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{--generate-missing-build-notes=[no|yes]}]
245 [@b{--target-help}] [@var{target-options}]
246 [@b{--}|@var{files} @dots{}]
249 @c Target dependent options are listed below. Keep the list sorted.
250 @c Add an empty line for separation.
254 @emph{Target AArch64 options:}
256 [@b{-mabi}=@var{ABI}]
260 @emph{Target Alpha options:}
262 [@b{-mdebug} | @b{-no-mdebug}]
263 [@b{-replace} | @b{-noreplace}]
264 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
265 [@b{-F}] [@b{-32addr}]
269 @emph{Target ARC options:}
270 [@b{-mcpu=@var{cpu}}]
271 [@b{-mA6}|@b{-mARC600}|@b{-mARC601}|@b{-mA7}|@b{-mARC700}|@b{-mEM}|@b{-mHS}]
278 @emph{Target ARM options:}
279 @c Don't document the deprecated options
280 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
281 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
282 [@b{-mfpu}=@var{floating-point-format}]
283 [@b{-mfloat-abi}=@var{abi}]
284 [@b{-meabi}=@var{ver}]
287 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
288 @b{-mapcs-reentrant}]
289 [@b{-mthumb-interwork}] [@b{-k}]
293 @emph{Target Blackfin options:}
294 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
301 @emph{Target CRIS options:}
302 [@b{--underscore} | @b{--no-underscore}]
304 [@b{--emulation=criself} | @b{--emulation=crisaout}]
305 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
306 @c Deprecated -- deliberately not documented.
311 @emph{Target D10V options:}
316 @emph{Target D30V options:}
317 [@b{-O}|@b{-n}|@b{-N}]
321 @emph{Target EPIPHANY options:}
322 [@b{-mepiphany}|@b{-mepiphany16}]
326 @emph{Target H8/300 options:}
330 @c HPPA has no machine-dependent assembler options (yet).
334 @emph{Target i386 options:}
335 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
336 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
340 @emph{Target IA-64 options:}
341 [@b{-mconstant-gp}|@b{-mauto-pic}]
342 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
344 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
345 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
346 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
347 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
351 @emph{Target IP2K options:}
352 [@b{-mip2022}|@b{-mip2022ext}]
356 @emph{Target M32C options:}
357 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
361 @emph{Target M32R options:}
362 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
367 @emph{Target M680X0 options:}
368 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
372 @emph{Target M68HC11 options:}
373 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}|@b{-mm9s12x}|@b{-mm9s12xg}]
374 [@b{-mshort}|@b{-mlong}]
375 [@b{-mshort-double}|@b{-mlong-double}]
376 [@b{--force-long-branches}] [@b{--short-branches}]
377 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
378 [@b{--print-opcodes}] [@b{--generate-example}]
382 @emph{Target MCORE options:}
383 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
384 [@b{-mcpu=[210|340]}]
388 @emph{Target Meta options:}
389 [@b{-mcpu=@var{cpu}}] [@b{-mfpu=@var{cpu}}] [@b{-mdsp=@var{cpu}}]
392 @emph{Target MICROBLAZE options:}
393 @c MicroBlaze has no machine-dependent assembler options.
397 @emph{Target MIPS options:}
398 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
399 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
400 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
401 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
402 [@b{-mfp64}] [@b{-mgp64}] [@b{-mfpxx}]
403 [@b{-modd-spreg}] [@b{-mno-odd-spreg}]
404 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
405 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
406 [@b{-mips32r3}] [@b{-mips32r5}] [@b{-mips32r6}] [@b{-mips64}] [@b{-mips64r2}]
407 [@b{-mips64r3}] [@b{-mips64r5}] [@b{-mips64r6}]
408 [@b{-construct-floats}] [@b{-no-construct-floats}]
409 [@b{-mignore-branch-isa}] [@b{-mno-ignore-branch-isa}]
410 [@b{-mnan=@var{encoding}}]
411 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
412 [@b{-mips16}] [@b{-no-mips16}]
413 [@b{-mmips16e2}] [@b{-mno-mips16e2}]
414 [@b{-mmicromips}] [@b{-mno-micromips}]
415 [@b{-msmartmips}] [@b{-mno-smartmips}]
416 [@b{-mips3d}] [@b{-no-mips3d}]
417 [@b{-mdmx}] [@b{-no-mdmx}]
418 [@b{-mdsp}] [@b{-mno-dsp}]
419 [@b{-mdspr2}] [@b{-mno-dspr2}]
420 [@b{-mdspr3}] [@b{-mno-dspr3}]
421 [@b{-mmsa}] [@b{-mno-msa}]
422 [@b{-mxpa}] [@b{-mno-xpa}]
423 [@b{-mmt}] [@b{-mno-mt}]
424 [@b{-mmcu}] [@b{-mno-mcu}]
425 [@b{-mcrc}] [@b{-mno-crc}]
426 [@b{-mginv}] [@b{-mno-ginv}]
427 [@b{-mloongson-mmi}] [@b{-mno-loongson-mmi}]
428 [@b{-minsn32}] [@b{-mno-insn32}]
429 [@b{-mfix7000}] [@b{-mno-fix7000}]
430 [@b{-mfix-rm7000}] [@b{-mno-fix-rm7000}]
431 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
432 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
433 [@b{-mdebug}] [@b{-no-mdebug}]
434 [@b{-mpdr}] [@b{-mno-pdr}]
438 @emph{Target MMIX options:}
439 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
440 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
441 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
442 [@b{--linker-allocated-gregs}]
446 @emph{Target Nios II options:}
447 [@b{-relax-all}] [@b{-relax-section}] [@b{-no-relax}]
452 @emph{Target NDS32 options:}
453 [@b{-EL}] [@b{-EB}] [@b{-O}] [@b{-Os}] [@b{-mcpu=@var{cpu}}]
454 [@b{-misa=@var{isa}}] [@b{-mabi=@var{abi}}] [@b{-mall-ext}]
455 [@b{-m[no-]16-bit}] [@b{-m[no-]perf-ext}] [@b{-m[no-]perf2-ext}]
456 [@b{-m[no-]string-ext}] [@b{-m[no-]dsp-ext}] [@b{-m[no-]mac}] [@b{-m[no-]div}]
457 [@b{-m[no-]audio-isa-ext}] [@b{-m[no-]fpu-sp-ext}] [@b{-m[no-]fpu-dp-ext}]
458 [@b{-m[no-]fpu-fma}] [@b{-mfpu-freg=@var{FREG}}] [@b{-mreduced-regs}]
459 [@b{-mfull-regs}] [@b{-m[no-]dx-regs}] [@b{-mpic}] [@b{-mno-relax}]
464 @emph{Target PDP11 options:}
465 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
466 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
467 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
471 @emph{Target picoJava options:}
476 @emph{Target PowerPC options:}
478 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
479 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
480 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
481 @b{-mbooke}|@b{-mpower4}|@b{-mpwr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
482 @b{-mpower7}|@b{-mpwr7}|@b{-mpower8}|@b{-mpwr8}|@b{-mpower9}|@b{-mpwr9}@b{-ma2}|
483 @b{-mcell}|@b{-mspe}|@b{-mspe2}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
484 [@b{-many}] [@b{-maltivec}|@b{-mvsx}|@b{-mhtm}|@b{-mvle}]
485 [@b{-mregnames}|@b{-mno-regnames}]
486 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
487 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
488 [@b{-msolaris}|@b{-mno-solaris}]
489 [@b{-nops=@var{count}}]
493 @emph{Target PRU options:}
496 [@b{-mno-warn-regname-label}]
500 @emph{Target RISC-V options:}
501 [@b{-fpic}|@b{-fPIC}|@b{-fno-pic}]
502 [@b{-march}=@var{ISA}]
503 [@b{-mabi}=@var{ABI}]
507 @emph{Target RL78 options:}
509 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
513 @emph{Target RX options:}
514 [@b{-mlittle-endian}|@b{-mbig-endian}]
515 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
516 [@b{-muse-conventional-section-names}]
517 [@b{-msmall-data-limit}]
520 [@b{-mint-register=@var{number}}]
521 [@b{-mgcc-abi}|@b{-mrx-abi}]
525 @emph{Target s390 options:}
526 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
527 [@b{-mregnames}|@b{-mno-regnames}]
528 [@b{-mwarn-areg-zero}]
532 @emph{Target SCORE options:}
533 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
534 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
535 [@b{-march=score7}][@b{-march=score3}]
536 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
540 @emph{Target SPARC options:}
541 @c The order here is important. See c-sparc.texi.
542 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Aleon}|@b{-Asparclet}|@b{-Asparclite}
543 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av8plusb}|@b{-Av8plusc}|@b{-Av8plusd}
544 @b{-Av8plusv}|@b{-Av8plusm}|@b{-Av9}|@b{-Av9a}|@b{-Av9b}|@b{-Av9c}
545 @b{-Av9d}|@b{-Av9e}|@b{-Av9v}|@b{-Av9m}|@b{-Asparc}|@b{-Asparcvis}
546 @b{-Asparcvis2}|@b{-Asparcfmaf}|@b{-Asparcima}|@b{-Asparcvis3}
547 @b{-Asparcvisr}|@b{-Asparc5}]
548 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}]|@b{-xarch=v8plusb}|@b{-xarch=v8plusc}
549 @b{-xarch=v8plusd}|@b{-xarch=v8plusv}|@b{-xarch=v8plusm}|@b{-xarch=v9}
550 @b{-xarch=v9a}|@b{-xarch=v9b}|@b{-xarch=v9c}|@b{-xarch=v9d}|@b{-xarch=v9e}
551 @b{-xarch=v9v}|@b{-xarch=v9m}|@b{-xarch=sparc}|@b{-xarch=sparcvis}
552 @b{-xarch=sparcvis2}|@b{-xarch=sparcfmaf}|@b{-xarch=sparcima}
553 @b{-xarch=sparcvis3}|@b{-xarch=sparcvisr}|@b{-xarch=sparc5}
556 [@b{--enforce-aligned-data}][@b{--dcti-couples-detect}]
560 @emph{Target TIC54X options:}
561 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
562 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
566 @emph{Target TIC6X options:}
567 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
568 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
569 [@b{-mpic}|@b{-mno-pic}]
573 @emph{Target TILE-Gx options:}
574 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
577 @c TILEPro has no machine-dependent assembler options
581 @emph{Target Visium options:}
582 [@b{-mtune=@var{arch}}]
586 @emph{Target Xtensa options:}
587 [@b{--[no-]text-section-literals}] [@b{--[no-]auto-litpools}]
588 [@b{--[no-]absolute-literals}]
589 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
590 [@b{--[no-]transform}]
591 [@b{--rename-section} @var{oldname}=@var{newname}]
592 [@b{--[no-]trampolines}]
596 @emph{Target Z80 options:}
597 [@b{-z80}] [@b{-r800}]
598 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
599 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
600 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
601 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
602 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
603 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
607 @c Z8000 has no machine-dependent assembler options
616 @include at-file.texi
619 Turn on listings, in any of a variety of ways:
623 omit false conditionals
626 omit debugging directives
629 include general information, like @value{AS} version and options passed
632 include high-level source
638 include macro expansions
641 omit forms processing
647 set the name of the listing file
650 You may combine these options; for example, use @samp{-aln} for assembly
651 listing without forms processing. The @samp{=file} option, if used, must be
652 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
655 Begin in alternate macro mode.
657 @xref{Altmacro,,@code{.altmacro}}.
660 @item --compress-debug-sections
661 Compress DWARF debug sections using zlib with SHF_COMPRESSED from the
662 ELF ABI. The resulting object file may not be compatible with older
663 linkers and object file utilities. Note if compression would make a
664 given section @emph{larger} then it is not compressed.
667 @cindex @samp{--compress-debug-sections=} option
668 @item --compress-debug-sections=none
669 @itemx --compress-debug-sections=zlib
670 @itemx --compress-debug-sections=zlib-gnu
671 @itemx --compress-debug-sections=zlib-gabi
672 These options control how DWARF debug sections are compressed.
673 @option{--compress-debug-sections=none} is equivalent to
674 @option{--nocompress-debug-sections}.
675 @option{--compress-debug-sections=zlib} and
676 @option{--compress-debug-sections=zlib-gabi} are equivalent to
677 @option{--compress-debug-sections}.
678 @option{--compress-debug-sections=zlib-gnu} compresses DWARF debug
679 sections using zlib. The debug sections are renamed to begin with
680 @samp{.zdebug}. Note if compression would make a given section
681 @emph{larger} then it is not compressed nor renamed.
685 @item --nocompress-debug-sections
686 Do not compress DWARF debug sections. This is usually the default for all
687 targets except the x86/x86_64, but a configure time option can be used to
691 Ignored. This option is accepted for script compatibility with calls to
694 @item --debug-prefix-map @var{old}=@var{new}
695 When assembling files in directory @file{@var{old}}, record debugging
696 information describing them as in @file{@var{new}} instead.
698 @item --defsym @var{sym}=@var{value}
699 Define the symbol @var{sym} to be @var{value} before assembling the input file.
700 @var{value} must be an integer constant. As in C, a leading @samp{0x}
701 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
702 value. The value of the symbol can be overridden inside a source file via the
703 use of a @code{.set} pseudo-op.
706 ``fast''---skip whitespace and comment preprocessing (assume source is
711 Generate debugging information for each assembler source line using whichever
712 debug format is preferred by the target. This currently means either STABS,
716 Generate stabs debugging information for each assembler line. This
717 may help debugging assembler code, if the debugger can handle it.
720 Generate stabs debugging information for each assembler line, with GNU
721 extensions that probably only gdb can handle, and that could make other
722 debuggers crash or refuse to read your program. This
723 may help debugging assembler code. Currently the only GNU extension is
724 the location of the current working directory at assembling time.
727 Generate DWARF2 debugging information for each assembler line. This
728 may help debugging assembler code, if the debugger can handle it. Note---this
729 option is only supported by some targets, not all of them.
731 @item --gdwarf-sections
732 Instead of creating a .debug_line section, create a series of
733 .debug_line.@var{foo} sections where @var{foo} is the name of the
734 corresponding code section. For example a code section called @var{.text.func}
735 will have its dwarf line number information placed into a section called
736 @var{.debug_line.text.func}. If the code section is just called @var{.text}
737 then debug line section will still be called just @var{.debug_line} without any
741 @item --size-check=error
742 @itemx --size-check=warning
743 Issue an error or warning for invalid ELF .size directive.
745 @item --elf-stt-common=no
746 @itemx --elf-stt-common=yes
747 These options control whether the ELF assembler should generate common
748 symbols with the @code{STT_COMMON} type. The default can be controlled
749 by a configure option @option{--enable-elf-stt-common}.
751 @item --generate-missing-build-notes=yes
752 @itemx --generate-missing-build-notes=no
753 These options control whether the ELF assembler should generate GNU Build
754 attribute notes if none are present in the input sources.
755 The default can be controlled by the @option{--enable-generate-build-notes}
761 Print a summary of the command-line options and exit.
764 Print a summary of all target specific options and exit.
767 Add directory @var{dir} to the search list for @code{.include} directives.
770 Don't warn about signed overflow.
773 @ifclear DIFF-TBL-KLUGE
774 This option is accepted but has no effect on the @value{TARGET} family.
776 @ifset DIFF-TBL-KLUGE
777 Issue warnings when difference tables altered for long displacements.
782 Keep (in the symbol table) local symbols. These symbols start with
783 system-specific local label prefixes, typically @samp{.L} for ELF systems
784 or @samp{L} for traditional a.out systems.
789 @item --listing-lhs-width=@var{number}
790 Set the maximum width, in words, of the output data column for an assembler
791 listing to @var{number}.
793 @item --listing-lhs-width2=@var{number}
794 Set the maximum width, in words, of the output data column for continuation
795 lines in an assembler listing to @var{number}.
797 @item --listing-rhs-width=@var{number}
798 Set the maximum width of an input source line, as displayed in a listing, to
801 @item --listing-cont-lines=@var{number}
802 Set the maximum number of lines printed in a listing for a single line of input
805 @item --no-pad-sections
806 Stop the assembler for padding the ends of output sections to the alignment
807 of that section. The default is to pad the sections, but this can waste space
808 which might be needed on targets which have tight memory constraints.
810 @item -o @var{objfile}
811 Name the object-file output from @command{@value{AS}} @var{objfile}.
814 Fold the data section into the text section.
816 @item --hash-size=@var{number}
817 Set the default size of GAS's hash tables to a prime number close to
818 @var{number}. Increasing this value can reduce the length of time it takes the
819 assembler to perform its tasks, at the expense of increasing the assembler's
820 memory requirements. Similarly reducing this value can reduce the memory
821 requirements at the expense of speed.
823 @item --reduce-memory-overheads
824 This option reduces GAS's memory requirements, at the expense of making the
825 assembly processes slower. Currently this switch is a synonym for
826 @samp{--hash-size=4051}, but in the future it may have other effects as well.
829 @item --sectname-subst
830 Honor substitution sequences in section names.
832 @xref{Section Name Substitutions,,@code{.section @var{name}}}.
837 Print the maximum space (in bytes) and total time (in seconds) used by
840 @item --strip-local-absolute
841 Remove local absolute symbols from the outgoing symbol table.
845 Print the @command{as} version.
848 Print the @command{as} version and exit.
852 Suppress warning messages.
854 @item --fatal-warnings
855 Treat warnings as errors.
858 Don't suppress warning messages or treat them as errors.
867 Generate an object file even after errors.
869 @item -- | @var{files} @dots{}
870 Standard input, or source files to assemble.
878 @xref{AArch64 Options}, for the options available when @value{AS} is configured
879 for the 64-bit mode of the ARM Architecture (AArch64).
884 The following options are available when @value{AS} is configured for the
885 64-bit mode of the ARM Architecture (AArch64).
888 @include c-aarch64.texi
889 @c ended inside the included file
897 @xref{Alpha Options}, for the options available when @value{AS} is configured
898 for an Alpha processor.
903 The following options are available when @value{AS} is configured for an Alpha
907 @include c-alpha.texi
908 @c ended inside the included file
915 The following options are available when @value{AS} is configured for an ARC
919 @item -mcpu=@var{cpu}
920 This option selects the core processor variant.
922 Select either big-endian (-EB) or little-endian (-EL) output.
924 Enable Code Density extenssion instructions.
929 The following options are available when @value{AS} is configured for the ARM
933 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
934 Specify which ARM processor variant is the target.
935 @item -march=@var{architecture}[+@var{extension}@dots{}]
936 Specify which ARM architecture variant is used by the target.
937 @item -mfpu=@var{floating-point-format}
938 Select which Floating Point architecture is the target.
939 @item -mfloat-abi=@var{abi}
940 Select which floating point ABI is in use.
942 Enable Thumb only instruction decoding.
943 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
944 Select which procedure calling convention is in use.
946 Select either big-endian (-EB) or little-endian (-EL) output.
947 @item -mthumb-interwork
948 Specify that the code has been generated with interworking between Thumb and
951 Turns on CodeComposer Studio assembly syntax compatibility mode.
953 Specify that PIC code has been generated.
961 @xref{Blackfin Options}, for the options available when @value{AS} is
962 configured for the Blackfin processor family.
967 The following options are available when @value{AS} is configured for
968 the Blackfin processor family.
972 @c ended inside the included file
979 See the info pages for documentation of the CRIS-specific options.
983 The following options are available when @value{AS} is configured for
986 @cindex D10V optimization
987 @cindex optimization, D10V
989 Optimize output by parallelizing instructions.
994 The following options are available when @value{AS} is configured for a D30V
997 @cindex D30V optimization
998 @cindex optimization, D30V
1000 Optimize output by parallelizing instructions.
1004 Warn when nops are generated.
1006 @cindex D30V nops after 32-bit multiply
1008 Warn when a nop after a 32-bit multiply instruction is generated.
1014 The following options are available when @value{AS} is configured for the
1015 Adapteva EPIPHANY series.
1018 @xref{Epiphany Options}, for the options available when @value{AS} is
1019 configured for an Epiphany processor.
1023 @c man begin OPTIONS
1024 The following options are available when @value{AS} is configured for
1025 an Epiphany processor.
1027 @c man begin INCLUDE
1028 @include c-epiphany.texi
1029 @c ended inside the included file
1037 @xref{H8/300 Options}, for the options available when @value{AS} is configured
1038 for an H8/300 processor.
1042 @c man begin OPTIONS
1043 The following options are available when @value{AS} is configured for an H8/300
1046 @c man begin INCLUDE
1047 @include c-h8300.texi
1048 @c ended inside the included file
1056 @xref{i386-Options}, for the options available when @value{AS} is
1057 configured for an i386 processor.
1061 @c man begin OPTIONS
1062 The following options are available when @value{AS} is configured for
1065 @c man begin INCLUDE
1066 @include c-i386.texi
1067 @c ended inside the included file
1072 @c man begin OPTIONS
1074 The following options are available when @value{AS} is configured for the
1080 Specifies that the extended IP2022 instructions are allowed.
1083 Restores the default behaviour, which restricts the permitted instructions to
1084 just the basic IP2022 ones.
1090 The following options are available when @value{AS} is configured for the
1091 Renesas M32C and M16C processors.
1096 Assemble M32C instructions.
1099 Assemble M16C instructions (the default).
1102 Enable support for link-time relaxations.
1105 Support H'00 style hex constants in addition to 0x00 style.
1111 The following options are available when @value{AS} is configured for the
1112 Renesas M32R (formerly Mitsubishi M32R) series.
1117 Specify which processor in the M32R family is the target. The default
1118 is normally the M32R, but this option changes it to the M32RX.
1120 @item --warn-explicit-parallel-conflicts or --Wp
1121 Produce warning messages when questionable parallel constructs are
1124 @item --no-warn-explicit-parallel-conflicts or --Wnp
1125 Do not produce warning messages when questionable parallel constructs are
1132 The following options are available when @value{AS} is configured for the
1133 Motorola 68000 series.
1138 Shorten references to undefined symbols, to one word instead of two.
1140 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1141 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1142 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1143 Specify what processor in the 68000 family is the target. The default
1144 is normally the 68020, but this can be changed at configuration time.
1146 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1147 The target machine does (or does not) have a floating-point coprocessor.
1148 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1149 the basic 68000 is not compatible with the 68881, a combination of the
1150 two can be specified, since it's possible to do emulation of the
1151 coprocessor instructions with the main processor.
1153 @item -m68851 | -mno-68851
1154 The target machine does (or does not) have a memory-management
1155 unit coprocessor. The default is to assume an MMU for 68020 and up.
1163 @xref{Nios II Options}, for the options available when @value{AS} is configured
1164 for an Altera Nios II processor.
1168 @c man begin OPTIONS
1169 The following options are available when @value{AS} is configured for an
1170 Altera Nios II processor.
1172 @c man begin INCLUDE
1173 @include c-nios2.texi
1174 @c ended inside the included file
1180 For details about the PDP-11 machine dependent features options,
1181 see @ref{PDP-11-Options}.
1184 @item -mpic | -mno-pic
1185 Generate position-independent (or position-dependent) code. The
1186 default is @option{-mpic}.
1189 @itemx -mall-extensions
1190 Enable all instruction set extensions. This is the default.
1192 @item -mno-extensions
1193 Disable all instruction set extensions.
1195 @item -m@var{extension} | -mno-@var{extension}
1196 Enable (or disable) a particular instruction set extension.
1199 Enable the instruction set extensions supported by a particular CPU, and
1200 disable all other extensions.
1202 @item -m@var{machine}
1203 Enable the instruction set extensions supported by a particular machine
1204 model, and disable all other extensions.
1210 The following options are available when @value{AS} is configured for
1211 a picoJava processor.
1215 @cindex PJ endianness
1216 @cindex endianness, PJ
1217 @cindex big endian output, PJ
1219 Generate ``big endian'' format output.
1221 @cindex little endian output, PJ
1223 Generate ``little endian'' format output.
1231 @xref{PRU Options}, for the options available when @value{AS} is configured
1232 for a PRU processor.
1236 @c man begin OPTIONS
1237 The following options are available when @value{AS} is configured for a
1240 @c man begin INCLUDE
1242 @c ended inside the included file
1247 The following options are available when @value{AS} is configured for the
1248 Motorola 68HC11 or 68HC12 series.
1252 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1253 Specify what processor is the target. The default is
1254 defined by the configuration option when building the assembler.
1256 @item --xgate-ramoffset
1257 Instruct the linker to offset RAM addresses from S12X address space into
1258 XGATE address space.
1261 Specify to use the 16-bit integer ABI.
1264 Specify to use the 32-bit integer ABI.
1266 @item -mshort-double
1267 Specify to use the 32-bit double ABI.
1270 Specify to use the 64-bit double ABI.
1272 @item --force-long-branches
1273 Relative branches are turned into absolute ones. This concerns
1274 conditional branches, unconditional branches and branches to a
1277 @item -S | --short-branches
1278 Do not turn relative branches into absolute ones
1279 when the offset is out of range.
1281 @item --strict-direct-mode
1282 Do not turn the direct addressing mode into extended addressing mode
1283 when the instruction does not support direct addressing mode.
1285 @item --print-insn-syntax
1286 Print the syntax of instruction in case of error.
1288 @item --print-opcodes
1289 Print the list of instructions with syntax and then exit.
1291 @item --generate-example
1292 Print an example of instruction for each possible instruction and then exit.
1293 This option is only useful for testing @command{@value{AS}}.
1299 The following options are available when @command{@value{AS}} is configured
1300 for the SPARC architecture:
1303 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1304 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1305 Explicitly select a variant of the SPARC architecture.
1307 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1308 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1310 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1311 UltraSPARC extensions.
1313 @item -xarch=v8plus | -xarch=v8plusa
1314 For compatibility with the Solaris v9 assembler. These options are
1315 equivalent to -Av8plus and -Av8plusa, respectively.
1318 Warn when the assembler switches to another architecture.
1323 The following options are available when @value{AS} is configured for the 'c54x
1328 Enable extended addressing mode. All addresses and relocations will assume
1329 extended addressing (usually 23 bits).
1330 @item -mcpu=@var{CPU_VERSION}
1331 Sets the CPU version being compiled for.
1332 @item -merrors-to-file @var{FILENAME}
1333 Redirect error output to a file, for broken systems which don't support such
1334 behaviour in the shell.
1339 @c man begin OPTIONS
1340 The following options are available when @value{AS} is configured for
1345 This option sets the largest size of an object that can be referenced
1346 implicitly with the @code{gp} register. It is only accepted for targets that
1347 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1349 @cindex MIPS endianness
1350 @cindex endianness, MIPS
1351 @cindex big endian output, MIPS
1353 Generate ``big endian'' format output.
1355 @cindex little endian output, MIPS
1357 Generate ``little endian'' format output.
1375 Generate code for a particular MIPS Instruction Set Architecture level.
1376 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1377 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1378 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1379 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips32r3},
1380 @samp{-mips32r5}, @samp{-mips32r6}, @samp{-mips64}, @samp{-mips64r2},
1381 @samp{-mips64r3}, @samp{-mips64r5}, and @samp{-mips64r6} correspond to generic
1382 MIPS V, MIPS32, MIPS32 Release 2, MIPS32 Release 3, MIPS32 Release 5, MIPS32
1383 Release 6, MIPS64, MIPS64 Release 2, MIPS64 Release 3, MIPS64 Release 5, and
1384 MIPS64 Release 6 ISA processors, respectively.
1386 @item -march=@var{cpu}
1387 Generate code for a particular MIPS CPU.
1389 @item -mtune=@var{cpu}
1390 Schedule and tune for a particular MIPS CPU.
1394 Cause nops to be inserted if the read of the destination register
1395 of an mfhi or mflo instruction occurs in the following two instructions.
1398 @itemx -mno-fix-rm7000
1399 Cause nops to be inserted if a dmult or dmultu instruction is
1400 followed by a load instruction.
1404 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1405 section instead of the standard ELF .stabs sections.
1409 Control generation of @code{.pdr} sections.
1413 The register sizes are normally inferred from the ISA and ABI, but these
1414 flags force a certain group of registers to be treated as 32 bits wide at
1415 all times. @samp{-mgp32} controls the size of general-purpose registers
1416 and @samp{-mfp32} controls the size of floating-point registers.
1420 The register sizes are normally inferred from the ISA and ABI, but these
1421 flags force a certain group of registers to be treated as 64 bits wide at
1422 all times. @samp{-mgp64} controls the size of general-purpose registers
1423 and @samp{-mfp64} controls the size of floating-point registers.
1426 The register sizes are normally inferred from the ISA and ABI, but using
1427 this flag in combination with @samp{-mabi=32} enables an ABI variant
1428 which will operate correctly with floating-point registers which are
1432 @itemx -mno-odd-spreg
1433 Enable use of floating-point operations on odd-numbered single-precision
1434 registers when supported by the ISA. @samp{-mfpxx} implies
1435 @samp{-mno-odd-spreg}, otherwise the default is @samp{-modd-spreg}.
1439 Generate code for the MIPS 16 processor. This is equivalent to putting
1440 @code{.module mips16} at the start of the assembly file. @samp{-no-mips16}
1441 turns off this option.
1444 @itemx -mno-mips16e2
1445 Enable the use of MIPS16e2 instructions in MIPS16 mode. This is equivalent
1446 to putting @code{.module mips16e2} at the start of the assembly file.
1447 @samp{-mno-mips16e2} turns off this option.
1450 @itemx -mno-micromips
1451 Generate code for the microMIPS processor. This is equivalent to putting
1452 @code{.module micromips} at the start of the assembly file.
1453 @samp{-mno-micromips} turns off this option. This is equivalent to putting
1454 @code{.module nomicromips} at the start of the assembly file.
1457 @itemx -mno-smartmips
1458 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1459 equivalent to putting @code{.module smartmips} at the start of the assembly
1460 file. @samp{-mno-smartmips} turns off this option.
1464 Generate code for the MIPS-3D Application Specific Extension.
1465 This tells the assembler to accept MIPS-3D instructions.
1466 @samp{-no-mips3d} turns off this option.
1470 Generate code for the MDMX Application Specific Extension.
1471 This tells the assembler to accept MDMX instructions.
1472 @samp{-no-mdmx} turns off this option.
1476 Generate code for the DSP Release 1 Application Specific Extension.
1477 This tells the assembler to accept DSP Release 1 instructions.
1478 @samp{-mno-dsp} turns off this option.
1482 Generate code for the DSP Release 2 Application Specific Extension.
1483 This option implies @samp{-mdsp}.
1484 This tells the assembler to accept DSP Release 2 instructions.
1485 @samp{-mno-dspr2} turns off this option.
1489 Generate code for the DSP Release 3 Application Specific Extension.
1490 This option implies @samp{-mdsp} and @samp{-mdspr2}.
1491 This tells the assembler to accept DSP Release 3 instructions.
1492 @samp{-mno-dspr3} turns off this option.
1496 Generate code for the MIPS SIMD Architecture Extension.
1497 This tells the assembler to accept MSA instructions.
1498 @samp{-mno-msa} turns off this option.
1502 Generate code for the MIPS eXtended Physical Address (XPA) Extension.
1503 This tells the assembler to accept XPA instructions.
1504 @samp{-mno-xpa} turns off this option.
1508 Generate code for the MT Application Specific Extension.
1509 This tells the assembler to accept MT instructions.
1510 @samp{-mno-mt} turns off this option.
1514 Generate code for the MCU Application Specific Extension.
1515 This tells the assembler to accept MCU instructions.
1516 @samp{-mno-mcu} turns off this option.
1520 Generate code for the MIPS cyclic redundancy check (CRC) Application
1521 Specific Extension. This tells the assembler to accept CRC instructions.
1522 @samp{-mno-crc} turns off this option.
1526 Generate code for the Global INValidate (GINV) Application Specific
1527 Extension. This tells the assembler to accept GINV instructions.
1528 @samp{-mno-ginv} turns off this option.
1530 @item -mloongson-mmi
1531 @itemx -mno-loongson-mmi
1532 Generate code for the Loongson MultiMedia extensions Instructions (MMI)
1533 Application Specific Extension. This tells the assembler to accept MMI
1535 @samp{-mno-loongson-mmi} turns off this option.
1539 Only use 32-bit instruction encodings when generating code for the
1540 microMIPS processor. This option inhibits the use of any 16-bit
1541 instructions. This is equivalent to putting @code{.set insn32} at
1542 the start of the assembly file. @samp{-mno-insn32} turns off this
1543 option. This is equivalent to putting @code{.set noinsn32} at the
1544 start of the assembly file. By default @samp{-mno-insn32} is
1545 selected, allowing all instructions to be used.
1547 @item --construct-floats
1548 @itemx --no-construct-floats
1549 The @samp{--no-construct-floats} option disables the construction of
1550 double width floating point constants by loading the two halves of the
1551 value into the two single width floating point registers that make up
1552 the double width register. By default @samp{--construct-floats} is
1553 selected, allowing construction of these floating point constants.
1555 @item --relax-branch
1556 @itemx --no-relax-branch
1557 The @samp{--relax-branch} option enables the relaxation of out-of-range
1558 branches. By default @samp{--no-relax-branch} is selected, causing any
1559 out-of-range branches to produce an error.
1561 @item -mignore-branch-isa
1562 @itemx -mno-ignore-branch-isa
1563 Ignore branch checks for invalid transitions between ISA modes. The
1564 semantics of branches does not provide for an ISA mode switch, so in
1565 most cases the ISA mode a branch has been encoded for has to be the
1566 same as the ISA mode of the branch's target label. Therefore GAS has
1567 checks implemented that verify in branch assembly that the two ISA
1568 modes match. @samp{-mignore-branch-isa} disables these checks. By
1569 default @samp{-mno-ignore-branch-isa} is selected, causing any invalid
1570 branch requiring a transition between ISA modes to produce an error.
1572 @item -mnan=@var{encoding}
1573 Select between the IEEE 754-2008 (@option{-mnan=2008}) or the legacy
1574 (@option{-mnan=legacy}) NaN encoding format. The latter is the default.
1577 @item --emulation=@var{name}
1578 This option was formerly used to switch between ELF and ECOFF output
1579 on targets like IRIX 5 that supported both. MIPS ECOFF support was
1580 removed in GAS 2.24, so the option now serves little purpose.
1581 It is retained for backwards compatibility.
1583 The available configuration names are: @samp{mipself}, @samp{mipslelf} and
1584 @samp{mipsbelf}. Choosing @samp{mipself} now has no effect, since the output
1585 is always ELF. @samp{mipslelf} and @samp{mipsbelf} select little- and
1586 big-endian output respectively, but @samp{-EL} and @samp{-EB} are now the
1587 preferred options instead.
1590 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1597 Control how to deal with multiplication overflow and division by zero.
1598 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1599 (and only work for Instruction Set Architecture level 2 and higher);
1600 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1604 When this option is used, @command{@value{AS}} will issue a warning every
1605 time it generates a nop instruction from a macro.
1611 The following options are available when @value{AS} is configured for
1617 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1618 The command-line option @samp{-nojsri2bsr} can be used to disable it.
1622 Enable or disable the silicon filter behaviour. By default this is disabled.
1623 The default can be overridden by the @samp{-sifilter} command-line option.
1626 Alter jump instructions for long displacements.
1628 @item -mcpu=[210|340]
1629 Select the cpu type on the target hardware. This controls which instructions
1633 Assemble for a big endian target.
1636 Assemble for a little endian target.
1645 @xref{Meta Options}, for the options available when @value{AS} is configured
1646 for a Meta processor.
1650 @c man begin OPTIONS
1651 The following options are available when @value{AS} is configured for a
1654 @c man begin INCLUDE
1655 @include c-metag.texi
1656 @c ended inside the included file
1661 @c man begin OPTIONS
1663 See the info pages for documentation of the MMIX-specific options.
1669 @xref{NDS32 Options}, for the options available when @value{AS} is configured
1670 for a NDS32 processor.
1672 @c ended inside the included file
1676 @c man begin OPTIONS
1677 The following options are available when @value{AS} is configured for a
1680 @c man begin INCLUDE
1681 @include c-nds32.texi
1682 @c ended inside the included file
1689 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1690 for a PowerPC processor.
1694 @c man begin OPTIONS
1695 The following options are available when @value{AS} is configured for a
1698 @c man begin INCLUDE
1700 @c ended inside the included file
1708 @xref{RISC-V-Options}, for the options available when @value{AS} is configured
1709 for a RISC-V processor.
1713 @c man begin OPTIONS
1714 The following options are available when @value{AS} is configured for a
1717 @c man begin INCLUDE
1718 @include c-riscv.texi
1719 @c ended inside the included file
1724 @c man begin OPTIONS
1726 See the info pages for documentation of the RX-specific options.
1730 The following options are available when @value{AS} is configured for the s390
1736 Select the word size, either 31/32 bits or 64 bits.
1739 Select the architecture mode, either the Enterprise System
1740 Architecture (esa) or the z/Architecture mode (zarch).
1741 @item -march=@var{processor}
1742 Specify which s390 processor variant is the target, @samp{g5} (or
1743 @samp{arch3}), @samp{g6}, @samp{z900} (or @samp{arch5}), @samp{z990} (or
1744 @samp{arch6}), @samp{z9-109}, @samp{z9-ec} (or @samp{arch7}), @samp{z10} (or
1745 @samp{arch8}), @samp{z196} (or @samp{arch9}), @samp{zEC12} (or @samp{arch10}),
1746 @samp{z13} (or @samp{arch11}), or @samp{z14} (or @samp{arch12}).
1748 @itemx -mno-regnames
1749 Allow or disallow symbolic names for registers.
1750 @item -mwarn-areg-zero
1751 Warn whenever the operand for a base or index register has been specified
1752 but evaluates to zero.
1760 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1761 for a TMS320C6000 processor.
1765 @c man begin OPTIONS
1766 The following options are available when @value{AS} is configured for a
1767 TMS320C6000 processor.
1769 @c man begin INCLUDE
1770 @include c-tic6x.texi
1771 @c ended inside the included file
1779 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1780 for a TILE-Gx processor.
1784 @c man begin OPTIONS
1785 The following options are available when @value{AS} is configured for a TILE-Gx
1788 @c man begin INCLUDE
1789 @include c-tilegx.texi
1790 @c ended inside the included file
1798 @xref{Visium Options}, for the options available when @value{AS} is configured
1799 for a Visium processor.
1803 @c man begin OPTIONS
1804 The following option is available when @value{AS} is configured for a Visium
1807 @c man begin INCLUDE
1808 @include c-visium.texi
1809 @c ended inside the included file
1817 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1818 for an Xtensa processor.
1822 @c man begin OPTIONS
1823 The following options are available when @value{AS} is configured for an
1826 @c man begin INCLUDE
1827 @include c-xtensa.texi
1828 @c ended inside the included file
1833 @c man begin OPTIONS
1836 The following options are available when @value{AS} is configured for
1837 a Z80 family processor.
1840 Assemble for Z80 processor.
1842 Assemble for R800 processor.
1843 @item -ignore-undocumented-instructions
1845 Assemble undocumented Z80 instructions that also work on R800 without warning.
1846 @item -ignore-unportable-instructions
1848 Assemble all undocumented Z80 instructions without warning.
1849 @item -warn-undocumented-instructions
1851 Issue a warning for undocumented Z80 instructions that also work on R800.
1852 @item -warn-unportable-instructions
1854 Issue a warning for undocumented Z80 instructions that do not work on R800.
1855 @item -forbid-undocumented-instructions
1857 Treat all undocumented instructions as errors.
1858 @item -forbid-unportable-instructions
1860 Treat undocumented Z80 instructions that do not work on R800 as errors.
1867 * Manual:: Structure of this Manual
1868 * GNU Assembler:: The GNU Assembler
1869 * Object Formats:: Object File Formats
1870 * Command Line:: Command Line
1871 * Input Files:: Input Files
1872 * Object:: Output (Object) File
1873 * Errors:: Error and Warning Messages
1877 @section Structure of this Manual
1879 @cindex manual, structure and purpose
1880 This manual is intended to describe what you need to know to use
1881 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1882 notation for symbols, constants, and expressions; the directives that
1883 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1886 We also cover special features in the @value{TARGET}
1887 configuration of @command{@value{AS}}, including assembler directives.
1890 This manual also describes some of the machine-dependent features of
1891 various flavors of the assembler.
1894 @cindex machine instructions (not covered)
1895 On the other hand, this manual is @emph{not} intended as an introduction
1896 to programming in assembly language---let alone programming in general!
1897 In a similar vein, we make no attempt to introduce the machine
1898 architecture; we do @emph{not} describe the instruction set, standard
1899 mnemonics, registers or addressing modes that are standard to a
1900 particular architecture.
1902 You may want to consult the manufacturer's
1903 machine architecture manual for this information.
1907 For information on the H8/300 machine instruction set, see @cite{H8/300
1908 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1909 Programming Manual} (Renesas).
1912 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1913 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1914 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1915 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1918 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1922 @c I think this is premature---doc@cygnus.com, 17jan1991
1924 Throughout this manual, we assume that you are running @dfn{GNU},
1925 the portable operating system from the @dfn{Free Software
1926 Foundation, Inc.}. This restricts our attention to certain kinds of
1927 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1928 once this assumption is granted examples and definitions need less
1931 @command{@value{AS}} is part of a team of programs that turn a high-level
1932 human-readable series of instructions into a low-level
1933 computer-readable series of instructions. Different versions of
1934 @command{@value{AS}} are used for different kinds of computer.
1937 @c There used to be a section "Terminology" here, which defined
1938 @c "contents", "byte", "word", and "long". Defining "word" to any
1939 @c particular size is confusing when the .word directive may generate 16
1940 @c bits on one machine and 32 bits on another; in general, for the user
1941 @c version of this manual, none of these terms seem essential to define.
1942 @c They were used very little even in the former draft of the manual;
1943 @c this draft makes an effort to avoid them (except in names of
1947 @section The GNU Assembler
1949 @c man begin DESCRIPTION
1951 @sc{gnu} @command{as} is really a family of assemblers.
1953 This manual describes @command{@value{AS}}, a member of that family which is
1954 configured for the @value{TARGET} architectures.
1956 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1957 should find a fairly similar environment when you use it on another
1958 architecture. Each version has much in common with the others,
1959 including object file formats, most assembler directives (often called
1960 @dfn{pseudo-ops}) and assembler syntax.@refill
1962 @cindex purpose of @sc{gnu} assembler
1963 @command{@value{AS}} is primarily intended to assemble the output of the
1964 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1965 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1966 assemble correctly everything that other assemblers for the same
1967 machine would assemble.
1969 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1972 @c This remark should appear in generic version of manual; assumption
1973 @c here is that generic version sets M680x0.
1974 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1975 assembler for the same architecture; for example, we know of several
1976 incompatible versions of 680x0 assembly language syntax.
1981 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1982 program in one pass of the source file. This has a subtle impact on the
1983 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1985 @node Object Formats
1986 @section Object File Formats
1988 @cindex object file format
1989 The @sc{gnu} assembler can be configured to produce several alternative
1990 object file formats. For the most part, this does not affect how you
1991 write assembly language programs; but directives for debugging symbols
1992 are typically different in different file formats. @xref{Symbol
1993 Attributes,,Symbol Attributes}.
1996 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1997 @value{OBJ-NAME} format object files.
1999 @c The following should exhaust all configs that set MULTI-OBJ, ideally
2001 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
2002 SOM or ELF format object files.
2007 @section Command Line
2009 @cindex command line conventions
2011 After the program name @command{@value{AS}}, the command line may contain
2012 options and file names. Options may appear in any order, and may be
2013 before, after, or between file names. The order of file names is
2016 @cindex standard input, as input file
2018 @file{--} (two hyphens) by itself names the standard input file
2019 explicitly, as one of the files for @command{@value{AS}} to assemble.
2021 @cindex options, command line
2022 Except for @samp{--} any command-line argument that begins with a
2023 hyphen (@samp{-}) is an option. Each option changes the behavior of
2024 @command{@value{AS}}. No option changes the way another option works. An
2025 option is a @samp{-} followed by one or more letters; the case of
2026 the letter is important. All options are optional.
2028 Some options expect exactly one file name to follow them. The file
2029 name may either immediately follow the option's letter (compatible
2030 with older assemblers) or it may be the next command argument (@sc{gnu}
2031 standard). These two command lines are equivalent:
2034 @value{AS} -o my-object-file.o mumble.s
2035 @value{AS} -omy-object-file.o mumble.s
2039 @section Input Files
2042 @cindex source program
2043 @cindex files, input
2044 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
2045 describe the program input to one run of @command{@value{AS}}. The program may
2046 be in one or more files; how the source is partitioned into files
2047 doesn't change the meaning of the source.
2049 @c I added "con" prefix to "catenation" just to prove I can overcome my
2050 @c APL training... doc@cygnus.com
2051 The source program is a concatenation of the text in all the files, in the
2054 @c man begin DESCRIPTION
2055 Each time you run @command{@value{AS}} it assembles exactly one source
2056 program. The source program is made up of one or more files.
2057 (The standard input is also a file.)
2059 You give @command{@value{AS}} a command line that has zero or more input file
2060 names. The input files are read (from left file name to right). A
2061 command-line argument (in any position) that has no special meaning
2062 is taken to be an input file name.
2064 If you give @command{@value{AS}} no file names it attempts to read one input file
2065 from the @command{@value{AS}} standard input, which is normally your terminal. You
2066 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
2069 Use @samp{--} if you need to explicitly name the standard input file
2070 in your command line.
2072 If the source is empty, @command{@value{AS}} produces a small, empty object
2077 @subheading Filenames and Line-numbers
2079 @cindex input file linenumbers
2080 @cindex line numbers, in input files
2081 There are two ways of locating a line in the input file (or files) and
2082 either may be used in reporting error messages. One way refers to a line
2083 number in a physical file; the other refers to a line number in a
2084 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
2086 @dfn{Physical files} are those files named in the command line given
2087 to @command{@value{AS}}.
2089 @dfn{Logical files} are simply names declared explicitly by assembler
2090 directives; they bear no relation to physical files. Logical file names help
2091 error messages reflect the original source file, when @command{@value{AS}} source
2092 is itself synthesized from other files. @command{@value{AS}} understands the
2093 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
2094 @ref{File,,@code{.file}}.
2097 @section Output (Object) File
2103 Every time you run @command{@value{AS}} it produces an output file, which is
2104 your assembly language program translated into numbers. This file
2105 is the object file. Its default name is @code{a.out}.
2106 You can give it another name by using the @option{-o} option. Conventionally,
2107 object file names end with @file{.o}. The default name is used for historical
2108 reasons: older assemblers were capable of assembling self-contained programs
2109 directly into a runnable program. (For some formats, this isn't currently
2110 possible, but it can be done for the @code{a.out} format.)
2114 The object file is meant for input to the linker @code{@value{LD}}. It contains
2115 assembled program code, information to help @code{@value{LD}} integrate
2116 the assembled program into a runnable file, and (optionally) symbolic
2117 information for the debugger.
2119 @c link above to some info file(s) like the description of a.out.
2120 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
2123 @section Error and Warning Messages
2125 @c man begin DESCRIPTION
2127 @cindex error messages
2128 @cindex warning messages
2129 @cindex messages from assembler
2130 @command{@value{AS}} may write warnings and error messages to the standard error
2131 file (usually your terminal). This should not happen when a compiler
2132 runs @command{@value{AS}} automatically. Warnings report an assumption made so
2133 that @command{@value{AS}} could keep assembling a flawed program; errors report a
2134 grave problem that stops the assembly.
2138 @cindex format of warning messages
2139 Warning messages have the format
2142 file_name:@b{NNN}:Warning Message Text
2146 @cindex file names and line numbers, in warnings/errors
2147 (where @b{NNN} is a line number). If both a logical file name
2148 (@pxref{File,,@code{.file}}) and a logical line number
2150 (@pxref{Line,,@code{.line}})
2152 have been given then they will be used, otherwise the file name and line number
2153 in the current assembler source file will be used. The message text is
2154 intended to be self explanatory (in the grand Unix tradition).
2156 Note the file name must be set via the logical version of the @code{.file}
2157 directive, not the DWARF2 version of the @code{.file} directive. For example:
2161 error_assembler_source
2167 produces this output:
2171 asm.s:2: Error: no such instruction: `error_assembler_source'
2172 foo.c:31: Error: no such instruction: `error_c_source'
2175 @cindex format of error messages
2176 Error messages have the format
2179 file_name:@b{NNN}:FATAL:Error Message Text
2182 The file name and line number are derived as for warning
2183 messages. The actual message text may be rather less explanatory
2184 because many of them aren't supposed to happen.
2187 @chapter Command-Line Options
2189 @cindex options, all versions of assembler
2190 This chapter describes command-line options available in @emph{all}
2191 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
2192 for options specific
2194 to the @value{TARGET} target.
2197 to particular machine architectures.
2200 @c man begin DESCRIPTION
2202 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
2203 you can use the @samp{-Wa} option to pass arguments through to the assembler.
2204 The assembler arguments must be separated from each other (and the @samp{-Wa})
2205 by commas. For example:
2208 gcc -c -g -O -Wa,-alh,-L file.c
2212 This passes two options to the assembler: @samp{-alh} (emit a listing to
2213 standard output with high-level and assembly source) and @samp{-L} (retain
2214 local symbols in the symbol table).
2216 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
2217 command-line options are automatically passed to the assembler by the compiler.
2218 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
2219 precisely what options it passes to each compilation pass, including the
2225 * a:: -a[cdghlns] enable listings
2226 * alternate:: --alternate enable alternate macro syntax
2227 * D:: -D for compatibility
2228 * f:: -f to work faster
2229 * I:: -I for .include search path
2230 @ifclear DIFF-TBL-KLUGE
2231 * K:: -K for compatibility
2233 @ifset DIFF-TBL-KLUGE
2234 * K:: -K for difference tables
2237 * L:: -L to retain local symbols
2238 * listing:: --listing-XXX to configure listing output
2239 * M:: -M or --mri to assemble in MRI compatibility mode
2240 * MD:: --MD for dependency tracking
2241 * no-pad-sections:: --no-pad-sections to stop section padding
2242 * o:: -o to name the object file
2243 * R:: -R to join data and text sections
2244 * statistics:: --statistics to see statistics about assembly
2245 * traditional-format:: --traditional-format for compatible output
2246 * v:: -v to announce version
2247 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
2248 * Z:: -Z to make object file even after errors
2252 @section Enable Listings: @option{-a[cdghlns]}
2262 @cindex listings, enabling
2263 @cindex assembly listings, enabling
2265 These options enable listing output from the assembler. By itself,
2266 @samp{-a} requests high-level, assembly, and symbols listing.
2267 You can use other letters to select specific options for the list:
2268 @samp{-ah} requests a high-level language listing,
2269 @samp{-al} requests an output-program assembly listing, and
2270 @samp{-as} requests a symbol table listing.
2271 High-level listings require that a compiler debugging option like
2272 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
2275 Use the @samp{-ag} option to print a first section with general assembly
2276 information, like @value{AS} version, switches passed, or time stamp.
2278 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
2279 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
2280 other conditional), or a true @code{.if} followed by an @code{.else}, will be
2281 omitted from the listing.
2283 Use the @samp{-ad} option to omit debugging directives from the
2286 Once you have specified one of these options, you can further control
2287 listing output and its appearance using the directives @code{.list},
2288 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
2290 The @samp{-an} option turns off all forms processing.
2291 If you do not request listing output with one of the @samp{-a} options, the
2292 listing-control directives have no effect.
2294 The letters after @samp{-a} may be combined into one option,
2295 @emph{e.g.}, @samp{-aln}.
2297 Note if the assembler source is coming from the standard input (e.g.,
2299 is being created by @code{@value{GCC}} and the @samp{-pipe} command-line switch
2300 is being used) then the listing will not contain any comments or preprocessor
2301 directives. This is because the listing code buffers input source lines from
2302 stdin only after they have been preprocessed by the assembler. This reduces
2303 memory usage and makes the code more efficient.
2306 @section @option{--alternate}
2309 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
2312 @section @option{-D}
2315 This option has no effect whatsoever, but it is accepted to make it more
2316 likely that scripts written for other assemblers also work with
2317 @command{@value{AS}}.
2320 @section Work Faster: @option{-f}
2323 @cindex trusted compiler
2324 @cindex faster processing (@option{-f})
2325 @samp{-f} should only be used when assembling programs written by a
2326 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
2327 and comment preprocessing on
2328 the input file(s) before assembling them. @xref{Preprocessing,
2332 @emph{Warning:} if you use @samp{-f} when the files actually need to be
2333 preprocessed (if they contain comments, for example), @command{@value{AS}} does
2338 @section @code{.include} Search Path: @option{-I} @var{path}
2340 @kindex -I @var{path}
2341 @cindex paths for @code{.include}
2342 @cindex search path for @code{.include}
2343 @cindex @code{include} directive search path
2344 Use this option to add a @var{path} to the list of directories
2345 @command{@value{AS}} searches for files specified in @code{.include}
2346 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
2347 many times as necessary to include a variety of paths. The current
2348 working directory is always searched first; after that, @command{@value{AS}}
2349 searches any @samp{-I} directories in the same order as they were
2350 specified (left to right) on the command line.
2353 @section Difference Tables: @option{-K}
2356 @ifclear DIFF-TBL-KLUGE
2357 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2358 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2359 where it can be used to warn when the assembler alters the machine code
2360 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2361 family does not have the addressing limitations that sometimes lead to this
2362 alteration on other platforms.
2365 @ifset DIFF-TBL-KLUGE
2366 @cindex difference tables, warning
2367 @cindex warning for altered difference tables
2368 @command{@value{AS}} sometimes alters the code emitted for directives of the
2369 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2370 You can use the @samp{-K} option if you want a warning issued when this
2375 @section Include Local Symbols: @option{-L}
2378 @cindex local symbols, retaining in output
2379 Symbols beginning with system-specific local label prefixes, typically
2380 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2381 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2382 such symbols when debugging, because they are intended for the use of
2383 programs (like compilers) that compose assembler programs, not for your
2384 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2385 such symbols, so you do not normally debug with them.
2387 This option tells @command{@value{AS}} to retain those local symbols
2388 in the object file. Usually if you do this you also tell the linker
2389 @code{@value{LD}} to preserve those symbols.
2392 @section Configuring listing output: @option{--listing}
2394 The listing feature of the assembler can be enabled via the command-line switch
2395 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2396 hex dump of the corresponding locations in the output object file, and displays
2397 them as a listing file. The format of this listing can be controlled by
2398 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2399 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2400 @code{.psize} (@pxref{Psize}), and
2401 @code{.eject} (@pxref{Eject}) and also by the following switches:
2404 @item --listing-lhs-width=@samp{number}
2405 @kindex --listing-lhs-width
2406 @cindex Width of first line disassembly output
2407 Sets the maximum width, in words, of the first line of the hex byte dump. This
2408 dump appears on the left hand side of the listing output.
2410 @item --listing-lhs-width2=@samp{number}
2411 @kindex --listing-lhs-width2
2412 @cindex Width of continuation lines of disassembly output
2413 Sets the maximum width, in words, of any further lines of the hex byte dump for
2414 a given input source line. If this value is not specified, it defaults to being
2415 the same as the value specified for @samp{--listing-lhs-width}. If neither
2416 switch is used the default is to one.
2418 @item --listing-rhs-width=@samp{number}
2419 @kindex --listing-rhs-width
2420 @cindex Width of source line output
2421 Sets the maximum width, in characters, of the source line that is displayed
2422 alongside the hex dump. The default value for this parameter is 100. The
2423 source line is displayed on the right hand side of the listing output.
2425 @item --listing-cont-lines=@samp{number}
2426 @kindex --listing-cont-lines
2427 @cindex Maximum number of continuation lines
2428 Sets the maximum number of continuation lines of hex dump that will be
2429 displayed for a given single line of source input. The default value is 4.
2433 @section Assemble in MRI Compatibility Mode: @option{-M}
2436 @cindex MRI compatibility mode
2437 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2438 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2439 compatible with the @code{ASM68K} assembler from Microtec Research.
2440 The exact nature of the
2441 MRI syntax will not be documented here; see the MRI manuals for more
2442 information. Note in particular that the handling of macros and macro
2443 arguments is somewhat different. The purpose of this option is to permit
2444 assembling existing MRI assembler code using @command{@value{AS}}.
2446 The MRI compatibility is not complete. Certain operations of the MRI assembler
2447 depend upon its object file format, and can not be supported using other object
2448 file formats. Supporting these would require enhancing each object file format
2449 individually. These are:
2452 @item global symbols in common section
2454 The m68k MRI assembler supports common sections which are merged by the linker.
2455 Other object file formats do not support this. @command{@value{AS}} handles
2456 common sections by treating them as a single common symbol. It permits local
2457 symbols to be defined within a common section, but it can not support global
2458 symbols, since it has no way to describe them.
2460 @item complex relocations
2462 The MRI assemblers support relocations against a negated section address, and
2463 relocations which combine the start addresses of two or more sections. These
2464 are not support by other object file formats.
2466 @item @code{END} pseudo-op specifying start address
2468 The MRI @code{END} pseudo-op permits the specification of a start address.
2469 This is not supported by other object file formats. The start address may
2470 instead be specified using the @option{-e} option to the linker, or in a linker
2473 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2475 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2476 name to the output file. This is not supported by other object file formats.
2478 @item @code{ORG} pseudo-op
2480 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2481 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2482 which changes the location within the current section. Absolute sections are
2483 not supported by other object file formats. The address of a section may be
2484 assigned within a linker script.
2487 There are some other features of the MRI assembler which are not supported by
2488 @command{@value{AS}}, typically either because they are difficult or because they
2489 seem of little consequence. Some of these may be supported in future releases.
2493 @item EBCDIC strings
2495 EBCDIC strings are not supported.
2497 @item packed binary coded decimal
2499 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2500 and @code{DCB.P} pseudo-ops are not supported.
2502 @item @code{FEQU} pseudo-op
2504 The m68k @code{FEQU} pseudo-op is not supported.
2506 @item @code{NOOBJ} pseudo-op
2508 The m68k @code{NOOBJ} pseudo-op is not supported.
2510 @item @code{OPT} branch control options
2512 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2513 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2514 relaxes all branches, whether forward or backward, to an appropriate size, so
2515 these options serve no purpose.
2517 @item @code{OPT} list control options
2519 The following m68k @code{OPT} list control options are ignored: @code{C},
2520 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2521 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2523 @item other @code{OPT} options
2525 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2526 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2528 @item @code{OPT} @code{D} option is default
2530 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2531 @code{OPT NOD} may be used to turn it off.
2533 @item @code{XREF} pseudo-op.
2535 The m68k @code{XREF} pseudo-op is ignored.
2540 @section Dependency Tracking: @option{--MD}
2543 @cindex dependency tracking
2546 @command{@value{AS}} can generate a dependency file for the file it creates. This
2547 file consists of a single rule suitable for @code{make} describing the
2548 dependencies of the main source file.
2550 The rule is written to the file named in its argument.
2552 This feature is used in the automatic updating of makefiles.
2554 @node no-pad-sections
2555 @section Output Section Padding
2556 @kindex --no-pad-sections
2557 @cindex output section padding
2558 Normally the assembler will pad the end of each output section up to its
2559 alignment boundary. But this can waste space, which can be significant on
2560 memory constrained targets. So the @option{--no-pad-sections} option will
2561 disable this behaviour.
2564 @section Name the Object File: @option{-o}
2567 @cindex naming object file
2568 @cindex object file name
2569 There is always one object file output when you run @command{@value{AS}}. By
2570 default it has the name @file{a.out}.
2571 You use this option (which takes exactly one filename) to give the
2572 object file a different name.
2574 Whatever the object file is called, @command{@value{AS}} overwrites any
2575 existing file of the same name.
2578 @section Join Data and Text Sections: @option{-R}
2581 @cindex data and text sections, joining
2582 @cindex text and data sections, joining
2583 @cindex joining text and data sections
2584 @cindex merging text and data sections
2585 @option{-R} tells @command{@value{AS}} to write the object file as if all
2586 data-section data lives in the text section. This is only done at
2587 the very last moment: your binary data are the same, but data
2588 section parts are relocated differently. The data section part of
2589 your object file is zero bytes long because all its bytes are
2590 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2592 When you specify @option{-R} it would be possible to generate shorter
2593 address displacements (because we do not have to cross between text and
2594 data section). We refrain from doing this simply for compatibility with
2595 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2598 When @command{@value{AS}} is configured for COFF or ELF output,
2599 this option is only useful if you use sections named @samp{.text} and
2604 @option{-R} is not supported for any of the HPPA targets. Using
2605 @option{-R} generates a warning from @command{@value{AS}}.
2609 @section Display Assembly Statistics: @option{--statistics}
2611 @kindex --statistics
2612 @cindex statistics, about assembly
2613 @cindex time, total for assembly
2614 @cindex space used, maximum for assembly
2615 Use @samp{--statistics} to display two statistics about the resources used by
2616 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2617 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2620 @node traditional-format
2621 @section Compatible Output: @option{--traditional-format}
2623 @kindex --traditional-format
2624 For some targets, the output of @command{@value{AS}} is different in some ways
2625 from the output of some existing assembler. This switch requests
2626 @command{@value{AS}} to use the traditional format instead.
2628 For example, it disables the exception frame optimizations which
2629 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2632 @section Announce Version: @option{-v}
2636 @cindex assembler version
2637 @cindex version of assembler
2638 You can find out what version of as is running by including the
2639 option @samp{-v} (which you can also spell as @samp{-version}) on the
2643 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2645 @command{@value{AS}} should never give a warning or error message when
2646 assembling compiler output. But programs written by people often
2647 cause @command{@value{AS}} to give a warning that a particular assumption was
2648 made. All such warnings are directed to the standard error file.
2652 @cindex suppressing warnings
2653 @cindex warnings, suppressing
2654 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2655 This only affects the warning messages: it does not change any particular of
2656 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2659 @kindex --fatal-warnings
2660 @cindex errors, caused by warnings
2661 @cindex warnings, causing error
2662 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2663 files that generate warnings to be in error.
2666 @cindex warnings, switching on
2667 You can switch these options off again by specifying @option{--warn}, which
2668 causes warnings to be output as usual.
2671 @section Generate Object File in Spite of Errors: @option{-Z}
2672 @cindex object file, after errors
2673 @cindex errors, continuing after
2674 After an error message, @command{@value{AS}} normally produces no output. If for
2675 some reason you are interested in object file output even after
2676 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2677 option. If there are any errors, @command{@value{AS}} continues anyways, and
2678 writes an object file after a final warning message of the form @samp{@var{n}
2679 errors, @var{m} warnings, generating bad object file.}
2684 @cindex machine-independent syntax
2685 @cindex syntax, machine-independent
2686 This chapter describes the machine-independent syntax allowed in a
2687 source file. @command{@value{AS}} syntax is similar to what many other
2688 assemblers use; it is inspired by the BSD 4.2
2693 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2697 * Preprocessing:: Preprocessing
2698 * Whitespace:: Whitespace
2699 * Comments:: Comments
2700 * Symbol Intro:: Symbols
2701 * Statements:: Statements
2702 * Constants:: Constants
2706 @section Preprocessing
2708 @cindex preprocessing
2709 The @command{@value{AS}} internal preprocessor:
2711 @cindex whitespace, removed by preprocessor
2713 adjusts and removes extra whitespace. It leaves one space or tab before
2714 the keywords on a line, and turns any other whitespace on the line into
2717 @cindex comments, removed by preprocessor
2719 removes all comments, replacing them with a single space, or an
2720 appropriate number of newlines.
2722 @cindex constants, converted by preprocessor
2724 converts character constants into the appropriate numeric values.
2727 It does not do macro processing, include file handling, or
2728 anything else you may get from your C compiler's preprocessor. You can
2729 do include file processing with the @code{.include} directive
2730 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2731 to get other ``CPP'' style preprocessing by giving the input file a
2732 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2733 Output, gcc info, Using GNU CC}.
2735 Excess whitespace, comments, and character constants
2736 cannot be used in the portions of the input text that are not
2739 @cindex turning preprocessing on and off
2740 @cindex preprocessing, turning on and off
2743 If the first line of an input file is @code{#NO_APP} or if you use the
2744 @samp{-f} option, whitespace and comments are not removed from the input file.
2745 Within an input file, you can ask for whitespace and comment removal in
2746 specific portions of the by putting a line that says @code{#APP} before the
2747 text that may contain whitespace or comments, and putting a line that says
2748 @code{#NO_APP} after this text. This feature is mainly intend to support
2749 @code{asm} statements in compilers whose output is otherwise free of comments
2756 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2757 Whitespace is used to separate symbols, and to make programs neater for
2758 people to read. Unless within character constants
2759 (@pxref{Characters,,Character Constants}), any whitespace means the same
2760 as exactly one space.
2766 There are two ways of rendering comments to @command{@value{AS}}. In both
2767 cases the comment is equivalent to one space.
2769 Anything from @samp{/*} through the next @samp{*/} is a comment.
2770 This means you may not nest these comments.
2774 The only way to include a newline ('\n') in a comment
2775 is to use this sort of comment.
2778 /* This sort of comment does not nest. */
2781 @cindex line comment character
2782 Anything from a @dfn{line comment} character up to the next newline is
2783 considered a comment and is ignored. The line comment character is target
2784 specific, and some targets multiple comment characters. Some targets also have
2785 line comment characters that only work if they are the first character on a
2786 line. Some targets use a sequence of two characters to introduce a line
2787 comment. Some targets can also change their line comment characters depending
2788 upon command-line options that have been used. For more details see the
2789 @emph{Syntax} section in the documentation for individual targets.
2791 If the line comment character is the hash sign (@samp{#}) then it still has the
2792 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2793 to specify logical line numbers:
2796 @cindex lines starting with @code{#}
2797 @cindex logical line numbers
2798 To be compatible with past assemblers, lines that begin with @samp{#} have a
2799 special interpretation. Following the @samp{#} should be an absolute
2800 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2801 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2802 new logical file name. The rest of the line, if any, should be whitespace.
2804 If the first non-whitespace characters on the line are not numeric,
2805 the line is ignored. (Just like a comment.)
2808 # This is an ordinary comment.
2809 # 42-6 "new_file_name" # New logical file name
2810 # This is logical line # 36.
2812 This feature is deprecated, and may disappear from future versions
2813 of @command{@value{AS}}.
2818 @cindex characters used in symbols
2819 @ifclear SPECIAL-SYMS
2820 A @dfn{symbol} is one or more characters chosen from the set of all
2821 letters (both upper and lower case), digits and the three characters
2827 A @dfn{symbol} is one or more characters chosen from the set of all
2828 letters (both upper and lower case), digits and the three characters
2829 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2835 On most machines, you can also use @code{$} in symbol names; exceptions
2836 are noted in @ref{Machine Dependencies}.
2838 No symbol may begin with a digit. Case is significant.
2839 There is no length limit; all characters are significant. Multibyte characters
2840 are supported. Symbols are delimited by characters not in that set, or by the
2841 beginning of a file (since the source program must end with a newline, the end
2842 of a file is not a possible symbol delimiter). @xref{Symbols}.
2844 Symbol names may also be enclosed in double quote @code{"} characters. In such
2845 cases any characters are allowed, except for the NUL character. If a double
2846 quote character is to be included in the symbol name it must be preceeded by a
2847 backslash @code{\} character.
2848 @cindex length of symbols
2853 @cindex statements, structure of
2854 @cindex line separator character
2855 @cindex statement separator character
2857 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2858 @dfn{line separator character}. The line separator character is target
2859 specific and described in the @emph{Syntax} section of each
2860 target's documentation. Not all targets support a line separator character.
2861 The newline or line separator character is considered to be part of the
2862 preceding statement. Newlines and separators within character constants are an
2863 exception: they do not end statements.
2865 @cindex newline, required at file end
2866 @cindex EOF, newline must precede
2867 It is an error to end any statement with end-of-file: the last
2868 character of any input file should be a newline.@refill
2870 An empty statement is allowed, and may include whitespace. It is ignored.
2872 @cindex instructions and directives
2873 @cindex directives and instructions
2874 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2875 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2877 A statement begins with zero or more labels, optionally followed by a
2878 key symbol which determines what kind of statement it is. The key
2879 symbol determines the syntax of the rest of the statement. If the
2880 symbol begins with a dot @samp{.} then the statement is an assembler
2881 directive: typically valid for any computer. If the symbol begins with
2882 a letter the statement is an assembly language @dfn{instruction}: it
2883 assembles into a machine language instruction.
2885 Different versions of @command{@value{AS}} for different computers
2886 recognize different instructions. In fact, the same symbol may
2887 represent a different instruction in a different computer's assembly
2891 @cindex @code{:} (label)
2892 @cindex label (@code{:})
2893 A label is a symbol immediately followed by a colon (@code{:}).
2894 Whitespace before a label or after a colon is permitted, but you may not
2895 have whitespace between a label's symbol and its colon. @xref{Labels}.
2898 For HPPA targets, labels need not be immediately followed by a colon, but
2899 the definition of a label must begin in column zero. This also implies that
2900 only one label may be defined on each line.
2904 label: .directive followed by something
2905 another_label: # This is an empty statement.
2906 instruction operand_1, operand_2, @dots{}
2913 A constant is a number, written so that its value is known by
2914 inspection, without knowing any context. Like this:
2917 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2918 .ascii "Ring the bell\7" # A string constant.
2919 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2920 .float 0f-314159265358979323846264338327\
2921 95028841971.693993751E-40 # - pi, a flonum.
2926 * Characters:: Character Constants
2927 * Numbers:: Number Constants
2931 @subsection Character Constants
2933 @cindex character constants
2934 @cindex constants, character
2935 There are two kinds of character constants. A @dfn{character} stands
2936 for one character in one byte and its value may be used in
2937 numeric expressions. String constants (properly called string
2938 @emph{literals}) are potentially many bytes and their values may not be
2939 used in arithmetic expressions.
2943 * Chars:: Characters
2947 @subsubsection Strings
2949 @cindex string constants
2950 @cindex constants, string
2951 A @dfn{string} is written between double-quotes. It may contain
2952 double-quotes or null characters. The way to get special characters
2953 into a string is to @dfn{escape} these characters: precede them with
2954 a backslash @samp{\} character. For example @samp{\\} represents
2955 one backslash: the first @code{\} is an escape which tells
2956 @command{@value{AS}} to interpret the second character literally as a backslash
2957 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2958 escape character). The complete list of escapes follows.
2960 @cindex escape codes, character
2961 @cindex character escape codes
2962 @c NOTE: Cindex entries must not start with a backlash character.
2963 @c NOTE: This confuses the pdf2texi script when it is creating the
2964 @c NOTE: index based upon the first character and so it generates:
2965 @c NOTE: \initial {\\}
2966 @c NOTE: which then results in the error message:
2967 @c NOTE: Argument of \\ has an extra }.
2968 @c NOTE: So in the index entries below a space character has been
2969 @c NOTE: prepended to avoid this problem.
2972 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2974 @cindex @code{ \b} (backspace character)
2975 @cindex backspace (@code{\b})
2977 Mnemonic for backspace; for ASCII this is octal code 010.
2980 @c Mnemonic for EOText; for ASCII this is octal code 004.
2982 @cindex @code{ \f} (formfeed character)
2983 @cindex formfeed (@code{\f})
2985 Mnemonic for FormFeed; for ASCII this is octal code 014.
2987 @cindex @code{ \n} (newline character)
2988 @cindex newline (@code{\n})
2990 Mnemonic for newline; for ASCII this is octal code 012.
2993 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2995 @cindex @code{ \r} (carriage return character)
2996 @cindex carriage return (@code{backslash-r})
2998 Mnemonic for carriage-Return; for ASCII this is octal code 015.
3001 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
3002 @c other assemblers.
3004 @cindex @code{ \t} (tab)
3005 @cindex tab (@code{\t})
3007 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
3010 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
3011 @c @item \x @var{digit} @var{digit} @var{digit}
3012 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
3014 @cindex @code{ \@var{ddd}} (octal character code)
3015 @cindex octal character code (@code{\@var{ddd}})
3016 @item \ @var{digit} @var{digit} @var{digit}
3017 An octal character code. The numeric code is 3 octal digits.
3018 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
3019 for example, @code{\008} has the value 010, and @code{\009} the value 011.
3021 @cindex @code{ \@var{xd...}} (hex character code)
3022 @cindex hex character code (@code{\@var{xd...}})
3023 @item \@code{x} @var{hex-digits...}
3024 A hex character code. All trailing hex digits are combined. Either upper or
3025 lower case @code{x} works.
3027 @cindex @code{ \\} (@samp{\} character)
3028 @cindex backslash (@code{\\})
3030 Represents one @samp{\} character.
3033 @c Represents one @samp{'} (accent acute) character.
3034 @c This is needed in single character literals
3035 @c (@xref{Characters,,Character Constants}.) to represent
3038 @cindex @code{ \"} (doublequote character)
3039 @cindex doublequote (@code{\"})
3041 Represents one @samp{"} character. Needed in strings to represent
3042 this character, because an unescaped @samp{"} would end the string.
3044 @item \ @var{anything-else}
3045 Any other character when escaped by @kbd{\} gives a warning, but
3046 assembles as if the @samp{\} was not present. The idea is that if
3047 you used an escape sequence you clearly didn't want the literal
3048 interpretation of the following character. However @command{@value{AS}} has no
3049 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
3050 code and warns you of the fact.
3053 Which characters are escapable, and what those escapes represent,
3054 varies widely among assemblers. The current set is what we think
3055 the BSD 4.2 assembler recognizes, and is a subset of what most C
3056 compilers recognize. If you are in doubt, do not use an escape
3060 @subsubsection Characters
3062 @cindex single character constant
3063 @cindex character, single
3064 @cindex constant, single character
3065 A single character may be written as a single quote immediately followed by
3066 that character. Some backslash escapes apply to characters, @code{\b},
3067 @code{\f}, @code{\n}, @code{\r}, @code{\t}, and @code{\"} with the same meaning
3068 as for strings, plus @code{\'} for a single quote. So if you want to write the
3069 character backslash, you must write @kbd{'\\} where the first @code{\} escapes
3070 the second @code{\}. As you can see, the quote is an acute accent, not a grave
3073 @ifclear abnormal-separator
3074 (or semicolon @samp{;})
3076 @ifset abnormal-separator
3078 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
3083 immediately following an acute accent is taken as a literal character
3084 and does not count as the end of a statement. The value of a character
3085 constant in a numeric expression is the machine's byte-wide code for
3086 that character. @command{@value{AS}} assumes your character code is ASCII:
3087 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
3090 @subsection Number Constants
3092 @cindex constants, number
3093 @cindex number constants
3094 @command{@value{AS}} distinguishes three kinds of numbers according to how they
3095 are stored in the target machine. @emph{Integers} are numbers that
3096 would fit into an @code{int} in the C language. @emph{Bignums} are
3097 integers, but they are stored in more than 32 bits. @emph{Flonums}
3098 are floating point numbers, described below.
3101 * Integers:: Integers
3109 @subsubsection Integers
3111 @cindex constants, integer
3113 @cindex binary integers
3114 @cindex integers, binary
3115 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
3116 the binary digits @samp{01}.
3118 @cindex octal integers
3119 @cindex integers, octal
3120 An octal integer is @samp{0} followed by zero or more of the octal
3121 digits (@samp{01234567}).
3123 @cindex decimal integers
3124 @cindex integers, decimal
3125 A decimal integer starts with a non-zero digit followed by zero or
3126 more digits (@samp{0123456789}).
3128 @cindex hexadecimal integers
3129 @cindex integers, hexadecimal
3130 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
3131 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
3133 Integers have the usual values. To denote a negative integer, use
3134 the prefix operator @samp{-} discussed under expressions
3135 (@pxref{Prefix Ops,,Prefix Operators}).
3138 @subsubsection Bignums
3141 @cindex constants, bignum
3142 A @dfn{bignum} has the same syntax and semantics as an integer
3143 except that the number (or its negative) takes more than 32 bits to
3144 represent in binary. The distinction is made because in some places
3145 integers are permitted while bignums are not.
3148 @subsubsection Flonums
3150 @cindex floating point numbers
3151 @cindex constants, floating point
3153 @cindex precision, floating point
3154 A @dfn{flonum} represents a floating point number. The translation is
3155 indirect: a decimal floating point number from the text is converted by
3156 @command{@value{AS}} to a generic binary floating point number of more than
3157 sufficient precision. This generic floating point number is converted
3158 to a particular computer's floating point format (or formats) by a
3159 portion of @command{@value{AS}} specialized to that computer.
3161 A flonum is written by writing (in order)
3166 (@samp{0} is optional on the HPPA.)
3170 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
3172 @kbd{e} is recommended. Case is not important.
3174 @c FIXME: verify if flonum syntax really this vague for most cases
3175 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
3176 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
3179 On the H8/300 and Renesas / SuperH SH architectures, the letter must be
3180 one of the letters @samp{DFPRSX} (in upper or lower case).
3182 On the ARC, the letter must be one of the letters @samp{DFRS}
3183 (in upper or lower case).
3185 On the HPPA architecture, the letter must be @samp{E} (upper case only).
3189 One of the letters @samp{DFRS} (in upper or lower case).
3192 One of the letters @samp{DFPRSX} (in upper or lower case).
3195 The letter @samp{E} (upper case only).
3200 An optional sign: either @samp{+} or @samp{-}.
3203 An optional @dfn{integer part}: zero or more decimal digits.
3206 An optional @dfn{fractional part}: @samp{.} followed by zero
3207 or more decimal digits.
3210 An optional exponent, consisting of:
3214 An @samp{E} or @samp{e}.
3215 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
3216 @c principle this can perfectly well be different on different targets.
3218 Optional sign: either @samp{+} or @samp{-}.
3220 One or more decimal digits.
3225 At least one of the integer part or the fractional part must be
3226 present. The floating point number has the usual base-10 value.
3228 @command{@value{AS}} does all processing using integers. Flonums are computed
3229 independently of any floating point hardware in the computer running
3230 @command{@value{AS}}.
3233 @chapter Sections and Relocation
3238 * Secs Background:: Background
3239 * Ld Sections:: Linker Sections
3240 * As Sections:: Assembler Internal Sections
3241 * Sub-Sections:: Sub-Sections
3245 @node Secs Background
3248 Roughly, a section is a range of addresses, with no gaps; all data
3249 ``in'' those addresses is treated the same for some particular purpose.
3250 For example there may be a ``read only'' section.
3252 @cindex linker, and assembler
3253 @cindex assembler, and linker
3254 The linker @code{@value{LD}} reads many object files (partial programs) and
3255 combines their contents to form a runnable program. When @command{@value{AS}}
3256 emits an object file, the partial program is assumed to start at address 0.
3257 @code{@value{LD}} assigns the final addresses for the partial program, so that
3258 different partial programs do not overlap. This is actually an
3259 oversimplification, but it suffices to explain how @command{@value{AS}} uses
3262 @code{@value{LD}} moves blocks of bytes of your program to their run-time
3263 addresses. These blocks slide to their run-time addresses as rigid
3264 units; their length does not change and neither does the order of bytes
3265 within them. Such a rigid unit is called a @emph{section}. Assigning
3266 run-time addresses to sections is called @dfn{relocation}. It includes
3267 the task of adjusting mentions of object-file addresses so they refer to
3268 the proper run-time addresses.
3270 For the H8/300, and for the Renesas / SuperH SH,
3271 @command{@value{AS}} pads sections if needed to
3272 ensure they end on a word (sixteen bit) boundary.
3275 @cindex standard assembler sections
3276 An object file written by @command{@value{AS}} has at least three sections, any
3277 of which may be empty. These are named @dfn{text}, @dfn{data} and
3282 When it generates COFF or ELF output,
3284 @command{@value{AS}} can also generate whatever other named sections you specify
3285 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
3286 If you do not use any directives that place output in the @samp{.text}
3287 or @samp{.data} sections, these sections still exist, but are empty.
3292 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
3294 @command{@value{AS}} can also generate whatever other named sections you
3295 specify using the @samp{.space} and @samp{.subspace} directives. See
3296 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3297 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3298 assembler directives.
3301 Additionally, @command{@value{AS}} uses different names for the standard
3302 text, data, and bss sections when generating SOM output. Program text
3303 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3304 BSS into @samp{$BSS$}.
3308 Within the object file, the text section starts at address @code{0}, the
3309 data section follows, and the bss section follows the data section.
3312 When generating either SOM or ELF output files on the HPPA, the text
3313 section starts at address @code{0}, the data section at address
3314 @code{0x4000000}, and the bss section follows the data section.
3317 To let @code{@value{LD}} know which data changes when the sections are
3318 relocated, and how to change that data, @command{@value{AS}} also writes to the
3319 object file details of the relocation needed. To perform relocation
3320 @code{@value{LD}} must know, each time an address in the object
3324 Where in the object file is the beginning of this reference to
3327 How long (in bytes) is this reference?
3329 Which section does the address refer to? What is the numeric value of
3331 (@var{address}) @minus{} (@var{start-address of section})?
3334 Is the reference to an address ``Program-Counter relative''?
3337 @cindex addresses, format of
3338 @cindex section-relative addressing
3339 In fact, every address @command{@value{AS}} ever uses is expressed as
3341 (@var{section}) + (@var{offset into section})
3344 Further, most expressions @command{@value{AS}} computes have this section-relative
3347 (For some object formats, such as SOM for the HPPA, some expressions are
3348 symbol-relative instead.)
3351 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3352 @var{N} into section @var{secname}.''
3354 Apart from text, data and bss sections you need to know about the
3355 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3356 addresses in the absolute section remain unchanged. For example, address
3357 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3358 @code{@value{LD}}. Although the linker never arranges two partial programs'
3359 data sections with overlapping addresses after linking, @emph{by definition}
3360 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3361 part of a program is always the same address when the program is running as
3362 address @code{@{absolute@ 239@}} in any other part of the program.
3364 The idea of sections is extended to the @dfn{undefined} section. Any
3365 address whose section is unknown at assembly time is by definition
3366 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3367 Since numbers are always defined, the only way to generate an undefined
3368 address is to mention an undefined symbol. A reference to a named
3369 common block would be such a symbol: its value is unknown at assembly
3370 time so it has section @emph{undefined}.
3372 By analogy the word @emph{section} is used to describe groups of sections in
3373 the linked program. @code{@value{LD}} puts all partial programs' text
3374 sections in contiguous addresses in the linked program. It is
3375 customary to refer to the @emph{text section} of a program, meaning all
3376 the addresses of all partial programs' text sections. Likewise for
3377 data and bss sections.
3379 Some sections are manipulated by @code{@value{LD}}; others are invented for
3380 use of @command{@value{AS}} and have no meaning except during assembly.
3383 @section Linker Sections
3384 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3389 @cindex named sections
3390 @cindex sections, named
3391 @item named sections
3394 @cindex text section
3395 @cindex data section
3399 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3400 separate but equal sections. Anything you can say of one section is
3403 When the program is running, however, it is
3404 customary for the text section to be unalterable. The
3405 text section is often shared among processes: it contains
3406 instructions, constants and the like. The data section of a running
3407 program is usually alterable: for example, C variables would be stored
3408 in the data section.
3413 This section contains zeroed bytes when your program begins running. It
3414 is used to hold uninitialized variables or common storage. The length of
3415 each partial program's bss section is important, but because it starts
3416 out containing zeroed bytes there is no need to store explicit zero
3417 bytes in the object file. The bss section was invented to eliminate
3418 those explicit zeros from object files.
3420 @cindex absolute section
3421 @item absolute section
3422 Address 0 of this section is always ``relocated'' to runtime address 0.
3423 This is useful if you want to refer to an address that @code{@value{LD}} must
3424 not change when relocating. In this sense we speak of absolute
3425 addresses being ``unrelocatable'': they do not change during relocation.
3427 @cindex undefined section
3428 @item undefined section
3429 This ``section'' is a catch-all for address references to objects not in
3430 the preceding sections.
3431 @c FIXME: ref to some other doc on obj-file formats could go here.
3434 @cindex relocation example
3435 An idealized example of three relocatable sections follows.
3437 The example uses the traditional section names @samp{.text} and @samp{.data}.
3439 Memory addresses are on the horizontal axis.
3443 @c END TEXI2ROFF-KILL
3446 partial program # 1: |ttttt|dddd|00|
3453 partial program # 2: |TTT|DDD|000|
3456 +--+---+-----+--+----+---+-----+~~
3457 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3458 +--+---+-----+--+----+---+-----+~~
3460 addresses: 0 @dots{}
3467 \line{\it Partial program \#1: \hfil}
3468 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3469 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3471 \line{\it Partial program \#2: \hfil}
3472 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3473 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3475 \line{\it linked program: \hfil}
3476 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3477 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3478 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3479 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3481 \line{\it addresses: \hfil}
3485 @c END TEXI2ROFF-KILL
3488 @section Assembler Internal Sections
3490 @cindex internal assembler sections
3491 @cindex sections in messages, internal
3492 These sections are meant only for the internal use of @command{@value{AS}}. They
3493 have no meaning at run-time. You do not really need to know about these
3494 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3495 warning messages, so it might be helpful to have an idea of their
3496 meanings to @command{@value{AS}}. These sections are used to permit the
3497 value of every expression in your assembly language program to be a
3498 section-relative address.
3501 @cindex assembler internal logic error
3502 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3503 An internal assembler logic error has been found. This means there is a
3504 bug in the assembler.
3506 @cindex expr (internal section)
3508 The assembler stores complex expression internally as combinations of
3509 symbols. When it needs to represent an expression as a symbol, it puts
3510 it in the expr section.
3512 @c FIXME item transfer[t] vector preload
3513 @c FIXME item transfer[t] vector postload
3514 @c FIXME item register
3518 @section Sub-Sections
3520 @cindex numbered subsections
3521 @cindex grouping data
3527 fall into two sections: text and data.
3529 You may have separate groups of
3531 data in named sections
3535 data in named sections
3541 that you want to end up near to each other in the object file, even though they
3542 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3543 use @dfn{subsections} for this purpose. Within each section, there can be
3544 numbered subsections with values from 0 to 8192. Objects assembled into the
3545 same subsection go into the object file together with other objects in the same
3546 subsection. For example, a compiler might want to store constants in the text
3547 section, but might not want to have them interspersed with the program being
3548 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3549 section of code being output, and a @samp{.text 1} before each group of
3550 constants being output.
3552 Subsections are optional. If you do not use subsections, everything
3553 goes in subsection number zero.
3556 Each subsection is zero-padded up to a multiple of four bytes.
3557 (Subsections may be padded a different amount on different flavors
3558 of @command{@value{AS}}.)
3562 On the H8/300 platform, each subsection is zero-padded to a word
3563 boundary (two bytes).
3564 The same is true on the Renesas SH.
3568 Subsections appear in your object file in numeric order, lowest numbered
3569 to highest. (All this to be compatible with other people's assemblers.)
3570 The object file contains no representation of subsections; @code{@value{LD}} and
3571 other programs that manipulate object files see no trace of them.
3572 They just see all your text subsections as a text section, and all your
3573 data subsections as a data section.
3575 To specify which subsection you want subsequent statements assembled
3576 into, use a numeric argument to specify it, in a @samp{.text
3577 @var{expression}} or a @samp{.data @var{expression}} statement.
3580 When generating COFF output, you
3585 can also use an extra subsection
3586 argument with arbitrary named sections: @samp{.section @var{name},
3591 When generating ELF output, you
3596 can also use the @code{.subsection} directive (@pxref{SubSection})
3597 to specify a subsection: @samp{.subsection @var{expression}}.
3599 @var{Expression} should be an absolute expression
3600 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3601 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3602 begins in @code{text 0}. For instance:
3604 .text 0 # The default subsection is text 0 anyway.
3605 .ascii "This lives in the first text subsection. *"
3607 .ascii "But this lives in the second text subsection."
3609 .ascii "This lives in the data section,"
3610 .ascii "in the first data subsection."
3612 .ascii "This lives in the first text section,"
3613 .ascii "immediately following the asterisk (*)."
3616 Each section has a @dfn{location counter} incremented by one for every byte
3617 assembled into that section. Because subsections are merely a convenience
3618 restricted to @command{@value{AS}} there is no concept of a subsection location
3619 counter. There is no way to directly manipulate a location counter---but the
3620 @code{.align} directive changes it, and any label definition captures its
3621 current value. The location counter of the section where statements are being
3622 assembled is said to be the @dfn{active} location counter.
3625 @section bss Section
3628 @cindex common variable storage
3629 The bss section is used for local common variable storage.
3630 You may allocate address space in the bss section, but you may
3631 not dictate data to load into it before your program executes. When
3632 your program starts running, all the contents of the bss
3633 section are zeroed bytes.
3635 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3636 @ref{Lcomm,,@code{.lcomm}}.
3638 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3639 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3642 When assembling for a target which supports multiple sections, such as ELF or
3643 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3644 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3645 section. Typically the section will only contain symbol definitions and
3646 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3653 Symbols are a central concept: the programmer uses symbols to name
3654 things, the linker uses symbols to link, and the debugger uses symbols
3658 @cindex debuggers, and symbol order
3659 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3660 the same order they were declared. This may break some debuggers.
3665 * Setting Symbols:: Giving Symbols Other Values
3666 * Symbol Names:: Symbol Names
3667 * Dot:: The Special Dot Symbol
3668 * Symbol Attributes:: Symbol Attributes
3675 A @dfn{label} is written as a symbol immediately followed by a colon
3676 @samp{:}. The symbol then represents the current value of the
3677 active location counter, and is, for example, a suitable instruction
3678 operand. You are warned if you use the same symbol to represent two
3679 different locations: the first definition overrides any other
3683 On the HPPA, the usual form for a label need not be immediately followed by a
3684 colon, but instead must start in column zero. Only one label may be defined on
3685 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3686 provides a special directive @code{.label} for defining labels more flexibly.
3689 @node Setting Symbols
3690 @section Giving Symbols Other Values
3692 @cindex assigning values to symbols
3693 @cindex symbol values, assigning
3694 A symbol can be given an arbitrary value by writing a symbol, followed
3695 by an equals sign @samp{=}, followed by an expression
3696 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3697 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3698 equals sign @samp{=}@samp{=} here represents an equivalent of the
3699 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3702 Blackfin does not support symbol assignment with @samp{=}.
3706 @section Symbol Names
3708 @cindex symbol names
3709 @cindex names, symbol
3710 @ifclear SPECIAL-SYMS
3711 Symbol names begin with a letter or with one of @samp{._}. On most
3712 machines, you can also use @code{$} in symbol names; exceptions are
3713 noted in @ref{Machine Dependencies}. That character may be followed by any
3714 string of digits, letters, dollar signs (unless otherwise noted for a
3715 particular target machine), and underscores.
3719 Symbol names begin with a letter or with one of @samp{._}. On the
3720 Renesas SH you can also use @code{$} in symbol names. That
3721 character may be followed by any string of digits, letters, dollar signs (save
3722 on the H8/300), and underscores.
3726 Case of letters is significant: @code{foo} is a different symbol name
3729 Symbol names do not start with a digit. An exception to this rule is made for
3730 Local Labels. See below.
3732 Multibyte characters are supported. To generate a symbol name containing
3733 multibyte characters enclose it within double quotes and use escape codes. cf
3734 @xref{Strings}. Generating a multibyte symbol name from a label is not
3735 currently supported.
3737 Each symbol has exactly one name. Each name in an assembly language program
3738 refers to exactly one symbol. You may use that symbol name any number of times
3741 @subheading Local Symbol Names
3743 @cindex local symbol names
3744 @cindex symbol names, local
3745 A local symbol is any symbol beginning with certain local label prefixes.
3746 By default, the local label prefix is @samp{.L} for ELF systems or
3747 @samp{L} for traditional a.out systems, but each target may have its own
3748 set of local label prefixes.
3750 On the HPPA local symbols begin with @samp{L$}.
3753 Local symbols are defined and used within the assembler, but they are
3754 normally not saved in object files. Thus, they are not visible when debugging.
3755 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols})
3756 to retain the local symbols in the object files.
3758 @subheading Local Labels
3760 @cindex local labels
3761 @cindex temporary symbol names
3762 @cindex symbol names, temporary
3763 Local labels are different from local symbols. Local labels help compilers and
3764 programmers use names temporarily. They create symbols which are guaranteed to
3765 be unique over the entire scope of the input source code and which can be
3766 referred to by a simple notation. To define a local label, write a label of
3767 the form @samp{@b{N}:} (where @b{N} represents any non-negative integer).
3768 To refer to the most recent previous definition of that label write
3769 @samp{@b{N}b}, using the same number as when you defined the label. To refer
3770 to the next definition of a local label, write @samp{@b{N}f}. The @samp{b}
3771 stands for ``backwards'' and the @samp{f} stands for ``forwards''.
3773 There is no restriction on how you can use these labels, and you can reuse them
3774 too. So that it is possible to repeatedly define the same local label (using
3775 the same number @samp{@b{N}}), although you can only refer to the most recently
3776 defined local label of that number (for a backwards reference) or the next
3777 definition of a specific local label for a forward reference. It is also worth
3778 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3779 implemented in a slightly more efficient manner than the others.
3790 Which is the equivalent of:
3793 label_1: branch label_3
3794 label_2: branch label_1
3795 label_3: branch label_4
3796 label_4: branch label_3
3799 Local label names are only a notational device. They are immediately
3800 transformed into more conventional symbol names before the assembler uses them.
3801 The symbol names are stored in the symbol table, appear in error messages, and
3802 are optionally emitted to the object file. The names are constructed using
3806 @item @emph{local label prefix}
3807 All local symbols begin with the system-specific local label prefix.
3808 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3809 that start with the local label prefix. These labels are
3810 used for symbols you are never intended to see. If you use the
3811 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3812 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3813 you may use them in debugging.
3816 This is the number that was used in the local label definition. So if the
3817 label is written @samp{55:} then the number is @samp{55}.
3820 This unusual character is included so you do not accidentally invent a symbol
3821 of the same name. The character has ASCII value of @samp{\002} (control-B).
3823 @item @emph{ordinal number}
3824 This is a serial number to keep the labels distinct. The first definition of
3825 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3826 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3827 the number @samp{1} and its 15th definition gets @samp{15} as well.
3830 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3831 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3833 @subheading Dollar Local Labels
3834 @cindex dollar local symbols
3836 On some targets @code{@value{AS}} also supports an even more local form of
3837 local labels called dollar labels. These labels go out of scope (i.e., they
3838 become undefined) as soon as a non-local label is defined. Thus they remain
3839 valid for only a small region of the input source code. Normal local labels,
3840 by contrast, remain in scope for the entire file, or until they are redefined
3841 by another occurrence of the same local label.
3843 Dollar labels are defined in exactly the same way as ordinary local labels,
3844 except that they have a dollar sign suffix to their numeric value, e.g.,
3847 They can also be distinguished from ordinary local labels by their transformed
3848 names which use ASCII character @samp{\001} (control-A) as the magic character
3849 to distinguish them from ordinary labels. For example, the fifth definition of
3850 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3853 @section The Special Dot Symbol
3855 @cindex dot (symbol)
3856 @cindex @code{.} (symbol)
3857 @cindex current address
3858 @cindex location counter
3859 The special symbol @samp{.} refers to the current address that
3860 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3861 .long .} defines @code{melvin} to contain its own address.
3862 Assigning a value to @code{.} is treated the same as a @code{.org}
3864 @ifclear no-space-dir
3865 Thus, the expression @samp{.=.+4} is the same as saying
3869 @node Symbol Attributes
3870 @section Symbol Attributes
3872 @cindex symbol attributes
3873 @cindex attributes, symbol
3874 Every symbol has, as well as its name, the attributes ``Value'' and
3875 ``Type''. Depending on output format, symbols can also have auxiliary
3878 The detailed definitions are in @file{a.out.h}.
3881 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3882 all these attributes, and probably won't warn you. This makes the
3883 symbol an externally defined symbol, which is generally what you
3887 * Symbol Value:: Value
3888 * Symbol Type:: Type
3890 * a.out Symbols:: Symbol Attributes: @code{a.out}
3893 * COFF Symbols:: Symbol Attributes for COFF
3896 * SOM Symbols:: Symbol Attributes for SOM
3903 @cindex value of a symbol
3904 @cindex symbol value
3905 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3906 location in the text, data, bss or absolute sections the value is the
3907 number of addresses from the start of that section to the label.
3908 Naturally for text, data and bss sections the value of a symbol changes
3909 as @code{@value{LD}} changes section base addresses during linking. Absolute
3910 symbols' values do not change during linking: that is why they are
3913 The value of an undefined symbol is treated in a special way. If it is
3914 0 then the symbol is not defined in this assembler source file, and
3915 @code{@value{LD}} tries to determine its value from other files linked into the
3916 same program. You make this kind of symbol simply by mentioning a symbol
3917 name without defining it. A non-zero value represents a @code{.comm}
3918 common declaration. The value is how much common storage to reserve, in
3919 bytes (addresses). The symbol refers to the first address of the
3925 @cindex type of a symbol
3927 The type attribute of a symbol contains relocation (section)
3928 information, any flag settings indicating that a symbol is external, and
3929 (optionally), other information for linkers and debuggers. The exact
3930 format depends on the object-code output format in use.
3934 @subsection Symbol Attributes: @code{a.out}
3936 @cindex @code{a.out} symbol attributes
3937 @cindex symbol attributes, @code{a.out}
3940 * Symbol Desc:: Descriptor
3941 * Symbol Other:: Other
3945 @subsubsection Descriptor
3947 @cindex descriptor, of @code{a.out} symbol
3948 This is an arbitrary 16-bit value. You may establish a symbol's
3949 descriptor value by using a @code{.desc} statement
3950 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3951 @command{@value{AS}}.
3954 @subsubsection Other
3956 @cindex other attribute, of @code{a.out} symbol
3957 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3962 @subsection Symbol Attributes for COFF
3964 @cindex COFF symbol attributes
3965 @cindex symbol attributes, COFF
3967 The COFF format supports a multitude of auxiliary symbol attributes;
3968 like the primary symbol attributes, they are set between @code{.def} and
3969 @code{.endef} directives.
3971 @subsubsection Primary Attributes
3973 @cindex primary attributes, COFF symbols
3974 The symbol name is set with @code{.def}; the value and type,
3975 respectively, with @code{.val} and @code{.type}.
3977 @subsubsection Auxiliary Attributes
3979 @cindex auxiliary attributes, COFF symbols
3980 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3981 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3982 table information for COFF.
3987 @subsection Symbol Attributes for SOM
3989 @cindex SOM symbol attributes
3990 @cindex symbol attributes, SOM
3992 The SOM format for the HPPA supports a multitude of symbol attributes set with
3993 the @code{.EXPORT} and @code{.IMPORT} directives.
3995 The attributes are described in @cite{HP9000 Series 800 Assembly
3996 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3997 @code{EXPORT} assembler directive documentation.
4001 @chapter Expressions
4005 @cindex numeric values
4006 An @dfn{expression} specifies an address or numeric value.
4007 Whitespace may precede and/or follow an expression.
4009 The result of an expression must be an absolute number, or else an offset into
4010 a particular section. If an expression is not absolute, and there is not
4011 enough information when @command{@value{AS}} sees the expression to know its
4012 section, a second pass over the source program might be necessary to interpret
4013 the expression---but the second pass is currently not implemented.
4014 @command{@value{AS}} aborts with an error message in this situation.
4017 * Empty Exprs:: Empty Expressions
4018 * Integer Exprs:: Integer Expressions
4022 @section Empty Expressions
4024 @cindex empty expressions
4025 @cindex expressions, empty
4026 An empty expression has no value: it is just whitespace or null.
4027 Wherever an absolute expression is required, you may omit the
4028 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
4029 is compatible with other assemblers.
4032 @section Integer Expressions
4034 @cindex integer expressions
4035 @cindex expressions, integer
4036 An @dfn{integer expression} is one or more @emph{arguments} delimited
4037 by @emph{operators}.
4040 * Arguments:: Arguments
4041 * Operators:: Operators
4042 * Prefix Ops:: Prefix Operators
4043 * Infix Ops:: Infix Operators
4047 @subsection Arguments
4049 @cindex expression arguments
4050 @cindex arguments in expressions
4051 @cindex operands in expressions
4052 @cindex arithmetic operands
4053 @dfn{Arguments} are symbols, numbers or subexpressions. In other
4054 contexts arguments are sometimes called ``arithmetic operands''. In
4055 this manual, to avoid confusing them with the ``instruction operands'' of
4056 the machine language, we use the term ``argument'' to refer to parts of
4057 expressions only, reserving the word ``operand'' to refer only to machine
4058 instruction operands.
4060 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
4061 @var{section} is one of text, data, bss, absolute,
4062 or undefined. @var{NNN} is a signed, 2's complement 32 bit
4065 Numbers are usually integers.
4067 A number can be a flonum or bignum. In this case, you are warned
4068 that only the low order 32 bits are used, and @command{@value{AS}} pretends
4069 these 32 bits are an integer. You may write integer-manipulating
4070 instructions that act on exotic constants, compatible with other
4073 @cindex subexpressions
4074 Subexpressions are a left parenthesis @samp{(} followed by an integer
4075 expression, followed by a right parenthesis @samp{)}; or a prefix
4076 operator followed by an argument.
4079 @subsection Operators
4081 @cindex operators, in expressions
4082 @cindex arithmetic functions
4083 @cindex functions, in expressions
4084 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
4085 operators are followed by an argument. Infix operators appear
4086 between their arguments. Operators may be preceded and/or followed by
4090 @subsection Prefix Operator
4092 @cindex prefix operators
4093 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
4094 one argument, which must be absolute.
4096 @c the tex/end tex stuff surrounding this small table is meant to make
4097 @c it align, on the printed page, with the similar table in the next
4098 @c section (which is inside an enumerate).
4100 \global\advance\leftskip by \itemindent
4105 @dfn{Negation}. Two's complement negation.
4107 @dfn{Complementation}. Bitwise not.
4111 \global\advance\leftskip by -\itemindent
4115 @subsection Infix Operators
4117 @cindex infix operators
4118 @cindex operators, permitted arguments
4119 @dfn{Infix operators} take two arguments, one on either side. Operators
4120 have precedence, but operations with equal precedence are performed left
4121 to right. Apart from @code{+} or @option{-}, both arguments must be
4122 absolute, and the result is absolute.
4125 @cindex operator precedence
4126 @cindex precedence of operators
4133 @dfn{Multiplication}.
4136 @dfn{Division}. Truncation is the same as the C operator @samp{/}
4142 @dfn{Shift Left}. Same as the C operator @samp{<<}.
4145 @dfn{Shift Right}. Same as the C operator @samp{>>}.
4149 Intermediate precedence
4154 @dfn{Bitwise Inclusive Or}.
4160 @dfn{Bitwise Exclusive Or}.
4163 @dfn{Bitwise Or Not}.
4170 @cindex addition, permitted arguments
4171 @cindex plus, permitted arguments
4172 @cindex arguments for addition
4174 @dfn{Addition}. If either argument is absolute, the result has the section of
4175 the other argument. You may not add together arguments from different
4178 @cindex subtraction, permitted arguments
4179 @cindex minus, permitted arguments
4180 @cindex arguments for subtraction
4182 @dfn{Subtraction}. If the right argument is absolute, the
4183 result has the section of the left argument.
4184 If both arguments are in the same section, the result is absolute.
4185 You may not subtract arguments from different sections.
4186 @c FIXME is there still something useful to say about undefined - undefined ?
4188 @cindex comparison expressions
4189 @cindex expressions, comparison
4194 @dfn{Is Not Equal To}
4198 @dfn{Is Greater Than}
4200 @dfn{Is Greater Than Or Equal To}
4202 @dfn{Is Less Than Or Equal To}
4204 The comparison operators can be used as infix operators. A true results has a
4205 value of -1 whereas a false result has a value of 0. Note, these operators
4206 perform signed comparisons.
4209 @item Lowest Precedence
4218 These two logical operations can be used to combine the results of sub
4219 expressions. Note, unlike the comparison operators a true result returns a
4220 value of 1 but a false results does still return 0. Also note that the logical
4221 or operator has a slightly lower precedence than logical and.
4226 In short, it's only meaningful to add or subtract the @emph{offsets} in an
4227 address; you can only have a defined section in one of the two arguments.
4230 @chapter Assembler Directives
4232 @cindex directives, machine independent
4233 @cindex pseudo-ops, machine independent
4234 @cindex machine independent directives
4235 All assembler directives have names that begin with a period (@samp{.}).
4236 The names are case insensitive for most targets, and usually written
4239 This chapter discusses directives that are available regardless of the
4240 target machine configuration for the @sc{gnu} assembler.
4242 Some machine configurations provide additional directives.
4243 @xref{Machine Dependencies}.
4246 @ifset machine-directives
4247 @xref{Machine Dependencies}, for additional directives.
4252 * Abort:: @code{.abort}
4254 * ABORT (COFF):: @code{.ABORT}
4257 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
4258 * Altmacro:: @code{.altmacro}
4259 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4260 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4261 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
4262 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, etc
4263 * Byte:: @code{.byte @var{expressions}}
4264 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4265 * Comm:: @code{.comm @var{symbol} , @var{length} }
4266 * Data:: @code{.data @var{subsection}}
4267 * Dc:: @code{.dc[@var{size}] @var{expressions}}
4268 * Dcb:: @code{.dcb[@var{size}] @var{number} [,@var{fill}]}
4269 * Ds:: @code{.ds[@var{size}] @var{number} [,@var{fill}]}
4271 * Def:: @code{.def @var{name}}
4274 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4280 * Double:: @code{.double @var{flonums}}
4281 * Eject:: @code{.eject}
4282 * Else:: @code{.else}
4283 * Elseif:: @code{.elseif}
4286 * Endef:: @code{.endef}
4289 * Endfunc:: @code{.endfunc}
4290 * Endif:: @code{.endif}
4291 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4292 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4293 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4295 * Error:: @code{.error @var{string}}
4296 * Exitm:: @code{.exitm}
4297 * Extern:: @code{.extern}
4298 * Fail:: @code{.fail}
4299 * File:: @code{.file}
4300 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4301 * Float:: @code{.float @var{flonums}}
4302 * Func:: @code{.func}
4303 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4305 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4306 * Hidden:: @code{.hidden @var{names}}
4309 * hword:: @code{.hword @var{expressions}}
4310 * Ident:: @code{.ident}
4311 * If:: @code{.if @var{absolute expression}}
4312 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4313 * Include:: @code{.include "@var{file}"}
4314 * Int:: @code{.int @var{expressions}}
4316 * Internal:: @code{.internal @var{names}}
4319 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4320 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4321 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4322 * Lflags:: @code{.lflags}
4323 @ifclear no-line-dir
4324 * Line:: @code{.line @var{line-number}}
4327 * Linkonce:: @code{.linkonce [@var{type}]}
4328 * List:: @code{.list}
4329 * Ln:: @code{.ln @var{line-number}}
4330 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4331 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4333 * Local:: @code{.local @var{names}}
4336 * Long:: @code{.long @var{expressions}}
4338 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4341 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4342 * MRI:: @code{.mri @var{val}}
4343 * Noaltmacro:: @code{.noaltmacro}
4344 * Nolist:: @code{.nolist}
4345 * Nops:: @code{.nops @var{size}[, @var{control}]}
4346 * Octa:: @code{.octa @var{bignums}}
4347 * Offset:: @code{.offset @var{loc}}
4348 * Org:: @code{.org @var{new-lc}, @var{fill}}
4349 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4351 * PopSection:: @code{.popsection}
4352 * Previous:: @code{.previous}
4355 * Print:: @code{.print @var{string}}
4357 * Protected:: @code{.protected @var{names}}
4360 * Psize:: @code{.psize @var{lines}, @var{columns}}
4361 * Purgem:: @code{.purgem @var{name}}
4363 * PushSection:: @code{.pushsection @var{name}}
4366 * Quad:: @code{.quad @var{bignums}}
4367 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4368 * Rept:: @code{.rept @var{count}}
4369 * Sbttl:: @code{.sbttl "@var{subheading}"}
4371 * Scl:: @code{.scl @var{class}}
4374 * Section:: @code{.section @var{name}[, @var{flags}]}
4377 * Set:: @code{.set @var{symbol}, @var{expression}}
4378 * Short:: @code{.short @var{expressions}}
4379 * Single:: @code{.single @var{flonums}}
4381 * Size:: @code{.size [@var{name} , @var{expression}]}
4383 @ifclear no-space-dir
4384 * Skip:: @code{.skip @var{size} [,@var{fill}]}
4387 * Sleb128:: @code{.sleb128 @var{expressions}}
4388 @ifclear no-space-dir
4389 * Space:: @code{.space @var{size} [,@var{fill}]}
4392 * Stab:: @code{.stabd, .stabn, .stabs}
4395 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4396 * Struct:: @code{.struct @var{expression}}
4398 * SubSection:: @code{.subsection}
4399 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4403 * Tag:: @code{.tag @var{structname}}
4406 * Text:: @code{.text @var{subsection}}
4407 * Title:: @code{.title "@var{heading}"}
4409 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4412 * Uleb128:: @code{.uleb128 @var{expressions}}
4414 * Val:: @code{.val @var{addr}}
4418 * Version:: @code{.version "@var{string}"}
4419 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4420 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4423 * Warning:: @code{.warning @var{string}}
4424 * Weak:: @code{.weak @var{names}}
4425 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4426 * Word:: @code{.word @var{expressions}}
4427 @ifclear no-space-dir
4428 * Zero:: @code{.zero @var{size}}
4431 * 2byte:: @code{.2byte @var{expressions}}
4432 * 4byte:: @code{.4byte @var{expressions}}
4433 * 8byte:: @code{.8byte @var{bignums}}
4435 * Deprecated:: Deprecated Directives
4439 @section @code{.abort}
4441 @cindex @code{abort} directive
4442 @cindex stopping the assembly
4443 This directive stops the assembly immediately. It is for
4444 compatibility with other assemblers. The original idea was that the
4445 assembly language source would be piped into the assembler. If the sender
4446 of the source quit, it could use this directive tells @command{@value{AS}} to
4447 quit also. One day @code{.abort} will not be supported.
4451 @section @code{.ABORT} (COFF)
4453 @cindex @code{ABORT} directive
4454 When producing COFF output, @command{@value{AS}} accepts this directive as a
4455 synonym for @samp{.abort}.
4460 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4462 @cindex padding the location counter
4463 @cindex @code{align} directive
4464 Pad the location counter (in the current subsection) to a particular storage
4465 boundary. The first expression (which must be absolute) is the alignment
4466 required, as described below.
4468 The second expression (also absolute) gives the fill value to be stored in the
4469 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4470 padding bytes are normally zero. However, on most systems, if the section is
4471 marked as containing code and the fill value is omitted, the space is filled
4472 with no-op instructions.
4474 The third expression is also absolute, and is also optional. If it is present,
4475 it is the maximum number of bytes that should be skipped by this alignment
4476 directive. If doing the alignment would require skipping more bytes than the
4477 specified maximum, then the alignment is not done at all. You can omit the
4478 fill value (the second argument) entirely by simply using two commas after the
4479 required alignment; this can be useful if you want the alignment to be filled
4480 with no-op instructions when appropriate.
4482 The way the required alignment is specified varies from system to system.
4483 For the arc, hppa, i386 using ELF, iq2000, m68k, or1k,
4484 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4485 alignment request in bytes. For example @samp{.align 8} advances
4486 the location counter until it is a multiple of 8. If the location counter
4487 is already a multiple of 8, no change is needed. For the tic54x, the
4488 first expression is the alignment request in words.
4490 For other systems, including ppc, i386 using a.out format, arm and
4491 strongarm, it is the
4492 number of low-order zero bits the location counter must have after
4493 advancement. For example @samp{.align 3} advances the location
4494 counter until it a multiple of 8. If the location counter is already a
4495 multiple of 8, no change is needed.
4497 This inconsistency is due to the different behaviors of the various
4498 native assemblers for these systems which GAS must emulate.
4499 GAS also provides @code{.balign} and @code{.p2align} directives,
4500 described later, which have a consistent behavior across all
4501 architectures (but are specific to GAS).
4504 @section @code{.altmacro}
4505 Enable alternate macro mode, enabling:
4508 @item LOCAL @var{name} [ , @dots{} ]
4509 One additional directive, @code{LOCAL}, is available. It is used to
4510 generate a string replacement for each of the @var{name} arguments, and
4511 replace any instances of @var{name} in each macro expansion. The
4512 replacement string is unique in the assembly, and different for each
4513 separate macro expansion. @code{LOCAL} allows you to write macros that
4514 define symbols, without fear of conflict between separate macro expansions.
4516 @item String delimiters
4517 You can write strings delimited in these other ways besides
4518 @code{"@var{string}"}:
4521 @item '@var{string}'
4522 You can delimit strings with single-quote characters.
4524 @item <@var{string}>
4525 You can delimit strings with matching angle brackets.
4528 @item single-character string escape
4529 To include any single character literally in a string (even if the
4530 character would otherwise have some special meaning), you can prefix the
4531 character with @samp{!} (an exclamation mark). For example, you can
4532 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4534 @item Expression results as strings
4535 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4536 and use the result as a string.
4540 @section @code{.ascii "@var{string}"}@dots{}
4542 @cindex @code{ascii} directive
4543 @cindex string literals
4544 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4545 separated by commas. It assembles each string (with no automatic
4546 trailing zero byte) into consecutive addresses.
4549 @section @code{.asciz "@var{string}"}@dots{}
4551 @cindex @code{asciz} directive
4552 @cindex zero-terminated strings
4553 @cindex null-terminated strings
4554 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4555 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4558 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4560 @cindex padding the location counter given number of bytes
4561 @cindex @code{balign} directive
4562 Pad the location counter (in the current subsection) to a particular
4563 storage boundary. The first expression (which must be absolute) is the
4564 alignment request in bytes. For example @samp{.balign 8} advances
4565 the location counter until it is a multiple of 8. If the location counter
4566 is already a multiple of 8, no change is needed.
4568 The second expression (also absolute) gives the fill value to be stored in the
4569 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4570 padding bytes are normally zero. However, on most systems, if the section is
4571 marked as containing code and the fill value is omitted, the space is filled
4572 with no-op instructions.
4574 The third expression is also absolute, and is also optional. If it is present,
4575 it is the maximum number of bytes that should be skipped by this alignment
4576 directive. If doing the alignment would require skipping more bytes than the
4577 specified maximum, then the alignment is not done at all. You can omit the
4578 fill value (the second argument) entirely by simply using two commas after the
4579 required alignment; this can be useful if you want the alignment to be filled
4580 with no-op instructions when appropriate.
4582 @cindex @code{balignw} directive
4583 @cindex @code{balignl} directive
4584 The @code{.balignw} and @code{.balignl} directives are variants of the
4585 @code{.balign} directive. The @code{.balignw} directive treats the fill
4586 pattern as a two byte word value. The @code{.balignl} directives treats the
4587 fill pattern as a four byte longword value. For example, @code{.balignw
4588 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4589 filled in with the value 0x368d (the exact placement of the bytes depends upon
4590 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4593 @node Bundle directives
4594 @section Bundle directives
4595 @subsection @code{.bundle_align_mode @var{abs-expr}}
4596 @cindex @code{bundle_align_mode} directive
4598 @cindex instruction bundle
4599 @cindex aligned instruction bundle
4600 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4601 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4602 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4603 disabled (which is the default state). If the argument it not zero, it
4604 gives the size of an instruction bundle as a power of two (as for the
4605 @code{.p2align} directive, @pxref{P2align}).
4607 For some targets, it's an ABI requirement that no instruction may span a
4608 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4609 instructions that starts on an aligned boundary. For example, if
4610 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4611 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4612 effect, no single instruction may span a boundary between bundles. If an
4613 instruction would start too close to the end of a bundle for the length of
4614 that particular instruction to fit within the bundle, then the space at the
4615 end of that bundle is filled with no-op instructions so the instruction
4616 starts in the next bundle. As a corollary, it's an error if any single
4617 instruction's encoding is longer than the bundle size.
4619 @subsection @code{.bundle_lock} and @code{.bundle_unlock}
4620 @cindex @code{bundle_lock} directive
4621 @cindex @code{bundle_unlock} directive
4622 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4623 allow explicit control over instruction bundle padding. These directives
4624 are only valid when @code{.bundle_align_mode} has been used to enable
4625 aligned instruction bundle mode. It's an error if they appear when
4626 @code{.bundle_align_mode} has not been used at all, or when the last
4627 directive was @w{@code{.bundle_align_mode 0}}.
4629 @cindex bundle-locked
4630 For some targets, it's an ABI requirement that certain instructions may
4631 appear only as part of specified permissible sequences of multiple
4632 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4633 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4634 instruction sequence. For purposes of aligned instruction bundle mode, a
4635 sequence starting with @code{.bundle_lock} and ending with
4636 @code{.bundle_unlock} is treated as a single instruction. That is, the
4637 entire sequence must fit into a single bundle and may not span a bundle
4638 boundary. If necessary, no-op instructions will be inserted before the
4639 first instruction of the sequence so that the whole sequence starts on an
4640 aligned bundle boundary. It's an error if the sequence is longer than the
4643 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4644 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4645 nested. That is, a second @code{.bundle_lock} directive before the next
4646 @code{.bundle_unlock} directive has no effect except that it must be
4647 matched by another closing @code{.bundle_unlock} so that there is the
4648 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4651 @section @code{.byte @var{expressions}}
4653 @cindex @code{byte} directive
4654 @cindex integers, one byte
4655 @code{.byte} expects zero or more expressions, separated by commas.
4656 Each expression is assembled into the next byte.
4658 @node CFI directives
4659 @section CFI directives
4660 @subsection @code{.cfi_sections @var{section_list}}
4661 @cindex @code{cfi_sections} directive
4662 @code{.cfi_sections} may be used to specify whether CFI directives
4663 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4664 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4665 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4666 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4667 directive is not used is @code{.cfi_sections .eh_frame}.
4669 On targets that support compact unwinding tables these can be generated
4670 by specifying @code{.eh_frame_entry} instead of @code{.eh_frame}.
4672 Some targets may support an additional name, such as @code{.c6xabi.exidx}
4673 which is used by the @value{TIC6X} target.
4675 The @code{.cfi_sections} directive can be repeated, with the same or different
4676 arguments, provided that CFI generation has not yet started. Once CFI
4677 generation has started however the section list is fixed and any attempts to
4678 redefine it will result in an error.
4680 @subsection @code{.cfi_startproc [simple]}
4681 @cindex @code{cfi_startproc} directive
4682 @code{.cfi_startproc} is used at the beginning of each function that
4683 should have an entry in @code{.eh_frame}. It initializes some internal
4684 data structures. Don't forget to close the function by
4685 @code{.cfi_endproc}.
4687 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4688 it also emits some architecture dependent initial CFI instructions.
4690 @subsection @code{.cfi_endproc}
4691 @cindex @code{cfi_endproc} directive
4692 @code{.cfi_endproc} is used at the end of a function where it closes its
4693 unwind entry previously opened by
4694 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4696 @subsection @code{.cfi_personality @var{encoding} [, @var{exp}]}
4697 @cindex @code{cfi_personality} directive
4698 @code{.cfi_personality} defines personality routine and its encoding.
4699 @var{encoding} must be a constant determining how the personality
4700 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4701 argument is not present, otherwise second argument should be
4702 a constant or a symbol name. When using indirect encodings,
4703 the symbol provided should be the location where personality
4704 can be loaded from, not the personality routine itself.
4705 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4706 no personality routine.
4708 @subsection @code{.cfi_personality_id @var{id}}
4709 @cindex @code{cfi_personality_id} directive
4710 @code{cfi_personality_id} defines a personality routine by its index as
4711 defined in a compact unwinding format.
4712 Only valid when generating compact EH frames (i.e.
4713 with @code{.cfi_sections eh_frame_entry}.
4715 @subsection @code{.cfi_fde_data [@var{opcode1} [, @dots{}]]}
4716 @cindex @code{cfi_fde_data} directive
4717 @code{cfi_fde_data} is used to describe the compact unwind opcodes to be
4718 used for the current function. These are emitted inline in the
4719 @code{.eh_frame_entry} section if small enough and there is no LSDA, or
4720 in the @code{.gnu.extab} section otherwise.
4721 Only valid when generating compact EH frames (i.e.
4722 with @code{.cfi_sections eh_frame_entry}.
4724 @subsection @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4725 @code{.cfi_lsda} defines LSDA and its encoding.
4726 @var{encoding} must be a constant determining how the LSDA
4727 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), the second
4728 argument is not present, otherwise the second argument should be a constant
4729 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4730 meaning that no LSDA is present.
4732 @subsection @code{.cfi_inline_lsda} [@var{align}]
4733 @code{.cfi_inline_lsda} marks the start of a LSDA data section and
4734 switches to the corresponding @code{.gnu.extab} section.
4735 Must be preceded by a CFI block containing a @code{.cfi_lsda} directive.
4736 Only valid when generating compact EH frames (i.e.
4737 with @code{.cfi_sections eh_frame_entry}.
4739 The table header and unwinding opcodes will be generated at this point,
4740 so that they are immediately followed by the LSDA data. The symbol
4741 referenced by the @code{.cfi_lsda} directive should still be defined
4742 in case a fallback FDE based encoding is used. The LSDA data is terminated
4743 by a section directive.
4745 The optional @var{align} argument specifies the alignment required.
4746 The alignment is specified as a power of two, as with the
4747 @code{.p2align} directive.
4749 @subsection @code{.cfi_def_cfa @var{register}, @var{offset}}
4750 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4751 address from @var{register} and add @var{offset} to it}.
4753 @subsection @code{.cfi_def_cfa_register @var{register}}
4754 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4755 now on @var{register} will be used instead of the old one. Offset
4758 @subsection @code{.cfi_def_cfa_offset @var{offset}}
4759 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4760 remains the same, but @var{offset} is new. Note that it is the
4761 absolute offset that will be added to a defined register to compute
4764 @subsection @code{.cfi_adjust_cfa_offset @var{offset}}
4765 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4766 value that is added/subtracted from the previous offset.
4768 @subsection @code{.cfi_offset @var{register}, @var{offset}}
4769 Previous value of @var{register} is saved at offset @var{offset} from
4772 @subsection @code{.cfi_val_offset @var{register}, @var{offset}}
4773 Previous value of @var{register} is CFA + @var{offset}.
4775 @subsection @code{.cfi_rel_offset @var{register}, @var{offset}}
4776 Previous value of @var{register} is saved at offset @var{offset} from
4777 the current CFA register. This is transformed to @code{.cfi_offset}
4778 using the known displacement of the CFA register from the CFA.
4779 This is often easier to use, because the number will match the
4780 code it's annotating.
4782 @subsection @code{.cfi_register @var{register1}, @var{register2}}
4783 Previous value of @var{register1} is saved in register @var{register2}.
4785 @subsection @code{.cfi_restore @var{register}}
4786 @code{.cfi_restore} says that the rule for @var{register} is now the
4787 same as it was at the beginning of the function, after all initial
4788 instruction added by @code{.cfi_startproc} were executed.
4790 @subsection @code{.cfi_undefined @var{register}}
4791 From now on the previous value of @var{register} can't be restored anymore.
4793 @subsection @code{.cfi_same_value @var{register}}
4794 Current value of @var{register} is the same like in the previous frame,
4795 i.e. no restoration needed.
4797 @subsection @code{.cfi_remember_state} and @code{.cfi_restore_state}
4798 @code{.cfi_remember_state} pushes the set of rules for every register onto an
4799 implicit stack, while @code{.cfi_restore_state} pops them off the stack and
4800 places them in the current row. This is useful for situations where you have
4801 multiple @code{.cfi_*} directives that need to be undone due to the control
4802 flow of the program. For example, we could have something like this (assuming
4803 the CFA is the value of @code{rbp}):
4813 .cfi_def_cfa %rsp, 8
4816 /* Do something else */
4819 Here, we want the @code{.cfi} directives to affect only the rows corresponding
4820 to the instructions before @code{label}. This means we'd have to add multiple
4821 @code{.cfi} directives after @code{label} to recreate the original save
4822 locations of the registers, as well as setting the CFA back to the value of
4823 @code{rbp}. This would be clumsy, and result in a larger binary size. Instead,
4835 .cfi_def_cfa %rsp, 8
4839 /* Do something else */
4842 That way, the rules for the instructions after @code{label} will be the same
4843 as before the first @code{.cfi_restore} without having to use multiple
4844 @code{.cfi} directives.
4846 @subsection @code{.cfi_return_column @var{register}}
4847 Change return column @var{register}, i.e. the return address is either
4848 directly in @var{register} or can be accessed by rules for @var{register}.
4850 @subsection @code{.cfi_signal_frame}
4851 Mark current function as signal trampoline.
4853 @subsection @code{.cfi_window_save}
4854 SPARC register window has been saved.
4856 @subsection @code{.cfi_escape} @var{expression}[, @dots{}]
4857 Allows the user to add arbitrary bytes to the unwind info. One
4858 might use this to add OS-specific CFI opcodes, or generic CFI
4859 opcodes that GAS does not yet support.
4861 @subsection @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4862 The current value of @var{register} is @var{label}. The value of @var{label}
4863 will be encoded in the output file according to @var{encoding}; see the
4864 description of @code{.cfi_personality} for details on this encoding.
4866 The usefulness of equating a register to a fixed label is probably
4867 limited to the return address register. Here, it can be useful to
4868 mark a code segment that has only one return address which is reached
4869 by a direct branch and no copy of the return address exists in memory
4870 or another register.
4873 @section @code{.comm @var{symbol} , @var{length} }
4875 @cindex @code{comm} directive
4876 @cindex symbol, common
4877 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4878 common symbol in one object file may be merged with a defined or common symbol
4879 of the same name in another object file. If @code{@value{LD}} does not see a
4880 definition for the symbol--just one or more common symbols--then it will
4881 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4882 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4883 the same name, and they do not all have the same size, it will allocate space
4884 using the largest size.
4887 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4888 an optional third argument. This is the desired alignment of the symbol,
4889 specified for ELF as a byte boundary (for example, an alignment of 16 means
4890 that the least significant 4 bits of the address should be zero), and for PE
4891 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4892 boundary). The alignment must be an absolute expression, and it must be a
4893 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4894 common symbol, it will use the alignment when placing the symbol. If no
4895 alignment is specified, @command{@value{AS}} will set the alignment to the
4896 largest power of two less than or equal to the size of the symbol, up to a
4897 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4898 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4899 @samp{--section-alignment} option; image file sections in PE are aligned to
4900 multiples of 4096, which is far too large an alignment for ordinary variables.
4901 It is rather the default alignment for (non-debug) sections within object
4902 (@samp{*.o}) files, which are less strictly aligned.}.
4906 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4907 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4911 @section @code{.data @var{subsection}}
4912 @cindex @code{data} directive
4914 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4915 end of the data subsection numbered @var{subsection} (which is an
4916 absolute expression). If @var{subsection} is omitted, it defaults
4920 @section @code{.dc[@var{size}] @var{expressions}}
4921 @cindex @code{dc} directive
4923 The @code{.dc} directive expects zero or more @var{expressions} separated by
4924 commas. These expressions are evaluated and their values inserted into the
4925 current section. The size of the emitted value depends upon the suffix to the
4926 @code{.dc} directive:
4930 Emits N-bit values, where N is the size of an address on the target system.
4934 Emits double precision floating-point values.
4936 Emits 32-bit values.
4938 Emits single precision floating-point values.
4940 Emits 16-bit values.
4941 Note - this is true even on targets where the @code{.word} directive would emit
4944 Emits long double precision floating-point values.
4947 If no suffix is used then @samp{.w} is assumed.
4949 The byte ordering is target dependent, as is the size and format of floating
4953 @section @code{.dcb[@var{size}] @var{number} [,@var{fill}]}
4954 @cindex @code{dcb} directive
4955 This directive emits @var{number} copies of @var{fill}, each of @var{size}
4956 bytes. Both @var{number} and @var{fill} are absolute expressions. If the
4957 comma and @var{fill} are omitted, @var{fill} is assumed to be zero. The
4958 @var{size} suffix, if present, must be one of:
4962 Emits single byte values.
4964 Emits double-precision floating point values.
4966 Emits 4-byte values.
4968 Emits single-precision floating point values.
4970 Emits 2-byte values.
4972 Emits long double-precision floating point values.
4975 If the @var{size} suffix is omitted then @samp{.w} is assumed.
4977 The byte ordering is target dependent, as is the size and format of floating
4981 @section @code{.ds[@var{size}] @var{number} [,@var{fill}]}
4982 @cindex @code{ds} directive
4983 This directive emits @var{number} copies of @var{fill}, each of @var{size}
4984 bytes. Both @var{number} and @var{fill} are absolute expressions. If the
4985 comma and @var{fill} are omitted, @var{fill} is assumed to be zero. The
4986 @var{size} suffix, if present, must be one of:
4990 Emits single byte values.
4992 Emits 8-byte values.
4994 Emits 4-byte values.
4996 Emits 12-byte values.
4998 Emits 4-byte values.
5000 Emits 2-byte values.
5002 Emits 12-byte values.
5005 Note - unlike the @code{.dcb} directive the @samp{.d}, @samp{.s} and @samp{.x}
5006 suffixes do not indicate that floating-point values are to be inserted.
5008 If the @var{size} suffix is omitted then @samp{.w} is assumed.
5010 The byte ordering is target dependent.
5015 @section @code{.def @var{name}}
5017 @cindex @code{def} directive
5018 @cindex COFF symbols, debugging
5019 @cindex debugging COFF symbols
5020 Begin defining debugging information for a symbol @var{name}; the
5021 definition extends until the @code{.endef} directive is encountered.
5026 @section @code{.desc @var{symbol}, @var{abs-expression}}
5028 @cindex @code{desc} directive
5029 @cindex COFF symbol descriptor
5030 @cindex symbol descriptor, COFF
5031 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
5032 to the low 16 bits of an absolute expression.
5035 The @samp{.desc} directive is not available when @command{@value{AS}} is
5036 configured for COFF output; it is only for @code{a.out} or @code{b.out}
5037 object format. For the sake of compatibility, @command{@value{AS}} accepts
5038 it, but produces no output, when configured for COFF.
5044 @section @code{.dim}
5046 @cindex @code{dim} directive
5047 @cindex COFF auxiliary symbol information
5048 @cindex auxiliary symbol information, COFF
5049 This directive is generated by compilers to include auxiliary debugging
5050 information in the symbol table. It is only permitted inside
5051 @code{.def}/@code{.endef} pairs.
5055 @section @code{.double @var{flonums}}
5057 @cindex @code{double} directive
5058 @cindex floating point numbers (double)
5059 @code{.double} expects zero or more flonums, separated by commas. It
5060 assembles floating point numbers.
5062 The exact kind of floating point numbers emitted depends on how
5063 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
5067 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
5068 in @sc{ieee} format.
5073 @section @code{.eject}
5075 @cindex @code{eject} directive
5076 @cindex new page, in listings
5077 @cindex page, in listings
5078 @cindex listing control: new page
5079 Force a page break at this point, when generating assembly listings.
5082 @section @code{.else}
5084 @cindex @code{else} directive
5085 @code{.else} is part of the @command{@value{AS}} support for conditional
5086 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
5087 of code to be assembled if the condition for the preceding @code{.if}
5091 @section @code{.elseif}
5093 @cindex @code{elseif} directive
5094 @code{.elseif} is part of the @command{@value{AS}} support for conditional
5095 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
5096 @code{.if} block that would otherwise fill the entire @code{.else} section.
5099 @section @code{.end}
5101 @cindex @code{end} directive
5102 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
5103 process anything in the file past the @code{.end} directive.
5107 @section @code{.endef}
5109 @cindex @code{endef} directive
5110 This directive flags the end of a symbol definition begun with
5115 @section @code{.endfunc}
5116 @cindex @code{endfunc} directive
5117 @code{.endfunc} marks the end of a function specified with @code{.func}.
5120 @section @code{.endif}
5122 @cindex @code{endif} directive
5123 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
5124 it marks the end of a block of code that is only assembled
5125 conditionally. @xref{If,,@code{.if}}.
5128 @section @code{.equ @var{symbol}, @var{expression}}
5130 @cindex @code{equ} directive
5131 @cindex assigning values to symbols
5132 @cindex symbols, assigning values to
5133 This directive sets the value of @var{symbol} to @var{expression}.
5134 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
5137 The syntax for @code{equ} on the HPPA is
5138 @samp{@var{symbol} .equ @var{expression}}.
5142 The syntax for @code{equ} on the Z80 is
5143 @samp{@var{symbol} equ @var{expression}}.
5144 On the Z80 it is an error if @var{symbol} is already defined,
5145 but the symbol is not protected from later redefinition.
5146 Compare @ref{Equiv}.
5150 @section @code{.equiv @var{symbol}, @var{expression}}
5151 @cindex @code{equiv} directive
5152 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
5153 the assembler will signal an error if @var{symbol} is already defined. Note a
5154 symbol which has been referenced but not actually defined is considered to be
5157 Except for the contents of the error message, this is roughly equivalent to
5164 plus it protects the symbol from later redefinition.
5167 @section @code{.eqv @var{symbol}, @var{expression}}
5168 @cindex @code{eqv} directive
5169 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
5170 evaluate the expression or any part of it immediately. Instead each time
5171 the resulting symbol is used in an expression, a snapshot of its current
5175 @section @code{.err}
5176 @cindex @code{err} directive
5177 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
5178 message and, unless the @option{-Z} option was used, it will not generate an
5179 object file. This can be used to signal an error in conditionally compiled code.
5182 @section @code{.error "@var{string}"}
5183 @cindex error directive
5185 Similarly to @code{.err}, this directive emits an error, but you can specify a
5186 string that will be emitted as the error message. If you don't specify the
5187 message, it defaults to @code{".error directive invoked in source file"}.
5188 @xref{Errors, ,Error and Warning Messages}.
5191 .error "This code has not been assembled and tested."
5195 @section @code{.exitm}
5196 Exit early from the current macro definition. @xref{Macro}.
5199 @section @code{.extern}
5201 @cindex @code{extern} directive
5202 @code{.extern} is accepted in the source program---for compatibility
5203 with other assemblers---but it is ignored. @command{@value{AS}} treats
5204 all undefined symbols as external.
5207 @section @code{.fail @var{expression}}
5209 @cindex @code{fail} directive
5210 Generates an error or a warning. If the value of the @var{expression} is 500
5211 or more, @command{@value{AS}} will print a warning message. If the value is less
5212 than 500, @command{@value{AS}} will print an error message. The message will
5213 include the value of @var{expression}. This can occasionally be useful inside
5214 complex nested macros or conditional assembly.
5217 @section @code{.file}
5218 @cindex @code{file} directive
5220 @ifclear no-file-dir
5221 There are two different versions of the @code{.file} directive. Targets
5222 that support DWARF2 line number information use the DWARF2 version of
5223 @code{.file}. Other targets use the default version.
5225 @subheading Default Version
5227 @cindex logical file name
5228 @cindex file name, logical
5229 This version of the @code{.file} directive tells @command{@value{AS}} that we
5230 are about to start a new logical file. The syntax is:
5236 @var{string} is the new file name. In general, the filename is
5237 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
5238 to specify an empty file name, you must give the quotes--@code{""}. This
5239 statement may go away in future: it is only recognized to be compatible with
5240 old @command{@value{AS}} programs.
5242 @subheading DWARF2 Version
5245 When emitting DWARF2 line number information, @code{.file} assigns filenames
5246 to the @code{.debug_line} file name table. The syntax is:
5249 .file @var{fileno} @var{filename}
5252 The @var{fileno} operand should be a unique positive integer to use as the
5253 index of the entry in the table. The @var{filename} operand is a C string
5256 The detail of filename indices is exposed to the user because the filename
5257 table is shared with the @code{.debug_info} section of the DWARF2 debugging
5258 information, and thus the user must know the exact indices that table
5262 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
5264 @cindex @code{fill} directive
5265 @cindex writing patterns in memory
5266 @cindex patterns, writing in memory
5267 @var{repeat}, @var{size} and @var{value} are absolute expressions.
5268 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
5269 may be zero or more. @var{Size} may be zero or more, but if it is
5270 more than 8, then it is deemed to have the value 8, compatible with
5271 other people's assemblers. The contents of each @var{repeat} bytes
5272 is taken from an 8-byte number. The highest order 4 bytes are
5273 zero. The lowest order 4 bytes are @var{value} rendered in the
5274 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
5275 Each @var{size} bytes in a repetition is taken from the lowest order
5276 @var{size} bytes of this number. Again, this bizarre behavior is
5277 compatible with other people's assemblers.
5279 @var{size} and @var{value} are optional.
5280 If the second comma and @var{value} are absent, @var{value} is
5281 assumed zero. If the first comma and following tokens are absent,
5282 @var{size} is assumed to be 1.
5285 @section @code{.float @var{flonums}}
5287 @cindex floating point numbers (single)
5288 @cindex @code{float} directive
5289 This directive assembles zero or more flonums, separated by commas. It
5290 has the same effect as @code{.single}.
5292 The exact kind of floating point numbers emitted depends on how
5293 @command{@value{AS}} is configured.
5294 @xref{Machine Dependencies}.
5298 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
5299 in @sc{ieee} format.
5304 @section @code{.func @var{name}[,@var{label}]}
5305 @cindex @code{func} directive
5306 @code{.func} emits debugging information to denote function @var{name}, and
5307 is ignored unless the file is assembled with debugging enabled.
5308 Only @samp{--gstabs[+]} is currently supported.
5309 @var{label} is the entry point of the function and if omitted @var{name}
5310 prepended with the @samp{leading char} is used.
5311 @samp{leading char} is usually @code{_} or nothing, depending on the target.
5312 All functions are currently defined to have @code{void} return type.
5313 The function must be terminated with @code{.endfunc}.
5316 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
5318 @cindex @code{global} directive
5319 @cindex symbol, making visible to linker
5320 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
5321 @var{symbol} in your partial program, its value is made available to
5322 other partial programs that are linked with it. Otherwise,
5323 @var{symbol} takes its attributes from a symbol of the same name
5324 from another file linked into the same program.
5326 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
5327 compatibility with other assemblers.
5330 On the HPPA, @code{.global} is not always enough to make it accessible to other
5331 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
5332 @xref{HPPA Directives, ,HPPA Assembler Directives}.
5337 @section @code{.gnu_attribute @var{tag},@var{value}}
5338 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
5341 @section @code{.hidden @var{names}}
5343 @cindex @code{hidden} directive
5345 This is one of the ELF visibility directives. The other two are
5346 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
5347 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5349 This directive overrides the named symbols default visibility (which is set by
5350 their binding: local, global or weak). The directive sets the visibility to
5351 @code{hidden} which means that the symbols are not visible to other components.
5352 Such symbols are always considered to be @code{protected} as well.
5356 @section @code{.hword @var{expressions}}
5358 @cindex @code{hword} directive
5359 @cindex integers, 16-bit
5360 @cindex numbers, 16-bit
5361 @cindex sixteen bit integers
5362 This expects zero or more @var{expressions}, and emits
5363 a 16 bit number for each.
5366 This directive is a synonym for @samp{.short}; depending on the target
5367 architecture, it may also be a synonym for @samp{.word}.
5371 This directive is a synonym for @samp{.short}.
5374 This directive is a synonym for both @samp{.short} and @samp{.word}.
5379 @section @code{.ident}
5381 @cindex @code{ident} directive
5383 This directive is used by some assemblers to place tags in object files. The
5384 behavior of this directive varies depending on the target. When using the
5385 a.out object file format, @command{@value{AS}} simply accepts the directive for
5386 source-file compatibility with existing assemblers, but does not emit anything
5387 for it. When using COFF, comments are emitted to the @code{.comment} or
5388 @code{.rdata} section, depending on the target. When using ELF, comments are
5389 emitted to the @code{.comment} section.
5392 @section @code{.if @var{absolute expression}}
5394 @cindex conditional assembly
5395 @cindex @code{if} directive
5396 @code{.if} marks the beginning of a section of code which is only
5397 considered part of the source program being assembled if the argument
5398 (which must be an @var{absolute expression}) is non-zero. The end of
5399 the conditional section of code must be marked by @code{.endif}
5400 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
5401 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
5402 If you have several conditions to check, @code{.elseif} may be used to avoid
5403 nesting blocks if/else within each subsequent @code{.else} block.
5405 The following variants of @code{.if} are also supported:
5407 @cindex @code{ifdef} directive
5408 @item .ifdef @var{symbol}
5409 Assembles the following section of code if the specified @var{symbol}
5410 has been defined. Note a symbol which has been referenced but not yet defined
5411 is considered to be undefined.
5413 @cindex @code{ifb} directive
5414 @item .ifb @var{text}
5415 Assembles the following section of code if the operand is blank (empty).
5417 @cindex @code{ifc} directive
5418 @item .ifc @var{string1},@var{string2}
5419 Assembles the following section of code if the two strings are the same. The
5420 strings may be optionally quoted with single quotes. If they are not quoted,
5421 the first string stops at the first comma, and the second string stops at the
5422 end of the line. Strings which contain whitespace should be quoted. The
5423 string comparison is case sensitive.
5425 @cindex @code{ifeq} directive
5426 @item .ifeq @var{absolute expression}
5427 Assembles the following section of code if the argument is zero.
5429 @cindex @code{ifeqs} directive
5430 @item .ifeqs @var{string1},@var{string2}
5431 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5433 @cindex @code{ifge} directive
5434 @item .ifge @var{absolute expression}
5435 Assembles the following section of code if the argument is greater than or
5438 @cindex @code{ifgt} directive
5439 @item .ifgt @var{absolute expression}
5440 Assembles the following section of code if the argument is greater than zero.
5442 @cindex @code{ifle} directive
5443 @item .ifle @var{absolute expression}
5444 Assembles the following section of code if the argument is less than or equal
5447 @cindex @code{iflt} directive
5448 @item .iflt @var{absolute expression}
5449 Assembles the following section of code if the argument is less than zero.
5451 @cindex @code{ifnb} directive
5452 @item .ifnb @var{text}
5453 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5454 following section of code if the operand is non-blank (non-empty).
5456 @cindex @code{ifnc} directive
5457 @item .ifnc @var{string1},@var{string2}.
5458 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5459 following section of code if the two strings are not the same.
5461 @cindex @code{ifndef} directive
5462 @cindex @code{ifnotdef} directive
5463 @item .ifndef @var{symbol}
5464 @itemx .ifnotdef @var{symbol}
5465 Assembles the following section of code if the specified @var{symbol}
5466 has not been defined. Both spelling variants are equivalent. Note a symbol
5467 which has been referenced but not yet defined is considered to be undefined.
5469 @cindex @code{ifne} directive
5470 @item .ifne @var{absolute expression}
5471 Assembles the following section of code if the argument is not equal to zero
5472 (in other words, this is equivalent to @code{.if}).
5474 @cindex @code{ifnes} directive
5475 @item .ifnes @var{string1},@var{string2}
5476 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5477 following section of code if the two strings are not the same.
5481 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5483 @cindex @code{incbin} directive
5484 @cindex binary files, including
5485 The @code{incbin} directive includes @var{file} verbatim at the current
5486 location. You can control the search paths used with the @samp{-I} command-line
5487 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5490 The @var{skip} argument skips a number of bytes from the start of the
5491 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5492 read. Note that the data is not aligned in any way, so it is the user's
5493 responsibility to make sure that proper alignment is provided both before and
5494 after the @code{incbin} directive.
5497 @section @code{.include "@var{file}"}
5499 @cindex @code{include} directive
5500 @cindex supporting files, including
5501 @cindex files, including
5502 This directive provides a way to include supporting files at specified
5503 points in your source program. The code from @var{file} is assembled as
5504 if it followed the point of the @code{.include}; when the end of the
5505 included file is reached, assembly of the original file continues. You
5506 can control the search paths used with the @samp{-I} command-line option
5507 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5511 @section @code{.int @var{expressions}}
5513 @cindex @code{int} directive
5514 @cindex integers, 32-bit
5515 Expect zero or more @var{expressions}, of any section, separated by commas.
5516 For each expression, emit a number that, at run time, is the value of that
5517 expression. The byte order and bit size of the number depends on what kind
5518 of target the assembly is for.
5522 On most forms of the H8/300, @code{.int} emits 16-bit
5523 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5530 @section @code{.internal @var{names}}
5532 @cindex @code{internal} directive
5534 This is one of the ELF visibility directives. The other two are
5535 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5536 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5538 This directive overrides the named symbols default visibility (which is set by
5539 their binding: local, global or weak). The directive sets the visibility to
5540 @code{internal} which means that the symbols are considered to be @code{hidden}
5541 (i.e., not visible to other components), and that some extra, processor specific
5542 processing must also be performed upon the symbols as well.
5546 @section @code{.irp @var{symbol},@var{values}}@dots{}
5548 @cindex @code{irp} directive
5549 Evaluate a sequence of statements assigning different values to @var{symbol}.
5550 The sequence of statements starts at the @code{.irp} directive, and is
5551 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5552 set to @var{value}, and the sequence of statements is assembled. If no
5553 @var{value} is listed, the sequence of statements is assembled once, with
5554 @var{symbol} set to the null string. To refer to @var{symbol} within the
5555 sequence of statements, use @var{\symbol}.
5557 For example, assembling
5565 is equivalent to assembling
5573 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5576 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5578 @cindex @code{irpc} directive
5579 Evaluate a sequence of statements assigning different values to @var{symbol}.
5580 The sequence of statements starts at the @code{.irpc} directive, and is
5581 terminated by an @code{.endr} directive. For each character in @var{value},
5582 @var{symbol} is set to the character, and the sequence of statements is
5583 assembled. If no @var{value} is listed, the sequence of statements is
5584 assembled once, with @var{symbol} set to the null string. To refer to
5585 @var{symbol} within the sequence of statements, use @var{\symbol}.
5587 For example, assembling
5595 is equivalent to assembling
5603 For some caveats with the spelling of @var{symbol}, see also the discussion
5607 @section @code{.lcomm @var{symbol} , @var{length}}
5609 @cindex @code{lcomm} directive
5610 @cindex local common symbols
5611 @cindex symbols, local common
5612 Reserve @var{length} (an absolute expression) bytes for a local common
5613 denoted by @var{symbol}. The section and value of @var{symbol} are
5614 those of the new local common. The addresses are allocated in the bss
5615 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5616 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5617 not visible to @code{@value{LD}}.
5620 Some targets permit a third argument to be used with @code{.lcomm}. This
5621 argument specifies the desired alignment of the symbol in the bss section.
5625 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5626 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5630 @section @code{.lflags}
5632 @cindex @code{lflags} directive (ignored)
5633 @command{@value{AS}} accepts this directive, for compatibility with other
5634 assemblers, but ignores it.
5636 @ifclear no-line-dir
5638 @section @code{.line @var{line-number}}
5640 @cindex @code{line} directive
5641 @cindex logical line number
5643 Change the logical line number. @var{line-number} must be an absolute
5644 expression. The next line has that logical line number. Therefore any other
5645 statements on the current line (after a statement separator character) are
5646 reported as on logical line number @var{line-number} @minus{} 1. One day
5647 @command{@value{AS}} will no longer support this directive: it is recognized only
5648 for compatibility with existing assembler programs.
5651 Even though this is a directive associated with the @code{a.out} or
5652 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5653 when producing COFF output, and treats @samp{.line} as though it
5654 were the COFF @samp{.ln} @emph{if} it is found outside a
5655 @code{.def}/@code{.endef} pair.
5657 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5658 used by compilers to generate auxiliary symbol information for
5663 @section @code{.linkonce [@var{type}]}
5665 @cindex @code{linkonce} directive
5666 @cindex common sections
5667 Mark the current section so that the linker only includes a single copy of it.
5668 This may be used to include the same section in several different object files,
5669 but ensure that the linker will only include it once in the final output file.
5670 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5671 Duplicate sections are detected based on the section name, so it should be
5674 This directive is only supported by a few object file formats; as of this
5675 writing, the only object file format which supports it is the Portable
5676 Executable format used on Windows NT.
5678 The @var{type} argument is optional. If specified, it must be one of the
5679 following strings. For example:
5683 Not all types may be supported on all object file formats.
5687 Silently discard duplicate sections. This is the default.
5690 Warn if there are duplicate sections, but still keep only one copy.
5693 Warn if any of the duplicates have different sizes.
5696 Warn if any of the duplicates do not have exactly the same contents.
5700 @section @code{.list}
5702 @cindex @code{list} directive
5703 @cindex listing control, turning on
5704 Control (in conjunction with the @code{.nolist} directive) whether or
5705 not assembly listings are generated. These two directives maintain an
5706 internal counter (which is zero initially). @code{.list} increments the
5707 counter, and @code{.nolist} decrements it. Assembly listings are
5708 generated whenever the counter is greater than zero.
5710 By default, listings are disabled. When you enable them (with the
5711 @samp{-a} command-line option; @pxref{Invoking,,Command-Line Options}),
5712 the initial value of the listing counter is one.
5715 @section @code{.ln @var{line-number}}
5717 @cindex @code{ln} directive
5718 @ifclear no-line-dir
5719 @samp{.ln} is a synonym for @samp{.line}.
5722 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5723 must be an absolute expression. The next line has that logical
5724 line number, so any other statements on the current line (after a
5725 statement separator character @code{;}) are reported as on logical
5726 line number @var{line-number} @minus{} 1.
5730 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5731 @cindex @code{loc} directive
5732 When emitting DWARF2 line number information,
5733 the @code{.loc} directive will add a row to the @code{.debug_line} line
5734 number matrix corresponding to the immediately following assembly
5735 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5736 arguments will be applied to the @code{.debug_line} state machine before
5739 The @var{options} are a sequence of the following tokens in any order:
5743 This option will set the @code{basic_block} register in the
5744 @code{.debug_line} state machine to @code{true}.
5747 This option will set the @code{prologue_end} register in the
5748 @code{.debug_line} state machine to @code{true}.
5750 @item epilogue_begin
5751 This option will set the @code{epilogue_begin} register in the
5752 @code{.debug_line} state machine to @code{true}.
5754 @item is_stmt @var{value}
5755 This option will set the @code{is_stmt} register in the
5756 @code{.debug_line} state machine to @code{value}, which must be
5759 @item isa @var{value}
5760 This directive will set the @code{isa} register in the @code{.debug_line}
5761 state machine to @var{value}, which must be an unsigned integer.
5763 @item discriminator @var{value}
5764 This directive will set the @code{discriminator} register in the @code{.debug_line}
5765 state machine to @var{value}, which must be an unsigned integer.
5767 @item view @var{value}
5768 This option causes a row to be added to @code{.debug_line} in reference to the
5769 current address (which might not be the same as that of the following assembly
5770 instruction), and to associate @var{value} with the @code{view} register in the
5771 @code{.debug_line} state machine. If @var{value} is a label, both the
5772 @code{view} register and the label are set to the number of prior @code{.loc}
5773 directives at the same program location. If @var{value} is the literal
5774 @code{0}, the @code{view} register is set to zero, and the assembler asserts
5775 that there aren't any prior @code{.loc} directives at the same program
5776 location. If @var{value} is the literal @code{-0}, the assembler arrange for
5777 the @code{view} register to be reset in this row, even if there are prior
5778 @code{.loc} directives at the same program location.
5782 @node Loc_mark_labels
5783 @section @code{.loc_mark_labels @var{enable}}
5784 @cindex @code{loc_mark_labels} directive
5785 When emitting DWARF2 line number information,
5786 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5787 to the @code{.debug_line} line number matrix with the @code{basic_block}
5788 register in the state machine set whenever a code label is seen.
5789 The @var{enable} argument should be either 1 or 0, to enable or disable
5790 this function respectively.
5794 @section @code{.local @var{names}}
5796 @cindex @code{local} directive
5797 This directive, which is available for ELF targets, marks each symbol in
5798 the comma-separated list of @code{names} as a local symbol so that it
5799 will not be externally visible. If the symbols do not already exist,
5800 they will be created.
5802 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5803 accept an alignment argument, which is the case for most ELF targets,
5804 the @code{.local} directive can be used in combination with @code{.comm}
5805 (@pxref{Comm}) to define aligned local common data.
5809 @section @code{.long @var{expressions}}
5811 @cindex @code{long} directive
5812 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5815 @c no one seems to know what this is for or whether this description is
5816 @c what it really ought to do
5818 @section @code{.lsym @var{symbol}, @var{expression}}
5820 @cindex @code{lsym} directive
5821 @cindex symbol, not referenced in assembly
5822 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5823 the hash table, ensuring it cannot be referenced by name during the
5824 rest of the assembly. This sets the attributes of the symbol to be
5825 the same as the expression value:
5827 @var{other} = @var{descriptor} = 0
5828 @var{type} = @r{(section of @var{expression})}
5829 @var{value} = @var{expression}
5832 The new symbol is not flagged as external.
5836 @section @code{.macro}
5839 The commands @code{.macro} and @code{.endm} allow you to define macros that
5840 generate assembly output. For example, this definition specifies a macro
5841 @code{sum} that puts a sequence of numbers into memory:
5844 .macro sum from=0, to=5
5853 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5865 @item .macro @var{macname}
5866 @itemx .macro @var{macname} @var{macargs} @dots{}
5867 @cindex @code{macro} directive
5868 Begin the definition of a macro called @var{macname}. If your macro
5869 definition requires arguments, specify their names after the macro name,
5870 separated by commas or spaces. You can qualify the macro argument to
5871 indicate whether all invocations must specify a non-blank value (through
5872 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5873 (through @samp{:@code{vararg}}). You can supply a default value for any
5874 macro argument by following the name with @samp{=@var{deflt}}. You
5875 cannot define two macros with the same @var{macname} unless it has been
5876 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5877 definitions. For example, these are all valid @code{.macro} statements:
5881 Begin the definition of a macro called @code{comm}, which takes no
5884 @item .macro plus1 p, p1
5885 @itemx .macro plus1 p p1
5886 Either statement begins the definition of a macro called @code{plus1},
5887 which takes two arguments; within the macro definition, write
5888 @samp{\p} or @samp{\p1} to evaluate the arguments.
5890 @item .macro reserve_str p1=0 p2
5891 Begin the definition of a macro called @code{reserve_str}, with two
5892 arguments. The first argument has a default value, but not the second.
5893 After the definition is complete, you can call the macro either as
5894 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5895 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5896 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5897 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5899 @item .macro m p1:req, p2=0, p3:vararg
5900 Begin the definition of a macro called @code{m}, with at least three
5901 arguments. The first argument must always have a value specified, but
5902 not the second, which instead has a default value. The third formal
5903 will get assigned all remaining arguments specified at invocation time.
5905 When you call a macro, you can specify the argument values either by
5906 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5907 @samp{sum to=17, from=9}.
5911 Note that since each of the @var{macargs} can be an identifier exactly
5912 as any other one permitted by the target architecture, there may be
5913 occasional problems if the target hand-crafts special meanings to certain
5914 characters when they occur in a special position. For example, if the colon
5915 (@code{:}) is generally permitted to be part of a symbol name, but the
5916 architecture specific code special-cases it when occurring as the final
5917 character of a symbol (to denote a label), then the macro parameter
5918 replacement code will have no way of knowing that and consider the whole
5919 construct (including the colon) an identifier, and check only this
5920 identifier for being the subject to parameter substitution. So for example
5921 this macro definition:
5929 might not work as expected. Invoking @samp{label foo} might not create a label
5930 called @samp{foo} but instead just insert the text @samp{\l:} into the
5931 assembler source, probably generating an error about an unrecognised
5934 Similarly problems might occur with the period character (@samp{.})
5935 which is often allowed inside opcode names (and hence identifier names). So
5936 for example constructing a macro to build an opcode from a base name and a
5937 length specifier like this:
5940 .macro opcode base length
5945 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5946 instruction but instead generate some kind of error as the assembler tries to
5947 interpret the text @samp{\base.\length}.
5949 There are several possible ways around this problem:
5952 @item Insert white space
5953 If it is possible to use white space characters then this is the simplest
5962 @item Use @samp{\()}
5963 The string @samp{\()} can be used to separate the end of a macro argument from
5964 the following text. eg:
5967 .macro opcode base length
5972 @item Use the alternate macro syntax mode
5973 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5974 used as a separator. eg:
5984 Note: this problem of correctly identifying string parameters to pseudo ops
5985 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5986 and @code{.irpc} (@pxref{Irpc}) as well.
5989 @cindex @code{endm} directive
5990 Mark the end of a macro definition.
5993 @cindex @code{exitm} directive
5994 Exit early from the current macro definition.
5996 @cindex number of macros executed
5997 @cindex macros, count executed
5999 @command{@value{AS}} maintains a counter of how many macros it has
6000 executed in this pseudo-variable; you can copy that number to your
6001 output with @samp{\@@}, but @emph{only within a macro definition}.
6003 @item LOCAL @var{name} [ , @dots{} ]
6004 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
6005 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
6006 @xref{Altmacro,,@code{.altmacro}}.
6010 @section @code{.mri @var{val}}
6012 @cindex @code{mri} directive
6013 @cindex MRI mode, temporarily
6014 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
6015 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
6016 affects code assembled until the next @code{.mri} directive, or until the end
6017 of the file. @xref{M, MRI mode, MRI mode}.
6020 @section @code{.noaltmacro}
6021 Disable alternate macro mode. @xref{Altmacro}.
6024 @section @code{.nolist}
6026 @cindex @code{nolist} directive
6027 @cindex listing control, turning off
6028 Control (in conjunction with the @code{.list} directive) whether or
6029 not assembly listings are generated. These two directives maintain an
6030 internal counter (which is zero initially). @code{.list} increments the
6031 counter, and @code{.nolist} decrements it. Assembly listings are
6032 generated whenever the counter is greater than zero.
6035 @section @code{.nops @var{size}[, @var{control}]}
6037 @cindex @code{nops} directive
6038 @cindex filling memory with no-op instructions
6039 This directive emits @var{size} bytes filled with no-op instructions.
6040 @var{size} is absolute expression, which must be a positve value.
6041 @var{control} controls how no-op instructions should be generated. If
6042 the comma and @var{control} are omitted, @var{control} is assumed to be
6045 Note: For Intel 80386 and AMD x86-64 targets, @var{control} specifies
6046 the size limit of a no-op instruction. The valid values of @var{control}
6047 are between 0 and 4 in 16-bit mode, between 0 and 7 when tuning for
6048 older processors in 32-bit mode, between 0 and 11 in 64-bit mode or when
6049 tuning for newer processors in 32-bit mode. When 0 is used, the no-op
6050 instruction size limit is set to the maximum supported size.
6053 @section @code{.octa @var{bignums}}
6055 @c FIXME: double size emitted for "octa" on some? Or warn?
6056 @cindex @code{octa} directive
6057 @cindex integer, 16-byte
6058 @cindex sixteen byte integer
6059 This directive expects zero or more bignums, separated by commas. For each
6060 bignum, it emits a 16-byte integer.
6062 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
6063 hence @emph{octa}-word for 16 bytes.
6066 @section @code{.offset @var{loc}}
6068 @cindex @code{offset} directive
6069 Set the location counter to @var{loc} in the absolute section. @var{loc} must
6070 be an absolute expression. This directive may be useful for defining
6071 symbols with absolute values. Do not confuse it with the @code{.org}
6075 @section @code{.org @var{new-lc} , @var{fill}}
6077 @cindex @code{org} directive
6078 @cindex location counter, advancing
6079 @cindex advancing location counter
6080 @cindex current address, advancing
6081 Advance the location counter of the current section to
6082 @var{new-lc}. @var{new-lc} is either an absolute expression or an
6083 expression with the same section as the current subsection. That is,
6084 you can't use @code{.org} to cross sections: if @var{new-lc} has the
6085 wrong section, the @code{.org} directive is ignored. To be compatible
6086 with former assemblers, if the section of @var{new-lc} is absolute,
6087 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
6088 is the same as the current subsection.
6090 @code{.org} may only increase the location counter, or leave it
6091 unchanged; you cannot use @code{.org} to move the location counter
6094 @c double negative used below "not undefined" because this is a specific
6095 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
6096 @c section. doc@cygnus.com 18feb91
6097 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
6098 may not be undefined. If you really detest this restriction we eagerly await
6099 a chance to share your improved assembler.
6101 Beware that the origin is relative to the start of the section, not
6102 to the start of the subsection. This is compatible with other
6103 people's assemblers.
6105 When the location counter (of the current subsection) is advanced, the
6106 intervening bytes are filled with @var{fill} which should be an
6107 absolute expression. If the comma and @var{fill} are omitted,
6108 @var{fill} defaults to zero.
6111 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
6113 @cindex padding the location counter given a power of two
6114 @cindex @code{p2align} directive
6115 Pad the location counter (in the current subsection) to a particular
6116 storage boundary. The first expression (which must be absolute) is the
6117 number of low-order zero bits the location counter must have after
6118 advancement. For example @samp{.p2align 3} advances the location
6119 counter until it a multiple of 8. If the location counter is already a
6120 multiple of 8, no change is needed.
6122 The second expression (also absolute) gives the fill value to be stored in the
6123 padding bytes. It (and the comma) may be omitted. If it is omitted, the
6124 padding bytes are normally zero. However, on most systems, if the section is
6125 marked as containing code and the fill value is omitted, the space is filled
6126 with no-op instructions.
6128 The third expression is also absolute, and is also optional. If it is present,
6129 it is the maximum number of bytes that should be skipped by this alignment
6130 directive. If doing the alignment would require skipping more bytes than the
6131 specified maximum, then the alignment is not done at all. You can omit the
6132 fill value (the second argument) entirely by simply using two commas after the
6133 required alignment; this can be useful if you want the alignment to be filled
6134 with no-op instructions when appropriate.
6136 @cindex @code{p2alignw} directive
6137 @cindex @code{p2alignl} directive
6138 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
6139 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
6140 pattern as a two byte word value. The @code{.p2alignl} directives treats the
6141 fill pattern as a four byte longword value. For example, @code{.p2alignw
6142 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
6143 filled in with the value 0x368d (the exact placement of the bytes depends upon
6144 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
6149 @section @code{.popsection}
6151 @cindex @code{popsection} directive
6152 @cindex Section Stack
6153 This is one of the ELF section stack manipulation directives. The others are
6154 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6155 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
6158 This directive replaces the current section (and subsection) with the top
6159 section (and subsection) on the section stack. This section is popped off the
6165 @section @code{.previous}
6167 @cindex @code{previous} directive
6168 @cindex Section Stack
6169 This is one of the ELF section stack manipulation directives. The others are
6170 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6171 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
6172 (@pxref{PopSection}).
6174 This directive swaps the current section (and subsection) with most recently
6175 referenced section/subsection pair prior to this one. Multiple
6176 @code{.previous} directives in a row will flip between two sections (and their
6177 subsections). For example:
6189 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
6195 # Now in section A subsection 1
6199 # Now in section B subsection 0
6202 # Now in section B subsection 1
6205 # Now in section B subsection 0
6209 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
6210 section B and 0x9abc into subsection 1 of section B.
6212 In terms of the section stack, this directive swaps the current section with
6213 the top section on the section stack.
6217 @section @code{.print @var{string}}
6219 @cindex @code{print} directive
6220 @command{@value{AS}} will print @var{string} on the standard output during
6221 assembly. You must put @var{string} in double quotes.
6225 @section @code{.protected @var{names}}
6227 @cindex @code{protected} directive
6229 This is one of the ELF visibility directives. The other two are
6230 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
6232 This directive overrides the named symbols default visibility (which is set by
6233 their binding: local, global or weak). The directive sets the visibility to
6234 @code{protected} which means that any references to the symbols from within the
6235 components that defines them must be resolved to the definition in that
6236 component, even if a definition in another component would normally preempt
6241 @section @code{.psize @var{lines} , @var{columns}}
6243 @cindex @code{psize} directive
6244 @cindex listing control: paper size
6245 @cindex paper size, for listings
6246 Use this directive to declare the number of lines---and, optionally, the
6247 number of columns---to use for each page, when generating listings.
6249 If you do not use @code{.psize}, listings use a default line-count
6250 of 60. You may omit the comma and @var{columns} specification; the
6251 default width is 200 columns.
6253 @command{@value{AS}} generates formfeeds whenever the specified number of
6254 lines is exceeded (or whenever you explicitly request one, using
6257 If you specify @var{lines} as @code{0}, no formfeeds are generated save
6258 those explicitly specified with @code{.eject}.
6261 @section @code{.purgem @var{name}}
6263 @cindex @code{purgem} directive
6264 Undefine the macro @var{name}, so that later uses of the string will not be
6265 expanded. @xref{Macro}.
6269 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
6271 @cindex @code{pushsection} directive
6272 @cindex Section Stack
6273 This is one of the ELF section stack manipulation directives. The others are
6274 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
6275 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6278 This directive pushes the current section (and subsection) onto the
6279 top of the section stack, and then replaces the current section and
6280 subsection with @code{name} and @code{subsection}. The optional
6281 @code{flags}, @code{type} and @code{arguments} are treated the same
6282 as in the @code{.section} (@pxref{Section}) directive.
6286 @section @code{.quad @var{bignums}}
6288 @cindex @code{quad} directive
6289 @code{.quad} expects zero or more bignums, separated by commas. For
6290 each bignum, it emits
6292 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
6293 warning message; and just takes the lowest order 8 bytes of the bignum.
6294 @cindex eight-byte integer
6295 @cindex integer, 8-byte
6297 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
6298 hence @emph{quad}-word for 8 bytes.
6301 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
6302 warning message; and just takes the lowest order 16 bytes of the bignum.
6303 @cindex sixteen-byte integer
6304 @cindex integer, 16-byte
6308 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
6310 @cindex @code{reloc} directive
6311 Generate a relocation at @var{offset} of type @var{reloc_name} with value
6312 @var{expression}. If @var{offset} is a number, the relocation is generated in
6313 the current section. If @var{offset} is an expression that resolves to a
6314 symbol plus offset, the relocation is generated in the given symbol's section.
6315 @var{expression}, if present, must resolve to a symbol plus addend or to an
6316 absolute value, but note that not all targets support an addend. e.g. ELF REL
6317 targets such as i386 store an addend in the section contents rather than in the
6318 relocation. This low level interface does not support addends stored in the
6322 @section @code{.rept @var{count}}
6324 @cindex @code{rept} directive
6325 Repeat the sequence of lines between the @code{.rept} directive and the next
6326 @code{.endr} directive @var{count} times.
6328 For example, assembling
6336 is equivalent to assembling
6344 A count of zero is allowed, but nothing is generated. Negative counts are not
6345 allowed and if encountered will be treated as if they were zero.
6348 @section @code{.sbttl "@var{subheading}"}
6350 @cindex @code{sbttl} directive
6351 @cindex subtitles for listings
6352 @cindex listing control: subtitle
6353 Use @var{subheading} as the title (third line, immediately after the
6354 title line) when generating assembly listings.
6356 This directive affects subsequent pages, as well as the current page if
6357 it appears within ten lines of the top of a page.
6361 @section @code{.scl @var{class}}
6363 @cindex @code{scl} directive
6364 @cindex symbol storage class (COFF)
6365 @cindex COFF symbol storage class
6366 Set the storage-class value for a symbol. This directive may only be
6367 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
6368 whether a symbol is static or external, or it may record further
6369 symbolic debugging information.
6374 @section @code{.section @var{name}}
6376 @cindex named section
6377 Use the @code{.section} directive to assemble the following code into a section
6380 This directive is only supported for targets that actually support arbitrarily
6381 named sections; on @code{a.out} targets, for example, it is not accepted, even
6382 with a standard @code{a.out} section name.
6386 @c only print the extra heading if both COFF and ELF are set
6387 @subheading COFF Version
6390 @cindex @code{section} directive (COFF version)
6391 For COFF targets, the @code{.section} directive is used in one of the following
6395 .section @var{name}[, "@var{flags}"]
6396 .section @var{name}[, @var{subsection}]
6399 If the optional argument is quoted, it is taken as flags to use for the
6400 section. Each flag is a single character. The following flags are recognized:
6404 bss section (uninitialized data)
6406 section is not loaded
6412 exclude section from linking
6418 shared section (meaningful for PE targets)
6420 ignored. (For compatibility with the ELF version)
6422 section is not readable (meaningful for PE targets)
6424 single-digit power-of-two section alignment (GNU extension)
6427 If no flags are specified, the default flags depend upon the section name. If
6428 the section name is not recognized, the default will be for the section to be
6429 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
6430 from the section, rather than adding them, so if they are used on their own it
6431 will be as if no flags had been specified at all.
6433 If the optional argument to the @code{.section} directive is not quoted, it is
6434 taken as a subsection number (@pxref{Sub-Sections}).
6439 @c only print the extra heading if both COFF and ELF are set
6440 @subheading ELF Version
6443 @cindex Section Stack
6444 This is one of the ELF section stack manipulation directives. The others are
6445 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
6446 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
6447 @code{.previous} (@pxref{Previous}).
6449 @cindex @code{section} directive (ELF version)
6450 For ELF targets, the @code{.section} directive is used like this:
6453 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
6456 @anchor{Section Name Substitutions}
6457 @kindex --sectname-subst
6458 @cindex section name substitution
6459 If the @samp{--sectname-subst} command-line option is provided, the @var{name}
6460 argument may contain a substitution sequence. Only @code{%S} is supported
6461 at the moment, and substitutes the current section name. For example:
6464 .macro exception_code
6465 .section %S.exception
6466 [exception code here]
6481 The two @code{exception_code} invocations above would create the
6482 @code{.text.exception} and @code{.init.exception} sections respectively.
6483 This is useful e.g. to discriminate between ancillary sections that are
6484 tied to setup code to be discarded after use from ancillary sections that
6485 need to stay resident without having to define multiple @code{exception_code}
6486 macros just for that purpose.
6488 The optional @var{flags} argument is a quoted string which may contain any
6489 combination of the following characters:
6493 section is allocatable
6495 section is a GNU_MBIND section
6497 section is excluded from executable and shared library.
6501 section is executable
6503 section is mergeable
6505 section contains zero terminated strings
6507 section is a member of a section group
6509 section is used for thread-local-storage
6511 section is a member of the previously-current section's group, if any
6512 @item @code{<number>}
6513 a numeric value indicating the bits to be set in the ELF section header's flags
6514 field. Note - if one or more of the alphabetic characters described above is
6515 also included in the flags field, their bit values will be ORed into the
6517 @item @code{<target specific>}
6518 some targets extend this list with their own flag characters
6521 Note - once a section's flags have been set they cannot be changed. There are
6522 a few exceptions to this rule however. Processor and application specific
6523 flags can be added to an already defined section. The @code{.interp},
6524 @code{.strtab} and @code{.symtab} sections can have the allocate flag
6525 (@code{a}) set after they are initially defined, and the @code{.note-GNU-stack}
6526 section may have the executable (@code{x}) flag added.
6528 The optional @var{type} argument may contain one of the following constants:
6532 section contains data
6534 section does not contain data (i.e., section only occupies space)
6536 section contains data which is used by things other than the program
6538 section contains an array of pointers to init functions
6540 section contains an array of pointers to finish functions
6541 @item @@preinit_array
6542 section contains an array of pointers to pre-init functions
6543 @item @@@code{<number>}
6544 a numeric value to be set as the ELF section header's type field.
6545 @item @@@code{<target specific>}
6546 some targets extend this list with their own types
6549 Many targets only support the first three section types. The type may be
6550 enclosed in double quotes if necessary.
6552 Note on targets where the @code{@@} character is the start of a comment (eg
6553 ARM) then another character is used instead. For example the ARM port uses the
6556 Note - some sections, eg @code{.text} and @code{.data} are considered to be
6557 special and have fixed types. Any attempt to declare them with a different
6558 type will generate an error from the assembler.
6560 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6561 be specified as well as an extra argument---@var{entsize}---like this:
6564 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6567 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6568 constants, each @var{entsize} octets long. Sections with both @code{M} and
6569 @code{S} must contain zero terminated strings where each character is
6570 @var{entsize} bytes long. The linker may remove duplicates within sections with
6571 the same name, same entity size and same flags. @var{entsize} must be an
6572 absolute expression. For sections with both @code{M} and @code{S}, a string
6573 which is a suffix of a larger string is considered a duplicate. Thus
6574 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6575 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6577 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6578 be present along with an additional field like this:
6581 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6584 The @var{GroupName} field specifies the name of the section group to which this
6585 particular section belongs. The optional linkage field can contain:
6589 indicates that only one copy of this section should be retained
6594 Note: if both the @var{M} and @var{G} flags are present then the fields for
6595 the Merge flag should come first, like this:
6598 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6601 If @var{flags} contains the @code{?} symbol then it may not also contain the
6602 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6603 present. Instead, @code{?} says to consider the section that's current before
6604 this directive. If that section used @code{G}, then the new section will use
6605 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6606 If not, then the @code{?} symbol has no effect.
6608 If no flags are specified, the default flags depend upon the section name. If
6609 the section name is not recognized, the default will be for the section to have
6610 none of the above flags: it will not be allocated in memory, nor writable, nor
6611 executable. The section will contain data.
6613 For ELF targets, the assembler supports another type of @code{.section}
6614 directive for compatibility with the Solaris assembler:
6617 .section "@var{name}"[, @var{flags}...]
6620 Note that the section name is quoted. There may be a sequence of comma
6625 section is allocatable
6629 section is executable
6631 section is excluded from executable and shared library.
6633 section is used for thread local storage
6636 This directive replaces the current section and subsection. See the
6637 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6638 some examples of how this directive and the other section stack directives
6644 @section @code{.set @var{symbol}, @var{expression}}
6646 @cindex @code{set} directive
6647 @cindex symbol value, setting
6648 Set the value of @var{symbol} to @var{expression}. This
6649 changes @var{symbol}'s value and type to conform to
6650 @var{expression}. If @var{symbol} was flagged as external, it remains
6651 flagged (@pxref{Symbol Attributes}).
6653 You may @code{.set} a symbol many times in the same assembly provided that the
6654 values given to the symbol are constants. Values that are based on expressions
6655 involving other symbols are allowed, but some targets may restrict this to only
6656 being done once per assembly. This is because those targets do not set the
6657 addresses of symbols at assembly time, but rather delay the assignment until a
6658 final link is performed. This allows the linker a chance to change the code in
6659 the files, changing the location of, and the relative distance between, various
6662 If you @code{.set} a global symbol, the value stored in the object
6663 file is the last value stored into it.
6666 On Z80 @code{set} is a real instruction, use
6667 @samp{@var{symbol} defl @var{expression}} instead.
6671 @section @code{.short @var{expressions}}
6673 @cindex @code{short} directive
6675 @code{.short} is normally the same as @samp{.word}.
6676 @xref{Word,,@code{.word}}.
6678 In some configurations, however, @code{.short} and @code{.word} generate
6679 numbers of different lengths. @xref{Machine Dependencies}.
6683 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6686 This expects zero or more @var{expressions}, and emits
6687 a 16 bit number for each.
6692 @section @code{.single @var{flonums}}
6694 @cindex @code{single} directive
6695 @cindex floating point numbers (single)
6696 This directive assembles zero or more flonums, separated by commas. It
6697 has the same effect as @code{.float}.
6699 The exact kind of floating point numbers emitted depends on how
6700 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6704 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6705 numbers in @sc{ieee} format.
6711 @section @code{.size}
6713 This directive is used to set the size associated with a symbol.
6717 @c only print the extra heading if both COFF and ELF are set
6718 @subheading COFF Version
6721 @cindex @code{size} directive (COFF version)
6722 For COFF targets, the @code{.size} directive is only permitted inside
6723 @code{.def}/@code{.endef} pairs. It is used like this:
6726 .size @var{expression}
6733 @c only print the extra heading if both COFF and ELF are set
6734 @subheading ELF Version
6737 @cindex @code{size} directive (ELF version)
6738 For ELF targets, the @code{.size} directive is used like this:
6741 .size @var{name} , @var{expression}
6744 This directive sets the size associated with a symbol @var{name}.
6745 The size in bytes is computed from @var{expression} which can make use of label
6746 arithmetic. This directive is typically used to set the size of function
6751 @ifclear no-space-dir
6753 @section @code{.skip @var{size} [,@var{fill}]}
6755 @cindex @code{skip} directive
6756 @cindex filling memory
6757 This directive emits @var{size} bytes, each of value @var{fill}. Both
6758 @var{size} and @var{fill} are absolute expressions. If the comma and
6759 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6764 @section @code{.sleb128 @var{expressions}}
6766 @cindex @code{sleb128} directive
6767 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6768 compact, variable length representation of numbers used by the DWARF
6769 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6771 @ifclear no-space-dir
6773 @section @code{.space @var{size} [,@var{fill}]}
6775 @cindex @code{space} directive
6776 @cindex filling memory
6777 This directive emits @var{size} bytes, each of value @var{fill}. Both
6778 @var{size} and @var{fill} are absolute expressions. If the comma
6779 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6784 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6785 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6786 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6787 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6795 @section @code{.stabd, .stabn, .stabs}
6797 @cindex symbolic debuggers, information for
6798 @cindex @code{stab@var{x}} directives
6799 There are three directives that begin @samp{.stab}.
6800 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6801 The symbols are not entered in the @command{@value{AS}} hash table: they
6802 cannot be referenced elsewhere in the source file.
6803 Up to five fields are required:
6807 This is the symbol's name. It may contain any character except
6808 @samp{\000}, so is more general than ordinary symbol names. Some
6809 debuggers used to code arbitrarily complex structures into symbol names
6813 An absolute expression. The symbol's type is set to the low 8 bits of
6814 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6815 and debuggers choke on silly bit patterns.
6818 An absolute expression. The symbol's ``other'' attribute is set to the
6819 low 8 bits of this expression.
6822 An absolute expression. The symbol's descriptor is set to the low 16
6823 bits of this expression.
6826 An absolute expression which becomes the symbol's value.
6829 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6830 or @code{.stabs} statement, the symbol has probably already been created;
6831 you get a half-formed symbol in your object file. This is
6832 compatible with earlier assemblers!
6835 @cindex @code{stabd} directive
6836 @item .stabd @var{type} , @var{other} , @var{desc}
6838 The ``name'' of the symbol generated is not even an empty string.
6839 It is a null pointer, for compatibility. Older assemblers used a
6840 null pointer so they didn't waste space in object files with empty
6843 The symbol's value is set to the location counter,
6844 relocatably. When your program is linked, the value of this symbol
6845 is the address of the location counter when the @code{.stabd} was
6848 @cindex @code{stabn} directive
6849 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6850 The name of the symbol is set to the empty string @code{""}.
6852 @cindex @code{stabs} directive
6853 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6854 All five fields are specified.
6860 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6861 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6863 @cindex string, copying to object file
6864 @cindex string8, copying to object file
6865 @cindex string16, copying to object file
6866 @cindex string32, copying to object file
6867 @cindex string64, copying to object file
6868 @cindex @code{string} directive
6869 @cindex @code{string8} directive
6870 @cindex @code{string16} directive
6871 @cindex @code{string32} directive
6872 @cindex @code{string64} directive
6874 Copy the characters in @var{str} to the object file. You may specify more than
6875 one string to copy, separated by commas. Unless otherwise specified for a
6876 particular machine, the assembler marks the end of each string with a 0 byte.
6877 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6879 The variants @code{string16}, @code{string32} and @code{string64} differ from
6880 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6881 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6882 are stored in target endianness byte order.
6888 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6889 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6894 @section @code{.struct @var{expression}}
6896 @cindex @code{struct} directive
6897 Switch to the absolute section, and set the section offset to @var{expression},
6898 which must be an absolute expression. You might use this as follows:
6907 This would define the symbol @code{field1} to have the value 0, the symbol
6908 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6909 value 8. Assembly would be left in the absolute section, and you would need to
6910 use a @code{.section} directive of some sort to change to some other section
6911 before further assembly.
6915 @section @code{.subsection @var{name}}
6917 @cindex @code{subsection} directive
6918 @cindex Section Stack
6919 This is one of the ELF section stack manipulation directives. The others are
6920 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6921 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6924 This directive replaces the current subsection with @code{name}. The current
6925 section is not changed. The replaced subsection is put onto the section stack
6926 in place of the then current top of stack subsection.
6931 @section @code{.symver}
6932 @cindex @code{symver} directive
6933 @cindex symbol versioning
6934 @cindex versions of symbols
6935 Use the @code{.symver} directive to bind symbols to specific version nodes
6936 within a source file. This is only supported on ELF platforms, and is
6937 typically used when assembling files to be linked into a shared library.
6938 There are cases where it may make sense to use this in objects to be bound
6939 into an application itself so as to override a versioned symbol from a
6942 For ELF targets, the @code{.symver} directive can be used like this:
6944 .symver @var{name}, @var{name2@@nodename}
6946 If the symbol @var{name} is defined within the file
6947 being assembled, the @code{.symver} directive effectively creates a symbol
6948 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6949 just don't try and create a regular alias is that the @var{@@} character isn't
6950 permitted in symbol names. The @var{name2} part of the name is the actual name
6951 of the symbol by which it will be externally referenced. The name @var{name}
6952 itself is merely a name of convenience that is used so that it is possible to
6953 have definitions for multiple versions of a function within a single source
6954 file, and so that the compiler can unambiguously know which version of a
6955 function is being mentioned. The @var{nodename} portion of the alias should be
6956 the name of a node specified in the version script supplied to the linker when
6957 building a shared library. If you are attempting to override a versioned
6958 symbol from a shared library, then @var{nodename} should correspond to the
6959 nodename of the symbol you are trying to override.
6961 If the symbol @var{name} is not defined within the file being assembled, all
6962 references to @var{name} will be changed to @var{name2@@nodename}. If no
6963 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6966 Another usage of the @code{.symver} directive is:
6968 .symver @var{name}, @var{name2@@@@nodename}
6970 In this case, the symbol @var{name} must exist and be defined within
6971 the file being assembled. It is similar to @var{name2@@nodename}. The
6972 difference is @var{name2@@@@nodename} will also be used to resolve
6973 references to @var{name2} by the linker.
6975 The third usage of the @code{.symver} directive is:
6977 .symver @var{name}, @var{name2@@@@@@nodename}
6979 When @var{name} is not defined within the
6980 file being assembled, it is treated as @var{name2@@nodename}. When
6981 @var{name} is defined within the file being assembled, the symbol
6982 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6987 @section @code{.tag @var{structname}}
6989 @cindex COFF structure debugging
6990 @cindex structure debugging, COFF
6991 @cindex @code{tag} directive
6992 This directive is generated by compilers to include auxiliary debugging
6993 information in the symbol table. It is only permitted inside
6994 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6995 definitions in the symbol table with instances of those structures.
6999 @section @code{.text @var{subsection}}
7001 @cindex @code{text} directive
7002 Tells @command{@value{AS}} to assemble the following statements onto the end of
7003 the text subsection numbered @var{subsection}, which is an absolute
7004 expression. If @var{subsection} is omitted, subsection number zero
7008 @section @code{.title "@var{heading}"}
7010 @cindex @code{title} directive
7011 @cindex listing control: title line
7012 Use @var{heading} as the title (second line, immediately after the
7013 source file name and pagenumber) when generating assembly listings.
7015 This directive affects subsequent pages, as well as the current page if
7016 it appears within ten lines of the top of a page.
7020 @section @code{.type}
7022 This directive is used to set the type of a symbol.
7026 @c only print the extra heading if both COFF and ELF are set
7027 @subheading COFF Version
7030 @cindex COFF symbol type
7031 @cindex symbol type, COFF
7032 @cindex @code{type} directive (COFF version)
7033 For COFF targets, this directive is permitted only within
7034 @code{.def}/@code{.endef} pairs. It is used like this:
7040 This records the integer @var{int} as the type attribute of a symbol table
7047 @c only print the extra heading if both COFF and ELF are set
7048 @subheading ELF Version
7051 @cindex ELF symbol type
7052 @cindex symbol type, ELF
7053 @cindex @code{type} directive (ELF version)
7054 For ELF targets, the @code{.type} directive is used like this:
7057 .type @var{name} , @var{type description}
7060 This sets the type of symbol @var{name} to be either a
7061 function symbol or an object symbol. There are five different syntaxes
7062 supported for the @var{type description} field, in order to provide
7063 compatibility with various other assemblers.
7065 Because some of the characters used in these syntaxes (such as @samp{@@} and
7066 @samp{#}) are comment characters for some architectures, some of the syntaxes
7067 below do not work on all architectures. The first variant will be accepted by
7068 the GNU assembler on all architectures so that variant should be used for
7069 maximum portability, if you do not need to assemble your code with other
7072 The syntaxes supported are:
7075 .type <name> STT_<TYPE_IN_UPPER_CASE>
7076 .type <name>,#<type>
7077 .type <name>,@@<type>
7078 .type <name>,%<type>
7079 .type <name>,"<type>"
7082 The types supported are:
7087 Mark the symbol as being a function name.
7090 @itemx gnu_indirect_function
7091 Mark the symbol as an indirect function when evaluated during reloc
7092 processing. (This is only supported on assemblers targeting GNU systems).
7096 Mark the symbol as being a data object.
7100 Mark the symbol as being a thread-local data object.
7104 Mark the symbol as being a common data object.
7108 Does not mark the symbol in any way. It is supported just for completeness.
7110 @item gnu_unique_object
7111 Marks the symbol as being a globally unique data object. The dynamic linker
7112 will make sure that in the entire process there is just one symbol with this
7113 name and type in use. (This is only supported on assemblers targeting GNU
7118 Note: Some targets support extra types in addition to those listed above.
7124 @section @code{.uleb128 @var{expressions}}
7126 @cindex @code{uleb128} directive
7127 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
7128 compact, variable length representation of numbers used by the DWARF
7129 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
7133 @section @code{.val @var{addr}}
7135 @cindex @code{val} directive
7136 @cindex COFF value attribute
7137 @cindex value attribute, COFF
7138 This directive, permitted only within @code{.def}/@code{.endef} pairs,
7139 records the address @var{addr} as the value attribute of a symbol table
7145 @section @code{.version "@var{string}"}
7147 @cindex @code{version} directive
7148 This directive creates a @code{.note} section and places into it an ELF
7149 formatted note of type NT_VERSION. The note's name is set to @code{string}.
7154 @section @code{.vtable_entry @var{table}, @var{offset}}
7156 @cindex @code{vtable_entry} directive
7157 This directive finds or creates a symbol @code{table} and creates a
7158 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
7161 @section @code{.vtable_inherit @var{child}, @var{parent}}
7163 @cindex @code{vtable_inherit} directive
7164 This directive finds the symbol @code{child} and finds or creates the symbol
7165 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
7166 parent whose addend is the value of the child symbol. As a special case the
7167 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
7171 @section @code{.warning "@var{string}"}
7172 @cindex warning directive
7173 Similar to the directive @code{.error}
7174 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
7177 @section @code{.weak @var{names}}
7179 @cindex @code{weak} directive
7180 This directive sets the weak attribute on the comma separated list of symbol
7181 @code{names}. If the symbols do not already exist, they will be created.
7183 On COFF targets other than PE, weak symbols are a GNU extension. This
7184 directive sets the weak attribute on the comma separated list of symbol
7185 @code{names}. If the symbols do not already exist, they will be created.
7187 On the PE target, weak symbols are supported natively as weak aliases.
7188 When a weak symbol is created that is not an alias, GAS creates an
7189 alternate symbol to hold the default value.
7192 @section @code{.weakref @var{alias}, @var{target}}
7194 @cindex @code{weakref} directive
7195 This directive creates an alias to the target symbol that enables the symbol to
7196 be referenced with weak-symbol semantics, but without actually making it weak.
7197 If direct references or definitions of the symbol are present, then the symbol
7198 will not be weak, but if all references to it are through weak references, the
7199 symbol will be marked as weak in the symbol table.
7201 The effect is equivalent to moving all references to the alias to a separate
7202 assembly source file, renaming the alias to the symbol in it, declaring the
7203 symbol as weak there, and running a reloadable link to merge the object files
7204 resulting from the assembly of the new source file and the old source file that
7205 had the references to the alias removed.
7207 The alias itself never makes to the symbol table, and is entirely handled
7208 within the assembler.
7211 @section @code{.word @var{expressions}}
7213 @cindex @code{word} directive
7214 This directive expects zero or more @var{expressions}, of any section,
7215 separated by commas.
7218 For each expression, @command{@value{AS}} emits a 32-bit number.
7221 For each expression, @command{@value{AS}} emits a 16-bit number.
7226 The size of the number emitted, and its byte order,
7227 depend on what target computer the assembly is for.
7230 @c on sparc the "special treatment to support compilers" doesn't
7231 @c happen---32-bit addressability, period; no long/short jumps.
7232 @ifset DIFF-TBL-KLUGE
7233 @cindex difference tables altered
7234 @cindex altered difference tables
7236 @emph{Warning: Special Treatment to support Compilers}
7240 Machines with a 32-bit address space, but that do less than 32-bit
7241 addressing, require the following special treatment. If the machine of
7242 interest to you does 32-bit addressing (or doesn't require it;
7243 @pxref{Machine Dependencies}), you can ignore this issue.
7246 In order to assemble compiler output into something that works,
7247 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
7248 Directives of the form @samp{.word sym1-sym2} are often emitted by
7249 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
7250 directive of the form @samp{.word sym1-sym2}, and the difference between
7251 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
7252 creates a @dfn{secondary jump table}, immediately before the next label.
7253 This secondary jump table is preceded by a short-jump to the
7254 first byte after the secondary table. This short-jump prevents the flow
7255 of control from accidentally falling into the new table. Inside the
7256 table is a long-jump to @code{sym2}. The original @samp{.word}
7257 contains @code{sym1} minus the address of the long-jump to
7260 If there were several occurrences of @samp{.word sym1-sym2} before the
7261 secondary jump table, all of them are adjusted. If there was a
7262 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
7263 long-jump to @code{sym4} is included in the secondary jump table,
7264 and the @code{.word} directives are adjusted to contain @code{sym3}
7265 minus the address of the long-jump to @code{sym4}; and so on, for as many
7266 entries in the original jump table as necessary.
7269 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
7270 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
7271 assembly language programmers.
7274 @c end DIFF-TBL-KLUGE
7276 @ifclear no-space-dir
7278 @section @code{.zero @var{size}}
7280 @cindex @code{zero} directive
7281 @cindex filling memory with zero bytes
7282 This directive emits @var{size} 0-valued bytes. @var{size} must be an absolute
7283 expression. This directive is actually an alias for the @samp{.skip} directive
7284 so in can take an optional second argument of the value to store in the bytes
7285 instead of zero. Using @samp{.zero} in this way would be confusing however.
7290 @section @code{.2byte @var{expression} [, @var{expression}]*}
7291 @cindex @code{2byte} directive
7292 @cindex two-byte integer
7293 @cindex integer, 2-byte
7295 This directive expects zero or more expressions, separated by commas. If there
7296 are no expressions then the directive does nothing. Otherwise each expression
7297 is evaluated in turn and placed in the next two bytes of the current output
7298 section, using the endian model of the target. If an expression will not fit
7299 in two bytes, a warning message is displayed and the least significant two
7300 bytes of the expression's value are used. If an expression cannot be evaluated
7301 at assembly time then relocations will be generated in order to compute the
7304 This directive does not apply any alignment before or after inserting the
7305 values. As a result of this, if relocations are generated, they may be
7306 different from those used for inserting values with a guaranteed alignment.
7308 This directive is only available for ELF targets,
7311 @section @code{.4byte @var{expression} [, @var{expression}]*}
7312 @cindex @code{4byte} directive
7313 @cindex four-byte integer
7314 @cindex integer, 4-byte
7316 Like the @option{.2byte} directive, except that it inserts unaligned, four byte
7317 long values into the output.
7320 @section @code{.8byte @var{expression} [, @var{expression}]*}
7321 @cindex @code{8byte} directive
7322 @cindex eight-byte integer
7323 @cindex integer, 8-byte
7325 Like the @option{.2byte} directive, except that it inserts unaligned, eight
7326 byte long bignum values into the output.
7331 @section Deprecated Directives
7333 @cindex deprecated directives
7334 @cindex obsolescent directives
7335 One day these directives won't work.
7336 They are included for compatibility with older assemblers.
7343 @node Object Attributes
7344 @chapter Object Attributes
7345 @cindex object attributes
7347 @command{@value{AS}} assembles source files written for a specific architecture
7348 into object files for that architecture. But not all object files are alike.
7349 Many architectures support incompatible variations. For instance, floating
7350 point arguments might be passed in floating point registers if the object file
7351 requires hardware floating point support---or floating point arguments might be
7352 passed in integer registers if the object file supports processors with no
7353 hardware floating point unit. Or, if two objects are built for different
7354 generations of the same architecture, the combination may require the
7355 newer generation at run-time.
7357 This information is useful during and after linking. At link time,
7358 @command{@value{LD}} can warn about incompatible object files. After link
7359 time, tools like @command{gdb} can use it to process the linked file
7362 Compatibility information is recorded as a series of object attributes. Each
7363 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
7364 string, and indicates who sets the meaning of the tag. The tag is an integer,
7365 and indicates what property the attribute describes. The value may be a string
7366 or an integer, and indicates how the property affects this object. Missing
7367 attributes are the same as attributes with a zero value or empty string value.
7369 Object attributes were developed as part of the ABI for the ARM Architecture.
7370 The file format is documented in @cite{ELF for the ARM Architecture}.
7373 * GNU Object Attributes:: @sc{gnu} Object Attributes
7374 * Defining New Object Attributes:: Defining New Object Attributes
7377 @node GNU Object Attributes
7378 @section @sc{gnu} Object Attributes
7380 The @code{.gnu_attribute} directive records an object attribute
7381 with vendor @samp{gnu}.
7383 Except for @samp{Tag_compatibility}, which has both an integer and a string for
7384 its value, @sc{gnu} attributes have a string value if the tag number is odd and
7385 an integer value if the tag number is even. The second bit (@code{@var{tag} &
7386 2} is set for architecture-independent attributes and clear for
7387 architecture-dependent ones.
7389 @subsection Common @sc{gnu} attributes
7391 These attributes are valid on all architectures.
7394 @item Tag_compatibility (32)
7395 The compatibility attribute takes an integer flag value and a vendor name. If
7396 the flag value is 0, the file is compatible with other toolchains. If it is 1,
7397 then the file is only compatible with the named toolchain. If it is greater
7398 than 1, the file can only be processed by other toolchains under some private
7399 arrangement indicated by the flag value and the vendor name.
7402 @subsection MIPS Attributes
7405 @item Tag_GNU_MIPS_ABI_FP (4)
7406 The floating-point ABI used by this object file. The value will be:
7410 0 for files not affected by the floating-point ABI.
7412 1 for files using the hardware floating-point ABI with a standard
7413 double-precision FPU.
7415 2 for files using the hardware floating-point ABI with a single-precision FPU.
7417 3 for files using the software floating-point ABI.
7419 4 for files using the deprecated hardware floating-point ABI which used 64-bit
7420 floating-point registers, 32-bit general-purpose registers and increased the
7421 number of callee-saved floating-point registers.
7423 5 for files using the hardware floating-point ABI with a double-precision FPU
7424 with either 32-bit or 64-bit floating-point registers and 32-bit
7425 general-purpose registers.
7427 6 for files using the hardware floating-point ABI with 64-bit floating-point
7428 registers and 32-bit general-purpose registers.
7430 7 for files using the hardware floating-point ABI with 64-bit floating-point
7431 registers, 32-bit general-purpose registers and a rule that forbids the
7432 direct use of odd-numbered single-precision floating-point registers.
7436 @subsection PowerPC Attributes
7439 @item Tag_GNU_Power_ABI_FP (4)
7440 The floating-point ABI used by this object file. The value will be:
7444 0 for files not affected by the floating-point ABI.
7446 1 for files using double-precision hardware floating-point ABI.
7448 2 for files using the software floating-point ABI.
7450 3 for files using single-precision hardware floating-point ABI.
7453 @item Tag_GNU_Power_ABI_Vector (8)
7454 The vector ABI used by this object file. The value will be:
7458 0 for files not affected by the vector ABI.
7460 1 for files using general purpose registers to pass vectors.
7462 2 for files using AltiVec registers to pass vectors.
7464 3 for files using SPE registers to pass vectors.
7468 @subsection IBM z Systems Attributes
7471 @item Tag_GNU_S390_ABI_Vector (8)
7472 The vector ABI used by this object file. The value will be:
7476 0 for files not affected by the vector ABI.
7478 1 for files using software vector ABI.
7480 2 for files using hardware vector ABI.
7484 @node Defining New Object Attributes
7485 @section Defining New Object Attributes
7487 If you want to define a new @sc{gnu} object attribute, here are the places you
7488 will need to modify. New attributes should be discussed on the @samp{binutils}
7493 This manual, which is the official register of attributes.
7495 The header for your architecture @file{include/elf}, to define the tag.
7497 The @file{bfd} support file for your architecture, to merge the attribute
7498 and issue any appropriate link warnings.
7500 Test cases in @file{ld/testsuite} for merging and link warnings.
7502 @file{binutils/readelf.c} to display your attribute.
7504 GCC, if you want the compiler to mark the attribute automatically.
7510 @node Machine Dependencies
7511 @chapter Machine Dependent Features
7513 @cindex machine dependencies
7514 The machine instruction sets are (almost by definition) different on
7515 each machine where @command{@value{AS}} runs. Floating point representations
7516 vary as well, and @command{@value{AS}} often supports a few additional
7517 directives or command-line options for compatibility with other
7518 assemblers on a particular platform. Finally, some versions of
7519 @command{@value{AS}} support special pseudo-instructions for branch
7522 This chapter discusses most of these differences, though it does not
7523 include details on any machine's instruction set. For details on that
7524 subject, see the hardware manufacturer's manual.
7528 * AArch64-Dependent:: AArch64 Dependent Features
7531 * Alpha-Dependent:: Alpha Dependent Features
7534 * ARC-Dependent:: ARC Dependent Features
7537 * ARM-Dependent:: ARM Dependent Features
7540 * AVR-Dependent:: AVR Dependent Features
7543 * Blackfin-Dependent:: Blackfin Dependent Features
7546 * CR16-Dependent:: CR16 Dependent Features
7549 * CRIS-Dependent:: CRIS Dependent Features
7552 * D10V-Dependent:: D10V Dependent Features
7555 * D30V-Dependent:: D30V Dependent Features
7558 * Epiphany-Dependent:: EPIPHANY Dependent Features
7561 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7564 * HPPA-Dependent:: HPPA Dependent Features
7567 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
7570 * IA-64-Dependent:: Intel IA-64 Dependent Features
7573 * IP2K-Dependent:: IP2K Dependent Features
7576 * LM32-Dependent:: LM32 Dependent Features
7579 * M32C-Dependent:: M32C Dependent Features
7582 * M32R-Dependent:: M32R Dependent Features
7585 * M68K-Dependent:: M680x0 Dependent Features
7588 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7591 * S12Z-Dependent:: S12Z Dependent Features
7594 * Meta-Dependent :: Meta Dependent Features
7597 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7600 * MIPS-Dependent:: MIPS Dependent Features
7603 * MMIX-Dependent:: MMIX Dependent Features
7606 * MSP430-Dependent:: MSP430 Dependent Features
7609 * NDS32-Dependent:: Andes NDS32 Dependent Features
7612 * NiosII-Dependent:: Altera Nios II Dependent Features
7615 * NS32K-Dependent:: NS32K Dependent Features
7618 * PDP-11-Dependent:: PDP-11 Dependent Features
7621 * PJ-Dependent:: picoJava Dependent Features
7624 * PPC-Dependent:: PowerPC Dependent Features
7627 * PRU-Dependent:: PRU Dependent Features
7630 * RISC-V-Dependent:: RISC-V Dependent Features
7633 * RL78-Dependent:: RL78 Dependent Features
7636 * RX-Dependent:: RX Dependent Features
7639 * S/390-Dependent:: IBM S/390 Dependent Features
7642 * SCORE-Dependent:: SCORE Dependent Features
7645 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7648 * Sparc-Dependent:: SPARC Dependent Features
7651 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7654 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7657 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7660 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7663 * V850-Dependent:: V850 Dependent Features
7666 * Vax-Dependent:: VAX Dependent Features
7669 * Visium-Dependent:: Visium Dependent Features
7672 * WebAssembly-Dependent:: WebAssembly Dependent Features
7675 * XGATE-Dependent:: XGATE Dependent Features
7678 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7681 * Xtensa-Dependent:: Xtensa Dependent Features
7684 * Z80-Dependent:: Z80 Dependent Features
7687 * Z8000-Dependent:: Z8000 Dependent Features
7694 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7695 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7696 @c peculiarity: to preserve cross-references, there must be a node called
7697 @c "Machine Dependencies". Hence the conditional nodenames in each
7698 @c major node below. Node defaulting in makeinfo requires adjacency of
7699 @c node and sectioning commands; hence the repetition of @chapter BLAH
7700 @c in both conditional blocks.
7703 @include c-aarch64.texi
7707 @include c-alpha.texi
7723 @include c-bfin.texi
7727 @include c-cr16.texi
7731 @include c-cris.texi
7736 @node Machine Dependencies
7737 @chapter Machine Dependent Features
7739 The machine instruction sets are different on each Renesas chip family,
7740 and there are also some syntax differences among the families. This
7741 chapter describes the specific @command{@value{AS}} features for each
7745 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7746 * SH-Dependent:: Renesas SH Dependent Features
7753 @include c-d10v.texi
7757 @include c-d30v.texi
7761 @include c-epiphany.texi
7765 @include c-h8300.texi
7769 @include c-hppa.texi
7773 @include c-i386.texi
7777 @include c-ia64.texi
7781 @include c-ip2k.texi
7785 @include c-lm32.texi
7789 @include c-m32c.texi
7793 @include c-m32r.texi
7797 @include c-m68k.texi
7801 @include c-m68hc11.texi
7805 @include c-s12z.texi
7809 @include c-metag.texi
7813 @include c-microblaze.texi
7817 @include c-mips.texi
7821 @include c-mmix.texi
7825 @include c-msp430.texi
7829 @include c-nds32.texi
7833 @include c-nios2.texi
7837 @include c-ns32k.texi
7841 @include c-pdp11.texi
7857 @include c-riscv.texi
7861 @include c-rl78.texi
7869 @include c-s390.texi
7873 @include c-score.texi
7881 @include c-sparc.texi
7885 @include c-tic54x.texi
7889 @include c-tic6x.texi
7893 @include c-tilegx.texi
7897 @include c-tilepro.texi
7901 @include c-v850.texi
7909 @include c-visium.texi
7913 @include c-wasm32.texi
7917 @include c-xgate.texi
7921 @include c-xstormy16.texi
7925 @include c-xtensa.texi
7937 @c reverse effect of @down at top of generic Machine-Dep chapter
7941 @node Reporting Bugs
7942 @chapter Reporting Bugs
7943 @cindex bugs in assembler
7944 @cindex reporting bugs in assembler
7946 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7948 Reporting a bug may help you by bringing a solution to your problem, or it may
7949 not. But in any case the principal function of a bug report is to help the
7950 entire community by making the next version of @command{@value{AS}} work better.
7951 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7953 In order for a bug report to serve its purpose, you must include the
7954 information that enables us to fix the bug.
7957 * Bug Criteria:: Have you found a bug?
7958 * Bug Reporting:: How to report bugs
7962 @section Have You Found a Bug?
7963 @cindex bug criteria
7965 If you are not sure whether you have found a bug, here are some guidelines:
7968 @cindex fatal signal
7969 @cindex assembler crash
7970 @cindex crash of assembler
7972 If the assembler gets a fatal signal, for any input whatever, that is a
7973 @command{@value{AS}} bug. Reliable assemblers never crash.
7975 @cindex error on valid input
7977 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7979 @cindex invalid input
7981 If @command{@value{AS}} does not produce an error message for invalid input, that
7982 is a bug. However, you should note that your idea of ``invalid input'' might
7983 be our idea of ``an extension'' or ``support for traditional practice''.
7986 If you are an experienced user of assemblers, your suggestions for improvement
7987 of @command{@value{AS}} are welcome in any case.
7991 @section How to Report Bugs
7993 @cindex assembler bugs, reporting
7995 A number of companies and individuals offer support for @sc{gnu} products. If
7996 you obtained @command{@value{AS}} from a support organization, we recommend you
7997 contact that organization first.
7999 You can find contact information for many support companies and
8000 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
8004 In any event, we also recommend that you send bug reports for @command{@value{AS}}
8008 The fundamental principle of reporting bugs usefully is this:
8009 @strong{report all the facts}. If you are not sure whether to state a
8010 fact or leave it out, state it!
8012 Often people omit facts because they think they know what causes the problem
8013 and assume that some details do not matter. Thus, you might assume that the
8014 name of a symbol you use in an example does not matter. Well, probably it does
8015 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
8016 happens to fetch from the location where that name is stored in memory;
8017 perhaps, if the name were different, the contents of that location would fool
8018 the assembler into doing the right thing despite the bug. Play it safe and
8019 give a specific, complete example. That is the easiest thing for you to do,
8020 and the most helpful.
8022 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
8023 it is new to us. Therefore, always write your bug reports on the assumption
8024 that the bug has not been reported previously.
8026 Sometimes people give a few sketchy facts and ask, ``Does this ring a
8027 bell?'' This cannot help us fix a bug, so it is basically useless. We
8028 respond by asking for enough details to enable us to investigate.
8029 You might as well expedite matters by sending them to begin with.
8031 To enable us to fix the bug, you should include all these things:
8035 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
8036 it with the @samp{--version} argument.
8038 Without this, we will not know whether there is any point in looking for
8039 the bug in the current version of @command{@value{AS}}.
8042 Any patches you may have applied to the @command{@value{AS}} source.
8045 The type of machine you are using, and the operating system name and
8049 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
8053 The command arguments you gave the assembler to assemble your example and
8054 observe the bug. To guarantee you will not omit something important, list them
8055 all. A copy of the Makefile (or the output from make) is sufficient.
8057 If we were to try to guess the arguments, we would probably guess wrong
8058 and then we might not encounter the bug.
8061 A complete input file that will reproduce the bug. If the bug is observed when
8062 the assembler is invoked via a compiler, send the assembler source, not the
8063 high level language source. Most compilers will produce the assembler source
8064 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
8065 the options @samp{-v --save-temps}; this will save the assembler source in a
8066 file with an extension of @file{.s}, and also show you exactly how
8067 @command{@value{AS}} is being run.
8070 A description of what behavior you observe that you believe is
8071 incorrect. For example, ``It gets a fatal signal.''
8073 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
8074 will certainly notice it. But if the bug is incorrect output, we might not
8075 notice unless it is glaringly wrong. You might as well not give us a chance to
8078 Even if the problem you experience is a fatal signal, you should still say so
8079 explicitly. Suppose something strange is going on, such as, your copy of
8080 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
8081 library on your system. (This has happened!) Your copy might crash and ours
8082 would not. If you told us to expect a crash, then when ours fails to crash, we
8083 would know that the bug was not happening for us. If you had not told us to
8084 expect a crash, then we would not be able to draw any conclusion from our
8088 If you wish to suggest changes to the @command{@value{AS}} source, send us context
8089 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
8090 option. Always send diffs from the old file to the new file. If you even
8091 discuss something in the @command{@value{AS}} source, refer to it by context, not
8094 The line numbers in our development sources will not match those in your
8095 sources. Your line numbers would convey no useful information to us.
8098 Here are some things that are not necessary:
8102 A description of the envelope of the bug.
8104 Often people who encounter a bug spend a lot of time investigating
8105 which changes to the input file will make the bug go away and which
8106 changes will not affect it.
8108 This is often time consuming and not very useful, because the way we
8109 will find the bug is by running a single example under the debugger
8110 with breakpoints, not by pure deduction from a series of examples.
8111 We recommend that you save your time for something else.
8113 Of course, if you can find a simpler example to report @emph{instead}
8114 of the original one, that is a convenience for us. Errors in the
8115 output will be easier to spot, running under the debugger will take
8116 less time, and so on.
8118 However, simplification is not vital; if you do not want to do this,
8119 report the bug anyway and send us the entire test case you used.
8122 A patch for the bug.
8124 A patch for the bug does help us if it is a good one. But do not omit
8125 the necessary information, such as the test case, on the assumption that
8126 a patch is all we need. We might see problems with your patch and decide
8127 to fix the problem another way, or we might not understand it at all.
8129 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
8130 construct an example that will make the program follow a certain path through
8131 the code. If you do not send us the example, we will not be able to construct
8132 one, so we will not be able to verify that the bug is fixed.
8134 And if we cannot understand what bug you are trying to fix, or why your
8135 patch should be an improvement, we will not install it. A test case will
8136 help us to understand.
8139 A guess about what the bug is or what it depends on.
8141 Such guesses are usually wrong. Even we cannot guess right about such
8142 things without first using the debugger to find the facts.
8145 @node Acknowledgements
8146 @chapter Acknowledgements
8148 If you have contributed to GAS and your name isn't listed here,
8149 it is not meant as a slight. We just don't know about it. Send mail to the
8150 maintainer, and we'll correct the situation. Currently
8152 the maintainer is Nick Clifton (email address @code{nickc@@redhat.com}).
8154 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
8157 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
8158 information and the 68k series machines, most of the preprocessing pass, and
8159 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
8161 K. Richard Pixley maintained GAS for a while, adding various enhancements and
8162 many bug fixes, including merging support for several processors, breaking GAS
8163 up to handle multiple object file format back ends (including heavy rewrite,
8164 testing, an integration of the coff and b.out back ends), adding configuration
8165 including heavy testing and verification of cross assemblers and file splits
8166 and renaming, converted GAS to strictly ANSI C including full prototypes, added
8167 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
8168 port (including considerable amounts of reverse engineering), a SPARC opcode
8169 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
8170 assertions and made them work, much other reorganization, cleanup, and lint.
8172 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
8173 in format-specific I/O modules.
8175 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
8176 has done much work with it since.
8178 The Intel 80386 machine description was written by Eliot Dresselhaus.
8180 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
8182 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
8183 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
8185 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
8186 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
8187 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
8188 support a.out format.
8190 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
8191 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
8192 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
8193 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
8196 John Gilmore built the AMD 29000 support, added @code{.include} support, and
8197 simplified the configuration of which versions accept which directives. He
8198 updated the 68k machine description so that Motorola's opcodes always produced
8199 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
8200 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
8201 cross-compilation support, and one bug in relaxation that took a week and
8202 required the proverbial one-bit fix.
8204 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
8205 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
8206 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
8207 PowerPC assembler, and made a few other minor patches.
8209 Steve Chamberlain made GAS able to generate listings.
8211 Hewlett-Packard contributed support for the HP9000/300.
8213 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
8214 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
8215 formats). This work was supported by both the Center for Software Science at
8216 the University of Utah and Cygnus Support.
8218 Support for ELF format files has been worked on by Mark Eichin of Cygnus
8219 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
8220 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
8221 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
8222 and some initial 64-bit support).
8224 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
8226 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
8227 support for openVMS/Alpha.
8229 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
8232 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
8233 Inc.@: added support for Xtensa processors.
8235 Several engineers at Cygnus Support have also provided many small bug fixes and
8236 configuration enhancements.
8238 Jon Beniston added support for the Lattice Mico32 architecture.
8240 Many others have contributed large or small bugfixes and enhancements. If
8241 you have contributed significant work and are not mentioned on this list, and
8242 want to be, let us know. Some of the history has been lost; we are not
8243 intentionally leaving anyone out.
8245 @node GNU Free Documentation License
8246 @appendix GNU Free Documentation License
8250 @unnumbered AS Index