1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
3 @c 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012
4 @c Free Software Foundation, Inc.
5 @c UPDATE!! On future updates--
6 @c (1) check for new machine-dep cmdline options in
7 @c md_parse_option definitions in config/tc-*.c
8 @c (2) for platform-specific directives, examine md_pseudo_op
10 @c (3) for object-format specific directives, examine obj_pseudo_op
12 @c (4) portable directives in potable[] in read.c
16 @macro gcctabopt{body}
19 @c defaults, config file may override:
24 @include asconfig.texi
29 @c common OR combinations of conditions
55 @set abnormal-separator
59 @settitle Using @value{AS}
62 @settitle Using @value{AS} (@value{TARGET})
64 @setchapternewpage odd
69 @c WARE! Some of the machine-dependent sections contain tables of machine
70 @c instructions. Except in multi-column format, these tables look silly.
71 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
72 @c the multi-col format is faked within @example sections.
74 @c Again unfortunately, the natural size that fits on a page, for these tables,
75 @c is different depending on whether or not smallbook is turned on.
76 @c This matters, because of order: text flow switches columns at each page
79 @c The format faked in this source works reasonably well for smallbook,
80 @c not well for the default large-page format. This manual expects that if you
81 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
82 @c tables in question. You can turn on one without the other at your
83 @c discretion, of course.
86 @c the insn tables look just as silly in info files regardless of smallbook,
87 @c might as well show 'em anyways.
91 @dircategory Software development
93 * As: (as). The GNU assembler.
94 * Gas: (as). The GNU assembler.
102 This file documents the GNU Assembler "@value{AS}".
104 @c man begin COPYRIGHT
105 Copyright @copyright{} 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
106 2000, 2001, 2002, 2006, 2007, 2008, 2009, 2010, 2011 Free Software Foundation,
109 Permission is granted to copy, distribute and/or modify this document
110 under the terms of the GNU Free Documentation License, Version 1.3
111 or any later version published by the Free Software Foundation;
112 with no Invariant Sections, with no Front-Cover Texts, and with no
113 Back-Cover Texts. A copy of the license is included in the
114 section entitled ``GNU Free Documentation License''.
120 @title Using @value{AS}
121 @subtitle The @sc{gnu} Assembler
123 @subtitle for the @value{TARGET} family
125 @ifset VERSION_PACKAGE
127 @subtitle @value{VERSION_PACKAGE}
130 @subtitle Version @value{VERSION}
133 The Free Software Foundation Inc.@: thanks The Nice Computer
134 Company of Australia for loaning Dean Elsner to write the
135 first (Vax) version of @command{as} for Project @sc{gnu}.
136 The proprietors, management and staff of TNCCA thank FSF for
137 distracting the boss while they got some work
140 @author Dean Elsner, Jay Fenlason & friends
144 \hfill {\it Using {\tt @value{AS}}}\par
145 \hfill Edited by Cygnus Support\par
147 %"boxit" macro for figures:
148 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
149 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
150 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
151 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
152 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
155 @vskip 0pt plus 1filll
156 Copyright @copyright{} 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
157 2000, 2001, 2002, 2006, 2007, 2008, 2009, 2010, 2011 Free Software Foundation,
160 Permission is granted to copy, distribute and/or modify this document
161 under the terms of the GNU Free Documentation License, Version 1.3
162 or any later version published by the Free Software Foundation;
163 with no Invariant Sections, with no Front-Cover Texts, and with no
164 Back-Cover Texts. A copy of the license is included in the
165 section entitled ``GNU Free Documentation License''.
172 @top Using @value{AS}
174 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}}
175 @ifset VERSION_PACKAGE
176 @value{VERSION_PACKAGE}
178 version @value{VERSION}.
180 This version of the file describes @command{@value{AS}} configured to generate
181 code for @value{TARGET} architectures.
184 This document is distributed under the terms of the GNU Free
185 Documentation License. A copy of the license is included in the
186 section entitled ``GNU Free Documentation License''.
189 * Overview:: Overview
190 * Invoking:: Command-Line Options
192 * Sections:: Sections and Relocation
194 * Expressions:: Expressions
195 * Pseudo Ops:: Assembler Directives
197 * Object Attributes:: Object Attributes
199 * Machine Dependencies:: Machine Dependent Features
200 * Reporting Bugs:: Reporting Bugs
201 * Acknowledgements:: Who Did What
202 * GNU Free Documentation License:: GNU Free Documentation License
203 * AS Index:: AS Index
210 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
212 This version of the manual describes @command{@value{AS}} configured to generate
213 code for @value{TARGET} architectures.
217 @cindex invocation summary
218 @cindex option summary
219 @cindex summary of options
220 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
221 see @ref{Invoking,,Command-Line Options}.
223 @c man title AS the portable GNU assembler.
227 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
231 @c We don't use deffn and friends for the following because they seem
232 @c to be limited to one line for the header.
234 @c man begin SYNOPSIS
235 @value{AS} [@b{-a}[@b{cdghlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
236 [@b{--compress-debug-sections}] [@b{--nocompress-debug-sections}]
237 [@b{--debug-prefix-map} @var{old}=@var{new}]
238 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
239 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
240 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
241 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
242 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
243 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
244 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
245 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
246 [@b{--size-check=[error|warning]}]
247 [@b{--target-help}] [@var{target-options}]
248 [@b{--}|@var{files} @dots{}]
250 @c Target dependent options are listed below. Keep the list sorted.
251 @c Add an empty line for separation.
254 @emph{Target AArch64 options:}
259 @emph{Target Alpha options:}
261 [@b{-mdebug} | @b{-no-mdebug}]
262 [@b{-replace} | @b{-noreplace}]
263 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
264 [@b{-F}] [@b{-32addr}]
268 @emph{Target ARC options:}
274 @emph{Target ARM options:}
275 @c Don't document the deprecated options
276 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
277 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
278 [@b{-mfpu}=@var{floating-point-format}]
279 [@b{-mfloat-abi}=@var{abi}]
280 [@b{-meabi}=@var{ver}]
283 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
284 @b{-mapcs-reentrant}]
285 [@b{-mthumb-interwork}] [@b{-k}]
289 @emph{Target Blackfin options:}
290 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
297 @emph{Target CRIS options:}
298 [@b{--underscore} | @b{--no-underscore}]
300 [@b{--emulation=criself} | @b{--emulation=crisaout}]
301 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
302 @c Deprecated -- deliberately not documented.
307 @emph{Target D10V options:}
312 @emph{Target D30V options:}
313 [@b{-O}|@b{-n}|@b{-N}]
317 @emph{Target EPIPHANY options:}
318 [@b{-mepiphany}|@b{-mepiphany16}]
322 @emph{Target H8/300 options:}
326 @c HPPA has no machine-dependent assembler options (yet).
330 @emph{Target i386 options:}
331 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
332 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
336 @emph{Target i960 options:}
337 @c see md_parse_option in tc-i960.c
338 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
340 [@b{-b}] [@b{-no-relax}]
344 @emph{Target IA-64 options:}
345 [@b{-mconstant-gp}|@b{-mauto-pic}]
346 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
348 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
349 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
350 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
351 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
355 @emph{Target IP2K options:}
356 [@b{-mip2022}|@b{-mip2022ext}]
360 @emph{Target M32C options:}
361 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
365 @emph{Target M32R options:}
366 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
371 @emph{Target M680X0 options:}
372 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
376 @emph{Target M68HC11 options:}
377 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}|@b{-mm9s12x}|@b{-mm9s12xg}]
378 [@b{-mshort}|@b{-mlong}]
379 [@b{-mshort-double}|@b{-mlong-double}]
380 [@b{--force-long-branches}] [@b{--short-branches}]
381 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
382 [@b{--print-opcodes}] [@b{--generate-example}]
386 @emph{Target MCORE options:}
387 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
388 [@b{-mcpu=[210|340]}]
391 @emph{Target MICROBLAZE options:}
392 @c MicroBlaze has no machine-dependent assembler options.
396 @emph{Target MIPS options:}
397 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
398 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
399 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
400 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
401 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
402 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
403 [@b{-mips64}] [@b{-mips64r2}]
404 [@b{-construct-floats}] [@b{-no-construct-floats}]
405 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
406 [@b{-mips16}] [@b{-no-mips16}]
407 [@b{-mmicromips}] [@b{-mno-micromips}]
408 [@b{-msmartmips}] [@b{-mno-smartmips}]
409 [@b{-mips3d}] [@b{-no-mips3d}]
410 [@b{-mdmx}] [@b{-no-mdmx}]
411 [@b{-mdsp}] [@b{-mno-dsp}]
412 [@b{-mdspr2}] [@b{-mno-dspr2}]
413 [@b{-mmt}] [@b{-mno-mt}]
414 [@b{-mmcu}] [@b{-mno-mcu}]
415 [@b{-mfix7000}] [@b{-mno-fix7000}]
416 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
417 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
418 [@b{-mdebug}] [@b{-no-mdebug}]
419 [@b{-mpdr}] [@b{-mno-pdr}]
423 @emph{Target MMIX options:}
424 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
425 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
426 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
427 [@b{--linker-allocated-gregs}]
431 @emph{Target PDP11 options:}
432 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
433 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
434 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
438 @emph{Target picoJava options:}
443 @emph{Target PowerPC options:}
445 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
446 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
447 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
448 @b{-mbooke}|@b{-mpower4}|@b{-mpr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
449 @b{-mpower7}|@b{-mpw7}|@b{-ma2}|@b{-mcell}|@b{-mspe}|@b{-mtitan}|@b{-me300}|@b{-mvle}|@b{-mcom}]
450 [@b{-many}] [@b{-maltivec}|@b{-mvsx}]
451 [@b{-mregnames}|@b{-mno-regnames}]
452 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
453 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
454 [@b{-msolaris}|@b{-mno-solaris}]
455 [@b{-nops=@var{count}}]
459 @emph{Target RX options:}
460 [@b{-mlittle-endian}|@b{-mbig-endian}]
461 [@b{-m32bit-ints}|@b{-m16bit-ints}]
462 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
466 @emph{Target s390 options:}
467 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
468 [@b{-mregnames}|@b{-mno-regnames}]
469 [@b{-mwarn-areg-zero}]
473 @emph{Target SCORE options:}
474 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
475 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
476 [@b{-march=score7}][@b{-march=score3}]
477 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
481 @emph{Target SPARC options:}
482 @c The order here is important. See c-sparc.texi.
483 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
484 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
485 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
490 @emph{Target TIC54X options:}
491 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
492 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
497 @emph{Target TIC6X options:}
498 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
499 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
500 [@b{-mpic}|@b{-mno-pic}]
504 @emph{Target TILE-Gx options:}
505 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
508 @c TILEPro has no machine-dependent assembler options
513 @emph{Target Xtensa options:}
514 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
515 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
516 [@b{--[no-]transform}]
517 [@b{--rename-section} @var{oldname}=@var{newname}]
522 @emph{Target Z80 options:}
523 [@b{-z80}] [@b{-r800}]
524 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
525 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
526 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
527 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
528 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
529 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
533 @c Z8000 has no machine-dependent assembler options
542 @include at-file.texi
545 Turn on listings, in any of a variety of ways:
549 omit false conditionals
552 omit debugging directives
555 include general information, like @value{AS} version and options passed
558 include high-level source
564 include macro expansions
567 omit forms processing
573 set the name of the listing file
576 You may combine these options; for example, use @samp{-aln} for assembly
577 listing without forms processing. The @samp{=file} option, if used, must be
578 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
581 Begin in alternate macro mode.
583 @xref{Altmacro,,@code{.altmacro}}.
586 @item --compress-debug-sections
587 Compress DWARF debug sections using zlib. The debug sections are renamed
588 to begin with @samp{.zdebug}, and the resulting object file may not be
589 compatible with older linkers and object file utilities.
591 @item --nocompress-debug-sections
592 Do not compress DWARF debug sections. This is the default.
595 Ignored. This option is accepted for script compatibility with calls to
598 @item --debug-prefix-map @var{old}=@var{new}
599 When assembling files in directory @file{@var{old}}, record debugging
600 information describing them as in @file{@var{new}} instead.
602 @item --defsym @var{sym}=@var{value}
603 Define the symbol @var{sym} to be @var{value} before assembling the input file.
604 @var{value} must be an integer constant. As in C, a leading @samp{0x}
605 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
606 value. The value of the symbol can be overridden inside a source file via the
607 use of a @code{.set} pseudo-op.
610 ``fast''---skip whitespace and comment preprocessing (assume source is
615 Generate debugging information for each assembler source line using whichever
616 debug format is preferred by the target. This currently means either STABS,
620 Generate stabs debugging information for each assembler line. This
621 may help debugging assembler code, if the debugger can handle it.
624 Generate stabs debugging information for each assembler line, with GNU
625 extensions that probably only gdb can handle, and that could make other
626 debuggers crash or refuse to read your program. This
627 may help debugging assembler code. Currently the only GNU extension is
628 the location of the current working directory at assembling time.
631 Generate DWARF2 debugging information for each assembler line. This
632 may help debugging assembler code, if the debugger can handle it. Note---this
633 option is only supported by some targets, not all of them.
635 @item --size-check=error
636 @itemx --size-check=warning
637 Issue an error or warning for invalid ELF .size directive.
640 Print a summary of the command line options and exit.
643 Print a summary of all target specific options and exit.
646 Add directory @var{dir} to the search list for @code{.include} directives.
649 Don't warn about signed overflow.
652 @ifclear DIFF-TBL-KLUGE
653 This option is accepted but has no effect on the @value{TARGET} family.
655 @ifset DIFF-TBL-KLUGE
656 Issue warnings when difference tables altered for long displacements.
661 Keep (in the symbol table) local symbols. These symbols start with
662 system-specific local label prefixes, typically @samp{.L} for ELF systems
663 or @samp{L} for traditional a.out systems.
668 @item --listing-lhs-width=@var{number}
669 Set the maximum width, in words, of the output data column for an assembler
670 listing to @var{number}.
672 @item --listing-lhs-width2=@var{number}
673 Set the maximum width, in words, of the output data column for continuation
674 lines in an assembler listing to @var{number}.
676 @item --listing-rhs-width=@var{number}
677 Set the maximum width of an input source line, as displayed in a listing, to
680 @item --listing-cont-lines=@var{number}
681 Set the maximum number of lines printed in a listing for a single line of input
684 @item -o @var{objfile}
685 Name the object-file output from @command{@value{AS}} @var{objfile}.
688 Fold the data section into the text section.
690 @kindex --hash-size=@var{number}
691 Set the default size of GAS's hash tables to a prime number close to
692 @var{number}. Increasing this value can reduce the length of time it takes the
693 assembler to perform its tasks, at the expense of increasing the assembler's
694 memory requirements. Similarly reducing this value can reduce the memory
695 requirements at the expense of speed.
697 @item --reduce-memory-overheads
698 This option reduces GAS's memory requirements, at the expense of making the
699 assembly processes slower. Currently this switch is a synonym for
700 @samp{--hash-size=4051}, but in the future it may have other effects as well.
703 Print the maximum space (in bytes) and total time (in seconds) used by
706 @item --strip-local-absolute
707 Remove local absolute symbols from the outgoing symbol table.
711 Print the @command{as} version.
714 Print the @command{as} version and exit.
718 Suppress warning messages.
720 @item --fatal-warnings
721 Treat warnings as errors.
724 Don't suppress warning messages or treat them as errors.
733 Generate an object file even after errors.
735 @item -- | @var{files} @dots{}
736 Standard input, or source files to assemble.
744 @xref{AArch64 Options}, for the options available when @value{AS} is configured
745 for the 64-bit mode of the ARM Architecture (AArch64).
750 The following options are available when @value{AS} is configured for the
751 64-bit mode of the ARM Architecture (AArch64).
754 @include c-aarch64.texi
755 @c ended inside the included file
763 @xref{Alpha Options}, for the options available when @value{AS} is configured
764 for an Alpha processor.
769 The following options are available when @value{AS} is configured for an Alpha
773 @include c-alpha.texi
774 @c ended inside the included file
781 The following options are available when @value{AS} is configured for
786 This option selects the core processor variant.
788 Select either big-endian (-EB) or little-endian (-EL) output.
793 The following options are available when @value{AS} is configured for the ARM
797 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
798 Specify which ARM processor variant is the target.
799 @item -march=@var{architecture}[+@var{extension}@dots{}]
800 Specify which ARM architecture variant is used by the target.
801 @item -mfpu=@var{floating-point-format}
802 Select which Floating Point architecture is the target.
803 @item -mfloat-abi=@var{abi}
804 Select which floating point ABI is in use.
806 Enable Thumb only instruction decoding.
807 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
808 Select which procedure calling convention is in use.
810 Select either big-endian (-EB) or little-endian (-EL) output.
811 @item -mthumb-interwork
812 Specify that the code has been generated with interworking between Thumb and
815 Specify that PIC code has been generated.
823 @xref{Blackfin Options}, for the options available when @value{AS} is
824 configured for the Blackfin processor family.
829 The following options are available when @value{AS} is configured for
830 the Blackfin processor family.
834 @c ended inside the included file
841 See the info pages for documentation of the CRIS-specific options.
845 The following options are available when @value{AS} is configured for
848 @cindex D10V optimization
849 @cindex optimization, D10V
851 Optimize output by parallelizing instructions.
856 The following options are available when @value{AS} is configured for a D30V
859 @cindex D30V optimization
860 @cindex optimization, D30V
862 Optimize output by parallelizing instructions.
866 Warn when nops are generated.
868 @cindex D30V nops after 32-bit multiply
870 Warn when a nop after a 32-bit multiply instruction is generated.
876 The following options are available when @value{AS} is configured for the
877 Adapteva EPIPHANY series.
880 @xref{Epiphany Options}, for the options available when @value{AS} is
881 configured for an Epiphany processor.
886 The following options are available when @value{AS} is configured for
887 an Epiphany processor.
890 @include c-epiphany.texi
891 @c ended inside the included file
899 @xref{i386-Options}, for the options available when @value{AS} is
900 configured for an i386 processor.
905 The following options are available when @value{AS} is configured for
910 @c ended inside the included file
917 The following options are available when @value{AS} is configured for the
918 Intel 80960 processor.
921 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
922 Specify which variant of the 960 architecture is the target.
925 Add code to collect statistics about branches taken.
928 Do not alter compare-and-branch instructions for long displacements;
935 The following options are available when @value{AS} is configured for the
941 Specifies that the extended IP2022 instructions are allowed.
944 Restores the default behaviour, which restricts the permitted instructions to
945 just the basic IP2022 ones.
951 The following options are available when @value{AS} is configured for the
952 Renesas M32C and M16C processors.
957 Assemble M32C instructions.
960 Assemble M16C instructions (the default).
963 Enable support for link-time relaxations.
966 Support H'00 style hex constants in addition to 0x00 style.
972 The following options are available when @value{AS} is configured for the
973 Renesas M32R (formerly Mitsubishi M32R) series.
978 Specify which processor in the M32R family is the target. The default
979 is normally the M32R, but this option changes it to the M32RX.
981 @item --warn-explicit-parallel-conflicts or --Wp
982 Produce warning messages when questionable parallel constructs are
985 @item --no-warn-explicit-parallel-conflicts or --Wnp
986 Do not produce warning messages when questionable parallel constructs are
993 The following options are available when @value{AS} is configured for the
994 Motorola 68000 series.
999 Shorten references to undefined symbols, to one word instead of two.
1001 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1002 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1003 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1004 Specify what processor in the 68000 family is the target. The default
1005 is normally the 68020, but this can be changed at configuration time.
1007 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1008 The target machine does (or does not) have a floating-point coprocessor.
1009 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1010 the basic 68000 is not compatible with the 68881, a combination of the
1011 two can be specified, since it's possible to do emulation of the
1012 coprocessor instructions with the main processor.
1014 @item -m68851 | -mno-68851
1015 The target machine does (or does not) have a memory-management
1016 unit coprocessor. The default is to assume an MMU for 68020 and up.
1023 For details about the PDP-11 machine dependent features options,
1024 see @ref{PDP-11-Options}.
1027 @item -mpic | -mno-pic
1028 Generate position-independent (or position-dependent) code. The
1029 default is @option{-mpic}.
1032 @itemx -mall-extensions
1033 Enable all instruction set extensions. This is the default.
1035 @item -mno-extensions
1036 Disable all instruction set extensions.
1038 @item -m@var{extension} | -mno-@var{extension}
1039 Enable (or disable) a particular instruction set extension.
1042 Enable the instruction set extensions supported by a particular CPU, and
1043 disable all other extensions.
1045 @item -m@var{machine}
1046 Enable the instruction set extensions supported by a particular machine
1047 model, and disable all other extensions.
1053 The following options are available when @value{AS} is configured for
1054 a picoJava processor.
1058 @cindex PJ endianness
1059 @cindex endianness, PJ
1060 @cindex big endian output, PJ
1062 Generate ``big endian'' format output.
1064 @cindex little endian output, PJ
1066 Generate ``little endian'' format output.
1072 The following options are available when @value{AS} is configured for the
1073 Motorola 68HC11 or 68HC12 series.
1077 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1078 Specify what processor is the target. The default is
1079 defined by the configuration option when building the assembler.
1081 @item --xgate-ramoffset
1082 Instruct the linker to offset RAM addresses from S12X address space into
1083 XGATE address space.
1086 Specify to use the 16-bit integer ABI.
1089 Specify to use the 32-bit integer ABI.
1091 @item -mshort-double
1092 Specify to use the 32-bit double ABI.
1095 Specify to use the 64-bit double ABI.
1097 @item --force-long-branches
1098 Relative branches are turned into absolute ones. This concerns
1099 conditional branches, unconditional branches and branches to a
1102 @item -S | --short-branches
1103 Do not turn relative branches into absolute ones
1104 when the offset is out of range.
1106 @item --strict-direct-mode
1107 Do not turn the direct addressing mode into extended addressing mode
1108 when the instruction does not support direct addressing mode.
1110 @item --print-insn-syntax
1111 Print the syntax of instruction in case of error.
1113 @item --print-opcodes
1114 Print the list of instructions with syntax and then exit.
1116 @item --generate-example
1117 Print an example of instruction for each possible instruction and then exit.
1118 This option is only useful for testing @command{@value{AS}}.
1124 The following options are available when @command{@value{AS}} is configured
1125 for the SPARC architecture:
1128 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1129 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1130 Explicitly select a variant of the SPARC architecture.
1132 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1133 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1135 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1136 UltraSPARC extensions.
1138 @item -xarch=v8plus | -xarch=v8plusa
1139 For compatibility with the Solaris v9 assembler. These options are
1140 equivalent to -Av8plus and -Av8plusa, respectively.
1143 Warn when the assembler switches to another architecture.
1148 The following options are available when @value{AS} is configured for the 'c54x
1153 Enable extended addressing mode. All addresses and relocations will assume
1154 extended addressing (usually 23 bits).
1155 @item -mcpu=@var{CPU_VERSION}
1156 Sets the CPU version being compiled for.
1157 @item -merrors-to-file @var{FILENAME}
1158 Redirect error output to a file, for broken systems which don't support such
1159 behaviour in the shell.
1164 The following options are available when @value{AS} is configured for
1165 a @sc{mips} processor.
1169 This option sets the largest size of an object that can be referenced
1170 implicitly with the @code{gp} register. It is only accepted for targets that
1171 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1173 @cindex MIPS endianness
1174 @cindex endianness, MIPS
1175 @cindex big endian output, MIPS
1177 Generate ``big endian'' format output.
1179 @cindex little endian output, MIPS
1181 Generate ``little endian'' format output.
1193 Generate code for a particular @sc{mips} Instruction Set Architecture level.
1194 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1195 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1196 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1197 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
1199 correspond to generic
1200 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
1201 and @samp{MIPS64 Release 2}
1202 ISA processors, respectively.
1204 @item -march=@var{CPU}
1205 Generate code for a particular @sc{mips} cpu.
1207 @item -mtune=@var{cpu}
1208 Schedule and tune for a particular @sc{mips} cpu.
1212 Cause nops to be inserted if the read of the destination register
1213 of an mfhi or mflo instruction occurs in the following two instructions.
1217 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1218 section instead of the standard ELF .stabs sections.
1222 Control generation of @code{.pdr} sections.
1226 The register sizes are normally inferred from the ISA and ABI, but these
1227 flags force a certain group of registers to be treated as 32 bits wide at
1228 all times. @samp{-mgp32} controls the size of general-purpose registers
1229 and @samp{-mfp32} controls the size of floating-point registers.
1233 Generate code for the MIPS 16 processor. This is equivalent to putting
1234 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1235 turns off this option.
1238 @itemx -mno-micromips
1239 Generate code for the microMIPS processor. This is equivalent to putting
1240 @code{.set micromips} at the start of the assembly file. @samp{-mno-micromips}
1241 turns off this option. This is equivalent to putting @code{.set nomicromips}
1242 at the start of the assembly file.
1245 @itemx -mno-smartmips
1246 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1247 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1248 @samp{-mno-smartmips} turns off this option.
1252 Generate code for the MIPS-3D Application Specific Extension.
1253 This tells the assembler to accept MIPS-3D instructions.
1254 @samp{-no-mips3d} turns off this option.
1258 Generate code for the MDMX Application Specific Extension.
1259 This tells the assembler to accept MDMX instructions.
1260 @samp{-no-mdmx} turns off this option.
1264 Generate code for the DSP Release 1 Application Specific Extension.
1265 This tells the assembler to accept DSP Release 1 instructions.
1266 @samp{-mno-dsp} turns off this option.
1270 Generate code for the DSP Release 2 Application Specific Extension.
1271 This option implies -mdsp.
1272 This tells the assembler to accept DSP Release 2 instructions.
1273 @samp{-mno-dspr2} turns off this option.
1277 Generate code for the MT Application Specific Extension.
1278 This tells the assembler to accept MT instructions.
1279 @samp{-mno-mt} turns off this option.
1283 Generate code for the MCU Application Specific Extension.
1284 This tells the assembler to accept MCU instructions.
1285 @samp{-mno-mcu} turns off this option.
1287 @item --construct-floats
1288 @itemx --no-construct-floats
1289 The @samp{--no-construct-floats} option disables the construction of
1290 double width floating point constants by loading the two halves of the
1291 value into the two single width floating point registers that make up
1292 the double width register. By default @samp{--construct-floats} is
1293 selected, allowing construction of these floating point constants.
1296 @item --emulation=@var{name}
1297 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1298 for some other target, in all respects, including output format (choosing
1299 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1300 debugging information or store symbol table information, and default
1301 endianness. The available configuration names are: @samp{mipsecoff},
1302 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1303 @samp{mipsbelf}. The first two do not alter the default endianness from that
1304 of the primary target for which the assembler was configured; the others change
1305 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1306 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1307 selection in any case.
1309 This option is currently supported only when the primary target
1310 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1311 Furthermore, the primary target or others specified with
1312 @samp{--enable-targets=@dots{}} at configuration time must include support for
1313 the other format, if both are to be available. For example, the Irix 5
1314 configuration includes support for both.
1316 Eventually, this option will support more configurations, with more
1317 fine-grained control over the assembler's behavior, and will be supported for
1321 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1328 Control how to deal with multiplication overflow and division by zero.
1329 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1330 (and only work for Instruction Set Architecture level 2 and higher);
1331 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1335 When this option is used, @command{@value{AS}} will issue a warning every
1336 time it generates a nop instruction from a macro.
1341 The following options are available when @value{AS} is configured for
1347 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1348 The command line option @samp{-nojsri2bsr} can be used to disable it.
1352 Enable or disable the silicon filter behaviour. By default this is disabled.
1353 The default can be overridden by the @samp{-sifilter} command line option.
1356 Alter jump instructions for long displacements.
1358 @item -mcpu=[210|340]
1359 Select the cpu type on the target hardware. This controls which instructions
1363 Assemble for a big endian target.
1366 Assemble for a little endian target.
1372 See the info pages for documentation of the MMIX-specific options.
1379 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1380 for a PowerPC processor.
1384 @c man begin OPTIONS
1385 The following options are available when @value{AS} is configured for a
1388 @c man begin INCLUDE
1390 @c ended inside the included file
1395 @c man begin OPTIONS
1397 See the info pages for documentation of the RX-specific options.
1401 The following options are available when @value{AS} is configured for the s390
1407 Select the word size, either 31/32 bits or 64 bits.
1410 Select the architecture mode, either the Enterprise System
1411 Architecture (esa) or the z/Architecture mode (zarch).
1412 @item -march=@var{processor}
1413 Specify which s390 processor variant is the target, @samp{g6}, @samp{g6},
1414 @samp{z900}, @samp{z990}, @samp{z9-109}, @samp{z9-ec}, @samp{z10},
1415 @samp{z196}, or @samp{zEC12}.
1417 @itemx -mno-regnames
1418 Allow or disallow symbolic names for registers.
1419 @item -mwarn-areg-zero
1420 Warn whenever the operand for a base or index register has been specified
1421 but evaluates to zero.
1429 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1430 for a TMS320C6000 processor.
1434 @c man begin OPTIONS
1435 The following options are available when @value{AS} is configured for a
1436 TMS320C6000 processor.
1438 @c man begin INCLUDE
1439 @include c-tic6x.texi
1440 @c ended inside the included file
1448 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1449 for a TILE-Gx processor.
1453 @c man begin OPTIONS
1454 The following options are available when @value{AS} is configured for a TILE-Gx
1457 @c man begin INCLUDE
1458 @include c-tilegx.texi
1459 @c ended inside the included file
1467 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1468 for an Xtensa processor.
1472 @c man begin OPTIONS
1473 The following options are available when @value{AS} is configured for an
1476 @c man begin INCLUDE
1477 @include c-xtensa.texi
1478 @c ended inside the included file
1483 @c man begin OPTIONS
1486 The following options are available when @value{AS} is configured for
1487 a Z80 family processor.
1490 Assemble for Z80 processor.
1492 Assemble for R800 processor.
1493 @item -ignore-undocumented-instructions
1495 Assemble undocumented Z80 instructions that also work on R800 without warning.
1496 @item -ignore-unportable-instructions
1498 Assemble all undocumented Z80 instructions without warning.
1499 @item -warn-undocumented-instructions
1501 Issue a warning for undocumented Z80 instructions that also work on R800.
1502 @item -warn-unportable-instructions
1504 Issue a warning for undocumented Z80 instructions that do not work on R800.
1505 @item -forbid-undocumented-instructions
1507 Treat all undocumented instructions as errors.
1508 @item -forbid-unportable-instructions
1510 Treat undocumented Z80 instructions that do not work on R800 as errors.
1517 * Manual:: Structure of this Manual
1518 * GNU Assembler:: The GNU Assembler
1519 * Object Formats:: Object File Formats
1520 * Command Line:: Command Line
1521 * Input Files:: Input Files
1522 * Object:: Output (Object) File
1523 * Errors:: Error and Warning Messages
1527 @section Structure of this Manual
1529 @cindex manual, structure and purpose
1530 This manual is intended to describe what you need to know to use
1531 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1532 notation for symbols, constants, and expressions; the directives that
1533 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1536 We also cover special features in the @value{TARGET}
1537 configuration of @command{@value{AS}}, including assembler directives.
1540 This manual also describes some of the machine-dependent features of
1541 various flavors of the assembler.
1544 @cindex machine instructions (not covered)
1545 On the other hand, this manual is @emph{not} intended as an introduction
1546 to programming in assembly language---let alone programming in general!
1547 In a similar vein, we make no attempt to introduce the machine
1548 architecture; we do @emph{not} describe the instruction set, standard
1549 mnemonics, registers or addressing modes that are standard to a
1550 particular architecture.
1552 You may want to consult the manufacturer's
1553 machine architecture manual for this information.
1557 For information on the H8/300 machine instruction set, see @cite{H8/300
1558 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1559 Programming Manual} (Renesas).
1562 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1563 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1564 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1565 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1568 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1572 @c I think this is premature---doc@cygnus.com, 17jan1991
1574 Throughout this manual, we assume that you are running @dfn{GNU},
1575 the portable operating system from the @dfn{Free Software
1576 Foundation, Inc.}. This restricts our attention to certain kinds of
1577 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1578 once this assumption is granted examples and definitions need less
1581 @command{@value{AS}} is part of a team of programs that turn a high-level
1582 human-readable series of instructions into a low-level
1583 computer-readable series of instructions. Different versions of
1584 @command{@value{AS}} are used for different kinds of computer.
1587 @c There used to be a section "Terminology" here, which defined
1588 @c "contents", "byte", "word", and "long". Defining "word" to any
1589 @c particular size is confusing when the .word directive may generate 16
1590 @c bits on one machine and 32 bits on another; in general, for the user
1591 @c version of this manual, none of these terms seem essential to define.
1592 @c They were used very little even in the former draft of the manual;
1593 @c this draft makes an effort to avoid them (except in names of
1597 @section The GNU Assembler
1599 @c man begin DESCRIPTION
1601 @sc{gnu} @command{as} is really a family of assemblers.
1603 This manual describes @command{@value{AS}}, a member of that family which is
1604 configured for the @value{TARGET} architectures.
1606 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1607 should find a fairly similar environment when you use it on another
1608 architecture. Each version has much in common with the others,
1609 including object file formats, most assembler directives (often called
1610 @dfn{pseudo-ops}) and assembler syntax.@refill
1612 @cindex purpose of @sc{gnu} assembler
1613 @command{@value{AS}} is primarily intended to assemble the output of the
1614 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1615 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1616 assemble correctly everything that other assemblers for the same
1617 machine would assemble.
1619 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1622 @c This remark should appear in generic version of manual; assumption
1623 @c here is that generic version sets M680x0.
1624 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1625 assembler for the same architecture; for example, we know of several
1626 incompatible versions of 680x0 assembly language syntax.
1631 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1632 program in one pass of the source file. This has a subtle impact on the
1633 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1635 @node Object Formats
1636 @section Object File Formats
1638 @cindex object file format
1639 The @sc{gnu} assembler can be configured to produce several alternative
1640 object file formats. For the most part, this does not affect how you
1641 write assembly language programs; but directives for debugging symbols
1642 are typically different in different file formats. @xref{Symbol
1643 Attributes,,Symbol Attributes}.
1646 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1647 @value{OBJ-NAME} format object files.
1649 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1651 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1652 @code{b.out} or COFF format object files.
1655 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1656 SOM or ELF format object files.
1661 @section Command Line
1663 @cindex command line conventions
1665 After the program name @command{@value{AS}}, the command line may contain
1666 options and file names. Options may appear in any order, and may be
1667 before, after, or between file names. The order of file names is
1670 @cindex standard input, as input file
1672 @file{--} (two hyphens) by itself names the standard input file
1673 explicitly, as one of the files for @command{@value{AS}} to assemble.
1675 @cindex options, command line
1676 Except for @samp{--} any command line argument that begins with a
1677 hyphen (@samp{-}) is an option. Each option changes the behavior of
1678 @command{@value{AS}}. No option changes the way another option works. An
1679 option is a @samp{-} followed by one or more letters; the case of
1680 the letter is important. All options are optional.
1682 Some options expect exactly one file name to follow them. The file
1683 name may either immediately follow the option's letter (compatible
1684 with older assemblers) or it may be the next command argument (@sc{gnu}
1685 standard). These two command lines are equivalent:
1688 @value{AS} -o my-object-file.o mumble.s
1689 @value{AS} -omy-object-file.o mumble.s
1693 @section Input Files
1696 @cindex source program
1697 @cindex files, input
1698 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1699 describe the program input to one run of @command{@value{AS}}. The program may
1700 be in one or more files; how the source is partitioned into files
1701 doesn't change the meaning of the source.
1703 @c I added "con" prefix to "catenation" just to prove I can overcome my
1704 @c APL training... doc@cygnus.com
1705 The source program is a concatenation of the text in all the files, in the
1708 @c man begin DESCRIPTION
1709 Each time you run @command{@value{AS}} it assembles exactly one source
1710 program. The source program is made up of one or more files.
1711 (The standard input is also a file.)
1713 You give @command{@value{AS}} a command line that has zero or more input file
1714 names. The input files are read (from left file name to right). A
1715 command line argument (in any position) that has no special meaning
1716 is taken to be an input file name.
1718 If you give @command{@value{AS}} no file names it attempts to read one input file
1719 from the @command{@value{AS}} standard input, which is normally your terminal. You
1720 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1723 Use @samp{--} if you need to explicitly name the standard input file
1724 in your command line.
1726 If the source is empty, @command{@value{AS}} produces a small, empty object
1731 @subheading Filenames and Line-numbers
1733 @cindex input file linenumbers
1734 @cindex line numbers, in input files
1735 There are two ways of locating a line in the input file (or files) and
1736 either may be used in reporting error messages. One way refers to a line
1737 number in a physical file; the other refers to a line number in a
1738 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1740 @dfn{Physical files} are those files named in the command line given
1741 to @command{@value{AS}}.
1743 @dfn{Logical files} are simply names declared explicitly by assembler
1744 directives; they bear no relation to physical files. Logical file names help
1745 error messages reflect the original source file, when @command{@value{AS}} source
1746 is itself synthesized from other files. @command{@value{AS}} understands the
1747 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1748 @ref{File,,@code{.file}}.
1751 @section Output (Object) File
1757 Every time you run @command{@value{AS}} it produces an output file, which is
1758 your assembly language program translated into numbers. This file
1759 is the object file. Its default name is
1767 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1769 You can give it another name by using the @option{-o} option. Conventionally,
1770 object file names end with @file{.o}. The default name is used for historical
1771 reasons: older assemblers were capable of assembling self-contained programs
1772 directly into a runnable program. (For some formats, this isn't currently
1773 possible, but it can be done for the @code{a.out} format.)
1777 The object file is meant for input to the linker @code{@value{LD}}. It contains
1778 assembled program code, information to help @code{@value{LD}} integrate
1779 the assembled program into a runnable file, and (optionally) symbolic
1780 information for the debugger.
1782 @c link above to some info file(s) like the description of a.out.
1783 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1786 @section Error and Warning Messages
1788 @c man begin DESCRIPTION
1790 @cindex error messages
1791 @cindex warning messages
1792 @cindex messages from assembler
1793 @command{@value{AS}} may write warnings and error messages to the standard error
1794 file (usually your terminal). This should not happen when a compiler
1795 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1796 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1797 grave problem that stops the assembly.
1801 @cindex format of warning messages
1802 Warning messages have the format
1805 file_name:@b{NNN}:Warning Message Text
1809 @cindex line numbers, in warnings/errors
1810 (where @b{NNN} is a line number). If a logical file name has been given
1811 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1812 the current input file is used. If a logical line number was given
1814 (@pxref{Line,,@code{.line}})
1816 then it is used to calculate the number printed,
1817 otherwise the actual line in the current source file is printed. The
1818 message text is intended to be self explanatory (in the grand Unix
1821 @cindex format of error messages
1822 Error messages have the format
1824 file_name:@b{NNN}:FATAL:Error Message Text
1826 The file name and line number are derived as for warning
1827 messages. The actual message text may be rather less explanatory
1828 because many of them aren't supposed to happen.
1831 @chapter Command-Line Options
1833 @cindex options, all versions of assembler
1834 This chapter describes command-line options available in @emph{all}
1835 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
1836 for options specific
1838 to the @value{TARGET} target.
1841 to particular machine architectures.
1844 @c man begin DESCRIPTION
1846 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1847 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1848 The assembler arguments must be separated from each other (and the @samp{-Wa})
1849 by commas. For example:
1852 gcc -c -g -O -Wa,-alh,-L file.c
1856 This passes two options to the assembler: @samp{-alh} (emit a listing to
1857 standard output with high-level and assembly source) and @samp{-L} (retain
1858 local symbols in the symbol table).
1860 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1861 command-line options are automatically passed to the assembler by the compiler.
1862 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1863 precisely what options it passes to each compilation pass, including the
1869 * a:: -a[cdghlns] enable listings
1870 * alternate:: --alternate enable alternate macro syntax
1871 * D:: -D for compatibility
1872 * f:: -f to work faster
1873 * I:: -I for .include search path
1874 @ifclear DIFF-TBL-KLUGE
1875 * K:: -K for compatibility
1877 @ifset DIFF-TBL-KLUGE
1878 * K:: -K for difference tables
1881 * L:: -L to retain local symbols
1882 * listing:: --listing-XXX to configure listing output
1883 * M:: -M or --mri to assemble in MRI compatibility mode
1884 * MD:: --MD for dependency tracking
1885 * o:: -o to name the object file
1886 * R:: -R to join data and text sections
1887 * statistics:: --statistics to see statistics about assembly
1888 * traditional-format:: --traditional-format for compatible output
1889 * v:: -v to announce version
1890 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1891 * Z:: -Z to make object file even after errors
1895 @section Enable Listings: @option{-a[cdghlns]}
1905 @cindex listings, enabling
1906 @cindex assembly listings, enabling
1908 These options enable listing output from the assembler. By itself,
1909 @samp{-a} requests high-level, assembly, and symbols listing.
1910 You can use other letters to select specific options for the list:
1911 @samp{-ah} requests a high-level language listing,
1912 @samp{-al} requests an output-program assembly listing, and
1913 @samp{-as} requests a symbol table listing.
1914 High-level listings require that a compiler debugging option like
1915 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1918 Use the @samp{-ag} option to print a first section with general assembly
1919 information, like @value{AS} version, switches passed, or time stamp.
1921 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1922 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1923 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1924 omitted from the listing.
1926 Use the @samp{-ad} option to omit debugging directives from the
1929 Once you have specified one of these options, you can further control
1930 listing output and its appearance using the directives @code{.list},
1931 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1933 The @samp{-an} option turns off all forms processing.
1934 If you do not request listing output with one of the @samp{-a} options, the
1935 listing-control directives have no effect.
1937 The letters after @samp{-a} may be combined into one option,
1938 @emph{e.g.}, @samp{-aln}.
1940 Note if the assembler source is coming from the standard input (e.g.,
1942 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1943 is being used) then the listing will not contain any comments or preprocessor
1944 directives. This is because the listing code buffers input source lines from
1945 stdin only after they have been preprocessed by the assembler. This reduces
1946 memory usage and makes the code more efficient.
1949 @section @option{--alternate}
1952 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1955 @section @option{-D}
1958 This option has no effect whatsoever, but it is accepted to make it more
1959 likely that scripts written for other assemblers also work with
1960 @command{@value{AS}}.
1963 @section Work Faster: @option{-f}
1966 @cindex trusted compiler
1967 @cindex faster processing (@option{-f})
1968 @samp{-f} should only be used when assembling programs written by a
1969 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1970 and comment preprocessing on
1971 the input file(s) before assembling them. @xref{Preprocessing,
1975 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1976 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1981 @section @code{.include} Search Path: @option{-I} @var{path}
1983 @kindex -I @var{path}
1984 @cindex paths for @code{.include}
1985 @cindex search path for @code{.include}
1986 @cindex @code{include} directive search path
1987 Use this option to add a @var{path} to the list of directories
1988 @command{@value{AS}} searches for files specified in @code{.include}
1989 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1990 many times as necessary to include a variety of paths. The current
1991 working directory is always searched first; after that, @command{@value{AS}}
1992 searches any @samp{-I} directories in the same order as they were
1993 specified (left to right) on the command line.
1996 @section Difference Tables: @option{-K}
1999 @ifclear DIFF-TBL-KLUGE
2000 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2001 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2002 where it can be used to warn when the assembler alters the machine code
2003 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2004 family does not have the addressing limitations that sometimes lead to this
2005 alteration on other platforms.
2008 @ifset DIFF-TBL-KLUGE
2009 @cindex difference tables, warning
2010 @cindex warning for altered difference tables
2011 @command{@value{AS}} sometimes alters the code emitted for directives of the
2012 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2013 You can use the @samp{-K} option if you want a warning issued when this
2018 @section Include Local Symbols: @option{-L}
2021 @cindex local symbols, retaining in output
2022 Symbols beginning with system-specific local label prefixes, typically
2023 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2024 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2025 such symbols when debugging, because they are intended for the use of
2026 programs (like compilers) that compose assembler programs, not for your
2027 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2028 such symbols, so you do not normally debug with them.
2030 This option tells @command{@value{AS}} to retain those local symbols
2031 in the object file. Usually if you do this you also tell the linker
2032 @code{@value{LD}} to preserve those symbols.
2035 @section Configuring listing output: @option{--listing}
2037 The listing feature of the assembler can be enabled via the command line switch
2038 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2039 hex dump of the corresponding locations in the output object file, and displays
2040 them as a listing file. The format of this listing can be controlled by
2041 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2042 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2043 @code{.psize} (@pxref{Psize}), and
2044 @code{.eject} (@pxref{Eject}) and also by the following switches:
2047 @item --listing-lhs-width=@samp{number}
2048 @kindex --listing-lhs-width
2049 @cindex Width of first line disassembly output
2050 Sets the maximum width, in words, of the first line of the hex byte dump. This
2051 dump appears on the left hand side of the listing output.
2053 @item --listing-lhs-width2=@samp{number}
2054 @kindex --listing-lhs-width2
2055 @cindex Width of continuation lines of disassembly output
2056 Sets the maximum width, in words, of any further lines of the hex byte dump for
2057 a given input source line. If this value is not specified, it defaults to being
2058 the same as the value specified for @samp{--listing-lhs-width}. If neither
2059 switch is used the default is to one.
2061 @item --listing-rhs-width=@samp{number}
2062 @kindex --listing-rhs-width
2063 @cindex Width of source line output
2064 Sets the maximum width, in characters, of the source line that is displayed
2065 alongside the hex dump. The default value for this parameter is 100. The
2066 source line is displayed on the right hand side of the listing output.
2068 @item --listing-cont-lines=@samp{number}
2069 @kindex --listing-cont-lines
2070 @cindex Maximum number of continuation lines
2071 Sets the maximum number of continuation lines of hex dump that will be
2072 displayed for a given single line of source input. The default value is 4.
2076 @section Assemble in MRI Compatibility Mode: @option{-M}
2079 @cindex MRI compatibility mode
2080 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2081 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2082 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
2083 configured target) assembler from Microtec Research. The exact nature of the
2084 MRI syntax will not be documented here; see the MRI manuals for more
2085 information. Note in particular that the handling of macros and macro
2086 arguments is somewhat different. The purpose of this option is to permit
2087 assembling existing MRI assembler code using @command{@value{AS}}.
2089 The MRI compatibility is not complete. Certain operations of the MRI assembler
2090 depend upon its object file format, and can not be supported using other object
2091 file formats. Supporting these would require enhancing each object file format
2092 individually. These are:
2095 @item global symbols in common section
2097 The m68k MRI assembler supports common sections which are merged by the linker.
2098 Other object file formats do not support this. @command{@value{AS}} handles
2099 common sections by treating them as a single common symbol. It permits local
2100 symbols to be defined within a common section, but it can not support global
2101 symbols, since it has no way to describe them.
2103 @item complex relocations
2105 The MRI assemblers support relocations against a negated section address, and
2106 relocations which combine the start addresses of two or more sections. These
2107 are not support by other object file formats.
2109 @item @code{END} pseudo-op specifying start address
2111 The MRI @code{END} pseudo-op permits the specification of a start address.
2112 This is not supported by other object file formats. The start address may
2113 instead be specified using the @option{-e} option to the linker, or in a linker
2116 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2118 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2119 name to the output file. This is not supported by other object file formats.
2121 @item @code{ORG} pseudo-op
2123 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2124 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2125 which changes the location within the current section. Absolute sections are
2126 not supported by other object file formats. The address of a section may be
2127 assigned within a linker script.
2130 There are some other features of the MRI assembler which are not supported by
2131 @command{@value{AS}}, typically either because they are difficult or because they
2132 seem of little consequence. Some of these may be supported in future releases.
2136 @item EBCDIC strings
2138 EBCDIC strings are not supported.
2140 @item packed binary coded decimal
2142 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2143 and @code{DCB.P} pseudo-ops are not supported.
2145 @item @code{FEQU} pseudo-op
2147 The m68k @code{FEQU} pseudo-op is not supported.
2149 @item @code{NOOBJ} pseudo-op
2151 The m68k @code{NOOBJ} pseudo-op is not supported.
2153 @item @code{OPT} branch control options
2155 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2156 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2157 relaxes all branches, whether forward or backward, to an appropriate size, so
2158 these options serve no purpose.
2160 @item @code{OPT} list control options
2162 The following m68k @code{OPT} list control options are ignored: @code{C},
2163 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2164 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2166 @item other @code{OPT} options
2168 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2169 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2171 @item @code{OPT} @code{D} option is default
2173 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2174 @code{OPT NOD} may be used to turn it off.
2176 @item @code{XREF} pseudo-op.
2178 The m68k @code{XREF} pseudo-op is ignored.
2180 @item @code{.debug} pseudo-op
2182 The i960 @code{.debug} pseudo-op is not supported.
2184 @item @code{.extended} pseudo-op
2186 The i960 @code{.extended} pseudo-op is not supported.
2188 @item @code{.list} pseudo-op.
2190 The various options of the i960 @code{.list} pseudo-op are not supported.
2192 @item @code{.optimize} pseudo-op
2194 The i960 @code{.optimize} pseudo-op is not supported.
2196 @item @code{.output} pseudo-op
2198 The i960 @code{.output} pseudo-op is not supported.
2200 @item @code{.setreal} pseudo-op
2202 The i960 @code{.setreal} pseudo-op is not supported.
2207 @section Dependency Tracking: @option{--MD}
2210 @cindex dependency tracking
2213 @command{@value{AS}} can generate a dependency file for the file it creates. This
2214 file consists of a single rule suitable for @code{make} describing the
2215 dependencies of the main source file.
2217 The rule is written to the file named in its argument.
2219 This feature is used in the automatic updating of makefiles.
2222 @section Name the Object File: @option{-o}
2225 @cindex naming object file
2226 @cindex object file name
2227 There is always one object file output when you run @command{@value{AS}}. By
2228 default it has the name
2231 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
2245 You use this option (which takes exactly one filename) to give the
2246 object file a different name.
2248 Whatever the object file is called, @command{@value{AS}} overwrites any
2249 existing file of the same name.
2252 @section Join Data and Text Sections: @option{-R}
2255 @cindex data and text sections, joining
2256 @cindex text and data sections, joining
2257 @cindex joining text and data sections
2258 @cindex merging text and data sections
2259 @option{-R} tells @command{@value{AS}} to write the object file as if all
2260 data-section data lives in the text section. This is only done at
2261 the very last moment: your binary data are the same, but data
2262 section parts are relocated differently. The data section part of
2263 your object file is zero bytes long because all its bytes are
2264 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2266 When you specify @option{-R} it would be possible to generate shorter
2267 address displacements (because we do not have to cross between text and
2268 data section). We refrain from doing this simply for compatibility with
2269 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2272 When @command{@value{AS}} is configured for COFF or ELF output,
2273 this option is only useful if you use sections named @samp{.text} and
2278 @option{-R} is not supported for any of the HPPA targets. Using
2279 @option{-R} generates a warning from @command{@value{AS}}.
2283 @section Display Assembly Statistics: @option{--statistics}
2285 @kindex --statistics
2286 @cindex statistics, about assembly
2287 @cindex time, total for assembly
2288 @cindex space used, maximum for assembly
2289 Use @samp{--statistics} to display two statistics about the resources used by
2290 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2291 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2294 @node traditional-format
2295 @section Compatible Output: @option{--traditional-format}
2297 @kindex --traditional-format
2298 For some targets, the output of @command{@value{AS}} is different in some ways
2299 from the output of some existing assembler. This switch requests
2300 @command{@value{AS}} to use the traditional format instead.
2302 For example, it disables the exception frame optimizations which
2303 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2306 @section Announce Version: @option{-v}
2310 @cindex assembler version
2311 @cindex version of assembler
2312 You can find out what version of as is running by including the
2313 option @samp{-v} (which you can also spell as @samp{-version}) on the
2317 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2319 @command{@value{AS}} should never give a warning or error message when
2320 assembling compiler output. But programs written by people often
2321 cause @command{@value{AS}} to give a warning that a particular assumption was
2322 made. All such warnings are directed to the standard error file.
2326 @cindex suppressing warnings
2327 @cindex warnings, suppressing
2328 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2329 This only affects the warning messages: it does not change any particular of
2330 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2333 @kindex --fatal-warnings
2334 @cindex errors, caused by warnings
2335 @cindex warnings, causing error
2336 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2337 files that generate warnings to be in error.
2340 @cindex warnings, switching on
2341 You can switch these options off again by specifying @option{--warn}, which
2342 causes warnings to be output as usual.
2345 @section Generate Object File in Spite of Errors: @option{-Z}
2346 @cindex object file, after errors
2347 @cindex errors, continuing after
2348 After an error message, @command{@value{AS}} normally produces no output. If for
2349 some reason you are interested in object file output even after
2350 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2351 option. If there are any errors, @command{@value{AS}} continues anyways, and
2352 writes an object file after a final warning message of the form @samp{@var{n}
2353 errors, @var{m} warnings, generating bad object file.}
2358 @cindex machine-independent syntax
2359 @cindex syntax, machine-independent
2360 This chapter describes the machine-independent syntax allowed in a
2361 source file. @command{@value{AS}} syntax is similar to what many other
2362 assemblers use; it is inspired by the BSD 4.2
2367 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2371 * Preprocessing:: Preprocessing
2372 * Whitespace:: Whitespace
2373 * Comments:: Comments
2374 * Symbol Intro:: Symbols
2375 * Statements:: Statements
2376 * Constants:: Constants
2380 @section Preprocessing
2382 @cindex preprocessing
2383 The @command{@value{AS}} internal preprocessor:
2385 @cindex whitespace, removed by preprocessor
2387 adjusts and removes extra whitespace. It leaves one space or tab before
2388 the keywords on a line, and turns any other whitespace on the line into
2391 @cindex comments, removed by preprocessor
2393 removes all comments, replacing them with a single space, or an
2394 appropriate number of newlines.
2396 @cindex constants, converted by preprocessor
2398 converts character constants into the appropriate numeric values.
2401 It does not do macro processing, include file handling, or
2402 anything else you may get from your C compiler's preprocessor. You can
2403 do include file processing with the @code{.include} directive
2404 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2405 to get other ``CPP'' style preprocessing by giving the input file a
2406 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2407 Output, gcc.info, Using GNU CC}.
2409 Excess whitespace, comments, and character constants
2410 cannot be used in the portions of the input text that are not
2413 @cindex turning preprocessing on and off
2414 @cindex preprocessing, turning on and off
2417 If the first line of an input file is @code{#NO_APP} or if you use the
2418 @samp{-f} option, whitespace and comments are not removed from the input file.
2419 Within an input file, you can ask for whitespace and comment removal in
2420 specific portions of the by putting a line that says @code{#APP} before the
2421 text that may contain whitespace or comments, and putting a line that says
2422 @code{#NO_APP} after this text. This feature is mainly intend to support
2423 @code{asm} statements in compilers whose output is otherwise free of comments
2430 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2431 Whitespace is used to separate symbols, and to make programs neater for
2432 people to read. Unless within character constants
2433 (@pxref{Characters,,Character Constants}), any whitespace means the same
2434 as exactly one space.
2440 There are two ways of rendering comments to @command{@value{AS}}. In both
2441 cases the comment is equivalent to one space.
2443 Anything from @samp{/*} through the next @samp{*/} is a comment.
2444 This means you may not nest these comments.
2448 The only way to include a newline ('\n') in a comment
2449 is to use this sort of comment.
2452 /* This sort of comment does not nest. */
2455 @cindex line comment character
2456 Anything from a @dfn{line comment} character up to the next newline is
2457 considered a comment and is ignored. The line comment character is target
2458 specific, and some targets multiple comment characters. Some targets also have
2459 line comment characters that only work if they are the first character on a
2460 line. Some targets use a sequence of two characters to introduce a line
2461 comment. Some targets can also change their line comment characters depending
2462 upon command line options that have been used. For more details see the
2463 @emph{Syntax} section in the documentation for individual targets.
2465 If the line comment character is the hash sign (@samp{#}) then it still has the
2466 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2467 to specify logical line numbers:
2470 @cindex lines starting with @code{#}
2471 @cindex logical line numbers
2472 To be compatible with past assemblers, lines that begin with @samp{#} have a
2473 special interpretation. Following the @samp{#} should be an absolute
2474 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2475 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2476 new logical file name. The rest of the line, if any, should be whitespace.
2478 If the first non-whitespace characters on the line are not numeric,
2479 the line is ignored. (Just like a comment.)
2482 # This is an ordinary comment.
2483 # 42-6 "new_file_name" # New logical file name
2484 # This is logical line # 36.
2486 This feature is deprecated, and may disappear from future versions
2487 of @command{@value{AS}}.
2492 @cindex characters used in symbols
2493 @ifclear SPECIAL-SYMS
2494 A @dfn{symbol} is one or more characters chosen from the set of all
2495 letters (both upper and lower case), digits and the three characters
2501 A @dfn{symbol} is one or more characters chosen from the set of all
2502 letters (both upper and lower case), digits and the three characters
2503 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2509 On most machines, you can also use @code{$} in symbol names; exceptions
2510 are noted in @ref{Machine Dependencies}.
2512 No symbol may begin with a digit. Case is significant.
2513 There is no length limit: all characters are significant. Multibyte characters
2514 are supported. Symbols are delimited by characters not in that set, or by the
2515 beginning of a file (since the source program must end with a newline, the end
2516 of a file is not a possible symbol delimiter). @xref{Symbols}.
2517 @cindex length of symbols
2522 @cindex statements, structure of
2523 @cindex line separator character
2524 @cindex statement separator character
2526 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2527 @dfn{line separator character}. The line separator character is target
2528 specific and described in the @emph{Syntax} section of each
2529 target's documentation. Not all targets support a line separator character.
2530 The newline or line separator character is considered to be part of the
2531 preceding statement. Newlines and separators within character constants are an
2532 exception: they do not end statements.
2534 @cindex newline, required at file end
2535 @cindex EOF, newline must precede
2536 It is an error to end any statement with end-of-file: the last
2537 character of any input file should be a newline.@refill
2539 An empty statement is allowed, and may include whitespace. It is ignored.
2541 @cindex instructions and directives
2542 @cindex directives and instructions
2543 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2544 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2546 A statement begins with zero or more labels, optionally followed by a
2547 key symbol which determines what kind of statement it is. The key
2548 symbol determines the syntax of the rest of the statement. If the
2549 symbol begins with a dot @samp{.} then the statement is an assembler
2550 directive: typically valid for any computer. If the symbol begins with
2551 a letter the statement is an assembly language @dfn{instruction}: it
2552 assembles into a machine language instruction.
2554 Different versions of @command{@value{AS}} for different computers
2555 recognize different instructions. In fact, the same symbol may
2556 represent a different instruction in a different computer's assembly
2560 @cindex @code{:} (label)
2561 @cindex label (@code{:})
2562 A label is a symbol immediately followed by a colon (@code{:}).
2563 Whitespace before a label or after a colon is permitted, but you may not
2564 have whitespace between a label's symbol and its colon. @xref{Labels}.
2567 For HPPA targets, labels need not be immediately followed by a colon, but
2568 the definition of a label must begin in column zero. This also implies that
2569 only one label may be defined on each line.
2573 label: .directive followed by something
2574 another_label: # This is an empty statement.
2575 instruction operand_1, operand_2, @dots{}
2582 A constant is a number, written so that its value is known by
2583 inspection, without knowing any context. Like this:
2586 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2587 .ascii "Ring the bell\7" # A string constant.
2588 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2589 .float 0f-314159265358979323846264338327\
2590 95028841971.693993751E-40 # - pi, a flonum.
2595 * Characters:: Character Constants
2596 * Numbers:: Number Constants
2600 @subsection Character Constants
2602 @cindex character constants
2603 @cindex constants, character
2604 There are two kinds of character constants. A @dfn{character} stands
2605 for one character in one byte and its value may be used in
2606 numeric expressions. String constants (properly called string
2607 @emph{literals}) are potentially many bytes and their values may not be
2608 used in arithmetic expressions.
2612 * Chars:: Characters
2616 @subsubsection Strings
2618 @cindex string constants
2619 @cindex constants, string
2620 A @dfn{string} is written between double-quotes. It may contain
2621 double-quotes or null characters. The way to get special characters
2622 into a string is to @dfn{escape} these characters: precede them with
2623 a backslash @samp{\} character. For example @samp{\\} represents
2624 one backslash: the first @code{\} is an escape which tells
2625 @command{@value{AS}} to interpret the second character literally as a backslash
2626 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2627 escape character). The complete list of escapes follows.
2629 @cindex escape codes, character
2630 @cindex character escape codes
2633 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2635 @cindex @code{\b} (backspace character)
2636 @cindex backspace (@code{\b})
2638 Mnemonic for backspace; for ASCII this is octal code 010.
2641 @c Mnemonic for EOText; for ASCII this is octal code 004.
2643 @cindex @code{\f} (formfeed character)
2644 @cindex formfeed (@code{\f})
2646 Mnemonic for FormFeed; for ASCII this is octal code 014.
2648 @cindex @code{\n} (newline character)
2649 @cindex newline (@code{\n})
2651 Mnemonic for newline; for ASCII this is octal code 012.
2654 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2656 @cindex @code{\r} (carriage return character)
2657 @cindex carriage return (@code{\r})
2659 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2662 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2663 @c other assemblers.
2665 @cindex @code{\t} (tab)
2666 @cindex tab (@code{\t})
2668 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2671 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2672 @c @item \x @var{digit} @var{digit} @var{digit}
2673 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2675 @cindex @code{\@var{ddd}} (octal character code)
2676 @cindex octal character code (@code{\@var{ddd}})
2677 @item \ @var{digit} @var{digit} @var{digit}
2678 An octal character code. The numeric code is 3 octal digits.
2679 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2680 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2682 @cindex @code{\@var{xd...}} (hex character code)
2683 @cindex hex character code (@code{\@var{xd...}})
2684 @item \@code{x} @var{hex-digits...}
2685 A hex character code. All trailing hex digits are combined. Either upper or
2686 lower case @code{x} works.
2688 @cindex @code{\\} (@samp{\} character)
2689 @cindex backslash (@code{\\})
2691 Represents one @samp{\} character.
2694 @c Represents one @samp{'} (accent acute) character.
2695 @c This is needed in single character literals
2696 @c (@xref{Characters,,Character Constants}.) to represent
2699 @cindex @code{\"} (doublequote character)
2700 @cindex doublequote (@code{\"})
2702 Represents one @samp{"} character. Needed in strings to represent
2703 this character, because an unescaped @samp{"} would end the string.
2705 @item \ @var{anything-else}
2706 Any other character when escaped by @kbd{\} gives a warning, but
2707 assembles as if the @samp{\} was not present. The idea is that if
2708 you used an escape sequence you clearly didn't want the literal
2709 interpretation of the following character. However @command{@value{AS}} has no
2710 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2711 code and warns you of the fact.
2714 Which characters are escapable, and what those escapes represent,
2715 varies widely among assemblers. The current set is what we think
2716 the BSD 4.2 assembler recognizes, and is a subset of what most C
2717 compilers recognize. If you are in doubt, do not use an escape
2721 @subsubsection Characters
2723 @cindex single character constant
2724 @cindex character, single
2725 @cindex constant, single character
2726 A single character may be written as a single quote immediately
2727 followed by that character. The same escapes apply to characters as
2728 to strings. So if you want to write the character backslash, you
2729 must write @kbd{'\\} where the first @code{\} escapes the second
2730 @code{\}. As you can see, the quote is an acute accent, not a
2731 grave accent. A newline
2733 @ifclear abnormal-separator
2734 (or semicolon @samp{;})
2736 @ifset abnormal-separator
2738 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2743 immediately following an acute accent is taken as a literal character
2744 and does not count as the end of a statement. The value of a character
2745 constant in a numeric expression is the machine's byte-wide code for
2746 that character. @command{@value{AS}} assumes your character code is ASCII:
2747 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2750 @subsection Number Constants
2752 @cindex constants, number
2753 @cindex number constants
2754 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2755 are stored in the target machine. @emph{Integers} are numbers that
2756 would fit into an @code{int} in the C language. @emph{Bignums} are
2757 integers, but they are stored in more than 32 bits. @emph{Flonums}
2758 are floating point numbers, described below.
2761 * Integers:: Integers
2766 * Bit Fields:: Bit Fields
2772 @subsubsection Integers
2774 @cindex constants, integer
2776 @cindex binary integers
2777 @cindex integers, binary
2778 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2779 the binary digits @samp{01}.
2781 @cindex octal integers
2782 @cindex integers, octal
2783 An octal integer is @samp{0} followed by zero or more of the octal
2784 digits (@samp{01234567}).
2786 @cindex decimal integers
2787 @cindex integers, decimal
2788 A decimal integer starts with a non-zero digit followed by zero or
2789 more digits (@samp{0123456789}).
2791 @cindex hexadecimal integers
2792 @cindex integers, hexadecimal
2793 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2794 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2796 Integers have the usual values. To denote a negative integer, use
2797 the prefix operator @samp{-} discussed under expressions
2798 (@pxref{Prefix Ops,,Prefix Operators}).
2801 @subsubsection Bignums
2804 @cindex constants, bignum
2805 A @dfn{bignum} has the same syntax and semantics as an integer
2806 except that the number (or its negative) takes more than 32 bits to
2807 represent in binary. The distinction is made because in some places
2808 integers are permitted while bignums are not.
2811 @subsubsection Flonums
2813 @cindex floating point numbers
2814 @cindex constants, floating point
2816 @cindex precision, floating point
2817 A @dfn{flonum} represents a floating point number. The translation is
2818 indirect: a decimal floating point number from the text is converted by
2819 @command{@value{AS}} to a generic binary floating point number of more than
2820 sufficient precision. This generic floating point number is converted
2821 to a particular computer's floating point format (or formats) by a
2822 portion of @command{@value{AS}} specialized to that computer.
2824 A flonum is written by writing (in order)
2829 (@samp{0} is optional on the HPPA.)
2833 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2835 @kbd{e} is recommended. Case is not important.
2837 @c FIXME: verify if flonum syntax really this vague for most cases
2838 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2839 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2842 On the H8/300, Renesas / SuperH SH,
2843 and AMD 29K architectures, the letter must be
2844 one of the letters @samp{DFPRSX} (in upper or lower case).
2846 On the ARC, the letter must be one of the letters @samp{DFRS}
2847 (in upper or lower case).
2849 On the Intel 960 architecture, the letter must be
2850 one of the letters @samp{DFT} (in upper or lower case).
2852 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2856 One of the letters @samp{DFRS} (in upper or lower case).
2859 One of the letters @samp{DFPRSX} (in upper or lower case).
2862 The letter @samp{E} (upper case only).
2865 One of the letters @samp{DFT} (in upper or lower case).
2870 An optional sign: either @samp{+} or @samp{-}.
2873 An optional @dfn{integer part}: zero or more decimal digits.
2876 An optional @dfn{fractional part}: @samp{.} followed by zero
2877 or more decimal digits.
2880 An optional exponent, consisting of:
2884 An @samp{E} or @samp{e}.
2885 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2886 @c principle this can perfectly well be different on different targets.
2888 Optional sign: either @samp{+} or @samp{-}.
2890 One or more decimal digits.
2895 At least one of the integer part or the fractional part must be
2896 present. The floating point number has the usual base-10 value.
2898 @command{@value{AS}} does all processing using integers. Flonums are computed
2899 independently of any floating point hardware in the computer running
2900 @command{@value{AS}}.
2904 @c Bit fields are written as a general facility but are also controlled
2905 @c by a conditional-compilation flag---which is as of now (21mar91)
2906 @c turned on only by the i960 config of GAS.
2908 @subsubsection Bit Fields
2911 @cindex constants, bit field
2912 You can also define numeric constants as @dfn{bit fields}.
2913 Specify two numbers separated by a colon---
2915 @var{mask}:@var{value}
2918 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2921 The resulting number is then packed
2923 @c this conditional paren in case bit fields turned on elsewhere than 960
2924 (in host-dependent byte order)
2926 into a field whose width depends on which assembler directive has the
2927 bit-field as its argument. Overflow (a result from the bitwise and
2928 requiring more binary digits to represent) is not an error; instead,
2929 more constants are generated, of the specified width, beginning with the
2930 least significant digits.@refill
2932 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2933 @code{.short}, and @code{.word} accept bit-field arguments.
2938 @chapter Sections and Relocation
2943 * Secs Background:: Background
2944 * Ld Sections:: Linker Sections
2945 * As Sections:: Assembler Internal Sections
2946 * Sub-Sections:: Sub-Sections
2950 @node Secs Background
2953 Roughly, a section is a range of addresses, with no gaps; all data
2954 ``in'' those addresses is treated the same for some particular purpose.
2955 For example there may be a ``read only'' section.
2957 @cindex linker, and assembler
2958 @cindex assembler, and linker
2959 The linker @code{@value{LD}} reads many object files (partial programs) and
2960 combines their contents to form a runnable program. When @command{@value{AS}}
2961 emits an object file, the partial program is assumed to start at address 0.
2962 @code{@value{LD}} assigns the final addresses for the partial program, so that
2963 different partial programs do not overlap. This is actually an
2964 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2967 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2968 addresses. These blocks slide to their run-time addresses as rigid
2969 units; their length does not change and neither does the order of bytes
2970 within them. Such a rigid unit is called a @emph{section}. Assigning
2971 run-time addresses to sections is called @dfn{relocation}. It includes
2972 the task of adjusting mentions of object-file addresses so they refer to
2973 the proper run-time addresses.
2975 For the H8/300, and for the Renesas / SuperH SH,
2976 @command{@value{AS}} pads sections if needed to
2977 ensure they end on a word (sixteen bit) boundary.
2980 @cindex standard assembler sections
2981 An object file written by @command{@value{AS}} has at least three sections, any
2982 of which may be empty. These are named @dfn{text}, @dfn{data} and
2987 When it generates COFF or ELF output,
2989 @command{@value{AS}} can also generate whatever other named sections you specify
2990 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2991 If you do not use any directives that place output in the @samp{.text}
2992 or @samp{.data} sections, these sections still exist, but are empty.
2997 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2999 @command{@value{AS}} can also generate whatever other named sections you
3000 specify using the @samp{.space} and @samp{.subspace} directives. See
3001 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3002 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3003 assembler directives.
3006 Additionally, @command{@value{AS}} uses different names for the standard
3007 text, data, and bss sections when generating SOM output. Program text
3008 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3009 BSS into @samp{$BSS$}.
3013 Within the object file, the text section starts at address @code{0}, the
3014 data section follows, and the bss section follows the data section.
3017 When generating either SOM or ELF output files on the HPPA, the text
3018 section starts at address @code{0}, the data section at address
3019 @code{0x4000000}, and the bss section follows the data section.
3022 To let @code{@value{LD}} know which data changes when the sections are
3023 relocated, and how to change that data, @command{@value{AS}} also writes to the
3024 object file details of the relocation needed. To perform relocation
3025 @code{@value{LD}} must know, each time an address in the object
3029 Where in the object file is the beginning of this reference to
3032 How long (in bytes) is this reference?
3034 Which section does the address refer to? What is the numeric value of
3036 (@var{address}) @minus{} (@var{start-address of section})?
3039 Is the reference to an address ``Program-Counter relative''?
3042 @cindex addresses, format of
3043 @cindex section-relative addressing
3044 In fact, every address @command{@value{AS}} ever uses is expressed as
3046 (@var{section}) + (@var{offset into section})
3049 Further, most expressions @command{@value{AS}} computes have this section-relative
3052 (For some object formats, such as SOM for the HPPA, some expressions are
3053 symbol-relative instead.)
3056 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3057 @var{N} into section @var{secname}.''
3059 Apart from text, data and bss sections you need to know about the
3060 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3061 addresses in the absolute section remain unchanged. For example, address
3062 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3063 @code{@value{LD}}. Although the linker never arranges two partial programs'
3064 data sections with overlapping addresses after linking, @emph{by definition}
3065 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3066 part of a program is always the same address when the program is running as
3067 address @code{@{absolute@ 239@}} in any other part of the program.
3069 The idea of sections is extended to the @dfn{undefined} section. Any
3070 address whose section is unknown at assembly time is by definition
3071 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3072 Since numbers are always defined, the only way to generate an undefined
3073 address is to mention an undefined symbol. A reference to a named
3074 common block would be such a symbol: its value is unknown at assembly
3075 time so it has section @emph{undefined}.
3077 By analogy the word @emph{section} is used to describe groups of sections in
3078 the linked program. @code{@value{LD}} puts all partial programs' text
3079 sections in contiguous addresses in the linked program. It is
3080 customary to refer to the @emph{text section} of a program, meaning all
3081 the addresses of all partial programs' text sections. Likewise for
3082 data and bss sections.
3084 Some sections are manipulated by @code{@value{LD}}; others are invented for
3085 use of @command{@value{AS}} and have no meaning except during assembly.
3088 @section Linker Sections
3089 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3094 @cindex named sections
3095 @cindex sections, named
3096 @item named sections
3099 @cindex text section
3100 @cindex data section
3104 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3105 separate but equal sections. Anything you can say of one section is
3108 When the program is running, however, it is
3109 customary for the text section to be unalterable. The
3110 text section is often shared among processes: it contains
3111 instructions, constants and the like. The data section of a running
3112 program is usually alterable: for example, C variables would be stored
3113 in the data section.
3118 This section contains zeroed bytes when your program begins running. It
3119 is used to hold uninitialized variables or common storage. The length of
3120 each partial program's bss section is important, but because it starts
3121 out containing zeroed bytes there is no need to store explicit zero
3122 bytes in the object file. The bss section was invented to eliminate
3123 those explicit zeros from object files.
3125 @cindex absolute section
3126 @item absolute section
3127 Address 0 of this section is always ``relocated'' to runtime address 0.
3128 This is useful if you want to refer to an address that @code{@value{LD}} must
3129 not change when relocating. In this sense we speak of absolute
3130 addresses being ``unrelocatable'': they do not change during relocation.
3132 @cindex undefined section
3133 @item undefined section
3134 This ``section'' is a catch-all for address references to objects not in
3135 the preceding sections.
3136 @c FIXME: ref to some other doc on obj-file formats could go here.
3139 @cindex relocation example
3140 An idealized example of three relocatable sections follows.
3142 The example uses the traditional section names @samp{.text} and @samp{.data}.
3144 Memory addresses are on the horizontal axis.
3148 @c END TEXI2ROFF-KILL
3151 partial program # 1: |ttttt|dddd|00|
3158 partial program # 2: |TTT|DDD|000|
3161 +--+---+-----+--+----+---+-----+~~
3162 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3163 +--+---+-----+--+----+---+-----+~~
3165 addresses: 0 @dots{}
3172 \line{\it Partial program \#1: \hfil}
3173 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3174 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3176 \line{\it Partial program \#2: \hfil}
3177 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3178 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3180 \line{\it linked program: \hfil}
3181 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3182 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3183 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3184 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3186 \line{\it addresses: \hfil}
3190 @c END TEXI2ROFF-KILL
3193 @section Assembler Internal Sections
3195 @cindex internal assembler sections
3196 @cindex sections in messages, internal
3197 These sections are meant only for the internal use of @command{@value{AS}}. They
3198 have no meaning at run-time. You do not really need to know about these
3199 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3200 warning messages, so it might be helpful to have an idea of their
3201 meanings to @command{@value{AS}}. These sections are used to permit the
3202 value of every expression in your assembly language program to be a
3203 section-relative address.
3206 @cindex assembler internal logic error
3207 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3208 An internal assembler logic error has been found. This means there is a
3209 bug in the assembler.
3211 @cindex expr (internal section)
3213 The assembler stores complex expression internally as combinations of
3214 symbols. When it needs to represent an expression as a symbol, it puts
3215 it in the expr section.
3217 @c FIXME item transfer[t] vector preload
3218 @c FIXME item transfer[t] vector postload
3219 @c FIXME item register
3223 @section Sub-Sections
3225 @cindex numbered subsections
3226 @cindex grouping data
3232 fall into two sections: text and data.
3234 You may have separate groups of
3236 data in named sections
3240 data in named sections
3246 that you want to end up near to each other in the object file, even though they
3247 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3248 use @dfn{subsections} for this purpose. Within each section, there can be
3249 numbered subsections with values from 0 to 8192. Objects assembled into the
3250 same subsection go into the object file together with other objects in the same
3251 subsection. For example, a compiler might want to store constants in the text
3252 section, but might not want to have them interspersed with the program being
3253 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3254 section of code being output, and a @samp{.text 1} before each group of
3255 constants being output.
3257 Subsections are optional. If you do not use subsections, everything
3258 goes in subsection number zero.
3261 Each subsection is zero-padded up to a multiple of four bytes.
3262 (Subsections may be padded a different amount on different flavors
3263 of @command{@value{AS}}.)
3267 On the H8/300 platform, each subsection is zero-padded to a word
3268 boundary (two bytes).
3269 The same is true on the Renesas SH.
3272 @c FIXME section padding (alignment)?
3273 @c Rich Pixley says padding here depends on target obj code format; that
3274 @c doesn't seem particularly useful to say without further elaboration,
3275 @c so for now I say nothing about it. If this is a generic BFD issue,
3276 @c these paragraphs might need to vanish from this manual, and be
3277 @c discussed in BFD chapter of binutils (or some such).
3281 Subsections appear in your object file in numeric order, lowest numbered
3282 to highest. (All this to be compatible with other people's assemblers.)
3283 The object file contains no representation of subsections; @code{@value{LD}} and
3284 other programs that manipulate object files see no trace of them.
3285 They just see all your text subsections as a text section, and all your
3286 data subsections as a data section.
3288 To specify which subsection you want subsequent statements assembled
3289 into, use a numeric argument to specify it, in a @samp{.text
3290 @var{expression}} or a @samp{.data @var{expression}} statement.
3293 When generating COFF output, you
3298 can also use an extra subsection
3299 argument with arbitrary named sections: @samp{.section @var{name},
3304 When generating ELF output, you
3309 can also use the @code{.subsection} directive (@pxref{SubSection})
3310 to specify a subsection: @samp{.subsection @var{expression}}.
3312 @var{Expression} should be an absolute expression
3313 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3314 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3315 begins in @code{text 0}. For instance:
3317 .text 0 # The default subsection is text 0 anyway.
3318 .ascii "This lives in the first text subsection. *"
3320 .ascii "But this lives in the second text subsection."
3322 .ascii "This lives in the data section,"
3323 .ascii "in the first data subsection."
3325 .ascii "This lives in the first text section,"
3326 .ascii "immediately following the asterisk (*)."
3329 Each section has a @dfn{location counter} incremented by one for every byte
3330 assembled into that section. Because subsections are merely a convenience
3331 restricted to @command{@value{AS}} there is no concept of a subsection location
3332 counter. There is no way to directly manipulate a location counter---but the
3333 @code{.align} directive changes it, and any label definition captures its
3334 current value. The location counter of the section where statements are being
3335 assembled is said to be the @dfn{active} location counter.
3338 @section bss Section
3341 @cindex common variable storage
3342 The bss section is used for local common variable storage.
3343 You may allocate address space in the bss section, but you may
3344 not dictate data to load into it before your program executes. When
3345 your program starts running, all the contents of the bss
3346 section are zeroed bytes.
3348 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3349 @ref{Lcomm,,@code{.lcomm}}.
3351 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3352 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3355 When assembling for a target which supports multiple sections, such as ELF or
3356 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3357 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3358 section. Typically the section will only contain symbol definitions and
3359 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3366 Symbols are a central concept: the programmer uses symbols to name
3367 things, the linker uses symbols to link, and the debugger uses symbols
3371 @cindex debuggers, and symbol order
3372 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3373 the same order they were declared. This may break some debuggers.
3378 * Setting Symbols:: Giving Symbols Other Values
3379 * Symbol Names:: Symbol Names
3380 * Dot:: The Special Dot Symbol
3381 * Symbol Attributes:: Symbol Attributes
3388 A @dfn{label} is written as a symbol immediately followed by a colon
3389 @samp{:}. The symbol then represents the current value of the
3390 active location counter, and is, for example, a suitable instruction
3391 operand. You are warned if you use the same symbol to represent two
3392 different locations: the first definition overrides any other
3396 On the HPPA, the usual form for a label need not be immediately followed by a
3397 colon, but instead must start in column zero. Only one label may be defined on
3398 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3399 provides a special directive @code{.label} for defining labels more flexibly.
3402 @node Setting Symbols
3403 @section Giving Symbols Other Values
3405 @cindex assigning values to symbols
3406 @cindex symbol values, assigning
3407 A symbol can be given an arbitrary value by writing a symbol, followed
3408 by an equals sign @samp{=}, followed by an expression
3409 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3410 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3411 equals sign @samp{=}@samp{=} here represents an equivalent of the
3412 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3415 Blackfin does not support symbol assignment with @samp{=}.
3419 @section Symbol Names
3421 @cindex symbol names
3422 @cindex names, symbol
3423 @ifclear SPECIAL-SYMS
3424 Symbol names begin with a letter or with one of @samp{._}. On most
3425 machines, you can also use @code{$} in symbol names; exceptions are
3426 noted in @ref{Machine Dependencies}. That character may be followed by any
3427 string of digits, letters, dollar signs (unless otherwise noted for a
3428 particular target machine), and underscores.
3432 Symbol names begin with a letter or with one of @samp{._}. On the
3433 Renesas SH you can also use @code{$} in symbol names. That
3434 character may be followed by any string of digits, letters, dollar signs (save
3435 on the H8/300), and underscores.
3439 Case of letters is significant: @code{foo} is a different symbol name
3442 Multibyte characters are supported. To generate a symbol name containing
3443 multibyte characters enclose it within double quotes and use escape codes. cf
3444 @xref{Strings}. Generating a multibyte symbol name from a label is not
3445 currently supported.
3447 Each symbol has exactly one name. Each name in an assembly language program
3448 refers to exactly one symbol. You may use that symbol name any number of times
3451 @subheading Local Symbol Names
3453 @cindex local symbol names
3454 @cindex symbol names, local
3455 A local symbol is any symbol beginning with certain local label prefixes.
3456 By default, the local label prefix is @samp{.L} for ELF systems or
3457 @samp{L} for traditional a.out systems, but each target may have its own
3458 set of local label prefixes.
3460 On the HPPA local symbols begin with @samp{L$}.
3463 Local symbols are defined and used within the assembler, but they are
3464 normally not saved in object files. Thus, they are not visible when debugging.
3465 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols:
3466 @option{-L}}) to retain the local symbols in the object files.
3468 @subheading Local Labels
3470 @cindex local labels
3471 @cindex temporary symbol names
3472 @cindex symbol names, temporary
3473 Local labels help compilers and programmers use names temporarily.
3474 They create symbols which are guaranteed to be unique over the entire scope of
3475 the input source code and which can be referred to by a simple notation.
3476 To define a local label, write a label of the form @samp{@b{N}:} (where @b{N}
3477 represents any positive integer). To refer to the most recent previous
3478 definition of that label write @samp{@b{N}b}, using the same number as when
3479 you defined the label. To refer to the next definition of a local label, write
3480 @samp{@b{N}f}---the @samp{b} stands for ``backwards'' and the @samp{f} stands
3483 There is no restriction on how you can use these labels, and you can reuse them
3484 too. So that it is possible to repeatedly define the same local label (using
3485 the same number @samp{@b{N}}), although you can only refer to the most recently
3486 defined local label of that number (for a backwards reference) or the next
3487 definition of a specific local label for a forward reference. It is also worth
3488 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3489 implemented in a slightly more efficient manner than the others.
3500 Which is the equivalent of:
3503 label_1: branch label_3
3504 label_2: branch label_1
3505 label_3: branch label_4
3506 label_4: branch label_3
3509 Local label names are only a notational device. They are immediately
3510 transformed into more conventional symbol names before the assembler uses them.
3511 The symbol names are stored in the symbol table, appear in error messages, and
3512 are optionally emitted to the object file. The names are constructed using
3516 @item @emph{local label prefix}
3517 All local symbols begin with the system-specific local label prefix.
3518 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3519 that start with the local label prefix. These labels are
3520 used for symbols you are never intended to see. If you use the
3521 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3522 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3523 you may use them in debugging.
3526 This is the number that was used in the local label definition. So if the
3527 label is written @samp{55:} then the number is @samp{55}.
3530 This unusual character is included so you do not accidentally invent a symbol
3531 of the same name. The character has ASCII value of @samp{\002} (control-B).
3533 @item @emph{ordinal number}
3534 This is a serial number to keep the labels distinct. The first definition of
3535 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3536 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3537 the number @samp{1} and its 15th definition gets @samp{15} as well.
3540 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3541 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3543 @subheading Dollar Local Labels
3544 @cindex dollar local symbols
3546 @code{@value{AS}} also supports an even more local form of local labels called
3547 dollar labels. These labels go out of scope (i.e., they become undefined) as
3548 soon as a non-local label is defined. Thus they remain valid for only a small
3549 region of the input source code. Normal local labels, by contrast, remain in
3550 scope for the entire file, or until they are redefined by another occurrence of
3551 the same local label.
3553 Dollar labels are defined in exactly the same way as ordinary local labels,
3554 except that they have a dollar sign suffix to their numeric value, e.g.,
3557 They can also be distinguished from ordinary local labels by their transformed
3558 names which use ASCII character @samp{\001} (control-A) as the magic character
3559 to distinguish them from ordinary labels. For example, the fifth definition of
3560 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3563 @section The Special Dot Symbol
3565 @cindex dot (symbol)
3566 @cindex @code{.} (symbol)
3567 @cindex current address
3568 @cindex location counter
3569 The special symbol @samp{.} refers to the current address that
3570 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3571 .long .} defines @code{melvin} to contain its own address.
3572 Assigning a value to @code{.} is treated the same as a @code{.org}
3574 @ifclear no-space-dir
3575 Thus, the expression @samp{.=.+4} is the same as saying
3579 @node Symbol Attributes
3580 @section Symbol Attributes
3582 @cindex symbol attributes
3583 @cindex attributes, symbol
3584 Every symbol has, as well as its name, the attributes ``Value'' and
3585 ``Type''. Depending on output format, symbols can also have auxiliary
3588 The detailed definitions are in @file{a.out.h}.
3591 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3592 all these attributes, and probably won't warn you. This makes the
3593 symbol an externally defined symbol, which is generally what you
3597 * Symbol Value:: Value
3598 * Symbol Type:: Type
3601 * a.out Symbols:: Symbol Attributes: @code{a.out}
3605 * a.out Symbols:: Symbol Attributes: @code{a.out}
3608 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3613 * COFF Symbols:: Symbol Attributes for COFF
3616 * SOM Symbols:: Symbol Attributes for SOM
3623 @cindex value of a symbol
3624 @cindex symbol value
3625 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3626 location in the text, data, bss or absolute sections the value is the
3627 number of addresses from the start of that section to the label.
3628 Naturally for text, data and bss sections the value of a symbol changes
3629 as @code{@value{LD}} changes section base addresses during linking. Absolute
3630 symbols' values do not change during linking: that is why they are
3633 The value of an undefined symbol is treated in a special way. If it is
3634 0 then the symbol is not defined in this assembler source file, and
3635 @code{@value{LD}} tries to determine its value from other files linked into the
3636 same program. You make this kind of symbol simply by mentioning a symbol
3637 name without defining it. A non-zero value represents a @code{.comm}
3638 common declaration. The value is how much common storage to reserve, in
3639 bytes (addresses). The symbol refers to the first address of the
3645 @cindex type of a symbol
3647 The type attribute of a symbol contains relocation (section)
3648 information, any flag settings indicating that a symbol is external, and
3649 (optionally), other information for linkers and debuggers. The exact
3650 format depends on the object-code output format in use.
3655 @c The following avoids a "widow" subsection title. @group would be
3656 @c better if it were available outside examples.
3659 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3661 @cindex @code{b.out} symbol attributes
3662 @cindex symbol attributes, @code{b.out}
3663 These symbol attributes appear only when @command{@value{AS}} is configured for
3664 one of the Berkeley-descended object output formats---@code{a.out} or
3670 @subsection Symbol Attributes: @code{a.out}
3672 @cindex @code{a.out} symbol attributes
3673 @cindex symbol attributes, @code{a.out}
3679 @subsection Symbol Attributes: @code{a.out}
3681 @cindex @code{a.out} symbol attributes
3682 @cindex symbol attributes, @code{a.out}
3686 * Symbol Desc:: Descriptor
3687 * Symbol Other:: Other
3691 @subsubsection Descriptor
3693 @cindex descriptor, of @code{a.out} symbol
3694 This is an arbitrary 16-bit value. You may establish a symbol's
3695 descriptor value by using a @code{.desc} statement
3696 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3697 @command{@value{AS}}.
3700 @subsubsection Other
3702 @cindex other attribute, of @code{a.out} symbol
3703 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3708 @subsection Symbol Attributes for COFF
3710 @cindex COFF symbol attributes
3711 @cindex symbol attributes, COFF
3713 The COFF format supports a multitude of auxiliary symbol attributes;
3714 like the primary symbol attributes, they are set between @code{.def} and
3715 @code{.endef} directives.
3717 @subsubsection Primary Attributes
3719 @cindex primary attributes, COFF symbols
3720 The symbol name is set with @code{.def}; the value and type,
3721 respectively, with @code{.val} and @code{.type}.
3723 @subsubsection Auxiliary Attributes
3725 @cindex auxiliary attributes, COFF symbols
3726 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3727 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3728 table information for COFF.
3733 @subsection Symbol Attributes for SOM
3735 @cindex SOM symbol attributes
3736 @cindex symbol attributes, SOM
3738 The SOM format for the HPPA supports a multitude of symbol attributes set with
3739 the @code{.EXPORT} and @code{.IMPORT} directives.
3741 The attributes are described in @cite{HP9000 Series 800 Assembly
3742 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3743 @code{EXPORT} assembler directive documentation.
3747 @chapter Expressions
3751 @cindex numeric values
3752 An @dfn{expression} specifies an address or numeric value.
3753 Whitespace may precede and/or follow an expression.
3755 The result of an expression must be an absolute number, or else an offset into
3756 a particular section. If an expression is not absolute, and there is not
3757 enough information when @command{@value{AS}} sees the expression to know its
3758 section, a second pass over the source program might be necessary to interpret
3759 the expression---but the second pass is currently not implemented.
3760 @command{@value{AS}} aborts with an error message in this situation.
3763 * Empty Exprs:: Empty Expressions
3764 * Integer Exprs:: Integer Expressions
3768 @section Empty Expressions
3770 @cindex empty expressions
3771 @cindex expressions, empty
3772 An empty expression has no value: it is just whitespace or null.
3773 Wherever an absolute expression is required, you may omit the
3774 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3775 is compatible with other assemblers.
3778 @section Integer Expressions
3780 @cindex integer expressions
3781 @cindex expressions, integer
3782 An @dfn{integer expression} is one or more @emph{arguments} delimited
3783 by @emph{operators}.
3786 * Arguments:: Arguments
3787 * Operators:: Operators
3788 * Prefix Ops:: Prefix Operators
3789 * Infix Ops:: Infix Operators
3793 @subsection Arguments
3795 @cindex expression arguments
3796 @cindex arguments in expressions
3797 @cindex operands in expressions
3798 @cindex arithmetic operands
3799 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3800 contexts arguments are sometimes called ``arithmetic operands''. In
3801 this manual, to avoid confusing them with the ``instruction operands'' of
3802 the machine language, we use the term ``argument'' to refer to parts of
3803 expressions only, reserving the word ``operand'' to refer only to machine
3804 instruction operands.
3806 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3807 @var{section} is one of text, data, bss, absolute,
3808 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3811 Numbers are usually integers.
3813 A number can be a flonum or bignum. In this case, you are warned
3814 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3815 these 32 bits are an integer. You may write integer-manipulating
3816 instructions that act on exotic constants, compatible with other
3819 @cindex subexpressions
3820 Subexpressions are a left parenthesis @samp{(} followed by an integer
3821 expression, followed by a right parenthesis @samp{)}; or a prefix
3822 operator followed by an argument.
3825 @subsection Operators
3827 @cindex operators, in expressions
3828 @cindex arithmetic functions
3829 @cindex functions, in expressions
3830 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3831 operators are followed by an argument. Infix operators appear
3832 between their arguments. Operators may be preceded and/or followed by
3836 @subsection Prefix Operator
3838 @cindex prefix operators
3839 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3840 one argument, which must be absolute.
3842 @c the tex/end tex stuff surrounding this small table is meant to make
3843 @c it align, on the printed page, with the similar table in the next
3844 @c section (which is inside an enumerate).
3846 \global\advance\leftskip by \itemindent
3851 @dfn{Negation}. Two's complement negation.
3853 @dfn{Complementation}. Bitwise not.
3857 \global\advance\leftskip by -\itemindent
3861 @subsection Infix Operators
3863 @cindex infix operators
3864 @cindex operators, permitted arguments
3865 @dfn{Infix operators} take two arguments, one on either side. Operators
3866 have precedence, but operations with equal precedence are performed left
3867 to right. Apart from @code{+} or @option{-}, both arguments must be
3868 absolute, and the result is absolute.
3871 @cindex operator precedence
3872 @cindex precedence of operators
3879 @dfn{Multiplication}.
3882 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3888 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3891 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3895 Intermediate precedence
3900 @dfn{Bitwise Inclusive Or}.
3906 @dfn{Bitwise Exclusive Or}.
3909 @dfn{Bitwise Or Not}.
3916 @cindex addition, permitted arguments
3917 @cindex plus, permitted arguments
3918 @cindex arguments for addition
3920 @dfn{Addition}. If either argument is absolute, the result has the section of
3921 the other argument. You may not add together arguments from different
3924 @cindex subtraction, permitted arguments
3925 @cindex minus, permitted arguments
3926 @cindex arguments for subtraction
3928 @dfn{Subtraction}. If the right argument is absolute, the
3929 result has the section of the left argument.
3930 If both arguments are in the same section, the result is absolute.
3931 You may not subtract arguments from different sections.
3932 @c FIXME is there still something useful to say about undefined - undefined ?
3934 @cindex comparison expressions
3935 @cindex expressions, comparison
3940 @dfn{Is Not Equal To}
3944 @dfn{Is Greater Than}
3946 @dfn{Is Greater Than Or Equal To}
3948 @dfn{Is Less Than Or Equal To}
3950 The comparison operators can be used as infix operators. A true results has a
3951 value of -1 whereas a false result has a value of 0. Note, these operators
3952 perform signed comparisons.
3955 @item Lowest Precedence
3964 These two logical operations can be used to combine the results of sub
3965 expressions. Note, unlike the comparison operators a true result returns a
3966 value of 1 but a false results does still return 0. Also note that the logical
3967 or operator has a slightly lower precedence than logical and.
3972 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3973 address; you can only have a defined section in one of the two arguments.
3976 @chapter Assembler Directives
3978 @cindex directives, machine independent
3979 @cindex pseudo-ops, machine independent
3980 @cindex machine independent directives
3981 All assembler directives have names that begin with a period (@samp{.}).
3982 The rest of the name is letters, usually in lower case.
3984 This chapter discusses directives that are available regardless of the
3985 target machine configuration for the @sc{gnu} assembler.
3987 Some machine configurations provide additional directives.
3988 @xref{Machine Dependencies}.
3991 @ifset machine-directives
3992 @xref{Machine Dependencies}, for additional directives.
3997 * Abort:: @code{.abort}
3999 * ABORT (COFF):: @code{.ABORT}
4002 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
4003 * Altmacro:: @code{.altmacro}
4004 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4005 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4006 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
4007 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, @code{.bundle_lock}, @code{.bundle_unlock}
4008 * Byte:: @code{.byte @var{expressions}}
4009 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4010 * Comm:: @code{.comm @var{symbol} , @var{length} }
4011 * Data:: @code{.data @var{subsection}}
4013 * Def:: @code{.def @var{name}}
4016 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4022 * Double:: @code{.double @var{flonums}}
4023 * Eject:: @code{.eject}
4024 * Else:: @code{.else}
4025 * Elseif:: @code{.elseif}
4028 * Endef:: @code{.endef}
4031 * Endfunc:: @code{.endfunc}
4032 * Endif:: @code{.endif}
4033 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4034 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4035 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4037 * Error:: @code{.error @var{string}}
4038 * Exitm:: @code{.exitm}
4039 * Extern:: @code{.extern}
4040 * Fail:: @code{.fail}
4041 * File:: @code{.file}
4042 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4043 * Float:: @code{.float @var{flonums}}
4044 * Func:: @code{.func}
4045 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4047 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4048 * Hidden:: @code{.hidden @var{names}}
4051 * hword:: @code{.hword @var{expressions}}
4052 * Ident:: @code{.ident}
4053 * If:: @code{.if @var{absolute expression}}
4054 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4055 * Include:: @code{.include "@var{file}"}
4056 * Int:: @code{.int @var{expressions}}
4058 * Internal:: @code{.internal @var{names}}
4061 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4062 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4063 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4064 * Lflags:: @code{.lflags}
4065 @ifclear no-line-dir
4066 * Line:: @code{.line @var{line-number}}
4069 * Linkonce:: @code{.linkonce [@var{type}]}
4070 * List:: @code{.list}
4071 * Ln:: @code{.ln @var{line-number}}
4072 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4073 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4075 * Local:: @code{.local @var{names}}
4078 * Long:: @code{.long @var{expressions}}
4080 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4083 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4084 * MRI:: @code{.mri @var{val}}
4085 * Noaltmacro:: @code{.noaltmacro}
4086 * Nolist:: @code{.nolist}
4087 * Octa:: @code{.octa @var{bignums}}
4088 * Offset:: @code{.offset @var{loc}}
4089 * Org:: @code{.org @var{new-lc}, @var{fill}}
4090 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4092 * PopSection:: @code{.popsection}
4093 * Previous:: @code{.previous}
4096 * Print:: @code{.print @var{string}}
4098 * Protected:: @code{.protected @var{names}}
4101 * Psize:: @code{.psize @var{lines}, @var{columns}}
4102 * Purgem:: @code{.purgem @var{name}}
4104 * PushSection:: @code{.pushsection @var{name}}
4107 * Quad:: @code{.quad @var{bignums}}
4108 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4109 * Rept:: @code{.rept @var{count}}
4110 * Sbttl:: @code{.sbttl "@var{subheading}"}
4112 * Scl:: @code{.scl @var{class}}
4115 * Section:: @code{.section @var{name}[, @var{flags}]}
4118 * Set:: @code{.set @var{symbol}, @var{expression}}
4119 * Short:: @code{.short @var{expressions}}
4120 * Single:: @code{.single @var{flonums}}
4122 * Size:: @code{.size [@var{name} , @var{expression}]}
4124 @ifclear no-space-dir
4125 * Skip:: @code{.skip @var{size} , @var{fill}}
4128 * Sleb128:: @code{.sleb128 @var{expressions}}
4129 @ifclear no-space-dir
4130 * Space:: @code{.space @var{size} , @var{fill}}
4133 * Stab:: @code{.stabd, .stabn, .stabs}
4136 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4137 * Struct:: @code{.struct @var{expression}}
4139 * SubSection:: @code{.subsection}
4140 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4144 * Tag:: @code{.tag @var{structname}}
4147 * Text:: @code{.text @var{subsection}}
4148 * Title:: @code{.title "@var{heading}"}
4150 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4153 * Uleb128:: @code{.uleb128 @var{expressions}}
4155 * Val:: @code{.val @var{addr}}
4159 * Version:: @code{.version "@var{string}"}
4160 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4161 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4164 * Warning:: @code{.warning @var{string}}
4165 * Weak:: @code{.weak @var{names}}
4166 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4167 * Word:: @code{.word @var{expressions}}
4168 * Deprecated:: Deprecated Directives
4172 @section @code{.abort}
4174 @cindex @code{abort} directive
4175 @cindex stopping the assembly
4176 This directive stops the assembly immediately. It is for
4177 compatibility with other assemblers. The original idea was that the
4178 assembly language source would be piped into the assembler. If the sender
4179 of the source quit, it could use this directive tells @command{@value{AS}} to
4180 quit also. One day @code{.abort} will not be supported.
4184 @section @code{.ABORT} (COFF)
4186 @cindex @code{ABORT} directive
4187 When producing COFF output, @command{@value{AS}} accepts this directive as a
4188 synonym for @samp{.abort}.
4191 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
4197 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4199 @cindex padding the location counter
4200 @cindex @code{align} directive
4201 Pad the location counter (in the current subsection) to a particular storage
4202 boundary. The first expression (which must be absolute) is the alignment
4203 required, as described below.
4205 The second expression (also absolute) gives the fill value to be stored in the
4206 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4207 padding bytes are normally zero. However, on some systems, if the section is
4208 marked as containing code and the fill value is omitted, the space is filled
4209 with no-op instructions.
4211 The third expression is also absolute, and is also optional. If it is present,
4212 it is the maximum number of bytes that should be skipped by this alignment
4213 directive. If doing the alignment would require skipping more bytes than the
4214 specified maximum, then the alignment is not done at all. You can omit the
4215 fill value (the second argument) entirely by simply using two commas after the
4216 required alignment; this can be useful if you want the alignment to be filled
4217 with no-op instructions when appropriate.
4219 The way the required alignment is specified varies from system to system.
4220 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
4221 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4222 alignment request in bytes. For example @samp{.align 8} advances
4223 the location counter until it is a multiple of 8. If the location counter
4224 is already a multiple of 8, no change is needed. For the tic54x, the
4225 first expression is the alignment request in words.
4227 For other systems, including ppc, i386 using a.out format, arm and
4228 strongarm, it is the
4229 number of low-order zero bits the location counter must have after
4230 advancement. For example @samp{.align 3} advances the location
4231 counter until it a multiple of 8. If the location counter is already a
4232 multiple of 8, no change is needed.
4234 This inconsistency is due to the different behaviors of the various
4235 native assemblers for these systems which GAS must emulate.
4236 GAS also provides @code{.balign} and @code{.p2align} directives,
4237 described later, which have a consistent behavior across all
4238 architectures (but are specific to GAS).
4241 @section @code{.altmacro}
4242 Enable alternate macro mode, enabling:
4245 @item LOCAL @var{name} [ , @dots{} ]
4246 One additional directive, @code{LOCAL}, is available. It is used to
4247 generate a string replacement for each of the @var{name} arguments, and
4248 replace any instances of @var{name} in each macro expansion. The
4249 replacement string is unique in the assembly, and different for each
4250 separate macro expansion. @code{LOCAL} allows you to write macros that
4251 define symbols, without fear of conflict between separate macro expansions.
4253 @item String delimiters
4254 You can write strings delimited in these other ways besides
4255 @code{"@var{string}"}:
4258 @item '@var{string}'
4259 You can delimit strings with single-quote characters.
4261 @item <@var{string}>
4262 You can delimit strings with matching angle brackets.
4265 @item single-character string escape
4266 To include any single character literally in a string (even if the
4267 character would otherwise have some special meaning), you can prefix the
4268 character with @samp{!} (an exclamation mark). For example, you can
4269 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4271 @item Expression results as strings
4272 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4273 and use the result as a string.
4277 @section @code{.ascii "@var{string}"}@dots{}
4279 @cindex @code{ascii} directive
4280 @cindex string literals
4281 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4282 separated by commas. It assembles each string (with no automatic
4283 trailing zero byte) into consecutive addresses.
4286 @section @code{.asciz "@var{string}"}@dots{}
4288 @cindex @code{asciz} directive
4289 @cindex zero-terminated strings
4290 @cindex null-terminated strings
4291 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4292 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4295 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4297 @cindex padding the location counter given number of bytes
4298 @cindex @code{balign} directive
4299 Pad the location counter (in the current subsection) to a particular
4300 storage boundary. The first expression (which must be absolute) is the
4301 alignment request in bytes. For example @samp{.balign 8} advances
4302 the location counter until it is a multiple of 8. If the location counter
4303 is already a multiple of 8, no change is needed.
4305 The second expression (also absolute) gives the fill value to be stored in the
4306 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4307 padding bytes are normally zero. However, on some systems, if the section is
4308 marked as containing code and the fill value is omitted, the space is filled
4309 with no-op instructions.
4311 The third expression is also absolute, and is also optional. If it is present,
4312 it is the maximum number of bytes that should be skipped by this alignment
4313 directive. If doing the alignment would require skipping more bytes than the
4314 specified maximum, then the alignment is not done at all. You can omit the
4315 fill value (the second argument) entirely by simply using two commas after the
4316 required alignment; this can be useful if you want the alignment to be filled
4317 with no-op instructions when appropriate.
4319 @cindex @code{balignw} directive
4320 @cindex @code{balignl} directive
4321 The @code{.balignw} and @code{.balignl} directives are variants of the
4322 @code{.balign} directive. The @code{.balignw} directive treats the fill
4323 pattern as a two byte word value. The @code{.balignl} directives treats the
4324 fill pattern as a four byte longword value. For example, @code{.balignw
4325 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4326 filled in with the value 0x368d (the exact placement of the bytes depends upon
4327 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4330 @node Bundle directives
4331 @section @code{.bundle_align_mode @var{abs-expr}}
4332 @cindex @code{bundle_align_mode} directive
4334 @cindex instruction bundle
4335 @cindex aligned instruction bundle
4336 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4337 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4338 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4339 disabled (which is the default state). If the argument it not zero, it
4340 gives the size of an instruction bundle as a power of two (as for the
4341 @code{.p2align} directive, @pxref{P2align}).
4343 For some targets, it's an ABI requirement that no instruction may span a
4344 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4345 instructions that starts on an aligned boundary. For example, if
4346 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4347 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4348 effect, no single instruction may span a boundary between bundles. If an
4349 instruction would start too close to the end of a bundle for the length of
4350 that particular instruction to fit within the bundle, then the space at the
4351 end of that bundle is filled with no-op instructions so the instruction
4352 starts in the next bundle. As a corollary, it's an error if any single
4353 instruction's encoding is longer than the bundle size.
4355 @section @code{.bundle_lock} and @code{.bundle_unlock}
4356 @cindex @code{bundle_lock} directive
4357 @cindex @code{bundle_unlock} directive
4358 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4359 allow explicit control over instruction bundle padding. These directives
4360 are only valid when @code{.bundle_align_mode} has been used to enable
4361 aligned instruction bundle mode. It's an error if they appear when
4362 @code{.bundle_align_mode} has not been used at all, or when the last
4363 directive was @w{@code{.bundle_align_mode 0}}.
4365 @cindex bundle-locked
4366 For some targets, it's an ABI requirement that certain instructions may
4367 appear only as part of specified permissible sequences of multiple
4368 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4369 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4370 instruction sequence. For purposes of aligned instruction bundle mode, a
4371 sequence starting with @code{.bundle_lock} and ending with
4372 @code{.bundle_unlock} is treated as a single instruction. That is, the
4373 entire sequence must fit into a single bundle and may not span a bundle
4374 boundary. If necessary, no-op instructions will be inserted before the
4375 first instruction of the sequence so that the whole sequence starts on an
4376 aligned bundle boundary. It's an error if the sequence is longer than the
4379 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4380 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4381 nested. That is, a second @code{.bundle_lock} directive before the next
4382 @code{.bundle_unlock} directive has no effect except that it must be
4383 matched by another closing @code{.bundle_unlock} so that there is the
4384 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4387 @section @code{.byte @var{expressions}}
4389 @cindex @code{byte} directive
4390 @cindex integers, one byte
4391 @code{.byte} expects zero or more expressions, separated by commas.
4392 Each expression is assembled into the next byte.
4394 @node CFI directives
4395 @section @code{.cfi_sections @var{section_list}}
4396 @cindex @code{cfi_sections} directive
4397 @code{.cfi_sections} may be used to specify whether CFI directives
4398 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4399 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4400 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4401 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4402 directive is not used is @code{.cfi_sections .eh_frame}.
4404 @section @code{.cfi_startproc [simple]}
4405 @cindex @code{cfi_startproc} directive
4406 @code{.cfi_startproc} is used at the beginning of each function that
4407 should have an entry in @code{.eh_frame}. It initializes some internal
4408 data structures. Don't forget to close the function by
4409 @code{.cfi_endproc}.
4411 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4412 it also emits some architecture dependent initial CFI instructions.
4414 @section @code{.cfi_endproc}
4415 @cindex @code{cfi_endproc} directive
4416 @code{.cfi_endproc} is used at the end of a function where it closes its
4417 unwind entry previously opened by
4418 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4420 @section @code{.cfi_personality @var{encoding} [, @var{exp}]}
4421 @code{.cfi_personality} defines personality routine and its encoding.
4422 @var{encoding} must be a constant determining how the personality
4423 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4424 argument is not present, otherwise second argument should be
4425 a constant or a symbol name. When using indirect encodings,
4426 the symbol provided should be the location where personality
4427 can be loaded from, not the personality routine itself.
4428 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4429 no personality routine.
4431 @section @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4432 @code{.cfi_lsda} defines LSDA and its encoding.
4433 @var{encoding} must be a constant determining how the LSDA
4434 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4435 argument is not present, otherwise second argument should be a constant
4436 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4439 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4440 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4441 address from @var{register} and add @var{offset} to it}.
4443 @section @code{.cfi_def_cfa_register @var{register}}
4444 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4445 now on @var{register} will be used instead of the old one. Offset
4448 @section @code{.cfi_def_cfa_offset @var{offset}}
4449 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4450 remains the same, but @var{offset} is new. Note that it is the
4451 absolute offset that will be added to a defined register to compute
4454 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4455 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4456 value that is added/substracted from the previous offset.
4458 @section @code{.cfi_offset @var{register}, @var{offset}}
4459 Previous value of @var{register} is saved at offset @var{offset} from
4462 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4463 Previous value of @var{register} is saved at offset @var{offset} from
4464 the current CFA register. This is transformed to @code{.cfi_offset}
4465 using the known displacement of the CFA register from the CFA.
4466 This is often easier to use, because the number will match the
4467 code it's annotating.
4469 @section @code{.cfi_register @var{register1}, @var{register2}}
4470 Previous value of @var{register1} is saved in register @var{register2}.
4472 @section @code{.cfi_restore @var{register}}
4473 @code{.cfi_restore} says that the rule for @var{register} is now the
4474 same as it was at the beginning of the function, after all initial
4475 instruction added by @code{.cfi_startproc} were executed.
4477 @section @code{.cfi_undefined @var{register}}
4478 From now on the previous value of @var{register} can't be restored anymore.
4480 @section @code{.cfi_same_value @var{register}}
4481 Current value of @var{register} is the same like in the previous frame,
4482 i.e. no restoration needed.
4484 @section @code{.cfi_remember_state},
4485 First save all current rules for all registers by @code{.cfi_remember_state},
4486 then totally screw them up by subsequent @code{.cfi_*} directives and when
4487 everything is hopelessly bad, use @code{.cfi_restore_state} to restore
4488 the previous saved state.
4490 @section @code{.cfi_return_column @var{register}}
4491 Change return column @var{register}, i.e. the return address is either
4492 directly in @var{register} or can be accessed by rules for @var{register}.
4494 @section @code{.cfi_signal_frame}
4495 Mark current function as signal trampoline.
4497 @section @code{.cfi_window_save}
4498 SPARC register window has been saved.
4500 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4501 Allows the user to add arbitrary bytes to the unwind info. One
4502 might use this to add OS-specific CFI opcodes, or generic CFI
4503 opcodes that GAS does not yet support.
4505 @section @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4506 The current value of @var{register} is @var{label}. The value of @var{label}
4507 will be encoded in the output file according to @var{encoding}; see the
4508 description of @code{.cfi_personality} for details on this encoding.
4510 The usefulness of equating a register to a fixed label is probably
4511 limited to the return address register. Here, it can be useful to
4512 mark a code segment that has only one return address which is reached
4513 by a direct branch and no copy of the return address exists in memory
4514 or another register.
4517 @section @code{.comm @var{symbol} , @var{length} }
4519 @cindex @code{comm} directive
4520 @cindex symbol, common
4521 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4522 common symbol in one object file may be merged with a defined or common symbol
4523 of the same name in another object file. If @code{@value{LD}} does not see a
4524 definition for the symbol--just one or more common symbols--then it will
4525 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4526 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4527 the same name, and they do not all have the same size, it will allocate space
4528 using the largest size.
4531 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4532 an optional third argument. This is the desired alignment of the symbol,
4533 specified for ELF as a byte boundary (for example, an alignment of 16 means
4534 that the least significant 4 bits of the address should be zero), and for PE
4535 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4536 boundary). The alignment must be an absolute expression, and it must be a
4537 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4538 common symbol, it will use the alignment when placing the symbol. If no
4539 alignment is specified, @command{@value{AS}} will set the alignment to the
4540 largest power of two less than or equal to the size of the symbol, up to a
4541 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4542 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4543 @samp{--section-alignment} option; image file sections in PE are aligned to
4544 multiples of 4096, which is far too large an alignment for ordinary variables.
4545 It is rather the default alignment for (non-debug) sections within object
4546 (@samp{*.o}) files, which are less strictly aligned.}.
4550 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4551 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4555 @section @code{.data @var{subsection}}
4557 @cindex @code{data} directive
4558 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4559 end of the data subsection numbered @var{subsection} (which is an
4560 absolute expression). If @var{subsection} is omitted, it defaults
4565 @section @code{.def @var{name}}
4567 @cindex @code{def} directive
4568 @cindex COFF symbols, debugging
4569 @cindex debugging COFF symbols
4570 Begin defining debugging information for a symbol @var{name}; the
4571 definition extends until the @code{.endef} directive is encountered.
4574 This directive is only observed when @command{@value{AS}} is configured for COFF
4575 format output; when producing @code{b.out}, @samp{.def} is recognized,
4582 @section @code{.desc @var{symbol}, @var{abs-expression}}
4584 @cindex @code{desc} directive
4585 @cindex COFF symbol descriptor
4586 @cindex symbol descriptor, COFF
4587 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4588 to the low 16 bits of an absolute expression.
4591 The @samp{.desc} directive is not available when @command{@value{AS}} is
4592 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4593 object format. For the sake of compatibility, @command{@value{AS}} accepts
4594 it, but produces no output, when configured for COFF.
4600 @section @code{.dim}
4602 @cindex @code{dim} directive
4603 @cindex COFF auxiliary symbol information
4604 @cindex auxiliary symbol information, COFF
4605 This directive is generated by compilers to include auxiliary debugging
4606 information in the symbol table. It is only permitted inside
4607 @code{.def}/@code{.endef} pairs.
4610 @samp{.dim} is only meaningful when generating COFF format output; when
4611 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4617 @section @code{.double @var{flonums}}
4619 @cindex @code{double} directive
4620 @cindex floating point numbers (double)
4621 @code{.double} expects zero or more flonums, separated by commas. It
4622 assembles floating point numbers.
4624 The exact kind of floating point numbers emitted depends on how
4625 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4629 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4630 in @sc{ieee} format.
4635 @section @code{.eject}
4637 @cindex @code{eject} directive
4638 @cindex new page, in listings
4639 @cindex page, in listings
4640 @cindex listing control: new page
4641 Force a page break at this point, when generating assembly listings.
4644 @section @code{.else}
4646 @cindex @code{else} directive
4647 @code{.else} is part of the @command{@value{AS}} support for conditional
4648 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
4649 of code to be assembled if the condition for the preceding @code{.if}
4653 @section @code{.elseif}
4655 @cindex @code{elseif} directive
4656 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4657 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
4658 @code{.if} block that would otherwise fill the entire @code{.else} section.
4661 @section @code{.end}
4663 @cindex @code{end} directive
4664 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4665 process anything in the file past the @code{.end} directive.
4669 @section @code{.endef}
4671 @cindex @code{endef} directive
4672 This directive flags the end of a symbol definition begun with
4676 @samp{.endef} is only meaningful when generating COFF format output; if
4677 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4678 directive but ignores it.
4683 @section @code{.endfunc}
4684 @cindex @code{endfunc} directive
4685 @code{.endfunc} marks the end of a function specified with @code{.func}.
4688 @section @code{.endif}
4690 @cindex @code{endif} directive
4691 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4692 it marks the end of a block of code that is only assembled
4693 conditionally. @xref{If,,@code{.if}}.
4696 @section @code{.equ @var{symbol}, @var{expression}}
4698 @cindex @code{equ} directive
4699 @cindex assigning values to symbols
4700 @cindex symbols, assigning values to
4701 This directive sets the value of @var{symbol} to @var{expression}.
4702 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
4705 The syntax for @code{equ} on the HPPA is
4706 @samp{@var{symbol} .equ @var{expression}}.
4710 The syntax for @code{equ} on the Z80 is
4711 @samp{@var{symbol} equ @var{expression}}.
4712 On the Z80 it is an eror if @var{symbol} is already defined,
4713 but the symbol is not protected from later redefinition.
4714 Compare @ref{Equiv}.
4718 @section @code{.equiv @var{symbol}, @var{expression}}
4719 @cindex @code{equiv} directive
4720 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4721 the assembler will signal an error if @var{symbol} is already defined. Note a
4722 symbol which has been referenced but not actually defined is considered to be
4725 Except for the contents of the error message, this is roughly equivalent to
4732 plus it protects the symbol from later redefinition.
4735 @section @code{.eqv @var{symbol}, @var{expression}}
4736 @cindex @code{eqv} directive
4737 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4738 evaluate the expression or any part of it immediately. Instead each time
4739 the resulting symbol is used in an expression, a snapshot of its current
4743 @section @code{.err}
4744 @cindex @code{err} directive
4745 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4746 message and, unless the @option{-Z} option was used, it will not generate an
4747 object file. This can be used to signal an error in conditionally compiled code.
4750 @section @code{.error "@var{string}"}
4751 @cindex error directive
4753 Similarly to @code{.err}, this directive emits an error, but you can specify a
4754 string that will be emitted as the error message. If you don't specify the
4755 message, it defaults to @code{".error directive invoked in source file"}.
4756 @xref{Errors, ,Error and Warning Messages}.
4759 .error "This code has not been assembled and tested."
4763 @section @code{.exitm}
4764 Exit early from the current macro definition. @xref{Macro}.
4767 @section @code{.extern}
4769 @cindex @code{extern} directive
4770 @code{.extern} is accepted in the source program---for compatibility
4771 with other assemblers---but it is ignored. @command{@value{AS}} treats
4772 all undefined symbols as external.
4775 @section @code{.fail @var{expression}}
4777 @cindex @code{fail} directive
4778 Generates an error or a warning. If the value of the @var{expression} is 500
4779 or more, @command{@value{AS}} will print a warning message. If the value is less
4780 than 500, @command{@value{AS}} will print an error message. The message will
4781 include the value of @var{expression}. This can occasionally be useful inside
4782 complex nested macros or conditional assembly.
4785 @section @code{.file}
4786 @cindex @code{file} directive
4788 @ifclear no-file-dir
4789 There are two different versions of the @code{.file} directive. Targets
4790 that support DWARF2 line number information use the DWARF2 version of
4791 @code{.file}. Other targets use the default version.
4793 @subheading Default Version
4795 @cindex logical file name
4796 @cindex file name, logical
4797 This version of the @code{.file} directive tells @command{@value{AS}} that we
4798 are about to start a new logical file. The syntax is:
4804 @var{string} is the new file name. In general, the filename is
4805 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4806 to specify an empty file name, you must give the quotes--@code{""}. This
4807 statement may go away in future: it is only recognized to be compatible with
4808 old @command{@value{AS}} programs.
4810 @subheading DWARF2 Version
4813 When emitting DWARF2 line number information, @code{.file} assigns filenames
4814 to the @code{.debug_line} file name table. The syntax is:
4817 .file @var{fileno} @var{filename}
4820 The @var{fileno} operand should be a unique positive integer to use as the
4821 index of the entry in the table. The @var{filename} operand is a C string
4824 The detail of filename indices is exposed to the user because the filename
4825 table is shared with the @code{.debug_info} section of the DWARF2 debugging
4826 information, and thus the user must know the exact indices that table
4830 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4832 @cindex @code{fill} directive
4833 @cindex writing patterns in memory
4834 @cindex patterns, writing in memory
4835 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4836 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4837 may be zero or more. @var{Size} may be zero or more, but if it is
4838 more than 8, then it is deemed to have the value 8, compatible with
4839 other people's assemblers. The contents of each @var{repeat} bytes
4840 is taken from an 8-byte number. The highest order 4 bytes are
4841 zero. The lowest order 4 bytes are @var{value} rendered in the
4842 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4843 Each @var{size} bytes in a repetition is taken from the lowest order
4844 @var{size} bytes of this number. Again, this bizarre behavior is
4845 compatible with other people's assemblers.
4847 @var{size} and @var{value} are optional.
4848 If the second comma and @var{value} are absent, @var{value} is
4849 assumed zero. If the first comma and following tokens are absent,
4850 @var{size} is assumed to be 1.
4853 @section @code{.float @var{flonums}}
4855 @cindex floating point numbers (single)
4856 @cindex @code{float} directive
4857 This directive assembles zero or more flonums, separated by commas. It
4858 has the same effect as @code{.single}.
4860 The exact kind of floating point numbers emitted depends on how
4861 @command{@value{AS}} is configured.
4862 @xref{Machine Dependencies}.
4866 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4867 in @sc{ieee} format.
4872 @section @code{.func @var{name}[,@var{label}]}
4873 @cindex @code{func} directive
4874 @code{.func} emits debugging information to denote function @var{name}, and
4875 is ignored unless the file is assembled with debugging enabled.
4876 Only @samp{--gstabs[+]} is currently supported.
4877 @var{label} is the entry point of the function and if omitted @var{name}
4878 prepended with the @samp{leading char} is used.
4879 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4880 All functions are currently defined to have @code{void} return type.
4881 The function must be terminated with @code{.endfunc}.
4884 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4886 @cindex @code{global} directive
4887 @cindex symbol, making visible to linker
4888 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4889 @var{symbol} in your partial program, its value is made available to
4890 other partial programs that are linked with it. Otherwise,
4891 @var{symbol} takes its attributes from a symbol of the same name
4892 from another file linked into the same program.
4894 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4895 compatibility with other assemblers.
4898 On the HPPA, @code{.global} is not always enough to make it accessible to other
4899 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4900 @xref{HPPA Directives, ,HPPA Assembler Directives}.
4905 @section @code{.gnu_attribute @var{tag},@var{value}}
4906 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
4909 @section @code{.hidden @var{names}}
4911 @cindex @code{hidden} directive
4913 This is one of the ELF visibility directives. The other two are
4914 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4915 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4917 This directive overrides the named symbols default visibility (which is set by
4918 their binding: local, global or weak). The directive sets the visibility to
4919 @code{hidden} which means that the symbols are not visible to other components.
4920 Such symbols are always considered to be @code{protected} as well.
4924 @section @code{.hword @var{expressions}}
4926 @cindex @code{hword} directive
4927 @cindex integers, 16-bit
4928 @cindex numbers, 16-bit
4929 @cindex sixteen bit integers
4930 This expects zero or more @var{expressions}, and emits
4931 a 16 bit number for each.
4934 This directive is a synonym for @samp{.short}; depending on the target
4935 architecture, it may also be a synonym for @samp{.word}.
4939 This directive is a synonym for @samp{.short}.
4942 This directive is a synonym for both @samp{.short} and @samp{.word}.
4947 @section @code{.ident}
4949 @cindex @code{ident} directive
4951 This directive is used by some assemblers to place tags in object files. The
4952 behavior of this directive varies depending on the target. When using the
4953 a.out object file format, @command{@value{AS}} simply accepts the directive for
4954 source-file compatibility with existing assemblers, but does not emit anything
4955 for it. When using COFF, comments are emitted to the @code{.comment} or
4956 @code{.rdata} section, depending on the target. When using ELF, comments are
4957 emitted to the @code{.comment} section.
4960 @section @code{.if @var{absolute expression}}
4962 @cindex conditional assembly
4963 @cindex @code{if} directive
4964 @code{.if} marks the beginning of a section of code which is only
4965 considered part of the source program being assembled if the argument
4966 (which must be an @var{absolute expression}) is non-zero. The end of
4967 the conditional section of code must be marked by @code{.endif}
4968 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4969 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4970 If you have several conditions to check, @code{.elseif} may be used to avoid
4971 nesting blocks if/else within each subsequent @code{.else} block.
4973 The following variants of @code{.if} are also supported:
4975 @cindex @code{ifdef} directive
4976 @item .ifdef @var{symbol}
4977 Assembles the following section of code if the specified @var{symbol}
4978 has been defined. Note a symbol which has been referenced but not yet defined
4979 is considered to be undefined.
4981 @cindex @code{ifb} directive
4982 @item .ifb @var{text}
4983 Assembles the following section of code if the operand is blank (empty).
4985 @cindex @code{ifc} directive
4986 @item .ifc @var{string1},@var{string2}
4987 Assembles the following section of code if the two strings are the same. The
4988 strings may be optionally quoted with single quotes. If they are not quoted,
4989 the first string stops at the first comma, and the second string stops at the
4990 end of the line. Strings which contain whitespace should be quoted. The
4991 string comparison is case sensitive.
4993 @cindex @code{ifeq} directive
4994 @item .ifeq @var{absolute expression}
4995 Assembles the following section of code if the argument is zero.
4997 @cindex @code{ifeqs} directive
4998 @item .ifeqs @var{string1},@var{string2}
4999 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5001 @cindex @code{ifge} directive
5002 @item .ifge @var{absolute expression}
5003 Assembles the following section of code if the argument is greater than or
5006 @cindex @code{ifgt} directive
5007 @item .ifgt @var{absolute expression}
5008 Assembles the following section of code if the argument is greater than zero.
5010 @cindex @code{ifle} directive
5011 @item .ifle @var{absolute expression}
5012 Assembles the following section of code if the argument is less than or equal
5015 @cindex @code{iflt} directive
5016 @item .iflt @var{absolute expression}
5017 Assembles the following section of code if the argument is less than zero.
5019 @cindex @code{ifnb} directive
5020 @item .ifnb @var{text}
5021 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5022 following section of code if the operand is non-blank (non-empty).
5024 @cindex @code{ifnc} directive
5025 @item .ifnc @var{string1},@var{string2}.
5026 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5027 following section of code if the two strings are not the same.
5029 @cindex @code{ifndef} directive
5030 @cindex @code{ifnotdef} directive
5031 @item .ifndef @var{symbol}
5032 @itemx .ifnotdef @var{symbol}
5033 Assembles the following section of code if the specified @var{symbol}
5034 has not been defined. Both spelling variants are equivalent. Note a symbol
5035 which has been referenced but not yet defined is considered to be undefined.
5037 @cindex @code{ifne} directive
5038 @item .ifne @var{absolute expression}
5039 Assembles the following section of code if the argument is not equal to zero
5040 (in other words, this is equivalent to @code{.if}).
5042 @cindex @code{ifnes} directive
5043 @item .ifnes @var{string1},@var{string2}
5044 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5045 following section of code if the two strings are not the same.
5049 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5051 @cindex @code{incbin} directive
5052 @cindex binary files, including
5053 The @code{incbin} directive includes @var{file} verbatim at the current
5054 location. You can control the search paths used with the @samp{-I} command-line
5055 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5058 The @var{skip} argument skips a number of bytes from the start of the
5059 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5060 read. Note that the data is not aligned in any way, so it is the user's
5061 responsibility to make sure that proper alignment is provided both before and
5062 after the @code{incbin} directive.
5065 @section @code{.include "@var{file}"}
5067 @cindex @code{include} directive
5068 @cindex supporting files, including
5069 @cindex files, including
5070 This directive provides a way to include supporting files at specified
5071 points in your source program. The code from @var{file} is assembled as
5072 if it followed the point of the @code{.include}; when the end of the
5073 included file is reached, assembly of the original file continues. You
5074 can control the search paths used with the @samp{-I} command-line option
5075 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5079 @section @code{.int @var{expressions}}
5081 @cindex @code{int} directive
5082 @cindex integers, 32-bit
5083 Expect zero or more @var{expressions}, of any section, separated by commas.
5084 For each expression, emit a number that, at run time, is the value of that
5085 expression. The byte order and bit size of the number depends on what kind
5086 of target the assembly is for.
5090 On most forms of the H8/300, @code{.int} emits 16-bit
5091 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5098 @section @code{.internal @var{names}}
5100 @cindex @code{internal} directive
5102 This is one of the ELF visibility directives. The other two are
5103 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5104 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5106 This directive overrides the named symbols default visibility (which is set by
5107 their binding: local, global or weak). The directive sets the visibility to
5108 @code{internal} which means that the symbols are considered to be @code{hidden}
5109 (i.e., not visible to other components), and that some extra, processor specific
5110 processing must also be performed upon the symbols as well.
5114 @section @code{.irp @var{symbol},@var{values}}@dots{}
5116 @cindex @code{irp} directive
5117 Evaluate a sequence of statements assigning different values to @var{symbol}.
5118 The sequence of statements starts at the @code{.irp} directive, and is
5119 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5120 set to @var{value}, and the sequence of statements is assembled. If no
5121 @var{value} is listed, the sequence of statements is assembled once, with
5122 @var{symbol} set to the null string. To refer to @var{symbol} within the
5123 sequence of statements, use @var{\symbol}.
5125 For example, assembling
5133 is equivalent to assembling
5141 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5144 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5146 @cindex @code{irpc} directive
5147 Evaluate a sequence of statements assigning different values to @var{symbol}.
5148 The sequence of statements starts at the @code{.irpc} directive, and is
5149 terminated by an @code{.endr} directive. For each character in @var{value},
5150 @var{symbol} is set to the character, and the sequence of statements is
5151 assembled. If no @var{value} is listed, the sequence of statements is
5152 assembled once, with @var{symbol} set to the null string. To refer to
5153 @var{symbol} within the sequence of statements, use @var{\symbol}.
5155 For example, assembling
5163 is equivalent to assembling
5171 For some caveats with the spelling of @var{symbol}, see also the discussion
5175 @section @code{.lcomm @var{symbol} , @var{length}}
5177 @cindex @code{lcomm} directive
5178 @cindex local common symbols
5179 @cindex symbols, local common
5180 Reserve @var{length} (an absolute expression) bytes for a local common
5181 denoted by @var{symbol}. The section and value of @var{symbol} are
5182 those of the new local common. The addresses are allocated in the bss
5183 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5184 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5185 not visible to @code{@value{LD}}.
5188 Some targets permit a third argument to be used with @code{.lcomm}. This
5189 argument specifies the desired alignment of the symbol in the bss section.
5193 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5194 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5198 @section @code{.lflags}
5200 @cindex @code{lflags} directive (ignored)
5201 @command{@value{AS}} accepts this directive, for compatibility with other
5202 assemblers, but ignores it.
5204 @ifclear no-line-dir
5206 @section @code{.line @var{line-number}}
5208 @cindex @code{line} directive
5209 @cindex logical line number
5211 Change the logical line number. @var{line-number} must be an absolute
5212 expression. The next line has that logical line number. Therefore any other
5213 statements on the current line (after a statement separator character) are
5214 reported as on logical line number @var{line-number} @minus{} 1. One day
5215 @command{@value{AS}} will no longer support this directive: it is recognized only
5216 for compatibility with existing assembler programs.
5219 Even though this is a directive associated with the @code{a.out} or
5220 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5221 when producing COFF output, and treats @samp{.line} as though it
5222 were the COFF @samp{.ln} @emph{if} it is found outside a
5223 @code{.def}/@code{.endef} pair.
5225 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5226 used by compilers to generate auxiliary symbol information for
5231 @section @code{.linkonce [@var{type}]}
5233 @cindex @code{linkonce} directive
5234 @cindex common sections
5235 Mark the current section so that the linker only includes a single copy of it.
5236 This may be used to include the same section in several different object files,
5237 but ensure that the linker will only include it once in the final output file.
5238 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5239 Duplicate sections are detected based on the section name, so it should be
5242 This directive is only supported by a few object file formats; as of this
5243 writing, the only object file format which supports it is the Portable
5244 Executable format used on Windows NT.
5246 The @var{type} argument is optional. If specified, it must be one of the
5247 following strings. For example:
5251 Not all types may be supported on all object file formats.
5255 Silently discard duplicate sections. This is the default.
5258 Warn if there are duplicate sections, but still keep only one copy.
5261 Warn if any of the duplicates have different sizes.
5264 Warn if any of the duplicates do not have exactly the same contents.
5268 @section @code{.list}
5270 @cindex @code{list} directive
5271 @cindex listing control, turning on
5272 Control (in conjunction with the @code{.nolist} directive) whether or
5273 not assembly listings are generated. These two directives maintain an
5274 internal counter (which is zero initially). @code{.list} increments the
5275 counter, and @code{.nolist} decrements it. Assembly listings are
5276 generated whenever the counter is greater than zero.
5278 By default, listings are disabled. When you enable them (with the
5279 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5280 the initial value of the listing counter is one.
5283 @section @code{.ln @var{line-number}}
5285 @cindex @code{ln} directive
5286 @ifclear no-line-dir
5287 @samp{.ln} is a synonym for @samp{.line}.
5290 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5291 must be an absolute expression. The next line has that logical
5292 line number, so any other statements on the current line (after a
5293 statement separator character @code{;}) are reported as on logical
5294 line number @var{line-number} @minus{} 1.
5297 This directive is accepted, but ignored, when @command{@value{AS}} is
5298 configured for @code{b.out}; its effect is only associated with COFF
5304 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5305 @cindex @code{loc} directive
5306 When emitting DWARF2 line number information,
5307 the @code{.loc} directive will add a row to the @code{.debug_line} line
5308 number matrix corresponding to the immediately following assembly
5309 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5310 arguments will be applied to the @code{.debug_line} state machine before
5313 The @var{options} are a sequence of the following tokens in any order:
5317 This option will set the @code{basic_block} register in the
5318 @code{.debug_line} state machine to @code{true}.
5321 This option will set the @code{prologue_end} register in the
5322 @code{.debug_line} state machine to @code{true}.
5324 @item epilogue_begin
5325 This option will set the @code{epilogue_begin} register in the
5326 @code{.debug_line} state machine to @code{true}.
5328 @item is_stmt @var{value}
5329 This option will set the @code{is_stmt} register in the
5330 @code{.debug_line} state machine to @code{value}, which must be
5333 @item isa @var{value}
5334 This directive will set the @code{isa} register in the @code{.debug_line}
5335 state machine to @var{value}, which must be an unsigned integer.
5337 @item discriminator @var{value}
5338 This directive will set the @code{discriminator} register in the @code{.debug_line}
5339 state machine to @var{value}, which must be an unsigned integer.
5343 @node Loc_mark_labels
5344 @section @code{.loc_mark_labels @var{enable}}
5345 @cindex @code{loc_mark_labels} directive
5346 When emitting DWARF2 line number information,
5347 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5348 to the @code{.debug_line} line number matrix with the @code{basic_block}
5349 register in the state machine set whenever a code label is seen.
5350 The @var{enable} argument should be either 1 or 0, to enable or disable
5351 this function respectively.
5355 @section @code{.local @var{names}}
5357 @cindex @code{local} directive
5358 This directive, which is available for ELF targets, marks each symbol in
5359 the comma-separated list of @code{names} as a local symbol so that it
5360 will not be externally visible. If the symbols do not already exist,
5361 they will be created.
5363 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5364 accept an alignment argument, which is the case for most ELF targets,
5365 the @code{.local} directive can be used in combination with @code{.comm}
5366 (@pxref{Comm}) to define aligned local common data.
5370 @section @code{.long @var{expressions}}
5372 @cindex @code{long} directive
5373 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5376 @c no one seems to know what this is for or whether this description is
5377 @c what it really ought to do
5379 @section @code{.lsym @var{symbol}, @var{expression}}
5381 @cindex @code{lsym} directive
5382 @cindex symbol, not referenced in assembly
5383 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5384 the hash table, ensuring it cannot be referenced by name during the
5385 rest of the assembly. This sets the attributes of the symbol to be
5386 the same as the expression value:
5388 @var{other} = @var{descriptor} = 0
5389 @var{type} = @r{(section of @var{expression})}
5390 @var{value} = @var{expression}
5393 The new symbol is not flagged as external.
5397 @section @code{.macro}
5400 The commands @code{.macro} and @code{.endm} allow you to define macros that
5401 generate assembly output. For example, this definition specifies a macro
5402 @code{sum} that puts a sequence of numbers into memory:
5405 .macro sum from=0, to=5
5414 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5426 @item .macro @var{macname}
5427 @itemx .macro @var{macname} @var{macargs} @dots{}
5428 @cindex @code{macro} directive
5429 Begin the definition of a macro called @var{macname}. If your macro
5430 definition requires arguments, specify their names after the macro name,
5431 separated by commas or spaces. You can qualify the macro argument to
5432 indicate whether all invocations must specify a non-blank value (through
5433 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5434 (through @samp{:@code{vararg}}). You can supply a default value for any
5435 macro argument by following the name with @samp{=@var{deflt}}. You
5436 cannot define two macros with the same @var{macname} unless it has been
5437 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5438 definitions. For example, these are all valid @code{.macro} statements:
5442 Begin the definition of a macro called @code{comm}, which takes no
5445 @item .macro plus1 p, p1
5446 @itemx .macro plus1 p p1
5447 Either statement begins the definition of a macro called @code{plus1},
5448 which takes two arguments; within the macro definition, write
5449 @samp{\p} or @samp{\p1} to evaluate the arguments.
5451 @item .macro reserve_str p1=0 p2
5452 Begin the definition of a macro called @code{reserve_str}, with two
5453 arguments. The first argument has a default value, but not the second.
5454 After the definition is complete, you can call the macro either as
5455 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5456 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5457 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5458 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5460 @item .macro m p1:req, p2=0, p3:vararg
5461 Begin the definition of a macro called @code{m}, with at least three
5462 arguments. The first argument must always have a value specified, but
5463 not the second, which instead has a default value. The third formal
5464 will get assigned all remaining arguments specified at invocation time.
5466 When you call a macro, you can specify the argument values either by
5467 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5468 @samp{sum to=17, from=9}.
5472 Note that since each of the @var{macargs} can be an identifier exactly
5473 as any other one permitted by the target architecture, there may be
5474 occasional problems if the target hand-crafts special meanings to certain
5475 characters when they occur in a special position. For example, if the colon
5476 (@code{:}) is generally permitted to be part of a symbol name, but the
5477 architecture specific code special-cases it when occurring as the final
5478 character of a symbol (to denote a label), then the macro parameter
5479 replacement code will have no way of knowing that and consider the whole
5480 construct (including the colon) an identifier, and check only this
5481 identifier for being the subject to parameter substitution. So for example
5482 this macro definition:
5490 might not work as expected. Invoking @samp{label foo} might not create a label
5491 called @samp{foo} but instead just insert the text @samp{\l:} into the
5492 assembler source, probably generating an error about an unrecognised
5495 Similarly problems might occur with the period character (@samp{.})
5496 which is often allowed inside opcode names (and hence identifier names). So
5497 for example constructing a macro to build an opcode from a base name and a
5498 length specifier like this:
5501 .macro opcode base length
5506 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5507 instruction but instead generate some kind of error as the assembler tries to
5508 interpret the text @samp{\base.\length}.
5510 There are several possible ways around this problem:
5513 @item Insert white space
5514 If it is possible to use white space characters then this is the simplest
5523 @item Use @samp{\()}
5524 The string @samp{\()} can be used to separate the end of a macro argument from
5525 the following text. eg:
5528 .macro opcode base length
5533 @item Use the alternate macro syntax mode
5534 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5535 used as a separator. eg:
5545 Note: this problem of correctly identifying string parameters to pseudo ops
5546 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5547 and @code{.irpc} (@pxref{Irpc}) as well.
5550 @cindex @code{endm} directive
5551 Mark the end of a macro definition.
5554 @cindex @code{exitm} directive
5555 Exit early from the current macro definition.
5557 @cindex number of macros executed
5558 @cindex macros, count executed
5560 @command{@value{AS}} maintains a counter of how many macros it has
5561 executed in this pseudo-variable; you can copy that number to your
5562 output with @samp{\@@}, but @emph{only within a macro definition}.
5564 @item LOCAL @var{name} [ , @dots{} ]
5565 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5566 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5567 @xref{Altmacro,,@code{.altmacro}}.
5571 @section @code{.mri @var{val}}
5573 @cindex @code{mri} directive
5574 @cindex MRI mode, temporarily
5575 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
5576 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
5577 affects code assembled until the next @code{.mri} directive, or until the end
5578 of the file. @xref{M, MRI mode, MRI mode}.
5581 @section @code{.noaltmacro}
5582 Disable alternate macro mode. @xref{Altmacro}.
5585 @section @code{.nolist}
5587 @cindex @code{nolist} directive
5588 @cindex listing control, turning off
5589 Control (in conjunction with the @code{.list} directive) whether or
5590 not assembly listings are generated. These two directives maintain an
5591 internal counter (which is zero initially). @code{.list} increments the
5592 counter, and @code{.nolist} decrements it. Assembly listings are
5593 generated whenever the counter is greater than zero.
5596 @section @code{.octa @var{bignums}}
5598 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5599 @cindex @code{octa} directive
5600 @cindex integer, 16-byte
5601 @cindex sixteen byte integer
5602 This directive expects zero or more bignums, separated by commas. For each
5603 bignum, it emits a 16-byte integer.
5605 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5606 hence @emph{octa}-word for 16 bytes.
5609 @section @code{.offset @var{loc}}
5611 @cindex @code{offset} directive
5612 Set the location counter to @var{loc} in the absolute section. @var{loc} must
5613 be an absolute expression. This directive may be useful for defining
5614 symbols with absolute values. Do not confuse it with the @code{.org}
5618 @section @code{.org @var{new-lc} , @var{fill}}
5620 @cindex @code{org} directive
5621 @cindex location counter, advancing
5622 @cindex advancing location counter
5623 @cindex current address, advancing
5624 Advance the location counter of the current section to
5625 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5626 expression with the same section as the current subsection. That is,
5627 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5628 wrong section, the @code{.org} directive is ignored. To be compatible
5629 with former assemblers, if the section of @var{new-lc} is absolute,
5630 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5631 is the same as the current subsection.
5633 @code{.org} may only increase the location counter, or leave it
5634 unchanged; you cannot use @code{.org} to move the location counter
5637 @c double negative used below "not undefined" because this is a specific
5638 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5639 @c section. doc@cygnus.com 18feb91
5640 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5641 may not be undefined. If you really detest this restriction we eagerly await
5642 a chance to share your improved assembler.
5644 Beware that the origin is relative to the start of the section, not
5645 to the start of the subsection. This is compatible with other
5646 people's assemblers.
5648 When the location counter (of the current subsection) is advanced, the
5649 intervening bytes are filled with @var{fill} which should be an
5650 absolute expression. If the comma and @var{fill} are omitted,
5651 @var{fill} defaults to zero.
5654 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5656 @cindex padding the location counter given a power of two
5657 @cindex @code{p2align} directive
5658 Pad the location counter (in the current subsection) to a particular
5659 storage boundary. The first expression (which must be absolute) is the
5660 number of low-order zero bits the location counter must have after
5661 advancement. For example @samp{.p2align 3} advances the location
5662 counter until it a multiple of 8. If the location counter is already a
5663 multiple of 8, no change is needed.
5665 The second expression (also absolute) gives the fill value to be stored in the
5666 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5667 padding bytes are normally zero. However, on some systems, if the section is
5668 marked as containing code and the fill value is omitted, the space is filled
5669 with no-op instructions.
5671 The third expression is also absolute, and is also optional. If it is present,
5672 it is the maximum number of bytes that should be skipped by this alignment
5673 directive. If doing the alignment would require skipping more bytes than the
5674 specified maximum, then the alignment is not done at all. You can omit the
5675 fill value (the second argument) entirely by simply using two commas after the
5676 required alignment; this can be useful if you want the alignment to be filled
5677 with no-op instructions when appropriate.
5679 @cindex @code{p2alignw} directive
5680 @cindex @code{p2alignl} directive
5681 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5682 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5683 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5684 fill pattern as a four byte longword value. For example, @code{.p2alignw
5685 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5686 filled in with the value 0x368d (the exact placement of the bytes depends upon
5687 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5692 @section @code{.popsection}
5694 @cindex @code{popsection} directive
5695 @cindex Section Stack
5696 This is one of the ELF section stack manipulation directives. The others are
5697 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5698 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5701 This directive replaces the current section (and subsection) with the top
5702 section (and subsection) on the section stack. This section is popped off the
5708 @section @code{.previous}
5710 @cindex @code{previous} directive
5711 @cindex Section Stack
5712 This is one of the ELF section stack manipulation directives. The others are
5713 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5714 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5715 (@pxref{PopSection}).
5717 This directive swaps the current section (and subsection) with most recently
5718 referenced section/subsection pair prior to this one. Multiple
5719 @code{.previous} directives in a row will flip between two sections (and their
5720 subsections). For example:
5732 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
5738 # Now in section A subsection 1
5742 # Now in section B subsection 0
5745 # Now in section B subsection 1
5748 # Now in section B subsection 0
5752 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
5753 section B and 0x9abc into subsection 1 of section B.
5755 In terms of the section stack, this directive swaps the current section with
5756 the top section on the section stack.
5760 @section @code{.print @var{string}}
5762 @cindex @code{print} directive
5763 @command{@value{AS}} will print @var{string} on the standard output during
5764 assembly. You must put @var{string} in double quotes.
5768 @section @code{.protected @var{names}}
5770 @cindex @code{protected} directive
5772 This is one of the ELF visibility directives. The other two are
5773 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5775 This directive overrides the named symbols default visibility (which is set by
5776 their binding: local, global or weak). The directive sets the visibility to
5777 @code{protected} which means that any references to the symbols from within the
5778 components that defines them must be resolved to the definition in that
5779 component, even if a definition in another component would normally preempt
5784 @section @code{.psize @var{lines} , @var{columns}}
5786 @cindex @code{psize} directive
5787 @cindex listing control: paper size
5788 @cindex paper size, for listings
5789 Use this directive to declare the number of lines---and, optionally, the
5790 number of columns---to use for each page, when generating listings.
5792 If you do not use @code{.psize}, listings use a default line-count
5793 of 60. You may omit the comma and @var{columns} specification; the
5794 default width is 200 columns.
5796 @command{@value{AS}} generates formfeeds whenever the specified number of
5797 lines is exceeded (or whenever you explicitly request one, using
5800 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5801 those explicitly specified with @code{.eject}.
5804 @section @code{.purgem @var{name}}
5806 @cindex @code{purgem} directive
5807 Undefine the macro @var{name}, so that later uses of the string will not be
5808 expanded. @xref{Macro}.
5812 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
5814 @cindex @code{pushsection} directive
5815 @cindex Section Stack
5816 This is one of the ELF section stack manipulation directives. The others are
5817 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5818 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5821 This directive pushes the current section (and subsection) onto the
5822 top of the section stack, and then replaces the current section and
5823 subsection with @code{name} and @code{subsection}. The optional
5824 @code{flags}, @code{type} and @code{arguments} are treated the same
5825 as in the @code{.section} (@pxref{Section}) directive.
5829 @section @code{.quad @var{bignums}}
5831 @cindex @code{quad} directive
5832 @code{.quad} expects zero or more bignums, separated by commas. For
5833 each bignum, it emits
5835 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5836 warning message; and just takes the lowest order 8 bytes of the bignum.
5837 @cindex eight-byte integer
5838 @cindex integer, 8-byte
5840 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5841 hence @emph{quad}-word for 8 bytes.
5844 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5845 warning message; and just takes the lowest order 16 bytes of the bignum.
5846 @cindex sixteen-byte integer
5847 @cindex integer, 16-byte
5851 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
5853 @cindex @code{reloc} directive
5854 Generate a relocation at @var{offset} of type @var{reloc_name} with value
5855 @var{expression}. If @var{offset} is a number, the relocation is generated in
5856 the current section. If @var{offset} is an expression that resolves to a
5857 symbol plus offset, the relocation is generated in the given symbol's section.
5858 @var{expression}, if present, must resolve to a symbol plus addend or to an
5859 absolute value, but note that not all targets support an addend. e.g. ELF REL
5860 targets such as i386 store an addend in the section contents rather than in the
5861 relocation. This low level interface does not support addends stored in the
5865 @section @code{.rept @var{count}}
5867 @cindex @code{rept} directive
5868 Repeat the sequence of lines between the @code{.rept} directive and the next
5869 @code{.endr} directive @var{count} times.
5871 For example, assembling
5879 is equivalent to assembling
5888 @section @code{.sbttl "@var{subheading}"}
5890 @cindex @code{sbttl} directive
5891 @cindex subtitles for listings
5892 @cindex listing control: subtitle
5893 Use @var{subheading} as the title (third line, immediately after the
5894 title line) when generating assembly listings.
5896 This directive affects subsequent pages, as well as the current page if
5897 it appears within ten lines of the top of a page.
5901 @section @code{.scl @var{class}}
5903 @cindex @code{scl} directive
5904 @cindex symbol storage class (COFF)
5905 @cindex COFF symbol storage class
5906 Set the storage-class value for a symbol. This directive may only be
5907 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5908 whether a symbol is static or external, or it may record further
5909 symbolic debugging information.
5912 The @samp{.scl} directive is primarily associated with COFF output; when
5913 configured to generate @code{b.out} output format, @command{@value{AS}}
5914 accepts this directive but ignores it.
5920 @section @code{.section @var{name}}
5922 @cindex named section
5923 Use the @code{.section} directive to assemble the following code into a section
5926 This directive is only supported for targets that actually support arbitrarily
5927 named sections; on @code{a.out} targets, for example, it is not accepted, even
5928 with a standard @code{a.out} section name.
5932 @c only print the extra heading if both COFF and ELF are set
5933 @subheading COFF Version
5936 @cindex @code{section} directive (COFF version)
5937 For COFF targets, the @code{.section} directive is used in one of the following
5941 .section @var{name}[, "@var{flags}"]
5942 .section @var{name}[, @var{subsection}]
5945 If the optional argument is quoted, it is taken as flags to use for the
5946 section. Each flag is a single character. The following flags are recognized:
5949 bss section (uninitialized data)
5951 section is not loaded
5961 shared section (meaningful for PE targets)
5963 ignored. (For compatibility with the ELF version)
5965 section is not readable (meaningful for PE targets)
5967 single-digit power-of-two section alignment (GNU extension)
5970 If no flags are specified, the default flags depend upon the section name. If
5971 the section name is not recognized, the default will be for the section to be
5972 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5973 from the section, rather than adding them, so if they are used on their own it
5974 will be as if no flags had been specified at all.
5976 If the optional argument to the @code{.section} directive is not quoted, it is
5977 taken as a subsection number (@pxref{Sub-Sections}).
5982 @c only print the extra heading if both COFF and ELF are set
5983 @subheading ELF Version
5986 @cindex Section Stack
5987 This is one of the ELF section stack manipulation directives. The others are
5988 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5989 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5990 @code{.previous} (@pxref{Previous}).
5992 @cindex @code{section} directive (ELF version)
5993 For ELF targets, the @code{.section} directive is used like this:
5996 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
5999 The optional @var{flags} argument is a quoted string which may contain any
6000 combination of the following characters:
6003 section is allocatable
6005 section is excluded from executable and shared library.
6009 section is executable
6011 section is mergeable
6013 section contains zero terminated strings
6015 section is a member of a section group
6017 section is used for thread-local-storage
6019 section is a member of the previously-current section's group, if any
6022 The optional @var{type} argument may contain one of the following constants:
6025 section contains data
6027 section does not contain data (i.e., section only occupies space)
6029 section contains data which is used by things other than the program
6031 section contains an array of pointers to init functions
6033 section contains an array of pointers to finish functions
6034 @item @@preinit_array
6035 section contains an array of pointers to pre-init functions
6038 Many targets only support the first three section types.
6040 Note on targets where the @code{@@} character is the start of a comment (eg
6041 ARM) then another character is used instead. For example the ARM port uses the
6044 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6045 be specified as well as an extra argument---@var{entsize}---like this:
6048 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6051 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6052 constants, each @var{entsize} octets long. Sections with both @code{M} and
6053 @code{S} must contain zero terminated strings where each character is
6054 @var{entsize} bytes long. The linker may remove duplicates within sections with
6055 the same name, same entity size and same flags. @var{entsize} must be an
6056 absolute expression. For sections with both @code{M} and @code{S}, a string
6057 which is a suffix of a larger string is considered a duplicate. Thus
6058 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6059 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6061 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6062 be present along with an additional field like this:
6065 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6068 The @var{GroupName} field specifies the name of the section group to which this
6069 particular section belongs. The optional linkage field can contain:
6072 indicates that only one copy of this section should be retained
6077 Note: if both the @var{M} and @var{G} flags are present then the fields for
6078 the Merge flag should come first, like this:
6081 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6084 If @var{flags} contains the @code{?} symbol then it may not also contain the
6085 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6086 present. Instead, @code{?} says to consider the section that's current before
6087 this directive. If that section used @code{G}, then the new section will use
6088 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6089 If not, then the @code{?} symbol has no effect.
6091 If no flags are specified, the default flags depend upon the section name. If
6092 the section name is not recognized, the default will be for the section to have
6093 none of the above flags: it will not be allocated in memory, nor writable, nor
6094 executable. The section will contain data.
6096 For ELF targets, the assembler supports another type of @code{.section}
6097 directive for compatibility with the Solaris assembler:
6100 .section "@var{name}"[, @var{flags}...]
6103 Note that the section name is quoted. There may be a sequence of comma
6107 section is allocatable
6111 section is executable
6113 section is excluded from executable and shared library.
6115 section is used for thread local storage
6118 This directive replaces the current section and subsection. See the
6119 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6120 some examples of how this directive and the other section stack directives
6126 @section @code{.set @var{symbol}, @var{expression}}
6128 @cindex @code{set} directive
6129 @cindex symbol value, setting
6130 Set the value of @var{symbol} to @var{expression}. This
6131 changes @var{symbol}'s value and type to conform to
6132 @var{expression}. If @var{symbol} was flagged as external, it remains
6133 flagged (@pxref{Symbol Attributes}).
6135 You may @code{.set} a symbol many times in the same assembly.
6137 If you @code{.set} a global symbol, the value stored in the object
6138 file is the last value stored into it.
6141 On Z80 @code{set} is a real instruction, use
6142 @samp{@var{symbol} defl @var{expression}} instead.
6146 @section @code{.short @var{expressions}}
6148 @cindex @code{short} directive
6150 @code{.short} is normally the same as @samp{.word}.
6151 @xref{Word,,@code{.word}}.
6153 In some configurations, however, @code{.short} and @code{.word} generate
6154 numbers of different lengths. @xref{Machine Dependencies}.
6158 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6161 This expects zero or more @var{expressions}, and emits
6162 a 16 bit number for each.
6167 @section @code{.single @var{flonums}}
6169 @cindex @code{single} directive
6170 @cindex floating point numbers (single)
6171 This directive assembles zero or more flonums, separated by commas. It
6172 has the same effect as @code{.float}.
6174 The exact kind of floating point numbers emitted depends on how
6175 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6179 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6180 numbers in @sc{ieee} format.
6186 @section @code{.size}
6188 This directive is used to set the size associated with a symbol.
6192 @c only print the extra heading if both COFF and ELF are set
6193 @subheading COFF Version
6196 @cindex @code{size} directive (COFF version)
6197 For COFF targets, the @code{.size} directive is only permitted inside
6198 @code{.def}/@code{.endef} pairs. It is used like this:
6201 .size @var{expression}
6205 @samp{.size} is only meaningful when generating COFF format output; when
6206 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6213 @c only print the extra heading if both COFF and ELF are set
6214 @subheading ELF Version
6217 @cindex @code{size} directive (ELF version)
6218 For ELF targets, the @code{.size} directive is used like this:
6221 .size @var{name} , @var{expression}
6224 This directive sets the size associated with a symbol @var{name}.
6225 The size in bytes is computed from @var{expression} which can make use of label
6226 arithmetic. This directive is typically used to set the size of function
6231 @ifclear no-space-dir
6233 @section @code{.skip @var{size} , @var{fill}}
6235 @cindex @code{skip} directive
6236 @cindex filling memory
6237 This directive emits @var{size} bytes, each of value @var{fill}. Both
6238 @var{size} and @var{fill} are absolute expressions. If the comma and
6239 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6244 @section @code{.sleb128 @var{expressions}}
6246 @cindex @code{sleb128} directive
6247 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6248 compact, variable length representation of numbers used by the DWARF
6249 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6251 @ifclear no-space-dir
6253 @section @code{.space @var{size} , @var{fill}}
6255 @cindex @code{space} directive
6256 @cindex filling memory
6257 This directive emits @var{size} bytes, each of value @var{fill}. Both
6258 @var{size} and @var{fill} are absolute expressions. If the comma
6259 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6264 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6265 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6266 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6267 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6275 @section @code{.stabd, .stabn, .stabs}
6277 @cindex symbolic debuggers, information for
6278 @cindex @code{stab@var{x}} directives
6279 There are three directives that begin @samp{.stab}.
6280 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6281 The symbols are not entered in the @command{@value{AS}} hash table: they
6282 cannot be referenced elsewhere in the source file.
6283 Up to five fields are required:
6287 This is the symbol's name. It may contain any character except
6288 @samp{\000}, so is more general than ordinary symbol names. Some
6289 debuggers used to code arbitrarily complex structures into symbol names
6293 An absolute expression. The symbol's type is set to the low 8 bits of
6294 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6295 and debuggers choke on silly bit patterns.
6298 An absolute expression. The symbol's ``other'' attribute is set to the
6299 low 8 bits of this expression.
6302 An absolute expression. The symbol's descriptor is set to the low 16
6303 bits of this expression.
6306 An absolute expression which becomes the symbol's value.
6309 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6310 or @code{.stabs} statement, the symbol has probably already been created;
6311 you get a half-formed symbol in your object file. This is
6312 compatible with earlier assemblers!
6315 @cindex @code{stabd} directive
6316 @item .stabd @var{type} , @var{other} , @var{desc}
6318 The ``name'' of the symbol generated is not even an empty string.
6319 It is a null pointer, for compatibility. Older assemblers used a
6320 null pointer so they didn't waste space in object files with empty
6323 The symbol's value is set to the location counter,
6324 relocatably. When your program is linked, the value of this symbol
6325 is the address of the location counter when the @code{.stabd} was
6328 @cindex @code{stabn} directive
6329 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6330 The name of the symbol is set to the empty string @code{""}.
6332 @cindex @code{stabs} directive
6333 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6334 All five fields are specified.
6340 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6341 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6343 @cindex string, copying to object file
6344 @cindex string8, copying to object file
6345 @cindex string16, copying to object file
6346 @cindex string32, copying to object file
6347 @cindex string64, copying to object file
6348 @cindex @code{string} directive
6349 @cindex @code{string8} directive
6350 @cindex @code{string16} directive
6351 @cindex @code{string32} directive
6352 @cindex @code{string64} directive
6354 Copy the characters in @var{str} to the object file. You may specify more than
6355 one string to copy, separated by commas. Unless otherwise specified for a
6356 particular machine, the assembler marks the end of each string with a 0 byte.
6357 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6359 The variants @code{string16}, @code{string32} and @code{string64} differ from
6360 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6361 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6362 are stored in target endianness byte order.
6368 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6369 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6374 @section @code{.struct @var{expression}}
6376 @cindex @code{struct} directive
6377 Switch to the absolute section, and set the section offset to @var{expression},
6378 which must be an absolute expression. You might use this as follows:
6387 This would define the symbol @code{field1} to have the value 0, the symbol
6388 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6389 value 8. Assembly would be left in the absolute section, and you would need to
6390 use a @code{.section} directive of some sort to change to some other section
6391 before further assembly.
6395 @section @code{.subsection @var{name}}
6397 @cindex @code{subsection} directive
6398 @cindex Section Stack
6399 This is one of the ELF section stack manipulation directives. The others are
6400 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6401 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6404 This directive replaces the current subsection with @code{name}. The current
6405 section is not changed. The replaced subsection is put onto the section stack
6406 in place of the then current top of stack subsection.
6411 @section @code{.symver}
6412 @cindex @code{symver} directive
6413 @cindex symbol versioning
6414 @cindex versions of symbols
6415 Use the @code{.symver} directive to bind symbols to specific version nodes
6416 within a source file. This is only supported on ELF platforms, and is
6417 typically used when assembling files to be linked into a shared library.
6418 There are cases where it may make sense to use this in objects to be bound
6419 into an application itself so as to override a versioned symbol from a
6422 For ELF targets, the @code{.symver} directive can be used like this:
6424 .symver @var{name}, @var{name2@@nodename}
6426 If the symbol @var{name} is defined within the file
6427 being assembled, the @code{.symver} directive effectively creates a symbol
6428 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6429 just don't try and create a regular alias is that the @var{@@} character isn't
6430 permitted in symbol names. The @var{name2} part of the name is the actual name
6431 of the symbol by which it will be externally referenced. The name @var{name}
6432 itself is merely a name of convenience that is used so that it is possible to
6433 have definitions for multiple versions of a function within a single source
6434 file, and so that the compiler can unambiguously know which version of a
6435 function is being mentioned. The @var{nodename} portion of the alias should be
6436 the name of a node specified in the version script supplied to the linker when
6437 building a shared library. If you are attempting to override a versioned
6438 symbol from a shared library, then @var{nodename} should correspond to the
6439 nodename of the symbol you are trying to override.
6441 If the symbol @var{name} is not defined within the file being assembled, all
6442 references to @var{name} will be changed to @var{name2@@nodename}. If no
6443 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6446 Another usage of the @code{.symver} directive is:
6448 .symver @var{name}, @var{name2@@@@nodename}
6450 In this case, the symbol @var{name} must exist and be defined within
6451 the file being assembled. It is similar to @var{name2@@nodename}. The
6452 difference is @var{name2@@@@nodename} will also be used to resolve
6453 references to @var{name2} by the linker.
6455 The third usage of the @code{.symver} directive is:
6457 .symver @var{name}, @var{name2@@@@@@nodename}
6459 When @var{name} is not defined within the
6460 file being assembled, it is treated as @var{name2@@nodename}. When
6461 @var{name} is defined within the file being assembled, the symbol
6462 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6467 @section @code{.tag @var{structname}}
6469 @cindex COFF structure debugging
6470 @cindex structure debugging, COFF
6471 @cindex @code{tag} directive
6472 This directive is generated by compilers to include auxiliary debugging
6473 information in the symbol table. It is only permitted inside
6474 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6475 definitions in the symbol table with instances of those structures.
6478 @samp{.tag} is only used when generating COFF format output; when
6479 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6485 @section @code{.text @var{subsection}}
6487 @cindex @code{text} directive
6488 Tells @command{@value{AS}} to assemble the following statements onto the end of
6489 the text subsection numbered @var{subsection}, which is an absolute
6490 expression. If @var{subsection} is omitted, subsection number zero
6494 @section @code{.title "@var{heading}"}
6496 @cindex @code{title} directive
6497 @cindex listing control: title line
6498 Use @var{heading} as the title (second line, immediately after the
6499 source file name and pagenumber) when generating assembly listings.
6501 This directive affects subsequent pages, as well as the current page if
6502 it appears within ten lines of the top of a page.
6506 @section @code{.type}
6508 This directive is used to set the type of a symbol.
6512 @c only print the extra heading if both COFF and ELF are set
6513 @subheading COFF Version
6516 @cindex COFF symbol type
6517 @cindex symbol type, COFF
6518 @cindex @code{type} directive (COFF version)
6519 For COFF targets, this directive is permitted only within
6520 @code{.def}/@code{.endef} pairs. It is used like this:
6526 This records the integer @var{int} as the type attribute of a symbol table
6530 @samp{.type} is associated only with COFF format output; when
6531 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6532 directive but ignores it.
6538 @c only print the extra heading if both COFF and ELF are set
6539 @subheading ELF Version
6542 @cindex ELF symbol type
6543 @cindex symbol type, ELF
6544 @cindex @code{type} directive (ELF version)
6545 For ELF targets, the @code{.type} directive is used like this:
6548 .type @var{name} , @var{type description}
6551 This sets the type of symbol @var{name} to be either a
6552 function symbol or an object symbol. There are five different syntaxes
6553 supported for the @var{type description} field, in order to provide
6554 compatibility with various other assemblers.
6556 Because some of the characters used in these syntaxes (such as @samp{@@} and
6557 @samp{#}) are comment characters for some architectures, some of the syntaxes
6558 below do not work on all architectures. The first variant will be accepted by
6559 the GNU assembler on all architectures so that variant should be used for
6560 maximum portability, if you do not need to assemble your code with other
6563 The syntaxes supported are:
6566 .type <name> STT_<TYPE_IN_UPPER_CASE>
6567 .type <name>,#<type>
6568 .type <name>,@@<type>
6569 .type <name>,%<type>
6570 .type <name>,"<type>"
6573 The types supported are:
6578 Mark the symbol as being a function name.
6581 @itemx gnu_indirect_function
6582 Mark the symbol as an indirect function when evaluated during reloc
6583 processing. (This is only supported on assemblers targeting GNU systems).
6587 Mark the symbol as being a data object.
6591 Mark the symbol as being a thead-local data object.
6595 Mark the symbol as being a common data object.
6599 Does not mark the symbol in any way. It is supported just for completeness.
6601 @item gnu_unique_object
6602 Marks the symbol as being a globally unique data object. The dynamic linker
6603 will make sure that in the entire process there is just one symbol with this
6604 name and type in use. (This is only supported on assemblers targeting GNU
6609 Note: Some targets support extra types in addition to those listed above.
6615 @section @code{.uleb128 @var{expressions}}
6617 @cindex @code{uleb128} directive
6618 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6619 compact, variable length representation of numbers used by the DWARF
6620 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
6624 @section @code{.val @var{addr}}
6626 @cindex @code{val} directive
6627 @cindex COFF value attribute
6628 @cindex value attribute, COFF
6629 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6630 records the address @var{addr} as the value attribute of a symbol table
6634 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6635 configured for @code{b.out}, it accepts this directive but ignores it.
6641 @section @code{.version "@var{string}"}
6643 @cindex @code{version} directive
6644 This directive creates a @code{.note} section and places into it an ELF
6645 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6650 @section @code{.vtable_entry @var{table}, @var{offset}}
6652 @cindex @code{vtable_entry} directive
6653 This directive finds or creates a symbol @code{table} and creates a
6654 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6657 @section @code{.vtable_inherit @var{child}, @var{parent}}
6659 @cindex @code{vtable_inherit} directive
6660 This directive finds the symbol @code{child} and finds or creates the symbol
6661 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6662 parent whose addend is the value of the child symbol. As a special case the
6663 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
6667 @section @code{.warning "@var{string}"}
6668 @cindex warning directive
6669 Similar to the directive @code{.error}
6670 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6673 @section @code{.weak @var{names}}
6675 @cindex @code{weak} directive
6676 This directive sets the weak attribute on the comma separated list of symbol
6677 @code{names}. If the symbols do not already exist, they will be created.
6679 On COFF targets other than PE, weak symbols are a GNU extension. This
6680 directive sets the weak attribute on the comma separated list of symbol
6681 @code{names}. If the symbols do not already exist, they will be created.
6683 On the PE target, weak symbols are supported natively as weak aliases.
6684 When a weak symbol is created that is not an alias, GAS creates an
6685 alternate symbol to hold the default value.
6688 @section @code{.weakref @var{alias}, @var{target}}
6690 @cindex @code{weakref} directive
6691 This directive creates an alias to the target symbol that enables the symbol to
6692 be referenced with weak-symbol semantics, but without actually making it weak.
6693 If direct references or definitions of the symbol are present, then the symbol
6694 will not be weak, but if all references to it are through weak references, the
6695 symbol will be marked as weak in the symbol table.
6697 The effect is equivalent to moving all references to the alias to a separate
6698 assembly source file, renaming the alias to the symbol in it, declaring the
6699 symbol as weak there, and running a reloadable link to merge the object files
6700 resulting from the assembly of the new source file and the old source file that
6701 had the references to the alias removed.
6703 The alias itself never makes to the symbol table, and is entirely handled
6704 within the assembler.
6707 @section @code{.word @var{expressions}}
6709 @cindex @code{word} directive
6710 This directive expects zero or more @var{expressions}, of any section,
6711 separated by commas.
6714 For each expression, @command{@value{AS}} emits a 32-bit number.
6717 For each expression, @command{@value{AS}} emits a 16-bit number.
6722 The size of the number emitted, and its byte order,
6723 depend on what target computer the assembly is for.
6726 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6727 @c happen---32-bit addressability, period; no long/short jumps.
6728 @ifset DIFF-TBL-KLUGE
6729 @cindex difference tables altered
6730 @cindex altered difference tables
6732 @emph{Warning: Special Treatment to support Compilers}
6736 Machines with a 32-bit address space, but that do less than 32-bit
6737 addressing, require the following special treatment. If the machine of
6738 interest to you does 32-bit addressing (or doesn't require it;
6739 @pxref{Machine Dependencies}), you can ignore this issue.
6742 In order to assemble compiler output into something that works,
6743 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6744 Directives of the form @samp{.word sym1-sym2} are often emitted by
6745 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6746 directive of the form @samp{.word sym1-sym2}, and the difference between
6747 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6748 creates a @dfn{secondary jump table}, immediately before the next label.
6749 This secondary jump table is preceded by a short-jump to the
6750 first byte after the secondary table. This short-jump prevents the flow
6751 of control from accidentally falling into the new table. Inside the
6752 table is a long-jump to @code{sym2}. The original @samp{.word}
6753 contains @code{sym1} minus the address of the long-jump to
6756 If there were several occurrences of @samp{.word sym1-sym2} before the
6757 secondary jump table, all of them are adjusted. If there was a
6758 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6759 long-jump to @code{sym4} is included in the secondary jump table,
6760 and the @code{.word} directives are adjusted to contain @code{sym3}
6761 minus the address of the long-jump to @code{sym4}; and so on, for as many
6762 entries in the original jump table as necessary.
6765 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6766 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6767 assembly language programmers.
6770 @c end DIFF-TBL-KLUGE
6773 @section Deprecated Directives
6775 @cindex deprecated directives
6776 @cindex obsolescent directives
6777 One day these directives won't work.
6778 They are included for compatibility with older assemblers.
6785 @node Object Attributes
6786 @chapter Object Attributes
6787 @cindex object attributes
6789 @command{@value{AS}} assembles source files written for a specific architecture
6790 into object files for that architecture. But not all object files are alike.
6791 Many architectures support incompatible variations. For instance, floating
6792 point arguments might be passed in floating point registers if the object file
6793 requires hardware floating point support---or floating point arguments might be
6794 passed in integer registers if the object file supports processors with no
6795 hardware floating point unit. Or, if two objects are built for different
6796 generations of the same architecture, the combination may require the
6797 newer generation at run-time.
6799 This information is useful during and after linking. At link time,
6800 @command{@value{LD}} can warn about incompatible object files. After link
6801 time, tools like @command{gdb} can use it to process the linked file
6804 Compatibility information is recorded as a series of object attributes. Each
6805 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
6806 string, and indicates who sets the meaning of the tag. The tag is an integer,
6807 and indicates what property the attribute describes. The value may be a string
6808 or an integer, and indicates how the property affects this object. Missing
6809 attributes are the same as attributes with a zero value or empty string value.
6811 Object attributes were developed as part of the ABI for the ARM Architecture.
6812 The file format is documented in @cite{ELF for the ARM Architecture}.
6815 * GNU Object Attributes:: @sc{gnu} Object Attributes
6816 * Defining New Object Attributes:: Defining New Object Attributes
6819 @node GNU Object Attributes
6820 @section @sc{gnu} Object Attributes
6822 The @code{.gnu_attribute} directive records an object attribute
6823 with vendor @samp{gnu}.
6825 Except for @samp{Tag_compatibility}, which has both an integer and a string for
6826 its value, @sc{gnu} attributes have a string value if the tag number is odd and
6827 an integer value if the tag number is even. The second bit (@code{@var{tag} &
6828 2} is set for architecture-independent attributes and clear for
6829 architecture-dependent ones.
6831 @subsection Common @sc{gnu} attributes
6833 These attributes are valid on all architectures.
6836 @item Tag_compatibility (32)
6837 The compatibility attribute takes an integer flag value and a vendor name. If
6838 the flag value is 0, the file is compatible with other toolchains. If it is 1,
6839 then the file is only compatible with the named toolchain. If it is greater
6840 than 1, the file can only be processed by other toolchains under some private
6841 arrangement indicated by the flag value and the vendor name.
6844 @subsection MIPS Attributes
6847 @item Tag_GNU_MIPS_ABI_FP (4)
6848 The floating-point ABI used by this object file. The value will be:
6852 0 for files not affected by the floating-point ABI.
6854 1 for files using the hardware floating-point with a standard double-precision
6857 2 for files using the hardware floating-point ABI with a single-precision FPU.
6859 3 for files using the software floating-point ABI.
6861 4 for files using the hardware floating-point ABI with 64-bit wide
6862 double-precision floating-point registers and 32-bit wide general
6867 @subsection PowerPC Attributes
6870 @item Tag_GNU_Power_ABI_FP (4)
6871 The floating-point ABI used by this object file. The value will be:
6875 0 for files not affected by the floating-point ABI.
6877 1 for files using double-precision hardware floating-point ABI.
6879 2 for files using the software floating-point ABI.
6881 3 for files using single-precision hardware floating-point ABI.
6884 @item Tag_GNU_Power_ABI_Vector (8)
6885 The vector ABI used by this object file. The value will be:
6889 0 for files not affected by the vector ABI.
6891 1 for files using general purpose registers to pass vectors.
6893 2 for files using AltiVec registers to pass vectors.
6895 3 for files using SPE registers to pass vectors.
6899 @node Defining New Object Attributes
6900 @section Defining New Object Attributes
6902 If you want to define a new @sc{gnu} object attribute, here are the places you
6903 will need to modify. New attributes should be discussed on the @samp{binutils}
6908 This manual, which is the official register of attributes.
6910 The header for your architecture @file{include/elf}, to define the tag.
6912 The @file{bfd} support file for your architecture, to merge the attribute
6913 and issue any appropriate link warnings.
6915 Test cases in @file{ld/testsuite} for merging and link warnings.
6917 @file{binutils/readelf.c} to display your attribute.
6919 GCC, if you want the compiler to mark the attribute automatically.
6925 @node Machine Dependencies
6926 @chapter Machine Dependent Features
6928 @cindex machine dependencies
6929 The machine instruction sets are (almost by definition) different on
6930 each machine where @command{@value{AS}} runs. Floating point representations
6931 vary as well, and @command{@value{AS}} often supports a few additional
6932 directives or command-line options for compatibility with other
6933 assemblers on a particular platform. Finally, some versions of
6934 @command{@value{AS}} support special pseudo-instructions for branch
6937 This chapter discusses most of these differences, though it does not
6938 include details on any machine's instruction set. For details on that
6939 subject, see the hardware manufacturer's manual.
6943 * AArch64-Dependent:: AArch64 Dependent Features
6946 * Alpha-Dependent:: Alpha Dependent Features
6949 * ARC-Dependent:: ARC Dependent Features
6952 * ARM-Dependent:: ARM Dependent Features
6955 * AVR-Dependent:: AVR Dependent Features
6958 * Blackfin-Dependent:: Blackfin Dependent Features
6961 * CR16-Dependent:: CR16 Dependent Features
6964 * CRIS-Dependent:: CRIS Dependent Features
6967 * D10V-Dependent:: D10V Dependent Features
6970 * D30V-Dependent:: D30V Dependent Features
6973 * Epiphany-Dependent:: EPIPHANY Dependent Features
6976 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6979 * HPPA-Dependent:: HPPA Dependent Features
6982 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6985 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6988 * i860-Dependent:: Intel 80860 Dependent Features
6991 * i960-Dependent:: Intel 80960 Dependent Features
6994 * IA-64-Dependent:: Intel IA-64 Dependent Features
6997 * IP2K-Dependent:: IP2K Dependent Features
7000 * LM32-Dependent:: LM32 Dependent Features
7003 * M32C-Dependent:: M32C Dependent Features
7006 * M32R-Dependent:: M32R Dependent Features
7009 * M68K-Dependent:: M680x0 Dependent Features
7012 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7015 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7018 * MIPS-Dependent:: MIPS Dependent Features
7021 * MMIX-Dependent:: MMIX Dependent Features
7024 * MSP430-Dependent:: MSP430 Dependent Features
7027 * NS32K-Dependent:: NS32K Dependent Features
7030 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7031 * SH64-Dependent:: SuperH SH64 Dependent Features
7034 * PDP-11-Dependent:: PDP-11 Dependent Features
7037 * PJ-Dependent:: picoJava Dependent Features
7040 * PPC-Dependent:: PowerPC Dependent Features
7043 * RL78-Dependent:: RL78 Dependent Features
7046 * RX-Dependent:: RX Dependent Features
7049 * S/390-Dependent:: IBM S/390 Dependent Features
7052 * SCORE-Dependent:: SCORE Dependent Features
7055 * Sparc-Dependent:: SPARC Dependent Features
7058 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7061 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7064 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7067 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7070 * V850-Dependent:: V850 Dependent Features
7073 * XGATE-Dependent:: XGATE Features
7076 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7079 * Xtensa-Dependent:: Xtensa Dependent Features
7082 * Z80-Dependent:: Z80 Dependent Features
7085 * Z8000-Dependent:: Z8000 Dependent Features
7088 * Vax-Dependent:: VAX Dependent Features
7095 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7096 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7097 @c peculiarity: to preserve cross-references, there must be a node called
7098 @c "Machine Dependencies". Hence the conditional nodenames in each
7099 @c major node below. Node defaulting in makeinfo requires adjacency of
7100 @c node and sectioning commands; hence the repetition of @chapter BLAH
7101 @c in both conditional blocks.
7104 @include c-aarch64.texi
7108 @include c-alpha.texi
7124 @include c-bfin.texi
7128 @include c-cr16.texi
7132 @include c-cris.texi
7137 @node Machine Dependencies
7138 @chapter Machine Dependent Features
7140 The machine instruction sets are different on each Renesas chip family,
7141 and there are also some syntax differences among the families. This
7142 chapter describes the specific @command{@value{AS}} features for each
7146 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7147 * SH-Dependent:: Renesas SH Dependent Features
7154 @include c-d10v.texi
7158 @include c-d30v.texi
7162 @include c-epiphany.texi
7166 @include c-h8300.texi
7170 @include c-hppa.texi
7174 @include c-i370.texi
7178 @include c-i386.texi
7182 @include c-i860.texi
7186 @include c-i960.texi
7190 @include c-ia64.texi
7194 @include c-ip2k.texi
7198 @include c-lm32.texi
7202 @include c-m32c.texi
7206 @include c-m32r.texi
7210 @include c-m68k.texi
7214 @include c-m68hc11.texi
7218 @include c-microblaze.texi
7222 @include c-mips.texi
7226 @include c-mmix.texi
7230 @include c-msp430.texi
7234 @include c-ns32k.texi
7238 @include c-pdp11.texi
7250 @include c-rl78.texi
7258 @include c-s390.texi
7262 @include c-score.texi
7267 @include c-sh64.texi
7271 @include c-sparc.texi
7275 @include c-tic54x.texi
7279 @include c-tic6x.texi
7283 @include c-tilegx.texi
7287 @include c-tilepro.texi
7303 @include c-v850.texi
7307 @include c-xgate.texi
7311 @include c-xstormy16.texi
7315 @include c-xtensa.texi
7319 @c reverse effect of @down at top of generic Machine-Dep chapter
7323 @node Reporting Bugs
7324 @chapter Reporting Bugs
7325 @cindex bugs in assembler
7326 @cindex reporting bugs in assembler
7328 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7330 Reporting a bug may help you by bringing a solution to your problem, or it may
7331 not. But in any case the principal function of a bug report is to help the
7332 entire community by making the next version of @command{@value{AS}} work better.
7333 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7335 In order for a bug report to serve its purpose, you must include the
7336 information that enables us to fix the bug.
7339 * Bug Criteria:: Have you found a bug?
7340 * Bug Reporting:: How to report bugs
7344 @section Have You Found a Bug?
7345 @cindex bug criteria
7347 If you are not sure whether you have found a bug, here are some guidelines:
7350 @cindex fatal signal
7351 @cindex assembler crash
7352 @cindex crash of assembler
7354 If the assembler gets a fatal signal, for any input whatever, that is a
7355 @command{@value{AS}} bug. Reliable assemblers never crash.
7357 @cindex error on valid input
7359 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7361 @cindex invalid input
7363 If @command{@value{AS}} does not produce an error message for invalid input, that
7364 is a bug. However, you should note that your idea of ``invalid input'' might
7365 be our idea of ``an extension'' or ``support for traditional practice''.
7368 If you are an experienced user of assemblers, your suggestions for improvement
7369 of @command{@value{AS}} are welcome in any case.
7373 @section How to Report Bugs
7375 @cindex assembler bugs, reporting
7377 A number of companies and individuals offer support for @sc{gnu} products. If
7378 you obtained @command{@value{AS}} from a support organization, we recommend you
7379 contact that organization first.
7381 You can find contact information for many support companies and
7382 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7386 In any event, we also recommend that you send bug reports for @command{@value{AS}}
7390 The fundamental principle of reporting bugs usefully is this:
7391 @strong{report all the facts}. If you are not sure whether to state a
7392 fact or leave it out, state it!
7394 Often people omit facts because they think they know what causes the problem
7395 and assume that some details do not matter. Thus, you might assume that the
7396 name of a symbol you use in an example does not matter. Well, probably it does
7397 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
7398 happens to fetch from the location where that name is stored in memory;
7399 perhaps, if the name were different, the contents of that location would fool
7400 the assembler into doing the right thing despite the bug. Play it safe and
7401 give a specific, complete example. That is the easiest thing for you to do,
7402 and the most helpful.
7404 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
7405 it is new to us. Therefore, always write your bug reports on the assumption
7406 that the bug has not been reported previously.
7408 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7409 bell?'' This cannot help us fix a bug, so it is basically useless. We
7410 respond by asking for enough details to enable us to investigate.
7411 You might as well expedite matters by sending them to begin with.
7413 To enable us to fix the bug, you should include all these things:
7417 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
7418 it with the @samp{--version} argument.
7420 Without this, we will not know whether there is any point in looking for
7421 the bug in the current version of @command{@value{AS}}.
7424 Any patches you may have applied to the @command{@value{AS}} source.
7427 The type of machine you are using, and the operating system name and
7431 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
7435 The command arguments you gave the assembler to assemble your example and
7436 observe the bug. To guarantee you will not omit something important, list them
7437 all. A copy of the Makefile (or the output from make) is sufficient.
7439 If we were to try to guess the arguments, we would probably guess wrong
7440 and then we might not encounter the bug.
7443 A complete input file that will reproduce the bug. If the bug is observed when
7444 the assembler is invoked via a compiler, send the assembler source, not the
7445 high level language source. Most compilers will produce the assembler source
7446 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
7447 the options @samp{-v --save-temps}; this will save the assembler source in a
7448 file with an extension of @file{.s}, and also show you exactly how
7449 @command{@value{AS}} is being run.
7452 A description of what behavior you observe that you believe is
7453 incorrect. For example, ``It gets a fatal signal.''
7455 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
7456 will certainly notice it. But if the bug is incorrect output, we might not
7457 notice unless it is glaringly wrong. You might as well not give us a chance to
7460 Even if the problem you experience is a fatal signal, you should still say so
7461 explicitly. Suppose something strange is going on, such as, your copy of
7462 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
7463 library on your system. (This has happened!) Your copy might crash and ours
7464 would not. If you told us to expect a crash, then when ours fails to crash, we
7465 would know that the bug was not happening for us. If you had not told us to
7466 expect a crash, then we would not be able to draw any conclusion from our
7470 If you wish to suggest changes to the @command{@value{AS}} source, send us context
7471 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
7472 option. Always send diffs from the old file to the new file. If you even
7473 discuss something in the @command{@value{AS}} source, refer to it by context, not
7476 The line numbers in our development sources will not match those in your
7477 sources. Your line numbers would convey no useful information to us.
7480 Here are some things that are not necessary:
7484 A description of the envelope of the bug.
7486 Often people who encounter a bug spend a lot of time investigating
7487 which changes to the input file will make the bug go away and which
7488 changes will not affect it.
7490 This is often time consuming and not very useful, because the way we
7491 will find the bug is by running a single example under the debugger
7492 with breakpoints, not by pure deduction from a series of examples.
7493 We recommend that you save your time for something else.
7495 Of course, if you can find a simpler example to report @emph{instead}
7496 of the original one, that is a convenience for us. Errors in the
7497 output will be easier to spot, running under the debugger will take
7498 less time, and so on.
7500 However, simplification is not vital; if you do not want to do this,
7501 report the bug anyway and send us the entire test case you used.
7504 A patch for the bug.
7506 A patch for the bug does help us if it is a good one. But do not omit
7507 the necessary information, such as the test case, on the assumption that
7508 a patch is all we need. We might see problems with your patch and decide
7509 to fix the problem another way, or we might not understand it at all.
7511 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
7512 construct an example that will make the program follow a certain path through
7513 the code. If you do not send us the example, we will not be able to construct
7514 one, so we will not be able to verify that the bug is fixed.
7516 And if we cannot understand what bug you are trying to fix, or why your
7517 patch should be an improvement, we will not install it. A test case will
7518 help us to understand.
7521 A guess about what the bug is or what it depends on.
7523 Such guesses are usually wrong. Even we cannot guess right about such
7524 things without first using the debugger to find the facts.
7527 @node Acknowledgements
7528 @chapter Acknowledgements
7530 If you have contributed to GAS and your name isn't listed here,
7531 it is not meant as a slight. We just don't know about it. Send mail to the
7532 maintainer, and we'll correct the situation. Currently
7534 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
7536 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
7539 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
7540 information and the 68k series machines, most of the preprocessing pass, and
7541 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
7543 K. Richard Pixley maintained GAS for a while, adding various enhancements and
7544 many bug fixes, including merging support for several processors, breaking GAS
7545 up to handle multiple object file format back ends (including heavy rewrite,
7546 testing, an integration of the coff and b.out back ends), adding configuration
7547 including heavy testing and verification of cross assemblers and file splits
7548 and renaming, converted GAS to strictly ANSI C including full prototypes, added
7549 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
7550 port (including considerable amounts of reverse engineering), a SPARC opcode
7551 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
7552 assertions and made them work, much other reorganization, cleanup, and lint.
7554 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
7555 in format-specific I/O modules.
7557 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
7558 has done much work with it since.
7560 The Intel 80386 machine description was written by Eliot Dresselhaus.
7562 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
7564 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
7565 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
7567 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
7568 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
7569 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
7570 support a.out format.
7572 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
7573 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
7574 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
7575 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
7578 John Gilmore built the AMD 29000 support, added @code{.include} support, and
7579 simplified the configuration of which versions accept which directives. He
7580 updated the 68k machine description so that Motorola's opcodes always produced
7581 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
7582 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
7583 cross-compilation support, and one bug in relaxation that took a week and
7584 required the proverbial one-bit fix.
7586 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
7587 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
7588 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
7589 PowerPC assembler, and made a few other minor patches.
7591 Steve Chamberlain made GAS able to generate listings.
7593 Hewlett-Packard contributed support for the HP9000/300.
7595 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
7596 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
7597 formats). This work was supported by both the Center for Software Science at
7598 the University of Utah and Cygnus Support.
7600 Support for ELF format files has been worked on by Mark Eichin of Cygnus
7601 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
7602 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
7603 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
7604 and some initial 64-bit support).
7606 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
7608 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
7609 support for openVMS/Alpha.
7611 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
7614 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
7615 Inc.@: added support for Xtensa processors.
7617 Several engineers at Cygnus Support have also provided many small bug fixes and
7618 configuration enhancements.
7620 Jon Beniston added support for the Lattice Mico32 architecture.
7622 Many others have contributed large or small bugfixes and enhancements. If
7623 you have contributed significant work and are not mentioned on this list, and
7624 want to be, let us know. Some of the history has been lost; we are not
7625 intentionally leaving anyone out.
7627 @node GNU Free Documentation License
7628 @appendix GNU Free Documentation License
7632 @unnumbered AS Index
projects / deliverable / binutils-gdb.git / blob