.section} directive is used in one of the following
6364 .section @var{name}[, "@var{flags}"]
6365 .section @var{name}[, @var{subsection}]
6368 If the optional argument is quoted, it is taken as flags to use for the
6369 section. Each flag is a single character. The following flags are recognized:
6373 bss section (uninitialized data)
6375 section is not loaded
6381 exclude section from linking
6387 shared section (meaningful for PE targets)
6389 ignored. (For compatibility with the ELF version)
6391 section is not readable (meaningful for PE targets)
6393 single-digit power-of-two section alignment (GNU extension)
6396 If no flags are specified, the default flags depend upon the section name. If
6397 the section name is not recognized, the default will be for the section to be
6398 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
6399 from the section, rather than adding them, so if they are used on their own it
6400 will be as if no flags had been specified at all.
6402 If the optional argument to the @code{.section} directive is not quoted, it is
6403 taken as a subsection number (@pxref{Sub-Sections}).
6408 @c only print the extra heading if both COFF and ELF are set
6409 @subheading ELF Version
6412 @cindex Section Stack
6413 This is one of the ELF section stack manipulation directives. The others are
6414 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
6415 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
6416 @code{.previous} (@pxref{Previous}).
6418 @cindex @code{section} directive (ELF version)
6419 For ELF targets, the @code{.section} directive is used like this:
6422 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
6425 @anchor{Section Name Substitutions}
6426 @kindex --sectname-subst
6427 @cindex section name substitution
6428 If the @samp{--sectname-subst} command-line option is provided, the @var{name}
6429 argument may contain a substitution sequence. Only @code{%S} is supported
6430 at the moment, and substitutes the current section name. For example:
6433 .macro exception_code
6434 .section %S.exception
6435 [exception code here]
6450 The two @code{exception_code} invocations above would create the
6451 @code{.text.exception} and @code{.init.exception} sections respectively.
6452 This is useful e.g. to discriminate between ancillary sections that are
6453 tied to setup code to be discarded after use from ancillary sections that
6454 need to stay resident without having to define multiple @code{exception_code}
6455 macros just for that purpose.
6457 The optional @var{flags} argument is a quoted string which may contain any
6458 combination of the following characters:
6462 section is allocatable
6464 section is a GNU_MBIND section
6466 section is excluded from executable and shared library.
6470 section is executable
6472 section is mergeable
6474 section contains zero terminated strings
6476 section is a member of a section group
6478 section is used for thread-local-storage
6480 section is a member of the previously-current section's group, if any
6481 @item @code{<number>}
6482 a numeric value indicating the bits to be set in the ELF section header's flags
6483 field. Note - if one or more of the alphabetic characters described above is
6484 also included in the flags field, their bit values will be ORed into the
6486 @item @code{<target specific>}
6487 some targets extend this list with their own flag characters
6490 Note - once a section's flags have been set they cannot be changed. There are
6491 a few exceptions to this rule however. Processor and application specific
6492 flags can be added to an already defined section. The @code{.interp},
6493 @code{.strtab} and @code{.symtab} sections can have the allocate flag
6494 (@code{a}) set after they are initially defined, and the @code{.note-GNU-stack}
6495 section may have the executable (@code{x}) flag added.
6497 The optional @var{type} argument may contain one of the following constants:
6501 section contains data
6503 section does not contain data (i.e., section only occupies space)
6505 section contains data which is used by things other than the program
6507 section contains an array of pointers to init functions
6509 section contains an array of pointers to finish functions
6510 @item @@preinit_array
6511 section contains an array of pointers to pre-init functions
6512 @item @@@code{<number>}
6513 a numeric value to be set as the ELF section header's type field.
6514 @item @@@code{<target specific>}
6515 some targets extend this list with their own types
6518 Many targets only support the first three section types. The type may be
6519 enclosed in double quotes if necessary.
6521 Note on targets where the @code{@@} character is the start of a comment (eg
6522 ARM) then another character is used instead. For example the ARM port uses the
6525 Note - some sections, eg @code{.text} and @code{.data} are considered to be
6526 special and have fixed types. Any attempt to declare them with a different
6527 type will generate an error from the assembler.
6529 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6530 be specified as well as an extra argument---@var{entsize}---like this:
6533 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6536 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6537 constants, each @var{entsize} octets long. Sections with both @code{M} and
6538 @code{S} must contain zero terminated strings where each character is
6539 @var{entsize} bytes long. The linker may remove duplicates within sections with
6540 the same name, same entity size and same flags. @var{entsize} must be an
6541 absolute expression. For sections with both @code{M} and @code{S}, a string
6542 which is a suffix of a larger string is considered a duplicate. Thus
6543 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6544 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6546 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6547 be present along with an additional field like this:
6550 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6553 The @var{GroupName} field specifies the name of the section group to which this
6554 particular section belongs. The optional linkage field can contain:
6558 indicates that only one copy of this section should be retained
6563 Note: if both the @var{M} and @var{G} flags are present then the fields for
6564 the Merge flag should come first, like this:
6567 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6570 If @var{flags} contains the @code{?} symbol then it may not also contain the
6571 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6572 present. Instead, @code{?} says to consider the section that's current before
6573 this directive. If that section used @code{G}, then the new section will use
6574 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6575 If not, then the @code{?} symbol has no effect.
6577 If no flags are specified, the default flags depend upon the section name. If
6578 the section name is not recognized, the default will be for the section to have
6579 none of the above flags: it will not be allocated in memory, nor writable, nor
6580 executable. The section will contain data.
6582 For ELF targets, the assembler supports another type of @code{.section}
6583 directive for compatibility with the Solaris assembler:
6586 .section "@var{name}"[, @var{flags}...]
6589 Note that the section name is quoted. There may be a sequence of comma
6594 section is allocatable
6598 section is executable
6600 section is excluded from executable and shared library.
6602 section is used for thread local storage
6605 This directive replaces the current section and subsection. See the
6606 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6607 some examples of how this directive and the other section stack directives
6613 @section @code{.set @var{symbol}, @var{expression}}
6615 @cindex @code{set} directive
6616 @cindex symbol value, setting
6617 Set the value of @var{symbol} to @var{expression}. This
6618 changes @var{symbol}'s value and type to conform to
6619 @var{expression}. If @var{symbol} was flagged as external, it remains
6620 flagged (@pxref{Symbol Attributes}).
6622 You may @code{.set} a symbol many times in the same assembly provided that the
6623 values given to the symbol are constants. Values that are based on expressions
6624 involving other symbols are allowed, but some targets may restrict this to only
6625 being done once per assembly. This is because those targets do not set the
6626 addresses of symbols at assembly time, but rather delay the assignment until a
6627 final link is performed. This allows the linker a chance to change the code in
6628 the files, changing the location of, and the relative distance between, various
6631 If you @code{.set} a global symbol, the value stored in the object
6632 file is the last value stored into it.
6635 On Z80 @code{set} is a real instruction, use
6636 @samp{@var{symbol} defl @var{expression}} instead.
6640 @section @code{.short @var{expressions}}
6642 @cindex @code{short} directive
6644 @code{.short} is normally the same as @samp{.word}.
6645 @xref{Word,,@code{.word}}.
6647 In some configurations, however, @code{.short} and @code{.word} generate
6648 numbers of different lengths. @xref{Machine Dependencies}.
6652 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6655 This expects zero or more @var{expressions}, and emits
6656 a 16 bit number for each.
6661 @section @code{.single @var{flonums}}
6663 @cindex @code{single} directive
6664 @cindex floating point numbers (single)
6665 This directive assembles zero or more flonums, separated by commas. It
6666 has the same effect as @code{.float}.
6668 The exact kind of floating point numbers emitted depends on how
6669 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6673 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6674 numbers in @sc{ieee} format.
6680 @section @code{.size}
6682 This directive is used to set the size associated with a symbol.
6686 @c only print the extra heading if both COFF and ELF are set
6687 @subheading COFF Version
6690 @cindex @code{size} directive (COFF version)
6691 For COFF targets, the @code{.size} directive is only permitted inside
6692 @code{.def}/@code{.endef} pairs. It is used like this:
6695 .size @var{expression}
6702 @c only print the extra heading if both COFF and ELF are set
6703 @subheading ELF Version
6706 @cindex @code{size} directive (ELF version)
6707 For ELF targets, the @code{.size} directive is used like this:
6710 .size @var{name} , @var{expression}
6713 This directive sets the size associated with a symbol @var{name}.
6714 The size in bytes is computed from @var{expression} which can make use of label
6715 arithmetic. This directive is typically used to set the size of function
6720 @ifclear no-space-dir
6722 @section @code{.skip @var{size} [,@var{fill}]}
6724 @cindex @code{skip} directive
6725 @cindex filling memory
6726 This directive emits @var{size} bytes, each of value @var{fill}. Both
6727 @var{size} and @var{fill} are absolute expressions. If the comma and
6728 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6733 @section @code{.sleb128 @var{expressions}}
6735 @cindex @code{sleb128} directive
6736 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6737 compact, variable length representation of numbers used by the DWARF
6738 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6740 @ifclear no-space-dir
6742 @section @code{.space @var{size} [,@var{fill}]}
6744 @cindex @code{space} directive
6745 @cindex filling memory
6746 This directive emits @var{size} bytes, each of value @var{fill}. Both
6747 @var{size} and @var{fill} are absolute expressions. If the comma
6748 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6753 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6754 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6755 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6756 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6764 @section @code{.stabd, .stabn, .stabs}
6766 @cindex symbolic debuggers, information for
6767 @cindex @code{stab@var{x}} directives
6768 There are three directives that begin @samp{.stab}.
6769 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6770 The symbols are not entered in the @command{@value{AS}} hash table: they
6771 cannot be referenced elsewhere in the source file.
6772 Up to five fields are required:
6776 This is the symbol's name. It may contain any character except
6777 @samp{\000}, so is more general than ordinary symbol names. Some
6778 debuggers used to code arbitrarily complex structures into symbol names
6782 An absolute expression. The symbol's type is set to the low 8 bits of
6783 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6784 and debuggers choke on silly bit patterns.
6787 An absolute expression. The symbol's ``other'' attribute is set to the
6788 low 8 bits of this expression.
6791 An absolute expression. The symbol's descriptor is set to the low 16
6792 bits of this expression.
6795 An absolute expression which becomes the symbol's value.
6798 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6799 or @code{.stabs} statement, the symbol has probably already been created;
6800 you get a half-formed symbol in your object file. This is
6801 compatible with earlier assemblers!
6804 @cindex @code{stabd} directive
6805 @item .stabd @var{type} , @var{other} , @var{desc}
6807 The ``name'' of the symbol generated is not even an empty string.
6808 It is a null pointer, for compatibility. Older assemblers used a
6809 null pointer so they didn't waste space in object files with empty
6812 The symbol's value is set to the location counter,
6813 relocatably. When your program is linked, the value of this symbol
6814 is the address of the location counter when the @code{.stabd} was
6817 @cindex @code{stabn} directive
6818 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6819 The name of the symbol is set to the empty string @code{""}.
6821 @cindex @code{stabs} directive
6822 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6823 All five fields are specified.
6829 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6830 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6832 @cindex string, copying to object file
6833 @cindex string8, copying to object file
6834 @cindex string16, copying to object file
6835 @cindex string32, copying to object file
6836 @cindex string64, copying to object file
6837 @cindex @code{string} directive
6838 @cindex @code{string8} directive
6839 @cindex @code{string16} directive
6840 @cindex @code{string32} directive
6841 @cindex @code{string64} directive
6843 Copy the characters in @var{str} to the object file. You may specify more than
6844 one string to copy, separated by commas. Unless otherwise specified for a
6845 particular machine, the assembler marks the end of each string with a 0 byte.
6846 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6848 The variants @code{string16}, @code{string32} and @code{string64} differ from
6849 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6850 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6851 are stored in target endianness byte order.
6857 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6858 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6863 @section @code{.struct @var{expression}}
6865 @cindex @code{struct} directive
6866 Switch to the absolute section, and set the section offset to @var{expression},
6867 which must be an absolute expression. You might use this as follows:
6876 This would define the symbol @code{field1} to have the value 0, the symbol
6877 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6878 value 8. Assembly would be left in the absolute section, and you would need to
6879 use a @code{.section} directive of some sort to change to some other section
6880 before further assembly.
6884 @section @code{.subsection @var{name}}
6886 @cindex @code{subsection} directive
6887 @cindex Section Stack
6888 This is one of the ELF section stack manipulation directives. The others are
6889 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6890 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6893 This directive replaces the current subsection with @code{name}. The current
6894 section is not changed. The replaced subsection is put onto the section stack
6895 in place of the then current top of stack subsection.
6900 @section @code{.symver}
6901 @cindex @code{symver} directive
6902 @cindex symbol versioning
6903 @cindex versions of symbols
6904 Use the @code{.symver} directive to bind symbols to specific version nodes
6905 within a source file. This is only supported on ELF platforms, and is
6906 typically used when assembling files to be linked into a shared library.
6907 There are cases where it may make sense to use this in objects to be bound
6908 into an application itself so as to override a versioned symbol from a
6911 For ELF targets, the @code{.symver} directive can be used like this:
6913 .symver @var{name}, @var{name2@@nodename}
6915 If the symbol @var{name} is defined within the file
6916 being assembled, the @code{.symver} directive effectively creates a symbol
6917 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6918 just don't try and create a regular alias is that the @var{@@} character isn't
6919 permitted in symbol names. The @var{name2} part of the name is the actual name
6920 of the symbol by which it will be externally referenced. The name @var{name}
6921 itself is merely a name of convenience that is used so that it is possible to
6922 have definitions for multiple versions of a function within a single source
6923 file, and so that the compiler can unambiguously know which version of a
6924 function is being mentioned. The @var{nodename} portion of the alias should be
6925 the name of a node specified in the version script supplied to the linker when
6926 building a shared library. If you are attempting to override a versioned
6927 symbol from a shared library, then @var{nodename} should correspond to the
6928 nodename of the symbol you are trying to override.
6930 If the symbol @var{name} is not defined within the file being assembled, all
6931 references to @var{name} will be changed to @var{name2@@nodename}. If no
6932 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6935 Another usage of the @code{.symver} directive is:
6937 .symver @var{name}, @var{name2@@@@nodename}
6939 In this case, the symbol @var{name} must exist and be defined within
6940 the file being assembled. It is similar to @var{name2@@nodename}. The
6941 difference is @var{name2@@@@nodename} will also be used to resolve
6942 references to @var{name2} by the linker.
6944 The third usage of the @code{.symver} directive is:
6946 .symver @var{name}, @var{name2@@@@@@nodename}
6948 When @var{name} is not defined within the
6949 file being assembled, it is treated as @var{name2@@nodename}. When
6950 @var{name} is defined within the file being assembled, the symbol
6951 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6956 @section @code{.tag @var{structname}}
6958 @cindex COFF structure debugging
6959 @cindex structure debugging, COFF
6960 @cindex @code{tag} directive
6961 This directive is generated by compilers to include auxiliary debugging
6962 information in the symbol table. It is only permitted inside
6963 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6964 definitions in the symbol table with instances of those structures.
6968 @section @code{.text @var{subsection}}
6970 @cindex @code{text} directive
6971 Tells @command{@value{AS}} to assemble the following statements onto the end of
6972 the text subsection numbered @var{subsection}, which is an absolute
6973 expression. If @var{subsection} is omitted, subsection number zero
6977 @section @code{.title "@var{heading}"}
6979 @cindex @code{title} directive
6980 @cindex listing control: title line
6981 Use @var{heading} as the title (second line, immediately after the
6982 source file name and pagenumber) when generating assembly listings.
6984 This directive affects subsequent pages, as well as the current page if
6985 it appears within ten lines of the top of a page.
6989 @section @code{.type}
6991 This directive is used to set the type of a symbol.
6995 @c only print the extra heading if both COFF and ELF are set
6996 @subheading COFF Version
6999 @cindex COFF symbol type
7000 @cindex symbol type, COFF
7001 @cindex @code{type} directive (COFF version)
7002 For COFF targets, this directive is permitted only within
7003 @code{.def}/@code{.endef} pairs. It is used like this:
7009 This records the integer @var{int} as the type attribute of a symbol table
7016 @c only print the extra heading if both COFF and ELF are set
7017 @subheading ELF Version
7020 @cindex ELF symbol type
7021 @cindex symbol type, ELF
7022 @cindex @code{type} directive (ELF version)
7023 For ELF targets, the @code{.type} directive is used like this:
7026 .type @var{name} , @var{type description}
7029 This sets the type of symbol @var{name} to be either a
7030 function symbol or an object symbol. There are five different syntaxes
7031 supported for the @var{type description} field, in order to provide
7032 compatibility with various other assemblers.
7034 Because some of the characters used in these syntaxes (such as @samp{@@} and
7035 @samp{#}) are comment characters for some architectures, some of the syntaxes
7036 below do not work on all architectures. The first variant will be accepted by
7037 the GNU assembler on all architectures so that variant should be used for
7038 maximum portability, if you do not need to assemble your code with other
7041 The syntaxes supported are:
7044 .type <name> STT_<TYPE_IN_UPPER_CASE>
7045 .type <name>,#<type>
7046 .type <name>,@@<type>
7047 .type <name>,%<type>
7048 .type <name>,"<type>"
7051 The types supported are:
7056 Mark the symbol as being a function name.
7059 @itemx gnu_indirect_function
7060 Mark the symbol as an indirect function when evaluated during reloc
7061 processing. (This is only supported on assemblers targeting GNU systems).
7065 Mark the symbol as being a data object.
7069 Mark the symbol as being a thread-local data object.
7073 Mark the symbol as being a common data object.
7077 Does not mark the symbol in any way. It is supported just for completeness.
7079 @item gnu_unique_object
7080 Marks the symbol as being a globally unique data object. The dynamic linker
7081 will make sure that in the entire process there is just one symbol with this
7082 name and type in use. (This is only supported on assemblers targeting GNU
7087 Note: Some targets support extra types in addition to those listed above.
7093 @section @code{.uleb128 @var{expressions}}
7095 @cindex @code{uleb128} directive
7096 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
7097 compact, variable length representation of numbers used by the DWARF
7098 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
7102 @section @code{.val @var{addr}}
7104 @cindex @code{val} directive
7105 @cindex COFF value attribute
7106 @cindex value attribute, COFF
7107 This directive, permitted only within @code{.def}/@code{.endef} pairs,
7108 records the address @var{addr} as the value attribute of a symbol table
7114 @section @code{.version "@var{string}"}
7116 @cindex @code{version} directive
7117 This directive creates a @code{.note} section and places into it an ELF
7118 formatted note of type NT_VERSION. The note's name is set to @code{string}.
7123 @section @code{.vtable_entry @var{table}, @var{offset}}
7125 @cindex @code{vtable_entry} directive
7126 This directive finds or creates a symbol @code{table} and creates a
7127 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
7130 @section @code{.vtable_inherit @var{child}, @var{parent}}
7132 @cindex @code{vtable_inherit} directive
7133 This directive finds the symbol @code{child} and finds or creates the symbol
7134 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
7135 parent whose addend is the value of the child symbol. As a special case the
7136 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
7140 @section @code{.warning "@var{string}"}
7141 @cindex warning directive
7142 Similar to the directive @code{.error}
7143 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
7146 @section @code{.weak @var{names}}
7148 @cindex @code{weak} directive
7149 This directive sets the weak attribute on the comma separated list of symbol
7150 @code{names}. If the symbols do not already exist, they will be created.
7152 On COFF targets other than PE, weak symbols are a GNU extension. This
7153 directive sets the weak attribute on the comma separated list of symbol
7154 @code{names}. If the symbols do not already exist, they will be created.
7156 On the PE target, weak symbols are supported natively as weak aliases.
7157 When a weak symbol is created that is not an alias, GAS creates an
7158 alternate symbol to hold the default value.
7161 @section @code{.weakref @var{alias}, @var{target}}
7163 @cindex @code{weakref} directive
7164 This directive creates an alias to the target symbol that enables the symbol to
7165 be referenced with weak-symbol semantics, but without actually making it weak.
7166 If direct references or definitions of the symbol are present, then the symbol
7167 will not be weak, but if all references to it are through weak references, the
7168 symbol will be marked as weak in the symbol table.
7170 The effect is equivalent to moving all references to the alias to a separate
7171 assembly source file, renaming the alias to the symbol in it, declaring the
7172 symbol as weak there, and running a reloadable link to merge the object files
7173 resulting from the assembly of the new source file and the old source file that
7174 had the references to the alias removed.
7176 The alias itself never makes to the symbol table, and is entirely handled
7177 within the assembler.
7180 @section @code{.word @var{expressions}}
7182 @cindex @code{word} directive
7183 This directive expects zero or more @var{expressions}, of any section,
7184 separated by commas.
7187 For each expression, @command{@value{AS}} emits a 32-bit number.
7190 For each expression, @command{@value{AS}} emits a 16-bit number.
7195 The size of the number emitted, and its byte order,
7196 depend on what target computer the assembly is for.
7199 @c on sparc the "special treatment to support compilers" doesn't
7200 @c happen---32-bit addressability, period; no long/short jumps.
7201 @ifset DIFF-TBL-KLUGE
7202 @cindex difference tables altered
7203 @cindex altered difference tables
7205 @emph{Warning: Special Treatment to support Compilers}
7209 Machines with a 32-bit address space, but that do less than 32-bit
7210 addressing, require the following special treatment. If the machine of
7211 interest to you does 32-bit addressing (or doesn't require it;
7212 @pxref{Machine Dependencies}), you can ignore this issue.
7215 In order to assemble compiler output into something that works,
7216 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
7217 Directives of the form @samp{.word sym1-sym2} are often emitted by
7218 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
7219 directive of the form @samp{.word sym1-sym2}, and the difference between
7220 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
7221 creates a @dfn{secondary jump table}, immediately before the next label.
7222 This secondary jump table is preceded by a short-jump to the
7223 first byte after the secondary table. This short-jump prevents the flow
7224 of control from accidentally falling into the new table. Inside the
7225 table is a long-jump to @code{sym2}. The original @samp{.word}
7226 contains @code{sym1} minus the address of the long-jump to
7229 If there were several occurrences of @samp{.word sym1-sym2} before the
7230 secondary jump table, all of them are adjusted. If there was a
7231 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
7232 long-jump to @code{sym4} is included in the secondary jump table,
7233 and the @code{.word} directives are adjusted to contain @code{sym3}
7234 minus the address of the long-jump to @code{sym4}; and so on, for as many
7235 entries in the original jump table as necessary.
7238 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
7239 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
7240 assembly language programmers.
7243 @c end DIFF-TBL-KLUGE
7245 @ifclear no-space-dir
7247 @section @code{.zero @var{size}}
7249 @cindex @code{zero} directive
7250 @cindex filling memory with zero bytes
7251 This directive emits @var{size} 0-valued bytes. @var{size} must be an absolute
7252 expression. This directive is actually an alias for the @samp{.skip} directive
7253 so in can take an optional second argument of the value to store in the bytes
7254 instead of zero. Using @samp{.zero} in this way would be confusing however.
7259 @section @code{.2byte @var{expression} [, @var{expression}]*}
7260 @cindex @code{2byte} directive
7261 @cindex two-byte integer
7262 @cindex integer, 2-byte
7264 This directive expects zero or more expressions, separated by commas. If there
7265 are no expressions then the directive does nothing. Otherwise each expression
7266 is evaluated in turn and placed in the next two bytes of the current output
7267 section, using the endian model of the target. If an expression will not fit
7268 in two bytes, a warning message is displayed and the least significant two
7269 bytes of the expression's value are used. If an expression cannot be evaluated
7270 at assembly time then relocations will be generated in order to compute the
7273 This directive does not apply any alignment before or after inserting the
7274 values. As a result of this, if relocations are generated, they may be
7275 different from those used for inserting values with a guaranteed alignment.
7277 This directive is only available for ELF targets,
7280 @section @code{.4byte @var{expression} [, @var{expression}]*}
7281 @cindex @code{4byte} directive
7282 @cindex four-byte integer
7283 @cindex integer, 4-byte
7285 Like the @option{.2byte} directive, except that it inserts unaligned, four byte
7286 long values into the output.
7289 @section @code{.8byte @var{expression} [, @var{expression}]*}
7290 @cindex @code{8byte} directive
7291 @cindex eight-byte integer
7292 @cindex integer, 8-byte
7294 Like the @option{.2byte} directive, except that it inserts unaligned, eight
7295 byte long bignum values into the output.
7300 @section Deprecated Directives
7302 @cindex deprecated directives
7303 @cindex obsolescent directives
7304 One day these directives won't work.
7305 They are included for compatibility with older assemblers.
7312 @node Object Attributes
7313 @chapter Object Attributes
7314 @cindex object attributes
7316 @command{@value{AS}} assembles source files written for a specific architecture
7317 into object files for that architecture. But not all object files are alike.
7318 Many architectures support incompatible variations. For instance, floating
7319 point arguments might be passed in floating point registers if the object file
7320 requires hardware floating point support---or floating point arguments might be
7321 passed in integer registers if the object file supports processors with no
7322 hardware floating point unit. Or, if two objects are built for different
7323 generations of the same architecture, the combination may require the
7324 newer generation at run-time.
7326 This information is useful during and after linking. At link time,
7327 @command{@value{LD}} can warn about incompatible object files. After link
7328 time, tools like @command{gdb} can use it to process the linked file
7331 Compatibility information is recorded as a series of object attributes. Each
7332 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
7333 string, and indicates who sets the meaning of the tag. The tag is an integer,
7334 and indicates what property the attribute describes. The value may be a string
7335 or an integer, and indicates how the property affects this object. Missing
7336 attributes are the same as attributes with a zero value or empty string value.
7338 Object attributes were developed as part of the ABI for the ARM Architecture.
7339 The file format is documented in @cite{ELF for the ARM Architecture}.
7342 * GNU Object Attributes:: @sc{gnu} Object Attributes
7343 * Defining New Object Attributes:: Defining New Object Attributes
7346 @node GNU Object Attributes
7347 @section @sc{gnu} Object Attributes
7349 The @code{.gnu_attribute} directive records an object attribute
7350 with vendor @samp{gnu}.
7352 Except for @samp{Tag_compatibility}, which has both an integer and a string for
7353 its value, @sc{gnu} attributes have a string value if the tag number is odd and
7354 an integer value if the tag number is even. The second bit (@code{@var{tag} &
7355 2} is set for architecture-independent attributes and clear for
7356 architecture-dependent ones.
7358 @subsection Common @sc{gnu} attributes
7360 These attributes are valid on all architectures.
7363 @item Tag_compatibility (32)
7364 The compatibility attribute takes an integer flag value and a vendor name. If
7365 the flag value is 0, the file is compatible with other toolchains. If it is 1,
7366 then the file is only compatible with the named toolchain. If it is greater
7367 than 1, the file can only be processed by other toolchains under some private
7368 arrangement indicated by the flag value and the vendor name.
7371 @subsection MIPS Attributes
7374 @item Tag_GNU_MIPS_ABI_FP (4)
7375 The floating-point ABI used by this object file. The value will be:
7379 0 for files not affected by the floating-point ABI.
7381 1 for files using the hardware floating-point ABI with a standard
7382 double-precision FPU.
7384 2 for files using the hardware floating-point ABI with a single-precision FPU.
7386 3 for files using the software floating-point ABI.
7388 4 for files using the deprecated hardware floating-point ABI which used 64-bit
7389 floating-point registers, 32-bit general-purpose registers and increased the
7390 number of callee-saved floating-point registers.
7392 5 for files using the hardware floating-point ABI with a double-precision FPU
7393 with either 32-bit or 64-bit floating-point registers and 32-bit
7394 general-purpose registers.
7396 6 for files using the hardware floating-point ABI with 64-bit floating-point
7397 registers and 32-bit general-purpose registers.
7399 7 for files using the hardware floating-point ABI with 64-bit floating-point
7400 registers, 32-bit general-purpose registers and a rule that forbids the
7401 direct use of odd-numbered single-precision floating-point registers.
7405 @subsection PowerPC Attributes
7408 @item Tag_GNU_Power_ABI_FP (4)
7409 The floating-point ABI used by this object file. The value will be:
7413 0 for files not affected by the floating-point ABI.
7415 1 for files using double-precision hardware floating-point ABI.
7417 2 for files using the software floating-point ABI.
7419 3 for files using single-precision hardware floating-point ABI.
7422 @item Tag_GNU_Power_ABI_Vector (8)
7423 The vector ABI used by this object file. The value will be:
7427 0 for files not affected by the vector ABI.
7429 1 for files using general purpose registers to pass vectors.
7431 2 for files using AltiVec registers to pass vectors.
7433 3 for files using SPE registers to pass vectors.
7437 @subsection IBM z Systems Attributes
7440 @item Tag_GNU_S390_ABI_Vector (8)
7441 The vector ABI used by this object file. The value will be:
7445 0 for files not affected by the vector ABI.
7447 1 for files using software vector ABI.
7449 2 for files using hardware vector ABI.
7453 @node Defining New Object Attributes
7454 @section Defining New Object Attributes
7456 If you want to define a new @sc{gnu} object attribute, here are the places you
7457 will need to modify. New attributes should be discussed on the @samp{binutils}
7462 This manual, which is the official register of attributes.
7464 The header for your architecture @file{include/elf}, to define the tag.
7466 The @file{bfd} support file for your architecture, to merge the attribute
7467 and issue any appropriate link warnings.
7469 Test cases in @file{ld/testsuite} for merging and link warnings.
7471 @file{binutils/readelf.c} to display your attribute.
7473 GCC, if you want the compiler to mark the attribute automatically.
7479 @node Machine Dependencies
7480 @chapter Machine Dependent Features
7482 @cindex machine dependencies
7483 The machine instruction sets are (almost by definition) different on
7484 each machine where @command{@value{AS}} runs. Floating point representations
7485 vary as well, and @command{@value{AS}} often supports a few additional
7486 directives or command-line options for compatibility with other
7487 assemblers on a particular platform. Finally, some versions of
7488 @command{@value{AS}} support special pseudo-instructions for branch
7491 This chapter discusses most of these differences, though it does not
7492 include details on any machine's instruction set. For details on that
7493 subject, see the hardware manufacturer's manual.
7497 * AArch64-Dependent:: AArch64 Dependent Features
7500 * Alpha-Dependent:: Alpha Dependent Features
7503 * ARC-Dependent:: ARC Dependent Features
7506 * ARM-Dependent:: ARM Dependent Features
7509 * AVR-Dependent:: AVR Dependent Features
7512 * Blackfin-Dependent:: Blackfin Dependent Features
7515 * CR16-Dependent:: CR16 Dependent Features
7518 * CRIS-Dependent:: CRIS Dependent Features
7521 * D10V-Dependent:: D10V Dependent Features
7524 * D30V-Dependent:: D30V Dependent Features
7527 * Epiphany-Dependent:: EPIPHANY Dependent Features
7530 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7533 * HPPA-Dependent:: HPPA Dependent Features
7536 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
7539 * IA-64-Dependent:: Intel IA-64 Dependent Features
7542 * IP2K-Dependent:: IP2K Dependent Features
7545 * LM32-Dependent:: LM32 Dependent Features
7548 * M32C-Dependent:: M32C Dependent Features
7551 * M32R-Dependent:: M32R Dependent Features
7554 * M68K-Dependent:: M680x0 Dependent Features
7557 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7560 * Meta-Dependent :: Meta Dependent Features
7563 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7566 * MIPS-Dependent:: MIPS Dependent Features
7569 * MMIX-Dependent:: MMIX Dependent Features
7572 * MSP430-Dependent:: MSP430 Dependent Features
7575 * NDS32-Dependent:: Andes NDS32 Dependent Features
7578 * NiosII-Dependent:: Altera Nios II Dependent Features
7581 * NS32K-Dependent:: NS32K Dependent Features
7584 * PDP-11-Dependent:: PDP-11 Dependent Features
7587 * PJ-Dependent:: picoJava Dependent Features
7590 * PPC-Dependent:: PowerPC Dependent Features
7593 * PRU-Dependent:: PRU Dependent Features
7596 * RISC-V-Dependent:: RISC-V Dependent Features
7599 * RL78-Dependent:: RL78 Dependent Features
7602 * RX-Dependent:: RX Dependent Features
7605 * S/390-Dependent:: IBM S/390 Dependent Features
7608 * SCORE-Dependent:: SCORE Dependent Features
7611 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7612 * SH64-Dependent:: SuperH SH64 Dependent Features
7615 * Sparc-Dependent:: SPARC Dependent Features
7618 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7621 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7624 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7627 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7630 * V850-Dependent:: V850 Dependent Features
7633 * Vax-Dependent:: VAX Dependent Features
7636 * Visium-Dependent:: Visium Dependent Features
7639 * WebAssembly-Dependent:: WebAssembly Dependent Features
7642 * XGATE-Dependent:: XGATE Dependent Features
7645 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7648 * Xtensa-Dependent:: Xtensa Dependent Features
7651 * Z80-Dependent:: Z80 Dependent Features
7654 * Z8000-Dependent:: Z8000 Dependent Features
7661 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7662 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7663 @c peculiarity: to preserve cross-references, there must be a node called
7664 @c "Machine Dependencies". Hence the conditional nodenames in each
7665 @c major node below. Node defaulting in makeinfo requires adjacency of
7666 @c node and sectioning commands; hence the repetition of @chapter BLAH
7667 @c in both conditional blocks.
7670 @include c-aarch64.texi
7674 @include c-alpha.texi
7690 @include c-bfin.texi
7694 @include c-cr16.texi
7698 @include c-cris.texi
7703 @node Machine Dependencies
7704 @chapter Machine Dependent Features
7706 The machine instruction sets are different on each Renesas chip family,
7707 and there are also some syntax differences among the families. This
7708 chapter describes the specific @command{@value{AS}} features for each
7712 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7713 * SH-Dependent:: Renesas SH Dependent Features
7720 @include c-d10v.texi
7724 @include c-d30v.texi
7728 @include c-epiphany.texi
7732 @include c-h8300.texi
7736 @include c-hppa.texi
7740 @include c-i386.texi
7744 @include c-ia64.texi
7748 @include c-ip2k.texi
7752 @include c-lm32.texi
7756 @include c-m32c.texi
7760 @include c-m32r.texi
7764 @include c-m68k.texi
7768 @include c-m68hc11.texi
7772 @include c-metag.texi
7776 @include c-microblaze.texi
7780 @include c-mips.texi
7784 @include c-mmix.texi
7788 @include c-msp430.texi
7792 @include c-nds32.texi
7796 @include c-nios2.texi
7800 @include c-ns32k.texi
7804 @include c-pdp11.texi
7820 @include c-riscv.texi
7824 @include c-rl78.texi
7832 @include c-s390.texi
7836 @include c-score.texi
7841 @include c-sh64.texi
7845 @include c-sparc.texi
7849 @include c-tic54x.texi
7853 @include c-tic6x.texi
7857 @include c-tilegx.texi
7861 @include c-tilepro.texi
7865 @include c-v850.texi
7873 @include c-visium.texi
7877 @include c-wasm32.texi
7881 @include c-xgate.texi
7885 @include c-xstormy16.texi
7889 @include c-xtensa.texi
7901 @c reverse effect of @down at top of generic Machine-Dep chapter
7905 @node Reporting Bugs
7906 @chapter Reporting Bugs
7907 @cindex bugs in assembler
7908 @cindex reporting bugs in assembler
7910 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7912 Reporting a bug may help you by bringing a solution to your problem, or it may
7913 not. But in any case the principal function of a bug report is to help the
7914 entire community by making the next version of @command{@value{AS}} work better.
7915 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7917 In order for a bug report to serve its purpose, you must include the
7918 information that enables us to fix the bug.
7921 * Bug Criteria:: Have you found a bug?
7922 * Bug Reporting:: How to report bugs
7926 @section Have You Found a Bug?
7927 @cindex bug criteria
7929 If you are not sure whether you have found a bug, here are some guidelines:
7932 @cindex fatal signal
7933 @cindex assembler crash
7934 @cindex crash of assembler
7936 If the assembler gets a fatal signal, for any input whatever, that is a
7937 @command{@value{AS}} bug. Reliable assemblers never crash.
7939 @cindex error on valid input
7941 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7943 @cindex invalid input
7945 If @command{@value{AS}} does not produce an error message for invalid input, that
7946 is a bug. However, you should note that your idea of ``invalid input'' might
7947 be our idea of ``an extension'' or ``support for traditional practice''.
7950 If you are an experienced user of assemblers, your suggestions for improvement
7951 of @command{@value{AS}} are welcome in any case.
7955 @section How to Report Bugs
7957 @cindex assembler bugs, reporting
7959 A number of companies and individuals offer support for @sc{gnu} products. If
7960 you obtained @command{@value{AS}} from a support organization, we recommend you
7961 contact that organization first.
7963 You can find contact information for many support companies and
7964 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7968 In any event, we also recommend that you send bug reports for @command{@value{AS}}
7972 The fundamental principle of reporting bugs usefully is this:
7973 @strong{report all the facts}. If you are not sure whether to state a
7974 fact or leave it out, state it!
7976 Often people omit facts because they think they know what causes the problem
7977 and assume that some details do not matter. Thus, you might assume that the
7978 name of a symbol you use in an example does not matter. Well, probably it does
7979 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
7980 happens to fetch from the location where that name is stored in memory;
7981 perhaps, if the name were different, the contents of that location would fool
7982 the assembler into doing the right thing despite the bug. Play it safe and
7983 give a specific, complete example. That is the easiest thing for you to do,
7984 and the most helpful.
7986 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
7987 it is new to us. Therefore, always write your bug reports on the assumption
7988 that the bug has not been reported previously.
7990 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7991 bell?'' This cannot help us fix a bug, so it is basically useless. We
7992 respond by asking for enough details to enable us to investigate.
7993 You might as well expedite matters by sending them to begin with.
7995 To enable us to fix the bug, you should include all these things:
7999 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
8000 it with the @samp{--version} argument.
8002 Without this, we will not know whether there is any point in looking for
8003 the bug in the current version of @command{@value{AS}}.
8006 Any patches you may have applied to the @command{@value{AS}} source.
8009 The type of machine you are using, and the operating system name and
8013 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
8017 The command arguments you gave the assembler to assemble your example and
8018 observe the bug. To guarantee you will not omit something important, list them
8019 all. A copy of the Makefile (or the output from make) is sufficient.
8021 If we were to try to guess the arguments, we would probably guess wrong
8022 and then we might not encounter the bug.
8025 A complete input file that will reproduce the bug. If the bug is observed when
8026 the assembler is invoked via a compiler, send the assembler source, not the
8027 high level language source. Most compilers will produce the assembler source
8028 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
8029 the options @samp{-v --save-temps}; this will save the assembler source in a
8030 file with an extension of @file{.s}, and also show you exactly how
8031 @command{@value{AS}} is being run.
8034 A description of what behavior you observe that you believe is
8035 incorrect. For example, ``It gets a fatal signal.''
8037 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
8038 will certainly notice it. But if the bug is incorrect output, we might not
8039 notice unless it is glaringly wrong. You might as well not give us a chance to
8042 Even if the problem you experience is a fatal signal, you should still say so
8043 explicitly. Suppose something strange is going on, such as, your copy of
8044 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
8045 library on your system. (This has happened!) Your copy might crash and ours
8046 would not. If you told us to expect a crash, then when ours fails to crash, we
8047 would know that the bug was not happening for us. If you had not told us to
8048 expect a crash, then we would not be able to draw any conclusion from our
8052 If you wish to suggest changes to the @command{@value{AS}} source, send us context
8053 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
8054 option. Always send diffs from the old file to the new file. If you even
8055 discuss something in the @command{@value{AS}} source, refer to it by context, not
8058 The line numbers in our development sources will not match those in your
8059 sources. Your line numbers would convey no useful information to us.
8062 Here are some things that are not necessary:
8066 A description of the envelope of the bug.
8068 Often people who encounter a bug spend a lot of time investigating
8069 which changes to the input file will make the bug go away and which
8070 changes will not affect it.
8072 This is often time consuming and not very useful, because the way we
8073 will find the bug is by running a single example under the debugger
8074 with breakpoints, not by pure deduction from a series of examples.
8075 We recommend that you save your time for something else.
8077 Of course, if you can find a simpler example to report @emph{instead}
8078 of the original one, that is a convenience for us. Errors in the
8079 output will be easier to spot, running under the debugger will take
8080 less time, and so on.
8082 However, simplification is not vital; if you do not want to do this,
8083 report the bug anyway and send us the entire test case you used.
8086 A patch for the bug.
8088 A patch for the bug does help us if it is a good one. But do not omit
8089 the necessary information, such as the test case, on the assumption that
8090 a patch is all we need. We might see problems with your patch and decide
8091 to fix the problem another way, or we might not understand it at all.
8093 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
8094 construct an example that will make the program follow a certain path through
8095 the code. If you do not send us the example, we will not be able to construct
8096 one, so we will not be able to verify that the bug is fixed.
8098 And if we cannot understand what bug you are trying to fix, or why your
8099 patch should be an improvement, we will not install it. A test case will
8100 help us to understand.
8103 A guess about what the bug is or what it depends on.
8105 Such guesses are usually wrong. Even we cannot guess right about such
8106 things without first using the debugger to find the facts.
8109 @node Acknowledgements
8110 @chapter Acknowledgements
8112 If you have contributed to GAS and your name isn't listed here,
8113 it is not meant as a slight. We just don't know about it. Send mail to the
8114 maintainer, and we'll correct the situation. Currently
8116 the maintainer is Nick Clifton (email address @code{nickc@@redhat.com}).
8118 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
8121 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
8122 information and the 68k series machines, most of the preprocessing pass, and
8123 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
8125 K. Richard Pixley maintained GAS for a while, adding various enhancements and
8126 many bug fixes, including merging support for several processors, breaking GAS
8127 up to handle multiple object file format back ends (including heavy rewrite,
8128 testing, an integration of the coff and b.out back ends), adding configuration
8129 including heavy testing and verification of cross assemblers and file splits
8130 and renaming, converted GAS to strictly ANSI C including full prototypes, added
8131 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
8132 port (including considerable amounts of reverse engineering), a SPARC opcode
8133 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
8134 assertions and made them work, much other reorganization, cleanup, and lint.
8136 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
8137 in format-specific I/O modules.
8139 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
8140 has done much work with it since.
8142 The Intel 80386 machine description was written by Eliot Dresselhaus.
8144 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
8146 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
8147 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
8149 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
8150 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
8151 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
8152 support a.out format.
8154 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
8155 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
8156 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
8157 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
8160 John Gilmore built the AMD 29000 support, added @code{.include} support, and
8161 simplified the configuration of which versions accept which directives. He
8162 updated the 68k machine description so that Motorola's opcodes always produced
8163 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
8164 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
8165 cross-compilation support, and one bug in relaxation that took a week and
8166 required the proverbial one-bit fix.
8168 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
8169 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
8170 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
8171 PowerPC assembler, and made a few other minor patches.
8173 Steve Chamberlain made GAS able to generate listings.
8175 Hewlett-Packard contributed support for the HP9000/300.
8177 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
8178 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
8179 formats). This work was supported by both the Center for Software Science at
8180 the University of Utah and Cygnus Support.
8182 Support for ELF format files has been worked on by Mark Eichin of Cygnus
8183 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
8184 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
8185 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
8186 and some initial 64-bit support).
8188 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
8190 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
8191 support for openVMS/Alpha.
8193 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
8196 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
8197 Inc.@: added support for Xtensa processors.
8199 Several engineers at Cygnus Support have also provided many small bug fixes and
8200 configuration enhancements.
8202 Jon Beniston added support for the Lattice Mico32 architecture.
8204 Many others have contributed large or small bugfixes and enhancements. If
8205 you have contributed significant work and are not mentioned on this list, and
8206 want to be, let us know. Some of the history has been lost; we are not
8207 intentionally leaving anyone out.
8209 @node GNU Free Documentation License
8210 @appendix GNU Free Documentation License
8214 @unnumbered AS Index
projects / deliverable / binutils-gdb.git / blob