1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright (C) 1991-2014 Free Software Foundation, Inc.
3 @c UPDATE!! On future updates--
4 @c (1) check for new machine-dep cmdline options in
5 @c md_parse_option definitions in config/tc-*.c
6 @c (2) for platform-specific directives, examine md_pseudo_op
8 @c (3) for object-format specific directives, examine obj_pseudo_op
10 @c (4) portable directives in potable[] in read.c
14 @macro gcctabopt{body}
17 @c defaults, config file may override:
22 @include asconfig.texi
27 @c common OR combinations of conditions
53 @set abnormal-separator
57 @settitle Using @value{AS}
60 @settitle Using @value{AS} (@value{TARGET})
62 @setchapternewpage odd
67 @c WARE! Some of the machine-dependent sections contain tables of machine
68 @c instructions. Except in multi-column format, these tables look silly.
69 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
70 @c the multi-col format is faked within @example sections.
72 @c Again unfortunately, the natural size that fits on a page, for these tables,
73 @c is different depending on whether or not smallbook is turned on.
74 @c This matters, because of order: text flow switches columns at each page
77 @c The format faked in this source works reasonably well for smallbook,
78 @c not well for the default large-page format. This manual expects that if you
79 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
80 @c tables in question. You can turn on one without the other at your
81 @c discretion, of course.
84 @c the insn tables look just as silly in info files regardless of smallbook,
85 @c might as well show 'em anyways.
89 @dircategory Software development
91 * As: (as). The GNU assembler.
92 * Gas: (as). The GNU assembler.
100 This file documents the GNU Assembler "@value{AS}".
102 @c man begin COPYRIGHT
103 Copyright @copyright{} 1991-2014 Free Software Foundation, Inc.
105 Permission is granted to copy, distribute and/or modify this document
106 under the terms of the GNU Free Documentation License, Version 1.3
107 or any later version published by the Free Software Foundation;
108 with no Invariant Sections, with no Front-Cover Texts, and with no
109 Back-Cover Texts. A copy of the license is included in the
110 section entitled ``GNU Free Documentation License''.
116 @title Using @value{AS}
117 @subtitle The @sc{gnu} Assembler
119 @subtitle for the @value{TARGET} family
121 @ifset VERSION_PACKAGE
123 @subtitle @value{VERSION_PACKAGE}
126 @subtitle Version @value{VERSION}
129 The Free Software Foundation Inc.@: thanks The Nice Computer
130 Company of Australia for loaning Dean Elsner to write the
131 first (Vax) version of @command{as} for Project @sc{gnu}.
132 The proprietors, management and staff of TNCCA thank FSF for
133 distracting the boss while they got some work
136 @author Dean Elsner, Jay Fenlason & friends
140 \hfill {\it Using {\tt @value{AS}}}\par
141 \hfill Edited by Cygnus Support\par
143 %"boxit" macro for figures:
144 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
145 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
146 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
147 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
148 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
151 @vskip 0pt plus 1filll
152 Copyright @copyright{} 1991-2014 Free Software Foundation, Inc.
154 Permission is granted to copy, distribute and/or modify this document
155 under the terms of the GNU Free Documentation License, Version 1.3
156 or any later version published by the Free Software Foundation;
157 with no Invariant Sections, with no Front-Cover Texts, and with no
158 Back-Cover Texts. A copy of the license is included in the
159 section entitled ``GNU Free Documentation License''.
166 @top Using @value{AS}
168 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}}
169 @ifset VERSION_PACKAGE
170 @value{VERSION_PACKAGE}
172 version @value{VERSION}.
174 This version of the file describes @command{@value{AS}} configured to generate
175 code for @value{TARGET} architectures.
178 This document is distributed under the terms of the GNU Free
179 Documentation License. A copy of the license is included in the
180 section entitled ``GNU Free Documentation License''.
183 * Overview:: Overview
184 * Invoking:: Command-Line Options
186 * Sections:: Sections and Relocation
188 * Expressions:: Expressions
189 * Pseudo Ops:: Assembler Directives
191 * Object Attributes:: Object Attributes
193 * Machine Dependencies:: Machine Dependent Features
194 * Reporting Bugs:: Reporting Bugs
195 * Acknowledgements:: Who Did What
196 * GNU Free Documentation License:: GNU Free Documentation License
197 * AS Index:: AS Index
204 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
206 This version of the manual describes @command{@value{AS}} configured to generate
207 code for @value{TARGET} architectures.
211 @cindex invocation summary
212 @cindex option summary
213 @cindex summary of options
214 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
215 see @ref{Invoking,,Command-Line Options}.
217 @c man title AS the portable GNU assembler.
221 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
225 @c We don't use deffn and friends for the following because they seem
226 @c to be limited to one line for the header.
228 @c man begin SYNOPSIS
229 @value{AS} [@b{-a}[@b{cdghlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
230 [@b{--compress-debug-sections}] [@b{--nocompress-debug-sections}]
231 [@b{--debug-prefix-map} @var{old}=@var{new}]
232 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
233 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--gdwarf-sections}]
234 [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
235 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
236 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
237 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
238 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
239 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
240 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
241 [@b{--size-check=[error|warning]}]
242 [@b{--target-help}] [@var{target-options}]
243 [@b{--}|@var{files} @dots{}]
245 @c Target dependent options are listed below. Keep the list sorted.
246 @c Add an empty line for separation.
249 @emph{Target AArch64 options:}
251 [@b{-mabi}=@var{ABI}]
255 @emph{Target Alpha options:}
257 [@b{-mdebug} | @b{-no-mdebug}]
258 [@b{-replace} | @b{-noreplace}]
259 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
260 [@b{-F}] [@b{-32addr}]
264 @emph{Target ARC options:}
270 @emph{Target ARM options:}
271 @c Don't document the deprecated options
272 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
273 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
274 [@b{-mfpu}=@var{floating-point-format}]
275 [@b{-mfloat-abi}=@var{abi}]
276 [@b{-meabi}=@var{ver}]
279 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
280 @b{-mapcs-reentrant}]
281 [@b{-mthumb-interwork}] [@b{-k}]
285 @emph{Target Blackfin options:}
286 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
293 @emph{Target CRIS options:}
294 [@b{--underscore} | @b{--no-underscore}]
296 [@b{--emulation=criself} | @b{--emulation=crisaout}]
297 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
298 @c Deprecated -- deliberately not documented.
303 @emph{Target D10V options:}
308 @emph{Target D30V options:}
309 [@b{-O}|@b{-n}|@b{-N}]
313 @emph{Target EPIPHANY options:}
314 [@b{-mepiphany}|@b{-mepiphany16}]
318 @emph{Target H8/300 options:}
322 @c HPPA has no machine-dependent assembler options (yet).
326 @emph{Target i386 options:}
327 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
328 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
332 @emph{Target i960 options:}
333 @c see md_parse_option in tc-i960.c
334 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
336 [@b{-b}] [@b{-no-relax}]
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{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
403 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
404 [@b{-mips64}] [@b{-mips64r2}]
405 [@b{-construct-floats}] [@b{-no-construct-floats}]
406 [@b{-mnan=@var{encoding}}]
407 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
408 [@b{-mips16}] [@b{-no-mips16}]
409 [@b{-mmicromips}] [@b{-mno-micromips}]
410 [@b{-msmartmips}] [@b{-mno-smartmips}]
411 [@b{-mips3d}] [@b{-no-mips3d}]
412 [@b{-mdmx}] [@b{-no-mdmx}]
413 [@b{-mdsp}] [@b{-mno-dsp}]
414 [@b{-mdspr2}] [@b{-mno-dspr2}]
415 [@b{-mmsa}] [@b{-mno-msa}]
416 [@b{-mmt}] [@b{-mno-mt}]
417 [@b{-mmcu}] [@b{-mno-mcu}]
418 [@b{-minsn32}] [@b{-mno-insn32}]
419 [@b{-mfix7000}] [@b{-mno-fix7000}]
420 [@b{-mfix-rm7000}] [@b{-mno-fix-rm7000}]
421 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
422 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
423 [@b{-mdebug}] [@b{-no-mdebug}]
424 [@b{-mpdr}] [@b{-mno-pdr}]
428 @emph{Target MMIX options:}
429 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
430 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
431 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
432 [@b{--linker-allocated-gregs}]
436 @emph{Target Nios II options:}
437 [@b{-relax-all}] [@b{-relax-section}] [@b{-no-relax}]
442 @emph{Target NDS32 options:}
443 [@b{-EL}] [@b{-EB}] [@b{-O}] [@b{-Os}] [@b{-mcpu=@var{cpu}}]
444 [@b{-misa=@var{isa}}] [@b{-mabi=@var{abi}}] [@b{-mall-ext}]
445 [@b{-m[no-]16-bit}] [@b{-m[no-]perf-ext}] [@b{-m[no-]perf2-ext}]
446 [@b{-m[no-]string-ext}] [@b{-m[no-]dsp-ext}] [@b{-m[no-]mac}] [@b{-m[no-]div}]
447 [@b{-m[no-]audio-isa-ext}] [@b{-m[no-]fpu-sp-ext}] [@b{-m[no-]fpu-dp-ext}]
448 [@b{-m[no-]fpu-fma}] [@b{-mfpu-freg=@var{FREG}}] [@b{-mreduced-regs}]
449 [@b{-mfull-regs}] [@b{-m[no-]dx-regs}] [@b{-mpic}] [@b{-mno-relax}]
454 @emph{Target PDP11 options:}
455 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
456 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
457 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
461 @emph{Target picoJava options:}
466 @emph{Target PowerPC options:}
468 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
469 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
470 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
471 @b{-mbooke}|@b{-mpower4}|@b{-mpwr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
472 @b{-mpower7}|@b{-mpwr7}|@b{-mpower8}|@b{-mpwr8}|@b{-ma2}|@b{-mcell}|@b{-mspe}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
473 [@b{-many}] [@b{-maltivec}|@b{-mvsx}|@b{-mhtm}|@b{-mvle}]
474 [@b{-mregnames}|@b{-mno-regnames}]
475 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
476 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
477 [@b{-msolaris}|@b{-mno-solaris}]
478 [@b{-nops=@var{count}}]
482 @emph{Target RX options:}
483 [@b{-mlittle-endian}|@b{-mbig-endian}]
484 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
485 [@b{-muse-conventional-section-names}]
486 [@b{-msmall-data-limit}]
489 [@b{-mint-register=@var{number}}]
490 [@b{-mgcc-abi}|@b{-mrx-abi}]
494 @emph{Target s390 options:}
495 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
496 [@b{-mregnames}|@b{-mno-regnames}]
497 [@b{-mwarn-areg-zero}]
501 @emph{Target SCORE options:}
502 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
503 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
504 [@b{-march=score7}][@b{-march=score3}]
505 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
509 @emph{Target SPARC options:}
510 @c The order here is important. See c-sparc.texi.
511 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
512 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
513 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
518 @emph{Target TIC54X options:}
519 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
520 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
525 @emph{Target TIC6X options:}
526 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
527 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
528 [@b{-mpic}|@b{-mno-pic}]
532 @emph{Target TILE-Gx options:}
533 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
536 @c TILEPro has no machine-dependent assembler options
541 @emph{Target Xtensa options:}
542 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
543 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
544 [@b{--[no-]transform}]
545 [@b{--rename-section} @var{oldname}=@var{newname}]
550 @emph{Target Z80 options:}
551 [@b{-z80}] [@b{-r800}]
552 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
553 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
554 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
555 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
556 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
557 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
561 @c Z8000 has no machine-dependent assembler options
570 @include at-file.texi
573 Turn on listings, in any of a variety of ways:
577 omit false conditionals
580 omit debugging directives
583 include general information, like @value{AS} version and options passed
586 include high-level source
592 include macro expansions
595 omit forms processing
601 set the name of the listing file
604 You may combine these options; for example, use @samp{-aln} for assembly
605 listing without forms processing. The @samp{=file} option, if used, must be
606 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
609 Begin in alternate macro mode.
611 @xref{Altmacro,,@code{.altmacro}}.
614 @item --compress-debug-sections
615 Compress DWARF debug sections using zlib. The debug sections are renamed
616 to begin with @samp{.zdebug}, and the resulting object file may not be
617 compatible with older linkers and object file utilities.
619 @item --nocompress-debug-sections
620 Do not compress DWARF debug sections. This is the default.
623 Ignored. This option is accepted for script compatibility with calls to
626 @item --debug-prefix-map @var{old}=@var{new}
627 When assembling files in directory @file{@var{old}}, record debugging
628 information describing them as in @file{@var{new}} instead.
630 @item --defsym @var{sym}=@var{value}
631 Define the symbol @var{sym} to be @var{value} before assembling the input file.
632 @var{value} must be an integer constant. As in C, a leading @samp{0x}
633 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
634 value. The value of the symbol can be overridden inside a source file via the
635 use of a @code{.set} pseudo-op.
638 ``fast''---skip whitespace and comment preprocessing (assume source is
643 Generate debugging information for each assembler source line using whichever
644 debug format is preferred by the target. This currently means either STABS,
648 Generate stabs debugging information for each assembler line. This
649 may help debugging assembler code, if the debugger can handle it.
652 Generate stabs debugging information for each assembler line, with GNU
653 extensions that probably only gdb can handle, and that could make other
654 debuggers crash or refuse to read your program. This
655 may help debugging assembler code. Currently the only GNU extension is
656 the location of the current working directory at assembling time.
659 Generate DWARF2 debugging information for each assembler line. This
660 may help debugging assembler code, if the debugger can handle it. Note---this
661 option is only supported by some targets, not all of them.
663 @item --gdwarf-sections
664 Instead of creating a .debug_line section, create a series of
665 .debug_line.@var{foo} sections where @var{foo} is the name of the
666 corresponding code section. For example a code section called @var{.text.func}
667 will have its dwarf line number information placed into a section called
668 @var{.debug_line.text.func}. If the code section is just called @var{.text}
669 then debug line section will still be called just @var{.debug_line} without any
672 @item --size-check=error
673 @itemx --size-check=warning
674 Issue an error or warning for invalid ELF .size directive.
677 Print a summary of the command line options and exit.
680 Print a summary of all target specific options and exit.
683 Add directory @var{dir} to the search list for @code{.include} directives.
686 Don't warn about signed overflow.
689 @ifclear DIFF-TBL-KLUGE
690 This option is accepted but has no effect on the @value{TARGET} family.
692 @ifset DIFF-TBL-KLUGE
693 Issue warnings when difference tables altered for long displacements.
698 Keep (in the symbol table) local symbols. These symbols start with
699 system-specific local label prefixes, typically @samp{.L} for ELF systems
700 or @samp{L} for traditional a.out systems.
705 @item --listing-lhs-width=@var{number}
706 Set the maximum width, in words, of the output data column for an assembler
707 listing to @var{number}.
709 @item --listing-lhs-width2=@var{number}
710 Set the maximum width, in words, of the output data column for continuation
711 lines in an assembler listing to @var{number}.
713 @item --listing-rhs-width=@var{number}
714 Set the maximum width of an input source line, as displayed in a listing, to
717 @item --listing-cont-lines=@var{number}
718 Set the maximum number of lines printed in a listing for a single line of input
721 @item -o @var{objfile}
722 Name the object-file output from @command{@value{AS}} @var{objfile}.
725 Fold the data section into the text section.
727 @kindex --hash-size=@var{number}
728 Set the default size of GAS's hash tables to a prime number close to
729 @var{number}. Increasing this value can reduce the length of time it takes the
730 assembler to perform its tasks, at the expense of increasing the assembler's
731 memory requirements. Similarly reducing this value can reduce the memory
732 requirements at the expense of speed.
734 @item --reduce-memory-overheads
735 This option reduces GAS's memory requirements, at the expense of making the
736 assembly processes slower. Currently this switch is a synonym for
737 @samp{--hash-size=4051}, but in the future it may have other effects as well.
740 Print the maximum space (in bytes) and total time (in seconds) used by
743 @item --strip-local-absolute
744 Remove local absolute symbols from the outgoing symbol table.
748 Print the @command{as} version.
751 Print the @command{as} version and exit.
755 Suppress warning messages.
757 @item --fatal-warnings
758 Treat warnings as errors.
761 Don't suppress warning messages or treat them as errors.
770 Generate an object file even after errors.
772 @item -- | @var{files} @dots{}
773 Standard input, or source files to assemble.
781 @xref{AArch64 Options}, for the options available when @value{AS} is configured
782 for the 64-bit mode of the ARM Architecture (AArch64).
787 The following options are available when @value{AS} is configured for the
788 64-bit mode of the ARM Architecture (AArch64).
791 @include c-aarch64.texi
792 @c ended inside the included file
800 @xref{Alpha Options}, for the options available when @value{AS} is configured
801 for an Alpha processor.
806 The following options are available when @value{AS} is configured for an Alpha
810 @include c-alpha.texi
811 @c ended inside the included file
818 The following options are available when @value{AS} is configured for
823 This option selects the core processor variant.
825 Select either big-endian (-EB) or little-endian (-EL) output.
830 The following options are available when @value{AS} is configured for the ARM
834 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
835 Specify which ARM processor variant is the target.
836 @item -march=@var{architecture}[+@var{extension}@dots{}]
837 Specify which ARM architecture variant is used by the target.
838 @item -mfpu=@var{floating-point-format}
839 Select which Floating Point architecture is the target.
840 @item -mfloat-abi=@var{abi}
841 Select which floating point ABI is in use.
843 Enable Thumb only instruction decoding.
844 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
845 Select which procedure calling convention is in use.
847 Select either big-endian (-EB) or little-endian (-EL) output.
848 @item -mthumb-interwork
849 Specify that the code has been generated with interworking between Thumb and
852 Specify that PIC code has been generated.
860 @xref{Blackfin Options}, for the options available when @value{AS} is
861 configured for the Blackfin processor family.
866 The following options are available when @value{AS} is configured for
867 the Blackfin processor family.
871 @c ended inside the included file
878 See the info pages for documentation of the CRIS-specific options.
882 The following options are available when @value{AS} is configured for
885 @cindex D10V optimization
886 @cindex optimization, D10V
888 Optimize output by parallelizing instructions.
893 The following options are available when @value{AS} is configured for a D30V
896 @cindex D30V optimization
897 @cindex optimization, D30V
899 Optimize output by parallelizing instructions.
903 Warn when nops are generated.
905 @cindex D30V nops after 32-bit multiply
907 Warn when a nop after a 32-bit multiply instruction is generated.
913 The following options are available when @value{AS} is configured for the
914 Adapteva EPIPHANY series.
917 @xref{Epiphany Options}, for the options available when @value{AS} is
918 configured for an Epiphany processor.
923 The following options are available when @value{AS} is configured for
924 an Epiphany processor.
927 @include c-epiphany.texi
928 @c ended inside the included file
936 @xref{H8/300 Options}, for the options available when @value{AS} is configured
937 for an H8/300 processor.
942 The following options are available when @value{AS} is configured for an H8/300
946 @include c-h8300.texi
947 @c ended inside the included file
955 @xref{i386-Options}, for the options available when @value{AS} is
956 configured for an i386 processor.
961 The following options are available when @value{AS} is configured for
966 @c ended inside the included file
973 The following options are available when @value{AS} is configured for the
974 Intel 80960 processor.
977 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
978 Specify which variant of the 960 architecture is the target.
981 Add code to collect statistics about branches taken.
984 Do not alter compare-and-branch instructions for long displacements;
991 The following options are available when @value{AS} is configured for the
997 Specifies that the extended IP2022 instructions are allowed.
1000 Restores the default behaviour, which restricts the permitted instructions to
1001 just the basic IP2022 ones.
1007 The following options are available when @value{AS} is configured for the
1008 Renesas M32C and M16C processors.
1013 Assemble M32C instructions.
1016 Assemble M16C instructions (the default).
1019 Enable support for link-time relaxations.
1022 Support H'00 style hex constants in addition to 0x00 style.
1028 The following options are available when @value{AS} is configured for the
1029 Renesas M32R (formerly Mitsubishi M32R) series.
1034 Specify which processor in the M32R family is the target. The default
1035 is normally the M32R, but this option changes it to the M32RX.
1037 @item --warn-explicit-parallel-conflicts or --Wp
1038 Produce warning messages when questionable parallel constructs are
1041 @item --no-warn-explicit-parallel-conflicts or --Wnp
1042 Do not produce warning messages when questionable parallel constructs are
1049 The following options are available when @value{AS} is configured for the
1050 Motorola 68000 series.
1055 Shorten references to undefined symbols, to one word instead of two.
1057 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1058 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1059 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1060 Specify what processor in the 68000 family is the target. The default
1061 is normally the 68020, but this can be changed at configuration time.
1063 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1064 The target machine does (or does not) have a floating-point coprocessor.
1065 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1066 the basic 68000 is not compatible with the 68881, a combination of the
1067 two can be specified, since it's possible to do emulation of the
1068 coprocessor instructions with the main processor.
1070 @item -m68851 | -mno-68851
1071 The target machine does (or does not) have a memory-management
1072 unit coprocessor. The default is to assume an MMU for 68020 and up.
1080 @xref{Nios II Options}, for the options available when @value{AS} is configured
1081 for an Altera Nios II processor.
1085 @c man begin OPTIONS
1086 The following options are available when @value{AS} is configured for an
1087 Altera Nios II processor.
1089 @c man begin INCLUDE
1090 @include c-nios2.texi
1091 @c ended inside the included file
1097 For details about the PDP-11 machine dependent features options,
1098 see @ref{PDP-11-Options}.
1101 @item -mpic | -mno-pic
1102 Generate position-independent (or position-dependent) code. The
1103 default is @option{-mpic}.
1106 @itemx -mall-extensions
1107 Enable all instruction set extensions. This is the default.
1109 @item -mno-extensions
1110 Disable all instruction set extensions.
1112 @item -m@var{extension} | -mno-@var{extension}
1113 Enable (or disable) a particular instruction set extension.
1116 Enable the instruction set extensions supported by a particular CPU, and
1117 disable all other extensions.
1119 @item -m@var{machine}
1120 Enable the instruction set extensions supported by a particular machine
1121 model, and disable all other extensions.
1127 The following options are available when @value{AS} is configured for
1128 a picoJava processor.
1132 @cindex PJ endianness
1133 @cindex endianness, PJ
1134 @cindex big endian output, PJ
1136 Generate ``big endian'' format output.
1138 @cindex little endian output, PJ
1140 Generate ``little endian'' format output.
1146 The following options are available when @value{AS} is configured for the
1147 Motorola 68HC11 or 68HC12 series.
1151 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1152 Specify what processor is the target. The default is
1153 defined by the configuration option when building the assembler.
1155 @item --xgate-ramoffset
1156 Instruct the linker to offset RAM addresses from S12X address space into
1157 XGATE address space.
1160 Specify to use the 16-bit integer ABI.
1163 Specify to use the 32-bit integer ABI.
1165 @item -mshort-double
1166 Specify to use the 32-bit double ABI.
1169 Specify to use the 64-bit double ABI.
1171 @item --force-long-branches
1172 Relative branches are turned into absolute ones. This concerns
1173 conditional branches, unconditional branches and branches to a
1176 @item -S | --short-branches
1177 Do not turn relative branches into absolute ones
1178 when the offset is out of range.
1180 @item --strict-direct-mode
1181 Do not turn the direct addressing mode into extended addressing mode
1182 when the instruction does not support direct addressing mode.
1184 @item --print-insn-syntax
1185 Print the syntax of instruction in case of error.
1187 @item --print-opcodes
1188 Print the list of instructions with syntax and then exit.
1190 @item --generate-example
1191 Print an example of instruction for each possible instruction and then exit.
1192 This option is only useful for testing @command{@value{AS}}.
1198 The following options are available when @command{@value{AS}} is configured
1199 for the SPARC architecture:
1202 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1203 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1204 Explicitly select a variant of the SPARC architecture.
1206 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1207 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1209 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1210 UltraSPARC extensions.
1212 @item -xarch=v8plus | -xarch=v8plusa
1213 For compatibility with the Solaris v9 assembler. These options are
1214 equivalent to -Av8plus and -Av8plusa, respectively.
1217 Warn when the assembler switches to another architecture.
1222 The following options are available when @value{AS} is configured for the 'c54x
1227 Enable extended addressing mode. All addresses and relocations will assume
1228 extended addressing (usually 23 bits).
1229 @item -mcpu=@var{CPU_VERSION}
1230 Sets the CPU version being compiled for.
1231 @item -merrors-to-file @var{FILENAME}
1232 Redirect error output to a file, for broken systems which don't support such
1233 behaviour in the shell.
1238 The following options are available when @value{AS} is configured for
1243 This option sets the largest size of an object that can be referenced
1244 implicitly with the @code{gp} register. It is only accepted for targets that
1245 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1247 @cindex MIPS endianness
1248 @cindex endianness, MIPS
1249 @cindex big endian output, MIPS
1251 Generate ``big endian'' format output.
1253 @cindex little endian output, MIPS
1255 Generate ``little endian'' format output.
1267 Generate code for a particular MIPS Instruction Set Architecture level.
1268 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1269 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1270 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1271 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
1272 @samp{-mips64r2} correspond to generic MIPS V, MIPS32, MIPS32 Release 2,
1273 MIPS64, and MIPS64 Release 2 ISA processors, respectively.
1275 @item -march=@var{cpu}
1276 Generate code for a particular MIPS CPU.
1278 @item -mtune=@var{cpu}
1279 Schedule and tune for a particular MIPS CPU.
1283 Cause nops to be inserted if the read of the destination register
1284 of an mfhi or mflo instruction occurs in the following two instructions.
1287 @itemx -mno-fix-rm7000
1288 Cause nops to be inserted if a dmult or dmultu instruction is
1289 followed by a load instruction.
1293 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1294 section instead of the standard ELF .stabs sections.
1298 Control generation of @code{.pdr} sections.
1302 The register sizes are normally inferred from the ISA and ABI, but these
1303 flags force a certain group of registers to be treated as 32 bits wide at
1304 all times. @samp{-mgp32} controls the size of general-purpose registers
1305 and @samp{-mfp32} controls the size of floating-point registers.
1309 Generate code for the MIPS 16 processor. This is equivalent to putting
1310 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1311 turns off this option.
1314 @itemx -mno-micromips
1315 Generate code for the microMIPS processor. This is equivalent to putting
1316 @code{.set micromips} at the start of the assembly file. @samp{-mno-micromips}
1317 turns off this option. This is equivalent to putting @code{.set nomicromips}
1318 at the start of the assembly file.
1321 @itemx -mno-smartmips
1322 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1323 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1324 @samp{-mno-smartmips} turns off this option.
1328 Generate code for the MIPS-3D Application Specific Extension.
1329 This tells the assembler to accept MIPS-3D instructions.
1330 @samp{-no-mips3d} turns off this option.
1334 Generate code for the MDMX Application Specific Extension.
1335 This tells the assembler to accept MDMX instructions.
1336 @samp{-no-mdmx} turns off this option.
1340 Generate code for the DSP Release 1 Application Specific Extension.
1341 This tells the assembler to accept DSP Release 1 instructions.
1342 @samp{-mno-dsp} turns off this option.
1346 Generate code for the DSP Release 2 Application Specific Extension.
1347 This option implies -mdsp.
1348 This tells the assembler to accept DSP Release 2 instructions.
1349 @samp{-mno-dspr2} turns off this option.
1353 Generate code for the MIPS SIMD Architecture Extension.
1354 This tells the assembler to accept MSA instructions.
1355 @samp{-mno-msa} turns off this option.
1359 Generate code for the MT Application Specific Extension.
1360 This tells the assembler to accept MT instructions.
1361 @samp{-mno-mt} turns off this option.
1365 Generate code for the MCU Application Specific Extension.
1366 This tells the assembler to accept MCU instructions.
1367 @samp{-mno-mcu} turns off this option.
1371 Only use 32-bit instruction encodings when generating code for the
1372 microMIPS processor. This option inhibits the use of any 16-bit
1373 instructions. This is equivalent to putting @code{.set insn32} at
1374 the start of the assembly file. @samp{-mno-insn32} turns off this
1375 option. This is equivalent to putting @code{.set noinsn32} at the
1376 start of the assembly file. By default @samp{-mno-insn32} is
1377 selected, allowing all instructions to be used.
1379 @item --construct-floats
1380 @itemx --no-construct-floats
1381 The @samp{--no-construct-floats} option disables the construction of
1382 double width floating point constants by loading the two halves of the
1383 value into the two single width floating point registers that make up
1384 the double width register. By default @samp{--construct-floats} is
1385 selected, allowing construction of these floating point constants.
1387 @item --relax-branch
1388 @itemx --no-relax-branch
1389 The @samp{--relax-branch} option enables the relaxation of out-of-range
1390 branches. By default @samp{--no-relax-branch} is selected, causing any
1391 out-of-range branches to produce an error.
1393 @item -mnan=@var{encoding}
1394 Select between the IEEE 754-2008 (@option{-mnan=2008}) or the legacy
1395 (@option{-mnan=legacy}) NaN encoding format. The latter is the default.
1398 @item --emulation=@var{name}
1399 This option was formerly used to switch between ELF and ECOFF output
1400 on targets like IRIX 5 that supported both. MIPS ECOFF support was
1401 removed in GAS 2.24, so the option now serves little purpose.
1402 It is retained for backwards compatibility.
1404 The available configuration names are: @samp{mipself}, @samp{mipslelf} and
1405 @samp{mipsbelf}. Choosing @samp{mipself} now has no effect, since the output
1406 is always ELF. @samp{mipslelf} and @samp{mipsbelf} select little- and
1407 big-endian output respectively, but @samp{-EL} and @samp{-EB} are now the
1408 preferred options instead.
1411 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1418 Control how to deal with multiplication overflow and division by zero.
1419 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1420 (and only work for Instruction Set Architecture level 2 and higher);
1421 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1425 When this option is used, @command{@value{AS}} will issue a warning every
1426 time it generates a nop instruction from a macro.
1431 The following options are available when @value{AS} is configured for
1437 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1438 The command line option @samp{-nojsri2bsr} can be used to disable it.
1442 Enable or disable the silicon filter behaviour. By default this is disabled.
1443 The default can be overridden by the @samp{-sifilter} command line option.
1446 Alter jump instructions for long displacements.
1448 @item -mcpu=[210|340]
1449 Select the cpu type on the target hardware. This controls which instructions
1453 Assemble for a big endian target.
1456 Assemble for a little endian target.
1465 @xref{Meta Options}, for the options available when @value{AS} is configured
1466 for a Meta processor.
1470 @c man begin OPTIONS
1471 The following options are available when @value{AS} is configured for a
1474 @c man begin INCLUDE
1475 @include c-metag.texi
1476 @c ended inside the included file
1481 @c man begin OPTIONS
1483 See the info pages for documentation of the MMIX-specific options.
1489 @xref{NDS32 Options}, for the options available when @value{AS} is configured
1490 for a NDS32 processor.
1492 @c ended inside the included file
1496 @c man begin OPTIONS
1497 The following options are available when @value{AS} is configured for a
1500 @c man begin INCLUDE
1501 @include c-nds32.texi
1502 @c ended inside the included file
1509 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1510 for a PowerPC processor.
1514 @c man begin OPTIONS
1515 The following options are available when @value{AS} is configured for a
1518 @c man begin INCLUDE
1520 @c ended inside the included file
1525 @c man begin OPTIONS
1527 See the info pages for documentation of the RX-specific options.
1531 The following options are available when @value{AS} is configured for the s390
1537 Select the word size, either 31/32 bits or 64 bits.
1540 Select the architecture mode, either the Enterprise System
1541 Architecture (esa) or the z/Architecture mode (zarch).
1542 @item -march=@var{processor}
1543 Specify which s390 processor variant is the target, @samp{g6}, @samp{g6},
1544 @samp{z900}, @samp{z990}, @samp{z9-109}, @samp{z9-ec}, @samp{z10},
1545 @samp{z196}, or @samp{zEC12}.
1547 @itemx -mno-regnames
1548 Allow or disallow symbolic names for registers.
1549 @item -mwarn-areg-zero
1550 Warn whenever the operand for a base or index register has been specified
1551 but evaluates to zero.
1559 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1560 for a TMS320C6000 processor.
1564 @c man begin OPTIONS
1565 The following options are available when @value{AS} is configured for a
1566 TMS320C6000 processor.
1568 @c man begin INCLUDE
1569 @include c-tic6x.texi
1570 @c ended inside the included file
1578 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1579 for a TILE-Gx processor.
1583 @c man begin OPTIONS
1584 The following options are available when @value{AS} is configured for a TILE-Gx
1587 @c man begin INCLUDE
1588 @include c-tilegx.texi
1589 @c ended inside the included file
1597 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1598 for an Xtensa processor.
1602 @c man begin OPTIONS
1603 The following options are available when @value{AS} is configured for an
1606 @c man begin INCLUDE
1607 @include c-xtensa.texi
1608 @c ended inside the included file
1613 @c man begin OPTIONS
1616 The following options are available when @value{AS} is configured for
1617 a Z80 family processor.
1620 Assemble for Z80 processor.
1622 Assemble for R800 processor.
1623 @item -ignore-undocumented-instructions
1625 Assemble undocumented Z80 instructions that also work on R800 without warning.
1626 @item -ignore-unportable-instructions
1628 Assemble all undocumented Z80 instructions without warning.
1629 @item -warn-undocumented-instructions
1631 Issue a warning for undocumented Z80 instructions that also work on R800.
1632 @item -warn-unportable-instructions
1634 Issue a warning for undocumented Z80 instructions that do not work on R800.
1635 @item -forbid-undocumented-instructions
1637 Treat all undocumented instructions as errors.
1638 @item -forbid-unportable-instructions
1640 Treat undocumented Z80 instructions that do not work on R800 as errors.
1647 * Manual:: Structure of this Manual
1648 * GNU Assembler:: The GNU Assembler
1649 * Object Formats:: Object File Formats
1650 * Command Line:: Command Line
1651 * Input Files:: Input Files
1652 * Object:: Output (Object) File
1653 * Errors:: Error and Warning Messages
1657 @section Structure of this Manual
1659 @cindex manual, structure and purpose
1660 This manual is intended to describe what you need to know to use
1661 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1662 notation for symbols, constants, and expressions; the directives that
1663 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1666 We also cover special features in the @value{TARGET}
1667 configuration of @command{@value{AS}}, including assembler directives.
1670 This manual also describes some of the machine-dependent features of
1671 various flavors of the assembler.
1674 @cindex machine instructions (not covered)
1675 On the other hand, this manual is @emph{not} intended as an introduction
1676 to programming in assembly language---let alone programming in general!
1677 In a similar vein, we make no attempt to introduce the machine
1678 architecture; we do @emph{not} describe the instruction set, standard
1679 mnemonics, registers or addressing modes that are standard to a
1680 particular architecture.
1682 You may want to consult the manufacturer's
1683 machine architecture manual for this information.
1687 For information on the H8/300 machine instruction set, see @cite{H8/300
1688 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1689 Programming Manual} (Renesas).
1692 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1693 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1694 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1695 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1698 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1702 @c I think this is premature---doc@cygnus.com, 17jan1991
1704 Throughout this manual, we assume that you are running @dfn{GNU},
1705 the portable operating system from the @dfn{Free Software
1706 Foundation, Inc.}. This restricts our attention to certain kinds of
1707 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1708 once this assumption is granted examples and definitions need less
1711 @command{@value{AS}} is part of a team of programs that turn a high-level
1712 human-readable series of instructions into a low-level
1713 computer-readable series of instructions. Different versions of
1714 @command{@value{AS}} are used for different kinds of computer.
1717 @c There used to be a section "Terminology" here, which defined
1718 @c "contents", "byte", "word", and "long". Defining "word" to any
1719 @c particular size is confusing when the .word directive may generate 16
1720 @c bits on one machine and 32 bits on another; in general, for the user
1721 @c version of this manual, none of these terms seem essential to define.
1722 @c They were used very little even in the former draft of the manual;
1723 @c this draft makes an effort to avoid them (except in names of
1727 @section The GNU Assembler
1729 @c man begin DESCRIPTION
1731 @sc{gnu} @command{as} is really a family of assemblers.
1733 This manual describes @command{@value{AS}}, a member of that family which is
1734 configured for the @value{TARGET} architectures.
1736 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1737 should find a fairly similar environment when you use it on another
1738 architecture. Each version has much in common with the others,
1739 including object file formats, most assembler directives (often called
1740 @dfn{pseudo-ops}) and assembler syntax.@refill
1742 @cindex purpose of @sc{gnu} assembler
1743 @command{@value{AS}} is primarily intended to assemble the output of the
1744 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1745 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1746 assemble correctly everything that other assemblers for the same
1747 machine would assemble.
1749 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1752 @c This remark should appear in generic version of manual; assumption
1753 @c here is that generic version sets M680x0.
1754 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1755 assembler for the same architecture; for example, we know of several
1756 incompatible versions of 680x0 assembly language syntax.
1761 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1762 program in one pass of the source file. This has a subtle impact on the
1763 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1765 @node Object Formats
1766 @section Object File Formats
1768 @cindex object file format
1769 The @sc{gnu} assembler can be configured to produce several alternative
1770 object file formats. For the most part, this does not affect how you
1771 write assembly language programs; but directives for debugging symbols
1772 are typically different in different file formats. @xref{Symbol
1773 Attributes,,Symbol Attributes}.
1776 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1777 @value{OBJ-NAME} format object files.
1779 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1781 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1782 @code{b.out} or COFF format object files.
1785 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1786 SOM or ELF format object files.
1791 @section Command Line
1793 @cindex command line conventions
1795 After the program name @command{@value{AS}}, the command line may contain
1796 options and file names. Options may appear in any order, and may be
1797 before, after, or between file names. The order of file names is
1800 @cindex standard input, as input file
1802 @file{--} (two hyphens) by itself names the standard input file
1803 explicitly, as one of the files for @command{@value{AS}} to assemble.
1805 @cindex options, command line
1806 Except for @samp{--} any command line argument that begins with a
1807 hyphen (@samp{-}) is an option. Each option changes the behavior of
1808 @command{@value{AS}}. No option changes the way another option works. An
1809 option is a @samp{-} followed by one or more letters; the case of
1810 the letter is important. All options are optional.
1812 Some options expect exactly one file name to follow them. The file
1813 name may either immediately follow the option's letter (compatible
1814 with older assemblers) or it may be the next command argument (@sc{gnu}
1815 standard). These two command lines are equivalent:
1818 @value{AS} -o my-object-file.o mumble.s
1819 @value{AS} -omy-object-file.o mumble.s
1823 @section Input Files
1826 @cindex source program
1827 @cindex files, input
1828 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1829 describe the program input to one run of @command{@value{AS}}. The program may
1830 be in one or more files; how the source is partitioned into files
1831 doesn't change the meaning of the source.
1833 @c I added "con" prefix to "catenation" just to prove I can overcome my
1834 @c APL training... doc@cygnus.com
1835 The source program is a concatenation of the text in all the files, in the
1838 @c man begin DESCRIPTION
1839 Each time you run @command{@value{AS}} it assembles exactly one source
1840 program. The source program is made up of one or more files.
1841 (The standard input is also a file.)
1843 You give @command{@value{AS}} a command line that has zero or more input file
1844 names. The input files are read (from left file name to right). A
1845 command line argument (in any position) that has no special meaning
1846 is taken to be an input file name.
1848 If you give @command{@value{AS}} no file names it attempts to read one input file
1849 from the @command{@value{AS}} standard input, which is normally your terminal. You
1850 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1853 Use @samp{--} if you need to explicitly name the standard input file
1854 in your command line.
1856 If the source is empty, @command{@value{AS}} produces a small, empty object
1861 @subheading Filenames and Line-numbers
1863 @cindex input file linenumbers
1864 @cindex line numbers, in input files
1865 There are two ways of locating a line in the input file (or files) and
1866 either may be used in reporting error messages. One way refers to a line
1867 number in a physical file; the other refers to a line number in a
1868 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1870 @dfn{Physical files} are those files named in the command line given
1871 to @command{@value{AS}}.
1873 @dfn{Logical files} are simply names declared explicitly by assembler
1874 directives; they bear no relation to physical files. Logical file names help
1875 error messages reflect the original source file, when @command{@value{AS}} source
1876 is itself synthesized from other files. @command{@value{AS}} understands the
1877 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1878 @ref{File,,@code{.file}}.
1881 @section Output (Object) File
1887 Every time you run @command{@value{AS}} it produces an output file, which is
1888 your assembly language program translated into numbers. This file
1889 is the object file. Its default name is
1897 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1899 You can give it another name by using the @option{-o} option. Conventionally,
1900 object file names end with @file{.o}. The default name is used for historical
1901 reasons: older assemblers were capable of assembling self-contained programs
1902 directly into a runnable program. (For some formats, this isn't currently
1903 possible, but it can be done for the @code{a.out} format.)
1907 The object file is meant for input to the linker @code{@value{LD}}. It contains
1908 assembled program code, information to help @code{@value{LD}} integrate
1909 the assembled program into a runnable file, and (optionally) symbolic
1910 information for the debugger.
1912 @c link above to some info file(s) like the description of a.out.
1913 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1916 @section Error and Warning Messages
1918 @c man begin DESCRIPTION
1920 @cindex error messages
1921 @cindex warning messages
1922 @cindex messages from assembler
1923 @command{@value{AS}} may write warnings and error messages to the standard error
1924 file (usually your terminal). This should not happen when a compiler
1925 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1926 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1927 grave problem that stops the assembly.
1931 @cindex format of warning messages
1932 Warning messages have the format
1935 file_name:@b{NNN}:Warning Message Text
1939 @cindex line numbers, in warnings/errors
1940 (where @b{NNN} is a line number). If a logical file name has been given
1941 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1942 the current input file is used. If a logical line number was given
1944 (@pxref{Line,,@code{.line}})
1946 then it is used to calculate the number printed,
1947 otherwise the actual line in the current source file is printed. The
1948 message text is intended to be self explanatory (in the grand Unix
1951 @cindex format of error messages
1952 Error messages have the format
1954 file_name:@b{NNN}:FATAL:Error Message Text
1956 The file name and line number are derived as for warning
1957 messages. The actual message text may be rather less explanatory
1958 because many of them aren't supposed to happen.
1961 @chapter Command-Line Options
1963 @cindex options, all versions of assembler
1964 This chapter describes command-line options available in @emph{all}
1965 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
1966 for options specific
1968 to the @value{TARGET} target.
1971 to particular machine architectures.
1974 @c man begin DESCRIPTION
1976 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1977 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1978 The assembler arguments must be separated from each other (and the @samp{-Wa})
1979 by commas. For example:
1982 gcc -c -g -O -Wa,-alh,-L file.c
1986 This passes two options to the assembler: @samp{-alh} (emit a listing to
1987 standard output with high-level and assembly source) and @samp{-L} (retain
1988 local symbols in the symbol table).
1990 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1991 command-line options are automatically passed to the assembler by the compiler.
1992 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1993 precisely what options it passes to each compilation pass, including the
1999 * a:: -a[cdghlns] enable listings
2000 * alternate:: --alternate enable alternate macro syntax
2001 * D:: -D for compatibility
2002 * f:: -f to work faster
2003 * I:: -I for .include search path
2004 @ifclear DIFF-TBL-KLUGE
2005 * K:: -K for compatibility
2007 @ifset DIFF-TBL-KLUGE
2008 * K:: -K for difference tables
2011 * L:: -L to retain local symbols
2012 * listing:: --listing-XXX to configure listing output
2013 * M:: -M or --mri to assemble in MRI compatibility mode
2014 * MD:: --MD for dependency tracking
2015 * o:: -o to name the object file
2016 * R:: -R to join data and text sections
2017 * statistics:: --statistics to see statistics about assembly
2018 * traditional-format:: --traditional-format for compatible output
2019 * v:: -v to announce version
2020 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
2021 * Z:: -Z to make object file even after errors
2025 @section Enable Listings: @option{-a[cdghlns]}
2035 @cindex listings, enabling
2036 @cindex assembly listings, enabling
2038 These options enable listing output from the assembler. By itself,
2039 @samp{-a} requests high-level, assembly, and symbols listing.
2040 You can use other letters to select specific options for the list:
2041 @samp{-ah} requests a high-level language listing,
2042 @samp{-al} requests an output-program assembly listing, and
2043 @samp{-as} requests a symbol table listing.
2044 High-level listings require that a compiler debugging option like
2045 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
2048 Use the @samp{-ag} option to print a first section with general assembly
2049 information, like @value{AS} version, switches passed, or time stamp.
2051 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
2052 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
2053 other conditional), or a true @code{.if} followed by an @code{.else}, will be
2054 omitted from the listing.
2056 Use the @samp{-ad} option to omit debugging directives from the
2059 Once you have specified one of these options, you can further control
2060 listing output and its appearance using the directives @code{.list},
2061 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
2063 The @samp{-an} option turns off all forms processing.
2064 If you do not request listing output with one of the @samp{-a} options, the
2065 listing-control directives have no effect.
2067 The letters after @samp{-a} may be combined into one option,
2068 @emph{e.g.}, @samp{-aln}.
2070 Note if the assembler source is coming from the standard input (e.g.,
2072 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
2073 is being used) then the listing will not contain any comments or preprocessor
2074 directives. This is because the listing code buffers input source lines from
2075 stdin only after they have been preprocessed by the assembler. This reduces
2076 memory usage and makes the code more efficient.
2079 @section @option{--alternate}
2082 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
2085 @section @option{-D}
2088 This option has no effect whatsoever, but it is accepted to make it more
2089 likely that scripts written for other assemblers also work with
2090 @command{@value{AS}}.
2093 @section Work Faster: @option{-f}
2096 @cindex trusted compiler
2097 @cindex faster processing (@option{-f})
2098 @samp{-f} should only be used when assembling programs written by a
2099 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
2100 and comment preprocessing on
2101 the input file(s) before assembling them. @xref{Preprocessing,
2105 @emph{Warning:} if you use @samp{-f} when the files actually need to be
2106 preprocessed (if they contain comments, for example), @command{@value{AS}} does
2111 @section @code{.include} Search Path: @option{-I} @var{path}
2113 @kindex -I @var{path}
2114 @cindex paths for @code{.include}
2115 @cindex search path for @code{.include}
2116 @cindex @code{include} directive search path
2117 Use this option to add a @var{path} to the list of directories
2118 @command{@value{AS}} searches for files specified in @code{.include}
2119 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
2120 many times as necessary to include a variety of paths. The current
2121 working directory is always searched first; after that, @command{@value{AS}}
2122 searches any @samp{-I} directories in the same order as they were
2123 specified (left to right) on the command line.
2126 @section Difference Tables: @option{-K}
2129 @ifclear DIFF-TBL-KLUGE
2130 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2131 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2132 where it can be used to warn when the assembler alters the machine code
2133 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2134 family does not have the addressing limitations that sometimes lead to this
2135 alteration on other platforms.
2138 @ifset DIFF-TBL-KLUGE
2139 @cindex difference tables, warning
2140 @cindex warning for altered difference tables
2141 @command{@value{AS}} sometimes alters the code emitted for directives of the
2142 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2143 You can use the @samp{-K} option if you want a warning issued when this
2148 @section Include Local Symbols: @option{-L}
2151 @cindex local symbols, retaining in output
2152 Symbols beginning with system-specific local label prefixes, typically
2153 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2154 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2155 such symbols when debugging, because they are intended for the use of
2156 programs (like compilers) that compose assembler programs, not for your
2157 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2158 such symbols, so you do not normally debug with them.
2160 This option tells @command{@value{AS}} to retain those local symbols
2161 in the object file. Usually if you do this you also tell the linker
2162 @code{@value{LD}} to preserve those symbols.
2165 @section Configuring listing output: @option{--listing}
2167 The listing feature of the assembler can be enabled via the command line switch
2168 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2169 hex dump of the corresponding locations in the output object file, and displays
2170 them as a listing file. The format of this listing can be controlled by
2171 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2172 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2173 @code{.psize} (@pxref{Psize}), and
2174 @code{.eject} (@pxref{Eject}) and also by the following switches:
2177 @item --listing-lhs-width=@samp{number}
2178 @kindex --listing-lhs-width
2179 @cindex Width of first line disassembly output
2180 Sets the maximum width, in words, of the first line of the hex byte dump. This
2181 dump appears on the left hand side of the listing output.
2183 @item --listing-lhs-width2=@samp{number}
2184 @kindex --listing-lhs-width2
2185 @cindex Width of continuation lines of disassembly output
2186 Sets the maximum width, in words, of any further lines of the hex byte dump for
2187 a given input source line. If this value is not specified, it defaults to being
2188 the same as the value specified for @samp{--listing-lhs-width}. If neither
2189 switch is used the default is to one.
2191 @item --listing-rhs-width=@samp{number}
2192 @kindex --listing-rhs-width
2193 @cindex Width of source line output
2194 Sets the maximum width, in characters, of the source line that is displayed
2195 alongside the hex dump. The default value for this parameter is 100. The
2196 source line is displayed on the right hand side of the listing output.
2198 @item --listing-cont-lines=@samp{number}
2199 @kindex --listing-cont-lines
2200 @cindex Maximum number of continuation lines
2201 Sets the maximum number of continuation lines of hex dump that will be
2202 displayed for a given single line of source input. The default value is 4.
2206 @section Assemble in MRI Compatibility Mode: @option{-M}
2209 @cindex MRI compatibility mode
2210 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2211 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2212 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
2213 configured target) assembler from Microtec Research. The exact nature of the
2214 MRI syntax will not be documented here; see the MRI manuals for more
2215 information. Note in particular that the handling of macros and macro
2216 arguments is somewhat different. The purpose of this option is to permit
2217 assembling existing MRI assembler code using @command{@value{AS}}.
2219 The MRI compatibility is not complete. Certain operations of the MRI assembler
2220 depend upon its object file format, and can not be supported using other object
2221 file formats. Supporting these would require enhancing each object file format
2222 individually. These are:
2225 @item global symbols in common section
2227 The m68k MRI assembler supports common sections which are merged by the linker.
2228 Other object file formats do not support this. @command{@value{AS}} handles
2229 common sections by treating them as a single common symbol. It permits local
2230 symbols to be defined within a common section, but it can not support global
2231 symbols, since it has no way to describe them.
2233 @item complex relocations
2235 The MRI assemblers support relocations against a negated section address, and
2236 relocations which combine the start addresses of two or more sections. These
2237 are not support by other object file formats.
2239 @item @code{END} pseudo-op specifying start address
2241 The MRI @code{END} pseudo-op permits the specification of a start address.
2242 This is not supported by other object file formats. The start address may
2243 instead be specified using the @option{-e} option to the linker, or in a linker
2246 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2248 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2249 name to the output file. This is not supported by other object file formats.
2251 @item @code{ORG} pseudo-op
2253 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2254 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2255 which changes the location within the current section. Absolute sections are
2256 not supported by other object file formats. The address of a section may be
2257 assigned within a linker script.
2260 There are some other features of the MRI assembler which are not supported by
2261 @command{@value{AS}}, typically either because they are difficult or because they
2262 seem of little consequence. Some of these may be supported in future releases.
2266 @item EBCDIC strings
2268 EBCDIC strings are not supported.
2270 @item packed binary coded decimal
2272 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2273 and @code{DCB.P} pseudo-ops are not supported.
2275 @item @code{FEQU} pseudo-op
2277 The m68k @code{FEQU} pseudo-op is not supported.
2279 @item @code{NOOBJ} pseudo-op
2281 The m68k @code{NOOBJ} pseudo-op is not supported.
2283 @item @code{OPT} branch control options
2285 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2286 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2287 relaxes all branches, whether forward or backward, to an appropriate size, so
2288 these options serve no purpose.
2290 @item @code{OPT} list control options
2292 The following m68k @code{OPT} list control options are ignored: @code{C},
2293 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2294 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2296 @item other @code{OPT} options
2298 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2299 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2301 @item @code{OPT} @code{D} option is default
2303 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2304 @code{OPT NOD} may be used to turn it off.
2306 @item @code{XREF} pseudo-op.
2308 The m68k @code{XREF} pseudo-op is ignored.
2310 @item @code{.debug} pseudo-op
2312 The i960 @code{.debug} pseudo-op is not supported.
2314 @item @code{.extended} pseudo-op
2316 The i960 @code{.extended} pseudo-op is not supported.
2318 @item @code{.list} pseudo-op.
2320 The various options of the i960 @code{.list} pseudo-op are not supported.
2322 @item @code{.optimize} pseudo-op
2324 The i960 @code{.optimize} pseudo-op is not supported.
2326 @item @code{.output} pseudo-op
2328 The i960 @code{.output} pseudo-op is not supported.
2330 @item @code{.setreal} pseudo-op
2332 The i960 @code{.setreal} pseudo-op is not supported.
2337 @section Dependency Tracking: @option{--MD}
2340 @cindex dependency tracking
2343 @command{@value{AS}} can generate a dependency file for the file it creates. This
2344 file consists of a single rule suitable for @code{make} describing the
2345 dependencies of the main source file.
2347 The rule is written to the file named in its argument.
2349 This feature is used in the automatic updating of makefiles.
2352 @section Name the Object File: @option{-o}
2355 @cindex naming object file
2356 @cindex object file name
2357 There is always one object file output when you run @command{@value{AS}}. By
2358 default it has the name
2361 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
2375 You use this option (which takes exactly one filename) to give the
2376 object file a different name.
2378 Whatever the object file is called, @command{@value{AS}} overwrites any
2379 existing file of the same name.
2382 @section Join Data and Text Sections: @option{-R}
2385 @cindex data and text sections, joining
2386 @cindex text and data sections, joining
2387 @cindex joining text and data sections
2388 @cindex merging text and data sections
2389 @option{-R} tells @command{@value{AS}} to write the object file as if all
2390 data-section data lives in the text section. This is only done at
2391 the very last moment: your binary data are the same, but data
2392 section parts are relocated differently. The data section part of
2393 your object file is zero bytes long because all its bytes are
2394 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2396 When you specify @option{-R} it would be possible to generate shorter
2397 address displacements (because we do not have to cross between text and
2398 data section). We refrain from doing this simply for compatibility with
2399 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2402 When @command{@value{AS}} is configured for COFF or ELF output,
2403 this option is only useful if you use sections named @samp{.text} and
2408 @option{-R} is not supported for any of the HPPA targets. Using
2409 @option{-R} generates a warning from @command{@value{AS}}.
2413 @section Display Assembly Statistics: @option{--statistics}
2415 @kindex --statistics
2416 @cindex statistics, about assembly
2417 @cindex time, total for assembly
2418 @cindex space used, maximum for assembly
2419 Use @samp{--statistics} to display two statistics about the resources used by
2420 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2421 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2424 @node traditional-format
2425 @section Compatible Output: @option{--traditional-format}
2427 @kindex --traditional-format
2428 For some targets, the output of @command{@value{AS}} is different in some ways
2429 from the output of some existing assembler. This switch requests
2430 @command{@value{AS}} to use the traditional format instead.
2432 For example, it disables the exception frame optimizations which
2433 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2436 @section Announce Version: @option{-v}
2440 @cindex assembler version
2441 @cindex version of assembler
2442 You can find out what version of as is running by including the
2443 option @samp{-v} (which you can also spell as @samp{-version}) on the
2447 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2449 @command{@value{AS}} should never give a warning or error message when
2450 assembling compiler output. But programs written by people often
2451 cause @command{@value{AS}} to give a warning that a particular assumption was
2452 made. All such warnings are directed to the standard error file.
2456 @cindex suppressing warnings
2457 @cindex warnings, suppressing
2458 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2459 This only affects the warning messages: it does not change any particular of
2460 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2463 @kindex --fatal-warnings
2464 @cindex errors, caused by warnings
2465 @cindex warnings, causing error
2466 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2467 files that generate warnings to be in error.
2470 @cindex warnings, switching on
2471 You can switch these options off again by specifying @option{--warn}, which
2472 causes warnings to be output as usual.
2475 @section Generate Object File in Spite of Errors: @option{-Z}
2476 @cindex object file, after errors
2477 @cindex errors, continuing after
2478 After an error message, @command{@value{AS}} normally produces no output. If for
2479 some reason you are interested in object file output even after
2480 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2481 option. If there are any errors, @command{@value{AS}} continues anyways, and
2482 writes an object file after a final warning message of the form @samp{@var{n}
2483 errors, @var{m} warnings, generating bad object file.}
2488 @cindex machine-independent syntax
2489 @cindex syntax, machine-independent
2490 This chapter describes the machine-independent syntax allowed in a
2491 source file. @command{@value{AS}} syntax is similar to what many other
2492 assemblers use; it is inspired by the BSD 4.2
2497 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2501 * Preprocessing:: Preprocessing
2502 * Whitespace:: Whitespace
2503 * Comments:: Comments
2504 * Symbol Intro:: Symbols
2505 * Statements:: Statements
2506 * Constants:: Constants
2510 @section Preprocessing
2512 @cindex preprocessing
2513 The @command{@value{AS}} internal preprocessor:
2515 @cindex whitespace, removed by preprocessor
2517 adjusts and removes extra whitespace. It leaves one space or tab before
2518 the keywords on a line, and turns any other whitespace on the line into
2521 @cindex comments, removed by preprocessor
2523 removes all comments, replacing them with a single space, or an
2524 appropriate number of newlines.
2526 @cindex constants, converted by preprocessor
2528 converts character constants into the appropriate numeric values.
2531 It does not do macro processing, include file handling, or
2532 anything else you may get from your C compiler's preprocessor. You can
2533 do include file processing with the @code{.include} directive
2534 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2535 to get other ``CPP'' style preprocessing by giving the input file a
2536 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2537 Output, gcc.info, Using GNU CC}.
2539 Excess whitespace, comments, and character constants
2540 cannot be used in the portions of the input text that are not
2543 @cindex turning preprocessing on and off
2544 @cindex preprocessing, turning on and off
2547 If the first line of an input file is @code{#NO_APP} or if you use the
2548 @samp{-f} option, whitespace and comments are not removed from the input file.
2549 Within an input file, you can ask for whitespace and comment removal in
2550 specific portions of the by putting a line that says @code{#APP} before the
2551 text that may contain whitespace or comments, and putting a line that says
2552 @code{#NO_APP} after this text. This feature is mainly intend to support
2553 @code{asm} statements in compilers whose output is otherwise free of comments
2560 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2561 Whitespace is used to separate symbols, and to make programs neater for
2562 people to read. Unless within character constants
2563 (@pxref{Characters,,Character Constants}), any whitespace means the same
2564 as exactly one space.
2570 There are two ways of rendering comments to @command{@value{AS}}. In both
2571 cases the comment is equivalent to one space.
2573 Anything from @samp{/*} through the next @samp{*/} is a comment.
2574 This means you may not nest these comments.
2578 The only way to include a newline ('\n') in a comment
2579 is to use this sort of comment.
2582 /* This sort of comment does not nest. */
2585 @cindex line comment character
2586 Anything from a @dfn{line comment} character up to the next newline is
2587 considered a comment and is ignored. The line comment character is target
2588 specific, and some targets multiple comment characters. Some targets also have
2589 line comment characters that only work if they are the first character on a
2590 line. Some targets use a sequence of two characters to introduce a line
2591 comment. Some targets can also change their line comment characters depending
2592 upon command line options that have been used. For more details see the
2593 @emph{Syntax} section in the documentation for individual targets.
2595 If the line comment character is the hash sign (@samp{#}) then it still has the
2596 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2597 to specify logical line numbers:
2600 @cindex lines starting with @code{#}
2601 @cindex logical line numbers
2602 To be compatible with past assemblers, lines that begin with @samp{#} have a
2603 special interpretation. Following the @samp{#} should be an absolute
2604 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2605 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2606 new logical file name. The rest of the line, if any, should be whitespace.
2608 If the first non-whitespace characters on the line are not numeric,
2609 the line is ignored. (Just like a comment.)
2612 # This is an ordinary comment.
2613 # 42-6 "new_file_name" # New logical file name
2614 # This is logical line # 36.
2616 This feature is deprecated, and may disappear from future versions
2617 of @command{@value{AS}}.
2622 @cindex characters used in symbols
2623 @ifclear SPECIAL-SYMS
2624 A @dfn{symbol} is one or more characters chosen from the set of all
2625 letters (both upper and lower case), digits and the three characters
2631 A @dfn{symbol} is one or more characters chosen from the set of all
2632 letters (both upper and lower case), digits and the three characters
2633 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2639 On most machines, you can also use @code{$} in symbol names; exceptions
2640 are noted in @ref{Machine Dependencies}.
2642 No symbol may begin with a digit. Case is significant.
2643 There is no length limit: all characters are significant. Multibyte characters
2644 are supported. Symbols are delimited by characters not in that set, or by the
2645 beginning of a file (since the source program must end with a newline, the end
2646 of a file is not a possible symbol delimiter). @xref{Symbols}.
2647 @cindex length of symbols
2652 @cindex statements, structure of
2653 @cindex line separator character
2654 @cindex statement separator character
2656 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2657 @dfn{line separator character}. The line separator character is target
2658 specific and described in the @emph{Syntax} section of each
2659 target's documentation. Not all targets support a line separator character.
2660 The newline or line separator character is considered to be part of the
2661 preceding statement. Newlines and separators within character constants are an
2662 exception: they do not end statements.
2664 @cindex newline, required at file end
2665 @cindex EOF, newline must precede
2666 It is an error to end any statement with end-of-file: the last
2667 character of any input file should be a newline.@refill
2669 An empty statement is allowed, and may include whitespace. It is ignored.
2671 @cindex instructions and directives
2672 @cindex directives and instructions
2673 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2674 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2676 A statement begins with zero or more labels, optionally followed by a
2677 key symbol which determines what kind of statement it is. The key
2678 symbol determines the syntax of the rest of the statement. If the
2679 symbol begins with a dot @samp{.} then the statement is an assembler
2680 directive: typically valid for any computer. If the symbol begins with
2681 a letter the statement is an assembly language @dfn{instruction}: it
2682 assembles into a machine language instruction.
2684 Different versions of @command{@value{AS}} for different computers
2685 recognize different instructions. In fact, the same symbol may
2686 represent a different instruction in a different computer's assembly
2690 @cindex @code{:} (label)
2691 @cindex label (@code{:})
2692 A label is a symbol immediately followed by a colon (@code{:}).
2693 Whitespace before a label or after a colon is permitted, but you may not
2694 have whitespace between a label's symbol and its colon. @xref{Labels}.
2697 For HPPA targets, labels need not be immediately followed by a colon, but
2698 the definition of a label must begin in column zero. This also implies that
2699 only one label may be defined on each line.
2703 label: .directive followed by something
2704 another_label: # This is an empty statement.
2705 instruction operand_1, operand_2, @dots{}
2712 A constant is a number, written so that its value is known by
2713 inspection, without knowing any context. Like this:
2716 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2717 .ascii "Ring the bell\7" # A string constant.
2718 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2719 .float 0f-314159265358979323846264338327\
2720 95028841971.693993751E-40 # - pi, a flonum.
2725 * Characters:: Character Constants
2726 * Numbers:: Number Constants
2730 @subsection Character Constants
2732 @cindex character constants
2733 @cindex constants, character
2734 There are two kinds of character constants. A @dfn{character} stands
2735 for one character in one byte and its value may be used in
2736 numeric expressions. String constants (properly called string
2737 @emph{literals}) are potentially many bytes and their values may not be
2738 used in arithmetic expressions.
2742 * Chars:: Characters
2746 @subsubsection Strings
2748 @cindex string constants
2749 @cindex constants, string
2750 A @dfn{string} is written between double-quotes. It may contain
2751 double-quotes or null characters. The way to get special characters
2752 into a string is to @dfn{escape} these characters: precede them with
2753 a backslash @samp{\} character. For example @samp{\\} represents
2754 one backslash: the first @code{\} is an escape which tells
2755 @command{@value{AS}} to interpret the second character literally as a backslash
2756 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2757 escape character). The complete list of escapes follows.
2759 @cindex escape codes, character
2760 @cindex character escape codes
2763 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2765 @cindex @code{\b} (backspace character)
2766 @cindex backspace (@code{\b})
2768 Mnemonic for backspace; for ASCII this is octal code 010.
2771 @c Mnemonic for EOText; for ASCII this is octal code 004.
2773 @cindex @code{\f} (formfeed character)
2774 @cindex formfeed (@code{\f})
2776 Mnemonic for FormFeed; for ASCII this is octal code 014.
2778 @cindex @code{\n} (newline character)
2779 @cindex newline (@code{\n})
2781 Mnemonic for newline; for ASCII this is octal code 012.
2784 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2786 @cindex @code{\r} (carriage return character)
2787 @cindex carriage return (@code{\r})
2789 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2792 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2793 @c other assemblers.
2795 @cindex @code{\t} (tab)
2796 @cindex tab (@code{\t})
2798 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2801 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2802 @c @item \x @var{digit} @var{digit} @var{digit}
2803 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2805 @cindex @code{\@var{ddd}} (octal character code)
2806 @cindex octal character code (@code{\@var{ddd}})
2807 @item \ @var{digit} @var{digit} @var{digit}
2808 An octal character code. The numeric code is 3 octal digits.
2809 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2810 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2812 @cindex @code{\@var{xd...}} (hex character code)
2813 @cindex hex character code (@code{\@var{xd...}})
2814 @item \@code{x} @var{hex-digits...}
2815 A hex character code. All trailing hex digits are combined. Either upper or
2816 lower case @code{x} works.
2818 @cindex @code{\\} (@samp{\} character)
2819 @cindex backslash (@code{\\})
2821 Represents one @samp{\} character.
2824 @c Represents one @samp{'} (accent acute) character.
2825 @c This is needed in single character literals
2826 @c (@xref{Characters,,Character Constants}.) to represent
2829 @cindex @code{\"} (doublequote character)
2830 @cindex doublequote (@code{\"})
2832 Represents one @samp{"} character. Needed in strings to represent
2833 this character, because an unescaped @samp{"} would end the string.
2835 @item \ @var{anything-else}
2836 Any other character when escaped by @kbd{\} gives a warning, but
2837 assembles as if the @samp{\} was not present. The idea is that if
2838 you used an escape sequence you clearly didn't want the literal
2839 interpretation of the following character. However @command{@value{AS}} has no
2840 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2841 code and warns you of the fact.
2844 Which characters are escapable, and what those escapes represent,
2845 varies widely among assemblers. The current set is what we think
2846 the BSD 4.2 assembler recognizes, and is a subset of what most C
2847 compilers recognize. If you are in doubt, do not use an escape
2851 @subsubsection Characters
2853 @cindex single character constant
2854 @cindex character, single
2855 @cindex constant, single character
2856 A single character may be written as a single quote immediately
2857 followed by that character. The same escapes apply to characters as
2858 to strings. So if you want to write the character backslash, you
2859 must write @kbd{'\\} where the first @code{\} escapes the second
2860 @code{\}. As you can see, the quote is an acute accent, not a
2861 grave accent. A newline
2863 @ifclear abnormal-separator
2864 (or semicolon @samp{;})
2866 @ifset abnormal-separator
2868 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2873 immediately following an acute accent is taken as a literal character
2874 and does not count as the end of a statement. The value of a character
2875 constant in a numeric expression is the machine's byte-wide code for
2876 that character. @command{@value{AS}} assumes your character code is ASCII:
2877 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2880 @subsection Number Constants
2882 @cindex constants, number
2883 @cindex number constants
2884 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2885 are stored in the target machine. @emph{Integers} are numbers that
2886 would fit into an @code{int} in the C language. @emph{Bignums} are
2887 integers, but they are stored in more than 32 bits. @emph{Flonums}
2888 are floating point numbers, described below.
2891 * Integers:: Integers
2896 * Bit Fields:: Bit Fields
2902 @subsubsection Integers
2904 @cindex constants, integer
2906 @cindex binary integers
2907 @cindex integers, binary
2908 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2909 the binary digits @samp{01}.
2911 @cindex octal integers
2912 @cindex integers, octal
2913 An octal integer is @samp{0} followed by zero or more of the octal
2914 digits (@samp{01234567}).
2916 @cindex decimal integers
2917 @cindex integers, decimal
2918 A decimal integer starts with a non-zero digit followed by zero or
2919 more digits (@samp{0123456789}).
2921 @cindex hexadecimal integers
2922 @cindex integers, hexadecimal
2923 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2924 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2926 Integers have the usual values. To denote a negative integer, use
2927 the prefix operator @samp{-} discussed under expressions
2928 (@pxref{Prefix Ops,,Prefix Operators}).
2931 @subsubsection Bignums
2934 @cindex constants, bignum
2935 A @dfn{bignum} has the same syntax and semantics as an integer
2936 except that the number (or its negative) takes more than 32 bits to
2937 represent in binary. The distinction is made because in some places
2938 integers are permitted while bignums are not.
2941 @subsubsection Flonums
2943 @cindex floating point numbers
2944 @cindex constants, floating point
2946 @cindex precision, floating point
2947 A @dfn{flonum} represents a floating point number. The translation is
2948 indirect: a decimal floating point number from the text is converted by
2949 @command{@value{AS}} to a generic binary floating point number of more than
2950 sufficient precision. This generic floating point number is converted
2951 to a particular computer's floating point format (or formats) by a
2952 portion of @command{@value{AS}} specialized to that computer.
2954 A flonum is written by writing (in order)
2959 (@samp{0} is optional on the HPPA.)
2963 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2965 @kbd{e} is recommended. Case is not important.
2967 @c FIXME: verify if flonum syntax really this vague for most cases
2968 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2969 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2972 On the H8/300, Renesas / SuperH SH,
2973 and AMD 29K architectures, the letter must be
2974 one of the letters @samp{DFPRSX} (in upper or lower case).
2976 On the ARC, the letter must be one of the letters @samp{DFRS}
2977 (in upper or lower case).
2979 On the Intel 960 architecture, the letter must be
2980 one of the letters @samp{DFT} (in upper or lower case).
2982 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2986 One of the letters @samp{DFRS} (in upper or lower case).
2989 One of the letters @samp{DFPRSX} (in upper or lower case).
2992 The letter @samp{E} (upper case only).
2995 One of the letters @samp{DFT} (in upper or lower case).
3000 An optional sign: either @samp{+} or @samp{-}.
3003 An optional @dfn{integer part}: zero or more decimal digits.
3006 An optional @dfn{fractional part}: @samp{.} followed by zero
3007 or more decimal digits.
3010 An optional exponent, consisting of:
3014 An @samp{E} or @samp{e}.
3015 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
3016 @c principle this can perfectly well be different on different targets.
3018 Optional sign: either @samp{+} or @samp{-}.
3020 One or more decimal digits.
3025 At least one of the integer part or the fractional part must be
3026 present. The floating point number has the usual base-10 value.
3028 @command{@value{AS}} does all processing using integers. Flonums are computed
3029 independently of any floating point hardware in the computer running
3030 @command{@value{AS}}.
3034 @c Bit fields are written as a general facility but are also controlled
3035 @c by a conditional-compilation flag---which is as of now (21mar91)
3036 @c turned on only by the i960 config of GAS.
3038 @subsubsection Bit Fields
3041 @cindex constants, bit field
3042 You can also define numeric constants as @dfn{bit fields}.
3043 Specify two numbers separated by a colon---
3045 @var{mask}:@var{value}
3048 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
3051 The resulting number is then packed
3053 @c this conditional paren in case bit fields turned on elsewhere than 960
3054 (in host-dependent byte order)
3056 into a field whose width depends on which assembler directive has the
3057 bit-field as its argument. Overflow (a result from the bitwise and
3058 requiring more binary digits to represent) is not an error; instead,
3059 more constants are generated, of the specified width, beginning with the
3060 least significant digits.@refill
3062 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
3063 @code{.short}, and @code{.word} accept bit-field arguments.
3068 @chapter Sections and Relocation
3073 * Secs Background:: Background
3074 * Ld Sections:: Linker Sections
3075 * As Sections:: Assembler Internal Sections
3076 * Sub-Sections:: Sub-Sections
3080 @node Secs Background
3083 Roughly, a section is a range of addresses, with no gaps; all data
3084 ``in'' those addresses is treated the same for some particular purpose.
3085 For example there may be a ``read only'' section.
3087 @cindex linker, and assembler
3088 @cindex assembler, and linker
3089 The linker @code{@value{LD}} reads many object files (partial programs) and
3090 combines their contents to form a runnable program. When @command{@value{AS}}
3091 emits an object file, the partial program is assumed to start at address 0.
3092 @code{@value{LD}} assigns the final addresses for the partial program, so that
3093 different partial programs do not overlap. This is actually an
3094 oversimplification, but it suffices to explain how @command{@value{AS}} uses
3097 @code{@value{LD}} moves blocks of bytes of your program to their run-time
3098 addresses. These blocks slide to their run-time addresses as rigid
3099 units; their length does not change and neither does the order of bytes
3100 within them. Such a rigid unit is called a @emph{section}. Assigning
3101 run-time addresses to sections is called @dfn{relocation}. It includes
3102 the task of adjusting mentions of object-file addresses so they refer to
3103 the proper run-time addresses.
3105 For the H8/300, and for the Renesas / SuperH SH,
3106 @command{@value{AS}} pads sections if needed to
3107 ensure they end on a word (sixteen bit) boundary.
3110 @cindex standard assembler sections
3111 An object file written by @command{@value{AS}} has at least three sections, any
3112 of which may be empty. These are named @dfn{text}, @dfn{data} and
3117 When it generates COFF or ELF output,
3119 @command{@value{AS}} can also generate whatever other named sections you specify
3120 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
3121 If you do not use any directives that place output in the @samp{.text}
3122 or @samp{.data} sections, these sections still exist, but are empty.
3127 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
3129 @command{@value{AS}} can also generate whatever other named sections you
3130 specify using the @samp{.space} and @samp{.subspace} directives. See
3131 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3132 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3133 assembler directives.
3136 Additionally, @command{@value{AS}} uses different names for the standard
3137 text, data, and bss sections when generating SOM output. Program text
3138 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3139 BSS into @samp{$BSS$}.
3143 Within the object file, the text section starts at address @code{0}, the
3144 data section follows, and the bss section follows the data section.
3147 When generating either SOM or ELF output files on the HPPA, the text
3148 section starts at address @code{0}, the data section at address
3149 @code{0x4000000}, and the bss section follows the data section.
3152 To let @code{@value{LD}} know which data changes when the sections are
3153 relocated, and how to change that data, @command{@value{AS}} also writes to the
3154 object file details of the relocation needed. To perform relocation
3155 @code{@value{LD}} must know, each time an address in the object
3159 Where in the object file is the beginning of this reference to
3162 How long (in bytes) is this reference?
3164 Which section does the address refer to? What is the numeric value of
3166 (@var{address}) @minus{} (@var{start-address of section})?
3169 Is the reference to an address ``Program-Counter relative''?
3172 @cindex addresses, format of
3173 @cindex section-relative addressing
3174 In fact, every address @command{@value{AS}} ever uses is expressed as
3176 (@var{section}) + (@var{offset into section})
3179 Further, most expressions @command{@value{AS}} computes have this section-relative
3182 (For some object formats, such as SOM for the HPPA, some expressions are
3183 symbol-relative instead.)
3186 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3187 @var{N} into section @var{secname}.''
3189 Apart from text, data and bss sections you need to know about the
3190 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3191 addresses in the absolute section remain unchanged. For example, address
3192 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3193 @code{@value{LD}}. Although the linker never arranges two partial programs'
3194 data sections with overlapping addresses after linking, @emph{by definition}
3195 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3196 part of a program is always the same address when the program is running as
3197 address @code{@{absolute@ 239@}} in any other part of the program.
3199 The idea of sections is extended to the @dfn{undefined} section. Any
3200 address whose section is unknown at assembly time is by definition
3201 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3202 Since numbers are always defined, the only way to generate an undefined
3203 address is to mention an undefined symbol. A reference to a named
3204 common block would be such a symbol: its value is unknown at assembly
3205 time so it has section @emph{undefined}.
3207 By analogy the word @emph{section} is used to describe groups of sections in
3208 the linked program. @code{@value{LD}} puts all partial programs' text
3209 sections in contiguous addresses in the linked program. It is
3210 customary to refer to the @emph{text section} of a program, meaning all
3211 the addresses of all partial programs' text sections. Likewise for
3212 data and bss sections.
3214 Some sections are manipulated by @code{@value{LD}}; others are invented for
3215 use of @command{@value{AS}} and have no meaning except during assembly.
3218 @section Linker Sections
3219 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3224 @cindex named sections
3225 @cindex sections, named
3226 @item named sections
3229 @cindex text section
3230 @cindex data section
3234 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3235 separate but equal sections. Anything you can say of one section is
3238 When the program is running, however, it is
3239 customary for the text section to be unalterable. The
3240 text section is often shared among processes: it contains
3241 instructions, constants and the like. The data section of a running
3242 program is usually alterable: for example, C variables would be stored
3243 in the data section.
3248 This section contains zeroed bytes when your program begins running. It
3249 is used to hold uninitialized variables or common storage. The length of
3250 each partial program's bss section is important, but because it starts
3251 out containing zeroed bytes there is no need to store explicit zero
3252 bytes in the object file. The bss section was invented to eliminate
3253 those explicit zeros from object files.
3255 @cindex absolute section
3256 @item absolute section
3257 Address 0 of this section is always ``relocated'' to runtime address 0.
3258 This is useful if you want to refer to an address that @code{@value{LD}} must
3259 not change when relocating. In this sense we speak of absolute
3260 addresses being ``unrelocatable'': they do not change during relocation.
3262 @cindex undefined section
3263 @item undefined section
3264 This ``section'' is a catch-all for address references to objects not in
3265 the preceding sections.
3266 @c FIXME: ref to some other doc on obj-file formats could go here.
3269 @cindex relocation example
3270 An idealized example of three relocatable sections follows.
3272 The example uses the traditional section names @samp{.text} and @samp{.data}.
3274 Memory addresses are on the horizontal axis.
3278 @c END TEXI2ROFF-KILL
3281 partial program # 1: |ttttt|dddd|00|
3288 partial program # 2: |TTT|DDD|000|
3291 +--+---+-----+--+----+---+-----+~~
3292 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3293 +--+---+-----+--+----+---+-----+~~
3295 addresses: 0 @dots{}
3302 \line{\it Partial program \#1: \hfil}
3303 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3304 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3306 \line{\it Partial program \#2: \hfil}
3307 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3308 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3310 \line{\it linked program: \hfil}
3311 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3312 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3313 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3314 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3316 \line{\it addresses: \hfil}
3320 @c END TEXI2ROFF-KILL
3323 @section Assembler Internal Sections
3325 @cindex internal assembler sections
3326 @cindex sections in messages, internal
3327 These sections are meant only for the internal use of @command{@value{AS}}. They
3328 have no meaning at run-time. You do not really need to know about these
3329 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3330 warning messages, so it might be helpful to have an idea of their
3331 meanings to @command{@value{AS}}. These sections are used to permit the
3332 value of every expression in your assembly language program to be a
3333 section-relative address.
3336 @cindex assembler internal logic error
3337 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3338 An internal assembler logic error has been found. This means there is a
3339 bug in the assembler.
3341 @cindex expr (internal section)
3343 The assembler stores complex expression internally as combinations of
3344 symbols. When it needs to represent an expression as a symbol, it puts
3345 it in the expr section.
3347 @c FIXME item transfer[t] vector preload
3348 @c FIXME item transfer[t] vector postload
3349 @c FIXME item register
3353 @section Sub-Sections
3355 @cindex numbered subsections
3356 @cindex grouping data
3362 fall into two sections: text and data.
3364 You may have separate groups of
3366 data in named sections
3370 data in named sections
3376 that you want to end up near to each other in the object file, even though they
3377 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3378 use @dfn{subsections} for this purpose. Within each section, there can be
3379 numbered subsections with values from 0 to 8192. Objects assembled into the
3380 same subsection go into the object file together with other objects in the same
3381 subsection. For example, a compiler might want to store constants in the text
3382 section, but might not want to have them interspersed with the program being
3383 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3384 section of code being output, and a @samp{.text 1} before each group of
3385 constants being output.
3387 Subsections are optional. If you do not use subsections, everything
3388 goes in subsection number zero.
3391 Each subsection is zero-padded up to a multiple of four bytes.
3392 (Subsections may be padded a different amount on different flavors
3393 of @command{@value{AS}}.)
3397 On the H8/300 platform, each subsection is zero-padded to a word
3398 boundary (two bytes).
3399 The same is true on the Renesas SH.
3402 @c FIXME section padding (alignment)?
3403 @c Rich Pixley says padding here depends on target obj code format; that
3404 @c doesn't seem particularly useful to say without further elaboration,
3405 @c so for now I say nothing about it. If this is a generic BFD issue,
3406 @c these paragraphs might need to vanish from this manual, and be
3407 @c discussed in BFD chapter of binutils (or some such).
3411 Subsections appear in your object file in numeric order, lowest numbered
3412 to highest. (All this to be compatible with other people's assemblers.)
3413 The object file contains no representation of subsections; @code{@value{LD}} and
3414 other programs that manipulate object files see no trace of them.
3415 They just see all your text subsections as a text section, and all your
3416 data subsections as a data section.
3418 To specify which subsection you want subsequent statements assembled
3419 into, use a numeric argument to specify it, in a @samp{.text
3420 @var{expression}} or a @samp{.data @var{expression}} statement.
3423 When generating COFF output, you
3428 can also use an extra subsection
3429 argument with arbitrary named sections: @samp{.section @var{name},
3434 When generating ELF output, you
3439 can also use the @code{.subsection} directive (@pxref{SubSection})
3440 to specify a subsection: @samp{.subsection @var{expression}}.
3442 @var{Expression} should be an absolute expression
3443 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3444 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3445 begins in @code{text 0}. For instance:
3447 .text 0 # The default subsection is text 0 anyway.
3448 .ascii "This lives in the first text subsection. *"
3450 .ascii "But this lives in the second text subsection."
3452 .ascii "This lives in the data section,"
3453 .ascii "in the first data subsection."
3455 .ascii "This lives in the first text section,"
3456 .ascii "immediately following the asterisk (*)."
3459 Each section has a @dfn{location counter} incremented by one for every byte
3460 assembled into that section. Because subsections are merely a convenience
3461 restricted to @command{@value{AS}} there is no concept of a subsection location
3462 counter. There is no way to directly manipulate a location counter---but the
3463 @code{.align} directive changes it, and any label definition captures its
3464 current value. The location counter of the section where statements are being
3465 assembled is said to be the @dfn{active} location counter.
3468 @section bss Section
3471 @cindex common variable storage
3472 The bss section is used for local common variable storage.
3473 You may allocate address space in the bss section, but you may
3474 not dictate data to load into it before your program executes. When
3475 your program starts running, all the contents of the bss
3476 section are zeroed bytes.
3478 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3479 @ref{Lcomm,,@code{.lcomm}}.
3481 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3482 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3485 When assembling for a target which supports multiple sections, such as ELF or
3486 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3487 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3488 section. Typically the section will only contain symbol definitions and
3489 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3496 Symbols are a central concept: the programmer uses symbols to name
3497 things, the linker uses symbols to link, and the debugger uses symbols
3501 @cindex debuggers, and symbol order
3502 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3503 the same order they were declared. This may break some debuggers.
3508 * Setting Symbols:: Giving Symbols Other Values
3509 * Symbol Names:: Symbol Names
3510 * Dot:: The Special Dot Symbol
3511 * Symbol Attributes:: Symbol Attributes
3518 A @dfn{label} is written as a symbol immediately followed by a colon
3519 @samp{:}. The symbol then represents the current value of the
3520 active location counter, and is, for example, a suitable instruction
3521 operand. You are warned if you use the same symbol to represent two
3522 different locations: the first definition overrides any other
3526 On the HPPA, the usual form for a label need not be immediately followed by a
3527 colon, but instead must start in column zero. Only one label may be defined on
3528 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3529 provides a special directive @code{.label} for defining labels more flexibly.
3532 @node Setting Symbols
3533 @section Giving Symbols Other Values
3535 @cindex assigning values to symbols
3536 @cindex symbol values, assigning
3537 A symbol can be given an arbitrary value by writing a symbol, followed
3538 by an equals sign @samp{=}, followed by an expression
3539 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3540 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3541 equals sign @samp{=}@samp{=} here represents an equivalent of the
3542 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3545 Blackfin does not support symbol assignment with @samp{=}.
3549 @section Symbol Names
3551 @cindex symbol names
3552 @cindex names, symbol
3553 @ifclear SPECIAL-SYMS
3554 Symbol names begin with a letter or with one of @samp{._}. On most
3555 machines, you can also use @code{$} in symbol names; exceptions are
3556 noted in @ref{Machine Dependencies}. That character may be followed by any
3557 string of digits, letters, dollar signs (unless otherwise noted for a
3558 particular target machine), and underscores.
3562 Symbol names begin with a letter or with one of @samp{._}. On the
3563 Renesas SH you can also use @code{$} in symbol names. That
3564 character may be followed by any string of digits, letters, dollar signs (save
3565 on the H8/300), and underscores.
3569 Case of letters is significant: @code{foo} is a different symbol name
3572 Multibyte characters are supported. To generate a symbol name containing
3573 multibyte characters enclose it within double quotes and use escape codes. cf
3574 @xref{Strings}. Generating a multibyte symbol name from a label is not
3575 currently supported.
3577 Each symbol has exactly one name. Each name in an assembly language program
3578 refers to exactly one symbol. You may use that symbol name any number of times
3581 @subheading Local Symbol Names
3583 @cindex local symbol names
3584 @cindex symbol names, local
3585 A local symbol is any symbol beginning with certain local label prefixes.
3586 By default, the local label prefix is @samp{.L} for ELF systems or
3587 @samp{L} for traditional a.out systems, but each target may have its own
3588 set of local label prefixes.
3590 On the HPPA local symbols begin with @samp{L$}.
3593 Local symbols are defined and used within the assembler, but they are
3594 normally not saved in object files. Thus, they are not visible when debugging.
3595 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols:
3596 @option{-L}}) to retain the local symbols in the object files.
3598 @subheading Local Labels
3600 @cindex local labels
3601 @cindex temporary symbol names
3602 @cindex symbol names, temporary
3603 Local labels help compilers and programmers use names temporarily.
3604 They create symbols which are guaranteed to be unique over the entire scope of
3605 the input source code and which can be referred to by a simple notation.
3606 To define a local label, write a label of the form @samp{@b{N}:} (where @b{N}
3607 represents any positive integer). To refer to the most recent previous
3608 definition of that label write @samp{@b{N}b}, using the same number as when
3609 you defined the label. To refer to the next definition of a local label, write
3610 @samp{@b{N}f}---the @samp{b} stands for ``backwards'' and the @samp{f} stands
3613 There is no restriction on how you can use these labels, and you can reuse them
3614 too. So that it is possible to repeatedly define the same local label (using
3615 the same number @samp{@b{N}}), although you can only refer to the most recently
3616 defined local label of that number (for a backwards reference) or the next
3617 definition of a specific local label for a forward reference. It is also worth
3618 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3619 implemented in a slightly more efficient manner than the others.
3630 Which is the equivalent of:
3633 label_1: branch label_3
3634 label_2: branch label_1
3635 label_3: branch label_4
3636 label_4: branch label_3
3639 Local label names are only a notational device. They are immediately
3640 transformed into more conventional symbol names before the assembler uses them.
3641 The symbol names are stored in the symbol table, appear in error messages, and
3642 are optionally emitted to the object file. The names are constructed using
3646 @item @emph{local label prefix}
3647 All local symbols begin with the system-specific local label prefix.
3648 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3649 that start with the local label prefix. These labels are
3650 used for symbols you are never intended to see. If you use the
3651 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3652 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3653 you may use them in debugging.
3656 This is the number that was used in the local label definition. So if the
3657 label is written @samp{55:} then the number is @samp{55}.
3660 This unusual character is included so you do not accidentally invent a symbol
3661 of the same name. The character has ASCII value of @samp{\002} (control-B).
3663 @item @emph{ordinal number}
3664 This is a serial number to keep the labels distinct. The first definition of
3665 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3666 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3667 the number @samp{1} and its 15th definition gets @samp{15} as well.
3670 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3671 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3673 @subheading Dollar Local Labels
3674 @cindex dollar local symbols
3676 @code{@value{AS}} also supports an even more local form of local labels called
3677 dollar labels. These labels go out of scope (i.e., they become undefined) as
3678 soon as a non-local label is defined. Thus they remain valid for only a small
3679 region of the input source code. Normal local labels, by contrast, remain in
3680 scope for the entire file, or until they are redefined by another occurrence of
3681 the same local label.
3683 Dollar labels are defined in exactly the same way as ordinary local labels,
3684 except that they have a dollar sign suffix to their numeric value, e.g.,
3687 They can also be distinguished from ordinary local labels by their transformed
3688 names which use ASCII character @samp{\001} (control-A) as the magic character
3689 to distinguish them from ordinary labels. For example, the fifth definition of
3690 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3693 @section The Special Dot Symbol
3695 @cindex dot (symbol)
3696 @cindex @code{.} (symbol)
3697 @cindex current address
3698 @cindex location counter
3699 The special symbol @samp{.} refers to the current address that
3700 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3701 .long .} defines @code{melvin} to contain its own address.
3702 Assigning a value to @code{.} is treated the same as a @code{.org}
3704 @ifclear no-space-dir
3705 Thus, the expression @samp{.=.+4} is the same as saying
3709 @node Symbol Attributes
3710 @section Symbol Attributes
3712 @cindex symbol attributes
3713 @cindex attributes, symbol
3714 Every symbol has, as well as its name, the attributes ``Value'' and
3715 ``Type''. Depending on output format, symbols can also have auxiliary
3718 The detailed definitions are in @file{a.out.h}.
3721 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3722 all these attributes, and probably won't warn you. This makes the
3723 symbol an externally defined symbol, which is generally what you
3727 * Symbol Value:: Value
3728 * Symbol Type:: Type
3731 * a.out Symbols:: Symbol Attributes: @code{a.out}
3735 * a.out Symbols:: Symbol Attributes: @code{a.out}
3738 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3743 * COFF Symbols:: Symbol Attributes for COFF
3746 * SOM Symbols:: Symbol Attributes for SOM
3753 @cindex value of a symbol
3754 @cindex symbol value
3755 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3756 location in the text, data, bss or absolute sections the value is the
3757 number of addresses from the start of that section to the label.
3758 Naturally for text, data and bss sections the value of a symbol changes
3759 as @code{@value{LD}} changes section base addresses during linking. Absolute
3760 symbols' values do not change during linking: that is why they are
3763 The value of an undefined symbol is treated in a special way. If it is
3764 0 then the symbol is not defined in this assembler source file, and
3765 @code{@value{LD}} tries to determine its value from other files linked into the
3766 same program. You make this kind of symbol simply by mentioning a symbol
3767 name without defining it. A non-zero value represents a @code{.comm}
3768 common declaration. The value is how much common storage to reserve, in
3769 bytes (addresses). The symbol refers to the first address of the
3775 @cindex type of a symbol
3777 The type attribute of a symbol contains relocation (section)
3778 information, any flag settings indicating that a symbol is external, and
3779 (optionally), other information for linkers and debuggers. The exact
3780 format depends on the object-code output format in use.
3785 @c The following avoids a "widow" subsection title. @group would be
3786 @c better if it were available outside examples.
3789 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3791 @cindex @code{b.out} symbol attributes
3792 @cindex symbol attributes, @code{b.out}
3793 These symbol attributes appear only when @command{@value{AS}} is configured for
3794 one of the Berkeley-descended object output formats---@code{a.out} or
3800 @subsection Symbol Attributes: @code{a.out}
3802 @cindex @code{a.out} symbol attributes
3803 @cindex symbol attributes, @code{a.out}
3809 @subsection Symbol Attributes: @code{a.out}
3811 @cindex @code{a.out} symbol attributes
3812 @cindex symbol attributes, @code{a.out}
3816 * Symbol Desc:: Descriptor
3817 * Symbol Other:: Other
3821 @subsubsection Descriptor
3823 @cindex descriptor, of @code{a.out} symbol
3824 This is an arbitrary 16-bit value. You may establish a symbol's
3825 descriptor value by using a @code{.desc} statement
3826 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3827 @command{@value{AS}}.
3830 @subsubsection Other
3832 @cindex other attribute, of @code{a.out} symbol
3833 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3838 @subsection Symbol Attributes for COFF
3840 @cindex COFF symbol attributes
3841 @cindex symbol attributes, COFF
3843 The COFF format supports a multitude of auxiliary symbol attributes;
3844 like the primary symbol attributes, they are set between @code{.def} and
3845 @code{.endef} directives.
3847 @subsubsection Primary Attributes
3849 @cindex primary attributes, COFF symbols
3850 The symbol name is set with @code{.def}; the value and type,
3851 respectively, with @code{.val} and @code{.type}.
3853 @subsubsection Auxiliary Attributes
3855 @cindex auxiliary attributes, COFF symbols
3856 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3857 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3858 table information for COFF.
3863 @subsection Symbol Attributes for SOM
3865 @cindex SOM symbol attributes
3866 @cindex symbol attributes, SOM
3868 The SOM format for the HPPA supports a multitude of symbol attributes set with
3869 the @code{.EXPORT} and @code{.IMPORT} directives.
3871 The attributes are described in @cite{HP9000 Series 800 Assembly
3872 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3873 @code{EXPORT} assembler directive documentation.
3877 @chapter Expressions
3881 @cindex numeric values
3882 An @dfn{expression} specifies an address or numeric value.
3883 Whitespace may precede and/or follow an expression.
3885 The result of an expression must be an absolute number, or else an offset into
3886 a particular section. If an expression is not absolute, and there is not
3887 enough information when @command{@value{AS}} sees the expression to know its
3888 section, a second pass over the source program might be necessary to interpret
3889 the expression---but the second pass is currently not implemented.
3890 @command{@value{AS}} aborts with an error message in this situation.
3893 * Empty Exprs:: Empty Expressions
3894 * Integer Exprs:: Integer Expressions
3898 @section Empty Expressions
3900 @cindex empty expressions
3901 @cindex expressions, empty
3902 An empty expression has no value: it is just whitespace or null.
3903 Wherever an absolute expression is required, you may omit the
3904 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3905 is compatible with other assemblers.
3908 @section Integer Expressions
3910 @cindex integer expressions
3911 @cindex expressions, integer
3912 An @dfn{integer expression} is one or more @emph{arguments} delimited
3913 by @emph{operators}.
3916 * Arguments:: Arguments
3917 * Operators:: Operators
3918 * Prefix Ops:: Prefix Operators
3919 * Infix Ops:: Infix Operators
3923 @subsection Arguments
3925 @cindex expression arguments
3926 @cindex arguments in expressions
3927 @cindex operands in expressions
3928 @cindex arithmetic operands
3929 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3930 contexts arguments are sometimes called ``arithmetic operands''. In
3931 this manual, to avoid confusing them with the ``instruction operands'' of
3932 the machine language, we use the term ``argument'' to refer to parts of
3933 expressions only, reserving the word ``operand'' to refer only to machine
3934 instruction operands.
3936 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3937 @var{section} is one of text, data, bss, absolute,
3938 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3941 Numbers are usually integers.
3943 A number can be a flonum or bignum. In this case, you are warned
3944 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3945 these 32 bits are an integer. You may write integer-manipulating
3946 instructions that act on exotic constants, compatible with other
3949 @cindex subexpressions
3950 Subexpressions are a left parenthesis @samp{(} followed by an integer
3951 expression, followed by a right parenthesis @samp{)}; or a prefix
3952 operator followed by an argument.
3955 @subsection Operators
3957 @cindex operators, in expressions
3958 @cindex arithmetic functions
3959 @cindex functions, in expressions
3960 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3961 operators are followed by an argument. Infix operators appear
3962 between their arguments. Operators may be preceded and/or followed by
3966 @subsection Prefix Operator
3968 @cindex prefix operators
3969 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3970 one argument, which must be absolute.
3972 @c the tex/end tex stuff surrounding this small table is meant to make
3973 @c it align, on the printed page, with the similar table in the next
3974 @c section (which is inside an enumerate).
3976 \global\advance\leftskip by \itemindent
3981 @dfn{Negation}. Two's complement negation.
3983 @dfn{Complementation}. Bitwise not.
3987 \global\advance\leftskip by -\itemindent
3991 @subsection Infix Operators
3993 @cindex infix operators
3994 @cindex operators, permitted arguments
3995 @dfn{Infix operators} take two arguments, one on either side. Operators
3996 have precedence, but operations with equal precedence are performed left
3997 to right. Apart from @code{+} or @option{-}, both arguments must be
3998 absolute, and the result is absolute.
4001 @cindex operator precedence
4002 @cindex precedence of operators
4009 @dfn{Multiplication}.
4012 @dfn{Division}. Truncation is the same as the C operator @samp{/}
4018 @dfn{Shift Left}. Same as the C operator @samp{<<}.
4021 @dfn{Shift Right}. Same as the C operator @samp{>>}.
4025 Intermediate precedence
4030 @dfn{Bitwise Inclusive Or}.
4036 @dfn{Bitwise Exclusive Or}.
4039 @dfn{Bitwise Or Not}.
4046 @cindex addition, permitted arguments
4047 @cindex plus, permitted arguments
4048 @cindex arguments for addition
4050 @dfn{Addition}. If either argument is absolute, the result has the section of
4051 the other argument. You may not add together arguments from different
4054 @cindex subtraction, permitted arguments
4055 @cindex minus, permitted arguments
4056 @cindex arguments for subtraction
4058 @dfn{Subtraction}. If the right argument is absolute, the
4059 result has the section of the left argument.
4060 If both arguments are in the same section, the result is absolute.
4061 You may not subtract arguments from different sections.
4062 @c FIXME is there still something useful to say about undefined - undefined ?
4064 @cindex comparison expressions
4065 @cindex expressions, comparison
4070 @dfn{Is Not Equal To}
4074 @dfn{Is Greater Than}
4076 @dfn{Is Greater Than Or Equal To}
4078 @dfn{Is Less Than Or Equal To}
4080 The comparison operators can be used as infix operators. A true results has a
4081 value of -1 whereas a false result has a value of 0. Note, these operators
4082 perform signed comparisons.
4085 @item Lowest Precedence
4094 These two logical operations can be used to combine the results of sub
4095 expressions. Note, unlike the comparison operators a true result returns a
4096 value of 1 but a false results does still return 0. Also note that the logical
4097 or operator has a slightly lower precedence than logical and.
4102 In short, it's only meaningful to add or subtract the @emph{offsets} in an
4103 address; you can only have a defined section in one of the two arguments.
4106 @chapter Assembler Directives
4108 @cindex directives, machine independent
4109 @cindex pseudo-ops, machine independent
4110 @cindex machine independent directives
4111 All assembler directives have names that begin with a period (@samp{.}).
4112 The rest of the name is letters, usually in lower case.
4114 This chapter discusses directives that are available regardless of the
4115 target machine configuration for the @sc{gnu} assembler.
4117 Some machine configurations provide additional directives.
4118 @xref{Machine Dependencies}.
4121 @ifset machine-directives
4122 @xref{Machine Dependencies}, for additional directives.
4127 * Abort:: @code{.abort}
4129 * ABORT (COFF):: @code{.ABORT}
4132 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
4133 * Altmacro:: @code{.altmacro}
4134 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4135 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4136 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
4137 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, @code{.bundle_lock}, @code{.bundle_unlock}
4138 * Byte:: @code{.byte @var{expressions}}
4139 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4140 * Comm:: @code{.comm @var{symbol} , @var{length} }
4141 * Data:: @code{.data @var{subsection}}
4143 * Def:: @code{.def @var{name}}
4146 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4152 * Double:: @code{.double @var{flonums}}
4153 * Eject:: @code{.eject}
4154 * Else:: @code{.else}
4155 * Elseif:: @code{.elseif}
4158 * Endef:: @code{.endef}
4161 * Endfunc:: @code{.endfunc}
4162 * Endif:: @code{.endif}
4163 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4164 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4165 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4167 * Error:: @code{.error @var{string}}
4168 * Exitm:: @code{.exitm}
4169 * Extern:: @code{.extern}
4170 * Fail:: @code{.fail}
4171 * File:: @code{.file}
4172 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4173 * Float:: @code{.float @var{flonums}}
4174 * Func:: @code{.func}
4175 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4177 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4178 * Hidden:: @code{.hidden @var{names}}
4181 * hword:: @code{.hword @var{expressions}}
4182 * Ident:: @code{.ident}
4183 * If:: @code{.if @var{absolute expression}}
4184 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4185 * Include:: @code{.include "@var{file}"}
4186 * Int:: @code{.int @var{expressions}}
4188 * Internal:: @code{.internal @var{names}}
4191 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4192 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4193 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4194 * Lflags:: @code{.lflags}
4195 @ifclear no-line-dir
4196 * Line:: @code{.line @var{line-number}}
4199 * Linkonce:: @code{.linkonce [@var{type}]}
4200 * List:: @code{.list}
4201 * Ln:: @code{.ln @var{line-number}}
4202 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4203 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4205 * Local:: @code{.local @var{names}}
4208 * Long:: @code{.long @var{expressions}}
4210 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4213 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4214 * MRI:: @code{.mri @var{val}}
4215 * Noaltmacro:: @code{.noaltmacro}
4216 * Nolist:: @code{.nolist}
4217 * Octa:: @code{.octa @var{bignums}}
4218 * Offset:: @code{.offset @var{loc}}
4219 * Org:: @code{.org @var{new-lc}, @var{fill}}
4220 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4222 * PopSection:: @code{.popsection}
4223 * Previous:: @code{.previous}
4226 * Print:: @code{.print @var{string}}
4228 * Protected:: @code{.protected @var{names}}
4231 * Psize:: @code{.psize @var{lines}, @var{columns}}
4232 * Purgem:: @code{.purgem @var{name}}
4234 * PushSection:: @code{.pushsection @var{name}}
4237 * Quad:: @code{.quad @var{bignums}}
4238 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4239 * Rept:: @code{.rept @var{count}}
4240 * Sbttl:: @code{.sbttl "@var{subheading}"}
4242 * Scl:: @code{.scl @var{class}}
4245 * Section:: @code{.section @var{name}[, @var{flags}]}
4248 * Set:: @code{.set @var{symbol}, @var{expression}}
4249 * Short:: @code{.short @var{expressions}}
4250 * Single:: @code{.single @var{flonums}}
4252 * Size:: @code{.size [@var{name} , @var{expression}]}
4254 @ifclear no-space-dir
4255 * Skip:: @code{.skip @var{size} , @var{fill}}
4258 * Sleb128:: @code{.sleb128 @var{expressions}}
4259 @ifclear no-space-dir
4260 * Space:: @code{.space @var{size} , @var{fill}}
4263 * Stab:: @code{.stabd, .stabn, .stabs}
4266 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4267 * Struct:: @code{.struct @var{expression}}
4269 * SubSection:: @code{.subsection}
4270 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4274 * Tag:: @code{.tag @var{structname}}
4277 * Text:: @code{.text @var{subsection}}
4278 * Title:: @code{.title "@var{heading}"}
4280 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4283 * Uleb128:: @code{.uleb128 @var{expressions}}
4285 * Val:: @code{.val @var{addr}}
4289 * Version:: @code{.version "@var{string}"}
4290 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4291 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4294 * Warning:: @code{.warning @var{string}}
4295 * Weak:: @code{.weak @var{names}}
4296 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4297 * Word:: @code{.word @var{expressions}}
4298 * Deprecated:: Deprecated Directives
4302 @section @code{.abort}
4304 @cindex @code{abort} directive
4305 @cindex stopping the assembly
4306 This directive stops the assembly immediately. It is for
4307 compatibility with other assemblers. The original idea was that the
4308 assembly language source would be piped into the assembler. If the sender
4309 of the source quit, it could use this directive tells @command{@value{AS}} to
4310 quit also. One day @code{.abort} will not be supported.
4314 @section @code{.ABORT} (COFF)
4316 @cindex @code{ABORT} directive
4317 When producing COFF output, @command{@value{AS}} accepts this directive as a
4318 synonym for @samp{.abort}.
4321 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
4327 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4329 @cindex padding the location counter
4330 @cindex @code{align} directive
4331 Pad the location counter (in the current subsection) to a particular storage
4332 boundary. The first expression (which must be absolute) is the alignment
4333 required, as described below.
4335 The second expression (also absolute) gives the fill value to be stored in the
4336 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4337 padding bytes are normally zero. However, on some systems, if the section is
4338 marked as containing code and the fill value is omitted, the space is filled
4339 with no-op instructions.
4341 The third expression is also absolute, and is also optional. If it is present,
4342 it is the maximum number of bytes that should be skipped by this alignment
4343 directive. If doing the alignment would require skipping more bytes than the
4344 specified maximum, then the alignment is not done at all. You can omit the
4345 fill value (the second argument) entirely by simply using two commas after the
4346 required alignment; this can be useful if you want the alignment to be filled
4347 with no-op instructions when appropriate.
4349 The way the required alignment is specified varies from system to system.
4350 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
4351 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4352 alignment request in bytes. For example @samp{.align 8} advances
4353 the location counter until it is a multiple of 8. If the location counter
4354 is already a multiple of 8, no change is needed. For the tic54x, the
4355 first expression is the alignment request in words.
4357 For other systems, including ppc, i386 using a.out format, arm and
4358 strongarm, it is the
4359 number of low-order zero bits the location counter must have after
4360 advancement. For example @samp{.align 3} advances the location
4361 counter until it a multiple of 8. If the location counter is already a
4362 multiple of 8, no change is needed.
4364 This inconsistency is due to the different behaviors of the various
4365 native assemblers for these systems which GAS must emulate.
4366 GAS also provides @code{.balign} and @code{.p2align} directives,
4367 described later, which have a consistent behavior across all
4368 architectures (but are specific to GAS).
4371 @section @code{.altmacro}
4372 Enable alternate macro mode, enabling:
4375 @item LOCAL @var{name} [ , @dots{} ]
4376 One additional directive, @code{LOCAL}, is available. It is used to
4377 generate a string replacement for each of the @var{name} arguments, and
4378 replace any instances of @var{name} in each macro expansion. The
4379 replacement string is unique in the assembly, and different for each
4380 separate macro expansion. @code{LOCAL} allows you to write macros that
4381 define symbols, without fear of conflict between separate macro expansions.
4383 @item String delimiters
4384 You can write strings delimited in these other ways besides
4385 @code{"@var{string}"}:
4388 @item '@var{string}'
4389 You can delimit strings with single-quote characters.
4391 @item <@var{string}>
4392 You can delimit strings with matching angle brackets.
4395 @item single-character string escape
4396 To include any single character literally in a string (even if the
4397 character would otherwise have some special meaning), you can prefix the
4398 character with @samp{!} (an exclamation mark). For example, you can
4399 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4401 @item Expression results as strings
4402 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4403 and use the result as a string.
4407 @section @code{.ascii "@var{string}"}@dots{}
4409 @cindex @code{ascii} directive
4410 @cindex string literals
4411 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4412 separated by commas. It assembles each string (with no automatic
4413 trailing zero byte) into consecutive addresses.
4416 @section @code{.asciz "@var{string}"}@dots{}
4418 @cindex @code{asciz} directive
4419 @cindex zero-terminated strings
4420 @cindex null-terminated strings
4421 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4422 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4425 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4427 @cindex padding the location counter given number of bytes
4428 @cindex @code{balign} directive
4429 Pad the location counter (in the current subsection) to a particular
4430 storage boundary. The first expression (which must be absolute) is the
4431 alignment request in bytes. For example @samp{.balign 8} advances
4432 the location counter until it is a multiple of 8. If the location counter
4433 is already a multiple of 8, no change is needed.
4435 The second expression (also absolute) gives the fill value to be stored in the
4436 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4437 padding bytes are normally zero. However, on some systems, if the section is
4438 marked as containing code and the fill value is omitted, the space is filled
4439 with no-op instructions.
4441 The third expression is also absolute, and is also optional. If it is present,
4442 it is the maximum number of bytes that should be skipped by this alignment
4443 directive. If doing the alignment would require skipping more bytes than the
4444 specified maximum, then the alignment is not done at all. You can omit the
4445 fill value (the second argument) entirely by simply using two commas after the
4446 required alignment; this can be useful if you want the alignment to be filled
4447 with no-op instructions when appropriate.
4449 @cindex @code{balignw} directive
4450 @cindex @code{balignl} directive
4451 The @code{.balignw} and @code{.balignl} directives are variants of the
4452 @code{.balign} directive. The @code{.balignw} directive treats the fill
4453 pattern as a two byte word value. The @code{.balignl} directives treats the
4454 fill pattern as a four byte longword value. For example, @code{.balignw
4455 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4456 filled in with the value 0x368d (the exact placement of the bytes depends upon
4457 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4460 @node Bundle directives
4461 @section @code{.bundle_align_mode @var{abs-expr}}
4462 @cindex @code{bundle_align_mode} directive
4464 @cindex instruction bundle
4465 @cindex aligned instruction bundle
4466 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4467 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4468 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4469 disabled (which is the default state). If the argument it not zero, it
4470 gives the size of an instruction bundle as a power of two (as for the
4471 @code{.p2align} directive, @pxref{P2align}).
4473 For some targets, it's an ABI requirement that no instruction may span a
4474 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4475 instructions that starts on an aligned boundary. For example, if
4476 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4477 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4478 effect, no single instruction may span a boundary between bundles. If an
4479 instruction would start too close to the end of a bundle for the length of
4480 that particular instruction to fit within the bundle, then the space at the
4481 end of that bundle is filled with no-op instructions so the instruction
4482 starts in the next bundle. As a corollary, it's an error if any single
4483 instruction's encoding is longer than the bundle size.
4485 @section @code{.bundle_lock} and @code{.bundle_unlock}
4486 @cindex @code{bundle_lock} directive
4487 @cindex @code{bundle_unlock} directive
4488 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4489 allow explicit control over instruction bundle padding. These directives
4490 are only valid when @code{.bundle_align_mode} has been used to enable
4491 aligned instruction bundle mode. It's an error if they appear when
4492 @code{.bundle_align_mode} has not been used at all, or when the last
4493 directive was @w{@code{.bundle_align_mode 0}}.
4495 @cindex bundle-locked
4496 For some targets, it's an ABI requirement that certain instructions may
4497 appear only as part of specified permissible sequences of multiple
4498 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4499 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4500 instruction sequence. For purposes of aligned instruction bundle mode, a
4501 sequence starting with @code{.bundle_lock} and ending with
4502 @code{.bundle_unlock} is treated as a single instruction. That is, the
4503 entire sequence must fit into a single bundle and may not span a bundle
4504 boundary. If necessary, no-op instructions will be inserted before the
4505 first instruction of the sequence so that the whole sequence starts on an
4506 aligned bundle boundary. It's an error if the sequence is longer than the
4509 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4510 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4511 nested. That is, a second @code{.bundle_lock} directive before the next
4512 @code{.bundle_unlock} directive has no effect except that it must be
4513 matched by another closing @code{.bundle_unlock} so that there is the
4514 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4517 @section @code{.byte @var{expressions}}
4519 @cindex @code{byte} directive
4520 @cindex integers, one byte
4521 @code{.byte} expects zero or more expressions, separated by commas.
4522 Each expression is assembled into the next byte.
4524 @node CFI directives
4525 @section @code{.cfi_sections @var{section_list}}
4526 @cindex @code{cfi_sections} directive
4527 @code{.cfi_sections} may be used to specify whether CFI directives
4528 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4529 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4530 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4531 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4532 directive is not used is @code{.cfi_sections .eh_frame}.
4534 @section @code{.cfi_startproc [simple]}
4535 @cindex @code{cfi_startproc} directive
4536 @code{.cfi_startproc} is used at the beginning of each function that
4537 should have an entry in @code{.eh_frame}. It initializes some internal
4538 data structures. Don't forget to close the function by
4539 @code{.cfi_endproc}.
4541 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4542 it also emits some architecture dependent initial CFI instructions.
4544 @section @code{.cfi_endproc}
4545 @cindex @code{cfi_endproc} directive
4546 @code{.cfi_endproc} is used at the end of a function where it closes its
4547 unwind entry previously opened by
4548 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4550 @section @code{.cfi_personality @var{encoding} [, @var{exp}]}
4551 @code{.cfi_personality} defines personality routine and its encoding.
4552 @var{encoding} must be a constant determining how the personality
4553 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4554 argument is not present, otherwise second argument should be
4555 a constant or a symbol name. When using indirect encodings,
4556 the symbol provided should be the location where personality
4557 can be loaded from, not the personality routine itself.
4558 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4559 no personality routine.
4561 @section @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4562 @code{.cfi_lsda} defines LSDA and its encoding.
4563 @var{encoding} must be a constant determining how the LSDA
4564 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4565 argument is not present, otherwise second argument should be a constant
4566 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4569 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4570 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4571 address from @var{register} and add @var{offset} to it}.
4573 @section @code{.cfi_def_cfa_register @var{register}}
4574 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4575 now on @var{register} will be used instead of the old one. Offset
4578 @section @code{.cfi_def_cfa_offset @var{offset}}
4579 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4580 remains the same, but @var{offset} is new. Note that it is the
4581 absolute offset that will be added to a defined register to compute
4584 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4585 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4586 value that is added/substracted from the previous offset.
4588 @section @code{.cfi_offset @var{register}, @var{offset}}
4589 Previous value of @var{register} is saved at offset @var{offset} from
4592 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4593 Previous value of @var{register} is saved at offset @var{offset} from
4594 the current CFA register. This is transformed to @code{.cfi_offset}
4595 using the known displacement of the CFA register from the CFA.
4596 This is often easier to use, because the number will match the
4597 code it's annotating.
4599 @section @code{.cfi_register @var{register1}, @var{register2}}
4600 Previous value of @var{register1} is saved in register @var{register2}.
4602 @section @code{.cfi_restore @var{register}}
4603 @code{.cfi_restore} says that the rule for @var{register} is now the
4604 same as it was at the beginning of the function, after all initial
4605 instruction added by @code{.cfi_startproc} were executed.
4607 @section @code{.cfi_undefined @var{register}}
4608 From now on the previous value of @var{register} can't be restored anymore.
4610 @section @code{.cfi_same_value @var{register}}
4611 Current value of @var{register} is the same like in the previous frame,
4612 i.e. no restoration needed.
4614 @section @code{.cfi_remember_state},
4615 First save all current rules for all registers by @code{.cfi_remember_state},
4616 then totally screw them up by subsequent @code{.cfi_*} directives and when
4617 everything is hopelessly bad, use @code{.cfi_restore_state} to restore
4618 the previous saved state.
4620 @section @code{.cfi_return_column @var{register}}
4621 Change return column @var{register}, i.e. the return address is either
4622 directly in @var{register} or can be accessed by rules for @var{register}.
4624 @section @code{.cfi_signal_frame}
4625 Mark current function as signal trampoline.
4627 @section @code{.cfi_window_save}
4628 SPARC register window has been saved.
4630 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4631 Allows the user to add arbitrary bytes to the unwind info. One
4632 might use this to add OS-specific CFI opcodes, or generic CFI
4633 opcodes that GAS does not yet support.
4635 @section @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4636 The current value of @var{register} is @var{label}. The value of @var{label}
4637 will be encoded in the output file according to @var{encoding}; see the
4638 description of @code{.cfi_personality} for details on this encoding.
4640 The usefulness of equating a register to a fixed label is probably
4641 limited to the return address register. Here, it can be useful to
4642 mark a code segment that has only one return address which is reached
4643 by a direct branch and no copy of the return address exists in memory
4644 or another register.
4647 @section @code{.comm @var{symbol} , @var{length} }
4649 @cindex @code{comm} directive
4650 @cindex symbol, common
4651 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4652 common symbol in one object file may be merged with a defined or common symbol
4653 of the same name in another object file. If @code{@value{LD}} does not see a
4654 definition for the symbol--just one or more common symbols--then it will
4655 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4656 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4657 the same name, and they do not all have the same size, it will allocate space
4658 using the largest size.
4661 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4662 an optional third argument. This is the desired alignment of the symbol,
4663 specified for ELF as a byte boundary (for example, an alignment of 16 means
4664 that the least significant 4 bits of the address should be zero), and for PE
4665 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4666 boundary). The alignment must be an absolute expression, and it must be a
4667 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4668 common symbol, it will use the alignment when placing the symbol. If no
4669 alignment is specified, @command{@value{AS}} will set the alignment to the
4670 largest power of two less than or equal to the size of the symbol, up to a
4671 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4672 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4673 @samp{--section-alignment} option; image file sections in PE are aligned to
4674 multiples of 4096, which is far too large an alignment for ordinary variables.
4675 It is rather the default alignment for (non-debug) sections within object
4676 (@samp{*.o}) files, which are less strictly aligned.}.
4680 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4681 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4685 @section @code{.data @var{subsection}}
4687 @cindex @code{data} directive
4688 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4689 end of the data subsection numbered @var{subsection} (which is an
4690 absolute expression). If @var{subsection} is omitted, it defaults
4695 @section @code{.def @var{name}}
4697 @cindex @code{def} directive
4698 @cindex COFF symbols, debugging
4699 @cindex debugging COFF symbols
4700 Begin defining debugging information for a symbol @var{name}; the
4701 definition extends until the @code{.endef} directive is encountered.
4704 This directive is only observed when @command{@value{AS}} is configured for COFF
4705 format output; when producing @code{b.out}, @samp{.def} is recognized,
4712 @section @code{.desc @var{symbol}, @var{abs-expression}}
4714 @cindex @code{desc} directive
4715 @cindex COFF symbol descriptor
4716 @cindex symbol descriptor, COFF
4717 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4718 to the low 16 bits of an absolute expression.
4721 The @samp{.desc} directive is not available when @command{@value{AS}} is
4722 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4723 object format. For the sake of compatibility, @command{@value{AS}} accepts
4724 it, but produces no output, when configured for COFF.
4730 @section @code{.dim}
4732 @cindex @code{dim} directive
4733 @cindex COFF auxiliary symbol information
4734 @cindex auxiliary symbol information, COFF
4735 This directive is generated by compilers to include auxiliary debugging
4736 information in the symbol table. It is only permitted inside
4737 @code{.def}/@code{.endef} pairs.
4740 @samp{.dim} is only meaningful when generating COFF format output; when
4741 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4747 @section @code{.double @var{flonums}}
4749 @cindex @code{double} directive
4750 @cindex floating point numbers (double)
4751 @code{.double} expects zero or more flonums, separated by commas. It
4752 assembles floating point numbers.
4754 The exact kind of floating point numbers emitted depends on how
4755 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4759 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4760 in @sc{ieee} format.
4765 @section @code{.eject}
4767 @cindex @code{eject} directive
4768 @cindex new page, in listings
4769 @cindex page, in listings
4770 @cindex listing control: new page
4771 Force a page break at this point, when generating assembly listings.
4774 @section @code{.else}
4776 @cindex @code{else} directive
4777 @code{.else} is part of the @command{@value{AS}} support for conditional
4778 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
4779 of code to be assembled if the condition for the preceding @code{.if}
4783 @section @code{.elseif}
4785 @cindex @code{elseif} directive
4786 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4787 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
4788 @code{.if} block that would otherwise fill the entire @code{.else} section.
4791 @section @code{.end}
4793 @cindex @code{end} directive
4794 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4795 process anything in the file past the @code{.end} directive.
4799 @section @code{.endef}
4801 @cindex @code{endef} directive
4802 This directive flags the end of a symbol definition begun with
4806 @samp{.endef} is only meaningful when generating COFF format output; if
4807 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4808 directive but ignores it.
4813 @section @code{.endfunc}
4814 @cindex @code{endfunc} directive
4815 @code{.endfunc} marks the end of a function specified with @code{.func}.
4818 @section @code{.endif}
4820 @cindex @code{endif} directive
4821 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4822 it marks the end of a block of code that is only assembled
4823 conditionally. @xref{If,,@code{.if}}.
4826 @section @code{.equ @var{symbol}, @var{expression}}
4828 @cindex @code{equ} directive
4829 @cindex assigning values to symbols
4830 @cindex symbols, assigning values to
4831 This directive sets the value of @var{symbol} to @var{expression}.
4832 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
4835 The syntax for @code{equ} on the HPPA is
4836 @samp{@var{symbol} .equ @var{expression}}.
4840 The syntax for @code{equ} on the Z80 is
4841 @samp{@var{symbol} equ @var{expression}}.
4842 On the Z80 it is an eror if @var{symbol} is already defined,
4843 but the symbol is not protected from later redefinition.
4844 Compare @ref{Equiv}.
4848 @section @code{.equiv @var{symbol}, @var{expression}}
4849 @cindex @code{equiv} directive
4850 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4851 the assembler will signal an error if @var{symbol} is already defined. Note a
4852 symbol which has been referenced but not actually defined is considered to be
4855 Except for the contents of the error message, this is roughly equivalent to
4862 plus it protects the symbol from later redefinition.
4865 @section @code{.eqv @var{symbol}, @var{expression}}
4866 @cindex @code{eqv} directive
4867 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4868 evaluate the expression or any part of it immediately. Instead each time
4869 the resulting symbol is used in an expression, a snapshot of its current
4873 @section @code{.err}
4874 @cindex @code{err} directive
4875 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4876 message and, unless the @option{-Z} option was used, it will not generate an
4877 object file. This can be used to signal an error in conditionally compiled code.
4880 @section @code{.error "@var{string}"}
4881 @cindex error directive
4883 Similarly to @code{.err}, this directive emits an error, but you can specify a
4884 string that will be emitted as the error message. If you don't specify the
4885 message, it defaults to @code{".error directive invoked in source file"}.
4886 @xref{Errors, ,Error and Warning Messages}.
4889 .error "This code has not been assembled and tested."
4893 @section @code{.exitm}
4894 Exit early from the current macro definition. @xref{Macro}.
4897 @section @code{.extern}
4899 @cindex @code{extern} directive
4900 @code{.extern} is accepted in the source program---for compatibility
4901 with other assemblers---but it is ignored. @command{@value{AS}} treats
4902 all undefined symbols as external.
4905 @section @code{.fail @var{expression}}
4907 @cindex @code{fail} directive
4908 Generates an error or a warning. If the value of the @var{expression} is 500
4909 or more, @command{@value{AS}} will print a warning message. If the value is less
4910 than 500, @command{@value{AS}} will print an error message. The message will
4911 include the value of @var{expression}. This can occasionally be useful inside
4912 complex nested macros or conditional assembly.
4915 @section @code{.file}
4916 @cindex @code{file} directive
4918 @ifclear no-file-dir
4919 There are two different versions of the @code{.file} directive. Targets
4920 that support DWARF2 line number information use the DWARF2 version of
4921 @code{.file}. Other targets use the default version.
4923 @subheading Default Version
4925 @cindex logical file name
4926 @cindex file name, logical
4927 This version of the @code{.file} directive tells @command{@value{AS}} that we
4928 are about to start a new logical file. The syntax is:
4934 @var{string} is the new file name. In general, the filename is
4935 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4936 to specify an empty file name, you must give the quotes--@code{""}. This
4937 statement may go away in future: it is only recognized to be compatible with
4938 old @command{@value{AS}} programs.
4940 @subheading DWARF2 Version
4943 When emitting DWARF2 line number information, @code{.file} assigns filenames
4944 to the @code{.debug_line} file name table. The syntax is:
4947 .file @var{fileno} @var{filename}
4950 The @var{fileno} operand should be a unique positive integer to use as the
4951 index of the entry in the table. The @var{filename} operand is a C string
4954 The detail of filename indices is exposed to the user because the filename
4955 table is shared with the @code{.debug_info} section of the DWARF2 debugging
4956 information, and thus the user must know the exact indices that table
4960 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4962 @cindex @code{fill} directive
4963 @cindex writing patterns in memory
4964 @cindex patterns, writing in memory
4965 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4966 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4967 may be zero or more. @var{Size} may be zero or more, but if it is
4968 more than 8, then it is deemed to have the value 8, compatible with
4969 other people's assemblers. The contents of each @var{repeat} bytes
4970 is taken from an 8-byte number. The highest order 4 bytes are
4971 zero. The lowest order 4 bytes are @var{value} rendered in the
4972 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4973 Each @var{size} bytes in a repetition is taken from the lowest order
4974 @var{size} bytes of this number. Again, this bizarre behavior is
4975 compatible with other people's assemblers.
4977 @var{size} and @var{value} are optional.
4978 If the second comma and @var{value} are absent, @var{value} is
4979 assumed zero. If the first comma and following tokens are absent,
4980 @var{size} is assumed to be 1.
4983 @section @code{.float @var{flonums}}
4985 @cindex floating point numbers (single)
4986 @cindex @code{float} directive
4987 This directive assembles zero or more flonums, separated by commas. It
4988 has the same effect as @code{.single}.
4990 The exact kind of floating point numbers emitted depends on how
4991 @command{@value{AS}} is configured.
4992 @xref{Machine Dependencies}.
4996 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4997 in @sc{ieee} format.
5002 @section @code{.func @var{name}[,@var{label}]}
5003 @cindex @code{func} directive
5004 @code{.func} emits debugging information to denote function @var{name}, and
5005 is ignored unless the file is assembled with debugging enabled.
5006 Only @samp{--gstabs[+]} is currently supported.
5007 @var{label} is the entry point of the function and if omitted @var{name}
5008 prepended with the @samp{leading char} is used.
5009 @samp{leading char} is usually @code{_} or nothing, depending on the target.
5010 All functions are currently defined to have @code{void} return type.
5011 The function must be terminated with @code{.endfunc}.
5014 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
5016 @cindex @code{global} directive
5017 @cindex symbol, making visible to linker
5018 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
5019 @var{symbol} in your partial program, its value is made available to
5020 other partial programs that are linked with it. Otherwise,
5021 @var{symbol} takes its attributes from a symbol of the same name
5022 from another file linked into the same program.
5024 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
5025 compatibility with other assemblers.
5028 On the HPPA, @code{.global} is not always enough to make it accessible to other
5029 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
5030 @xref{HPPA Directives, ,HPPA Assembler Directives}.
5035 @section @code{.gnu_attribute @var{tag},@var{value}}
5036 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
5039 @section @code{.hidden @var{names}}
5041 @cindex @code{hidden} directive
5043 This is one of the ELF visibility directives. The other two are
5044 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
5045 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5047 This directive overrides the named symbols default visibility (which is set by
5048 their binding: local, global or weak). The directive sets the visibility to
5049 @code{hidden} which means that the symbols are not visible to other components.
5050 Such symbols are always considered to be @code{protected} as well.
5054 @section @code{.hword @var{expressions}}
5056 @cindex @code{hword} directive
5057 @cindex integers, 16-bit
5058 @cindex numbers, 16-bit
5059 @cindex sixteen bit integers
5060 This expects zero or more @var{expressions}, and emits
5061 a 16 bit number for each.
5064 This directive is a synonym for @samp{.short}; depending on the target
5065 architecture, it may also be a synonym for @samp{.word}.
5069 This directive is a synonym for @samp{.short}.
5072 This directive is a synonym for both @samp{.short} and @samp{.word}.
5077 @section @code{.ident}
5079 @cindex @code{ident} directive
5081 This directive is used by some assemblers to place tags in object files. The
5082 behavior of this directive varies depending on the target. When using the
5083 a.out object file format, @command{@value{AS}} simply accepts the directive for
5084 source-file compatibility with existing assemblers, but does not emit anything
5085 for it. When using COFF, comments are emitted to the @code{.comment} or
5086 @code{.rdata} section, depending on the target. When using ELF, comments are
5087 emitted to the @code{.comment} section.
5090 @section @code{.if @var{absolute expression}}
5092 @cindex conditional assembly
5093 @cindex @code{if} directive
5094 @code{.if} marks the beginning of a section of code which is only
5095 considered part of the source program being assembled if the argument
5096 (which must be an @var{absolute expression}) is non-zero. The end of
5097 the conditional section of code must be marked by @code{.endif}
5098 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
5099 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
5100 If you have several conditions to check, @code{.elseif} may be used to avoid
5101 nesting blocks if/else within each subsequent @code{.else} block.
5103 The following variants of @code{.if} are also supported:
5105 @cindex @code{ifdef} directive
5106 @item .ifdef @var{symbol}
5107 Assembles the following section of code if the specified @var{symbol}
5108 has been defined. Note a symbol which has been referenced but not yet defined
5109 is considered to be undefined.
5111 @cindex @code{ifb} directive
5112 @item .ifb @var{text}
5113 Assembles the following section of code if the operand is blank (empty).
5115 @cindex @code{ifc} directive
5116 @item .ifc @var{string1},@var{string2}
5117 Assembles the following section of code if the two strings are the same. The
5118 strings may be optionally quoted with single quotes. If they are not quoted,
5119 the first string stops at the first comma, and the second string stops at the
5120 end of the line. Strings which contain whitespace should be quoted. The
5121 string comparison is case sensitive.
5123 @cindex @code{ifeq} directive
5124 @item .ifeq @var{absolute expression}
5125 Assembles the following section of code if the argument is zero.
5127 @cindex @code{ifeqs} directive
5128 @item .ifeqs @var{string1},@var{string2}
5129 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5131 @cindex @code{ifge} directive
5132 @item .ifge @var{absolute expression}
5133 Assembles the following section of code if the argument is greater than or
5136 @cindex @code{ifgt} directive
5137 @item .ifgt @var{absolute expression}
5138 Assembles the following section of code if the argument is greater than zero.
5140 @cindex @code{ifle} directive
5141 @item .ifle @var{absolute expression}
5142 Assembles the following section of code if the argument is less than or equal
5145 @cindex @code{iflt} directive
5146 @item .iflt @var{absolute expression}
5147 Assembles the following section of code if the argument is less than zero.
5149 @cindex @code{ifnb} directive
5150 @item .ifnb @var{text}
5151 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5152 following section of code if the operand is non-blank (non-empty).
5154 @cindex @code{ifnc} directive
5155 @item .ifnc @var{string1},@var{string2}.
5156 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5157 following section of code if the two strings are not the same.
5159 @cindex @code{ifndef} directive
5160 @cindex @code{ifnotdef} directive
5161 @item .ifndef @var{symbol}
5162 @itemx .ifnotdef @var{symbol}
5163 Assembles the following section of code if the specified @var{symbol}
5164 has not been defined. Both spelling variants are equivalent. Note a symbol
5165 which has been referenced but not yet defined is considered to be undefined.
5167 @cindex @code{ifne} directive
5168 @item .ifne @var{absolute expression}
5169 Assembles the following section of code if the argument is not equal to zero
5170 (in other words, this is equivalent to @code{.if}).
5172 @cindex @code{ifnes} directive
5173 @item .ifnes @var{string1},@var{string2}
5174 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5175 following section of code if the two strings are not the same.
5179 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5181 @cindex @code{incbin} directive
5182 @cindex binary files, including
5183 The @code{incbin} directive includes @var{file} verbatim at the current
5184 location. You can control the search paths used with the @samp{-I} command-line
5185 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5188 The @var{skip} argument skips a number of bytes from the start of the
5189 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5190 read. Note that the data is not aligned in any way, so it is the user's
5191 responsibility to make sure that proper alignment is provided both before and
5192 after the @code{incbin} directive.
5195 @section @code{.include "@var{file}"}
5197 @cindex @code{include} directive
5198 @cindex supporting files, including
5199 @cindex files, including
5200 This directive provides a way to include supporting files at specified
5201 points in your source program. The code from @var{file} is assembled as
5202 if it followed the point of the @code{.include}; when the end of the
5203 included file is reached, assembly of the original file continues. You
5204 can control the search paths used with the @samp{-I} command-line option
5205 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5209 @section @code{.int @var{expressions}}
5211 @cindex @code{int} directive
5212 @cindex integers, 32-bit
5213 Expect zero or more @var{expressions}, of any section, separated by commas.
5214 For each expression, emit a number that, at run time, is the value of that
5215 expression. The byte order and bit size of the number depends on what kind
5216 of target the assembly is for.
5220 On most forms of the H8/300, @code{.int} emits 16-bit
5221 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5228 @section @code{.internal @var{names}}
5230 @cindex @code{internal} directive
5232 This is one of the ELF visibility directives. The other two are
5233 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5234 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5236 This directive overrides the named symbols default visibility (which is set by
5237 their binding: local, global or weak). The directive sets the visibility to
5238 @code{internal} which means that the symbols are considered to be @code{hidden}
5239 (i.e., not visible to other components), and that some extra, processor specific
5240 processing must also be performed upon the symbols as well.
5244 @section @code{.irp @var{symbol},@var{values}}@dots{}
5246 @cindex @code{irp} directive
5247 Evaluate a sequence of statements assigning different values to @var{symbol}.
5248 The sequence of statements starts at the @code{.irp} directive, and is
5249 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5250 set to @var{value}, and the sequence of statements is assembled. If no
5251 @var{value} is listed, the sequence of statements is assembled once, with
5252 @var{symbol} set to the null string. To refer to @var{symbol} within the
5253 sequence of statements, use @var{\symbol}.
5255 For example, assembling
5263 is equivalent to assembling
5271 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5274 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5276 @cindex @code{irpc} directive
5277 Evaluate a sequence of statements assigning different values to @var{symbol}.
5278 The sequence of statements starts at the @code{.irpc} directive, and is
5279 terminated by an @code{.endr} directive. For each character in @var{value},
5280 @var{symbol} is set to the character, and the sequence of statements is
5281 assembled. If no @var{value} is listed, the sequence of statements is
5282 assembled once, with @var{symbol} set to the null string. To refer to
5283 @var{symbol} within the sequence of statements, use @var{\symbol}.
5285 For example, assembling
5293 is equivalent to assembling
5301 For some caveats with the spelling of @var{symbol}, see also the discussion
5305 @section @code{.lcomm @var{symbol} , @var{length}}
5307 @cindex @code{lcomm} directive
5308 @cindex local common symbols
5309 @cindex symbols, local common
5310 Reserve @var{length} (an absolute expression) bytes for a local common
5311 denoted by @var{symbol}. The section and value of @var{symbol} are
5312 those of the new local common. The addresses are allocated in the bss
5313 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5314 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5315 not visible to @code{@value{LD}}.
5318 Some targets permit a third argument to be used with @code{.lcomm}. This
5319 argument specifies the desired alignment of the symbol in the bss section.
5323 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5324 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5328 @section @code{.lflags}
5330 @cindex @code{lflags} directive (ignored)
5331 @command{@value{AS}} accepts this directive, for compatibility with other
5332 assemblers, but ignores it.
5334 @ifclear no-line-dir
5336 @section @code{.line @var{line-number}}
5338 @cindex @code{line} directive
5339 @cindex logical line number
5341 Change the logical line number. @var{line-number} must be an absolute
5342 expression. The next line has that logical line number. Therefore any other
5343 statements on the current line (after a statement separator character) are
5344 reported as on logical line number @var{line-number} @minus{} 1. One day
5345 @command{@value{AS}} will no longer support this directive: it is recognized only
5346 for compatibility with existing assembler programs.
5349 Even though this is a directive associated with the @code{a.out} or
5350 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5351 when producing COFF output, and treats @samp{.line} as though it
5352 were the COFF @samp{.ln} @emph{if} it is found outside a
5353 @code{.def}/@code{.endef} pair.
5355 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5356 used by compilers to generate auxiliary symbol information for
5361 @section @code{.linkonce [@var{type}]}
5363 @cindex @code{linkonce} directive
5364 @cindex common sections
5365 Mark the current section so that the linker only includes a single copy of it.
5366 This may be used to include the same section in several different object files,
5367 but ensure that the linker will only include it once in the final output file.
5368 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5369 Duplicate sections are detected based on the section name, so it should be
5372 This directive is only supported by a few object file formats; as of this
5373 writing, the only object file format which supports it is the Portable
5374 Executable format used on Windows NT.
5376 The @var{type} argument is optional. If specified, it must be one of the
5377 following strings. For example:
5381 Not all types may be supported on all object file formats.
5385 Silently discard duplicate sections. This is the default.
5388 Warn if there are duplicate sections, but still keep only one copy.
5391 Warn if any of the duplicates have different sizes.
5394 Warn if any of the duplicates do not have exactly the same contents.
5398 @section @code{.list}
5400 @cindex @code{list} directive
5401 @cindex listing control, turning on
5402 Control (in conjunction with the @code{.nolist} directive) whether or
5403 not assembly listings are generated. These two directives maintain an
5404 internal counter (which is zero initially). @code{.list} increments the
5405 counter, and @code{.nolist} decrements it. Assembly listings are
5406 generated whenever the counter is greater than zero.
5408 By default, listings are disabled. When you enable them (with the
5409 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5410 the initial value of the listing counter is one.
5413 @section @code{.ln @var{line-number}}
5415 @cindex @code{ln} directive
5416 @ifclear no-line-dir
5417 @samp{.ln} is a synonym for @samp{.line}.
5420 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5421 must be an absolute expression. The next line has that logical
5422 line number, so any other statements on the current line (after a
5423 statement separator character @code{;}) are reported as on logical
5424 line number @var{line-number} @minus{} 1.
5427 This directive is accepted, but ignored, when @command{@value{AS}} is
5428 configured for @code{b.out}; its effect is only associated with COFF
5434 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5435 @cindex @code{loc} directive
5436 When emitting DWARF2 line number information,
5437 the @code{.loc} directive will add a row to the @code{.debug_line} line
5438 number matrix corresponding to the immediately following assembly
5439 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5440 arguments will be applied to the @code{.debug_line} state machine before
5443 The @var{options} are a sequence of the following tokens in any order:
5447 This option will set the @code{basic_block} register in the
5448 @code{.debug_line} state machine to @code{true}.
5451 This option will set the @code{prologue_end} register in the
5452 @code{.debug_line} state machine to @code{true}.
5454 @item epilogue_begin
5455 This option will set the @code{epilogue_begin} register in the
5456 @code{.debug_line} state machine to @code{true}.
5458 @item is_stmt @var{value}
5459 This option will set the @code{is_stmt} register in the
5460 @code{.debug_line} state machine to @code{value}, which must be
5463 @item isa @var{value}
5464 This directive will set the @code{isa} register in the @code{.debug_line}
5465 state machine to @var{value}, which must be an unsigned integer.
5467 @item discriminator @var{value}
5468 This directive will set the @code{discriminator} register in the @code{.debug_line}
5469 state machine to @var{value}, which must be an unsigned integer.
5473 @node Loc_mark_labels
5474 @section @code{.loc_mark_labels @var{enable}}
5475 @cindex @code{loc_mark_labels} directive
5476 When emitting DWARF2 line number information,
5477 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5478 to the @code{.debug_line} line number matrix with the @code{basic_block}
5479 register in the state machine set whenever a code label is seen.
5480 The @var{enable} argument should be either 1 or 0, to enable or disable
5481 this function respectively.
5485 @section @code{.local @var{names}}
5487 @cindex @code{local} directive
5488 This directive, which is available for ELF targets, marks each symbol in
5489 the comma-separated list of @code{names} as a local symbol so that it
5490 will not be externally visible. If the symbols do not already exist,
5491 they will be created.
5493 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5494 accept an alignment argument, which is the case for most ELF targets,
5495 the @code{.local} directive can be used in combination with @code{.comm}
5496 (@pxref{Comm}) to define aligned local common data.
5500 @section @code{.long @var{expressions}}
5502 @cindex @code{long} directive
5503 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5506 @c no one seems to know what this is for or whether this description is
5507 @c what it really ought to do
5509 @section @code{.lsym @var{symbol}, @var{expression}}
5511 @cindex @code{lsym} directive
5512 @cindex symbol, not referenced in assembly
5513 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5514 the hash table, ensuring it cannot be referenced by name during the
5515 rest of the assembly. This sets the attributes of the symbol to be
5516 the same as the expression value:
5518 @var{other} = @var{descriptor} = 0
5519 @var{type} = @r{(section of @var{expression})}
5520 @var{value} = @var{expression}
5523 The new symbol is not flagged as external.
5527 @section @code{.macro}
5530 The commands @code{.macro} and @code{.endm} allow you to define macros that
5531 generate assembly output. For example, this definition specifies a macro
5532 @code{sum} that puts a sequence of numbers into memory:
5535 .macro sum from=0, to=5
5544 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5556 @item .macro @var{macname}
5557 @itemx .macro @var{macname} @var{macargs} @dots{}
5558 @cindex @code{macro} directive
5559 Begin the definition of a macro called @var{macname}. If your macro
5560 definition requires arguments, specify their names after the macro name,
5561 separated by commas or spaces. You can qualify the macro argument to
5562 indicate whether all invocations must specify a non-blank value (through
5563 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5564 (through @samp{:@code{vararg}}). You can supply a default value for any
5565 macro argument by following the name with @samp{=@var{deflt}}. You
5566 cannot define two macros with the same @var{macname} unless it has been
5567 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5568 definitions. For example, these are all valid @code{.macro} statements:
5572 Begin the definition of a macro called @code{comm}, which takes no
5575 @item .macro plus1 p, p1
5576 @itemx .macro plus1 p p1
5577 Either statement begins the definition of a macro called @code{plus1},
5578 which takes two arguments; within the macro definition, write
5579 @samp{\p} or @samp{\p1} to evaluate the arguments.
5581 @item .macro reserve_str p1=0 p2
5582 Begin the definition of a macro called @code{reserve_str}, with two
5583 arguments. The first argument has a default value, but not the second.
5584 After the definition is complete, you can call the macro either as
5585 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5586 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5587 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5588 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5590 @item .macro m p1:req, p2=0, p3:vararg
5591 Begin the definition of a macro called @code{m}, with at least three
5592 arguments. The first argument must always have a value specified, but
5593 not the second, which instead has a default value. The third formal
5594 will get assigned all remaining arguments specified at invocation time.
5596 When you call a macro, you can specify the argument values either by
5597 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5598 @samp{sum to=17, from=9}.
5602 Note that since each of the @var{macargs} can be an identifier exactly
5603 as any other one permitted by the target architecture, there may be
5604 occasional problems if the target hand-crafts special meanings to certain
5605 characters when they occur in a special position. For example, if the colon
5606 (@code{:}) is generally permitted to be part of a symbol name, but the
5607 architecture specific code special-cases it when occurring as the final
5608 character of a symbol (to denote a label), then the macro parameter
5609 replacement code will have no way of knowing that and consider the whole
5610 construct (including the colon) an identifier, and check only this
5611 identifier for being the subject to parameter substitution. So for example
5612 this macro definition:
5620 might not work as expected. Invoking @samp{label foo} might not create a label
5621 called @samp{foo} but instead just insert the text @samp{\l:} into the
5622 assembler source, probably generating an error about an unrecognised
5625 Similarly problems might occur with the period character (@samp{.})
5626 which is often allowed inside opcode names (and hence identifier names). So
5627 for example constructing a macro to build an opcode from a base name and a
5628 length specifier like this:
5631 .macro opcode base length
5636 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5637 instruction but instead generate some kind of error as the assembler tries to
5638 interpret the text @samp{\base.\length}.
5640 There are several possible ways around this problem:
5643 @item Insert white space
5644 If it is possible to use white space characters then this is the simplest
5653 @item Use @samp{\()}
5654 The string @samp{\()} can be used to separate the end of a macro argument from
5655 the following text. eg:
5658 .macro opcode base length
5663 @item Use the alternate macro syntax mode
5664 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5665 used as a separator. eg:
5675 Note: this problem of correctly identifying string parameters to pseudo ops
5676 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5677 and @code{.irpc} (@pxref{Irpc}) as well.
5680 @cindex @code{endm} directive
5681 Mark the end of a macro definition.
5684 @cindex @code{exitm} directive
5685 Exit early from the current macro definition.
5687 @cindex number of macros executed
5688 @cindex macros, count executed
5690 @command{@value{AS}} maintains a counter of how many macros it has
5691 executed in this pseudo-variable; you can copy that number to your
5692 output with @samp{\@@}, but @emph{only within a macro definition}.
5694 @item LOCAL @var{name} [ , @dots{} ]
5695 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5696 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5697 @xref{Altmacro,,@code{.altmacro}}.
5701 @section @code{.mri @var{val}}
5703 @cindex @code{mri} directive
5704 @cindex MRI mode, temporarily
5705 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
5706 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
5707 affects code assembled until the next @code{.mri} directive, or until the end
5708 of the file. @xref{M, MRI mode, MRI mode}.
5711 @section @code{.noaltmacro}
5712 Disable alternate macro mode. @xref{Altmacro}.
5715 @section @code{.nolist}
5717 @cindex @code{nolist} directive
5718 @cindex listing control, turning off
5719 Control (in conjunction with the @code{.list} directive) whether or
5720 not assembly listings are generated. These two directives maintain an
5721 internal counter (which is zero initially). @code{.list} increments the
5722 counter, and @code{.nolist} decrements it. Assembly listings are
5723 generated whenever the counter is greater than zero.
5726 @section @code{.octa @var{bignums}}
5728 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5729 @cindex @code{octa} directive
5730 @cindex integer, 16-byte
5731 @cindex sixteen byte integer
5732 This directive expects zero or more bignums, separated by commas. For each
5733 bignum, it emits a 16-byte integer.
5735 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5736 hence @emph{octa}-word for 16 bytes.
5739 @section @code{.offset @var{loc}}
5741 @cindex @code{offset} directive
5742 Set the location counter to @var{loc} in the absolute section. @var{loc} must
5743 be an absolute expression. This directive may be useful for defining
5744 symbols with absolute values. Do not confuse it with the @code{.org}
5748 @section @code{.org @var{new-lc} , @var{fill}}
5750 @cindex @code{org} directive
5751 @cindex location counter, advancing
5752 @cindex advancing location counter
5753 @cindex current address, advancing
5754 Advance the location counter of the current section to
5755 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5756 expression with the same section as the current subsection. That is,
5757 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5758 wrong section, the @code{.org} directive is ignored. To be compatible
5759 with former assemblers, if the section of @var{new-lc} is absolute,
5760 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5761 is the same as the current subsection.
5763 @code{.org} may only increase the location counter, or leave it
5764 unchanged; you cannot use @code{.org} to move the location counter
5767 @c double negative used below "not undefined" because this is a specific
5768 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5769 @c section. doc@cygnus.com 18feb91
5770 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5771 may not be undefined. If you really detest this restriction we eagerly await
5772 a chance to share your improved assembler.
5774 Beware that the origin is relative to the start of the section, not
5775 to the start of the subsection. This is compatible with other
5776 people's assemblers.
5778 When the location counter (of the current subsection) is advanced, the
5779 intervening bytes are filled with @var{fill} which should be an
5780 absolute expression. If the comma and @var{fill} are omitted,
5781 @var{fill} defaults to zero.
5784 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5786 @cindex padding the location counter given a power of two
5787 @cindex @code{p2align} directive
5788 Pad the location counter (in the current subsection) to a particular
5789 storage boundary. The first expression (which must be absolute) is the
5790 number of low-order zero bits the location counter must have after
5791 advancement. For example @samp{.p2align 3} advances the location
5792 counter until it a multiple of 8. If the location counter is already a
5793 multiple of 8, no change is needed.
5795 The second expression (also absolute) gives the fill value to be stored in the
5796 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5797 padding bytes are normally zero. However, on some systems, if the section is
5798 marked as containing code and the fill value is omitted, the space is filled
5799 with no-op instructions.
5801 The third expression is also absolute, and is also optional. If it is present,
5802 it is the maximum number of bytes that should be skipped by this alignment
5803 directive. If doing the alignment would require skipping more bytes than the
5804 specified maximum, then the alignment is not done at all. You can omit the
5805 fill value (the second argument) entirely by simply using two commas after the
5806 required alignment; this can be useful if you want the alignment to be filled
5807 with no-op instructions when appropriate.
5809 @cindex @code{p2alignw} directive
5810 @cindex @code{p2alignl} directive
5811 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5812 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5813 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5814 fill pattern as a four byte longword value. For example, @code{.p2alignw
5815 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5816 filled in with the value 0x368d (the exact placement of the bytes depends upon
5817 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5822 @section @code{.popsection}
5824 @cindex @code{popsection} directive
5825 @cindex Section Stack
5826 This is one of the ELF section stack manipulation directives. The others are
5827 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5828 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5831 This directive replaces the current section (and subsection) with the top
5832 section (and subsection) on the section stack. This section is popped off the
5838 @section @code{.previous}
5840 @cindex @code{previous} directive
5841 @cindex Section Stack
5842 This is one of the ELF section stack manipulation directives. The others are
5843 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5844 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5845 (@pxref{PopSection}).
5847 This directive swaps the current section (and subsection) with most recently
5848 referenced section/subsection pair prior to this one. Multiple
5849 @code{.previous} directives in a row will flip between two sections (and their
5850 subsections). For example:
5862 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
5868 # Now in section A subsection 1
5872 # Now in section B subsection 0
5875 # Now in section B subsection 1
5878 # Now in section B subsection 0
5882 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
5883 section B and 0x9abc into subsection 1 of section B.
5885 In terms of the section stack, this directive swaps the current section with
5886 the top section on the section stack.
5890 @section @code{.print @var{string}}
5892 @cindex @code{print} directive
5893 @command{@value{AS}} will print @var{string} on the standard output during
5894 assembly. You must put @var{string} in double quotes.
5898 @section @code{.protected @var{names}}
5900 @cindex @code{protected} directive
5902 This is one of the ELF visibility directives. The other two are
5903 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5905 This directive overrides the named symbols default visibility (which is set by
5906 their binding: local, global or weak). The directive sets the visibility to
5907 @code{protected} which means that any references to the symbols from within the
5908 components that defines them must be resolved to the definition in that
5909 component, even if a definition in another component would normally preempt
5914 @section @code{.psize @var{lines} , @var{columns}}
5916 @cindex @code{psize} directive
5917 @cindex listing control: paper size
5918 @cindex paper size, for listings
5919 Use this directive to declare the number of lines---and, optionally, the
5920 number of columns---to use for each page, when generating listings.
5922 If you do not use @code{.psize}, listings use a default line-count
5923 of 60. You may omit the comma and @var{columns} specification; the
5924 default width is 200 columns.
5926 @command{@value{AS}} generates formfeeds whenever the specified number of
5927 lines is exceeded (or whenever you explicitly request one, using
5930 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5931 those explicitly specified with @code{.eject}.
5934 @section @code{.purgem @var{name}}
5936 @cindex @code{purgem} directive
5937 Undefine the macro @var{name}, so that later uses of the string will not be
5938 expanded. @xref{Macro}.
5942 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
5944 @cindex @code{pushsection} directive
5945 @cindex Section Stack
5946 This is one of the ELF section stack manipulation directives. The others are
5947 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5948 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5951 This directive pushes the current section (and subsection) onto the
5952 top of the section stack, and then replaces the current section and
5953 subsection with @code{name} and @code{subsection}. The optional
5954 @code{flags}, @code{type} and @code{arguments} are treated the same
5955 as in the @code{.section} (@pxref{Section}) directive.
5959 @section @code{.quad @var{bignums}}
5961 @cindex @code{quad} directive
5962 @code{.quad} expects zero or more bignums, separated by commas. For
5963 each bignum, it emits
5965 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5966 warning message; and just takes the lowest order 8 bytes of the bignum.
5967 @cindex eight-byte integer
5968 @cindex integer, 8-byte
5970 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5971 hence @emph{quad}-word for 8 bytes.
5974 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5975 warning message; and just takes the lowest order 16 bytes of the bignum.
5976 @cindex sixteen-byte integer
5977 @cindex integer, 16-byte
5981 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
5983 @cindex @code{reloc} directive
5984 Generate a relocation at @var{offset} of type @var{reloc_name} with value
5985 @var{expression}. If @var{offset} is a number, the relocation is generated in
5986 the current section. If @var{offset} is an expression that resolves to a
5987 symbol plus offset, the relocation is generated in the given symbol's section.
5988 @var{expression}, if present, must resolve to a symbol plus addend or to an
5989 absolute value, but note that not all targets support an addend. e.g. ELF REL
5990 targets such as i386 store an addend in the section contents rather than in the
5991 relocation. This low level interface does not support addends stored in the
5995 @section @code{.rept @var{count}}
5997 @cindex @code{rept} directive
5998 Repeat the sequence of lines between the @code{.rept} directive and the next
5999 @code{.endr} directive @var{count} times.
6001 For example, assembling
6009 is equivalent to assembling
6018 @section @code{.sbttl "@var{subheading}"}
6020 @cindex @code{sbttl} directive
6021 @cindex subtitles for listings
6022 @cindex listing control: subtitle
6023 Use @var{subheading} as the title (third line, immediately after the
6024 title line) when generating assembly listings.
6026 This directive affects subsequent pages, as well as the current page if
6027 it appears within ten lines of the top of a page.
6031 @section @code{.scl @var{class}}
6033 @cindex @code{scl} directive
6034 @cindex symbol storage class (COFF)
6035 @cindex COFF symbol storage class
6036 Set the storage-class value for a symbol. This directive may only be
6037 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
6038 whether a symbol is static or external, or it may record further
6039 symbolic debugging information.
6042 The @samp{.scl} directive is primarily associated with COFF output; when
6043 configured to generate @code{b.out} output format, @command{@value{AS}}
6044 accepts this directive but ignores it.
6050 @section @code{.section @var{name}}
6052 @cindex named section
6053 Use the @code{.section} directive to assemble the following code into a section
6056 This directive is only supported for targets that actually support arbitrarily
6057 named sections; on @code{a.out} targets, for example, it is not accepted, even
6058 with a standard @code{a.out} section name.
6062 @c only print the extra heading if both COFF and ELF are set
6063 @subheading COFF Version
6066 @cindex @code{section} directive (COFF version)
6067 For COFF targets, the @code{.section} directive is used in one of the following
6071 .section @var{name}[, "@var{flags}"]
6072 .section @var{name}[, @var{subsection}]
6075 If the optional argument is quoted, it is taken as flags to use for the
6076 section. Each flag is a single character. The following flags are recognized:
6079 bss section (uninitialized data)
6081 section is not loaded
6087 exclude section from linking
6093 shared section (meaningful for PE targets)
6095 ignored. (For compatibility with the ELF version)
6097 section is not readable (meaningful for PE targets)
6099 single-digit power-of-two section alignment (GNU extension)
6102 If no flags are specified, the default flags depend upon the section name. If
6103 the section name is not recognized, the default will be for the section to be
6104 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
6105 from the section, rather than adding them, so if they are used on their own it
6106 will be as if no flags had been specified at all.
6108 If the optional argument to the @code{.section} directive is not quoted, it is
6109 taken as a subsection number (@pxref{Sub-Sections}).
6114 @c only print the extra heading if both COFF and ELF are set
6115 @subheading ELF Version
6118 @cindex Section Stack
6119 This is one of the ELF section stack manipulation directives. The others are
6120 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
6121 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
6122 @code{.previous} (@pxref{Previous}).
6124 @cindex @code{section} directive (ELF version)
6125 For ELF targets, the @code{.section} directive is used like this:
6128 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
6131 The optional @var{flags} argument is a quoted string which may contain any
6132 combination of the following characters:
6135 section is allocatable
6137 section is excluded from executable and shared library.
6141 section is executable
6143 section is mergeable
6145 section contains zero terminated strings
6147 section is a member of a section group
6149 section is used for thread-local-storage
6151 section is a member of the previously-current section's group, if any
6154 The optional @var{type} argument may contain one of the following constants:
6157 section contains data
6159 section does not contain data (i.e., section only occupies space)
6161 section contains data which is used by things other than the program
6163 section contains an array of pointers to init functions
6165 section contains an array of pointers to finish functions
6166 @item @@preinit_array
6167 section contains an array of pointers to pre-init functions
6170 Many targets only support the first three section types.
6172 Note on targets where the @code{@@} character is the start of a comment (eg
6173 ARM) then another character is used instead. For example the ARM port uses the
6176 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6177 be specified as well as an extra argument---@var{entsize}---like this:
6180 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6183 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6184 constants, each @var{entsize} octets long. Sections with both @code{M} and
6185 @code{S} must contain zero terminated strings where each character is
6186 @var{entsize} bytes long. The linker may remove duplicates within sections with
6187 the same name, same entity size and same flags. @var{entsize} must be an
6188 absolute expression. For sections with both @code{M} and @code{S}, a string
6189 which is a suffix of a larger string is considered a duplicate. Thus
6190 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6191 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6193 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6194 be present along with an additional field like this:
6197 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6200 The @var{GroupName} field specifies the name of the section group to which this
6201 particular section belongs. The optional linkage field can contain:
6204 indicates that only one copy of this section should be retained
6209 Note: if both the @var{M} and @var{G} flags are present then the fields for
6210 the Merge flag should come first, like this:
6213 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6216 If @var{flags} contains the @code{?} symbol then it may not also contain the
6217 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6218 present. Instead, @code{?} says to consider the section that's current before
6219 this directive. If that section used @code{G}, then the new section will use
6220 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6221 If not, then the @code{?} symbol has no effect.
6223 If no flags are specified, the default flags depend upon the section name. If
6224 the section name is not recognized, the default will be for the section to have
6225 none of the above flags: it will not be allocated in memory, nor writable, nor
6226 executable. The section will contain data.
6228 For ELF targets, the assembler supports another type of @code{.section}
6229 directive for compatibility with the Solaris assembler:
6232 .section "@var{name}"[, @var{flags}...]
6235 Note that the section name is quoted. There may be a sequence of comma
6239 section is allocatable
6243 section is executable
6245 section is excluded from executable and shared library.
6247 section is used for thread local storage
6250 This directive replaces the current section and subsection. See the
6251 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6252 some examples of how this directive and the other section stack directives
6258 @section @code{.set @var{symbol}, @var{expression}}
6260 @cindex @code{set} directive
6261 @cindex symbol value, setting
6262 Set the value of @var{symbol} to @var{expression}. This
6263 changes @var{symbol}'s value and type to conform to
6264 @var{expression}. If @var{symbol} was flagged as external, it remains
6265 flagged (@pxref{Symbol Attributes}).
6267 You may @code{.set} a symbol many times in the same assembly.
6269 If you @code{.set} a global symbol, the value stored in the object
6270 file is the last value stored into it.
6273 On Z80 @code{set} is a real instruction, use
6274 @samp{@var{symbol} defl @var{expression}} instead.
6278 @section @code{.short @var{expressions}}
6280 @cindex @code{short} directive
6282 @code{.short} is normally the same as @samp{.word}.
6283 @xref{Word,,@code{.word}}.
6285 In some configurations, however, @code{.short} and @code{.word} generate
6286 numbers of different lengths. @xref{Machine Dependencies}.
6290 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6293 This expects zero or more @var{expressions}, and emits
6294 a 16 bit number for each.
6299 @section @code{.single @var{flonums}}
6301 @cindex @code{single} directive
6302 @cindex floating point numbers (single)
6303 This directive assembles zero or more flonums, separated by commas. It
6304 has the same effect as @code{.float}.
6306 The exact kind of floating point numbers emitted depends on how
6307 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6311 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6312 numbers in @sc{ieee} format.
6318 @section @code{.size}
6320 This directive is used to set the size associated with a symbol.
6324 @c only print the extra heading if both COFF and ELF are set
6325 @subheading COFF Version
6328 @cindex @code{size} directive (COFF version)
6329 For COFF targets, the @code{.size} directive is only permitted inside
6330 @code{.def}/@code{.endef} pairs. It is used like this:
6333 .size @var{expression}
6337 @samp{.size} is only meaningful when generating COFF format output; when
6338 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6345 @c only print the extra heading if both COFF and ELF are set
6346 @subheading ELF Version
6349 @cindex @code{size} directive (ELF version)
6350 For ELF targets, the @code{.size} directive is used like this:
6353 .size @var{name} , @var{expression}
6356 This directive sets the size associated with a symbol @var{name}.
6357 The size in bytes is computed from @var{expression} which can make use of label
6358 arithmetic. This directive is typically used to set the size of function
6363 @ifclear no-space-dir
6365 @section @code{.skip @var{size} , @var{fill}}
6367 @cindex @code{skip} directive
6368 @cindex filling memory
6369 This directive emits @var{size} bytes, each of value @var{fill}. Both
6370 @var{size} and @var{fill} are absolute expressions. If the comma and
6371 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6376 @section @code{.sleb128 @var{expressions}}
6378 @cindex @code{sleb128} directive
6379 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6380 compact, variable length representation of numbers used by the DWARF
6381 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6383 @ifclear no-space-dir
6385 @section @code{.space @var{size} , @var{fill}}
6387 @cindex @code{space} directive
6388 @cindex filling memory
6389 This directive emits @var{size} bytes, each of value @var{fill}. Both
6390 @var{size} and @var{fill} are absolute expressions. If the comma
6391 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6396 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6397 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6398 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6399 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6407 @section @code{.stabd, .stabn, .stabs}
6409 @cindex symbolic debuggers, information for
6410 @cindex @code{stab@var{x}} directives
6411 There are three directives that begin @samp{.stab}.
6412 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6413 The symbols are not entered in the @command{@value{AS}} hash table: they
6414 cannot be referenced elsewhere in the source file.
6415 Up to five fields are required:
6419 This is the symbol's name. It may contain any character except
6420 @samp{\000}, so is more general than ordinary symbol names. Some
6421 debuggers used to code arbitrarily complex structures into symbol names
6425 An absolute expression. The symbol's type is set to the low 8 bits of
6426 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6427 and debuggers choke on silly bit patterns.
6430 An absolute expression. The symbol's ``other'' attribute is set to the
6431 low 8 bits of this expression.
6434 An absolute expression. The symbol's descriptor is set to the low 16
6435 bits of this expression.
6438 An absolute expression which becomes the symbol's value.
6441 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6442 or @code{.stabs} statement, the symbol has probably already been created;
6443 you get a half-formed symbol in your object file. This is
6444 compatible with earlier assemblers!
6447 @cindex @code{stabd} directive
6448 @item .stabd @var{type} , @var{other} , @var{desc}
6450 The ``name'' of the symbol generated is not even an empty string.
6451 It is a null pointer, for compatibility. Older assemblers used a
6452 null pointer so they didn't waste space in object files with empty
6455 The symbol's value is set to the location counter,
6456 relocatably. When your program is linked, the value of this symbol
6457 is the address of the location counter when the @code{.stabd} was
6460 @cindex @code{stabn} directive
6461 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6462 The name of the symbol is set to the empty string @code{""}.
6464 @cindex @code{stabs} directive
6465 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6466 All five fields are specified.
6472 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6473 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6475 @cindex string, copying to object file
6476 @cindex string8, copying to object file
6477 @cindex string16, copying to object file
6478 @cindex string32, copying to object file
6479 @cindex string64, copying to object file
6480 @cindex @code{string} directive
6481 @cindex @code{string8} directive
6482 @cindex @code{string16} directive
6483 @cindex @code{string32} directive
6484 @cindex @code{string64} directive
6486 Copy the characters in @var{str} to the object file. You may specify more than
6487 one string to copy, separated by commas. Unless otherwise specified for a
6488 particular machine, the assembler marks the end of each string with a 0 byte.
6489 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6491 The variants @code{string16}, @code{string32} and @code{string64} differ from
6492 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6493 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6494 are stored in target endianness byte order.
6500 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6501 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6506 @section @code{.struct @var{expression}}
6508 @cindex @code{struct} directive
6509 Switch to the absolute section, and set the section offset to @var{expression},
6510 which must be an absolute expression. You might use this as follows:
6519 This would define the symbol @code{field1} to have the value 0, the symbol
6520 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6521 value 8. Assembly would be left in the absolute section, and you would need to
6522 use a @code{.section} directive of some sort to change to some other section
6523 before further assembly.
6527 @section @code{.subsection @var{name}}
6529 @cindex @code{subsection} directive
6530 @cindex Section Stack
6531 This is one of the ELF section stack manipulation directives. The others are
6532 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6533 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6536 This directive replaces the current subsection with @code{name}. The current
6537 section is not changed. The replaced subsection is put onto the section stack
6538 in place of the then current top of stack subsection.
6543 @section @code{.symver}
6544 @cindex @code{symver} directive
6545 @cindex symbol versioning
6546 @cindex versions of symbols
6547 Use the @code{.symver} directive to bind symbols to specific version nodes
6548 within a source file. This is only supported on ELF platforms, and is
6549 typically used when assembling files to be linked into a shared library.
6550 There are cases where it may make sense to use this in objects to be bound
6551 into an application itself so as to override a versioned symbol from a
6554 For ELF targets, the @code{.symver} directive can be used like this:
6556 .symver @var{name}, @var{name2@@nodename}
6558 If the symbol @var{name} is defined within the file
6559 being assembled, the @code{.symver} directive effectively creates a symbol
6560 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6561 just don't try and create a regular alias is that the @var{@@} character isn't
6562 permitted in symbol names. The @var{name2} part of the name is the actual name
6563 of the symbol by which it will be externally referenced. The name @var{name}
6564 itself is merely a name of convenience that is used so that it is possible to
6565 have definitions for multiple versions of a function within a single source
6566 file, and so that the compiler can unambiguously know which version of a
6567 function is being mentioned. The @var{nodename} portion of the alias should be
6568 the name of a node specified in the version script supplied to the linker when
6569 building a shared library. If you are attempting to override a versioned
6570 symbol from a shared library, then @var{nodename} should correspond to the
6571 nodename of the symbol you are trying to override.
6573 If the symbol @var{name} is not defined within the file being assembled, all
6574 references to @var{name} will be changed to @var{name2@@nodename}. If no
6575 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6578 Another usage of the @code{.symver} directive is:
6580 .symver @var{name}, @var{name2@@@@nodename}
6582 In this case, the symbol @var{name} must exist and be defined within
6583 the file being assembled. It is similar to @var{name2@@nodename}. The
6584 difference is @var{name2@@@@nodename} will also be used to resolve
6585 references to @var{name2} by the linker.
6587 The third usage of the @code{.symver} directive is:
6589 .symver @var{name}, @var{name2@@@@@@nodename}
6591 When @var{name} is not defined within the
6592 file being assembled, it is treated as @var{name2@@nodename}. When
6593 @var{name} is defined within the file being assembled, the symbol
6594 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6599 @section @code{.tag @var{structname}}
6601 @cindex COFF structure debugging
6602 @cindex structure debugging, COFF
6603 @cindex @code{tag} directive
6604 This directive is generated by compilers to include auxiliary debugging
6605 information in the symbol table. It is only permitted inside
6606 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6607 definitions in the symbol table with instances of those structures.
6610 @samp{.tag} is only used when generating COFF format output; when
6611 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6617 @section @code{.text @var{subsection}}
6619 @cindex @code{text} directive
6620 Tells @command{@value{AS}} to assemble the following statements onto the end of
6621 the text subsection numbered @var{subsection}, which is an absolute
6622 expression. If @var{subsection} is omitted, subsection number zero
6626 @section @code{.title "@var{heading}"}
6628 @cindex @code{title} directive
6629 @cindex listing control: title line
6630 Use @var{heading} as the title (second line, immediately after the
6631 source file name and pagenumber) when generating assembly listings.
6633 This directive affects subsequent pages, as well as the current page if
6634 it appears within ten lines of the top of a page.
6638 @section @code{.type}
6640 This directive is used to set the type of a symbol.
6644 @c only print the extra heading if both COFF and ELF are set
6645 @subheading COFF Version
6648 @cindex COFF symbol type
6649 @cindex symbol type, COFF
6650 @cindex @code{type} directive (COFF version)
6651 For COFF targets, this directive is permitted only within
6652 @code{.def}/@code{.endef} pairs. It is used like this:
6658 This records the integer @var{int} as the type attribute of a symbol table
6662 @samp{.type} is associated only with COFF format output; when
6663 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6664 directive but ignores it.
6670 @c only print the extra heading if both COFF and ELF are set
6671 @subheading ELF Version
6674 @cindex ELF symbol type
6675 @cindex symbol type, ELF
6676 @cindex @code{type} directive (ELF version)
6677 For ELF targets, the @code{.type} directive is used like this:
6680 .type @var{name} , @var{type description}
6683 This sets the type of symbol @var{name} to be either a
6684 function symbol or an object symbol. There are five different syntaxes
6685 supported for the @var{type description} field, in order to provide
6686 compatibility with various other assemblers.
6688 Because some of the characters used in these syntaxes (such as @samp{@@} and
6689 @samp{#}) are comment characters for some architectures, some of the syntaxes
6690 below do not work on all architectures. The first variant will be accepted by
6691 the GNU assembler on all architectures so that variant should be used for
6692 maximum portability, if you do not need to assemble your code with other
6695 The syntaxes supported are:
6698 .type <name> STT_<TYPE_IN_UPPER_CASE>
6699 .type <name>,#<type>
6700 .type <name>,@@<type>
6701 .type <name>,%<type>
6702 .type <name>,"<type>"
6705 The types supported are:
6710 Mark the symbol as being a function name.
6713 @itemx gnu_indirect_function
6714 Mark the symbol as an indirect function when evaluated during reloc
6715 processing. (This is only supported on assemblers targeting GNU systems).
6719 Mark the symbol as being a data object.
6723 Mark the symbol as being a thead-local data object.
6727 Mark the symbol as being a common data object.
6731 Does not mark the symbol in any way. It is supported just for completeness.
6733 @item gnu_unique_object
6734 Marks the symbol as being a globally unique data object. The dynamic linker
6735 will make sure that in the entire process there is just one symbol with this
6736 name and type in use. (This is only supported on assemblers targeting GNU
6741 Note: Some targets support extra types in addition to those listed above.
6747 @section @code{.uleb128 @var{expressions}}
6749 @cindex @code{uleb128} directive
6750 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6751 compact, variable length representation of numbers used by the DWARF
6752 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
6756 @section @code{.val @var{addr}}
6758 @cindex @code{val} directive
6759 @cindex COFF value attribute
6760 @cindex value attribute, COFF
6761 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6762 records the address @var{addr} as the value attribute of a symbol table
6766 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6767 configured for @code{b.out}, it accepts this directive but ignores it.
6773 @section @code{.version "@var{string}"}
6775 @cindex @code{version} directive
6776 This directive creates a @code{.note} section and places into it an ELF
6777 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6782 @section @code{.vtable_entry @var{table}, @var{offset}}
6784 @cindex @code{vtable_entry} directive
6785 This directive finds or creates a symbol @code{table} and creates a
6786 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6789 @section @code{.vtable_inherit @var{child}, @var{parent}}
6791 @cindex @code{vtable_inherit} directive
6792 This directive finds the symbol @code{child} and finds or creates the symbol
6793 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6794 parent whose addend is the value of the child symbol. As a special case the
6795 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
6799 @section @code{.warning "@var{string}"}
6800 @cindex warning directive
6801 Similar to the directive @code{.error}
6802 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6805 @section @code{.weak @var{names}}
6807 @cindex @code{weak} directive
6808 This directive sets the weak attribute on the comma separated list of symbol
6809 @code{names}. If the symbols do not already exist, they will be created.
6811 On COFF targets other than PE, weak symbols are a GNU extension. This
6812 directive sets the weak attribute on the comma separated list of symbol
6813 @code{names}. If the symbols do not already exist, they will be created.
6815 On the PE target, weak symbols are supported natively as weak aliases.
6816 When a weak symbol is created that is not an alias, GAS creates an
6817 alternate symbol to hold the default value.
6820 @section @code{.weakref @var{alias}, @var{target}}
6822 @cindex @code{weakref} directive
6823 This directive creates an alias to the target symbol that enables the symbol to
6824 be referenced with weak-symbol semantics, but without actually making it weak.
6825 If direct references or definitions of the symbol are present, then the symbol
6826 will not be weak, but if all references to it are through weak references, the
6827 symbol will be marked as weak in the symbol table.
6829 The effect is equivalent to moving all references to the alias to a separate
6830 assembly source file, renaming the alias to the symbol in it, declaring the
6831 symbol as weak there, and running a reloadable link to merge the object files
6832 resulting from the assembly of the new source file and the old source file that
6833 had the references to the alias removed.
6835 The alias itself never makes to the symbol table, and is entirely handled
6836 within the assembler.
6839 @section @code{.word @var{expressions}}
6841 @cindex @code{word} directive
6842 This directive expects zero or more @var{expressions}, of any section,
6843 separated by commas.
6846 For each expression, @command{@value{AS}} emits a 32-bit number.
6849 For each expression, @command{@value{AS}} emits a 16-bit number.
6854 The size of the number emitted, and its byte order,
6855 depend on what target computer the assembly is for.
6858 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6859 @c happen---32-bit addressability, period; no long/short jumps.
6860 @ifset DIFF-TBL-KLUGE
6861 @cindex difference tables altered
6862 @cindex altered difference tables
6864 @emph{Warning: Special Treatment to support Compilers}
6868 Machines with a 32-bit address space, but that do less than 32-bit
6869 addressing, require the following special treatment. If the machine of
6870 interest to you does 32-bit addressing (or doesn't require it;
6871 @pxref{Machine Dependencies}), you can ignore this issue.
6874 In order to assemble compiler output into something that works,
6875 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6876 Directives of the form @samp{.word sym1-sym2} are often emitted by
6877 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6878 directive of the form @samp{.word sym1-sym2}, and the difference between
6879 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6880 creates a @dfn{secondary jump table}, immediately before the next label.
6881 This secondary jump table is preceded by a short-jump to the
6882 first byte after the secondary table. This short-jump prevents the flow
6883 of control from accidentally falling into the new table. Inside the
6884 table is a long-jump to @code{sym2}. The original @samp{.word}
6885 contains @code{sym1} minus the address of the long-jump to
6888 If there were several occurrences of @samp{.word sym1-sym2} before the
6889 secondary jump table, all of them are adjusted. If there was a
6890 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6891 long-jump to @code{sym4} is included in the secondary jump table,
6892 and the @code{.word} directives are adjusted to contain @code{sym3}
6893 minus the address of the long-jump to @code{sym4}; and so on, for as many
6894 entries in the original jump table as necessary.
6897 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6898 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6899 assembly language programmers.
6902 @c end DIFF-TBL-KLUGE
6905 @section Deprecated Directives
6907 @cindex deprecated directives
6908 @cindex obsolescent directives
6909 One day these directives won't work.
6910 They are included for compatibility with older assemblers.
6917 @node Object Attributes
6918 @chapter Object Attributes
6919 @cindex object attributes
6921 @command{@value{AS}} assembles source files written for a specific architecture
6922 into object files for that architecture. But not all object files are alike.
6923 Many architectures support incompatible variations. For instance, floating
6924 point arguments might be passed in floating point registers if the object file
6925 requires hardware floating point support---or floating point arguments might be
6926 passed in integer registers if the object file supports processors with no
6927 hardware floating point unit. Or, if two objects are built for different
6928 generations of the same architecture, the combination may require the
6929 newer generation at run-time.
6931 This information is useful during and after linking. At link time,
6932 @command{@value{LD}} can warn about incompatible object files. After link
6933 time, tools like @command{gdb} can use it to process the linked file
6936 Compatibility information is recorded as a series of object attributes. Each
6937 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
6938 string, and indicates who sets the meaning of the tag. The tag is an integer,
6939 and indicates what property the attribute describes. The value may be a string
6940 or an integer, and indicates how the property affects this object. Missing
6941 attributes are the same as attributes with a zero value or empty string value.
6943 Object attributes were developed as part of the ABI for the ARM Architecture.
6944 The file format is documented in @cite{ELF for the ARM Architecture}.
6947 * GNU Object Attributes:: @sc{gnu} Object Attributes
6948 * Defining New Object Attributes:: Defining New Object Attributes
6951 @node GNU Object Attributes
6952 @section @sc{gnu} Object Attributes
6954 The @code{.gnu_attribute} directive records an object attribute
6955 with vendor @samp{gnu}.
6957 Except for @samp{Tag_compatibility}, which has both an integer and a string for
6958 its value, @sc{gnu} attributes have a string value if the tag number is odd and
6959 an integer value if the tag number is even. The second bit (@code{@var{tag} &
6960 2} is set for architecture-independent attributes and clear for
6961 architecture-dependent ones.
6963 @subsection Common @sc{gnu} attributes
6965 These attributes are valid on all architectures.
6968 @item Tag_compatibility (32)
6969 The compatibility attribute takes an integer flag value and a vendor name. If
6970 the flag value is 0, the file is compatible with other toolchains. If it is 1,
6971 then the file is only compatible with the named toolchain. If it is greater
6972 than 1, the file can only be processed by other toolchains under some private
6973 arrangement indicated by the flag value and the vendor name.
6976 @subsection MIPS Attributes
6979 @item Tag_GNU_MIPS_ABI_FP (4)
6980 The floating-point ABI used by this object file. The value will be:
6984 0 for files not affected by the floating-point ABI.
6986 1 for files using the hardware floating-point with a standard double-precision
6989 2 for files using the hardware floating-point ABI with a single-precision FPU.
6991 3 for files using the software floating-point ABI.
6993 4 for files using the hardware floating-point ABI with 64-bit wide
6994 double-precision floating-point registers and 32-bit wide general
6999 @subsection PowerPC Attributes
7002 @item Tag_GNU_Power_ABI_FP (4)
7003 The floating-point ABI used by this object file. The value will be:
7007 0 for files not affected by the floating-point ABI.
7009 1 for files using double-precision hardware floating-point ABI.
7011 2 for files using the software floating-point ABI.
7013 3 for files using single-precision hardware floating-point ABI.
7016 @item Tag_GNU_Power_ABI_Vector (8)
7017 The vector ABI used by this object file. The value will be:
7021 0 for files not affected by the vector ABI.
7023 1 for files using general purpose registers to pass vectors.
7025 2 for files using AltiVec registers to pass vectors.
7027 3 for files using SPE registers to pass vectors.
7031 @node Defining New Object Attributes
7032 @section Defining New Object Attributes
7034 If you want to define a new @sc{gnu} object attribute, here are the places you
7035 will need to modify. New attributes should be discussed on the @samp{binutils}
7040 This manual, which is the official register of attributes.
7042 The header for your architecture @file{include/elf}, to define the tag.
7044 The @file{bfd} support file for your architecture, to merge the attribute
7045 and issue any appropriate link warnings.
7047 Test cases in @file{ld/testsuite} for merging and link warnings.
7049 @file{binutils/readelf.c} to display your attribute.
7051 GCC, if you want the compiler to mark the attribute automatically.
7057 @node Machine Dependencies
7058 @chapter Machine Dependent Features
7060 @cindex machine dependencies
7061 The machine instruction sets are (almost by definition) different on
7062 each machine where @command{@value{AS}} runs. Floating point representations
7063 vary as well, and @command{@value{AS}} often supports a few additional
7064 directives or command-line options for compatibility with other
7065 assemblers on a particular platform. Finally, some versions of
7066 @command{@value{AS}} support special pseudo-instructions for branch
7069 This chapter discusses most of these differences, though it does not
7070 include details on any machine's instruction set. For details on that
7071 subject, see the hardware manufacturer's manual.
7075 * AArch64-Dependent:: AArch64 Dependent Features
7078 * Alpha-Dependent:: Alpha Dependent Features
7081 * ARC-Dependent:: ARC Dependent Features
7084 * ARM-Dependent:: ARM Dependent Features
7087 * AVR-Dependent:: AVR Dependent Features
7090 * Blackfin-Dependent:: Blackfin Dependent Features
7093 * CR16-Dependent:: CR16 Dependent Features
7096 * CRIS-Dependent:: CRIS Dependent Features
7099 * D10V-Dependent:: D10V Dependent Features
7102 * D30V-Dependent:: D30V Dependent Features
7105 * Epiphany-Dependent:: EPIPHANY Dependent Features
7108 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7111 * HPPA-Dependent:: HPPA Dependent Features
7114 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
7117 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
7120 * i860-Dependent:: Intel 80860 Dependent Features
7123 * i960-Dependent:: Intel 80960 Dependent Features
7126 * IA-64-Dependent:: Intel IA-64 Dependent Features
7129 * IP2K-Dependent:: IP2K Dependent Features
7132 * LM32-Dependent:: LM32 Dependent Features
7135 * M32C-Dependent:: M32C Dependent Features
7138 * M32R-Dependent:: M32R Dependent Features
7141 * M68K-Dependent:: M680x0 Dependent Features
7144 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7147 * Meta-Dependent :: Meta Dependent Features
7150 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7153 * MIPS-Dependent:: MIPS Dependent Features
7156 * MMIX-Dependent:: MMIX Dependent Features
7159 * MSP430-Dependent:: MSP430 Dependent Features
7162 * NDS32-Dependent:: Andes NDS32 Dependent Features
7165 * NiosII-Dependent:: Altera Nios II Dependent Features
7168 * NS32K-Dependent:: NS32K Dependent Features
7171 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7172 * SH64-Dependent:: SuperH SH64 Dependent Features
7175 * PDP-11-Dependent:: PDP-11 Dependent Features
7178 * PJ-Dependent:: picoJava Dependent Features
7181 * PPC-Dependent:: PowerPC Dependent Features
7184 * RL78-Dependent:: RL78 Dependent Features
7187 * RX-Dependent:: RX Dependent Features
7190 * S/390-Dependent:: IBM S/390 Dependent Features
7193 * SCORE-Dependent:: SCORE Dependent Features
7196 * Sparc-Dependent:: SPARC Dependent Features
7199 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7202 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7205 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7208 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7211 * V850-Dependent:: V850 Dependent Features
7214 * XGATE-Dependent:: XGATE Features
7217 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7220 * Xtensa-Dependent:: Xtensa Dependent Features
7223 * Z80-Dependent:: Z80 Dependent Features
7226 * Z8000-Dependent:: Z8000 Dependent Features
7229 * Vax-Dependent:: VAX Dependent Features
7236 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7237 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7238 @c peculiarity: to preserve cross-references, there must be a node called
7239 @c "Machine Dependencies". Hence the conditional nodenames in each
7240 @c major node below. Node defaulting in makeinfo requires adjacency of
7241 @c node and sectioning commands; hence the repetition of @chapter BLAH
7242 @c in both conditional blocks.
7245 @include c-aarch64.texi
7249 @include c-alpha.texi
7265 @include c-bfin.texi
7269 @include c-cr16.texi
7273 @include c-cris.texi
7278 @node Machine Dependencies
7279 @chapter Machine Dependent Features
7281 The machine instruction sets are different on each Renesas chip family,
7282 and there are also some syntax differences among the families. This
7283 chapter describes the specific @command{@value{AS}} features for each
7287 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7288 * SH-Dependent:: Renesas SH Dependent Features
7295 @include c-d10v.texi
7299 @include c-d30v.texi
7303 @include c-epiphany.texi
7307 @include c-h8300.texi
7311 @include c-hppa.texi
7315 @include c-i370.texi
7319 @include c-i386.texi
7323 @include c-i860.texi
7327 @include c-i960.texi
7331 @include c-ia64.texi
7335 @include c-ip2k.texi
7339 @include c-lm32.texi
7343 @include c-m32c.texi
7347 @include c-m32r.texi
7351 @include c-m68k.texi
7355 @include c-m68hc11.texi
7359 @include c-metag.texi
7363 @include c-microblaze.texi
7367 @include c-mips.texi
7371 @include c-mmix.texi
7375 @include c-msp430.texi
7379 @include c-nds32.texi
7383 @include c-nios2.texi
7387 @include c-ns32k.texi
7391 @include c-pdp11.texi
7403 @include c-rl78.texi
7411 @include c-s390.texi
7415 @include c-score.texi
7420 @include c-sh64.texi
7424 @include c-sparc.texi
7428 @include c-tic54x.texi
7432 @include c-tic6x.texi
7436 @include c-tilegx.texi
7440 @include c-tilepro.texi
7456 @include c-v850.texi
7460 @include c-xgate.texi
7464 @include c-xstormy16.texi
7468 @include c-xtensa.texi
7472 @c reverse effect of @down at top of generic Machine-Dep chapter
7476 @node Reporting Bugs
7477 @chapter Reporting Bugs
7478 @cindex bugs in assembler
7479 @cindex reporting bugs in assembler
7481 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7483 Reporting a bug may help you by bringing a solution to your problem, or it may
7484 not. But in any case the principal function of a bug report is to help the
7485 entire community by making the next version of @command{@value{AS}} work better.
7486 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7488 In order for a bug report to serve its purpose, you must include the
7489 information that enables us to fix the bug.
7492 * Bug Criteria:: Have you found a bug?
7493 * Bug Reporting:: How to report bugs
7497 @section Have You Found a Bug?
7498 @cindex bug criteria
7500 If you are not sure whether you have found a bug, here are some guidelines:
7503 @cindex fatal signal
7504 @cindex assembler crash
7505 @cindex crash of assembler
7507 If the assembler gets a fatal signal, for any input whatever, that is a
7508 @command{@value{AS}} bug. Reliable assemblers never crash.
7510 @cindex error on valid input
7512 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7514 @cindex invalid input
7516 If @command{@value{AS}} does not produce an error message for invalid input, that
7517 is a bug. However, you should note that your idea of ``invalid input'' might
7518 be our idea of ``an extension'' or ``support for traditional practice''.
7521 If you are an experienced user of assemblers, your suggestions for improvement
7522 of @command{@value{AS}} are welcome in any case.
7526 @section How to Report Bugs
7528 @cindex assembler bugs, reporting
7530 A number of companies and individuals offer support for @sc{gnu} products. If
7531 you obtained @command{@value{AS}} from a support organization, we recommend you
7532 contact that organization first.
7534 You can find contact information for many support companies and
7535 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7539 In any event, we also recommend that you send bug reports for @command{@value{AS}}
7543 The fundamental principle of reporting bugs usefully is this:
7544 @strong{report all the facts}. If you are not sure whether to state a
7545 fact or leave it out, state it!
7547 Often people omit facts because they think they know what causes the problem
7548 and assume that some details do not matter. Thus, you might assume that the
7549 name of a symbol you use in an example does not matter. Well, probably it does
7550 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
7551 happens to fetch from the location where that name is stored in memory;
7552 perhaps, if the name were different, the contents of that location would fool
7553 the assembler into doing the right thing despite the bug. Play it safe and
7554 give a specific, complete example. That is the easiest thing for you to do,
7555 and the most helpful.
7557 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
7558 it is new to us. Therefore, always write your bug reports on the assumption
7559 that the bug has not been reported previously.
7561 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7562 bell?'' This cannot help us fix a bug, so it is basically useless. We
7563 respond by asking for enough details to enable us to investigate.
7564 You might as well expedite matters by sending them to begin with.
7566 To enable us to fix the bug, you should include all these things:
7570 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
7571 it with the @samp{--version} argument.
7573 Without this, we will not know whether there is any point in looking for
7574 the bug in the current version of @command{@value{AS}}.
7577 Any patches you may have applied to the @command{@value{AS}} source.
7580 The type of machine you are using, and the operating system name and
7584 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
7588 The command arguments you gave the assembler to assemble your example and
7589 observe the bug. To guarantee you will not omit something important, list them
7590 all. A copy of the Makefile (or the output from make) is sufficient.
7592 If we were to try to guess the arguments, we would probably guess wrong
7593 and then we might not encounter the bug.
7596 A complete input file that will reproduce the bug. If the bug is observed when
7597 the assembler is invoked via a compiler, send the assembler source, not the
7598 high level language source. Most compilers will produce the assembler source
7599 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
7600 the options @samp{-v --save-temps}; this will save the assembler source in a
7601 file with an extension of @file{.s}, and also show you exactly how
7602 @command{@value{AS}} is being run.
7605 A description of what behavior you observe that you believe is
7606 incorrect. For example, ``It gets a fatal signal.''
7608 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
7609 will certainly notice it. But if the bug is incorrect output, we might not
7610 notice unless it is glaringly wrong. You might as well not give us a chance to
7613 Even if the problem you experience is a fatal signal, you should still say so
7614 explicitly. Suppose something strange is going on, such as, your copy of
7615 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
7616 library on your system. (This has happened!) Your copy might crash and ours
7617 would not. If you told us to expect a crash, then when ours fails to crash, we
7618 would know that the bug was not happening for us. If you had not told us to
7619 expect a crash, then we would not be able to draw any conclusion from our
7623 If you wish to suggest changes to the @command{@value{AS}} source, send us context
7624 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
7625 option. Always send diffs from the old file to the new file. If you even
7626 discuss something in the @command{@value{AS}} source, refer to it by context, not
7629 The line numbers in our development sources will not match those in your
7630 sources. Your line numbers would convey no useful information to us.
7633 Here are some things that are not necessary:
7637 A description of the envelope of the bug.
7639 Often people who encounter a bug spend a lot of time investigating
7640 which changes to the input file will make the bug go away and which
7641 changes will not affect it.
7643 This is often time consuming and not very useful, because the way we
7644 will find the bug is by running a single example under the debugger
7645 with breakpoints, not by pure deduction from a series of examples.
7646 We recommend that you save your time for something else.
7648 Of course, if you can find a simpler example to report @emph{instead}
7649 of the original one, that is a convenience for us. Errors in the
7650 output will be easier to spot, running under the debugger will take
7651 less time, and so on.
7653 However, simplification is not vital; if you do not want to do this,
7654 report the bug anyway and send us the entire test case you used.
7657 A patch for the bug.
7659 A patch for the bug does help us if it is a good one. But do not omit
7660 the necessary information, such as the test case, on the assumption that
7661 a patch is all we need. We might see problems with your patch and decide
7662 to fix the problem another way, or we might not understand it at all.
7664 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
7665 construct an example that will make the program follow a certain path through
7666 the code. If you do not send us the example, we will not be able to construct
7667 one, so we will not be able to verify that the bug is fixed.
7669 And if we cannot understand what bug you are trying to fix, or why your
7670 patch should be an improvement, we will not install it. A test case will
7671 help us to understand.
7674 A guess about what the bug is or what it depends on.
7676 Such guesses are usually wrong. Even we cannot guess right about such
7677 things without first using the debugger to find the facts.
7680 @node Acknowledgements
7681 @chapter Acknowledgements
7683 If you have contributed to GAS and your name isn't listed here,
7684 it is not meant as a slight. We just don't know about it. Send mail to the
7685 maintainer, and we'll correct the situation. Currently
7687 the maintainer is Nick Clifton (email address @code{nickc@@redhat.com}).
7689 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
7692 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
7693 information and the 68k series machines, most of the preprocessing pass, and
7694 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
7696 K. Richard Pixley maintained GAS for a while, adding various enhancements and
7697 many bug fixes, including merging support for several processors, breaking GAS
7698 up to handle multiple object file format back ends (including heavy rewrite,
7699 testing, an integration of the coff and b.out back ends), adding configuration
7700 including heavy testing and verification of cross assemblers and file splits
7701 and renaming, converted GAS to strictly ANSI C including full prototypes, added
7702 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
7703 port (including considerable amounts of reverse engineering), a SPARC opcode
7704 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
7705 assertions and made them work, much other reorganization, cleanup, and lint.
7707 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
7708 in format-specific I/O modules.
7710 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
7711 has done much work with it since.
7713 The Intel 80386 machine description was written by Eliot Dresselhaus.
7715 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
7717 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
7718 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
7720 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
7721 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
7722 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
7723 support a.out format.
7725 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
7726 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
7727 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
7728 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
7731 John Gilmore built the AMD 29000 support, added @code{.include} support, and
7732 simplified the configuration of which versions accept which directives. He
7733 updated the 68k machine description so that Motorola's opcodes always produced
7734 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
7735 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
7736 cross-compilation support, and one bug in relaxation that took a week and
7737 required the proverbial one-bit fix.
7739 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
7740 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
7741 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
7742 PowerPC assembler, and made a few other minor patches.
7744 Steve Chamberlain made GAS able to generate listings.
7746 Hewlett-Packard contributed support for the HP9000/300.
7748 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
7749 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
7750 formats). This work was supported by both the Center for Software Science at
7751 the University of Utah and Cygnus Support.
7753 Support for ELF format files has been worked on by Mark Eichin of Cygnus
7754 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
7755 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
7756 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
7757 and some initial 64-bit support).
7759 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
7761 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
7762 support for openVMS/Alpha.
7764 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
7767 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
7768 Inc.@: added support for Xtensa processors.
7770 Several engineers at Cygnus Support have also provided many small bug fixes and
7771 configuration enhancements.
7773 Jon Beniston added support for the Lattice Mico32 architecture.
7775 Many others have contributed large or small bugfixes and enhancements. If
7776 you have contributed significant work and are not mentioned on this list, and
7777 want to be, let us know. Some of the history has been lost; we are not
7778 intentionally leaving anyone out.
7780 @node GNU Free Documentation License
7781 @appendix GNU Free Documentation License
7785 @unnumbered AS Index